An imprint of Elsevier Inc.

2008, Elsevier Inc. All rights reserved.

First published 2008

No part of this publication may be reproduced, stored in a retrieval

system, or transmitted in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, without the prior
permission of the Publishers. Permissions may be sought directly from
Elseviers Health Sciences Rights Department, 1600 John F. Kennedy
Boulevard, Suite 1800, Philadelphia, PA 19103-2899, USA: phone: (+1)
215 239 3804; fax: (+1) 215 239 3805; or, e-mail: healthpermissions@
elsevier.com. You may also complete your request on-line via the Elsevier
homepage (http://www.elsevier.com), by selecting Support and contact
and then Copyright and Permission.

ISBN: 978-1-4160-4271-6

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data

A catalog record for this book is available from the Library of Congress

Notice
Medical knowledge is constantly changing. Standard safety precautions must
be followed, but as new research and clinical experience broaden our
knowledge, changes in treatment and drug therapy may become necessary
or appropriate. Readers are advised to check the most current product
information provided by the manufacturer of each drug to be administered
to verify the recommended dose, the method and duration of administra-
tion, and contraindications. It is the responsibility of the practitioner, relying
on experience and knowledge of the patient, to determine dosages and the
best treatment for each individual patient. Neither the Publisher nor the
author assume any liability for any injury and/or damage to persons or
property arising from this publication.
The Publisher

Printed in China
Last digit is the print number: 9 8 7 6 5 4 3 2 1

Working together to grow

libraries in developing countries
www.elsevier.com | www.bookaid.org | www.sabre.org

Commissioning Editor: Rolla Couchman

Development Editor: Sharon Nash
Editorial Assistant: Sarah Penny
Project Manager: Jess Thompson
Design Manager: Charles Gray
Illustration Manager: Gillian Richards
Illustrator: Debbie Maizels & Barking Dog
Marketing Manager(s) (UK/USA): Kathy Neely/Clara Toombs
Preface
ERCP PAST, PRESENT AND FUTURE

ERCP past
1
Since its initial description over 32 decades ago, ERCP has evolved from a diagnostic to a therapeutic procedure.
Such an evolution, however, has required developments in technology to include marketing of a side-viewing
duodenoscope with an elevator to facilitate cannulation, use of therapeutic instruments with larger accessory chan-
nels, and creation and marketing of a wide variety of endoscopic accessories. In addition to technology evolution,
there have been evolutions in techniques and training to bring us to ERCP present. From the latter standpoint,
ERCP is currently recognized as an advanced therapeutic procedure usually requiring an additional year of training
beyond a standard 3-year gastroenterology training program.

ERCP present
With the exception of performing concomitant sphincter of Oddi manometry (SOM), diagnostic ERCP has largely
been supplanted by non-invasive imaging to include abdominal CT, MRI-MRCP and EUS. Therapeutic ERCP was
initially described over 30 years ago when Classen and Kawai independently described endoscopic sphincterotomy
in Germany and Japan, respectively.
From a diagnostic standpoint, ERCP is still used to dene the etiology of acute relapsing pancreatitis (divisum,
anomalous PB union, annular pancreas, SOD . . .), to differentiate chronic pancreatitis from intraductal papillary
mucus secreting neoplasm (IPMN), to dene the presence or absence of common bile duct stones in jaundice,
cholangitis, or acute pancreatitis and to distinguish benign from malignant biliary strictures. From the latter
standpoint, however, ERCP remains an imperfect technology, and meta-analyses of series reviewing brush cytology
or endoscopic biopsy to distinguish benign from malignant biliary stenoses suggest only a 3050% positivity, even
in the setting of malignancy. ERCP is also used diagnostically to dene the etiology of smoldering pancreatitis to
include diagnosis of ductal disruption and sphincter edema.
Currently, ERCP has evolved into a primary therapeutic modality. For the past three decades, therapeutic
maneuvers have been centered on, and facilitated by, endoscopic sphincterotomy. Despite changes in sphinctero-
tome design (pull, push, needle-knife, long-nose, 2030 mm wire, monolament vs braided cutting wire, standard
vs rapid-exchange system, single/double/triple lumen technology), technique of use (pure cut, blended current,
pulsed current/Erbe generator) and length and direction of incision, enlarging the pancreaticobiliary orice to
facilitate access to strictures, stones, and ductal disruptions remains the most common therapy undertaken with
ERCP. While balloon dilation of the biliary sphincter may theoretically offer comparable access to biliary sphinc-
terotomy without the potential risk of bleeding and perforation complications, the rates of procedural pancreatitis,
often severe, are prohibitive without the use of adjunctive measures such as pancreatic duct stent placement. As
such, the use of balloon dilation of the papilla has been relegated to bit parts in our therapeutic armamentarium,
most commonly in the setting of coagulopathy and Billroth-II anatomy.
Changes that have occurred over the past decade in the endoscopic treatment of biliary stones have included
improved mechanical lithotriptors and electrohydraulic and laser lithotripsy under direct cholangioscopic visualiza-
tion. Changes in malignant stricture therapy include evolution from plastic prostheses to self-expandable metal
stents (SEMS). For benign strictures, dilating balloons have become stronger and have been marketed with larger
diameters. Treatment has evolved from placement of a single polyethylene prosthesis to 2, 3, 4, and even 5 stents
in parallel. Studies are currently underway looking at the use of completely covered 810 mm SEMS in an attempt
to treat a subgroup of these patients with a single endoscopic manipulation. The treatments of postoperative biliary
leaks have become routine and endoscopic papillectomy has replaced surgical resection for most ampullary
adenomas.
On the pancreatic side, obstructing calculi have become a major indication for pancreatic endotherapy although
approximately 50% of such patients require pretreatment with extracorporeal shock wave lithotripsy (ESWL) prior
to removal. Obstructing strictures are also routinely being treated, although results remain more nebulous than
those reported for benign biliary stenosis and there appears to be discordance between resolution of the stricture
(2030%) and symptomatic improvement (6780%). Treatment of ductal disruption has now become standard
therapy in high-volume ERCP centers. The latter include amenable pancreaticocutaneous stulas (external stula)
and internal stulas (pseudocyst, high amylase pleural effusion, pancreatic ascites, pancreaticoenteric or pancre-
aticobiliary stula). Data are clear that stulas are much more likely to close if a transpapillary stent bridges the
area of duct disruption. In contrast, a glandular disconnection (disconnected gland syndrome) in which the major

ix
 PREFACE

component of ongoing glandular secretions is from the disconnected upstream duct is unlikely to close unless there is concomi-
tant transgastric or transenteric drainage of an associated uid collection.

ERCP future
There are multiple possible scenarios for ERCP in the future. These include widespread application of techniques or technology
that are currently used only in high volume centers. Examples include semi-routine use of cholangioscopy to rule out retained
CBD stones or to better dene the etiology of biliary strictures. The latter will likely require disposable cholangioscopes. On the
other hand, widespread application of durable video cholangioscopes or pancreaticoscopes is potentially feasible.
Other procedures that are utilized infrequently may become commonplace if technologic and reimbursement issues can be
solved. These include endoscopic delivery of brachytherapy or photodynamic therapy (PDT) to patients with unresectable, hilar
cholangiocarcinoma. They include a more widespread approach to the endotherapy of pancreatic necrosis, treating not only the
ductal disruption that is invariable in severe necrosis but also its consequences, to include drainage of associated uid collections
and debridement of necrotic debris.
It is possible that ERCP in the future will become aggressively more therapeutic. An example might be injection of a litholytic
agent into the pancreatic duct in patients with chronic calcic pancreatitis. Prostaglandin inhibitors as well as long-acting neuro-
modulators may also be injected intraductally in painful chronic pancreatitis. Stent placement may evolve to solve the problems
of occlusion by bacterial biolm or mucosal hyperstasia with plastic and metal prostheses, respectively. It is possible that intra-
ductal injection of chemotherapeutic or immunomodulating drugs may play a bit part or a major role in the future management
of pancreaticobiliary malignancies. Certainly work will continue on ways to minimize procedurally-related pancreatitis, the major
complication of ERCP. The latter will include modications of technique, prophylactic PD stents and pre- or intra-procedure
injection of drugs, intravenously or directly into the pancreatic duct.
On the other hand, protenomics promises what conventional tumor markers (CA19-9, CEA, alterations in p53, K-Ras . . .) have
not delivered: early diagnosis of PB malignancy or denition of an extremely high risk patient group who perhaps require more,
as opposed to fewer, diagnostic procedures.
The future of ERCP is like, The check is in the mail. It may or may not arrive. Even if it does, the check may be so delayed
as to prove almost worthless. The evolution of ERCP, of course, will not occur in a vacuum and its future will be contingent upon
parallel advancements in imaging and lab testing and breakthrough technology or techniques in other disciplines such as lapa-
roscopic or transgastric surgery, more effective and less toxic chemoirradiation, or even alternative endoscopic therapies delivered
by EUS. Nevertheless, we predict a bright decade for ERCP and suggest that even at its 50th anniversary, it will remain a vibrant
technology and technique in the care of patients with PB disease.
It is with this timeline and cumulative perspective that the editors have assembled this ERCP text. Many of the worlds experts
have contributed chapters and video material in an attempt to make this book not only relevant, but comprehensive, for any
endoscopist who utilizes ERCP in their practice. Omissions, if any, are ours. Technology changes rapidly and reusable cholan-
gioscopes (e.g. Spy ScopeTM) will have been introduced by the time this text is published and it is likely that other technology may
be rendered obsolete. Clinical trials may fuel or dampen our enthusiasm for one procedure or another. Despite this, the editors
have done their best to produce a text that will put both evolving technology and current techniques into clinical perspective. After
all, ERCP remains a clinical tool to facilitate patient care and improve clinical outcomes. It is our hope and intention that this
book facilitates and improves that care.

Todd H. Baron, MD
Richard A. Kozarek, MD
David L. Carr-Locke, MD

x
List of Contributors
Dr Douglas G. Adler Dr Todd H. Baron
Assistant Professor of Medicine, Director of Professor of Medicine
Therapeutic Endoscopy Mayo Clinic College of Medicine
Huntsman Cancer Center Rochester MN
University of Utah School of Medicine USA
Salt Lake City UT
USA Dr Suryaprakash Bhandari
Associate Consultant
Dr Sushil K. Ahlawat Institute of Advanced Endoscopy
Assistant Professor of Medicine Jaslok Hospital
Department of Medicine, Gastroenterology Mumbai
Division India
Georgetown University School of Medicine
Georgetown University Hospital Dr Kenneth F. Binmoeller
Washington DC Director, Interventional Endoscopy Services
USA California Pacic Medical Center
San Francisco CA
Dr Jawad Ahmad USA
Assistant Professor of Medicine and Co-Medical
Director of Liver Transplantation Dr William R. Brugge
University of Pittsburgh Associate Professor of Medicine, Harvard Medical
Pittsburgh PA School
USA Director of Endoscopy, GI Unit
Massachusetts General Hospital
Dr Firas H. Al-Kawas Boston MA
Professor of Medicine and Chief of Endoscopy USA
Gastroenterology Division
Georgetown University Hospital Dr Jonathan M. Buscaglia
Washington DC Therapeutic Endoscopy Fellow
USA Division of Gastroenterology and Hepatology
The Johns Hopkins University School of Medicine
Dr Raed M. Alsulaiman Baltimore MD
Assistant Professor of Medicine and Consultant USA
Internist, Gastroenterologist
King Faisal University Dr David L. Carr-Locke
King Fahd Hospital of The University Director
Alkhobar The Endoscopy Institute
Kingdom of Saudia Arabia Brigham & Womens Hospital
Boston MA
Dr Tiing Leong Ang USA
Consultant Gastroenterologist
Department of Medicine Dr Annie On-On Chan
Changi General Hospital Associate Professor of Medicine
Singapore Department of Medicine
University of Hong Kong
Dr John Baillie Queen Mary Hospital
Professor of Medicine Hong Kong
Division of Gastroenterology
Wake Forest University Health Sciences Dr Suyi Chang
Winston-Salem NC Gastroenterologist
USA Kaiser Permanente
Walnut Creek CA
Dr Alan Barkun USA
Professor of Medicine and Chairholder the Douglas
G. Kinnear Chair in Gastroenterology Dr Ann M. Chen
Director, Division of Gastroenterology Clinical Instructor
McGill University and the McGill University University of California, Irvine
Health Centre Comprehensive Digestive Disease Center
Montral, Qubec Orange CA
Canada USA

xi
 LIST OF CONTRIBUTORS

Professor Guido Costamagna Dr Joseph E. Geenen

Professor of Surgery and Gastroenterology
Digestive Endoscopy Unit
Universita Cattolica A Gemelli
Roma
Italy
Dr Giovanni D. De Palma
Chief, Section of Diagnostic and Therapeutic
Endoscopy
Department of Surgery and Advanced
Technologies
University of Naples Federico II
Naples
Italy
Clinical Professor of Medicine
Medical College of Wisconsin
Director, Pancreatic Biliary Center
St Lukes Medical Center
Milwaukee WI
USA
Dr Gregory G. Ginsberg
Professor of Medicine
Director of Endoscopic Services
Gastroenterology Division
Hospital of the University of Pennsylvania
Philadelphia PA
USA

Dr Jacques Deviere Dr Jaquelina M. Gobelet

Professor of Medicine Fellow of The Latin American Advanced
Chairman, Department of Gastroenterology Gastrointestinal Endoscopy Training Center
and Hepatopancreatology Santiago
ULBHpital Erasme Chile
Brussels
Belgium Dr Khean-Lee Goh
Professor of Medicine
Dr James J. Farrell Head of Gastroenterology and Hepatology
Director of Pancreaticobiliary Endoscopy and Chief of GI Endoscopy
Endoscopic Ultrasound Faculty of Medicine
Division of Digestive Disease University of Malaya
UCLA School of Medicine Kuala Lumpur
Los Angeles CA Malaysia
USA
Dr Nalini M Guda
Dr Paul Fockens Clinical Assistant Professor of Medicine
Professor of Gastrointestinal Endoscopy University of Wisconsin School of Medicine and
Director of Endoscopy Public Health
Academic Medical Center Pancreatic Biliary Center
University of Amsterdam St Lukes Medical Center
Amsterdam Milwaukee WI
Netherlands USA

Dr Victor L. Fox Dr Robert H. Hawes

Director GI Procedure Unit and Endoscopy Professor of Medicine
Program Division of Gastroenterology/Hepatology
Childrens Hospital Boston Digestive Disease Center
Harvard Medical School Medical University of South Carolina
Boston MA Charleston SC
USA USA

Dr James T. Frakes Dr Jrgen Hochberger

Clinical Professor of Medicine Chefarzt Med Klink IIIGastroenterologie
University of Illinois College of Medicine at Intervent. Endoskopie
Rockford St Bernward Krankenhaus
and Rockford Gastroenterology Associates Hildesheim
Ltd Germany
Rockford IL
USA Dr Kunal Jajoo
Advanced Endoscopy Fellow
Dr Martin L. Freeman The Endoscopy Institute
Professor of Medicine, University of Brigham and Womens Hospital
Minnesota Boston MA
Director, Pancreaticobiliary Endoscopy USA
Fellowship
University of Minnesota and Hennepin Dr Ann Marie Joyce
County Medical Center Gastroenterologist
Minneapolis MN Gastroenterology Department
USA Lahey Clinic
Burlington MA
USA

xii
 LIST OF CONTRIBUTORS

Dr Anthony N. Kalloo Dr Chi Leung Liu

Professor of Medicine Honorary Professor of Surgery
Director, Division of Gastroenterology and Department of Surgery
Hepatology University of Hong Kong
The Johns Hopkins University School of Medicine Queen Mary Hospital
Baltimore MD Hong Kong
USA
Dr Simon K. Lo
Dr Peter B. Kelsey Director of Endoscopy, Cedars-Sinai Medical Center
Assistant Professor of Medicine Clinical Professor of Medicine
Harvard Medical School David Geffen School of Medicine at UCLA
Gastrointestinal Unit Division of Digestive Diseases
Massachusetts General Hospital Cedars-Sinai Medical Center
Boston MA Los Angeles CA
USA USA

Dr Michael B. Kimmey Dr Jrgen Maiss

Professor of Medicine Department of Medicine I
Division of Gastroenterology University Erlangen-Nrnberg
University of Washington Medical Center Erlangen
Seattle WA Germany
USA
Dr Amit Maydeo
Dr Michael L. Kochman Director, Institute of Advanced Endoscopy
Professor of Medicine and Professor of Medicine in Chief, Department of Endoscopy and EUS
Surgery Jaslok Hospital
Gastroenterology Division Mumbai
Hospital of the University of Pennsylvania India
Philadelphia PA
USA Dr Detlev Menke
Leading Consultant
Dr Tadashi Kodama Department of Medicine III
Chief of Gastroenterology St. Bernward Hospital
Kyoto Shimogamo Hospital Hildesheim
Kyoto Germany
Japan
Dr Desiree E. Morgan
Dr Tatsuya Koshitani Associate Professor, Diagnostic Radiology
Chief of Endoscopy Unit Department of Radiology
Department of Gastroenterology University of Alabama at Birmingham
Kyoto City Hospital Birmingham AB
Kyoto USA
Japan
Dr Claudio Navarrete
Richard A. Kozarek Director of Minimally Invasive Surgery
Executive Director, Digestive Disease Institute Clinica Alemana-Universidad del Desarrollo
Virginia Mason Medical Center Chair of The Latin American Advanced
Clinical Professor of Medicine Gastrointestinal Endoscopy Training Center
University of Washington Santiago
Seattle WA Chile
USA
Dr Horst Neuhaus
Dr Yuk Tong Lee Professor of Medicine
Honorary Clinical Associate Professor Department of Gastroenterology
The Chinese University of Hong Kong Evangelisches Krankenhaus Dsseldorf
Prince of Wales Hospital Dsseldorf
Shatin, NT Germany
Hong Kong
Dr Wai Choung Ong
Dr Glen A. Lehman Asian Institute of Gastrenterology
Indiana University Medical Center Jubilee Hills
Indianapolis IN Hyderabad
USA India

Dr Joseph W. Leung Dr Wai Choung Ong

Mr & Mrs C.W. Law Professor of Medicine Asian Institute of Gastrenterology
Division of Gastroenterology and Hepatology Jubilee Hills
UC Davis Medical Center Hyderabad
Sacramento CA India
USA
xiii
 LIST OF CONTRIBUTORS

Dr William M. Outlaw Dr Chan-Sup Shim

Fellow Professor of Medicine
Division of Gastroenterology Digestive Disease Center
Wake Forest University Health Sciences Soon Chun Hyang University
Winston-Salem NC Seoul
USA Korea

Dr Bret T. Petersen Dr Lalit Shimpi

Professor of Medicine Professor of Medicine
Division of Gastroenterology and Hepatology Department of Gastroenterology & GI Endoscopy
Mayo Clinic College of Medicine Jehangir Hospital in association with Apollo
Rochester MN Hospitals
USA Pune
India
Dr Jeffrey L. Ponsky
Department of Surgery Dr Hardeep M. Singh
University Hospital of Cleveland Fellow, Department of Gastroenterology
Cleveland OH Kaiser Permanente Los Angeles Medical Center
USA Los Angeles CA
USA
Dr G. Venkat Rao
Asian Institute of Gastroenterology Dr Adam Slivka
Jubilee Hills Professor of Medicine
Hyderabad University of Pittsburgh
India Pittsburgh PA
USA
Dr Nageshwar Reddy
Asian Institute of Gastroenterology Dr Nib Soehendra
Jubilee Hills Professor of Surgery, Director, Department of
Hyderabad Interdisciplinary Endoscopy
India University Medical Center
Hamburg-Eppendorf
Dr Banerjee Rupa Germany
Asian Institute of Gastroenterology
Jubilee Hills Dr Ellert J. van Soest
Hyderabad Department of Gastroenterology and Hepatology
India University of Amsterdam
Amsterdam
Dr Stefan Seewald The Netherlands
Professor of Gastroenterology
Department of Interdisciplinary Endoscopy Dr Geoffrey Spencer
University Medical Center Instructor of Medicine
Hamburg-Eppendorf Division of Gastroenterology
Germany Hospital of the University of Pennsylvania
Philadelphia PA
Dr Dong Wan Seo USA
Associate Professor of Medicine and Director of
Endoscopy Unit Dr Joseph Sung
Division of Gastroenterology, Department of Professor of Medicine
Internal Medicine Director, Institute of Digestive Disease
Asan Medical Center, University of Ulsan College The Chinese University of Hong Kong
of Medicine Prince of Wales Hospital
Seoul Shatin, NT
South Korea Hong Kong

Dr Syed G.Shah Dr Paul R. Tarnasky

Consultant Gastroenterologist
Department of Gastroenterology
Pinderelds General Hospital
Wakeeld
UK
Clinical Associate Professor of Medicine
University of Texas Southwestern
Methodist Dallas Medical Center
Dallas TX
USA

Dr Stuart Sherman Dr Yoshihide Tatsumi

Professor of Medicine and Radiology Chief, Department of Gastrointestinal Diseases
Clinical Director of Gastroenterology and Hepatology Matsushita Health Care Center
Director of ERCP Moriguchi
Indiana University Medical Center Osaka
Indianapolis IN Japan
USA

xiv
 LIST OF CONTRIBUTORS

Dr Mark Topazian Dr Benjamin Chun-Yu Wong

Associate Professor of Medicine Professor of Medicine
Mayo College of Medicine Department of Medicine
Rochester MN University of Hong Kong
USA Queen Mary Hospital
Hong Kong
Dr Eduardo Valdivieso
Professor of Surgery Ponticia Universidad Dr Gregory Zuccaro Jr
Javeriana, Colombia Director, Center for Endoscopy and Diseases of the
Fellow of The Latin American Advanced Pancreas and Bile Ducts
Gastrointestinal Endoscopy Training Center Department of Gastroenterology and Hepatology
Santiago Cleveland Clinic
Chile Cleveland OH
USA
Dr James L Watkins
Division of Gastroenterology and
Hepatology
Indiana University Medical Center
Indianapolis IN
USA

xv
 Dedication
To our families, friends and colleagues who tolerate us
To our teachers and students who taught us everything we know
To our patients who amaze us and
To our contributors who helped us,
We dedicate this to you.
Acknowledgements
Virginia Mason Medical Center:
Mayo Clinic
Brigham & Womens Hospital:
Hannah Scott, Word-Processing
Terrance King, AV Support
Donna Smith, Practice Support
Mindy K. Parker, Secretarial Support
Sandra Healey, Secretarial Support
xvii
 SECTION 1 GENERAL TOPICS
Chapter
Medicolegal Issues
1 James T. Frakes
INTRODUCTION TO MEDICOLEGAL ISSUES
The medicolegal environment
Medicine is an imprecise science, inuenced by the vagaries and
unpredictable nature of biologic systems and the art of interpersonal
relationships. Human illnesses are, from the outset, adverse out-
comes of life, and it is often difcult for physicians to correct or
mitigate these illnesses. Furthermore, the techniques, tools, and
technology available to aid in this task often have associated inade-
quacies or risks. Therefore, restoring biologic function to its former
healthy state is oftentimes incomplete, sometimes unsuccessful, and
occasionally complicated by iatrogenic injury. Negative or adverse
outcomes may include cognitive or technical failures, ineffective
therapies, complications of therapy, high costs and extended hospi-
talizations, and missed work and life activities. Any or all of these
may lead to patient dissatisfaction and a desire to assign blame and
seek compensation.
It is in this environment of personal illness and fear, limited
medical art and science, patient dissatisfaction, and legal avenues
for redress that medicolegal issues arise. Especially in the United
States, there has been a steady increase in both lawsuits led for
medical malpractice and the size of awards granted for damages.1
Physicians and insurance companies generally blame unrealistic
patient expectations, avaricious trial lawyers and inappropriately
high jury awards for the increased number of lawsuits, which in turn
lead to high malpractice insurance rates, diminished access to
certain types of medical care and the costly practice of defensive
medicine. Alternatively, attorneys and some patients blame true
medical negligence, high medical costs, inadequate policing of
incompetent physicians, and poor nancial management by insur-
ance companies for the worsening medicolegal climate.
It is therefore appropriate for physicians to study medicolegal
issues, especially in their specialty areas of practice, to optimize
patient outcomes, limit patient harm and dissatisfaction, and mini-
mize the risk of malpractice litigation.
Medicolegal issues in gastroenterology and
gastrointestinal endoscopy
Gastroenterologists, like all physicians, have reason to be concerned
about malpractice litigation. But, data specic to gastroenterology
and endoscopy are scarce. Many commercial insurance carriers are
unwilling to share data on claims and awards, regarding this as
proprietary information and fearing possible adverse publicity or
stimulation of even greater malpractice litigation activity. One asso-
ciation of insurers, the Physician Insurers Association of America
(PIAA), pools information from approximately 30 physician-owned
or managed insurance companies in the United States and periodi-
cally publishes data. These data probably provide the best picture of
medical malpractice claims currently available. It is interesting to
note that despite the increasingly complex nature of gastroenterol-
ogy and gastrointestinal endoscopy, gastroenterologists are sued less
often than most other specialty groups, ranking 21st of 28 specialty
groups in the number of claims reported in the PIAA data.2 Gastro-
enterologists accounted for about 2% of claims and 1.4% of PIAA
indemnity payments. About one of ve gastroenterology claims
resulted in indemnity payment. The highest indemnity payment has
been $2.9 million, almost triple the largest gastroenterology payment
reported in 1999.3
According to PIAA data, despite the fact that most gastro-
enterologists spend most of their time performing procedures, the
basis for claims in gastroenterology from 1985 to 2004 involved
cognitive issues 60% of the time, including consultations, diagnostic
interviews, evaluations, medication prescriptions, injections
and vaccinations. Procedural misadventures accounted for about
40% of the claims during that time period, including procedures
on the hollow gut or biliary tract, including endoscopic retrograde
cholangiopancreatography (ERCP). The vast majority of these
endoscopy cases involved perforation of the gut, while a much
smaller percentage included pancreatitis, hemorrhage, dental
injury or falling from the bed while sedated. Issues involving
informed consent as a secondary allegation occurred about 17% of
the time.
Medicolegal issues in ERCP
Because ERCP is one of the most technically difcult procedures
performed by gastrointestinal endoscopists and because compli-
cations, sometimes severe, are more common than with other
endoscopic procedures, ERCP might be expected to account for a
disproportionate number of claims against gastroenterologists.
However, the relative risk of litigation arising from ERCP, despite
the signicantly higher rate of complications, appears to be less than
twice that of signicantly less complex procedures such as exible
sigmoidoscopy or gastroscopy.4 In one Canadian database, ERCP-
related events accounted for only about 6% of gastrointestinal-related
legal actions from 1990 to 1997.5 One possible explanation for this
seeming paradox is the more intensive informed consent processes
for ERCP when compared with simpler endoscopic procedures. This
hints at the importance of adequate informed consent as a risk
management strategy.
After a discussion of legal principles of importance in medical
practice, specic risk management strategies for ERCP will be out-
lined below, including sound medical practice, avoidance of compli-
cations, and truly informed consent.
3
 SECTION 1 GENERAL TOPICS
IMPORTANT LEGAL PRINCIPLES IN
MEDICAL PRACTICE
Elements of a malpractice case: the principles
of tort law
The most common form of a medical malpractice action falls under
the principles of tort law, a civil wrong rather than a criminal
action. Such civil wrongs are generally compensated by monetary
redress. To succeed in a medical malpractice action, the plaintiff
must prove four basic legal elements by a preponderance of evidence
(the fact at issue is more probable than not) rather than proving
beyond a reasonable doubt as in criminal actions.4,6,7 These four
basic elements that must be proven are:
1. The physician owed a duty of care to the patient.
2. The physician breached that duty by violating the applicable
standard of care.
3. The breach of duty caused an injury.
4. The patients injury is compensable (damages).
Duty
The physicians duty to the patient arises from the physicianpatient
relationship. If there is no physicianpatient relationship there is no
malpractice risk. The relationship is usually created through an
ofce visit, hospital visit or performance of a procedure, but may be
created without actual face-to-face meeting between the physician
and patient. For example, an appointment for an ofce visit or endo-
scopic procedure or the prescribing of a colon cleansing agent prior
to colonoscopy might create such a physicianpatient relationship.
Clearly dening a physicians duty or role in the management of a
patient, thereby limiting the scope of the duty, might help to reduce
subsequent liability. Once established, the physician-patient rela-
tionship continues until ofcially and appropriately terminated by
the patient or physician.
Breach of duty
The duty of the physician once the physicianpatient relationship
has been established is to practice within a reasonable standard of
care. Failure to meet the standard of care constitutes negligence and
is the central issue in most medical malpractice litigation. While
often difcult to dene, a reasonable standard of care is usually
established with the aid of expert witnesses acting as medicolegal
consultants.
Causation
To be successful in a medical malpractice suit, the plaintiff must
prove that substandard care was the proximate (substantial rather
than minor) cause of injury.
Injury
Further, to succeed, the plaintiff patient must establish that some
type of physical or psychological injury occurred. Having shown that
a breach of duty caused an injury, one or more of three types of
damages might be awarded in the form of monetary payments.
These include general damages for pain and suffering; special
damages for past, present and future medical expenses, loss of
income, wages and prots; and punitive damages for gross
negligence such as intentional harm, conscious indifference, or
fraud. Punitive damages are generally not covered by malpractice
insurance.
4
STANDARDS OF CARE
The general concept
The standard of care is a legal concept describing the duty that physi-
cians must fulll in their care of patients.4,7,8 A failure to practice
within the standard of care is a breach of that duty, one of the four
central elements of a malpractice case. The standard of care is
usually established through expert testimony, published data and
accepted practice guidelines. Of these, the most important in court
is expert testimony. Expert testimony seeks to establish a standard
of care reecting the practice that is customary among competent
gastroenterologists in good professional standing who are practicing
with reasonable diligence, and should reect the current practice at
the time of the injury. Simply stated, the standard of care is good
patient care. It is not dened as optimal or best medical practice
exhibited by only a few noted experts in the eld but rather what
would be expected from a reasonable peer under the same
circumstances.
Majority and minority standards
Because there are often many ways to manage a clinical problem,
more than one standard of care may be applicable for evaluating or
treating a condition. Practicing the majority standard, or most
commonly taken approach by most peers, is usually the most defen-
sible method of practice. A less common approach, the minority
standard, may be acceptable, but should be explained in terms of
why a strategy differing from the usual was employed. As stated in
one recent publication, physicians who practice in ignorance of the
majority standard do so at their peril.7
Guidelines
Guidelines, or so-called practice parameters, developed by specialty
societies, federal agencies, or panels of experts may be useful in
establishing standards of care. These professional guidelines are
often widely available and provide consensus statements codifying
professional custom which may then form the actual basis for a legal
standard of care. The validity of such guidelines stems from the
sponsoring organizations expertise and prestige, the nature and
purpose of the guideline, conicting views held by other authorities,
and the direct applicability of the guideline to the case under
consideration.
While it might be tempting to assume that clinical guidelines
would reduce malpractice risk by helping physicians understand a
consensus of good care, in reality such guidelines are more likely
to be used in malpractice litigation by the plaintiff as evidence that
the physician failed to meet the standard of care.
Expert testimony
The most common method of establishing a standard of care and
subsequently a breach of duty is to rely on expert testimony from a
medical witness. Such an expert witness should be appropriately
licensed and board certied and should have been practicing in the
medical specialty area in question for at least 3 of the previous 5
years.9,10 For such testimony the expert should receive reasonable
compensation not based on a contingency fee. The opinion of the
medical expert should be unbiased and non-emotive, and as such it
should not matter whether the expert is retained by the plaintiff or
the defendant. Expert testimony requires a review of the medical
record and an opinion regarding the patients care. Such an opinion

may be given in a variety of formats including afdavit, deposition
or even testimony in court. The expert medical witness provides
an important service to patients, physician colleagues, and the
courts provided that such opinions are thoughtful, accurate and
unbiased.
Vicarious liability
While most medical malpractice actions are taken against an indi-
vidual directly involved with an alleged wrongdoing, there is also a
legal concept which allows liability to be extended beyond someone
who directly caused an injury to persons on whose behalf that person
may have acted. This may mean that physicians may be held liable
not only for their own actions but also for the actions of others in
some type of subordinate role. Such liability is referred as vicarious
liability.11,12
Respondeat superior
Respondeat superior is a legal principle that holds a master respon-
sible for the wrongdoings of his servants. These masterservant
denitions have evolved to include employeremployee relation-
ships, corporate agent relationships, and teacherstudent relation-
ships.1215 These relationships may apply to preceptors, proctors,
administrators or employers. Such a concept allows blame to be
shared among doctors, trainees, nurses, institutions, etc. and may
provide additional nancially responsible defendants, some with
potentially greater resources than the original defendant, to share
the liability for an injury.
Preceptor
The concept of a preceptor as a teacher or instructor in the area of
gastrointestinal endoscopy is central to the training of young physi-
cians new to the specialty and to practicing physicians acquiring new
skills. Such a preceptor endoscopist might be found vicariously
liable for current or future acts of his trainee. More to the point of
ERCP, a supervising endoscopist might be held liable for part of the
damages arising from a trainee learning the procedure, an experi-
enced colleague acquiring new skills, or either in future misadven-
tures. The degree of liability attributable to each of the principals
would depend on many factors, including knowledge on the part of
the patient that the procedure would be performed by a trainee, the
experience of the trainee and whether the trainee was performing
the procedure within an appropriate standard of care such that the
procedure was done for reasonable indications and with appropriate
skill. With regard to future injuries after completion of training, lia-
bility would hinge on the appropriateness of training and the veracity
of credentials.
Proctor
A physician who observes and monitors another physician, particu-
larly one seeking privileges, is known as a proctor. Proctors have no
duty to the patient and therefore no liability since their role is simply
to assess the capabilities of the physician being monitored. If the
proctor becomes involved in the care of the patient, however, this
could change. To avoid such entanglement, the proctor should not
interfere with the proctored physician, should have a thorough
understanding of proctoring and hospital endoscopy privileges,
should not offer advice or interact with the patient, should only
report to the hospital or regulatory committee, and in the event
he/she witnesses substandard medical care which is harmful to
the patient, should consider contacting an appropriate superior,
Chapter 1 Medicolegal Issues
asking the proctored physician to discontinue the questionable
actions, or, as a last resort, intervening with careful appropriate
documentation.
Employer liability
A physician may be held responsible for an adverse outcome attrib-
utable to substandard service by ofce staff such as violations in
patient condentiality, violations in sterile technique, or failure to
provide appropriate training and supervision to ensure the proper
functioning of ofce staff.
Administrator
If a physician acts in an administrative capacity in an endoscopy unit
or gastroenterology division, he/she has a duty to patients receiving
care in that unit. Failure to develop policies and procedures that
ensure a safe environment and comply with state and federal regula-
tions may constitute vicarious liability. Such responsibilities may
include the acquisition and maintenance of endoscopic equipment,
privileging, infection control and workplace safety. Further, if the
responsible director knew or should have known that an unskilled
physician was practicing in the unit and did not take appropriate
corrective actions, vicarious liability could exist.
Hospital liability
Hospitals may be held responsible for the mistakes of a hospital-
based physician employed by that institution or for inadequate over-
sight provided by endoscopy unit or gastroenterology division
directors. They may also incur vicarious liability for improperly privi-
leging physicians who are inadequately trained to perform a certain
service.13,15
Summary
In summary, the gastroenterologist or gastrointestinal endoscopist
may incur liability for mistakes of individuals whom they supervise
even if they themselves were unaware of the improper actions
and even after their immediate supervisory role had ended. All of
the aforementioned roles of preceptor, proctor, employer and ad-
ministrator should be approached with care and forethought. An
understanding of potential vicarious liability may allow better risk
management strategies to minimize exposure to liability.
INFORMED CONSENT
Introduction
Medical malpractice actions most commonly involve the tort of
negligence wherein a physician is felt to have practiced below the
standard of care (breach of duty). There is, however, a common
and independent cause of malpractice action involving the failure to
obtain informed consent.1619 This is often a secondary allegation
led along with an allegation of practicing below the standard of
care.
Theory of informed consent
The ethical and legal requirement to obtain informed consent prior
to a procedure comes from the concept of personal patient autonomy
and is rooted in the theory of patient self-determination. Against
such a backdrop, the courts have found that a persons right to self-
determination warrants that a physician obtain informed consent.
The competent patient, after receiving appropriate disclosure of
5
 SECTION 1 GENERAL TOPICS
material risks of the procedure in question, and understanding the
risks, benets and alternative approaches, then makes a voluntary
and uncoerced informed decision of whether to proceed.
Early on, the consent process operated under a provider-based
standard (professional standard of disclosure) whereby the physician
was expected to disclose information about the treatment that rea-
sonable physicians believed relevant and that reasonable physicians
generally disclosed to their patients in similar circumstances. More
recently, however, courts have been moving toward a patient-based
standard which mandates that a treating physician disclose as much
information as a reasonable patient would wish to know.
The essential elements of informed consent include:
1. The nature and character of the proposed procedure, preferably
in non-technical terms.
2. The reason or indication for the procedure.
3. The likely benets of the procedure.
4. The material risks and complications of the procedure, including
their relative incidences and severities.
5. The alternatives to the procedure, including those which may be
more or less hazardous than the one proposed, and the alternative
of no treatment.
The consent process should also include an assessment of the indi-
viduals competence to understand the information presented and
the opportunity for patients to ask questions.
Obtaining informed consent is a process that includes more than
placing a signature on a standardized consent form. It involves
mutual communication and decision making and can advance the
physicianpatient relationship. It can also be a risk management
tool, transferring responsibility for risk to the patient who has under-
stood and accepted that even competently performed procedures can
have adverse outcomes.
Material risks
One essential element of disclosure is discussion of the risks and
potential complications of the procedure. These risks should include
procedure-specic material risks, those that a reasonable patient
would want to know in order to make an appropriate decision. The
four elements of risk that the physician needs to consider include:
1. The nature of the risk.
2. The magnitude of the risk.
3. The probability that the risk may occur.
4. The timing of the risk, whether contemporaneous with the pro-
cedure, post-procedure or delayed.
Deciding what material should be disclosed is often not easy. One
authoritative text on informed consent states: The physician must
walk a ne line between providing pertinent risk information and
overwhelming the patient with frightening statistics. Providing too
much extraneous information may be as likely to impair informed
decision making as providing too little.20
Controversial areas
The trend toward a patient-oriented standard of disclosure has
allowed for a broader interpretation of the material risks. What a
reasonable patient would wish to know in making an appropriate
decision might now argue for pertinent disclosure of the experience
level and personal complication history of the specic physician,
rather than national averages, as well as any pertinent economic
interests of the physician. This question of the endoscopists
personal experience might be especially applicable to complicated
6
endoscopic procedures such as ERCP. In a legal case involving
difcult and risky brain surgery, a physician was found liable for not
informing the patient regarding his lack of experience.18
Exceptions to informed consent
There are several exceptions to the informed consent process. These
include:
1. emergencies where the patient is incapacitated to a degree that
consent cannot be obtained and delay would put the patient at
risk;
2. waiver of the right of self-determination, where the patient assigns
that right to the physician;
3. therapeutic privilege, where the physician believes that informed
consent would be detrimental to the patient, usually in an emo-
tional sense;
4. legal mandate, whereby the court orders the patient to undergo
medical therapy without the patients consent;
5. incompetency, where the patient is unable to make a decision and
that responsibility is given to the patients legal guardian.
Informed refusal
This doctrine holds that a patient who refuses a procedure or treat-
ment must do so in a well informed way and that it is the physicians
duty to ensure that such refusal is informed.
Legal consequences of failing to obtain
informed consent
Failure to obtain informed consent is most often a secondary
allegation attached to a charge of practicing below the standard
of care. However, it can be an independent cause of malpractice
action alleging that even though the injury was not due to substan-
dard care, the patient would have refused the treatment or procedure
if material risks had been known. The plaintiff would have to show,
however, that a reasonable person in the same position would not
have undergone the procedure knowing that a small risk of injury
existed.
If there was absolutely no consent obtained for medical treat-
ment, or if the treatment went well beyond the scope of consent, a
charge of battery could be lodged. Although rare, a charge of battery
is a criminal charge and is not covered by most malpractice
insurance.
RISK MANAGEMENT STRATEGIES FOR ERCP
Introduction
Risk management is a process designed to identify reasons for poor
patient outcomes and to suggest corrective actions to prevent both
patient injury and malpractice risk. The formal process includes
dening situations that place the patient and physician at risk, deter-
mining the likelihood and signicance of these circumstances,
applying risk management strategies to individual cases and devel-
oping preventive measures. The following risk management
strategies are generalizable to all medical practice and all gastro-
intestinal endoscopic procedures, but have been adapted to apply
specically to ERCP. They include sound medical practice, in-
cluding proper training; rigorous privileging; understanding and
avoiding complications and lawsuits; and dealing with lawsuits once
led.

Sound medical practice
The best defense against adverse outcomes and malpractice suits is
good medical practice. The rst step in developing sound medical
practice is attaining competence through proper training.
Competence
Competence is dened as the minimal level of skill, knowledge,
and/or expertise derived through training and experience that is
required to safely and prociently perform a task or procedure.21
Simply stated, this means having been trained adequately to develop
endoscopic skills and acquire the knowledge required not only to
recommend and perform endoscopic procedures, but to interpret
and correctly manage the ndings of these procedures. Thus com-
petence assures a safe, accurate, technically safe procedure. Con-
versely, incompetently performed endoscopy, particularly ERCP, can
result in patient injury, incorrect or missed diagnoses, and incom-
plete procedures, which in turn lead to missed or delayed diagnoses,
additional procedures and other testing for the same condition.
There are two components to ensuring competence: training and
subsequent demonstration of competence.
Training
The American Society for Gastrointestinal Endoscopy (ASGE) has
promulgated guidelines to ensure adequate training in gastrointes-
tinal endoscopy, patient monitoring and sedation and analgesia.22
Training in ERCP should take place within the context of a global
clinical training program in the elds of adult or pediatric gastroen-
terology or general surgery. Training obtained outside of a formal
training program must conform to the same guidelines as for formal
fellowship or residency training, and short courses on endoscopy, or
ERCP in particular, are not adequate to attain competence. Once
training is complete, competence should be assessed and evaluated
by the training program director who can provide documentation of
the individuals competence. Alternatively, direct observation by an
impartial credentialed endoscopist might also sufce for an endos-
copist who received training outside of a formal training program.
Complex diagnostic and therapeutic procedures are employed
less frequently than standard procedures and are more prone to have
complications and adverse outcomes. These advanced endoscopic
procedures, including ERCP, require greater skill and are concen-
trated in the hands of fewer individuals. The ASGE recommends
that trainees wishing to acquire advanced endoscopic skills should
have rst completed a standard endoscopy training program during
an approved GI fellowship (or equivalent training) and then receive
further training in specic advanced procedures. Training in ERCP
is covered elsewhere in this book (Chapter 7), and the information
will not be recounted in detail here, but available data suggest that
at least 180200 ERCPs are required for the usual trainee to achieve
competence.23,24 In addition, expert endoscopists are generally
expected to perform at a 95100% technical success level, and
current research supports establishing a standard of 8090% techni-
cal success before trainees are deemed competent in a specic skill.21
This type of training in ERCP is necessary to assure good patient
outcomes and minimize complications and failed procedures which
could subsequently result in malpractice litigation.
Continued competence
Ensuring continued competence is an additional safeguard against
adverse patient outcomes and malpractice litigation. Maintaining
Chapter 1 Medicolegal Issues
clinical and endoscopic skills in ERCP requires an ongoing effort.
This effort includes staying current with GI literature concerning
ERCP, engaging in continuing medical education activities and
achieving familiarity with new endoscopic technologies. In addition,
the endoscopist performing ERCP must also maintain an adequate
case volume to maintain expert procedural skills. In general, studies
have shown a correlation between higher case volumes and greater
technical success.2426 Furthermore, higher ERCP volume has been
associated with lower complication rates, especially with respect to
severe complications. Also, there is probably an important effect of
lifetime experience. Some experienced individuals may, based on
that lifetime experience, be able to maintain high success rates and
low complication rates despite performing only a modest volume of
ERCPs.26
New technology
ERCP has been a robust eld for the development of new endoscopic
techniques and procedures. New techniques require new skills both
major and minor. A major skill is a new technique or procedure that
involves a high level of complexity. Such techniques require formal
training within a training program or through the guidance of a
preceptor before competence can be assessed. Within the realm of
ERCP, these major skills require interventions beyond standard
biliary stone removal and simple stent placement. Minor skills
include new non-experimental developments that are merely minor
extensions of an accepted and widely available technique or proce-
dure. These techniques for ERCP require limited education and
practical exposure such as that which can be obtained from short
courses, training videos, CD-ROMs, DVDs and interactive computer
programs.
A few caveats regarding malpractice exposure are needed with
regard to new endoscopic technology. Endoscopists wishing to
acquire new technologies should not overestimate the need for these
nor overestimate their own skill level in seeking to acquire such
techniques. Who is experienced enough to appropriately incorpo-
rate new or more advanced skills into ones practice and how is the
training acquired? No rigid standards apply, but my personal guide-
line for experience and skill would be three years of experience
beyond training, 95% technical success in gaining access to the
desired duct, and a low personal complication rate compared to
national averages. Further, there should be a compelling clinical
need in ones practice and a persuasive lack of an alternative expert
to refer to. Skills should be acquired in a formal training program
or through a hands-on preceptorship with an experienced expert. It
would not be wise to undertake a newly described but inadequately
assessed technique or to engage in cases which are risky or difcult
early in the endoscopists experience. Ignoring these caveats will
expose patients to adverse outcomes and endoscopists to malpractice
litigation.
A further caveat relates to vicarious liability. The expert
endoscopist should not train the unprepared novice endoscopist to
take on complex difcult tasks before that trainee has the necessary
training and experience to safely acquire these skills. Furthermore,
training less-than-expert ERCP endoscopists for a technically dif-
cult and seldom needed procedure exposes patients, endoscopists,
and trainers to lawsuits, including lawsuits involving vicarious
liability. These procedures should probably be conducted at advanced
centers for complex or high risk cases, and ERCP, particularly
with advanced techniques, should be concentrated among fewer
7
 SECTION 1 GENERAL TOPICS
endoscopists who would thereby perform these procedures more
frequently.
Ensuring competence through privileging
Privileging is the process by which an institution authorizes
an individual to perform a specic procedure.21 These privileges
should be determined separately for each type of procedure, particu-
larly for advanced procedures such as ERCP. The privileging
process includes a review of credentials provided by the training
program or trainer, and a review of the training experience and case
load for each procedure for which privileges are requested. Finally,
an actual observed level of competency should be assured through
proctoring, particularly for advanced procedures. Additionally, insti-
tutions should have guidelines on recredentialing and reprivileging
which ensure continued competence in all procedures, but parti-
cularly the more complex advanced procedures such as ERCP.
Hospitals have a duty to exercise due care in granting privileges to
physicians and they expose themselves to liability for granting spe-
cialized privileges to the poorly trained or inexperienced.12,13,15 This
vicarious liability extends not only to the hospital, but also to indi-
viduals in administrative roles such as the director of the endoscopy
unit.
Peer review
Peer review is intended to identify problems related to adverse out-
comes, prevent their recurrence and help in the reprivileging process.
Ideally, this should be done in a non-threatening manner, but should
be a formal process with a written record. Each physician should
have a personal compilation of his/her own adverse events for com-
parison with peers.7,21,27 Patients have a right to know, in general
terms, the physicians outcome prole.
Interpersonal skills
Another component of sound medical practice, in addition to cogni-
tive and procedural competence and careful privileging, is the inter-
personal skill of the practitioner. This must be developed during
training and polished once in practice. Effective communication
with the patient and family and with other physicians and healthcare
providers is a critical element in good healthcare and risk manage-
ment. Conversely, patient dissatisfaction with physician interper-
sonal and communication skills can be a major factor in the decision
to le a lawsuit.28,29 Displaying a positive caring attitude and com-
municating honestly, beginning with the initial interview and
extending to informed consent and beyond, are critical in forming
a healthy physicianpatient relationship. This good patient rapport
decreases the likelihood of lawsuits. It also helps dene the role of
the physician, limiting the physician duty, and transferring some of
the responsibility for an adverse event to the informed patient.
UNDERSTANDING WHY COMPLICATIONS AND
LAWSUITS OCCUR
ERCP has evolved from a diagnostic tool to a predominately thera-
peutic procedure for a variety of pancreaticobiliary problems. It is a
complicated and sometimes dangerous procedure which can result
in a wide range of short-term complications including pancreatitis,
hemorrhage, perforation and infection. These complications and
other adverse outcomes can lead to patient injury, dissatisfaction and
lawsuits.
8
COMPLICATIONS OF ERCP
Complications of ERCP and sphincterotomy have been covered in
excellent detail by Freeman elsewhere in this book (Chapter 6). Dis-
cussions of risk management strategies, however, require reviewing
some of the information presented elsewhere.
The complications of ERCP and sphincterotomy can vary widely
in different clinical circumstances. These complications appear to
be related primarily to patient-related factors, including the indica-
tion for the procedure, and to the technical skill of the endoscopist.26
The risk factors for overall complications include suspected sphinc-
ter of Oddi dysfunction and technique-related factors such as dif-
cult cannulation, the use of precut sphincterotomy in inexperienced
hands, failure to achieve biliary drainage, and use of percutaneous
transhepatic biliary access. All of these risk factors result in higher
rates of complications for endoscopists who have low case
volumes.
Patient related risk factors for post-ERCP pancreatitis include
younger age, female sex, suspected sphincter of Oddi dysfunction,
prior post-ERCP pancreatitis, absence of jaundice and absence of
chronic pancreatitis. Furthermore, many of the patient-related risk
factors are cumulative. Technique-related risk factors include dif-
cult cannulation, multiple contrast injections of the pancreatic duct,
performance of pancreatic sphincterotomy, balloon dilation of the
biliary sphincter and precut sphincterotomy in inexperienced
hands.
Risk factors for hemorrhage include coagulopathy, acute cholan-
gitis, bleeding during the procedure, resumption of anticoagulation
immediately after the procedure, and endoscopist inexperience.
Although rare, perforation during ERCP is probably more
common with Billroth II anatomy, with needle-knife access sphinc-
terotomy, and in patients suspected of having sphincter of Oddi
dysfunction. Furthermore, the sequelae of perforations are some-
times magnied by delays in recognition and management. Cholan-
gitis complicating ERCP results mainly from failed or incomplete
biliary drainage, or from failure to give preprocedure antibiotics in
the patient with biliary obstruction.
Why are lawsuits led?
Data regarding lawsuits involving endoscopy are scarce and usually
relate to the more common endoscopic procedures. Information on
the total volume and causes of ERCP-related lawsuits is even more
scarce. However, an analysis of a personal series of reviews of 59
cases of alleged ERCP malpractice is particularly instructive as to the
reasons for ERCP lawsuits.30 Half of the cases involved pancreatitis;
16 sustained perforation after sphincterotomy (8 of which involved
precutting), and 10 had severe biliary tract infection; 15 of the
patients died. In this series, the most common allegation, in 32
cases, was that the procedure or associated therapeutic maneuver
was not indicated. This was also a secondary allegation in another
16 patients. These allegations were based on inadequate or weak
evidence for biliary or pancreatic pathology. The second commonest
allegation, in 19 cases, was that procedures were performed negli-
gently. This was a secondary allegation in 4 additional cases. Poor
post-ERCP care was alleged in a total of 15 cases and inadequate
informed consent was cited as a secondary allegation in 15. The
author and accompanying editorials3032 concluded that ERCP carries
a risk of life-threatening complications, that it should be performed
only for good indications and after non-invasive techniques have
been exhausted, and only with truly informed consent. Also,

precutting for marginal indications is particularly hazardous. Finally,
it was stated that Speculative ERCP, sphincterotomy, and precuts,
are high risk for patients and for practitioners.
STRATEGIES TO AVOID COMPLICATIONS
AND LAWSUITS
A risk-factor assessment may help the endoscopist decide whether
or not to perform ERCP and, if ERCP is undertaken, aids in making
decisions regarding the techniques to be utilized.
Indications
Patients with marginal indications for ERCP are most likely to
develop complications,26 or as stated by Peter Cotton in a 2001 edito-
rial, ERCP is most dangerous for people who need it least.33 Based
on his more recent analysis of lawsuit reviews,30 he would likely add
and they are most likely to sue. As a risk management strategy,
marginal cases should be avoided and alternative imaging
techniques such as endoscopic ultrasonography (EUS), magnetic
resonance cholangiopancreatography (MRCP), or intraoperative
cholangiography should be utilized instead. Although marginal
cases are admittedly difcult to dene, some clinical settings that
t this description include patients with vague abdominal pain
without objective evidence of biliary or pancreatic pathology, mildly
abnormal serum liver chemistries without jaundice, precholecystec-
tomy, or suspected sphincter of Oddi dysfunction.
Technique
Useful technique-related risk management strategies exist in addi-
tion to patient- or indication-related risk factors for complications,
particularly for post-ERCP pancreatitis. These include avoiding mul-
tiple injections of the pancreatic duct and avoiding either pancreatic
sphincterotomy or precut sphincterotomy in inexperienced hands.
Regarding post-sphincterotomy hemorrhage, attention to pre-
existing coagulopathy is important, as is delaying restarting anti-
coagulants for at least three days after sphincterotomy. Avoiding
sudden precipitous zipper cuts is also important, as is injecting
the edges of the sphincterotomy if any bleeding is detected or
coagulopathy is present. With cholangitis, care should be taken not
to inject an obstructed segment if the endoscopist is unprepared
to establish drainage. Some feel that it is helpful to aspirate the
obstructed or infected segment before injecting contrast. Careful
stent selection to maximize successful drainage is also important,
as are pre- and post-procedure antibiotics. The risk of perforation
can be lowered by avoiding repeated hook and pull maneuvers
with the duodenoscope, difcult anatomies (Billroth II or Roux-en-Y
gastroenterostomies, minor papilla cannulations, short intramural
segments), and riskier pathologies (sphincter of Oddi dysfunction,
multiple large stones or snare removal of ampullary neoplasms).
Of course, good technique is important for the sphincterotomy
including appropriate length, good control and orienting the cut in
the 11:00 to 1:00 oclock position. One of the big problems with
perforation is delayed diagnosis, so a high index of suspicion and
early evaluation with CT scanning with oral contrast is important.
Surgery should be considered for large leaks or uid collections, in
patients with severe pain, guarding or crepitus, with signs or symp-
toms of major sepsis, or in the clinical situation of deterioration or
no improvement within the rst 24 hours. Risk of cardiovascular
complications is decreased by remembering that these are the
Chapter 1 Medicolegal Issues
leading cause of death from ERCP.34 One should not be hesitant
to use an anesthesiologist for assistance in appropriate clinical
situations.
As a further risk management strategy, if ERCP is performed in
high-risk patients such as younger patients, women with normal
bilirubin or possible sphincter of Oddi dysfunction, prolonged
attempts at cannulation and other high-risk maneuvers should be
avoided and the placement of a pancreatic stent should be consid-
ered in appropriately experienced hands.26 Pancreatic stents appear
to be particularly helpful in preventing post-ERCP pancreatitis in
patients with sphincter of Oddi dysfunction and in those who have
undergone pancreatic sphincterotomy or precut access sphinctero-
tomy. However, such stenting has not been proven to be safe or
effective in less experienced hands. Furthermore, needle-knife
access sphincterotomy is probably best avoided entirely by the less
experienced.
A nal word on technique concerns contrast allergies. The appro-
priate risk management strategy is simply to have and adhere to a
specic policy regarding contrast allergies. Short-cutting the policy
regarding contrast allergies simply for patient convenience may well
result in an adverse outcome and would be extremely hard to defend
in court. Either not having a policy (ignorance of the problem or
having not dealt with it) or ignoring a well-thought-out policy would
be indefensible.
Informed consent
As described earlier, obtaining informed consent is a process which
extends beyond the actual signing of a signature on a standardized
form. It includes assessing the competence of the individual, disclo-
sure of appropriate information to allow an informed decision, and
ensuring that the decision made by the patient is voluntary. It is a
communication and decision-making process which can cement the
physicianpatient relationship and serve as a risk management tool.
Within the realm of gastrointestinal endoscopy, this management
of risk is especially important with ERCP. Transferring some respon-
sibility for the appreciable risks to the patient, who has understood
and accepted that even competently performed ERCP can result in
severe adverse outcomes, is a signicant risk management tool. The
possibility of severe adverse outcomes with ERCP is so real that they
cannot be glossed over or minimized if the patient is to be truly
informed and the decision making is to be truly shared. Some even
advocate disclosing personal experience level and outcomes in the
consent process.27
Documentation
Good documentation is an essential risk management tool and also
a component of good medical care. The old adage If it isnt written
down, it didnt happen is often true in litigation. If a discussion is
well documented in the notes, a court will generally accept that
something was discussed or did occur. Conversely, lack of documen-
tation may persuade a court or jury that a conversation did not occur
if the patient denies it. Written documentation has generally been
viewed as adequate, although some have advocated videotaping the
consent process as an extreme method of documenting exactly what
occurred.7 Documentation of informed consent is extremely impor-
tant with ERCP. A simple note afrming that discussions occurred
regarding the nature of the procedure, alternatives, risks (enumerat-
ing the major risks), sedation, and an opportunity for questions goes
far in assuring that the physician has fullled this part of the duty
to the patient and that the patient has participated in the decision-
9
 SECTION 1 GENERAL TOPICS
making process, thereby accepting some of the risk of the
procedure.
Managing complications
ERCP is difcult and complex. Complications will occur even
with good patient selection and good technique. The risk is real
and unavoidable. Complications happen with all endoscopies
and it is important that they be handled well. The fact that a
complication has occurred is not malpractice in and of itself.
The failure to make a timely diagnosis of the complication could
be, however. It is important that the endoscopist recognize the
possibility of post-ERCP complications and be attentive to their
signs and symptoms. Early diagnosis and treatment are often vitally
important.
As important as early diagnosis and treatment is communication
with the patient and family. A full explanation of the situation and
the plan of treatment is necessary. Empathy, compassion, and
honesty are essential. But this does not require self-denigration or
an unnecessary admission of wrongdoing. Complications are a
known risk of ERCP and informed consent has allowed the patient
to accept and share in that risk.
During the treatment of complications, it is important to main-
tain contact with the patient and family so as not to engender feel-
ings of abandonment and further to ensure that all necessary
consultations are obtained expeditiously. Although often personally
uncomfortable for the physician, such continued contact with the
patient and family promotes good medical care, demonstrates com-
passion and helps manage legal risk. Finally, it is important to
inform the insurance carrier following any major complication or
signicantly litigious situation.
A SUMMARY OF RECOMMENDATIONS
Freemans analysis has suggested that avoidance of marginal indica-
tions for ERCP may be the single best way to avoid complications.26
Cottons article on ERCP-related lawsuits suggests that avoidance of
marginal indications also may be the single best way to avoid medical
litigation.30 These insights plus recent opinions on medicolegal
issues31,32 suggest the following risk management strategies to avoid
ERCP-related complications and lawsuits:
1. Practice within a reasonable standard of care and avoid risky
cases. See Box 1.1.
2. Use good ERCP technique, avoiding advanced ERCP techniques
if inexperienced and utilizing pancreatic duct stents for high risk
patients if appropriately experienced.
3. Utilize informed consent as a communication and decision-
making process to transfer some responsibility for risk to the
patient, who has understood and accepted that even competently
performed procedures can have adverse outcomes.
4. Employ good documentation of clinical events, decision making
and informed consent.
5. Be vigilant to assure early recognition and management of com-
plications, communicating honestly and with empathy and com-
passion with the patient and family, and maintaining contact with
them throughout the post-complication period.
Drs. Freeman and Cotton have persuasively demonstrated
that to minimize the risk of ERCP-related complications and
lawsuits one should avoid ERCP that is marginally indicated, too
difcult for ones experience level or if ERCP is performed
infrequently.
10
BOX 1.1 RISKY CLINICAL SITUATIONS
FOR ERCP-RELATED COMPLICATIONS
AND LAWSUITS
Marginal cases
Vague abdominal pain without objective evidence of biliary
or pancreatic pathology
Mildly abnormal serum liver chemistries without jaundice in
the pre-cholecystectomy patient
Suspected sphincter of Oddi dysfunction
Technically difcult cases
Past history of difcult cannulation or post-ERCP pancreatitis
Precut (access) sphincterotomy
Minor papilla cannulation
Percutaneous biliary access
Billroth II or roux-en-y gastroenterostomy
Obstruction at the biliary bifurcation
Infrequent cases
Sphincter of Oddi manometry
Pancreatic therapeutics
Large or intrahepatic stones
Snare ampullectomy
LIMITING FINANCIAL AND PSYCHOLOGICAL
COSTS OF LAWSUITS
Insurance
Every physician should have adequate indemnity coverage for law-
suits that occur now or in the future. Keeping adequate malpractice
insurance is important to ensure that injured patients have access
to compensation if they are inappropriately injured and to safeguard
the nancial well-being of the individual physician and institution.
If a lawsuit is led
The possibility of being sued is unavoidable. Nothing, not even
superb medical practices, will absolutely protect one from a lawsuit.
The severity of adverse patient outcomes is a more important predic-
tor of plaintiff success in winning a malpractice claim than whether
the physician was indeed negligent.35 Sympathetic juries and judges
often wish to reward an injured plaintiff irrespective of actual negli-
gence. It is important that physicians, particularly those involved in
high-risk procedures such as ERCP, accept realistic goals and realize
that lawsuits may occur.
It is also important to remember that medical litigation is a long
and stressful process for physicians and their families. Psychologic
support is important, but casual discussion of lawsuits is inadvisable
since discussions with friends and colleagues can be elicited through
subpoena. It is also heartening to note that physicians win the major-
ity of malpractice suits. In Cottons recent analysis of 59 ERCP
lawsuit cases, where the nal outcome was available in 40, 16 were
withdrawn, 14 were settled, and of the 10 going to trial, half were
decided for the plaintiff and half for the defense.30
Obviously if a lawsuit is led, the physicians insurance carrier
must be notied and it is the responsibility of the physician to aid
the insurance company in the defense of the case.

Physicians should become educated in the litigation process.
Depositions generally benet plaintiffs and are used as tools to gain
information to help press cases against defendants. During deposi-
tion, physicians should be truthful but volunteer as little information
as possible. Excessive elaboration can only harm the defendant. In
deposition and at trial demeanor is important. Arrogance, anger, or
dismissive behavior will only reect poorly upon the physician and
engender sympathy for the plaintiff. It is also important to be well
prepared for any type of testimony. Malpractice attorneys are usually
intelligent and well versed. Physicians should be similarly knowl-
edgeable and well prepared and should be deliberate in their testi-
mony, pausing before every answer to ensure a proper and accurate
response.
SUMMARY
ERCP is a complex and difcult procedure with signicant risks
for adverse patient outcomes and for medical litigation. It is
important for endoscopists performing ERCP to understand
liability issues related to ERCP, including vicarious liability, and
to understand the legal principles important in medical
practice, including the elements of a malpractice case,
REFERENCES
1. Feld AD, Walta D. Malpractice, tort reform, and you: an
introduction to risk management. Am J Gastroenterol 2004;
99(2):977980.
2. Physician Insurers Association of America: PIAA risk management
review: Gastroenterology. Rockville, MD, Research Department,
Physician Insurers Association of America, 2005.
3. Physician Insurers Association of America: A risk management
review of malpractice claims: gastroenterology summary report.
Rockville, MD, Research Department, Physician Insurers
Association of America, 2000.
4. Gerstenberger PD. Malpractice in gastrointestinal endoscopy.
Gastrointest Endosc Clin N Am 1995; 5(2):395389.
5. Springer J. ERCP: new standards, new challenges. Can J
Gastroenterol 2002; 16(3):175176.
6. American Society for Gastrointestinal Endoscopy. Medical
Malpractice. In: Petrini JL, Feld AD, Gerstenberger PD, et al. (eds.)
Risk management for the GI endoscopist. Oak Brook, IL:
American Society for Gastrointestinal Endoscopy; 2001:1013.
7. Norton I, Schroeder KW. Reporting, documentation, and risk
management. In: Ginsberg GG, Kochman M, Norton I. Gostout CJ,
eds. Clinical Gastrointestinal Endoscopy. London: Elsevier Science;
2005:107112.
8. American Society for Gastrointestinal Endoscopy. Standards of
care. In: Petrini JL, Feld AD, Gerstenberger PD, et al. (eds.) Risk
management for the GI endoscopist. Oak Brook, IL: American
Society for Gastrointestinal Endoscopy; 2001:23.
9. American College of Physicians. Guidelines for the physician
expert witness. Ann Intern Med 1990; 113:789.
10. Greene ML. Medical-legal consultation in gastroenterology.
Gastrointest Endosc Clin N Am 1995; 5(2):403419.
11. Gerstenberger PD. Risk management for the endoscopy center.
Gastrointest Endosc Clin N Am 2002; 12(2):367384.
12. American Society for Gastrointestinal Endoscopy. Vicarious
liability. In: Petrini JL, Feld AD, Gerstenberger PD, et al. (eds.) Oak
Chapter 1 Medicolegal Issues
standards of care and informed consent. The pancreaticobiliary
endoscopist must realize that the deviations from a reasonable
standard of care most likely to lead to ERCP related medical
litigation include procedural indications, procedural technique,
post-ERCP care and issues of informed consent. With this
understanding, the endoscopist performing ERCP can
formulate and adopt risk management strategies to improve
patient safety, satisfaction and outcomes while minimizing the
risk of litigation. These strategies include practicing within a
reasonable standard of care, focusing attention on the
informed consent process and documentation, and
understanding specic patient-related and technique-related
risk factors for complications and lawsuits (Box 1.1).
Acknowledgements
The author is pleased to acknowledge Martin L. Greene, MD for help
with access to data from the Physicians Insurers Association of
America, Arnold M. Rosen, MD for constructive suggestions in
editing the manuscript, and Brenda Paulson for preparation of the
manuscript.
Brook, IL: American Society for Gastrointestinal Endoscopy;
2001:1921.
13. Sundermeyer MS, Murphy PA. Prevention of hospital liability for
granting privileges to unqualied physicians. Gastrointest Endosc
Clin N Am 1995; 5(2):433445.
14. Plumeri PA. Endoscopic training directors: a few legal and ethical
considerations. Gastrointest Endosc Clin N Am 1995; 5(2):447455.
15. American Society for Gastrointestinal Endoscopy and American
College of Gastroenterology. Hospital liability update. In: Faigel
DO, Baron TH, Lewis B, et al. (eds.) Ensuring competence in
endoscopy. Oak Brook, IL: American Society for Gastrointestinal
Endoscopy; 2005:1531.
16. Lofft AL. Informed consent for endoscopy. Gastrointest Endosc
Clin N Am 1995; 5(2):457470.
17. American Society for Gastrointestinal Endoscopy. Informed
consent. In: Petrini JL, Feld AD, Gerstenberger PD, et al. (eds.)
Oak Brook, IL: American Society for Gastrointestinal Endoscopy;
2001:79.
18. Feld AD. Informed consent: not just for procedures anymore.
Am J Gastroenterol 2004; 99(6):977980.
19. Berg JW, Appelbaum PS, Lidz CW, et al. Informed consent: legal
theory and clinical practice. Oxford: Oxford Press; 2001:3227.
20. Berg JW, Appelbaum PS, Lidz CW, et al. Informed consent:
legal theory and clinical practice. Oxford: Oxford Press;
2001:5354.
21. American Society for Gastrointestinal Endoscopy and American
College of Gastroenterology. Ensuring competence in endoscopy:
a primer. In: Faigel DO, Baron TH, Lewis B, et al. (eds.) Ensuring
competence in endoscopy. Oak Brook, IL: American Society for
Gastrointestinal Endoscopy; 2005:515.
22. American Society for Gastrointestinal Endoscopy. ASGE policy
and procedures manual for gastrointestinal endoscopy:
guidelines for training and practice. Online. Available:
http://www.asge.org. 16 Oct 2005.
11
 SECTION 1 GENERAL TOPICS
23. Jowell PS, Baille J, Branch S, et al. Qualitative assessment of
procedural competence: a prospective study of training in
endoscopic retrograde cholangiopancreatography. Ann Intern
Med 1996; 125:983989.
24. Freeman ML. Training and competence in gastrointestinal
endoscopy. Reviews in Gastroenterological Disorders 2001;
1(2):7385.
25. Petersen BT. ERCP outcomes: dening the operators,
experience and environments. Gastrointest Endosc 2002;
55(7):953958.
26. Freeman ML. Adverse outcomes of endoscopic retrograde
cholangiopancreatography. Reviews in Gastroenterological
Disorders 2002; 2(4):147168.
27. Cotton PB. How many times have you done this procedure,
doctor? Am J Gastroenterol 2002; 97(3):522523.
28. Hickson GB, Federspiel CF, Pichert JW, et al. Patient complaints
and malpractice risk. JAMA 2002; 287:29512957.
12
29. Levinson W, Roter DL, Mullooly JP, et al. Physician-patient
communication: the relationship with malpractice claims among
primary care physicians and surgeons. JAMA 1997; 277:553559.
30. Cotton PB. Analysis of 59 ERCP lawsuits; mainly about indications.
Gastrointest Endosc 2006;63:378382.
31. Peterini JL. Fools rush in . . . Gastrointestinal Endosc 2006;63:
383384.
32. Frakes JT. The ERCP-related lawsuit: Best avoid it! Gastrointest
Endosc 2006;63:385388.
33. Cotton P. ERCP is most dangerous for people who need it least.
Gastrointest Endosc 2001; 54:535536.
34. Freeman ML. Sedation and monitoring for gastrointestinal
endoscopy. In: Yamada T, ed. Textbook of Gastroenterology.
Philadelphia, PA: Lippincott, Williams & Wilkins; 1999:26552667.
35. Brennan TA, Sox CM, Burstin HR. Relation between negligent
adverse events and the outcomes of medical-malpractice
litigation. N Engl J Med 1996; 335(26):19631967.
 SECTION 1 GENERAL TOPICS
Chapter
The ERCP Room
2 Michael B. Kimmey
INTRODUCTION
Physicians, staff and patients all appreciate the benets of an ERCP
procedure performed in a pleasant and efcient working environ-
ment. A well-planned procedure room can contribute to a successful
and comfortable procedure for the patient and the medical staff.
In this chapter, the components of the ideal ERCP room will be
reviewed.
Most ERCPs are performed in the hospital, usually in the outpa-
tient department. This facilitates doing the procedure on hospital-
ized patients and allows ready access to inpatient hospital facilities
for patients who need observation following the procedure. Although
ERCPs can be performed in ambulatory surgery centers, current
reimbursement policies in the United States do not include ERCP
as an approved procedure for this setting. In addition, the higher
capital costs for radiographic equipment and the ongoing costs of
single-use ERCP accessories discourage the performance of ERCPs
in this environment.
The ERCP room can be a part of a gastrointestinal endoscopy unit
where other endoscopic procedures are performed, or may be part
of the hospitals radiology department. The choice of location is
often dependent on the number of ERCP cases done at the facility.1
The efciencies of having the ERCP room in the endoscopy unit
should not be overlooked. These include time savings for the physi-
cians, nurses and technical personnel as well as reduction in time
spent transporting patients and equipment. The nancial benet of
these time savings is likely to justify the capital cost of the uoros-
copy equipment for endoscopy units doing over 300400 procedures
per year. In addition, uoroscopy can be used for other endoscopic
procedures such as dilation of complex strictures, placing esopha-
geal and enteral stents, and facilitating overtube placement for
enteroscopy.
DESCRIPTION OF THE ROOM
The ERCP room should be large enough to accommodate all of the
necessary equipment to do ERCP as well as a comfortable space for
the patient, physician and two endoscopy assistants. When ERCP
needs to be performed under deep sedation or general anesthesia
due to patient intolerance of conscious sedation, there should also
be sufcient space for anesthesia equipment and personnel. If uo-
roscopy is performed using a C-arm, there needs to be sufcient
room around the patient table to allow the image intensier arm to
move without being encumbered by other equipment in the room.
All of these factors add up to a need for the ERCP room to be
50100% larger than procedure rooms where routine endoscopy is
performed. A room size of approximately 300 square feet has been
recommended by some authorities.2,3
The ERCP room is best congured around the location of the
uoroscopy equipment and patient table. This is usually placed cen-
trally in the room to allow access to both sides of the table and to
allow adequate room for movement of the uoroscopy units C-arm.
A more complete description of the options for uoroscopy
equipment, including the use of spatially xed or mobile units can
be found below. Figure 2.1 shows a generic diagram of a typical
ERCP room layout and an actual ERCP room in use is shown in
Figure 2.2.
The patient table should be at least 30 inches wide to accommo-
date large patients and also allow safe patient repositioning during
the procedure. It should also be sturdy enough to accommodate
heavy patients, preferably to at least 400 pounds. There should be
room under the head of the table for the assistants knees while
sitting at the patients head and attending the patients airway. There
should be adequate shielding below the table to reduce exposure of
personnel to scattered radiation. Tables that are xed to the uoro-
scope should have a movable top for positioning the patient under
the uoroscope. Some physicians also prefer to have the table tilt
to facilitate gravitational movement of contrast and air bubbles,
although modern ERCP techniques and accessories have made this
requirement less important than previously. Controls for table move-
ment should be readily accessible to the endoscopist who should be
able to move the table top while simultaneously holding the endo-
scope, if an assistant or x-ray technologist is not available.
Appropriate placement of uoroscopic and endoscopic monitors
is very important. Monitors can be mounted on movable carts or
booms that are xed to the ceiling of the room. Poor monitor place-
ment has been identied as a major contributor to endoscopist
musculoskeletal injuries.4 Monitors should be set so that the endos-
copists eye level is three-quarters toward the top of the monitor
screen. Since not all endoscopists are the same height, provision for
adjusting monitor position is necessary. Monitors should be placed
on the opposite side of the patient, generally at a level above the head
of the table directly in front of the endoscopist so that neck rotation
and strain are avoided. Monitors should not restrict movement of
the C-arm of the uoroscopic unit and should not be obstructed by
the airway nurse or anesthesiologist. Ideally, endoscopic and uoro-
scopic image monitors should be mounted side by side. A second
uoroscopic monitor that displays the last captured radiographic
image is optional. An additional monitor that can be used inter-
changeably for intraductal ultrasound, cholangioscopy and sphincter
of Oddi manometry is also appropriate for units that perform these
specialized procedures (Fig. 2.3). Older monitors built with cathode
ray tube technology are being increasingly replaced with at screen
monitors as the quality of these smaller at screen monitors is
rapidly improving.
Vital sign monitors are usually placed at a lower level for primary
viewing by the seated endoscopy assistant or nurse, although most
13
 SECTION 1 GENERAL TOPICS

Accessory cart A B

B
Monitors

Endoscope
processor
Patient C
D
C-arm

Sink Fig. 2.3 This photograph shows A a ceiling-mounted boom

holding monitors for uoroscopy, B the endoscopic image, C acces-
sories such as endoscopic ultrasound, and D patient vital signs.

Lead aprons

Storage cabinets

Fig. 2.1 A schematic diagram of a typical ERCP room is shown,

indicating the positions of personnel including A the endoscopist,
B airway nurse, C accessory assistant, and endoscopic trainee,
D radiology technician.

Fig. 2.4 Endoscopic processors and the endoscope are shown on

this boom.

endoscopists also prefer to be able to observe vital signs intermit-

tently when there are any concerns raised by the assistant.
B The endoscope processor, light source and electrosurgery genera-
D tor are generally kept together on a cart or boom that is behind the
A C endoscopist (Fig. 2.4). The top shelf of the cart is best left open to
use as a working surface. The endoscope can either be laid on top
of the cart when not in use, or hung on the side of the cart. A water
pump for irrigation is another useful tool that can be kept on this
cart or stand.
There are a number of types of electrosurgery generators that can
be used during ERCP. Monopolar current is commonly used for
Fig. 2.2 An ERCP is being performed by A the endoscopist,
assisted by B the airway nurse, C accessory assistant, and D GI sphincterotomy; however, bipolar current should also be available
trainee. for the control of bleeding. Generators that contain both monopolar

14
 Chapter 2 The ERCP Room

and bipolar cautery are preferred to reduce the number of generators
on the cart and conserve space. Recently, an electrosurgical genera-
tor that uses a microprocessor to automatically adjust current has
become popular in ERCP rooms, and may reduce the frequency of
zipper cuts and endoscopically visible but not necessarily clinically
signicant post-sphincterotomy bleeding.5
RADIOLOGIC IMAGING EQUIPMENT
Radiologic imaging will also be addressed in Chapter 3. ERCP
involves the use of uoroscopy to monitor contrast injections and to
guide the use of accessory devices that are inside the pancreatic or
biliary ducts. Image capture and storage capability is also necessary
to document radiographic ndings for patient care and follow-up.
The key components of the radiographic equipment include the x-
ray tube, image intensier, monitor, table, and hardware for image
capture and storage. These components will be described and their
critical specications discussed. A comparison of commercially
available radiographic equipment can be found in Table 2.1.
Fixed versus portable equipment
The rst decision to be made about the radiographic equipment is
whether the uoroscope and image intensier are xed to the patient
table or whether a portable or mobile radiographic unit is used in
conjunction with a freely movable table or stretcher for the patient.
The main advantage of a portable radiographic unit is that it is gener-
ally of lower cost than a xed unit. This allows mobility to other parts
of the GI suite and hospital so that the capital costs can be shared
and its use maximized. Hence, it is an option to be considered for
low volume ERCP units. In addition, patients can be moved in and
out of the ERCP room on the same stretcher that they lie on during
the procedure. The main disadvantage of a mobile radiographic unit
is it requires a separate operator or technician to be present during
the ERCP procedure. The portable unit or patient stretcher must be
moved during a procedure, and imaging parameters changed using
controls that are not easily accessible to the endoscopist. In addition,
radiation exposure to staff and patients is substantially higher with
the mobile units due to difculties providing adequate shielding and
generally higher power outputs.6 These portable units formerly had
inferior image quality compared to xed units, however equipment
is currently available that provides image quality that is comparable
to xed units.
Digital vs conventional imaging
Most modern x-ray imaging units use digital techniques for image
manipulation and capture, although older conventional radiographic
units are still used by some hospitals. Digital imaging has several
advantages over conventional analog imaging including better image
quality, the capability of image manipulation after acquisition, and
easier and less expensive image storage and transfer. In addition,
radiation exposure is reduced because less radiation is needed to
produce an image. Image quality in pancreatography has been
shown to be better for digital over conventional imaging, especially
for evaluating side branch abnormalities.7
Imaging components and specications
The tube that generates the x-rays is usually placed under the xed
table or at the bottom of the mobile C-arm. X-rays then go up
through the patient and are either scattered or penetrate through to
be captured by the image intensier which is placed over the patient.
The size of the image is determined by the diameter of the image
intensier. For ERCP, a 12 or 14 inch image intensier is ideal so
that the entire biliary tree or pancreatic duct can be visualized on
one screen without patient or uoroscope movement. The image
intensier is also the most important component for achieving
optimal image resolution, an important feature for seeing small
diameter guidewires.8
Staff exposure to radiation is mostly due to scatter of the x-ray
beam as it passes through the patient.911 This exposure can be
greatly reduced with adequate shielding around the image intensi-
er and also below the table top (Fig. 2.5). Thin lead shields can be
custom tted to most parts of the image intensier and table to block
the scattered radiation.
Images can be captured on radiographic spot lms or increas-
ingly as digital images. The former requires manual placement of
x-ray cassettes and subsequent lm development or conversion to
digital images. Newer units capture digital images directly, reducing
the need for x-ray technologists and making the images available for
viewing immediately.
Controls
Ideally, all controls for uoroscopy and image capture should be
within close reach of the endoscopist. This allows the endoscopist
to perform all aspects of the ERCP procedure without assistance
from a radiologist or technician. Newer equipment makes this quite

Image
PACSb intensier
Minimum Mobile C-arm Digital Compatible diameter
Manufacturer Model room size (Y/N) (Y/N) (Y/N) (Y/N) (inches) Resolution List pricea
General Electric/OEC 9900 elite Variable Y Y Y Y 12 1000 1000 $179 600
Omega CS-50 16 20 N Y Y Y 14 1024 1024 $650 000
Philips Pulsera Variable Y Y Y Y 12 1000 1000 $175 000
Philips Easy Diagnost Eleva 15 21 N N Y Y 15 1024 1024 $511 600
Philips MultiDiagnost Eleva 19 22 N Y Y Y 15 1024 1024 $933 000
Siemens Iso-C 10 10 Y Y Y Y 12 1024 1024 $150 000
Siemens Sireskop SD 16 20 N N Y Y 16 1024 1024 $435 000
Siemens Artis MP 16 20 N Y Y Y 16 1024 1024 $735 000

Table 2.1 Comparison of radiographic equipment used for ERCP

a
Manufacturers list price in October 2005; actual cost will vary with local purchasing contracts.
b
Picture Archiving and Computerized Storage system.

15
 SECTION 1 GENERAL TOPICS
Fig. 2.5 Lead shields mounted on A the image intensier and
B patient table reduce radiation exposure to staff from scattered
radiation.
feasible and many busy units no longer involve radiology personnel
during the procedure. Older units and mobile units usually require
the presence of a technologist to set up the examination, make
changes to x-ray parameters during the procedure and in some cases
move the mobile C-arm unit.
Image capture and storage
Most current hospitals store radiographic images using a Picture
Archiving and Computerized Storage (PACS) system. Images
obtained during ERCP are ideally transferred to the PACS system
for access by referring physicians, and in some settings, review and
interpretation by a radiologist. Using the central hospital PACS
system is preferred over local storage within the endoscopy unit for
ease of access and to avoid the need for endoscopy unit quality con-
trols for storage and retrieval of images. A local computer hard drive
to store images temporarily and a writable CD drive can be useful
for downloading images for educational conferences, however.
Radiographic hardware
Fixed uoroscopy units usually have housing units for the x-ray tube
and associated hardware. These units can take up considerable space
in the procedure room which needs to be included in planning room
dimensions. In addition, a console for entering individual patient
data and setting imaging parameters is usually placed outside the
ERCP room or behind a leaded glass shield to reduce radiation
exposure to technical personnel. If placed in a separate room, the
presence of a window for the technologist or assistant to see into the
room is desirable (Fig. 2.6).
Equipment and device storage
Accessories are integral to the practice of ERCP. A range of
accessories are needed for ERCP (Table 2.2; see also Chapter 4) and
should be readily available during each procedure. There are a
number of options for storage of these accessories. Mobile carts are
available for this purpose or customized cabinets or shelves within
the room can be used (Fig. 2.7). Accessories should be clearly labeled
for quick access during a procedure. A regularly updated inventory
16
Fig. 2.6 A console for entering patient data is outside the ERCP
room. The procedure can be viewed by technical personnel
through a leaded glass window, thereby reducing their radiation
exposure.
system is also essential to avoid the problem of not having a needed
accessory for a specic procedure.Endoscopes are generally stored
with the other endoscopes for the endoscopy unit, outside of the
procedure room. The endoscopes should hang free and not be coiled
within an instrument case. Air and suction valves should be removed
from the endoscopes during storage.
Lead aprons and thyroid collars should be stored near a door,
either inside or outside of the room. This minimizes disruption of
uoroscopy when someone enters the room and needs to put on
lead shielding.
MISCELLANEOUS ISSUES
Each staff member involved with an ERCP procedure should have a
radiation dosimeter. Provision should be made for storage of these
dosimeters outside of the procedure room to avoid recording of scat-
tered radiation that is not associated with the individual to whom
the dosimeter is assigned. The dosimeter should be in a convenient
location to serve as a reminder to staff to wear the dosimeter during
each procedure.Drugs used during the procedure should be readily
available. Sedative medications are usually checked out for each
individual patient from a central medication storage area. Emer-
gency drugs, such as atropine, naloxone, and umazenil need to be
readily available for each case and not be locked up in an area that
delays access to them in an emergent situation.
 Chapter 2 The ERCP Room

Essential accessories for all ERCP rooms Additional accessories for specialty centers
Electrosurgery generator Manometry catheters
Ground pads Intraductal ultrasound catheters
Guidewires (0.0180.035 inch) Intraductal baby scopes
Cannulas Electrohydraulic or laser lithotriptor bers
Sphincterotomes Argon plasma coagulation probes
Stents
Plasticrange of diameters (310 French) and lengths (315 cm)
Self-expanding metal
Stone retrieval balloons (8.515 mm balloon diameters)
Stone retrieval baskets
Mechanical lithotriptor and catheters
Dilating balloonsrange of diameters (410 mm) and lengths (26 cm)
Dilating catheters
Brush cytology catheters
Intraductal biopsy forceps
Snares and foreign body forceps for stent retrieval
Injection needles
Multipolar electrocautery probes

Table 2.2 Accessories for the ERCP room

Fig. 2.7 A cart holding the accessories commonly used during

ERCP should be stored nearby the procedure table.

REFERENCES
1. Lennard-Jones JE, Williams CB, Axon A, et al. The Working Party
of the Clinical Services Committee of the British Society of
Gastroenterology Provision of gastrointestinal endoscopy and
related services for a district general hospital. Gut 1991;
32:95105.
2. Sivak MV Jr, Manoy R, Rich ME. The endoscopy unit. Chapter in:
Gastroenterologic Endoscopy, 2nd edn. Edited by MV Sivak;
Philadlelphia: WB Saunders, 1999.
3. Gostout CJ, Ott BJ, Burton D, et al. Design of the endoscopy
procedure room. Gastrointest Endosc Clin N Am; 1993:509524.

17
 SECTION 1 GENERAL TOPICS
4. OSullivan S, Bridge G, Ponich T. Musculoskeletal injuries among
ERCP endoscopists in Canada. Can J Gastroenterol 2002;
16:369374.
5. Perini RF, Sadurski R, Cotton, PB, et al. Post-sphincterotomy
bleeding after the introduction of microprocessor-controlled
electrosurgery: does the new technology make the difference?
Gastrointest Endosc 2005; 61:5357.
6. Johlin FC, Pelsang RE, Greenleaf M. Phantom study to determine
radiation exposure to medical personnel involved in ERCP
uoroscopy and its reduction through equipment and behavior
modications. Am J Gastroenterol 2002; 97:893897.
7. Uomo G, deRitis R, Rabitti PG, et al. Does endoscopic digital
pancreatography constitute an advance in pancreatic imaging?
Gastrointest Endosc 1998; 48:6771.
18
8. Bushong SC. Radiologic science for technologists: physics,
biology, and protection. 8th edn. St Louis: Mosby, 2004.
9. Chen MYM, Van Swearingen FL, Mitchell R, et al. Radiation
exposure during ERCP: effect of a protective shield. Gastrointest
Endosc 1996; 43:15.
10. Campbell N, Sparrow K, Fortier M, et al. Practical radiation safety
and protection for the endoscopist during ERCP. Gastrointest
Endosc 2002; 55:552557.
11. Buls N, Pages J, Mana F, et al. Patient and staff exposure during
endoscopic retrograde cholangiopancreatography. Br J Radiol
2002; 75:435443.
 SECTION 1 GENERAL TOPICS
Chapter
Radiologic Issues in ERCP
3 Desiree E. Morgan
The scope and practice of ERCP has changed dramatically over the
past 10 years. With the advent of MRCP, in academic and in private
settings alike, the proportion of therapeutic to diagnostic ERCP cases
has greatly increased,1,2 This shift in practice patterns heightens the
importance of communication between the endoscopist and radiolo-
gist in order to provide the patient with the best possible interpreta-
tion of images acquired during ERCP. It cannot be stressed enough
that the interaction between physicians leads to the most accurate
and consistent reporting,3 whether the patient is undergoing a diag-
nostic or a therapeutic ERCP examination. This chapter will focus
primarily on techniques of image acquisition during ERCP, using
examples of pathology (discussed in depth elsewhere in the book) to
demonstrate imaging principles.
Prior to the ERCP procedure, review of other imaging studies
(CT, MRI, or ultrasound) is often helpful to plan and expedite the
case. If therapeutic endoscopic interventions are intended, availabil-
ity of the pancreatobiliary surgeon or vascular/interventional radiol-
ogist for treatment of potential complications or participation in
combined procedures is desirable. The patients history of medical
allergies, including contrast material, should be ascertained during
the consent process.
ERCPs may be performed in dedicated rooms with xed uoros-
copy (conventional or digital system) units, or with C-arm uoro
units in endoscopy suites or outpatient centers. While the benets
of lower expense, convenience and versatility for C-arm units make
them desirable to some users, the cost of dramatically increased x-
ray dose to patients and ERCP personnel must be considered.4 Our
ERCPs take place in dedicated digital uoroscopy rooms within an
endoscopy suite located in our hospital. The images acquired in the
endoscopy uoro rooms are transmitted to and archived within the
Department of Radiology PACS (picture archiving and communica-
tion system), and the digital images are available throughout the
hospital and clinics of our medical center once they are transmitted
to the PACS. Fluoroscopy rooms designed for ERCP should be large
enough to accommodate the video endoscope machinery, supply
cart, patient monitoring devices, and the conventional or digital x-ray
device. Digital systems allow on the y change in magnication patients, limiting uoro time is critical. Also, to reduce dose to the
factors and rapid image acquisition generally ranging from two to
seven lms per second, features often helpful during ductal evalua-
tion. Exchange of x-ray lm cassettes in conventional uoroscopy
units adds time both for acquisition and development of images, and
generally requires the assistance of a radiology technologist.
Most xed uoroscopy units, whether digital or conventional,
have the x-ray beam source in the table and an image intensier in
the tower. Consideration of the x-ray source is important when
shielding the patient is necessary, such as during ERCP in pregnant
patients. During most ERCPs, the patient is generally positioned in
either a prone or left lateral decubitus (LLD) position. Changes in
the patient position are key to visualization of both normal ducts and
duct pathology. In this chapter, the patient position will be described
relative to the tabletop rather than the image intensier. For example,
left anterior oblique (LAO) refers to prone patient with left side
angled down against the table top and right side angled up.
Radiation dose is important to patients and ERCP personnel, and
should be monitored with quality assurance testing of the equip-
ment and monthly quantication of exposure measured by radiation
badges worn by the endoscopist, his or her technical assistants, and
the nurses involved in the procedure. The amount of exposure to
personnel varies according to their position relative to the patient.
In a controlled, phantom study of radiation doses to personnel
during ERCP, Johlin et al.4 described that the largest doses are
received by the person at the head of the table, generally the nurse
who monitors the patient and administers drugs. The next highest
dose is received by the endoscopist, who stands at the right hand
corner of the uoro table, and the lowest dose is received by
the assistant who stands alongside the endoscopist at the level of the
patients abdomen. The low dose received by the assistant at the
patients abdomen is explained by the use of vertically oriented lead
drapes that attach to the uoro tower and diminish the amount of
scatter radiation. Some uoro units are equipped with right angled
vertical lead drapes which can be maneuvered to shield both the
head and side of the tower. Alternatively, some centers utilize a lead
bead curtain at the head of the table to diminish dose to personnel
in this position. Other means that should be employed to decrease
dose to patients and personnel alike during ERCP include lowering
the tower as close to the patient as possible during the procedure
(anytime the uoro is on as well as when making exposures), colli-
mating the x-ray beam manually (using buttons generally located on
the tower), and standing back from the table. The scatter radiation
is inversely proportional to the distance squared from the x-ray
beam, so even stepping back a small amount is helpful. Conversely,
leaning over the head of the patient while uoro is on greatly
increases the dose to the nurse.4 Finally, measures to conserve expo-
sure in the rst place by using minimal uoro time or pulsed uoro
will decrease the dose to all persons in the ERCP room. For pregnant
fetus, shielding of the mothers abdomen (from the direction of the
x-ray source) with lead aprons is warranted.
While the above discussion pertains to ERCPs performed with
either xed uoroscopy or C-arm equipment, it should be stressed
that the amount of scatter radiation produced by most C-arm units
is of the order of 100-fold that of xed equipment. This is due to lack
of shielding at the x-ray source and lack of shielding of the x-ray
beam that enters the patient, as well as issues of xed distance from
the x-ray source to the patient and enhanced uoro features which
double the energy of the x-ray beam for higher resolution. Readers
are referred to several articles in the radiology literature5,6 and
an excellent therapeutic ERCP specic article on this subject,7
19
 SECTION 1 GENERAL TOPICS
and are strongly urged to pay particular attention to these issues
when purchasing equipment for their endoscopy centers.
During the ERCP procedure, the orientation of the uoroscopic
image on the image intensier tower may be viewed in a number
of ways. Some endoscopists prefer to view the uoroscopic image
just as it is obtained, that is, with the prone patients left side on the
right side of the image intensier screen and the cephalad portion
projecting at the bottom of the screen (Fig. 3.1). The head of the
ERCP table equals the foot of the x-ray uoro table; in this sce-
nario, what you see is what you get. Most prefer to ip the image so
that the anatomy appears upright and in anatomic position on the
image intensier (Fig. 3.2). But when the uoro tower is moved, one
must remember this inverse relation to achieve the desired location
of the x-ray beam in the patients body. The orientation of the image
on the tower may easily be changed during the exam.
Contrast preferences also vary among endoscopists. Some prefer
to dilute contrast to half strength when looking for stones. Others
use full strength contrast injected slowly while looking stringently
for lling defects, or employ the technique of chasing the initial
contrast injection with saline to achieve a lesser opacity through
which stones may become more evident. Still others use full strength
Fig. 3.1 Direct projection of ERCP image on uoroscopic tower.
Early injection into the CBD in this patient with Mirizzi Syndrome
and cholelithiasis demonstrates orientation without adjustment.
The patients head is at the foot of the table, and he is prone, so
that the right side projects to the left of the screen.
Fig. 3.2 Vertically and horizontally ipped projection of ERCP
image on uoroscopic tower. Later injection into the CBD in the
same patient (Fig. 3.1) with Mirizzi Syndrome and cholelithiasis dem-
onstrates orientation with adjustment so that it is viewed in ana-
tomic position with the patients head at the top and his left side
on the left of the screen. Note stone (arrow) impacted in cystic duct
approximately 1 cm above its insertion into the common hepatic
duct.
20
contrast with the argument that the bile in a potentially obstructed
or dilated system (Fig. 3.3) will dilute the contrast enough to
preclude having to do so proactively. In addition, the use of full-
strength contrast allows earlier detection of pancreatic duct injec-
tion, minimizing the volume and potentially reducing the risk of
pancreatitis.
Also, with conventional radiographic lm screen combinations,
the exposure creates a white duct on black background image.
With digital images, the black on white images that appear similar
to the uoroscopic image can be lmed or viewed as such, or con-
verted to a standard white on black appearance on PACS or lm,
if preferred. In my opinion, retroperitoneal and free air (Fig. 3.4) are
easier to detect with white on black images, as are small stones
(Fig. 3.5), but to my knowledge, there has been no formal, controlled
perception study to support this theory.
The sequence of images obtained during ERCP should tell the
story of the examination, whether diagnostic or therapeutic. Initially,
a scout radiograph obtained with the patient prone reveals any resid-
ual contrast, calcications, tubes, drains, or stents already in place,
and any other material that may obscure the region of interest during
contrast injection (Fig. 3.6). Most gallstones are nonradiopaque and
will not be seen on the scout radiograph, however pancreatic stones
frequently may (Fig. 3.7). Once contrast is administered, radiopaque
stones in either ductal system may be obscured (see Fig. 3.7). The
Fig. 3.3 Spot radiograph demonstrating adequate visualization of
multiple small faceted gallstones (arrows) in a dilated Type 1 cho-
ledochal cyst.
A B
Fig. 3.4 Subhepatic free intraperitoneal air (arrow) after biliary
sphincterotomy comparing A black on white to B white on black
image presentation. Note large amount of air in the extrahepatic
duct after sphincterotomy.
 Chapter 3 Radiologic Issues in ERCP

A B A

Fig. 3.5 Small distal common bile duct stone (arrow) in dilated
system comparing A black on white to B white on black image
presentation.

Fig. 3.6 Scout radiograph in a patient about to undergo ERCP. The

radiopaque oral contrast from abdominal CT performed 12 hours
earlier is located in the hepatic exure of the colon, potentially
interfering with visualization of the extrahepatic bile duct. Because
this patient was being evaluated for suspected post traumatic (gun
shot wound) bile leak from the intrahepatic ducts more superiorly,
the exam was performed (see Fig. 3.20).

diagnostic images should include early lling as well as full duct

opacication and generally are acquired with a nine inch or six inch C
image intensier eld of view mode. These overview images of the
bile and or pancreatic ducts should be supplemented with spot
radiographs of abnormal or suspicious ndings. The spot images
may be obtained at different degrees of magnication (six or 4.5 inch
mode) for emphasis. Delayed lms often are critical for assessing
biliary drainage (Fig. 3.8) or lack thereof, or for demonstrating small
calculi. A nal lm obtained with a large image intensier eld of
view (12 or 15 inch) is helpful to evaluate potential retroperitoneal
(Fig. 3.9) or intraperitoneal air.
For calculation of actual duct size with digital or standard radio-
graphic images, knowledge of the scope caliber enables a simple
proportion to be created to determine the exact magnication for
each image (Fig. 3.10). In a hypothetical example, if an older 13.5 mm
caliber therapeutic endoscope is used during an ERCP where there
is biliary dilation above a strictured region, the exact degree of duct
dilation may be calculated by measuring the dilated portion of the
duct as well as the scope on the image. The following simple propor- Fig. 3.7 Severe chronic pancreatitis. A Innumerable pancreatic cal-
tion is then set up: cications are present on the ERCP scout radiograph. B Following
injection of contrast into the main pancreatic duct, the radiopaque
17 (measured caliber of scope 25 (measured caliber of intraductal calcications are obscured by contrast. Parenchymal cal-
cications should project apart from the opacied duct lumen;
on lm) = dilated duct) however in markedly dilated ductal systems the sheer caliber of the
13.5 mm (actual scope caliber) X mm opacied duct may obscure even parenchymal calcication. C Cor-
responding abdominal CT image through the body of the pancreas
shows parenchymal and ductal calcications in this patient.
21
 SECTION 1 GENERAL TOPICS
Fig. 3.8 ERCP drainage lm. Spot radiograph after stent insertion
above the level of obstruction by the impacted cystic duct stone in
this patient with Mirizzi syndrome (same patient as Figs 3.1 and 3.2)
reveals emptying of the right and extrahepatic bile ducts. If the
patient is turned to a supine position and the head of the table tilted
up, drainage of the left intrahepatic ducts may also be veried.
Fig. 3.9 Retroperitoneal air after sphincterotomy. Final lm
obtained after scope removal reveals air outlining the right upper
renal pole (arrow) and right adrenal gland (arrowhead), indicating
duodenal perforation.
22
Fig. 3.10 Calculation of individualized magnication factor during
ERCP. The caliber of the upstream CBD (arrow) may be set up in a
proportion to the measured scope caliber (arrowhead) on the lm
to correct for magnication produced by short object-to-image
distance of the x-ray beam.
The hypothetical duct measures 19.9 mm. If the scope was a
newer 11.5 mm model, the caliber of the duct would be 16.9 mm.
Solving for X alleviates the need to calculate pixel correction
measurements (for digitally acquired spot images) or to estimate
magnication on standard radiographic spot images. This is also
important when calculating the length from the papilla to a stricture
in order to select the appropriate length stent, although the use of a
ruled catheter or wire withdrawal may be more accurate for selecting
stents than use of x-ray lm measurements8 (see Chapter 16 for
further details). If a new or different size scope is utilized for a par-
ticular procedure, this information should be communicated to the
radiologist to insure accuracy of measurements.
BILE DUCT OPACIFICATION
The distal portion of the common bile duct is readily opacied by
injection of the major papilla or ampulla of Vater. The normal
caliber of the injected extrahepatic bile duct ranges from 3 mm up
to 9 mm, with the larger normal caliber more typically seen in older
individuals912 and in patients who have undergone cholecystec-
tomy,1216 although the wider caliber does not signicantly increase
in the postoperative period.9 This is opposed to the relatively smaller
upper limits of normal for duct caliber as measured by cross sec-
tional imaging such as ultrasound or CT, where there is not active
injection occurring to distend the duct,1316 yet the same trends in
larger normal calibers in older and post-cholecystectomy patients are
seen.12,15,16 Variability of insertion of the cystic duct leads to different
degrees of obliquity required for adequate visualization. In the
case of a long common channel between the CBD and cystic duct

(Fig. 3.11), the best position to identify entrance of the cystic duct
or to visualize potential cystic duct stones is typically left anterior
oblique (Fig. 3.12).
The best position to identify the conuence of right and left
hepatic ducts, particularly important in the evaluation of hilar tumors
(Fig. 3.13), is right anterior oblique. With the patient prone or in left
lateral decubitus position, there is preferential opacication of the
left intrahepatic ducts (Fig. 3.14A) due to gravity. Placing the patient
A Changing the patients position (Fig. 3.15) in attempts to visualize
B Inadequate lling may result in overestimation of strictures, and the
C D
Fig. 3.11 Long common cystic duct/ common bile duct channel,
ERCP/CT correlation. A ERCP spot radiograph revealing a large
amount of amorphous stone debris in the long cystic duct and
common duct which overly one another in a straight anterior to
posterior (AP) direction. BD Corresponding successively caudal
axial CT images obtained after ERCP reveal the relationship of the
cystic duct and common hepatic duct in this patient, with the wall
between the lumens well depicted on B (arrow) and C by the pres-
ence of residual biliary contrast. The wall ends approximately 1 cm
above the ampulla on D.
Fig. 3.12 ERCP depiction of long common cystic duct/common
hepatic duct channel with left anterior oblique projection. When
the prone patient is angled with the left side down on the uoro
table, the wall between the cystic duct and common bile duct
lumens is better seen. The distal, smooth common bile duct stric-
ture is due to a nonhyperfunctioning neuroendocrine tumor of the
pancreatic head in this patient.
Chapter 3 Radiologic Issues in ERCP
in a right decubitus position or supine position helps to opacify the
right-sided ducts. The posterior segmental branches may be best
seen with the patient supine. In some patients, it is necessary to tilt
the head of the table down to opacify the right intrahepatic ducts,
and other endoscopic measures such as proximal CHD injection or
balloon occlusion injected may help to opacify all of the IHD concur-
rently (Fig. 3.14B). If there is complete opacication of the intrahe-
patic duct system and the cystic duct and gallbladder do not ll
despite changing the patients position, cystic duct obstruction
should be suspected.17
all intrahepatic ducts is especially critical in patients with strictures
near the conuence, as is injecting near the stricture (Fig. 3.16).
shouldered feature of malignant strictures may not be ideally dem-
onstrated without adequate duct lling (Fig. 3.17). Likewise the char-
acteristics of benign strictures are better delineated with adequate
duct lling (Fig. 3.18). Changing the patients position may also help
to correctly identify the source of bile leak if aberrant or overlapping
ductal structures are present (Fig. 3.19).
As in the case of injecting near a stricture to better characterize
its extent and character, injecting near the site of a bile leak
Fig. 3.13 Hilar stricture. The very tight stricture (arrow) of the right
main bile duct is well seen in the right anterior oblique projection
in this patient with a Klatskin tumor. The patient underwent right
hepatectomy and L hepaticojejunostomy with biliary margin free
of tumor.
A B
Fig. 3.14 Preferential ow of contrast due to gravity. A Early lling
of the common bile duct with the patient in prone position reveals
stone (arrow) impacted at cystic duct remnant. B after stone retrieval,
the entire intrahepatic duct system is opacied during injection
using balloon occlusion technique.
23
 SECTION 1 GENERAL TOPICS
A B
Fig. 3.15 Hilar strictures due to lymphoma. A The long, relatively
smooth stricture (arrowheads) of the main left and right bile ducts
in the hilum is well seen on the prone image. There is also dilation
of the posterior segmental branch duct (arrow) which drains aber-
rantly into the left duct. B With the patient obliqued, the strictured
region (white arrow) of the aberrant duct is now evident.
A B
Fig. 3.16 Primary sclerosing cholangitis. A Injection of contrast in
the mid common bile duct reveals partial visualization of the tight
strictures of the right and left ducts in the porta, with some reux
of contrast into the cystic duct and gallbladder. B Injection within
the right bile duct just above the conuence demonstrates improved
visualization of the diffusely strictured ductal system in this patient
with primary sclerosing cholangitis. The left ducts were not
opacied further during ERCP, making it difcult to exclude
cholangiocarcinoma.
A B
Fig. 3.17 Shouldered stricture. A Early injection of the distal
common bile duct demonstrates narrowing in the superior pancre-
atic head. B With further lling the persistent shouldering (arrow)
of the malignant stricture is better depicted. Note the layering
of contrast in the underlled duct above the stricture on A, under-
estimating the extent of upstream dilation, better seen on B.
Gallstones are present within the gallbladder.
24
Fig. 3.18 Chronic pancreatitis stricture. Opacication of the bile
and pancreatic ducts reveals a smooth conical stricture of the distal
common bile duct and upstream dilation in a pattern typical of
inammatory stricture. There is dilation of the main and side branch
pancreatic ducts, and a stricture is present in the mid body region
of the main pancreatic duct in this patient with severe chronic
pancreatitis.
A B
Fig. 3.19 Post cholecystectomy bile leak. A Prone image during
lling of the common and intrahepatic ducts reveals contrast within
the drain in the right upper quadrant. Both the stapled end of the
cystic duct and an aberrant right branch overlie the drain. No early
lm was obtained to determine the site of leak; however, with the
patient obliqued slightly in B, contrast appeared to extravasate from
the aberrant duct rather than the cystic duct stump. Note air bubbles
in distal CBD after stent placement in B.
(Fig. 3.20) or obstructing stone (see Figure 3.2) is important to fully
understand the pathology. In the case of the endoscope obscur-
ing portions of the bile duct (generally this occurs in the supra-
pancreatic portion of the extrahepatic duct) (Fig. 3.21), attempts to
change scope position to allow direct visualization of the duct may
be necessary. For long bile duct strictures, obtaining orthogonal view
spot images may help characterize the stricture and demonstrate
extrinsic effect. The same is true for intrahepatic bile duct abnor-
malities produced by hepatic parenchymal disease such as polycystic
liver disease (Fig. 3.22) or cirrhosis. Drainage lms may be facili-
tated by tilting the head of the table up for several minutes prior to
image exposure (see Fig. 3.8).
PANCREATIC DUCT OPACIFICATION
The pancreas is oriented with the head and tail located relatively
more posteriorly within the patient compared to the neck and body
 Chapter 3 Radiologic Issues in ERCP

A B

D E

Fig. 3.20 Post traumatic bile leak. A Initial ERCP injection into the bile duct shows extravasation from two right-sided intrahepatic ducts.
B After stent placement the ducts have emptied, with residual intrahepatic extravasated contrast remaining near the drain. C Axial CT image
reveals the disrupted hepatic parenchyma due to gun shot wound. D Two weeks later, repeat ERCP with injection into the extrahepatic
CBD reveals no extravasation. E Injection directly into the affected right duct structures (arrow) further characterizes the ducts as sealed.

Fig. 3.22 Deviation of intrahepatic bile ducts due to polycystic

Fig. 3.21 Endoscope obscuring majority of the malignant stricture
of the common bile duct in a patient with pancreatic adenocarci-
noma involving the superior head region. Note adjacent stricture
of the main pancreatic duct. See Fig. 3.31 for biopsy image of same
patient.
liver disease. Injection into the right biliary duct system demon-
strates a long, smooth stricture of the extrahepatic and central
right-sided ducts, with superior displacement and draped appear-
ance of peripheral ducts in a patient with large intrahepatic paren-
chymal cysts.

region. Thus, with injection at the level of the papilla, contrast must
travel against gravity, or posteriorly in the patient, to reach the tail
region when the patient is prone on the uoroscopic table. Changing
the patient to the supine position will often allow more prompt
visualization of the duct in the pancreatic tail (Fig. 3.23). The main
pancreatic duct is approximately 20 cm long and variable in caliber.17
In general, the caliber of the duct is greatest in the downstream or
head region next to the papilla, tapering continually towards the
upstream or tail region. Normal caliber of the injected pancreatic
duct in general is 4 mm in the head, 3 mm in the body, and 2 mm
in the tail,17,18 though larger diameters are considered normal in
advanced age.19 If the patient has a pronounced anterior to posterior
curve of the pancreas within the abdomen (readily noted with CT or
MR axial images), combined RAO and LAO images may best lay out
the duct for complete visualization. Because the normal pancreatic
duct drains rapidly, image acquisition during active contrast injec-
tion at a rate of two to three images per second may avoid repeated
injections and allow complete visualization of all portions including
pathologic regions of the duct. This may be critical in that repeated
pancreatic injections have been shown to increase risk for post ERCP
pancreatitis.2022
In general, it is the focal change in caliber of either the bile or
pancreatic duct that indicates pathology and warrants further image
acquisition. If the catheter is placed farther into the duct up to the

25
 SECTION 1 GENERAL TOPICS
Fig. 3.23 Complete visualization of the main pancreatic duct. Injec-
tion into the distal main pancreatic duct reveals normal smooth
transition from largest caliber in the head to smallest caliber in the
tail.
A B
Fig. 3.24 Chronic pancreatitis. A Early lling of the main pancreatic
duct shows dilation in the tail region, upstream from a smooth
stricture (arrow) in the proximal body region. B With further injec-
tion and repositioning of the scope, there is better demonstration
of the stricture in the body (white arrow), as well as side branch
dilation and a second focal inammatory stricture of the main pan-
creatic duct in the neck region (black arrow).
A B
Fig. 3.25 Acinarization during pancreatic duct injection. A Early
lling of the main pancreatic duct reveals a loop variation in the
head region and obscuration of the pancreatic duct by the scope.
B With further injection under pressure there is acinarization of
contrast and lling of the tight malignant stricture (arrowheads) in
the pancreatic neck region, accompanied by upstream dilation in a
patient with pancreatic adenocarcinoma.
location of a suspected stricture or lling defect, the interpreting
physician can be relatively certain that the affected area is visualized
adequately (Fig. 3.24). Although most often not intentionally pro-
duced, when acinarization occurs distal or downstream to a stricture
(Fig. 3.25), one may be sure that the diminished caliber or obstruc-
tion of the duct is not due to technical factors.
26
It is also important to remember that when contrast passes
through a strictured region to ll a more dilated proximal duct
segment, there will be dilution of the contrast and sometimes layer-
ing (see Fig. 3.17) of the contrast,3 making estimation of the
length of stricture or the degree of upstream dilation potentially
inaccurate. In addition, extravasation from the duct upstream to the
strictured region, as is often seen in patients with chronic pancre-
atitis who have a dominant stricture and chronic pseudocyst
formation, may only be seen with deep cannulation and injection
immediately adjacent to the affected portion of the main pancreatic
duct. Likewise to show that a leak has sealed, injection of contrast
near the site of prior extravasation (see Fig. 3.20) during follow-up
studies is helpful to exclude technical factors such as underlling of
the affected duct that may produce false negative results. If extra-
vasation occurs from either the pancreatic or bile duct, the size of
the cavity (if contained rather than freely extravasated contrast is
present) may be underestimated on ERCP due to the relative limited
volume of contrast injected, and is better assessed with cross-
sectional imaging (Fig. 3.26).
The evaluation of the lling defects within either duct may be
difcult without good communication between the endoscopist and
radiologist. Distinguishing between air bubbles that tend to be
round or oval (Fig. 3.27) and small gallstones or pancreatic concre-
tions that tend to be angular or faceted (see Fig. 3.3) but which may
also be round, can generally be accomplished by changing patient
position.23,24 With the head of the table tilted up, gallstones will
proceed downward within the duct system, and air bubbles will rise
upward (Fig. 3.28).
Air bubbles are to be expected after sphincterotomy. Sometimes,
such a large amount of air enters the duct that identication of
stones is no longer possible (see Fig. 3.4). In the case of precut
sphincterotomy to access the duct, there will be no initial injection
to clearly document the presence of stones prior to the introduction
of air into the duct system. Paying particular attention to the shape
and movement of intraductal lling defects may help distinguish
between the two, even in this circumstance. When lling defects are
removed without documentation of their presence on the images,
there is no way for an accurate interpretation to occur after the fact.
Since the most common indication for endoscopic retrograde
cholangiography remains suspected biliary obstruction17 and for
endoscopic sphincterotomy continues to be choledocholithiasis,
documentation of stones on images obtained prior to stone removal
and/or communication of real-time endoscopic ndings help to
ensure consistent reporting. In addition, the performance of sphinc-
terotomy must be communicated Since there is a slightly higher risk
of perforation during ERCP when sphincterotomy is performed,2527
the presence of retroperitoneal and free air should be sought more
stringently in those patients. On the other hand, the radiologist
cannot assume that a sphincterotomy has occurred because the
instrument is documented on an image; the sphincterotome may
simply be used to cannulate the major papilla in patients with a dif-
cult entry angle.28 Similarly, pancreatic guidewire or stent place-
ment may be performed solely to facilitate biliary cannulation29,30
and images documenting this event do not imply pancreatic disease.
Communication of the endoscopists thoughts during a procedure
is helpful to avoid errors related to misinterpretation of these phe-
nomena and to provide the most accurate report.
In the case of minor papilla injection or therapeutic maneuver
(minor papillotomy and/or stent placement), the long scope position
(Fig. 3.29) typically employed to access the more proximal papilla
 Chapter 3 Radiologic Issues in ERCP

A B C

D E

Fig. 3.26 Extravasation from pancreatic duct. A Early lm obtained during ERP demonstrates smooth distal main pancreatic duct and
sidebranches in the head region. B and C With deeper cannulation and further injection, extravasation into an amorphous cavity is present
in the mid body region. D After pancreatic duct stent is placed and the patient is placed supine, the contrast spreads out further within
the cavity. E Accompanying axial CT image demonstrates cystic collection in the mid body region in this patient with duct disruption due
to evolving subacute pancreatitis.

A B

Fig. 3.27 Gas bubbles in main pancreatic duct. With wire exchange
and deep cannulation there is introduction of air and formation of
smooth, round bubbles in the main pancreatic duct.
C

may be a hint to the interpreting radiologist that the duct of Santorini

has been injected, even if the major papilla and typical features of
pancreas divisum (Fig. 3.30) are not documented elsewhere in the
procedure. But again, for the most effective reporting, this informa-
tion should be rapidly communicated by the endoscopist.
Some therapeutic maneuvers and their respective images speak
for themselves such as biopsy (Fig. 3.31), stone extraction (Fig. 3.32)
or stent placement (Fig. 3.33) so long as an image is obtained to
document the event. However, in other circumstances such as
in patients with sphincter of Oddi dyskinesia, both the bile and
pancreatic ducts may be morphologically normal, and the diagnosis
Fig. 3.28 Gallstones versus air bubbles. A Initial injection into a
cannot be made on the basis of images alone. Communication type 4 choledochal cyst demonstrates multiple lling defects (arrow)
of manometric measurements, if acquired, and clinical factors in in the distal aspect of the dilated duct. B After sphincterotomy,
these patients is critical to insure an accurate interpretation of the several rounded lling defects suggestive of air bubbles are more
radiographs. clearly seen within the more superior aspect of the duct. C With
the head of the table tilted up, there is downward migration of the
In summary, the modern-day practice of ERCP involves a large stones/debris (arrow); the rounded air bubbles have risen and are
percentage of therapeutic procedures and frequently does not permit no longer evident.

27
 SECTION 1 GENERAL TOPICS
Fig. 3.29 Long scope position. Spot radiograph obtained during
injection of the minor papilla demonstrates long scope position
often, but not always, utilized to cannulate the minor papilla in
patients with pancreas divisum. Note changes of mild chronic
pancreatitis.
Fig. 3.30 Pancreas divisum. Injection of the major papilla in short
scope position reveals concurrent visualization of bile duct and
typical arborizing ventral pancreatic ducts in a patient with pancreas
divisum.
A B
Fig. 3.32 Stone extraction. A Spot image reveals mild common duct dilation with rounded 7 mm stone and balloon inated above the
stone prior to sweeping the duct. B After unsuccessful balloon sweeping, a wire basket was placed around the stone and in C retrieved
successfully.
28
the radiologist to be in the endouoroscopy suite concurrently with
the endoscopist; therefore communication is critical to the rendering
of adequate radiologic reporting and concurrence with endoscopic
ndings. Some centers are equipped with inter-departmental inter-
coms and videomonitors to allow real-time discussion between
the endoscopist and radiologist while they are physically located in
different rooms. In the case of more remote (both spatial and tem-
poral) endoscopy/radiology cooperation, voice recognition dictation
systems that allow immediate reporting, digital medical record
archiving, and hospital system-wide accessibility to documents
enable rapid reporting of ERCP but cannot replace real-time com-
munication. If real-time exchange is not feasible, then careful docu-
mentation of normal or abnormal structures and image documentation
of endotherapeutic maneuvers during ERCP are of paramount
importance!
Fig. 3.31 Endoscopic biopsy. Spot radiograph reveals open
forceps deployed to region of malignant biliary stricture for tissue
sampling of suspected tumor.
C

A B
Fig. 3.33 Restenting through tumor overgrowth. A Injection into the lumen of the distal common bile duct in a pancreatic carcinoma
patient who underwent wire mesh stent placement three months earlier for palliative biliary drainage reveals near occlusion of the lumen
due to tumor ingrowth. B After deep wire cannulation and injection of dilated upstream bile ducts, C, a longer stent was successfully
placed through the occluded stent, re-establishing biliary drainage.
REFERENCES
1. Cotton PB. Evaluating ERCP is important but difcult. Gut 2002;
51:287289.
2. NIH state-of-the-science statement on endoscopic retrograde
cholangiopancreatography (ERCP) for diagnosis and therapy.
NIH Consensus Sci Statements. 2002; 19(1):126.
3. Technical Considerations in Imaging. In: Taylor AJ, Bohorfoush AG
III (eds) Interpretation of ERCP with Associated Digital Imaging
Correlation. Philadelphia, Lippincott-Raven, 1977: pp 124.
4. Johnlin FC, Pelsang RE, Greenleaf MG. Phantom study to
determine radiation exposure to medical personnel involved in
ERCP uoroscopy and its reduction through equipment and
behavior modications. Am J Gastroenterol 2002; 97:893897.
5. Cousin AJ, Lawdahl RB, Chakraborty DP, et al. The case for
radioprotective eyewear/facewear: practical implications and
suggestions. Invest Radiol 1987; 22:747750.
6. Boone JM, Levin DC. Radiation exposure to angiographers under
different uoroscopic imaging conditions. Radiology 1991;
180:861861.
7. Heyd RL, Kopecky KK, Sherman S, et al. Radiation exposure to
patients and personnel during interventional ERCP at a teaching
institution. Gastrointest Endosc 1996; 44:287292.
8. Seibert DG. Biliary stricture measurement and stent selection. Am
J Gastroenterol. 1997; 92:15101514.
9. Mageed AW, Ross B, Johnson AG. The preoperatively normal bile
duct does not dilate after cholecystectomy: results of a ve year
study. Gut 1999; 45:741743.
10. Mahour G, Wakim K, Ferris D. The common bile duct in man: its
diameter and circumference. Ann Surg 1967; 165:415419.
11. Edholm P, Jonsson G. Bile duct stones related to age and duct
width. Acta Churg Scand 1962; 124:7579.
12. Kaim A, Steinke K, Frank M, et al. Diameter of the common bile
duct in the elderly patient: measurement by ultrasound. Eur
Radiol. 1998; 8:14131415.
13. Morgan B, Rathod A, Crozier A, et al. Biliary distensibility during
per-operative cholangiography as compared with preoperative
ultrasound: a four year follow up study. Clinical Radiology 1996;
51:338340.
Chapter 3 Radiologic Issues in ERCP
C
14. Sauerbrei EE, Cooperberg PL, Gordon P, et al. The discrepancy
between radiographic and sonographic bile-duct measurements.
Radiology 1980; 137:751755.
15. Bowie JD. What is the upper limit of normal for the common bile
duct on ultrasound: how much do you want it to be? Am J
Gastroenterol. 2000; 95(4):897900.
16. Daradeh S, Tarawneh E, Al-Hadidy A. Factors affecting
common bile duct diameter. Hepatogastroenterology 2005;
52:16591661.
17. Schoeman MN, Huibregtse K, Reeders JWAJ. Endoscopic
Cholangiography and Pancreatography (ERCP), In: Van Leeuwen
DJ, Reeders JWAJ, Ariyana J (eds), Imaging in Hepatobiliary and
Pancreatic Disease: A Practical Clinical Approach. London, W. B.
Saunders, 2000: pp 309332.
18. Ohnuma N, Takashi H, Tanabe M, et al. The role of ERCP in biliary
atresia. Gastrointest Endosc 1997; 45:365370.
19. Hastier P, Buckley MJ, Dumas R, et al. A study of the effect of age
on pancreatic duct morphology. Gastrointest Endosc. 1998;
48:5357.
20. Cheng CL, Sherman S, Watkins JL, et al. Risk factors for post-ERCP
pancreatitis: a prospective multicenter study. Am J Gastroenterol.
2006; 101:139147.
21. Testoni PA. Why the incidence of post-ERCP pancreatitis
varies considerably? Factors affecting the diagnosis and
the incidence of their complication. J Pancreas 2002;
3:195201.
22. Freeman ML. Post-ERCP pancreatitis: patient and technique
related risk factors. J Pancreas 2002; 3:169176.
23. Thompson WM, Halvorsen RA, Foster WL, et al. Optimal
cholangiographic technique for detecting bile duct stones.
AJR 1986; 146:537.
24. Train JS, Novick A, Dan SJ, et al. Radiolucency in the common
bile duct simulating a gallstone. AJR 1987; 148:136.
25. Siegel J, Ben-Zvi J, Pullano W, Cooperman A. Effectiveness
of endoscopic drainage for pancreas divisum: endoscopic
and surgical results in 31 patients. Endoscopy 1990;
22:129133.
29
 SECTION 1 GENERAL TOPICS
26. Suissa A, Yassin K, Lavy A, et al. Outcome and early complications
of ERCP: a prospective single center study.
Hepatogastroenterology 2005; 62:352355.
27. Masci E, Toto G, Mariani A, et al. Complications of diagnostic and
therapeutic ERCP: a prospective multicenter study. Am J
Gastroenterol 2001; 96:417423.
28. Gulliver D, Cotton P, Baillie J. Anatomic variants and artifacts in
ERCP interpretation. AJR 1991; 156:975980.
30
29. Maeda S, Hayashi H, Hosokawa O, et al. Prospective randomized
pilot trial of selective biliary cannulation using pancreatic
guidewire placement. Endoscopy 2003; 35:721724.
30. Goldberg E, Titus M, Haluszka O et al. Pancreatic-duct stent
placement facilitates difcult common bile duct cannulation.
Gastrointest Endosc 2005; 62:592596.
 SECTION 1 GENERAL TOPICS
Chapter
Endoscopes, Guidewires
4 and Accessories
Sushil K. Ahlawat and Firas H. Al-Kawas
INTRODUCTION
ERCP (endoscopic retrograde cholangiopancreatography) has
become the preferred technique for the management of patients
with a variety of benign and malignant pancreaticobiliary disorders.
Success and safety of the procedure depends to a large extent on the
indication of the procedure, skills of the examiner and an organized
and functional ERCP unit. In addition to a dedicated ERCP room
and a uoroscopy unit, essential equipment for ERCP includes a
duodenoscope and a variety of ancillary devices or accessories. A
growing range of ERCP accessories have been developed to support
the increasing demands and complexity of therapeutic ERCP. This
chapter describes current and emerging accessories that are cur-
rently available to use during diagnostic and therapeutic ERCP.
ENDOSCOPES
Duodenoscopes
Side viewing video endoscopes are equipped with an elevator and
are used routinely for diagnostic and therapeutic ERCP procedures.
The elevator facilitates cannulation and placement of some accesso-
ries (Fig. 4.1), while large diameter working channel of therapeutic
duodenoscopes (4.2 and 4.8 mm) allow the use of large accessories.
Many current ERCP endoscopes combine a large, therapeutic
channel with a standard size insertion tube. Smaller, 7.4 mm pedi-
atric duodenoscopes with a 2.2 mm channel are available for exami-
nation in neonates. Unfortunately, the small channel limits the use
of the endoscope to mostly diagnostic purposes since the use of
smaller accessories restricts the therapeutic potential of this endo-
scope. In general, the standard adult duodenoscope can be used in
most children above the age of two. A jumbo-size duodenoscope
(5.5 mm channel) was previously available as a mother/baby scope
system. However, this system was difcult to manipulate and is now
rarely used.
Forward-viewing endoscopes
Upper endoscopes, colonoscopes, and enteroscopes are occasionally
used in patients with altered anatomy such as previous choledocho-
duodenostomy, Billroth II gastrectomy, or in patients with hepatico-
jejunostomy. Since conventional forward viewing endoscopes do not
have an elevator, they are limited with respect to control of accesso-
ries during cannulation or therapy. In addition, visualization of the
ampulla may be limited. A prototype oblique-viewing endoscope is
currently available (Pentax Medical Montvale, NJ) with a viewing
angle of 45 and a eld of view of 130, a 3.8-mm-diameter instru-
ment channel, accessory elevator, and a working length of 120 cm.
These features provide an adequate angle of approach and allow
intubation of difcult bowel loops and performance of complex
therapeutic maneuvers. Initial experience with this oblique-viewing
therapeutic endoscope allowed duct access and therapy in two
patients with surgically altered anatomy.1 In both patients, prior
multiple attempts at ERCP with standard duodenoscope and colo-
noscope had failed.
ACCESSORIES
Accessories are devices or pharmacologic agents that help the endos-
copist accomplish his diagnostic or therapeutic intent. Cannulation
of the desired duct is a prerequisite to successful diagnostic and
therapeutic ERCP. A variety of devices are currently available to gain
duct access. In particular, the use of sphincterotomes/guidewires
and precut sphincterotomes has increased our ability to achieve
biliary cannulation.
Cannulas
Standard ERCP cannulas are 5 Fr (French) to 7 Fr catheters, with a
straight, tapered or a rounded tip that can accept up to a 0.035-inch
guidewire (Fig. 4.2A). Use of a triple-lumen device or the attachment
of a side-arm adaptor will allow contrast injection without removing
the guidewire. The use of tapered (4.5 Fr-4 Fr-3.5 Fr) or ultra-tapered
(5 Fr-4 Fr-3 Fr) tip catheters may allow better duct access in some
patients. However, some tapered tip cannulas will only accommo-
date a smaller-caliber guidewire (= 0.025-inch). No published studies
have directly compared cannulation success rates between standard
and tapered catheters. Tapered tip catheters may be associated with
a higher risk of submucosal injection.
Standard cannulas with or without guidewires are limited in their
ability to vary the angle of approach to the papilla. The Swing-tip
catheter (Olympus America Inc., Lehigh Valley, PA) (Fig. 4.2B) over-
comes the limitations of conventional catheters and offers the endos-
copist the ability to bend the cannula tip in either direction, thereby
facilitating biliary cannulation2,3 or selective entry into the right or
left hepatic ducts. The Cremer needle tip catheter (Cook Endoscopy,
Winston-Salem, NC) is 1.8 mm in diameter and has a metal needle
tip design that facilitates minor papilla cannulation (Figs 4.3a and
4.3b). The standard pancreaticobiliary manometry catheter is a
water-perfused 5 Fr catheter with a tip diameter of 3.5 Fr and is used
during sphincter of Oddi motility studies (Fig. 4.4A). A variety of
catheter types can also be used. Some manometry catheters have a
longer nose to help maintain catheter position. The standard cath-
eter has three side ports spaced 2 mm apart for simultaneous pres-
sure measurement. The Lehman (Cook Endoscopy) catheter sacrices
31
 SECTION 1 GENERAL TOPICS
Fig. 4.1 4.2 mm channel duodenoscope (courtesy of Pentax
Medical, Montvale, NJ).
A B
Fig. 4.2 A ERCP cannulas (courtesy of Cook Endoscopy,
Winston-Salem, NC). B Swing-tip ERCP cannula (courtesy of
Olympus America Inc., Lehigh Valley, PA).
B pancreatic sphincterotomy, pure cutting current is often used to
A formation.
Fig. 4.3 A Cremer cannula (courtesy of Cook Endoscopy,
Winston-Salem, NC). B Endoscopic view showing Cremer cannula
and minor papilla.
one port for aspiration of water out of the pancreatic duct during
infusion to prevent overlling of the pancreatic duct and thereby
reduce the risk of post procedure pancreatitis.4 Standard water per-
fused motility recording systems are used for sphincter of Oddi
manometry (Figs 4.4B and 4.4C). Recently a micro transducer cath-
32
eter became available. Early data suggest that this catheter is associ-
ated with a lower risk of post procedure pancreatitis.5
Sphincterotomes
Pull type (Erlangen) sphincterotomes were originally designed for
the performance of sphincterotomy. They consist of a Teon cathe-
ter containing a continuous wire loop with 23 cm of exposed wire
exiting at a variable distance from the tip (Fig. 4.5A). In precut
sphincterotomes, the cutting wire extends to the tip (Fig. 4.5A).
The other end of the wire is insulated and connected to an
electrosurgical unit. Over the last decade or so, endoscopists have
recognized the need to angle the catheter upward to selectively
enter the bile duct.6 Subsequently, prospective randomized trials
comparing standard catheters with sphincterotomes have shown a
cannulation rate of 8497% with sphincterotomes compared with
6267% with standard catheter.7,8 Therefore, in many institutions,
sphincterotomes have become the primary cannulation device in
ERCP.
Currently, sphincterotomes are available with single, double and
triple lumens. Double lumen sphincterotomes allow for the intro-
duction of a wire to facilitate cannulation or to accomplish a thera-
peutic task. Contrast can be injected after removing the wire or by
adding a side arm. Triple lumen sphincterotomes allow injection of
contrast without removing the wire. Unfortunately, because of the
small size of the injection lumen, contrast ow is slow and difcult
for the assistant because of the force required during infusion. The
use of a small syringe facilitates injection. A sphincterotome is avail-
able that incorporates a combination of cutting and balloon stone
extraction (Boston Scientic, Natick, MA); however, adding the
lumen increases the catheter diameter and tip size which may make
cannulation more difcult.
When sphincterotomy is performed, a variety of generator
currents can be used: cutting, auto-cut, coagulation or blended.
Limited data suggest that the use of a pure cutting current is associ-
ated with a lower risk of pancreatitis after sphincterotomy9 while the
use of an auto-cut is associated with a lower risk of bleeding during
sphincterotomy. This autocut feature has essentially eliminated the
Zipper cut phenomenon after sphincterotomy. When performing
reduce the risk of pancreatic duct injury and subsequent stricture
Rotatable sphincterotomes have recently been introduced to offer
the endoscopist the ability to change the orientation of the sphinc-
terotomes. Preliminary data suggest that this may be useful in
improving cannulation, especially in patients with unusually ori-
ented or distorted papillae,10 or in patients with Billroth II anatomy.
Rotatable sphincterotomes may also help orient the cutting wire
during sphincterotomy. However, no published data is available to
support this theory.
Access sphincterotomes
Precut or access sphincterotomy refers to a variety of endoscopic
techniques used to gain access to the bile or pancreatic duct after
conventional methods of cannulation have failed (see Chapter 9).
Needle-knife and precut sphincterotomes are the two most com-
monly used devices to gain access into the bile duct. The needle-
knife was rst described by Kees Huibregtse in 1981 and is essentially
a bare wire that protrudes 45 mm from the end of a Teon catheter
(Fig. 4.5B). Several papers have discussed the use of this device.

A B
Fig. 4.4 A Endoscopic view showing motility catheter. B Manometry tracing (normal
sphincter of Oddi pressure). C Sphincter of Oddi motility recording system.
A
B
Fig. 4.5 A Standard and precut sphincterotome (courtesy of Cook
Endoscopy, Winston-Salem, NC). B Needle-knife sphincterotome
(courtesy of Cook Endoscopy, Winston-Salem, NC).
Newer versions include additional lumen for wire (double lumen)
and both wire and contrast (triple lumen) introduction. The precut
sphincterotome was rst reported by Nib Soehendra and the
Hamburg group in 1996 (Fig. 4.5A). This sphincterotome allows
papillary roof incision.11 A double lumen version has been recently
Chapter 4 Endoscopes, Guidewires and Accessories
Duodenal baseline
introduced.12 It has the advantage of having separate lumens for
contrast and guidewire. Biliary sphincterotomy can be immediately
completed by using the same instrument and is facilitated by having
a preloaded slippery wire. Data comparing different precut tech-
niques are limited.13
Guidewires
Guidewires are the cornerstone of diagnostic and therapeutic ERCP.
During ERCP, guidewires are used for achieving and maintaining
access and placing and exchanging devices. Examples include dif-
cult cannulation, sphincterotomy, navigating strictures, stricture
dilation, tissue sampling, stent placement, mini-scope placement,
and manometery.
Ideal guidewire characteristics for gaining access to lumen
differ from those for advancement of accessories. Guidewires with
slippery and exible leading tips are generally used to gain access
through tight biliary strictures. On the other hand, stiff and taut
guidewires are best used for advancement of devices such as biliary
33
 SECTION 1 GENERAL TOPICS

stents or dilators. Stiff and taut wires also minimize lateral deviation Guidewires are advanced under uoroscopic monitoring through
and facilitate forward axial transmission of forces. Friction can aid ERCP catheter or sphincterotome that imparts stiffness and direc-
in maintaining wire tensions but it hinders both wire and device tion. Guidewire passage is easier after ushing water through dry
movement. A variety of guidewires are currently available (Table 4.1) or contrast-lled devices because of minimal friction. Moistening of
and these vary in materials, length, diameter, and design to optimize exposed portions of hydrophilic wires prevents drying and sticking.
performance.14 Maintenance of wire position is critical for safe and effective use
In general, three guidewire designs are available for ERCP appli- of over the wire accessories such as dilators and stents. The risk of
cations: (1) Monolament wires are designed for rigidity and are
made of stainless steel. (2) Coiled wires are stiff and exible and
have an inner monolament core and outer spiral coil made of stain-
less steel. Inner core and outer spiral coil provide stiffness and ex-
ibility respectively. Most coiled wires are Teon painted in order to
A B
minimize resistance and are optimal for traversing tortuous biliary
strictures because of their improved trackability resulting from
combination of stiffness and exibility. (3) Coated or sheathed wires
have a monolament core made of stainless steel or nitinol and an
outer sheath made of Teon, polyurethane, or another lubricious
polymer. The outer sheath material can be designed to improve
radiopacity, slipperiness, and electrical insulation properties. Coated
wire tip exibility depends on the taper of the inner core. Many wires
have platinum tipped core to improve uoroscopic visualization. The
conguration of the guidewire can be straight or angled (J-shaped)
(Fig. 4.6A). Some wires have graduated or continuous markings for
Fig. 4.6 A Straight and angled tip guidewires (courtesy of Boston
visual endoscopic measurement or movement detection. Most wires Scientic, Natick, MA). B Endoscopic view showing the markings
are only minimally steerable in the radial direction. on the guidewires.

Wire type/name
(manufacturer) Diameter (inch) Length (cm) Core material Sheath material Tip material
MONOFILAMENT
Axcess 21 (CE) 0.021 480 Nitinol None Platinum
Cope (CE) 0.018 480 SS None Platinum

COILED
Standard Wires (CE) 0.018, 0.025, 0.035, 0.038 400480 SS Stainless coil, Teon Stainless tapered
painted core + coil

COATED
Tracer (CE) 0.035 260480 Nitinol Teon Platinum, hydrophilic
Protector Plus (CE) 0.035 480 Nitinol Teon Platinum
Roadrunner (CE) 0.018 480 Nitinol Teon Platinum
Zebra (BS) 0.025, 0.035, 0.038 260, 450 Nitinol Teon Platinum, endoglide
Jagwirea (BS) 0.035, 0.025 450 Nitinol Teon Platinum, hydrophilic
Hydra (BS) 0.035 260, 450 Nitinol Endoglide coating Tungsten, hydrophilic
Glidewire (BS) 0.018, 0.025, 0.035 450, 260 Nitinol Polyurethane hydrophilic Platinum
coat on entire length
Geenen Endotorque (CE) 0.035 450 SS Teon Platinum
Pathnder (BS) 0.018 450 Nitinol Endoglide Platinum, hydrophilic
LinearGuideV (O) 0.035 270, 450 Nitinol Polytetrauoroethylene Hydrophilic
X wirea (CM) 0.035, 0.025 260, 450 Nitinol Hydrophilic Nitinol

Table 4.1 Currently available Guidewires for ERCP applications (adapted with permission from reference no. 14)
a
Available in a stiff version.
SS = Stainless Steel.
CE = Cook Endoscopy, Winston-Salem, NC.
BS = Boston Scientic, Natick, MA.
CM = ConMed, Utica, NY.
O = Olympus America Inc., Lehigh Valley, PA.

34

wire displacement can be minimized by using guidewires that have
graduated or continuous markings or movement detection printed
distance markers and movement guides (Fig. 4.6b). In addition, xa-
tion of the proximal end (outside of the patient) on an immobile
accessory device can also lower the risk of wire dislodgement.
Currently available guidewire types include conventional,
hydrophilic, and hybrid, ranging from 0.018 to 0.035 inch in
diameter and 260 to 480 cm in length.15 These are summarized in
Table 4.1. Wire lengths above 400 cm are used for exchange of
devices. Only coated wires should be used during electro-cautery
applications.
Data is limited regarding the relative efcacy of specic wires for
ERCP applications. Clinical experience suggests that coated and
hydrophilic wires improve the ultimate success of those ERCP
procedures requiring access through difcult papillae or strictures.
Completely hydrophilic wires are subject to inadvertent displace-
ment from ducts or strictures and can make the catheter exchange
difcult. However, newer combination wires such as Jagwire, Hydra
jagwire (Boston Scientic, Natick, MA), FX, X (ConMed, Utica, NY),
and Metro (Cook Endoscopy, Winston-Salem, NC) may provide the
best combination of a slippery tip with a stiffer shaft (Fig. 4.6a)
Limited data suggest that biliary cannulation using a guidewire
through a sphincterotome lowers the risk of post-ERCP pancreatitis,
presumably because of less pancreatic injection.16 Teon coated
wires are least expensive. Hybrid wires are more user friendly but
more expensive. A useful and detailed review of guidewires can be
found in a recent Americal Society for Gastrointestinal Endoscopy
(ASGE) technology assessment report.14
Wire safety
Perforation and failed device placement are the two main wire-
related risks in the pancreas or biliary tree during ERCP. Applying
excessive force below a stricture or at an acute angle can result in
wire-related perforation. Loss of wire tension or access from a stric-
ture while using rigid devices such as biliary dilators can also result
in perforation. Wire-guided sphincterotomy can transmit signicant
electric current from the cutting wire through standard Teon-
painted guidewires into the bile duct. Intact, coated wires are
effectively insulated against transmission of short circuits or induced
A B Guidewire stripping from the catheter. (A and B Cour-
Chapter 4 Endoscopes, Guidewires and Accessories
currents. All damaged wires are potential sources of dangerous
currents.
Exchange assistance devices
Multiple devices are frequently required for a successful therapeutic
ERCP. Frequently, an exchange or series of exchanges over a previ-
ously placed guidewire is required to introduce subsequent device(s).
Several exchange assistance devices have been developed by differ-
ent manufacturers to facilitate exchange of over-the-wire accessories
in order to reduce reliance on an expert assistant and to allow
the use of a short (260 cm) wire. These devices may reduce uoros-
copy time and increase efciency.17 Potential problems with exchange
assistance devices include a restriction in the choice of accessories,
difculty in re-using the same accessory during the procedure, and
cost.
Rapid Exchange Biliary System
The Rapid Exchange (RX) Biliary System (Boston Scientic, Natick,
MA) is based on a monorail design that provides the endoscopist
with control over the guidewire and subsequent exchanges. The
system is composed of three integral units: a guidewire locking
device (Fig. 4.7A), a specially designed RX catheter, and a 260-cm
guidewire. A locking device secures the position of the guidewire
during exchange of over-the-wire accessories, advancement of acces-
sories, and manipulation. The locking device can accommodate mul-
tiple guidewires that can be secured at any time and thus allow for
multiple therapeutic interventions. Cannulation devices (i.e. cannula,
sphincterotome) have a proximal open-channel (beginning at 5 to
30 cm from the tip) that allows the guidewire to exit at this point
rather than at the hub of the endoscope. Once cannulation is
achieved, the wire is separated from the catheter (Fig. 4.7B) and is
secured in the guidewire locking device at the suction cap. A variety
of RX accessories are available to use with this system.
Potential benets of RX Biliary System include shorter total pro-
cedural and post cannulation times and a reduction of uoroscopy
exposure.18 However, the cost may be higher than the standard
equipment and the endoscopist may be limited to using RX acces-
sories. Cost-benet studies using the RX Biliary System are not
available.
Fig. 4.7 Rapid exchange system. A Locking device.
tesy of Boston Scientic, Natick, MA).
35
 SECTION 1 GENERAL TOPICS
A B
Fig. 4.8 Fusion system A Fusion catheter. B Biopsy valve with locking mechanism. (A and B Courtesy of Cook Endoscopy, Winston-
Salem, NC).
Fusion system
This system is made by Cook Endoscopy (Winston-Salem, NC). It
consists of either a double lumen catheter or a triple lumen sphinc-
terotome. The design of this system facilitates the exchange of
accessories without removing the guidewire or exchanging the
initially placed catheter/sphincterotome over the full length of the
guidewire.
The main difference between this new system and the conven-
tional design is that a side hole is placed at 6 cm from the tip of the
catheter (or any accessories from this line of products except the
stent introducer system in which the side hole is placed at 2.5 cm)
(Fig. 4.8A). The total length of the guidewire is 185 cm and the
length of most accessories is 220 cm. To provide proper control of
these much shorter accessories and guidewire, the system utilizes a
special disposable biopsy valve with a locking mechanism to anchor
the guidewire while performing exchanges (Fig. 4.8B).
A major advantage of the Fusion system is in the ability to place
multiple stents without removing the guidewire. With this system,
the guidewire is left within the bile duct across the stricture or
papilla, and this facilitates deployment of subsequent stents without
concerns about losing access across the stricture. In situations where
intervention requires the use of standard length or conventional
accessories, a standard length guidewire can be inserted through the
end of the catheter or sphincterotome after removing the inner
nylon stylet, and exchange can be performed in the usual manner.
Controlled data regarding the efciency with Fusion system is not
available.
V-system
The Olympus V-system (Olympus America Inc., Lehigh Valley, PA)
integrates Olympus endoscopes and endotherapy devices. This
design offers the option of guidewire manipulation by the physician
or by the assistant and may allow easier exchange of catheters using
a short guidewire. The V-endoscope has an increased elevator angle
and V-groove that allows the endoscopist to lock the wire when the
elevator is closed. This endoscope design may also enhance selective
biliary cannulation capability. The V-system features a C-Hook, V-
markings, and V-sheath for device control in addition to the V-
groove on the elevator of the duodenoscope (Fig. 4.9A).
The C-Hook attaches the device to the endoscope just below the
biopsy port (Fig. 4.9B) and allows a choice of control of the device
36
A B
Fig. 4.9 V-system A V-scope tip. B Hook. (A and B Courtesy of
Olympus America Inc., Lehigh Valley, PA).
by the physician or the assistant. V-markings are present on the
proximal portion of all V-system devices. When the V-marking on
the accessory device reaches the biopsy port of the endoscope, the
tip of the catheter has reached the endoscope elevator V-groove
indicating that raising the elevator at that point would lock the
guidewire in the V-groove. The V-sheath design allows the guidewire
sheath and the injection/handle sheath to be separated offering the
choice of control by the endoscopist or the assistant. The V-scope
and V-system accessories can also be used with long 0.035 and
smaller guidewires and with ERCP accessories from other device
manufacturers. Initial evaluation using this system (V-scope) has
shown improved reliability of guidewire xation;19 however, limited
data exist on efciency of catheter/guidewire exchanges.20
Drainage devices
Drainage devices include stents and nasobiliary drains. Stents are
used for a variety of purposes and are available in various materials
and congurations. Nasobiliary and nasopancreatic drains are infre-
quently used in the US.
Plastic stents
Plastic stents are made of polyethylene or Teon and are available
in varying size, shapes and length for biliary and pancreatic patholo-
gies. A pusher tube is used to place plastic stents over a guidewire
with or without a guiding catheter. Delivery systems are available for
plastic stents that combine the pushing and guiding catheters. The
standard stent delivery system for 10 Fr comprise a 0.035 inch guide-
wire (480 cm) and a 6 Fr radio-opaque Teon (260 cm in length)

guiding catheter with a tapered tip to facilitate cannulation and a
pusher tube. Some guiding catheters have two metal rings (placed
7 cm apart) at the distal end that helps in identication and measure-
ment of the stricture length. The pusher tube is made of Teon (8,
10, and 11.5 Fr) and used for positioning the stent during
deployment.
Most plastic stents are made of radiopaque polyethylene and are
available in sizes varying from 3 to 11.5 Fr. They also vary in length
and conguration. There is no inner catheter for the 37 Fr stent
delivery systems. Straight Amsterdam type stents are predomi-
nantly used for biliary drainage (Fig. 4.10a). Based on Poiseuilles
law there is a clear relationship between stent diameter and duration
of stent patency (Fig. 4.11).21 A straight conguration also appears
to improve stent patency. Attempts to improve stent patency by
eliminating side holes, changing stent material or coating the inner
surface with a hydrophilic substance have generally not been suc-
cessful.22,23 Double pigtail congurations (Fig. 4.10A) help anchor
A
B cannulation. In patients with difcult strictures, maintaining wire
Fig. 4.10 A Straight and double pigtail stents (courtesy of Olympus
America Inc., Lehigh Valley, PA). B Single pigtail stent (courtesy of
Cook Endoscopy, Winston-Salem, NC).
Patency rate of prostheses
12F
10F
Prostheses diameter*
Multiple
7F
7F
1 2 3 4
Patency rate (months)
* 3 French = 1mm
Fig. 4.11 Bar graph showing relationship between stent diameter
and the duration of functional patency (adapted from reference #21
with permission 2004 American Society for Gastrointestinal
Endoscopy).
Chapter 4 Endoscopes, Guidewires and Accessories
the stent to prevent upward or downward migration. These stents
are frequently used to maintain drainage in patients with difcult
bile duct stones and in some patients with hilar strictures.
Single pigtail stents (Fig. 4.10B) are frequently used in the pan-
creatic duct to prevent inward migration. Limited data suggest that
smaller stents (i.e. 3 and 4 Fr) will lead to less damage when used
in a normal pancreatic duct and that elimination of side holes and
aps may prolong patency and promote spontaneous migration of
pancreatic stents.24 New pancreatic stents have become available
recently that are constructed to have running channels and no side
holes (GI Supply, Camp Hill, PA). Limited data are available in refer-
ence to their superiority over currently used stents.25
Stents are usually removed using snares, baskets and foreign
body forceps. Large bore (10 Fr) stents can be removed through the
channel of a therapeutic endoscope with the aid of a snare. Smaller
stents, i.e. 3 Fr and 5 Fr pancreatic stents can also be removed via
the working channel of the endpscope using a foreign body forceps
(e.g. rat tooth forceps). The Soehendra stent retriever (Cook Endos-
copy, Winston-Salem, NC) consists of a screw-tipped wire-guided
device that allows stent removal while maintaining guidewire
position. It is also available with an extended tip design to facilitate
access can also be accomplished by using a wire and a snare.26
Self-expandable metal stents
Self-expandable metal stents (SEMS) were introduced to prolong
stent patency over plastic stents. SEMS expand to 810 mm in diam-
eter and do not occlude from bacterial biolm. In the US, commonly
available SEMS have an open mesh design and include the Wallstent
(Boston Scientic, Natick, MA), the Spiral Z-Stent and Zilver stent
(Cook Endoscopy, Winston-Salem, NC), and Flexxus (ConMed,
Utica, NY) (Table 4.2, Figs 4.12A and 4.12B). SEMS are made of
stainless steel or nitinol, a nickeltitanium alloy that provides a high
degree of exibility and is kink resistant. However nitinol is less
radiopaque than stainless steel and additional radiopaque (gold or
platinum) markers are added to the stents to improve radiopacity in
order to facilitate proper positioning during deployment. Covered
SEMS are also available. One example is the Wallstent (Boston Sci-
entic, Natick, MA) which has a polymer (Permalune) coating on
the inside of the stent except for the proximal and distal 1 cm. This
membrane is designed to prevent tumor in growth and prolong stent
patency.27
The delivery system for preloaded self-expanding metal stents
(SEMS) varies in design (Table 4.2). The wire mesh metal stents are
collapsed and constrained on a 6/6.5 Fr introducer catheter by an
Wallstent a Luminexx a Zilver a
(BS) (CM) (CE)
Material Elgiloy Nitinol Nitinol
5 6 7 Length (cm) 4/6/8/10 4/6/8/10 4/6/8
Diameter (mm) 8/10 8/10 6, 8, 10
Stent foreshortening Y N N
Introducer Diameter (Fr) 7.5/8 6/7.5 7
Table 4.2 Self-expandable biliary metal stents
a
According to the manufacturer Magnetic Resonance Imaging compatible.
CE = Cook Endoscopy, Winston-Salem, NC.
BS = Boston Scientic, Natick, MA.
CM = ConMed, Utica, NY.
37
 SECTION 1 GENERAL TOPICS
A A
B
Fig. 4.12 Self-expandable metal stent. A Endoscopic view of the
stent. B Fluoroscopy image of the stent.
A
B
Fig. 4.13 A Cytobrush (courtesy of Cook Endoscopy, Winston-
Salem, NC). B Fluoroscopy image showing biliary biopsy forceps
(arrow).
8/8.5 Fr overlying plastic sheath. Smaller 7/7.5 Fr introducer systems
are also available. The entire system is advanced over the guidewire
through the endoscope channel and passed under uoroscopic guid-
ance across the stricture using radiopaque markers. The Wallstent
delivery system allows recapture and repositioning of the stent
before reaching the 80% marker. Major limitations of currently avail-
able SEMS are cost and the difculty in removing uncovered stents
after placement.
Nasobiliary and pancreatic drainage catheters
Nasobiliary drainage catheters are used for temporary drainage of
the biliary tree and are available as 250 cm long 5 Fr to 7 Fr diameter
catheters with 5 or 9 sideports that facilitate drainage ow. Multiple
tip congurations are available. Nasopancreatic drainage catheters
are 5 Fr in diameter and may be used to drain the main pancreatic
duct after sphincterotomy or to irrigate and drain pancreatic pseu-
docysts. Biliary and pancreatic indwelling drainage catheters are
placed over the wire using a 0.035 inch guidewire. A nasal transfer
tube is needed for rerouting the tube from the mouth to the nose.
A connecting tube is needed for irrigation and drainage.
Tissue sampling devices
Brush cytology devices are available as single or multiple lumen
systems. Using the single-lumen cytology system, cell loss is inevi-
table because the brush is pulled back through the whole length of
the catheter. It is useful to aspirate bile from the catheter to collect
any dislodged cells within the catheter to improve the diagnostic
yield. Double-lumen cytology brush systems are preferable (Fig.
4.13a) and allow the guidewire and brush to pass through two
38
B
Fig. 4.14 A Fluoroscopy image of dilator balloon. B Soehendra
dilator (courtesy of Cook Endoscopy, Winston-Salem, NC).
B
A
Fig. 4.15 A Stent retriever (courtesy of Cook Endoscopy, Winston-
Salem, NC). B Endoscopic view showing stent retrieval (arrow).
separate lumens so that access is not lost. In addition, this design
minimizes cell loss by eliminating the need to pull back the brush
through the entire length of the catheter. Biliary biopsy forceps
(Olympus America Inc., Lehigh Valley, PA and ConMed, Utica, NY)
are useful for selectively obtaining biopsy specimens from the bile
duct under uoroscopy (Fig. 4.13b).28
Stricture dilation devices
In general, pancreaticobiliary dilation can be accomplished using
balloons (Fig. 4.14A) or bougies (Fig. 4.14B). Balloon dilators are
made of non-compliant polyethylene and are available in different
sizes and lengths: 4, 6, 8, or 10 mm in diameter and 24 cm long.
Balloons are passed over a guidewire through the accessory channel
of the endoscope. A radiopaque band proximal to the taper indicates
the point of maximal dilation. Soehendra dilators (Cook Endoscopy,
Winston-Salem, NC) are standard-shaped bougies that are available
in 611.5 Fr diameters. These are passed over a guidewire, the
10 Fr and 11.5 Fr size dilator require the use of a large accessory
channel.
Threaded-Tip Soehendra Stent Retrievers have also been used
to dilate very tight pancreaticobiliary strictures that otherwise allow
only passage of a guidewire. The wire-guided screw-tipped device is
used to negotiate high-grade strictures (Figs 4.15A and 4.15B). A
modied device is now commercially available as a dilator (Cook
Endoscopy, Winston-Salem, NC).
Stone extraction accessories
Accessories useful for stone extraction include double-lumen balloon
catheters, wire baskets and mechanical lithotriptors. The stone

extraction balloon (Fig. 4.16A) consists of a 56.8 Fr double-lumen
catheter with a balloon at the tip (818 mm size). Multi-size stone
extraction balloons are currently available (8.5, 12, and 15). Prior to
insertion into the endoscope it is useful to ensure that the balloon
inates correctly. The balloon catheter can be inserted over a guide-
wire or directly into the desired duct without guidewire.
Stones can also be removed using a wire basket (Fig. 4.16B). The
basket is shaped such that the wires open like a trap to engage the
stones. Basket function varies depending on the number of wires.
Newer design baskets can be advanced over a preplaced wire allow-
ing the basket to reach difcult areas (Trapezoid Basket, Boston
Scientic, Natick, MA and Flower basket, Olympus America Inc.,
Lehigh Valley, PA). The Trapezoid basket has a handle that is
designed to provide the endoscopist with the ability to perform
mechanical lithotripsy. Unfortunately, the small basket size limits
its efcacy in capturing and crushing large (>1.5 cm) stones.
Mechanical lithotriptors
Lithotripsy wire baskets facilitate removal of large (>1.5 cm) common
duct stones by crushing the stones before extraction. The original
B sizes.32 They have a small working channel (1.2 mm) that allows
A tations include the fragility of these devices, the small working
Fig. 4.16 A Endoscopic view of stone extractor balloon. B Stone
extractor basket (courtesy of Olympus America Inc., Lehigh Valley,
PA).
A
B
Fig. 4.17 A Soehendra mechanical lithotriptor Handle (courtesy of Cook Endoscopy, Winston-Salem, NC). B Through the scope
mechanical lithotriptor (courtesy of Olympus America Inc., Lehigh Valley, PA). C Fluoroscopy image of mechanical lithotriptor.
Chapter 4 Endoscopes, Guidewires and Accessories
Soehendra lithotriptor (Cook Endoscopy, Winston Salem, NC)
requires cutting the handle of the basket and removing the endo-
scope prior to stone fragmentation. This device consists of a 14 Fr
metal sheath and a self-locking crank handle (Fig. 4.17A). The litho-
triptor can be used with most standard stone extraction baskets.
Another mechanical lithotriptor is a pre-assembled through-the-
scope lithotripsy basket which can be inserted through a therapeutic
duodenoscope (Fig. 4.17B and 4.17C) (BML lithotripsy baskets by
Olympus Medical Inc., Lehigh Valley, PA). This device is available
in disposable and reusable versions. A single-piece disposable
mechanical lithotriptor with the basket, metal sheath and crank
handle is also available (Monolith, Boston Scientic, Natick, MA).
In one report, the disposal lithotripter was easy to use and its per-
formance was comparable to a standard reusable lithotriptor.29
Cholangioscopes
Duodenoscope-assisted cholangiopancreatoscopy (Figs 4.18A and
4.18B) allows for the direct visualization of the biliary and pancreatic
ducts. In the past, a dedicated motherdaughter system was required.
Currently, a variety of electronic and beroptic miniscopes are
available that can be passed through a 4.2 mm channel therapeutic
duodenoscope for direct visualization of the biliary and pancreatic
duct.3031 These instruments are now available in 8 Fr and 9 Fr
passage of small diameter forceps and bers for tissue acquisition
and for the application of laser and electrohydraulic lithotripsy. Limi-
channel and the need for two endoscopists (one for each endoscope).
Recently, a homemade support system was reported, allowing one
endoscopist to use the system.33 These devices are discussed in more
detail in Chapters 20 and 21.
Ultrasound probes
Increased availability of high frequency ultrasound probes have
made it possible for experts to use these devices for evaluating biliary
strictures. Endoscopic ultrasound probes are introduced free hand
or over the wire (Fig. 4.19) through the working channel of the
duodenoscope allowing for real time evaluation of biliary strictures
and surrounding vascular structures. Limited data suggest that these
39
 SECTION 1 GENERAL TOPICS
Fig. 4.18 A Cholangioscope (courtesy of Pentax Medical, Mont-
vale, NJ). B Fluoroscopy image of cholangioscope.
Fig. 4.19 Intraductal ultrasound probe (courtesy of Olympus
America Inc., Lehigh Valley, PA).
probes can enhance our ability to distinguish between benign and
malignant biliary strictures.34 Patient selection, operator experience
and cost continue to be limiting factors before wider use of this
technology is advocated.35
OTHER ACCESSORIES
Pharmacologic and chemical agents are not considered accessories
in the classic denition. However, secretin injection, with or without
the use of methylene blue has been reported to facilitate cannulation
of the pancreatic duct, especially in patients with pancreas divisum
(Fig. 4.20).36,37 These agents are also helpful in identifying the pan-
creatic duct opening after biliary sphincterotomy or endoscopic
ampullectomy. Glucagon and hyoscyamine often are used to relax
motility and have been found to be of similar efcacy but were never
compared with placebo in a randomized controlled trial.38
Accessories for use in patients with
altered anatomy
Standard accessories are designed for use with a duodenoscope and
are usually 200 to 260 cm in length. However, in patients with surgi-
cally altered anatomy such as Roux-en-Y gastroenteric anastomosis,
the standard ERCP scope may not be able to reach the ampulla, and
the use of a longer scope such as pediatric colonoscope or entero-
scope may be needed. Some accessories such as balloons, sphinc-
terotomes and push catheters are available in longer versions for
this purpose (Cook Endoscopy, Winston-Salem, NC). In addition,
40
Fig. 4.20 Endoscopic view showing identication of pancreatic
duct orice using methylene blue spray (arrow).
reverse sphincterotomes are available for use in patients with Bill-
roth II type anatomy. In most patients, however, sphincterotomy can
be performed using standard accessories such as needle-knife. The
endoscopist should make sure that appropriate accessories are avail-
able before initiating ERCP.
Single vs reusable accessories
The choice between single use, disposable and reusable accessories
for ERCP depends on various medical and economic factors.39 Addi-
tionally, there are liability issues that may result from reusing single
use devices. Several studies have evaluated reuse of ERCP accesso-
ries. Lee et al. found that a reusable sphincterotome could be safely
and efciently used.40 A disposable sphincterotome was cost-
effective after 2.2 uses and a reusable sphincterotome after 7.9 uses.
A recent study also found reusable sphincterotomes and stone
extraction baskets to be safe and cost-effective when compared with
single-use devices.41 According to a recent ASGE guideline on dis-
posable endoscopic accessories, the selection of reusable or dispos-
able devices must be based upon local purchase costs, reprocessing
costs and abilities, storage and disposable facilities, and personal
preferences.39
Storage of accessories
A special room with a uoroscopy unit offers the advantage of a
better oor plan, organization and ready access to stored accessories
that are required for the procedure (see Chapter 2). The room is
organized to facilitate the use of needed equipment such as
endoscope/processor, monitors, and the uoroscopy unit. The uo-
roscopy and endoscopy monitors should be placed side by side at
the endoscopist eye level to avoid the need for repeated turning,
which can displace scope position and additionally puts body strain
on the endoscopist. The dedicated ERCP room should be large
enough to house and store accessories in cabinets that are properly
labeled and easily accessible to the assistant at the time of procedure.
Easily retrievable accessories at the time of procedure will increase
procedure efciency by reducing time lost in looking for accessories
when needed.
RADIATION EXPOSURE
ERCP relies on the use of uoroscopy but the risks associated with
radiation exposure to patients and to personnel during the procedure
are not well documented (see Chapter 3). A recent prospective study
suggested that personnel as well as patients may be exposed to radia-
tion doses that equate to an estimated additional lifetime fatal cancer

risk of 1 in 35007000.42 Radiation exposure to the personnel is
proportionally related to the distance and duration of uoroscopy.
Higher voltage and lower current for uoroscopy is used to mini-
mize radiation exposure to the personnel. Various other strategies
that are used to minimize radiation exposure include protective lead
shielding, the use of digital imaging, and an under-couch X ray
emitter tube. Doubling the distance from the source reduces the
radiation dose received by a factor of 4. Therefore, the operator
should be vigilant in avoiding prolonged use of uoroscopy and
should stand as far back as possible from the patient during expo-
sure. Exposure to patients particularly those at high risk such as
young patients and pregnant women (discussed further in Chapter
23) can be minimized by shielding the pelvic area with a lead-lined
apron. In addition, obtaining hard copy uoroscopic images in lieu
of spot radiographs can further reduce exposure.43 Ongoing quality
assurance programs should be instituted with hospital radiation
safety ofcers.
CONCLUSIONS
The natural history of an endoscopic device involves an initial learn-
ing curve followed by rapid adoption and then a relatively slow phase
of stability followed by an incremental innovation (Fig. 4.21).44 ERCP
accessories have also followed this natural history. Major advances
have been made over the last decade. In general, many new acces-
sories are in fact evolutions of old ones as a result of innovations
by endoscopists in collaboration with product engineers and special-
ists. In addition, several products developed for other intra-luminal
interventions such as vascular, cardiac and urologic diseases have
similar applications in ERCP. Guidewires and expandable metal
stents are examples. Therefore, many new accessories are in fact
REFERENCES
1. Law N and Freeman ML. ERCP by using a prototype oblique-
viewing endoscope in patients with surgically altered anatomy.
Gastrointest Endosc 2004; 59:724728.
2. Igarashi Y, Tada T, Shimura J, et al. A new cannula with a exible
tip (swing tip) may improve the success rate of endoscopic
retrograde cholangiopancreatography. Endoscopy 2002;
34:628631.
3. Laasch HU, Tringali A, Wilbraham L, et al. Comparison of standard
and steerable catheter for bile duct cannulation in ERCP.
Endoscopy 2003; 35:669674.
4. Sherman S, Troiano FP, Hawes RH, Lehman GA. Sphincter of Oddi
manometery: decreased risk of clinical pancreatitis with the use of
modied aspirating catheter. Gastrointest Endosc 1990;
36:462466.
5. Wehrmann T, Stergiou N, Schmitt T, et al. Reduced risk for
pancreatitis after endoscopic micro transducer manometery of
the sphincter of Oddi: A randomized comparison with the
perfusion manometery technique. Endoscopy 2003; 35: 472477.
6. Rossos PG, Kortan P, Haber G. Selective common bile duct
cannulation can be simplied by the use of a standard
papillotome. Gastrointest Endosc 1993; 39:6769.
7. Schwacha H, Allgaier HP, Deibert P, et al. A sphincterotome-
based technique for selective transpapillary common bile duct
cannulation. Gastrointest Endosc 2000; 52: pp 387391.
8. Cortas GA, Mehta SN, Abraham NS, et al. Selective cannulation of
the common bile duct: a prospective randomized trial comparing
Chapter 4 Endoscopes, Guidewires and Accessories
Stability Incremental innovation
Simpler
Cheaper
Rapid adoption
Learning curve:
safety and efficacy issues
Time
Fig. 4.21 Graph illustrating the natural history of an endoscopic
device (adapted from reference # 44 with permission 2004
American Gastroenterological Association).
new applications or modications of available products. This
means that many products are approved using 510K pre-market
notication to the Food and Drug Administration. Therefore, many
products may not have rigorous pre-marketing evaluation and their
use may be based on word of mouth, personal experience or at best,
case series. Major limitations continue to be cost compared to reim-
bursement and the lack of cost-effectiveness and pre-marketing
studies.45,46
standard catheters with sphincterotomes. Gastrointest Endosc
1999; 50:775779.
9. Elta GH, Barnett JL, Wille RT, et al. Pure cut electrocautery current
for sphincterotomy causes less post-procedure pancreatitis than
blended current.Gastrointest Endosc. 1998 Feb; 47(2):149153.
10. Shah RJ, Antillon MR, Springer EW, et al. A new rotatable
papillotome (RP) in complex therapeutic ERCP: indications for use
and results. [abstract] Gastrointest Endosc (2003) 57: AB206.
11. Binmoeller KF, Seifert H, Gerke H, et al. Papillary roof incision
using the Erlangen-type pre-cut papillotome to achieve selective
bile duct cannulation. Gastrointest Endosc (1996) 44:689695.
12. Uchida N, Tsutsui K, Kamata H, et al. Precutting using a noseless
papillotome with independent lumens for contrast material and
guidewire. J Gastroenterol Hepatol 2005; 20:947950.
13. Al-Kawas FH. Biliary access during endoscopic retrograde
cholangiopancreatography: How to precut and a word of
caution! J Gastroenterol Hepatol 2005; 20:805806.
14. American Society of Gastrointestinal Endoscopy (ASGE).
Guidewires in gastrointestinal endoscopy. Gastrointest Endosc
1998; 47:579583.
15. Jacob L, Geenen JE. ERCP guide wires. Gastrointest Endosc 1996;
43:5760.
16. Lella F, Bagnolo F, Colombo E, et al. A simple way of avoiding
post-ERCP pancreatitis. Gastrointest Endosc 2004; 59:830834.
17. Farrell RJ, Howell DA, Pleskow DA. New technology for
endoscopic retrograde cholangiopancreatography: improving
41
 SECTION 1 GENERAL TOPICS
safety, success, and efciency. Gastrointest Endoscopy Clin N Am
2003; 13:539559.
18. Aliperti G, Branch S, Geisman R, et al. Comparison of new rapid
exchange technique with standard device during ERCP. A
multicenter trial. Digestive Disease Week, May 1619, Orlando, FL.
19. Beilstein MC, Ahmad NA, Kochman ML, et al. Initial evaluation of a
duodenoscope modied to allow guidewire xation during ERCP.
Gastrointest Endosc 2004; 60:284287.
20. Joyce AM, Ahmad NA, Kochman ML, et al. Multicenter
comparative trial of the V-System for therapeutic ERCP.
Gastrointest Endosc 2005; 61:AB208.
21. Siegel JH, Pullano W, Kodsi B, et al. Optimal palliation of
malignant bile duct obstruction: Experience with endoscopic 12
French prostheses. Endoscopy 1988; 20:137141.
22. Catalano M, Geenen J, Lehman G, et al. Tannenbaum: Teon
stents versus traditional polyethelene stents for treatment of
malignant biliary stricture. Gastointest Endosc 2002; 55: 354358.
23. Costamagna G, Mutignani M, Rotondano G, et al. Hydrophilic
hydromer-coated polyurethane stents versus uncoated stents in
malignant biliary obstruction: a randomized trial. Gastrointest
Endosc. 2000; 51(1):811.
24. Rashdan A, Fogel EL, McHenry L Jr, et al. Improved stent
characteristics for prophylaxis of post-ERCP pancreatitis. Clin
Gastroenterol Hepatol. 2004; 2(4):322329.
25. Raju GS, Gomez G, Xiao SY, et al. Determinants of pancreatic
injury induced by short-term indwelling stents and modulation of
the same by a novel stent design. Gastrointestinal Endoscopy
2004; 59(5):106.
26. Tarnasky PR, Morris J, Hawes RH, et al. Snare beside-a-wire biliary
stent exchange: a method that maintains access across biliary
strictures. Gastrointest Endosc 1996; 44:185187.
27. Isayama H, Komatsu Y, Tsujino T, et al. A prospective randomized
study of covered versus uncovered diamond stents for the 42. Naidu LS, Singhal S, Preece DE, et al. Radiation exposure to
management of distal malignant biliary obstruction. Gut. 2004;
53(5):729734.
28. Tamada K, Higashizawa T, Tomiyama T, et al. Ropeway-type bile
duct biopsy forceps with a slit for a guidewire. Gastrointest
Endosc 2001; 53:8992.
29. Sorbi D, Van Os EC, Aberger FJ, et al. Clinical application of a new
disposable lithotripter: a prospective multicenter study.
Gastrointest Endosc 1999; 49:210213.
30. Shim CS, Neuhaus H, Tamada K. Direct cholangioscopy.
Endoscopy 2003; 35:752758.
31. Kodama T, Tatsumi Y, Kozarek RA, et al. Direct pancreatoscopy.
Endoscopy 2002; 34:653660.
32. Kodama T, Tatsumi Y, Sato H, et al. Initial experience with a new
peroral electronic pancreatoscope with an accessory channel.
Gastrointest Endosc. 2004; 59:895900.
42
33. Farrell JJ, Bounds BC, Al-Shalabi S, et al. Single-operator
duodenoscope-assisted cholangioscopy is an effective alternative
in the management of choledocholithiasis not removed by
conventional methods, including mechanical lithotripsy.
Endoscopy. 2005; 37(6):542547.
34. Levey MJ, Vasquez-Sequerios E, Wiersema MJ. Evaluation of
pancreatobiliary ductal system by intraductal US. Gastrointest
Endosc 2002; 55:397408.
35. Jha R, Al-Kawas FH. Nothing quite new is perfect. How good is
IDUS in patients with isolated biliary strictures? Am J
Gastroenterol 2004; 99:16901691.
36. Devereaux BM, Fein S, Purich E et al. A new synthetic porcine
secretin for facilitation of cannulation of the dorsal pancreatic
duct at ERCP in patients with pancreas divisum: a multicenter,
randomized, double-blind comparative study. Gastrointest Endosc
2003; 57:643647.
37. Park SH, de Bellis M, McHenry L et al. Use of methylene blue to
identify the minor papilla or its orice in patients with pancreas
divisum. Gastrointest Endosc 2003; 57:358363.
38. Lahoti S, Catalano MF, Geenen JE, et al. A prospective, double-
blind trial of L-hyoscyamine versus glucagon for the inhibition of
small intestinal motility during ERCP. Gastrointest Endosc 1997;
46:139142.
39. ASGE Technology Status Evaluation Report: disposable
endoscopic accessories. Gastrointest Endosc 2005; 62:
477479.
40. Lee RM, Vida F, Koarek RA, et al. In vitro and in vivo evaluation of
a reusable double-channel sphincterotome. Gastrointest Endosc
1999; 49:477482.
41. Prat F, Spieler JF, Paci S, et al. Reliability, cost-effectiveness, and
safety of reuse of ancillary devices for ERCP. Gastrointest Endosc
2004; 60:246252.
personnel performing endoscopic retrograde
cholangiopancreatography. Postgrad Med J 2005; 81:
660662.
43. Axelrad AM, Fleischer DE, Strack LL, et al. Performance of
ERCP for symptomatic choledocholithiasis during
pregnancy: techniques to increase safety and improve
patient management. Am J Gastroenterol 1994; 89(1):
109112.
44. Pasricha PJ. The future of therapeutic endoscopy. Clin
Gastroenterol Hepatol. 2004; 2(4):286289.
45. Narain MA, Cockel R. How should endoscopic accessories be
selected: trial and error? Gut 2000; 46:305306.
46. Ganz RA. The development and implementation of new
endoscopic technology: what are the challenges? Gastrointest
Endosc 2004; 60:592598.
 SECTION 1 GENERAL TOPICS
Chapter
Sedation and Analgesia for ERCP
5 Gregory Zuccaro Jr
INTRODUCTION
From the technical perspective, endoscopic retrograde cholangio-
pancreatography (ERCP) is unlike routine upper endoscopy and
colonoscopy. The instrument itself is side viewing rather than
forward viewing. The procedure goal is not to simply reach the
cecum or duodenum, but rather to cannulate small openings with
even smaller devices. The time to completion is often signicantly
longer. We now require trainees to spend extra time acquiring these
advanced skills before we consider them competent to perform this
demanding procedure. However, in most units, sedation and anal-
gesia is provided in a fashion similar to routine procedures. This
chapter will address issues of sedation and analgesia for ERCP, with
emphasis on those aspects that might distinguish it from routine
endoscopy.
An increasing number of endoscopists have chosen to have
another provider (i.e. a nurse anesthetist or anesthesiologist) provide
sedation for their patients, including ERCP. In some units, deep
sedation (dened below) is administered, and in others general anes-
thesia is routine. The comments in this chapter are directed toward
those units and endoscopists that provide sedation and analgesia
themselves.
PHYSIOLOGY OF SEDATION AND
ANALGESIA FOR ERCP
As endoscopists we describe the sedation and analgesia we provide
as conscious sedation. Not only is that term oxymoronic, but it
further incorrectly implies that all sedation and analgesia for endos-
copy is the same. To explore this fully, it is essential to review the
American Society of Anesthesiologists statement on the Continuum
of Sedation.1 In this statement, four discrete levels of sedation are
described (Table 5.1).
For most routine endoscopy, it is assumed that what we incor-
rectly and imprecisely refer to as conscious sedation actually corre-
sponds to the ASAs category of moderate sedation. However, note
the differences between moderate sedation, and the next category,
referred to as deep sedation. In moderate sedation, the patient is able
to make a purposeful response to verbal or verbal plus gentle
tactile stimulation; for example, the patient might be sleeping, but
arouses with calling of his/her name or gentle tap on the shoulder,
and can give a thumbs up or other positive response when asked.
Due to the length and difculty of some ERCP procedures, the level
of sedation is often further along the continuum. In the denition
for deep sedation, repeated or even painful stimulation may be nec-
essary to obtain a response. Many endoscopists will admit that for
at least part of the time, their ERCP patients responsiveness better
ts the deep rather than moderate category. We have found that,
during serial assessments of level of sedation during ERCP, 85% of
patients are deeply sedated for a portion of the procedure.2 Obvi-
ously, responsiveness alone during an ERCP procedure is not of
high importance, in that the patients cooperation or movement is
not necessary to complete the procedure. Rather, it is the cardiopul-
monary correlates of this level of responsiveness that are essential
to consider.
It must be recognized that sedation is a continuum, and that it
is impossible for the endoscopist to target a specic level of sedation
with accuracy. More frequently, patients undergoing sedation for
gastrointestinal endoscopy may go from minimal to moderate to
deep sedation during the same procedure. A tenet of the ASA guide-
lines is that skills must be present to rescue patients from a level of
sedation deeper than the intended target; i.e. if moderate sedation
is the goal, skills must be present to effectively manage any situation
that might arise in deep sedation, thereby bringing the patient back
to a state of moderate sedation. Similarly, if the goal is provision of
deep sedation, skills must be present to effectively manage any situ-
ation that might arise in general anesthesia.
SAFETY OF SEDATION AND ANALGESIA
Before we discuss the specic actions appropriate for deep sedation
during ERCP, we should consider what is known about the safety of
our sedation practice. It may be argued that the safety of existing
endoscopic practice is not entirely known. Our colleagues in anes-
thesiology have devoted considerable time dening the safety of
general anesthesia, and it is helpful to consider their literature as a
prelude to analyzing ours. Death does occur as a result of anesthesia.
Prior to the 1980s, there were approximately 2 deaths per 10 000
anesthesias administered. The Institute of Medicine has asserted
that this rate has fallen to approximately 1 death per 200 000 to
300 000 anesthesias administered.3 This dramatic change is attrib-
uted to improved monitoring, development and implementation of
practice guidelines, and other systematic approaches to reducing
errors.47 To corroborate this, Legasse reviewed mortality rates from
anesthesia published by 21 different investigators, and further exam-
ined anesthesia experience in a suburban vs urban hospital system
between 1992 and 1994. Preventable deaths persist despite an
ongoing quality improvement process. Peer review determined that
human error contributed to 2.6% of deaths in the suburban system,
and 4.7% of deaths in the urban hospital system.8 Not all deaths
related were to error; among other factors patients preoperative ASA
physical status (Box 5.1) correlated with anesthesia mortality.9
A few conclusions may be drawn from reviewing the anesthesia
literature:
Morbidity and mortality rates, and factors that contribute to them,
can be identied in the literature.
43
 SECTION 1 GENERAL TOPICS

Minimal sedation Moderate sedation/analgesia

Responsiveness
Airway
Spontaneous Ventilation
Cardiovascular Function
(anxiolysis) (conscious sedation) Deep sedation/analgesia General anesthesia
Normal response to Purposeful response to Purposeful response Unarousable even with
verbal stimulation verbal or tactile stimulation following repeated or painful stimulus
painful stimulation
Unaffected No intervention required Intervention may Intervention often
be required required
Unaffected Adequate May be inadequate Frequently inadequate
Unaffected Usually maintained Usually maintained May be impaired

Table 5.1 ASA levels of sedation (reproduced with permission from the American Society of Anesthesiologists)

BOX 5.1 ASA PHYSICAL CLASSIFICATION Conventional sedation Adverse effects Death
10
SYSTEM Silvis NR 0.001%
Quine11 0.07% 0.03%
Arrowsmith12 0.5% 0.03%
(reproduced with permission from the American Society of Sieg13 0.008% 0%
Anesthesiologists) Propofol

Class 1: Patient has no organic, physiologic, biochemical or Walker37 0.3% 0

Rex38 0.2% 0
psychiatric disturbance. The pathologic process for which
Huss39 0.3% 0
operation is to be performed is localized and does not entail
Clark40 0.38% 0
systemic disturbance.
Table 5.2 Large case series reecting mortality and severe
Class 2: Mild to moderate systemic disturbance caused either cardiopulmonary adverse eventsa related to sedation and
by the condition to be treated surgically or by other analgesia for endoscopy
a
pathophysiologic processes. including aspiration, laryngospasm, respiratory arrest.

Class 3: Severe, systemic disturbance or disease from whatever

cause, even though it may not be possible to dene the
degree of disability with nality.
Class 4: Severe systemic disorders that are already life
threatening, not always correctable by operation.
Class 5: The moribund patient who has little chance of survival
but is submitted to operation in desperation.
As expected, sicker patients are at increased risk of adverse out-
comes from anesthesia.
Morbidity and mortality rates can be improved with better monitor-
ing, training, and implementation of practice guidelines and
systems to reduce error.
Despite all efforts to improve anesthesias safety record, human
error continues to cause death.
How do endoscopists fare with respect to sedation and analgesia?
Considering the millions of endoscopic procedures performed every
year, safety data for sedation and analgesia are relatively scant (Table
5.2). Most clinical series and controlled trials focus on objective
endpoints such as oxygen desaturation, use of reversal agents, or
subjective endpoints like patient satisfaction. Many of these studies
are not large enough to allow relevant observations on morbidity and
mortality. In the 1974 ASGE endoscopy survey, three deaths were
attributed to sedation (0.001%).10 In the early 1990s Quine surveyed
36 hospitals where upper GI endoscopic procedures were performed
and reported death likely attributable to sedation in 0.03%, and
cardiopulmonary arrest in 0.07%. Midazolam and diazepam were
equally likely to have been the benzodiazepine utilized. There was a
strong correlation between higher doses of benzodiazepines and
lack of patient monitoring with adverse outcomes.11 Arrowsmith
analyzed data from over 21 000 procedures, and determined a death
rate of 0.03% related to sedation and analgesia and 0.5% due to car-
diorespiratory arrest.12 Complications were similar with midazolam
and diazepam. In distinction, Sieg et al. found no deaths and 16
signicant cardiopulmonary adverse events related to sedation
(0.008%) among 190 000 EGD and colonoscopy procedures per-
formed in several facilities.13
PRACTICE GUIDELINES FOR SEDATION AND
ANALGESIA BY NON-ANESTHESIOLOGISTS
The American Society of Anesthesiologists have developed practice
guidelines for the provision of sedation and analgesia by non-
anesthesiologists, which may apply to all settings outside the operat-
ing room, including the emergency department, pulmonary and
cardiology suites, and the gastroenterology lab.14 These guidelines
were developed by a rigorous analytic process, in an attempt to create
evidence-based guidelines. A series of statements regarding all
aspects of care for sedated patients was generated, focusing on a
clinical intervention and the desired outcome (example: availability
of an individual solely dedicated to patient monitoring and safety
improves clinical efcacy and/or reduces adverse outcomes). Then,
the available evidence from the literature for each statement was

44

judged as supportive (sufcient information from adequately designed
studies to describe a statistically signicant relationship between an
intervention and a clinical outcome), suggestive (evidence from case
reports and descriptive studies provides a directional assessment of
the relationship between an intervention and clinical outcome, but
the type of information does not permit a statistical assessment of
signicance), equivocal (while studies exist, no clear direction can be
determined for clinical outcomes related to an intervention), incon-
clusive (while studies exist they are not useful to judge a relationship
between an intervention and a clinical outcome), insufcient (too
few studies to assess a relationship) and silent (no studies were
identied).
In addition to an assessment of the literature, a group of consul-
tants from all specialties administering sedation was assembled, and
their opinion sought on the validity of each of the statements, rated
on a scale of 1 (strongly disagree) to 5 (strongly agree).
Summarized below is a portion of the literature assessment,
consultant opinion, and recommendations for moderate and deep
sedation for each aspect of patient care discussed in the ASA guide-
lines, followed by selected comments, focusing specically on those
aspects most relevant to ERCP:
Patient evaluation
Statement: Pre procedure patient evaluation improves clinical efcacy
and/or reduces adverse outcomes.
Literature: Insufcient
Consultants: Strongly agree
Summary of recommendations: Elements of the pre-sedation history
should be established prior to the procedure, including
abnormalities of any major organ system, previous
experiences with sedation/anesthesia/surgery, current
medications, history of allergic reactions, fasting inter-
val prior to the procedure or test, and history of tobacco,
alcohol, or substance abuse. Elements of the physical
exam include vital signs, auscultation of heart and
lungs, evaluation of the airway, head and neck. Labora-
tory testing should be guided by the patients underly-
ing conditions, the procedure to be performed, and the
likelihood that results of any test might inuence the
management of sedation.
Comments: The airway consists of those structures through which
air passes during ventilation, from the nose to the
alveoli. A detailed discussion of the anatomy, physiol-
ogy and response to sedation of the entire airway is
beyond the scope of this review. There are, however,
salient facts with which individuals administering
sedation and analgesia must be familiar. Adverse out-
comes in patients receiving sedation and analgesia are
frequently due to airway difculties.1517 In particular,
the upper airway is the weak link where airway
obstruction is most likely to occur. With sedation,
obstruction of the airway may occur due to the tongue
falling posteriorly, causing obstruction at the level of
the oropharynx.18,19 Analysis of sedated patients identi-
ed decreased diameter of the pharynx at the level of
the soft palate and epiglottis as a potential cause of
airway obstruction.20 Airway patency is controlled by a
complex neuromuscular system. There are two main
inuences on the control of upper airway muscles. The
rst involves receptors throughout the upper and lower
Chapter 5 Sedation and Analgesia for ERCP
airway which respond to mechanical and chemical
stimuli by affecting tone of the upper airway muscles;
this mechanism may help to prevent aspiration. The
second inuence on upper airway muscles is related to
the level of wakefulness of the patient (e.g. snoring
reects an increased airway resistance). Agents used in
moderate or deep sedation affect the upper airway
muscles more than the diaphragm and produce more
depressant effects on airway function than on diaphrag-
matic function.2123 Topical anesthetics may contribute
to airway obstruction by removing airway mediated
reex brainstem stimulation.24
Given the potential for ventilatory problems sur-
rounding the upper airway, the ASA as well as individ-
ual practitioners have focused considerable attention
on recognition of patients who are at greater risk of
airway failure before any sedative or analgesic agent is
administered.14 This evaluation is intended to identify
patients in whom positive pressure ventilation, with or
without tracheal intubation, may be difcult or impos-
sible. The patient history should identify prior prob-
lems with anesthesia or sedation, including difcult
airway. Patients with sleep apnea or habitual stridorous
snoring are also at high risk for airway problems related
to sedation. Other risk factors include advanced rheu-
matologic or osteoarthritic cervical spine disease, and
chromosomal abnormalities such as trisomy 21. Head
and neck ndings potentially associated with difculty
in positive pressure ventilation include a short neck
with limited neck extension, decreased hyoid-mental
distance (<3 cm in an adult), presence of a neck
mass, evidence of cervical spine disease or trauma,
tracheal deviation, or dysmorphic facial features (as in
Pierre-Robin syndrome). Additional physical ndings
that raise the possibility of difcult positive pressure
ventilation include a small mouth opening (<3 cm in
adult), a high, arched palate, macroglossia, tonsillar
hypertrophy, a non-visible uvula, micrognathia, retrog-
nathia, trismus or signicant malocclusion of the jaw.
Langeron has noted the following factors predicting
difcult mask ventilation: presence of a beard, BMI
>26, lack of teeth, age >55 years and a history of
snoring.25
The presence of any of these ndings on history or
physical examination does not necessarily imply that
only an anesthesiologist may safely provide sedation or
analgesia for endoscopy; rather, they are signals that
the use and amount of sedation, type of procedure
contemplated, or other factors might be modied to
minimize risk to the patient. In extreme examples, or
when there is signicant doubt regarding safety,
involvement of an anesthesiologist is reasonable.
Pre procedure preparation
Statement: Pre procedure patient preparation (e.g. counseling, fasting)
improves clinical efcacy and/or reduces adverse
outcomes.
Literature: Insufcient
Consultants: Agree for moderate sedation, strongly agree for deep
sedation
45
 SECTION 1 GENERAL TOPICS
Summary of recommendations: Informed consent should be obtained
in all cases. Patients should fast for the appropriate
time preprocedure so as to ensure complete gastric
emptying. Patients receiving sedation should be aware
of the proscription against driving while the effects of
sedation persist. Post-procedure instructions should be
discussed prior to sedation and written instructions
provided.
Monitoring
Statement: Patient monitoring (i.e. level of consciousness, respiratory
function, hemodynamics) at regular intervals improves
clinical efcacy and/or reduces adverse outcomes.
Literature: Insufcient
Consultants: Strongly agree
Summary of recommendations: Monitoring of patient response to
verbal commands (for moderate sedation) or for stron-
ger stimuli (for deep sedation) should be routine. All
patients should be monitored by pulse oximetry. Ven-
tilatory function should be monitored by observation
and/or auscultation. Blood pressure and pulse should
be monitored at 5-minute intervals during the proce-
dure. EKG monitoring should be performed in all
patients undergoing deep sedation.
Comments: Just as has occurred in the eld of anesthesiology,
automated advance monitoring techniques have been
increasingly utilized in gastrointestinal endoscopy.
Pulse oximetry has become a de facto standard of care
during sedation and analgesia for endoscopy, owing to
the evidence that clinical observation alone is inaccu-
rate in the detection of hypoxemia and that supplemen-
tal oxygen can minimize the degree of desaturation and
hopefully its deleterious effects. To date, neither pulse
oximetry nor supplemental oxygen administration has
yet been shown to decrease the severity or incidence
of cardiopulmonary complications. It is important to
point out that pulse oximetry does not measure alveolar
hypoventilation, which is measured by hypercapnea or
a rise in arterial carbon dioxide pressure. Furthermore,
it must be remembered that a pulse oximeter is not an
apnea monitor; the early stages of signicant hypoven-
tilation may be concurrent with a normal pulse oxime-
try, particularly if the patient is receiving supplemental
oxygen. Although oxygen administration may prevent
hypoxemia and its deleterious effects, it will not detect
the development of hypercapnea. Deleterious conse-
quences of alveolar hypoventilation include myocardial
depression, acidosis, intracranial hypertension, narco-
sis, and arterial hypertension or hypotension.
Transcutaneous CO2 monitoring (PtcCO2) is a
non-invasive method for measuring arterial CO2. An
electrode is placed on the skin, which is heated to arte-
rialize the microcirculation. CO2 then diffuses through
the skin and into an electrolyte solution between the
skin/electrode interface, which produces carbonic acid.
A pH reading is then rendered and the CO2 value is
obtained via the Henderson-Hasselbach equation.
Nelson et al. utilized this technique in 395 patients
undergoing ERCP and concluded that it did prevent
severe CO2 retention better than standard monitoring.26
46
Nevertheless, this is a cumbersome technique not typi-
cally utilized in gastrointestinal endoscopy.
Capnography is based on the principle that carbon
dioxide absorbs light in the infrared region of the elec-
tromagnetic spectrum. Quantication of the absorp-
tion leads to the generation of a curve, which represents
a real-time display of the patients respiratory activity
(Fig. 5.1A5.1C). Capnography has been utilized to
allow the safe titration of propofol by a qualied gastro-
enterologist during ERCP and EUS.27
Bispectral index (BIS) monitoring represents a
complex mathematical evaluation of electroencephalo-
graphic parameters of frontal cortex activity, corre-
sponding to varying levels of sedation. The BIS scale
varies from 0 to 100 (0, no cortical activity or coma;
4060, unconscious; 7090, varying levels of conscious
sedation, 100, fully awake). Theoretically this index
should reect the same level of sedation regardless of
the medications used, except for ketamine. In a pre-
liminary observational study, involving 50 patients
undergoing ERCP, colonoscopy, and upper endoscopy,
BIS levels were found to correlate with a commonly
used score for the degree of sedation.28 A BIS range of
7585 demonstrated a probability of >96% that the
patient would exhibit an acceptable sedation score.
However, there was increasing variability of the BIS
score with deeper levels of sedation. Additionally, there
was no correlation between the BIS score and standard
physiologic parameters such as pulse oximetry, blood
pressure or heart rate. The BIS algorithm employed for
this study was validated only for deeper levels of seda-
tion, and therefore, may not be a sensitive indicator of
moderate levels of sedation and analgesia that are uti-
lized during endoscopy. Although only pulse oximetry
is currently routinely utilized, these other advanced
monitoring techniques merit further study, particularly
given the increased frequency of undertaking deep
sedation for endoscopic procedures.
Availability of an individual responsible for
patient monitoring
Statement: Availability of an individual who is dedicated solely to
patient monitoring and safety improves clinical efcacy
and/or reduces adverse outcomes.
Literature: Silent
Consultants: Agree for moderate sedation, strongly agree for deep
sedation
Summary of recommendations: An individual (not the endoscopist)
should be present to monitor the sedated patient.
During deep sedation, this person should have no other
responsibilities. For moderate sedation, this individual
may perform minor, interruptible tasks provided that
adequate mechanical and personal monitoring is
maintained.
Comments: Fundamental to this discussion is the choice each unit
makes as to the level of sedation they provide. If a unit
contends that they provide moderate sedation, then
their patients should meet those criteria, particularly
with respect to level of responsiveness. If a unit
 Chapter 5 Sedation and Analgesia for ERCP

A
mmHg
60

30

0
CO2

B
mmHg
60

30

0
CO2

mmHg
60

30

0
CO2

SPO2

Fig. 5.1 A Capnography tracing reecting normal respiration. B Capnography tracing showing an interruption in regular respiration,
followed by resumption of normal breathing. C Simultaneous EKG, capnography and pulse oximetry tracings.

acknowledges deep sedation, then the appropriate level
of monitoring must take place, both personal and
automated.
Comments: Propofol is a highly lipophilic compound formulated in
Anesthetic induction agents used for sedation and
analgesia (e.g. propofol)
Statement: Intravenous sedation/analgesic medications specically
designed to be used for general anesthesia (e.g. propofol)
improve clinical efcacy and/or reduce adverse outcomes.
Literature: Suggestive for moderate sedation, insufcient for deep
sedation
Consultants: Equivocal
Summary of recommendations: Patients receiving propofol should be
monitored as for deep sedation, even if moderate seda-
tion is the target level. Practitioners should be qualied
to rescue patients from any level of sedation, including
general anesthesia.
a glycerol, egg phosphatide and soybean emulsion. It
is a substituted phenolic compound not related struc-
turally to barbiturates, narcotics, or benzodiazepines.
Due to its lipophilic nature, immediately upon admin-
istration it distributes rst to peripheral tissues and the
central nervous system. It is metabolized in the liver,
and a conjugated metabolite is excreted in the urine. It
is rapidly cleared from the blood, both due to its distri-
bution to the central nervous system and peripheral
tissues, and due to effective hepatic metabolism. The
presence of liver disease or renal disease does not

47
 SECTION 1 GENERAL TOPICS
appear to signicantly alter the effective dosing of this
agent.29,30
Propofol possesses sedative and amnesic, but very
limited analgesic properties. It may be utilized for mod-
erate and deep sedation for minor procedures such as
endoscopy, deep sedation in intensive care settings,
and for induction and maintenance of general anesthe-
sia. Similar to benzodiazepines, the sedative effects of
propofol are mediated primarily through gamma-ami-
nobutyric acid (GABA) receptors in the brain. The
binding of GABA to its receptors, located in the cere-
brum and cerebral cortex, leads to decreased ability of
neurons to generate an action potential. Propofol binds
to receptors which increase GABA afnity for its recep-
tor, thereby producing decreased cognition, sensory,
memory and motor function.31 Propofol does not work
on the exact same GABA receptors as the benzodiaze-
pines, and therefore pharmacologic antagonists for
benzodiazepines do not reverse the effects of pro-
pofol.32 Sedation and amnesia are dose-dependent;
however, propofol produces less amnesia than mid-
azolam at equivalent sedative doses.33,34
There may be considerable adverse hemodynamic
effects from propofol, particularly at doses required for
deep sedation or general anesthesia. Cerebral blood
ow may decrease by 50%, and systemic blood pressure
by a third. Cardiac output may fall without a compensa-
tory rise in pulse rate.35 These effects may be accentu-
ated with concomitant use of a parenteral narcotic.
Effects on blood pressure and ventilatory effort are
accentuated in the elderly. It has been suggested that
maintaining a constant blood level by continuous
infusion of propofol will result in less hemodynamic
compromise compared to periodic bolus infusion with
resultant peak levels.36
There are in fact multiple clinical series and ran-
domized controlled trials that support the use of pro-
pofol by gastroenterologist-nursing teams. An attribute
of a large case series is that some comment can be
made regarding safety and efcacy, and while smaller
in patient number, controlled trials offer a comparison
between two or more sedation and analgesia regimens.
In the majority of these case series and controlled
trials, propofol was administered in the periodic bolus
technique.
The largest case series where propofol was admin-
istered by an endoscopist-nurse team was published by
Walker et al.37 Deep sedation was achieved in the major-
ity of the 9152 cases. Due in large part to the rigorous
training the care givers received, propofol sedation was
quite safe, with no reported deaths, and only seven
REFERENCES
1. American Society of Anesthesiologists Website; www.asahq.
org.2004.
2. Patel S, Vargo JJ, Khandwala F, et al. Deep sedation
occurs frequently during elective endoscopy with
Meperidine and Midazolam: Am J Gastroenterol 2005;
100(12):26892695.
48
instances of respiratory compromise which responded
to temporary support with bag and mask ventilation.
Rex, et al., reported a similar series of 2000 patients
receiving propofol sedation with periodic bolus by an
endoscopist-nursing team.38 Again, propofol delivery
was quite safe in the hands of this highly trained and
focused team, with no deaths and only four patients
required temporary bag and mask ventilatory support.
Heuss, et al., provided propofol for 2574 patients
undergoing a variety of endoscopic procedures and
found a similar safety prole.39 Clarke et al. also report
a similar safety prole in over 28 000 patients.40 In
these, the largest case series of endoscopist-nurse
administered propofol, most patients requiring ventila-
tory support were undergoing upper endoscopy. Rex
has speculated that increased bolus dosing to facilitate
esophageal intubation contributes to the increased like-
lihood of transient ventilatory insufciency.41
Several prospective, randomized controlled trials
have compared endoscopist-nurse administered propo-
fol with conventional parenteral sedation with narcotic
and benzodiazepine combinations.4245 In our experi-
ence with gastroenterologist-nurse administered pro-
pofol for ERCP and EUS, we found that while patient
satisfaction was similar, recovery time was signicantly
shortened with propofol compared to conventional
sedation, and patients had a higher recovery of baseline
activity level and dietary intake 24 hours after the pro-
cedure.42 While the study was powered only to detect a
difference in recovery times, physiologic monitoring
parameters showed a trend of actually being better in
the propofol group as compared to the conventional
sedation group. That said, it is clear from our experi-
ence that propofol sedation requires a highly motivated
and trained team of gastroenterologists and nurses.
Moreover, in most units, if propofol is given as a sole
sedation agent, it is more likely to be administered by
an anesthesiologist or similarly skilled individual.
SUMMARY
Our data indicate that a signicant number of patients
undergoing ERCP will at some point of the procedure be
deeply sedated. Both personal and automated monitoring of
these patients should be appropriate for this level of sedation.
The ability to rescue a patient from any level of sedation
should be available, if necessary, as spelled out by the ASA
guideline.
3. Committee on Quality of Health Care in America IOM: To Err is
Human: Building a Safer Health System. Edited by Kohn L,
Corrigan J, Donaldson M. Washington, National Academy Press,
1999; 241.
4. Sentinel events: approaches to error reduction and prevention.
J Comm J Qual Improv 1998; 24(4):175186.

5. Gaba DM. Anesthesiology as a model for safety in health care.
BJM 2000; 320:785788.
6. Stoelting R. APSF response to IOM medical error report.
Anesthesia Patient Safety Foundation Newsletter 2000;
15(1):1.
7. Cooper JB, Gaba DM, Liang B, et al. The National Patient Safety
Foundation agenda for research and development in patient
safety. Med Gen Med 2000; E38.
8. Lagasse R, Anesthesia safety; model or myth? Anesthesiology
2002; 09:16091617.
9. American Society of Anesthesiologists Website; www.asahq.org
2004.
10. Silvis SE, Nebel O, Rogers G, et al. Endoscopic complications:
Results of the 1974 American Society for Gastrointestinal
Endoscopy survey. JAMA 1976; 235:928930.
11. Quine MA, Bell GD, McCloy RF, et al. Prospective audit of upper
gastrointestinal endoscopy in two regions of England: safety,
stafng, and sedation methods. Gut 1995; 36:462467.
12. Arrowsmith JB, Gerstman BB, Fleischer DE, et al. Results from the
American Society for Gastrointestinal Endoscopy/U.S. Food and
Drug Administration collaborative study on complication rates
and drug use during gastrointestinal endoscopy. Gastrointestinal
Endoscopy 1991; 37:421427.
13. Sieg A, Hachmoeller-Eisenbach U, Eisenbach T. Prospective
evaluation of complications in outpatient GI endoscopy: a survey
among German gastroenterologists. Gastrointestinal Endoscopy
2001; 53:620627.
14. Gross JB, Bailey PL, Caplan RA, et al. Practice guidelines for
sedation and analgesia by non-anesthesiologists. Anesthesiology
2002; 96:10041017.
15. Keenan RL, Boyan CP. Cardiac arrest due to anesthesia, JAMA
1985; 252(16):2373.
16. Caplan RA, Posner KL, Ward RJ, et al. Adverse respiratory events
in anesthesia: a closed claims analysis, Anesthesiology 1990;
72:828.
17. Cheney FW, Posner KL, Caplan RA. Adverse respiratory events
infrequently leading to malpractice suits: a closed claims analysis,
Anesthesiology 1991; 75:932.
18. Morikawa S, Safar P, DeCarlo J. Inuence of the head-jaw position
upon upper airway patency, Anesthesiology 1961; 22:265.
19. Safar P, Escarraga LA, Chang F. Upper airway obstuction in the
unconscious patient, J Appl Physiol 1959; 14:760.
20. Shorten GD, Opie NJ, Graziotti P, et al. Assessment of upper
airway anatomy in awake, sedated and anaesthetised patients
using magnetic resonance imaging, Anaesth Intensive Care 1994;
22:165.
21. Hwang J-C, St. John WM, Bartlett D Jr. Respiratory-related
hypoglossal nerve activity: inuence of anesthetics, J Appl Physiol
1983; 55:785.
22. Nishino T et al. Comparison of changes in the hypoglossal and
the phrenic nerve activity in response to increasing depth of
anesthesia in cats. Anesthesiology 1984; 60:19.
23. Hillman DR, Platt PR, Eastwood PR. The upper airway during
anesthesia. BR J Anaesth 2003; 91(1):3139.
24. Deegan PC, Mulloy E, McNicholas WT. Topical oropharyngeal
anesthesia in patients with obstructive sleep apnea. Am J Respir
Crit Care med 1995; 151(4):11081112.
25. Langeron O, Masso E, Huraux C, et al. Prediction of difcult mask
ventilation. Anesthesiology,2000 May; 92(5):12171218.
26. Nelson DB, Freeman ML, Silvis SE, Cass OW, Yashke PN, et al. A
randomized, controlled trial of transcutaneous carbon dioxide
monitoring during ERCP. Gastrointest Endosc 2000;
51:288295.
27. Vargo JJ, Zuccaro G, Dumot JA, et al. Gastroenterologist-
administered propofol for therapeutic upper endoscopy with
Chapter 5 Sedation and Analgesia for ERCP
graphic assessment of respiratory activity: a case series.
Gastrointest Endosc 2000; 52:250255.
28. Bower AL, Ripepi A, Dilger J, Boparai N, Brody FJ, Ponsky JL.
Bispectral index monitoring of sedation during endoscopy.
Gastrointest Endosc 2000; 52:192196.
29. Servin F, Desmonts JM, Haberer JP, et al. Pharmacokinetics and
protein binding of propofol in patients with cirrhosis.
Anesthesiology 1988; 69(6):887891.
30. Ickx B, Cockshott ID, Barvais L, et al. Propofol infusion for
induction and maintenance of anaesthesia in patients with end-
stage renal disease. Br J Anaesth 1998; 81(6):854860.
31. Trapani G, Altomare C, Sanna E, et al. Propofol in anesthesia.
Mechanism of action, structure-activity relationships, and drug
delivery. Curr Med Chem 2000; 7:249271.
32. Hales TG, Lambert JJ. The actions of propofol on inhibitory amino
acid receptors of bovine adrenomedullary chromafn cells and
rodent central neurons. Br J Pharmacol 1991; 104:619628.
33. Peduto VA, Concas A, Santoro G, et al. Biochemical and
electrophysiologic evidence that propofol enhances GABAergic
transmission in the rat brain. Anesthesiology 1991; 75:10001009.
34. Whitehead C, Sanders LD, Oldroyd G, et al. The subjective effects
of low dose propofol; a double blind study to evaluate
dimensions of sedation and consciousness with low-dose
propofol. Anaesthes 1994; 490496.
35. Smith I, Monk TG, White PF, et al. Propofol infusion during
regional anesthesia: sedative, amnestic, and anxiolytic properties.
Anesth Analg 1994; 79:313319.
36. Karski JM, Tearsdale SJ, Boylan J, et al. Propofol for continuous
intravenous sedation after aortocoronary bypass graft surgery.
Dose nding study [abstract]. Can J Anaesth 1994; 41(pt 2):A17.
37. Walker JA, McIntyre RD, Scleinitz PF, et al. Nurse-administered
propofol sedation without anesthesia specialists in 9152
endoscopic cases in an ambulatory surgery center. Am J
Gastroenterol 2003; 98:17441750.
38. Rex DK, Sipe BW, Kinser KM, et al. Safety of propofol
administered by registered nurses with gastroenterologist
supervision in 2000 endoscopic cases. Am J Gastroenterol 2002;
97:11591163.
39. Heuss LT, Schieper P, Drewe J, et al. Risk stratication and safe
administration of propofol by registered nurses supervised by
the gastroenterologist: a prospective observational study of more
than 2000 cases. Gastrointest Endosc 2003; 57:664671.
40. Clarke AC, Chiragakis L, Hillman LC, et al. Sedation for endoscopy:
the safe use of propofol by general practitioner sedationists. Med
J Aust 2002; 176:158161.
41. Rex DK, Overley CA, Walker J. Registered nurse-administered
propofol sedation for upper endoscopy and colonoscopy: why?
when? how? Reviews in Gastroenterological Disorders 2003;
3:7080.
42. Vargo JJ, Zuccaro G, Dumot J, et al. Gastroenterologist-
administered propofol versus meperidine and midazolam for
ERCP and EUS: a randomized controlled trial with cost
effectiveness analysis. Gastroenterology 2002; 123:816.
43. Sipe BW, Rex DK, Latinovich D, et al. Propofol versus midazolam/
meperidine for outpatient colonoscopy: administration by nurses
supervised by endoscopists. Gastrointest Endosc 2002;
55:815825.
44. Wehrmann T, Grotkamp J, Stergiou N, et al.
Electroencephalogram monitoring facilitates sedation with
propofol for routine ERCP: a randomized, controlled trial.
Gastrointest Endosc 2002; 56:817824.
45. Ulmer BJ, Hansen JJ, Overly CA, et al. Propofol versus midazolam/
fentanyl for outpatient colonoscopy: administration by nurses
supervised by endoscopists. Clin Gastroenterol Hepatol 2003;
1:425423.
49
 SECTION 1 GENERAL TOPICS
Chapter
Complications of ERCP: Prediction,
6 Prevention and Management
Martin L. Freeman
INTRODUCTION
ERCP has evolved from a diagnostic modality to a primarily thera-
peutic procedure for pancreatic as well as biliary disorders. ERCP
alone or with associated biliary and pancreatic instrumentation and
therapy can cause a variety of short-term complications including
pancreatitis, hemorrhage, perforation, cardiopulmonary events, and
others (Box 6.1). These complications can range from minorwith
one or two additional hospital days followed by full recovery, to
severe and devastating with permanent disability or death. Compli-
cations may cause the endoscopist signicant anxiety and exposure
to medical malpractice claims.
Major advances in complications of ERCP have occurred in
several areas: standardized consensus-based denitions of complica-
tions;1 large scale multicenter multivariate analyses that have allowed
clearer identication of patient and technique-related risk factors for
complications;27 and introduction of new devices and techniques to
minimize risks of ERCP.
DEFINITIONS OF COMPLICATIONS, ADVERSE
EVENTS, UNPLANNED EVENTS AND OTHER
NEGATIVE OUTCOMES
In 1991, standardized consensus denitions for complications of
sphincterotomy were introduced1 (Table 6.1). Severity is graded pri-
marily on number of hospital days and type of intervention required
to treat the complication. This classication allows uniform assess-
ment of outcomes of ERCP and sphincterotomy in various settings.
Beyond immediate complications, there is an increasing awareness
of the entire spectrum of negative (as well as positive) outcomes
including technical failures, ineffectiveness of the procedure in
resolving the presenting complaint, long-term sequelae, costs,
extended hospitalization, and patient (dis)satisfaction. Accordingly,
the terminology has evolved from complications to adverse
events to unplanned events. Adverse events must be viewed in
context of the entire clinical outcome: a successful procedure with a
minor or even a moderate complication may sometimes be a prefer-
able outcome to a failed procedure attempt without any obvious
complication: failure at ERCP usually leads to a repeated ERCP, or
to an alternative percutaneous or surgical procedure which may
result in signicant additional morbidity, hospitalization and cost.
ANALYSES OF COMPLICATION RATES
Reported complication rates vary widely, even between prospective
studies. In two large prospective studies, pancreatitis rates ranged
between 0.74% for diagnostic and 1.4% for therapeutic ERCP respec-
tively in one study5 compared with 5.1% (about 7 times higher) for
diagnostic ERCP and 6.9% (5 times higher) for therapeutic ERCP in
another prospective study.3 Reasons for such variation include: (1)
denitions used; (2) thoroughness of detection; (3) patient-related
factors; (4) procedural variables, such as use of pancreatic stents, or
extent of therapy. For all these reasons, it should not be assumed
that a lower complication rate at one center necessarily reects better
quality of practice.
Most recent studies have utilized multivariate analysis as a tool
to identify and quantify the effect of multiple potentially confound-
ing risk factors, but these are not infallible as many potentially key
risk factors were not examined in most studies, and some are over-
tted (too many predictor variables for too few outcomes). Only a
limited number of studies have included more than 1000 patients.
Tables 6.2, 6.3 and 6.4 show a summary of risk factors for complica-
tions of ERCP and sphincterotomy based on published multivariate
analyses.
OVERALL COMPLICATIONS OF ERCP
AND SPHINCTEROTOMY
Most prospective series report an overall short-term complication
rate for ERCP and/or sphincterotomy of about 5 to 10%.27 There is
a particularly high rate of complications for sphincter of Oddi dys-
function (up to 20% or more, primarily pancreatitis, with up to 4%
severe complications), and a very low complication rate for routine
bile duct stone extraction, especially in tandem with laparoscopic
cholecystectomy (under 5% in most).2 Sphincterotomy bleeding
occurs primarily in patients with bile duct stones, and cholangitis
mostly in patients with malignant biliary obstruction.
Summaries of multivariate analyses of risk factors for overall
complications of ERCP and sphincterotomy are shown in Table 6.2.
Although relevant studies are heterogeneous and sometimes omit
potentially key risk factors, several patterns emerge (Table 6.2): (1)
Indication of suspected sphincter of Oddi dysfunction was a signi-
cant risk factor whenever examined. (2) technical factors, likely
linked to the skill or experience of the endoscopist, were found to
be signicant risk factors for overall complications. These technical
factors include difcult cannulation, use of precut or access papil-
lotomy to gain bile duct entry, failure to achieve biliary drainage, and
use of simultaneous or subsequent percutaneous biliary drainage
for otherwise failed endoscopic cannulation. In turn, the ERCP case
volume of the endoscopists or medical centers, when examined, has
always been a signicant factor in complications by both univariate
or multivariate analysis.27 (3) Death from ERCP is rare (less than
0.5%), but most often related to cardiopulmonary complications,
51
 SECTION 1 GENERAL TOPICS

highlighting the need for the endoscopist to pay attention to issues
of safety during sedation and monitoring.
Notably, risk factors found not to be signicant are the following:
(1) older age or increased number of coexisting medical condi-
tionson the contrary, younger age generally increases the risk
both by univariate and multivariate analysis; (2) smaller bile duct
diameter, in contrast to previous observations; and (3) anatomic
obstacles such as periampullary diverticulum or Billroth II gas-
trectomy, although they do increase technical difculty for the
endoscopist.27

PANCREATITIS
BOX 6.1 COMPLICATIONS OF ERCP Pancreatitis is the most common complication of ERCP, with
reported rates varying from 1% to 40%, with a rate of about 5% being
Pancreatitis most typical. In the consensus classication, pancreatitis is dened
as clinical syndrome consistent with pancreatitis (i.e. new or wors-
Hemorrhage ened abdominal pain) with an amylase at least three times normal
at more than 24 hours after the procedure, and requiring more than
Perforation one night of hospitalization1 (Table 6.1). Some events are difcult to
classify in the consensus denitions, such as patients with post-
Cholangitis procedural abdominal pain and elevation of amylase to just under
three times normal, or those with dramatic amylase elevations but
Cholecystitis minimal symptoms that are not clearly suggestive of clinical pancre-
atitis. There are many potential mechanisms of injury to the pan-
Stent-related creas during ERCP and endoscopic sphincterotomy: mechanical,
chemical, hydrostatic, enzymatic, microbiologic, and thermal.
Cardiopulmonary Although the relative contribution of these mechanisms to post-
ERCP is not known, recent multivariate analyses have helped to
Miscellaneous identify the clinical patient and procedure-related factors that are
independently associated with pancreatitis.

Mild Moderate Severe

Pancreatitis Clinical pancreatitis, amylase at least three
times normal at more that 24 h after
the procedure, requiring admission or
prolongation of planned admission to
23 days
Bleeding Clinical (i.e. not just endoscopic) evidence
of bleeding, hemoglobin drop <3 g, no
transfusion
Perforation Possible, or only very slight leak of uid
or contrast, treatable by uids and
suction for 3 days
Infection (cholangitis) >38C for 2448 hr
Pancreatitis requiring hospitalization
of 410 days
Transfusion (4 units or less), no
angiographic intervention or
surgery
Any denite perforation treated
medically 410 days
Febrile or septic illness requiring
more than 3 days of hospital
treatment or percutaneous
intervention
Hospitalization for more than
10 days, pseudocyst, or
intervention (percutaneous
drainage or surgery)
Transfusion 5 units or more, or
intervention (angiographic
or surgical)
Medical treatment for more
than 10 days, or intervention
(percutaneous or surgical)
Septic shock or surgery

Table 6.1 Consensus denitions for the major complications of ERCP

Any intensive care unit admission after a procedure grades the complication as severe.
Other rarer complications can be graded by length of needed hospitalization.

Denitea Maybeb Noc

Suspected sphincter of Oddi dysfunction Young age Comorbid illness burden
Cirrhosis Pancreatic contrast injection Small CBD diameter
Difcult cannulation Failed biliary drainage Female sex
Precut sphincterotomy Trainee involvement Billroth II
Percutaneous biliary access Periampullary diverticulum
Lower ERCP case volume

Table 6.2 Risk factors for overall complications of ERCP in multivariate analyses
a
signicant by multivariate analysis in most studies.
b
signicant by univariate analysis only in most studies.
c
not signicant by multivariate analysis in any study.

52
 Chapter 6 Complications of ERCP: Prediction, Prevention and Management
Denitea
Suspected sphincter of Oddi dysfunction
Young age
Normal bilirubin
History of post-ERCP pancreatitis
Difcult or failed cannulation
Pancreatic duct injection
Pancreatic sphincterotomy (especially minor papilla)
Balloon dilation of intact biliary sphincter
Precut sphincterotomy
Table 6.3 Risk factors for post-ERCP pancreatitis in multivariate analyses
a
signicant by multivariate analysis in most studies.
b
signicant by univariate analysis only in most studies.
c
not signicant by multivariate analysis in any study.
Denitea Maybeb
Coagulopathy Cirrhosis
Anticoagulation <3 days after ES Dilated CBD
Cholangitis prior to ERCP CBD stone
Bleeding during ES Periampullary diverticulum
Lower ERCP case volume Precut sphincterotomy
Table 6.4 Risk factors for hemorrhage after endoscopic sphincterotomy in multivariate analyses
a
signicant by multivariate analysis in most studies.
b
signicant by univariate analysis only in most studies.
c
not signicant by multivariate analysis in any study.
Patient-related risk factors for
post-ERCP pancreatitis
The risk of post-ERCP pancreatitis is determined at least as much
by the characteristics of the patient as by endoscopic techniques or
maneuvers (Table 6.3). Patient-related predictors found to be signi-
cant in one or more major studies include younger age, indication
of suspected sphincter of Oddi dysfunction, history of previous post-
ERCP pancreatitis, and absence of elevated serum bilirubin.28
Women may have increased risk, but it is difcult to sort out the
contribution of sphincter of Oddi dysfunction, a condition that
occurs almost exclusively in women. In one meta-analysis, female
gender was clearly a risk,8 and women account for a majority of cases
of severe or fatal post-ERCP pancreatitis.9 Patients with multiple risk
factors have a dramatically enhanced risk.3
Sphincter of Oddi dysfunction, most often suspected in women
with post-cholecystectomy abdominal pain, poses a formidable risk
for pancreatitis after any kind of ERCP whether diagnostic, mano-
metric or therapeutic. Suspicion of sphincter of Oddi dysfunction
independently triples the risk of post-ERCP pancreatitis to about
1030%. The reason for heightened susceptibility in these patients
remains unknown. Contrary to widely held opinion that sphincter
of Oddi manometry is the culprit, recent multivariate analyses show
that empirical biliary sphincterotomy or even diagnostic ERCP has
similarly high risk.3 With the widespread use of aspiration instead
of conventional perfusion manometry catheters, the risk of manom-
etry has probably been reduced to that of cannulation with any other
ERCP accessory. Most previous studies linking manometry with risk
have been from tertiary centers in which manometry is always per-
formed in patients with suspected sphincter of Oddi dysfunction,
Maybeb Noc
Female sex Small CBD diameter
Acinarization Sphincter of Oddi manometry
Absence of CBD stone Biliary sphincterotomy
Lower ERCP case volume
Trainee involvement
Noc
ASA or NSAID
Ampullary tumor
Longer length sphincterotomy
Extension of prior ES
thus losing the ability to separate the contribution of risk from the
procedure from that of the patient. Two studies specically com-
pared risk of post-ERCP pancreatitis in patients having ERCP for
suspected sphincter of Oddi dysfunction with and without sphincter
of Oddi manometry and found no detectable independent effect of
manometry on risk.2,10 Absence of a stone in patients with suspected
choledocholithiasis has been found to be a potent single risk factor
for post-ERCP pancreatitis in patients suspected of having stones,
thus tting into the category of possible sphincter of Oddi dysfunc-
tion. These observations point out the danger of performing diag-
nostic ERCP to look for bile duct stones in women with recurrent
post-cholecystectomy pain, as there is generally a low probability of
nding stones in such patients, and a high risk of causing pancre-
atitis. It is an erroneous and potentially dangerous assumption that
merely avoiding sphincter of Oddi manometry will signicantly
reduce risk.
History of previous post-ERCP pancreatitis has been found to be
a potent risk factor (odds ratio = 2.0 to 5.4),3,7 and warrants special
caution. Advanced chronic pancreatitis, on the other hand, confers
some immunity against ERCP-pancreatitis, perhaps because of
atrophy and decreased enzymatic activity.3 Pancreas divisum is only
a risk factor if minor papilla cannulation is attempted.
Despite many early studies suggesting small bile duct diameter
to be a risk factor for pancreatitis, most recent studies have shown
no independent inuence of duct size on risk; small duct diameter
may have been a surrogate marker for sphincter of Oddi dysfunction
in the earlier studies utilizing only univariate analysis. ERCP for
removal of bile duct stones has been found to be relatively safe with
respect to pancreatitis rates (<4%) in multicenter studies regardless
of bile duct diameter.2 Neither the presence of periampullary
53
 SECTION 1 GENERAL TOPICS
diverticula nor Billroth II gastrectomy have been found to inuence
risk of pancreatitis.2
Technique-related risk factors for
post-ERCP pancreatitis
Technical factors have long been recognized to be important in
causing post-ERCP pancreatitis. Papillary trauma induced by dif-
cult cannulation has a negative effect that is independent of the
number of pancreatic duct contrast injections, which is also a risk
factor.2,3,6 Pancreatitis occurred in one study after 2.5% of ERCP in
which there was no pancreatic duct contrast injection at all.3 Aci-
narization of the pancreas, although undesirable, is probably less
important than generally thought and has not been found to be sig-
nicant in two recent studies.3,7
Overall, risk of pancreatitis is generally similar for diagnostic and
therapeutic ERCP.27 Performance of biliary sphincterotomy does
not appear to add signicant independent risk of pancreatitis to
ERCP,3,7 a nding that is contrary to widely held opinion. This is
probably not due to the safety of sphincterotomy, but rather to
the risk of diagnostic ERCP. Pancreatic sphincterotomy of any
kind,3 including minor papilla sphincterotomy7 was found to be a
signicant risk factor for pancreatitis, although the risk of severe
pancreatitis has been very low (less than 1%), perhaps because
nearly all of these patients had pancreatic drainage via a pancreatic
stent.
Precut or access papillotomy to gain access to the common bile
duct is controversial with respect to risk of pancreatitis and other
complications. Use among endoscopists varies from under 5% to as
many as 30% of cases.11 There are many variations on precut tech-
nique: standard needle-knife inserted at the papillary orice and
cutting upwards; needle-knife stulotomy starting the incision
above the papillary orice and then cutting either up or down; use
of a pull-type sphincterotome wedged in the papillary orice, or into
the pancreatic duct. Any of the above techniques has the potential
to lacerate and injure the pancreatic sphincter, and precut tech-
niques have been uniformly associated with a higher risk of pancre-
atitis in multicenter studies involving endoscopists with varied
experience, with precut sphincterotomy found signicant as a uni-
variate or multivariate risk factor for post-ERCP pancreatitis and/or
overall complications.2,5 In contrast, many series from tertiary refer-
ral centers have found complication rates no different than for stan-
dard sphincterotomy, suggesting that risk of precut sphincterotomy
is highly operator-dependent.11 In one study, endoscopists perform-
ing more than one sphincterotomy a week averaged 90% immediate
bile duct access after precutting, versus only 50% for lower volume
endoscopists, a success rate which hardly justies the risk of
complications.2
Comparative studies of precut with standard sphincterotomy are
hard to interpret because indications and settings may be very dif-
ferent, with precut preferentially performed in lower risk situations
such as obstructive jaundice, and prominent papillae. In addition,
increasing use of pancreatic stents in series from tertiary centers
may have neutralized the otherwise higher risk of precut sphincter-
otomy.7 Complications of precut sphincterotomy vary with the indi-
cation for the procedure, occurring in as many as 30% of patients
with sphincter of Oddi dysfunction in older studies without use of
pancreatic stents.2 Paradoxically, in patients with sphincter of Oddi
dysfunction, needle-knife sphincterotomy over a pancreatic stent
placed early in the procedure has been shown to be substantially
54
safer than conventional pull-type sphincterotomy without a pancre-
atic stent.12
There has been controversy as to whether increased risk of precut
sphincterotomy is due to the technique itself or due to prolonged
cannulation attempts that often precede its use. A randomized trial
at one tertiary center found no signicant difference in pancreatitis
or overall complication rates between prolonged cannulation and
early use of precut without a pancreatic duct stent.13
Risk of post-ERCP pancreatitis escalates in patients with multiple
risk factors.3 The interactive effect of multiple risk factors is reected
in the prole of patients developing severe post-ERCP pancreatitis.
In one study, females with a normal serum bilirubin had a 5% risk
of pancreatitis; with addition of difcult cannulation risk rose to
16%; with further addition of suspected sphincter of Oddi dysfunc-
tion (i.e. no stone found), the risk rose to 42%.3 In two different
studies, nearly all of the patients who developed severe pancreatitis
were young to middle-aged women with recurrent abdominal pain,
a normal serum bilirubin, and with no biliary obstructive pathol-
ogy.3,9 These observations emphasize the importance of tailoring the
approach of ERCP to the individual patient.
One recent study has clearly shown that trainee participation adds
independent risk of pancreatitis.7 In contrast, most multicenter
studies have failed to show a signicant correlation between endos-
copists ERCP case volumes and pancreatitis rates.2,3,5 It is possible
that none of the participating endoscopists in those studies reached
the threshold volume of ERCP above which pancreatitis rates would
diminish (perhaps greater than 250500 cases per year). However,
most American endoscopists average less than two ERCPs per
week,3 and the reported rates of pancreatitis from the highest volume
tertiary referral centers in the US are often relatively higher than
those in private practices. All of these observations suggest that case
mix is at least as important as expertise in determining risk of post-
ERCP pancreatitis.
Specic techniques to reduce risk of
post-ERCP pancreatitis
It stands to reason that the most expeditious method of cannulation
will likely be the safest. Use of a papillotome or steerable catheter
for biliary cannulation has been prospectively compared to a stan-
dard catheter in a number of randomized trials.11 Although all
showed signicantly higher success with the sphincterotome, there
was no difference in rates of pancreatitis or other complications.
Another randomized trial did show signicant reduction of pancre-
atitis risk when a guidewire was used in conjunction with a papillo-
tome, as opposed to a papillotome alone.14
Pancreatic stent placement can reduce risk of post-ERCP pancre-
atitis in a number of settings (Table 6.5), and is widely performed
at many advanced centers for this purpose (Fig. 6.1). Specic situa-
tions where placement of a pancreatic stent has been shown to
reduce risk include biliary sphincterotomy for sphincter of Oddi
dysfunction, pancreatic sphincterotomy, precut sphincterotomy,
balloon-dilation of the biliary sphincter, and endoscopic ampullec-
tomy, and probably after difcult cannulation.1520 A meta-analysis
suggests that use of pancreatic stents in high-risk patients reduced
rates of pancreatitis by about two thirds, with virtual elimination of
severe post-ERCP pancreatitis.18 While effective in high-risk cases,
placement of pancreatic stents is usually unnecessary regardless of
cannulation difculty in older, jaundiced patients especially if they
have a pancreatic duct obstructed by cancer.
 Chapter 6 Complications of ERCP: Prediction, Prevention and Management
Setting Benet Evidence
Biliary sphincterotomy for SOD yes RCT
Pancreatic sphincterotomy for SOD trend RCT (abstract)
Biliary balloon dilation for stone trend retrospective
case-control
Precut (access) biliary yes RCT (abstract)
sphincterotomy
High risk including difcult yes/trend RCT 2 in up to 80% of patients with normal ducts, using conventional 5
cannulation
Endoscopic ampullectomy yes/trend RCT, Retrospective
case-control
Table 6.5 Pancreatic stents to reduce risk of post-ERCP
pancreatitis
RCT = randomized controlled trial.
A B
C D
Fig. 6.1 Placement of pancreatic stent to reduce risk of post-ERCP
pancreatitis. A A guidewire passed to body of pancreatic duct
around genu. B 4 Fr single pigtail 9 cm long unanged pancreatic
stent placed. C Endoscopic view of guidewire in pancreatic duct
after biliary sphincterotomy. D 4 Fr single pigtail pancreatic stent
placed with drainage of pancreatic juice.
Pancreatic stenting has limitations as a strategy to reduce risk as
many endoscopists and their assistants are unfamiliar with their
placement and may have a substantial failure rate, leaving the patient
worse off than if no attempt was made.20 Small caliber wires (0.018
or 0.025) are often required, and techniques for deep insertion of
such guidewires may be unfamiliar to many endoscopists. Small
tortuous ducts and ansa pancreaticus (360 alpha loop) may post a
challenge even for the most experienced endoscopist. A technique
has been described which avoids deep wire passage and allows uni-
versal success at placing stents in difcult anatomy;20 a small caliber
nitinol tipped wire can be knuckled inside the main pancreatic duct
just beyond the sphincter and allow delivery of a small caliber short
stent without need for passing the wire around tight turns in a tortu-
ous duct.
Unfortunately, pancreatic stents may cause problems. They may
migrate inside the pancreatic duct, especially stents of straight con-
guration without a pigtail on the duodenal end, and those with dual
inner anges. This complication can largely be avoided by use of a
single pigtail on the duodenal end. The major concern about pan-
creatic stents is the potential to cause ductal or parenchymal injury
or even perforation; duct and parenchymal injury has been reported
French or greater polyethylene stents, and sometimes leads to severe
ductal stenosis and relapsing pancreatitis.21,22 Strategies to avoid this
complication included use of smaller caliber stents (3 or 4 French),
which have been shown to be associated with dramatically lower
rates of duct injury,23 and use of stents made of softer materials.
Pancreatic stents placed for prevention of post-ERCP pancreatitis in
normal ducts should be documented to pass by x-ray or removed
within a few weeks. If they have not passed by then, they need to be
removed.
Balloon-dilation of the biliary sphincter has been introduced as
an alternative to sphincterotomy for the extraction of bile duct stones.
Although trials from overseas have shown complications to be equiv-
alent to or less than for sphincterotomy, balloon dilation has been
associated with a markedly increased risk of pancreatitis in the US,
resulting in two deaths in one study,24 and with a higher risk of
pancreatitis by meta-analysis of pooled studies.25 In general, balloon
dilation of the intact biliary sphincter for extraction of bile duct
stones is not recommended unless there is a relative contraindica-
tion to sphincterotomy such as coagulopathy or need for early anti-
coagulation. In contrast, balloon dilation performed after biliary
sphincterotomy to facilitate large stone extraction may be relatively
safe and may reduce need for excessively large sphincterotomy and
its associated risk of perforation or bleeding.
Thermal injury is thought to play some role in causing pancre-
atitis after biliary and pancreatic sphincterotomy. A number of
randomized trials have compared the impact of pure cutting versus
blended current, with mixed results but generally lower rates of
pancreatitis using the pure cut current.15,26 Automated current deliv-
ery systems programmed to deliver a specic tissue effect are now
widely used. None of the available studies suggest a signicant dif-
ference in rates of pancreatitis between these units compared with
blended current, so that it is not yet clear whether automated current
delivery systems provide the same benet for prevention of pancre-
atitis as do those using pure cutting current.
Pharmacological agents
Many pharmacological agents have been investigated as potential
agents to reduce post-ERCP pancreatitis, but results have generally
been mixed or negative. In meta-analyses of randomized controlled
trials, gabexate (a protease inhibitor) or somatostatin has been
found to be marginally effective but only if given over an extended
infusion (up to 12 hours after ERCP), while shorter infusions (less
than 4 hours) are generally ineffective.15 Neither of these agents is
available in the United States, and the cost-effectiveness of pro-
longed infusions severely limits the cost-effectiveness and practical-
ity of these agents. More promising agents in single pilot studies
have included NSAID, secretin, and Ulistatin, a long-acting protease
inhibitor.15 However each of these agents has been found to be
effective in only a single trial, and it seems likely that these agents
may join the ranks of other promising agents whose efcacy did
not hold up under further multicenter trials. Agents shown not
to be effective include interleukin 10, octreotide, corticosteroids,
55
 SECTION 1 GENERAL TOPICS
allopurinol, platelet-activating factor inhibitors, heparin, and use of
non-ionic contrast.15 Thus it appears that at this time, meaningful
pharmacologic prophylaxis against post-ERCP pancreatitis is cur-
rently not feasible.
Prevention and treatment of
post-ERCP pancreatitis
The single most important way to avoid post-ERCP pancreatitis is to
avoid performing ERCP for marginal indications, especially in
patients at higher risk of complications. Paradoxically, the risk is
often higher and potential benet of therapy lower in marginally
indicated ERCP than for patients with obstructive jaundice. ERCP
should generally be avoided when the probability of nding stones
or other obstructive pathology is low and other methods are avail-
able, or situations in which the risk/benet ratio of conventional
diagnostic or biliary therapeutic ERCP is excessive (such as sus-
pected sphincter of Oddi dysfunction). Alternative imaging tech-
niques such as intraoperative laparoscopic cholangiography, MRCP
and endoscopic ultrasound are safer alternatives for excluding
obstructive biliary pathology. Patients who have negative evaluation
by these alternative techniques, but who are still suspected to have
a pancreatic or biliary cause for recurrent symptoms, are probably
best served by referral to a tertiary ERCP center capable of advanced
techniques for diagnosis including endoscopic ultrasound, advanced
therapeutics including pancreatic endotherapy, for near-certain
ability to place pancreatic stents.
Once the decision has been reached to proceed with ERCP, can-
nulation and sphincterotomy techniques should be tailored to the
risk prole of that individual. In low risk cases such as elderly
patients with obstructive jaundice, manipulation is generally well
tolerated, and whatever techniques are effective at gaining bile duct
access and drainage are reasonable. In high-risk cases, manipulation
should be minimized, and placement of a pancreatic stent consid-
ered. Placement of pancreatic stents is recommended in most
patients with suspected sphincter dysfunction, history of post-ERCP
pancreatitis, difcult cannulation, or prior to precut sphincterotomy
with unclear papillary anatomy or other risk factors. For pancreatic
stent insertion, size of stent should be tailored to the caliber and
course of the pancreatic duct, with small caliber (3 to 5 French)
pancreatic stents that are generally either short (23cm), or long
(710cm) and unanged. Use of precut sphincterotomy in high-risk
patients is probably best performed primarily by experts, probably
best done early rather than late in the procedure, and after placement
of a pancreatic stent in high-risk circumstances or unclear papillary
anatomy.
Treatment of post-ERCP pancreatitis is like that for any other
cause of acute pancreatitis. Early recognition of impending post-
ERCP pancreatitis can be facilitated by checking serum amylase or
other enzymes within a few hours after the procedure in patients
who are at high risk or who have abdominal pain. If serum amylase
or lipase is normal, probability of developing pancreatitis is very low
and the patient can be considered for same-day discharge if other-
wise reasonable. On the other hand, if the pancreatic enzymes are
signicantly elevated, premature same-day discharge may be avoided,
and pre-emptive hospitalization for observation, fasting and vigor-
ous intravenous hydration initiated. Severely ill patients should be
hospitalized in the intensive care unit with help obtained from other
specialists in managing the patient.
56
HEMORRHAGE
Bleeding seen endoscopically during sphincterotomy is often
reported as a complication, but of itself does not represent an adverse
outcome to the patient. Some degree of bleeding, ranging from
oozing to severe bleeding, is seen at the time of sphincterotomy in
about 1030% of cases. Clinically signicant hemorrhage is dened
in the consensus criteria (Table 6.1) as clinical evidence of bleeding
such as melena or hematemesis, with or without an associated fall
in hemoglobin, or requirement for secondary intervention such as
endoscopy or blood transfusion, and occurs in 0.52% of sphincter-
otomies.2 Clinical presentation is generally delayed from 1 to as
many as 10 days after sphincterotomy.2
Risk factors for hemorrhage after sphincterotomy
For clinically signicant hemorrhage (Table 6.4), risk factors include
any degree of bleeding during the procedure, presence of any coagu-
lopathy or thrombocytopenia (including hemodialysis-associated
coagulation disorders), initiation of anticoagulant therapy within
three days after ES, and relatively low case-volume on the part of the
endoscopist (performance of not more than one sphincterotomy per
week), which may reect less precise control of the incision or less
effective endoscopic control of bleeding once it has occurred.2 Factors
that do not appear to raise risk include use of aspirin or nonsteroidal
anti-inammatory drugs, making a longer incision, or enlarging a
previous sphincterotomy.2 The effect of newer antiplatelet agents
such as Plavix is unknown. Some studies have shown additional
risk factors for signicant hemorrhage to include use of precut
sphincterotomy.
Methods to prevent and treat hemorrhage
Bleeding after sphincterotomy can mostly be avoided by avoiding
sphincterotomy in patients with risk factors such as coagulopathy.
Once sphincterotomy is undertaken, risk can be minimized by cor-
rection of any coagulopathies, withholding anticoagulant medica-
tions for as many as three days afterwards, and by use of meticulous
endoscopic technique. Prophylactic injection of the sphincterotomy
site with epinephrine or even a sclerosing agent in patients with
coagulopathy may reduce risk of hemorrhage. Newer computerized
tissue-effect electrocautery units have been shown to reduce risk of
immediate bleeding but have not as yet been shown to decrease the
incidence of clinically signicant hemorrhage.
Once hemorrhage occurs, either immediately during sphincter-
otomy, or delayed, it can generally be controlled with endoscopic
therapy via injection of dilute epinephrine. Balloon-tamponade
using standard occlusion balloons may allow temporary control of
bleeding and improve visualization of the bleeding site. Thermal
therapy such as bipolar coagulation or clipping can follow (Fig. 6.2).
Caution should be taken to avoid thermal injury or clip placement
over the pancreatic sphincter, especially if the bleeding site is on the
right-hand wall of the sphincterotomy incision. Rarely, angiography
or surgery is required for refractory bleeding.
PERFORATION
Perforation may occur within the bowel wall by the endoscope,
extension of a sphincterotomy incision beyond the intramural
portion of the bile or pancreatic duct with retroperitoneal leakage,
 Chapter 6 Complications of ERCP: Prediction, Prevention and Management
A B
C D
Fig. 6.2 Endoscopic injection and clipping of sphincterotomy
bleed. A Bleeding from left edge of sphincterotomy. B Injection of
epinephrine. C Positioning of endoscopic clip. D Final placement of
two clips on left edge of sphincterotomy with hemostasis.
Fig. 6.3 Fluoroscopic image after biliary sphincterotomy with
large retroperitoneal perforation. A nasobiliary drain has been
placed, and large amounts of retroperitoneal air outlining the right
kidney are apparent, with contrast tracking into retroperitoneum
around the nasobiliary drain. As the leak was recognized immedi-
ately and was large and ongoing, this patient was managed with
urgent operative intervention with oversew of the perforation but
without duodenotomy, and was discharged home 5 days later.
or occur at any location due to extramural passage or migration of
guidewires or stents (Figs 6.3, 6.4A, 6.4B, and 6.5). Perforation is
now reported in less than 1% of ERCP and sphincterotomies.27 Risk
factors for sphincterotomy perforation have been difcult to quantify
due to the rarity of perforation. It is probable that bowel perforation
is more common in patients with Billroth II anatomy, and sphinc-
terotomy perforation after needle-knife precut techniques, and in
patients with suspected sphincter of Oddi dysfunction, all situations
where control and extent of the required incision is uncertain.27,28
A B
Fig. 6.4 A Distal migration of a biliary stent with perforation of
the opposite wall of the duodenum. The stent was placed ve days
prior for hilar tumor obstruction. The patient presenting with an
acute abdomen. B CT scan showing tip of stent and air in retro-
peritoneum anterior to the right kidney.
Fig. 6.5 CT scan obtained immediately after biliary sphincterot-
omy perforation shows air in subcutaneous tissues, free intraperi-
toneal and retroperitoneal air. This patient developed crepitus
during ERCP with sphincterotomy and lithotripsy for large stone. A
nasobiliary drain was placed. No contrast extravasation was dem-
onstrated either through nasobiliary drain or by CT, suggesting a
favorable course for nonoperative management. This patient was
treated with nasobiliary and nasogastric drainage, antibiotics, and
recovered fully without further intervention. (Courtesy Dr Oliver
Cass.)
Treatment of post-ERCP perforation varies with the type and
severity of the leak and clinical manifestations. Bowel wall per-
forations must generally be treated surgically, while guidewire or
stent-related perforations can usually be treated endoscopically by
providing adequate ductal drainage.28 Keys to avoiding perforation
during sphincterotomy are to limit the length of cutting wire in
contact with the tissue and to use stepwise incisions. If perforation
is suspected during a sphincterotomy, injection of a small amount
of contrast while pulling the papillotome through the incision over
a guidewire will conrm or exclude extravasation and allow proactive
treatment. Endoscopic clipping may be attempted in order to close
a denite leak.29 In most cases, a nasobiliary and/or nasopancreatic
drain should be placed (depending on the sphincter cut), and the
patient treated with nasogastric suction, intravenous antibiotics,
strict fasting, and in-hospital observation. The importance of early
recognition and endoscopic drainage of suspected perforations is
57
 SECTION 1 GENERAL TOPICS
supported by the observation that nearly all patients with immediate
recognition and endoscopic drainage did well with conservative
management, in comparison with poor outcomes including need for
surgery and some mortality in patients with delayed recognition.29
Once a perforation of any kind is suspected, a CT scan of the
abdomen should be obtained to assess for contrast leakage and any
retroperitoneal or intraperitoneal air (Fig. 6.5). If the leak is sizeable
and ongoing as suggested by contrast extravasation, or the patients
clinical condition deteriorates, prompt drainage via surgery or the
percutaneous route is advisable (Fig. 6.3).
CHOLANGITIS AND CHOLECYSTITIS
Cholangitis (ascending bile duct infection) and cholecystitis (gall-
bladder infection) are potential complications or sequelae of ERCP
and/or sphincterotomy. Risk factors for cholangitis after ERCP
and sphincterotomy consist primarily of failed or incomplete biliary
drainage27 and use of combined percutaneous-endoscopic proce-
dures.2 Other risk factors may include jaundice especially if due to
malignancy, and operator inexperience.2 Several studies have shown
that prophylactic antibotics can reduce the rate of bacteremia, but
few studies have shown a reduction in clinical sepsis following
ERCP, and a meta-analysis concluded that there was no clinical
benet to routine administration of antibiotics.30 Thus the principal
recommendation regarding prevention and treatment of cholangitis
is obtaining successful and complete biliary drainage.
LONG-TERM COMPLICATIONS/SEQUELAE
Recent studies have shown that if the gallbladder is left intact after
sphincterotomy, both early and late cholecystitis occur more fre-
quently than previously thought. Not surprisingly, both cholecystitis
and recurrent bile duct stones are more common if the gallbladder
left in situ contains stones. There is increasing concern about poten-
tial long-term sequelae of various components of endoscopic therapy,
including endoscopic biliary and pancreatic sphincterotomy. These
include recurrent stone formation, possibly resulting from sphinc-
terotomy stenosis, or bacterobilia due to duodenal-biliary reux, or
sine-materia cholangitis. Recurrent stones and other biliary prob-
lems may occur in from 6% to 24% of patients undergoing long-
term follow-up. Recurrent pancreatitis, presumably due to thermal
injury to the pancreatic sphincter, may occur after biliary sphincter-
otomy. The long-term effects of pancreatic sphincterotomy, which
is increasingly performed in patients with and without chronic pan-
creatitis, are largely unknown.
OPERATOR EXPERIENCE AND COMPLICATIONS
The effect of endoscopic expertise on outcome of ERCP is difcult
to evaluate but is likely profound. Simple comparisons of complica-
tion rates of ERCP between centers can be misleading, since the case
mix, intent of the procedure, and success rates at achieving biliary
and pancreatic duct access vary widely. A number of studies have
evaluated operator factors in complications of ERCP. Lower ERCP
case volume, dened variably, was signicantly associated with
higher overall complications by univariate and multivariate analysis
in all studies which have evaluated that risk factor. In one study,
endoscopists who performed more than one sphincterotomy per
58
week had somewhat lower rates of overall complications (8% vs
11%), but substantially lower rates of severe complications (0.9% vs
2.3%);2 in a multivariate model utilizing only information available
prior to ERCP, lower procedure volume was one of only three vari-
ables which predicted complications of sphincterotomy.2 Lower case
volume was signicantly associated with higher rates of hemorrhage
after sphincterotomy in two studies.2,5 In contrast, lower ERCP case
volume has not consistently been found to correlate with rates of
post-ERCP pancreatitis, suggesting the importance of case-mix in
determining this complication.
The available data probably underestimate the inuence of opera-
tor experience on outcomes of ERCP, since high-volume endosco-
pists attempt higher-risk cases, and also have higher success rates
at duct access. In one study endoscopists averaging more than 100
ERCP cases per year had 96.5% success at bile duct access compared
with 91.5% for lower volume endoscopists.3 In two other studies,
rates for failure and complications of ERCP by higher volume endos-
copists were signicantly lower than those of lesser-volume endos-
copists.2,5 Failure to complete ERCP may have as much negative
impact on patients as complications in terms of cost, need for further
interventions, and extension of hospital stay.
It is not known what minimum volume of cases is required in
order to maintain prociency, but probably in excess of 100 cases
per year to sustain good outcomes for routine biliary therapy, and
200250 cases per year for advanced pancreatic techniques. A minor-
ity of endoscopists in the US achieve such volumes of ERCP. The
data suggest that outcomes will be optimal if fewer endoscopists
perform more ERCP. It is not feasible or palatable to suggest that
all ERCP be performed at advanced centers. Rather, adequate
training and ongoing case volume should be a prerequisite for per-
forming ERCP in practice. Larger groups should concentrate all
their ERCP to a few dedicated individuals rather than dilute the
experience, and smaller groups who are unable to sustain adequate
volumes should consider contracting their ERCP work out to more
experienced individuals. Endoscopists who perform limited amounts
of complex ERCP should be amenable to prompt referral to a special-
ized center of potentially complex cases including difcult biliary
problems, all pancreatic therapeutics, and most cases of suspected
sphincter of Oddi dysfunction. The key is for each endoscopist to
nd the optimal balance between risk and benet for the individual
patient and their own individual expertise and experience
(Box 6.2).
BOX 6.2 STRATEGIES TO REDUCE
COMPLICATIONS OF ERCP
Improved training
Education of endoscopists regarding risk factors
Avoidance of marginally indicated ERCP
Referral to advanced centers for complex or high-risk cases
Fewer endoscopists performing more ERCP
 Chapter 6 Complications of ERCP: Prediction, Prevention and Management
REFERENCES
1. Cotton PB, Lehman G, Vennes JA, et al. Endoscopic
sphincterotomy complications and their management: an attempt
at consensus. Gastrointest Endosc 1991; 37:383391.
2. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med 1996;
335:909918.
3. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for post-
ERCP pancreatitis: a prospective, multicenter study. Gastrointest
Endosc 2001; 54:425434.
4. Masci E, Toti G, Mariani A, et al. Complications of diagnostic and
therapeutic ERCP: a prospective multicenter study. Am J
Gastroenterol. 2001; 96:417423.
5. Loperdo S, Angelini G, Benedetti G, et al. Major early
complications from diagnostic and therapeutic ERCP: a
prospective multicenter study. Gastrointest Endosc 1998; 48:110.
6. Vandervoort J, Soetikno RM, Tham TC, et al. Risk factors for
complications after performance of ERCP. Gastrointest Endosc.
2002; 56:652656.
7. Cheng C, Sherman S, Watkins JL, et al. Risk factors for post-ERCP
pancreatitis: a prospective multicenter study. American Journal of
Gastroenterology 2005 (in press).
8. Masci E, Mariani A, Curioni S, et al. Risk factors for pancreatitis
following endoscopic retrograde cholangiopancreatography: a
meta-analysis. Endoscopy 2003; 35:830834.
9. Trap R, Adamsen S, Hart-Hansen O, et al. Severe and fatal
complications after diagnostic and therapeutic ERCP: a
prospective series of claims to insurance covering public
hospitals. Endoscopy 1999; 31:125130.
10. Singh P, Gurudu SR, Davidoff S, et al. Sphincter of Oddi
manometry does not predispose to post-ERCP acute pancreatitis.
Gastrointest Endosc. 2004; 59:499505.
11. Freeman ML, Guda NL. Cannulation techniques for ERCP: a
review of reported techniques. Gastrointest Endosc 2005;
61:112125.
12. Fogel EL, Eversman D, Jamidar P, et al. Sphincter of Oddi
dysfunction: pancreaticobiliary sphincterotomy with pancreatic
stent placement has a lower rate of pancreatitis than biliary
sphincterotomy alone. Endoscopy. 2002; 34:280285.
13. Tang SJ, Haber GB, Kortan P, et al. Precut papillotomy versus
persistence in difcult biliary cannulation: a prospective
randomized trial. Endoscopy. 2005; 37:5865.
14. Lella F, Bagnolo F, Colombo E, et al. A simple way of avoiding
post-ERCP pancreatitis. Gastrointest Endosc. 2004; 59:830834.
15. Freeman ML, Guda NM. Prevention of post-ERCP pancreatitis: a
comprehensive review. Gastrointest Endosc 2004; 59:845684.
16. Tarnasky, P, Palesch, Y, Cunningham J, et al. Pancreatic stenting
prevents pancreatitis after biliary sphincterotomy in patients with
sphincter of Oddi dysfunction. Gastroenterology 1998;
115:15181524.
17. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent
prevent post-ERCP pancreatitis? A prospective randomized study.
Gastrointest Endosc 2003; 57:291294.
18. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic
stent placement reduce the risk of post-ERCP acute pancreatitis?
A meta-analysis of controlled trials. Gastrointest Endosc. 2004;
60:544550.
19. Harewood GC, Pochron NL, Gostout CJ. Prospective, randomized,
controlled trial of prophylactic pancreatic stent placement for
endoscopic snare excision of the duodenal ampulla. Gastrointest
Endosc. 2005; 62:367370.
20. Freeman ML, Overby CS, Qi DF. Pancreatic stent insertion:
consequences of failure, and results of a modied technique to
maximize success. Gastrointest Endosc. 2004; 59:814.
21. Smith MT, Sherman S, Ikenberry SO, et al. Alternations in
pancreatic ductal morphology following polyethylene pancreatic
stent therapy. Gastrointest Endosc. 1996; 44:268275.
22. Kozarek RA. Pancreatic stents can induce ductal changes
consistent with chronic pancreatitis. Gastrointest Endosc. 1990;
36:9395.
23. Rashdan A, Fogel EL, McHenry L Jr, et al. Improved stent
characteristics for prophylaxis of post-ERCP pancreatitis. Clin
Gastroenterol Hepatol. 2004; 2:322329.
24. Disario JA, Freeman ML, Bjorkman DJ, et al. Endoscopic balloon
dilation compared with sphincterotomy for extraction of bile duct
stones. Gastroenterology 2004; 127:12911299.
25. Baron TH, Harewood GC. Endoscopic balloon dilation of the
biliary sphincter compared to endoscopic biliary sphincterotomy
for removal of common bile duct stones during ERCP: a
metaanalysis of randomized, controlled trials. Am J Gastroenterol.
2004; 99:14551460.
26. Elta GH, Barnett JL, Wille RT, et al. Pure cut electrocautery current
for sphincterotomy causes less post-procedure pancreatitis than
blended current. Gastrointest Endosc 1998; 47:149153.
27. Enns R, Eloubeidi MA, Mergener K. ERCP-related perforations: risk
factors and management. Endoscopy. 2002; 34:293298.
28. Howard TJ, Tan T, Lehman GA, et al. Classication and
management of perforations complicating endoscopic
sphincterotomy. Surgery. 1999; 126(4):658663.
29. Baron TH, Gostout CJ, Herman L. Hemoclip repair of a
sphincterotomy-induced duodenal perforation. Gastrointest
Endosc. 2000; 52:566568.
30. Harris A, Chan AC, Torres-Viera C, et al. Meta-analysis of antibiotic
prophylaxis in endoscopic retrograde cholangiopancreatography
(ERCP). Endoscopy. 1999; 9:718724.
59
 SECTION 1 GENERAL TOPICS
Chapter
ERCP Training
7 Jrgen Hochberger, Detlev Menke, and Jrgen Maiss
INTRODUCTION
Diagnostic and interventional endoscopy is in a state of continuous
technological advancement and in the past few decades the latter has
replaced surgery for many gastrointestinal disorders. One example
is the development of endoscopic retrograde cholangiopancreato-
graphy (ERCP) for the nonsurgical treatment of common bile duct
stones.13 Endoscopic therapy for common bile duct stones can be
as simple as sphincterotomy and as complex as the use of lithotripsy
devices.46 Prociency in all aspects of ERCP requires several years
of practical training and continuous renement of knowledge.711 A
standardized mandatory teaching program for gastrointestinal (GI)
endoscopy has not been established so far. Historically, endoscopic
training has consisted primarily of learning by doing under the
supervision of an experienced endoscopist.12,13
With the advent of non-invasive tests such as magnetic resonance
cholangiopancreatography (MRCP) and endoscopic ultrasound
(EUS), ERCP has moved from a diagnostic to an almost purely
therapeutic procedure.14,15 This creates a new challenge to the educa-
tion of young endoscopists, since ERCP procedures are becoming
more concentrated in large or mid volume endoscopy centers while
the number of ERCPs performed in smaller hospitals is decreasing.
These smaller hospitals are often located in rural areas and provide
limited ERCP services such as sphincterotomy, stone extraction and
stent implantation. Threshold numbers of ERCP procedures that
must be performed by trainees for credentialing were published in
the Gastroenterology Core Curriculum in 1996.12 This document
indicated that fellows had to complete 100 ERCPs, including 25
therapeutic cases (20 sphincterotomies and 5 stent placement cases).
Jowell et al. found a minimum number of 180200 ERCPs needed
to be performed before a trainee can be considered competent for
non-supervised ERCP (Fig. 7.1).16
Approximately 80100 ERCPs per endoscopist per year seems
necessary to maintain sufcient competence for biliary procedures.
More than 250 ERCPs per endoscopist per year seems mandatory
for developing and maintaining expertise level in complex therapeu-
tic procedures in the pancreas.17 ERCP volume plays a role in com-
plication rates. In some studies a minimum of 4050 sphincterotomies
(EST) per endoscopist per year is associated with a lower complica-
tion rate than when performed by endoscopists who perform less.18,19
Rabenstein showed that both the number of ERCPs and ESTs per-
formed in the past as well as the number of ERCPs currently per-
formed by the endoscopist seem to inuence success and complication
rates.20 Finally, objective outcomes and medicolegal concerns play
an increasing role in daily GI practice.7,21,22 In light of all of these
issues, this chapter will cover the training options for the beginner
as well as for the practicing gastroenterologist to acquire or maintain
ERCP skills.
CLINICAL TRAINING IN ERCP
Prior to acquiring the skills necessary for the performance of ERCP
in a safe, effective, and comfortable manner, the endoscopist must
rst understand the indications, risks, and limitations of the pro-
cedure. In addition to this knowledge, training and prociency in
manual and technical skills are other aspects of a competent endos-
copist. To this end, the American Society for Gastrointestinal Endos-
copy (ASGE) published a new core curriculum for training in ERCP
in March 2006.23
In most fellowship training programs traditional ERCP training
follows education in diagnostic gastroscopy and colonoscopy and is
often begun when the trainee has been introduced to polypectomy,
hemostasis or EUS training as part of a learning pyramid.24
Fellows begin their ERCP training by observing the procedure
and/or assisting the primary endoscopist. In Europe this initial expe-
rience may involve maneuvering x-ray equipment during the proce-
dure or in some institutions the trainee may perform the duties of
the assisting endoscopy nurse in order to learn how to properly
handle catheters, guidewires and other accessories. Accompanying
learning aids include the review of video material, ERCP atlases, or
interactive computer programs. In most cases the rst practical steps
to learning ERCP involve understanding how to maneuver a side-
viewing endoscope by passing the endoscope during the early stages
of the procedure. This involves incrementally learning how to intu-
bate the esophagus, maneuver along the lesser curvature of the
stomach, appreciate the setting sun phenomenon at the passage
of the pylorus, maneuver around the superior duodenal angle, and
nally to bring the endoscope in an appropriate short position in
front of the papilla. It is often easier for the trainee to begin this
process by placing the patient in the left lateral decubitus position
with the patients left arm behind their back instead of starting pri-
marily in a prone position.
Now that most ERCPs are performed for therapeutic purposes, it
is controversial if cannulation is the most appropriate step for the
trainee to learn after he or she is able to competently maneuver the
duodenoscope to the papilla. For example, it is well known that
routine stent exchange in the setting of a prior sphincterotomy
requires a lower number of procedures (60) to obtain competence
than cannulation of a native papilla (180200), and is associated with
a lower risk prole.16,25 Patients with benign biliary strictures, chronic
obstructive pancreatitis, and recurrent bile duct stones in the setting
of prior sphincterotomy may be good cases for the trainee to perform
in the early stages of his/her ERCP experience.
Schutz and Abbot developed an ERCP grading scale based upon
procedural difculty. In a single-center study they used benchmarks
such as cannulation rates to gauge competency in attempted proce-
dures. A modication of this score was adopted by the ASGE as part
61
 SECTION 1 GENERAL TOPICS

1 Cholangiography 1 Pancreatography
.90 .90
.80 .80
Probability of achieving

Probability of achieving
.70 .70
acceptable score

acceptable score
.60 .60
.50 .50
.40 .40
.30 .30
.20 .20
n=155 167 135 120 158 184 133 112 73 6 n=120 106 108 93 119 151 103 96 59 6
.10 .10
0 0
20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200
ERCPs, n ERCPs, n

Deep common bile duct cannulation Deep pancreatic duct cannulation

1 1
.90 .90
.80 .80
Probability of achieving

Probability of achieving

.70 .70
acceptable score

acceptable score

.60 .60
.50 .50
.40 .40
.30 .30
.20 .20
.10 n= 89 119 89 80 119 113 92 81 55 6 .10 n= 20 22 22 23 26 36 34 17 21 2
0 0
20 40 60 80 100 120 140 160 180 200 20 40 60 80 100 120 140 160 180 200
ERCPs, n ERCPs, n

from Jowell, PS et al Ann Intern Med 1996; 125:983989

Fig. 7.1 Probability (95% Cls) of achieving an acceptable score for cholangiography, pancreatography, deep pancreatic cannulation and
deep biliary cannulation during fellows training in ERCP determined by Jowell et al. for 17 GI fellows during 1450 ERCP procedures.16

of their quality-assessment document.26,23 Absolute numbers of pro-
cedures partially performed by a fellow may not realistically reect
competency. Where possible, trainee logbook records should specify
particular skills completed by the fellow (cannulation, sphincterot-
omy, stent placement, tissue sampling), as well as indicate cases that
the trainee completed without assistance.
ASGE guidelines for advanced endoscopic training state that
most fellows require at least 180 cases to achieve competency, with
at least half of these cases being therapeutic. It must be emphasized
that performance of these numbers of procedures does not automati-
cally bestow competency, rather that competency can be assessed after
this number of procedures has been performed. Though nearly all
GI training programs offer some exposure to ERCP, not all of the
trainees may ultimately perform ERCP after the completion of their
training. All fellows should at least develop an understanding of the
diagnostic and therapeutic role of the procedure, including indica-
tions, contraindications, and possible complications. This exposure
is generally accomplished within the context of a 3-year gastroenter-
ology fellowship training program.23 The decision by a program
director as to whether to train one or more fellows each year to
achieve sufcient competency will depend in some measure on the
volume of ERCPs performed at the institution and the availability of
experts in ERCP to supervise the training of fellows. With data from
Jowell et al. to suggest that well over 200 cases are required for most
trainees to consistently cannulate the desired duct, programs with a
limited case volume will have to weigh their training objectives with
what is feasible. For example, with an annual volume of 400 cases
and three fellows, it would be reasonable to have one fellow perform
300 or more cases and provide the other two with an exposure to
ERCP, rather than have all three individuals equally share cases, with
a low likelihood that any of the three would reach competency by
the end of the fellowship.
Trainees who elect to pursue additional training in ERCP in order
to attain procedural competence should have completed at least 18
months of a standard gastroenterology training program as per
the Gastroenterology Core Curriculum. The minimum duration of

62

training required to achieve advanced technical and cognitive skills
is usually 12 months. This period of advanced training may be
incorporated into the standard three-year fellowship program or may
be completed during an additional year dedicated to advanced endo-
scopic procedures.23
TRAINING MODELS AND SIMULATORS
There have been a number of initiatives aiming to improve endo-
scopic training.23,2730 Courses using plastic dummies for gastroscopy
and colonoscopy are to be mentioned in this context. However, these
models have the disadvantage of allowing minimal interventions
to be performed and they could not be established for ERCP
training.13
Computer simulators
Apart from mechanical simulators, different computer simulation
systems have been developed.31,13,24 In 1990 Christopher Williams
introduced a computer simulator for colonoscopy and ERCP.29 Wil-
liams stated that in 1982 his group had already adapted a simple
electronic video game for training leftright hand coordination. The
next generation simulator was connected to an MS-DOS computer
with friction brakes to provide mechanical limitations. The provoca-
tion of audible signals, such as patient groans and patient protest,
when undue force or insufation was used, was already integrated
at that time.29 In the mid-1980s an interesting, realistic computer-
assisted simulator, the Robotics Interactive Endoscopy Simulation
(RIES) System was introduced for esophagogastroduodenoscopy
(EGD) and ERCP.3234 His aim was to integrate a functioning endo-
scope into an interactive environment, thus creating a realistic visual
appearance. The system reached a high level of technical sophistica-
tion with functioning sphincters of the papilla of Vater and tactile
feedback. In addition to catheter movements a virtual sphincterot-
omy could be performed.33 However, at that time computer simula-
tors required expensive platforms and therefore were not widely
accepted. Nowadays, because of the enormous evolution in electron-
ics and computers, the capabilities of any standard personal com-
puter system are far higher than that of the high-end computers of
10 years ago. This may be the main reason for the development of
different endoscopy computer simulation models seen in the late
1990s and early 2000s.The rst of these models had been the Sim-
bionix GI-MentorTM in the form of a dummy (Mr. Silverman).35
The current model GI-Mentor IITM (Simbionix Corporation, Cleve-
land, Ohio), as well as the AccuTouchTM computer simulator (Immer-
sion Medical Inc., Gaithersburg, MD) not only allows simulation of
different diagnostic and interventional procedures at different levels
of difculty but also includes didactic teaching modules with
anatomy and pathology atlases.13,31 Both systems create a relatively
realistic virtual endoscopic environment. ERCP modules with paral-
lel x-ray and endoscopic simulations, virtual sphincterotomy, stone
extraction, etc. have been implemented.
Aabakken et al. described their experience using the Simbionix
simulator in a one-day training program involving 33 participants.
Of the participants who completed a questionnaire, 85% rated the
simulator training as somewhat or very useful in their education;
61% of the trainees indicated that it would have a potential role in
the training and re-certication of physicians. Bar-Meir reported
similar results in which the GI-Mentor was used during a workshop.
This workshop involved 71 gastroenterologists with more than one
year of endoscopic experience.
Chapter 7 ERCP Training
Modern medical simulator technology has been shown to distin-
guish the upper endoscopy skill level between novices and experts.36,37
Similar results have been shown for sigmoidoscopy and colonos-
copy.3840 However, a prospective randomized trail including 9 simu-
lator-trained and 7 bedside-trained residents in internal medicine
failed to show a signicant benet for the computer simulator group
in sigmoidoscopy training.41 However, Sedlack and Kolars from
Mayo Clinic, Rochester, MN, could prove the value of prior computer
simulator training in the education of fellows in colonoscopy and
subsequently developed a special course program for fellowship
training.42,43 In a prospective study that applied computer-based colo-
noscopy simulation, 4 novice fellows received 6 hours of simulator-
based training, compared to 4 novice fellows without training.
Simulator-trained fellows outperformed traditionally trained fellows
during their initial 15 colonoscopies in all performance aspects
except for time of insertion (p < 0.05). Three parameters (depth of
insertion, independent completion, and ability to identify land-
marks) demonstrated a continued advantage up to 30 colonoscopies.
Beyond 30 procedures, there was no difference in the performance
of the two groups.43
Unfortunately, validation of the use of additional computer simu-
lator training compared to clinical training alone does not exist for
ERCP training. However, there is data concerning the use of three
different ERCP training models by participants and tutors who
attended an ASGE ERCP workshop, as discussed below.44
ERCP TRAINING IN LIVE ANIMALS
Since the beginning of the 1990s anesthetized pigs and dogs have
been used in systematic endoscopy training courses, especially for
ERCP techniques (Fig. 7.2).34,4547 The major advantages of using live
animals for training are the natural tissue sensation, elasticity, and
realistic tactile feedback which occurs with the use of organs similar
to those found in humans. Substantial restrictions to the use of
animals include: ethical considerations, animal welfare, concern for
cleanliness, need for additional endoscopes designated for animal
Fig. 7.2 ERCP hands-on training in the pig. Animal training courses
have been established since the beginning of the 1990s for ERCP
training in specialized training institutions. However, they require a
considerable logistic and nancial effort.
63
 SECTION 1 GENERAL TOPICS

use, and cost. Furthermore, the procedures must be performed in easy handling. CompactEASIE is a modied lightweight version
special animal facilities which may mandate separate permission for (weight 15 kg) developed in 1998 and focuses exclusively on inter-
animal experiments, and the procedures often require veterinary ventional endoscopic applications.13,58 This model uses a specially
and anesthesiology support. If one does not allow an adequate prepared porcine upper gastrointestinal organ package (esophagus,
amount of fasting time, the pig stomach remains lled with food, stomach and duodenum) and includes the common bile duct, gall-
impairing endoscopic visualization. The anatomy of the pigs upper bladder and liver (Fig. 7.2) for use as an ERCP training device.13,59,60
gastrointestinal tract is relatively similar to that of the human, An advantage of the lightweight model is the nearly radiotranslucent
though there are some differences that impact ERCP training. There ground plate which can be easily positioned under x-ray equipment
are separate papillae for the bile duct and pancreatic duct. The biliary (Figs 7.3 and 7.4).
papilla is located about 1.52 cm distal to the pylorus at the roof of For ERCP interventions such as sphincterotomy and stent place-
the duodenal bulb. The pancreatic papilla is located more distally ment the hepatobiliary system with liver, extrahepatic bile ducts and
and is often difcult to nd due to its small size and deep location
in the duodenum. A polyp-like structure at the pylorus, called the
torus pylorus, resembles a papilla with an impacted stone and can
be used for practicing needle-knife techniques. Problems encoun-
tered during examination of the upper gastrointestinal tract of pigs
include the often dilated stomach and the excessive distance to the
pylorus because of the long snout. During ERCP training courses,
perforation of the bile duct not infrequently occurs, necessitating
sacrice of the animal earlier than anticipated.
An evaluation of participants and tutors of three different training
models for ERCP during an ASGE ERCP hands-on workshop includ-
ing the animal model are outlined below.44

EX VIVO PORCINE TISSUE MODELS

(COMPACTEASIETM, ERLANGER
ENDOTRAINERTM ETC.)
The use of a pig stomach for diagnostic gastroscopy the training has
been widely available for many years.48 In 1997 Hochberger and
Neumann presented the rst generation of training models that
used specially prepared pig specimens for training interventional
exible GI endoscopy.4951 Training in endoscopic ulcer hemostasis Fig. 7.3 Training set up using the compactEASIE ex vivo animal
could be demonstrated by reliably reproducing spurting arterial part simulator for ERCP training under uoroscopic control.
bleeding. Since then the model has been used in training more than
30 interventional techniques.24,5052 Additionally a colon model for
stricture management, proctoscopic interventions, and EMR in the
lower GI tract has been developed.53,54
The Erlangen Active Training Simulator for Interventional
Endoscopy (EASIE) was developed by integrating an endoscopic
environment into the surgical Biosimulation Model of Neumann.55,56
The original model of Neumann was a relatively heavy model which
was not suitable for therapeutic exible endoscopic interventions.
It had been primarily designed for teaching laparoscopic and open
surgical procedures. The 30 kg simulator consists of a rotatable
plastic thorax-abdomen dummy. Upper gastrointestinal organ pack-
ages, obtained from ordinary slaughtering processes as used in the
meat industry, are thoroughly cleaned and placed into a special
simulator mold. Similar to the POP simulator of Szinicz et al.,57 a
roller pump can be used to drive an articial blood circulation with
citrated and diluted blood through the arteries of previously heparin-
ized organs of parenchymal resections. According to the idea of
Hochberger, this perfusion system was used for the rst time to
simulate arterial spurting bleeding in hollow gastrointestinal
organs.51 This was achieved by sewing segments of porcine splenic
arteries into the anterior wall of the stomach and connecting them
to an articial blood circuit.
Fig. 7.4 Organ package including upper GI tract with esophagus,
The CompactEASIE is a simplied version of the original stomach, duo-denum, biliary system and liver mounted on the
Biosimulation Model (EndoTrainer) and is specially designed for CompactEASIE ex vivo animal part simulator.

64

A B
C D
Fig. 7.5AD Cannulation and sphincterotomy in the ex vivo
porcine model. The biliary papilla is located at the transition of parts
I and II of major side of the porcine duodenum. It has a relatively
long intraduodenal precourse. Once in place, training for cannula-
tion, sphincterotomy and plastic and metal stent placement as well
as many other different procedures in the bile duct can be under-
taken (from left to right above and below). The separately drained
pancreatic papilla is usually to small, deep in the duodenum and
mostly not suited for training purposes.
the gallbladder are dissected and added to the upper GI tract.
However, in the pig, as mentioned above, the pancreatic duct and
biliary system are drained via separate duodenal papillae. The larger,
biliary papilla is usually located at the roof of the duodenal bulb and
is quite at. The pancreatic papilla, located deep in the descending
duodenum or even the fourth (horizontal) portion, is small, difcult
to cannulate, and it is generally not used for training. Additional
articial papillae can be implanted in the duodenum or stomach.
Endoscopists, especially beginners in ERCP, have to rst adapt to
the pig anatomy, as is the case in live pig training courses. It is pos-
sible to perform conventional endoscopic sphincterotomy (EST) as
well as needle-knife techniques (Fig. 7.5). Techniques that can be
performed include basic accessory exchange and advanced tech-
niques such as selective cannulation of the left and right hepatic
ducts and intrahepatic segments. Furthermore, the cystic duct can
be used to demonstrate guidewire steering and catheter manipula-
tion. Unilateral and bilateral hilar stent placement can be performed
using plastic and metal stents. Bilateral hilar metal stent placement
as well as retrieval of proximally migrated plastic biliary stents is part
of expert training programs. Bile duct stone can be simulated by
inserting 35 mm long pieces of 8.5 Fr plastic stents into the bile
duct. Extraction techniques can then be demonstrated using bal-
loons and baskets after performing EPT.
Recently Matthes and Cohen reported on an interesting creation
of a so-called neo-papilla.61 Analogous to the Grund Interphant
Chapter 7 ERCP Training
Fig. 7.6 Neo-papilla as developed by Matthes and Cohen in 2006
trying to overcome the natural differences of human and porcine
anatomy. They created an articial papilla made from the muscular
structure of a chicken heart with common drainage of an arti-
cial biliary and pancreatic duct made from porcine vessels. The
neo-papilla is located contrary to the pigs papilla at the minor
curvature and serves especially for cannulation and sphincterotomy
training.61
or Susi simulator, a chicken heart is used to simulate the muscles
of the sphincter apparatus.13,62 As opposed to the natural duodenal
papilla of the pig they were able to simulate the human anatomical
situation by integrating a bile duct and a pancreatic duct (Fig. 7.6).
Furthermore, the papilla could be situated more distally into the
duodenum as in the human.61
All organs used for these simulations are subject to veterinary
inspection and comply with the pertinent food hygiene regulations.
The organ packages must be specically prepared and adapted to the
topics and objectives of the course in which they are utilized. The
organs from the recently slaughtered animals can be stored for
several months in sealed plastic bags at a temperature of about
18C. The organs are thawed the night prior to the training
session.
Since 2002, simple versions of the original CompactEASIE and
EndoTrainer simulators have become available. A company, Ham-
merhead Design (Mt. Pleasant, SC), developed a simple two-part
plastic mold similar to the CompactEASIE simulator. However
instead of screw pins, this simulator uses a exible net suspended
over the specimen to keep the stomach in position on the mold. The
ASGE also developed a simulator mold similar to the Compact-
EASIE model called EndoTrainer X which also uses a plastic net
that is xed over the specimen.
Neumann et al. published a pilot study using the Erlanger Endo-
Trainer. An experienced endoscopist and nurse used the simulator
to demonstrate positioning of the duodenoscope, cannulation,
guidewire insertion, sphincterotomy, stone extraction and plastic
stent placement.63 In 2001, Maiss et al. presented their experience
with interventional ERCP training using the CompactEASIE biologi-
65
 SECTION 1 GENERAL TOPICS

cal simulator.64 They analyzed nine structured training courses from Since then, different levels of training for beginners, advanced
March 1999 to July 2001 using the CompactEASIE ex vivo simula- endoscopists and experts have been developed using the Compact-
tor. The courses were designed for team-training groups of three EASIE simulator in hands on workshops in Hildesheim and Erlan-
doctors and three nurses per simulator. All courses were performed gen, Germany (Table 7.3).
at the endoscopy unit of the Department of Medicine I, University
Hospital in Erlangen, Germany. In total, 188 participants were
trained. The workshop structure included three 30-minute theoreti- ARTIFICIAL TISSUE MODELS
cal lectures and video demonstrations concerning an introduction to
ERCP, sphincterotomy techniques, treatment of stones and biliary As previously mentioned Grund and colleagues at the University of
strictures as well as pathological ndings at ERCP (pitfalls, tips and Tbingen, Germany developed an innovative static model using
tricks). Three blocks of 1.5 h, 1.5 h and 1 h of hands-on training were animal parts.13,62 These simulators have also been utilized for spe-
integrated in the course. After demonstration by the tutor, the train- cic ERCP techniques. The models employ a plastic bile duct and a
ees were supervised in performing papillary cannulation, sphincter- papilla made from a chicken heart to allow training of sphincterot-
otomy techniques, guidewire exchange techniques, stone extraction omy techniques. The model is not commercially available and there
and stenting. Only advanced trainees received simulator experience are no published data validating its use in ERCP training.
in needle-knife techniques and metal stent placement. Each group
was instructed by an experienced endoscopist and GI assistant. At
the end of the course a standardized questionnaire served for evalu- COMPARISON OF DIFFERENT ERCP
ation of the course. 132 trainees (78 doctors, 53 nurses, 1 unspeci- TRAINING MODELS
ed) of the 188 trainees (70%) completed the questionnaire. Nearly
all trainees (97%) rated the training as excellent or good. The evalu- Sedlack et al. compared three different ERCP training models used
ations of the participants concerning the single techniques trained at an ASGE advanced ERCP training course.44 A live, anesthetized
are listed in Table 7.1. Evaluations of the realism of the anatomical pig model was compared to the CompactEASIE ex-vivo simulator
environment, the optical impression and the tactile feedback are and to the Simbionix GI Mentor ERCP computer module. Ten course
listed in Table 7.2. participants and 10 experienced faculty practiced biliary cannulation
and other interventional maneuvers for 2030 minutes per model
and then evaluated model parameters. A 7-point Likert scale (1 = very
unrealistic, 7 = very realistic was used to assess (1) tissue pliability,
(2) papillary anatomy, (3) visual realism, (4) cannulation realism, and
Technique Excellent Good Insufcient No statement (5) overall ERCP experience. A similar scale was used to assess the
Cannulation 66% 27% 2% 5% utility of each model for teaching basic or advanced ERCP techniques
Sphincterotomy 60% 30% 2% 8% (1 = not useful at all, 7 = very useful). The CompactEASIE harvested
Stent implantation 57% 26% 1% 16% porcine model scored the highest for realism and usefulness in
(plastic) teaching basic and advanced ERCP skills. The scores for the com-
Stent implantation 16%a 8%a 4%a 72%a puter model were signicantly lower (p < 0.05) than those of the live
(metal) and CompactEASIE porcine models in nearly all areas except for
Needle-knife 36%a 20%a 3%a 41%a
papillary anatomy. Course faculty favored the CompactEASIE har-
vested pig model, whereas the course participants favored the live pig
Table 7.1 Realism of ERCP techniques trained in the harvested pig
model. Analyzing the results of the combined group (tutors plus
specimen CompactEASIETM simulator. Results of the evaluation of
nine hands-on training workshops on interventional ERCP trainees) the harvested pig organ model was rated most promising.
obtained from 132 participants (of 188 total) between March 1999 The ex vivo model seems best suited to translate didactic contents
and July 2001 and experience from the experienced tutor to the learner, while offer-
a
hands-on training for experts only, therefore 72% and 41% of questionnaires with ing a relatively realistic endoscopic environment. This type of train-
no statement. ing model offers a compromise between live animal courses located
in special training facilities and computer simulator training. Pro-
spective trails comparing the use of simulators and/or live animals
to clinical ERCP training are needed before they can be recom-
mended as standard training curricula.
Criteria Excellent Good Insufcient No statement
Anatomical 37% 53% 3% 7%
environment DATA SUPPORTING THE ROLE OF ERCP
Visual feel 48% 45% 2% 5%
Tactile feedback 35% 49% 3% 13%
TRAINING ON SIMULATORS
Since 1997, regular training workshops on interventional endoscopic
Table 7.2 Realism of the anatomical environment, visual
impression and tactile feedback using the harvested pig specimen techniques using the CompactEASIE-simulator have been estab-
CompactEASIETM simulator in an ERCP setting (see Figure 7.2). lished at the Department of Medicine at the University of Erlangen-
Results of the evaluation of nine hands-on training workshops on Nuremberg in Germany and recently in Hildesheim, Germany, as
interventional ERCP training obtained from 132 participants (of well as in numerous international teaching centers and national
188 total) in workshops between March 1999 and July 2001 society endoscopy courses throughout the world. The EASIE group

66

ERCPBeginners
Cannulation, standard
Guide wire exchange techniques (standard or rapid exchange,
etc.)
Spincterotomy (guide wire)
Stone extraction
basket
basket along guide wire
balloon
Plastic stents
ERCPAdvanced
Stricture management
dilatation
bougienage
Selective intrahepatic cannulation left / right
Cannulation of the difcult papilla with
sphincterotome
steerable catheters (swing tip etc.)
guidewire
Needle-knife sphincterotomy
Complication management
post-sphincterotomy bleeding, injection
post-sphincterotomy bleeding, clips
Metal stents
distal common bile duct
hilar, unilateral
Difcult bile duct stones
mechanical lithotripsy, emergency device (e.g. Soehendra)
mechanical lithotripsy, through the channel device (e.g.
Olympus)
avoidance and management of complications
ERCP in Billroth II anatomy with side viewing scope
endoscopic techniques for negotiating entero-enteric
anastomoses
Table 7.3 ERCP techniques trained according to level of education in current workshops in Hildesheim and Erlangen, Germany using
the CompactEASIETM harvested specimen simulator
has propagated the EASIE Team Training Concept for the simul-
taneous training of doctors and nurses in different interventional
endoscopic techniques.13,59,60,65
Simulator training in interventional endoscopy provides an effec-
tive opportunity for endoscopy trainees to gain considerable experi-
ence in ERCP techniques without time limitation and patient risk.
In a prospective, randomized trial using the EASIE simulator for
hemostasis training, trainees in New York City achieved signicant
improvement in their performance of multiple skills on the simula-
tor after only three workshops.66 It appears that a structured educa-
tional program with access to simulator training in addition to
supervised patient care would also enhance ERCP education. Such
efforts should benet patients by improving the skill level of trainees
before they perform actual procedures and could result in a lower
complication rate when trainees are involved with actual cases,
leading to better patient outcomes. The results of the actual hemo-
stasis cases performed in the New York study highlight this poten-
tial.66 After the pilot project in New York, a second program which
involved the training of 35 gastroenterology fellows from 25 French
universities conrmed these results.67
Chapter 7 ERCP Training
cannulation
sphincterotomy
stenting
stone extraction
ERCP in Billroth II anatomy with pediatric colonoscope
or gastroscope without elevator
endoscopic techniques for negotiating entero-enteric
anastomoses
cannulation
sphincterotomy
stenting
stone extraction
ERCPExperts
Cholangioscopy
Laser lithotripsy
smart laser lithotripsy via balloon
smart laser lithotripsy via steerable catheter etc.
ERCP with Double Balloon Enteroscope in Billroth II anatomy or
after entero-enteric anastomoses
Roux-en-Y intubation, gaining access to the papilla
india ink and Lipiodol marking of afferent loop
cannulation with a 200 cm instrument
sphincterotomy, instruments, strategy
stenting
stone extraction
laser lithotripsy
complication management (injection hemostasis, clipping)
Metal stents
bilateral hilar stenting, Y side-to side
bilateral hilar stenting, Y through mesh
combined biliary and enteral metal stenting
ACQUIRING TEACHING SKILLS AS TUTOR FOR
HANDS-ON WORKSHOPS
The benet of the trainee from hands-on workshops is likely a com-
bination of the amount of unsupervised time using the model,
expert instruction, high faculty-to-student ratios, and formal skill
evaluations with opportunity for feedback.52 To make simulator
training accessible to more physicians, an expanded number of
experts need to receive training on how to use these models to teach
others. Potential instructors need to know how to set up the equip-
ment, how to conduct the workshops, and how to evaluate the train-
ees using the model. The training of tutors who are available to a
wide geographic distribution should enable greater local simulator
availability. Only by doing so can simulator-based workshops be
integrated into standard endoscopy education. An additional benet
of focusing efforts to develop the educational skills of endoscopy
instructors is to promote uniformity in endoscopy education.68,69
For hemostasis training we conducted a pilot study that examined
the feasibility of short train-the-trainer sessions to achieve these
goals. Seven senior endoscopists without prior EASIE simulator
67
 SECTION 1 GENERAL TOPICS
experience were enrolled in this study to serve as the New Tutors
Group. Five expert endoscopists with EASIE team training experi-
ence instructed the New Tutors Group in a one-day train-the-trainer
session. The next day, eight additional gastroenterology fellows
attended a one-day hemostasis workshop conducted by the New
Tutor Group and underwent pre-/post-training evaluations on the
simulator by their instructors.
In nearly all parameters assessed, fellows under the direction of
newly trained trainers made signicant progress in just one day.
Tutors trained in this manner were able to provide an educational
experience similar to that provided by experts who have conducted
many hands-on workshops. The same has not yet been proven for
ERCP training.
MAINTAINING SKILLS IN ERCP
There is little doubt that the knowledge gained from hands-on
courses decreases over time. Little is known about the volume of
ERCP cases needed to maintain skills acquired during these sessions
in order to continue to achieve good outcomes. While simulator train-
ing has the potential to facilitate the maintenance of ERCP skills as
well as to teach individuals in practice how to use new devices, there
is no data that conrms this benet. In addition, there is no data of
how often such refresher courses would be needed. Train-the-trainer
sessions for ex vivo ERCP simulator training as successfully put
REFERENCES
1. Hochberger J, Tex S, Maiss J, et al. Management of difcult
common bile duct stones. Gastrointest Endosc Clin N Am 2003;
13:623634.
2. Carr-Locke DL. Cholelithiasis plus choledocholithiasis: ERCP rst,
what next? Gastroenterology 2006; 130:270272.
3. NIH state-of-the-science statement on endoscopic retrograde
cholangiopancreatography (ERCP) for diagnosis and therapy. NIH
Consens State Sci Statements 2002; 19:126.
4. Hochberger J, Bayer J, May A, et al. Laser lithotripsy of difcult
bile duct stones: results in 60 patients using a rhodamine 6G dye
laser with optical stone tissue detection system. Gut 1998;
43:823829.
5. Hochberger J, Bayer J, Maiss J, et al. Clinical results with a new
frequency-doubled, double pulse Nd:YAG laser (FREDDY) for
lithotripsy in complicated choledocholithiasis. Biomed Tech (Berl)
1998; 43 Suppl:172.
6. Kozarek R. Role of ERCP in acute pancreatitis. Gastrointest Endosc
2002; 56:S231236.
7. Baron TH, Petersen BT, Mergener K, et al. Quality indicators for
endoscopic retrograde cholangiopancreatography. Am J
Gastroenterol 2006; 101:892897.
8. Renewal of endoscopic privileges: guidelines for clinical
application. From the ASGE. American Society for Gastrointestinal
Endoscopy. Gastrointest Endosc 1999; 49:823825.
9. Principles of training in gastrointestinal endoscopy. From the
ASGE. American Society for Gastrointestinal Endoscopy.
Gastrointest Endosc 1999; 49:845853.
10. Guidelines for credentialing and granting privileges for
gastrointestinal endoscopy. American Society for Gastrointestinal
Endoscopy. Gastrointest Endosc 1998; 48:679682.
11. Position statement. Maintaining competency in endoscopic skills.
American Society for Gastrointestinal Endoscopy. Gastrointest
Endosc 1995; 42:620621.
68
into practice for EASIE hands-on hemostasis courses would are
needed.52,70
OPEN QUESTIONS AND PERSPECTIVES FOR
ERCP TRAINING IN THE FUTURE
In 2003 Kowalski et al. performed a survey concerning ERCP train-
ing among US gastroenterology fellows. In a short questionnaire,
they assessed training program, personal ERCP experience, percep-
tions regarding training adequacy, and post-training practice plans.
Graduating fellows performed a median of 140 ERCPs and 35
sphincterotomies during training, with an associated median
comfort level for independently performing sphincterotomy of 7.5
on a scale of 1 to 10. The median estimated success rate for inde-
pendent free cannulation was 75%. Based on non-parametric corre-
lation and regression analysis, 180 ERCPs would be necessary to
achieve a free cannulation rate of 80% and 69 sphincterotomies to
achieve a comfort level of 8 on a scale of 1 to 10. Only 36% of fellows
achieved the number of procedures and cannulation success. Sixty-
four percent of fellows did not achieve procedural competence and
33% reported inadequate ERCP training. Nevertheless, 91% of
fellows said they expected to perform unsupervised ERCP after train-
ing. The authors concluded that the majority of graduating fellows
did not achieve an acceptable success rate during training, yet still
intended to perform ERCP after training.
12. Training the gastroenterologist of the future: the
gastroenterology core curriculum. The Gastroenterology
Leadership Council. Gastroenterology 1996; 110:12661300.
13. Hochberger J, Maiss J, Magdeburg B, et al. Training simulators and
education in gastrointestinal endoscopy: current status and
perspectives in 2001. Endoscopy 2001; 33:541549.
14. Carr-Locke DL, Conn MI, Faigel DO, et al. Technology status
evaluation: magnetic resonance cholangiopancreatography:
November 1998. From the ASGE. American Society for
Gastrointestinal Endoscopy. Gastrointest Endosc 1999;
49:858861.
15. Mergener K, Kozarek RA. Therapeutic pancreatic endoscopy.
Endoscopy 2005; 37:201207.
16. Jowell PS, Baillie J, Branch MS, et al. Quantitative assessment of
procedural competence. A prospective study of training in
endoscopic retrograde cholangiopancreatography. Ann Intern
Med 1996; 125:983989.
17. Freeman ML. Adverse outcomes of endoscopic retrograde
cholangiopancreatography: avoidance and management.
Gastrointest Endosc Clin N Am 2003; 13:775898, xi.
18. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med 1996;
335:909918.
19. Huibregtse K. Complications of endoscopic sphincterotomy and
their prevention. N Engl J Med 1996; 335:961963.
20. Rabenstein T, Hahn EG. Post-ERCP pancreatitis: is the endoscopists
experience the major risk factor? Jop 2002; 3:177187.
21. Cotton PB. Analysis of 59 ERCP lawsuits; mainly about indications.
Gastrointest Endosc 2006; 63:378382; quiz 464.
22. Cotton PB. Evaluating ERCP is important but difcult. Gut 2002;
51:287289.
23. Chutkan RK, Ahmad AS, Cohen J, et al. ERCP core curriculum.
Gastrointest Endosc 2006; 63:361376.

24. Hochberger J, Maiss J, Hahn EG. The use of simulators for training
in GI endoscopy. Endoscopy 2002; 34:727729.
25. Jowell PS. Endoscopic retrograde cholangiopancreatography:
toward a better understanding of competence. Endoscopy 1999;
31:755757.
26. Schutz SM, Abbott RM. Grading ERCPs by degree of difculty: a
new concept to produce more meaningful outcome data.
Gastrointest Endosc 2000; 51:535539.
27. Axon AT, Aabakken L, Malfertheiner P, et al. Recommendations of
the ESGE workshop on Ethics in Teaching and Learning
Endoscopy. First European Symposium on Ethics in
Gastroenterology and Digestive Endoscopy, Kos, Greece, June
2003. Endoscopy 2003; 35:761764.
28. Waye JD. Teaching basic endoscopy. Gastrointest Endosc 2000;
51:375377.
29. Williams CB, Baillie J, Gillies DF, et al. Teaching gastrointestinal
endoscopy by computer simulation: a prototype for colonoscopy
and ERCP. Gastrointest Endosc 1990; 36:4954.
30. Frakes JT. An evaluation of performance after informal training in
endoscopic retrograde sphincterotomy. Am J Gastroenterol 1986;
81:512515.
31. Gerson LB, Van Dam J. Technology review: the use of simulators
for training in GI endoscopy. Gastrointest Endosc 2004;
60:9921001.
32. Noar MD. Endoscopy simulation: a brave new world? Endoscopy
1991; 23:147149.
33. Noar MD. Robotics interactive endoscopy simulation of ERCP/
sphincterotomy and EGD. Endoscopy 1992; 24 Suppl 2:539541.
34. Noar MD, Soehendra N. Endoscopy simulation training devices.
Endoscopy 1992; 24:159166.
35. Bar-Meir S. A new endoscopic simulator. Endoscopy 2000;
32:898900.
36. Moorthy K, Munz Y, Jiwanji M, et al. Validity and reliability of a
virtual reality upper gastrointestinal simulator and cross validation
using structured assessment of individual performance with video
playback. Surg Endosc 2004; 18:328333.
37. Ferlitsch A, Glauninger P, Gupper A, et al. Evaluation of a virtual
endoscopy simulator for training in gastrointestinal endoscopy.
Endoscopy 2002; 34:698702.
38. Datta V, Mandalia M, Mackay S, et al. The PreOp exible
sigmoidoscopy trainer. Validation and early evaluation of a virtual
reality based system. Surg Endosc 2002; 16:14591463.
39. MacDonald J, Ketchum J, Williams RG, et al. A lay person versus a
trained endoscopist: can the preop endoscopy simulator detect a
difference? Surg Endosc 2003; 17:896898.
40. Sedlack RE, Kolars JC. Validation of a computer-based
colonoscopy simulator. Gastrointest Endosc 2003; 57:214218.
41. Gerson LB, Van Dam J. A prospective randomized trial comparing
a virtual reality simulator to bedside teaching for training in
sigmoidoscopy. Endoscopy 2003; 35:569575.
42. Sedlack RE, Kolars JC. Colonoscopy curriculum development and
performance-based assessment criteria on a computer-based
endoscopy simulator. Acad Med 2002; 77:750751.
43. Sedlack RE, Kolars JC. Computer simulator training enhances the
competency of gastroenterology fellows at colonoscopy: results
of a pilot study. Am J Gastroenterol 2004; 99:3337.
44. Sedlack R, Petersen B, Binmoeller K, et al. A direct comparison
of ERCP teaching models. Gastrointest Endosc 2003; 57:
886890.
45. Gholson CF, Provenza JM, Silver RC, et al. Endoscopic retrograde
cholangiography in the swine: a new model for endoscopic
training and hepatobiliary research. Gastrointest Endosc 1990;
36:600603.
46. Gholson CF, Provenza JM, Doyle JT, et al. Endoscopic retrograde
sphincterotomy in swine. Dig Dis Sci 1991; 36:14061409.
Chapter 7 ERCP Training
47. Noar MD. An established porcine model for animate
training in diagnostic and therapeutic ERCP. Endoscopy 1995;
27:7780.
48. Freys SM, Heimbucher J, Fuchs KH. Teaching upper
gastrointestinal endoscopy: the pig stomach. Endoscopy 1995;
27:7376.
49. Hochberger J, Neumann M, Hohenberger W, et al. Neuer
Endoskopie-Trainer fr die therapeutische exible Endoskopie. Z
Gastroenterol 1997; 35:722723 (AB).
50. Hochberger J, Neumann M, Hohenberger W, Hahn EG. [EASIE-
Erlangen Education Simulation Model for Interventional
Endoscopya new bio-training model for surgical endoscopy].
Biomed Tech (Berl) 1997; 42 Suppl:334.
51. Hochberger J, Neumann M, Maiss J, et al. EASIEErlangen Active
Simulator for Interventional Endoscopya new bio-simulation-
modelrst experiences gained in training workshops.
Gastrointest Endosc 1998; 47 (Suppl.):AB116.
52. Matthes K. Simulator training in endoscopic hemostasis.
Gastrointest Endosc Clin N Am 2006; 16:511527.
53. Maiss J, Matthes K, Naegel A, et al. Der coloEASIE-SimulatorEin
neues Trainingsmodell fr die interventionelle Kolo- und
Rektoskopie (The coloEASIE-Simulatora new training model for
interventional colonoscopy and rectoscopy). Endo heute 2005;
18:190193.
54. Hochberger J, Maiss J. Currently available simulators: ex vivo
models. Gastrointest Endosc Clin N Am 2006; 16:435449.
55. Neumann M, Hochberger J, Felzmann T, et al. Part 1. The Erlanger
endo-trainer. Endoscopy 2001; 33:887890.
56. Neumann M, Stangl T, Auenhammer G, et al. Laparoscopic
cholecystectomy. Training on a bio-simulation model with
learning success documented using score-cards. Chirurg 2003;
74:208213.
57. Szinicz G, Beller S, Bodner W, et al. Simulated operations by
pulsatile organ-perfusion in minimally invasive surgery. Surg
Laparosc Endosc 1993; 3:315317.
58. Maiss J, Hildebrand V, Bayer J, et al. miniEASIEEin neues, auf die
Belange der interventionellen Endoskopie reduziertes
Trainingsmodell. Endoskopie heute 1999; 12:53.
59. Maiss J, Ngel A, Tex S, et al. EASIE-team-training ERCP
Experiences with a new training concept for interventional ERCP.
Endoscopy 2000; 32:E65.
60. Hochberger J, Maiss J, Ngel A, et al. Polypectomy, endoscopic
staining techniques, mucosectomyA new structured team
training course in a close to reality endoscopy simulator (EASIE).
Endoscopy 2000; 32:E23.
61. Matthes K, Cohen J. The Neo-Papilla: a new modication of
porcine ex vivo simulators for ERCP training (with video).
Gastrointest Endosc 2006; 65:in press.
62. Lange V, Grund KE. Education in intraluminal endoscopy
experiences up to now. Chirurg 2001; 72: suppl 164165.
63. Neumann M, Mayer G, Ell C, et al. The Erlangen Endo-Trainer: life-
like simulation for diagnostic and interventional endoscopic
retrograde cholangiography. Endoscopy 2000; 32:906910.
64. Maiss J, Tex S, Naegel A, et al. EASIE Team Training ERCP
Experiences with a new training concept for interventional ERCP.
Endoscopy 2001; 33, Suppl. 1:AB 2119 (Abstr.).
65. Hochberger J, Maiss J, Neumann M, et al. EASIE-Team-Training in
Endoscopic HemostasisAcceptance of a systematic training in
interventional endoscopy by 134 trainees. Gastrointest Endosc
1999; 49:AB143.
66. Hochberger J, Matthes K, Maiss J, et al. Training with the
compactEASIE biologic endoscopy simulator signicantly
improves hemostatic technical skill of gastroenterology fellows: a
randomized controlled comparison with clinical endoscopy
training alone. Gastrointest Endosc 2005; 61:204215.
69
 SECTION 1 GENERAL TOPICS
67. Maiss J, Wiesnet J, Proeschel A, et al. Objective benet of a 1-day
training course in endoscopic hemostasis using the
compactEASIE endoscopy simulator. Endoscopy 2005;
37:552558.
68. Waye JD, Leicester RJ. Teaching endoscopy in the new
millennium. Gastrointest Endosc 2001; 54:671673.
69. Ladas SD, Malfertheiner P, Axon A. An introductory course for
training in endoscopy. Dig Dis 2002; 20:242245.
70. Matthes K, Cohen J, Kochman ML, et al. Efcacy and costs of a
one-day hands-on EASIE endoscopy simulator train-the-trainer
workshop. Gastrointest Endosc 2005; 62:921927.
71. Kalloo AN, Singh VK, Jagannath SB, et al. Flexible transgastric
peritoneoscopy: a novel approach to diagnostic and therapeutic
interventions in the peritoneal cavity. Gastrointest Endosc 2004;
60:114117.
70
72. Rattner D, Kalloo A. ASGE/SAGES Working Group on Natural
Orice Translumenal Endoscopic Surgery. October 2005. Surg
Endosc 2006; 20:329333.
73. Kantsevoy SV, Jagannath SB, Niiyama H, et al. Endoscopic
gastrojejunostomy with survival in a porcine model. Gastrointest
Endosc 2005; 62:287292.
74. Hochberger J, Lamade W. Transgastric surgery in the abdomen:
the dawn of a new era? Gastrointest Endosc 2005; 62: 293296.
75. Jagannath SB, Kantsevoy SV, Vaughn CA, et al. Peroral transgastric
endoscopic ligation of fallopian tubes with long-term survival in a
porcine model. Gastrointest Endosc 2005; 61:449453.
76. Pai R, Fong D, Fishman D, et al. Natural Orice Transluminal
Endoscopic Surgery (NOTES) Cholecystectomy: A Transcolonic
Survival Study in a Porcine Model. Gastrointest Endosc 2006; 63:
AB229, 623 (Abstr.) Abstract #623.
 SECTION 2 TECHNIQUES
Chapter
Cannulation of the Major and
8 Minor Papilla
Douglas G. Adler, Bret T. Petersen, and Todd H. Baron
CANNULATION OF THE MAJOR PAPILLA
Introduction
As the vast majority of ERCP procedures involve selective cannula-
tion of the biliary tree and/or main pancreatic duct, command of
techniques to access these ducts via the major duodenal papilla are
fundamental to clinical success. Because most ERCPs are performed
with therapeutic intent, failure to cannulate produces global proce-
dural failure, emphasizing the importance of cannulation.
In the course of training in therapeutic ERCP, learning to can-
nulate the desired duct consumes the most time, effort, and energy
on the part of both the trainee and the instructor. In clinical practice,
more time will likely be spent on cannulation than on any other
maneuver. Efforts during and after training should focus not only
on successful cannulation, but also on efcient, safe, and minimally
traumatic cannulation.
Intubation
Esophageal intubation with a duodenoscope is most easily done with
the instrument held horizontally and parallel to the examination
table with the patients neck slightly exed. Often only upward tip
deection (backward tension on the large wheel) combined with
gentle advancement is required to pass through the hypopharynx to
the level of the upper esophageal sphincter, although some right or
left tip deection (via backward rotation of the lateral wheel) can
often be helpful. Despite the assumption that intubation is done
blindly, evaluation of the oral cavity, epiglottis, vocal cords and the
pharynx can all be accomplished with side viewing instruments.
Once the scope has been advanced to the upper esophageal sphincter
(UES), gentle pressure combined with subtle tip deection will often
be adequate for esophageal intubation to proceed. Rotation and
slight downward tip deection (forward on the large wheel) will
allow esophagoscopy to be performed, if required. This is easier and
has less potential for complication when the endoscope is at the
gastroesophageal junction. If detailed evaluation of the esophagus is
not indicated, the duodenoscope can be advanced in a neutral posi-
tion while examining the esophageal wall for landmarks such as the
Z-line and the transition to gastric mucosa.
Passage through the stomach is usually straightforward. Imme-
diately upon entering the stomach, leftward endoscope torque and
brief tip extension (large wheel forward) combined with endoscope
advancement and air insufation will allow visualization of the
fundus and a portion of the gastric body. A lowering of the left hand
to the level of the table and simultaneous partial extension of the left
elbow (dropping the left hand) will also facilitate this initial maneu-
ver. Once the gastric body has been identied the endoscope is
advanced, following the direction of the gastric folds, allowing it to
pass along the surface of the greater curvature of the stomach
towards the antrum. During this advancement, the endoscopist
should slowly raise the left hand back to the vertical position so that
the endoscopic image rotates in a clockwise manner and the inci-
sura, when seen, appears to be horizontal. The posterior gastric wall
will be rightward in the endoscopic eld. Identication of the pylorus
often occurs at this time. Advancement at this point with the endo-
scope tip in a neutral position will bring the pylorus (and a limited
view of the post-bulbar duodenum) into full view. This image will
be replaced with further advancement with the classic setting sun
as the endoscope approaches and ultimately traverses the pylorus,
resulting in a close view of the duodenal bulb with its characteristic
mucosa. During attempts at traversal through the pylorus, some
upward tip deection combined with lateral tip deection can be
helpful if the tip of the endoscope repeatedly rolls off of the
pylorus.
Passage through the stomach can sometimes be difcult when
the patient is in the prone position. Flexure of the patients right
knee, combined with rotation of the patients shoulders to approxi-
mate a left-lateral decubitus position, is often helpful if difculty is
encountered. Inexperienced endoscopists will often over-ex the tip
of the endoscope upon arrival into the stomach, which can result in
retroexion in the fundus. Gastric transit can also be difcult in
patients with a J-shaped stomach or in obese patients. It should be
emphasized that the duodenoscope is designed to easily pass through
the stomach and that if repeated attempts at passages through the
stomach fail it is rarely the fault of the instrument. Often, a return
to rst principles via withdrawal of the endoscope to the distal esoph-
agus, removal of excessive air, and careful re-intubation of the
stomach will allow successful passage.
Once past the pylorus, tip extension (large wheel forward) will
provide a global view of the duodenal bulb and will show the proper
route for further advancement to the second portion of the duode-
num. Gentle advancement combined with tip exion will allow the
instrument to follow the curve of the duodenum to the long endo-
scope position, with the shaft of the endoscope still against the
greater curvature of the stomach (Fig. 8.1). The maneuver will leave
the tip of the endoscope at the end of the second or beginning of
the third portion of the duodenum. The standard shortening maneu-
ver consists of full rightward deection of the lateral wheel and full
upward of the large wheel (often locking both) followed by with-
drawal and simultaneous clockwise torque applied to the endoscope
73
 SECTION 2 TECHNIQUES
Fig. 8.1 Illustration when duodenoscope is in the long position.
Note endoscope is against greater curvature of the stomach.
A Fig. 8.3 A Illustration of separate openings
B separate openings. Bile can be seen around the
shaft by the right hand. The tip exion helps to help bow the
duodenal wall and position the duodenoscope under the
papilla.
In some patients, only a long endoscope position (with the
endoscope shaft following the greater curvature of the stomach) will
allow adequate alignment and orientation to the major papilla to be
obtained. For patients under conscious sedation, the long endoscope
position may be difcult to tolerate for long periods, due to gastric
distension.
Evaluation of the papilla and initial positioning
Upon arrival in the 2nd duodenum and following the standard short-
ening maneuver (with the shaft of the endoscope now ush against
the lesser curvature of the stomach), the endoscopist will usually be
in a position to readily identify the major papilla. Rarely, the papilla
can be difcult to identify at all as it may be obscured behind a fold,
distorted or ablated in patients with malignancy, or indistinct in
patients with bowel wall edema (as can be encountered in patients
with acute pancreatitis or severe hypoalbuminemia). In some
patients with aberrant anatomy, the papilla can be located proxi-
mally, just beyond the apex of the duodenal bulb, or distally, near
the junction of the second and third duodenum. Once identied, a
brief evaluation of the papilla allows assessment of the type and size
of the papilla and any unusual features that may affect cannulation
74
CBD
PD
CC
Fig. 8.2 Illustration of typical anatomy of the ampulla of Vater. A
short common channel (CC) is present. The septum (S) separates
the common bile duct (CBD) and pancreatic duct (PD).
for the CBD and PD. B Actual papilla with
biliary opening.
specically or the procedure as a whole (periampullary diverticula,
visibly impacted stones, etc.).
Most patients with native papillary anatomy will have a single
ampullary orice that provides access to a common channel that
further bifurcates via a septum into the distal common bile duct
(CBD) and the main pancreatic duct (PD), also known as the duct
of Wirsung (Fig. 8.2). A minority of patients will have two separate
orices (Figs 8.3A and 8.3B), but this may not always be apparent
on initial evaluation as these two separate openings may be in close
proximity. While the overall size and shape of the papilla can some-
times provide information about the possible length of the intraduo-
denal segment of the distal CBD or the angles at which the CBD and
PD diverge from the common channel, in general, the intra-papillary
anatomy can be difcult to dene via inspection alone.
As a general rule of thumb, time spent on proper positioning
prior to attempts at cannulation is time well spent. A thorough
examination of the papilla from several angles with the wheel locks
off and with variations in torque applied at the endoscope head, the
positions of the main and lateral wheels, and subtle alterations in
the depth of endoscopic intubation will often provide the endosco-
pist with several possible positions from which to attempt cannula-
tion. At the same time, these maneuvers will usually lead to the
selection of a presumed optimal position from which to rst attempt
cannulation.

Standard catheters
Catheters and sphincterotomes preloaded with guidewires
Placement of pancreatic guidewire or stent followed by biliary
cannulation
Precut access sphincterotomy
Needle-knife (with or without pancreatic duct stent placement)
Freehand starting at orice
Freehand starting above orice stulotomy
Transpancreatic
Rendezvous techniques
Table 8.1 Biliary cannulation methods
Opinions differ regarding the optimal distance from the tip of the
endoscope to the papilla, but 23 cms of distance is commonly used.
This middle-length distance allows for changes in knob and eleva-
tor position and depth of endoscopic intubation to have a clearly
visible effect on the view of the major papilla. In addition, when
using sphincterotomes, this distance allows adequate room for
proper bowing to occur. Attempting to cannulate from an extremely
close position, while affording an excellent view of the ampullary
orice, may not allow endoscopic maneuvers to signicantly change
the view of the papilla and will not allow for signicant bowing of a
sphincterotome, as these devices need at least partial extension out
of the accessory channel to function properly. Cannulating from a
long distance from the papilla, although valuable in certain situa-
tions, may require excessive cannula or sphincterotome extension,
decrease the mechanical advantage of extended accessories, and
therefore limit the usefulness of these devices.
Cannulation of the common bile duct
There are a variety of methods used to cannulate the common bile
duct (Table 8.1).
Personal preferences vary regarding whether cannulation should
be attempted with a cannula or a sphincterotome. This decision is
usually based on the indication for the procedure as a whole, and
whether the procedure will require a sphincterotomy. Regardless of
which device is rst selected to cannulate the biliary tree, several
principles universally apply.
When facing the major papilla en face, the biliary orice will
nearly always be located in the left upper quadrant, corresponding
to the 9 oclock to the 12 oclock position (Fig. 8.4). The angle of the
distal CBD can often be extrapolated, at least partially, by evaluating
the angle and direction of the intraduodenal portion of the distal
CBD. This angle, combined with the presumed location of the biliary
orice in the major papilla, will help to guide attempts at cannula-
tion. Using these two pieces of information, one can project an
imaginary line into the lumen showing the direction the CBD would
take if it extended beyond the papilla. The endoscopist should
manipulate the endoscope and cannulating device to move along
this imaginary line into the CBD.
Typically, approaches to the biliary orice directed perpendicu-
larly to the duodenal wall will not produce access to the biliary tree,
although this is frequently attempted by the novice endoscopist as
it may not be appreciated that the axis is incorrect. Examination of
the catheter position under uoroscopy will show the inappropriate
axis. An upward sweeping motion of the catheter is required to allow
the device tip to mirror the true direction of the distal CBD (Fig. 8.5).
Once the tip of the selected device has partially engaged the orice
Chapter 8 Cannulation of the Major and Minor Papilla
Bile duct location
Fig. 8.4 Illustration of the location of the CBD opening on the
papilla when viewed en face.
at its most superior point, and hopefully above the septum of the
common channel separating the CBD and the PD, direct access may
be obtained by applying gentle pressure to the cannulating device.
Other maneuvers, such as gentle relaxation of the elevator, a subtle
withdrawal of the device and/or the endoscope, and reduced bowing
(if a sphincterotome is being used) will often allow deep entry into
the CBD in this situation. Simultaneous un-bowing of the sphinc-
terotome in conjunction with pulling the scope shaft more tightly
along the lesser curve of the stomach is often a very effective tech-
nique to achieve selective CBD cannulation once the tip of the
sphincterotome has been inserted into the papillary orice.
Although simple in conception, cannulation of the CBD can often
be quite difcult, even for experienced endoscopists. Patients can
have aberrant papillary or duodenal anatomy (see below) that com-
plicates attempts at cannulation. They may have seemingly normal
anatomy, yet access to the CBD may not be obtained via standard
maneuvers. Over time, a set of more advanced techniques have been
developed that can assist in a variety of situations.
Standard techniques for biliary cannulation
Cannulation with a sphincterotome
With the rise of noninvasive techniques to evaluate the pancreatico-
biliary tree, ERCP has evolved into an almost purely therapeutic
technique. As such, initial attempts at cannulation are frequently
performed with a sphincterotome rather than a standard biliary
catheter. Sphincterotomes can be bowed at their tips by placing the
cutting wire under tension. Use of a sphincterotome can often facili-
tate rapid access to the CBD as the angle of the distal duct can be
more closely approximated from within the duodenal lumen (Figs
8.5A and 8.5B). Sphincterotomes are available in many different
specications. Different sphincterotomes from various manufactur-
ers often have very different mechanical properties (catheter shaft
thickness, degree of bowing, length of cutting wire, catheter tip
diameter, length, and degree of tapering, etc) and can be designed
to accept wires of different diameters (most commonly 0.025 inches
or 0.035 inches) which can also affect how the devices operate (see
Chapter 4). For example, a sphincterotome with a 30 mm length of
cutting wire will often ex to a greater extent than a sphincterotome
with a 20 mm length of cutting wire, but will need to be more fully
advanced into the lumen to allow complete bowing. Likewise, a
75
 SECTION 2 TECHNIQUES
A B
sphincterotome designed to accommodate a 0.035 wire is likely to
be stiffer than a device for a 0.025 wire. Although some sphinctero-
tomes are specically designed to be rotated clockwise or counter-
clockwise in a controlled manner to assist in obtaining an adequate
orienta-tion to the papilla, most sphincterotomes can be rotated to
some degree.
It should be stressed that for a sphincterotome to properly and
fully bow, the cutting wire must be advanced out of the endoscope
and into the lumen. Attempts at bowing with the cutting wire par-
tially within the endoscope channel will limit the ability of the device
to ex. Bowing with the cutting wire in the endoscope channel can
sometimes be valuable as it allows the sphincterotome to ex in a
dynamic manner when advanced out of the channel while simulta-
neously applying tension to the cutting wire.
There have been few studies comparing sphincterotomes to stan-
dard catheters for initial biliary cannulation.16 In these studies
sphincterotomes provided faster and easier access to the biliary tree,
although no large scale studies have been done on this topic to date.
Guidewire assisted cannulation
If attempts at cannulation with a standard catheter or a sphinctero-
tome are unsuccessful, a guidewire can be inserted through either
device to assist in cannulation (Fig. 8.6). Guidewire cannulation can
be performed via several different techniques. The guidewire can
be advanced into the papilla by itself with the accessory in contact
with the papilla. Conversely, the cannula or sphincterotome can be
advanced into the ampullary orice for several millimeters prior to
attempts at guidewire advancement. Both of techniques are per-
formed under endoscopic and uoroscopic guidance to allow assess-
ment of extraluminal guidewire position. Most currently available
guidewires have a soft, hydrophilic segment as their leading segment,
facilitating passage of the guidewire through a tortuous common
channel.
Guidewire cannulation has several advantages: it is relatively
atraumatic, the guidewire itself does not generate signicant hydro-
static pressure (thus minimizing the risk of post-ERCP pancreatitis
if the wire accesses the pancreatic duct) and the position of the
guidewire (and thus the duct cannulated) can be easily detected via
uoroscopy. Guidewire cannulation carries a small risk pancreatitis
if the guidewire is forcefully advanced into a side branch of the
pancreatic duct causing trauma. Guidewire-induced perforation
typically occurs within the papillary structures themselves. If the
76
Fig. 8.5 A Cannulation of the bile duct
using a standard catheter. Note steep trajec-
tory that the biliary axis takes. B Bowed
sphincterotome facilitates biliary cannula-
tion by achieving the steep axis.
Fig. 8.6 Biliary cannulation using a guidewire.
sphincterotome (with or without a guidewire) exits the intraduode-
nal segment of the bile duct, either purposefully or accidentally, a
unique form of precut sphincterotomy may be performed by cutting
the tissue between the papillary orice and where the guidewire
exits.7 The biliary orice is then exposed (Figs 8.7A and 8.7B).
Guidewire cannulation has been evaluated formally in only a
limited manner. In the only study of its kind, Lella et al. compared
guidewire cannulation to traditional methods of cannulation in 400
patients when performed by one endoscopist.8 In this study, both
techniques had a very high success rate (approximately 98% for
either method) but there were no cases of post-ERCP pancreatitis in
the guidewire cannulation group compared to eight cases in the
control group. This was likely due to the lack of hydrostatic pressure
incurred by the pancreatic duct during guidewire cannulation.
Pancreatic duct stent placement to facilitate
biliary cannulation
If attempts at biliary cannulation consistently result in entry to the
pancreatic duct, placement of a pancreatic duct stent can often facili-
tate access to the CBD (Figs 8.8A and 8.8B). Pancreatic duct stent
placement has several potential advantages. First, pancreatic duct
stents appear to reduce the risk of post-ERCP pancreatitis, likely by
ensuring pancreatic ductal decompression and drainage. This is
especially true in patients who have experienced repeated pancreatic
duct injections.911 Secondly, pancreatic duct stents may also help to
block the pancreatic duct orice from further attempts at cannula-

A B
Fig. 8.7 A The tip of a sphincterotome has passed from the papil-
lary orice through the intraduodenal portion of the bile duct
during attempted biliary cannulation. B After severing the intra-
duodenal portion.
B practiced for several years, only one randomized study and one small
A
Fig. 8.8 A Endoscopic view of biliary cannulation above the pre-
viously placed pancreatic duct stent. B Fluoroscopic view of the
same patient. Pancreatic duct stent is seen (arrow).
tion, both minimizing further pancreatic manipulation and facilitat-
ing cannulation of the biliary orice. Lastly, the direction at which
the pancreatic duct stent exits the papilla may give clues, via extrapo-
lation, about the optimal angle to further approach the papilla to
obtain biliary access. This latter benet is most useful when the
anatomy is distorted by severe duodenal edema or by tumor. Pan-
creatic stent placement to facilitate biliary cannulation has not been
well studied, but the limited available data supports the efcacy and
safety of this approach.12
The optimal type, size, and conguration of a pancreatic stent
that would both facilitate biliary access and minimize the risk of
post-ERCP pancreatitis are unknown. One study has suggested that
3Fr pancreatic stents without internal aps are superior to short
length 5Fr pancreatic stents for preventing post-ERCP pancreatitis,13
but many physicians use 5Fr stents routinely. Stents used in this
setting without internal aps will typically fall out spontaneously
after several days, but can pass prematurely immediately following
placement. Internal aps prevent premature stent dislodgement, but
may not pass spontaneously and require another endoscopic proce-
dure for removal. Smaller stents and those without internal aps,
however, may become dislodged when placed prior to attempted
biliary cannulation.
Chapter 8 Cannulation of the Major and Minor Papilla
Guidewire placement in the pancreatic duct to facilitate
biliary access
If attempts at biliary cannulation result in repeated pancreatic duct
cannulation, another option to facilitate biliary cannulation is a tech-
nique that involves the use of two separate guidewires. A guidewire
is advanced to the mid- to distal pancreatic duct, and the cannula or
sphincterotome is removed leaving the guidewire in place. The
device is preloaded with a second guidewire and biliary cannulation
is re-attempted (Fig. 8.9AC) (b). The angle of the rst wire emanat-
ing from the pancreatic orice can give clues that can assist in biliary
cannulation. In addition, attempts at biliary cannulation, when
observed under uoroscopy, will readily reveal whether or not the
second guidewire is simply being advanced into the pancreas.
Leaving a guidewire in the pancreatic duct, even for protracted
amounts of time, does not generate intraductal hydrostatic pressure
and does not appear to be associated with an increased risk of post-
ERCP pancreatitis. Although this technique has been known and
case series have been performed to evaluate this technique formally,
both with very encouraging results.1416
Precut biliary sphincterotomy and
associated techniques
The term precut refers to the action of performing a sphincterot-
omy (the cut) prior to obtaining biliary access (the pre-). The
term is often used, both in print and in speech, interchangeably with
the term needle-knife sphincterotomy. In fact, the term precut
refers rather to a group of high-risk techniques to obtain biliary
access if standard maneuvers are unsuccessful. These techniques
are collectively termed access sphincterotomy techniques, and carry
a greater risk of complications. These techniques will be summa-
rized briey here and are discussed in detail in Chapter 9.
Needle-knife sphincterotomy
Needle-knife sphincterotomy refers to the use of a catheter capable
of delivering electrosurgical current through a wire to directly dissect
the major papilla in vivo to obtain ductal access. The standard needle-
knife design incorporates a bare wire that can be extended through
a modied biliary cannula. These devices do not routinely have the
capability to bow as do most sphincterotomes. The most common
way this device is used involves beginning the sphincterotomy at the
level of the biliary orice and moving in a cephalad direction (Fig.
8.10A and 8.10B). Another variation on this technique involves
beginning the incision above the level of the ampullary orice and
attempting to cut directly into the CBD via the creation of a biliary
stula; this approach is referred to as precut stulotomy (Fig. 8.11).
Once the CBD is accessed, a standard sphincterotome can be used
to extend the sphincterotomy, if needed.1723
Because the endoscopist is cutting into the papilla and the duo-
denal wall without the benet of a guidewire in the CBD, complica-
tions such as bleeding, perforation, and pancreatitis can easily
develop. Complications can occur in as many as one-third of all
needle-knife sphincterotomies, emphasizing the risk involved with
this approach, although most reports using this technique suggest
complication rates approximating 1518% (signicantly higher than
seen when sphincterotomy is performed with standard sphinctero-
tomes). Complication rates may not decrease despite both increased
experience and success in obtaining biliary access with needle-knife
sphincterotomy.24 Placement of a pancreatic duct stent prior to
77
 SECTION 2 TECHNIQUES

A B

Fig. 8.9 A Illustration of double wire technique. B Endoscopic view of sphincterotome

alongside pancreatic guidewire. C Fluoroscopic view of double wire technique.

A B Fig. 8.10 Illustration of precut sphincter-

otomy technique using a needle-knife
beginning at the papillary orice. A Without
pancreatic duct stent. B With pancreatic
duct stent.

Fig. 8.11 A Illustration of precut stulotomy

A B technique. B Successful precut stulotomy. The cut
was performed with a needle-knife and is well above
the papillary orice (arrow).

78

Fig. 8.12 Illustration of transpancreatic septotomy technique.
performing needle-knife sphincterotomy may decrease the risk of
post-ERCP pancreatitis by preventing trauma or obstruction of the
pancreatic orice due to swelling.
Needle-knife sphincterotomy is a high-risk technique used to
obtain ductal access, and is not intended for indiscriminate use. In
general, prior to attempting needle-knife sphincterotomy, legitimate
attempts at cannulation with standard techniques should have been
attempted. Failed biliary cannulation in the setting of clinical situa-
tions such as cholangitis, a visibly impacted stone at the level of the
ampulla, and obstructive jaundice due to malignancy are all appro-
priate for cannulation attempts using this technique. However, not
all patients are ideal candidates for needle-knife sphincterotomy. In
clinically stable patients undergoing ERCP for chronic abdominal
pain or for abnormal liver function tests of unclear etiology, needle-
knife sphincterotomy may not be warranted due to the associated
risks.
Endoscopic transpancreatic papillary septotomy
Endoscopic transpancreatic papillary septotomy is a technique
whereby biliary access is attempted via the use of a traction-type
sphincterotome with its tip located in the pancreatic duct, but with
the endoscope and the sphincterotome oriented for a biliary sphinc-
terotomy. This allows the cut to begin in the pancreatic duct but to
traverse the septum between this structure and the CBD, ultimately
resulting in biliary access (Fig. 8.12). Results with this technique
have been mixed, with some practitioners having a very high success
rate with low rates of complications, while others have incurred
complications at a level comparable to that of standard needle-knife
sphincterotomy.2528 Endoscopic transpancreatic papillary septotomy
is not widely performed at the present time as it is unclear if it offers
any signicant benet over standard needle-knife sphincterotomy.
Cannulation in patients with periampullary
duodenal diverticula
Duodenal diverticula are common in the general population. Those
occurring in proximity to or the papilla are sometimes known as
periampullary diverticula, while diverticula directly involving the
major papilla are sometimes referred to as ampullary diverticula.
This terminology is not standardized and many authors simply use
the term periampullary diverticula in a catch-all manner. Data
Chapter 8 Cannulation of the Major and Minor Papilla
A B
Fig. 8.13 A Endoscopic view of papilla within diverticulum. Note
papillary orice (arrow) is facing the wrong direction. The intraduo-
denal portion of the bile duct is seen (arrowhead). B After can-
nulation, the papilla orients properly.
suggest that, although asymptomatic in many patients, ampullary
diverticula can actually predispose patients to pancreatobiliary
disease, most commonly stone disease.2931 Common bile duct can-
nulation may sometimes be easier in the presence periampullary
diverticula if accessible. The challenge is gaining the appropriate axis
for access. In some cases, the ampulla abuts the diverticula, while
in others the papilla is partially or completely located within the
diverticula. The resultant distortion in the anatomy may render stan-
dard approaches and landmarks useless. The relative locations of
the biliary and pancreatic orices within the ampulla can even be
reversed due to rotation of the ductal system within the ampulla. In
addition, the ampullary orice may be located eccentrically within
the diverticula, and may be oriented away from an endoscopically
approachable position (Fig. 8.13A and 8.13B) and, in rare cases, may
not even be identiable endoscopically.
Cannulation in patients with duodenal diverticula can range from
straightforward to extremely challenging, and in general the pres-
ence of ampullary diverticula will add a layer of complexity to can-
nulation of either duct. In one retrospective analysis of ERCP in 72
patients with periampullary diverticula, difculty in cannulation was
encountered in 79% of cases as compared with 10% of cases in
patients without diverticula (p < 0.001).32
If the ampulla is located within the diverticula and oriented away
from the endoscope, an endoscopic clip can be applied to the edge
of the diverticula to provide a more open view of its interior while
at the same time often everting the major papilla.33 Sometimes two
devices can be advanced side-by-side through a large caliber acces-
sory channel and used together to manipulate the diverticula. Ultra-
thin forceps advanced alongside a catheter or sphincterotome can
apply traction or tension to diverticula, whereas conventional forceps
can be used in conjunction with a cannulating guidewire. Similarly,
the use of two separate endoscopes simultaneously to gain access in
patient with an intradiverticular ampulla has been reported.34 Pan-
creatic duct stenting of patients with diverticula may facilitate access
at cannulation with routine devices or needle-knife sphinctero-
tomes.35 Given the tortuous nature of the CBD and PD in these
patients, guidewire cannulation can be extremely helpful in these
patients. Narrow (0.025 or less), hydrophilic guidewires may be able
to follow the course of an irregular duct and allow deep biliary or
pancreatic access. Although standard biliary sphincterotomy in the
79
 SECTION 2 TECHNIQUES
setting of a diverticulum does not carry a higher complication rate,
a needle-knife precut sphincterotomy may be particularly dangerous
if the intraduodenal portion of the bile duct cannot be clearly delin-
eated. On the other hand, in the presence of a pantaloon diverticu-
lum, the intraduodenal anatomy may be more apparent than in the
usual situation, making needle-knife precut sphincterotomy easier.
Repeat ERCP
If cannulation fails, an alternative is consideration for referral of the
patient to a more experienced endoscopist. In one study of 113
patients referred to a tertiary institution for repeat ERCP after a
failed prior attempt at another hospital, successful cannulation of
the desired duct was achieved in 96% of attempts.36 Thus, second
attempt ERCP is generally worthwhile if clinically indicated and
ERCP expertise is geographically available. A more recent study
from the same institution conrmed these results.37 Yet another
option is to repeat the ERCP on another day. Only one study pub-
lished from a teaching institution addressed the success rate of
repeat ERCP by the same endoscopist after a failed initial attempt.38
At repeat endoscopy access to the desired duct was achieved in
87.5% of the patients. A needle-knife was used in 21% and no com-
plications occurred with repeat ERCP.
Rendezvous techniques
In some situations, endoscopic attempts at cannulation will be
unsuccessful despite all of the above approaches. In patients in
whom internal biliary drainage is still desired, one can achieve this
goal via a so-called rendezvous procedure. The procedure can be
performed several ways, but essentially entails having an interven-
tional radiologist obtain percutaneous biliary access. This access can
be obtained either via a peripheral duct or via a transcystic approach
(through a percutaneous cholecystostomy). A catheter or a guidewire
is passed in an antegrade manner into the bile duct and through the
ampulla and into the duodenum.3943 In some situations, the trans-
ampullary guidewire or catheter can be placed during a surgical
procedure, most commonly during cholecystectomy.44
Once the guidewire is passed into the duodenum, endoscopic
access can be achieved by grasping the guidewire with a polypectomy
snare or a biopsy forceps and withdrawing it through the accessory
channel. Standard cannulas and/or sphincterotomes are then
advanced over the wire to allow endoscopic access. If a small diam-
eter percutaneous drainage tube (5Fr or less) has been placed in the
duodenum instead of a guidewire, it is often a simple manner to
cannulate next to the tube to access the biliary tree. After accessing
the bile duct, any and all required therapeutic maneuvers can be
performed (sphincterotomy, stone extraction, transpapillary stent-
ing, etc). The percutaneous catheter or guidewire can be removed
after endoscopic biliary access and drainage are secured. If larger
catheters were placed percutaneously, they may need to be left in
place until a mature tract has formed. The decision and timing of
catheter removal should be coordinated with the interventional radi-
ology team.
Rather than to pass a wire antegrade as described above, one
alternative is to endoscopically cannulate alongside the drain. After
successful cannulation, the drain is progressively withdrawn during
the procedure to allow retrograde therapeutic intervention. This has
been referred to as parallel cannulation.45
There is no consensus on whether it is superior for patients to
have a percutaneous placement of a guidewire alone or a formal
drainage catheter. In general, if ERCP is to be performed immedi-
80
ately following a percutaneous biliary access procedure, a guidewire
alone or a low-prole (3Fr) catheter can be placed. Unprotected
guidewires have been associated with laceration of the liver capsule
and parenchyma during manipulation, and placement without a
small diameter protective sheath should be discouraged. If ERCP
and internalization of biliary drainage is to be delayed, placement of
a larger caliber drainage tube (10Fr or greater) seems warranted.
Endoscopic rendezvous without the need for percutaneous access
has been described through side-to-side choledochoduodenal anas-
tomoses to facilitate cannulation of the bile duct for the treatment
of sump syndrome.46
A modern variation on this theme is to perform an EUS-guided
rendezvous procedure.4749 In this situation, a linear echoendoscope
is used to identify a bile duct (which, unlike a blood vessel, has no
Doppler ow signature). The bile duct selected for access can be an
intrahepatic or extrahepatic duct. Using an FNA needle, biliary
access is obtained and cholangiography can be performed. Under
uoroscopy, a guidewire is advanced through the FNA needle into
the bile duct in an antegrade manner and, ultimately, through the
ampulla to allow endoscopic cannulation to proceed as described
above. In some cases, ERCP may be obviated if drainage, via a stent,
can be achieved via EUS alone.
Cannulation of the pancreatic duct
When facing the major papilla en face, the pancreatic orice will
nearly always be located in the mid to right lower quadrant, corre-
sponding to the 1 oclock to the 5 oclock position (Fig. 8.14). In
contrast to cannulation of the common bile duct, cannulation of
the ventral pancreatic duct (Duct of Wirsung) is usually performed
with the catheter oriented in a more straight-on direction (Fig. 8.15).
Cannulation of the pancreatic duct is usually achieved in a
short endoscope position, with a more anterior facing endoscope
lens view (leftward rotation of the lateral wheel). The endoscope
should be at or slightly above the level of the major papilla (as
opposed to the bile duct where a below the papilla position is
favorable). The middle distance between the tip of the endoscope
and the papilla is (23 cm) described for biliary cannulation is often
useful in pancreatic cannulation, although a closer approach is
sometimes required.
Pancreatic duct location
Fig. 8.14 Illustration of the location of the PD opening on the
papilla when viewed en face.

Fig. 8.15 Cannulation of the pancreatic duct using a standard cath-
eter. Note at trajectory that the pancreatic axis takes.
Fig. 8.16 Prior small biliary sphincterotomy (arrow). The pancre-
atic duct is cannulated separately.
Guidewire cannulation of the pancreatic duct is recommended as
it is much less likely to produce hydrostatic overpressure in the
pancreatic duct, thus minimizing the risk of post-ERCP pancreatitis.
Smaller diameter guidewires (0.025 or less) are recommended by
some endoscopists for pancreatic guidewire cannulation, as are
hydrophilic guidewires, although many modern guidewires have a
hydrophilic tip that is quite helpful for pancreatic cannulation. More
aggressive techniques, such as needle-knife sphincterotomy, are
rarely required for pancreatic duct access. In patients who have
undergone prior biliary sphincterotomy, the pancreatic orice is
usually easier to identify as a separate opening located below and to
the right (as the endoscopist faces the papilla) of the biliary sphinc-
terotomy (Fig. 8.16). However, in some instances it may be more
difcult to identify. The orice may be located anteriorly or within
the prior biliary sphincterotomy in a patient with a long combined
sphincter (>10 mm). It is often helpful to look for residual ampullary
structures such as mucosal fronds, as the pancreatic orice is usually
within the connes of this ampullary remnant. Although most com-
monly used to facilitate cannulation of the minor papilla in patients
with pancreas divisum, intravenous secretin can assist in cannula-
tion of the main pancreatic duct as well.50
Chapter 8 Cannulation of the Major and Minor Papilla
Establish diagnosis of pancreas divisum
Endoscopic therapy for patients with pancreas divisum
Minor papilla sphincterotomy
Stone extraction
Stent placement
Endoscopic therapy for non-divisum patients in whom major
papilla cannulation fails
Table 8.2 Indications for minor papilla cannulation
Fig. 8.17 Minor papilla almost adjacent to major papilla.
CANNULATION OF THE MINOR PAPILLA
Standard techniques
Indications for cannulation of the minor papilla are listed in Table
8.2. Although cannulation of the minor papilla can be performed to
conrm the diagnosis of pancreas divisum, non-invasive imaging
studies such as thin-slice coronal CT,51 secretin-MRCP52 and EUS53
can provide the diagnosis of pancreas divisum with a high degree of
accuracy (see Chapter 42). If cannulation of the minor papilla is
performed only for diagnosis of pancreas divisum, deep cannulation
is often not required. However, as in most other aspects of ERCP,
cannulation of the minor papilla is undertaken in order to perform
therapeutic interventions, and deep cannulation is mandatory. Addi-
tionally, in non-divisum patients, cannulation of the minor papilla
can allow therapeutic interventions when cannulation of the main
pancreatic duct cannot be achieved.54 Minor papilla sphincterotomy
is discussed in detail in Chapter 15. This chapter will discuss can-
nulation of the minor papilla.
A standard duodenoscope is used. Because a long endoscope
position may be required to achieve an en face position of the minor
papilla, a more exible, smaller caliber diagnostic scope may
provide an advantage over a more rigid therapeutic channel endo-
scope, though this has not been formally studied.
The rst step for minor papilla cannulation is to locate the minor
papilla. It is located proximally to the major papilla on the medial
wall. The distance from the major papilla is variable, being imme-
diately adjacent or several centimeters cephalad (Fig. 8.17). In addi-
tion, the minor papilla may be very small or quite prominent (Fig.
8.18). Rarely, the minor papilla is within a diverticulum. Careful
81
 SECTION 2 TECHNIQUES
Fig. 8.18 Prominent minor papilla in patient with pancreas divisum
and acute recurrent pancreatitis.
Fig. 8.19 Probing of mucosal folds to expose the minor papilla.
inspection is needed, and in some cases probing of the folds can
allow identication (Fig. 8.19), although care must be taken not to
traumatize the duodenum as edematous and friable tissue may
make identication of the minor papilla particularly difcult. The
minor papilla may be recognized by carefully withdrawing the endo-
scope from the position of the major papilla. However, the long
endoscope position is frequently useful to allow identication and
initial cannulation because of an improved endoscopic view, as well
as an improved uoroscopic view, because the endoscope is not over
the distal (head) pancreatic duct (Fig. 8.20). If the endoscope is
already in a shortened position, it may need to be brought back into
the stomach and re-advanced so in the long position. It must be
stressed that once cannulation of the minor papilla is achieved, or
at least a guidewire has been positioned well into the main pancre-
atic duct, it is almost always advantageous to shorten the endoscope
because of improved mechanical advantage and patient comfort.
Secretin stimulates secretion of water and bicarbonate from pan-
creatic ductal cells and has been used for many years as an adjunct
82
Fig. 8.20 Long endoscope position with diagnostic cannulation
of the minor papilla using a metal tipped catheter. The endoscope
is not overlying the main pancreatic duct. Note prior biliary stent.
to minor papilla cannulation, since the output of juices can be identi-
ed endoscopically. Secretin is available from several sources. Secre-
FloTM (RepliGen Corp., Waltham, MA, USA) is a synthetic porcine
secretin administered at a dose of 1 clinical unit/kg (200 nanograms/
kg) and is FDA approved for aiding in minor papilla cannulation. It
has been shown in a randomized, placebo-controlled, double-blind,
comparative trial in patients with pancreas divisum and was found
to signicantly improve minor papilla cannulation rates and to
shorten cannulation time.55 ChiRhoStimTM (ChiRhoClin, Inc.,
ChiRhoClin, Inc, Burtonsville, MD, USA) is a puried synthetic
peptide with an amino acid sequence identical to the naturally occur-
ring hormone and has also been shown to improve cannulation of
the minor papilla as compared to placebo.56 It is administered at a
dose of 0.2 mcg/kg over one minute. It is now reimbursable with
CPT code 43271. An advantage of this agent is its prolonged stability
with 24 months in freezer, one year in refrigerator, and six months
at room temperature.
In the absence of severe chronic pancreatitis and/or complete
ductal obstruction, pancreatic output should occur within several
minutes.
In patients in whom cannulation is unsuccessful, minor papilla
cannulation may be facilitated by spraying dilute dyes such as meth-
ylene blue57 and indigo carmine58over the duodenal mucosa in the
region suspected to contain the minor papilla. The output of pan-
creatic juice from the minor papilla minor papilla orice creates an
enlarging spot as pancreatic juice washes away the dye (Fig. 8.21).
In patients with incomplete pancreas divisum, the dye can be mixed
with contrast and injected into the major papilla. The minor papilla
is then identied by dye decompressing through the accessing
papilla.
Minor papilla cannulation to obtain a dorsal ductogram is facili-
tated by the use of a needle tip catheter specically designed for this
indication (ERCP-1-CRAMER, Cook Endoscopy, Winston-Salem,
NC).59 This catheter does not accept a guidewire, but the nose of the
catheter dilates the opening and further facilitates catheter and
guidewire based cannulation. Some endoscopists prefer to use ultra-
tapered 3Fr catheters and 0.018 inch wires (Roadrunner, Cook
Endoscopy; Glidewire, Boston Scientic).60,61 We prefer to identify
the minor papilla with a needle-tip catheter as mentioned above.

Fig. 8.21 Minor papilla identication (arrow) after spraying with
dilute methylene blue.
Fig. 8.22 Minor papilla cannulation using a 5Fr catheter preloaded
with a 0.035 hydrophilic wire.
Once it is accurately identied and a pancreatogram is obtained, the
catheter is left in place to dilate the orice until the team is com-
pletely prepared with the catheter used for deep cannulation. When
ready, the metal tip catheter is removed and a 0.035 inch Terumo
wire (Glidewire, Boston Scientic) preloaded into a 5Fr catheter is
quickly passed into the channel and used to probe the papillary
opening (Fig. 8.22). Our success rate approximates 100% without
the need for precut sphincterotomy. The 0.018 wire is frequently
too oppy for manipulation. On rare instances, however, the wire
will pass into the duct, although the catheter will not. In this case,
the guidewire is left in place and a needle-knife is passed alongside
the guidewire. A precut is preformed by following the path of the
wire in order to enlarge the opening to the minor papilla.62 An alter-
native to this is the use a biliary dilation catheter (PassageTM, Boston
Scientic), which accepts an 0.035-inch guidewire. In some cases,
an en face orientation to the minor papilla cannot be achieved for
initial cannulation. In this setting, use of a sphincterotome to bow
Chapter 8 Cannulation of the Major and Minor Papilla
A B
C D
Fig. 8.23 Rendezvous in a patient with incomplete divisum. A A
prominent minor papilla is seen. B A guidewire catheter is passed
through the major papilla out through the minor papilla. C A cath-
eter is passed over the guidewire to dilate the minor papilla. D The
minor papilla is easily cannulated using a sphincterotome.
into the correct axis may be useful. This may also be preferable if it
is anticipated that minor papilla sphincterotomy using a pull sphinc-
terotome is to be performed (see Chapter 15).
Advanced techniques
In patients with incomplete divisum who require endoscopic therapy
of the minor papilla and/or dorsal duct, a rendezvous technique can
be employed. After cannulation of the major papilla, a guidewire is
passed from the ventral duct into the dorsal duct and out though the
minor papilla. On occasion, a catheter will follow the wire and act
as a dilator to allow subsequent retrograde cannulation of the minor
papilla (Fig. 8.23A8.23D).
One novel method of minor papilla cannulation is the use of
EUS. EUS can be used two ways, both of which involve needle
puncture of the main pancreatic duct using a linear array echoendo-
scope. The rst method is to pass a guidewire antegrade into and
through the minor papilla (e.g. rendezvous).63 The endoscope is
withdrawn and a standard duodenoscope is inserted. The wire
exiting the minor papilla is used to facilitate cannulation. The other
technique involves injecting contrast mixed with dilute methylene
blue into the main pancreatic duct. Again, the echoendoscope is
withdrawn and exchanged for a side-viewing duodenoscope. Methy-
lene blue is seen exiting and dening the orice of the minor papilla,
facilitating cannulation.64
When all the above methods have been exhausted and attempted
cannulation has failed, true minor papilla precut techniques can
allow successful cannulation. However, these techniques are best
reserved for experts in therapeutic endoscopy, since severe complica-
83
 SECTION 2 TECHNIQUES
tions can occur. The technique for precutting the minor papilla is
discussed in Chapter 15.
Using the techniques described in this chapter, expert endosco-
pists can achieve successful minor papilla ductography in 90100%
REFERENCES
1. Schwacha H, Allgaier HP, Deibert P, et al. A sphincterotome-
based technique for selective transpapillary common bile duct
cannulation. Gastrointest Endosc. 2000 Sep; 52(3):387391.
2. Cortas GA, Mehta SN, Abraham NS, et al. Selective cannulation of
the common bile duct: a prospective randomized trial comparing
standard catheters with sphincterotomes. Gastrointest Endosc.
1999 Dec; 50(6):775779.
3. Karamanolis G, Katsikani A, Viazis N, et al. A prospective cross-
over study using a sphincterotome and a guidewire to increase
the success rate of common bile duct cannulation. World J
Gastroenterol. 2005 Mar 21; 11(11):16491652.
4. Laasch HU, Tringali A, Wilbraham L, et al. Comparison of standard
and steerable catheters for bile duct cannulation in ERCP.
Endoscopy. 2003 Aug; 35(8):669674.
5. Silverman WB. Tapered-tip, triple-lumen papillotome/cannula
facilitates cannulation yet accepts standard guidewires.
Gastrointest Endosc. 1997 Nov; 46(5):471472.
6. Rossos PG, Kortan P, Haber G. Selective common bile duct
cannulation can be simplied by the use of a standard
papillotome. Gastrointest Endosc. 1993 JanFeb; 39(1):6769.
7. Burdick JS, London A, Thompson DR. Intramural incision
technique. Gastrointest Endosc. 2002 Mar; 55(3):425427.
8. Lella F, Bagnolo F, Colombo E, Bonassi U. A simple way of
avoiding post-ERCP pancreatitis. Gastrointest Endosc. 2004;
59:830834.
9. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic
stent placement reduce the risk of post-ERCP acute pancreatitis?
A meta-analysis of controlled trials. Gastrointest Endosc. 2004 Oct;
60(4):544550.
10. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent
prevent post-ERCP pancreatitis? A prospective randomized study.
Gastrointest Endosc. 2003 Mar; 57(3):291294.
11. Tarnasky PR, Palesch YY, Cunningham JT, et al. Pancreatic stenting
prevents pancreatitis after biliary sphincterotomy in patients with
sphincter of Oddi dysfunction. Gastroenterology. 1998 Dec;
115(6):151824.
12. Goldberg E, Titus M, Haluszka O, et al. Pancreatic-duct stent
placement facilitates difcult common bile duct cannulation.
Gastrointest Endosc. 2005 Oct; 62(4):592596.
13. Rashdan A, Fogel EL, McHenry L Jr, et al. Improved stent
characteristics for prophylaxis of post-ERCP pancreatitis. Clin
Gastroenterol Hepatol. 2004 Apr; 2(4):322329.
14. Maeda S, Hayashi H, Hosokawa O, et al. Prospective
randomized pilot trial of selective biliary cannulation using
pancreatic guide-wire placement. Endoscopy. 2003 Sep;
35(9):721724.
15. Draganov P, Devonshire DA, Cunningham JT. A new technique to
assist in difcult bile duct cannulation at the time of endoscopic
retrograde cholangiopancreatography. JSLS. 2005 AprJune;
9(2):218221.
16. Saad M. Biliary cannulation and pancreatic guide-wire placement.
Endoscopy. 2004 Aug; 36(8):743.
17. Huibregtse K, Katon RM, Tytgat GN. Precut papillotomy via ne-
needle-knife papillotome: a safe and effective technique.
Gastrointest Endosc. 1986; 32:403405.
84
of patients with pancreas divisum without the use of minor papilla
precut and/or EUS. Moreover, deep cannulation of the minor papilla
is achievable in most of these patients as well as individuals without
divisum.
18. Dowsett JF, Polydorou AA, Vaira D, et al. Needle-knife papillotomy:
how safe and how effective?. Gut. 1990; 31:905908.
19. Foutch PG. A prospective assessment of results for needle-knife
papillotomy and standard endoscopic sphincterotomy.
Gastrointest Endosc. 1995; 41:2532.
20. Bruins SW, Schoeman MN, DiSario JA, et al. Needle-knife
sphincterotomy as a precut procedure: a retrospective evaluation
of efcacy and complications. Endoscopy. 1996; 28:334339.
21. Gholson CF, Favrot D. Needle-knife papillotomy in a university
referral practice. Safety and efcacy of a modied technique. J Clin
Gastroenterol. 1996; 23:177180.
22. Kasmin FE, Cohen D, Batra S, et al. Needle-knife sphincterotomy
in a tertiary referral center: efcacy and complications.
Gastrointest Endosc. 1996; 44:4853.
23. Mavrogiannis C, Liatsos C, Romanos A, et al. Needle-knife
stulotomy versus needle-knife precut papillotomy for the
treatment of common bile duct stones. Gastrointest Endosc. 1999;
50(3):334339.
24. Harewood GC, Baron TH. An assessment of the learning curve for
precut biliary sphincterotomy. Am J Gastroenterol. 2002;
97:17081712.
25. Goff JS. Long-term experience with the transpancreatic sphincter
pre-cut approach to biliary sphincterotomy. Gastrointest Endosc.
1999 Nov; 50(5):642645.
26. Catalano MF, Linder JD, Geenen JE. Endoscopic transpancreatic
papillary septotomy for inaccessible obstructed bile ducts:
Comparison with standard pre-cut papillotomy. Gastrointest
Endosc. 2004 Oct; 60(4):557561.
27. Kahaleh M, Tokar J, Mullick T, et al. Prospective evaluation of
pancreatic sphincterotomy as a precut technique for biliary
cannulation. Clin Gastroenterol Hepatol. 2004 Nov; 2(11):971977.
28. Baron TH. Complications of endoscopic biliary sphincterotomy:
the rst cut is the deepest. Clin Gastroenterol Hepatol. 2004 Nov;
2(11):968970.
29. Zoepf T, Zoepf DS, Arnold JC, et al. The relationship between
juxtapapillary duodenal diverticula and disorders of the
biliopancreatic system: analysis of 350 patients. Gastrointest
Endosc. 2001 Jul; 54(1):5661.
30. Naranjo-Chavez J, Schwarz M, Leder G, et al. Ampullary but not
periampullary duodenal diverticula are an etiologic factor for
chronic pancreatitis. Dig Surg. 2000; 17(4):358363.
31. Tham TC, Kelly M. Association of periampullary duodenal
diverticula with bile duct stones and with technical success of
endoscopic retrograde cholangiopancreatography. Endoscopy.
2004 Dec; 36(12):10501053.
32. Rajnakova A, Goh PM, Ngoi SS, et al. ERCP in patients with
periampullary diverticulum. Hepatogastroenterology. 2003
MayJune; 50(51):625628.
33. Scotiniotis I, Ginsberg GG. Endoscopic clip-assisted biliary
cannulation: externalization and xation of the major papilla from
within a duodenal diverticulum using the endoscopic clip xing
device. Gastrointest Endosc. 1999 Sep; 50(3):431436.
34. Kulling D, Haskell E. Double endoscope method to access
intradiverticular papilla. Gastrointest Endosc. 2005 Nov;
62(5):811812.

35. Fogel E, Sherman S, Lehman GA. Increased selective biliary
cannulation rates in the setting of periampullary diverticula: main
pancreatic duct stent placement followed by pre-cut biliary
sphincterotomy. Gastrointest Endosc. 1998; 47:396400.
36. Kumar S, Sherman S, Hawes RH, et al. Success and yield of
second attempt ERCP. Gastrointest Endosc. 1995 May;
41(5):445447.
37. Choudari CP, Sherman S, Fogel EL, et al. Success of ERCP at a
referral center after a previously unsuccessful attempt.
Gastrointest Endosc. 2000 Oct; 52(4):478483.
38. Ramirez FC, Dennert B, Sanowski RA. Success of repeat ERCP
by the same endoscopist. Gastrointest Endosc. 1999 Jan;
49(1):5861.
39. Calvo MM, Bujanda L, Heras I, et al. The rendezvous technique for
the treatment of choledocholithiasis. Gastrointest Endosc. 2001
Oct; 54(4):511513.
40. Petzold V, Rosch T, Born P. Combined endoscopic and
percutaneous transhepatic approach in postsurgical common bile
duct occlusion. Dtsch Med Wochenschr. 2001 Oct 26;
126(43):11972000. German.
41. Monkemuller KE, Linder JD, Fry LC. Modied rendezvous
technique for biliary cannulation. Endoscopy. 2002 Nov;
34(11):936.
42. Dickey W. Parallel cannulation technique at ERCP rendezvous.
Gastrointest Endosc. 2006 Apr; 63(4):686687.
43. Shlansky-Goldberg RD, Ginsberg GG, Cope C. Percutaneous
puncture of the common bile duct as a rendezvous procedure to
cross a difcult biliary obstruction. J Vasc Interv Radiol. 1995 Nov
Dec; 6(6):943946.
44. Lella F, Bagnolo F, Rebuffat C, et al. Use of the laparoscopic-
endoscopic approach, the so-called rendezvous technique, in
cholecystocholedocholithiasis: a valid method in cases with
patient-related risk factors for post-ERCP pancreatitis. Surg Endosc.
2006 Mar; 20(3):419423.
45. Dickey W. Parallel cannulation technique at ERCP rendezvous.
Gastrointest Endosc. 2006 Apr; 63(4):686687.
46. Peterson BT. Biliary rendezvous or solo combined procedure
for therapy of sump syndrome. Gastrointest Endosc. 1996 Feb;
43(2 Pt 1):176177.
47. Kahaleh M, Wang P, Shami VM, et al. EUS-guided transhepatic
cholangiography: report of 6 cases. Gastrointest Endosc. 2005
Feb; 61(2):307313.
48. Kahaleh M, Hernandez AJ, Tokar J, et al. Interventional EUS-guided
cholangiography: evaluation of a technique in evolution.
Gastrointest Endosc. 2006 Jul; 64(1):5259.
49. Lai R, Freeman ML. Endoscopic ultrasound-guided bile duct
access for rendezvous ERCP drainage in the setting of
intradiverticular papilla. Endoscopy. 2005 May; 37(5):487489.
Chapter 8 Cannulation of the Major and Minor Papilla
50. Devereaux BM, Lehman GA, Fein S, et al. Facilitation of pancreatic
duct cannulation using a new synthetic porcine secretin. Am J
Gastroenterol. 2002 Sep; 97(9):22792281.
51. Itoh S, Takada A, Satake H, et al. Diagnostic value of multislice
computed tomography for pancreas divisum: assessment with
oblique coronal reconstruction images. J Comput Assist Tomogr.
2005 JulAug; 29(4):452460.
52. Hellerhoff KJ, Helmberger H 3rd, Rosch T, et al. Dynamic MR
pancreatography after secretin administration: image quality and
diagnostic accuracy. AJR Am J Roentgenol. 2002 Jul;
179(1):121129.
53. Lai R, Freeman ML, Cass OW, et al. Accurate diagnosis of pancreas
divisum by linear-array endoscopic ultrasonography. Endoscopy.
2004 Aug; 36(8):705709.
54. Song MH, Kim MH, Lee SK, et al. Endoscopic minor papilla
interventions in patients without pancreas divisum. Gastrointest
Endosc. 2004 June; 59(7):901905.
55. Devereaux BM, Lehman GA, Fein S, et al. Facilitation of pancreatic
duct cannulation using a new synthetic porcine secretin. Am J
Gastroenterol. 2002 Sep; 97(9):22792281.
56. Devereaux BM, Fein S, Purich E, et al. A new synthetic porcine
secretin for facilitation of cannulation of the dorsal pancreatic
duct at ERCP in patients with pancreas divisum: a multicenter,
randomized, double-blind comparative study. Gastrointest
Endosc. 2003 May; 57(6):643647.
57. Park SH, de Bellis M, McHenry L, et al. Use of methylene blue to
identify the minor papilla or its orice in patients with pancreas
divisum. Gastrointest Endosc. 2003 Mar; 57(3):358363.
58. Ohshima Y, Tsukamoto Y, Naitoh Y, et al. Function of the minor
duodenal papilla in pancreas divisum as determined by
duodenoscopy using indigo carmine dye and a pH sensor. Am J
Gastroenterol. 1994 Dec; 89(12):21882191.
59. Dunham F, Deltenre M, Jeanmart J, et al. Special catheters for
E.R.C.P. Endoscopy. 1981 Mar; 13(2):8185.
60. Schleinitz PF, Katon RM. Blunt tipped needle catheter for
cannulation of the minor papilla. Gastrointest Endosc. 1984 Aug;
30(4):263266.
61. Khalid A, Slivka A. Pancreas Divisum. Curr Treat Options
Gastroenterol. 2001 Oct; 4(5):389399.
62. Wilcox CM, Monkemuller KF. Wire-assisted minor papilla precut
papillotomy. Gastrointest Endosc. 2001 Jul; 54(1):8386.
63. Will U, Meyer F, Manger T, et al. Endoscopic ultrasound-assisted
rendezvous maneuver to achieve pancreatic duct drainage in
obstructive chronic pancreatitis. Endoscopy. 2005 Feb;
37(2):171173.
64. Dewitt J, McHenry L, Fogel E, et al. EUS-guided methylene blue
pancreatography for minor papilla localization after unsuccessful
ERCP. Gastrointest Endosc. 2004 Jan; 59(1):133136.
85
 SECTION 2 TECHNIQUES
Chapter
Access (Precut) Sphincterotomy:
9 Conceptual Philosophy and
Technical Details
Amit Maydeo, Suryaprakash Bhandari, and Hardeep Singh
INTRODUCTION
With the advent of imaging modalities like MRCP and EUS, there
are very few indications left for diagnostic endoscopic retrograde
cholangiopancreatography (ERCP). However, with the develop-
ment of better endoscopic technologies including specialized can-
nulae, hydrophilic wires and endoprostheses, the indications and
possibilities of therapeutic ERCP have dramatically expanded. The
ability to selectively cannulate the bile duct and/or the pancreatic
duct is a prerequisite before being able to perform further endo-
scopic therapy in the biliary and pancreatic system. The anatomy
and the appearance of the ampulla of Vater is however not uniform
and varies in every patient. Ampullae can be at or bulging; exposed
or covered by duodenal folds; distorted or inamed and untouched
or previously tampered with. In fact, the ampulla of Vater is usually
the biggest hurdle in performing a therapeutic ERCP for the begin-
ner endoscopist.
The fact remains that in spite of all the advances in technology
and technique, there continue to be some instances when even the
most experienced endoscopist will be unable to cannulate the desired
duct. It is when the standard techniques of cannulation fail that one
needs to resort to specialized techniques of ductal cannulation.
Precut, or access sphincterotomy as it is popularly called, is primarily
designed for gaining access into the biliary or pancreatic duct when
the conventional methods of selective cannulation fail.
In this chapter we will focus predominantly on the conceptual
philosophy behind the techniques of precut accessotomy and
describe in detail the different techniques and tools which are com-
monly used. We will also review the existing published literature
regarding the usefulness and complications of these techniques
when used in different clinical situations.
INDICATION FOR PRECUT
As a general rule, precut accessotomy should be performed
when standard techniques of selective cannulation fail. However,
the denition of cannulation failure is itself unclear and the
decision to resort to a precut usually depends on a number of
factors including the indication of the procedure, anatomy of the
ampulla/ duodenum and most importantly the training, comfort
and threshold of the endoscopist. Though some of these factors
are subjective, most endoscopists agree that precut should not
be done for purely diagnostic purposes and until a sincere attempt
to cannulate has been made by an experienced and skilled
endoscopist.
The use of precut accessotomy varies from center to center
(Table 9.1). Binmoeller et al. described employing a precut in 38%
of 327 patients,1 while Foutch reports using precut in only 11% of
456 patients.2 One perspective is that trauma to the ampulla and
pancreatic duct should be minimized in order to avoid complica-
tions, and therefore precut should be considered early in the proce-
dure. Others prefer to leave precut as a last resort or rescue
method, to be employed only when multiple exhaustive attempts at
standard cannulation have failed.
There seems to be a general agreement however that
multiple unsuccessful attempts at cannulation tend to increase
complications such as pancreatitis and, therefore, it is better to
perform a precut accessotomy early before damaging the ampulla/
ductal epithelium further. Many experienced endoscopists there-
fore resort to a precut if a few attempts (35) of selective cannula-
tion fail.
There are some clinical situations when precut is the pre-
ferred method of initial cannulation. In patients with an impacted
gallstone at the ampulla of Vater, it may be easier to precut over
the stone, rather than attempting to cannulate around the stone
through the papillary orice. In patients with a previous Billroth
II operation it may be better to cannulate the bile duct and place a
stent over which a precut of the ampulla is done. Similarly stenting
followed by precut may be the preferred method of cannulating and
cutting the minor papilla in symptomatic patients with pancreas
divisum.
We favor utilizing precut accessotomy early in the course of the
procedure. Typically, we allow 35 targeted attempts at wire-guided
cannulation using a smooth-tipped, short-nosed, short-wire sphinc-
terotome (Clever cut, Olympus Optical Co. Japan) and a 0.032-inch,
J-shaped glide wire (Terumo Corp., Japan). We avoid unnecessary
probing and traumatization of the ampulla and the ductal epithe-
lium. If this fails to provide access in 35 attempts and the ampullary
anatomy is suitable for a precut, we proceed with the accessotomy
technique.
TECHNIQUE OF PRECUT ACCESSOTOMY
Conceptual philosophy of the technique
Siegel rst reported the technique of precut papillotomy in 1980,
using a standard sphincterotome inserted into the papillary orice
to expose the bile duct.3 Since then many expert endoscopists of the
world have utilized the precut using either the modied Erlangen
sphincterotome or the needle-knife. Different techniques have been
87
 SECTION 2 TECHNIQUES

described using these instruments with the cut being made from Once the ampullary anatomy is clearly imagined, the cut should
below upwards or from above downwards. be made in a layer by layer manner starting from the ampullary
Regardless of the type of instrument used for the precut, the basic orice. For bile duct entry, the cut is made slowly from the orice
principle is to unroof the ampulla of Vater by cutting open its super- upwards and somewhat towards the left. Whichever instrument is
cial layers and thereby exposing the biliary epithelium. It is manda- used for the cut, each stroke of the cut has to be supercial and
tory that the duodenal peristalsis is suppressed with an antiperistaltic should be a clean incision using a pure cut or blended current. It is
drug and the patient is well sedated for a clear, quiet and accurate imperative that all the strokes of the cut are made in one direction
visualization of the ampulla of Vater. This period of proper and only without changing the direction with each stroke. Similarly it is
careful inspection of the ampulla is of prime importance to plan the mandatory that the depth of each stroke has to be supercial
precut accessotomy rather than immediately starting the cut. If the and not deep. Once the supercial layer of the mucosa is cut, the
ampulla is covered with duodenal folds, one needs to lift the over- bulge of the bile duct is usually evident with a whitish covering.
hanging folds and inspect the ampulla properly before beginning This needs to be further cut supercially in the same direction until
the cut. During this inspection, it is necessary to appreciate the ori- the Salmon pink epithelium of the bile duct is exposed. The bile
entation and direction of the bulging portion of the ampulla which duct can be carefully probed using a delicate instrument like an
usually indicates the bile duct direction below the surface. ERCP cannula or a soft-tipped sphincterotome along with a glide
The ampulla of Vater is covered by the duodenal mucosa on its wire (Fig. 9.2).
outermost part with the submucosa below it and then the muscular During the precut, some oozing may take place. This is likely to
layer covering the bile duct and pancreatic ducts (Fig. 9.1). There- obscure vision and thereby prevent further progress of the cut. In
fore, though the biliary and pancreatic ducts open nally at the this case, it is a good idea to spray diluted epinephrine solution
ampullary orice, the bile duct actually courses upwards and towards (1 : 20000) on the cut surface of the ampulla until the oozing stops.
the left just below the ampullary roof, whereas the pancreatic duct If this is not enough to control the bleeding, direct coagulation of
courses towards the right and in a more straight direction. There- the bleeding spot can be done after careful inspection of the bleed-
fore, for entering the biliary or pancreatic ducts, unroong of the ing area.
supercial ampullary layers needs to be done in the directions of the If, in spite of an adequately made cut in the proper direction, the
required ducts. It is mandatory therefore to imagine the ampullary bile duct bulge is not seen, one should carefully inspect to see if
and ductal anatomy precisely before beginning the precut. the cut has been made on either side of the bulge rather than on the
bulge itself. In this case it may be a good idea to identify the pan-
creatic orice rst and then look for the biliary bulge above and on
the left of the pancreatic orice. In rare instances, the biliary orice
Pre-cut % of is located on the right of the pancreatic orice instead of the left.
Author Year Technique (n) Pre-cut
THE TECHNIQUE OF NEEDLE-KNIFE
Akashi 2004 Transpancreatic 172 10
PRECUT ACCESSOTOMY
Goff 1999 Transpancreatic 51 26
Rabenstein 1997 Needle-knife 694 33 The needle-knife consists of a catheter with a straight retractable
Binmoeller 1996 Short nosed 123 38 diathermy wire at the end. When current is applied, the needle can
Erlangen type be used to make a cut through the papillary mucosa to gain access
sphincterotome to the bile duct. There are a variety of different approaches to precut
Foutch 1995 Needle-knife 52 11 accessotomy when using a needle-knife. The most common method
Huibregtse 1986 Needle-knife 190 19 is the free-hand needle-knife technique. With this technique, the
needle-knife is gently inserted into the papillary orice and the inci-
Table 9.1 Percentage of precut accessotomy utilizing different sion is made superiorly toward the 11 oclock position. This upward
techniques motion is directed through manipulation of the elevator, large wheel,

A B Fig. 9.1 The ampulla of Vater, conceptual

anatomy of this complex organ. The papil-
lary folds A are like the peels of an onion
B. The CBD and PD orices lie hidden in the
core of the onion. The CBD direction is left
and upward, the PD is right and straight.
Cannula getting stuck in the folds is the com-
monest problem.

88
 Chapter 9 Access (Precut) Sphincterotomy: Conceptual Philosophy and Technical Details

A B C

D E F

Fig. 9.2 Serial images showing layer by layer deroong of the papilla using a needle-knife. A The rst mucosal layer being cut with
needle-knife. Note the direction of the cut is towards 11 oclock position. B The second submucosal layer being cut with the needle-knife.
C The third muscle layer being cut with the needle-knife. DF Exposed Salmon pink epithelium of the bile duct with free owing bile.

or by applying upward torque on the shaft of the endoscope. Usually
the incision should be approximately 5 mm in length, but the exact
size of the cut would depend on the size and anatomy of the papilla.
Huibregtse advises taking a few practice swings in order to map
out the path of the cut, prior to performing the precut.4 Current is
applied only while the wire is moving within the tissue, so as to avoid
coagulation and edema of the papilla. Blended current was used
most often in large case series on needle-knife papillotomy.2,4,5 One
small trial by Baron et al. showed no difference in complication rates
in patients undergoing needle-knife papillotomy with blended versus
pure cutting current.6 After the initial cut is made, further cuts are
made in the same direction, in an effort to slowly dissect down
toward the bile duct. Once the bile duct orice is exposed, the needle
is withdrawn and the needle-knife catheter can be used to carefully
enter the duct. Alternatively, a tapered tip catheter or a exible
guidewire can be used to enter the duct. When the duct has been
successfully cannulated, therapeutic maneuvers can be performed
in the same manner as with normal cannulation.
There is much data to support the use of needle-knife papillotomy
as a safe and effective method for selective cannulation of the bile
duct. Rabenstein et al. published a large retrospective review of the
benets and risks of needle-knife papillotomy in 694 patients.5 In
this study, the author achieved a successful initial cannulation in
70% of patients, and with a second attempt, was able to cannulate
the duct in 85% of patients. The most common complications were
bleeding and pancreatitis. The overall complication rate was the
same in the group that underwent needle-knife papillotomy com-
pared to those that underwent conventional sphincterotomy (7%).
Other studies have yielded similar results.2,6,7 However, whether
needle-knife papillotomy confers an increased risk of pancreatitis
remains controversial. A meta-analysis performed by Masci et al.
found precut papillotomy to be associated with an increased risk of
post-ERCP pancreatitis (Odds Ratio 2.71).8 It is likely that the dis-
crepancy in these results indicates that the risks of needle-knife
papillotomy depend on proper patient selection and on the experi-
ence of the endoscopist.
Another way of gaining access using a needle-knife is by perform-
ing a suprapapillary stulotomy. Typically this is done by inserting
the needle a few millimeters above the papillary orice in the 11
oclock orientation, and cutting in a downward motion toward the
orice. The motion is again repeated slowly in an effort to extend
the depth of the cut and eventually expose the bile duct. A variation
of this technique involves puncturing the papilla above the orice
and cutting in an upward direction. Once the duct is exposed, the
duct is gently entered in the same manner. The study by Baron
et al. and a large study by OConnor et al. of over 500 patients
demonstrated suprapapillary stulotomy to be a safe and effective
maneuver for gaining access to the CBD.6,9 However, one limitation
of this method is that the created incision cannot be as large as a
conventional precut, which may make therapeutic maneuvers more
difcult. Mavrogiannis et al. found that although success rates of
selective cannulation were comparable between patients undergoing
suprapapillary stulotomy versus needle-knife papillotomy, stone
extraction was more difcult in the stulotomy group.10
A nal method employs inserting a stent into the pancreatic duct
and then using the needle-knife to cut over the stent to perform a
papillotomy or stulotomy. In this technique, the stent has two roles.
First, if the pancreatic duct has been repeatedly injected and manip-
ulated while attempting cannulation, placement of a temporary stent
may help prevent post-ERCP pancreatitis.11 Secondly, a stent placed

89
 SECTION 2 TECHNIQUES

in the pancreatic duct may serve as a guide for performing the Eventual % of Precut
precut. Although we do not routinely place pancreatic stents prior Author Year Technique success (%) complications
to performing a precut, there is some rationale for this technique.
Akashi 2004 Transpancreatic 95 10
Goff 1999 Transpancreatic 98 2
TRACTION PAPILLOTOME TECHNIQUES Rabenstein 1997 Needle-knife 85 7
Binmoeller 1996 Papillary roof 100 5
Another approach to performing a precut employs the use of a trac- incision
tion papillotome. This technique was pioneered by Binmoeller and Foutch 1995 Needle-knife 90 6
Soehendra, who describe performing a papillary roof incision Huibregtse 1986 Needle-knife 84 3
using an Erlangen-type precut papillotome in order to achieve selec-
tive cannulation of the bile duct.1 Proponents of this technique claim Table 9.2 Percentage of cannulation success and complication
that traction papillotomes allow for a more controlled incision rates with different methods of precut
compared to a needle-knife. This type of papillotome has a one cm
monolament diathermy wire and a short tip that extends beyond
the cutting wire. The leading tip of this specialized papillotome is special circumstances and by an experienced endoscopist. Moreover,
inserted into the papillary orice. The instrument is then swung in a pancreatic duct stent should be inserted post-sphincterotomy to
the 11 oclock direction as current is delivered, and the incision is minimize procedural pancreatitis.
typically 5 to 10 mm in length. Although the authors originally
described using pure cutting current at 4050 W, blended current CONCLUSION
can also be used. As with a needle-knife papillotome, the incision is
repeated in the same direction to gradually and carefully llet open There is extensive data indicating that precut techniques can con-
the roof of the papilla and expose the biliary orice. Once the duct tribute to successful selective cannulation of the bile duct, but there
is identied, the orice is probed with the tip of the papillotome until are some associated risks (Table 9.2). Typical risks ascribed to precut
cannulation is achieved. Although we are not aware of any studies accessotomy include pancreatitis, perforation and bleeding. The data
comparing traction papillotomy to needle-knife, success and compli- regarding the frequency of these complications is variable, and there
cation rates appear comparable to those seen in large studies of remains some controversy whether these complications are the
needle-knife papillotomy.1,2,5,6 result of the precut, or whether they are the result of multiple failed
A variation of this method is trans-pancreatic precut sphincterot- attempts at standard cannulation. Also, rates of complications may
omy, originally described by Goff in 1995.12 This technique involves vary with the technique of precut used. Nevertheless, there is agree-
inserting the traction papillotome into the pancreatic duct. Next the ment that precut accessotomy is a technique that should only be
papillotome is used to make an incision toward the common bile utilized in the proper clinical situation. It is only after sincere and
duct, cutting the inter-ductal septum and unroong the papilla in careful attempts at standard cannulation have failed, that precut
the process. A retrospective study by Goff including 51 patients accessotomy should be considered. Furthermore, it should be
found the rate of success and complications to be comparable to emphasized that precut should not be a substitute for sound endo-
conventional methods.13 Unfortunately, a larger study by Akashi et al. scopic technique. Endoscopists performing this procedure should
published in 2004 found a higher incidence of complications and not only have extensive experience in performing the various tech-
lower success rate with the trans-pancreatic technique.14 It is our niques of precut, but should also be well-versed in how to manage
opinion, that if this maneuver is used, it should only be done in possible endoscopic complications.

REFERENCES
1. Binmoeller KF et al. Papillary roof incision using the Erlangen-
type pre-cut papillotome to achieve selective bile duct
cannulation. Gastrointestinal Endoscopy, 1996; 44:689695.
2. Foutch PG. A prospective assessment of results of needle-knife
papillotomy and standard endoscopic sphincterotomy.
gastrointestinal endoscopy, 1995; 41:2532.
3. Siegel JH. Precut papillotomy: a method to improve success of
ERCP and papillotomy. Endoscopy, 1980; 12:130133.
4. Larkin CJ, Huibregtse K. Precut sphincterotomy: indications,
pitfalls, and complications. Current Gastroenterology Reports,
2001; 3:147153.
5. Rabenstein T et al. Benets and risks of needle-knife papillotomy.
Gastrointestinal Endoscopy, 1997; 46:207211.
6. Abu-Hamda EM, Baron TH, et al. A retrospective comparison of
outcomes using three different precut needle-knife techniques for
biliary cannulation. J of Clinical Gastroenterology, 2005;
39:717721.
7. Huibregtse K et al. Precut papillotomy via ne-needle-knife
papillotome: a safe and effective technique. Gastrointestinal
Endoscopy. 1986; 32:403405.
8. Masci E et al. Risk factors for pancreatitis following endoscopic
retrograde cholangio-pancreatography: a meta-analysis.
Endoscopy, 2003; 35:830834.
9. OConnor HJ et al. Suprapapillary stulosphincterotomy at ERCP: a
prospective study. Endoscopy, 1997; 29:266270.
10. Mavrogiannis C et al. Needle-knife stulotomy versus needle-
knife precut sphincterotomy for the treatment of commmon bile
duct stones. Gastrointestinal Endoscopy. 1999; 50:334339.
11. Freeman ML, Guda NM. Prevention of post-ERCP pancreatitis: a
comprehensive review. Gastrointestinal Endoscopy, 2004;
59:845864.
12. Goff JS. Common bile duct pre-cut sphincterotomy:
transpancreatic sphincter approach. Gastrointestinal Endoscopy
1995; 41:502505.
13. Goff JS. Long-term experience with the trans-pancreatic sphincter
pre-cut approach to biliary sphincterotomy. Gastrointestinal
Endoscopy 1999; 50:642645.
14. Akashi R et al. Pancreatic sphincter precutting to gain selective
access to the common bile duct: a series of 172 patients.
Endoscopy, 2004; 36:405410.

90
 SECTION 2 TECHNIQUES
Chapter
Sphincter of Oddi Manometry
10 Nalini M Guda and Joseph E Geenen
INTRODUCTION AND SCIENTIFIC BASIS
Sphincter of Oddi was rst described in 1887 by Ruggero Oddi as a
distinct anatomical and physiological entity. The sphincter is com-
posed of smooth muscles encircling the distal end of the common
bile duct and the main (ventral) pancreatic duct. The muscles are
arranged in both a circular and a gure of eight conguration and
are about 410 mm in length. The main function of the sphincter is
to regulate bile and exocrine pancreatic secretions and to prevent
reux of duodenal contents into the ducts. In addition to a basal
pressure the sphincter of Oddi has a phasic contractile activity. The
basal pressure is the predominant mechanism which regulates the
release of biliary and pancreatic secretions into the duodenum.
Phasic contractions are thought to be important in preventing the
duodenal reux into the ducts. Phasic contractions are related to the
migratory motor complex of the duodenum. Neural regulation does
not appear to play an important role since even in post-transplant
patients the sphincter of Oddi function appears preserved. Chem-
ical regulation is important. Cholecystokinin and secretin cause
sphincter relaxation. Patients with abnormalities in the contraction
of the sphincter of Oddi are labeled as having sphincter of Oddi
dysfunction. It is also known as papillitis, papillary stenosis, post
cholecystectomy syndrome, papillary spasm, biliary dyskinesia,
tachyoddia etc. Detailed clinical presentation of those with sphin-
cter of Oddi dysfunction is described elsewhere in this book
(Chapter 34).
Sphincter of Oddi manometry provides data regarding the basal
pressure of the sphincter, the frequency of contractions and the
amplitude of contractions occurring at the sphincter. Manometry of
the sphincter of Oddi is performed at the time of ERCP for those
with a clinical suspicion of biliary or pancreatic type of pain thought
to be related to either dyskinesia or stenosis of the biliary and/or
pancreatic sphincters. Sustained basal pressures (>40 mm of Hg) on
manometry are thought to be due to sphincter of Oddi stenosis and
sphincterotomy in these situations has been shown to result in relief
of symptoms.1,2 Sphincter of Oddi Manometry (SOM), though inva-
sive, is considered the gold standard for measurement of biliary
motility.3 Though there is some concern whether a few minutes of
pressure observations would reect the pressure dynamics over 24
hours, currently this is the accepted method of measurement. There
are other non-invasive means of detecting the sphincter of Oddi
dysfunction: Morphine-Prostigmin Provocative Test (Nardi Test),
Ultrasonographic Assessment of Extrahepatic Bile Duct and Main
Pancreatic Duct Diameter After Secretory Stimulation and Quantita-
tive Hepatobiliary Scintigraphy. There are also limited data regard-
ing the use of secretin stimulated MRCP and EUS in the evaluation
of SO dysfunction. In a pilot study of 18 patients with idiopathic
acute recurrent pancreatitis there appeared to be a high concordance
rate between ERCP and secretin MRCP in diagnosis of SO dysfunc-
tion.4 In a similar study secretin was injected and pancreatic duct
diameter was subsequently measured by ultrasound in patients with
idiopathic acute recurrent pancreatitis. Those with dilated duct
diameter beyond 20 minutes were suspected to have sphincter of
oddi dysfunction. All patients underwent SO manometry in 37
days. Using SO manometry as the gold standard the sensitivity and
specicity of secretin stimulated ultrasound in diagnosis of SO
dysfunction was 88% and 82%.5 These techniques have not been
evaluated in larger series of patients and in those without pancreati-
tis. At the current time none of the non-invasive methods have
superior operating characteristics compared to the conventional
SOM (Table 10.1).6
TECHNIQUE
Patient preparation
As with all endoscopic procedures patients should be fasting for at
least 6 hours. Drugs that either stimulate or relax the sphincter of
Oddi should be avoided at least 12 hours prior to the scheduled
manometry. Drugs that are thought to stimulate the sphincter
include narcotic analgesics and other cholinergic agents. Drugs that
relax the sphincter include nitrates, glucagon, calcium channel
blockers and other anti-cholinergic agents. Midazolam in doses
>2 mg has been shown to reduce the basal sphincter pressure.7
Benzodiazepines and Meperidine (Demerol) when given at a dose
of <1 mg/kg body weight do not appear to alter the sphincter pres-
sures and are considered acceptable for sedation.8 Propofol sedation
has been shown to have no effect on the sphincter of Oddi basal
pressures.9 Similarly use of general anesthesia for sedation has not
been shown to have any effect on the sphincter of Oddi.10 Patients
could be sedated either by conscious sedation or by anesthesia
depending on the patients tolerance. There has been an increasing
trend to perform these procedures either with propofol or general
anesthesia. The patient is placed usually in a prone position as with
any standard ERCP procedures.
Preparation of equipment
The hydraulic capillary infusion pump should be lled with 500 cc
of sterile water and sealed tightly. The valve on the nitrogen tank
attached to the perfusion system should be opened to a pressure of
100 PSI. The low pressure valve on the machine should be opened
to 7 PSI. Gently tap the transducer dome to remove any air. Initially
the transducers are attached to the connecting catheters and the
connecting catheters are in turn connected to the triple lumen cath-
eters. Turn on the perfusion valve. Prime the triple lumen and
the duodenal catheters (if used). It is recommended that the perfu-
sion system should be switched on for at least 15 minutes before
91
 SECTION 2 TECHNIQUES
the procedure. Care should be taken to avoid any air bubbles (Fig.
10.1).
The duodenoscope is introduced into the duodenum as with
any standard ERCP procedure. The duct of interest is initially can-
nulated with a catheter and guidewire. Catheters with 5 Fr diameter
are generally used. A triple lumen sphincterotome would be useful
since ease of initial cannulation with a sphincterotome has been
demonstrated and moreover if sphincterotomy needs to be done it
would result in cost savings and reduce need for additional equip-
ment. After selective cannulation one can check for the location of
the catheter by gently aspirating. Presence of bilious material con-
rms presence in the bile duct (Fig. 10.2) and a clear aspirate is
suggestive of pancreatic duct cannulation (Fig. 10.3). If needed,
contrast could be injected into the duct for better visualization of the
anatomy and to rule out any other structural problem. Injection of
contrast into the bile duct prior to sphincter of Oddi manometry
has been shown to increase the mean pressure but not the basal
sphincter pressure.11 Such evaluation has not been performed in the
pancreatic duct. It is suggested that, prior to manometry, if a chol-
angiogram or pancreatogram were to be routinely obtained, it might
obviate manometry if another etiology or anatomic abnormality was
seen which could explain a patients symptoms. We routinely inject
contrast material into the bile or pancreatic ducts to evaluate the
biliary and pancreatic duct anatomy and to rule out any other struc-
Diagnostic modality Yes Maybe No
ProvocativeNardi test x communication between the endoscopist and the technician/nurse
Pain with contrast injection x assisting with manometry. Information regarding the position of
CBD dilation/delayed drainage x the catheter, number of black marks visible, duodenal contractions
LFT abnormalities x
DISIDA scan x
US fatty meal x
SO manometry x technician/nurse should communicate the average baseline pres-
Table 10.1 Tests used in evaluating SOD
DISIDA: diisopropyl iminodiacetic acid (nuclear medicine scan).
92
tural causes including strictures, stones etc., which could explain the
symptoms and obviate manometry. Moreover injection of contrast
also helps in passing the guidewire deep into the duct without
passing the wire into side branches and potentially causing duct
perforation in case of pancreatic duct cannulation. Duct congura-
tions including ansa pancreaticus will be seen easily if initial con-
trast injection is made prior to advancing the wire (Figs 10.4 and
10.5).
After injection of contrast into the duct a 0.018 inch guidewire is
passed deep into the duct of interest and the cannula is exchanged
if one is to perform wire-guided manometry. Over the guidewire a
manometry catheter (monorail) is introduced deep into the biliary
or pancreatic ducts. We recommend introducing the catheter deep
into the duct and trying not to touch any of the walls of the duct
since this could result in false results. Pressure recordings are
usually obtained within the duct as prior studies have shown that
those with SO dysfunction have higher intraductal pressures. Prior
to introducing the manometry catheter care should be taken to
obtain a duodenal pressure recording which is usually set as zero
pressure. The recording should be equal in all channels and estab-
lishes a baseline value. Historically an extra catheter was attached to
the duodenoscope shaft and served to provide continuous intralumi-
nal duodenal pressure. The motility catheter itself should be with-
drawn carefully 12 mm increments while pressures are measured.
It is recommended to obtain a reading at every black mark while
withdrawing the catheter and to record pressures at each station for
at least two minutes. Phasic waves should be recorded separately
from the basal pressures. It is very important to have a continuous
and patient activity should be communicated by the endoscopist
to the technician/nurse performing the recordings. Similarly the
sure, phasic waves and any interference in the recording to the
endoscopist. It is a team effort and communication is the key to
accurate manometry (Figs 10.6, 10.7 and 10.8 demonstrate normal
Fig. 10.1 Photograph of Arndorfer Perfu-
sion Pump.
 Chapter 10 Sphincter of Oddi Manometry

Fig. 10.4 Cholangiogram prior to performing SO manometry.

Fig. 10.2 Photograph of catheter aspirating bile.

Fig. 10.5 Pancreatogram prior to performing SO manometry.

Fig. 10.3 Photograph showing clear aspirate from the duct sug-
gesting pancreatic duct.

duodenal pressure, elevated biliary sphincter pressure and phasic

contractions).
It is recommended to study pressures in both biliary and pancre-
atic ducts. Data from several studies have shown that often times
the pressures are abnormal only in one segment (sphincter choledo-
chus or pancreaticus) of the sphincter of Oddi in 3565% of the
patients.6 Data also suggest that those with pancreatitis often have
abnormal pancreatic sphincter basal pressures while those with
biliary type of pain and abnormal liver functions have higher basal
pressures of the bile duct sphincter.12 It is recommended that the
manometric abnormality be seen for at least 30 seconds and should
be seen at least in two or more pull throughs. If, however, the basal
pressures are clearly normal or abnormal, one can limit to a single
pull through. The interobserver variability in manometric readings
is good. There is controversy if the measurement of the average of
the basal measurements from the three ports in the standard cathe-
ter and the two ports in the modied catheter should be used to
make a determination of pressure. We perform a station pull through
Fig. 10.6 Manometry tracing showing the basal duodenal
at least two times. We average the recordings and prefer to see ele- pressure.
vated pressures at more than one station and in both the pull through
maneuvers.

93
 SECTION 2 TECHNIQUES
Fig. 10.7 Manometry tracing showing elevated biliary pressure.
2 mm 2mm
1 mm 20mm
0.018*
guide wire
Parameter Abnormal values
Basal pressure >40 mm of Hg
Phasic wave amplitude >350 mm of Hg
Phasic wave frequency >8/min
Phasic wave retrograde propogation >50% total
Table 10.2 Abnormalities in SO Manometry
Interpretation of the Manometry recordings: Before interpreta-
tion of readings is done, care should be taken to establish a basic
duodenal pressure recording. Typically this is the average of the
three recordings when the triple lumen catheter is placed freely in
the duodenum through the duodenoscope prior to cannulation. The
elevator should be in the down position and the catheter should not
touch any duodenal wall to avoid any errors. (Table 10.2)
We currently use a triple lumen Arndorfer pneumohydraulic
capillary perfusion system (Arndorfer Medical Specialties, Green-
dale, Wisconsin, USA) (Fig. 10.9). The Arndorfer catheter is per-
fused at 0.25 ml/minute with distilled water. Efcacy of perfusion
with physiological solutions has not been established. The perfusion
catheter has three side holes and pressure should be recorded at all
the three side holes.13 Sacricing one of the ports to provide for
aspiration has been shown to reduce the incidence of pancreatitis
when performing manometry of the pancreatic duct but not the bile
duct.14,15 Aspiration cannot be done with wire-guided mano. Perfu-
sion at a lower rate could accurately measure the basal sphincter but
the accuracy of phasic waves is unreliable.
Once pressure measurements have been made, one has to decide
whether to do a sphincterotomy. Techniques of sphincterotomy are
94
Fig. 10.8 Manometry tracing showing phasic contractions.
Fig. 10.9 Schematic representation of the Arn-
dorfer triple lumen manometry catheter.
described elsewhere in this book (see Chapters 12 and 14). For
sphincterotomy one must have deep cannulation of the desired duct.
Sphincterotomy is performed by a traction sphincterotome over a
guidewire. The cutting wire should track up to the middle of the
papilla and in biliary sphincterotomy should be oriented between
the 10 oclock and 12 oclock position. Automated electrical genera-
tors delivering pulse current reduce excessive rapid cutting zipper
effect. There is debate over the superiority of the type of current in
prevention of bleeding and post-ERCP pancreatitis. In a meta-analy-
sis it was shown that pure cut current was associated with a slightly
increased risk of immediate post-sphincterotomy bleeding; however,
there were no signicant differences in the rates of delayed bleeding
or pancreatitis.16 We recommend using pure cutting current for
pancreatic sphincterotomy and blended current for biliary sphinc-
terotomy using standard electrical generators.
Irrespective of whether the pressure is elevated or not or whether
biliary and/or pancreatic sphincterotomy is performed, it is a stan-
dard practice now to place a stent in the pancreatic duct to reduce
the risk of post-ERCP pancreatitis post-manometry (Fig. 10.10). This
has been documented well in various studies.17,18 It is not clear at
this time whether a small caliber 3 Fr stent is better than a larger
bore stent. Preliminary data indicate that a modied 5 Fr straight
stent with the inner ange removed is associated with lower rates of
pancreatitis compared to the 3 Fr pigtail stents.19 Insertion of 3 Fr
pigtail stents is a little more difcult and one can use only a wire of
0.018 inch diameter.
Catheters with microtransducers at the tips of the catheters are
available and these can record real-time data when cannulation is
achieved. Since there is no perfusion involved with these systems
they may reduce the risk of post-ERCP pancreatitis. These catheters

Fig. 10.10 Radiograph of the 3 Fr pancreatic stent to prevent post-
ERCP pancreatitis.
Fig. 10.11 Picture of the new manometry catheter with Dent
sleeve.
are stiffer than the regular manometry catheters and are difcult to
cannulate. Its usage is not widespread and data are limited.20
Sleeve catheters were recently developed. They have the advan-
tage of perfusion of 0.04 ml/min and the uid is collected back in
the sleeve. The sleeve also helps stabilize the catheter in the sphinc-
ter zone without touching the side walls of the duct and providing
erroneous values. This is not yet commercially available and clinical
studies are in progress evaluating the efcacy of these catheters21
(Fig. 10.11).
Indications for sphincter of Oddi manometry
Pain if suspected to be pancreatobiliary in nature with or without
liver function abnormality (GeenenHogan classication)
Idiopathic recurrent pancreatitis
EQUIPMENT
The following is the standard equipment which is required for
manometry.
1. Standard diagnostic or therapeutic duodenoscope
2. Hydraulic capillary Infusion system
3. Manometry catheter
4. Recording system: Dynagraph/computerized/solid state system
Catheters
We prefer the Arndorfer catheter since it is a validated manometry
catheter (Arndorfer Medical Specialties, Greendale, Wisconsin,
USA). The pressure recordings can be done with addition of soft-
Chapter 10 Sphincter of Oddi Manometry
ware to available esophageal manometry equipment. We currently
use PolygramNet (Medtronic, Minneapolis, MN) to record and
assess sphincter pressures. A manometry catheter with one port
sacriced to aid suction of the perfused saline is also available
(Lehman SO Manometry catheter, Cook Endoscopy, Winston-Salem,
NC). The catheter is 5 Fr in diameter and has a 5 mm or 1.5 cm tip.
Both catheters are relatively easy to use for cannulation. While an
Arndorfer catheter is an over-the-wire catheter, one could use the
Lehman catheter without the wire or alternatively, it could be intro-
duced over the wire. We use the Arndorfer infusion pump (Arndor-
fer Medical Specialties, Greendale, Wisconsin, USA).
COMPLICATIONS
As with any ERCP, pancreatitis remains the foremost complication
of pancreaticobiliary manometry (see also Chapter 6). It is the patient
characteristics that predispose these individuals to increased fre-
quency of post-ERCP pancreatitis. Manometry by itself does not pose
any greater risk.22 Techniques that possibly reduce the risk of pan-
creatitis include initial injection of contrast into the pancreatic duct
so that when the wire is introduced into the main duct repeated
manipulations into side branches and duct disruption is avoided.
One needs to perform a quick station pull through. There is concern
of over-perfusing with saline for manometry and hence an aspiration
catheter is thought to be benecial for pancreatic manometry because
of the ability to continuously aspirate saline.14 If wire-guided manom-
etry is done rst in the pancreatic duct, it is probably benecial to
leave the wire in the pancreatic duct and cannulate the bile duct
alongside of the wire, to do biliary manometry. This will make
sphincterotomy/stent placement in the pancreatic duct easier after
biliary manometry. The wire might facilitate drainage of the pancre-
atic duct as well by keeping the PD opening patent. Care should be
taken with pancreatic duct sphincterotomy. Unlike biliary sphincter-
otomy the length of sphincterotomy is relatively small, approximat-
ing to 56 mm.
It is now recommended to place a small caliber pancreatic stent
in most if not all patients with suspected sphincter of Oddi dysfunc-
tion who undergo pancreatic instrumentation of any kind, including
those with normal pancreatic manometry.18,23 Three, 4, or 5 French
stents have been used for this purpose. If a 3 Fr pigtail stent is placed
one has to use 0.018 wire in the pancreatic duct. Larger diameter
wires will not allow for deployment of 3 Fr stents. Currently the
pigtail end of the catheter is not marked. We recommend using a
permanent marker to make a circumferential mark at the pigtail end
of the stent so that when the stent is out of the scope one will know
the end of the length of the stent to be placed in the pancreatic duct
rather than deploying the entire stent in the pancreatic duct. Once
the marking is visible, the stent is pushed out into the duodenum
by lowering the elevator and moving the up/down wheel to the down
position. Once the entire stent is out of the elevator of the scope, the
inner wire is pulled out while keeping the pusher catheter in place.
Care should be taken not to raise the elevator during this operation
(tapered tip ERCP catheter) as there is a risk of pushing the pigtail
portion of the stent into the pancreatic duct. If one is not comfortable
and familiar with 3 Fr pigtail stents we recommend using a straight
stent with inner ange removed so that the stent would migrate
distally into the duodenum spontaneously similar to the unanged
3 Fr stent.
Bleeding both immediate and delayed is another major complica-
tion of sphincterotomy and is addressed in the chapter of sphincter-
95
 SECTION 2 TECHNIQUES
otomy. If bleeding is seen, one could inject dilute epinephrine or
sometimes even contrast material. A submucosal injection is
required for tamponade effect. This could be done with the cannulat-
ing catheter itself and an injection with an injection needle catheter
might not be necessary. In difcult situations hemoclips have been
used. Bleeding is usually seen towards the apex of the sphincterot-
omy. In rare instances of continued major bleeding, octreotide infu-
sion to reduce the splanchnic circulation might be benecial. The
risk of bleeding is especially high if one extends an existing sphinc-
terotomy or in the presence of a periampullary diverticulum, and
caution should be exercised.
Papillary stenosis following sphincterotomy is a delayed compli-
cation more commonly seen with pancreatic sphincterotomy.
REFERENCES
1. Geenen JE, Hogan WJ, Dodds WJ, et al. The efcacy of endoscopic
sphincterotomy after cholecystectomy in patients with sphincter-
of-Oddi dysfunction. N Engl J Med 1989; 320(2):8287.
2. Toouli J, Roberts-Thomson IC, Kellow J, et al. Manometry based
randomised trial of endoscopic sphincterotomy for sphincter of
Oddi dysfunction. Gut 2000; 46(1):98102.
3. Lans JL, Parikh NP, Geenen JE. Application of sphincter of Oddi
manometry in routine clinical investigations. Endoscopy 1991;
23(3):139143.
4. Mariani A, Curioni S, Zanello A, et al. Secretin MRCP and
endoscopic pancreatic manometry in the evaluation of sphincter
of Oddi function: a comparative pilot study in patients with
idiopathic recurrent pancreatitis. Gastrointest Endosc 2003;
58(6):847852.
5. Di F, V, Brunori MP, Rigo L, et al. Comparison of ultrasound-
secretin test and sphincter of Oddi manometry in patients with
recurrent acute pancreatitis. Dig Dis Sci 1999; 44(2):336340.
6. Sherman S, Lehman GA. Sphincter of Oddi dysfunction: diagnosis
and treatment. JOP 2001; 2(6):382400.
7. Fazel A, Burton FR. The effect of midazolam on the normal
sphincter of Oddi: a controlled study. Endoscopy 2002;
34(1):7881.
8. Sherman S, Gottlieb K, Uzer MF, et al. Effects of meperidine on
the pancreatic and biliary sphincter. Gastrointest Endosc 1996;
44(3):239242.
9. Goff JS. Effect of propofol on human sphincter of Oddi. Dig Dis
Sci 1995; 40(11):23642367.
10. Sherman S, Hawes RH, Madura JA, et al. Comparison of
intraoperative and endoscopic manometry of the sphincter of
Oddi. Surg Gynecol Obstet 1992; 175(5):410418.
11. Blaut U, Sherman S, Fogel E, et al. Inuence of cholangiography
on biliary sphincter of Oddi manometric parameters. Gastrointest
Endosc 2000; 52(5):624629.
12. Raddawi HM, Geenen JE, Hogan WJ, et al. Pressure measurements
from biliary and pancreatic segments of sphincter of Oddi.
Comparison between patients with functional abdominal pain,
biliary, or pancreatic disease. Dig Dis Sci 1991; 36(1):7174.
13. Toouli J, Roberts-Thomson IC, Dent J, et al. Manometric disorders
in patients with suspected sphincter of Oddi dysfunction.
Gastroenterology 1985; 88(5 Pt 1):12431250.
96
Post-sphincter of Oddi manometry patients should be monitored
in the recovery area for at least 46 hours. In case of any complaints
of signicant pain or nausea, one should check serum amylase and
lipase to help rule out procedural pancreatitis. Four-hour serum
amylase and lipase levels have been shown to be predictive of post-
ERCP pancreatitis.24 Patients could have pain only without pancre-
atitis after SO manometry and it might be reasonable to watch them
overnight or to provide adequate analgesia.25 In case of suspected
pancreatitis, we recommend adequate hydration and we routinely
administer 12 liters of intravenous uids over 68 hours. Analgesia
is important. Patient-controlled analgesia might be benecial.
For detailed description of post-ERCP complications and man-
agement of the same, please refer to Chapter 5.
14. Sherman S, Troiano FP, Hawes RH, et al. Sphincter of Oddi
manometry: decreased risk of clinical pancreatitis with use of a
modied aspirating catheter. Gastrointest Endosc 1990;
36(5):462466.
15. Sherman S, Hawes RH, Troiano FP, et al. Pancreatitis following bile
duct sphincter of Oddi manometry: utility of the aspirating
catheter. Gastrointest Endosc 1992; 38(3):347350.
16. Guda N, Partington S, Freeman M. Does the type of current (pure
vs blended) matter for post ERCP complications: a meta analysis.
Gastrointest.Endosc. 61(5). 2005. Ref Type: Abstract.
17. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent
prevent post-ERCP pancreatitis? A prospective randomized study.
Gastrointest Endosc 2003; 57(3):291294.
18. Freeman ML, Guda NM. Prevention of post-ERCP pancreatitis: a
comprehensive review. Gastrointest Endosc 2004; 59(7):
845864.
19. Thomas M, Catalano MF, Geenen JE. A prospective comparative
stent study for prophylaxis of post ERCP pancreatitis: comparison
of 5 Fr straight stent vs 3 FR pigtail stent. Gastrointest.Endosc 61(5),
196. 41-2005. Ref Type: Abstract.
20. Wehrmann T, Stergiou N, Schmitt T, et al. Reduced risk for
pancreatitis after endoscopic microtransducer manometry of the
sphincter of Oddi: a randomized comparison with the perfusion
manometry technique. Endoscopy 2003; 35(6):472477.
21. Craig AG, Omari T, Lingenfelser T, et al. Development of a sleeve
sensor for measurement of sphincter of Oddi motility. Endoscopy
2001; 33(8):651657.
22. Singh P, Gurudu SR, Davidoff S, et al. Sphincter of Oddi
manometry does not predispose to post-ERCP acute pancreatitis.
Gastrointest Endosc 2004; 59(4):499505.
23. Tarnasky P, Cunningham J, Cotton P et al. Pancreatic sphincter
hypertension increases the risk of post-ERCP pancreatitis.
Endoscopy 1997; 29(4):252257.
24. Testoni PA, Bagnolo F, Caporuscio S, et al. Serum amylase
measured four hours after endoscopic sphincterotomy is a
reliable predictor of postprocedure pancreatitis. Am J
Gastroenterol 1999; 94(5):12351241.
25. Wong GS, Teoh N, Dowsett JD, et al. Complications of sphincter of
Oddi manometry: biliary-like pain versus acute pancreatitis. Scand
J Gastroenterol 2005; 40(2):147153.
 SECTION 2 TECHNIQUES
Chapter
Balloon Dilation of the Papilla
11 Chan-Sup Shim
INTRODUCTION
Endoscopic biliary sphincterotomy (EST) has become a cornerstone
of therapeutic endoscopic retrograde cholangiopancreatography
(ERCP) of the biliary tree. The most common indication for EST is
to enlarge the access of the bile duct for stone extraction. The possi-
ble adverse consequences of EST include bacterial colonization and
chronic inammation of the biliary tree, the clinical relevance of
which is poorly understood.1 While increased incidence of primary
choledocholithiasis is an acceptable and relatively harmless long-
term result of EST, some experts express concern about the risk of
biliary malignancy. Longitudinal studies of patients who have had
biliodigestive anastomoses and surgical sphincteroplasty suggest an
incidence of late bile duct cancer between 5.6 and 7.4%.2,3
Endoscopic papillary balloon dilation (EPBD) is an alternative to
EST for removing bile duct stones.47 In an effort to avoid permanent
destruction of the biliary sphincter, EPBD seemed to be an attractive
alternative to early investigators, such as Staritz and Meyer zum
Buschenfelde, who rst reported it in 1983.8 In this procedure, a
balloon is inated to enlarge the opening of the bile duct at the level
of the biliary sphincter. The main theoretical advantage of this
technique is that it does not involve cutting the biliary sphincter.
Therefore, acute complications such as bleeding and perforation
should be less likely, and the function of the biliary sphincter is also
preserved.5
The enthusiasm for the potential advantages of EPBD over EST
for the avoidance of short-term complications of bleeding and
perforation, while preserving the biliary sphincter and possibly
reducing the long-term sequelae of EST was soon dampened by
reports of serious post-procedure pancreatitis.9 Therefore, EPBD was
nearly abandoned as a treatment for bile duct stones, but its use was
revived with the development of laparoscopic cholecystectomy. With
several groups reporting favorable results using EPBD for stone
extraction, conservation of the biliary sphincter regained popularity
in the 1990s. In 1995, Mac Mathuna et al. reported good results with
EPBD for treating bile duct stones in 100 consecutively treated
patients.4,5
The results of subsequent randomized, controlled trials compar-
ing EST to EPBD are conicting. Some authors have reported an
increased incidence of post-procedure pancreatitis, while others
have not, and an argument has been presented against EPBD and
its failure to provide adequate access for extracting difcult (large or
multiple) bile duct stones.6,7,10
The nal success rates for EST and EPBD are comparable; the
reported success rates of stone removal are 8199% for EPBD 4,6,7,10
and 8598% for EST.6,7 Randomized trials comparing EPBD with
EST suggest that EPBD is at least as effective as EST in patients with
small to moderate-sized bile duct stones.5,6,1018
TECHNIQUE
During the informed consent process, the risks and benets of
EPBD compared to EST should be discussed with patients and
consent obtained if EPBD is being considered. Preprocedural anti-
biotics are administered as appropriate. The procedure is performed
using a standard duodenoscope. During the procedure, identica-
tion of candidates for whom the EPBD is indicated can be simplied
by comparing the bile duct stone size to the diameter of the duode-
noscope on the same radiographic image; patients with stones that
have a diameter equal to or less than that of the duodenoscope are
considered eligible. After diagnostic ERCP and selective bile duct
cannulation a standard 0.025 or 0.035 inch guidewire is inserted into
the bile duct. After removing the cannula, an 8-mm balloon-tipped
catheter (Fig. 11.1) (Hurricane RX dilation balloon or Wire-guided
CRETM balloon; Boston Scientic, Natick, MA, USA; balloon length
3 cm, maximum inated outer diameter 8 mm) is passed over the
guidewire, positioned across the papilla, and inated with diluted
contrast medium at a pressure of 8 atm (Fig. 11.2). The balloon is
expanded slowly with a mixture of contrast medium and saline
(50/50) paying close attention to the waist of the balloon. When the
waist disappears, the ination is stopped. Care must be taken to
avoid rapid application of excessive pressure (Fig. 11.2). The dilation
is maintained for 1530 seconds. When the bile duct is less than
8 mm in diameter a 6-mm 2-cm balloon can be used. Other dilator
balloons can also be used (e.g. Hurricane Rx dilation balloon; Boston
Scientic, Natick, MA, USA, PET balloon; ConMed Endoscopic
Technologies, Billerica, MA, USA, Quantum balloon; Cook Endos-
copy Winston-Salem, NC, USA). The smaller balloons pass readily
through a diagnostic duodenoscope, such as an Olympus JF-240,
whereas the 24 Fr balloon requires a biopsy channel of at least
3.2 mm.
After papillary dilation, the stones are removed using Dormia
baskets and/or retrieval balloon catheters (Fig. 11.2). Mechanical
lithotripsy (BML-3Q-1, -4Q-1; Olympus Medical System Corpora-
tion, Tokyo, Japan) can be used to fragment stones if they are over
10 mm in diameter as determined from the cholangiogram. If the
stones are too large to engage within a mechanical lithotripsy basket,
an electrohydraulic lithotripter can be in the next session used to
crush the stones after introducing a baby cholangioscope (CHF-
BP30; Olympus Medical Systems, Tokyo, Japan) through the dilated
papilla. The initial ERCP session is performed within 60 min, and
if complete clearance of the stones fails, a biliary stent or nasobiliary
catheter can be inserted to prevent stone impaction. Data from
outside the US suggest that pancreatic duct stent placement is not
required after EPBD is performed.
The optimal size of the balloon in EPBD has not yet been estab-
lished. Most studies have reported results with 8-mm diameter
97
 SECTION 2 TECHNIQUES

Fig. 11.1 Dilating kit consisting of 8-, 10-, 12-, 15-, 18-, and 20-mm
balloons, syringe/gauge assembly and ination handle.

dilation balloons, although some studies have used balloons up to

15 mm in diameter.8 Theoretically, the larger the balloon, the easier
it is to extract large stones, although more complications after dila-
tion may be likely, and pancreatitis or damage to the sphincter may
be expected. However, there is insufcient information to support
these concerns. In contrast, Staritz et al. reported no complications,
even after using 15-mm balloons.8
Other points of EPBD to be claried are the optimal pressure for
dilation, the optimal duration of dilation, and the optimal number
of dilations. Other operators have reported the use of different pro- D
tocols, such as a maximal pressure of less than 1.5 atmospheres,5 a
duration of 4560 seconds after the disappearance of the balloon
waist,7 or repeated dilation protocols.4 It is still unknown which
protocol is best for the complete removal of stones and the preserva- C
tion of sphincter function. To answer these questions, more studies
that focus on the various EPBD techniques are needed.

INDICATIONS FOR AND LIMITATIONS OF

ENDOSCOPIC BALLOON DILATION
In the recent meta-analysis by Baron et al., the incidence of bleeding
was signicantly less after EPBD compared to EST.16 Clinically sig- E
nicant post-EST bleeding occurs in 25% of EST patients.17,18 In
addition, patients with coagulopathy and those requiring anticoagu-
lation within 3 days of the procedure are at increased risk for bleed-
ing.17 Thus, transient discontinuation of anticoagulation, correction
of coagulopathy with fresh frozen plasma, or platelet transfusion are
frequently used to avoid bleeding after EST, though these measures
might be inadequate to prevent it. EPBD provides a useful alterna-
tive to EST in such cases. No articles have described bleeding after
EPBD.4,6,7,10 In light of this, EPBD should be considered a viable
alternative to EST in patients with an underlying coagulopathy or
the need for anticoagulation following EST, as such patients have a Fig. 11.2 Endoscopic balloon dilation in a patient with multiple
higher incidence of post-EST bleeding.17 EPBD may signicantly small CBD stones. A The endoscopic cholangiogram demonstrated
reduce the risk of bleeding compared to EST in patients with multiple stones in the common bile duct. After diagnostic ERCP, a
0.035-inch guidewire was passed through the ERCP catheter into
advanced cirrhosis and coagulopathy. In these patients, EPBD is the common bile duct, and the catheter was removed. BD A
recommended over EST for treating choledocholithiasis.19 Another balloon-tipped catheter is inserted into the common bile duct over
population in which EPBD may be an attractive option is those the guidewire. The balloon is inated once it is located across the
patients who refuse blood transfusion for religious reasons, and papilla. The biliary sphincter can be seen as a waist in the balloon.
C The biliary sphincter is considered adequately dilated if the waist
patients with difcult anatomy that prevents safe orientation of the has disappeared completely. E After removing the balloon and
papillotome for EST (e.g. prior Billroth II gastrectomy, Fig. 11.3; or guidewire, stones are extracted using a Dormia basket.
intradiverticular location of the papilla, Fig. 11.4).14
Bergman et al. reported a randomized trial of EPBD and EST for
removing bile duct stones in patients with a prior Billroth II gastrec-
tomy.14 Compared to patients with a normal anatomy, patients with
prior Billroth II gastrectomy had a signicantly increased risk of

98
 Chapter 11 Balloon Dilation of the Papilla

A B C

D E F

G H I

J K L

Fig. 11.3 Serial endoscopic images AH and retrograde cholangiograms IL show endoscopic papillary balloon dilation of the biliary
sphincter in a patient with two bile duct stones and prior Billroth II gastrectomy. A Endoscopy shows an image of the papilla, upside-down.
B Two lling defects are seen on the cholangiogram. CF The balloon is advanced over a guidewire and is inated with diluted contrast.
G Transient oozing of blood is observed at the papilla after deating the balloon, but it does not develop into serious hemorrhage.
HL A stone is removed with a basket catheter.

bleeding after EST. Early complications occurred in 19% of the
patients who underwent EPBD as compared to 39% of the patients
who underwent EST. Endoscopic stone removal in patients with a
prior Billroth II gastrectomy and Billroth II anastomosis poses one
of the great challenges to the biliary endoscopist. Several methods
and instruments have been developed to enable EST in Billroth II
patients.10 Currently, the most widely accepted technique consists of
a needle-knife sphincterotomy over a previously inserted endopros-
thesis.20 Compared to standard EST in the normal anatomic situa-
tion, all of these techniques are more demanding and probably

99
 SECTION 2 TECHNIQUES

A B C

D E F

Fig. 11.4 Endoscopic images of EPBD in a patient with a CBD stone and a papilla with an intradiverticular location. A The major papilla
is located in the diverticulum. B Small EST is performed. C,D After cannulation of a guidewire, the balloon catheter is inated over the
guidewire up to 15 mm. E,F After removing the balloon, the stone is evacuated with a basket catheter through the widely opened
papilla.

associated with a smaller sphincterotomy incision, less successful Odds Likelihood

stone removal, and a higher rate of acute complications.14 When EST Variable ratio 95% CI ratio c2 P Value
is used for such patients, careful consideration must be given to the a
Age 0.98 0.951.00 3.88 0.049
direction and length of the incision, and a high level of skill is
required to avoid severe complications. With EPBD, however, once CBD diameter (mm)
a catheter is inserted into the common bile duct, the balloon catheter <12 3.24 1.208.72 8.14 0.017
12 1
is simply inserted and the balloon is inated. Therefore, patients
with Billroth II anatomy apppear to be especially suited for stone Number of stones
removal using EPBD. 2 2.08 0.904.80 7.5 0.024
>2 1
LIMITATIONS OF EPBD Size of stone (mm)
<10 10.0 1.1444.64 17.72 <0.001
10 1
The success of stone removal and procedure time varies between
EPBD and EST. Vlavianos et al. performed a univariate logistic
Table 11.1 Parameters predicting success of EPBD from the
regression analysis assessing the success of bile duct stone removal
univariate analysis
after EPBD. The following parameters were analyzed: sex, age, ran- a
The success decreases with increasing age. CI, condence interval; CBD, common
domization, presentation with jaundice, acute cholangitis or acute bile duct.
pancreatitis, diameter of the common bile duct (CBD) on the initial
cholangiogram, number of stones, and size of the largest stone. Of
these, age, diameter of the CBD, and size and number of stones were
signicantly associated with success (Table 11.1). Multivariate logis- stone size or number. Patients were excluded if CBD stones had a
tic regression analysis showed that only the size of the largest stone diameter = 12,21 = 14,22 = 15,10 or = 20 mm,23 or if there were more
was an independent predictor of success for duct clearance. On than ve23 or ten10 stones. The limitation of EPBD for extracting
average, these patients had a 12-mm diameter CBD and up to two large bile duct stones is highlighted by the more frequent need for
10-mm stones. These were taken as cut-off points in the statistical mechanical lithotripsy as an adjunctive procedure. This likely length-
analysis.11 ens the procedure time. Indeed, in three studies used in this analy-
Larger stones are more difcult to remove using EPBD because sis, the protocol called for the use of mechanical lithotripsy in the
the biliary opening is enlarged to a greater degree with EST. In fact, EPBD group if the diameter of any CBD stone was = 8,5 = 11,12 or
in many patients in the studies examined in the analysis by Baron = 12 mm.24 Therefore, EPBD may be more technically difcult to
et al. comparing EPBD to EST, patients were excluded based on perform and more time-consuming than EST.21

100
 Chapter 11 Balloon Dilation of the Papilla

In an analysis classifying the patients according to stone size,
both treatment approaches ultimately achieved similar success rates
and needed similar numbers of treatment sessions for patients with
stones less than 10 mm in diameter. For patients with stones over
10 mm in diameter, EPBD required a signicantly greater mean
number of treatment sessions than EST due to the technical dif-
culty in retrieving large stones after EPBD.
EPBD is a possible alternative to EST, especially in patients with
impaired hemostasis. However, large stones may be difcult to
remove with EPBD alone. Therefore, ideal patients for selecting
EPBD over EST are those with a limited number of CBD stones ( 3),
CBD stones with a maximum diameter = 10 mm, and minimally
dilated bile ducts.10,11,23,24 It is also important to use extreme caution
when EPBD is applied in the following clinical settings: the presence
of severe acute cholangitis, a history of previous or ongoing acute
pancreatitis, age = 50 years, and difcult biliary cannulation,25 espe-
cially because of reports describing fatal pancreatitis in younger
patients.26 In patients with severe cholangitis, one should consider
placing a biliary stent to ensure adequate drainage if EPBD is per-
formed. In the other clinical settings mentioned, placement of a
pancreatic duct stent could be considered to prevent post-ERCP
pancreatitis.
COMPLICATIONS
Early complications, dened as those occurring within 24 h of
the procedure, are pancreatitis, bleeding, infection (cholangitis or
cholecystitis), and perforation. The meta-analysis of randomized,
controlled trials by Baron et al. showed the early complication rate
of EPBD was comparable to EST for removing common bile duct
stones during ERCP.16 Overall, the early complication rates were
similar in EPBD and EST, 10.5 vs 10.3%, p = 0.9 (Table 11.1). Of
note, the bleeding rate was higher in the EST group (2.0 vs 0%, p =
0.001), while the rates of infection (2.7% for EPBD vs 3.6% for EST,
p = 0.3) and perforation (0.4 vs 0.4%, p = 1.0) were similar (Table
11.2). The rate of pancreatitis was higher in the EPBD group (7.4 vs
4.3%, p = 0.05) (Table 11.2). One patient death occurred in each
group, yielding a mortality rate of 0.2%.
In those patients suspected of having developed immediate
EPBD-related complications, complete blood counts, liver en-
zymes, and serum amylase are measured within 24 h after the
procedure. Abdominal roentgenograms, US, and CT are obtained
if needed.
Hemorrhage is one of the most common and serious complica-
tions of EST, and the presence of coagulopathy is one of the risk
factors for hemorrhage. In the meta-analysis by Baron et al.,16 bleed-
ing was clearly reduced when EPBD was performed as compared to
EST for the removal of CBD stones, but nearly all comparative
studies of EPBD and EST excluded patients with coagulopathy and
liver disorders. When bleeding occurs in cirrhotic patients, they may
also develop further complications, such as hepatic failure. Komatsu
et al. treated 24 cirrhotic patients with CBD stones using EPBD.7
Although hemostasis was impaired due to liver dysfunction, no
bleeding occurred and all patients responded well to treatment. In
particular, no complications were seen in four patients with Child-
Pugh class C, or in six patients with severe coagulopathy. The rate
of EST-related hemorrhage was 30% (6/20), whereas the rate for
EPBD-related hemorrhage was 0% (p = 0.009). Regarding the rates
of hemorrhage in relation to Child-Pugh class, most (n = 5) of the
bleeding complications occurred in patients with Child-Pugh class
C cirrhosis, while bleeding occurred in only one patient with Child-
Pugh B cirrhosis.19 Based upon these results, EPBD appears to be
the preferred strategy in patients with CBD stones and an underlying
coagulopathy and those who require full anticoagulation within 72 h

Vlavianos11 Fujita 22 Arnold 23 Minami 5 Bergman 6 Ochi 10 Natsui 24 Yasuda 12 Total

Pancreatitis (EPBD) 5/103 15/138 6/30 2/20 7/101 0/55 4/70 2/35 41/552
(1 severe) (2 severe) (2 severe) 7.4%a
Pancreatitis (ES) 1/99 4/144 3/30 2/20 7/101 2/55 3/70 2/35 24/554
(severe) (1 severe) 4.3%a
Bleed (EPBD) 0 0 0 0 0 0 0 0 0/552
0%b
Bleed (ES) 0/99 2/144 2/30 0/20 4/101 0/55 2/70 1/35 11/554
(1 severe) 2.0%b
Infection (EPBD) 2/103 4/138 3/30 0/20 4/101 0/55 2/70 0/35 15/552
2.7%c
Infection (ES) 1/99 11/144 0/30 0/20 5/101 0/55 3/70 0/35
20/554
3.6%c
Perforation (EPBD) 0/103 0/138 0/30 0/20 2/101 0/55 0/70 0/35 2/552
0.4%
Perforation (ES) 0/99 0/144 0/30 0/20 1/101 1/55 0/70 0/35 2/554
0.4%
Death (EPBD) 0 0 0 0 1 0 0 0 1/552
Death (ES) 1 0 0 0 0 0 0 0 1/554

Table 11.2 Procedure-related complications based on prospective trials comparing Endoscopic Balloon Dilation (EPBD) and Endoscopic
Sphincterotomy (EST) for treating choledocholithiasis
a
p = 0.05 (pancreatitis in EPBD vs. EST. bp = 0.001 (bleeding in EPBD vs. EST. cp = 0.3 (infection in EPBD vs. EST.

101
 SECTION 2 TECHNIQUES

of stone removal27 since these patients are at higher risk for post-
sphincterotomy bleeding following EST.17
Procedure-related pancreatitis needs to be addressed. In recent
years, ve prospective randomized controlled trials of EPBD versus
EST have been performed.6,11,22,23 The frequency and severity of post-
EPBD pancreatitis are summarized in Table 11.3.28 Of these, the
Dutch6 and United Kingdom11 studies showed similar efcacy and
safety between the two methods. Their incidence of pancreatitis was
similar in patients undergoing EPBD and EST, in the range of 5 to
7%. In the Japanese study,22 the rate of pancreatitis was slightly
higher with EPBD than with EST; however, there were no reports of
severe pancreatitis and all patients recovered with conservative treat-
ment. In addition, Mac Mathuna et al.4 and Komatsu et al.7 con-
ducted large-scale studies, but they were not randomized controlled
trials; the incidence rates of pancreatitis were 5 and 7%, respectively.
No severe pancreatitis or fatalities occurred (Table 11.3).
The mechanism of post-EPBD hyperamylasemia and pancreatitis
is not clear, although it seems to be multifactorial. Balloon compres-
sion of the papilla or the pancreatic duct orice may provoke peri-
papillary edema or sphincter of Oddi spasm.15,23 Bile duct cannulation
per se or transpapillary manipulation (stone extraction, nasobiliary
drainage) may also induce edema or spasm. The peri-papillary
edema or spasm may in turn obstruct the ow of pancreatic juice
and eventually induce pancreatic edema or pancreatitis associated
with hyperamylasemia.15,23 Contrast medium injection or cannula-
tion of the pancreatic duct is also likely to have some effect on the
pancreas and pancreatic secretions.29
Several studies have examined the risk factors for post-ERCP
pancreatitis. The most important risk factors were related to either
patient characteristics or the ease of cannulation. In the study by
Bergman et al.15, none of these factors was associated with EPBD-
induced pancreatitis. However, a univariate analysis revealed that
stone size, distal bile duct stricture, mechanical lithotripsy, multiple
endoscopic sessions, additional EST, and endoscopic nasobiliary
drainage after EPBD were signicant risk factors for the develop-
ment of acute pancreatitis after EPBD (Table 11.4), although in the
multivariate analysis, mechanical lithotripsy was the only risk factor
for pancreatitis after EPBD (Table 11.5).30.31
Yasuda et al. reported an increased rate of hyperamylasemia after
EPBD in patients who underwent mechanical lithotripsy as com-
pared to patients in whom stones were fragmented with extracorpo-
real shock-wave lithotripsy (ESWL), a fact that might be explained
by the reduced manipulation of the papillary complex in the latter
group.12 Sugiyama et al. reported an incidence rate of more than 30%
in patients with a history of pancreatitis,25 which may limit the use
of EPBD for treating CBD stones in the setting of acute pancreatitis.
Younger patients have a higher risk of post-ERCP pancreatitis,32
though many of the EPBD studies have not included patients
younger than 50 years old, the very group for whom concern exists
on the long-term complications of EST. In addition, some studies

PANCREATITIS SEVERITY

Investigator (Ref. No.) Year No. of patients Patients with pancreatitis (%) Mild Moderate Severe Death
Bergman (6) 1997 101 7 (7.0) 5 0 2 0
DiSario (26) 2004 117 18 (15.4) 7 5 6 2
Arnold (23) 2001 30 6 (20.0) 4 0 2 0
Vlavianos (11) 2003 103 5 (4.9) 2 2 1 0
Fujita (22) 2003 144 15 (10.4) 12 3 0 0
Tsujino (28) 2004 304 15 (5.0) 8 7 0 0

Table 11.3 Frequency and severity of post-EPBD pancreatitis in randomized studies

Risk factor Patients with pancreatitis (11/156) Patients without pancreatitis (145/156) p
Pre-endoscopic papillary balloon dilation related
Stone size (10 mm) 9 (81.8) 54 (48.6) 0.036
Distal bile duct stricture 3 (27.3) 6 (5.4) 0.008
Age <60 years 4 (36.4) 50 (45) 0.83
Female sex 7 (63.6) 50 (45) 0.24
Previous cholecystectomy 4 (36.4) 29 (26.1) 0.46
Periampullary diverticulum 4 (36.4) 32 (28.8) 0.51
Bile duct diameter (mean SD [mm]) 17.1 3.9 14.8 4.5 0.31
Multiple stones (2) 7 (63.7) 62 (55.8) 0.73
Pancreatic contrast injection (1) 7 (63.7) 62 (55.9) 0.71
Post-endoscopic papillary balloon dilation related
Mechanical lithotripsy 7 (63.6) 25 (22.5) 0.003
Multiple endoscopic session (2) 6 (54.5) 25 (22.5) 0.02
Additional EST 3 (27.3) 2 (1.8) 0.00
ENBD after EPBD 4 (36.4) 16 (13.5) 0.046

Table 11.4 Risk factors for pancreatitis after endoscopic papillary balloon dilation in a univariate analysis

102

Adjusted
Risk factor odds ratio 95% CI P each).16
Mechanical lithotripsy 5.25 1.2921.31 0.02
Multiple ERCP session (2) 2.81 0.6711.86 0.16
Stone size (10 mm) 1.24 0.275.84 0.35
Multiple stone (2) 1.44 0.1612.65 0.74
Female sex 1.92 0.487.63 0.36
Age 60 years 0.58 0.437.82 0.68
Table 11.5 Risk factors for pancreatitis after endoscopic papillary
balloon dilation in a multivariate analysis
have used gabexate mesylate, a protease inhibitor, to prevent post-
ERCP pancreatitis.22 Although gabexate mesylate was given to both
the EST and EPBD groups, it might have reduced the severity of
pancreatitis in the EPBD group.33 Several studies have suggested
that pancreatic duct stent placement reduces the risk of post-ERCP
pancreatitis in high-risk patients.29,30,32 Recently, pancreatic duct
stent placement has been used in a non-randomized comparative
trial of patients undergoing EPBD to remove CBD stones.33 Although
no signicant difference was seen in the rate of pancreatitis when
pancreatic duct stents were placed, pancreatic duct stent placement
might be useful for preventing post-ERCP pancreatitis in young
patients undergoing EPBD.
Acute pancreatitis is usually mild and occurs in approximately
6% of patients after EPBD for bile duct stone extraction. EPBD often
results in hyperamylasemia (25%), although this is usually clinically
inconsequential. Hyperamylasemia, however, may represent pan-
creatic irritation or latent pancreatic injury. Particular care is neces-
sary when EPBD is performed on younger patients, those with a
history of pancreatitis, and patients with a non-dilated bile duct, or
when cannulation is difcult, given the high frequency of hyperamy-
lasemia. While performing EPBD, careful attention should also be
paid to gentle handling of the papilla, avoiding unnecessary pan-
creatic opacication, and the injection of contrast medium into the
pancreatic duct. In addition to pancreatic duct stent placement after
EPBD, potential safeguards for prevention of pancreatitis after EPBD
include gradual ination of the balloon at a low pressure, intrave-
nous infusion of isosorbide dinitrate34 and temporary placement of
a nasobiliary drainage catheter, which may act by preventing pan-
creatic duct obstruction from residual stones or papillary edema.35
Infection is another potential problem that can be challenging to
the endoscopist. The incidence of cholangitis and cholecystitis
appears to be higher following EST than in EPBD, although in the
analysis by Baron et al. the difference did not reach statistical signi-
cance. In patients with severe cholangitis, one should consider
placing a biliary stent to ensure adequate drainage if EPBD is per-
formed. Currently, cholecystectomy is recommended in patients
with gallstones after EST, as there is a high rate of acute cholecystitis.
Several authors have reported that the incidence of cholecystitis is
signicantly lower after EPBD than after EST.6,10 Natsui et al. diag-
nosed acute cholecystitis in 25.0% of EST patients and in 0% of EPBD
patients who had gallbladder stones after endoscopic treatment.24 It
is hasty to conclude that cholecystectomy is unnecessary for all
patients with gallbladder stones after EPBD, as most gallbladder
stones migrating to the bile duct are probably less likely to pass spon-
taneously into the duodenum and more likely to cause cholangitis.
Perforation is a very rare, but potentially fatal, complication fol-
lowing EPBD. In the meta-analysis by Baron et al., there was no
Chapter 11 Balloon Dilation of the Papilla
difference between EST and EPBD in rates of perforation (0.4%
Recurrent bile duct stones are a late complication of EPBD and
EST. Several studies have shown that 320% of patients who under-
went EST developed recurrent bile duct stones during a median
period of 915 years.36 The permanent destruction of the sphincter
mechanism by EST results in ascending bacterial infection of the
biliary tract, which might be involved in the formation of brown
pigment stones. In contrast, manometric studies have shown that
EPBD does not completely restore the sphincter function to its intact
state, but could preserve it better than EST.5,10,12 Therefore, we would
expect EPBD to reduce the recurrence of bile duct stones as com-
pared to EST in patients without an intact gallbladder.37
LARGE BALLOON DILATION AFTER MINIMAL
BILIARY SPHINCTEROTOMY (EST)
EPBD was rst initiated for the purpose of extracting common bile
duct stones while minimizing damage to the sphincter of Oddi.
According to the stone size, both EPBD and EST approaches ulti-
mately achieve similar success rates and need similar numbers of
treatment sessions for patients with stones less than 10 mm in
diameter. However, for patients with stones over 10 mm in diame-
ter, EPBD requires a signicantly greater mean number of treatment
sessions than EST. This is due to the technical difculty in retrieving
large stones after EPBD, as the papillary orice cannot be enlarged
to the same extent as after EST. The limitation of EPBD for extract-
ing large bile duct stones is highlighted by the more frequent need
for mechanical lithotripsy as an adjunctive procedure, which likely
lengthens procedure time.
To overcome the limitations of conventional EPBD, large balloon
dilation after minimal biliary sphincterotomy has been devised.
Balloon dilation after minimal EST is effective for retrieving large
biliary stones without the use of mechanical lithotripsy (Figs 11.5
and 11.6). Although EST with a large incision may be effective
in reducing the need for mechanical lithotripsy, a large incision
has a higher risk of perforation and possibly a higher risk of
bleeding than standard EST. This innovative, novel method
incorporating slow dilation of the papilla to a large diameter, can
provide a larger opening than a large EST (Fig. 11.5) and prevents
perforation and bleeding. This method of stone retrieval is easy to
perform and can effectively treat large or multiple bile duct stones
(Fig. 11.6).
The balloons employed in this technique are larger than those
currently used for standard endoscopic balloon dilation of the bile
duct (Fig. 11.1), but similar to those used in preliminary studies of
balloon dilation of the papillary sphincter.7,9 Although the technical
aspects of EST are well described, the criteria for creating a sphinc-
terotomy that is adequate for removing large stones have not been
established. It is usually impossible to evaluate the adequacy of a
sphincterotomy based on anatomic parameters alone. Data from
an earlier era when sphincterotomy was performed at laparotomy
indicate that sphincteroplasty was generally superior to sphincter-
otomy in terms of adequacy.38 Moreover, complete elimination of
sphincter function by EST is not always possible. Biliary manometric
studies after EST conrm that sphincterotomy is incomplete in most
cases.39
EST has been performed after endoscopic balloon (810 mm)
dilation in cases when dilation was inadequate.5 Therefore, this
103
 SECTION 2 TECHNIQUES

A B C

D E

Fig. 11.5 A case of large balloon dilation after minimal EST in a patient with multiple large extrahepatic bile duct stones. A Retrograde
cholangiogram shows multiple large stones that completely ll the extrahepatic bile duct. B,C After minimal EST, a large balloon is inated
up to 18 mm over the guidewire and through the sphincterotomized papilla. D The papillary orice is dilated fully and the bile duct mucosa
is readily seen. E The cholangiogram shows a completely evacuated extrahepatic bile duct after sweeping out the multiple stones with
the balloon catheter.

A B C

D E F

Fig. 11.6 A huge stone is impacted at the bile duct bifurcation. A,B After sphincterotomy, large balloon dilation was performed up to
18 mm. C,D Removal with a large basket catheter and mechanical lithotriptor failed, and retrieval of the large stone was attempted with
a retrieval balloon catheter C. The stone was pulled out with a retrieval balloon catheter C and extracted from the papilla D. E,F The huge
stone (4.5 2.0 cm) was nally evacuated without crushing.

104

technique is unique, as balloon dilation is performed after EST,
using large-diameter balloons, to extract bile duct stones that resisted
removal with conventional techniques. Mechanical lithotripsy is
another alternative for removing such stones, with an overall success
rate of over 80%,40 although this procedure can be lengthy and
requires a second session in 30% of cases.41
Using a therapeutic duodenoscope (TJF 240; Olympus
Medical System, Tokyo, Japan), the endoscope is advanced to the
duodenum. It is important to use a duodenoscope with a large
working channel (4.2 mm in diameter), for easier passage of large
balloons. The difference from conventional EPBD is that EST is
performed before the balloon catheter is inserted. In most cases, a
major EST is not required and a minimal is sufcient. This is
because the purpose of EST is not to dilate the sphincter of Oddi
(SO), but to direct the direction of SO dilation. When using a large
balloon catheter to dilate the SO without EST, it is difcult to predict
the direction in which the SO will dilate. Therefore, by performing
a minimal EST, the direction of papilla dilation can be predicted.
Another reason for minimal EST is to prevent post-procedure pan-
creatitis by minimizing the peri-papillary edema after dilating the
papilla.
After EST, a guidewire is inserted into the bile duct and a balloon
catheter is guided over the wire. The diameter of the balloon catheter
should be 1520 mm. A balloon catheter that was initially developed
for dilation in pyloric stenosis (Wire-guided CRETM balloon; Boston
Scientic, Natick, MA, USA) can be useful (Fig. 11.1).
The diameter of the balloon catheter is determined by the size of
the bile duct stone and the size of the bile duct proximal to the
tapered segment. EST with a small incision up to the pancreatic
orice is performed over a guidewire. Endoscopic papillary dilation
is performed slowly with a large balloon (maximum of 20 mm in
diameter) to match the size of the bile duct. Approximately 1 min of
balloon dilation time is sufcient.
Dilation with large-diameter balloons is performed at the same
session after EST. Over-the-guidewire type balloons for esophageal/
pyloric dilation are used. The balloon catheters are passed over a
guidewire and positioned at the biliary orice; the middle portion of
the balloon is gradually lled with diluted contrast medium under
endoscopic and uoroscopic guidance to maintain the correct posi-
tion and to observe the gradual disappearance of the waist in the
balloon, which is taken to indicate progressive dilation of the orice.
Once the waist has disappeared, the balloon is kept in position for
2045 seconds, after which it is deated and removed. A standard
stone basket or retrieval balloon catheter is then used to remove the
stones. In a few cases, the waist in the balloon decreases, but does
not disappear completely, in this case keeping the balloon in place
for over 45 seconds may be useful. Stones are then retrieved from
the bile duct with a retrieval balloon catheter or a stone basket. After
stone retrieval, washing the bile duct with normal saline may help
to detect any remaining stones.
A dilation time lasting less than 1 minute may actually induce
bleeding, which may be attributable to insufcient compression
time by the balloon. After dilating the papilla for 1 min, the balloon
catheter is removed and a basket is inserted to remove the stone. If
the stone is too large to pass through the papilla, mechanical litho-
tripsy can be used. In those patients in whom stone clearance is still
impossible, a nasobiliary drain or biliary stent can be placed and
another session can be performed at a later time using large diame-
ter balloons.
Chapter 11 Balloon Dilation of the Papilla
Ersoz et al. reported that dilation with a large-diameter
balloon after EST was useful for clearing bile duct stones in
patients with a tapered distal bile duct. By using a larger balloon,
the distal duct can be shaped into a near square, facilitating
stone removal.
Patients with distal CBD stenosis or a narrow common bile
duct are at risk of bleeding, perforation, or bile duct injury after
complete balloon dilation. Therefore, it seems prudent to avoid
excessive dilation in patients with risk of balloon dilation beyond
the width of the common bile duct. In one study this technique
was successful in 89% of patients. In the remaining 11%, stone
removal was achieved easily after mechanical lithotripsy. EST
followed by dilation with a large-diameter balloon was also effective
for clearing 95% of large-diameter (1528 mm) bile duct stones
(Fig. 11.6), with mechanical lithotripsy required in only two
patients.42
In 19 of 24 patients, extrahepatic bile duct stones were removed
without endoscopic mechanical lithotripsy (EML) after dilating the
ampulla.43 Stone retrieval was successful in all cases without the
need to crush large stones up to 14 3 mm.44
Standard EST is the classical treatment modality for extrahepatic
bile duct stones. However, for large bile duct stones (usually >15 mm
in diameter), EML is used to break the large stones into small frag-
ments. If multiple large bile duct stones are present, repeated EML
is needed to remove the extrahepatic bile duct stones. However, if
the ampulla can be dilated widely (Fig. 11.5), such large stones can
be removed without the use of EML (Figs 11.5 and 11.6). Patients in
whom bile duct stones cannot be cleared because of a tapered distal
bile duct and patients with large, square, or barrel-shaped stones
would benet from this procedure.
The most common complications of this procedure are mild
cholangitis, pancreatitis, bleeding, and perforation. Complications
occurred in 15.5% of patients in one study,40 with most (10.3%)
being mild and self-limiting. Moderately severe bleeding
developed in three patients (5.2%), which was attributed to EST,
and all recovered without the need for surgery. Perforation did
not occur in any patient who underwent dilation with a large
diameter balloon.45 Mild pancreatitis developed in two patients
(3.4%). Theoretically, the risk of pancreatitis by large balloon
dilation after minimal sphincterotomy is less than balloon dilation
alone. It is probable that after EST, the force exerted by balloon
dilation is directed more toward the common bile duct than
the pancreatic orice. Minimal EST before a large balloon dilation
might decrease the risk of pancreatitis as compared to dilation
alone.
Bleeding occurs in 25% of patients undergoing EST to remove
bile duct stones.46,47 In contrast, no signicant bleeding has been
observed after endoscopic balloon dilation.6,7 The bleeding rate (9%)
reported by Ersoz et al. was higher than the rate reported for stan-
dard EST and EPBD.40 Therefore, bleeding is a potentially important
complication, particularly in patients with a tapered distal bile duct.
Studies of larger series of patients are warranted to determine the
frequency of this complication.
In conclusion, large balloon dilation after minor EST may be
helpful in the case of bile duct stones that are difcult to extract
using EST and conventional techniques. This procedure could
reduce the sessions of EML and shorten the procedure time, and
thus serve as an effective treatment modality for multiple large
extrahepatic bile duct stones.
105
 SECTION 2 TECHNIQUES
REFERENCES
1. Bergman JJ, van Berkel AM, Groen AK, et al. Biliary manometry,
bacterial characteristics, bile composition, and histologic changes
fteen to seventeen years after endoscopic sphincterotomy.
Gastrointest Endosc. 1997; 45(5):400405.
2. Hakamada K, Sasaki M, Endoh M, et al. Late development of bile
duct cancer after sphincteroplasty: a ten- to twenty-two-year
follow-up study. Surgery. 1997; 121(5):488492.
3. Tocchi A, Mazzoni G, Liotta G, et al. Late development of bile duct
cancer in patients who had biliary-enteric drainage for benign
disease: a follow-up study of more than 1000 patients. Ann Surg.
2001; 234(2):210214.
4. Mac Mathuna P, White P, Clarke E, et al. Endoscopic balloon
sphincteroplasty (papillary dilation) for bile duct stones: efcacy,
safety, and follow-up in 100 patients. Gastrointest Endosc 1995;
42:468474.
5. Minami A, Nakatsu T, Uchida N, et al. Papillary dilation vs
sphincterotomy in endoscopic removal of bile duct stones. A
randomized trial with manometric function. Dig. Dis. Sci. 1995;
40:25502554.
6. Bergman JJ, Rauws EAJ, Fockens P, et al. Randomized trial
of endoscopic balloon dilation versus endoscopic
sphincterotomy for removal of bile duct stones. Lancet 1997;
349:11241129.
7. Komatsu Y, Kawabe T, Toda N, et al. Endoscopic papillary balloon
dilation for the management of common bile duct stones:
experience of 226 cases. Endoscopy 1998; 30:1217.
8. Staritz M, Ewe K, Meyer zum Buschenfelde KH. Endoscopic
papillary dilation (EPD) for the treatment of common bile duct
stones and papillary stenosis. Endoscopy. 1983; 15 Suppl
1:197198.
9. Kozarek RA. Balloon dilation of the sphincter of Oddi. Endoscopy
1988; 20:207210.
10. Ochi Y, Mukawa K, Kiyosawa K, et al. Comparing the treatment
outcomes of endoscopic papillary dilation and endoscopic
sphincterotomy for removal of bile duct stones. J Gastroenterol
Hepatol. 1999; 14(1):9096.
11. Vlavianos P, Chopra K, Mandalia S, et al. Endoscopic balloon
dilatation versus endoscopic sphincterotomy for the removal of
bile duct stones: a prospective randomized trial. Gut 2003;
52:11651169.
12. Yasuda I, Tomita E, Enya M, et al. Can endoscopic papillary
balloon dilation really preserve sphincter of Oddi function? Gut.
2001; 49(5):686691.
13. Cho YD, Moon JH, Shim CS, et al. Prospective analysis of
endoscopic papillary balloon dilation and endoscopic
sphincterotomy for removal of common bile duct stones. Korean
J Gastroenterol 1999; 33:258267 [Korean].
14. Bergman JJ, van Berkel AM, Bruno MJ, et al. A randomized trial of
endoscopic balloon dilation and endoscopic sphincterotomy for
removal of bile duct stones in patients with a prior Billroth II
gastrectomy. Gastrointest Endosc. 2001; 53(1):1926.
15. Bergman JJ, van Berkel AM, Bruno MJ, et al. Is endoscopic balloon
dilation for removal of bile duct stones associated with an
increased risk of pancreatitis or a higher rate of
hyperamylasemia? Endoscopy. 2001; 33:416420.
16. Baron TH, Harewood GC. Endoscopic balloon dilation of the
biliary sphincter compared to endoscopic biliary sphincterotomy
for removal of common bile duct stones during ERCP: A
metaanalysis of randomized, controlled trials. Am J Gastroenterol
2004; 99:14551460.
17. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med. 1996; 26;
335(13):909918.
106
18. Rabenstein T, Roggenbuck S, Framke B, et al. Complications of
endoscopic sphincterotomy: Can heparin prevent acute
pancreatitis after ERCP?. Gastrointest Endosc 2002; 55:476483.
19. Park DH, Kim MH, Lee SK, et al. Endoscopic sphincterotomy vs
endoscopic papillary dilatation for choledocholithiasis in patients
with liver cirrhosis and coagulopathy. Gastrointest Endosc 2004;
60:180185.
20. Van Buuren HR, Boender J, Nix GA, et al. Needle-knife
sphincterotomy guided by a biliary endoprosthesis in Billroth II
gastrectomy patients. Endoscopy. 1995; 27(3):229232.
21. Minami A, Maeta T, Kohi F, et al. Endoscopic papillary dilation by
balloon and isosorbide dinitrate drip infusion for removing bile
duct stones. Scand J Gastroenterol 1998; 33:765768.
22. Fujita N, Maguchi H, Komatsu Y, et al. Endoscopic sphincterotomy
and endoscopic papillary balloon dilation for bile duct stones: A
prospective randomized controlled multicenter trial. Gastrointest
Endosc 2003; 57:151155.
23. Arnold JC, Benz C, Martin WR, et al. Endoscopic papillary balloon
dilation vs. sphincterotomy for removal of common bile duct
stones: A prospective randomized pilot study. Endoscopy 2001;
33:563567.
24. Natsui M, Narisawa R, Motoyama H, et al. What is an appropriate
indication for endoscopic papillary balloon dilation? Eur J
Gastroenterol Hepatol 2002; 14:635640.
25. Sugiyama M, Izumisato Y, Abe N, et al. Predictive factors for acute
pancreatitis and hyperamylasemia after endoscopic papillary
balloon dilation. Gastrointest Endosc. 2003; 57(4):531535.
26. Disario JA, Freeman ML, Bjorkman DJ, et al. Endoscopic balloon
dilation compared with sphincterotomy for extraction of bile duct
stones. Gastroenterology. 2004 Nov; 127(5):12911299.
27. Huibregtse K. Biliary sphincter balloon dilation; who, when and
how? Can J Gastroenterol. 1999; 13(6):499500.
28. Tsujino T, Isayama H, Komatsu Y, et al. Risk factors for pancreatitis
in patients with common bile duct stones managed by
endoscopic papillary balloon dilatation. Am J gastroenterol 2005;
100:3942.
29. Cristoforidis E, Goulimaris I, Kanellos I, et al. Post-ERCP pancreatitis
and hyperamylasemia: patient-related and operative risk factors.
Endoscopy 2002; 34:286292.
30. Moon JH, Cho YD, Shim CS, et al. Risk factors for pancreatitis after
endoscopic papillary balloon dilation. Kor J Gastroentrol. 2001;
38(Suppl.2):192.
31. Shim CS. Endoscopic papillary balloon dilatation for removal of
common bile duct stones: a review. Dig Endosc 2003; 15:16.
32. Mehta SN, Pavone E, Barkun JS, et al. Predictors of post-ERCP
complications in patients with suspected choledocholithiasis.
Endoscopy 1998; 30:457463.
33. Cavallini G, Tottobello A, Frulloni L, et al. Gabexate for the
prevention of pancreatic damage related to endoscopic
retrograde cholangiopancreatography. Gabexate in digestive
endoscopy Italian Group. N Engl J Med 1996; 335:919923.
34. Nakagawa H, Ohara K. Safeguards against acute pancreatitis
associated with endoscopic papillary balloon dilatation. J
Hepatobiliary Pancreat Surg. 2006; 13(2):7579.
35. Sato D, Shibahara T, Miyazaki K, et al. Efcacy of endoscopic
nasobiliary drainage for the prevention of pancreatitis after
papillary balloon dilatation: a pilot study. Pancreas. 2005 Jul;
31(1):9397.
36. Tanaka M, Takanata S, Konomi H, et al. Long-term consequence
of endoscopic sphincterotomy for bile duct stones. Gastrointest
Endosc 1998; 48:465469.
37. Komatsu Y, Fujita N, Magucho H, et al. Prospective randomized
trial of endoscopic sphincterotomy compared with endoscopic

papillary balloon dilatation for bile duct stones: late complications
after stone removal. Gastrointest Endosc 2005; 61(5):AB209.
38. Kozloff L, Joseph WL. Transduodenal sphincteroplasty for biliary
tract disease, Am J Surg 1975; 41:125130.
39. Wehrman T, Wiewer K, Lembrecke B, et al. Effect of endoscopic
sphincterotomy on Oddi manometry results in patients with or
without papillary stenosis. Gastroenterol 1995; 33:662668.
40. Cipoletta L, Costamagna G, Bianco MA, et al. Endoscopic
mechanical lithotripsy of difcult common bile duct stones. Br J
Surg 1997; 84:14071409.
41. Sorbi D, Van Os EC, Aberger FJ, et al. Clinical application of a new
disposable lithotripter: a prospective multicenter study.
Gastrointest Endosc 1999; 49:210213.
42. Ersoz G, Tekesin O, Ozutemiz AO, et al. Biliary sphincterotomy
plus dilation with a large balloon for bile duct stones that are
difcult to extract. Gastrointest Endosc. 2003 Feb; 57(2):156159.
Chapter 11 Balloon Dilation of the Papilla
43. Yoo BM. Large balloon-lithotripsy (LB-L) in patients with large
extrahepatic bile duct tones. Gastrointest Endosc 2005; 61:
AB244.
44. Minami A, Okuyama T, Hitose S. Small sphincterotomy combined
papillary dilatation with large balloon permits retrieval of large
stones without lithotripsy second report. Gastrointest Endosc
2005; 61:AB213.
45. Hwang JH, Kim YG, Lee KC, et al. Endoscopic sphincterotomy
plus endoscopic papillary large balloon dilatation for large
bile duct stones. Korean J Gastrointest Endosc 2006;
32:184189.
46. Leung JW, Chan FK, Sung JJ. Endoscopic sphincterotomy induced
hemorrhage: a study of risk factors and the role of epinephrine
injection. Gastrointest Endosc 1995; 42:550554.
47. Freeman ML. Complications of endoscopic biliary
sphincterotomy: a review. Endoscopy 1997; 29:288297.
107
 SECTION 2 TECHNIQUES
Chapter
Biliary Sphincterotomy
12 Horst Neuhaus
INTRODUCTION
Diagnostic ERCP has been increasingly replaced by MRCP and
endoscopic ultrasound which are comparable in accuracy but are
non-invasive or less invasive, respectively. ERCP is now mainly per-
formed for therapeutic pancreaticobiliary interventions. Endoscopic
sphincterotomy (EST) of the biliary sphincter is used for the treat-
ment of disorders of the papilla of Vater or to facilitate adjunctive
procedures in the bile duct. Since the introduction of EST in 1973,
a variety of complementary methods have been developed for the
management of ductal obstruction. They have become invaluable
tools for minimally invasive therapy of biliary diseases and have
gained widespread acceptance throughout the world. The need for
EST depends on the indication. Data from several prospective, mul-
ticenter trials have allowed the determination of the clinical, ana-
tomical and technical parameters and their relation to the efcacy
and safety of EST. Outcomes following EST can be affected by ductal
cannulation prior to EST, by subsequent therapeutic interventions,
and by the expertise of the endoscopist.
DESCRIPTION OF THE TECHNIQUE
Premedication, duodenoscopy and the approach to the papilla are
the same as for diagnostic ERC, as discussed in Chapters 5 and 8.
The use of a therapeutic endoscope is recommended since the large
diameter instrumentation channel allows for insertion of large
diameter stents and accessories required for therapeutic interven-
tions. The initial use of a standard sphincterotome for deep bile duct
cannulation is recommended for several reasons. First, when it is
anticipated that a sphincterotomy will be needed, exchange to a
sphincterotome from another catheter is avoided. Secondly, it allows
variable upward tip deection in order to introduce the tip of the
catheter into the biliary orice; the tip deection is then relaxed to
achieve deep cannulation. Steerable catheters are comparably effec-
tive and safe for this technique but their additional use does not
seem to be cost-effective for routine use.1 Two randomized con-
trolled trials showed success rates of 84% and 97%, respectively, for
primary cannulation with sphincterotomes which were superior
to the use of standard catheters with no signicant differences
in safety.2,3
Instruments (see also Chapter 4)
The type of the sphincterotome should be selected according to
the individual anatomic situation and the preference of the endos-
copist. According to a prospective US multicenter trial standard
reusable devices are satisfactory for most cases of EST and yield cost
savings as compared to disposable triple lumen sphincterotomes.4
However these results are not directly transferable to other health
care systems and an individual cost analysis is recommended
for each center. Tapered devices, which require smaller wires
(0.025 or less) can be easier to insert into the papilla but are also
more prone to cause tissue trauma and contrast inltration than
those with a more blunted tip. A recently developed ultra-smooth
tapered rounded polished tip may overcome this potential problem
(Fig. 12.1).
Modern sphincterotomes provide a lumen for insertion of a
guidewire and an integrated hub for contrast injection. These devices
allow repeated injection of contrast media without need for guide-
wire removal. This approach can be very helpful in difcult cannula-
tion or in targeting ductal strictures with the wire under uoroscopic
guidance. Sphincterotomes with a preloaded guidewire are conve-
nient for the assistant and may accelerate the procedure. Recently
developed technologies which utilize short-length guidewires seem
to be additionally helpful since they reduce the over-the-wire
exchange to a short part of the total device length while locking the
wire. They offer the option of guidewire manipulation by the endos-
copist which can be advantageous depending on the expertise of the
operator and the assistant. Sphincterotomes also differ in cutting
wire length, wire characteristics and shaft stiffness. A short, 20 mm
cutting wire can be precisely controlled but it tends to draw the
cutting direction toward the 2 oclock position. Contact with the
sphincter may be inadequate, thus making the cutting difcult in
some situations. However, the advantage of shorter cutting wires
(e.g. exposed length of 25 mm) over a 30 mm wire device is the
reduction of risk of an uncontrolled large cut when inserted too deep
into the bile duct. In addition, the proximal part of a long cutting
wire may come into contact with the elevator of the duodenoscope
or duodenal wall; the former causes wire breakage when electocau-
tery current is applied. This problem can be overcome by use of a
sphincterotome that is coated on the proximal part of the cutting
wire (Fig. 12.2). A thin monolament wire provides a clean, sharp
cut but may break more easily than a braided wire during application
of electrocautery. There are no formal trials that compare the efcacy
and safety of these different devices. As experience grows, each
endoscopist will develop preferences for a limited array of standard
accessories depending on expertise, skill of assistants and patient
selection. The use of special sphincterotomes can be limited to par-
ticular cases. A thin device with a 4.0 Fr ultra taper tip is useful after
failed cannulation using standard techniques in the setting of a
small papilla, suspicion of a narrow ductal orice or difculty in
achieving proper cutting orientation. The latter problem can be over-
come by the use of a rotatable sphincterotome. This instrument
utilizes a specially designed handle that allows for controlled tip
rotation. Sphincterotomes with a tip length of more than 5 mm can
occasionally be helpful when there is difcult access to the papillary
orice as seen. in patients with a juxtapapillary duodenal diverti-
culum or altered surgical anatomy. Push-type or sigmoid shaped
109
 SECTION 2 TECHNIQUES
Fig. 12.1 Smooth, tapered, rounded, polished tip of a sphinctero-
tome with colored markers that allow for determination of the
depth of insertion.
Fig. 12.2 Sphincterotome with a coated proximal portion of the
cutting wire avoids direct contact with the elevator of the duode-
noscope and to the duodenal wall.
A B
Fig. 12.3 A Approach to the papillary orice with the tip of a
sphincterotome. B Bowing the sphincterotome to allow its inser-
tion towards 11 oclock into the opening of the common bile duct.
sphincterotomes have been developed for EST in patients with Bill-
roth II anatomy.
Procedure
The papilla is usually approached with the sphincterotome from a
distance so that its precurved distal part is evident upon exiting
the endoscope. Alternatively the tip of the sphincterotome is gently
introduced into the papillary orice. A short, straight position of the
duodenoscope facilitates a precise control of the device. Subsequent
bowing of the tip usually allows its insertion towards 11 oclock into
the opening of the common bile duct (Figs 12.3A, 12.3B). Straighten-
ing the tip and gently withdrawing the endoscope results in further
anchoring the tip of the device within the common bile duct. It can
be then further advanced to achieve deep cannulation. Such a
smooth, direct approach may fail and the procedure can become
difcult and frustrating. Careful injection of contrast may allow
visualization of the biliary anatomy for targeted guidance of the tip
110
A B
Fig. 12.4 A Passage of a guidewire with a radiopaque hydrophilic
tip in the direction of the bile duct under uoroscopic guidance
prior to contrast injection. B Careful injection of dilute contrast
medium allows conrmation of common bile duct cannulation; a
lling defect in the distal bile duct is seen.
of the sphincterotome or insertion of a guidewire. However repeated
injections may induce edema of the papilla and increase the likeli-
hood of post-ERCP pancreatitis. Alternatively, a guidewire can be
gently passed in the direction of the bile duct under endoscopic and
uoroscopic guidance without contrast injection. For this approach
guidewires with a soft tip, preferably hydrophilic, should be used to
reduce the risk of ductal injury (Figs 12.4A, 12.4B). A recent large
randomized trial compared biliary cannulation with a guidewire
through a sphincterotome with a sphincterotome alone. The success
rates were comparably high for both methods but the rate of pan-
creatitis was signicantly lower in the guidewire group.5 However
all procedures were performed by a single endoscopist and these
results require conrmation in a multicenter trial. Cannulation of
the bile duct with a sphincterotome with or without use of a guide-
wire should succeed in approximately 90% of the cases. Failures are
mainly caused by a difcult access to the papilla because of anatomi-
cal variations or previous gastroduodenal surgery. Ampullary tumors
or impacted stones creating a bulging papilla can also impair the
approach to the papillary orice or deep ductal cannulation. For
these cases lifting the roof of the papilla with the adjustable tip of
the sphincterotome and use of a hydrophilic guidewire are helpful
in selective cannulation and controlled cutting (Figs 12.5A, 12.5B).
When cannulation fails, a variety of additional techniques includ-
ing precut sphincterotomy and rendezvous procedures have been
established. Their appropriate use allows access to the biliary system
in nearly all cases. However, the risk of complications increases in
particular when access papillotomy is performed, which should
therefore be limited to experienced endoscopists. Availability and
local expertise will determine how long to persist in difcult biliary
cannulation and when to switch to precut techniques.6 Details are
described in Chapters 8 and 9.
After deep cannulation has been conrmed by contrast injection,
a guidewire should be advanced to the proximal biliary system in
order to secure ductal access for subsequent maneuvers and exchange
of accessories. The tip of the sphincterotome is then slightly bowed
so that it is in contact with the roof of the papilla. Not more than
5 mm of the cutting wire should be inside the papilla so that only a
small amount of tissue will be cauterized. This approach improves
the cutting action and avoids a rapid large incision (zipper) which
can occur when pure cutting current is used. Newer electrosurgical
generators have largely eliminated this complication. Most sphinc-
terotomes have endoscopically visible markers at the distal part of
the catheter which allow one to determine the depth of insertion of

A B
C D
E F
G H
Fig. 12.5 A Bulging papilla due to an impacted bile duct
stone. B The orice of the papilla is located at the distal, bottom of
the bulging papilla: the roof is lifted by bending the sphincterotome
after tip insertion into the orice, the cutting wire is bowing towards
3 oclock. C Fluoroscopy indicating a short, straight position of the
duodenoscope corresponding to the endoscopic image of Figure
12.5B. D Fluoroscopy after rotation of the dial of the duodeno-
scope to the left while advancing the scope slightly toward the long
scope position; the corresponding endoscopic view is seen in Figure
12.5E. E Due to the maneuver shown in Figure 12.5D the papilla has
been placed to the left side and the cutting wire is bowing towards
11 oclock; only a few millimeters of cutting wire are inside the papilla
indicating an excellent position for cutting. F Sphincterotomy with
sequential cutting by use of the ENDO-CUT mode; the cut can be
extended a few more millimeters to the visible junction between the
duodenal wall and the intraduodenal portion of the papilla; the lumen
of the common bile duct can already be seen above the black marker
of the sphincterotome. G Fluoroscopy showing balloon extraction
of a small bile duct stone. The balloon is positioned just above the
stone. The stone is removed by pulling downwards in the direction
of the long axis of the common bile duct. H Endoscopic view after
extraction of the intact bile duct stone (located at the right side of the
papilla). The papilla is edematous after previous stone impaction.
Chapter 12 Biliary Sphincterotomy
the cutting wire into the bile duct (Fig. 12.1). It is generally believed
that orientation of the cutting wire in the range between the 11
oclock and 1 oclock position, reduces the likelihood of bleeding and
perforation. In spite of great efforts by the accessory manufacturers
to develop sphincterotomes with cutting wires that automatically cut
along these directions, sphincterotomes may still orient in the 3
oclock direction (Fig. 12.5B, 12.5C). To overcome this problem, the
papilla has to be placed on the left side and along the 10 or 11 oclock
position. This maneuver can be achieved by rotating the left-right
dial to the left while advancing the duodenoscope slightly into the
long endoscope position (Fig. 12.5D, 12.5E).
The choice of electrosurgical current for EST is a source of con-
troversy. The combination of high cutting current blended with a
low coagulation current is most frequently used. Creation of edema-
tous, extensively whitened or blackened tissue during EST is evi-
dence of suboptimal cutting and may predispose to post-procedural
pancreatitis or sphincter scarring. A previous trial indicated that
pure-cut electrocautery current is safer than blended current in
terms of post-ERCP pancreatitis without increasing the bleeding
rate.7 Similar results were obtained in a prospective, controlled trial
which found an incidence of mild post-EST pancreatitis in 3.2% of
patients who underwent EST with pure-cutting current compared to
12.9% in groups randomized to EST with blended current or pure-
cutting current initially followed by blended current.8 These dif-
ferences were statistically different, although the study has been
criticized because of the extraordinarily high incidence of pancreati-
tis after standard EST with blended current. In contrast, a recent
randomized, controlled trial in 246 patients did not demonstrate a
signicant difference in the frequency of post-procedural pancreati-
tis between those patients who underwent EST with pure-cutting
current or blended current.9 There was a signicant increase in
minor bleeding episodes in the pure-cut group with equal numbers
of delayed episodes of bleeding in each arm. In another series, the
incision was begun using cutting current and nished using blended
current but was not associated with a decrease in the risk of post-
ERCP pancreatitis.10 An alternative electrocautery option is the use
of the ENDO CUT mode of the ERBE electrosurgical generator in
which cutting and coagulation current are alternated by the intrinsic
software. A potential advantage of this method is a stepwise cutting
action which allows precise control of the direction and length of the
incision. This replaces the technique whereby current is applied in
short pulses controlled by foot pedal activation, and which is recom-
mended for pure-cut or blended current to reduce the risk of a
zipper cut. A large retrospective analysis suggested that the micro-
processor-controlled EST is associated with a signicantly lower
frequency of intra-procedural bleeding but had no impact on clini-
cally evident hemorrhage.11
The size of EST can vary and depends on the diameter of the
distal portion of the common bile duct as well as the indication for
sphincterotomy. During cutting, pressure on the papillary roof is
provided by upward lifting of the slightly bowed sphincterotome.
EST should be continued only when the wire can be clearly seen and
when it is directed between the 11 and 1 oclock position. Guidance
and repositioning of the device should be mainly controlled with the
tip and the shaft of the endoscope which is maneuvered with the
right hand of the operator like the handle of a knife. A small incision
seems to be appropriate for placement of an endoprosthesis in
malignant biliary obstruction whereas complete splitting of the
sphincter should be attempted for treatment of bile duct stones
and sphincter of Oddi dysfunction (SOD) to decrease the risk of
111
 SECTION 2 TECHNIQUES
A B
Fig. 12.6 A Papilla in a patient with bile duct stones; slight bulging
of the roof suggests that the proximal border of the intraduodenal
portion is above the mucosal fold. B After completion of EST the
lumen of the common bile duct can be seen above the inserted
guidewire; there does not appear to be any residual roof of the
papilla and further extension of cutting could cause duodenal
perforation.
recurrences and EST-related stenosis of the papilla. However, the
correlation between the length of cutting and the incidence of early
or late complications of EST has not been determined in formal
trials. Biliary sphincterotomy should be limited to the junction
between the duodenal wall and the intraduodenal portion of the
papilla of Vater which is sometimes difcult to determine since
there is no reliable endoscopic landmark. The incision should be
also nished if the inner lumen of the bile duct is completely visible
or the bended tip of the sphincterotome can be pulled through the
papilla without any resistance (Figs 12.5F12.5H, 12.6A, 12.6B). In
view of modern techniques of lithotripsy, a large hazardous EST is
rarely required. If a wide opening to the common bile duct is needed
it may be safer to perform a small to moderate sized incision and
then dilate the incision with a balloon catheter rather than to increase
the size of the opening by cutting (Figs 12.7A12.7D) (see Chapter
11). Compared with other indications, the risk of complications of
EST is usually low in patients with a dilated common bile duct and
in the presence of ductal stones, especially when the papilla is large
and protruding due to an impacted stone.
Extension of a previous biliary sphincterotomy may be required
for the treatment of recurrent bile duct stones or recurrence of
symptoms after SOD. The technique of EST in this setting does not
differ from that when EST is performed initially. Data from anec-
dotal reports and small case series were controversial regarding the
risk of hemorrhage after extension of a previous EST and suggested
that the incidence of bleeding was increased during a post-sphinc-
terotomy period of approximately one week due to a resultant
increased vascularity. However recent large prospective studies did
not nd that extension of a previous EST is an independent risk
factor for hemorrhage.12,13
EST in patients with difcult anatomy
Juxtapapillary duodenal diverticula are found in 1015% of patients
undergoing ERCP. Depending on the location of the papilla, can-
nulation of the bile duct can be difcult and may require special
techniques such as insertion of the tip of the endoscope into the
diverticulum, use of sphincterotomes with a long nose or pulling the
papilla out of a diverticulum with biopsy forceps or a second cathe-
ter. After successful deep cannulation it is strongly recommended
that EST is performed over a guidewire. This approach facilitates
112
A
C
D
Fig. 12.7 A Papillary orice after EST in a patient with multiple
giant bile duct stones. Mechanical lithotripsy was required followed
by removal of large fragments. B Balloon dilation of the papilla
after EST using an 18 mm diameter balloon. C Fluoroscopy cor-
responding to Figure 12.7B showing the fully inated balloon lled
with diluted contrast media. D Smooth extraction of large stone
fragments with a basket catheter through the widely open sphinc-
terotomy orice.
cutting in the direction of the bile duct which may otherwise be dif-
cult to determine due to the altered anatomy. Moderate bending of
the tip of the sphincterotome with a few millimeters of the cutting
wire inside the papilla exposes the papillary roof to allow for a con-
trolled incision. Any direction of the cutting wire towards the base
of the diverticulum should be avoided (Figs 12.8A, 12.8B). Data from
a recent large, retrospective analysis suggested that bleeding after
EST was an independent risk factor associated with juxtapapillary
duodenal diverticula.14
The approach to patients with post-surgical anatomy is discussed
further in Chapter 24. Use of a duodenosope in patients with Billroth
II anatomy allows a better visualization of the papillary roof and
improves the maneuverability of accessories because of the avail-
ability of the elevator. Precurved sphincterotomes should not be
used for initial cannulation of the bile duct in these patients because
they direct the tip of the catheter towards the pancreatic duct orice.
This is because the papilla is rotated 180 degrees as compared to
native anatomy. A straight, new ERCP cannula and a straight guide-
wire with a hydrophilic tip aim in the direction of the bile duct and
facilitate entry into the biliary orice. After successful placement of
guidewire, a rotatable, push-type or sigmoid shaped sphincterotome
can be used for EST. In spite of use of these special accessories the
correct direction of the cutting wire towards the desired 5 oclock
position (in this situation) can remain difcult because of the
reversed anatomic orientation. It is usually easier to place a 7 Fr
biliary endoprosthesis and to sever the papillary roof with a needle
knife incision using the endoprosthesis as a guide. A similar
approach is required when performing EST in the setting of
 Chapter 12 Biliary Sphincterotomy

A B

Fig. 12.8 A Papilla at a 6 oclock position on the rim of a juxtapapillary diverticulum; the bridge above the papilla inside the diverticulum
represents the papillary roof; there is no clear landmark of the proximal border of the intraduodenal portion. B The blue marker of the
sphincterotome is visible during EST and indicates that only a few millimeters of the cutting wire are inside the papilla; passing the sphinc-
terotome over a guidewire and inserting only a small amount of cutting wire into the duct during EST allows precise control of cutting;
the correct direction alongside the bridge will avoid cutting towards the base of the diverticulum; slight bulging of the papillary roof
above the wire indicates that the incision can be extended just beyond the mucosal fold.

loop gastrojejunostomy in patients with duodenal or pyloric
obstruction.
Reaching the papilla with a duodenoscope in patients with a
Roux-en-Y anatomy can be very difcult. The approach can be better
achieved with a pediatric colonoscope, a push enteroscope or a
double-balloon endoscope. However therapeutic interventions are
impaired because of the limited transmission of manipulations via
the long insertion tube and the limited array of available accessories
for these endoscopes.
If the papilla is inaccessible due to altered anatomy or can not be
cannulated even after precut techniques, the rendezvous method
should be considered. A percutaneous transhepatic tract is estab-
lished with placement of a 7 Fr catheter into the common bile duct.
Thereafter duodenoscopy is repeated. A percutaneously inserted
400 cm long guidewire is passed antegrade through the papilla and
grasped endoscopically with a snare passed through the endoscope.
In cases where there is a long afferent loop this technique allows
one to pull the tip of the endoscope towards the papilla by applying
percutaneous tension on the guidewire while simultaneously apply-
ing tension on the wire exiting the endoscope. The sphincterotome
is passed over the guidewire for subsequent EST. In selected cases,
the sphincterotome can be passed percutaneously for the perfor-
mance of antegrade sphincterotomy under endoscopic retrograde
visualization.15 In rare cases, even the rendezvous method does not
allow one to endoscopically reach the papilla. Antegrade sphin-
cterotomy under percutaneous transhepatic cholangioscopic and
uoroscopic guidance can be performed, though this technique is
potentially hazardous and should be restricted to centers with large
expertise in percutaneous transhepatic interventions.
Alternatives to EST
Balloon sphincteroplasty is discussed in detail in Chapter 11. Com-
pared with EST, endoscopic balloon dilation of the biliary sphincter
(EBD) offers the theoretical advantage of sphincter preservation in
particular in young patients with bile duct stones. A meta-analysis
of several randomized, controlled trials of EBD versus EST for biliary
stones showed no signicant difference between the success rates
(94.3% vs 96.5%) and overall complication rates (10.5% vs 10.3%) of
either technique.16 Compared to EST, EBD caused less bleeding (0%
vs 2.0%, p = 0.001) but a higher rate of post-ERCP pancreatitis (7.4%
vs. 4.3%, p = 0.05). In addition, patients undergoing EBD were more
likely to require mechanical lithotripsy for bile duct clearance (20.9%
vs 14.8%, p = 0.014). This meta-analysis was performed prior to a
recent publication of a large prospective US multicenter randomized
trial comparing EBD with EST for extraction of bile duct stones.17
The success rates were comparably high in both groups but the
overall morbidity (17.9% vs 3.3%) and severe morbidity (6.8% and
0%) were signicantly higher in patients who had undergone EBD.
There were two deaths (1.7%) due to pancreatitis following dilation
and none for sphincterotomy. The complication rate of EBD was
higher in the US trial than in previous series, which may be explained
by a selection of younger patients who had a mean age of 47 years.
Technical details of EBD may also inuence the clinical outcome.
The procedure has not yet been standardized in terms of the ina-
tion pressure, the duration of ination, and the number of dilations.
The biliary sphincter function may be preserved but the clinical rel-
evance of this potential advantage remains undetermined. There are
limited data on the long-term outcome after EBD. According to a
recent trial the frequency of stone recurrence was higher after EBD
compared to EST at mid-term evaluation, but over the long term the
estimated probability of stone recurrence tended to be higher in
patients who had undergone EST.18
Based on these data, EBD can not be recommended as a routine
procedure at this time. However, it should be considered as a valu-
able alternative to EST in patients with coagulopathies and those in
whom the endoscopic approach is difcult (e.g. Billroth II anatomy
or a duodenal diverticulum).19 For each individual patient, the risks
of pancreatitis and bleeding must be weighed when choosing
between EBD and EST, especially in younger patients.
INDICATIONS
Well-established indications for EST are common bile duct stones,
acute cholangitis, severe acute biliary pancreatitis, palliation of
ampullary malignancies, facilitation of biliary stent placement, and
treatment of SOD (Box 12.1). A variety of randomized, controlled

113
 SECTION 2 TECHNIQUES
BOX 12.1 INDICATIONS FOR BILIARY
ENDOSCOPIC SPHINCTEROTOMY
Common bile duct stones
Palliation of obstruction due to malignant ampullary neoplasm
as alternative to stent placement (selected cases)
Facilitation of biliary stent placement (especially multiple stents)
for malignant or benign common bile duct obstruction
Sphincter of Oddi dyskinesia (SOD), benign papillary stenosis
Biliary leaks
Miscellaneous conditions (choledochocele, sump syndrome,
biliary parasites)
Access for peroral choledochoscopya
During endoscopic resection of ampullary neoplasmsa
Access for cannulation of the pancreatic duct after failure of
standard cannulation techniquesa
a
Poor evidence to support sphincterotomy for these indications.
trials demonstrated the efcacy and safety of EST for these biliary
diseases. Choledocholithiasis is still one of the major indications for
biliary sphincterotomy in order to allow endoscopic stone extraction
using basket or balloon catheters. The success rate of ductal clear-
ance for standard procedures is approximately 90% depending on
patient selection. Adjunctive techniques for intracorporeal or extra-
corporeal lithotripsy further increase the clearance rates.20 A recent
trial demonstrated that EST can be safely and effectively performed
for the treatment of bile duct stones even in patients 90 years of age
or older.21 Previous restrictions of EST to elderly patients or those
with previous cholecystectomy are, however, no longer valid. In a
prospective, US multicenter trial on EST a group of 487 patients
underwent sphincterotomy for bile duct stones within 30 days of
laparoscopic cholecystectomy. They were signicantly younger (on
average 51 versus 64 years) and the common bile duct was smaller
in diameter (8.7 mm vs 10.0 mm) compared to a group of 1113
patients with their gallbladder in situ or with previous cholecystec-
tomy who received EST for the same indication. The complication
rate of EST was signicantly lower in the former group (4.9% versus
9.5%).12 These results demonstrate that EST can be safely performed
in young patients with bile duct stones shortly before or after lapa-
roscopic cholecystectomy.
Acute cholangitis due to choledocholithiasis or ductal stenosis
can be effectively treated by EST in conjunction with additional
procedures such as removal of ductal stones or placement of drain-
age catheters or endoprostheses. Early EST has also been established
in patients with severe acute biliary pancreatitis (see Chapter 31). A
recent meta-analysis of four randomized, controlled trials showed
signicantly lower morbidity and mortality rates in patients who
underwent EST compared to those who were treated conservatively.22
According to this trial 8 patients need to be treated by EST to avoid
one severe complication and 26 to prevent one death.
Another established indication of EST is biliary sphincterotomy
as an initial therapeutic step before dilation and stent placement for
palliation of malignant biliary obstruction. However sphincterotomy
is not obligatory for this indication unless multiple large bore stents
are inserted (see Chapter 16). Nevertheless a small cut seems to be
114
safe and is frequently performed to facilitate access to the biliary
system for scheduled exchanges of plastic prostheses.
Biliary sphincterotomy has become the method of choice for
treatment of patients with documented SOD (see Chapter 34). Can-
nulation and cutting can be more difcult in patients with SOD
compared to other indications for sphincterotomy due to small size
of the papilla and a narrow orice. Atraumatic cannulation of the
papilla, careful manipulation of accessories, and precise control of
the sphincterotomy incision while cutting over a guidewire is man-
datory to minimize trauma of the tissue. For these reasons, and
because of the increased risks, only experienced endoscopists should
perform EST for SOD.
The level of evidence is lower for a variety of other indications for
EST that are listed in Box 12.1. Most of these indications have been
established on the basis of prospective, uncontrolled trials or appro-
priate retrospective analyses. Randomized, controlled trials seem to
be difcult to perform for many of these biliary disorders for a
variety of reasons.23,24
CONTRAINDICATIONS
Contraindications to ERCP and EST include an uncooperative or
unstable patient, inability of the patient to provide informed consent,
uncorrected coagulopathy, and a newly created gastrointestinal
anastomosis. Contrast hypersensitivity is not considered as a contra-
indication to EST, but prophylactic intravenous application of
corticosteroids may be considered. Preprocedure coagulation studies
are strongly recommended and coagulopathy must be corrected
before sphincterotomy. The presence of liver cirrhosis and use of
aspirin or other nonsteroidal anti-inammatory drugs do not appear
to be important predictors of bleeding.12 However antiplatelet drugs
such as clopidogrel and ticlopidin should be interrupted for at least
seven days before elective sphincterotomy depending on the indi-
vidual clinical risks. EST using a pull sphincterotome should not be
performed if proper positioning of the sphincterotome with its tip
in the bile duct can not be documented. Cutting should be avoided
if the position of cutting wire can not be seen or if the tip of the
sphincterotome is bowing in a the wrong direction because of dif-
cult anatomy. If these problems can not be resolved with changing
the position of the device or other maneuvers, then balloon dilation
of the biliary sphincter should be considered as an alternative to EST.
The indication for EST should be reconsidered for the individual
case if the level of evidence is fair or poor.
COMPLICATIONS AND THEIR MANAGEMENT
Chapter 6 covers complications of ERCP in detail. A large, prospec-
tive US multicenter trial was published on early complications fol-
lowing EST (Table 12.1). In this trial Freeman et al. reported a total
complication rate of 9.8% in 2347 patients.12 Acute pancreatitis was
the most frequent major complication of EST and was seen in 5.4%
of all cases. Hemorrhage, perforation, cholangitis and cholecystitis
occur less frequently and may have decreased compared with pre-
vious reports. Other prospective, multicenter trials of ERCP and
related risk factors have been published, though in contrast to the
study by Freeman et al. diagnostic procedures were also included.
The reported data do not allow a separate analysis of the EST-related
morbidity.2527
Various risk factors, prophylactic measurements, early recogni-
tion and appropriate treatment should be considered in order to
 Chapter 12 Biliary Sphincterotomy

Type of Incidence Severe Fatal Adjusted odds ratioa (95% CI)

complication (%) complications complications
Anticoagulation 3 days after EST 5.1 (1.616.7)
Pancreatitis 5.4 0.4 <0.1 Coagulopathy before EST 3.3 (1.57.2)
Hemorrhage 2.0 0.5 0.1 Cholangitis before EST 2.6 (1.44.9)
Perforation 0.3 0.2 <0.1
Mean case volume of the endoscopist
Cholangitis 1.0 0.1 <0.1
1 EST / week 2.2 (1.14.2)
Cholecystitis 0.5 0.1 <0.1
Bleeding during EST 1.7 (1.22.7)
Miscellaneous 1. 1 0.3 0.2
Total 9.8 1 .6 0.4
Table 12.3 Risk factors for EST-related hemorrhage12
a
Signicant in a multivariate analysis.
Table 12.1 Complications of biliary endoscopic sphincterotomy in
2347 patients12

Authors Freeman et al. (12) Rabenstein et al. (31)

Adjusted odds ratioa (95% CI) Patients for EST 2347 1335

Suspected SOD 5.1 (2.79.2) Complication rates

Precut EST 4.3 (1.710.9) Endoscopists case volume
Difculty of cannulation 2.4 (1.15.4) Low (1 per week) 11.1%a 9.3%b
Younger age 2.1 (1.43.3) High (>1 per week) 8.4%a 5.6%b
Repeated pancreatic duct injection 1.4 (1.01.8)
Table 12.4 Complications related to endoscopists case volume
Table 12.2 Risk factors for EST related pancreatitis12 a,b
p < 0.05.
a
Signicant in a multivariate analysis.

analysis of ve controlled trials. A total of 481 patients were included

decrease the risks of EST. In an individual patient it may be difcult
to determine if complications were caused by EST, by bile duct can-
nulation or by adjunctive therapeutic interventions.
EST-related pancreatitis
The denition of pancreatitis varies between different studies.12,25,26
According to a consensus denition, ERCP-related pancreatitis is
diagnosed in patients with new or worsened abdominal pain and a
serum amylase or lipase that is three or more times the upper limits
of normal 24 hours after the procedure and requires at least two days
of hospitalization.28 In most series the reported rates of post-ERCP
pancreatitis range from 1% to 7%.2527 EST-related risk factors for
pancreatitis that were identied in a multivariate analysis of the
study by Freeman et al. are summarized in Table 12.2.12 These
parameters and preventive measurements should be considered
when patients are selected for EST to reduce the risk of unnecessary
pancreatitis.
Pharmacologic prophylaxis may be considered in high-risk can-
didates for ERCP. Data from a recent meta-analysis suggests that the
incidence of pancreatitis after ERCP can be reduced by the admin-
istration of either somatostatin or gabexate mesilate.29 In spite of
these results pharmacologic prophylaxis is rarely performed in clini-
cal practice because of its limited efcacy and high costs. In addition,
a high number of cases have to be treated to avoid one case of
pancreatitis.
A variety of electrophysical factors can inuence the efcacy and
safety of EST. They include different waveforms and power settings
of modern electrosurgical generators but also the cutting wire
contact length. The clinical relevance of these factors is less clear as
was previously discussed in this chapter. In view of conicting
results the selection of electrosurgical current for biliary EST should
be based primarily on endoscopist preference.
The impact of prophylactic pancreatic stent placement on pre-
vention of post-ERCP pancreatitis was recently studied in a meta-
who had at least one of the following risk factors for pancreatitis:
suspected and/or conrmed SOD, difcult cannulation, or EBD for
stone extraction within the same procedure. The incidence of post-
ERCP pancreatitis was signicantly lower in patients who had
received a temporary pancreatic stent compared to those in whom a
stent was not placed (5.8% vs 15.5%, p = 0.003).30 The analysis indi-
cated that 10 patients have to be treated to avoid one case of pancre-
atitis. Although these data do not allow a separate analysis for EST,
pancreatic stent placement is strongly recommended for patients
undergoing EST with risk factors for post-ERCP pancreatitis.
Treatment of EST-related pancreatitis does not differ from the
management of pancreatitis of other etiologies. It includes short-
term parenteral nutrition, antibiotics in cases of pancreatic necrosis
determined by CT scan, analgesia, and management of related
complications. Repeat ERCP should be considered in patients
with residual bile duct stones, ongoing obstructive jaundice or
cholangitis.
EST-related hemorrhage
Clinically relevant hemorrhage can be dened as the presence of
melena, hematochezia, or hematemesis associated with a hemoglo-
bin decrease of at least 2 g/dl or the need for blood transfusions. The
incidence of EST-related bleeding in prospective trials ranges from
0.8% to 2%.12,25,26 In approximately half of the cases hemorrhage is
delayed and occurs at 24 hours, though it can occur up to one week
or more after EST. Risk factors for hemorrhage include coagulopa-
thy before EST, anticoagulation within three days after EST, cholan-
gitis before EST, and bleeding during EST (Table 12.3).12 In addition,
EST performed by endoscopists who perform less than approxi-
mately one EST per week is associated with a higher rate of bleeding
as compared to those performed by endoscopists with higher EST
volumes. The impact of the expertise of the endoscopist on the pro-
cedure-related morbidity was also reported in a recent retrospective
analysis of 1335 ESTs (Table 12.4).31 A multivariate analysis of an

115
 SECTION 2 TECHNIQUES

Italian multicenter trial found precut procedures and obstruction of

the papillary orice were risk factors for EST-related hemorrhage.25
The pattern of bleeding following EST during the procedure did not
seem to predict the risk of late bleeding.32
Oozing bleeding after EST frequently stops spontaneously and
further therapeutic interventions can usually be performed without
treatment of bleeding. However, ongoing hemorrhage and/or pul-
satile bleeding, which arises from an aberrant branch of the retro-
duodenal artery, require endoscopic hemostasis. Repeated duodenal
irrigation is mandatory for endoscopic localization of the bleeding
site and to facilitate application of appropriate hemostatic interven-
tions. The cutting wire of a sphincterotome can be used to apply
pure coagulation to the apex of the bleeding site. Endoscopic injec-
tion of saline-epinephrine solution or brin glue at the proximal
edge of the incision site is usually effective in achieving hemostatis
in the setting of severe hemorrhage.33,34 Multipolar or bipolar elec-
trocoagulation is another option to treat hemorrhage but should be
carefully performed at an appropriate distance from the pancreatic
orice; alternatively pancreatic drainage can be secured with a pan-
creatic duct stent. Placement of conventional hemoclips through a
side-viewing endoscope is technically difcult but can be performed
with a therapeutic channel duodenoscope.35 Bending of the tip and
lifting of the elevator of the instrumentation channel should be
minimized to allow release of the clips from the rigid applicator tip.
Placement of a biliary endoprosthesis or a nasobiliary drain should
be considered if interventions used to treat bleeding cause obstruc- Fig. 12.9 Plain abdominal x-ray showing air in the retroperitoneal
tion of the common bile duct. space parallel to the spine; perforation was caused by a forceful
stone extraction after fracture of a basket during mechanical litho-
In rare cases endoscopic management of EST-related hemor- tripsy; the distal part of the broken basket can be seen in the
rhage fails; angiographic transcatheter embolization or even lapa- common hepatic duct; lling of the biliary system and the duode-
rotomy may be required, though the latter is associated with num with contrast injection via a nasobiliary probe demonstrates
signicant morbidity and mortality. no leakage and justies a conservative therapeutic approach.

EST-related perforation
Perforation of the bile duct or the pancreatic duct with the sphinc-
terotome, guidewire, balloon, or other accessories can usually be
managed by endoscopic or percutaneous placement of drainage
catheters or stents into the bile duct and pancreatic duct. EST-related
retroduodenal perforations are uncommon and mainly caused by
zipper cutting which can be avoided by limited insertion of the
cutting wire into the papilla and use of modern controlled-cut elec-
trosurgical generators. Perforation occurs when EST is performed
beyond the duodenal wall. Reported perforation rates for ERCP are
0.3% to 0.6% and those related specically to EST are 0.3%.12,36,37 A
univariate analysis of a recent retrospective trial indicated that EST,
SOD, a dilated common bile duct and biliary stricture dilation are
risk factors for perforation.37 Most cases are diagnosed during the
ERCP procedure by the nding of free intra-abdominal or retroperi-
toneal air on uoroscopy, or post-procedurally on plain abdominal
x-rays. Clinical presentation is variable and can be mild. Perforation
should be considered in case of ongoing abdominal pain, signs of
peritonitis, fever, leukocytosis, and an elevated C-reactive protein.
Abdominal CT scan with luminal contrast into the duodenum is the
method of choice for the diagnosis and stratication for manage-
ment. Conservative treatment with temporary parenteral nutrition
and administration of antibiotics is appropriate if an ongoing leak
is excluded. Otherwise an interdisciplinary approach should be
undertaken.38 A nasoduodenal tube and a nasobiliary or percutane-
ous transhepatic drain are useful to prevent gastric, pancreatic, and
biliary uid from entering into the retroperitoneal space (Fig. 12.9).
Percutaneous drainage with large-bore tubes is indicated in cases of
abscess formation in the retroperitoneum. Laparotomy is usually
required when these measurements fail and/or there is evidence of
sepsis or peritonitis.
EST-related cholangitis
Prophylactic administration of antibiotics to prevent cholangitis is
recommended in patients with incomplete biliary drainage, particu-
larly those patients with proximal stenoses due to hilar tumors.
Nasobiliary catheters or endoprostheses should be placed if there is
any evidence of incomplete bile duct clearance after EST or adjunc-
tive interventions for biliary stones. Repeated ERCP may be needed
in those cases of delayed cholangitis due to biliary obstruction. Con-
tinuous bile duct irrigation via nasobiliary drains may be useful for
management of patients with purulent cholangitis.
Long-term consequences of EST
Five studies comprising a large number of patients, a high follow-up
rate, and follow-up periods of more than six years after EST, were
recently reviewed.39 The overall rate of late symptoms which can be
attributed to EST ranges from 6% to 24%, with the rate being nearly
10% in three of the ve trials. Common bile duct stones and papil-
lary stenosis were the most common complications. Stones can
usually be removed after extension of the previous EST or after
balloon dilation. Papillary stenosis can be managed by extension of
the sphincterotomy. Cautery-induced distal bile duct strictures can
be managed by balloon dilation and placement of one or more biliary
stents.40 Cholangitis caused by reux of duodenal contents into the

116

biliary system is rare but can occasionally require creation of a
surgery for a biliodigestive anastomosis. A correlation between the
size of EST and these late complications can not be determined from
the current literature. Concerns about long-term biliary carcinogenic
risks following sphincterotomy were not demonstrated in a large
case controlled Scandinavian study.41
RELATIVE COSTS
The costbenet ratio of EST depends on its indication. In view of
high success rates, low procedure-related morbidity, and frequent
REFERENCES
1. Laasch HU, Tringali A, Wilbraham L, et al. Comparison of standard
and steerable catheters for bile duct cannulation in ERCP.
Endoscopy 2003; 35:669674.
2. Schwacha H, Allgaier HP, Deibert P, et al. A sphincterotome-
based technique for selective transpapillary common bile duct
cannulation. Gastrointest Endosc 2000; 52:387391.
3. Cortas GA, Mehta SN, Abraham NS, et al. Selective cannulation of
the common bile duct: a prospective randomized trial comparing
standard catheters with sphincterotomes. Gastrointest Endosc
1999; 50:775779.
4. Canard JM, Cellier C, Houcke P, et al. Prospective multicenter
study comparing a standard reusable sphincterotome with a
disposable triple-lumen sphincterotome. Gastrointest Endosc
2000; 51:704707.
5. Lella F, Bagnolo F, Colombo E, et al. A simple way of
avoiding post-ERCP pancreatitis. Gastrointest Endosc 2004;
59:830834.
6. Tang SJ, Haber GB, Kortan P, et al. Precut papillotomy versus
persistence in difcult biliary cannulation: a prospective
randomized trial. Endoscopy 2005; 37:5865.
7. Elta GH, Barnett JL, Wille RT, et al. Pure cut electrocautery current
for sphincterotomy causes less post-procedure pancreatitis than
blended current. Gastrointest Endosc 1998; 47:149153.
8. Stefanidis G, Karamanolis G, Viazis N, et al. A comparative study of
postendoscopic sphincterotomy complications with various types
of electrosurgical current in patients with choledocholithiasis.
Gastrointest Endosc 2003; 57:192197.
9. MacIntosh DG, Love J, Abraham Ns. Endoscopic sphincterotomy
by using pure-cut electrosurgical current and the risk of post-
ERCP pancreatitis: a prospective randomized trial. Gastrointest
Endosc 2004; 60:551556.
10. Gorelick A, Cannon M, Barnett J, et al. First cut, then blend: a
electrocautery technique affecting bleeding at sphincterotomy.
Endoscopy 2001; 33:976980.
11. Perini RF, Sadurski R, Cotton PB, et al. Post-sphincterotomy
bleeding after the introduction of microprocessor-controlled
electrosurgery: does the new technique make the difference?
Gastrointest Endosc 2005; 61:5357.
12. Freeman ML, Nelson DB, Sherman S, et al. Complications
of endoscopic biliary sphincterotomy. NEJM 1996; 335:
909918.
13. Mavrogiannis C, Liatos C, Papanikolaou IS, et al. Safety of
extension of a previous endoscopic sphincterotomy: a
prospective study. Am J Gastroenterol 2003; 98:7276.
14. Zoepf T, Zoepf DS, Arnold JC, et al. The relationship between
juxtapapillary duodenal diverticula and disorders of the
biliopancreatic system : analysis of 350 patients. Gastrointest
Endosc 2001; 54:5661.
Chapter 12 Biliary Sphincterotomy
avoidance of surgery EST seems to be cost-effective. The choice
of reusable or disposable sphincterotomes can have a signicant
impact on procedural cost but this practice varies from institution
to institution and country to country.4 The use of simple and
inexpensive devices is not necessarily cost-effective if these devices
have a high failure rate and need to be exchanged for sphinctero-
tomes that allow more options such as guidewire insertion and
contrast injection. Depending on the local situation limiting
therapeutic ERCP to high volume centers should be cost-effective
because of the associated higher success rates and lower complica-
tion rates.12,31
15. Born P, Sandschin W, Rsch T. Percutaneous antegrade
sphincterotomy under endoscopic retrograde control: report of
two cases. Endoscopy 2002; 34:512513.
16. Baron TH, Harewood GC. Endoscopic balloon dilation of the
biliary sphincter compared to endoscopic biliary sphincterotomy
for removal of bile duct stones during ERCP: a metaanalysis of
randomized, controlled trials. Am J Gastroenterol 2004;
99:14551460.
17. DiSario JA, Freeman ML, Bjorkman DJ, et al. Endoscopic balloon
dilation compared with sphincterotomy for extraction of bile duct
stones. Gastroenterology 2004; 127:12911299.
18. Tanaka S, Sawayama T, Yoshioka T. Endoscopic papillary balloon
dilation and endoscopic sphincterotomy for bile duct stones:
long-term outcomes in a prospective randomized controlled trial.
Gastrointest Endosc 2004; 59:614618.
19. Bergman JJGHM, van Berkel AM, Bruno MJ, et al. Is endoscopic
balloon dilation for removal of bile duct stones associated with
an increased risk for pancreatitis or a higher rate of
hyperamylasemia? Endoscopy 2001; 33:416420.
20. Neuhaus H. Endoscopic and percutaneous treatment of difcult
bile duct stones. Endoscopy 2003; 35:S31S43.
21. Sugiyama M, Atomi Y. Endoscopic sphincterotomy for bile duct
stones in patients 90 years of age and older. Gastrointest Endosc
2000; 52:187191.
22. Sharma VK, Howden CW. Metaanalysis of randomized controlled
trials of endoscopic retrograde cholangiography and endoscopic
sphincterotomy for the treatment of acute biliary pancreatitis. Am
J Gastroenterol 1999; 94:32113214.
23. Costamagna G, Pandol M, Mutignani M, et al. Long-term results
of endoscopic management of postoperative bile duct strictures
with increasing numbers of stents. Gastrointest Endosc 2001;
54:162168.
24. Bergman JJGHM, Burgemeister L, Bruno MJ, et al. Long-term
follow-up after biliary stent placement for postoperative bile duct
stenosis. Gastrointest Endosc 2001; 54:154161.
25. Masci E, Toti G, Mariani A, et al. Complications of diagnostic and
therapeutic ERCP : a prospective multicenter study. Am J
Gastroenterol 2001; 96:417423.
26. Loperdo S, Angelini G, Benedetti G, et al. Major early
complications from diagnostic and therapeutic ERCP: a
prospective multicenter study. Gastrointest Endosc 1998;
48:110.
27. Freeman ML, DiSario JA, Nelson DB, et al. Risk factors for post-
ERCP pancreatitis: a prospective, multicenter study. Gastrointest
Endosc 1998; 48:110.
28. Cotton PB, Lehman G, Venes J, et al. Endoscopic sphincterotomy
complications and their management: an attempt at consensus.
Gastrointest Endosc 1991; 37:383393.
117
 SECTION 2 TECHNIQUES
29. Andriulli A, Leandro G, Niro G, et al. Pharmacologic treatment can
prevent pancreatic injury after ERCP: a meta-analysis. Gastrointest
Endosc 2000; 51:17.
30. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic
stent placement reduce the risk of post-ERCP acute pancreatitis?
A meta-analysis of controlled trials. Gastrointest Endosc 2004;
60:544550.
31. Rabenstein T, Schneider HT, Nicklas M, et al. Impact of skill and
experience of the endoscopist on the outcome of endoscopic
sphincterotomy techniques. Gastrointest Endosc 1999;
50:628636.
32. Wilcox CM, Canakis J, Mnkemller MD, et al. Patterns of bleeding
after endoscopic sphincterotomy. The subsequent risk of
bleeding, and the role of epinephrine injection. Am J
Gastroenterol 2004; 99:244248.
33. Vsconez C, Llach J, Bordas JM, et al. Injection treatment of
hemorrhage induced by endoscopic sphincterotomy. Endoscopy
1998; 1:3739.
34. Matsushita M, Hajiro K, Takakuwa H, et al. Effective hemostatic
injection above the bleeding site for uncontrolled bleeding after
endoscopic sphincterotomy. Gastrointest Endosc 2000;
51:628636.
118
35. Baron TH, Norton ID, Herman L. Endoscopic hemoclip placement
for post-sphincterotomy bleeding. Gastrointest Endosc 2000;
52:662.
36. Mallery JS, Baron TH, Dominitz JA, et al. Complications of ERCP.
Gastrointest Endosc 2003; 57:633638.
37. Enns R, Eloubeidi MA, Mergener K, et al. ERCP-related
perforations: risk factors and management. Endoscopy 2002;
34:293298.
38. Stapfer M, Selby R, Stain S, et al. Management of duodenal
perforation after endoscopic retrograde
cholangiopancreatography and sphincterotomy. Ann Surg 2000;
232:191198.
39. Prat F. The long term consequences of endoscopic
sphincterotomy. Acta Gastro-Enterologica Belgica 2000;
63:395396.
40. Pozsar J, Sahin P, Laszlo F, et al. Endoscopic treatment of
sphincterotomy-associated distal common bile duct strictures by
using sequential insertion of multiple plastic stents. Gastrointest
Endosc 2005; 62:8591.
41. Karlson BM, Ekbom A, Arvidsson D, et al. Population based study
of cancer risk and relative survival following sphincterotomy for
stones in the common bile dcut. Br J Surg 1997; 84:12351238.
 SECTION 2 TECHNIQUES
Chapter
Stone Extraction
13 Suyi Chang and Joseph W. Leung
INTRODUCTION
Following endoscopic sphincterotomy, the majority of stones <1 cm
will pass spontaneously.1 Nonetheless, it is current standard practice
to attempt stone extraction and clear the bile duct to avoid stone
impaction and subsequent risk of cholangitis. Extraction of stones
can usually be achieved with balloon catheter or wire basket.
However, large stones, particularly those over 2 cm, may be difcult
to remove and will require stone fragmentation prior to removal with
baskets and/or balloons. Methods of stone fragmentation include
mechanical lithotripsy, intraductal electrohydraulic or laser litho-
tripsy, and extracorporeal shock wave lithotripsy (ESWL). In the case
where endoscopic stone extraction fails, surgery or chemical dissolu-
tion of the stone25 should be considered. Permanent stenting can
be used as an alternative for biliary drainage in cases of large un-
extractable stones to prevent cholangitis.611
Initial cholangiograms are usually obtained using 60% (normal)
contrast, and early lling images should be carefully analyzed
for stones, which are often seen as lling defects with a meniscus
sign (Fig. 13.1A, 13.1B). However, if the bile duct is dilated, 30%
(half normal) contrast should be used as a large amount of dense
contrast can mask the small stones in a dilated bile duct. In patients
with suspected intrahepatic stones or stone above a stricture,
an occlusion cholangiogram may be necessary to inject contrast
above the stricture/obstruction to visualize the stone(s) (Fig. 13.1C).
There is however a risk of causing cholangitis if contrast is
injected into an obstructed system increasing the intrabiliary
pressure.
In order to achieve successful stone extraction, it is of prime
importance to assess the stone size relative to the size of the sphinc-
terotomy and distal common bile duct (i.e. the exit passage). The
sphincterotomy should be of adequate size to allow passage of the
stone. One method of gauging sphincterotomy size is to pull a fully
bowed sphincterotome across the cut papilla. A generous sphincter-
otomy should allow easy passage of the bowed sphincterotome. In
addition, appropriate accessories should be available to handle any
foreseeable complications.
Dilation of a bile duct stricture (see Chapter 30) may be necessary
for stones that occur above a bile duct stricture or intrahepatic
stones. This can be achieved using Gruntzig type balloons, which
are low compliance balloons that can be inated to a xed diameter.
Dilute contrast should be used for inating the balloon to a pre-
determined pressure as recommended by the manufacturer. The
balloon has radiopaque markers that help with the positioning, and
the maximum diameter of the balloon used should be gauged based
on the diameter of the normal portion of the bile duct. The balloon
is inated and the presence or disappearance of the waist formation
on the balloon is noted. This will determine the ease of stone extrac-
tion through the stricture. If the stricture cannot be opened, stone
fragmentation is necessary prior to removal. Balloon dilation or
sphincteroplasty after a small initial sphincterotomy has been used
to facilitate removal of a large stone and to avoid the risk of bleeding
and perforation from a large sphincterotomy.
BALLOON STONE EXTRACTION
Key points
Bile duct stones may be removed by inating a balloon catheter
above the stone and pulling back the catheter until the stone
reaches the ampulla. The stone is then expelled by traction and
downward deection of scope tip.
Ensure that an adequate sphincterotomy is performed for ease of
stone extraction.
If indicated, size the exit passage before attempted stone extraction
to avoid stone impaction
Monitor balloon size during traction removal of the stone.
Avoid overinating the balloon to minimize false resistance.
Indications/contraindications
Common bile duct stones and intrahepatic stones may be removed
using a balloon catheter.
Useful for removing small to medium stones.
Large stones may be difcult to remove with a balloon and may
result in stone impaction.
Complications
Balloon rupture may occur.
Impacted stone due to relative inadequacy of the sphincterotomy.
Introduction
Extraction balloons are available in different sizes, ranging from 8
to 15 mm. The size of the balloon can be adjusted by injecting a
varying amount of air to ll the balloon and regulating the volume
with a two-way stopcock (Fig. 13.2A). It may be helpful also to size
the exit passage prior to stone extraction in order to avoid stone
impaction. The balloon is inated to the widest diameter of the
common bile duct below the stone and pulled back gently to deter-
mine if there is any resistance to traction removal of the balloon,
noting any signicant deformity of the balloon at the same time. In
cases of chronic pancreatitis where the retropancreatic portion of the
bile duct may be compressed or xed, the balloon may deform or
become sausage shaped rather than retaining its rounded appear-
ance. This may indicate likely resistance to stone passage. Triple
lumen balloon allows the catheter to go over a guidewire and main-
tain access to the biliary system, in addition to injection of contrast
(Fig. 13.2B). However, triple-lumen balloon shafts may be stiffer
than regular double-lumen balloons. The tip of the balloon catheter
119
 SECTION 2 TECHNIQUES
A B
Fig. 13.1 A Irregular distal bile duct stone, seen as a lling defect. Denser contrast below stone demonstrates a good meniscus
sign. B Distal common bile duct stone formed around a surgical clip. C Occlusion cholangiogram showing inated balloon in mid-
CBD, normal gallbladder and a segment of right hepatic duct lled with stones.
A
B
Fig. 13.2 A Variable diameter stone extraction balloon. Balloon
(courtesy of Boston Scientic, Natick, MA) is inated to 12 mm in
diameter and held in position with 2-way stopcock. B Triple
lumen stone extraction balloon (courtesy of Cook Endoscopy,
Winston-Salem, NC) showing balloon inated and held in position
with two-way stopcock, with a separate lumen for injection of
contrast and passage of guidewire.
120
C
may be curled gently before introduction into the scope to facilitate
cannulation.
Description of technique
Once the catheter is within the bile duct, the balloon is inated above
the stone and pulled back gently until the stone is at the level of the
papilla. The scope should be aligned so that the axis of traction is in
the same axis as the bile duct. The tip of the scope is then angled
upwards against the sphincterotomy. While maintaining gentle trac-
tion on the balloon catheter at the level of the biopsy valve, the tip
of the scope is deected downwards, expelling the stone from the
sphincterotomy (Fig. 13.3). If there is resistance, the tip of the scope
is again angled upwards with steady traction applied to the catheter,
and the ip down movement is repeated to remove the stone. It may
be necessary to maintain the traction on the balloon as the stone is
slowly eased out of the bile duct. If necessary, the scope tip can be
angled downwards and rotated to the right to exert more traction
force to expel the stone. It is important to remember that an overin-
ated balloon may give rise to resistance as it is being pulled down,
and it may be necessary to deate the balloon slowly to conform to
the size of the bile duct. If a number of stones are present, it would
be appropriate to remove the most distal stone rst and then work
up the bile duct.
Care should also be taken to avoid overinating the balloon in
the bile duct as this will stretch the bile duct and cause pain to the
patient. The balloon should be inated or deated accordingly to
adjust the size to t the diameter of the bile duct. In cases where the
balloon goes over a short guidewire which is locked to the scope,
excessive scope movement during stone extraction may dislodge the
wire. It may be necessary to pull the balloon gently and avoid excess
tip deection of the scope to prevent dislodging the guidewire. An
alternative is to insert more guidewire into the bile duct to maintain

Fig. 13.3 Stone extracted with a balloon.
access before stone extraction. Nonetheless, it should be easy
to recannulate the bile duct in the presence of an adequate
sphincterotomy.
Indications/contraindications
The advantage of using an extraction balloon is that the inated
balloon fully occludes the lumen of the bile duct, facilitating removal
of small stones and debris. In addition, an occlusion cholangiogram
can be performed at the same time to ensure complete clearance of
the bile duct. Furthermore, the balloon catheter can be inserted over
a guidewire, allowing access to the intrahepatic ducts and removal
of intrahepatic stones.
Complications and management
Complications may arise with using the balloon. If the balloon is
pulled too hard against the stone, rupture of the balloon may occur.
If the stone is too large for the sphincterotomy, the balloon may
deform and slip out, leaving the stone impacted at the lower end of
the common bile duct or at the level of the papilla. In order to free
an impacted stone, it may be necessary to push it back using a stiffer
accessory, including a biopsy forceps. Alternatively, it may be possi-
ble to extend the sphincterotomy if a standard sphincterotome can
be inserted past the stone. Another option is to use a needle knife
to cut onto the bulging intraduodenal portion of the distal bile duct
and papilla in order to free the stone (Fig. 13.4). A balloon may also
be inated below an impacted stone in the distal duct and contrast
injected under pressure to push the stone back up the proximal duct.
In the event that the stone is impacted within the bile duct, it is
important to ensure drainage of the biliary system by placing an
Chapter 13 Stone Extraction
Fig. 13.4 Impacted ampullary stone. Note bulging distal bile
duct. Initial precut is followed by standard sphincterotomy, result-
ing in spontaneous expulsion of stone after completion of
sphincterotomy.
indwelling stent or a nasobiliary drain to avoid subsequent cholan-
gitis if the stone is not removed.
Relative cost
Balloons range in price from US$100 to $150 (list price), depending
on the manufacturer and whether they are double-lumen or triple-
lumen catheters.
BASKET STONE EXTRACTION
Key points
The basket should be lled with dilute contrast to facilitate localiza-
tion of stone. Avoid injecting excess contrast to reduce risk of dis-
placing stone upwards into the intrahepatic system.
Open basket above the stone and pull back to engage the stone.
Half close the basket to facilitate traction removal of stone.
Avoid using basket to engage the stone in intrahepatic system as
this may further displace the stone.
Use an occlusion balloon over a guidewire to pull an intrahepatic
stone back into the common duct for ease of stone engagement
and extraction.
If extraction fails, be prepared to free the engaged stone or proceed
with mechanical lithotripsy to avoid basket and stone impaction.
Indications/contraindications
Medium to larger size stones may be more easily removed with a
basket.
Complications
Migration of stones into intrahepatic ducts.
Impaction of stone and basket.
Introduction
Wire baskets are frequently used for stone extraction. A variety of
baskets are available in different sizes and congurations, which
allow engagement of stones varying from 5 mm up to 3 cm. However,
stones larger than 2 cm often cannot be extracted intact and will
require fragmentation prior to removal. The 4-wire Dormia basket
is the most commonly used type (Fig. 13.5). It is hexagonal in shape
and made of braided steel wires or nitinol wires. The stone is engaged
between the wires when the basket is closed, and the stone removed
by traction removal of the basket. However, small stones sometimes
121
 SECTION 2 TECHNIQUES
Fig. 13.5 Opened wire-guided basket (Olympus America, Lehigh
Valley, PA) with the guidewire. Note the much larger gaps between
the wires.
A
B intrahepatic stones or stones that may have migrated into the intra-
Fig. 13.6 A An opened ower basket (Olympus America, Lehigh
Valley, PA). Note the smaller mesh size on the upper part of the
basket. B Partially closed ower basket showing smaller mesh
size, which is better for trapping small stones.
are difcult to capture with standard baskets which have large gaps
between the wires. The Olympus ower basket has a modied
design in that the top part of the basket is further divided into eight
wires, giving rise to smaller mesh size for improved stone engage-
ment. Small stones are thus more easily trapped than with the
regular 4-wire basket (Fig. 13.6).
Spiral baskets are also available and may be used to remove rela-
tively small stones (Fig. 13.7). In the spiral conguration, the wires
spring close around the stone as the basket is pushed open. However,
spiral baskets are not designed for lithotripsy. Other larger and
stronger baskets are available which are made for mechanical litho-
tripsy and are designed to engage large stones. The basket wires are
much stronger and can be used to mechanically crush a large stone.
Traction or tension can be applied to the wires either manually or
122
Fig. 13.7 Opened helical basket (Cook Endoscopy).
with the help of a special crank handle that is used to tighten the
basket wires around the stone to break the stone.
In general, contrast can be injected through the basket to outline
the stones in the bile duct. When a single lumen basket is used, the
basket should be opened slightly to allow free ow of dilute contrast.
A double-lumen basket allows both injection of contrast through the
basket channel and passage of a guidewire through a separate
channel. Alternatively, the basket can be advanced over a previously
positioned guidewire. This is especially helpful for the removal of
hepatic ducts.
Description of technique
After the stones are visualized on the cholangiogram, a closed basket
is inserted into the bile duct and advanced beyond the stone. After
a fresh sphincterotomy or balloon sphincteroplasty, it is important
that the basket be inserted in the correct axis in the bile duct to avoid
dissecting a false plane, which may result in submucosal trauma or
even a perforation. Once in the bile duct, the basket is opened gently
above the stone. If necessary, the basket can be opened in the intra-
hepatic duct and pulled back to engage the stone. Care should be
taken to avoid opening the basket below the stone, as the opened
basket wires may push the stone further up the bile duct or into the
intrahepatic system. The opened basket is pulled back gently and
jiggled alongside the stone to engage it. After the stone is trapped,
the basket is closed gently to avoid losing the stone (Fig. 13.8A). The
scope is then pushed further into the second and third part of the
duodenum to straighten the axis of the basket and the bile duct to
ensure proper stone engagement. Steady traction is applied to the
basket catheter at the level of the biopsy valve and the basket is
withdrawn together with the stone until it reaches the distal bile duct
or at the level of the sphincterotomy. Traction is applied to the basket
catheter and at the same time, the tip of the scope is angled down
and rotated gently to the right, pulling the stone out of the bile duct
(Fig. 13.8B). In most situations, small to medium-sized stones can
be removed easily. If the stone does not come out immediately, it is
worth repeating the same maneuver while maintaining steady trac-
tion on the basket to ease the stone out of the bile duct.
During stone extraction, pulling with an opened basket is less
effective, as the loose basket wires tend to cut across the sphincter-

A
B
Fig. 13.8 A Stone engaged within a basket. B Stone removed
with a wire basket.
otomy rather than along the axis of the common bile duct. This also
tends to pull the stone against the sphincterotomy and results in
bruising the tissue. Closing the basket gently allows the wires to
come together, allowing more effective extraction force to be trans-
mitted along the basket catheter for stone removal. However, the
basket should not be closed too tightly around the stone to avoid
embedding the wires onto the surface of the stone. This is especially
important in the case of a large stone, as stone and basket impaction
Chapter 13 Stone Extraction
may occur if the stone is too large to be removed and the trapped
stone cannot be freed from the basket.
Indications/contraindications
The advantage that the wire basket offers over the extraction balloon
is that it provides more effective traction and therefore is helpful for
removing medium size to large stones. However, smaller stones or
stone fragments may not be easily engaged by the wires. In addition,
intrahepatic stones may be difcult to access due to the smaller
caliber of the intrahepatic ducts and constraints in opening the
basket. In these situations, use of an extraction balloon is
preferred.
Complications and management
During basket stone extraction, the stone(s) may migrate up into
the intrahepatic duct. Catching a migrated stone within the
intrahepatic duct can be challenging. The best method is to avoid
using the basket. A balloon and guidewire may be used to selectively
cannulate the respective segment containing the stone. The balloon
is advanced over the guidewire beyond the stone and inated. The
stone is then pulled back into the common hepatic duct or the
common bile duct before further attempts are made to remove the
stone, either using the same balloon catheter or exchanging for a
basket. Use of a wire-guided basket is also helpful in removing
stones from the intrahepatic ducts, although these baskets tend to
be rather stiff and manipulation in the intrahepatic ducts can be
somewhat difcult.
If basket stone extraction fails, it may be necessary to free the
engaged stone to avoid basket and stone impaction. This may be
achieved by gently advancing the basket and stone up towards the
bifurcation and opening the basket so that the wires bend back on
themselves. In this fashion, the wires are opened and the stone can
be dropped from the basket. The basket is then closed slowly above
the stone to avoid re-engaging it when the opened basket is pulled
back. Once the basket is closed, it can be removed. Further stone
extraction may require extension of the sphincterotomy or stone
fragmentation using a mechanical lithotripter.
One potential complication that may arise is impaction of the
basket and stone within the bile duct or at the level of the papilla
due to large stone size and inadequate sphincterotomy or inability
to enlarge the sphincterotomy. In rare instances, stone and basket
impaction have occurred at the level of the head of the pancreas due
to a narrowed distal common duct. In these cases, emergent mechan-
ical lithotripsy may be attempted using a Soehendra lithotripter.
Relative cost
List prices for baskets vary from $150$350, depending on the
design. Some are disposable while others are reusable.
MECHANICAL LITHOTRIPSY
Key points
Always have a device available in case of unexpected stone/basket
impaction.
The stone is captured within a wire basket and the metal sheath is
advanced over the basket up to the stone. Closing the basket
crushes the stone against the tip of the metal sheath.
Ensure that the tip of the metal sheath is in contact with stone
before lithotripsy.
123
 SECTION 2 TECHNIQUES
Fig. 13.9 Failed basket stone extraction. Three large stones in
common bile duct. Basket is too small for the stones.
Turn the crank handle slowly and allow time for the basket wires
to cut through the stone and to avoid breaking the basket.
Indications/contraindications
Mechanical lithotripsy is used when there is failure of standard
balloon or basket stone extraction due to large stones.
May also be used in the case of impacted stone and basket.
Complications
Excess traction on the wires against a very hard stone may lead to
breakage of the basket wires.
Failure can occur if there is stone impaction that prevents proper
opening of the large basket around the stone.
Introduction
When standard balloon or basket stone extraction fails (Fig. 13.9),
fragmentation of the stone within the bile duct may become neces-
sary, particularly if the stone is larger than 2 cm, or if there is a
discrepancy between the stone size and the exit passage, i.e., distal
common bile duct stricture, small sphincterotomy, or after balloon
sphincteroplasty. In some cases, extension of the sphincterotomy
is not feasible and may increase the risk of bleeding and perfora-
tion. It may be possible to use balloon dilation of a partially cut
sphincterotomy in order to gain a bigger opening to facilitate
removal of large stones. Mechanical lithotripsy is accomplished by
capturing a stone within a wire basket and advancing a metal
sheath over the basket catheter up to the stone. The stone is then
crushed by forceful traction of the basket wires and stone against
the metal sheath.
Description of technique
Several lithotripters are available, but they fall into two main catego-
ries. One requires cutting of the basket handle and removal of the
endoscope prior to lithotripsy, and is frequently used on an emer-
gency basis when unexpected stone and basket impaction occurs
(life saver). The Soehendra lithotripter (Cook Endoscopy, Winston-
124
Fig. 13.10 The Soehendra lithotripter (courtesy of Cook Endos-
copy, Winston-Salem, NC) consists of a 10 Fr metal sheath that goes
through the instrument channel of the duodenoscope and a self-
locking crank handle.
A B
C D
Fig. 13.11 For demonstration purpose, an articial stone is used.
A Stone is engaged in the basket. The basket handle is cut and
Teon sheath is then removed. B Metal sheath of the Soehendra
lithotripter inserted over basket wire. C The metal sheath is
inserted through the scope channel and advanced to the level of
the stone. The cut end of the basket wires is inserted through the
shaft of the crank handle. D The basket wires are tied to the shaft
of the crank handle and tension is applied as the handle is slowly
wound to crush the stone.
Salem, NC) consists of a 14 Fr metal sheath and a self-locking crank
handle (Fig. 13.10). The apparatus is compatible with standard
extraction baskets or large lithotripsy baskets with stronger wires. It
requires cutting of the basket handle to allow removal of the duode-
noscope. The metal sheath is then inserted over the basket catheter.
There are grooves at the tip of the metal sheath and to avoid the bare
basket wires from getting stuck, a tape may be applied to the end of
the metal sheath. This also helps to avoid injury to the posterior
pharynx during insertion. It is helpful to retain the Teon sheath
initially to facilitate passage of the wires through this 14 Fr metal
sheath before removing it. The metal sheath is advanced all the way
up to the stone under uoroscopy control. The Teon sheath is
removed and the cut ends of the basket wires are then inserted into
the shaft of the crank handle. The metal sheath is attached to the
handle by a luer lock. The wires are then tied to the handle and trac-
tion applied slowly by winding the crank handle. The basket is closed
and drawn into the metal sheath as forceful traction is applied,
crushing the stone against the tip of the metal sheath in the process
(Fig. 13.11). It is important to remember that standard baskets are

B lithotripsy basket. B In a separate case, a
A stone. C The metal sheath of the litho-
not designed for lithotripsy, and if traction is applied too quickly the
basket wires may break rather than the stone. Winding the crank
handle slowly and allowing time for the basket wires to cut through
serves to break up the stone and avoid the complication of a broken
basket and stone impaction.
The latest improvement in design for the Soehendra lithotripter
involves using a smaller 10 Fr sheath, which can be inserted through
the scope channel over the basket wires after cutting off the basket
handle. In this case, it is not necessary to remove the duodenoscope.
It is used specically with the Webb basket (Cook Endoscopy),
which has smaller wires. This through-the-scope set-up may not be
compatible with other standard baskets. However, unlike other TTS
lithotripsy baskets, the basket cannot be reused for repeat stone
fragmentation because the handle is cut.
The other type of lithotripter is specically designed to be used
through-the-scope (TTS) without the need for scope removal, such
as the BML lithotripsy baskets (Olympus Co., Tokyo, Japan) or the
Trapezoid basket (Microvasive, Boston Scientic, Natick, MA). These
are used in a more elective setting, when a large stone, above a
stricture or difcult stone removal is anticipated.
The BML lithotripter is a three-layer device consisting of a strong
four-wire basket within a Teon sheath and an overlying metal
sheath. The larger lithotripsy basket or BML-3Q equivalent (BML-
201) has a slightly thicker metal sheath and requires a scope with a
4.2 mm working channel. It allows injection of contrast because of
the larger diameter. The smaller basket or BML-4Q equivalent (BML-
202, 203) goes through a 3.2 mm channel, but injection of contrast
Chapter 13 Stone Extraction
Fig. 13.12 A Large stone engaged in BML
BML lithotripsy basket is used to engage the
tripter is advanced over the Teon sheath,
and basket wires tightened around the stone
in mid CBD for stone fragmentation.
A B
Fig. 13.13 A Through-the-scope BML lithotripter (courtesy of
Olympus America Inc., Melville, NY) consists of a wire basket within
a Teon catheter, an overlying metal sheath, and the crank
handle. B With the stone engaged in the basket, the metal sheath
is advanced over the Teon catheter up to the basket. Basket is
closed by turning the wheel and the stone is crushed against the
metal sheath.
is more difcult. The BML lithotripter has a reusable version and a
disposable version. The reusable system is assembled by inserting
the Teon sheath through the metal sheath, then loading the basket
into the Teon catheter. The wires are soldered together onto a shaft
that is connected to the crank handle. Contrast can be injected via
the Teon catheter. The opening and closing of the basket is con-
trolled with the handle. Stone engagement is performed initially
with the Teon catheter and basket. The metal sheath is then
advanced over the Teon catheter up to the level of the stone only
when lithotripsy is required (Figs 13.12, 13.13).
125
 SECTION 2 TECHNIQUES
A B
Fig. 13.14 A The wires of a lithotripsy basket can be deformed as
a result of crushing a hard stone. B It is necessary to reshape the
basket wires to ensure they open properly for repeat stone
engagement.
Although a general shaking and jiggling movement of the basket
can be tried to engage the stone, very often the large stone(s) may
compress the basket wire making stone engagement difcult. It may
be useful to rotate the basket wires around the stone by rotational
movement of the tip of the scope (even with the help of the metal
sheath to transmit a stronger force) in order to engage the stone.
There is a locking mechanism (notches on the metal handle) for the
metal sheath to ensure proper engagement of the stone before litho-
tripsy. It is very important to make sure that the metal sheath is
locked in the correct position. This is done by keeping the device
straight and avoiding any looping. If the metal sheath is not locked
properly, i.e. with a short segment of the Teon catheter exposed
between the stone and the tip of the metal sheath, mechanical litho-
tripsy will not be effective. If there is recoil or a clicking noise heard
when turning the control wheel, lithotripsy is not working, and the
position of the metal sheath will need to be adjusted. With the stone
properly engaged in the basket and the metal sheath correctly posi-
tioned, traction is applied to the wires by turning the control wheel
to crush the stone. As the control wheel does not have a built-in
self-locking mechanism, traction should be applied slowly and con-
tinuously to allow time for the wires to cut through the stone. The
reusable system can be taken apart after lithotripsy for cleaning and
sterilization. The disposable version comes with the lithotripsy
basket, Teon catheter and metal sheath all built into one and is
meant for use in one patient session only. If a large and hard stone
is crushed, always remove the basket to examine the wires as they
are often deformed due to stone breakage (Fig. 13.14). It is necessary
to reshape the basket wire either by hand or with a pair of hemostats
to ensure subsequent proper basket opening to trap the stone frag-
ments for repeat lithotripsy. As a large stone will give rise to large
fragments, these will still require further fragmentation before they
can be removed easily.
In general, mechanical lithotripsy is very effective in breaking
large stones. Repeated stone fragmentation may be required in the
case of a very large stone. The reported success rate of mechanical
lithotripsy for large stones has ranged from 85% to 90%.1216
The Trapezoid basket (Microvasive, Boston Scientic) is made of
nitinol wire and comes with a coated and rather exible metal sheath
(Fig. 13.15). The design of the basket also allows it to go over a guide-
wire inserted through a separate channel. This is especially handy
with the Rapid Exchange system (Boston Scientic). The exibility of
the sheath also allows free cannulation with the basket for stone
engagement. If stone fragmentation is not necessary, stone extraction
can be done in the usual manner because of the exible sheath. When
126
A B
Fig. 13.15 A Opened trapezoid basket (courtesy of Boston
Scientic, Natick, MA) with handle. Note the plastic-covered metal
sheath. The plastic handle can be tted to a cranking device if litho-
tripsy is required. B Opened trapezoid basket inserted over a
guidewire.
stone fragmentation is indicated, the handle of an insufator (similar
to a caulk gun design) can be tted to the handle of the basket and
traction is then applied to the basket wires to break the stone. This
design offers the benet of having selective cannulation over a guide-
wire and the potential option for stone fragmentation.
Indications/contraindications
When stones are too large to be removed with standard balloons or
wire baskets, mechanical lithotripsy is indicated for stone fragmen-
tation prior to removal. It is also a useful tool in the case of stone
and basket impaction.17
Complications and management
The Soehendra lithotripter may be used as a lifesaver measure in
cases of stone and basket impaction. However, the wires of the stan-
dard basket are relatively soft, and may break in the duodenum result-
ing in retained broken basket and stone in the bile duct. Surgical
exploration may be needed to remove the impacted stone and basket.
The Olympus TTS lithotripsy basket set-up is designed to break
at the connection between the basket and the crank handle and
alternatively, the basket wires are also designed to break at the tip to
prevent having a broken basket around an impacted stone in the bile
duct. In the unexpected event of the basket breaking at the connec-
tion point, Olympus has developed a special metal sheath that
resembles the Soehendra set-up. This metal sheath can be inserted
over the broken basket after removal of the endoscope. Stone frag-
mentation can be performed as with the Soehendra lithotripter. It is
not advisable to use the Cook Endoscopy Soehendra lithotripter
handle and adapt it onto a broken Olympus lithotripsy basket as it
is not truly compatible. If that is necessary, it is important to retain
the Teon catheter around the basket wires to offer support for the
metal sheath. The design of the metal sheath with the disposable
BML system is different and if tension is applied to the wires without
the Teon catheter, the coils of the metal sheath buckle thus making
lithotripsy ineffective.
TTS mechanical lithotripsy is successful in the majority of cases
because these baskets are stronger. Over 80% of large (>2 cm) stones
have been fragmented in reported series giving a common duct
clearance rate of over 95%. The main reason for failure is the stone
is too large for the size of the basket but even then, if part of the
stone can be engaged, stone fragmentation can be attempted to
reduce the size of the stone allowing for proper engagement for
complete stone fragmentation. Mechanical lithotripsy fails when
there is stone impaction or if there is not enough room within the

bile duct to open the basket to engage the stone. In this situation,
temporary stenting can be tried to insure drainage of the biliary
system. Subsequent spontaneous stone fragmentation has been
observed in 30% of cases possibly due to the friction between the
stone and the stent or to possible dissolution effects of improved
bile ow.
Although relatively uncommon, perforation of the bile duct may
occur due to the relative stiffness of the basket when the wires are
tightened. In addition, pancreatitis may result from forceful removal
of an impacted stone and basket, causing injury to the pancreatic
orice.
INTRADUCTAL ELECTROHYDRAULIC
LITHOTRIPSY
Key points
It is preferable to use the mother and baby scope system for direct
visual control of intraductal lithotripsy.
Despite the angulations control, positioning of the baby scope is
coordinated by the operator manipulating the mother scope.
Both stone and probe are immersed in uid medium (normal
saline), and with the probe directly against the stone, electrical
discharge from the probe generates shock waves which result in
stone fragmentation.
An adequate sphincterotomy is necessary to allow drainage from
the bile duct while the baby scope is in position to avoid raising
the intra-biliary pressure.
Use an indwelling nasobiliary catheter to ush the bile duct to
improve visualization and to remove stone fragments during
lithotripsy.
Indications/contraindications
May be used in the setting of failed lithotripsy due to large or
impacted stone.
Complications
Bile duct injury and perforation.
Introduction
Mechanical lithotripsy is so effective and successful that there is less
and less need for intraductal lithotripsy except in very special cir-
cumstances, and the set-up is mostly available only in academic
units. Intraductal lithotripsy can be conducted under uoroscopy
using a special balloon that centers the probe in the bile duct, or
alternatively under direct visual control using the mother and baby
scope system.
Description of technique
There are two forms of intraductal lithotripsy (IDL)electrohydrau-
lic (EHL)1820 or laser lithotripsy.21,22 Each of these types of intraductal
lithotripsy is best performed under direct visual control using the
mother and baby scope system. An older Olympus system consists
of a jumbo size duodenoscope with a 5.5 mm channel and the baby
scope is 4.7 mm in diameter with a 1.7 mm channel. The larger
instrument channel allows easier passage of the lithotripsy probe.
The newer baby scope is smaller with a 3.2 mm diameter and a
1 mm instrument channel. This can be inserted through a therapeu-
tic duodenoscope of 4.2 mm instrument channel. However, because
of the small instrument channel, it can only accommodate the small
lithotripsy probe or the smaller laser probe.
Chapter 13 Stone Extraction
After an adequate sphincterotomy or combined sphincterotomy
and balloon sphincteroplasty, the baby scope is introduced into the
common bile duct with the help of the mother scope. Despite
the two way (up/down) tip deection of the baby scope, most of the
positioning of the baby scope is achieved by manipulating the mother
scope (see Chapter 21). Two endoscopists are required to conduct
this examination and it is best to project the image of both scopes
onto two monitors to facilitate the coordination between the two
operators. A uid medium (normal saline) is required for electrohy-
draulic lithotripsy and this can be introduced through the channel
of the baby scope. However for practical purposes, we prefer to insert
a nasobiliary catheter which allows more efcient lling and also
ushing of the system with saline to remove the stone fragments
and debris generated as a result of stone fragmentation.
The electrohydraulic lithotripter consists of a power generator
(e.g. Walz, Germany) by which the power or energy can be preset
based on the size of the probe (3 or 4.5 Fr) used (see Chapter 21).
The tip of the probe has a pair of bipolar electrodes, which when
activated in a uid medium, will generate a shock wave to fragment
the stone. The frequency of discharge of the shock wave can be
preset on the machine and activated by a foot switch either as a
single pulse or continuous discharge. Depending on the instrument
channel of the baby scope, two sizes of probes (3 Fr and 4.5 Fr) can
be used. For the more oppy 3 Fr probe, it may be helpful to preload
this before inserting the baby scope into the bile duct. Examination
with the baby scope is facilitated by lling the bile duct with saline.
Care is taken not to overll the system to avoid the risk of cholangi-
tis. The probe is placed in contact with the stone, under direct visu-
alization with the baby scope as well as on uoroscopy. The stone
and probe are immersed with saline lled through the nasobiliary
catheter using an infusion pump. The foot switch is activated and
stone fragmentation is performed under direct visual control. Stone
fragments generated during lithotripsy can obscure the view and this
can be cleared by ushing more saline and suctioning with the
mother scope. Effective stone fragmentation can be demonstrated
under uoroscopy by injecting contrast into the bile duct. Subse-
quent stone extraction is performed with basket after removal of the
baby scope, followed by an occlusion cholangiogram to document
ductal clearance. Since stone fragments are generated during litho-
tripsy, it may be helpful to insert the stone extraction basket deep in
the bile duct and irrigate the duct with saline ushed through the
basket (while applying suction with the scope). This helps to remove
the remaining stone fragments from the common duct. If the jumbo
size duodenoscope is used for mother and baby examination, it may
be useful to change back to a regular therapeutic scope for ease of
manipulation. In most cases, it may be advisable to insert an indwell-
ing stent or nasobiliary catheter to decompress the biliary system
and to allow the stone fragments to settle and avoid the risk of stone
impaction and cholangitis. A repeat ERCP is performed to remove
remaining stone fragments for complete duct clearance.
Indications/contraindications
When mechanical lithotripsy fails to remove the stone, particularly
large stones that are difcult to capture with the lithotripsy basket,
or an impacted stone, intraductal electrohydraulic lithotripsy is an
effective option. The reported success rates in these cases are excel-
lent.1820 However, the cost of the equipment, the lack of general
availability of mother-baby scope system, and the requirement for
two endoscopists with expertise in the use of the equipment can be
prohibitive for general use.
127
 SECTION 2 TECHNIQUES
Complications and management
Although highly effective, one of the major problems is difculty
with targeting of the shock waves and inadvertent bile duct injury
and perforation. The procedure should be performed under direct
endoscopic and uoroscopic guidance by endoscopists with experi-
ence in the use of the equipment.
CONCLUSION
It is customary to clear the bile duct of stones because of the risk of
obstruction, cholangitis and pancreatitis. Sphincterotomy and
REFERENCES
1. Lauri A, Horton RC, Davidson BR, et al. Endoscopic extraction of
bile duct stones: management related to stone size. Gut. 1993;
34:17181721.
2. Palmer KR, Hoffman AF. Intraductal monooctanoin for the direct
dissolution of bile duct stones. Experiences in 343 patients. Gut
1986; 27:196202.
3. Kaye GL, Summereld JA, McIntyre N, et al. Methyl tert butyl ether
dissolution therapy for common bile duct stones. J Hepatol 1990;
10:337340.
4. Neoptolemos JP, Hall C, OConnor HJ, et al. Methyl tert butyl
ether for treating bile duct stones: the British experience. Br J
Surg 1990; 77:3235.
5. Stock SE, Carlson GL, Lavelle MI, et al. Treatment of common bile
duct stones using monooctanoin. Br J Surg 1992; 79:653654.
6. Cotton PB, Forbes A, Leung JWC, et al. Endoscopic stenting for
long-term treatment of large bile duct stones: 2- to 5-year follow-
up. Gastrointest Endosc 1987; 33:411412.
7. Foutch PG, Harlan J, Sanowski RA. Endoscopic placement of
biliary stents for treatment of high-risk geriatric patients with
common duct stones. Am J Gastroenterol 1989; 84:527529.
8. Van Steenbergen W, Pelemans W, Fevery J. Endoscopic biliary
endoprosthesis in elderly patients with large bile duct stones:
long-term follow-up. J Am Geriatr Soc 1992; 40:5760.
9. Siegel JH, Yatto RP. Biliary endoprosthesis for the management of
retained common bile duct stones. Am J Gastroenterol 1984;
79:5054.
10. Maxton DG, Tweedle DE, Martin DF. Retained common bile duct
stones after endoscopic sphincterotomy: temporary and long-
term treatment with biliary stenting. Gut 1995; 36:446449.
11. Peters R, Macmathuna P, Lombard M, et al. Management of
common bile duct stones with a biliary endoprosthesis. Report
on 40 cases. Gut 1992; 33:14121415.
128
sphincteroplasty facilitate access into the biliary system. Basket and
balloon catheters are useful to remove stones up to 1.5 cm in diam-
eter. The use of wire-guided basket or balloon catheters allows
proper access into the intrahepatic system to remove intrahepatic
stones or migrated stones. Stones above a bile duct stricture will
require balloon dilation of the stricture prior to successful removal.
The use of mechanical lithotripsy to break up the stones facilitates
duct clearance of large stones or stones above a stricture. Mechanical
lithotripsy is very effective in achieving common duct clearance and
the use of intraductal lithotripsy is limited to 5% of very difcult
ductal stones.
12. Riemann JF, Seuberth K, Demling L. Clinical application of a new
mechanical lithotripter for smashing common bile duct stones.
Endoscopy 1982; 14:226.
13. Riemann JF, Seuberth K, Demling. Mechanical lithotripsy
of common bile duct stones. Gastrointest Endosc 1985;
31:207210.
14. Schneider MU, Matek W, Bauer R, et al. Mechanical lithotripsy of
bile duct stones in 209 patientsEffect of technical advances.
Endoscopy 1988; 20:248253.
15. Chung SCS, Leung JWC, Leong HT, et al. Mechanical lithotripsy of
large common bile duct stones with a basket. Br J Surg 1991;
78:14481450.
16. Shaw MJ, Mackie RD, Moore JP, et al. Results of a multicenter trial
using a mechanical lithotripter for the treatment of large bile duct
stones. Am J Gastroenterol 1993; 88:730733.
17. Schutz SM, Chinea C, Friedrichs P. Successful endoscopic removal
of a severed, impacted Dormia basket. Am J Gastroenterol 1997;
92:679681.
18. Leung JW, Chung SC. Electrohydraulic lithotripsy with peroral
choledochoscopy. BMJ 1989; 299:595597.
19. Siegel JH, Ben-Avi JS, Pullano WE. Endoscopic electrohydraulic
lithotripsy. Gastrointest Endosc 1990; 36:134136.
20. Hixson LJ, Fennerty MB, Jaffee PE, et al. Peroral
cholangioscopy with intracorporeal electrohydraulic
lithotripsy for choledocholithiasis. Am J Gastroenterol 1992;
87:296299.
21. Ell C, Hochberger J, May A, et al. Laser lithotripsy of common bile
duct stones. Gut 1988; 29:746751.
22. Neuhaus H, Hoffmann W, Gottlieb K, et al. Endoscopic
lithotripsy of bile duct stones using a new laser with
automatic stone recognition. Gastrointest Endosc 1994;
40:708715.
 SECTION 2 TECHNIQUES
Chapter
Pancreatic Sphincterotomy
14 Jonathan M. Buscaglia and Anthony N. Kalloo
INTRODUCTION
Since its initial application in 1974, endoscopic biliary sphincterot-
omy has revolutionized the approach to patients with biliary tract
diseases.1 Using biliary sphincterotomy in conjunction with other
techniques such as stent placement, balloon and basket extraction
of stones, and stricture dilatation, biliary sphincterotomy has become
the standard of care for problems that were once only remedied by
surgical procedures. Endoscopic therapy for pancreatic disorders has
not advanced quite as rapidly, however. The main reason for this
seems to be the longstanding fear of inducing pancreatitis in an
organ that frequently expresses its dislike for simple manipulation
of the papilla and sphincter alone.2 Historically, pancreatitis and its
associated complications have prevented some endoscopists from
attempting to apply therapeutic techniques similar to those used in
treating biliary tract disorders. In addition, clear-cut indications for
endoscopic therapy of the pancreas have been much more difcult
to dene due to a paucity of well-designed clinical trials justifying
its use. Most of the techniques that have been used in previous
studies were performed on small numbers of patients, and in expert
centers only. The majority of studies have been retrospective in
design without control groups. There has been a deciency in studies
that utilize randomization and directly compare endoscopic therapy
with either surgical or medical therapy.2
It is on this background that we begin to discuss endoscopic
pancreatic sphincterotomy. This technique is the cornerstone of
endoscopic therapy of the pancreas, and it provides initial access to
the main pancreatic duct.3 Once access to the duct is obtained, it
may be used as a single therapeutic maneuver (e.g. to treat pancre-
atic type sphincter of Oddi dysfunction), or in series with other
endoscopic therapeutic techniques such as the placement of a stent
across a ductal stricture.4 In chronic pancreatitis, for example, pan-
creatic sphincterotomy not only decreases pressure within the main
pancreatic duct, but it may be used to help facilitate the extraction
of calculi and protein plugs.1
The ensuing chapter will focus on the endoscopic techniques
and the equipment used by most experts who regularly perform
pancreatic sphincterotomy. The indications and contraindications
for this technique as well as the evidence that supports its basis-
will also be discussed. Furthermore, the complications associated
with pancreatic sphincterotomy and their management strategies
will be outlined. Lastly, we will briey discuss any existing litera-
ture associated with the costs and cost savings of pancreatic
sphincterotomy.
ENDOSCOPIC PANCREATIC SPHINCTEROTOMY
Preparation
As with all endoscopic procedures, it is imperative to obtain valid
informed consent before beginning.5 This becomes of paramount
importance when discussing with patients and their families the
potential risks involved in performing an ERCP with pancreatic
sphincterotomy, as the complication rates are greater when com-
pared with routine upper endoscopy. Routine blood work, including
a CBC and coagulation parameters are usually veried before the
procedure. Aspirin, NSAIDs, warfarin, and other anticoagulants are
withheld for several days before and after the procedure, if possible.5
Antibiotics are frequently given 3060 minutes before the procedure
in an effort to prevent procedure-related infection such as biliary
sepsis. The data supporting antibiotic prophylaxis prior to pancreatic
or biliary sphincterotomy is sparse and very limited. Only a few
studies have attempted to investigate this issue, and most of them
have had small numbers with poorly dened endpoints.6 Nonethe-
less, many endoscopists will utilize antibiotic prophylaxis before
(and frequently following) a planned endoscopic sphincterotomy.
Coverage with a second- or third-generation cephalosporin appears
sufcient. Broader spectrum antibiotics such as piperacillin/tazo-
bactam, or vancomycin and gentamicin if there is a penicillin allergy,
may be warranted in some instances.
Equipment
Pancreatic sphincterotomy, like biliary sphincterotomy, is performed
with a standard side-viewing duodenoscope, the appropriate sphinc-
terotome (papillotome), and an electrosurgical generator.7 There are
a variety of options in terms of commercially available electrosurgi-
cal generators. Most have both monopolar and bipolar options, and
they offer pure cutting, pure coagulation, and blended (cut/coag)
current modes.5 These are the same generators that are used when
performing polypectomies. Newer models even allow for cutting
options that incrementally splice the sphincter muscle in 1 mm seg-
ments, while informing the endoscopist with an audible alarm at the
end of each segment. This attempts to ensure that the sphincterot-
omy is performed in a careful, stepwise fashion without producing
an unintentional exaggerated cut. However, there is no data that
veries the effectiveness of this method.
There is little evidence supporting the use of one current mode
over the others during a sphincterotomy. Some data, however, sug-
gests that pure cutting current may be associated with less post-
ERCP pancreatitis when compared to blended current.8 Also, the
129
 SECTION 2 TECHNIQUES
pure cutting technique is thought to cause less brosis, thus helping
to diminish the chance of developing future papillary stenosis. As a
result, some endoscopists advocate using only the pure cutting mode
when performing a pancreatic sphincterotomy. It is unclear as to
whether or not there are truly more bleeding complications associ-
ated with this option, as some have suggested.
An enormous variety of sphincterotomes and guidewires are now
commercially available for use in pancreatic sphincterotomy. For a
more detailed description of all endoscopic guidewires and ERCP
accessories, refer to Chapter 4. The original Demling-Classen or
Erlangen pull-type sphincterotome (bowstring design) is still the
most popular choice for performing a pancreatic sphincterotomy
(Fig. 14.1). There are several variations in this type of sphinctero-
tome; they are all based upon differences in the length of the exposed
cutting wire, the number of additional lumens, and differences in
the length of the nose of the instrument.5,7 A shorter nose (5
8 mm beyond the wire) is sometimes more convenient for cannula-
tion of the major papilla prior to sphincterotomy. It allows for easier
engagement between the sphincterotome tip and the papilla without
much interference from the cutting wire. Once the tip is positioned
inside the papillary orice, tension on the wire may be used to bow
the tip into the correct axis and correctly align the instrument for
eventual sphincterotomy.7 Sphincterotomes with longer noses (2
5 cm beyond the wire) lose their ability to bow since most of the
cutting wire is retained inside the duodenoscope until deep cannula-
tion is attained. The advantage of this type of sphincterotome is the
ability to maintain cannulation while the wire is being withdrawn
during sphincterotomy. However, now with triple-lumen cannulas
and soft, exible tip guidewires which cause less injury to the pan-
creatic duct, a guidewire can be easily left in place within the pan-
creatic duct in order to maintain cannulation while performing a
pancreatic sphincterotomy (wire-guided sphincterotomy).
Standard sphincterotomes have a 5 Fr to 7 Fr catheter tip that can
accept a 0.035-inch guidewire.9 Use of a triple-lumen cannula allows
for the placement of a preloaded guidewire and simultaneous
contrast injection without having to remove the guidewire. When
Wire pulled
A Pull type
B Needle knife
Fig. 14.1 Standard sphincterotomes used in pancreatic sphincter-
otomy. (Reproduced by permission of Division of Gastroenterol-
ogy and Hepatology, Johns Hopkins Hospital.)
130
manipulating the papilla and the most proximal portions of the main
pancreatic duct prior to sphincterotomy, many endoscopists prefer
an ultra-tapered catheter tip (5 Fr-4 Fr-3 Fr) for easier cannulation.
These sphincterotomes utilize smaller caliber guidewires; frequently
down to 0.018-inch in diameter. Conversely, a special 3F cannula
can be passed down through the channel of a standard sphinctero-
tome to produce the effect of a tapered tip catheter.9
Endoscopic pancreatic sphincterotomy is performed after deep
cannulation of the main pancreatic duct with a guidewire.10 There
are several different types of commercially available guidewires that
may be used to perform this technique. Such congurations of
guidewires include conventional, nitinol, hydrophilic, and hybrid.
The range in wire diameter is from 0.018-inch to 0.035-inch.11 When
performing a wire-guided pancreatic sphincterotomy, the hydro-
philic-coated wires with soft and oppy tips may be helpful in pre-
venting trauma to the main pancreatic duct or its side branches.10
As mentioned above, this deep wire-guided technique in pancreatic
sphincterotomy has obviated longer-nose sphincterotomes. Main-
taining adequate cannulation of the papilla in this manner is less
traumatic and more secure.
The endoscopic technique
The main principles involved in pancreatic sphincterotomy are very
much like those of biliary sphincterotomy. They involve wire-guided
cannulation of the duct prior to cutting, and they utilize a slow and
stepwise approach that relies on accurate identication of anatomi-
cal landmarks. There are essentially two different types of tech-
niques that are used by most expert endoscopists when performing
this procedure. The rst approach, and the more popular one, is
performed while using a standard pull-type sphincterotome. The
second approach uses an endoscopic needle-knife to cut the sphinc-
ter muscle after placement of a pancreatic duct stent. Both tech-
niques have their advantages and disadvantages, and the details
surrounding each approach will be discussed below. In addition, we
will briey describe the technique of precut or access pancreatic
sphincterotomy in those instances when the endoscopist is faced
with a difcult pancreatic cannulation. Sphincterotomy of the minor
papilla will be discussed separately in Chapter 15.
As with both biliary and pancreatic sphincterotomy, the key to
success starts with accurate cannulation of the correct duct. This is,
at times, the greatest hurdle for novice therapeutic endoscopists. In
general, selective cannulation of the pancreatic duct is easier than
that of the biliary duct, assuming there has not been a previous
sphincterotomy. The reason for this is directly related to the ana-
tomical axis of each duct in relation to the wall of the duodenum.
Although the main pancreatic duct may be tortuous with multiple
side branches, the most proximal portion of the duct courses away
from the papilla at a 90 angle from the duodenal wall (Fig. 14.2). It
then runs more towards the right and straight inside.12 The most
distal part of the common bile duct, on the other hand, assumes
more of an acute angle in its relationship with the duodenal wall. It
extends in an upward and leftward direction from the papillary
orice. This provides for a more difcult cannulation, depending on
how acute the angle of take-off is from the papilla.
One must remember that a cross-sectional view of the inner
portion of the ampulla of Vater (the part furthest from the duodenal
lumen) is similar to a double-eyed onion, with each duct running
off in its own unique direction.12 The most proximal portion of the
ampulla, however, is a single orice that leads from the lumen of
the duodenum into a common channel. This common channel then

Fig. 14.2 Position of the main pancreatic duct and the distal common bile duct in relation to the major papilla. The pancreatic duct runs
90 degrees perpendicular from the duodenal side wall. (Reproduced by permission of Division of Gastroenterology and Hepatology,
Johns Hopkins Hospital.)
merges with the so called double-eyed onion. The length of this
channel is variable, but usually ranges between 1 and 10 mm.7
Within this channel are several folds of papillary mucosa that may
often act as obstacles to selective cannulation with the sphinctero-
tome. Therefore, accurate cannulation depends on nding the
correct axis with the catheter tip before the guidewire can be pushed
all the way out into the main pancreatic duct. Approaching the papil-
lary orice with the right orientation allows the endoscopist to nd
the correct axis.
When aiming for the pancreatic duct, the catheter should enter
the orice perpendicular to the duodenal wall. Then, in order to tra-
verse the correct plane, the catheter tip should be advanced along
the oor of the orice to nd the pancreatic duct. This is in contrast sphincterotomy, the cutting direction is in the 11 oclock to 1 oclock
to biliary cannulation, in which the catheter tip is aimed at the roof
of the papillary orice to nd the distal common bile duct.7 The only
way to ultimately assure correct cannulation is to inject contrast and
verify ones position uoroscopically. In order to limit the risk of
pancreatitis, as little contrast as possible should be used in this
situation.
The importance of correct ductal axis is paramount. When faced
with a difcult cannulation of the pancreatic duct, one may have to
lower the catheter tip even further once advancing along the oor of
the papillary orice. This can be achieved by lowering the elevator
on the duodenoscope, thus pressing down on the oor.7 The injec-
tion of contrast is done simultaneously in an effort to correctly
identify the pancreatic duct.
Chapter 14 Pancreatic Sphincterotomy
Pull-type sphincterotomy
After successful pancreatic cannulation and advancement of the
guidewire into the main pancreatic duct, conrmation of position is
usually obtained with a contrast pancreatogram. Assuming a clear
indication for sphincterotomy has been established, this part of the
procedure is most often performed with a pull-type sphincterotome
(as mentioned above). Like biliary sphincterotomy, the incision
should be hot and slow.7 It should be directed towards the 1 to 2
oclock position with the very distal part of the cutting wire.5,10 In
other words, most of the cutting wire should be visible outside the
papillary orice. Note that the direction of the cut is very different
from that of a biliary sphincterotomy (Fig. 14.3a14.3d). In biliary
position (preferably the 12 oclock position). The sphincterotome is
slightly bowed while the cutting wire is walked up the roof of the
papilla in a stepwise fashion.7 In pancreatic sphincterotomy, the
same principles apply, but the direction is more towards the right,
guiding the cutting wire along the oor of the papillary orice.
The actual incision should be performed using the pure cutting
current with the electrosurgical generator. This prevents further
damage to the pancreas and limits the possible future development
of brosis and papillary stenosis.10,13 The length of the cut is gener-
ally between 5 and 10 mm. Larger diameter ducts require longer cuts
in order to achieve the largest possible access. Once the sphincter-
otomy has been completed, a temporary pancreatic stent is usually
left in place for a short period of time in order to help facilitate
131
 SECTION 2 TECHNIQUES

Fig. 14.3 A View of the major papilla before pancreatic sphincterotomy. B Sphincterotomy performed with the cutting wire angled
towards the 2 oclock position.

132
 Chapter 14 Pancreatic Sphincterotomy

Fig. 14.3 Continued C Placement of a pancreatic stent following sphincterotomy. D Completion of the sphincterotomy with the stent
in good position. (Reproduced by permission of Division of Gastroenterology and Hepatology, Johns Hopkins Hospital.)

133
 SECTION 2 TECHNIQUES
adequate drainage from the duct. The edema that ensues following
a pancreatic sphincterotomy can cause ductal obstruction and even-
tual pancreatitis.14 This policy of placing a pancreatic stent after every
pancreatic sphincterotomy, however, is not universal. Some expert
endoscopists do not feel the need to perform this step. Moreover,
the types of stents that are chosen and the desired duration of use
have also been debated.15
Early in the era of pancreatic sphincterotomy, many endoscopists
advocated always performing this procedure in concert with a prior
biliary sphincterotomy. Biliary sphincterotomy done immediately
before pancreatic sphincterotomy is felt by some to allow for easier
identication of clear anatomical landmarks, thus making it a safer
and more effective procedure. It may provide better exposure of the
pancreaticobiliary septum, and therefore allow improved access to
the desired pancreatic tissue.16 Also, this technique prevents the rare
possibility of biliary complications following a primary pancreatic
cut.1 This includes inadvertent damage to the distal bile duct, as well
as possible biliary obstruction due to edema adjacent to the biliary
duct orice. Many expert endoscopists recommend a biliary sphinc-
terotomy before a pancreatic sphincterotomy in cases of cholangitis
or obstructive jaundice, a common bile duct diameter >12 mm, or
an alkaline phosphatase level > twice normal.10 It may also be done
when it is needed to obtain better access to the main pancreatic
duct.17
When performing a pancreatic sphincterotomy after a biliary
sphincterotomy, the anatomical landmarks are different. Part of the
papilla has already been lleted open for this procedure, and so the
pancreatic orice is usually seen at the 5 oclock position near
the rightward margin of the sphincterotomy.1 Transient opening of
the pancreatic duct will allow for better visualization and more accu-
rate cutting. This may be achieved by gently sucking air into the
duodenum with the scope. Once the orice is correctly identied
and cannulated, the sphincterotomy can be carried out in a similar
fashion as described above. See Box 14.1.
Needle-knife sphincterotomy
An alternative method to pancreatic sphincterotomy utilizes an
endoscopic needle-knife instead of a standard pull-type sphinctero-
tome (Fig. 14.1).16 Cutting with the needle-knife is done only after
BOX 14.1 THE PULL-TYPE
SPHINCTEROTOME TECHNIQUE
Direction of the incision is along the 12 oclock position
Pure cutting current should be used
Incision should be slow and hot
Majority of the cutting wire should be outside the papillary
orice
Length of the incision is generally between 5 and 10 mm
A pancreatic duct stent is inserted after completion
May or may not be preceded by a biliary sphincterotomy
134
placement of a pancreatic duct stent. The tip of the needle-knife is
placed at the most proximal portion of pancreatic sphincter tissue
that is overlying the stent. While using the stent as a guide to direct
the cut along the plane of the pancreatic duct, the needle-knife tip
is advanced over the top of the stent and down its longitudinal axis
(Fig. 14.4a14.4b). Incision length is similar to that of sphincterot-
omy with a pull-type sphincterotome; that is, the length is generally
between 5 and 10 mm. Many experts believe that a prior biliary
sphincterotomy is especially helpful before utilizing this needle-
knife technique.16 Good exposure of the pancreaticobiliary septum
allows for better tissue access and more effective septotomy.
There are a few limitations to this technique, however. The abso-
lute prerequisite of pancreatic duct stent placement makes it a tech-
nique that may not be feasible if a stent cannot be placed. For
example, in chronic pancreatitis, it may be very difcult to insert a
stent without rst removing ductal calculi.10 Furthermore, utilizing
the pull-type sphincterotome technique allows a more complete
assessment of sphincterotomy completeness. The endoscopist can
reassess the incision and extend the cut, if necessary, with the
sphincterotome wire. This is not possible with the needle-knife and
stent technique. Lastly, many endoscopists nd it simpler and faster
to perform the sphincterotomy without having to rst place a pan-
creatic stent (see Box 14.2).
Despite the fact that pancreatic sphincterotomy is performed by
only two different techniques, survey questionnaires show that there
is truly a lack of expert consensus in terms of which is the better
approach. A recent survey of 14 expert endoscopists in nine US
centers showed that six of the 14 gastroenterologists either always
or often use the pull-type sphincterotome technique, while 7 out
of 14 always or often use the needle-knife technique.15 Eight
physicians always perform a biliary sphincterotomy prior to pan-
creatic sphincterotomy, and only two of 14 use pure cutting current
during the procedure (Table 14.1). Almost all endoscopists insert a
pancreatic stent after sphincterotomy, as it lowers the likelihood of
post-ERCP pancreatitis.14 However, which types of stents are used
and how long to leave them in place is quite variable among those
who perform pancreatic sphincterotomy on a regular basis.15 Ques-
tions surrounding these differences among techniques can only be
answered with future randomized trials that examine both the short-
term and long-term outcomes of each. See Box 14.3.
Precut pancreatic sphincterotomy
The precut pancreatic sphincterotomy refers to an endoscopic tech-
nique that allows one to gain access to the pancreatic duct without
performing prior deep cannulation. It is usually done when access
to the duct is blocked in some manner (e.g. an impacted stone).9,12
Once the pancreatic duct is nally accessed, conventional pancreatic
sphincterotomy is then able to be performed. Generally, this tech-
nique is not utilized as often as the precut biliary sphincterotomy
since a difcult pancreatic duct cannulation is encountered far less
often than a difcult biliary cannulation. The pancreatic precut is
done in a manner which is very similar to that of the biliary precut
sphincterotomy (see Chapter 12). Most endoscopists will use a free-
hand needle-knife to perform the precut, although there are several
options for this technique.9,18 In the case of a stone that is obstructing
the pancreatic orice, for example, a needle-knife can be used to cut
the papillary mucosa lying directly over the stone. Once the stone is
released and the obstruction is relieved, the pancreatic duct can be
cannulated in the usual manner to prepare for a conventional pan-
creatic sphincterotomy.
 Chapter 14 Pancreatic Sphincterotomy

Fig. 14.4 A Sphincterotomy performed with a needle-knife sphincterotome. A pancreatic stent is placed before the sphincter tissue is
cut. The stent acts as a guide, directing the cut along the plane of the pancreatic duct. Notice the angle of the cut is in the 2 oclock posi-
tion down towards the stent. B Completion of the needle-knife sphincterotomy. (Reproduced by permission of Division of Gastroen-
terology and Hepatology, Johns Hopkins Hospital.)

INDICATIONS FOR PANCREATIC
SPHINCTEROTOMY
Unlike endoscopic biliary sphincterotomy, literature that describes
and validates the indications for pancreatic sphincterotomy is sparse.
There are several reasons for this disparity. First, pancreatic sphinc-
terotomy appears to be mainly performed at specialized referral
centers. Physicians performing this procedure usually have years of
experience in therapeutic biliary and pancreatic endoscopy. In order
to perform these advanced endoscopic procedures with adequate
prociency, the endoscopist must typically practice in an environ-

135
 SECTION 2 TECHNIQUES

BOX 14.2 THE NEEDLE-KNIFE TECHNIQUE 1. EPS as primary therapy

A pancreatic duct stent is always inserted beforehand
The needle-knife is advanced over the top of the stent
Pure cutting current should be used
Length of incision is between 5 and 10 mm
Frequently preceded by a biliary sphincterotomy
Always Often Sometimes
PTS 3 3 7 1
NK 1 6 5 2
EBS 8 4 1 1
PC 2
BC 12
PS 12 2
Table 14.1 Differences in technique of pancreatic sphincterotomy
based on a survey of 14 expert endoscopists. Modied from
Alsolaiman et al.15 with permission
PTS = pull-type sphincterotome, NK = needle-knife technique, EBS = endoscopic
biliary sphincterotomy before pancreatic sphincterotomy, PC = pure cutting current,
BC = blended current, PS = pancreatic stent placement afterwards.
BOX 14.3 CONTROVERSIES
SURROUNDING PANCREATIC
SPHINCTEROTOMY
Pull-type sphincterotome technique vs. needle-knife
technique
Biliary sphincterotomy before pancreatic sphincterotomy?
Blended current vs. pure cutting current
Pancreatic stent vs. no stent after sphincterotomy
If stent, what type of stent? How long should the stent stay
in place?
ment which yields a relatively high volume of ERCP (a workload
volume which is not seen at most centers). This is usually a larger
academic or referral center capable of handling all the possible
complications associated with this procedure. Furthermore, it is the
relatively high likelihood of complications seen with pancreatic
sphincterotomy that creates a general uneasiness among endosco-
pists, thus contributing to an overall decreased number of physi-
cians performing this technique. As a result, there have been fewer
published studies over the years that outline the indications, out-
comes, and safety of pancreatic sphincterotomy. It is on this back-
ground that we discuss the indications for this technique.
sphincter of Oddi dysfunction (SOD)
pancreatic SOD
biliary SOD unresponsive to biliary sphincterotomy
chronic pancreatitis with papillary stenosis/stricture
pancreas divisum (EPS of the minor papilla)
2. EPS to facilitate a further intervention
chronic pancreatitis with ductal strictures or stones treated
with pancreatic stents and/or stone removal
pancreatic pseudocysts treated with transpapillary drainage
resection of an ampullary adenoma
pancreatic stula treated with stent placement
pancreatic disease due to malignancy
primary pancreatic cancer causing strictures, stones,
pseudocysts
metastatic disease to the pancreas causing strictures, stones,
Never pseudocysts
Table 14.2 Indications for endoscopic pancreatic
sphincterotomy (EPS)
Pancreatic sphincterotomy may be indicated for a variety of dis-
eases and disease-related manifestations that involve the pancreas.
In general, it is easiest to think about the indications for pancreatic
sphincterotomy in terms of primary therapy and secondary therapy
(Table 14.2). In other words, this technique may be performed by
itself as the primary treatment modality (i.e. for the treatment of
pancreatic sphincter of Oddi dysfunction); or it may be utilized as a
secondary treatment modality to facilitate a further intervention (i.e.
better access to the main pancreatic duct before dilating a down-
stream dominant stricture). Overall, there is far more data available
regarding the use of pancreatic sphincterotomy in conjunction with
an additional intervention (secondary therapy) than for using this
technique alone (primary therapy).4 Much of the following will con-
centrate on the indications of pancreatic sphincterotomy as a primary
therapy.
Pancreatic sphincterotomy as primary therapy
Pancreas divisum and sphincter of Oddi dysfunction
Most of the literature describing pancreatic sphincterotomy as the
primary endoscopic therapy of choice is concentrated on the area of
pancreas divisum and sphincterotomy of the minor papilla. This
topic is covered at length in Chapter 15. Pancreatic sphincterotomy
has been shown to provide primary therapeutic benet in patients
with pancreatic-type sphincter of Oddi dysfunction (SOD). A brief
review of this disorder is necessary in order to better understand the
role of pancreatic sphincterotomy as its main treatment modality.
SOD is a benign obstruction to the ow of bile or pancreatic juice
at the level of the pancreaticobiliary junction.19 It is due to functional
dyskinesia or hypertension of the biliary and/or pancreatic portion
of the sphincter. It results in transient noncalculous obstruction,
causing abdominal pain or pancreatitis. It can be seen at any age,
but it is most commonly encountered in middle-aged women. SOD
should always be suspected in those patients who are postcholecys-
tectomy and experiencing biliary-type abdominal pain and/or bouts
of acute recurrent pancreatitis. At the present moment, the gold-
standard from making the diagnosis of SOD is biliary or pancreatic
sphincter manometry (Fig. 14.5A14.5B). Sphincter of Oddi manom-
etry involves passing a pressure-sensing catheter through a duode-

136
 Chapter 14 Pancreatic Sphincterotomy

Fig. 14.5AB Pancreatic sphincter of Oddi manometry. The tip of the pressure-sensing catheter lies within the proximal pancreatic (B)
and bile (B) ducts. (Reproduced by permission of Division of Gastroenterology and Hepatology, Johns Hopkins Hospital.)

noscope into the bile duct or pancreatic duct. Pressures can be
measured from both portions of the sphincter (biliary and pancre-
atic) as the catheter is slowly pulled back and positioned within each
one of the sphincter zones (Fig. 14.6).19 Elevated pressures may be
due to either sphincter muscle dyskinesia or structural stenosis.
In pancreatic-type sphincter of Oddi dysfunction, there are essen-
tially three criteria used in making the diagnosis: (a) pancreatic-type
pain, (b) amylase/lipase >1.52.0 times normal, (c) pancreatic duct
diameter >6 mm in the head, or >5 mm in the body (Table 14.3).
Type 1 pancreatic SOD has all three components. Type 2 SOD has
pancreatic-type pain, plus (b) or (c). Type 3 SOD has just pancreatic-
type pain. In terms of biliary-type SOD, the criteria are very similar
but involve the use of serum liver function tests and delayed drain-
age of contrast from the biliary tree during ERCP (Table 14.4).
Isolated pancreatic-type SOD may be seen in 1520% of all
patients with acute recurrent pancreatitis of unknown etiology.19 It
has been estimated to occur in 25% of all patients undergoing
manometry for suspected SOD. The overall clinical response rate of
endoscopic sphincterotomy for SOD (biliary and pancreatic) ranges
between 55 and 95%. Patients with Type 1 pancreatic SOD are most
likely to benet from a pancreatic sphincterotomy. Several studies
have shown that these patients may experience a signicant reduc-

137
 SECTION 2 TECHNIQUES
Fig. 14.6 Biliary sphincter of Oddi manometry. (Reproduced by permission of Gastroenterology and Hepatology, Johns Hopkins
Hospital.)
(a) pancreatic-type pain
(b) amylase/lipase >1.52.0 Xs normal
(c) pancreatic duct diameter >6 mm in the head, or >5 mm in the
body
Type 1 pancreatic-type SOD = (a), (b), (c)
Type 2 pancreatic-type SOD = (a), plus (b) or (c)
Type 3 pancreatic-type SOD = (a) only
Table 14.3 Modied Milwaukee classication for pancreatic-type
sphincter of Oddi dysfunction. Adapted from Novak and Al-
Kawas19 by permission of BC Decker Inc.
(a) biliary-type pain (ROME criteria)
(b) abnormal AST or alkaline phosphatase >2 Xs normal, on two
or more occasions
(c) delayed drainage of contrast from the common bile duct on
ERCP >45 minutes, and a dilated common bile duct >12 mm
Type 1 biliary-type SOD = (a), (b), (c)
Type 2 biliary-type SOD = (a), plus (b) or (c)
Type 3 biliary-type SOD = (a) only
Table 14.4 Milwaukee classication of biliary-type sphincter of
Oddi dysfunction. Adapted from Novak and Al-Kawas19 by
permission of BC Decker Inc.
tion in pain and clinical episodes of pancreatitis. Type 2 pancreatic
SOD may also achieve benet from a pancreatic sphincterotomy, but
most experts prefer to document abnormal pancreatic manometry
before undertaking sphincterotomy. In addition, more recent studies
have suggested a clinical benet from pancreatic sphincterotomy in
138
those patients who have persistent pain despite prior biliary
sphincterotomy.20
Chronic pancreatitis
A pancreatic sphincterotomy alone is frequently used as the primary
treatment modality in moderate to severe chronic pancreatitis. The
rationale for treating chronic pancreatitis with endoscopic therapy is
based on the principle of decreasing pancreatic intraductal pressure.
In moderate to severe disease, the development of ductal stones,
protein plugs, and ductal strictures may occur. Each of these can
cause partial or complete obstruction to the ow of pancreatic juice
out into the duodenum, resulting in permanent alterations to the
duct morphology (Figs 14.7A14.7B, 14.8A14.8B). Ductal obstruc-
tion leads to tissue hypertension, and thus tissue ischemia. Karanja
et al. demonstrated a reduction of pancreatic blood ow after ligation
of the main pancreatic duct (thereby producing intraductal hyperten-
sion) in a feline model of pancreatitis.21 The reduction of blood ow
was partially reversed after relief of the main duct obstruction. It is
strongly believed that the symptom of pain in chronic pancreatitis
is directly due to this parenchymal ischemia.1
Another consequence of obstruction of the main pancreatic duct
is secondary obstruction of the smaller side branch ducts. This ulti-
mately causes parenchymal atrophy. As the tissue begins to atrophy,
the pancreas loses its ability to perform both its endocrine and exo-
crine functions. A therapeutic intervention that could minimize
intraductal pressure might help to prevent this dangerous cascade
of events, thus diminishing pain and preserving pancreatic function.
This is the basis behind sphincterotomy in chronic pancreatitis.
Few studies have specically examined the role of pancreatic
sphincterotomy as the sole endoscopic therapy in chronic pancreati-
Chapter 14 Pancreatic Sphincterotomy

Fig. 14.7 Changes to the ductal

morphology seen in moderate
severity chronic pancreatitis. (Repro-
duced by permission of Division of
Gastroenterology and Hepatology,
Johns Hopkins Hospital.)

Fig. 14.8ab Changes to the ductal

morphology seen in severe chronic
pancreatitis. (Reproduced by per-
mission of Division of Gastroenter-
ology and Hepatology, Johns
Hopkins Hospital.)

139
 SECTION 2 TECHNIQUES
tis. Most studies that have investigated this topic have done so in the
context of additional endoscopic interventions. That is, the sphinc-
terotomy is often performed in conjunction with a further interven-
tion (i.e. stent placement or stricture dilatation). Studies in this area
need to be examined closely in order to separate those patients who
received a sphincterotomy alone versus those who received a sphinc-
terotomy in concert with an additional endoscopic technique. This
is often difcult, especially if the authors have not clearly distin-
guished between the two groups. Nonetheless, several studies have
attempted to evaluate the safety and long-term results of pancreatic
sphincterotomy in chronic pancreatitis.
Ell et al. described pancreatic sphincterotomy in 118 patients with
chronic pancreatitis.22 Eighty percent of the patients underwent a
standard pull-type sphincterotomy, while 20% underwent a needle-
knife technique. Overall, 98% of the sphincterotomies performed
were successful, and the complication rate was only 4.2% (four cases
of moderate pancreatitis, one case of severe bleeding). The results
in terms of pain relief were not examined in this study, however.
Okolo et al. retrospectively analyzed 55 patients who had a pan-
creatic sphincterotomy.23 Forty patients (73%) underwent the proce-
dure for the indication of symptomatic chronic pancreatitis. The goal
of the study was to assess the long-term efcacy of sphincterotomy
with pain relief being the primary endpoint. After a median follow-
up of 16 months, 60% of all patients reported a signicant improve-
ment in their pain scores.
Papillary stenosis appears to be a clear-cut indication for pancre-
atic sphincterotomy alone in those patients with symptomatic
chronic pancreatitis. Without signicant ductal abnormalities distal
to the papilla that require some additional form of intervention,
sphincterotomy can be condently utilized as the primary endo-
scopic therapy of choice in these patients. Similarly, mucinous
ductal ectasia involving the proximal main pancreatic duct for pan-
creatic sphincterotomy has also been proposed as potentially efca-
cious in patients with recurrent pancreatitis.4
Pancreatic sphincterotomy as secondary therapy
Pancreatic sphincterotomy is commonly performed in concert with
other endoscopic techniques such as stent placement or balloon dila-
tation of the main duct. In this setting, the purpose of the sphinc-
terotomy is to help facilitate the primary therapy (i.e. removal of
stones from the duct or dilatation of a ductal stricture). There are
several diseases and conditions in which pancreatic sphincterotomy
is used in this manner (Table 14.2). The decision to cut the sphincter
in these situations is based on sound clinical judgment by the endos-
copist, and whether or not he feels that the risk of a sphincterotomy
is outweighed by the potential benet that may be gained by aiding
the primary therapy.
In moderate to severe chronic pancreatitis, ductal strictures and
stones are often times the norm. Frequently, their location within
the main duct may be very distal to the papilla. Therefore sphincter-
otomy alone may not be sufcient. Stone removal or stricture dilata-
tion may therefore be the main goal of ERCP for certain patients.
Pancreatic sphincterotomy may be needed before the procedure for
better access to the duct (precut), or it can be used simply to help
reduce intraductal hypertension and allow for easier ow of juice
and calculous debris into the duodenum. This also holds true, for
example, when treating pancreatic pseudocysts by means of a trans-
papillary approach. For those pseudocysts that communicate with
the main pancreatic duct, a stent is placed within the duct in order
to bridge the stulous connection.24 A pancreatic sphincterotomy in
140
this setting also helps to reduce intraductal pressures and facilitate
ow out towards the papilla.
Other clinical scenarios for which sphincterotomy has been pro-
posed as secondary therapy include stent placement prior to surgery
for mucinous ductal ectasia, as well as stent placement in the treat-
ment of a pancreatic stula.4 Pancreatic sphincterotomy may also be
used in concert with a pancreatic stent following the resection of an
ampullary adenoma. Here, the purpose of the sphincterotomy (and
the stent) is to reduce the risk of post-procedural pancreatitis due to
periampullary edema. Finally, sphincterotomy is often indicated for
the palliative treatment of strictures, stones, and pseudocysts in
malignant obstruction of the pancreas.
COMPLICATIONS OF PANCREATIC
SPHINCTEROTOMY
Although the rst endoscopic pancreatic sphincterotomy was per-
formed almost 30 years ago, the technique has not been used nearly
as often as biliary sphincterotomy.25 The reason for this is partly due
to past uncertainty about its indications, and also concerns over the
relatively high likelihood of complications related to this procedure.26
When discussing the complications associated with pancreatic
sphincterotomy, it must be remembered that studies which have
evaluated this topic are generally small in number and have a small
number of participants. They are usually performed at expert referral
centers only, and they most often do not have control groups.2 Fur-
thermore, most of the studies report on pancreatic sphincterotomy
as it is used to facilitate other endoscopic maneuvers, such as pan-
creatic stent placement, balloon dilatation, or stone removal. There-
fore it is often difcult to decipher which maneuver is truly
responsible for the complication. For example, is the resultant pan-
creatitis due to the stricture dilatation alone, or the sphincterotomy
that was rst required to access the duct? These are the types of
issues that complicate the literature in this area of study. It is on this
background in which we discuss the complications that are associ-
ated with pancreatic sphincterotomy.
In general, there are essentially three different types of complica-
tions associated with pancreatic sphincterotomy: early, late, and
stent-related complications (Table 14.5).26 Early complications are
usually recognized within the rst 72 hours after the procedure, but
Early complications (<3 months, typically <72 hours)
Pancreatitis
Severe bleeding
Perforation
Pancreatic and/or biliary sepsis
Late complications (>3 months)
Papillary stenosis
Proximal pancreatic duct strictures
Stent-related complications (variable timing)
Ductal and parenchymal changes
Stone formation
Infection
Ductal perforation
Stent migration
Stent occlusion
Duodenal erosion
Table 14.5 Complications of pancreatic sphincterotomy

Author (ref ) n Pancreatitis Total complications
Kozarek et al.17 56 4 (7.1%) 6 (10.7%)
Esber et al.27 236 33 (14%) 37 (15.7%)
Parsons et al.28 31 1 (3.2%) 1 (3.2%)
Table 14.6 Recent studies reporting early complications of
pancreatic sphincterotomy. Adapted from Sherman and Lehman26
with permission
often within the rst few hours. They include pancreatitis, severe
bleeding, perforation, and pancreatic or biliary sepsis. Late complica-
tions are encountered at least three months after the procedure, and
this category mainly consists of papillary stenosis and proximal
ductal strictures. On the other hand, there are several complications
that are stent-related. The timing of their occurrence is variable. They
include pancreatic ductal and parenchymal changes, stone forma-
tion, infection, ductal perforation, stent migration, stent occlusion
(causing pain and/or pancreatitis), and duodenal erosion.
Within the last 12 years, there have been three major studies that
have examined the rates of complication associated with pancreatic
sphincterotomy (Table 14.6).17,27,28 In a study by Kozarek et al. 56
patients underwent a pancreatic sphincterotomy. Fifty-four (96%)
patients had chronic pancreatitis and two patients had acute recur-
rent pancreatitis. The indications for the sphincterotomy were as
follows: obstructing ductal calculi,26 ductal disruption and leak,12
sphincter stenosis,10 and dominant stricture.8,17 Forty-seven patients
had a pull-type sphincterotomy, and 33 of these patients also had a
pancreatic stent placed after the sphincterotomy. Nine patients had
a needle-knife sphincterotomy over an existing pancreatic stent.
Early complications occurred in 10.7% of the patients, and they
included pancreatitis (four patients, or 7.1%) and cholangitis (two
patients, or 3.6%). Late complications, however, occurred in 30% of
the patients; 14% with papillary stenosis and 16% with asymptom-
atic ductal changes (thought to be due to the stent placement).
Esber et al. reported the complications of pancreatic sphincterot-
omy in 236 consecutive patients.27 A pull-type sphincterotomy was
performed in 123, and 87 patients in this group also had a stent
placed following the sphincterotomy. Needle-knife sphincterotomy
over a pancreatic stent was performed in 113 patients. Seventy-four
percent of the patients had a sphincterotomy for the purposes of
treating pancreatic-type SOD, while 26% had chronic pancreatitis
and the procedure was performed to facilitate an additional endo-
scopic maneuver such as removal of stones, stricture biopsy, etc.
Overall, post-ERCP pancreatitis occurred in 14% (mild in 76%, mod-
erate in 21%, and severe in 3%). Other various complications occurred
in only 1.7% of the cases. The rate of pancreatitis was 15.5% in the
patients with pancreatic-type SOD and 9.7% in patients with chronic
pancreatitis. It was suggested that the lower rate of post-ERCP pan-
creatitis in chronic pancreatitis patients was due to all the periductal
brosis and scarring. In other words, the limited amount of nearby
healthy pancreatic parenchyma offers some protection against the
injury that occurs after a pancreatic sphincterotomy.17,26
Parsons et al. evaluated the complication rate of performing a
stentless pancreatic sphincterotomy.26 In 31 patients, sphincterot-
omy was done with a pull-type sphincterotome followed by the place-
ment of a nasopancreatic tube. All the tubes were removed within
24 hours of placement. Post-ERCP pancreatitis was observed in one
patient (3.2%), and there were no other complications seen such as
perforation, bleeding, or sepsis.
Chapter 14 Pancreatic Sphincterotomy
Overall, the rate of pancreatitis following a pancreatic sphincter-
otomy appears to be approximately 1012%, with a total early com-
plication rate (perforation, bleeding, etc.) between 10% and 15%.
Pancreatitis occurs more frequently in those patients with pancre-
atic-type SOD, rather than those who have it performed for problems
associated with chronic pancreatitis. Thorough data concerning the
use of pancreatic stents in the prevention pancreatitis following a
pull-type sphincterotomy is somewhat lacking. Sherman et al.
showed that a pancreatic stent used with needle-knife sphincterot-
omy may limit the frequency of post-procedural pancreatitis in SOD
patients.29 The problem, however, is that if the stent is left in place
for too long, it may begin to induce unwanted ductal and parenchy-
mal changes itself. Also, patients must undergo an additional pro-
cedure to have this endoprosthesis removed unless a 3 Fr prosthesis
is routinely employed.
Pancreatitis is the most concerning potential complication for
those endoscopists who perform pancreatic sphincterotomy. This is
mainly because it appears to be the complication over which we have
the least amount of control, and also because its effect may be very
severe and sometimes lethal. The decision to place a stent following
sphincterotomy is made on a case-by-case basis. Factors weighed in
the decision include the perceived risk of early pancreatitis versus
the potential for late complications and the need for an additional
procedure.
THE COST OF PANCREATIC SPHINCTEROTOMY
There is little published information regarding the cost and cost
savings of pancreatic sphincterotomy. We assume that endoscopic
therapy for diseases such as chronic pancreatitis and sphincter of
Oddi dysfunction ultimately reduce long-term costs by decreasing
the frequency of hospitalizations and the number of required surgi-
cal interventions; yet there is truly a lack of data in the current
medical literature pertaining to this area of interest. The reason for
this may be the enormity and complexity of such a study. In order
to yield useful and effective information, it would need to be con-
ducted over many years, and preferably in several different centers.
Studies with such difcult endpoints and complex variables are less
likely to be initiated by the majority of investigators.
There are some studies, however, that have examined the ability
to safely perform endoscopic pancreatic sphincterotomy on an out-
patient basis. Presumably, the importance of this idea centers around
the reduction of unnecessary overnight observational admissions;
thereby reducing the overall costs associated with therapeutic pan-
creatic endoscopy. Tham et al. reviewed 190 patients undergoing
planned outpatient therapeutic ERCP.30 Five patients had a pancre-
atic sphincterotomy alone, and 28 patients had pancreatic stent
insertion. Admission was necessary in 31 patients (16%). Five of the
31 patients (3% overall) were admitted from home following a
median interval of 24 hours after discharge. The other 26 patients
(13% overall) were admitted directly from the endoscopy unit because
of obvious, discernable post-procedural complications. In the 219
consecutive inpatients undergoing ERCP, there was an overall com-
plication rate of 13%. The authors claim that a policy of selective
outpatient therapeutic ERCP, with admission reserved for those with
established or suspected complication, appears to be safe and reduces
health care costs.30 More studies in this area are needed to assess
issues of cost and safety in outpatient therapeutic pancreatic
endoscopy.
141
 SECTION 2 TECHNIQUES
REFERENCES
1. Cremer M, Deviere J. Chronic pancreatitis. In: Testoni PA,
Tittobello A, eds. Endoscopy in pancreatic disease: diagnosis and
therapy. Chicago: Mosby-Wolfe; 1997:99112.
2. Mergener K, Kozarek RA. Therapeutic pancreatic endoscopy.
Endoscopy 2005; 37(3):201207.
3. Howell DA, Holbrook RF, Bosco JJ, et al. Endoscopic needle
localization of pancreatic pseudocysts before transmural drainage.
Gastrointest Endosc 1993; 39:693698.
4. Elton E, Howell DA, Parsons WG, et al. Endoscopic pancreatic
sphincterotomy: indications, outcomes, and a safe stentless
technique. Gastrointest Endosc 1998; 47(3):240249.
5. Shields SJ, Carr-Locke DL. Sphincterotomy techniques and risks.
In: Cotton PB, Hawes RH, eds. Gastrointestinal endoscopy clinics
of North America. Philadelphia: Saunders; 1996:1742.
6. Batovsky M, Valko L, Paulen P, et al. Prophylactic effects of
ceftriaxone in patients after endoscopic papillosphincterotomy.
Bratisl Lek Listy 1999; 100(3):164167.
7. Cotton PB, Williams CB. Therapeutic ERCP. In: Practical
gastrointestinal endoscopy, 3rd edition. London: Blackwell;
1990:118156.
8. Elta GH, Barnett JL, Brown KA. Pure cut electrocautery current for
sphincterotomy causes less post-procedure pancreatitis than
blended current (abstr). Gastrointest Endosc 1995; 41:A400.
9. Freeman ML, Guda NM. ERCP cannulation: a review of reported
techniques. Gastrointest Endosc 2005; 61(1):112125.
10. Delhaye M, Matos C, Deviere J. Endoscopic management of
chronic pancreatitis. In: Chuttani R, Pleskow DK. Gastrointestinal
clinics of North America. Philadelphia: Saunders; 2003:717742.
11. Jacob L, Geenan JE. ERCP guide wires. Gastrointest Endosc 1996;
43:5760.
12. Maydeo A, Borkar D. Techniques of selective cannulation and
sphincterotomy. Endoscopy 2003; 35(S1):S19S23.
13. Deviere J, Delhaye M. Pancreatic duct stones management.
Gastrointest Endosc Clin N Am 1998; 8:163179.
14. Sherman S, Lehman GA, Hawes RH, et al. Pancreatic ductal stones:
frequency of successful endoscopic removal and improvement in
symptoms. Gastrointest Endosc 1991; 37:511517.
15. Alsolaiman M, Cotton P, Hawes R, et al. Techniques of pancreatic
sphincterotomy: lack of expert consensus. Gastrointest Endosc
2004; 59(5):AB210.
16. Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic pancreatic duct
sphincterotomy: indications, technique, and analysis of results.
Gastrointest Endosc 1994; 40(5):592598.
142
17. Kim MH, Myung SJ, Kim YS, et al. Routine biliary sphincterotomy
may not be indispensable for endoscopic pancreatic
sphincterotomy. Endoscopy 1998; 30:697701.
18. Freeman ML. Precut (access) sphincterotomy. Techniques in
Gastrointestinal Endoscopy 1999; 1:4048.
19. Novack DJ, Al-Kawas F. Endoscopic management of bile duct
obstruction and sphincter of Oddi dysfunction. In: Bayless TM,
Diehl AM, eds. Advanced therapy in gastroenterology and liver
disease. Hamilton: B.C. Decker; 2005:766773.
20. Touli J, Roberts-Thompson IC, Kellow J, et al. Mannometry based
randomized trial of endoscopic sphincterotomy for sphincter of
Oddi dysfunction. Gut 2000; 46:98102.
21. Karanja N, Widdison AL, Leung F, et al. Compartment syndrome
in experimental chronic pancreatitis: effect of decompressing the
main pancreatic duct. Br J Surg 1994; 81:259264.
22. Ell C, Rabenstein T, Schneider HT, et al. Safety and efcacy of
pancreatic sphincterotomy in chronic pancreatitis. Gastrointest
Endosc 1998; 48(3):244249.
23. Okolo PI, Pasricha PJ, Kalloo AN. What are the long-term results
of endoscopic pancreatic sphincterotomy? Gastrointest Endosc
2000; 52(1):1519.
24. Buscaglia JM, Jagannath SB. Endoscopic decompression of
pancreatic pseudocysts. Practical Gastro 2005; 29(3):7483.
25. Cremer M, Deviere J, Delhaye M, et al. Non-surgical management
of severe chronic pancreatitis. Scand J Gastroenterol 1990;
25(S175):7784.
26. Sherman S, Lehman GA. Complications of endoscopic pancreatic
sphincterotomy. In: Testoni PA, Tittobello A, eds. Endoscopy in
pancreatic disease: diagnosis and therapy. Chicago: Mosby-Wolfe;
1997:167171.
27. Esber E, Sherman S, Earle D, et al. Complications of major papilla
pancreatic sphincterotomy: a review of 236 patients. Gastrointest
Endosc 1995; 43:405A.
28. Parsons WG, Howell DA, Qasseem T, et al. Pancreatic duct
sphincterotomy without stenting. Gastrointest Endosc 1995;
41:427A.
29. Sherman S, Eversman D, Earle D, et al. Sphincterotomy by
needle knife over pancreatic stent technique lowers the
post-procedure pancreatitis frequency and severity in sphincter
of Oddi dysfunction patients. Gastrointest Endosc 1996;
43:413A.
30. Tham TC, Vandervoort J, Wong RC, et al. Therapeutic ERCP in
outpatients. Gastrointest Endosc 1997; 45(3):225230.
 SECTION 2 TECHNIQUES
Chapter
Minor Papilla Endoscopic
15 Sphincterotomy
James L. Watkins and Glen A. Lehman
INTRODUCTION
The decision to perform a minor papilla endoscopic sphincterotomy
is a serious one. This chapter assumes indications, patient selection,
pre-ERCP evaluation and informed consent have all been appropri-
ately addressed. The largest category of candidate patients will be
those with idiopathic recurrent chronic pancreatitis and pancreas
divisum.1
Endoscopic visualization aspects of ERCP have improved greatly
in the past decade. This now makes minor papilla and minor papilla
orice identication a simpler task. Accessory tools for cannulation
and sphincterotomy have similarly advanced. Before addressing
minor papilla therapy, usually diagnostic ventral (major papilla) pan-
creatography and dorsal pancreatic ductography via minor papilla
cannulation will have been done as described in Chapter 8.
This chapter is oriented toward minor papilla therapy of pancreas
divisum or incomplete pancreas divisum. The same techniques
apply if minor endoscopic sphincterotomy is needed in patients with
normal main duct/accessory duct anatomy.
The opinions here are based mainly on our team experience
gained from greater than 700 minor papilla sphincterotomies. Where
other case series or scientic data exist, these will be discussed.
INDICATIONS/CONTRAINDICATIONS TO MINOR
PAPILLA SPHINCTEROTOMY
Indications for minor papilla sphincterotomy are similar to indica-
tions for major papilla sphincterotomy (Table 15.1). Most patients
who need minor papilla sphincterotomy will have pancreas divisum.
Table 15.2 reviews contraindications to minor papilla therapy. The
clinician is reminded that errors in patient selection and technique
may cause serious complications including pancreatitis, perforation,
or stricture formation.
SEDATION AND SUPPLEMENTAL DRUGS
Sedation/analgesia for minor papilla cases is the same as for more
difcult or long duration major papilla cases. Approximately 7% of
our cases are done with general anesthesia. Minor papilla cases are
more tedious and a quiet cooperative patient is mandatory.
Drugs to stop motility are needed. Glucagon is routinely given in
0.250.5 mg increments and repeated as often as needed. Secretin
(ChiRhoClin, Inc., Silver Spring, MD) needs to be readily available
(in room, not a pharmacy trip away). Currently available Secretin
gives vigorous pancreatic juice ow approximately 510 minutes
after administration. This Secretin effect usually lasts greater than
30 minutes. Methylene blue (Faulding Puerto Rico, Inc., Aguadillia,
Puerto Rico) should be available to identify the minor papilla or its
orice. It should also be in the endoscopy room for all pancreas
cases.2
ACCESSORIES/GUIDEWIRES
In general, similar accessories are used for major and minor papilla
endoscopic sphincterotomy, i.e. pull sphincterotome or needle-knife.
The exception may be in the very small (or stenotic) minor papilla
orice which requires dilation with a highly tapered push catheter
(e.g. 3-4-5 Fr, Cook Endoscopy, Winston-Salem, NC) (Fig. 15.3)
before passage of a pull sphincterotome.
Minor papilla deep cannulation usually requires an 0.0180.021
diameter guidewire (e.g. RoadRunner, Cook Endoscopy, Winston-
Salem, NC). A very soft tip guidewire is preferred for atraumatic
deep cannulation. The guidewire is preferably passed to the mid
body of the dorsal duct of the pancreas (46 cm into the duct) before
even touching the minor papilla with the catheter or sphinctero-
tome. This helps to avoid minor papilla orice trauma which makes
subsequent cannulation even more difcult.
Pull-type sphincterotome
Any standard wire-guided sphincterotome can be used if the papilla
is rst dilated with a push catheter. We use a 45 Fr diameter short
tipped (23 mm tip extending beyond cutting wire) sphincterotome
and prefer a 2025 mm length cutting wire, but a 30 mm long wire
is acceptable (Fig. 15.4).
Needle-knife
We prefer a needle-knife with a 0.012 diameter cutting wire. The
retractable wire extends 4 mm beyond a 3 Fr insulated coating (Fig.
15.5).
Needle-knives with large diameter cutting wires are discouraged
as they coagulate too much and cut slowly, which may contribute to
stenosis of the pancreatic duct. A needle-knife can be converted from
any commercially available endoscopic pull sphincterotome by
cutting off the plastic tip at the point where the distal cutting wire
inserts into the sphincterotome. This frees the residual bare wire
which is then used for cutting.
Cautery unit
There are no studies which address optimal cautery features for
minor papilla sphincterotomy.3 Our personal experience has favored
use of ERBE model ICC350 (ERBE USA Inc., Marietta, GA) with
setting of Effect 3 and watt setting of 150. If pure cutting current is
143
 SECTION 2 TECHNIQUES

Pancreas divisum with recurrent or disabling symptoms (to

improve dorsal ductal juice ow).
Pancreas divisum with dorsal duct dilation and parenchymal
atrophy (even though without pancreatic symptoms).
Provide access to dorsal ductal system, e.g. for stone removal;
biopsy.
Permit placement of stents larger than 5 French.
Facilitate mucus exiting minor papilla, e.g. mucinous tumor
(Figure 15.1).
To aid guidewire passage out major papilla (guidewire passage
into minor papilla and then out major papilla back into the
duodenum) in patient with failed cannulation of major papilla
and anticipated need for major papilla/main duct at major
papilla/pancreatic therapy.

Table 15.1 Indications for minor papilla sphincterotomy

Fig. 15.2 Minor papilla seen at 6 oclock position (arrow) of

diverticulum. Minor papilla endoscopic sphincterotomy was not
attempted.
Pancreas divisum with minimal or vague symptoms (especially
when associated with normal dorsal duct diameter on CT scan
or MRCP)
Coagulopathy
Inadequate endoscopic view of minor papilla and surrounding
duodenal wall (Figure 15.2)
Operator inexperience
Inadequate informed consent

Table 15.2 Contraindications to minor papilla sphincterotomy

Fig. 15.3 Various tapered-tip catheters and guidewires used in

minor papilla therapy.

Fig. 15.1 Widely patent minor papilla associated with intraductal

mucin producing neoplasm. Minor papilla sphincterotomy done to
facilitate mucus ow. Clinical pancreatitis did not recur while patient
awaited resective surgery.

Fig. 15.4 Short-nosed pull-type sphincterotome.

desired, the setting is Effect 1 with the computer regulated endocut

feature (which alternates cutting and coagulation current) being
shut off. Pure cutting current may induce less scar formation after
sphincterotomy but this has not been evaluated in prospective trials.
Pure cutting current has not decreased pancreatitis rates when used
for major papilla work.4
Stents associated with minor
papilla sphincterotomy
Nearly all needle-knife sphincterotomies are done over a previously
inserted plastic pancreatic stent. Although we have used a variety of
stents ranging from 3 to 7 Fr in diameter,5 we strongly recommend

144
 Chapter 15 Minor Papilla Endoscopic Sphincterotomy

Fig. 15.5 Typical taper-tipped guidewire used in minor papilla

therapy. Needle-knife with 4 mm long cutting wire. Needle-knife
sphincterotome.

Fig. 15.6 Guidewire in minor papilla. Note major papilla in left

lower photo. Planned minor papilla sphincterotomy should pro-
gress along 1011 oclock course to the junction of papillary mound
initial use of 3 Fr stents as they adequately decompress the dorsal and at wall.
duct and induce minimal ductal reaction.68 Unanged stents are
routinely used and spontaneously pass out of the minor papilla in
90% of cases over two weeks. Larger diameter stents are now dis-
couraged as they are unnecessary for prevention of post-ERCP pan-
creatitis and are more hazardous to the dorsal duct.

Choosing between pull sphincterotome and

needle-knife sphincterotome
A pull sphincterotome is preferred if (1) the angle of approach to the
minor papilla is difcult for any reason, e.g. duodenal inammation,
variant anatomy; (2) duodenal motility is difcult to control; (3)
Insertion of multiple pancreatic stents is desired after endoscopic
sphincterotomy.
Needle-knife sphincterotomy performed over a pancreatic stent
has several advantages over the pull sphincterotomy method. The
needle-knife method has the ability to cut more delicately and pre-
cisely without the restriction incurred by the pull sphincterotome.
We suggest that the application of excessive cautery to the terminal
pancreatic duct is less likely (but unproven) with needle-knife tech-
Fig. 15.7 Minor papilla sphincterotomy in setting of redundant
nique. Pull sphincterotomy more commonly passes >5 mm cutting folds. The cut was in the 1112 oclock direction. The upper margin
wire into the orice and extends the thermal effect into the pancreas, of the papillary mound is less certain.
not just the duodenal wall sphincter zone. However, a recent
database review from our unit showed that >80% of the last 100
minor papilla sphincterotomies were done by the pull technique. begins, the foot pedal is continuously activated until the cut is nearly
complete. The last 1 mm cut is performed again with 12 taps on
Minor papilla endoscopic sphincterotomy the cautery activation foot pedal. If no cutting is initiated (only white
techniquepull type char) after 23 taps followed by 1 second of continuous activation,
Once a 0.0180.025 inches guidewire is securely advanced intraduct- then all cutting is stopped and the remainder of the sphincterotomy
ally to at least the mid body of the dorsal duct, the sphincterotome is performed using the needle-knife method. The cut is extended
is advanced with the cutting wire oriented toward the 1011 oclock into the 1011 oclock direction as directed by the intraductal wire.
position. This is often the unadjusted orientation after sphinctero- Pulling back on the bowed sphincterotomy helps expose the avail-
tome passage. Alternatively, this may require the long scope posi- able cutting space (Fig. 15.8).
tion. The pull back short scope position tends to cut more toward The sphincterotome is bowed so as to achieve moderate tension
the 1112 oclock position. The latter cut tends to produce more on the tissue. This allows more rapid cutting with less char. Inade-
coagulation and probably results in an overall smaller orice. The quate wire contact results in inadequate cutting and may result in
papilla is positioned in the lower to mid visual eld on the video excess coagulation. Precise cutting requires a relatively dry eld.
monitor (Figs 15.615.7). This allows for better view of the target Repeated uid aspiration may be required. Excessive bile, pancreatic
tissue. The pull endscopic sphincterotome is then positioned with juice or other uid in contact with the cutting wire will divert current
34 mm of cutting wire into the orice. The cut is initiated with 23 away from the target tissue (Fig. 15.9). Such a pool of uid may boil
half-second taps on the cautery activation foot pedal. If cutting and broadly coagulate the surrounding tissue. This may result in

145
 SECTION 2 TECHNIQUES

A B

Fig. 15.8 A Pulling out (back) on pull sphincterotome helps to

better delineate papillary mound and cutting zone. B Pull out
view accentuating the minor papilla mound and better dening the
safe cutting zone.

Fig. 15.10 Minor papilla sphincterotomy by pull technique

sequence. With placement of 5 French stent, our current recom-
mendation is to use small caliber stents, e.g. 3 French.

Fig. 15.9 Minor papilla after ~80% of mound has been cut toward
the 1011 oclock position. After suctioning bilious uid, the upper
rim of mound was cut.

serious stricture formation. Fluid pooling is especially problematic

if secretin was given to aid orice identication. Pancreatic stent
placement with juice diversion and needle-knife sphincterotomy are
then recommended.
The goal is to cut to the upper rim of the minor papilla mound
per se. Alternatively, some of our team members prefer to stop
1 mm short of the minor papilla-at wall junction (Figs 15.10, 15.11).
We attempt to orient the cutting wire perpendicular to the duodenal
wall. This helps to limit excessive thermal effect to the intramural
portion and avoid ductal injury. If the minor papilla mound is very
small, the cut may extend 12 mm cephalad to the mound. With a
small minor papilla, the completed cut minor papilla allows an
approximately 1/41/3 bowed sphincterotome to pull through
the sphincter. For a larger papilla, a 1/31/2 bowed 25 mm wire
sphincterotome will pull through the sphincterotomy with only mild
resistance. To view the ductal lumen per se is usually only possible
with large papilla cases.
After pull endoscopic sphincterotomy, a temporary pancreatic
stent is placed in nearly all cases. Occasionally, in advanced chronic
pancreatitis, with very dilated dorsal pancreatic duct (>8 mm), stent
Fig. 15.11 View of minor papilla 19 months after endoscopic
sphincterotomy. Note no papillary mound remains. The cut
extended up to the at wall.
placement is not needed. Over the last 15 years, stent experience has
now led to preferred use of very small diameter (3 Fr) 68 cm long
polyethylene stents with no intraductal barb.5 We prefer a 3/4 pigtail
for the duodenal end but double barbs in duodenum are acceptable.
The intraductal length serves as a friction anchor (without barb) yet
the small diameter causes less ductitis than larger diameters. We
have abandoned 23 cm 5 Fr stents as the intraductal ends cause
more focal ductitis (and occasionally strictures). This occurs espe-
cially when the rm stent tip abuts the wall perpendicularly at an
angulation in the duct.

146

A follow-up plain abdominal x-ray must be obtained in approxi-
mately two weeks to conrm spontaneous stent passage. Residual
intraductal stents (510% of patients) require endoscopy with stent
extraction. Patients who refuse follow-up are problematic. For
patients who are predicted to be unreliable at follow-up, alternatives
include: (1) hospitalization and stent removal 23 days later (if not
passed spontaneously), or (2) placement of a nasopancreatic catheter
for a duration of 23 days. The minimal interval for effective post
sphincterotomy protective pancreatic stenting has not been dened
but we suspect it is 23 days.
Needle-knife sphincterotomy over
pancreatic duct stent
After wire passage deep into the dorsal duct, a pancreatic stent is
placed as noted above. Care is taken during nal stent delivery (nal
guidewire withdrawal) to maneuver the pigtail toward the descend-
ing duodenum (downward) (Fig. 15.12). This aids minor papilla
viewing during cutting and at times lifts the minor papilla up for
easier cutting. For cutting, the minor papilla is positioned in the
lower to mid visual eld on the video screen. This allows better view
of the upper rim of the minor papilla.
The needle-knife should be maneuvered over the cutting zone of
the minor papilla orice to the upper extent of the minor papilla
mound in a dry run fashion being sure that the accessory channel
A B
Fig. 15.12 A Pigtail stent placement with downward placement of
pigtail. This facilitates cephalad view of cutting space. B Cutting
with needle-knife over stent.
Chapter 15 Minor Papilla Endoscopic Sphincterotomy
elevator permits a full extent excursion of this distance. The cut is
then initiated (usually start at orice then cephalad) by lightly
hooking (embedding) the terminal 12 mm of needle-knife wire into
the minor papilla orice (Fig 15.13). Light, but denite cephalad
tension is placed on the wire and current activation done. We gener-
ally use continuous foot pedal activation with the ERBE cautery
generator while sweeping the needle-knife cephalad over nearly the
full extent of the minor papilla mound. This sequence is repeated
cutting another 12 mm deeper into the mound and exposing more
intramural stent. The cut tissue tends to retract exposing the deeper
tissue. The cut is extended cephalad until the upper mm of
the minor papilla mound is divided. The needle-knife catheter
(with needle retracted) is then passed alongside the stent into the
pancreatic duct. If the orice does not permit this, cutting 12 mm
of the tissue touching the upper rim of the stent is done until can-
nulation permitted. The above cutting is done in the 1011 oclock
directionprecisely over the stent. An alternative cutting method
is to start at the apex of the anticipated cut and cut downward onto
the stent. This more precisely denes the most cephalad extent of
the cut. Unintentionally cutting more cephalad than desired can
more easily be avoided.
PRECUT SPHINCTEROTOMY TECHNIQUE
Precutting refers to cutting the minor papilla without prior deep
cannulation, wire passage, or stent placement (Fig. 15.14). Precut-
ting is more risky than the above techniques because if cannulation
fails there is no protective pancreatic duct stent. Careful risk:benet
analysis should therefore precede any decision to initiate precut
sphincterotomy. Three subcategories exist: (1) minor papilla orice
evident; (2) no minor papilla orice seen; (3) Santorinicele present.
Precut minor papillaorice seen
In such cases, the orice usually can be entered 12 mm even if
not deeply cannulated. In most cases, Secretin (ChiRhoClin, Inc.
Silver Spring, MD) 0.2 mcg/kg IV has already been given to aid
orice identication. The needle-knife cutting wire is extended 1
2 mm and impacted in the orice cephalad rim. Cutting is then done
toward the 1011 oclock direction in identical fashion as to needle-
knife over stent. We prefer to cut all the way to the rim of the papilla.
A second cut is then made 12 mm deeper into the base of the rst
incision or slightly left or right of the rst incision. If secretin was
Fig. 15.13 Diagram of needle-knife over
pancreatic stent technique.
147
 SECTION 2 TECHNIQUES
A B
C D
Fig. 15.14 A Minor papilla with orice seen (arrow) as pink dot in
blue background. Secretin was given 10 minutes prior to aid orice
identication. B After precut minor papilla sphincterotomy.
C Deep cannulation achieved and sphincterotomy completed
by pull technique. D 3 French plastic stent placed after
sphincterotomy.
Fig. 15.15 Initially minor papilla not found. The duodenum was
then washed with a methylene blue solution. Dot of pancreatic
juice seen at minor papilla orice after Secretin stimulation
(arrow).
not previously given, or its effect has worn off, repeat dosing is done
and the surface of the precut area is washed with a solution of 1 ml
methylene blue in 9 ml water with a few drops of simethicone solu-
tion (antifoaming properties). Pancreatic juice ow, if present, will
be seen as a tiny clear uid spot or stream amidst the blue stained
background mucosa (Fig. 15.15). Deep cannulation is then gently
148
Fig. 15.16 Santorinicele. Dorsal ductogram showing saccular dila-
tion of downstream terminal end of duct.
attempted with a soft tip guidewire (e.g. Metro 0.0210.025 inches
or a blunt tip guidewire such as 0.035 Teon-coated stainless steel
coil wrapped wire (Cook Endoscopy, Winston-Salem, NC)). Gentle
probing is mandatory and helps avoid bleeding, which may obscure
landmarks. Probing should be toward the 1011 oclock direction
varying from 45 to 90 degrees (perpendicular) to the wall. If cannula-
tion fails and the precut area maintains a clear view, additional
shallow cuts of 34 mm length and 12 mm depth can be made on
each side of the original cut or into the base of the rst cuts. Secretin
use and gentle probing continue until successful ductal entry or ter-
mination of the procedure. With ductal entry, completion sphincter-
otomy is done with either pull technique or needle-knife over stent
technique as needed above. If no ductal entry is achieved, repeat
cannulation attempt is best delayed at least 4 weeks until complete
healing of the precut area is achieved, if the clinical conditions
permit such waiting time.
Precutno orice seen
When considering precutting and no orice (and usually no juice
ow) can be identied, review carefully whether pancreas divisum
or need for minor papilla therapy really exists. If suspicion/need
remains high, proceed as follows:
Precut identically as above except start on the lower rim of the
minor papilla or the point on the minor papilla that the orice seems
most likely. Cut 23 mm length, or preferably to the upper rim of
the minor papilla. Make 1or 2 cuts, probe gently, administer Secretin
if needed, apply methlylene blue, cut 12 more shallow incisions,
etc. All other aspects are as described in the preceding section above.
Occasionally, aps of tissue between incisions will need removal
with small biopsy forceps.
Precutwith Santorinicele
Approximately 15% of minor papillae have saccular dilation of the
terminal dorsal duct beneath the duodenal mucosa and papillary
mound (Fig. 15.16).8 This has been named a Santorinicele (after the
accessory duct of Santorini). In such cases, the mound (bulge) is
usually prominent, especially after contrast lling of the dorsal
ductal system or after Secretin stimulation. Five to ten minutes after
Secretin, the bulge is usually maximal and may have a bluish color.
The wall thickness, (distance between the duodenal lumen and sac-
cular lumen) is usually less than 2 mm. Therefore a 23 mm long
12 mm deep needle-knife incision over the dome (center) of the
bulge will usually enter the ductal lumen. Occasionally a second
deeper incision is needed. Completion sphincterotomy can be done

A B
Fig. 15.17 A Major and minor papilla orices two years after
sphincterotomies. Slit-like minor papilla orice (upper arrow) seen
with no residual mound. B Due to recurrent pancreatitis episode,
a 4 mm balloon dilation done and minimal further cutting done.
Fig. 15.18 Re-do minor papilla sphincterotomy case. After pull
sphincterotome extension of cut to the at duodenal wall, two 5
French internally unanged plastic stents placed.
by pull or needle-knife technique. The nal sphincterotomy length
is often larger (58 mm long) if the bulge was large.
Re-do minor papilla sphincterotomy
Re-do casesin which there has been a previous minor papilla
sphincterotomyare more difcult, as the papillary mound has
usually been nearly or fully ablated (Fig. 15.11). If residual cutting
space is clearly present, we proceed with sphincterotomyeither
pull or needle-knife. If there is no obvious residual cutting space
seen, we prefer to dilate the orice using a push dilator to 7 Fr (or
larger if the dorsal duct is >4 mm) or with a 4 mm balloon. After
dilation, a small cut zone may become evident. If intramural cutting
space is then seen, pull or needle-knife sphincterotomy extensions
can be done with the cut extending 12 mm up onto the at wall
(Fig. 15.17). In re-do cases we commonly place two stents of 35 Fr
diameter, then cut 23 mm further from the upper rim (Fig. 15.18).
Care is taken to cut tissue and coagulate minimally. Excess coagula-
tion current probably contributes to scar formation and re-stenosis.
The optimal follow-up duration of stenting is unknown. We prefer
unanged stents and check for passage one to two months later
using a plain abdominal radiograph. If there is no cutting space,
dilation followed by stenting alone is done.
Precise methods to detect when the minor papilla orice is still
too tight have not been dened. We often use manometry tech-
niques identical to those for the major papilla and use a pressure of
40 mmHg as the cut-off for upper limits of normal for basal sphincter
Chapter 15 Minor Papilla Endoscopic Sphincterotomy
Fig. 15.19 Standard 5 Fr triple lumen aspiration type manometry
catheter (courtesy of Cook Endoscopy, Winston-Salem, NC) used
to measure minor papilla sphincter pressure in this re-evaluation
case. Prior sphincterotomy was 2 years ago.
pressure9 (Fig. 15.19). The rationale for use of the same cut-off
values is that the pancreatic parenchyma probably does not tolerate
secretion against a barrier >40 mmHg, whether through the major
or minor papilla. Alternatively, the orice can be probed with a 5, 6,
and 7 Fr tapered catheter over a guidewire. If more than minimal
resistance to passage is detected (similar to passage of esophageal
dilators) by the 6 Fr catheter, the manometry will usually detect
>40 mmHg basal sphincter pressure.
OUTCOMES
Efcacy of minor papilla therapy is variable according to the underly-
ing disease state1015 (Table 15.3). Patients with acute recurrent pan-
creatitis are most likely to benet from minor papilla intervention
with 77% of 164 patients in Table 15.3 followed up for more than
two years having no or fewer hospitalizations from pancreatitis
ares. We offer minor papilla therapy to patients with a clinical
course of acute recurrent pancreatitis if they have had at least two
hospitalizations. Therefore, we usually do not offer minor papilla
therapy after a single bout of pancreatitis (unless the CT scan shows
evidence of dorsal duct dilation/obstruction/or stones). Long-term
(1020 years) outcomes are awaited. Patients with chronic pain alone
(but often with endoscopic ultrasound evidence of chronic pancre-
atitis) and patients with chronic pancreatitis changes by ERCP or
computed tomography have improvement rates of 3040% (Table
15.3). These latter patients are often disabled and on chronic narcot-
ics. Such desperate patients make us more willing to offer trial
therapy even if results are suboptimal. Preliminary results from a
prospective randomized trial of medical vs endoscopic therapy for
pain only patients with pancreas divisum has been reported.16 This
trial is ongoing.
Complications of minor papilla sphincterotomy
Acute complications of minor endoscopic sphincterotomy are similar
to major endoscopic sphincterotomy except perforations are less
frequent. Pancreatitis rates are similar to other high-risk groups
such as idiopathic pancreatitis and sphincter of Oddi dysfunction.
Table 15.4 summarizes our large series.
Management of complications is nearly identical to that of major
papilla sphincterotomy. We have little experience with management
of perforations, but would expect most to be small size and easily
149
 SECTION 2 TECHNIQUES

ACUTE RECURRENT CHRONIC

PANCREATITIS PAIN ONLY PANCREATITIS

Mean
Author/year Therapy follow-up (mo) n % improved* n % improved* n % improved*
Coleman/1994 MiES/stent 23 9 78 5 0 20 60
Sherman/1994 MiES 28 0 16 44 0
Kozarek/1995 MiES/stent 20 15 73 5 20 19 32
Heyries/2002 MiES/stent 39 24 92 0 0
Bierig/2003 MiES 19 16 94 7 43 16 38
Linder/2003 MiES NG (range 1120) 38 58 12 0 4 25
Borak/2005 MiES 44 62 89 29 69 23 43
Total 28 164 77 74 33 82 41

Table 15.3 Selected series of endoscopic minor papilla therapy for pancreas divisum
MiES = Minor papilla sphincterotomy.
% of patients improved assessed by overall clinical status.
NG = Not given.
W:P:MiPaRxPD

Pancreatitis Total 13.2% aware of one center which does so (without published supportive
data).
Mild 10.9% The most important long-term complication of minor papilla
Moderate 1.9%
sphincterotomy is re-stenosis. This occurred in nearly 20% of one
Severe 0.4%
tallied series.18 Whether this represents natural healing of a congeni-
Bleeding Total 1.0% tally small orice or scar formation after cautery and manipulation
Mild .6% is unknown. Such re-stenosis can be difcult to manage, as the ste-
Moderate .1% notic zone may extend 25 mm outside the duodenal wall and into
Severe .3% the head of the pancreas. A randomized trial of steroid injection
(20 mg triamcinolone) into the sphincterotomy zone numerically
Perforationc Total .2%
(not statistically signicantly) decreased the re-stenosis rate from 23
Mild .1% to 15% over a mean follow-up period of 47 months.18 Cutting space
Moderate .1% is often not present for either extension of the sphincterotomy or
surgical sphincteroplasty. We offer such patients a series of sequen-
Table 15.4 Complicationsa of a minor papilla sphincterotomyb in tial plastic stents or surgical decompression (usually lateral pancre-
778 procedures in pancreas divisum (from reference no. 17 with
aticojejunostomy).19 Our goal is to progressively dilate the narrowing
permission)
a
Denitions per referenceCotton et al.17 until at least two 6 French stents can be placed side by side, left in
b
Database search of 12 year experience at Indiana University Medical Center. situ for 12 months, and subsequently removed (if not passed spon-
c
Intraprocedure only (does not include delayed stent induced perforations). taneously). Outcome from a series of such patients has not yet been
tallied. Trials of pure cutting current sphincterotomy to prevent
managed with pancreatic stent and nasoduodenal decompression. re-stenosis are awaited.
Immediate bleeding during cutting will resolve spontaneously in
>80% of cases. If bleeding occurs during needle-knife over stent
technique, the pancreatic duct is already stented and protected. This SUMMARY
allows use of epinephrine 1 : 10 000 injection (0.52 ml/injection)
into/around the bleeding site. If bleeding persists, bipolar coaptive
The techniques of minor papilla sphincterotomy have been
cautery is done at setting of 1520 watts. Care is taken to cauterize
derived largely from biliary and major papilla pancreatic
focally (not diffusely) in order to limit subsequent scar formation.
sphincterotomy. Therapeutic efcacy and complication rates
Persistent bleeding is usually a reection of a coagulopathy which
are similar to major papilla techniques. It is recommended that
may then need correction.
these techniques be performed by endoscopists at referral
Pancreatic stenting has decreased (but not eliminated) the fre-
centers with large ERCP volumes. Long-term outcomes from
quency and severity of post-ERCP pancreatitis, including for minor
such intervention and comparative randomized trials with
papilla settings. If pancreatitis occurs after a stent is placed, pancre-
stenting alone20 or surgical sphincteroplasty would be of
atitis is mild and resolves in 13 days. More severe pancreatitis
interest.
warrants a CT scan to exclude simultaneous post-sphincterotomy
perforation or stent-related ductal perforation. Such stent perfora-
tions may occur at sharp angulations of the main duct or out a side
branch. Such stents need urgent endoscopic stent removal or pull Acknowledgement
back. Pancreatitis may occur with early (>24 hr) spontaneous stent The authors are grateful to Joyce Eggleston for the technical compila-
passage. In this situation we do not usually replace the stent but are tion of this document.

150

REFERENCES
1. Fischer M, Fogel EL, McHenry L, et al. ERCP/manometry in
1108 idiopathic pancreatitis patients. Gastrointest Endosc 2005;
61:190A.
2. Park S-H, de Bellis M, McHenry L, et al. Use of methylene blue to
identify the minor papilla or its orice in patients with pancreas
divisum. Gastrointest Endosc 2003; 57:358363.
3. Alsolaiman M, Cotton, P, Hawes R, et al. Techniques for pancreatic
sphincterotomy: Lack of expert consensus. Gastrointest Endosc
2004; 59:AB210.
4. Gorelick A, Cannon M, Barnett J, et al. First cut, then blend: an
electrocautery technique affecting bleeding at sphincterotomy:
Randomized controlled trial. Endoscopy 2001; 33(11):976980.
5. Rashdan A, Fogel EL, McHenry L, et al. Improved stent
characteristics for prophylaxis of post-ERCP pancreatitis. Clin
Gastrointest Hepatol 2004; 2:322329.
6. Smith M, Ikenberry S, Uzer M, et al. Alterations in pancreatic
ductal morphology following pancreatic stent therapy.
Gastrointest Endosc 1996; 44:268275.
7. Sherman S, Alvarez C, Robert M, et al. Polyethylene pancreatic
duct stent-induced changes in the normal dog pancreas.
Gastrointest Endosc 1993; 39:658664.
8. Eisen G, Schutz S, Metzler D, et al. Santorinicele: new evidence for
obstruction in pancreas divisum. Gastrointest Endosc 1994;
40:7376.
9. Fogel EL, Sherman S, Kalayci C, et al. Manometry in native minor
papillae and post minor papilla therapy: experience at a tertiary
referral center. Gastrointest Endosc 1999; 49:187A.
10. Coleman SD, Cotton PB. Endoscopic accessory sphincterotomy
and stenting in pancreas divisum. Gastrointest Endosc 1993;
39:312.
Chapter 15 Minor Papilla Endoscopic Sphincterotomy
11. Linder JD, Bukeirat FA, Geenen JE, et al. Long-term response to
pancreatic duct stent placement in symptomatic patients with
pancreas divisum. Gastrointest Endosc 2003; 57:208A.
12. Lehman GA, Sherman S, Nisi R, et al. Pancreas divisum: results of
minor papilla sphincterotomy. Gastrointest Endosc 1993; 39:18.
13. Heyries L, Barthet M, Delvasto C, et al. Long-term results of
endoscopic management of pancreas divisum with recurrent
acute pancreatitis. Gastrointest Endosc 2002; 55:376381.
14. Bierig L, Chen YK, Shah RJ. Patient outcomes following minor
papilla endotherapy (MPE) for pancreas divisum (PD). Gastrointest
Endosc 2006; 63:313A.
15. Borak G Alsolaimon M, Holt E, et al. Pancreas divisum: long-term
follow up after endoscopic therapy. Gastrointest Endosc 2005;
61:149A.
16. Sherman S, Hawes R, Nisi R, et al. Randomized controlled trial of
minor papilla sphincterotomy (MiES) in pancreas divisum (Pdiv)
patients with pain only. Gastrointest Endosc 1994; 40:125A.
17. Cotton PB, Lehman GA, Vennes J, et al. Endoscopic
sphincterotomy complications and their management: an attempt
at consensus. Gastrointest Endosc 1991; 37(3):383393.
18. Toth TG, Sherman S, Fogel EL, et al. Does intrapapillary steroid
injection improve the efcacy of minor sphincterotomy in
pancreas divisum? Gastrointest Endosc 2001; 53:60A.
19. Madura JA, Canal DF, Lehman GA. Wall stent-enhanced lateral
pancreaticojejunostomy for small-duct pancreatitis. Arch Surg
2003; 138:644650.
20. Ertan A. Long-term results after endoscopic pancreatic stent
placement without pancreatic papillotomy in acute recurrent
pancreatitis due to pancreas divisum. Gastrointest Endosc 2000;
52:914.
151
 SECTION 2 TECHNIQUES
Chapter
Plastic Pancreatic and Biliary Stents:
16 Concepts and Insertion Techniques
Todd H. Baron and Jeffrey L. Ponsky
INTRODUCTION AND SCIENTIFIC BASIS
The use of plastic biliary stents for drainage of the bile duct was
described over two decades ago1 and plastic stents are now used in
the biliary tree for a variety of therapeutic indications.2 Applications
in pancreatic disease have also developed.2 These stents are used for
malignant and benign conditions and have proven reliable and safe
in decompression of the biliary tree. Palliative insertion of biliary
stents relieves distal biliary obstruction as effectively as surgical
bypass.3 Plastic stents are available in a variety of congurations and
lengths and are composed of Teon, polyethylene, or polyurethane
(Tables 16.1 and 16.2).2 Common congurations are straight, single
pigtail, or double pigtail (Fig. 16.1A16.1B). All plastic stents have
limited patency due to occlusion with debris and biolm (Fig. 16.2)46
and require periodic replacement when long-term drainage is
required. Nearly all stents of the same diameter have similar patency
rates. One 10Fr stent with a unique double layer design (Fig. 16.3)
was shown in one study to have prolonged patency as compared to
standard stent design.7 Plastic stents have been demonstrated to be
easy to insert, effective for decompression, and inexpensive to use.
Almost all plastic stents are hollow tubes. Side holes are present
to a variable degree, but uniformly present in pancreatic duct stents
to allow side branches to drain (Fig. 16.4). Recently, a star-shaped
stent with a limited central lumen (Fig. 16.5) has become available
for both biliary8 and pancreatic insertion (Viaduct, GI Supply,
Camp Hill, PA).9 The central lumen allows only a guidewire and is
inserted without an inner guiding catheter even at 10 Fr diameter
(see stent systems below). A new S-shaped pancreatic stent has been
introduced which may have less potential for migration and a pro-
longed patency.10 Finally, a new biliary stent with an antireux valve
(wind-sock) has recently become available to prevent stent occlusion
due to food and vegetable material (Cook Endoscopy, Winston-
Salem, NC) (Fig. 16.6). This stent may have improved patency over
conventional large bore 10 Fr stents.11
STENT SYSTEMS
A variety of stent systems are available as discussed in Chapter 4.
Stents of less than 8.5 Fr diameter are placed directly over a guide-
wire using a pusher tube or sphincterotome. Stents of greater than
8.5 Fr diameter typically come with an inner guiding catheter which
passes over the guidewire (Fig. 16.7); the stent and pusher tube are
then passed over the inner guiding catheter (Fig. 16.8). The inner
guiding catheter promotes stability and rigidity which are necessary
in passing through tight strictures.
Endoscope requirements
For stents of 7 Fr in diameter a diagnostic channel endoscope can
be used. However, nearly all modern duodenoscopes are equipped
with a therapeutic channel (4.2 mm) which can accommodate stents
up to 11.5 Fr in diameter.
Description of technique: biliary
Because 10 Fr stents have a patency that is superior to 7 Fr stents,
it is recommended that all patients with malignant disease who are
undergoing plastic stent placement receive 10 Fr stents, if possible,
so as to limit the number of endoscopic procedures required for
long-term palliation.
Distal biliary obstruction
The approach to distal biliary strictures is slightly more straightfor-
ward than for hilar tumors and will be discussed separately. After
successful deep cannulation of the biliary tree contrast is introduced
to clearly elucidate the margins of the stricture to allow the appropri-
ate stent length to be chosen. The stricture is traversed with a guide-
wire. It is important to pass the wire well proximal to the stricture
to prevent wire loss and to provide mechanical advantage. In general,
a biliary sphincterotomy is not required for successful single 10 Fr
stent insertion.12 If multiple stents are to be placed, however (for
example in patients with benign disease in whom multiple stents
are required), a biliary sphincterotomy is required.
The guidewire is left in place. For placement of a single 10 Fr
stent across a distal biliary stricture it is rarely necessary to dilate the
stricture, since the mechanical advantage is great enough to over-
come resistance. In cases of uncertainty, a 10 Fr dilating catheter
(e.g. Soehendra dilator, Cook Endoscopy, Winston-Salem, NC) can
be passed. If it traverses the stricture, then a 10 Fr stent will also
traverse the stricture. Otherwise, hydrostatic balloon dilation can be
performed. When the insertion of multiple stents is planned, stric-
ture dilation is essential. In this setting, additional guidewires may
be placed prior to placement of the rst stent or may be passed
alongside the rst stent after its placement.
When multiple stents are placed, it may be useful to place a
slightly longer stent rst as the friction of the second stent insertion
may result in upward movement of the stent. If the rst stent is too
short it may disappear into the duct. This is usually of no conse-
quence assuming that the stent is still across the stricture. The
stents length should be selected based upon the distance from
the papilla to the proximal edge of the stricture plus an additional
2 cm. Excessively long stents should be avoided as migration
tends to occur until the proximal end of the stent impacts the top
of the stricture; meanwhile, the distal end of the stent may then
153
 SECTION 2 TECHNIQUES

MANUFACTURER/SHAPE

ConMed Hobbs Medicalb Microvasive Olympus Cook Endoscopy

ACS DP ACS DP ACS DP ACS DP ACS DP
Size (F)a
5
6
7
8.5
10
11.5
12
Length (cm)
1
2.5
3
4
4.5
5
6
6.5
7
8
8.5
9
10
10.5
11
12
12.5
13
14
15
>15
Material
Nylon
Polyethylene
Polyurethane c
Teon
Two layer
Operator centered system Price No No Yes Yes Yes
Stent 60 40 69 4547 57
With delivery system 115130 86 119159 117198 123
With operator centered system N/A N/A 139 N/A 123

Table 16.1 Plastic biliary stents (from reference no. 2 with permission)
ACS, Angled, curved, or straight; DP, double pigtail.
a
Stents >10 Fr require a 4.2 mm channel duodenoscope.
b
Hobbs Medical did not disclose their stent material for this review.
c
Covered with hydromer coating.

impact the opposite duodenal wall causing perforation (Figs. 16.9A
and 16.9B). As a rule of thumb, most pancreatic cancers producing
biliary obstruction will be adequately managed with 5 or 7 cm long
stents. Measuring of the stricture can be achieved in several ways.
One way is during withdrawal of the initial cannulating catheter.
When the catheter is at the proximal end of the stricture, the endos-
copist holds the catheter just outside of the biopsy port; the catheter
is then withdrawn until it is seen just outside of the papilla. The
distance from the endoscopists ngers to the biopsy port is mea-
sured, which is the minimal length of stent required to cross the
lesion. Another way is to use the radiograph to measure the length
from the top of the stricture to the tip of the endoscope when pressed
against the papilla; the diameter of the endoscope serves as the
comparison measuring point to account for the magnication factor.
The following equation can be used to solve for the unknown (stric-
ture length, Figure 16.10A16.10B):

154
 Chapter 16 Plastic Pancreatic and Biliary Stents: Concepts and Insertion Techniques

MANUFACTURER/SHAPE

GI Supply Hobbs Medical Olympus Cook Endoscopy

Feature S SP S SP S SP S SP
Size (F)
3 5, 7, 9
4
5 315
7
Length (cm)
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Material Polyurethane Not availablea Polyethylene Polyethylene
Price ($) (stent/preloaded) 58 40 49 57/123

Table 16.2 Pancreatic stents (from reference no. 2 with permission)

S, Straight; SP, single pigtail.
a
Hobbs Medical did not disclose their stent material for this review.

A B

Fig. 16.2 Endoscopic photo of occluded 10 Fr stent exiting the

bile duct.

Fig. 16.1 Various stents: A Straight 10 Fr biliary stent (courtesy

of Olympus America Inc., Melville, NY). B Double pigtail 10 Fr
stent (courtesy of Cook Endoscopy, Winston-Salem, NC).

155
 SECTION 2 TECHNIQUES

Fig. 16.3 Double-layer design (Olympus).

Fig. 16.7 Cook Endoscopy stent system showing typical 10 Fr

design. Inner guiding catheter (arrows), stent (blue) and pusher
tube (arrowheads) are seen.

Actual stricture length (X) Actual endoscope diameter

=
Measured stricture length Measured endoscope diameter

Finally, uoroscopic markers separated by a known distance are

Fig. 16.4 Pancreatic duct stent, Cook Endoscopy. Note side
holes.
Fig. 16.5 Magnied photograph of a sagittal section of a Viaduct
stent. The ow is through the six channels (C) rather than through
the central guidewire lumen (L).
Fig. 16.6 Antireux stent (MarathonTM, Cook Endoscopy). The anti-
reux wind sock (white) is seen at the end of the stent.
available on some catheters and guidewires and used as a reference
point to the stricture and papilla. At this point, the stent is placed.
The tapered end is the proximal end. Depending on the type of stent
system, either the inner guiding catheter alone or the inner guiding
catheter with the stent, are advanced over the guidewire. It is impor-
tant not to allow the wire to pass too proximally into the biliary tree
during this advancement, since this could cause hepatic capsule or
intrahepatic ductal injury. On the other hand, excessive traction on
the wire may result in wire loss. The stent is then advanced over the
guide catheter by advancing the pusher tube, which is a somewhat
larger bore and stiffer catheter which approximates the diameter of
the stent. During advancement, the elevator should remain closed.
When the stent impacts the elevator, the elevator is opened slightly
to allow it to emerge from the endoscope channel. The elevator is
closed to direct the stent upward and into the papilla. It is imperative
to maintain a short endoscope position as close as possible to the
papilla to maintain maximal mechanical advantage. Using a series
of small movements in which the elevator is sequentially lowered to
allow advancement of the stent, and then closed to advance the stent
in a ratchet-like manner, the stent is advanced into the bile duct.
Upward tip deection as well as withdrawing the endoscope to
further shorten it also provides forward motion to the stent. It is
important to not allow more than a minimal amount of the stent to
be advanced out of the endoscope into the duodenum; excessive
length between the endoscope tip and papillary orice decreases the
mechanical advantage to forward advancement. To facilitate forward
movement of the stent, the endoscopy assistant must provide trac-
tion on the inner guiding catheter. Once optimal stent position is
achieved, the inner guiding catheter and guidewire are removed
while the endoscopist maintains forward pressure with the pusher
tube against the stent to prevent distal dislodgement. If additional
contrast is needed to assess drainage or intrahepatic anatomy above
the stent, the guidewire is removed prior to removing the inner
guiding catheter to allow the injection. The guide catheter and
pusher tube are then removed as described above. The process is
repeated for additional stent placement.

156
 Chapter 16 Plastic Pancreatic and Biliary Stents: Concepts and Insertion Techniques
A B
Fig. 16.9 Endoscopic photos of distally migrated 11.5 Fr biliary
stent impacted against the duodenal wall opposite the major papilla.
A Before removal and B after removal, a small defect is seen.
In patients with short, distal bile strictures in whom multiple
stents are to be placed (for example chronic pancreatitis with biliary
stricture, post-sphincterotomy papillary stenosis) three to four 10 Fr
5 cm stents can be placed at one time on the inner guiding catheter.
Once the rst stent is in place (Fig. 16.11), the inner guiding catheter
and guidewire are withdrawn just enough to release the rst stent;
Fig. 16.8 Illustration of 10 Fr stent system
with stent placed for relief of malignant
distal biliary obstruction.
the duct is then cannulated alongside the rst stent with the second
stent, guidewire, and inner guiding catheter. The process is contin-
ued until all stents are deployed. Alternatively, the stents can be
placed one by one alongside each stent (Fig. 16.12). Newer stent
delivery systems such as the Fusion (Cook Endoscopy) facilitate
placement of multiple stents, since they allow intraductal exchange
whereby the wire remains across the stricture between stent
placements.
Stents for irretrievable bile duct stones
In the absence of a stricture, pigtail stents (Fig. 16.1b) may be prefer-
able to straight stents when placed in a dilated biliary tree because
they are less likely to migrate distally and completely out of the bile
duct. Pigtail stents are placed slightly differently than straight stents
in that if the distal end of the stent is against the papilla, too much
stent has been advanced into the duct to allow the pigtail to form in
the duodenum. The stent should be advanced until the portion of
the stent that is just proximal to the distal pigtail (identied by apply-
ing indelible marker prior to placement if a visible marker is not
already on the stent) is visible. The stent is then advanced while
simultaneously withdrawing the endoscope so that the pigtail is
deployed into the duodenum.
157
 SECTION 2 TECHNIQUES
Fig. 16.10 Measurement on the radiograph to calculate stent
length. A The measurement from the top of the stricture to
endoscope tip when positioned at the papilla (bracket) compared
to the diameter of the endoscope (arrowheads) was 7 : 1. B Since
the diameter of the endoscope was 11.5 mm, a 9 cm stent was
placed.
A B
Fig. 16.11 Insertion of Multiple 10 Fr stents. A The rst stent
(1) is being pushed by the second stent (2) since the actual pusher
tube is still well above the multiple stents loaded onto the catheter.
B Final result of four 10 Fr stents, all placed with one passage of
the stent introducer system.
Hilar biliary obstruction
Hilar biliary obstruction differs from distal obstruction in two ways.
Although a sphincterotomy is not needed for placement of a unila-
teral biliary stent, limited data suggest that hilar stent placement for
obstruction carries a higher risk of pancreatitis than for distal
obstruction, and pancreatitis, which may be prevented by perform-
ing a biliary sphincterotomy.13 Secondly, stricture dilation is fre-
quently required because of the loss of mechanical advantage as the
resistance of the stricture is away from the tip of the endoscope. Both
of these maneuvers become mandatory when bilateral stents are
placed (Fig. 16.13A16.13C).
In general, stents used for hilar tumors are 1215 cm in length,
since the average distance to the bifurcation is 9 cm. Stents that may
158
Fig. 16.12 Additional stent insertion. Passage of the catheter
alongside the initial stents in order to recannulate and place addi-
tional stents.
be adequate to cross the stricture but too short to be anchored in
the intrahepatic system are more prone to migrate distally. If it is
determined that bilateral stents are to be placed, there are two
options for guidewire placement. One way is the placement of two
wires, one in each intrahepatic system, prior to placing either stent
(Fig. 16.1). The other way is to place the rst stent, recannulate the
bile duct alongside this stent, and pass the guidewire into the oppo-
site intrahepatic system. There are proponents of both methods,
with advantages being the lack of friction within the endoscope
channel between the rst stent (if 10 Fr) and its larger pusher tube
and the second guidewire within the endoscope channel. This can
be overcome by using a 0.025 guidewire as the second wire. It is
important to note that it may not be possible to place bilateral 10 Fr
stents during the rst session. In that case it may be best to place
two 7 Fr stents or one 10 Fr and one 7 Fr stent, then upsize one or
both at the second session one month later.
Pancreatic duct stent insertion
Pancreatic duct stent placement does not usually require the perfor-
mance of a pancreatic sphincterotomy, especially since these stents
have a small caliber (37 Fr). Rarely, 10 Fr stents are placed and even
then a sphincterotomy for stent placement alone is usually unneces-
sary. The diameter of the stent chosen is dependent on the indica-
tion as well as the size of the main pancreatic duct. As mentioned
previously, smaller diameter stents are passed over the guidewire
without an inner guiding catheter. Similar to the biliary stenting
process, the site of the pathology is identied, a wire passed into
the tail and dilation performed, if necessary. The stent is selected
and advanced into place with a pusher tube over the guidewire,
although most 5, 6, or 7 Fr stents can be pushed into place using a
standard 5 Fr catheter or sphincterotome. The wire is removed while
keeping the pusher tube in position as mentioned. The pusher tube
is then removed, leaving the end of the stent extruding from the
papilla. Single pigtail stents with the pigtail in the duodenum are
commonly employed in the pancreatic duct to avoid inward migra-
tion, which can be difcult to retrieve.
Small caliber plastic stents (35 Fr) are now typically used for
prevention of post-ERCP pancreatitis in patients at high risk (e.g.
sphincter of Oddi dysfunction, ampullectomy) and/or the perfor-
 Chapter 16 Plastic Pancreatic and Biliary Stents: Concepts and Insertion Techniques

A B

Fig. 16.13 Bilateral stent placement for hilar

cholangiocarcinoma. A Malignant stricture involves left (arrow-
head) and right hepatic (arrow) ducts. B Balloon dilation is
performed of left hepatic duct stricture (arrowhead); note wire
is in right intrahe-patic (arrow). C Successful bilateral stent
placement.

mance of high-risk interventions (precut biliary sphincterotomy,
pancreatic sphincterotomy)14 (Fig. 16.14). These stents are expected
to pass spontaneously within 14 days and thereby minimize pan-
creatic ductal injury.
Drainage of pancreatic uid collections
Double pigtail stents are placed transgastrically or transduodenally
when transmural drainage of pancreatic uid collections is under-
taken.15 Usually two stents are placed across the wall into the collec-
tion (Fig. 16.15). Although straight stents can be used, they may be
a source of delayed bleeding as the end within the collection impacts
the wall as the collection collapses.16 Therefore double pigtail stents
are preferred. They are inserted as mentioned above for the biliary
tree. It is important to note that the proximal ends of some of the
10 Fr stents are tapered and do not allow an inner guiding catheter
to pass. The tapered portion may need to be severed to allow an inner
guiding catheter pusher tube to pass through the stent. In addition,
one must be especially careful that an excessive length of stent is
not passed into a pseudocyst since the entire stent can be lost
during deployment.

159
 SECTION 2 TECHNIQUES

Fig. 16.14 3 Fr pancreatic duct stent placed for prevention of post-

ERCP pancreatitis. Arrows denote ends of stent.
Fig. 16.16 Plastic biliary stent (arrowhead) passed through
occluded metal biliary stent (arrows) which had been placed for
palliation of pancreatic carcinoma.

Fig. 16.15 Two 10 Fr stents placed transduodenally to drain a

pancreatic pseudocyst.

Indications and contraindications

Biliary indications
Malignant biliary obstruction is the most frequent indication for the
use of plastic stents. Distal obstruction is most commonly due to
pancreatic carcinoma. Mid to proximal malignant obstruction may
be due to primary cancer of the biliary tree (gallbladder or cholan-
giocarcinoma) or from invasion or obstruction of the duct by adja-
cent malignant metastatic lymph nodes. Plastic stents may be used
to relieve obstruction of previously placed metal stents17 (Fig. 16.16).
In general, distal bile duct tumors are more effectively palliated with
plastic stents than are hilar tumors.
Benign strictures can frequently be managed by the use of plastic
stents. Causes of benign obstruction include post-sphincterotomy
stenosis, chronic pancreatitis, post-surgical injury, ischemia, and
anastomotic strictures after liver transplantation. Indeed, in addition
to acute relief of obstruction, serial stenting and dilation with increas-
ing numbers of stents appears to be more effective than single stents
for such pathology1820 (Fig. 16.17). Biliary leaks and stulae after
biliary surgery, cholecystectomy, or trauma are also effectively treated
by short-term stent placement across the papilla21 (Fig. 16.18). In
most of these latter indications short-length stents are effective and
do not need to cross the leak site. The elimination of sphincter pres-
Fig. 16.17 Fluoroscopic image after placement of ve stents for
treatment of benign distal bile duct stricture.
sure promotes ow away from the leak into the duodenum, promot-
ing closure. For more complex leaks and major leaks of the common
bile duct, it may be necessary to traverse the leak site.
Pancreatic indications
Plastic stents have been used for relief of pancreatic duct obstruction
in the setting of chronic pancreatitis. In this setting there may refrac-
tory pain or pancreatic leaks, with resultant pancreatic ascites or
pseudocysts.22 Occasionally malignant pancreatic duct obstruction
will result in pancreatitis or contribute to disabling pain. Placement
of pancreatic stents may be effective in this setting (Fig. 16.19).23 As
previously mentioned, temporary stent placement is useful in the
prevention of post-ERCP pancreatitis in selected patients. In the
setting of severe acute pancreatitis, pancreatic duct leaks and disrup-

160
 Chapter 16 Plastic Pancreatic and Biliary Stents: Concepts and Insertion Techniques

A
B C

Fig. 16.18 Placement of biliary stent for treatment of post-cholecystectomy cystic duct leak. A Active leak is seen. B Fluoroscopic
image taken immediately after placement of 10 Fr biliary stent. C Follow-up cholangiogram several weeks later showing closure of
leak.

Fig. 16.19 Placement plastic pancreatic stent in patient with unre-

B sectable pancreatic cancer, intractable pain, fever, and
hyperamylasemia. A Stricture (arrowheads) and dilated main
A pancreatic duct (arrows). B Immediately after placement of stent.
Signicant improvement in pain was achieved.

A C

Fig. 16.20 Placement of pancreatic stent for treatment of post-splenectomy pancreatic duct leak. A Active leak is seen.
B Fluoroscopic image taken immediately after placement of 7 Fr pancreatic stent to tail. C Follow-up pancreatogram several weeks
later showing closure of leak.

tions may contribute to the poor outcome of these patients; pan-
creatic stent placement may improve the clinical course in a subset
of these patients.24 In patients with traumatic pancreatic ductal
injury, plastic stents may be effective in bridging the injured duct
and permitting resolution of the leak. Post surgical pancreatic duct
leaks (distal pancreatectomy, inadvertent surgical injury) can occur
and are effectively treated with pancreatic stents (Fig. 16.20).25
Finally, a variety of plastic stent congurations have been useful in
transpapillary and transmural drainage of pancreatic uid collec-
tions (see Chapter 45).15

161
 SECTION 2 TECHNIQUES

Complications
When sphincterotomy is performed to facilitate stent insertion, com-
plications such as hemorrhage (Fig. 16.21) or perforation may
occur.26 When placed into the bile duct, problems caused by the stent
itself include cholangitis, frequently due to stent occlusion, and
cholecystitis as a result of cystic duct obstruction.27 Occlusion of a
biliary stent secondary to deposition of bacterial biolm and/or plant
material (Fig. 16.2) is the most commonly encountered complication
of plastic stents and occurs in about 30% of cases with resulting
jaundice and cholangitis. When placed into the pancreatic duct, stent
occlusion may cause pancreatitis. Stent migration, into or out of the
bile duct, occurs in up to 5% of cases and may result in recurrent
obstruction and cholangitis. Pancreatic stent migration into the duct
can be difcult to retrieve due to their small diameter and the small
size of the pancreatic duct. If not retrieved, permanent ductal damage
can occur. Even when left in for a few weeks in a planned situation,
pancreatic duct stents can induce ductal damage similar to chronic Fig. 16.21 Endoscopically visible vessel (arrow) identied from
post-sphincterotomy bleeding following biliary stent placement.
pancreatitis.28 (Figs 16.22A16.22C). Uncommon complications Heater probe therapy was applied and no further bleeding
include perforation of the duodenum if distal migration occurs (Fig. ensued.
16.9);29 such perforations may be occult until the stent is removed

B
A
C

Fig. 16.22 Pancreatic duct stent-induced ductal damage after treatment of post-tail resection pancreatic duct leak. A Active leak is seen
at tail (arrow). B Fluoroscopic image taken immediately after placement of short 7 Fr pancreatic stent. C Follow-up pancreatogram
several weeks later showing stricture (arrow) at site where the stent end was in contact with the duct.

and the hole opened. A variety of rare complications have been
reported following migration completely out of the bile duct, such
as bowel obstruction,30 and intestinal perforation.31
Relative cost
Plastic stents provide rapid palliation of biliary obstruction, and
shorten hospital stay when compared to surgical bypass. In many
cases, stent placement obviates major surgical intervention. The cost
of a plastic stent is less than $100 and is far less than an expandable
metal stent, the cost of which may exceed $1800 contingent upon
manufacturer and presence or absence of a covering. Metal stents
have a signicantly longer patency than plastic stents, although if
the patient does not survive long enough, this benet will not be
realized. Therefore, in patients with distal malignancy who have an
anticipated life expectancy less than three to four months, plastic
stents are more cost-effective.3,32 CPT codes and ambulatory payment
classications in the US for placement and/or removal of plastic
biliary stents are available in a recent review.2

REFERENCES
1. Soehendra N, Reynders-Frederix. Palliative bile duct drainage- a
new endoscopic method of introducing a transpapillary drain.
Endoscopy 1980; 12:811.
2. Somogyi L, Chuttani R, Crofe J, et al. Biliary and pancreatic stents.
Gastrointest Endosc. 2006 June; 63(7):910919.
3. Moss AC, Morris E, Mac Mathuna P. Palliative biliary stents for
obstructing pancreatic carcinoma. Cochrane Database Syst Rev.
2006 Apr 19(2):CD004200.
4. van Berkel AM, van Marle J, Groen AK, et al. Mechanisms of biliary
stent clogging: confocal laser scanning and scanning electron
microscopy. Endoscopy. 2005 Aug; 37(8):729734.
5. Farnbacher MJ, Voll RE, Faissner R, et al. Composition of clogging
material in pancreatic endoprostheses. Gastrointest Endosc. 2005
June; 61(7):862866.
6. Weickert U, Venzke T, Konig J, et al. Why do bilioduodenal plastic
stents become occluded? A clinical and pathological investigation

162
 Chapter 16 Plastic Pancreatic and Biliary Stents: Concepts and Insertion Techniques
on 100 consecutive patients. Endoscopy. 2001 Sep;
33(9):786790.
7. Tringali A, Mutignani M, Perri V, et al. A prospective, randomized
multicenter trial comparing DoubleLayer and polyethylene stents
for malignant distal common bile duct strictures. Endoscopy.
2003 Dec; 35(12):992997.
8. Raju GS, Sud R, Elfert AA, et al. Biliary drainage by using stents
without a central lumen: a pilot study. Gastrointest Endosc. 2006
Feb; 63(2):317320.
9. Raju GS, Gomez G, Xiao SY, et al. Effect of a novel pancreatic stent
design on short-term pancreatic injury in a canine model.
Endoscopy. 2006 Mar; 38(3):260265.
10. Ishihara T, Yamaguchi T, Seza K, et al. Efcacy of s-type stents for
the treatment of the main pancreatic duct stricture in patients with
chronic pancreatitis. Scand J Gastroenterol. 2006 June;
41(6):744750.
11. Reddy DN, Banerjee R, Choung OW. Antireux biliary stents: are
they the solution to stent occlusions? Curr Gastroenterol Rep.
2006 Apr; 8(2):156160.
12. Giorgio PD, Luca LD. Comparison of treatment outcomes
between biliary plastic stent placements with and without
endoscopic sphincterotomy for inoperable malignant common
bile duct obstruction. World J Gastroenterol. 2004 Apr 15;
10(8):12121214.
13. Tarnasky PR, Cunningham JT, Hawes RH, et al. Transpapillary
stenting of proximal biliary strictures: does biliary sphincterotomy
reduce the risk of postprocedure pancreatitis? Gastrointest
Endosc. 1997 Jan; 45(1):4651.
14. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic
stent placement reduce the risk of post-ERCP acute pancreatitis?
A meta-analysis of controlled trials. Gastrointest Endosc. 2004 Oct;
60(4):544550.
15. Baron TH. Endoscopic drainage of pancreatic uid collections and
pancreatic necrosis. Gastrointest Endosc Clin N Am. 2003 Oct;
13(4):743764.
16. Cahen D, Rauws E, Fockens P, et al. Endoscopic drainage of
pancreatic pseudocysts: long-term outcome and procedural
factors associated with safe and successful treatment. Endoscopy.
2005 Oct; 37(10):977983.
17. Tham TC, Carr-Locke DL, Vandervoort J, et al. Management of
occluded biliary Wallstents.Gut. 1998 May; 42(5):703707.
18. Baron TH. Endoscopic therapy with multiple plastic stents for
benign biliary strictures due to chronic calcic pancreatitis: the
good, the bad, and the ugly. J Clin Gastroenterol. 2004 Feb;
38(2):9698.
19. Pozsar J, Sahin P, Laszlo F, et al. Endoscopic treatment of
sphincterotomy-associated distal common bile duct strictures by
using sequential insertion of multiple plastic stents. Gastrointest
Endosc. 2005 Jul; 62(1):8591.
20. Kuzela L, Oltman M, Sutka J, et al. Prospective follow-up of
patients with bile duct strictures secondary to laparoscopic
cholecystectomy, treated endoscopically with multiple
stents. Hepatogastroenterology. 2005 SepOct; 52(65):
13571361.
21. Kaffes AJ, Hourigan L, De Luca N, et al. Impact of endoscopic
intervention in 100 patients with suspected postcholecystectomy
bile leak. Gastrointest Endosc. 2005 Feb; 61(2):269275.
22. Delhaye M, Arvanitakis M, Bali M, et al. Endoscopic
therapy for chronic pancreatitis. Scand J Surg. 2005; 94(2):
143153.
23. Costamagna G, Mutignani M. Pancreatic stenting for malignant
ductal obstruction. Dig Liver Dis. 2004 Sep; 36(9):635638.
24. Lau ST, Simchuk EJ, Kozarek RA, et al. A pancreatic ductal leak
should be sought to direct treatment in patients with acute
pancreatitis. Am J Surg. 2001 May; 181(5):411415.
25. Le Moine O, Matos C, Closset J, et al. Endoscopic management
of pancreatic stula after pancreatic and other abdominal
surgery. Best Pract Res Clin Gastroenterol. 2004 Oct;
18(5):957975.
26. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med. 1996 Sep 26;
335(13):909918.
27. Dolan R, Pinkas H, Brady PG. Acute cholecystitis after palliative
stenting for malignant obstruction of the biliary tree. Gastrointest
Endosc. 1993 MayJune; 39(3):447449.
28. Kozarek RA. Pancreatic stents can induce ductal changes
consistent with chronic pancreatitis. Gastrointest Endosc. 1990
MarApr; 36(2):9395.
29. Melita G, Curro G, Iapichino G, et al. Duodenal perforation
secondary to biliary stent dislocation: a case report and review of
the literature. Chir Ital. 2005 MayJune; 57(3):385388.
30. Simpson D, Cunningham C, Paterson-Brown S. Small bowel
obstruction caused by a dislodged biliary stent. J R Coll Surg
Edinb. 1998 June; 43(3):203.
31. Storkson RH, Edwin B, Reiertsen O, et al. Gut perforation caused
by biliary endoprosthesis. Endoscopy. 2000 Jan; 32(1):8789.
32. Levy MJ, Baron TH, Gostout CJ, et al. Palliation of malignant
extrahepatic biliary obstruction with plastic versus expandable
metal stents: An evidence-based approach. Clin Gastroenterol
Hepatol. 2004 Apr; 2(4):273285.
163
 SECTION 2 TECHNIQUES
Chapter
Expandable Stent Insertion
17 Ann Marie Joyce and Gregory G. Ginsberg
Introduction
Expandable biliary stents are used primarily for the palliation of
malignant biliary obstruction. There are two main categories of
biliary stents: xed-diameter plastic stents (FDPS) and self-expanding
metallic stents (SEMS). FDPS, introduced in 1980, preceded their
SEMS counterparts and are discussed in detail in Chapter 16.
While FDPS are a safe and effective means to overcome biliary
stenoses, they eventually become occluded.1 Stent occlusion is
attributed to biolm formation such that under even ideal circum-
stances, FDPS occlusion occurs in 30% and 50% of patients within
three and six-months, respectively.2 Bile ow rate is impacted on
by the stent lumen diameter. The internal diameter of an FDPS
is limited by the accessory channel size of the duodenoscope.
Because the diameter of the accessory channel of a therapeutic
duodenoscope is 3.2 mm, FDPSs are available with internal diame-
ters up to 12 Fr. SEMS were developed to overcome this limitation
as they deliver a larger diameter stent (10 mm) via a small diameter
(7.5 Fr) delivery device. Because malignant biliary obstruction
is typically associated with a survival of less than one year,
SEMS are intended to yield lifelong palliation of obstructive
symptoms.35
This chapter reviews the indications for SEMS placement, SEMS
selection, types of SEMS available, techniques for SEMS placement,
SEMS-related complications, and complication avoidance and
management.
INDICATIONS
SEMS are indicated for palliation of malignant biliary obstruction.
The most common cause of malignant biliary obstruction is pancre-
atic adenocarcinoma arising in the head or genu region of the pan-
creas. The management of distal biliary obstruction is discussed in
more detail in Chapter 27 by Barkun. Other causes of malignant
biliary stenoses are cholangiocarcinoma, ampullary carcinoma, gall-
bladder cancer, and extrinsic compression associated with lymph-
adenopathy due to lymphoma and metastatic carcinoma. Patients
with malignant biliary obstruction typically present with jaundice.
Left unchecked, biliary obstruction may lead to the development of
pruritis, pain, cholangitis, hepatic synthetic dysfunction, and malab-
sorption. Without therapy the mean survival for patients presenting
with malignant biliary obstruction is less than 200 days. Because
most patients have advanced disease at the time of presentation,
operative resection with curative intent is only possible in 1015%
of cases.6,7 Therefore, palliation of symptoms is a major component
of management in the majority of patients with malignant biliary
obstruction.
Options for palliation of malignant biliary obstruction include
operative bypass, percutaneous drainage, and/or endoscopic stent-
ing. In a prospective, randomized trial of endoscopic stenting with
FDPSs versus operative bypass the two compared equivalently with
respect to alleviation of obstruction and relief of jaundice, and favor-
ably with respect to hospital stay and procedure-related complica-
tions and mortality.8 However, late recurrence of jaundice occurred
more commonly in the FDPS group owing to stent occlusion. Versus
percutaneous drainage, endoscopic stenting compares equivalently
for alleviation of jaundice, and superiorly with respect to procedure-
related morbidity and 30-day mortality.9
Both FDPS and SEMS can be used for the palliation of malignant
biliary obstruction. FDPSs are safe and effective. They are relatively
less expensive compared to SEMS, and can be removed and replaced
if they become occluded. To restate, the main drawback of FDPS is
their variable rate of occlusion. Symptoms of stent occlusion include
recurrence of jaundice and/or ascending cholangitis. Two strategies
have been employed related to FDPS occlusion: (1) prophylactic
exchange, and (2) expectant management. The former involves elec-
tive stent removal and exchange at three-month intervals in order to
reduce the risk of cholangitis and need for emergency exchange
while the latter relies on watchful waiting and nimble responsive-
ness should stent occlusion occur. Further discussion on FDPS can
be found in Chapter 16 (Baron and Ponsky).
SEMS are intended, owing to their larger internal diameter, to
extend the duration of patency, thereby reducing the need for and
frequency of reintervention. As such, their increased cost may be
offset by a reduction in episodes of cholangitis and need for elective
and emergency interventions, including hospitalization.
A multicenter randomized study that compared a SEMS (Wall-
stent) to 10 Fr FDPS was performed by the US Wallstent study
group10 (Table 17.1). Early stent occlusion occurred in about 3% of
the FDPS group as compared to none in the Wallstent group.
Early occlusion was attributed to sludge accumulation in the plastic
stent groups. During long-term follow-up, the probability of
stent occlusion was 2.8 times greater for plastic stents than for
Wallstents. Tumor ingrowth or overgrowth contributed to about
14% of the Wallstent occlusions, but this did not affect the plastic
stent group. The overall complication rate was signicantly lower in
the Wallstent group than in the FDPS group (20% vs 31%, p < 0.05).
In the plastic stent group there was a higher number of procedures
performed which resulted in a higher cost as compared to the Wall-
stent group. Another prospective, randomized trial was performed
to compare the patency and determine the cost-effectiveness of
SEMS versus FDPS.4 This study demonstrated that the median
period of patency was longer for the metal stent group (33%) after
a median of 273 days compared with 54% after a median of 126 days
(p = 0.006). Two subsequent prospective, randomized trials11,12 also
reported longer duration of patency with Wallstents compared to
165
 SECTION 2 TECHNIQUES

Occlusion on
No. of patients Drainage (%) rate (%) Stent patency (days) p value
Author FDPS SEMS FDPS SEMS FDPS SEMS FDPS SEMS
Davids (4) 49 56 95 96 54 33 126a 273a p = 0.006
Carr-Locke (10) 78 86 95 98 13 13 62a 111a
Knyrim (5) 31 31 100 100 36 22 140a 189b p = 0.035
Kaassis (11) 59 59 63 20 165a Not reached p = 0.007

Table 17.1 Randomized trials comparing FDPS versus SEMS for palliation of malignant strictures of the bile duct
a
median.
b
mean.

COMPLICATIONS

Author, year No. of Tumor

(reference) pts Stent occlusion in growth Migration Cholecystitis Pancreatitis Cholangitis
Smits et al. 1995 (20) 22 2 0 1
Born et al. 1996 (21) 10 3 1
Miyayama et al. 1997 (22) 19 1 0 1 1 (related to tumor) 0 2
Rossi et al. 1997 (23) 21 7 4 4
Hausegger et al. 1998 (24) 30 11 2
Shim et al. 1998 (25) 21 2 2
Nakamura et al. 2002 (26) 13 1 1
Han et al. 2002 (27) 10 0 0 3 0 0
Isayama et al. 2002 (28) 21 3 0 2 1
Schoder et al. 2002 (29) 42 6 0 0 3 1
Bezzi et al. 2002 (30) 26 4 0 0 3 0
Han et al. 2003 (31) 13 3 2 0 0 0
Isayama et al. 2004 (32) 57 8 0 2 5
Miyayama et al. 2004 (33) 22 3 1 3 1
Nakai et al. 2005 (34) 69 7 0 4 4 4 5
Kahaleh et al. 2005 (35) 80 12 5 3 5

Table 17.2 Review of available studies of the complications of covered SEMS

FDPS. Along with the extended patency rate with Wallstents there
were fewer accumulated hospital days in this group (p < 0.05).
Conventionally, the use of SEMS is limited to patients with con-
rmed, non-operable, malignant biliary obstruction. Our practice,
and the practice of many, has been to place FDPS for initial manage-
ment of suspected malignant biliary obstruction. We have relatively
deferred the use of SEMS until there was evidence of occlusion of
the initially placed FDPS, a performance status suggesting a greater
than 6-month survival, a conrmed tissue diagnosis, and completed
staging to conrm non-operability. Our most common application
of SEMS has been in patients with biopsy proven, inoperable malig-
nant biliary obstruction that have developed occlusion of a previously
placed FDPS with a life-expectancy of greater than 36 months.
However, recent studies suggest broader application for SEMS.
Concerns about the use of SEMS prior to conrming a tissue
diagnosis of cancer and afrming unresectability have been con-
tested. In patients with potentially resectable pancreatic cancer,
patients have improved survival with neoadjuvant chemotherapy
followed by operative resection.1317 In a small group of patients,
Wasan et al. demonstrated that SEMS can be useful for relief of
biliary obstruction in resectable pancreatic carcinoma.18 While there
had been concern about SEMS complicating operative resection, this
has not been borne out in clinical practice. One cost analysis con-
cluded that initial SEMS placement provided equal or superior ef-
cacy and reduced overall costs compared to FDPS placement.19
Covered SEMS share the same indications as their uncovered
precursors, though they are not used in hilar or intrahepatic ducts
because of blockage to the contralateral intrahepatic system if ipsi-
lateral intrahepatic branches. The anticipated advantage of covered
SEMS is the retardation of tissue (malignant or hyperplastic)
in-growth contribution to stent occlusion. However, there is consid-
erably less published and unpublished experience with covered
SEMS. Limited studies of covered SEMS have raised concerns of
higher rates of stent migration, cystic duct obstruction, and pancre-
atitis (Table 17.2)2036 Covered stents may be particularly well suited
for reconstitution of an occluded indwelling uncovered SEMS.
Because they are considered non-removable SEMS have generally
been considered contraindicated for management of benign biliary
strictures. While SEMS remain patent longer than FDPS, the dura-
bility of their biotolerance is not indenite. Endoscopic removal of
occluded uncovered SEMS cannot be reliably achieved as compared
with covered SEMS.37 Long-term results from ongoing trials evaluat-
ing the application of covered SEMS in the management of benign
bile duct disease are awaited with interest.38,39

166
 Chapter 17 Expandable Stent Insertion

tip to allow ease of passage. The SEMS is deployed by withdrawing

TYPES OF SELF-EXPANDING STENTS
There are a variety of SEMS used in the palliation of malignant
biliary obstruction (Table 17.3). Commercially available SEMS vary
moderately in design, delivery, conguration, and sizes. There are
few studies comparing the different stents. The available uncovered
stents include: Wallstent (Boston Scientic, Natick, MA), Zilver
stent (Cook Endoscopy, Winston-Salem, NC), Diamond stent (Boston
Scientic, Natick, MA), and Flexxus stent (ConMed, Billerica, MA).
Covered stents include the covered Wallstent (Boston Scientic,
Natick, MA) and Viabil stent (W.L. Gore, Flagstaff, AZ). To decrease
the occlusion of expandable stents by tumor ingrowth covered stents
have been introduced. These stents vary slightly but all are deployed
through a duodenoscope.
Wallstent
The Wallstent is the original SEMS and is considered the industry
standard (Fig. 17.1). Most of the published literature on SEMS
applies to the biliary Wallstent. It is a braided stainless steel mesh
with soft barbed ends. The Wallstent is available in 40, 60 and 80mm
lengths. The available diameters of the fully expanded Wallstent are
8 and 10 mm. The delivery device has an outside diameter of 7.5
Fr and consists of an 0.035-inch guidewire compatible introducer
catheter, on which the compressed SEMS is constrained by a
hydrophilic-coated outer sheath. The delivery device has a tapered
the outer sheath releasing the SEMS in the desired location. The
Wallstent is radiopaque and there are four radiopaque markers on
the delivery device to guide precision deployment. The stent can be
recaptured, if need be, and repositioned up until 90% of full stent
release. Wallstents can be deployed entirely within the bile duct
or in transpapillary position. There is 33% foreshortening of the
Wallstent post-deployment. Transpapillary positioned uncovered
Wallstents may be reliably removed within 12 to 24 hours after inser-
tion. Subsequently, the stent becomes embedded into the bile duct
wall and it is more difcult, if not impossible, to remove.
Diamond Ultraex stent
The Ultraex Diamond stent is made of nitinol, a nickel-titanium
alloy that provides a high degree of exibility (Fig. 17.2). It is con-
structed as a laser-welded single knitted wire. The interstices of the
lattice work are larger compared to those of the Wallstent. This may
more easily permit cannulation of the interstices and dilation for
placement of another stent to create a Y conguration; this may
be potentially helpful in the palliation of hilar strictures. The delivery
device is similar to that of the Wallstent. The outer sheath measures
3 mm (8.5 Fr) in diameter. The stent is available in 4, 6, and 8 cm
in length and 10 mm in diameter. Once the deployment has com-
menced the stent cannot be recaptured. There is little foreshorten-
ing. There are radiopaque markers to assist with the accurate
positioning of the stent, however; the stent itself is less visible

Type Delivery system (Fr) Metal Length (cm) Diameter (mm) List price
Wallstent 7.5 Steel 4,6,8 8,10 $1450
Diamond Ultraex stent 9.25 Nitinol 4,6,8 10 $1300
Gianturco-Rosch Z stent 12 Stainless steel 4,6,8 10 N/A
Spiral Z stent 8.5 Stainless steel 5.7,7.5 10 N/A
Za stent 8.5 Nitinol 4,6,8 10 N/A
Zilver stent 7 Nitinol 4,6,8 6,8,10 $1250
Flexxus 7.5 Nitinol 4,6,8,10 8,10 $1525$1735
Covered Wallstent 8 Steel 4,6,8 8,10 $1650
Viabil 10 Nitinol 4,6,8,10 8,10

Table 17.3 Characteristics of SEMS

Fig. 17.1 Wallstent Fig. 17.2 Diamond Ultraex stent

167
 SECTION 2 TECHNIQUES

radiographically compared to the Wallstent. Radial expansion forces Flexxus

are purportedly similar. The Flexxus stent (formerly Memotherm and Luminexx) is a highly
Four studies have been published which compared the Ultraex exible nitinol stent with ared ends (Fig. 17.4). The stent is a laser-
Diamond stent with Wallstent for palliation of malignant biliary cut single piece of nitinol. Similar to the Diamond stent, the inter-
strictures.4043 While one reported equivalency, three others reported stices of the lattice work are large enough to permit cannulation of
inferior performance of the Ultraex Diamond as compared to the the interstices and dilation for placement of another stent to create
Wallstent. a Y conguration for palliation of hilar strictures. There are four
Tantalum markers on each end to provide improved uoroscopic
Z stent imaging. The predeployment delivery diameter is 7.5 Fr and the
There have been multiple iterations of the Z stent. The original post-deployment diameters are 8 and 10 mm with lengths of 40 mm,
Gianturco-Rosch Z stent was a stainless steel wire bent in a con- 60 mm, 80 mm, and 100 mm. The release mechanism is unique and
tinuous Z shaped pattern forming a cylinder. This was modied by employs a pistol-grip handle which withdraws the constraining
stringing together individual cages by adding small eyelets making sheath and allows stepwise, controlled release. During deployment,
the stent more exible and compressible. This is known as the there is no foreshortening and the stent cannot be re-constrained.
Spiral Z stent The introducer is similar in diameter to the Wallstent There are no studies available to compare the Flexxus with the other
but the stent lengths vary. The Spiral Z stent is available in 5.7 cm stents previously mentioned.
and 7.5 cm lengths and 10 mm in diameter. There are silver radi-
opaque markers along the length of the stent. Another iteration of Covered SEMS
the design, the Za-stent, incorporates nitinol in place of stainless SEMS have a longer patency rate as compared with plastic stents but
steel making the stent more exible. The available lengths of the they may still eventually become occluded. Plastic stents typically
Za-stent are 4, 6 and 8 cm with a diameter of 10 mm. There are become occluded with sludge and/or biolm whereas SEMS become
gold radiopaque markers in the middle and at the end of the Za-stent occluded with tumor overgrowth or ingrowth, ingrowth of benign
for uoroscopic visualization. The Zilver stent (Fig. 17.3) is one epithelial hyperplasia, and/or sludge accumulation. Covered SEMS
piece of nitinol compared to many pieces of nitinol threaded
together (Za). The gold radiopaque markers are at the proximal
and distal end of the stents. The introducer diameter is 7 Fr, which
is the smallest on the market. The release mechanism is similar to
that of the Wallstent. All forms of the Z stent including the newest
edition, Zilver stent, are non-shortening facilitating accurate
deployment.
A multi-center trial comparing the Wallstent with Spiral Z stent
was performed by Shah et al. and included 145 patients.44 There were
64 patients in the Z stent group and 68 in the Wallstent group. There
was a 100% success in the placement of the stents. There were 8
occlusions in the Z-stent group and 13 in the Wallstent group (p =
0.3). The calculated median patency rates for the Z-stent and the
Wallstent were 152 days and 154 days, respectively (p = 0.9). Accord-
ing to this study, the two stents appeared comparable.

Fig. 17.3 Zilver stent Fig. 17.4 Flexxus stent

168
 Chapter 17 Expandable Stent Insertion

were developed to overcome ingrowth through the SEMS interstices.

The covered Wallstent has a polyurethane covering (Fig. 17.5). The TECHNIQUES FOR SEMS PLACEMENT
delivery system and deployment are the same as the uncovered
version, though the predeployment diameter is slightly larger at 8 Duodenoscope
Fr. The initial stents were completely covered but modications have We preferentially use a therapeutic duodenoscope (accessory
been made so that 5 mm of each end is uncovered to allow embed- channel4.2mm) for most ERCP. Standard diagnostic duodeno-
ding into the tissue to decrease the rate of migration. Published scopes with a 3.2 mm accessory channel do permit insertion of most
studies of the covered Wallstent have yielded mixed results with
respect to improvement in patency rates (Table 17.4).2036
Furthermore, complications from covered Wallstents included
adherent debris, migration (6%) and cholecystitis (12%).
Another covered stent for use is the Viabil (Fig. 17.6). The non-
porous polytetrauoroethylene (ePFTE) and uorinated ethylene
propylene (FEP) covering prevents tumor ingrowth. Outside this
lining there is a nitinol stent with radiopaque rings on either end.
There are anchoring ns along the nitinol stent to prevent migra-
tion. The stent is available with or without holes in the covering. The
holes are along the proximal end of the stent to prevent occlusion
of other duct branches. The stent is available in a variety of sizes
with a 10 Fr delivery system and is available through ConMed
(Billerica, MA, USA).

Other SEMS
Obscure, pirated, and boutique stents have been developed and mar- Fig. 17.6 Viabil stents: without holes (top) and with holes
keted to limited extents around the globe. However these are not (bottom).
readily available in most markets. The Y Stent (Fig. 17.7) is an
example, intended for palliation of hilar strictures.

Fig. 17.5 Polyurethane covered Wallstent Fig. 17.7 Y stent

No. Median
of stent Stent Tumor
Author pts patency occlusion ingrowth Migration Cholecystitis Pancreatitis
(reference) U C U C U C U C U C U C U C
Shim et al. 1998 (25) 26 21 233 267 6 2 6 2
Nakamura et al. 2002 (26) 10 13 >402 >470 2 1 2 0 1 1 0 1 0
Miyayama et al. 2004 (33) 19 22 14 3 1 1 2
Isayama et al. 2004 (32) 55 57 193a 225a 21 8 16 0 0 1 0 2 1 5
Smits et al. 1995 (20) 24 22 3 2 2 0 0 1

Table 17.4 Covered (C) vs uncovered (U) SEMS studies

a
mean.

169
 SECTION 2 TECHNIQUES

SEMS. In patients with pending or existing gastric outlet or duode-
nal obstruction, a forward-viewing endoscope may be necessary for
initial inspection and stricture dilation. When considering dual
enteral and biliary stenting, we place the biliary SEMS rst, followed
by enteral stent placement.
Cholangiogram
A good quality cholangiogram using undiluted contrast will dene
the length, localization, and conguration of the biliary obstruction
and is important for selection of the appropriate stent. MRCP or CT
scan with pancreaticobiliary protocol may be valuable in the pre-
ERCP evaluation of suspected hilar obstruction when selected uni-
lateral drainage versus bilateral or multisegmental drainage may be
indicated, as discussed in Chapter 27.45
Sphincterotomy
A biliary sphincterotomy is not obligatory for SEMS placement in
either the supra- or transpapillary positions. Assertions that trans-
papillary SEMS placement without preplacement biliary sphincter-
otomy increases the risk of pancreatitis are unfounded, and avoidance
of sphincterotomy may reduce the risk of procedure-related
complications.
Dilation
Routine stricture dilation to facilitate SEMS placement is not
required. SEMS radial forces are sufcient to permit full or near-full
expansion within 48 hours; therefore post-deployment dilation is
also not routinely performed. In rare instances the stricture may
need to be dilated to allow the passage of the SEMS delivery
device.
SEMS shortening needs to be taken into consideration to enhance
precision of placement.
Use of guidewire
Guidewires are commonly used to traverse malignant bile duct stric-
tures, facilitate catheter insertion, and maintain access during device
exchange. The SEMS delivery device is passed over the guidewire
and positioned within the stricture. Larger diameter (0.035), nitinol,
hydrophilic coated wires are preferred as they enhance device
exchange. If two or more stents are going to be placed to palliate a
hilar stricture in a side-by-side fashion, multiple guidewires are used
to maintain access to the specic segments.
SEMS positioning
Biliary SEMS may be placed in a suprapapillary or transpapillary
position. This positioning is at the discretion of the endoscopist.
In a suprapapillary position (Fig. 17.8) the sphincter of Oddi
remains intact, assuming sphincterotomy was not performed. The
potential advantage of this approach is that it prevents free reux of
the duodenal contents into the bile duct which may contribute to
stent occlusion. However, there is insufcient data to support this
notion for SEMS46 and when this concept was tested with FDPS, no
difference in stent occlusion rates was observed. Suprapapillary
SEMS placement is most commonly performed for strictures of
the hilar region or common hepatic duct with which SEMS length
is insufcient to traverse the ampulla. In a study of 59 patients,
Liu et al.47 demonstrated that an inside-stent was possible in one-
third of patients presenting with malignant obstructive jaundice. In

Stent selection
The endoscopist and technician should be sufciently familiar with
the delivery and deployment mechanisms and post-deployment per-
formance characteristics of a SEMS when considered for use. This
includes guidewire and accessory channel compatibility, device
preparation, insertion and deployment mechanisms, radiographic
markings, SEMS shortening, recapture and reposition capabilities,
and a shared communications terminology.
Size matters. While the 10 mm diameter SEMS is used most
commonly, SEMS length selection is individualized and dependent
on the length and location of the stricture and the intention of
supra- or transpapillary placement. Careful measurements best
ensure optimal outcomes. Deployed and fully expanded SEMS
should extend a minimum of 10 mm beyond the proximal and
distal aspects of the stenosis to retard tumor overgrowth. One
should prevent placement of the stent where the ends of the SEMS
butt up against the bile duct side wall. The SEMS length should
be determined at the time of the ERCP cholangiography. Experi-
enced endoscopists commonly estimate the length based on the
dimensions of the superimposed duodenoscope diameter (see
Chapters 3 and 16). The length of the stricture may also be deter-
mined during the out exchange of a catheter or guidewire. The
catheter tip is positioned at the desired proximal extent of the stent.
Demarcating this position with the ngers on the catheter sheath
at the accessory channel port, the catheter is then withdrawn until
the tip is positioned at the desired distal extent of the SEMS. The
distance is then determined by the length of catheter withdrawn
from the accessory channel port. Catheters and guidewires
with designated length markings can also be used. The degree of Fig. 17.8 Suprapapillary placement of SEMS

170

this patient group there was 2 cm between the papilla and the distal
end of the stricture. The potential disadvantage of suprapapillary
SEMS placement is that, should it be required for management of
SEMS occlusion, recannulation of the stent lumen may prove
challenging.
Transpapillary SEMS placement (Fig. 17.9) is typically used for
common bile duct obstruction. Transpapillary placed SEMS should
extend 510 mm into the duodenal lumen. This extent permits ease
of subsequent recannulation. Transpapillary SEMS that extend
much further increase the risk for mechanical trauma to the oppos-
ing duodenal wall with potential for development of ulceration,
bleeding and perforation. Transpapillary SEMS placement does not
promote pancreatic duct obstruction or pancreatitis.
Endoscopic and uoroscopic guides
Once access has been obtained and the appropriate stent selected
then the SEMS can be delivered and deployed. One or both ports of
the stent introducer (depending on the stent) should be lubricated
with a ush of normal saline to ease advancement over the guidewire
and withdrawal of the outer sheath. The tip of the duodenoscope
should remain in close proximity to the papilla and thus minimal
wire is seen endoscopically. This close position helps to prevent
accidental loss of guidewire access when the rather stiff SEMS intro-
ducer catheter is being inserted into the ampullary orice. The stent
is passed over a guidewire into the bile duct and advanced through
the stricture. In a coordinated effort between the endoscopist and
the assistant, the guidewire is maintained stationary while the intro-
ducer catheter is inserted over it. The arrangement and designation
Fig. 17.9 Transpapillary SEMS placement
Chapter 17 Expandable Stent Insertion
of endoscopic and uoroscopic marking vary between commercially
available SEMS and the endoscopist and assistant must be familiar
with these representations. Fluoroscopic markings commonly des-
ignate the predeployment and approximate post-deployment proxi-
mal and distal ends of the SEMS. If the stent will be placed
transpapillary, the distal margin of the SEMS can be seen endoscopi-
cally and positioned so as to deploy with the desired length extending
from the ampulla. We generally maintain uoroscopic and endo-
scopic guidance throughout deployment to ensure accurate SEMS
placement. It should be noted that during deployment, SEMS may
have a tendency to move upstream. To counter this, during deploy-
ment, the endoscopist may need to apply resistance to or graded
withdrawal of the introducer catheter in order to achieve precise
placement.
Deployment
With the SEMS introducer apparatus in the desired position, the
outer sheath is incrementally withdrawn by the technician to release
the stent. The proximal end (with respect to the liver) of the stent
will gradually open as the outer sheath is withdrawn (Fig. 17.10).
The stent position may be adjusted distally by withdrawing the
entire apparatus. For more proximal position readjustment, the par-
tially released SEMS must be recaptured, if possible, by advancing
the outer sheath. The extent of deployment to which the stent may
be recaptured varies between products. Withdrawing the outer
sheath fully deploys the SEMS. The introducer catheter and guide-
wire are then removed. Care should be taken to ensure that the
obturator tip of the introducer catheter does not catch the SEMS
171
 SECTION 2 TECHNIQUES
B
A endoscopist and interventional radiologist. A guidewire is placed
C D
Fig. 17.10 Deployment of SEMS. A Cholangiogram demonstrat-
ing a markedly dilated proximal bile duct with a distal stricture.
B SEMS introducer advanced over the wire with the proximal end
of the stent above the proximal end of the stricture. C Initial
withdrawal of the outer sheath. D Full deployment of the SEMS.
risking displacement. When using SEMS for palliation of tight ste-
nosis, incomplete radial expansion of the deployed stent is not
uncommonly observed. Incomplete expansion may resist withdrawal
of the introducer catheter, particularly at the level of the obturator
tip, risking dislodgement or malpositioning of the SEMS. To over-
come this, the outer sheath may be incrementally re-advanced over
the introducer catheter as it is withdrawn. This permits withdrawal
of the stent delivery system while applying resistance to the deployed
SEMS thereby avoiding unintended dislodgement. The elevator lever
should be relaxed during deployment to allow smooth withdrawal of
the outer sheath. Post-deployment SEMS cannot be readjusted proxi-
mally. A rat-tooth forceps may be used to adjust the position more
distally.
172
Rendezvous technique
When ERCP cannulation of the bile duct cannot be achieved, SEMS
placement may be performed in a coordinated effort between the
into the bile duct through the percutaneous transhepatic approach.
The guidewire is advanced into the duodenum. The guidewire is
grasped in the duodenum by a biopsy forceps or snare and pulled
through the accessory channel of the duodenoscope. A catheter or
stent introducer is then advanced over the guidewire through the
accessory channel and eventually into the bile duct. This combined
technique has been successful in 80% of patients with malignant
bile duct obstruction.48
Hilar stricture
Management of hilar tumors is also discussed in Chapter 28. Chol-
angiocarcinoma, gallbladder carcinoma, and portal hepatic lymph
nodes can lead to an obstruction of the bile duct at the level of the
hilum of the liver. These tumors are associated with poor prognoses
and death is usually caused by cholangitis or liver failure. This
patient population tends to present with advanced disease; as such,
most of the tumors are unresectable. Palliation of hilar obstructions
provides a greater challenge than common bile duct lesions. While
unilateral stenting is effective for the relief of jaundice, bilateral
stenting may be required to palliate cholangitis. In a prospective
study of 61 patients with hilar strictures, unilateral stent placement
was achieved in 97% of patients.49 Jaundice resolved in 86.9% of
patients. Cholangitis developed in three patients within the rst
week of the stent placement but all of these patients were success-
fully treated with antibiotics. There was one patient that developed
a liver abscess and required percutaneous interventions on the con-
tralateral side.50 It is now generally recommended that both the left
and right intrahepatic ducts be drained during the procedure if both
sides are opacied during a cholangiogram (Fig. 17.11). A preproce-
dure MRCP may help to plan selective drainage with metallic stents.45
If there is one dominant side identied by MRCP then selective
cannulation with a catheter and a guidewire should be attempted
with intentional avoidance of pressure injection of contrast. Once
the catheter is in the desired location then measured cholangiogra-
phy should be performed. If only one side is opacied then single
stent can be placed. If there is opacication of both the right and left
biliary systems, bilateral stenting should be pursued. Bilateral SEMS
may be placed alongside one another or the second SEMS may be
deployed through the mesh of the initial SEMS (Fig. 17.12). Two
wires are used to maintain access to the right and left biliary systems
while placing bilateral stents in a side-by-side fashion. Those wires
must remain separated and secure while placing the stents. It can
sometimes be difcult to place the second stent beside the rst stent.
Hookey demonstrated that the placement of a tem-porary plastic
stent in the CBD prevents the full expansion of the rst stent to
facilitate placement of the second SEMS.51 When a second stent is
being placed through the interstices of the rst stent, sometimes the
interstices need to be balloon dilated. If obstructed intrahepatic
ducts remain inadequately drained then percutaneous drainage may
be required adjunctively.
Duodenal obstruction
Up to 1020% of patients with pancreatic and ampullary tumors
develop duodenal or gastric outlet obstruction.52 Enteral stenting is
an effective means of palliation of malignant gastroduodenal obstruc-
tion symptoms.5357 Most of these patients have concurrent or
 Chapter 17 Expandable Stent Insertion

A B C

D E

Fig. 17.11 Bilateral SEMS for a hilar stricture. A Cholangiogram revealed dilated left and right intrahepatic ducts with a hilar stricture.
B Wire placed in the left intrahepatic system. C Catheter placed in right intrahepatic system to facilitate placement of a second
wire. D SEMS placed over wire into the right intrahepatic system with a wire in the left intrahepatic system. E Stents placed in both
the left and right intrahepatic ducts. The stent in the left intrahepatic duct was placed in a suprapapillary position.

A B Fig. 17.12 Positioning of SEMS in hilar

strictures. A Side by side placement of
stents. B One stent through the interstices
of another stent creating a Y formation.
Redrawn from Ahmad NA, Ginsberg GG.
Gastrointestinal Endoscopy 2001;3(2):93
102 with permission.

pending bile duct obstruction and many will already have a bile duct
stent in place. Owing to their prolonged patency, a biliary SEMS is
recommended prior to placement of an enteral stent. This may be
performed at the same setting. While feasible in individual cases, it
is apt to be more difcult to place a biliary stent through the inter-
stices of a previously placed enteral stent. Sequential placement of
biliary and enteral SEMS was reported on 17 patients. The duodenal
stricture was traversed with the duodenoscope after dilation. All
patients had resolution of jaundice and 16 of the 17 patients
had relief of nausea and vomiting. During the follow-up two patients
had recurrence of biliary obstruction and two different patients had
recurrence of duodenal obstruction. Three cases were secondary to
tumor ingrowth and the fourth case was secondary to migration of
the duodenal stent.58
COMPLICATIONS
SEMS are generally safe and effective for the palliation of malignant
biliary obstruction. The complications related to SEMS include those
associated with ERCP and immediate and delayed SEMS malfunc-
tion. The generic ERCP associated complications include perfora-
tion, pancreatitis, and sedation-related cardiorespiratory compromise
and are discussed in more detail in Chapter 6.
Immediate causes of SEMS malfunction include device failure,
deployment failure, and malpositioning of the stent. Failure to ade-
quately lubricate the delivery device channels may inhibit withdrawal
of the outer sheath. Excessive articulation of the elevator lever may
similarly retard withdrawal of the outer sheath and has led to separa-
tion of the delivery apparatus.

173
 SECTION 2 TECHNIQUES
Malpositioning of the SEMS is usually attributable to operator
error. If the location of the stent is not satisfactory then a second
stent may be placed at the same setting. When in the transpapillary
position, another option is to completely remove the stent. Recently
placed SEMS may be repositioned more distally or completely
removed with a grasping forceps or a snare. For SEMS removal, the
stent should be grasped and pulled to the tip of the endoscope, and
the endoscope withdrawn under uoroscopic guidance. The stent
should be not pulled into the accessory channel of the duodenoscope
because the sharp ends can damage the channel. This is best accom-
plished within 48 hours of insertion as tissue hyperplasia leads to
embedding of the stent within the bile duct wall. Covered SEMS may
be more readily removed remote to the time of insertion.37
Other early complications, as dened as occurring within one
week of the stent placement, include cholangitis, hemobilia, and
perforation. Ineffective drainage of segments opacied during chol-
angiography increases the risk of cholangitis. Patients with malig-
nant hilar strictures or coexisting primary sclerosing cholangitis are
at greatest risk. Administration of prophylactic antibiotics may be
benecial to prevent or manage cholangitis in selected patients.
Persistent evidence of cholangitis while on antibiotics requires
repeat instrumentation either through a retrograde or antegrade
approach.
SEMS placement into friable bile duct tumors may induce hemo-
bilia. This bleeding is typically not clinically signicant and resolves
spontaneously. However, retained clot may result in recurrence of
biliary obstruction. The retained blood clot may be cleared with
catheter irrigation or a stone retrieval balloon.
A malpositioned SEMS may perforate the bile duct wall or
produce ulceration, bleeding, and perforation of the opposing duo-
denal wall. There are reports of using argon beam plasma coagula-
tion to shorten stents of excessive length within the duodenum that
have caused complications.59 Studies in animals suggest this is safe
to surrounding tissues.60
The most common late complication related to SEMS is stent
occlusion. The stent may become occluded by tumor ingrowth or
overgrowth, tissue hyperplasia or biliary sludge. Biliary sludge is the
most common cause of occlusion of plastic stents but this occurs
less commonly with SEMS given their larger diameter. With SEMS,
the most common cause of occlusion is tumor ingrowth. The tumor
grows through the interstices of the stent. There have been two
studies that have investigated the management of an occluded stent.
Tham et al.61 performed a multicenter trial that included 38 patients
with occluded stents. Occlusions were managed by insertion of
another metallic stent in 19, insertion of a plastic stent in 20 and
mechanical cleaning in 5. There was no statistical difference in
revised stent patency among the different treatment groups. It was
noted that it is more cost-effective to place a plastic stent for the
management of an occluded metallic stent. A similar single-center
REFERENCES
1. Schmassmann A, von Guten E, Knuchel J, et al. Wallstents versus
plastic stents in malignant biliary obstruction: Effects of stent
patency of the rst and second stent on patient compliance and
survival. Am J Gastroenterol 1996; 91:654659.
2. Leung J, Ling T, Kung J, et al. The role of bacteria in the blockage
of biliary stents. Gastrointest Endosc 1988; 34:1922.
3. OBrien S, Hateld A, Craig P, et al. A three year follow up of self
expanding metal stents in the endoscopic palliation of longterm
174
study was performed by Bueno et al.62 In this study, placement of a
second Wallstent or plastic stent was equally effective. A covered
Wallstent may be chosen for the management of an occluded SEMS.
Patient outcome was better if they had a distal bile duct occlusion
as opposed to a proximal bile duct occlusion. The means of SEMS
revision should be individualized and take into consideration the
overall prognosis of the patient.
Once deployed, migration of non-covered SEMS is rare. Migra-
tion of covered SEMS occurred in 6% of patients in the study by
Kahaleh et al.35 Initial iterations of covered stents were 100% covered
which appeared to increase the risk of SEMS migration. Subsequent
modications have left the proximal and distal ends of the stent
uncovered. This allows some degree of embedding into the bile duct
wall to retard migration. Cholecystitis has occurred in 2.9% to
12% of patients with covered biliary SEMS owing to cystic duct
obstruction.35
COST
SEMS vs Plastic
Overall biliary SEMS have longer patency than FDPS. The major
limiting factor to SEMS is the higher cost. The cost-effectiveness of
the SEMS becomes apparent when multiple FDPS exchanges need
to be performed. A randomized, prospective study by Kaassis et al.11
demonstrated that the number of additional days of hospitalization,
days of antibiotics and the number of ERCPs performed on the
group with the FDPS was higher compared to the group that received
the SEMS. Davids et al.4 demonstrated a 28% reduction in ERCPs
per patient with the use of SEMS. From those two studies as well as
two studies using decision analysis with Markov modeling,63,64 initial
SEMS placement is cost-effective in patients with survival of greater
than six months. In Kaassis, et al.11 patients with liver metastases
had a poorer prognosis with less than six months to live so would
be an ideal group for the placement of plastic stents.
CONCLUSION
Malignant pancreaticobiliary disease presents at an advanced stage
and has a poor prognosis. Endoscopic palliative techniques have
largely replaced surgical bypass. SEMS are effective for the palliation
of malignant biliary obstruction. SEMS have more durable patency
versus FDPS. While the upfront cost is greater for SEMS, compared
to FDPS, they require few reinterventions which make SEMS more
cost-effective. Covered SEMS have been more recently introduced
and have been shown to decrease the rate of tumor ingrowth as
compared to uncovered SEMS, however, they have been associated
with increased frequency of complications. Covered SEMS are being
studied for non-malignant applications.
survivors of malignant biliary obstruction. Gut 1995; 36:
618621.
4. Davids PHP, Groen AK, Rauws EA, et al. Randomized trial of self-
expanding metal stents versus polyethylene stents for distal
malignant biliary obstruction. Lancet 1992; 340:14881492.
5. Knyrim K, Wagner HJ, Pausch J, et al. A prospective, randomized,
controlled trial of metal stents for malignant obstruction of the
common bile duct. Endoscopy 1993; 25:207212.

6. Warshaw AL, Fernandez-del Castillo C. Pancreatic carcinoma. N
Engl J Med 1992; 326:455465.
7. Rosewicz S, Wiedenmann B. Pancreatic carcinoma. Lancet 1997;
349:485489.
8. Smith AC, Dowsett JF, Russell RCG, et al. Randomised trial of
endoscopic stenting versus surgical bypass in malignant low bile
duct obstruction. Lancet 1994; 344:16551660.
9. Speer AG, Cotton PB, Russell RCG, et al. Randomised trial of
endoscopic versus percutaneous stent insertion in malignant
obstructive jaundice. Lancet 1987; 1:5762.
10. Carr-Locke DL, Ball TJ, Connors PJ, et al. Multicenter randomized
trial of Wallstent biliary endoprosthesis versus plastic stents.
Gastrointest Endosc 1993; 39:A310.
11. Kaassis M, Boyer J, Dumas R, et al. Plastic or metal stents for
malignant stricture of the common bile duct? Results of a
randomized prospective study. Gastrointest Endosc 2003;
57:178182.
12. Prat F, Chapat O, Ducot B, et al. A randomized trial of endoscopic
drainage methods for inoperable malignant strictures of the
common bile duct. Gastrointest Endosc 1998; 47:17.
13. Pisters PWT, Hudec W, Lee JE, et al. Preoperative chemoradiation
for patients with pancreatic cancer: Toxicity of endobiliary stents. J
Clin Oncol 2000; 18(4):860867.
14. Spitz FR, Abbruzzese JL, Lee JE, et al. Preoperative and
postoperative chemoradiation strategies in patients treated with
pancreaticoduodenectomy for adenocarcinoma of the pancreas. J
Clin Oncol 1997; 15:928937.
15. Yeo CJ, Abrams RA, Grochow LB, et al. Pancreaticoduodenectomy
for pancreatic carcinoma: Postoperative adjuvant chemoradiation
improves survival-A prospective, single-institution experience.
Ann Surg 1997; 225:621633.
16. Klinkenbijl JHG, Jeckel J, Sahmoud T, et al. Radiotherapy and 5-FU
after curative resection for cancer of the pancreas and
periampullary region: A phase III trial of the EORTC
Gastrointestinal Tract Cancer Cooperative Group. Ann Surg 1999;
230:776784.
17. Pendurthi TK, Hoffman JP, Ross E, et al. Preoperative versus
postoperative chemoradiation for patients with resected
pancreatic adenocarcinoma. Am Surg 1998; 64:686692.
18. Wasan SM, Ross WA, Staerkel GA, et al. Use of expandable
metallic biliary stents in resectable pancreatic cancer. Am J
Gastroenterol 2005; 100:20562061.
19. Chen VK, Arguedas MR, Baron TH. Expandable metal biliary stents
before pancreaticoduodenectomy for pancreatic cancer: A
Monte-Carlo decision analysis. Clinical Gastroenterology and
Hepatology. 2005; 3:12291237.
20. Smits ME, Rauws EAJ, Tytgat GNJ, et al. Preliminary results of a
prospective randomized study of partially covered Wallstents vs
noncovered Wallstents. Gastrointest Endosc 1995; 41:AB484.
21. Born P, Neuhaus H, Rosch T, et al. Initial experience with a new,
partially covered Wallstent for malignant biliary obstruction.
Endoscopy 1996; 28:699702.
22. Miyayama S, Matsui O, Terayama N, et al. Covered Gianturco
stents for malignant biliary obstruction: preliminary clinical
evaluation. Journal of Vascular and Interventional Radiology. 1997;
8(4):641648.
23. Rossi P, Bezzi M, Salvatori FM, et al. Clinical experience with
covered Wallstents for biliary malignancies: 23-month follow-up.
Cardiovasc Intervent Radiol 1997; 20:441447.
24. Hausegger KA, Thurnher S, Bodendorfer G, et al. Treatment of
malignant biliary obstruction with polyurethane-covered
Wallstents. Am J Roentgenol. 1998; 170:403408.
25. Shim CS, Lee YH, Bong HK, et al. Preliminary results of a new
covered biliary metal stent for malignant biliary obstruction.
Endoscopy 1998; 30:345350.
Chapter 17 Expandable Stent Insertion
26. Nakamura T, Hirai R, Kitagawa M, et al. Treatment of common
bile duct obstruction by pancreatic cancer using various stents:
single-center experience. Cardiovasc Intervent Radiol 2002;
25:373380.
27. Han YM, Hwang SB, Lee ST, et al. Polyurethane-covered self-
expandable nitinol stent for malignant biliary obstruction:
preliminary results. Cardiovasc Intervent Radiol 2002; 25:381387.
28. Isayama H, Komatsu Y, Tsujino T, et al. Polyurethane-covered
metal stent for management of distal malignant biliary
obstruction. Gastrointest Endosc 2002; 55:366370.
29. Schoder M, Rossi P, Uacker R, et al. Malignant biliary obstruction:
treatment with ePTFE-FEP-covered endoprostheses-initial
technical and clinical experiences in a multicenter trial. Radiology
2002; 225(1):3542.
30. Bezzi M, Zolovkins A, Cantisani V, et al. New ePTFE/FEP-covered
stent in the palliative treatment of malignant biliary obstruction. J
Vasc Interv Radiol 2002; 13:581589.
31. Han YM, Jin GY, Lee SO, et al. Flared polyurethane-covered self-
expandable nitinol stent for malignant biliary obstruction. J Vasc
Interv Radiol 2003; 14:12911301.
32. Isayama H, Komatsu Y, Tsujino T, et al. A prospective randomized
study of covered versus uncovered diamond stents for the
management of distal malignant biliary obstruction. Gut 2004;
53:729734.
33. Miyayama S, Matsui O, Akakura Y, et al. Efcacy of covered
metallic stents in the treatment of unresectable malignant biliary
obstruction. Cardiovasc Intervent Radiol 2004; 27:349354.
34. Nakai Y, Isayama H, Komatsu Y, et al. Efcacy and safety of the
covered Wallstent in patients with distal malignant biliary
obstruction. Gastrointest Endosc 2005; 62:742748.
35. Kahaleh M, Tokar J, Conaway MR, et al. Efcacy and complications
of covered Wallstents in malignant distal biliary obstruction.
Gastrointest Endosc 2005; 61:528533.
36. Yoon WJ, Lee JK, Lee KH, et al.. A comparison of covered and
uncovered Wallstents for the management of distal malignant
biliary obstruction. Gastrointest Endosc 2006; 63(7):9961000.
37. Familiari P, Bulajic M, Mutignani M, et al. Endoscopic removal of
malfunctioning biliary self-expandable metallic stents. Gastrointest
Endosc 2005; 62:903910.
38. Kahaleh M, Brock A, De La Rue SA, et al. Temporary placement of
covered self expandable metal stents (SEMS) in benign biliary
strictures: preliminary data. Gastrointest Endosc 2005; 61:AB208.
39. Baron TH, Poterucha JJ. Insertion and removal of covered
expandable metal stents for closure of complex biliary leaks. Clin
Gastroenterol Hepatol 2006; 4(3):381386.
40. Dumonceau JM, Cremer M, Auroux J, et al. A comparison of
Ultraex Diamond stents and Wallstents for palliation of distal
malignant biliary strictures. Am J Gastroenterol 2000; 95:670676.
41. Rajiman I, Amin V, Siddique I, et al. The use of the Diamond stent
(DS) in the treatment of malignant bile duct stricture. Gastrointest
Endosc 1999; 49:AB235.
42. Seecoomar LF, Cohen SA, Kasmin FE, et al. Preliminary
experience with the Ultraex Diamond stent for management of
malignant biliary obstruction. Gastrointest Endosc 1999; 49:AB236.
43. Siqueira E, Martin JA, Vargas, et al. Prospective evaluation of a
new metal stent for treating malignant biliary obstruction.
Gastrointest Endosc 1999; 49:AB236.
44. Shah RJ, Howell DA, Desilets DJ, et al. Multicenter randomized trial
of the spiral Z-stent compared with the Wallstent for malignant
biliary obstruction. Gastrointest Endosc 2003; 57:830836.
45. Freeman ML, Overby C. Selective MRCP and CT-targeted drainage
of malignant hilar biliary obstruction with self-expanding metallic
stents. Gastrointest Endosc 2003; 58:4149.
46. Pedersen FM, Lassen AT, Schaffalitzky de Muckadell OB.
Randomized trial of stent placement above and across the
175
 SECTION 2 TECHNIQUES
sphincter of Oddi in malignant bile duct obstruction. Gastrointest
Endosc 1998; 48:574579.
47. Liu Q, Khay G, Cotton PB. Feasibility of stent placement above the
sphincter of Oddi (inside-stent) for patients with malignant
biliary obstruction. Endoscopy 1998; 30:687690.
48. Dowsett JF, Vaira D, Hateld AR, et al. Endoscopic biliary therapy
using the combined percutaneous and endoscopic technique.
Gastroenterol 1989; 96:11801186.
49. Deviere J, Baize M, de Toeuf J, et al. Long term follow-up of
patients with hilar malignant stricture treated by endoscopic
internal biliary drainage. Gastrointest Endosc 1988; 34:95101.
50. DePalma GD, Pezzullo A, Rega M, et al. Unilateral placement of
metallic stents for malignant hilar obstructions: A prospective
study. Gastrointest Endosc 2003; 58:5053.
51. Hookey LC, Le Moine O, Deviere J. Use of a temporary plastic
stent to facilitate the placement of multiple self-expanding metal
stents in malignant biliary hilar strictures. Gastrointest Endosc
2005; 62(4):605609.
52. Watanapa P, Williamson RC. Surgical palliation for pancreatic
cancer: developments during the past two decades. Br J Surg
1992; 102:608613.
53. Maetani I, Ogawa S, Hoshi H, et al. Self-expanding metal stents for
palliative treatment of malignant biliary and duodenal stenosis.
Endoscopy 1994; 26:7014.
54. Soetikno RM, Carr-Locke DL. Expandable metal stents for gastric
outlet, duodenal, and small intestinal obstruction. Gastrointest
Endosc Clin North A 1999; 9:447458.
55. Soetikno RM, Lichtenstein DR, Vadervoort J, et al. Palliation of
malignant gastric outlet obstruction using an endoscopically
placed Wallstent. Gastrointest Endosc 1998; 47:267270.
176
56. Yates III MR, Morgan DE, Baron TH. Palliation of malignant gastric
and small intestinal strictures with self-expandable metal stents.
Endoscopy 1998; 30:266272.
57. Yim HB, Jacobson BC, Saltzman JR, et al. Clinical outcome of the
use of enteral stents for palliation of patients with malignant
upper GI obstruction. Gastrointest Endosc 2001; 53:329332.
58. Kaw M, Singh S, Gagneja H. Clinical outcome of simultaneous
self-expandable metal stents for palliation of malignant biliary
and duodenal obstruction. Surgical Endoscopy 2003;
17:457461.
59. Vanbiervliet G, Piche T, Caroli-Bosc FX, et al. Endoscopic argon
plasma trimming of biliary and gastrointestinal metallic stents.
Endoscopy 2005; 37(5):434438.
60. Chen YK, Jakribettuu V, Springer EW, et al. Safety and efcacy of
argon plasma coagulation trimming of malpositioned and
migrated biliary metal stents: a controlled study in the porcine
model. Am J Gastroenterol 2006; 101:16.
61. Tham TCK, Carr-Locke DL, Vandervoort J, et al. Management of
occluded biliary Wallstents. Gut 1998; 42:703707.
62. Bueno JT, Gerdes H, Kurtz RC. Endoscopic management of
occluded biliary Wallstents: a cancer center experience.
Gastrointest Endosc 2003; 58:879884.
63. Arguedas MR, Heudebert GH, Stinnett AA, et al. Biliary stents
in malignant obstructive jaundice due to pancreatic carcinoma:
a cost effective anaylsis. Am J Gastroenterol 2002;
97:897904.
64. Yeoh KG, Zimmerman MJ, Cunningham JT, et al. Comparative
costs of metal versus plastic biliary stent strategies of malignant
obstructive jaundice by decision analysis. Gastrointest Endosc
1999; 49:466471.
 SECTION 2 TECHNIQUES
Chapter
Stent Removal: Migrated and
18 Non-Migrated
Tiing Leong Ang, Stefan Seewald and Nib Soehendra
BOX 18.1 SUMMARY OF KEY
TECHNIQUES
Techniques that do not maintain duct access:
Direct grasping with a basket, rat-tooth forceps or snare
Use of an inated balloon catheter to provide lateral traction
Techniques that maintain duct access:
The occluded or proximally migrated stent is rst cannulated
using a guidewire and catheter. The stent is then retrieved
with retrieval devices such as:
Soehendra stent retriever: The retriever is inserted over
the guidewire and pushed into close proximity with the
stent. Its screw-tip is then rotated into the distal end of the
stent, thus anchoring the stent retriever. The stent is then
pulled out through the accessory channel with continued
duct access maintained by the guidewire.
Other techniques: 1. A partially opened small snare is
passed over the guidewire. The snare is fully opened
around the distal tip of the stent which is grasped over the
distal aps. The stent is removed through the accessory
channel while the guidewire is maintained in position.
2. A 4 mm diameter, 2.5 cm long dilating balloon is
advanced over the guidewire into the stent (size > 10 Fr) and
inated. The stent is withdrawn by pulling the inated
balloon.
INTRODUCTION AND SCIENTIFIC BASIS
Plastic endoprostheses have become an established therapy for a
variety of biliary and pancreatic disorders (see Chapter 16). Endo-
scopic biliary stent placement is used for indications such as drain-
age of malignant or benign biliary obstruction, biliary stricture
dilatation and biliary leakage. Pancreatic stents are used for trans-
papillary drainage of pseudocysts, treatment of pancreatic stulae,
relief of pancreatic duct obstruction from stricture or malignancy,
as well as prophylaxis against post-ERCP pancreatitis.1 These stents
may be temporary, such as in the case of stricture dilatation or pre-
vention of post-ERCP pancreatitis, or intended for long-term treat-
ment. Stent removal will be required once the therapeutic endpoint
is achieved, and when stent occlusion or migration occurs because
of complications such as cholangitis, pancreatitis, duodenal wall
injury or perforation.
Self-expandable metallic stents (SEMS) are increasingly being
used for palliative treatment of malignant biliary obstruction. As
compared to plastic stents, the duration of stent patency is longer
because of the larger luminal diameter. SEMS have also been shown
to be more cost-effective compared to plastic stents for patients who
survive more than 46 months.2 The placement of uncovered SEMS
is generally regarded as permanent due to the fact that they become
deeply embedded in the bile duct wall and induce a marked tissue
reaction making endoscopic removability difcult, if not impossible.
Covered SEMS were developed primarily to extend stent patency by
preventing tumor ingrowth. The covering prevents deep embedding
of the stent into the biliary wall, which, although increasing the risk
of migration, allows them to be potentially endoscopically remov-
able. SEMS are generally only used for palliation of malignant biliary
stenosis with intent to remain in place permanently. SEMS place-
ment in benign biliary strictures has been described3 but it is not
generally accepted. Although uncovered SEMS are generally not
removable, cases of successful extraction have been reported.4
Plastic stents may be removed by directly grasping them
with accessories such as rat-tooth forceps, polypectomy snares, or
stone retrieval baskets. Other methods of direct removal involve
cannulation of the stent lumen with a guidewire followed by inser-
tion of retrieval devices such as the Soehendra stent retriever or a
dilating balloon. Plastic stents can also be removed indirectly by
providing lateral traction force, such as with the use of a stone
retrieval or dilating balloon. SEMS extraction is more difcult
than plastic stents. Covered SEMS can be removed by directly
snaring the distal end within the duodenum, whereas uncovered
SEMS can be removed by grasping and breaking individual wire
laments in order to unravel the stent, though this is not always
technically feasible.
INDICATIONS/CONTRAINDICATIONS OF
STENT REMOVAL
BOX 18.2 INDICATIONS AND
CONTRAINDICATIONS
Indications:
Occluded stents
Migrated stents
177
 SECTION 2 TECHNIQUES
Therapeutic endpoint achieved
To facilitate duct access for further endoscopic treatment
Contraindications:
Presence of severe co-morbid illnesses such that endoscopy
cannot be safely performed
In situations such as terminal malignancies where the clinical
outcome would not be altered
Uncovered self-expandable metallic stents (relative
contraindication)
Indications
1. Occluded stents: Occluded biliary stents need to be removed and
replaced in order to facilitate continued biliary drainage. Occluded
stents lead to jaundice and cholangitis. If occluded pancreatic duct
stents are not replaced, acute pancreatitis may result. If not
removed in a timely fashion chronic pancreatic duct changes may
also occur.
2. Migrated stents: Completely distally migrated biliary and pancre-
atic stents usually pass spontaneously and uneventfully out of the
gastrointestinal tract into the stool. However, if distal migration
of a stent is incomplete, injury to the contra-lateral duodenal wall
may arise; duodenal wall perforation has been reported in this
context. Usually a distally migrated stent may be simply grasped
and removed. Frequently, the distal end is too far to be reached
with a snare or basket and can be grasped with a rat-tooth forceps
at the point where the stent is exiting the papilla. 7 Fr stents can
be grasped with a diagnostic forceps while a therapeutic forceps
is required to grasp 10 Fr stents. Grasping the ap is usually not
effective, since it usually breaks away from the stent. Proximal
migration has been reported to occur at an incidence rate of 4.9%
and 5.2% for biliary and pancreatic stents, respectively.5 Unlike
distally migrated stents, proximally migrated stents are more dif-
cult to retrieve since they cannot be directly endoscopically visu-
alized; guidance is provided by uoroscopy which does not provide
three-dimensional orientation.
3. Therapeutic purpose achieved: When the treatment endpoint has
been achieved, such as after successful dilatation of benign biliary
strictures including anastomotic strictures following liver trans-
plantation,6 those due to chronic pancreatitis,7 and after resolution
of biliary8 or pancreatic9 stula, the stent should be removed.
Temporary pancreatic stents placed after endoscopic papillectomy
or pancreatic duct sphincterotomy require removal if they do not
pass spontaneously.
4. To provide access for further endoscopic interventions: In some situa-
tions stents are inserted for temporary ductal drainage and require
removal before denitive treatment can be carried out. For
example, insertion of a temporary biliary stent may be performed
if bile duct calculi cannot be extracted. Stents require removal in
order to perform additional interventions such as mechanical
lithotripsy or electrohydraulic lithotripsy. Temporary stents have
also been inserted to facilitate healing of biliary and pancreatic
stulae; if these stulae are persistent, sealing may be attempted
with the help of N-butyl-2-cyanoacrylate.10,11 In these instances, in
order to provide adequate duct access for the passage of accessory
devices, the temporary stents must rst be removed.
178
Contraindications
1. Severe comorbid illnesses: In the presence of severe comorbid
medical illnesses, hemodynamic instability, or pulmonary insuf-
ciency, where sedation is hazardous or the patient is unstable,
ERCP and stent removal are contraindicated. In terminally ill
patients, with a migrated stent but who are asymptomatic,
attempted ERCP and stent removal should probably be avoided
since it will not be expected to inuence the clinical course of the
patient.
2. Self-expandable metallic stent: The placement of an uncovered
SEMS is generally regarded as permanent. Uncovered SEMS may
occlude because of tumor ingrowth and/or tissue hyperplasia.
Attempted removal of uncovered SEMS may lead to complica-
tions such as bleeding and perforation because they become
deeply embedded into the tissue soon after placement. Hence,
removal of uncovered SEMS is considered a contraindication,
though in certain circumstances the benets of removal outweigh
the potential risks. The presence of a coagulopathy, however, is
considered a contraindication to removal of uncovered SEMS.
TECHNIQUES OF STENT REMOVAL
Plastic stents
Overview of techniques for plastic stent removal
These techniques include direct grasping with an accessory device
such as rat-tooth forceps, stone retrieval basket, or polypectomy
snare. Indirect techniques utilize traction with an inated occlusion
or dilating balloon that has been advanced alongside the stent fol-
lowed by withdrawal. Direct traction techniques involve cannulation
of the stent lumen with a guidewire followed by insertion of a dilat-
ing balloon or screw extractor. These latter methods allow access to
the biliary tree to be maintained as the stent is withdrawn facilitating
subsequent insertion of a new stent or performance of other thera-
peutic maneuvers. Maintaining duct access with a guidewire is
crucial if the intent is to replace the stent, especially if selective duct
cannulation or traversal of a stricture was initially difcult to achieve.
Migrated stents are generally more difcult to remove compared to
non-migrated stents. With distal migration, the end may be impacted
against the contra-lateral duodenal wall or out of reach of the endo-
scope. Much of the published data on removal of proximally migrated
stents consist of case reports. Three larger case series have been
reported. In these series successful stent removal was achieved in
8590% of cases (Table 18.1).1214 In addition, novel techniques for
the removal of migrated stents have also been described.15,16
Common accessories used
In general, a therapeutic duodenoscope with a working channel of
4.2 mm diameter is preferred. This allows withdrawal of the retrieved
stents (up to 10 Fr) through the accessory channel without having
to withdraw the endoscope. Endoscopic accessories that are com-
monly used for stent removal include the following: polypectomy
snare (e.g. ReSnare (a reusable polypectomy snare), Cook Endos-
copy), rat-tooth forceps (e.g. FG-14P-1, FG-8L-1; Olympus Optical
Co, Tokyo, Japan), Dormia basket (e.g. FG-22Q-1, Olympus Optical
Co, Tokyo, Japan), biliary dilatation balloon (e.g. CRETM wire guided
biliary balloon dilators, Boston Scientic, Natick, MA, USA; BB-1,
Olympus Optical Co, Tokyo, Japan; QBIDTM, Cook Endoscopy),
extraction balloon catheters (e.g. B5-2Q, Olympus Optical Co, Tokyo,
Japan; DASH extraction balloon, Cook Endoscopy), Teon coated
 Chapter 18 Stent Removal: Migrated and Non-Migrated

Biliary Pancreatic Success rate of Management of

1st author stent stent endoscopic retrieval failed retrieval
12
Tarnasky 1995 44 0 38/44 (86%) Radiological
Surgery
Follow-up
13
Lahoti 1998 33 26 Biliary: 28 (85%) Insertion of 2nd stent
Pancreatic: 21 (80%) Surgery
Follow-up
14
Chaurasia 1999 41 0 37/41 (90%) Insertion of 2nd stent
Total 118 26 124/144 (86%)

Table 18.1 Results of endoscopic removal of proximally migrated stents

Fig. 18.1 Soehendra Stent Retriever (courtesy of Cook Endos-

copy, Winston-Salem, NC, USA).

Fig. 18.3 Grasping an occluded biliary stent with a Dormia

basket.

if a diagnostic duodenoscope with a 2.8 mm accessory channel is

Fig. 18.2 Soehendra Stent Retriever with extended curved plastic
tip facilitating cannulation of the stent (courtesy of Cook Endos-
copy, Winston-Salem, NC, USA).
stainless steel and nitinol tracer guidewires (Cook Endoscopy,
Winston-Salem, NC, USA), Zebra wire (Boston Scientic, Natick,
MA, USA), Soehendra universal catheter and Soehendra stent
retriever (Cook Endoscopy, Winston-Salem, NC, USA) (Figs 18.1
and 18.2). Most of these accessories are readily available from a
variety of medical device companies (Chapter 4).
SPECIFIC TECHNIQUES
1. Direct grasping of stent
The standard method of removing non-migrated or distally migrated
stents involves grasping the distal intra-duodenal portion of the stent
with a polypectomy snare, stone retrieval basket (Fig. 18.3) or rat-
tooth forceps (Fig. 18.4), followed either by withdrawing the endo-
scope completely out of the patient or by depositing the stent in the
stomach for subsequent removal at the end of the procedure. Alter-
natively, the stent may be withdrawn through the accessory channel
of the endoscope, if the diameter of the stent allows. For example,
used, a 7 Fr stent (2.3 mm diameter) may be extracted through the
channel, but a 10 Fr stent (3.3 mm) cannot; similarly, if a therapeutic
duodenoscope with a 4.2 mm accessory channel is used, a 10 Fr
stent may be extracted through the endoscope if it is captured at its
distal end, but if caught in mid-shaft, it tends to fold upon itself,
such that the total diameter is 6.6 mm; in this scenario, the duode-
noscope would have to be withdrawn. If the distally migrated stent
has penetrated the duodenal wall, then neither stone retrieval baskets
nor snares are suitable and a rat-tooth forceps would be required.
This technique of direct grasping may also be utilized for proximally
migrated stents, but in the presence of non-dilated ducts, passage of
accessories may be difcult and the forceps may not have enough
room to adequately open. A rat-tooth forceps can be used to grasp
the distal end of the proximally migrated stent. In addition, a basket
or snare can be used to capture the stent from either the proximal
or distal end initially, but on removal only the distal end should be
grasped in order to avoid perforation.17
If subsequent stent replacement is required, or if the access to
the duct needs to be maintained, the biliary or pancreatic duct must
be cannulated with a guidewire and the stent removed leaving the
wire in place.
2. Indirect technique using balloon traction
To retrieve a proximally migrated stent there are two variations of
this indirect technique in which either a single or double lumen
balloon catheter is passed over a guidewire that has been passed
alongside the stent. Either a stone extraction balloon or a dilatation

179
 SECTION 2 TECHNIQUES

Fig. 18.4 Retrieving an occluded biliary stent with a rat-tooth

forceps.

Fig. 18.5 Removal of proximally migrated biliary stent using a

dilating balloon. The dilating balloon (arrow) and distal end of the
stent (arrowhead) can be seen.
balloon may be used. The balloon is advanced so that it is located
alongside and at the midpoint of the stent. The balloon is then
inated and gradually withdrawn. The indirect traction provided by
the balloon catheter causes the stent to pass distally (Fig. 18.5). An
alternative is to pass the balloon catheter to a point just above the
proximal end of the stent. The balloon is then inated and the stent
retrieved by gradually pulling the balloon catheter distally. The size
of the balloon used depends on the size of the duct; for instance, in
a dilated common bile duct, one may be able to use an 8 or 11.5 mm
dilating balloon, whereas in a non-dilated pancreatic duct, a smaller
size dilatation balloon (4 mm) is used.18

3. Techniques utilizing stent cannulation

A variety of techniques which utilize stent cannulation may be used
for stent retrieval. The stent lumen is cannulated with a guidewire
loaded into a catheter or sphincterotome; the latter allows for upward
deection into the stent. When applying this technique to proximally
migrated stents, the lumen can be difcult to cannulate; use of a
hydrophilic guidewire may be helpful to access the stent lumen. The
Soehendra stent retriever was developed to maintain duct access
during stent removal. This is especially useful when replacing stents
in patients in whom the initial cannulation and stent placement
were very difcult, such as in the case of perihilar cholangiocarci-
noma and multiple stents (Figs 18.618.12).
It can also be used to retrieve a proximally migrated stent upon
initial guidewire cannulation of the stent (Figs 8.138.21). It is a
wire-guided metal spiral retrieval device 200 cm long with a screw
tip 3 to 4 mm long that is rotated into the inner lumen of the stent.19 Fig. 18.6 A patient with perihilar cholangiocarcinoma. Four biliary
Exact alignment of the inner lumen of the stent with the retrieval stents were inserted in order to drain different liver segments; the
insertion of two of these stents required a rendezvous procedure
device is needed for this device to self-thread into the stent. Once after initial unsuccessful ERCP. The gure shows initial cannulation
the stent retriever is securely anchored to the end of the stent, the of one of the occluded stents using a 7 Fr universal catheter.

180
 Chapter 18 Stent Removal: Migrated and Non-Migrated

Fig. 18.8 The catheter was withdrawn and the guidewire left
within the stent.
Fig. 18.7 A guidewire had been successfully inserted through the
occluded stent.

Fig. 18.10 The screw tip of Soehendra stent retriever was rotated
into the distal end of the stent.

Fig. 18.9 Insertion of the Soehendra stent retriever over the

guidewire.

Fig. 18.11 The occluded stent was then removed by the Soehen-
dra stent retriever.

181
 SECTION 2 TECHNIQUES

Fig. 18.12 After the stent had been retrieved, the guidewire was
left in place to facilitate subsequent reinsertion of a new stent.

Fig. 18.15 The glidewire was used to direct the tip of the universal
catheter into the distal end of the migrated stent.

Fig. 18.13 X-ray picture of a proximally migrated biliary stent.

Fig. 18.16 The glidewire had been withdrawn, and the tip of the
universal catheter was inserted in the distal end of the migrated
stent.

Fig. 18.14 Cannulation of the proximally migrated biliary stent Fig. 18.17 A standard guidewire was then inserted through the
using Terumo glidewire guided by a universal catheter. universal catheter into the lumen of the migrated stent.

182

Fig. 18.18 The universal catheter was then withdrawn leaving the
guidewire in place.
Fig. 18.19 The Soehendra stent retriever had been inserted over
the guidewire and its tip had been rotated into the end of the
proximally migrated stent.
Fig. 18.20 X-ray picture of the stent being withdrawn by the
Soehendra stent retriever.
Chapter 18 Stent Removal: Migrated and Non-Migrated
Fig. 18.21 The migrated stent had now been pulled out through
the major papilla.
stent is withdrawn through the biopsy channel. The guidewire is
maintained during exchange as when any device is exchanged.
Depending on the size of the stent, one would need to select an
appropriately sized stent retriever. Plastic stents made of polyethyl-
ene become softer after a few months and so the tip of the retriever
may not anchor rmly into the end of the stent, resulting in failure
of stent retrieval. In such a situation, a larger size stent retriever (e.g.
11.5 Fr retriever for a 10 Fr stent) should be used. Newer stents such
as the Viaduct (GI Supply, Camp Hill, PA) and the Marathon
antireux stent (Cook Endoscopy, Winston-Salem, NC) (Chapter 16)
are not amenable to this type of removal.
Another method of stent removal using the stent cannulation
technique is passage of a 4 mm standard biliary dilating balloon into
the stent (Fig. 18.22). The balloon is advanced over the guidewire
into the stent and inated. The amount of ination necessary to
provide enough traction is less than for stricture dilatation; 4 atmo-
spheres of pressure is usually adequate. The balloon ination port
is then locked and the stent is withdrawn by withdrawing the balloon
while the wire is kept in place. This method can only be used for
stents 10 Fr and larger.20 For removal of proximally migrated 5 Fr
pancreatic stents, interventional cardiology dilating balloons, such
as a 3 mm angioplasty balloon mounted on 2.5 Fr catheter, can be
used.15
Yet another technique of removal over the wire utilizes a standard
polypectomy snare (Figs 18.2318.25).2122 The stent lumen is can-
nulated with a standard guidewire as mentioned above. A partially
opened snare (e.g. 5 Fr mini-snare, snare diameter 5 mm, snare
length 1.5 cm, WCMSR-1; Cook Endoscopy, Winston-Salem, NC) is
then passed over the guidewire and through the accessory channel.
Once in view, the snare is fully opened and advanced to encircle the
distal tip of the stent. The snare is closed to grasp the stent and the
stent is withdrawn through the biopsy channel while the guidewire
position is maintained.22
183
 SECTION 2 TECHNIQUES

A B Fig. 18.22 Balloon over the guidewire

technique. A guidewire was rst inserted
into the stent. Next a balloon catheter
was inserted over the guidewire into the
stent. The balloon would then be inated
(either at the proximal end of the stent
A or within the stent lumen B) and used
to pull the stent down, while keeping the
guidewire in place for subsequent stent
reinsertion.

Fig. 18.23 This was a patient with pancreas divisum who had been Fig. 18.24 The distal end of the pancreatic stent was then captured
treated with pancreatic stenting due to dorsal duct disruption fol- by the snare.
lowing pancreatitis. A guidewire had already been inserted through
the stent, and a partially opened snare was being advanced over
the guidewire toward the stent.

184

4. Novel techniques
In addition to the use of a rat-tooth forceps, the lasso technique
had also been reported as a means of retrieving distally migrated
plastic stents which have impacted against and perforated the con-
tralateral duodenal wall.23 A 0.035 inch guidewire was passed behind
the stent and the tip of the guidewire grasped by a polypectomy snare
in front, thus looping the stent. The stent is then winched into close
proximity with the endoscope and the entire apparatus retrieved
together with the stent (Fig. 18.26). Instead of a snare, a retrieval
basket had also been used for this purpose.16
In a case report the removal of proximally migrated 10 Fr biliary
stent in which the distal end had become embedded in the wall of
the common bile duct was described. A 0.035-inch Glidewire (Boston
Scientic Corp, Natick, Mass.) was advanced beyond the proximal
end of the stent within the common hepatic duct and then looped
at the bifurcation with the tip of the guidewire returning toward the
papilla. A snare was inserted through the accessory channel with the
wire passing through the loop of the snare. The end of the Glidewire
protruding from the papilla was grasped with the snare, tightening
the wire and lassoing the stent. With further traction, the stent could
be kinked and withdrawn.24
Some patients undergoing liver transplantation have a T-tube
inserted across the choledochocholedochostomy to ensure anasto-
Fig. 18.25 The pancreatic stent was then removed by the snare
while the guidewire remained in place.
A B
Duodenoscope
Snare over stent and
guidewire underneath
Fig. 18.26 The lasso technique for removing a distally migrated stent that had impacted against the contralateral duodenal wall. A A
snare was inserted over the stent while a guidewire was inserted below the stent. B The guidewire was then captured by the snare,
creating a lasso over the distal end of the stent which was then winched into close proximity with the endoscope, thus freeing the
impacted distal end.
Chapter 18 Stent Removal: Migrated and Non-Migrated
motic patency. A stent modied from a 5 Fr latex urinary catheter
may be used intraductally to maintain anastomotic patency, and
being without aps, it is intended to pass spontaneously within
weeks. A case had been reported in which such a latex stent did not
pass spontaneously and was found to be folded back on itself within
the bile duct. In this instance, ERC was performed and a guidewire
was inserted without performing papillotomy. A rotatable rat-toothed
forceps (Olympus America Corp, Lehigh Valley, PA) was inserted
through the working channel and manipulated around the guide-
wire such that the guidewire was entrapped in the notch when the
forceps was closed. The forceps was then introduced easily into the
common bile duct in tandem with the guidewire, then disengaged,
allowing the catheter to be grasped and removed without disrupting
the architecture of the papilla.25
Special considerations
Newer plastic stents that have become available such as the Viaduct
stent and antireux stent (Chapter 16) may pose special problems
for removal. The Viaduct has a small central lumen that only allows
passage of a guidewire; thus balloon passage and stent retriever
passage into the lumen are not possible. With proximal migration,
they may be more difcult to grasp. In patients with the antireux
stent, the wind-sock on the distal end makes it nearly impossible to
cannulate the stent lumen. In addition in the event of proximal
migration, the wind-sock may tear if grasped with a rat-tooth
forceps.
SELF-EXPANDABLE METALLIC STENTS
Overview of techniques for metallic stent removal
Unlike plastic stents, the techniques used for removal of SEMS
involve direct grasping of the stent. For removal of uncovered SEMS
embedded into the mucosa, initial fragmentation of the individual
wires in order to cause the stent to collapse may be required. Covered
stents may be removed by grasping the stent with a snare.
Specic techniques
For an uncovered SEMS which is embedded in the bile duct mucosa,
snare removal will be impossible. An option then would be the initial
extraction of individual wire laments from the distal end of the
A lasso is formed as the snare
grasps the guidewire, and is
used to winch the tip of the
stent from the duodenal wall
185
 SECTION 2 TECHNIQUES
SEMS using a hot biopsy forceps so that the stent architecture is
destroyed, thus allowing the stent to collapse. Once the stent has
collapsed, it may then be possible to extract it using a snare. It may
be necessary to coagulate the lumen within the stent rst before
stent extraction in order to reduce the risk of bleeding. If a SEMS
has migrated distally and impacted upon the opposite duodenal wall,
it may be necessary to attempt cutting the stent with argon plasma
coagulation rst before extraction with a snare. In one series SEMS
removal was successful in 17 out of 18 patients. Among these suc-
cessfully removed SEMS, 13 were covered and 4 uncovered stents.4
In a larger series, endoscopic removal of 39 SEMS (13 uncovered
and 26 covered) was attempted. It was successful in 29 SEMS
(74.3%). Covered SEMS were effectively removed more frequently
than uncovered ones: 24 of 26 (92.3%) and 5 of 13 (38.4%), respec-
tively (p < 0.05). No major complications were recorded. The pres-
ence of a stent covering was the only factor predictive of successful
stent extraction. The presence of diffuse and severe ingrowth was
the main feature limiting SEMS removal.26
For removal of covered SEMS in which the distal end is lying free
in the duodenum beyond the papilla, the distal end is grasped with
a polypectomy snare as far proximally as possible based upon the
amount of stent remaining within the duodenum. The snare is then
closed, such that the distal end of the SEMS becomes compressed,
whereupon the SEMS elongates and collapses over a short distance
from the point of compression. Either the entire stent can be with-
drawn, or the stent can be withdrawn through the accessory channel
of the endoscope without injury to the patient or damage to the
instrument.
In rare situations SEMS can migrate proximally above the stric-
ture and may be free oating. Stent retrieval may be possible with
either a basket27 or a rat-tooth forceps.28 An interesting method of
removal of distally migrated covered Wallstent which had impacted
on the contralateral duodenal wall had been reported. A closed
biopsy forceps (FB-21K-1, Olympus Optical Co. Ltd, Tokyo, Japan)
was advanced through the stent mesh and opened within the stent
to form an anchor. The endoscope was then withdrawn together with
the opened forceps. During endoscope withdrawal, the stent folded
upon itself in the mid-body and was easily transported from the
duodenum to the stomach. Once inside the stomach one end of the
stent was caught with a snare and the stent was removed by complete
withdrawal of the endoscope.29
COMPLICATIONS AND THEIR MANAGEMENT
BOX 18.3 COMPLICATIONS AND
CONTROVERSIES
Complications:
General complications related to endoscopy and ERCP
Duct injury with bleeding or perforation during retrieval of
proximally migrated stents
Injury to the cardia and distal esophagus when the
unprotected stent is pulled out simultaneously with the
endoscope
186
Failure of stent extraction:
Options include insertion of another stent to maintain duct
drainage, surgical removal of the stent and observation in the
context of an asymptomatic patient.
Controversies:
The placement of a SEMS is generally regarded as permanent.
This is due to the metallic mesh being deeply embedded in
the biliary wall. Attempted removal may be technically
impossible, or may result in bleeding and perforation. Reports
of successful endoscopic removal of SEMS have been
published, especially of the covered type. Hence the presence
of SEMS should no longer be viewed as an absolute
contraindication for stent removal.
General complications
The usual complications of sedation, cardiopulmonary events,
perforation and pancreatitis may occur. Proper patient selection
prior to performing the procedure and close clinical monitoring
during the procedure are important in reducing the risk of
complications.
Complications related specically to stent removal
Although guided by uoroscopy, the retrieval of a proximally
migrated plastic stent is essentially a blind procedure. When acces-
sories such as forceps are used, there is a possibility of accidentally
grasping the ductal epithelium instead of the stent which may cause
ductal injuries such as bleeding and perforation. In addition, if the
stent is grasped at its proximal end instead of its distal end, it may
kink and rotate during withdrawal and potentially cause ductal
injury. Thus it is important to capture the distal end of the stent
before applying traction. In the very rare event of bile duct perfora-
tion during stent removal, the insertion of a new biliary stent may
allow the perforation to close.
One must be careful if the stent and endoscope are withdrawn
together because the distal end of the stent may tear the cardia or
esophageal wall. The elevator of the working channel must be com-
pletely opened. In addition, when using a stone retrieval basket or
snare, allowing a sufcient length between the captured end of the
stent and the distal tip of the duodenoscope will allow a more exible
and smooth retrieval, as the axis of the snare may be aligned with
that of the esophageal lumen.
The removal of an uncovered SEMS, if it is possible, is associated
with an increased risk of bleeding and perforation. Therefore in
patients with uncovered SEMS one should seriously consider
whether removal is absolutely necessary. Options to manage the
stent include placement of another stent, be it a plastic stent or a
SEMS, through the occluded stent.
In the event that endoscopic stent removal is not possible,
management options include surgical removal, insertion of another
stent (if continued duct drainage is required), or observation if
the patient remains asymptomatic. Surgical removal should be
considered in healthy patients with proximally migrated pancreatic
duct stents to minimize the possibility of irreversible ductal
injury.

RELATIVE COSTS AND CHOICE
OF TECHNIQUE
There are no studies that have compared one technique of stent
removal to another. In addition, there are no cost-effectiveness
studies. It is unlikely that such studies will be done. All of the acces-
sories mentioned in this chapter are likely already available and their
use for stent removal should not lead to increased costs. The choice
REFERENCES
1. Soehendra N, Binmoeller KF, Seifert H, et al. Therapeutic
endoscopy: color atlas of operative techniques for the
gastrointestinal tract. 2nd edn. Stuttgart: Georg Thieme Verlag;
2005:30155.
2. Arguedas MR, Heudebert GH, Stinnett AA, et al. Biliary stents in
malignant obstructive jaundice due to pancreatic carcinoma: a
cost-effectiveness analysis. Am J Gastroenterol 2002; 97:898904.
3. Dumonceau JM, Deviere J, Delhaye M, et al. Plastic and metal
stents for postoperative benign bile duct strictures: the best and
the worst. Gastrointest Endosc 1998; 47:817.
4. Kahaleh M, Tokar J, Le T, et al. Removal of self-expandable
metallic Wallstents. Gastrointest Endosc 2004; 60:640644.
5. Johanson JF, Schmalz MJ, Geenen JE. Incidence and risk factors for
biliary and pancreatic stent migration. Gastrointest Endosc 1992;
38:341346.
6. Shah JN, Ahmad NA, Shetty K, et al. Endoscopic management of
biliary complications after adult living donor liver. transplantation.
Am J Gastroenterol. 2004; 99:12911295.
7. Draganov P, Hoffman B, Marsh W, et al. Long-term outcome in
patients with benign biliary strictures treated endoscopically with
multiple stents. Gastrointest Endosc. 2002; 55:680686.
8. Costamagna G, Mutignani M, Ingrosso M, et al. Endoscopic
treatment of postsurgical external pancreatic stulas. Endoscopy.
2001; 33:317322.
9. Sherman S, Shaked A, Cryer HM, et al. Endoscopic management
of biliary stulas complicating liver transplantation and other
hepatobiliary operations. Ann Surg. 1993; 218:167175.
10. Seewald S, Groth S, Sriram PV, et al. Endoscopic treatment of
biliary leakage with N-butyl-2 cyanoacrylate. Gastrointest Endosc.
2002; 56:916919.
11. Seewald S, Brand B, Groth S, et al. Endoscopic sealing of
pancreatic stula by using N-butyl-2-cyanoacrylate. Gastrointest
Endosc. 2004; 59:463470.
12. Tarnasky PR, Cotton PB, Baillie J; et al. Proximal migration of
biliary stents: attempted endoscopic retrieval in forty-one patients.
Gastrointest Endosc 1995; 42:513519.
13. Lahoti S, Catalano MF, Geenen JE, et al. Endoscopic retrieval of
proximally mgrated biliary and pancreatic stents: experience of a
large referral center. Gastrointest Endosc 1998; 47:486491.
14. Chaurasia OP, Rauws EAJ, Fockens P, et al. Endoscopic techniques
for retrieval of proximally migrated biliary stents: the Amsterdam
experience. Gastrointest Endosc 1999; 50:780785.
15. Baron TH, Dean LS, Morgan DE, et al. Proximal migration of a
pancreatic duct stent: endoscopic retrieval using interventional
cardiology accessories. Gastrointest Endosc 1999; 50:124125.
Chapter 18 Stent Removal: Migrated and Non-Migrated
of a particular technique depends on the specic circumstances
requiring stent removal.
In patients in whom the goal is to remove and not replace a plastic
stent, one should simply grasp the stent with a polypectomy snare,
rat-tooth forceps, or stone retrieval basket. If the intent is to replace
a stent it is advisable to initially cannulate the stent with a guidewire
and remove the stent while maintaining duct access by leaving the
guidewire in place.
16. Mergener K, Baillie J. Retrieval of distally migrated, impacted
biliary endoprostheses using a novel guidewire/basket lasso
technique. Gastrointest Endosc 1999; 50:9395.
17. Sharara AI, Leung JW. Endoscopic extraction of proximally
migrated biliary endoprostheses using a grasping rat-tooth
forceps. Gastrointest Endosc 1995; 41:619620.
18. Horwhat JD, Jowell P, Branch S, et al. Proximal migration of a 3
French pancreatic stent in a patient with pancreas divisum:
suggested technique for successful retrieval. JOP. J Pancreas
(Online) 2005; 6:178184.
19. Soehendra N, Maydeo NA, Eckmann B, et al. A new technique for
replacing obstructed biliary endoprostheses. Endoscopy 1990;
22:271272.
20. Martin DF. Wire-guided balloon assisted endoscopic biliary stent
exchange. Gut 1991; 32:15621564.
21. Sherman S, Hawes RH, Uzer MF, et al. Endoscopic stent exchange
using a guidewire and snare. Gastrointest Endosc 1993;
39:794799.
22. Bohorfoush AG, Ballinger PJ, Hogan WJ. A new method for
exchange of endoprostheses in the biliary and pancreatic ducts.
Gastrointest Endosc 1993; 39:799802.
23. Smith FC, OConnor HJ, Downing R. An endoscopic technique for
stent recovery used after duodenal perforation by a biliary stent.
Endoscopy 1991; 23:244245.
24. Vandervoort J, Carr-Locke DL, Tham TCK, et al. A new technique
to retrieve an intrabiliary stent: a case report. Gastrointest Endosc
1999; 49:800802.
25. Mulcahy HE, Cunningham JT. A guidewire-assisted technique for
removing retained biliary stents with rat-toothed forceps during
endoscopic retrograde cholangiography. Gastrointest Endosc
2001; 53:386387.
26. Familiari P, Bulajic M, Mutignani M, et al. Endoscopic removal of
malfunctioning biliary self-expandable metallic stents. Gastrointest
Endosc. 2005 Dec; 62(6):903910.
27. Mo LR, Tsai CC, Yueh SK, et al. Endoscopic retrieval of a self-
expanding stainless steel stent from the common bile duct.
Endoscopy 1994; 26:562563.
28. Baron TH, Blackard WG, Morgan DE. Endoscopic removal of a
oating biliary Gianturco Z stent ve years after placement for a
benign anastomotic stricture in a liver transplant patient.
Gastrointest Endosc 1997; 46:8082.
29. Matsushita M, Takakuwa H, Nishio A, et al. Open biopsy forceps
technique for endoscopic removal of distally migrated and
impacted biliary metallic stents. Gastrointest Endosc 2003;
58:924927.
187
 SECTION 2 TECHNIQUES
Chapter
Papillectomy/Ampullectomy
19 Ann M. Chen and Kenneth F. Binmoeller
INTRODUCTION
Ampullary tumors
Tumors of the ampulla of Vater include adenocarcinomas, lympho-
mas, neuroendocrine tumors, lipomas, bromas, leiomyomas, and
hamartomas, but adenomas are the most common. Ampullary ade-
nomas are premalignant neoplasms and occur sporadically in about
0.04 to 0.12% of the general population based on autopsy series.1,2
The incidence is increased among patients with hereditary polyposis
syndromes, such as familial adenomatous polyposis (FAP) and
hereditary non-polyposis colorectal cancer (HNPCC). In patients
with FAP, 40100% will develop duodenal adenomas and the relative
risk of papillary cancer is over 100 times that of the general
population.3
Surgical treatments
Surgical treatment of ampullary tumors includes transduodenal
local resection (LR) and pancreaticoduodenectomy (PD). For PD, the
possible benet of a low recurrence rate is weighed against a high
risk of complications. Surgical-related morbidity of PD can be as
high as 5063% and mortality ranging from 09% with the higher
rates reported in patients with malignant disease.47 Although LR is
generally associated with lower morbidity (1427%)46 and mortality
rates (04%),57 recurrence rates can be as high as 1732%48 and are notorious for missing carcinomatous foci harbored within papil-
thus endoscopic surveillance after surgery is still required.
Endoscopic ampullectomy/papillectomy
Endoscopic ampullectomy was described in 1983 by Suzuki et al.9
and the rst large case series was reported in 1993 by Binmoeller et
al.10 Over the last decade, many more studies have been published
showing high success rates with low morbidity and mortality risks.
Endoscopic therapy is therefore gaining increasing acceptance for
the treatment of ampullary tumors.
TECHNIQUE
BOX 19.1 SUMMARY OF THE TECHNIQUE
Staging of ampullary tumors and denition of the biliary and
pancreatic ductal anatomy are achieved with high accuracy
by endoscopic ultrasound.
Endoscopic retrograde cholangiopancreatography prior to
ampullectomy is indicated if there is suspicion of intraductal
involvement.
Using a duodenoscope, endoscopic ampullectomy is
performed in a similar manner to snare polypectomy.
Thermal ablation may be used as adjunctive therapy.
Although not denitive, current data supports prophylactic
biliary sphincterotomy and pancreatic stent placement to
prevent ductal stenosis, cholangitis, and pancreatitis.
Local tumor recurrences do occur and therefore follow-up
surveillance duodenoscopy is essential.
Endoscopic evaluations
1. Conventional endoscopy
Evaluation of the ampulla of Vater requires a high degree of suspi-
cion and is best performed with a duodenoscope. Large periampul-
lary diverticula can hinder endoscopic exam (Fig. 19.1). Tumors of
the major ampulla can vary in appearance from slight enlargement
to granular or polypoid, with or without ulceration (Fig. 19.2). Unfor-
tunately, the presence or absence of malignant transformation
cannot be determined by appearance alone and endoscopic biopsies
lary adenomas.5,11,12 Rattner et al.6 found a sensitivity of only 42% for
detection of malignancy by endoscopic biopsies and a specicity of
79% with a poor positive predictive value of 50% and negative predic-
tive value of 73%. Therefore a negative biopsy does not rule out
presence of cancer and a minimum of six biopsies to increase his-
tologic yield has been recommended.10,13 It is important to be aware
that frequency of malignant foci in an adenoma of the papilla gures
around 2630%.4,14
2. Endoscopic ultrasound
Endoscopic ultrasound (EUS) should be performed whenever pos-
sible. It is useful in assessing tumor size, depth of invasion relative
to the duodenal wall layers, and involvement of the pancreatic or
biliary ducts (Fig. 19.3). The radial scanning echoendoscope has
been shown to be valuable for identifying patients with disease local-
ized to the muscularis propria in whom endoscopic resection may
be appropriate for cure. In one study,6 EUS correctly staged preop-
eratively 9 of 12 ampullary lesions (3 of 5 villous adenomas, 3 of 3
adenocarcinomas, 1 carcinoid, and 2 of 2 adenocarcinomas arising
in a villous adenoma). Size of the lesion did not correlate with the
presence of malignancy. Cannon et al.15 reported similar N-stage
accuracy by EUS (68%) compared to CT (59%) and MRI (77%) but
superior T-stage accuracy (78% for EUS vs 24% and 46% for CT
and MRI, respectively). Prior sphincterotomy and the presence of a
189
 SECTION 2 TECHNIQUES

transpapillary endobiliary stent, however, can decrease EUS T-stage Endoscopic ampullectomy
accuracy. One group16 found intraductal ultrasound to be even better There is currently no consensus as to how endoscopic ampullectomy
than conventional EUS for T-staging (87% vs 63%) but more studies should be performed. Randomized trials are lacking. In general,
are needed to conrm these results. ampullectomy is accomplished in a manner similar to snare polyp-
ectomy using a standard duodenoscope (JF-140, TJF-140; Olympus
3. Endoscopic retrograde cholangiopancreatography America, Inc., Leigh Valley, PA). The application of adjuvant thermal
Endoscopic retrograde cholangiopancreatography (ERCP) is helpful therapy, sphincterotomy, and prophylactic stent placement is vari-
in detecting intraductal tumor extension not visualized by computed able (Table 19.1) and will be discussed further later in this section.
tomography (CT) scan (Fig. 19.4). Opacication of both the pancre-
atic and common bile duct should be attempted and surgical referral
should be considered if evidence of ductal involvement is found. In
addition, pancreatography can dene the presence of pancreas
divisum and obviate the need for pancreatic stent placement. While A B
EUS is generally performed before resection to increase staging
accuracy, ERCP is performed at the time of ampullectomy to exclude
intraductal tumor involvement, to perform sphincterotomy, and to
place a pancreatic duct stent. If EUS is not available or the ndings
are equivocal, then ERCP should be performed before ampullectomy
to assess for intraductal extension.

A B D

Fig. 19.2 Endoscopic appearances of tumors of the ampulla of

Fig. 19.1 Ampullary mass obscured by periampullary diverticu-
lum. Ampullary adenoma hidden in a diverticulum A can be appre-
ciated after eversion of the surrounding mucosa B.
Vater. A Villous adenoma with granular surface. B Bilobed
tubular adenoma in a patient with FAP. C T1 submucosal carci-
noid tumor seen as a prominent polypoid mass. D Well-differnti-
ated ampullary adenocarcinoma with bleeding and supercial
ulcerations. FAP, Familial adenomatous polyposis.

A B C

Fig. 19.3 EUS examination of ampullary tumors. A Large ampullary mass with tumor extension into a dilated common bile duct.
B Hyperechoic stage T2 ampullary mass (black arrow) is seen invading into the wall of the duodenum with extension (white arrow) into
the muscularis mucosa (black arrowheads denote intact muscularis). C Large ampullary adenoma. As compared to b, the muscularis is
intact (arrowheads).

190
 Chapter 19 Papillectomy/Ampullectomy

Although no data exist on prophylactic antibiotics for endoscopic
ampullectomy, the procedure is similar to other submucosal resec-
tions and we routinely give antibiotics prior to and for three days
after the procedure.
1. Snare excision
Various types of standard monopolar diathermic polypectomy snares
have been used. We prefer a more rigid snare such as the 20 mm
oval snare with spiral wires (SnareMaster, Olympus America Inc.,
Lehigh Valley, PA). In our experience, a stiffer wire can be more
easily positioned parallel to the plane of dissection and perpendicu-
lar to the catheter for a uniform excision to the level of the muscu-
laris propria (Fig. 19.5). Blended cut is more commonly used than
pure cut current to decrease the risks of bleeding.
No studies have shown a decreased incidence of tumor recur-
rence when en bloc (Fig. 19.6) rather than piecemeal endoscopic
ampullectomy is performed. However, as a general rule of oncologic
surgery, en bloc resection is preferred because of higher likelihood
of complete tumor excision and improved histologic analysis of the
resected margins. The downside of en bloc resection is that it can
be technically more difcult to perform and may incur higher risks
of bleeding and perforation, especially when lesions are large or
extremely sessile. Piecemeal excision is therefore sometimes neces-
sary, although complete tumor removal may require several ses-
sions. Vital dye staining with indigocarmine or methylene blue may
help to delineate the tumor margins prior to resection (Fig. 19.7).
Common bile duct Pancreatic duct

A1 A2

Ampulla of Vater

B Major duodenal papilla

Fig. 19.5 Snare ampullectomy. Schematic drawing showing the

ideal plane of dissection during snare ampullectomy.

A B

Fig. 19.4 ERCP examination of ampullary tumors. A1 Routine

endoscopic visualization of the ampulla after prior sphinctero-
tomy did not reveal obvious residual adenomatous tissue but
balloon sweep of the bile duct during ERCP A2 showed intra-
ductal adenomatous growth. B Intraductal lling defect after con-
trast injection of the common bile duct seen here is consistent with Fig. 19.6 En bloc ampullectomy. A Ampullary adenoma
ampullary ductal tumor extension. ERCP, endoscopic retrograde ensnared in entirety with gentle pressure on the catheter before
cholangiopancreatography. snare closure. B Appearance of same lesion after resection.

Author, y Sphincterotomya Thermal therapy, Pancreatic stent, Biliary stent,

(reference no.) N FAP (biliary, pancreatic) (APC/E/YAG) size (Fr) size (Fr)
Binmoeller, 1993 (10) 25 NA 9 (36%), 0 NA 1 (4%), 7 0
Desilets, 2001 (27) 13 7 (54%) 13 (100%), 13 (100%) 7 (54%), (5/2/0) 11 (85%), 5 3 (23%), NA
Zadorova, 2001 (29) 16 1 (6%) 16 (100%), 8 (50%) 3 (19%), (3/0/0) 6 (38%), 7 3 (19%), 10
Norton, 2002 (26) 26 15 (58%) 26 (100%), 2 (8%) 12 (46%), (2/10/0) 10 (38%), 5 NA
Catalano, 2004 (22) 103 31 (30%) NA 35 (34%), (18/14/3) 91 (88%), 57 34 (33%), 7-10
Cheng, 2004 (30) 55 14 (25%) 2 (4%), 0 NA 41 (75%), 35 NA

Table 19.1 Techniques of endoscopic ampullectomy in published series

FAP, Familial adenomatous polyposis; NA, not available; APC, argon plasma coagulation; E, electrocautery; YAG, YAG laser.
a
Pre- or post-ampullectomy.

191
 SECTION 2 TECHNIQUES
A1 A2
B1 B2
Fig. 19.7 Vital dye staining. A1 Ampullary adenoma with indis-
tinct borders. A2 Same lesion with well-dened borders after
methylene blue staining. B1 A small recurrent adenoma after
prior ampullectomy is better visualized with B2 Indigocarmine
staining.
The role of submucosal injection with saline or dilute epinephrine
prior to ampullectomy remains unclear. Extrapolating from the prac-
tice of submucosal injection prior to mucosectomy, there is the theo-
retical benet of a reduction in risk of perforation and bleeding. The
inability to obtain a cleavage plane with saline injection may be useful
in predicting the presence of malignancy in ampullary neoplasia.17
However, ampullary tumors differ from mucosal neoplasms in that
the bile and pancreatic ducts are embedded in the tissue. A submu-
cosal injection will fail to raise the tissue at the site of ductal insertion.
Furthermore, submucosal injections may increase the risk of proce-
dure-induced pancreatitis. We concur with the ndings of Harewood
et al.18 that submucosal lifts deter complete resection of ampullary
adenomas down to the sphincteric musculature and hinder subse-
quent access to both pancreatic and biliary ducts. In our practice we
do not perform submucosal injection prior to ampullectomy.
Recently novel techniques to facilitate endoscopic ampullectomy
have been proposed. Aiura et al.19 reported successful en bloc resec-
tion of ampullary tumors less than 2 cm in size with the use of an
intraductal balloon-catheter for traction. Another group in Korea20
inserted a guidewire into the pancreatic duct preampullectomy to
maintain ductal access for stent placement immediately after snare
resection. More study is needed to conrm feasibility and safety of
these inventive methods.
The retrieval of all resected specimens for surgical pathology is
essential for detection of small malignant foci. Specimens should be
retrieved immediately after resection with a snare, basket, or net
(Fig. 19.8). Glucagon administered just before resection may be
helpful in preventing loss of tissue downstream due to small bowel
peristalsis. While some authors suggest pinning the specimen on
polystyrene plates for orientation, our pathologists have not found
this to be necessary.
192
Fig. 19.8 Retrieval of specimen with a net after en bloc ampullary
resection.
2. Thermal ablation
Although thermal ablation was initially used as primary therapy with
acceptable success,21 it is now more commonly applied as adjunctive
treatment. When used alone, thermal ablation precludes complete
histopathologic evaluation and may risk incomplete treatment. Cata-
lano et al.22 reported similar overall success rates among patients
who had adjunctive ablation compared with those who did not (81%,
30 of 37 vs 78%, 52 of 66, respectively). However adjunctive thermal
ablation has not been addressed consistently in most studies, so its
absolute benet is inconclusive.
Thermal modalities to achieve ablation include the Nd:YAG
laser,2325 bipolar and monopolar coagulation,23,2527 photodynamic
therapy,24 and argon plasma coagulation.2529 We use argon plasma
coagulation only when visible residual adenomatous tissue remains
after snare excision. Treatment can also be palliative for patients
with non-operable neoplasms. We apply a 7 Fr argon plasma probe
at a power of 60W and ow rate of 1.21.8 L/ml for ablation. For
intraductal lesions, a multipolar probe is used at a power of 2530W.
If placement of pancreatic and biliary stents is to be performed, both
should be placed before thermal ablation to potentially decrease
complications of stenosis by protecting the exposed pancreatic and
biliary epithelia (Fig. 19.9).
3. Preresection sphincterotomy
There are no data to support any benecial effect of sphincterotomy,
either biliary or pancreatic, prior to ampullectomy. It has been
argued that biliary sphincterotomy may allow a more complete exci-
sion of the ampulla by facilitating access to tissues in the biliary
orice and thereby increase diagnostic accuracy.24,27 It has also been
proposed that biliary and pancreatitic sphincterotomy may decrease
the risks of cholangitis and pancreatitis, respectively. Arguments

A B
Fig. 19.9 Adjuvant thermal therapy. A Pancreatic and biliary
stents placed after endoscopic ampullectomy and before thermal
ablation B with APC. APC, argon plasma coagulation.
against preresection sphincterotomy include increased risks of
bleeding, perforation, and tumor seeding. The resultant thermal
injury from sphincterotomy may impede accurate histopathologic
interpretation of the resected specimen. Moreover, if a biliary or
pancreatic stent is placed routinely after preresection sphincterot-
omy, ensuing ampullectomy will likely be in piecemeal fashion
rather than en bloc.
4. Postampullectomy sphincterotomy and
stent placement
After ampullectomy, separate orices for the biliary and pancreatic
ducts can usually be readily identied. Mixing contrast with methy-
lene blue during pancreatography prior to ampullary resection or
administering secretin after resection may facilitate the identica-
tion of the ductal orice in cases where localization is difcult.
Cholangiopancreatography is performed to dene the ductal anatomy
as described perviously.
The role of prophylactic sphincterotomy to minimize complica-
tions such as pancreatitis, cholangitis, and papillary stenosis remains
controversial. Sphincterotomy will expose the opening to the distal
duct and may thereby allow the detection of intraductal involvement.
This benet will need to be weighed against the risks of bleeding
and perforation. Our protocol is to perform a biliary sphincterotomy
in all patients, and pancreatic sphincterotomy in patients with any
suspicion for pancreatic duct involvement.
Pancreatic stent placement after ampullectomy may be performed
to minimize the risk of post-ERCP pancreatitis but supportive data
is not denitive. A retrospective study by Norton et al.26 found pan-
creatitis developed in 2 of 10 (20%) patients with pancreatic duct
stent placement and 2 of 18 (11%) patients without a stent, but this
difference was not statistically signicant (p = 0.5). In a larger study
using 57 Fr pancreatic stents, Catalano et al.22 also found that both
acute pancreatitis and papillary stenosis occurred more frequently
in patients without stents (17% vs 3% for pancreatitis and 8% vs 3%
for stenosis). However, the overall number of patients with pancre-
atitis (5 of 103) and papillary stenosis (3 of 103) was small and no
randomization was done, making interpretation of the results dif-
cult. Similarly, Cheng et al.30 reported 4 of 41 patients (10%) with
35 Fr pancreatic stents placed prophylactically developed pancreati-
tis compared to 1 of 4 (25%) patients without stents. This was again
Chapter 19 Papillectomy/Ampullectomy
Fig. 19.10 Prophylactic 5 Fr pancreatic stent placement and biliary
sphincterotomy after endoscopic ampullectomy.
not statistically signicant (p = 0.33). Furthermore, the only two
patients in this study who developed biliary and pancreatic stenosis
were among those who had undergone both prophylactic biliary
sphincterotomy and pancreatic stent insertions.
The only prospective, randomized controlled trial of prophylactic
pancreatic stent placement after endoscopic ampullectomy was
reported by Harewood et al.18 All patients underwent biliary sphinc-
terotomy, and pancreatic stents were inserted immediately after
ampullectomy without pancreatic sphincterotomy. Results showed
that single-anged, 3 or 5 cm long, 5 Fr stents reduced pancreatitis
(3 of 19 patients in the unstented group vs 0 of 10 patients in the
stented group, p = 0.02). The study was terminated early in response
to institutional review board concerns about the risk of pancreatitis
in the unstented group. It is important to note that the number of
patients enrolled was smaller than the intended 25 patients in each
arm for a power of 80% to detect a 25% difference in rates of pan-
creatitis, hence a single episode of pancreatitis in the stented group
would have resulted in a nonsignicant p value.31 Also of interest is
that pancreatic stents were endoscopically removed after 24 hours
which is much sooner than the general practice of allowing sponta-
neous stent migration over 1 to 2 weeks and further raises questions
regarding the role of stents in these patients. Further larger scale
prospective studies are needed to prove prophylactic pancreatic
stenting decreases complications of endoscopic ampullectomy,
nonetheless current data from difcult or invasive ERCP does
support empiric stent placement. In our earlier practice,10 pancreatic
stents were inserted only if delayed drainage was noted. We have
subsequently changed to routine stenting after observing a high rate
of postampullectomy pancreatitis in patients without stenting. Typi-
cally a 5 cm long, 57 Fr polyethylene stent without an intraductal
ange is used (Fig. 19.10). An abdominal x-ray is obtained 12 weeks
193
 SECTION 2 TECHNIQUES

later to conrm spontaneous stent migration. A stent that remains

Contraindications to ampullectomy include lesions with
in situ is removed endoscopically without need for ductography.
advanced intraductal involvement or local metastasis.
In contrast to pancreatitis, cholangitis after endoscopic ampul-
Patients unwilling to undergo postresection surveillance
lectomy is a rare occurrence and therefore biliary stenting is not
should also be advised against the procedure.
routinely performed. One exception is in the event that thermal
therapy is required, biliary stent placement should be considered to
With improvements in technique and increasing experience,
ensure adequate drainage and to minimize the risk of stenosis of
indications for endoscopic ampullectomy will likely expand
the biliary orice. We also place a 10 Fr biliary stent if there is evi-
and continue to be modied.
dence of poor bile duct drainage despite biliary sphincterotomy.

Postampullectomy surveillance
Surveillance duodenoscopy applies predominately to adenomatous
Appropriate therapeutic strategy for ampullary tumors is determined
tumors, but the same principle can be used to monitor for the recur-
by the patients general health, tumor characteristics, and experience
rence of rare non-adenomatous lesions. Duodenoscopy should be
of the endoscopist. In general, patients with benign ampullary
performed with multiple biopsies from the site of the resected
tumors are candidates for endoscopic mucosectomy. Exclusion cri-
ampulla, even in the absence of macroscopic recurrence. There are
teria10,27,30 for endoscopic resection include:
no standard guidelines regarding appropriate intervals for post-
1. histologic documentation of coexistent carcinoma;
ampullectomy surveillance. Generally, small (<3 cm) sporadic
2. suspected intraductal tumor extension on EUS or ERCP;
benign adenomas with clean resection margins are re-evaluated at
3. tumor size >4 cm;
3 months initially, then at 6 to 12 months intervals for at least 2 years
4. endoscopic ndings suspicious for malignancy (i.e. indurated
if eradication is conrmed.10,27,30 Some authors recommend extend-
mass, presence of ulceration, excessive friability, or spontaneous
ing initial examinations to 6 months intervals for at least 2 years and
bleeding);
then only when clinically indicated if two consecutive biopsies have
5. poor patient compliance with follow-up;
shown no residual adenomatous tissue.22 In patients with FAP who
6. absence of endoscopic expertise.
require routine surveillance for duodenal polyps, duodenoscopy
It is important to note that the above list of exclusion criteria is not
should be performed indenitely every year after initial negative
absolute and may be modied as more studies are published. Ampul-
biopsies at 6 and 12 months.22,28
lary adenomas up to 7 cm in diameter29 and lesions with intraductal
For large (>3 cm) sporadic adenomas excised in a piecemeal
growth32 have been successfully managed by snare resection. Cases
fashion and for lesions with involved resected margins, repeat
of early T1 ampullary adenocarcinoma have also been treated endo-
ampullectomy with possible thermal ablation as well as ERCP should
scopically.33 Patients who are poor candidates for surgery or who
be performed at 2 to 3 month intervals until ablation is complete.10,22
refuse surgery may also be considered for endoscopic resection and/
Subsequently, follow-up duodenoscopy with biopsies and ERCP
or ablation despite unfavorable tumor characteristics.
should be performed every 6 months for a minimum of 2 years.
Surgery should be considered if intraductal tumor extension is sus-
pected or if invasive carcinoma is found on biopsies. COMPLICATIONS

INDICATIONS AND CONTRAINDICATIONS

BOX 19.3 COMPLICATIONS

BOX 19.2 INDICATIONS AND Complications occur in between 8% and 35% of reported
cases but are usually mild and managed endoscopically.
CONTRAINDICATIONS
These include acute pancreatitis, cholangitis, bleeding,
papillary stenosis, and perforation.
In general, endoscopic resection is indicated in patients
who have benign ampullary lesions less than 4 cm in size Death as a result of endoscopic ampullectomy is very rare.
without evidence of intraductal involvement. These criteria
may be changing, however, as successful resection of larger
lesions and lesions with early intraductal invasion and/or
malignant transformation is increasingly reported in recent
literature.
BOX 19.4 CONTROVERSIES

Endoscopic therapy may be considered in patients with Controversy exists regarding adequate endoscopic
ampullary malignancies who refuse surgery or are not treatment of a lesion with potential for malignant
surgical candidates. transformation and possible undetected neoplastic foci.

Endoscopic ampullectomy should be performed by Currently, there is no randomized, controlled, prospective

appropriately trained and experienced endoscopists because study comparing surgery with endoscopic resection for
of potential serious complications. ampullary tumors.

194
 Chapter 19 Papillectomy/Ampullectomy

bringing the elevator down again when deploying or closing the

The rates of submucosal injection prior to ampullectomy,
hemoclip (again blindly).
thermal ablation of resection margins post-ampullectomy,
sphincterotomy, and prophylactic stent placement remain
debatable.
SUCCESS
Complete tumor eradication is achieved in greater than 85% of
Optimal strategy for postampullectomy surveillance is also
endoscopic ampullectomies for benign adenomas, although several
unknown.
treatment sessions may be necessary (Fig. 19.11 and Table 19.3).
Recurrences, however, do occur and average around 20% in reported
case series. Recurrent lesions are usually benign and most can be
retreated endoscopically. Outcome is inuenced by tumor histology,
Overall complications of endoscopic ampullectomy range between 8 size, and genetic disposition. In a retrospective case series, Catalano
and 35%. Most commonly, mild cases of acute pancreatitis (515%) et al.22 reported higher overall success rates for endoscopic treatment
and bleeding (016%) are reported (Table 19.2). Perforations26,30 and of sporadic lesions compared to genetically determined lesions
orice stenosis26,30 are much less common, and death17,25,34 from (86%, 63 of 72 vs 67%, 20 of 31). Age greater than 48 years, lesion
ampullary resection is rare. size of 24 mm or less, and male gender were also associated with
Patients with FAP and those with severe dysplasia or adenocarci- higher rates of successful resection. The presence of intraductal
noma may have higher complication rates. In a prospective study35
of 25 patients with FAP and routine pancreatic stenting after ampul-
lectomy, 14% developed pancreatitis. However, all of these patients
had adjunctive thermal ablation and some patients had intraductal
ablation of adenomatous tissue so that meaningful conclusions can A
not be made from this data. B
In most case series, complications were managed without surgi-
cal intervention. Pancreatic and biliary stenosis resulting from endo-
scopic ampullary resections can be treated with sphincterotomy,26
stents,22 and/or balloon dilation.30 Acute pancreatitis usually resolves
with conservative therapy while postampullectomy bleeding can
be controlled with endoscopic epinephrine injection, electrocautery,
or hemoclips and rarely requires blood transfusions or emboli-
zations. Norton et al.26 described one patient with duodenal per-
foration who was treated endoscopically with a hemoclip. Applying
clips through a duodenoscope is technically challenging owing
to the angulation at the end of the working channel. The key points Fig. 19.11 Eradication with endoscopic ampullectomy. A Large
to remember are to keep the elevator down when opening the clip 3.5 cm benign extraductal adenoma. B Same lesion after 4 endo-
scopic resections. Multiple biopsies of the ampulla resection site
(thus opening the clip blindly), lifting the elevator to bring the revealed benign duodenal tissue. Scarring from previous excisions
opened clip into view for proper positioning at the target site, then can be seen near the orice.

Author, y Bleeding; Perforation; Papillary stenosis; Overall

(reference no.) N Pancreatitisa management (n) management (n) management (n) morbidity mortality
Binmoeller, 1993 (10) 25 3 (12%) 2 (8%); epinephrine 0 0 5 (20%) 0%
injection
Desilets, 2001 (27) 13 1 (18%) 0 0 0 1 (8%) 0%
Zadorova, 2001 (29) 16 2 (13%) 2 (13%); epinephrine 0 0 4 (25%) 0%
injection
Norton, 2002 (26) 26 4 (15%) 2 (8%); epinephrine 1 (4%); hemoclip 2 (8%) 9 (35%) 0%
injection sphincterotomy
Catalano, 2004 (22) 103 5 (5%) 2 (2%) 0 3 (3%) 10 (10%) 0%
electrocautery (1) sphincterotomy
and hemoclip (1) and stent (2),
surgery (1)
Cheng, 2004 (30) 55 5 (9%) 9 (16%); 1 (2%); 2 (4%); 17 (31%) 0%
nonspecied conservative sphincterotomy
endoscopic (6) and dilation
transfusion only (3)

Table 19.2 Complications of endoscopic ampullectomy in published series

a
All cases of pancreatitis were mild to moderate and managed.

195
 SECTION 2 TECHNIQUES

Mean Mean Recurrences

Author, y follow-up procedures Malignant Intraductal Complete managed by Recurrence
(reference no.) N FAP (mos) per patient foci growth eradication Recurrences endoscopy type
Binmoeller, 25 NA 37 1.1 0 2 NA 6/23a (26%) 4/6 (67%) benign
1993 (10) adenomas,
1 with
intraductal
growth
Desilets, 13 7 (54%) 19 2.7 0 1 12/13 (92%) 0/12a (0%)
2001 (27)
Zadorova, 16 1 (6%) NA 1.4 0 0 16/16 3/16 (19%) 2/3 (67%) benign
2001 (29) (100%) adenomas,
1 with
intraductal
growth
Norton, 26 15 (58%) 9 1.1 1 0 NA 2/20b (10%) 2/2 (100%) extraductal
2002 (26) benign
adenomas
Catalano, 103 31 (30%) 36 1.8 6 0 93/103c 21/103c 11/21 (52%) NA
2004 (22) (90%) (20%)
Cheng, 55 13 (33%) 30 1.3 7 6 37/39d 9/27d (33%) 7/7 (100%) extraductal
2004 (30) (95%) benign
adenomas

Table 19.3 Outcome of endoscopic ampullectomy in published series

FAP, Familial adenomatous polyposis; NA, not available.
a
Excludes lesions with ductal (all sent to surg).
b
Excludes lesion with malignant foci and 5 missing follow-up data.
c
Includes lesions with malignant foci (n = 6).
d
Excludes ductal extension, malig foci, and nl histology (n = 3). Recurrence rate also excludes 12 missing follow-up data.

tumor extension also affects outcome. In a prospective study of
benign tumors excised endoscopically, Bohnacker et al.32 found com-
parable recurrence rates between patients with and without intra-
ductal growth (14%, 4 of 31 vs 15%, 11 of 78). However, long-term
success rate was signicantly higher in the group without ductal
involvement (83% vs 46%, p < 0.001).
RELATIVE COST SAVINGS
Mean length of hospital stay ranges from 1113 days following surgi-
cal transduodenal LR and 1523 days following PD.5,6 By compari-
son, endoscopic ampullectomy is usually performed using conscious
sedation alone and on an outpatient basis with two hours of post-
procedure observation before discharge. As described earlier, the
rates of morbidity and mortality after endoscopic therapy are signi-
cantly lower than those of surgical alternatives and recurrence rates
are similar to those after LR. Complete eradication after endoscopic
resection of ampullary adenomas is eventually successful in greater
than 85% and the number of procedures required range from only
1.1 to 2.7 per patient (Table 19.3) which still translates to signicant
savings over surgery. Finally, in patients with FAP, recurrence rates
are high even after surgical excision36 and these patients will still
require lifelong surveillance duodenoscopy for evaluation of duode-
nal polyps.
There is currently no denitive data suggesting improvement in
overall survival outcome when small ampullary adenomas in the
setting of FAP are aggressively resected. In a study36 of FAP patients
surveyed by duodenoscopy with biopsy for >10 years, the histological
grade of dysplasia increased in only 3 of 12 subjects who initially
had adenoma suggesting that the natural history of these adenomas
may be rather static. However, given that endoscopic biopsy surveil-
lance is known for high false-negative detection rates for malignant
foci and that there are currently no factors stratifying patients who
are likely to progress to ampullary adenocarcinoma, small ampullary
adenomas should probably still be resected when found in patients
with FAP.

REFERENCES
1. Baker H, Caldwell D. Lesions of the ampulla of Vater. Surgery
1947; 21:523531.
2. Shapiro P, Lifvendahl R. Tumors of the extrahepatic bile ducts.
Ann Surg 1931; 94:6179.
3. Offerhaus G, Giardiello F, Krush A, et al. The risk of upper
gastrointestinal cancer in familial adenomatous polyposis.
Gastroenterology 1992; 102:19801982.
4. Cahen D, Fockens P, de Wit L, et al. Local resection or
pancreaticoduodenectomy for villous adenoma of the ampulla of
Vater diagnosed before operation. Br J Surg 1997; 84(7):948951.
5. de Castro S, van Heek N, Kuhlmann K, et al. Surgical management
of neoplasms of the ampulla of Vater: Local resection or
pancreatoduodenectomy and prognostic factors for survival.
Surgery 2004; 136(5):9941002.

196

6. Rattner D, Fernandez-del Castillo C, Brugge W, et al. Dening the
criteria for local resection of ampullary neoplasms. Arch Surg
1996; 131(4):366371.
7. Beger H, Treitschke F, Gansauge F, et al. Tumor of the ampulla of
Vater: Experience with radical resection in 171 consecutively
treated patients. Arch Surg 1999; 134(5):526532.
8. Farnell M, Sakorafas G, Sarr M, et al. Villous tumors of the
duodenum: Reappraisal of local versus extended resection. J
Gastrointest Surg 2000; 4:1321.
9. Suzuki K, Kantou U, Murakami Y. Two cases with ampullary cancer
who underwent endoscopic excision. Prog Dig Endosc 1983;
23:236239.
10. Binmoeller K, Boaventura S, Ramsperger K, et al. Endoscopic
snare excision of benign adenomas of the papilla of Vater.
Gastrointest Endosc 1993; 39:127131.
11. Seifert B, Schulte F, Stolte M. Adenoma and carcinoma of the
duodenum and papilla of Vater: a clinicopathologic study. Am J
Gastroenterol 1992; 87:3742.
12. Yamaguchi K, Enjoji M, Kitamura K. Endoscopic biopsy has limited
accuracy in the diagnosis of ampullary tumors. Gastrointest
Endosc 1990; 36:588592.
13. Shemesh E, Nass S, Czerniak A. Endoscopic sphincterotomy
and endoscopic fulguration in the management of adenoma
of the papilla of Vater. Surg Gynecol Obstet 1989;
169:445448.
14. Heidecke C, Rosenberg R, Bauer M, et al. Impact of grade of
dysplasia in villous adenomas of Vaters papilla. World J Surg
2002; 26:709714.
15. Cannon M, Carpenter S, Elta G, et al. EUS compared with CT,
magnetic resonance imaging, and angiography and the inuence
of biliary stenting on staging accuracy of ampullary neoplasms.
Gastroentest Endo 1999; 50(1):2733.
16. Menzel J, Hoepffner N, Sulkowski U, et al. Polypoid tumors of the
major duodenal papilla: preoperative staging with intraductal US,
EUS, and CTa prospective, histopathologically controlled study.
Gastroentest Endo 1999; 49(3):349357.
17. Kahaleh M, Shami V, Brock A, et al. Factors predictive of
malignancy and endoscopic resectability in ampullary neoplasia.
Am J Gastroenterol 2004; 99:23352339.
18. Harewood G, Pochron N, Gostout C. Prospective, randomized,
controlled trial of prophylactic pancreatic stent placement for
endoscopic snare excision of the duodenal ampulla. Gastrointest
Endosc 2005; 62(3):367370.
19. Aiura K, Imaeda H, Kitajima M, et al. Balloon-catheter-assisted
endoscopic snare papillectomy for benign tumors of the major
duodenal papilla. Gastrointest Endosc 2003; 57(6):743747.
20. Moon J, Cha S, Cho Y, et al. Wire-guided endoscopic snare
papillectomy for tumors of the major duodenal papilla.
Gastrointest Endosc 2005; 61(3):461466.
Chapter 19 Papillectomy/Ampullectomy
21. Saurin J, Chavillon A, Napoleon B, et al. Long-term follow-up of
patients with endoscopic treatment of sporadic adenomas of the
papilla of Vater. Endoscopy 2003; 35:402406.
22. Catalano M, Linder J, Chak A, et al. Endoscopic management of
adenoma of the major duodenal papilla. Gastrointest Endosc
2004; 59(2):225232.
23. Ponnudurai R, Martin J, Haber G, et al. Endoscopic snare
ampullectomy for resection of benign ampullary neoplasm in 25
patients. Gastroentest Endo 2000; 51(4):part 2:AB2213.
24. Ponchon T, Berger F, Chavaillon A, et al. Contribution of
endoscopy to diagnosis and treatment of tumors of the ampulla
of Vater. Cancer 1989; 64:161167.
25. Martin J, Haber G, Kortan P, et al. Endoscopic snare ampullectomy
for resection of benign ampullary neoplasms [abstract].
Gastrointest Endosc 1997; 45:AB458.
26. Norton I, Gostout C, Baron T, et al. Safety and outcome of
endoscopic snare excision of the major duodenal papilla.
Gastrointest Endosc 2002; 56:239243.
27. Desilets D, Dy R, Ku P, et al. Endoscopic management of tumors
of the major duodenal papilla: rened techniques to improve
outcome and avoid complications. Gastrointest Endosc 2001;
54(2):202208.
28. Martin J, Haber G. Ampullary adenoma: clinical manifestations,
diagnosis, and treatment. Gastrointest Endoscopy Clin N Am
2003; 13:649669.
29. Zadorova Z, Dvorak M, Hajer J. Endoscopic therapy of benign
tumors of the papilla of Vater. Endoscopy 2001; 33:345347.
30. Cheng C, Sherman S, Fogel E, et al. Endoscopic snare
papillectomy for tumors of the duodenal papillae. Gastrointest
Endosc 2004; 60:757764.
31. Baille J. Endoscopic ampullectomy: does pancreatic stent
placement make it safer? Gastrointest Endosc 2005;
62(3):371373.
32. Bohnacker S, Seitz U, Nguyen D, et al. Endoscopic snare
resection of benign ampullary tumors of the duodenal papilla
without and with intraductal growth. Gastrointest Endosc 2005;
62(4):551560.
33. Ito K, Fujita N, Noda Y, et al. Case of early ampullary cancer
treated by endoscopic papillectomy. Dig Endosc 2004;
16(2):157161.
34. Churrani R, Cretu D, Pleskow D, et al. Efcacy, safety, and
outcome of endoscopic snare ampullectomy for 31 ampullary
adenomas. Gastrointest Endosc 2002; 55(5):AB166.
35. DiSario J, Giampaolo A, Samowitz W, et al. Endoscopic
management of ampulla of Vater lesions in familial adenomatous
polyposis. Gut 2002; 51:A278.
36. Heiskanen I, Kellokumpu I, Jarvinen H. Management of duodenal
adenomas in 98 patients with familial adenomatous polyposis.
Endoscopy 1999; 31:412416.
197
 SECTION 2 TECHNIQUES
Chapter
Pancreatoscopy
20 Tadashi Kodama, Yoshihide Tatsumi and Tatsuya Koshitani
INTRODUCTION AND SCIENTIFIC BASIS
Various techniques imaging the pancreas have recently been devel-
oped. Among those methods such as computed tomography (CT),
magnetic resonance imaging (MRI) and transabdominal or endo-
scopic ultrasonography, magnetic resonance cholangiopancreatog-
raphy (MRCP) has emerged as an alternative diagnostic technology,
yielding images at times comparable to endoscopic retrograde chol-
angiopancreatography (ERCP). Although ERCP is one of the estab-
lished gold standards for the diagnosis of pancreatic cancer, it, as
well as MRCP, may fail to differentiate the origin of a lling defect
or a stenosis in the main pancreatic duct, because of inability to
directly visualize the duct lumen. This desire of direct vision of the
main pancreatic duct gave birth to the idea of direct pancreatoscopy.
The pancreatoscope was rst developed by the Japanese group of
Takekoshi et al. in 1975.1 It was a small caliber berscope (baby
scope) that could pass through the accessory channel of a duodeno-
scope (mother scope). Although the idea was very attractive and a
few investigators studied its feasibility, pancreatoscopy was unpopu-
lar because of instrument fragility and suboptimal visibility as well
as its relatively large diameter relative to the duodenal papilla. The
description of intraductal papillary mucinous tumor (IPMT) by
another Japanese investigator, Ohashi et al.,2 made an impact on
pancreatoscopy in the early 1980s. Since then, the pancreatoscope
has been reassessed as a useful method to diagnose IPMT by virtue
of characteristic endoscopic ndings resembling salmon eggs.
Initial pancreatoscope prototypes had only optic image ber
bundles without any channel or tip deection. Subsequent devices
of more than 3 or 4 mm in diameter were then developed with one-
or two-way tip deection and with an accessory channel. To insert
the scope into a non-dilated pancreatic duct, an ultrathin pancreato-
scope (0.8 mm in diameter) was also developed by decreasing the
number of optical bers. These thin or ultrathin beroptic scopes
had the same problem of low visibility to different degrees, contin-
gent upon the numbers of optical bundles. Recently a peroral elec-
tronic pancreatoscope (PEPS), the smallest known electronic
endoscope, was developed using the latest high quality television
technology. With the above advancements in technology and
imaging, peroral pancreatoscopy has been assessed and considered
to be a useful technique, although most cases continue to be per-
formed in Advanced Centers of Endoscopic Excellence.
DESCRIPTION OF TECHNIQUE
Equipment
Pancreatoscope
To date, there are two types of commercially available beroptic
pancreatoscopes.3 These scopes can be purchased from several endo-
scopic manufactures. Video pancreatoscopes manufactured by
Olympus Medical Systems (Tokyo, Japan) are commercially avail-
able in Japan.
The most commonly used beroptic pancreatoscope has an outer
diameter of more than 3 mm. It has both tip angulation and an
accessory channel (more than 1.2 mm diameter) which allows biopsy
and lithotripsy under direct visualization.46 Pentax Co. (Tokyo,
Japan) offers a 2.7 mm-miniscope with tip angulation, but its small
0.75 mm forceps channel is not large enough to reliably allow biopsy
and brush cytology. A much smaller scope lacking an angulation
mechanism, with an outer diameter of 1.67 mm and 0.55 mm acces-
sory channel, has also been reported.7
A second type of device is known as the ultrathin pancreatoscope.
This berscope has an outer diameter of 0.75 or 0.8 mm and con-
tains 30006000 bundles of image bers.4,6,8 Since the ultrathin pan-
creatoscope can be inserted through the usual ERCP catheter, it can
be applied easily at the time of ERCP. Although it has no tip angula-
tion or accessory channel, cytology sampling as well as injection
and aspiration of saline are available through the outer cannula
(Figs 20.1A, 20.1B).9
Specications for representative beroptic pancreatoscopes in
Japan are compared in Table 20.1, although a number of other
scopes are available from other manufacturers.
The peroral electronic pancreatoscope (PEPS) was rst described
by Kodama et al. in 1999.10 A new prototype was developed with a
newly designed 50 000-pixel interline charge-coupled device (CCD)
(about 1 mm square in size) (Matsushita Electronics, Osaka, Japan)
in cooperation with Olympus Optical Co. (Tokyo, Japan). The proto-
type PEPS (XPF-22EY) has an outer diameter of 2.1 mm and a bi-
directional angle function without an accessory channel. After its
prototype success, Olympus has developed an enhanced version that
has an accessory channel (XCHF-BP240).11 At present, Olympus
offers two types of commercial electronic pancreatoscopes in Japa-
nese markets. The rst type (CHF-BP260) is a commercial model of
XCHF-BP240 with a small 0.5 mm accessory channel for a guide-
wire (Figs 20.2A20.2E). The second type (CHF-B260) is a larger
scope with a larger accessory channel with use for biopsy and litho-
tripsy under direct visualization. These scopes use a eld sequential
imaging system. Both types of electronic pancreatoscopes using a
simultaneous imaging system have also been developed in the US,
but they are not commercially available. The specications for the
prototypes and these new commercial electronic pancreatoscopes
are compared in Table 20.2.
Mother scope
Direct pancreatoscopy can be safely performed by means of a
mother-baby method. A baby scope with an outer diameter less
than 3.1 mm can be inserted through the channel of a standard
therapeutic duodenoscope. Since larger baby scopes (e.g. with an
199
 SECTION 2 TECHNIQUES

A B

Fig. 20.1 An ultrathin beroptic pancreatoscope, PF-8P (Olympus) and guide catheter. A Overview. B Ultrathin pancreatoscope with
a guide catheter inserted through an accessory channel of duodenoscope (TJF-240).

Tip diameter Channel Angulation up/down View

Manufacturer (mm) (mm) (degrees) (degrees)
CHF-BP30 Olympus 3.1 1.2 160/130 80
PF-8P Olympus 0.8 (1.8a) None none 80
FCP-9P Pentax 3.0 1.2 90/90 90
FCP-8P Pentax 2.7 0.75 90/90 90

Table 20.1 Specications of representative beroptic pancreatoscopes in Japan

a
Outer diameter of guide catheter.

Tip Angulation
diameter Channel up/down View
Imaging (mm) (mm) (degrees) (degrees)
XPF-22EY Simultaneous 2.1 None 120/120 80
(not commercially available)
XCHF-BP240 Field sequential 2.6 0.5 90/90 90
(not commercially available)
CHF-BP260 Field sequential 2.6 0.5 70/70 90
CHF-B260 Field sequential 3.4 1.2 70/70 90

Table 20.2 Specications of prototype and commercially available electronic pancreatoscopes

PEPS, Peroral electronic pancreatoscope.
All electronic pancreatoscopes are manufactured by Olympus Medical Systems Co. at present.

outer diameter more than 4 mm) require a special mother scope tial color television system that is utilized in another ultrathin
which proved expensive and hard to control; they are no longer video endoscope adapted for upper GI endoscopy (e.g. GIF-N230;
commercialized. Olympus).12

Light source and image processor
Direct pancreatoscopy requires a second light source for a baby
scope as well as one for a mother scope. Video pancreatoscopes
also need suitable processors. The prototype (XPF-22EY) works
with a processor modied from a CV-140 processor, using a simul-
taneous color television system. Both new commercial electronic
pancreatoscopes (CHF-BP260 and CHF-B260) and XCHF-BP240
work with a CV-240 or CV-260 processor (Fig. 20.3) using a sequen-
Image converter for beroptic pancreatoscope
Connecting a special video converter (OVC-200) (Fig. 20.4) with a
processor (CV-200 or later) to the head of a conventional beroptic
pancreatoscope facilitates the endoscopic procedure and improves
imaging.8 This converter enables viewing the image on a television
monitor as a sequential electronic endoscope image, although its
image quality is inferior to that of the electronic pancreatoscope. In
the US markets, this video converter is also available as OVC-140

200
 Chapter 20 Pancreatoscopy

A B

C D

Fig. 20.2 Peroral electronic pancreatoscope with a small acces-

sory channel, CHF-BP260 (Olympus). A Overview. B Control
section with an up/down angulation control knob and accessory
channel port. C Deection part of pancreatoscope. D Distal
chip of pancreatoscope (note channel). E Deection section of
pancreatoscope entirely out of an accessory channel of a duode-
nosocpe (TJF-240).

201
 SECTION 2 TECHNIQUES
Fig. 20.3 CHF-BP260 with a CV-260 processor (Olympus America),
using a sequential color television system.
Fig. 20.4 A special video converter, OVC-200 (Olympus) that
enables viewing the image of beroptic pancreatoscope on a tele-
vision monitor. OVC-200 connected with the head part of the
ultrathin beroptic pancreatoscope, PF-8P (Olympus).
202
with CV-160 (Olympus America), processor using a simultaneous
imaging system.
ENDOSCOPIC PROCEDURE
Procedure timing
In our institutions, MRCP is preferable for the initial diagnostic
evaluation of the pancreatic duct. MRCP is usually followed by ERCP
and selective pancreatoscopy when a pancreatic lesion is suspected.
Enough preparation time for direct pancreatoscopy should be con-
sidered if it is undertaken at the time of an initial diagnostic ERCP,
since its application requires signicant time.
Dilatation of the duodenal papilla prior to insertion
When baby scopes with an outer diameter of more than 2.5 mm are
utilized, endoscopic sphincterotomy (EST) of the pancreatic sphinc-
ter is usually needed. EST may be skipped, if the papilla is dilated
in the pathological situation like IPMT (Figs 20.5A, 20.5B). Intrave-
nous administration of a nitric oxide donor such as isosorbide dini-
trate (5 mg/hr) can facilitate the insertion of the scope without EST
even with some larger pancreatoscopes. With the prototype elec-
tronic pancreatoscope (XPF-22EY; outer diameter 2.1 mm) and the
enhanced version with an accessory channel (XCHF-BP240; outer
diameter 2.6 mm), insertion of the scope has been performed
without sphincterotomy with relatively high success rates.10,11
However, the success rate of XCHF-BP240 appears to be inferior to
that of XPF-22EY as a consequence of larger outer diameter.
Insertion of the scope
All pancreatoscopes need a careful insertion procedure because of
their fragility. They are easily damaged by acute angulation over the
elevator of the duodenoscope. After introduction of a mother scope
into the second portion of the duodenum, the main duodenal papilla
is viewed and conventional pancreatography performed.
Pancreatoscopes without an accessory channel are inserted
directly through the papilla, at times alongside of a previously placed
guidewire. Cooperation of two experienced endoscopists for the
mother and baby scope, using a combination of tip angulation and
movement of the duodenoscope, is the key for successful deep inser-
tion (Fig. 20.6).10
Pancreatoscopes which have an accessory channel usually are
inserted over a guidewire placed into the main pancreatic duct.
However, when using the enhanced PEPS with an accessory channel
(XCHF-BP240, CHF-BP260), direct insertion through the papilla is
needed, because it cannot be inserted over any available guidewire,
to date. However, once inserted into the main pancreatic duct, a thin
guidewire through the accessory channel of the instrument can
facilitate deep insertion to the targeted point like other scopes
with an accessory channel (Fig. 20.7)11 Another type of commercial
pancreatoscope, CHF-B260, can be inserted over a 0.035 inch
guidewire.
An ultrathin pancreatoscope is inserted through the guide
cannula. In this setting, a cannula is advanced over a guidewire
under uoroscopy deep into the main pancreatic duct. The guide-
wire is then removed and an ultrathin pancreatoscope is inserted
through the cannula. After the scope has emerged several millime-
ters from the tip of the cannula, the image of the pancreatic duct is
obtained. Observation should be done only during withdrawal of the
cannula of the scope from the duct. Pushing the ultrathin pancre-

A lary mucinous tumor (IPMT) in the main
B were observed at ERCP. PEPS was easily
Fig. 20.6 Electronic pancreatoscopy with a second experienced
endoscopist who handles the baby scope.
atoscope system may damage both scope and mucosa. If investiga-
tion at the distal part of the duct is needed again, the cannula should
be advanced again over a guidewire without the scope.8
Improving visualization
To obtain a clear image, washing with saline is necessary during
every procedure. Techniques to wash the pancreatic duct lumina are
different for each pancreatoscope mentioned above.
When a pancreatoscope without an accessory channel like
XPF-22EY is used, washing is needed through an ERCP catheter
prior to direct pancreatoscopy, if necessary.10 This method has
some limitation when abundant mucus lls the lumen. Secretin
(100 clinical units) has also been injected intravenously to stimulate
the exocrine function of the pancreas and thus improve visualization
with the initial cases utilizing the XPF-22EY miniscope.13 However,
this method cannot be utilized in Japan at present, because the
drug is currently not available from domestic pharmaceutical
companies.
When using a pancreatoscope that has an accessory channel,
washing with saline in the pancreatic duct can be done through
the channel under direct vision. The advantage of this method is the
Chapter 20 Pancreatoscopy
Fig. 20.5 Observation of intraductal papil-
pancreatic duct by visualized PEPS. A Abun-
dant mucin in dilated papilla. B Filling
defects in the dilated main pancreatic duct
inserted without dilatation procedure of
papilla.
Fig. 20.7 Endoscopic image obtained with XCHF-BP240 (Olympus)
during insertion over a guidewire.
ability to obtain clear images even with remarkably thick and
mucinous pancreatic juice as seen in some cases of IPMT. The
XCHF-BP240 pancreatoscope is signicantly superior to the XPF-
22EY when washing under direct visualization (Figs 20.8AC).11
With any debriding, care must be taken not to use excessive force
with irrigation, resulting in acinar rupture and procedural
pancreatitis. With an ultrathin pancreatoscope, washing can be
also performed under direct visualization through the narrow
lumen between the ultrathin scope and the guide catheter. However,
saline ow is sometimes too weak to wash out thick mucus. If
thick mucus precludes visualization, irrigation through the
guide catheter is preferable after temporary removal of the
pancreatoscope.
Biopsy and cytology sampling
A sample of pancreatic juice for cytology can be obtained under
direct vision when the pancreatoscope has an accessory channel.
With an ultrathin pancreatoscope, sampling can be also performed
under direct visualization through the narrow lumen between the
ultrathin scope and the guide catheter. The diagnosis of in situ car-
cinoma of the pancreas has been reported by sampling under direct
203
 SECTION 2 TECHNIQUES

A B
C

Fig. 20.8 Endoscopic images of intraductal papillary mucinous tumors (IPMT) obtained with XCHF-BP240 (Olympus). Note the effect of
washing under direct visualization. Rinsing with saline is sufcient to wash out abundant mucin. After rinsing, an image of egg-shaped
tumor of the pancreatic duct was clearly visualized. A Before washing. B During washing. C After washing. Reprinted from Gastro-
intest Endosc, 59, Kodama T et al., Initial experience with a new peroral electronic pancreatoscope with an accessory channel, 895900,
2004 with permission from American Society for Gastrointestinal Endoscopy.

vision using the ultrathin scope.9 However, when the juice is
extremely turbid or mucinous, it is often impossible to collect a
pancreatic juice sample through the narrow lumen of the ultrathin
scope system or the 0.5 mm channel of an XCHF-BP240.11
Brush cytology catheters and biopsy forceps can be used if the
pancreatoscope has a relatively large accessory channel of 1.2 or
1.7 mm.4,5 When ultrathin pancreatoscopes or other pancreatoscopes
with smaller channels are used for examination, transpapillary brush
cytology or biopsy is needed under uoroscopy after baby scope
removal. In such cases, the pancreatoscopic ndings are referred to
decide the target point in the main pancreatic duct.
PROCEDURAL SUCCESS RATES
It is important to select a scope prior to examination contingent
upon the diameter of the duct and purpose of the study. Most endos-
copists will not have access to multiple instruments.
With a peroral electronic pancreatoscope (PEPS), a high resolu-
tion image is obtained by an ultra-miniature interline charge-coupled
device (CCD). Olympus XPF-22EY, XCHF-BP240 and CHF-BP260
electronic baby scopes are designed to be inserted through the
papilla without pancreatic duct sphincterotomy. They can be inserted
deep into the tail of the non-dilated pancreatic duct unless the duct
is too crooked or small in its previously noted diameter. The CHF-
B260 is designed to function for biopsy and lithotripsy under direct
visualization. However, its application requires a pancreatic sphinc-
terotomy and it cannot be inserted into a non-dilated or extremely
angulated pancreatic duct.
Kodama et al.13 reported the initial use of the XPF-22EY inserted
into the pancreatic duct without sphincterotomy. It was successfully
inserted into the predetermined target point and the pancreatic duct
lesion could be visualized in 42 (75%) of 56 cases. Of the 42 cases,
the PEPS reached the pancreatic head in 10 cases, the pancreatic
body in 15 cases, and the pancreatic tail in 17 cases. In the 14
remaining cases, 6 failures were related to failure to intubate
the papilla and the other 8 to passing the pancreatoscope beyond
the genu of the pancreatic duct. In our preliminary experience, the
XCHF-BP240 could be inserted successfully into the pancreatic or
bile duct without sphincterotomy in 9 of the 11 patients (82%).
Observation of a predetermined target and juice collection with
direct visualization was successful in 8 of 9 patients (89%).11
With the 3.1 mm beroptic pancreatoscope, angulation of the
endoscope facilitates the visualization of the pancreatic duct and
target biopsy or brush cytology is possible. However, this instrument
cannot be inserted into a non-dilated or angulated pancreatic duct.
Sphincterotomy of the pancreatic sphincter is usually needed unless
the orice of the papilla is patulous.
The 3.1 mm pancreatoscope has been reported to perform suc-
cessful pancreatoscopy after a pancreatic duct sphincterotomy in 16
(89%) of 18 cases by Riemann et al.4 Two failures occurred because
of tight strictures in the setting of calcic pancreatitis. In this
study, 5 cases of intraductal cystadenoma and 2 cases of pancreatic
adenocarcinoma were all diagnosed by pancreatoscopic images. His-
tologic conrmation of the above cases was successfully obtained by
biopsy under direct visualization. Two cases of segmental pancreati-
tis were also diagnosed by pancreatoscopic images (smooth stric-
tures in conjunction with negative cytology). With a comparably
small sized pancreatoscope, Jung et al.5 also reported that the pan-
creatic duct lesion could be observed in 15 (83%) of 18 patients with
various pancreatic disorders. The author had a comment on the
limitation of biopsy with this instrument, specically, that angula-
tion in a narrow pancreatic duct often precluded passage of a small
biopsy forceps.
With an ultrathin pancreatoscope, a screening examination is
possible for a non-dilated pancreatic duct without sphincterotomy,
but lack of angulation of the endoscope limits the visualization of
the pancreatic duct. Cytology under direct visualization of the pan-
creatic duct is possible.
Tajiri et al.8 have reported that an ultrathin pancreatoscope was
able to reach a predetermined target point and the pancreatic duct
lesion noted at ERCP could be observed in 42 (81%) of 52 cases
examined. In the 10 inadequate studies, 5 were related to failure to

204

intubate the papilla and the other 5 to failure to pass the pancreato-
scope beyond the genu of the pancreatic duct. Yamao, et al.6 also
reported with the same type scope that 22 of the 35 pancreatic cancer
cases (63%) were observed, although insertion was successful in all
cases. The failure to diagnose 13 cases was because of tapering ste-
nosis or obstruction of the main pancreatic duct with asymmetrical
deviation. The observation rate was increased to 75% when pancre-
atic cancers were less than 2 cm. On the other hand, an observation
rate of benign stenoses of the pancreatic duct and IPMT were
reported to be 16 of 20 (80%) and 57 of 60 (95%), respectively.
ELECTRONIC PANCREATOSCOPIC FINDINGS IN
THE VARIOUS TYPES OF PANCREATIC DISEASES
With electronic pancreatoscopes, various pancreatic diseases were
diagnosed from the endoscopic ndings more precisely compared
with previous reports using beroptic scopes. Furthermore, ne
capillary vessels on the surface of the pancreatic duct mucosa was
rst clearly observed by PEPS.
Fig. 20.9 Endoscopic image of the pancreatic duct of a healthy
control subject obtained with PEPS system. Network of ne vessels
clearly visualized. Reprinted from Gastrointest Endosc, 49, Kodama
T et al., Pancreatoscopy for the next generation: development
of the peroral electronic pancreatoscope system, 366371, 1999
with permission from American Society for Gastrointestinal Endos-
copy and Elsevier.
A B
Chapter 20 Pancreatoscopy
With PEPS, in normal cases, smooth pancreatic duct walls with
white to pink color, and clear conuences of side branches, are
observed. Fine capillary vessels are clearly visualized on the surface
of the pancreatic duct (Fig. 20.9). In cases with chronic pancreatitis,
protein plugs and calcied stones are clearly observed in the main
pancreatic duct (Figs 20.10, 20.11A, 20.11B). On the mucosal surface
of the pancreatic duct, whitish, rough, scar-like, or erythematous
mucosa is observed. Fine capillary vessels on the surface of the
pancreatic duct are frequently blurred (Figs 20.12A, 12B).
Smooth ductal stenoses with scar formation are also observed
(Figs 20.13A, 20.13B). In cases with advanced cancers, friable mucosa
with erythema and erosive changes around the stenosis (Fig. 20.14)
are observed as diagnostic ndings of pancreatic cancer, while in
some cases a compressed pancreatic duct wall covered with normal
epithelium is observed. In IPMT cases, the characteristic nding of
papillary tumors reported by other investigators is visualized with
extreme clarity as are mucosal excrescences resembling clustered
salmon eggs. Endoscopic ndings of IPMT are further discussed in
the Indications section of this chapter.
Fig. 20.10 Endoscopic image of the pancreatic duct of a patient
with chronic pancreatitis obtained with PEPS system. Fine protein
plugs within the pancreatic duct. Reprinted from Gastroenterol
Endosc, 44, Kodama T, Present and future of the peroral electronic
pancreatoscope (PEPS), 310, 2002 with permission from Japan
Gastroenterological Endoscopy Society and Elsevier.
Figs. 20.11A, B Endoscopic images of the
pancreatic duct of a patient with chronic
pancreatitis obtained with PEPS system. A
white stone within the pancreatic duct. The
ne detail of the texture of the stone is
clearly seen. Reprinted from Gastrointest
Endosc, 49, Kodama T et al., Pancreatoscopy
for the next generation: development of the
peroral electronic pancreatoscope system,
366371, 1999 with permission from Ameri-
can Society for Gastrointestinal Endoscopy
and Elsevier.
205
 SECTION 2 TECHNIQUES
B pancreatic duct in a patient with chronic pan-
A
B creatic duct of a patient with chronic pancre-
A image. B Smooth stenosis of the main
Fig. 20.14 Endoscopic image of the pancreatic duct of a patient
with pancreatic cancer obtained with PEPS system. Friable mucosa
with erythema and erosive changes around the stenosis of the
pancreatic duct clearly visualized. Reprinted from Gastroenterol
Endosc, 44, Kodama T, Present and future of the peroral electronic
pancreatoscope (PEPS), 310, 2002 with permission from Japan
Gastroenterological Endoscopy Society and Elsevier.
206
Figs. 20.12A, B Endoscopic images of the
creatitis obtained with PEPS system. Vague
network of ne vessels on the rough-sur-
faced pancreatic duct. Reprinted from Gas-
trointest Endosc, 49, Kodama T et al.,
Pancreatoscopy for the next generation:
development of the peroral electronic pan-
creatoscope system, 366371, 1999 with
permission from American Society for Gas-
trointestinal Endoscopy and Elsevier.
Fig. 20.13 Endoscopic image of the pan-
atitis obtained with PEPS system.
A Remarkable stenosis in the main pancre-
atic duct was observed in ERCP
pancreatic duct was noted with the PEPS.
INDICATIONS
Necessary indications
Major indications for direct pancreatoscopy include lling defects
on ERCP of uncertain etiology and duct cut-off or stricture of
uncertain origin. Direct pancreatoscopy can improve the diagnosis
of IPMT from its characteristic endoscopic ndings. Direct pancre-
atoscopy can also differentiate pancreatic cancers from chronic
pancreatitis if the cancerous lesion is within the main pancreatic
duct. However, nal diagnosis should be assured by virtue of
directed histology or cytology. Intraoperative pancreatoscopy is very
useful in some cases of IPMT to determine the resection line of
the pancreas, although frozen section of the margin remains man-
datory. The high quality image of PEPS has the potential to improve
diagnostic accuracy of IPMT or pancreatic cancer. It also has the
potential to improve visual detection of tumor invasion within the
pancreatic duct.
Intraductal papillary mucinous tumor (IPMT)
The investigation of possible IPMT cases is one of the best indica-
tions for pancreatoscopy. IPMT has been recently recognized as a
unique pancreatic tumor with an indolent biologic behavior and
 Chapter 20 Pancreatoscopy

favorable prognosis. The tumor spreads along the pancreatic duct

A B
replacing the normal epithelium and includes a broad spectrum of
histopathologic disorders such as hyperplasia, adenoma, and adeno-
carcinoma. The possible diagnosis of IPMT is usually established by
ERCP or MRCP ndings.14 They include dilated, mucus-lled papilla,
lling defects in the main pancreatic duct, and dilated main and
branch pancreatic ducts in the absence of remarkable obstructing
ductal strictures. CT and transabdominal or endoscopic ultrasonog-
raphy are also useful for detecting cystic lesions and tumors within
the cysts.15 However, denite diagnosis of IPMT is possible if the
characteristic appearance of papillary tumors are observed during
pancreatoscopy. A biopsy to differentiate various kinds of histology
can be taken from lesions under direct vision when a pancreatoscope
with an accessory channel is used. Fig. 20.15 Endoscopic images of the pancreatic duct of a patient
with intraductal papillary mucinous tumors (IPMT) obtained with
Even without biopsy, pancreatoscopy has been reported to be PEPS system. Various shapes of tumors were observed inside the
useful in the differentiation of benign IPMT of the pancreas from dilated pancreatic duct. This case proved to be adenocarcinoma
more dysplastic lesions. It is often combined with intraductal ultra- histologically. A numerous tumors with taller villous projections
and dilatation of capillary vessels. B granular or nodular tumors.
sonography (IDUS) which can measure the size of papillary lesions Reprinted from Gastrointest Endosc, 52, Koshitani T et al., Clinical
and dene potential invasion. Hara, et al. classied the endoscopic application of the peroral electronic pancreatoscope for the inves-
ndings of IPMT into 5 groups and described a sh-egg-like type tigation of intraductal mucin-hypersecreting neoplasm, 9599,
with prominent vessels, villous type and vegetative type, as often 2000 with permission from American Society for Gastrointestinal
Endoscopy and Elsevier.
malignant.16,17 Pancreatoscopy also provides valuable information in
assessing the extent of the lesion and multicentric lesions. It is
helpful at selecting the best surgical procedure in IPMT. Kaneko,
et al.18,19 reported that intraoperative pancreatoscopy of IPMT was
useful in determining the surgical resection line.
These ndings were more clearly conrmed by investigators
using PEPS.20 With this instrument the characteristic appearance of
papillary tumors can be visualized with unsurpassed clarity, even
though the distance between each element of the tumor and the
distal end of the endoscope varied. PEPS made it possible to make
a more detailed morphologic assessment of the tumor and its grade
of malignancy. In the case of adenocarcinoma, PEPS revealed a
variety of tumors, from granular or nodular tumors to higher villous
tumors with dilatation of capillary vessels (Figs 20.15A, 20.15B).
Papillary tumors resembling salmon eggs with reddish inner color
(venous dilatation) were also observed with adenocarcinoma (Fig.
20.16), while papillary tumors resembling salmon eggs with whitish
inner color were recognized in adenomas (Fig. 20.17A). These nd-
ings supported the results of the previous studies concerning papil-
lary adenoma or adenocarcinoma in IPMT.
PEPS information on the extent of the intraductal growth of
the tumor was also reported to be better than conventional pancre-
atoscopy.20 PEPS clearly visualized even small papillary projections
near larger tumors, and the border of the lesion with the normal
mucosa was well identied (Figs 20.17A, 20.17B). In a single case
of branch-duct type of IPMT, where the main tumor existed in a
dilated branch duct off the main pancreatic duct, the PEPS visual-
ized papillary tumors spreading from the orice of the dilated branch
duct.
Fig. 20.16 Endoscopic images of the pancreatic duct of a patient
Differentiation of stenosis of the main pancreatic duct with intraductal papillary mucinous tumors (IPMT) obtained with
(benign or malignant) PEPS system. Papillary tumors resembling salmon eggs with dilata-
tion of capillary vessels. This case was proved to be adenocarci-
From the initial prototype of pancreatoscopy, many trials have been noma. Reprinted from Gastroenterol Endosc, 44, Kodama T, Present
performed to distinguish focal or chronic pancreatitis from pancre- and future of the peroral electronic pancreatoscope (PEPS), 310,
atic cancer. Many investigators have described smooth stenoses with 2002 with permission from Japan Gastroenterological Endoscopy
or without scar formation or mucosal edema observed in chronic Society and Elsevier.
pancreatitis, while lesions with friable erythematous mucosa and
erosive changes are more common in pancreatic cancer. Our PEPS
ndings of pseudotumorous pancreatitis that was differentiated

207
 SECTION 2 TECHNIQUES
A creatic duct of a patient with intraductal pap-
B
from pancreatic cancer is supportive of this consensus.21 However,
smooth stenoses without friable erythematous mucosa and erosive
changes do not always imply a benign lesion. Yamao et al.6 reported
that the sensitivity and specicity of coarse mucosa or friability for
pancreatic cancer were 59% and 88% (coarse mucosa), and 50% and
100% (friability), respectively, in 38 cases that were followed up for
more than 2 years. A compressed pancreatic duct wall covered with
normal epithelium can be observed with an extrinsic malignancy
that does not involve the epithelium of the pancreatic duct at the
distal site of the stenosis. Miyakawa et al.22 classied endoscopic
cancerous changes into two groups: supercial and compressed.
They reported that malignant cells were detected histopathologi-
cally in 41% of the supercial type and in none of the compressed
type, based on transpapillary biopsy or brush cytology. Although
pancreatic duct compression by an extrinsic pancreatic cancer is
steeper than compression in chronic pancreatitis, the necessity of
surgical treatment cannot be decided from the pancreatoscopic
ndings alone. Clinical diagnosis of pancreatic cancer should be
carefully considered with other multiple imaging modalities if biopsy
or cytology is negative.
Appropriate indications
Using an ultrathin pancreatoscope, the usefulness of direct pancre-
atoscopy for the early detection of pancreatic cancer has been
reported in small series. Detection of pancreatic cancer using PEPS
is being investigated,23 looking for early stages of pancreatic
cancer.
Direct pancreatoscopy with PEPS image quality may be consid-
ered in patients who are suspected of having chronic pancreatitis,
equivocal by other imaging modalities including endoscopic ultra-
sound in this setting. Pancreatoscopic images obtained by PEPS
could suggest the clinical diagnosis of chronic pancreatitis from its
characteristic endoscopic ndings.
Early detection of pancreatic cancer
Because most pancreatic cancers are derived from pancreatic duct
epithelium, pancreatoscopy can contribute to the diagnosis of early
pancreatic cancer if located within the main pancreatic duct.
However, reports of in situ carcinoma of the pancreas found by
pancreatoscopy are very rare except for that associated with IPMT.
208
Fig. 20.17 Endoscopic images of the pan-
illary mucinous tumors (IPMT) obtained with
PEPS system. A Papillary tumors resem-
bling salmon eggs with whitish inner
color. B Small papillary projections near
larger tumors. Reprinted from Gastrointest
Endosc, 52, Koshitani T et al., Clinical applica-
tion of the peroral electronic pancreato-
scope for the investigation of intraductal
mucin-hypersecreting neoplasm, 9599,
2000 with permission from American
Society for Gastrointestinal Endoscopy and
Elsevier.
This is partly because the detection of early pancreatic cancer by
pancreatoscopy is a limited method for advanced centers of endo-
scopic excellence. Moreover, it is uncertain how frequently ductal
adenocarcinoma begins its growth in the main pancreatic duct.
Uehara et al.9 reported diagnosis of in situ carcinoma of the pancreas
using the ultrathin peroral pancreatoscope and pancreatoscopic
cytology. Out of 11 cases of in situ carcinoma diagnosed in the surgi-
cally resected specimen, they observed endoscopic ndings of irreg-
ular, nodular or papillary mucosa of the main pancreatic duct in 10
cases. They also collected pancreatic juice from abnormal sites of
the pancreatic duct seen by peroral pancreatoscopy through the
narrow lumen between the scope and the guide catheter. They diag-
nosed all the 10 cases as carcinoma by this pancreatoscopic cytology.
They concluded that pancreatoscopic cytology was useful for locating
and diagnosing in situ carcinoma when compared to pancreatic juice
cytology obtained by catheter at the time of ERCP. However a more
comprehensive study in high-risk patients could reveal the sensitiv-
ity and specicity for diagnosing minute pancreatic cancer with
pancreatoscopy in the future.
Further investigation of chronic pancreatitis
Endoscopic ndings of protein plugs and calcied stones oating in
turbid pancreatic juice or scar formation on the mucosa have been
reported by many investigators in patients with chronic pancreati-
tis.24 These ndings were further studied with PEPS in 36 patients
with chronic pancreatitis who were classied as having equivocal to
marked ductographic changes by ERP according to the Cambridge
criteria.25 With the high quality image of PEPS, these plugs consisted
of ne granular or thread-like protein oating in the lumen or
coating the epithelium on occasion. In equivocal cases, ductal con-
tents were turbid and calculi co-existing with protein plugs appeared
to be relatively soft, while in advanced chronic pancreatitis calculi
had a rough surface and were generally calcied. Although the vas-
cular markings in the pancreatic ductal mucosa were clearly observed
in normal patients, those in chronic pancreatitis were generally
indistinct. However, the vascular markings tended to be visible again
in advanced stages of chronic pancreatitis. Normal patients had
smooth capillary reticulation markings, whereas changes in the
visible vascular net such as disruption, stenosis, irregularity, rear-
rangement and stretching were observed in chronic pancreatitis.

Whitish mucosa co-existing with a rough surface and scar formation
of the ductal wall were observed in advanced subjects with vascular
changes, while the vascular markings had disappeared and edema-
tous white mucosa was observed in equivocal cases. From these
ndings, PEPS images appear to be particularly helpful with a
patient complaining of symptoms consistent with chronic pancreati-
tis, whose ERCP image is equivocal relative to the Cambridge criteria
of chronic pancreatitis. Further study may establish new diagnostic
criteria for chronic pancreatitis by PEPS images.26
Inappropriate indications
When the lesion is limited to a side branch of the pancreatic duct,
direct pancreatoscopy cannot reach the lesion. Alternative diagnostic
modalities in such a case are endoscopic or intraductal ultrasonog-
raphy (EUS or IDUS) combined with pancreatic juice cytology.27
Multiple molecular biological analyses of the pancreatic juice, includ-
ing K-ras oncogene mutations and telomerase activity, are being
investigated as an adjunct to cytology.28
COMPLICATIONS
Acute pancreatitis occurring with or without sphincterotomy is the
main complication of pancreatoscopy. Reported complications of
pancreatoscopy are relatively low (012%), partially because the pro-
cedure is usually undertaken at Advanced Centers of Endoscopic
Excellence. Moreover, patients with chronic pancreatitis are less
susceptible to endoscopic complications than patients with normal
pancreatic ducts.
Mild pancreatitis after sphincterotomy was reported by Riemann
et al.4 in 1 of 16 cases (6%) in one series using the 3.1 mm pancre-
atoscope. This pancreatitis rapidly improved with standard therapy.
In this study, the total complication rate of pancreatoscopy was
reported to be 2.6%, including 20 cases studied with an ultrathin
pancreatoscope which is probably less invasive. Yamao et al.,6 in
turn, reported that using the 3.4 mm pancreatoscope and ultrathin
pancreatoscope, mild pancreatitis (lasting 12 days) occurred in 4 of
33 patients (12%). In 3 of these 4 patients, transpapillary biopsy
specimens or cytology specimens (brushings) were obtained as well
as pancreatoscopic investigation.
With use of an ultrathin pancreatoscope, Tajiri et al.8 reported
that 2 of 52 patients (4%) developed acute pancreatitis. Clinical
symptoms and biochemical abnormalities improved completely
within 7 days. With use of 3.2 mm or 1.67 mm pancreatoscopes,
Hara et al.17 reported that mild to moderate pancreatitis occurred in
Imaging
CHF-BP30 Fiberoptic
PF-8P Fiberoptic
FCP-9P Fiberoptic
FCP-8P Fiberoptic
CHF-BP260 Field sequential
CHF-B260 Field sequential
Video Converter CV-140 Simultaneous
Processor CV-160 Simultaneous
Table 20.3 Comparison of prices of pancreatoscopes and processors in US marketsa
a
Prices are based on data, September 2005.
Chapter 20 Pancreatoscopy
4 of 60 patients (7%). All patients recovered with conservative treat-
ment. Using the XPF-22EY PEPS, 1 of 56 patients developed moder-
ate acute pancreatitis with abdominal pain and elevation of serum
amylase after the procedure. Pancreatitis improved with standard
therapy, thus giving a total complication rate of 1.8%.13 With the
enhanced version, XCHF-BP240, our preliminary experiences in 9
cases, including 2 cases of cholangioscopy, revealed no signicant
complications.11
Management of complications
To avoid possible complications, patients are usually hospitalized for
one day after the procedure and receive peri-procedural intravenous
antibiotics. Drugs (gabexate mesilate, nafamostat mesilate, ulina-
statin) which inhibit the activation of pancreatic enzymes are also
used in Japan to minimize procedurally related pancreatitis.13 A
pancreatic duct stent is sometimes placed following endoscopic pro-
cedure to prevent acute pancreatitis, especially after sphincterotomy.
Standard therapy of intravenous antibiotics and pancreatic enzyme
inhibitors is usually enough to treat the mild acute pancreatitis
which occasionally occurs after a pancreatoscopy procedure.
RELATIVE COST
Pancreatoscopes are relatively expensive instruments for special
usage in the pancreatobiliary area. Comparison of prices for pancre-
atoscopes, video converter and the image processor in US markets
are summarized in Table 20.3. CHF-BP30, FCP-9P, FCP-8P, CHF-
BP260 and CHF-B260 need a therapeutic duodenoscope (channel
diameter 4.2 mm) as a mother scope, whereas PF-8P can be used
through an ERCP catheter and can be used with a diagnostic
duodenoscope.
Since all pancreatoscopes are fragile, they require careful inser-
tion procedures. Inspection of the scope before and after the proce-
dure should be performed to nd minor damage of the scope as
early as possible. In the case of electronic pancreatoscopes, the
deection part is covered with only one layer of very thin rubber to
make the diameter small. This is the most fragile part which needs
to be checked after each use. It is also important to check that this
fragile part is entirely out of the accessory channel of the mother
scope when the elevator of the duodenoscope is lifted (Fig. 20.2e).
In our experience, the XPF-22EY and XCHF-BP240 needed repairs
of the deection part after about 5 procedures. Even with the most
careful procedure, signicant maintenance costs can be expected for
all miniscopes.
Diameter Channel Price
(mm) (mm) (US dollars)
3.1 1.2 20,900
0.8 None not available
3.0 1.2 23,700
2.7 0.75 23,700
2.6 0.5 not available
3.4 1.2 not available
9,500
19,700
209
 SECTION 2 TECHNIQUES
CONCLUSION
Although the problem of suboptimal visualization has nally been
resolved by development of electronic pancreatoscopes, other prob-
lems such as instrument fragility and relative large diameter remain
REFERENCES
1. Takekoshi T, Maruyama M, Sugiyama N. Retrograde pancreatic
cholangioscopy [in Japanese]. Gastroenterol Endosc 1975;
17:678683.
2. Ohashi K, Murakami Y, Maruyama M, et al. Four cases of mucous
secreting cancer of the pancreas on specic ndings of the
papilla Vater [in Japanese]. Prog Dig Endosc 1982; 20:348351.
3. Technology Committee of American Society for Gastrointestinal
Endoscopy. Technology status evaluation: duodenoscope-assisted
cholangiopancreatoscopy. Gastrointest Endosc 1999; 50:
943945.
4. Riemann JF, Kohler B. Endoscopy of the pancreatic duct: value
of different endoscope types. Gastrointest Endosc 1993;
39:367370.
5. Junk M, Zipf A, Schoonbroodt D et al. Is pancreatoscopy of any
benet in clarifying the diagnosis of pancreatic duct lesions?
Endoscopy 1998; 30:273280.
6. Yamao K, Ohashi K, Nakamura T et al. Efcacy of peroral
pancreatoscope in the diagnosis of pancreatic diseases.
Gastrointest Endosc 2003; 57:205209.
7. zkan H, Saisho H, Yamaguchi T et al. Clinical usefulness of a new
miniscope in the diagnosis of pancreatic disease. Gastrointest
Endosc 1995; 42:480485.
8. Tajiri H, Kobayashi M, Niwa H, et al. Clinical application of an ultra-
thin pancreatoscope using a sequential video converter.
Gastrointest Endosc 1993; 39:371374.
9. Uehara H, Nakaizumi A, Tatsuta M. Diagnosis of carcinoma in situ
of the pancreas by peroral pancreatoscopy and pancreatoscopic
cytology. Cancer 1997; 79:454461.
10. Kodama T, Sato H, Horii Y, et al. Pancreatoscopy for the next
generation: development of the peroral electronic
pancreatoscope system. Gastrointest Endosc 1999; 49:366371.
11. Kodama T, Tatsumi Y, Sato H, et al. Initial experience with a new
peroral electronic pancreatoscope with an accessory channel.
Gastrointest Endosc 2004; 59:895900.
12. Technology Committee of American Society for Gastrointestinal
Endoscopy. Technology status evaluation: ultrathin endoscopes
esophagogastroduodenoscopy. Gastrointest Endosc 2000;
51:786789.
13. Kodama T, Koshitani T, Sato H, et al. Electronic pancreatoscopy
for the diagnosis of pancreatic diseases. Am J Gastroenterol 2002;
97:617622.
14. Telford JJ, Carr-Locke DL. The role of ERCP and pancreatoscopy in
cystic and intraductal tumors. Gastrointest Endosc Clin N Am
2002; 12:747757.
210
problematic in the search for the ideal scope.23,29 Continued research
into the design of the pancreatoscope, which could improve durabil-
ity and success rates, may make this technology more popular in the
future.
15. Yamao K, Okubo K, Sawaka A, et al. Endoluminal ultrasonography
in the diagnosis of pancreatic diseases. Abdom Imaging 2003;
28:545555.
16. Yamaguchi T, Hara T, Tsuyuguchi T, et al. Peroral pancreatoscopy
in the diagnosis of mucin-producing tumors of the pancreas.
Gastrointest Endosc. 2000; 52:6773.
17. Hara T, Yamaguchi T, Ishihara T, et al. Diagnosis and patient
management of intraductal papillary-mucinous tumor of the
pancreas by using peroral pancreatoscopy and intraductal
ultrasonography. Gastroenterology 2002; 122:3443.
18. Kaneko T, Nakao A, Nomoto S, et al. Intraoperative
pancreatoscopy with the ultrathin pancreatoscope for mucin-
producing tumors of the pancreas. Arch Surg 1998; 133:263267.
19. Kanazumi N, Nakao A, Kaneko T, et al. Surgical treatment of
intraductal papillary-mucinous tumors of the pancreas.
Hepatogastroenterology 2001; 48:967971.
20. Koshitani T, Kodama T, Sato H, et al. Clinical application of the
peroral electronic pancreatoscope for the investigation of
intraductal mucin-hypersecreting neoplasm. Gastrointest Endosc
2000; 52:9599.
21. Kodama T, Abe M, Sato H, et al. A case of pseudotumorous
pancreatitis that presented unique pancreatoscopic ndings with
the peroral electronic pancreatoscope. J Gastroenterol Hepatol
2003; 18:108111.
22. Miyakawa H, Suga T, Murashima Y, et al. The role of peroral
pancreatoscopy. [in Japanese]. Endosc Dig 1993; 5:943948.
23. Kozarek RA, Kodama T, Tatsumi Y. Direct cholangioscopy and
pancreatoscopy. Gastrointest Endosc Clin N Am 2003;
13:593607.
24. Kozarek RA. Direct pancreatoscopy in chronic pancreatitis. Dig
Endosc 1990; 2:15.
25. Kodama T, Imamura Y, Satoh H, et al. Feasibility study using a
new small electronic pancreatoscope: description of ndings in
chronic pancreatitis. Endoscopy 2003; 35:305310.
26. DiMagno MJ, DiMagno EP. Chronic pancreatitis. Curr Opin
Gastroenterol 2003; 19:451457.
27. Fujita N, Noda Y, Kobayashi G, et al. Endoscopic approach to
early diagnosis of pancreatic cancer. Pancreas 2004; 28:279281.
28. Myung SJ, Kim MH, Kim YS, et al. Telomerase activity in pure
pancreatic juice for diagnosis of pancreatic cancer may be
complementary to K-ras mutation. Gastrointest Endosc 2000;
51:708713.
29. Kodama T, Tatsumi Y, Kozarek RA, et al. Direct pancreatoscopy.
Endoscopy 2002; 34:653660.
 SECTION 2 TECHNIQUES
Chapter
Cholangioscopy
21 Peter B. Kelsey
INTRODUCTION
The prospect of visualizing the biliary tree during ERCP has allured
gastroenterologists for decades.1 Signicant mechanical challenges
have resulted in a design evolution of both rigid and exible endo-
scopes employing beroptic and video technology. The nomencla-
ture has likewise evolved through a spectrum of descriptive terms
including cholangioscopy, cholangioscopy, duodenoscope-assisted
cholangiopancreatoscopy, and peroral cholangioscopy. The duct can
now be accessed through a variety of approaches: percutaneously
through a transhepatic route, through a choledochotomy, or the
cystic duct created intraoperatively, or through the papilla of Vater
via a duodenoscope. This chapter will specically review the duode-
noscope assisted approach to cholangioscopy. In this two endoscope
system, the supporting duodenoscope and the cholangioscope are
often referred to as the mother and baby scope, respectively.
DESCRIPTION OF THE TECHNIQUE
Preprocedure room set-up
If cholangioscopy is to become a truly meaningful adjunct for the
interventionalist, the equipment needs to be readily available as the
indications arise. A giant stone may be encountered that is simply
too large to fragment by standard techniques or a stricture suspi-
cious for malignancy may require direct visualization for tissue
sampling. Under these circumstances, the ability to perform
cholangioscopy as an adjunct to ERCP may both clarify the
diagnosis, and facilitate appropriate therapy. The result may both
optimize patient care quality and minimize the need for
reintervention.
For many years, prototype cholangioscopes have been sited in
various academic institutions around the world. Currently, in the
United States, the limited assortment of cholangioscopes available
for purchase is listed in Table 21.1. In general, the smaller the outer
diameter of the cholangioscope, the greater the maneuverability
within the bile duct. The smaller size, however, leaves less room for
important features such as a sufciently large working channel.
The optimal ERCP suite design places the cholangioscope, its
power supply, and accessories in close proximity to the endoscopist.
Because of the intimate spatial working relationships between the
mother and baby scope, the cholangioscope components are set up
on or adjacent to the ERCP processor cart. These components
include the light generator and image processor as well as the ush-
ing and suctioning equipment. Depending on the manufacturer and
the model, the light source, image processing hardware and air/uid
pump are available either as individual components or combined in
a single unit (Fig. 21.1).
Likewise strategic placement of the monitors is critical to permit
simultaneous viewing of the video images from the mother scope,
the baby scope and uoroscopic unit during cholangioscopy. Ideally,
the three monitors are clustered and mounted at eye level in front
of the endoscopist. In new unit designs, these three monitors are
mounted on articulating arms to maximize the operators comfort
and ergonomics. Toggle switches or footpedals permit easy switch-
ing between other video output sources such as EUS, or a micros-
copy unit. Extra monitors are positioned for the assisting staff.
Digital recording of the cholangioscopic exam provides a permanent
record of the exam and most importantly permits post-procedure
review of the ndings (Box 21.1).
A variety of accessories are used during routine cholangioscopy.
A special adapter is attached to the opening of the mother scopes
instrument channel through which the cholangioscope is passed.
This adapter is designed to prevent crimping of the baby scope as it
is maneuvered during the procedure. As copious ushing and suc-
tioning is routine during cholangioscopy, the saline irrigant, connec-
tor tubing, stopcocks, and syringes need to be available. Some
processors are equipped with irrigation and suctioning components.
Finally, there are a variety of accessories specic to cholangioscopy
such as cytology brushes, biopsies forceps, snares, and electrohy-
draulic lithotripsy accessories that need to be organized and readily
available.
TECHNIQUE: DIAGNOSTIC
The step up from basic interventional ERCP to cholangioscopy
requires the development of several new skills. One major challenge
of cholangioscopy is the coordination and handling of the two-
endoscope system using either the single or the two operator
technique. The two person technique requires two experienced
endoscopists to be present during the exam, one to handle the mother
scope and the other to operate the baby scope with all of its accesso-
ries. Alternatively, the single operator technique requires a single
interventionalist who manages both endoscopes. The endoscopists
left hand is used to control the mother duodenoscope scope. The baby
scope is secured to the endoscopist by a breast plate (see Fig. 21.2).
The endoscopists right hand is then free to manage the controls
and accessories of both the baby scope as well as the mother scope.
The single operator technique has several advantages over the dual
operator technique. First, cholangioscopy requires a carefully cho-
reographed coordination of movement between the two endoscopes
to both successfully maneuver the baby scope and to minimize the
likelihood of damage due to crimping at its insertion into the mother
scope. This coordination is more easily performed by a single person
than by two trying to synchronize their motions. Second, obligating
two physicians to be present during an entire exam is an inefcient
use of manpower, especially when the exam can be performed
211
 SECTION 2 TECHNIQUES

Feature Olympus Pentax Pentax

Model CHF BP 30 FCP9N FCP8P
Working Channel 1.2 mm 1.2 0.75
Guidewire size 0.035 in 0.035 0.025
Outer Diameter 3.4 mm 3.1 2.8
Mother Scope 4.2 mm 4.2 3.8
Channel Diameter
Field of View 90 90 90
Depth of Field 1-50 1-50 1-50
Tip Deection 160 up/ 130 down 160/130 160/130
Working Length 187 cm 190 190

Table 21.1 Cholangioscopes available in the US

Fig. 21.2 The baby scope secured to the endoscopist by a breast

plate.

mother scope and can be immobilized to a breastplate or waistband

strapped to the endoscopist.

CANNULATION
There are three considerations when cannulating the ampulla; the
need for a papillotomy, the need for guidewire assistance, and the
actual maneuvering required to advance the baby scope up the bile
duct without injury to the instrument or the patient.
First, in almost all cases, the presence of a papillotomy greatly
facilitates the ease of the exam. In many instances, however, a papil-
lotomy will have already been performed before the decision has
been made to refer the patient for cholangioscopy. The patient with
Fig. 21.1 Cholangioscopy set-up for routine use. giant, recalcitrant choledocholithiasis has often already undergone
one or more failed prior attempts during which a papillotomy has
been performed to either remove smaller debris or to permit the use
of balloons or lithotripsy catheters. Cholangioscopy has been suc-
BOX 21.1 ROOM SET-UP REQUIREMENTS cessfully performed following balloon ampullary dilation without
papillotomy.2 Likewise the development of the ultrathin caliber
FOR CHOLANGIOSCOPY endoscopes permits scope passage through supporting catheters
without the need for a sphincterotomy.3 These ultrathin scopes, still
Cholangioscope not widely used, do not have an instrument channel for tissue sam-
Light source and video Processor pling or for interventions. The papilla can, however, be dilated using
Breastplate (for single operator technique) either catheter or balloon dilators to permit passage of the larger
Accessories cholangioscopes over a guidewire into the biliary tree. Balloon dila-
Cytology brush tion of the papilla carries a dened risk of pancreatitis.
Biopsy forceps Cannulation can be performed either over a guidewire or free
Electrohydraulic Lithotripsy hand. The guidewire technique is recommended until the endosco-
EHL power source pist becomes experienced with cholangioscopy. Once the guidewire
EHL probes is passed up the bile duct, the cholangioscope is back-loaded over
Monitors (2) for cholangioscopic image the wire using caution not to damage the channel as it angles and
exits near the scope handle. A straw or other similar device can be
used to intercept the advancing wire to safely deect the wire tip out
of the scope. The wire now acts as a rail over which the cholangio-
efciently by one physician alone. Finally, since the cholangioscope scope can be passed out from the mother scope and up through the
functions more as a catheter than as a true endoscope, the baby papillotomy. As the cholangioscope approaches the papilla, the assis-
scope handle remains essentially motionless during the cholangios- tant provides gentle traction on the guidewire to pull the baby scope
copy. Most of the steering and maneuvering of the baby scope actu- upward and into the papillary orice. This upward deection mini-
ally comes from the mother. Since the baby scope is not designed mizes the need to raise the elevator. It is the lifting up of the elevator
to tolerate torque, its position should remain xed relative to the that causes the most damage to these fragile scopes either by tearing

212

the bending rubber at the distal end of the insertion tube or by actu-
ally crimping the insertion tube. Once the baby scope tip penetrates
the ampullary orice, it can be further advanced up the bile duct by
gently tugging back on the mother scope. There is often a soft pop
as the baby scope moves up the biliary tree. The baby scope can then
be advanced up the bile duct in one to two centimeter increments
followed by a gentle lifting of the elevator. This sequence of motions,
similar to the technique used to advance a biliary stent, is repeated
until the scope is in a desired and secured position.
The free hand cannulation technique can be accomplished, in
most cases, by rst positioning the mother scope close to the
ampulla. The baby scope tip is then advanced through the mother
scope elevator. With the tip turned upward, and using the mother
scope elevator, the baby scope is lifted up to the papillotomy site.
The large dial of the mother scope is turned fully towards the endos-
copist which lifts the tip of the mother scope upward and introduces
the baby scope into the papilla. Once the tip of the baby scope is fully
committed into the papillary orice, a gentle tugging back on the
mother scope, pulling it slightly outward from the patients mouth,
will often pop the baby scope up through the papillary region and
into the free biliary lumen.
Fluoroscopy is employed during cannulation to observe both the
angle of the cholangioscopes deecting tip and the trajectory of its
insertion tube as it passes up the bile duct. The cholangioscopes
deecting tip should not knuckle or bow as it heads up the bile duct.
Ideally, the trajectory of the tip and the insertion tube should be
straight and should point towards the bifurcation. This will lessen
the likelihood of inadvertent cannulation of the pancreatic duct.
Fluoroscopic observation is also used to guide the depth of the
cholangioscopes insertion.
Preparing the bile duct for observation
With the cholangioscope positioned in the biliary system and the
guidewire removed, the lumen of the bile duct comes into view. This
view is often obscured by bile, mucus, and debris. Clearing the duct
of this debris is necessary for an accurate diagnostic exam and
usually requires several minutes of irrigation using a saline lavage.
Irrigation with volumes from 5 to 25 cc saline can be used with each
ush, depending on the duct diameter. Because of the small caliber
of the cholangioscopes working channel, the viscosity of bile, and
the frequent presence of small particles of stone and other debris,
suctioning of the irrigant from the duct is a tedious but necessary
process. Lavaging the bile duct until it is clear facilitates optimal
visualization and is necessary prior to any therapeutic maneuvers.
During this cleaning process, uid and debris tend to accumulate
in the intestinal lumen and track up into the stomach. Comparing
the amount of lavage uid instilled to that amount suctioned will
help minimize the aspiration risk. When this difference approaches
100 cc, the endoscopist might consider removal of the baby scope
and withdrawing the mother scope into the stomach to aspirate the
uid that has accumulated there. Constant suctioning of this uid
from the duodenum should be performed through the mother scope
during the procedure. This is a counter-intuitive maneuver for most
experienced interventionalists who are trained to constantly insuf-
ate the duodenal lumen with air to maintain good luminal
visualization.
Maneuvering the cholangioscope
Once positioned in the bile duct and with the guidewire removed,
inspection can begin. Optimal visualization occurs under an aqueous
Chapter 21 Cholangioscopy
medium and thus saline is constantly pulsed or infused during the
procedure. Sterile water can likewise be used safely.4 The amount of
irrigant used should be sufcient to permit clear visualization care-
fully avoiding high pressure injections and excess accumulation of
irrigant in the intestinal lumen. The cholangioscope is advanced up
the bile duct with the same technique used to advance a biliary stent.
With the elevator lowered, the cholangioscope is advanced 23 cm.
The elevator is raised slightly and the endoscopists left thumb turns
the mother scopes large dial towards the endoscopist lifting up the
tip of the mother scope towards the ampulla, thus pushing the baby
scope up the bile duct. The elevator is then lowered, the large dial
turned away, and the process is repeated.
The challenge at this point is to keep the baby scope centered in
the duct lumen, permitting good visualization. The baby scope tends
to slide along the bile duct wall, smearing the mucosa across the
viewing lens, thus obscuring the view. The crux of cholangioscopy
is the challenge of maintaining a centered position of the baby scope
within the bile duct. The insertion tube of the baby scope is not
designed to respond to torque, and thus little control actually comes
from manipulating the baby scope itself. Most of the baby scope
control comes from the handling of the mother scope.
Several principles may help the beginner cholangioscopist. Once
the cholangioscope is maneuvered up the bile duct, ne tuning of
its position lengthwise in the duct is best achieved by slight move-
ments of the mother scope up and down the duodenum past the
ampulla. These position changes will similarly move the cholangio-
scope up and down the bile duct. Lateral movements within the bile
duct are more difcult. Rolling and torquing the mother scope and
use of the duodenoscopes small dial will translate into some baby
scope lateral movement. Pulling or pushing on the mother scope
will affect how straight or curved the baby scope lies within the bile
duct and will alter its orientation. The baby scope tip deection dial
offers additional positioning changes.
INDICATIONS FOR CHOLANGIOSCOPY
The primary indications for cholangioscopy include the evaluation
and management of biliary strictures, lling defects, and difcult
choledocholithiasis (Table 21.2).
Evaluation of bile duct lesions
Conventional ERCP has an excellent track record in the diagnosis
and management of well-dened bile duct abnormalities such as
choledocholithiasis and bile leaks. This technique has fared less
well in the accurate diagnosis of two, not so well-dened classes
of bile duct lesions: biliary strictures and biliary lling defects.
Though there is some overlap between these two ndings, it may
be useful to consider them separately. A lling defect relates to
ERCP Diagnosis Cholangioscopy offers
Strictures Improved diagnostic accuracy
Tissue acquisition under visualization
Filling defects Improved diagnostic accuracy
Denitive therapy
Stones, refractory EHL
Duct clearance
Table 21.2 Indications for cholangioscopy
213
 SECTION 2 TECHNIQUES

Fig. 21.3 Intraluminal lling defects due to stones.

Fig. 21.4 Nodular lesion with friable mucosa.

the uoroscopic appearance of something that actually lies within
the bile duct lumen such as a stone or a polypoid tumor. Conversely,
a stricture implies a narrowing of the duct lumen due to either
thickening of the wall or compression from extrinsic pathology. The
increased wall thickening can be intrinsic to the wall such as with a
cholangiocarcinoma and might therefore have some associated
mucosal defects that could be detected by direct visualization. Com-
pression, on the other hand, due to extrinsic disease, such as nodal
metastasis may result in a narrowing of the lumen, but the epithelial
lining of the bile duct wall in the region of the stricture may retain
a normal appearance. Cholangiocarcinoma, for example, can spread
through the biliary tree in the submucosal layers and obstruct by
compression. The overlying mucosa, however, may have a com-
pletely unremarkable appearance.
It has now been clearly established that cholangioscopy can
improve the accuracy in the diagnosis of biliary lling defects.57 One Fig. 21.5 Papillary mucinous epithelium.
recent experience examined the impact of cholangioscopy on the
diagnostic accuracy of ERCP when there was uncertainty in the
ERCP diagnosis.8 Patients were excluded if they had obvious stone
disease or classic biliary obstruction due to malignancy in the head
of the pancreas. In this study, 91 consecutive patients were evaluated of bile duct adenocarcinoma is the inltrating type. This is the most
by ERCP supplemented by biopsy/brush cytology when indicated. difcult to diagnose because of the paucity of specic cholangio-
There were 76 strictures and 21 lling defects in the study group. scopic characteristics. The overlying mucosa is bland and whitish
Of the patients with the 21 lling defects, ERCP with biopsy or brush with minimal neovascularization. The authors described several
cytology was able to correctly identify the 8 malignant lesions and other less common bile duct tumors including biliary papillomato-
the 9 benign tumors. ERCP did not, however, correctly diagnose the sis, mucin-hypersecreting cholangiocarcinoma, and biliary cyst ade-
four cases of stone disease. In these patients, the stones were adher- nocarcinoma. Biliary papillomatosis looks similar to the papillary
ent to the bile duct wall and had the appearance of a mass. Cholan- adenocarcinoma except that it is a multifocal disease with areas of
gioscopy, on the other hand was able to make the correct diagnosis normal intervening mucosa. The mucin-hypersecreting cholangio-
using direct visualization alone in all 21 patients. The four patients carcinoma is similar to the papillary type adenocarcinoma except
with stone disease were easily identied and treated with stone that according to the authors, the biliary ducts may be dilated and
removal (Fig. 21.3). mucin lled.
While the cholangioscopic differentiation between stone and These cholangioscopic criteria for malignancy were tested pro-
tissue appears straightforward, the same is not true in the ability to spectively on 76 patients with biliary strictures of unknown type and
differentiate malignant from benign strictures on the basis of direct compared to the accuracy of ERCP with biopsy alone. In this study,
visualization alone. Early cholangioscopic experience reported mor- ERCP with tissue sampling had a sensitivity of 58% and a specicity
phologic characteristics that claimed to distinguish malignant from of 100% (Fig. 21.6). The addition of cholangioscopy to this group of
benign tissue with an accuracy that approached 95%.9 Several fea- patients did increase the sensitivity to100% but dropped the specic-
tures were identied as being accurate predictors of malignancy ity to 87%. The loss of specicity was due to the incorrect diagnosis
including tumor neovascularization, a dense papillary pattern, and of malignancy in 5 patients on the basis of the cholangioscopic
friable nodularity. It was observed that bile duct adenocarcinomas appearance of neovascularization and the presence of a tumor vessel.
had three patterns: nodular, papillary, and inltrative. Nodular These ve false positives were found instead to have chronic pan-
lesions were bulky and eccentric with an overlying friable mucosa creatitis in two patients and one patient each with primary sclerosing
(Fig. 21.4). There was often neovascularization with tissue friability cholangitis, autoimmune pancreatitis, and a peri-biliary cyst.
and oozing. The papillary type (Fig. 21.5) was identied by a high Another potential advantage of cholangioscopy in the evaluation
density pattern of papilla or sh egg type mucosa. There is often of biliary strictures is the opportunity to obtain pathologic material
luminal mucin, and blood obscuring the visual eld. The third type under direct observation. Brushings and biopsy forceps can be

214

Fig. 21.6 Tissue sampling under direct visualization.
steered directly onto the area of suspicion and the sample obtained
under visual guidance. Whether this approach actually increases the
yield of correct diagnosis has not been studied. Though it may seem
advantageous to obtain tissue in this fashion, there are several rec-
ognized difculties. First, is the issue of access. Negotiating the
angulation and turns of the bifurcation and the small biliary radicals
remains a challenge given the limitation of maneuverability and
steerability of the current cholangioscopes. Once in a tight space, it
may not be possible to deect the scope tip to obtain a specimen
from abnormal appearing tissue off to one side. And nally, the cups
on the biopsy forceps are quite small, approximately 1 mm in diam-
eter and thus the tissue yield is likewise small and frequently insuf-
cient when biopsying hard or brotic tissue.
Brush cytology, however, is often feasible. This can be performed
by removing the polyethylene sleeve from any inexpensive, dispos-
able brush, and passing the unsheathed brush through the cholan-
gioscopes channel. After obtaining the specimen, the brush can be
withdrawn 12 cm inside the baby scope. With the baby scope then
removed from the duodenoscope, the brush is advanced out of the
cholangioscopes tip and the specimen swiped on to frosted slides
as per routine.
Electrohydraulic lithotripsy
Fragmentation of giant or recalcitrant stones is one of the primary
indications for interventional cholangioscopy. Candidate stones are
usually too large to be trapped in a mechanical lithotripsy basket
or are adherent to the bile duct wall and thus can not be easily
manipulated. In these circumstances, fragmentation using electro-
hydraulic lithotripsy (EHL) is an efcient and highly successful
technique. Traditionally, endoscopists have been resigned to long-
term stenting of large, recalcitrant stones. It has been recently dem-
onstrated, however, that when compared to stone removal using
EHL, long-term stenting is associated with higher long-term com-
plications such as cholangitis,10 and thus most patients with such
stone burden should be considered for denitive fragmentation
therapy.
EHL was originally designed as an industrial mining tool. The
modication for endoscopy employs a ber with two embedded
electrodes. A power generator delivers a high voltage electrical
Chapter 21 Cholangioscopy
current creating a spark across the two electrodes at the tip of the
ber. High frequency discharges cause rapid expansion of the uid-
stone interface generating shock waves that fragment the stone. This
technique has been successfully applied using percutaneous, surgi-
cal and transampullary routes to the bile duct using either cholan-
gioscopic or uoroscopic guidance.
To perform EHL, the cholangioscope must be rst positioned in
front of the target stone. Achieving a satisfactory position is critical
to the safe deployment of the probe. When the EHL probe projects
from the cholangioscope, it must hit directly on the target stone and
not touch or travel adjacent to the biliary epithelium. In addition,
the contact interface between the probe and the stone must be in an
aqueous environment for the shock wave to be transmitted and
effect fragmentation. In patients with a capacious bile duct and a
generous papillotomy, the bile duct may drain too rapidly to perform
EHL. The patient must then be rolled to place the biliary tree in a
dependent orientation. Before ring, the probe tip should be in close
apposition to the stone. The duct is lavaged to sweep away frag-
mented debris to maintain good visualization. The probe should be
aimed at a single target on the stone, chipping away at a focus until
the stone cleaves. Often the outer coating of a stone is more durable
and requires more EHL pulses. Once chipped, continued ring at
the same spot rapidly enlarges the defect and fragments the stone.
As the target stone mass is reduced in size, the cholangioscope is
advanced up the duct to the next target. Fragmentation is then
repeated until all of the target stones are fragmented or the visual
eld is obscured by debris. It is common to overestimate the degree
of fragmentation that has occurred during a round of EHL dis-
charges. With the cholangioscope removed from the bile duct, the
stone fragments are swept out using standard balloon and basket
techniques. Often, by simply fragmenting the lower stones in a
packed duct, the remaining stones can then be more easily removed
by standard maneuvers. The second round of EHL, if needed, pro-
ceeds more quickly as the duct is less tightly packed and there is
more room to maneuver the cholangioscope. Once the bile duct has
been cleared of debris and stones, the baby scope can be reintro-
duced into the bile duct to check for large retained fragments and
for unsuspected strictures.
The success of EHL has been reported in a number of series.1113
In one large study, the power generator (Lithotripter Elgin, Il) was
congured at settings of 100 watts, a frequency of 6 shots per second
and 8 shots per pulse.13 In the management of recalcitrant stone
disease that has failed standard ERCP techniques, cholangioscopic
directed EHL is 90100% successful in complete stone eradica-
tion.14 While the number of procedures required to achieve this
success has ranged broadly from 1 to 13 exams, the experienced
cholangioscopists can expect to achieve complete duct clearance in
one exam of under 2 hours duration in the majority of patients.13
Cholangioscopy without uoroscopy
Occasionally, urgent biliary exploration is indicated when uoros-
copy is either unavailable or not practical. The ability to examine the
bile duct without uoroscopy has proven useful in three clinical situ-
ations: in morbidly obese patients whose features may not conform
to standard uoroscopy units, in patients during their rst trimester
of pregnancy and nally in those patients too critically ill to be trans-
ported to a uoroscopy unit.
The technique of cholangioscopy without uoroscopy requires
few modications from the technique described above. Routine
biliary cannulation is performed using a papillotome and a wire.
215
 SECTION 2 TECHNIQUES

Once a duct has been deeply cannulated, the guidewire is removed.

The observation of bile tracking up the lumen of the papillotome
as the guidewire is withdrawn conrms the position of the papillo-
tome in the bile duct. The guidewire is then advanced back up
through the papillotome into the biliary system. Sphincterotomy
is performed and the papillotome is removed leaving the wire in
the biliary tree. The cholangioscope is passed down through the
mother scope over the guidewire and a wire-guided cannulation of
the bile duct is performed. Once the baby scope is positioned in the
region of the bifurcation, the guidewire is removed. Bile and debris
can then be ushed from the duct allowing inspection of the biliary
tree.
Several important interventional manipulations of the biliary
tree can be performed without the use of uoroscopy. Small stones
can be removed under direct visualization using either a balloon or
a wire-guided basket. Large stones are usually stented, to be more
denitively managed at a later date when the patient has stabilized. Fig. 21.7 Cholangioscopy without uoroscopy.
Cholangioscopic directed EHL without the use of uoroscopic
assistance has not yet been reported, but may become feasible as
baby scope technology evolves. Mass lesions encountered during
cholangioscopy without uoroscopy can be brushed or biopsied
under direct visualization. Stenting of strictures or obstructing
lesions in the common hepatic duct or common bile duct requires
a modication from the standard stenting procedure. In determin-
ing the appropriate stent length, it is necessary to rst pass the CONTRAINDICATIONS TO CHOLANGIOSCOPY
baby scope above the level of the obstruction. A guidewire is passed
into the free space above the obstruction. The baby scope is slowly Cholangioscopy can be performed in most situations where ERCP
withdrawn down the length of the bile duct to the level of the is indicated. While there are no absolute contraindications to chol-
ampulla, carefully measuring this length as it is pulled out of the angioscopy, there are several noteworthy areas of caution. Coagulo-
instrument channel of the mother scope; this denes the appropri- pathic patients may not safely undergo sphincterotomy, thus
ate stent length. With the guidewire positioned above the stricture, preventing passage of many of the currently employed cholan-
the cholangioscope is removed, and the plastic biliary stent is gioscopes. The risk of bleeding due to EHL induced tissue
passed up the guidewire. Direct visualization of the guidewire is injury might be increased. In patients with ascending cholangitis,
lost as the stent passes out of the working channel of the mother the risk of inducing bacteremia during cholangioscopy might be
scope. Care must be taken at this point to prevent accident removal increased.
of the guidewire to a level below the obstructing lesion. The obser-
vation at this point of bile ow through the stent is reassuring of
correct stent placement although this nding can also be occasion- COMPLICATIONS
ally observed if the stent has been accidentally placed up the cystic
duct (Fig. 21.7). The reported complications of cholangioscopy include bacteremia,
aspiration, bleeding, and pancreatitis. Bacteremia as determined by
serial blood cultures in the minutes following cholangioscopy can
be demonstrated to occur in 15% of patients, but cholangitis is clini-
cally relevant in only a minority of these situations.20 Prophylactic
Therapy of malignant bile duct lesions antibiotics do not appear to be of benet following cholangioscopy
There is increasing interest in the targeting of therapy against a and stone removal in the surgical setting.21 Bacteremia can result
variety of malignant and pre-malignant bile duct lesions using chol- from over-distention of the biliary tree during irrigation. This may
angioscopic assistance. Photodynamic therapy of cholangiocarci- be more likely if the ampulla forms a tight seal around the insertion
noma has been performed by both the percutaneous15,16 and per-oral tube of the cholangioscope, preventing the venting of excess irrigant
route.17 The technique appears safe with minimal complications but into the duodenum. As a rule, irrigation can be safely performed
its long-term clinical effectiveness remains to be evaluated in multi- with a volume equal to or less than that amount of bile aspirated
centered trials. In general, these patients still require stenting to from the obstructed ductal system. Patients suspected of having
maintain duct patency and the goal of the PDT therapy is likely pal- cholangitis should receive peri-procedure antibiotics. Aspiration
liative rather than curative. occurs as a consequence of the irrigating uid that accumulates in
Biliary papillomatosis is an uncommon condition of multifocal the stomach.22 Aspiration of these contents can be prevented by fre-
papillary lesions of the bile duct. Patients can present with obstruc- quent suctioning of the gastric contents The reported complications
tion and may require transplantation. Transhepatic cholangioscopy following EHL stone fragmentation during cholangioscopy include
using a variety of ablative therapies has been successful in control- cholangitis, bleeding, pancreatitis,23 and perforation. Self-limited
ling disease progression.18,19 bleeding occurs when the EHL pulses are discharged in close apposi-

216

tion to the bile duct wall.24 There are no reports of uncontrolled
bleeding. Bile duct leaks due to EHL injury have been reported.
Perforation can occur at either the papillotomy site or at the site of
an errant EHL discharge.
RELATIVE COST
There are no published cost comparisons between the technique of
cholangioscopy and alternative techniques.
REFERENCES
1. Kozarek RA. Direct cholangioscopy and pancreatoscopy at time
of endoscopic retrograde cholangiopancreatography. Am J
Gastroenterol 1988; 83:5557.
2. Minami A, Nakatsu T, Uchida N, et al. Papillary dilation vs
sphincterotomy in endoscopic removal of bile duct stones. A
randomized trial with manometric function. Dig Dis Sci 1995;
40:25502554.
3. Soda K, Shitou K, Yoshida Y, et al. Peroral cholangioscopy using
new ne-caliber exible scope for detailed examination without
papillotomy. Gastrointest Endosc 1996; 43:233238.
4. Sheen-Chen SM, Chou FF. Is sterile water irrigation safe during
postoperative cholangioscopy? A prospective trial. Eur J Surg
1996; 162:801804.
5. Siddique I, Galati J, Ankoma-Sey V, et al. The role of
cholangioscopy in the diagnosis and management of biliary tract
diseases. Gastrointest Endosc 1999; 50:6773.
6. Seo DW, Kim MH, Lee SK, et al. Usefulness of cholangioscopy in
patients with focal stricture of the intrahepatic duct unrelated to
intrahepatic stones. Gastrointest Endosc 1999; 49:204209.
7. Seo DW, Lee SK, Yoo KS, et al. Cholangioscopic ndings in bile
duct tumors. Gastrointest Endosc 2000; 52:630634.
8. Fukuda Y, Tsuyuguchi T, Sakai Y, et al. Diagnostic utility of peroral
cholangioscopy for various bile-duct lesions. Gastrointest Endosc
2005; 62:374382.
9. Nimura Y, Kamiya J, Hayakawa N, et al. Cholangioscopic
differentiation of biliary strictures and polyps. Endoscopy 1989; 21
Suppl 1:351356.
10. Hui CK, Lai KC, Ng M, et al. Retained common bile duct stones: a
comparison between biliary stenting and complete clearance of
stones by electrohydraulic lithotripsy. Aliment Pharmacol Ther
2003; 17:289296.
11. Adamek HE, Maier M, Jakobs R, et al. Management of retained
bile duct stones: a prospective open trial comparing
extracorporeal and intracorporeal lithotripsy. Gastrointest Endosc
1996; 44:4047.
12. Binmoeller KF, Bruckner M, Thonke F, et al. Treatment of difcult
bile duct stones using mechanical, electrohydraulic and
extracorporeal shock wave lithotripsy. Endoscopy 1993;
25:201206.
Chapter 21 Cholangioscopy
SUMMARY
Cholangioscopy is an important adjunct to interventional
ERCP. It has been documented to improve the diagnostic
accuracy in the assessment of biliary strictures. Cholangioscopy
in conjunction with electrohydraulic lithotripsy can fragment
and eradicate stones refractory to standard interventional
ERCP techniques. There are few complications.
13. Farrell JJ, Bounds BC, Al-Shalabi S, et al. Single-operator
duodenoscope-assisted cholangioscopy is an effective alternative
in the management of choledocholithiasis not removed by
conventional methods, including mechanical lithotripsy.
Endoscopy 2005; 37:542547.
14. Arya N, Nelles SE, Haber GB, et al. Electrohydraulic lithotripsy in
111 patients: a safe and effective therapy for difcult bile duct
stones. Am J Gastroenterol 2004; 99:23302334.
15. Shim CS, Moon JH, Cho YD, et al. The role of extracorporeal
shock wave lithotripsy combined with endoscopic management
of impacted cystic duct stones in patients with high surgical risk.
Hepatogastroenterology 2005; 52:10261029.
16. Wiedmann MW, Caca K. General principles of photodynamic
therapy (PDT) and gastrointestinal applications. Curr Pharm
Biotechnol 2004; 5:397408.
17. Harewood GC, Baron TH, Rumalla A, et al. Pilot study to assess
patient outcomes following endoscopic application of
photodynamic therapy for advanced cholangiocarcinoma. J
Gastroenterol Hepatol 2005; 20:415420.
18. Gunven P, Gorsetman J, Ohlsen H, et al. Six-year recurrence free
survival after intraluminal iridium-192 therapy of human bilobar
biliary papillomatosis. A case report. Cancer 2000; 89:6973.
19. Meng WC, Lau WY, Choi CL, et al. Laser therapy for multiple
biliary papillomatosis via cholangioscopy. Aust N Z J Surg 1997;
67:664666.
20. Chen MF, Jan YY. Bacteremia following postoperative
choledochoberscopya prospective study.
Hepatogastroenterology 1996; 43:586589.
21. Sheen-Chen SM, Chou FF. Postoperative cholangioscopy: is
routine antibiotic prophylaxis necessary?A prospective
randomized study. Surgery 1994; 115:170175.
22. Schebesta AG, Sporr D, OLeary J, et al. Gastric aspiration
associated with operative cholangioscopy. Anaesth Intensive Care
1983; 11:257258.
23. Sheen-Chen SM, Eng HL. Acute pancreatitis following
choledochoscopic stone extraction for hepatolithiasis. Med Sci
Monit 2003; 9:CS13CS15.
24. Fan ST, Choi TK, Wong J. Electrohydraulic lithotripsy for biliary
stones. Aust N Z J Surg 1989; 59:217221.
217
 SECTION 2 TECHNIQUES
Chapter
ERCP in Children
22 Victor L. Fox
INTRODUCTION
Endoscopic retrograde cholangiopancreatography (ERCP) was intro-
duced into pediatric medicine in the mid to late 1970s following
initial experience in adult patients. It is now routinely used for the
diagnosis and treatment of biliary tract and pancreatic diseases in
children who are referred to major medical centers worldwide.13
Although expertise remains concentrated among endoscopists with
advanced training in adult medicine, pediatric specialists collaborate
closely in patient selection, in pre- and post-procedural manage-
ment, and in the periodic appraisal of the role of ERCP in current
pediatric practice.4 In high volume tertiary pediatric referral centers,
ERCP is sometimes performed by expert pediatric endoscopists
working alone or in consultation with adult medicine colleagues.
Major differences between adult and pediatric ERCP relate to
alternative approaches to patient preparation and sedation, to techni-
cal constraints of equipment that is not optimally designed for small
children and infants, and to the rarity of biliary and pancreatic
pathology in children.
DESCRIPTION OF TECHNIQUE
Procedure setting
In modern practice, most patients undergo ERCP with the potential
for immediate therapeutic intervention. Therefore, the setting in
which the procedure is conducted should include appropriate equip-
ment and staff to proceed with available therapies and support for
complications that might arise. Although interventional ERCP can
be performed safely on an ambulatory basis in children, overnight
hospital admission for observation is often advisable given the risk
for post-procedure pancreatitis, and rarer complications of bleeding,
infection, or cardiorespiratory compromise. Immediate access to
subspecialty consultation by pediatric anesthesiologists, surgeons,
and radiologists is essential to provide optimal and comprehensive
team management. Ideally, recovery nurses with experience in
postoperative recognition and management of complications that
occur in children should be available to expedite supportive
interventions.
Endoscopist
Pediatric ERCP is best performed by an endoscopist with advanced
technical skills and sufcient breadth of clinical experience to achieve
an optimal outcome for the child. This may require a collaborative
effort involving both adult and pediatric medicine specialists.
Therapeutic ERCP requires that an endoscopist achieve selective
deep cannulation of the desired (biliary or pancreatic) duct with
>90% success to enable essential interventions including dilation,
stent placement, sphincterotomy, and stone extraction. Since experi-
ence with >200 cases is required by the average trainee to achieve
this rate of technical success5 and the volume of pediatric cases is
relatively small even in tertiary care facilities, pediatric specialists
usually require either supplemental training with adult patients or
a very long training period to achieve initial competence. The volume
of cases required to maintain competence is less well studied in
endoscopists specializing in pediatrics than those specializing in
adults. Complication rates in the latter have been shown to correlate
with case volume and complexity.6 While advanced endoscopic skills
reside most often within adult medicine centers of excellence, tech-
nical skill alone does not benet children in the absence of special-
ized clinical knowledge and experience, since technical success does
not necessarily equate with optimal clinical outcome. Both adult and
pediatric medicine trained endoscopists must consider these factors
and the availability of alternative management options before
embarking on ERCP in pediatric patients.
Sedation
Most pediatric gastroenterologists prefer general anesthesia or deep
sedation for technically challenging procedures in children. There
has also been a trend toward more frequent use of deep sedation in
adults undergoing particularly uncomfortable or lengthy endoscopic
procedures. Although ERCP may be performed successfully with
intravenous (IV) sedation in children, especially in cooperative ado-
lescents, general anesthesia with endotracheal intubation affords
safer airway management with assured analgesia and hypnosis for
as much time as is necessary to complete a potentially lengthy or
difcult procedure.
Fluoroscopy
Fluoroscopy for pediatric ERCP may be performed using a xed
table in a dedicated uoroscopy suite or a portable C-arm in a sepa-
rate procedure room. The advantages of the C-arm device are porta-
bility, lower cost, and easier oblique imaging. Modern digital devices
provide excellent image quality. The x-ray equipment should be
adjusted to accommodate the smaller body of a young child and
reduce the radiation dose rate. Shielding of reproductive organs is
important and should be performed in all patients. Good uoro-
scopic technique by the examiner can minimize radiation exposure
to the child and to personnel (see also Chapter 3). The following
rules or principles will help advance this goal: (1) position the child
so that the beam takes the shortest distance through the body, i.e.
avoid unnecessary oblique projection; (2) position the image intensi-
er or receptor above the patient; (3) minimize the distance of the
intensier and maximize the distance of the x-ray tube to the childs
body; (4) use the least magnication necessary and utilize eld col-
limators to focus on areas of interest; (5) avoid the use of a grid; (6)
minimize beam-on time and use the slowest pulse rates that produce
219
 SECTION 2 TECHNIQUES

acceptable imaging for a given task. The assistance of a radiation

technologist with pediatric experience for equipment set-up and the
availability of a radiologist with pediatric training for consultation
can be very important in achieving the goals outlined above. Either
low osmolar, non-ionic, or high osmolar water soluble contrast
media in the range of 150 to 300 mg per ml may be used.

Supplemental medications
Drug dosing for children is usually based on units per kilogram body
weight ranging up to maximum adult doses. In addition to endocar-
ditis prophylaxis, antibiotics are generally used in the setting of
high-grade biliary or pancreatic duct obstruction, biliary or pancreatic
duct disruption, and pancreatic pseudocyst. Ampicillin/sulbactam
(100200 mg/kg/d IV divided every 6 hours, maximum 4 grams sul-
bactam/d) or a broad spectrum cephalosporin such as cefazolin
(50100 mg/kg/d IV divided every 8 hours, maximum 6 grams/d),
or a uoroquinolone such as ciprooxacin (2030 mg/kg/d IV
divided every 12 hours, maximum 800 mg/d) are usually adequate.
Intravenous glucagon can be used to briey reduce duodenal contrac-
tions during cannulation. A dose of 0.5 mg IV is appropriate for most
ages and can be repeated. Intravenous secretin 0.2 mcg/kg may be
used to facilitate successful cannulation of the minor papilla.
Endoscopic equipment
Children of all ages and sizes, including full-term neonates, can
undergo diagnostic and therapeutic ERCP using duodenoscopes that
are commercially available. Standard diagnostic duodenoscopes with
insertion tube diameters in the range of 1112 mm can be used
effectively in children older than 2 years and with difculty between
12 years of age.7 These endoscopes generally have operating chan-
nels that will accommodate catheters and stents up to 78 Fr, which
is adequate for most interventions. While therapeutic duodeno-
scopes containing operating channels in excess of 4 mm are needed
to place 10 Fr stents, such large endoprostheses are rarely needed
in young children. These larger endoscopes are easily used in
adolescents.
Neonates and infants require a small diameter instrument in the
range of 78 mm that will pass easily through the pylorus and allow
effective positioning of the tip adjacent to the major papilla.8 Cur-
rently, only two duodenoscopes are available specically for use in
small infants: the PJF 160 (Olympus America, Inc Lehigh Valley,
PA) and the ED-2370K (Pentax Medical Company, Montvale, NJ).
Both endoscopes have a maximum distal tip diameter of approxi-
mately 7.5 mm, an operating channel diameter of approximately
2.0 mm, and an elevator. Most diagnostic and therapeutic maneu-
vers are possible with these endoscopes, although the repertoire of
available accessories that will t through the small operating channel
is quite limited. Sphincterotomy, stone extraction, and temporary
stent placement have all been successfully performed in very young
infants using these endoscopes9 (Fox, unpublished) (Fig. 22.1). Cath-
eter tips that taper to a diameter of 34 Fr are helpful in order to
selectively cannulate biliary and pancreatic ducts in infants. Deep
advancement of commercially available catheters into these ducts is
not always physically possible in young infants with normal anatomy
due to the ne caliber of these structures at this age (Fig. 22.2).
Technique
The techniques for ERCP in children are the same as for adult
patients. The procedure can be conducted with the child either prone
or supine on the examining table, although prone positioning is
Fig. 22.1 Placement of nasobiliary stent in 4-month-old infant fol-
lowing sphincterotomy for impacted stone.
most comfortable for the endoscopist. The basic endoscopic maneu-
vers are more technically challenging in children because side-
viewing endoscopes and accessories have not been optimally
designed to work in a narrow lumen and through a narrow operating
channel, respectively. In small children and infants the endoscope
tip is forced into a position in close proximity to the papilla, allowing
very little of the cannula to extend out into view and increasing the
difculty of achieving optimal position for selective bile duct can-
nulation. Although pre-curved cannulas tapering to 3 Fr at the tip
are available (Glo-Tip, GT-5-4-3, Cook Endoscopy, Winston-Salem,
NC), selective biliary cannulation is more easily achieved using a
double-lumen tapered tip, pull type sphincterotome with a short,
20 mm cutting wire (Mini-tomeTM pc, MT-20, Cook Endoscopy).
Tightening the short cutting wire increases angulation of the cathe-
ter tip within a short working distance. Also, by starting the proce-
dure with a sphincterotome, the endoscopist can proceed directly
with therapy when indicated. The UTS-15 (Cook Endoscopy), a 5 Fr
sphinctertome that tapers to 4 Fr at the tip, accepts a 0.021 diameter
wire guide, and has a 15 mm braided cutting wire, may be used for
infants. Wire-guided access, using a soft-tipped, hydrophilic, narrow
gauge wire, may be used if free cannulation proves too difcult (Fig.
22.2). Stiff catheters such as those used for stricture dilation or stone
retrieval are more likely to require wire-guided entry. Also, in young
infants a soft-wire retrieval basket (Memory Baskets, MSB5-2x4,
Cook Endoscopy) will enter the duct more easily if partially opened
since the basket wires are more exible when extending out from
the stiffer plastic sheath. Alternatively, the endoscope can be placed
in the long position, which may achieve a more favorable position
in front of the major papilla, similar to the technique used for can-
nulation of the minor papilla. Tip control is not optimal, however,
with the endoscope in this position.
Although ultra thin duodenoscopes have 2.0 mm operating
channels that will accept accessory catheters of 5 Fr diameter, the

220
 Chapter 22 ERCP in Children

A B C

D E F

G H I

Fig. 22.2 Former 28-week gestation neonate, now a 3-month-old, 3 kg infant with severe cholestasis.
A Transabdominal ultrasound showed fusiform distension of the CBD containing debris suggestive of
choledochal cyst. BC Endoscopic views of major and minor papillae. DF Initial and completed
sphincterotomy and emerging sludge. G Tiny normal pancreatic duct. H Wire-guided access for
deep bile duct cannulation. IJ Cholangiogram during and after basket extraction of sludge.

catheters tend to bind in the channel when the endoscope is bent.
In addition, air and uid cannot be evacuated easily when an acces-
sory is within the channel. These ultrathin endoscopes are also less
apt to maintain a stable scope position due to increased exibility.
Assistance is sometimes needed to maintain torque on the insertion
tube.
Another caveat to consider when instrumenting a small child or
infant is the fragility of the soft tissues. Repeated impaction of cathe-
ters or wire guides against a small papilla in a young infant can render
the structure unrecognizable due to traumatic edema. Also, a false
tract can be created using surprisingly little force while advancing a
catheter or wire into the ampulla of Vater or into the minor papilla.

221
 SECTION 2 TECHNIQUES

INDICATIONS AND CONTRAINDICATIONS

Diagnostic and therapeutic indications

While the necessity of diagnostic ERCP has been diminished by
advances in magnetic resonance cholangiopancreatography (MRCP),
early or subtle diagnostic ndings are still best resolved with direct
contrast injection. This is particularly true for young children who
cannot cooperate with breath-holding sequences required for MRI
or high-resolution CT in the setting of conditions that may require
ne spatial resolution such as early sclerosing cholangitis, bile duct
paucity syndromes and neonatal biliary atresia, anomalous junction
of the biliary and pancreatic ducts, and pancreas divisum. Continued
improvement in non-invasive imaging techniques may eventually
remove this limitation.
The main indication for ERCP in children, as in adults, is for
potential therapeutic intervention of known or suspected structural
abnormalities in order to relieve obstruction, divert leakage, or drain
a collection.

Biliary indications
Neonatal cholestasis
Neonatal cholestasis is the only biliary condition unique to pediatrics
in which purely diagnostic cholangiography has a role. The most
common causes of neonatal cholestasis are idiopathic neonatal hepa-
titis and total parenteral nutrition, both of which are frequently
encountered in neonates compromised by premature birth, congeni-
tal abnormalities requiring surgical intervention, or other acute ill-
nesses of the newborn. These conditions are characterized by
intrahepatic features of hepatocellular and canalicular dysfunction
rather than bile duct obstruction. However, they can sometimes be
difcult to distinguish from ductal obstruction due to inspissated
bile (seen with cystic brosis or idiopathic causes), bile duct paucity
(e.g. Alagilles syndrome), or obliteration of the duct due to biliary
atresia (BA). Of these conditions, the correct diagnosis is most
urgently needed for BA since surgical intervention by portoenteros-
tomy (Kasai procedure), substantially reduces long-term morbidity
and mortality if undertaken early (less than 8 weeks). If left untreated,
BA leads to liver failure and organ transplantation or death by 12
years of age.
The role of ERCP in the diagnosis of BA remains controversial
and is most helpful when the diagnosis is thought to be unlikely
but cannot be denitely excluded without cholangiography (Fig.
22.3). In this situation, an unnecessary exploratory laparotomy can
be avoided. A specialized ultra thin diameter duodenoscope must
be used in neonates. The endoscopic ndings that suggest BA
include: absence of visible bile within the duodenum, partial lling
of the bile duct with abnormal termination, and failure to ll the
bile duct despite lling of the pancreatic duct (Fig. 22.4).10 A high
level of skill and condence in technical prociency is required to
make this diagnosis with certainty. In the largest neonatal series of
suspected biliary atresia, ERCP predicted the correct diagnosis in
all but 2 of 147 infants in whom biliary cannulation was success-
ful.11 Complete lling of the bile duct denitely excludes the diag-
nosis of BA. However, most pediatric gastroenterologists and
hepatologists rely on a combination of clinical presentation, serum
chemistry prole, ultrasonography, biliary scintigraphy, and liver
histology rather than ERCP to select infants with cholestasis for
surgical exploration, intraoperative cholangiography, and antici-
pated portoenterostomy.
Fig. 22.3 Normal intrahepatic and extrahepatic bile ducts and
small gallbladder in 10-week, 4.3 kg infant who has cholestasis
and cystic brosis. Biliary atresia was suspected based on non-
excreting biliary scintigraphy and liver biopsy with suggestive
histopathology.
Cholelithiasis and choledocholithiasis
Choledocholithiasis, usually associated with cholelithiasis, is the pre-
dominant indication for ERCP in children. Black pigment bilirubi-
nate material is usually found in infants and young children with
cholelithiasis while light-colored cholesterol stones are more typical
in adolescent patients without an underlying hemolytic disorder
such as sickle cell disease or spherocytosis. Stringer and colleagues
recently reported detailed analysis of the chemical composition of
gallstones in a series of 20 children ranging in age from 0.3 to 13.9
years.12 Ten had black pigment stones, 2 had cholesterol stones, 1
had brown pigment stones, and 7 (35%) had calcium carbonate
stones, a form uniquely found in children.
Asymptomatic neonatal cholelithiasis may resolve spontane-
ously13 and even symptomatic choledocholithiasis can clear without
the need for aggressive intervention.14 Therefore, a brief period of
supportive care with dietary fasting, IV uids, and antibiotics can be
justied to avoid unnecessary invasive therapy. Otherwise, symp-
tomatic small stones and impacted sludge can be denitively treated
endoscopically without resorting to surgical intervention or more
involved percutaneous transhepatic techniques. Sphincterotomy
with removal of stone and/or sludge material can be undertaken
even in very young infants with appropriate equipment9 (Fig. 22.5).
Balloon sphincteroplasty rather than sphincterotomy might seem an
appealing alternative in young children since the long-term effects
of sphincterotomy performed in childhood are unknown. However,
there are no data beyond individual case reports on the outcome of
balloon sphincteroplasty in children and there is no reason to expect
a lower rate of complications from this technique in children com-
pared with adults.15,16

222

Type 1 Type 2
Type 3A Type 3B
Some pediatric surgeons advocate intraoperative cholangiogra-
phy and laparoscopic CBD exploration at the time of cholecystec-
tomy for primary therapy of choledocholithiasis, thereby avoiding
potential complications of ERCP.1719 With this approach, therapeu-
tic ERCP is reserved for situations in which intraductal calculi cannot
be easily cleared at the time of surgery. However, there are no data
to support routine cholecystectomy in young children with choledo-
cholithiasis, especially if there are no residual stones within the
gallbladder. Small residual gallstones are expected to pass spontane-
ously after endoscopic sphincterotomy alone. In this situation, a
team approach is needed to offer a thoughtful and coordinated man-
agement plan to the family.
Choledochal anomalies
Choledochal anomalies include cystic malformations of the bile duct
and anomalous junctions between the bile duct and pancreatic duct
and these conditions often co-exist. Choledochal cyst (described in
more detail in Chapters 36 and 42) is a descriptive term used when
there is segmental rounded or fusiform distension of the bile duct.
An anatomic classication scheme proposed by Todani20 subcatego-
rizes the condition into types 1 through 5 depending on the shape
and location. Distal CBD obstruction in infants due to choledocho-
lithiasis can induce fusiform distension mimicking Type I chole-
dochal cyst (Figs 22.2 and 22.6). ERCP is, therefore, important in
Chapter 22 ERCP in Children
Fig. 22.4 Schematic representation of
cholangiographic patterns in infants with
biliary atresia. Light green indicates non-
opacied or atretic segments (adapted
from reference 11 with permission of Taylor
and Francis).
this situation to clarify the diagnosis and render therapy that relieves
the obstruction, avoiding unnecessary surgical resection of the bile
duct. When anomalous pancreatobiliary union (APBU) accompanies
a cystic deformity of the bile duct, stenosis at the junction is often
present and surgery becomes a more reasonable treatment option
for denitive decompression and to reduce the long-term risk of later
onset biliary cancer (Fig. 22.7). APBU may also be found incidentally
without associated cystic changes in the bile duct or in association
with acute recurrent pancreatitis.
Biliary strictures and leaks
Most pediatric biliary strictures are due to sclerosing cholangitis.
However, patients with advanced changes of PSC with a dominant
stricture that is amenable to therapeutic endoscopic dilation are the
exception rather than the rule. The typical picture is one of subtle,
diffusely irregular intrahepatic ducts with alternating thin and thick
caliber. Since the bile duct epithelium is diffusely inamed extend-
ing to the papilla, there is often minimal dilation of the CBD due to
mild functional papillary obstruction (Fig. 22.8). A characteristic
observation when PSC is present is underlling of the intrahepatic
ducts despite deep cannulation to the level of the common hepatic
duct. Injection through an inated balloon catheter permits optimal
lling of the intrahepatic branches. Sclerosing cholangitis in chil-
dren is associated with chronic inammatory bowel disease (ulcer-
223
 SECTION 2 TECHNIQUES

A B

C D

E F

Fig. 22.5 A 4-month-old, 4.8 kg infant presenting with acholic stool and jaundice 6 weeks after surgery for complex congenital heart
disease. Endoscopic views of a darkened bulging papilla, cannulation with sphincterotome, removal of impacted soft pigment stone, and
retrieval basket within completed sphincterotomy are seen AD. Cholangiograms showing dilated intra- and extraheptic ducts and small
lling defects within common bile duct E, and retrieval basket open within common bile duct F.

ative colitis and Crohns disease) and is the most common hepatic
complication of primary immunodeciency disorders.21 Patients
with PSC may present with clinical features that are indistinguish-
able from autoimmune hepatitis and the distinction is identied
during cholangiography.22
Although the double duct sign is considered ominous for the
presence of malignancy when seen in adults, it is usually due to a
benign process when seen in children. Imaging with CT or EUS
is recommended in order to exclude a rare tumor within the pancre-
atic head. When the double duct sign is present in children, the
head of the pancreas appears either normal or slightly swollen due
to acute or subacute pancreatitis. Patients present with pain and
obstructive jaundice due to extrinsic compression of the distal
common bile duct as it traverses the head of the pancreas (Fig. 22.9).

224
 Chapter 22 ERCP in Children

Fig. 22.6 A Two-year-old child presenting with pancreatitis,

obstructive jaundice, and cystic dilation of CBD on ultrasound.
B Cholangiogram shows dilation of the CBD extending to the
ampulla. Biliary sphincterotomy and stone extraction resolved the
problem.

Fig. 22.7 Four-year-old child presenting with recurrent pancreati-

tis and biliary obstruction. Cholangiogram shows an anomalous
pancreatobiliary union, distal biliary stricture, and dilated CBD. This
was successfully treated with bile duct resection and Roux-en-Y
anastomosis.

A B

Fig. 22.8 16-year-old female with Crohns disease and PSC conrmed by liver biopsy. A Cholangiogram shows slightly dilated common
bile duct and poor lling of intrahepatic ducts despite deep cannulation. B Contrast injected proximal to inated balloon catheter lls
intrahepatic ducts revealing diffusely irregular pattern.

225
 SECTION 2 TECHNIQUES

Fig. 22.9 Stricturing of the CBD and pancreatic duct within the
pancreatic head due to subacute pancreatitis in an 11-year-old male
with autism who presented with jaundice.

Dilation of the stenosis and placement of one or more temporary

7 Fr stents will relieve the obstruction while awaiting resolution of
the pancreatitis. Biliary brush cytology and/or bile duct biopsy may Fig. 22.10 Choledochal anastomotic stricture and impacted stone
be helpful to assess for malignancy, although the yield is quite low. in an adolescent male following liver transplantation.
Biliary strictures following pediatric liver transplantation are
frequently inaccessible by endoscopic cholangiography since few
children undergo whole organ transplantation with duct-to-duct
anastomosis. For example, in the case of biliary atresia, the most
common indication for liver transplantation in childhood, a biliary
enterostomy is created. Many other children receive a split organ
graft due to a shortage of age-matched whole organ donors and the
use of living related donor grafts. Strictures at the biliary enteric
anastomoses are usually due to ischemia and require radiologic
or surgical intervention. Anastomotic strictures after duct-to-duct
anastomosis are managed the same as in adult patients using
dilation and stent therapy (Fig. 22.10). Strictures near the junction
of the main left and right hepatic ducts have been reported
rarely in children and are presumed to be congenital in origin
(Fig. 22.11).23 These have been successfully managed both surgically
and endoscopically.
Biliary leaks occur in children with laceration of the liver after
blunt abdominal trauma and also after abdominal surgery such as
cholecystectomy. ERCP can be used to simultaneously conrm the
source and to treat the leak by transpapillary stent placement (Fig.
22.12).24 The diameter of the stent is based upon the ductal
diameter.

Unusual biliary infections

Human immunodeciency virus (HIV)-associated cholangiopathy
has been described in children.25 As in adults, the biliary abnormali-
ties include irregularities of contour and caliber of the intrahepatic
and extrahepatic ducts and papillary stenosis. The changes may
result from concomitant infection with opportunistic organisms
such as cytomegalovirus and Cryptosporidium parvum. Ascariasis
infestation may be the most prevalent biliary infection worldwide,
although concentrated within tropical climates. Among 214 children
admitted to hospital in northern India for management of hepato-
biliary and pancreatic ascariasis, 20 (9%) underwent endoscopic and
7 (4%) surgical intervention.26
Fig. 22.11 Presumed congenital stricture of the common hepatic
duct presenting with advanced cirrhosis and portal hypertension in
a 10-year-old child.
Sphincter of Oddi dysmotility
Sphincter of Oddi dysmotility is sometimes considered in children
with unexplained biliary colic-like pain. Although no normal
manometric values have been established for children, some
experts apply adult normal data and perform interventions such as

226
 Chapter 22 ERCP in Children

biliary sphincterotomy when basal pressure exceeds 40mm Hg.
Improvement following sphincterotomy has been reported in small
numbers of patients but no controlled outcome data exist for
children.2,27
Pancreatic indications
Acute pancreatitis
ERCP is rarely indicated in the setting of acute pancreatitis. As in
adults, ERCP is most helpful in the setting of biliary pancreatitis
when there is evidence of choledocholithiasis and severe cholangitis.
Biliary pancreatitis is a commonly seen but underreported entity in
childhood.2830 Drainage of the bile duct may abruptly improve the
childs condition without necessarily improving the pancreatitis.
Another acute indication for ERCP is known or suspected pancreatic
trauma, where disruption of the pancreatic duct is questioned. In
this case, urgent ERCP is recommended to investigate the integrity
of the main duct and attempt to place a transpapillary stent that
crosses a site of free duct leakage31 (Fig. 22.13). Endoscopic therapy
is not indicated for contained intraparenchymal leaks, which usually
heal spontaneously (Fig. 22.14). In tropical countries, parasites,

A B

Fig. 22.12 A Contrast leaking from cystic duct stump in 15-year-old female with biliary leak after laparoscopic cholecystectomy. B Leak
healed after temporary placement of short transpapillary biliary stent.

A
B

Fig. 22.13 A Laceration at junction of head and body of the pancreas in 8-year-old boy after fall from a go-cart. B Pancreatogram
reveals free extravasation from a disruption that could not be crossed with a wire guide. A distal pancreatectomy was performed after
complete pancreatic transection was conrmed during exploratory laparotomy.

227
 SECTION 2 TECHNIQUES

Fig. 22.14 Pancreatogram shows faint intraparenchymal extrava-

sation of contrast in head of pancreas in 9-year-old girl after a lap
belt injury that occurred during an automobile accident.

Fig. 22.15 Anomalous pancreatobiliary union with long common

channel treated with sphincterotomy in 14-year-old girl presenting
with recurrent acute pancreatitis.
primarily Ascaris lumbricoides, are an important cause of acute
pancreatitis in children.26 Endoscopy is indicated to remove obstruct-
ing worms that fail to pass after drug therapy.

Persistent, recurrent, and chronic pancreatitis

ERCP is an important investigative procedure in children with pan-
creatitis that is unrelenting after initial onset or recurs without an
identiable cause. However, a careful history and comprehensive
medical evaluation along with non-invasive imaging should precede
the performance of ERCP in order to avoid unnecessary risk. Occult
drug exposures, underlying metabolic and autoimmune disorders,
and newly recognized genetic disorders may be revealed and obviate
the need for direct pancreatography. Increasingly, children previ-
ously labeled with the diagnosis of idiopathic relapsing pancreatitis
have been reassigned a specic diagnosis without the need for inva-
sive testing. Complete analysis of the genes for CFTR, SPINK1, and
Fig. 22.16 18-month-old infant presenting with recurrent acute
PRSSI, which can identify mutations associated with recurrent and pancreatitis and persistent pancreatic cyst. Pancreatogram revealed
chronic pancreatic disease, has recently become possible using com- intestinal duplication cyst in continuity with main pancreatic duct
mercially available assays (Ambry Genetics, Aliso Viejo, CA). The that was surgically excised and histologically conrmed.
nding of mutations in one or more of these genes may call into
question formerly assigned diagnoses such as pancreas divisum or
sphincter of Oddi dysfunction, for which causality to pancreatitis
remains controversial. Sphincter of Oddi dysfunction has been reported as a cause of
Developmental anomalies involving the pancreas have been recurrent pancreatitis in children, and improves after endoscopic
reported in association with recurrent attacks of pancreatitis. These pancreatic sphincterotomy.2,27,32 As with biliary manometry, basal
include complete and partial pancreas divisum, anomalous pancre- pressures considered normal for adults have been used for normal
aticobiliary junction (Fig. 22.15), and enteric duplications (Fig. baseline values in children. There are no controlled studies compar-
22.16). Reports of endoscopic therapy with sphincterotomy of ing endoscopic sphincterotomy to sham or placebo therapy in
the minor and major papillae, respectively for the rst two children. The largest series to date2 has not yet reported outcome
entities in children are limited but indicate potential benecial data.
outcomes. Communication of the pancreatic duct with cystic dupli- Altered morphology of the main pancreatic duct and its side
cations may be conrmed endoscopically and help guide surgical branches is seen with the progressive disease of chronic pancreatitis
intervention. from various causes (Fig. 22.17). The range of changes is the same

228

A B
Fig. 22.17 Repeat pancreatograms six years apart showing progressive dilation of the main duct in a child with chronic pancreatitis.
as seen in adult patients including ectatic side branches, contour
irregularities and dilation of the main duct, and occasional lling
defects consisting of protein plugs and stones. Pancreatic sphincter-
otomy has been advocated in the setting of a dilated main pancreatic
duct in symptomatic children who have failed to respond to medical
therapy. There are no controlled studies, but short-term improve-
ment following sphincterotomy has been reported.
Pseudocysts that are causing clinical symptoms due to compres-
sion of adjacent structures may be drained endoscopically using
transpapillary, transmural, or combination approaches, as discussed
in Chapter 45. Experience in children is limited to case reports and
small series.33,34
COMPLICATIONS
Pancreatitis is the most common complication of ERCP in chil-
dren. Rates have ranged from 3% to 17% with higher rates associ-
ated with therapeutic procedures.35,36 The highest rates of post-ERCP
pancreatitis in children were reported by Cheng and colleagues2 in
patients who underwent sphincterotomy for SOD: 30% with biliary
sphincterotomy alone, 25% with biliary sphincterotomy followed by
placement of a temporary pancreatic duct stent, and 20% with pan-
creatic sphincterotomy followed by placement of a pancreatic duct
stent. Other complications such as bleeding, perforation, and infec-
Chapter 22 ERCP in Children
tion occur rarely in the larger pediatric ERCP series.2,4 Rates of
complications after sphincterotomy in children appear comparable
to those in adults.37 Cheng et al.2 reported minor acute bleeding
treated with epinephrine injection in 5 patients in their series of
245 therapeutic ERCPs in children less than 18 years of age, includ-
ing 100 selective biliary and 22 dual biliary and pancreatic
sphincterotomies.
The incidence of delayed complications following sphincterotomy
in early childhood has not been reported. Complications of sedation
or anesthesia, although relatively rare, especially when administered
by an anesthesiologist, must also be considered as part of the total
risk to children undergoing ERCP.
RELATIVE COSTS
There are no published data comparing the cost of ERCP with alter-
native diagnostic and therapeutic approaches such as percutaneous
transhepatic cholangiography and direct surgical exploration. Anes-
thesia contributes a large fraction of the total cost for each approach.
Surgical fees likely exceed those of either endoscopy or interven-
tional radiology. The expense of maintaining costly specialized endo-
scopes and a reasonable inventory of accessories can be prohibitive
for many pediatric facilities. Sharing equipment and accessories
with a busy adult medicine service is most cost-effective.
229
 SECTION 2 TECHNIQUES
REFERENCES
1. Pfau PR, Chelimsky GG, Kinnard MF, et al. Endoscopic retrograde
cholangiopancreatography in children and adolescents. J Pediatr
Gastroenterol Nutr 2002 Nov; 35(5):619623.
2. Cheng CL, Fogel EL, Sherman S, et al. Diagnostic and therapeutic
endoscopic retrograde cholangiopancreatography in children: a
large series report. J Pediatr Gastroenterol Nutr 2005 Oct;
41(4):445453.
3. Keil R, Snajdauf J, Stuj J, et al. Endoscopic retrograde
cholangiopancreatography in infants and children. Indian J
Gastroenterol 2000 OctDec; 19(4):175177.
4. Fox VL, Werlin SL, Heyman MB. Endoscopic retrograde
cholangiopancreatography in children. Subcommittee on
Endoscopy and Procedures of the Patient Care Committee of the
North American Society for Pediatric Gastroenterology and
Nutrition. J Pediatr Gastroenterol Nutr 2000 Mar; 30(3):335342.
5. Jowell PS, Baillie J, Branch S, et al. Quantitative assessment of
procedural competence: a prospective study of training in
retrograde cholangiopancreatography. Ann Intern Med 1996;
125:983989.
6. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med 1996;
335:909918.
7. Teng R, Yokohata K, Utsunomiya N, et al. Endoscopic retrograde
cholangiopancreatography in infants and children. J Gastroenterol
2000; 35(1):3942.
8. Rocca R, Castellino F, Daperno M, et al. Therapeutic ERCP in
paediatric patients. Dig Liver Dis 2005 May; 37(5):357362.
9. Wilkinson ML, Clayton PT. Sphincterotomy for jaundice in a
neonate. J Pediatr Gastroenterol Nutr 1996; 23:507509.
10. Guelrud M, Jaen D, Mendoza S, et al. ERCP in the diagnosis of
extrahepatic biliary atresia. Gastrointest Endosc 1991 SepOct;
37(5):522526.
11. Guelrud M, Carr-Locke DL, Fox VL. ERCP in Pediatric Practice:
Diagnosis and Treatment. Oxford: Isis Medical Media Ltd; 1997.
12. Stringer MD, Taylor DR, Soloway RD. Gallstone composition: are
children different? Journal of Pediatrics 2003; 142:435440.
13. Brown DL, Teele RL, Doubilet PM, et al. Echogenic material in the
fetal gallbladder: sonographic and clinical observations. Radiology
1992; 182:7376.
14. St-Vil D, Yazbec S, Luks FI, et al. Cholelithiasis in newborns and
infants. Journal of Pediatric Surgery 1992; 27:13051307.
15. Baron TH, Harewood GC. Endoscopic balloon dilation of the
biliary sphincter compared to endoscopic biliary sphincterotomy
for removal of common bile duct stones during ERCP: a
metaanalysis of randomized, controlled trials. Am J Gastroenterol
2004 Aug; 99(8):14551460.
16. Disario JA, Freeman ML, Bjorkman DJ, et al. Endoscopic balloon
dilation compared with sphincterotomy for extraction of bile duct
stones. Gastroenterology 2004 Nov; 127(5):12911299.
17. Bonnard A, Seguier-Lipszyc E, Liguory C, et al. Laparoscopic
approach as primary treatment of common bile duct stones in
children. J Pediatr Surg 2005 Sep; 40(9):14591463.
18. Mah D, Wales P, Njere I, et al. Management of suspected
common bile duct stones in children: role of selective
intraoperative cholangiogram and endoscopic retrograde
cholangiopancreatography. J Pediatr Surg 2004 June; 39(6):808
812; discussion -12.
19. Vrochides DV, Sorrells DL, Jr., Kurkchubasche AG, et al. Is there a
role for routine preoperative endoscopic retrograde
230
cholangiopancreatography for suspected choledocholithiasis in
children? Arch Surg 2005 Apr; 140(4):359361.
20. Todani T, Watanabe Y, Narusue M, et al. Classication, operative
procedures and review of thirty seven cases including cancer
arising from choledochal cyst. Am J Surg 1977; 134:263269.
21. Rodrigues F, Davies EG, Harrison P, et al. Liver disease in children
with primary immunodeciencies. J Pediatr 2004 Sep;
145(3):333339.
22. Gregorio GV, Portmann B, Karani J, et al. Autoimmune hepatitis/
sclerosing cholangitis overlap syndrome in childhood: a 16-year
prospective study. Hepatology 2001 Mar; 33(3):544553.
23. Chapoy PR, Kendall RS, Fonkalsrud E, et al. Congenital stricture of
the common hepatic duct: an unusual case without jaundice.
Gastroenterology 1981; 80:380383.
24. Sharpe RP, Nance ML, Stafford PW. Nonoperative management
of blunt extrahepatic biliary duct transection in the pediatric
patient: case report and review of the literature. J Pediatr Surg
2002 Nov; 37(11):16121616.
25. Yabut B, Werlin SL, Havens P, et al. Endoscopic retrograde
cholangiopancreatography in children with HIV infection. J
Pediatr Gastroenterol Nutr 1996 Dec; 23(5):624627.
26. Malik AH, Saima BD, Wani MY. Management of hepatobiliary and
pancreatic ascariasis in children of an endemic area. Pediatr Surg
Int 2006 Feb; 22(2):164168.
27. Varadarajulu S, Wilcox CM. Endoscopic management of sphincter
of Oddi dysfunction in children. J Pediatr Gastroenterol Nutr
2006; 42:526530.
28. Sutton R, Cheslyn-Curtis S. Acute gallstone pancreatitis in
childhood. Ann R Coll Surg Engl 2001 Nov; 83(6):406408.
29. Akel S, Khalifeh M, Makhlouf Akel M. Gallstone pancreatitis in
children: atypical presentation and review. Eur J Pediatr 2005 Aug;
164(8):482485.
30. Nowak A, Kohut M, Nowakowska-Dulawa E, et al. Acute biliary
pancreatitis in a 9-year-old child treated with endoscopic
sphincterotomy. Dig Liver Dis 2003 Sep; 35(9):656659.
31. Canty TG, Sr., Weinman D. Management of major pancreatic duct
injuries in children. J Trauma 2001 June; 50(6):10011007.
32. Guelrud M, Morera C, Rodriguez M, et al. Sphincter of Oddi
dysfunction in children with recurrent pancreatitis and anomalous
pancreaticobiliary union: an etiologic concept. Gastrointest
Endosc 1999 Aug; 50(2):194199.
33. Haluszka O, Campbell A, Horvath K. Endoscopic management of
pancreatic pseudocyst in children. Gastrointest Endosc 2002 Jan;
55(1):128131.
34. Cahen D, Rauws E, Fockens P, et al. Endoscopic drainage of
pancreatic pseudocysts: long-term outcome and procedural
factors associated with safe and successful treatment. Endoscopy
2005 Oct; 37(10):977983.
35. Brown CW, Werlin SL, Geenen JE, et al. The diagnostic and
therapeutic role of endoscopic retrograde
cholangiopancreatography in children. J Pediatr Gastroenterol
Nutr 1993 Jul; 17(1):1923.
36. Guelrud M, Mujica C, Jaen D, et al. The role of ERCP in the
diagnosis and treatment of idiopathic recurrent pancreatitis in
children and adolescents. Gastrointest Endosc 1994 JulAug;
40(4):428436.
37. Varadarajulu S, Wilcox CM, Hawes RH, et al. Technical outcomes
and complications of ERCP in children. Gastrointest Endosc 2004
Sep; 60(3):367371.
 SECTION 2 TECHNIQUES
Chapter
ERCP in Pregnancy
23 William M Outlaw and John Baillie
OVERVIEW
Fear of causing fetal and maternal injury or death by procedure-
related pancreatitis delayed the introduction of ERCP for biliary
complications of pregnancy until 1990. Since then, there have been
seven case series (total of 57 patients) and numerous (>30) single
case reports of ERCP in pregnancy. No case of severe ERCP-related
pancreatitis (or other severe complication of the procedure) has been
reported, and all but one reported case has been successful. Post-
ERCP pancreatitis (PEP) and bleeding were reported in 8% and 4%
of the 57 case series patients, while pre-eclampsia, premature birth
and spontaneous abortion were observed in 4%, 2% and 2% of these
same patients. The physiologic changes of pregnancy have profound
implications for many aspects of ERCP, ranging from the type of
drugs than can be used for sedation to positioning the patient and
shielding the fetus from ionizing radiation. Single cases successfully
managed by ERCP during pregnancy include spontaneous bile duct
rupture with bile peritonitis and relief of obstructive jaundice from
pancreatic adenocarcinoma have been reported. Overall, ERCP with
biliary sphincterotomy and clearance of bile duct stones has an
acceptably low morbidity in pregnant women, with no reported
maternal mortality. It is not clear if one reported spontaneous abor-
tion and two premature labors noted in the case series literature
were causally related. ERCP can be recommended for managing the
complications of bile duct stones in pregnancy. However, balloon
dilation of the papilla (balloon sphincteroplasty) should be avoided
due to its tendency to cause pancreatitis.
INTRODUCTION
Endoscopic retrograde cholangiopancreatography (ERCP) rapidly
advanced as a therapeutic specialty in 1974 with the advent of endo-
scopic sphincterotomy (ES), reported simultaneously by Kawai1 and
Classen.2
However, acceptance of ERCP during pregnancy occurred later
than other therapeutic interventions. Based on dated surgical litera-
ture which suggested a signicant spontaneous abortion rate when
laparotomy was performed in the rst trimester of pregnancy,
endoscopists were frightened to attempt ERCP during pregnancy in
any clinical situation. It was felt that a severe attack of pancreatitis
following ERCP might result in fetal, and possibly even maternal,
death. There was also signicant concern about the teratogenicity of
uoroscopy during the rst trimester of pregnancy. In 1990, however,
a group from Duke University Medical Center (including one of us
(JB)) reported its experience of ERCP during pregnancy in ve
patients. These ERCPs allowed the avoidance of surgery.3 All ve
procedures were uncomplicated, the pregnancies lasted to full term,
and ve healthy babies were born. This landmark paper emboldened
other endoscopists to perform ERCP in pregnant women, and this
is now considered accepted practice. Moreover, a recent case series4
comparing medical and surgical treatment of biliary stones in preg-
nancy failed to support the concern about fetal loss following surgery
in pregnancy. Cholelithiasis and choledocholithiasis (Fig. 23.1) in
pregnancy are common, and their complications, including biliary
obstruction, cholangitis and gallstone pancreatitis (Fig. 23.2) are
appropriately and safely managed by ERCP with ES. To the best of
this authors knowledge, no pregnant woman has died as a result of
ERCP. There have been no studies comparing standard endoscopic
sphincterotomy with so-called balloon sphincteroplasty (Fig. 23.3) in
pregnant women, nor are there likely to be given the propensity of
the latter to cause pancreatitis. It is the authors opinion that balloon
sphincterotomy should never be used in pregnancy. Biliary obstruc-
tion in the setting of uncorrectable coagulopathy in a pregnant
woman is more appropriatelyand safelymanaged by stenting to
ensure drainage (Fig. 23.4). Elective sphincterotomy can be per-
formed at a later date, when normal coagulation has been restored.
INDICATION
The indication for ERCP during pregnancy must be well-established
as necessary and critical to the safety of the mother and fetus: preg-
nancy is not the appropriate time to be assessing possible Type III
sphincter of Oddi dysfunction by biliary manometry. Table 23.1 lists
the ASGEs recommendations regarding Indications for Endoscopy
in pregnancy. Table 23.2 lists the ASGEs Guiding Principles for
Endoscopy in Pregnancy. Although choledocholithiasis (Fig. 23.1) is
the commonest biliary pathology requiring ERCP in pregnancy, in
certain parts of the world parasites (e.g. Fasciola, Clonorchis, Ascaris)
can cause biliary obstruction and pancreatitis. These parasites are
best treated by a combination of ERCP with ES for duct clearance
and specic chemotherapeutic agents5 (see below).
RADIATION PROTECTION
Considerable attention has been devoted to shielding the fetus from
ionizing radiation during ERCP. Radiation protection is also dis-
cussed in Chapter 3. Three types of radiation exposure occur:
primary, secondary and leakage.6 The primary radiation beam is
focused and directed through the area of interest. Personnel (doctors,
nurses, technicians) in the procedure room are exposed to secondary
(or scatter) radiation. Leakage is leakage radiation from the radia-
tion source itself. It has been estimated that the typical radiation
dose to the fetus during ERCP is 3.1 mGy.7 This compares with
1.75 mGy for an upper GI contrast series, 7.8 mGy for a PET scan,
15.4 mGy for a pelvic CT scan and 37.0 mGy for an intravenous
urogram (IVU). Protecting the rst and second trimester fetus from
irradiation is usually straightforward; a lead apron is applied to the
231
 SECTION 2 TECHNIQUES
Fig. 23.1 Endoscopic retrograde cholangiogram showing
common bile duct stones.
Fig. 23.2 Duodenal papilla obstructed by a stone in a patient with
gallstone pancreatitis.
appropriate part, shielding the mothers abdomen. The endoscopist
must conrm which direction the x-rays travel in the uoroscopy
system being employed to ensure that the shielding is appropriately
applied; this may dictate placing a lead apron or mat underneath the
patient. In the third trimesterespecially close to termit may be
impossible to avoid exposing the fetus to a small amount of radiation
from screening, but we have no data to suggest that this has caused
any detectable harm, especially since the fetus is well-formed by
then. As a rule, taking hard copy uoroscopic images should be
232
Fig. 23.3 Dilation (balloon sphincteroplasty) of the duodenal
papilla
Fig. 23.4 Plastic stent draining purulent bile from the bile duct in
a patient whose severe coagulopathy was a contraindication to
endoscopic sphincterotomy
Signicant or continuing GI bleeding
Severe or refractory nausea and vomiting or abdominal pain
Dysphagia or odynophagia
Strong suspicion of a colonic mass
Severe diarrhea with negative evaluation
Biliary pancreatitis, choledocholithiasis or cholangitis
Biliary or pancreatic ductal injury
Table 23.1 Indications for endoscopy in pregnancy (ASGE)
avoided, as this signicantly increases radiation exposure. Fluoros-
copy time is kept to an absolute minimum; short bursts of uo-
roscopy are used to conrm the seating of the papillotome in the
bile duct and conrmation of the underlying pathology followed by
as little contrast injection as possible. Typically, less than 60 seconds

Always have a strong indication, particularly in high-risk
pregnancy
Defer endoscopy to the second trimester whenever possible
Use the lowest effective dose of sedative medication
Whenever possible, use category A or B drugs
Minimize procedure time
Position pregnant patients in left pelvic tilt or left lateral position
to avoid vena caval or aortic compression
The presence of fetal heart sounds should be conrmed before
sedation is begun and again at the end of the procedure
Obstetric support should be available in the event of a
pregnancy-related complication
Endoscopy is contraindicated in the presence of obstetric
complications, such as placental abruption, imminent delivery,
ruptured membranes or eclampsia
Table 23.2 General principles guiding endoscopy in pregnancy
of uoroscopy time are required. However, published reports cite
uoroscopy time ranging from 8 seconds8 to 3.2 minutes.9 A stan-
dard radiation dosimeter is applied to the abdominal wall overlying
the fundus of the gravid uterus to check that the fetus has been
adequately shielded. Radiation doses are typically negligible.
Reported measured fetal exposures range from 40 to 310 mrad (4
31 mGy)well within established safe limits.9,10 (Note: 1 mGy =
100 mrad.)
One novel method of ERCP that avoids radiation exposure alto-
gether was described by Simmons et al.11 This technique involves
wire-guided bile duct cannulation without uoroscopic guidance.
Once the wire has passed proximally, the catheter is advanced and
entry into the presumed duct is conrmed by aspirating bile into the
catheter. After a biliary sphincterotomy is performed stone retrieval
is achieved using a stone retrieval balloon. The ballon is passed
proximally into the bile duct, inated, and swept through the sphinc-
terotomy. Using this technique, successful ERCP and stone extrac-
tion were described in six cases.11 Neither uoroscopy nor spot
radiographs (hard copy images) were used during these proce-
dures. The mean ERCP time was 16 minutes (range 1025 minutes).
Though useful, it is likely that this technique will be used by only
the most experienced endoscopists and for straightforward cases.
POSITIONING
Positioning the pregnant mother is rarely a problem. However,
women in the later stages of pregnancy cannot lie prone, the stan-
dard position for ERCP imaging. A semi-prone or left lateral position
is acceptable in such circumstances (Fig. 23.5). The endoscopist
performing the procedure should be familiar with the altered endo-
scopic and radiologic appearances when pregnant patients are so
positioned. The supine position is not acceptable in the third trimes-
ter of pregnancy: the supine hypotensive syndrome occurs in 15
20% of pregnant women at term.12 Uterine blood ow is compromised
in the supine position due to inferior vena cava (IVC), and some-
times aortic, compression by the gravid uterus. This phenomenon
is exacerbated by the use of vasodilators.
As amniotic uid can conduct electrical current to the fetus,13 it
is recommended that the grounding pad for electrocautery should
Chapter 23 ERCP in Pregnancy
Fig. 23.5 The patient is lying in the left semi-prone position with
lead mats below her and a lead apron draped over the lower
abdomen. A dosimeter (not seen) is placed over the fundus of the
uterus on the mothers abdomen to measure fetal radiation
exposure.
be placed in such a way as to avoid interposing the uterus on an
imaginary line between the electrical device and the pad. Bipolar
electrocautery is preferred to minimize the risk of stray currents.
PHYSIOLOGIC CHANGES OF PREGNANCY
Although endoscopists rarely consider the physiologic changes of
pregnancy, these are important and impact directly on sedation,
positioning, uid balance, prevention of gastroesophageal reux
disease (GERD), etc.14 Pregnancy causes a hypercoagulable state;
cardiac output increases 50% during pregnancy; oncotic pressure is
reduced, making pulmonary edema more likely in the setting of
uid overload; oxygen consumption increases 60% and functional
residual (lung) capacity is reduced 20% at term, due to the presence
of the large gravid uterus in the abdomen (this reduction in lung
capacity is reduced a further 30% in the supine position); renal blood
ow increases 75% (glomerular ltration rate 50%) during preg-
nancy, which has implications for drug clearance; loss of tone in the
lower esophageal sphincter (LES) during pregnancy occurs in 22%
of women in the rst trimester and increases to nearly 80% in the
third trimester. Due to the risk of gastroesophageal reux and aspira-
tion of gastric contents, it is recommended that pregnant women
undergo rapid induction for general anesthesia with cricoid pressure
and endotracheal intubation.
SEDATION
The sedation used for ERCP in pregnancy is similar to that used for
moderate (conscious) sedation and general anesthesia, except that
benzodiazepines are generally avoided due to their long half-life.15
Sustained use of diazepam in early pregnancy has been associated
with cleft palate.16 In later pregnancy, it has been suggested that
diazepam use may result in neurobehavioral disorders.17 Midazolam
has not been reported to be associated with congenital abnormali-
ties. However, due to concerns about benzodiazepines in general,
midazolam is usually avoided, especially in the rst trimester. Pro-
233
 SECTION 2 TECHNIQUES
pofolTM (DiprivanTM and others, 2,6-bis(1-methylethyl-phenol)) is
considered safe in pregnancy. However, if PropofolTM is used, it
should be noted that this agent has no analgesic effect. The fetal
heart rate should be assessed before and after ERCP to ensure that
the fetus remains viable and undistressed by the procedure. Stan-
dard monitoring (i.e. pulse oximetry, automated blood pressure
monitoring, and capnography) is employed for the mother. The
safety of commonly used medications for endoscopy during preg-
nancy is covered in more detail in the ASGE Guideline for Endos-
copy in Pregnant and Lactating Women.18
OUTCOMES
Table 23.3 summarizes the collective outcome of case series consist-
ing of = 3 patients. Reports of one or two cases, and those lumping
together ERCP and non-ERCP endoscopy in pregnancy,11,1924 have
been excluded. Seven case series3,911,15,19 comprising a total of 57
patients undergoing ERCP during pregnancy have been reported.
The median stage of pregnancy in which these patients presented
with symptomatic choledocholithiasis (the main indication for
ERCP) was mid-to-late second trimester. Complications were post-
ERCP pancreatitis (PEP) 8%, post-sphincterotomy bleeding (4%),
and pre-eclampsia (4%). There were two premature births (at 35
weeks, one with severe growth retardation and pulmonary immatu-
rity) (4%) and one spontaneous abortion (2%).
MANAGEMENT OF RELATED BILIARY DISORDERS
OF PREGNANCY
Thaggard et al.25 reported a rare case of spontaneous bile duct
rupture during pregnancy leading to bile peritonitis that was
managed endoscopically. Allmedinger et al.26 describe the use of
percutaneous cholecystostomy to manage acute cholecystitis in two
women during the third trimester of pregnancy. Both delivered
normal babies at term and underwent uneventful laparoscopic
Author Number of pts
Kaleleh (ref 12) 17
Gupta (ref 10) 18
Jamidar (ref 16) 23
(6 centers; total of 29 ERCPs)
Tham (ref 11) 15
Baillie (ref 6) 5 1-1st trimester
Barthel (ref 20) 3 n/a
Simmons (ref 21) 6 6-30 weeks
Total 57
Table 23.3 ERCP during pregnancy: series of 2 patients
PEP-post-ERCP pancreatitis.
234
cholecystectomy (LC) within 3 months. Sungler at al.,27 and Fried-
man and Friedman,28 report the combination of ERCP and LC for
managing gallstone disease in seven pregnancies, all with successful
outcome. Diettrich et al.29 identied 34/1100 pregnant women
patients who underwent laparotomy within 6 weeks of delivery.
Twenty-six patients had spontaneous vaginal deliveries (SVDs) and
eight had Cesarian sections. There were no complications resulting
from any of these procedures. Blackbourne et al.30 reported a most
unusual case of a patient diagnosed with pancreatic adenocarcinoma
at 17 weeks gestation in whom a biliary stent was successfully placed
to relieve biliary obstruction. The patient subsequently underwent
pancreaticoduodenectomy (Whipple procedure).
Hewitt et al.31 described their experience of managing three
patients with choledochal cysts (CDC) during pregnancy. One CDC
that was conservatively managed ruptured, with fetal loss and a
prolonged hospitalization. Based upon this case proactive drainage
of CDC should be considered in order to limit the risk of spontane-
ous rupture in the peripartum period.
Routine worming of women of childbearing age is recom-
mended in areas where Ascaris is endemic. Despite this, Shah
et al.32 describe an interesting experience of 15 cases of symptomatic
biliary Ascariasis seen over a 10-year period. Ten of these patients
were in their third trimester of pregnancy. The condition was most
reliably diagnosed by transabdominal ultrasound. Nine of the
15(60%) patients responded to conservative treatment. Among the
remaining six cases, endoscopic extraction of Ascaris worms was
successful in four cases and open biliary surgery was required in
two cases. Of the pregnant patients, one suffered a spontaneous
abortion at 12 weeks and another went into premature labor.
In summary, ERCP is a safe and effective modality primarily for
the management of choledocholithiasis.3336
Acknowledgement
The authors are grateful to Dr Dick Kozarek (Seattle, WA) for his
helpful critique of the rst draft of this manuscript.
Mean gestation (weeks) Complications/outcomes
18.6 PEP-1, Bleed-1
Pre-eclampsia-2
17 PEP-1, Bleed-1
Preterm delivery-1
15-1st trimester PEP-1
8-2nd trimester Spontaneous abortion 3 months
6-3rd trimester after ERCP
28.9 None
None
4-2nd trimester
PEP-1
2 premature births
mid-late 2nd trimester PEP-4 (8%)
Bleed-2 (4%)
Pre-eclampsia-2 (4%)
Spontaneous abortion 1 (2%)
Premature birth 2 (4%)

REFERENCES
1. Kawai K, Akasaka Y, Murakami K, et al. Endoscopic
sphincterotomy of the ampulla of Vater. Gastrointest Endosc
1974; 20:148151.
2. Classen M, Demling L. Endoscopic sphincterotomy of the papilla
of Vater and extraction of stones from the choledochal duct.
Dtsch Med Wochenschr 1974; 99(11):496497.
3. Baillie J, Cairns SR, Putnam WS, et al. Endoscopic management of
choledocholithiasis during pregnancy. Surg Gynecol Obstet 1990;
171(1):14.
4. Lu EJ, Curet MJ, El-Sayed YY, et al. Medical versus surgical
management of biliary tract disease in pregnancy. Am J Surg
2004; 188(6):755759.
5. Shah OJ, Robanni I, Khan F, et al. Management of biliary Ascariasis
in pregnancy. World J Surg 2005; 29(10):12941298.
6. http://www.sgna.org/resources/guidelines/guideline2_print.html.
7. http://brighamrad.harvard.edu/education/fetaldose/diag-
exposure.html.
8. Gupta R, Tandan M, Lakhtakia S, et al. Safety of therapeutic ERCP
in pregnancyan Indian experience. Indian J Gastroenterol 2005;
24(4):161163.
9. Tham TC, Vandervoort J, Wong RC, et al. Safety of ERCP during
pregnancy, Am J Gastroenterol 2003; 98(2):308311.
10. Kahaleh M, Hartwell GD, Arseneau KO, et al. Safety and efcacy of
ERCP in pregnancy. Gastrointest Endosc 2004; 60(2):287292.
11. Simmons DC, Tarnasky PR, Rivera-Alsina ME, et al. Endoscopic
retrograde cholangiopancreatography (ERCP) in pregnancy
without the use of radiation. Am J Obstet Gynecol 2004;
190(5):14671469.
12. Holmes F. Incidence of supine hypotensive syndrome in late
pregnancy. A clinical study of 500 subjects. J Obstet Gynaecol Br
Emp 1960; 67:254258.
13. Einarson A, Bailey B, Inocencion G, et al. Accidental electric shock
in pregnancy: a prospective cohort study. Am J Obstst Gynecol
1997; 176:678681.
14. Cappell MS. Sedation and analgesia for gastrointestinal
endoscopy during pregnancy. Gastrointest Endosc Clin N Am
2006; 16(1):131.
15. Jamidar PA, Beck GJ, Hoffman BJ, et al. Endoscopic retrograde
cholangiopancreatography in pregnancy. Am J Gastroenterol
1995; 90(8):12631267.
16. Ornoy A, Arnon J, Shectman S, et al. Is benzodiazepine use during
pregnancy really teratogenic? Reprod Toxicol 1998; 12:511515.
17. Laegreid L, Olegard R, Wahlstrom J, et al. Teratogenic effects of
benzodiazepine use during pregnancy. J Pediatr 1989;
114:126131.
18. ASGE Standards of Practice Committee. Guidelines for endoscopy
in pregnant and lactating women. Gastrointest Endosc 2005;
61(3):357362.
19. Barthel JS, Chowdhury T, Miedema BW. Endoscopic
sphincterotomy for the treatment of gallstone pancreatitis in
pregnancy. Surg Endosc 1998; 12(5):394399.
Chapter 23 ERCP in Pregnancy
20. Rahmin MG, Hitscherich R, Jacobson IM. ERCP for symptomatic
choledocholithiasis in pregnancy. Am J Gastroenterol 1994;
89(9):16011602.
21. Uomo G, Manes G, Picciotto FP, et al. Endoscopic treatment of
acute biliary pancreatitis in pregnancy. J Clin Gastroenterol 1994;
18(3):250252.
22. Cohen SA, Kasmin FE, Siegel JH. ERCP during pregnancy. Am J
Gastroenterol 2003; 98(2):237238.
23. Goldschmidt M, Wolf L, Shires T. Treatment of symptomatic
choledocholithiasis during pregnancy. Gastrointest Endosc 1993;
39(6):812814.
24. Quan WL, Chia CK, Yim HB. Safety of endoscopical
procedures during pregnancy. Singapore Med J 2006;
47(6):525528.
25. Thaggard WG, Johnson PN, Baron TH. Endoscopic management
of spontaneous bile duct perforation and bile peritonitis
complicating term pregnancy (case report). Am J Gastroenterol
1995; 90(11):20542055 (case report).
26. Allmendinger N, Hallisey MJ, Okhi SK, et al. Percutaneous
cholecystostomy of acute cholecystitis of pregnancy. Obstet
Gynecol 1995; 86(4 pt 2):653654.
27. Sungler P, Heinerman PM, Steiner H, et al. Laparoscopic
cholecystectomy and interventional endoscopy for gallstone
complications of pregnancy. Surg Endosc 2000; 14(3):267271.
28. Friedman RL, Friedman IH. Acute cholecystitis with calculous
biliary obstruction in the gravid patient. Management by ERCP,
papillotomy, stone extraction and laparoscopic cholecystectomy.
Surg Endosc 1995; 9(8):910913.
29. Diettrich NA, Kaplan G. Surgical considerations in the
contemporary management of biliary tract disease in the post-
partum period. Am J Surg 1998; 176(3):251253.
30. Blackbourne LH, Jones RS, Catalano CJ, et al. Pancreatic
adenocarcinoma in the pregnant patient: case report and review
of the literature. Cancer 1997; 79(9):17761779.
31. Hewitt PM, Krige JE, Bornmann PC, et al. Choledochal cyst in
pregnancy: a therapeutic dilemma. J Am Coll Surg 1995;
181(3):237240.
32. Shah OJ, Robanni I, Khan F, et al. Management of biliary ascariasis
in pregnancy. World J Surg 2005; 29(10):12941298.
33. Axelrad AM, Fleischer DE, Strack LL, et al. Performance of ERCP
for symptomatic choledocholithiasis during pregnancy:
techniques to increase safety and improve patient management.
Am J Gastroenterol 1994; 89(1):109112.
34. Baillie J. ERCP during pregnancy. Am J Gastroenterol 2003;
98(2):237238.
35. Menees S, Elta G. Endoscopic retrograde
cholangiopancreatography during pregnancy. Gastrointest
Endosc Clin N Am 2006; 16(1):4157.
36. Cappell MS. The fetal safety and clinical efcacy of gastrointestinal
endoscopy during pregnancy (review). Gastrointest Endosc Clin
North Am 2003; 32(1):123179 (review).
235
 SECTION 2 TECHNIQUES
Chapter
ERCP in Surgically Altered Anatomy
24 Simon K. Lo
ERCP is generally considered the technically most difcult endo-
scopic procedure. But it can be made more challenging if the gas-
trointestinal or pancreaticobiliary anatomy is modied. A thorough
understanding of a surgically altered anatomy is essential to mini-
mize complications and to enhance the chance of a successful
outcome. In virtually all these cases, careful preprocedure planning
is mandatory (Table 24.1).
SURGERY THAT MAY IMPACT THE
PERFORMANCE OR INTERPRETATION OF ERCP
Many operations reroute the upper gastrointestinal tract and may
require a special equipment or extreme familiarity in order to reach
the pancreatic and biliary systems. Some surgery removes or alters
a portion of the bile duct or pancreas but does not create any hard-
ship to performing ERCP. Conversely, there are those improbable
surgical consequences that would not allow any endoscopic access
of the biliary tract regardless of experience, skill and equipment. We
will explain most of these operations and attempt to bring up rele-
vant points pertaining to ERCP.
ESOPHAGEAL RESECTION
Mostly done for esophageal neoplasm or pre-malignant conditions,
up to 22% of esophageal resections may be accompanied by high
esophageal anastomosis strictures.1 Additionally, small diverticuli
can form proximal to these anastomoses. In passing a duodenoscope,
care must be taken to avoid forcing it through a diverticulum or an
anastomotic stricture. If resistance is encountered, an end-viewing
upper endoscope should be used to inspect the esophageal anatomy
carefully. Esophageal resections also result in the stomach being
brought up above the diaphragm and turned into a tubular sack that
ends with a pyloroplasty. Once the duodenoscope has passed through
the pylorus, the approach to the major papilla requires slightly more
clockwise rotation or a longer insertion of the endoscope than usual.
This is because of the more proximal location of the duodenum and
the lack of a competent pylorus to hold the endoscope in place.
GASTRIC RESECTION
There are many forms of gastric resection, ranging from a Billroth
I where there is very little loss of the volume of the stomach to a
total gastrectomy. As a result, the impact of gastric resection on
ERCP can be either minimal or profound.
Billroth I
In a Billroth I surgery, only the antrum and pylorus are removed
and the stomach is attached to the duodenum along its greater
curvature (Fig. 24.1). Scope passage into the duodenum is typically
easier than usual, but the papilla is harder to visualize. As expected,
both the major and minor papillae are more proximally located
than usual. In the short-scope position, the papilla is seen following
an exaggerated clockwise rotation of the endoscope. However,
anchoring of the endoscope is difcult without the pylorus and
achieving a stable scope position for deliberate cannulation can
be quite difcult. In this situation, working in the long-scope posi-
tion may be more desirable for bile duct cannulation because the
papillary orice is better visualized and the scope is more stably
situated.
Billroth II
Before proton pump inhibitors were introduced, peptic ulcer surgery
was common. A Billroth II operation involves an antrectomy and
creation of a gastrojejunostomy. The result is an end (stomach)-to-
side (jejunum) anastomosis with an afferent and an efferent limb
(Figs 24.2A, 24.2B) next to each other beyond the line of anastomo-
sis. The afferent limb travels proximally and ends at the duodenal
stump whereas the efferent limb restores continuity with the rest of
the gastrointestinal tract. The major papilla is obviously located near
the duodenal stump and is seen with the papillary orice facing the
endoscope. There are several challenges that the endoscopist has to
deal with when performing ERCP in a patient with Billroth II
anatomy. It begins with the choice of endoscope. Once it was believed
that end-viewing endoscopes were best used to navigate the small
bowel and pass cannulas into the biliary orice. Many experienced
biliary endoscopists switched to using the side-viewing duodeno-
scopes to take advantage of the elevator and better viewing of the
papilla. But in a prospective randomized study of 45 patients in
Korea, no difference was noted between the side-viewing and end-
viewing endoscopes in success of cannulation and sphincterotomy.2
In fact, the end-viewing endoscopes were safer to use. Regardless of
what endoscope to use, Billroth II ERCP is one of the more difcult
procedures. In a study involving 185 Billroth II ERCP procedures,
the failure rate was 34%.3
The afferent lumen cannot be identied by visual inspection
alone, though it is commonly believed that it is the more awkwardly
located orice. The afferent limb can be attached to the stomach
along either the lesser (antiperistaltic) (Fig. 24.2A) or greater (isope-
ristaltic) curvature (Fig. 24.2B). Scope passage with an end-viewing
device is rather intuitive, and the major challenge is in visualizing
and cannulating the papilla. Even though data suggests there is no
difference in technical success between side-viewing and end-
viewing, endoscopies, in reality it is far easier to perform cannula-
tion and therapies using a duodenoscope.
When using a duodenoscope, the difculty starts with gaining
entry into the afferent lumen. It is even more difcult if the afferent
limb is sutured to the lesser curvature after it has exited the
237
 SECTION 2 TECHNIQUES

Understand the prior surgery thoroughly

Choose the proper endoscope
Standard therapeutic duodenoscope
Thin-caliber diagnostic duodenoscope
Pediatric duodenoscope
Diagnostic upper endoscope
Therapeutic upper endoscope
Pediatric (variable stiffness) colonoscope
Therapeutic colonoscope
Push enteroscope
Double balloon enteroscope
Linear EUS scope
Position the patient properly
Prone
Supine
Left lateral
Left oblique
Prepare accessories
Standard accessories
Non-precurved catheters
Nasobiliary drains
Specialty accessories (e.g. for Billroth II)
Long-length accessories
Anesthesia
Conscious sedation
Propofol anesthesia
General anesthesia Fig. 24.1 Billroth I gastrectomy. Antrectomy is followed by con-
nection of the stomach to the duodenum in the end-to-end fashion.
The lesser curvature side of the cut end of the stomach is closed
Table 24.1 Pre-procedure planning in situations that involve to allow creation of the gastroduodenostomy.
surgically altered anatomy

A B

Fig. 24.2 A Billroth II gastrectomy with an antiperistaltic gastrojejunostomy anastomosis. In this case, the afferent limb is accessed through
the stomal orice located near the lesser curvature. B Billroth II gastrectomy with an isoperistaltic anastomosis, where the afferent limb
is attached to the greater curvature.
238
 Chapter 24 ERCP in Surgically Altered Anatomy

gastrojejunostomy, as this form of operation creates a xed and sig-
nicantly angulated point of entry4 (Figs 24.3A, 24.3B). When a
duodenoscope is used, the technique to enter the orice is the same
as for the pylorus. When the lumen can only be visualized in the
retroexed position, gliding the scope along gently toward it rarely
works. Suctioning out excess gastric air may make the gliding a little
easier. Once the tip of the scope has reached the orice, the scope
should be gradually pulled back and straightened in order to advance
further. There is the technique of entering the lumen by rotating the
scope 180 degrees at the orice and pointing the tip of the scope
down until the small bowel lumen is clearly in sight. This technique
can theoretically be done with the scope facing the opening rather
than backing in. However, the endoscope would be blinded by
mucosal lining and visual inspection during the maneuvering would
become impossible. Hand compression of the mid-abdomen or
extending a polypectomy snare into the intended lumen has been
reported to add to the success of intubation of a difcult intestinal
orice.5 Even in a tertiary biliary center, failure to enter the afferent
limb has been reported to be as high as 10%.3
Once the side-viewing duodenoscope has securely engaged in a
jejunal lumen, passage forward is safe and effective by constantly
orientating the bowel in the 6 oclock position (Figs 24.4A). This

A B

Fig. 24.3 A Endoscopic appearance of a typical Billroth II gastrojejunostomy. Viewing from the stomach, the two jejunal limbs are located
at the extreme right and left of this image. B Billroth II gastrectomy with the greater curvature side of the stomach attached to the
jejunum in an end-to-side fashion. The lesser curvature side of the stomach is closed surgically. Some surgeons choose to suture the
jejunum onto this area to protect the suture line. If the afferent limb is tagged down in this manner, then endoscope entry into this limb
may be quite difcult.

Fig. 24.4 A The typical view of the jejunal

A B lumen when a side-viewing duodenoscope
is being advanced. Note the upper half of
the lumen should always be kept in the 6
oclock position. B A similar duodeno-
scopic view of the distal lumen in the 6
oclock position. Note the upper lumen
always represents a retroexed view. An
attempt to pass the scope toward the 12
oclock direction would cause either perfo-
ration or the scope to fold backward.

239
 SECTION 2 TECHNIQUES

would simulate the view of driving a car in a long tunnel. It is A

common to see two lumens when a duodenoscope is partially
retroexed and it creates confusion regarding where to advance the
endoscope. A good rule of thumb is to orientate the two lumens
along the vertical midline and the lower lumen is the one that the
endoscope should enter (Fig. 24.4B). Natural intestinal redundancy
and tortuosity rarely allow unimpeded forward advancement. Rather,
successful passages require a combination of gentle rotation, dial
redirection, pulling and pushing. Care must be taken to minimize
sudden or forceful manipulations, as perforation may result. In one
series, perforation occurred in 5% and was mainly from manipula-
tions in the afferent limb.3 In another study, 18% of the cases were
complicated by jejunal perforation.2 Usage of a small caliber and
soft, older duodenoscope may reduce the chance of traumatizing the
intestinal wall. Minimizing air insufation helps keep the lumen
straight and the bowel wall soft and stretchable. Rotating the patient
is occasionally effective in getting around seemingly improbable
turns. These measures all may contribute to a safe and successful
procedure. With experience and special care, an acceptable risk of
perforation can be achieved.6 After passing some distance, it is wise
to take a uoroscopic picture to conrm that the scope has passed
or is passing though the transverse duodenum (Fig. 24.5). If the
scope is in the pelvis, then it is likely to be in the efferent limb and
should be withdrawn to search for the other intestinal orice. Some
afferent limbs seem to be longer and more tortuous to reach the
papilla than others. This impression is indeed correct, as the afferent
loop may be created in the antecolic fashion over the transverse
colon (Figs 24.6A, 24.6B).
On the way to the proximal duodenum, an anastomosis that con- B
nects the afferent to the efferent limbs may be encountered. This
Braun procedure is a modication of the Billroth II operation to
reduce bile reux into the stomach or to lessen the chance of duo-

Fig. 24.6 A Retrocolic construction of the Billroth II gastrojejunos-

tomy. The afferent limb is relatively short in this case. B Antecolic
Fig. 24.5 Fluoroscopy shows the duodenoscope is facing the Billroth II gastrojejunostomy. The afferent limb is signicantly longer
right direction and crossing the transverse duodenum. than that in 24.6a.

240
 Chapter 24 ERCP in Surgically Altered Anatomy

denal obstruction (Fig. 24.7). It should not inuence the endoscopic perform sphincterotomy in this setting.8 However, most endosco-
passage other than by creating confusion to the endoscopist. On pists seem to prefer needle-knife sphincterotomy over a biliary stent
some occasions the duodenal stump appears as a blind sac with a because it avoids injury to the pancreatic sphincter and allows
distinctly at and smooth mucosa and is reached without the major gradual, unhurried tissue cutting.4,9 Balloon sphincter dilation, com-
papilla being noticed. The minor and then major papilla should be monly done with an 8 mm balloon, is technically easy to perform. A
readily identied upon gradual withdrawal of the endoscope (Figs randomized study showed balloon sphincter dilation was as effective
24.8A24.8C). as sphincterotomy to facilitate stone extraction and had fewer com-
The major papilla is almost always found near the 12 oclock
position of the duodenum when a duodenoscope is used (Fig. 24.8C).
If the intestinal lumen is kept in view ahead of the endoscope, then
the bile duct leads away from the endoscope straight ahead or slightly
to the right (Figs 24.9A, 24.10). In order to keep the cannulating
catheter or guidewire tangentially to the duodenal wall for biliary
access, the papilla should not be approached up close (Fig. 24.8C).
Rather, the scope should be pulled back slightly with its elevator
partially lowered for catheter passage. Conversely, the pancreatic
duct is easier to cannulate by advancing the scope close to the papilla
and keeping the elevator in a lifted position (Fig. 24.8B). Some
endoscopists prefer to use straight-tip catheters for biliary cannula-
tion,5 while others like to use straight guidewires. However, perhaps
the most effective way in entering the bile duct is to use a catheter
that has been bent in an S-shape (Fig. 24.9B). A cap-assisted approach
has been described to improve cannulation of a Billroth II papilla
when an end-viewing endoscope is used7 (Fig. 24.11). On some occa-
sions when pancreas divisum is suspected, the minor papilla should
be correctly identied for cannulation. As a general rule, it is located
slightly further away from (cephalad to) and to the left of the major
papilla (Figs 24.8A, 24.10).
Biliary sphincterotomy is accomplished either by the use of a
Billroth II sphincterotome or a needle-knife to cut over a biliary stent.
A Billroth II push sphincterotome is designed virtually opposite to a
conventional traction sphincterotome. The cutting wire is loosened
to form a half loop over a straight sphincterotome catheter. The wire
loop is then pushed forward to cut the papilla hood along the 12
oclock position. Sphincterotomy done in this manner is slightly less
well controlled than in the normal setting because of the pushing
motion and suboptimal visualization of the proximal edge of the
Fig. 24.7 A Braun modication of a Billroth II operation. Here the
papillary mound. A modied sphincterotome that forces its tip into afferent and efferent limbs are connected via a side-to-side
an S-shape when the cutting wire tightens may also be used to anastomosis.

A B C

Fig. 24.8 A The minor papilla, which is quite prominent in this case, is located more cephalad to and to the left side of the major papilla.
Further up the afferent lumen is the duodenal stump. B As the scope is pulled back from the duodenal stump, the major papilla is seen
up close and perpendicular to the duodenoscope. This position favors cannulation of the pancreatic duct. C The duodenoscope is
pulled back further and farther away from the major papilla. This position favors cannulation of the bile duct.

241
 SECTION 2 TECHNIQUES
A B
Minor papilla Major papilla
Fig. 24.10 A schematic illustration of the relationship of the major
and minor papillae and the directions of the bile duct (yellow
arrow) and pancreatic duct (blue arrow).
Fig. 24.11 A balloon dilator is being used to perform sphinctero-
plasty using an end-viewing endoscope. Note a short, soft cap has
been tted to the tip of the scope. This cap is believed to improve
the ability to cannulate the papilla.
242
Fig. 24.9 A The major papilla is usually
located at the 12 oclock position. Here the
guidewire points in the direction of the bile
duct. B This S-shaped biliary cannula is
best used for intubating the bile duct.
Fig. 24.12 A typical Roux-en-Y gastrojejunostomy.
plications.9 Additionally, there was no more pancreatitis with this
treatment than sphincterotomy.
Roux-en-Y gastrectomy
Created to reduce reux of pancreatic and biliary uids into the
stomach following a partial gastrectomy, the gastric outlet is con-
structed similar to that of a Billroth II surgery. However, this end-
to-side anastomosis leads to a very short blind stump and a long,
efferent limb (Fig. 24.12). The jejunum of the efferent limb extends

around 40 cm before a jejunojejunostomy anastomosis is found. At
this point two or three lumens (Figs 24.13A, 24.13B) will be identi-
ed, depending on whether the two jejunal limbs are connected
end-to-side (Fig. 24.13C) or side-to-side (Fig. 24.13D). If done side-
to-side, one of the three outlets is a short, blind stump. If the afferent
limb is correctly entered, the endoscope will travel up the proximal
jejunum, the ligament of Treitz, transverse duodenum and nally
the descending duodenum. This long distance to travel makes it
nearly impossible for a 125 cm-long duodenoscope to reach the
major papilla. Many longer-length endoscopes have been used to
perform ERCP in this setting, including pediatric or adult versions
of colonoscopes and push enteroscopes.10 A special oblique-viewing
endoscope has been reported to be potentially useful for this
purpose.11 Double-balloon enteroscopes can reach as far as the
cecum per-orally and were introduced in the United States in 2004
(Figs 24.14AD). ERCP, including therapeutic maneuvers, has been
reported to be successful in 5 of 6 patients when using this new form
of enteroscope.12
The challenge of performing ERCP in a Roux-en-Y gastrectomy
lies not just in traveling a great length and recognizing the proper
intestinal lumen but also in selectively cannulating the bile duct and
pancreatic duct. All end-viewing endoscopes have the inherent dif-
culty in identifying the major papilla because of its location along
A B
C D
Chapter 24 ERCP in Surgically Altered Anatomy
the interior aspect of the duodenal C-loop. Even when it is seen,
cannulation is extraordinarily difcult because of awkward orienta-
tions and unstable positioning for cannulation (Figs 24.15AC).
In the hands of ERCP experts, the success rate is a mere 67%.13
Yamamoto and colleagues reported the success of performing
diagnostic and therapeutic ERCP with a double-balloon enteroscope
in ve patients.12 Interestingly, the authors tted a small plastic cap
on the tip of the enteroscope to enable cannulation. It is uncertain
if and how this plastic cap helps with the procedure. Given the
difculty and frequent failure of performing ERCP in this post-
operative anatomy, it is best to refer this type of cases to a tertiary
biliary center or choose an alternative method such as a transhepatic
study. Performing ERCP with the intent to evaluate and treat a
pancreatic condition is particularly problematic as a transhepatic
procedure is ineffective. An intraoperative transjejunal ERCP method
has been reported to overcome this difculty. At laparotomy, an
enterotomy is made at 20 cm distal to the ligament of Treitz to allow
passage of a gas-sterilized duodenoscope to advance up the afferent
limb.14
Total gastrectomy
Done usually for treatment of gastric cancer, a total gastrectomy
leads to the creation of an end-to-side esophagojejunostomy. One
Fig. 24.13 A A schematic illustration of 3
lumens at the point of a jejunojejunostomy
anastomosis. The single distal lumen on the
same side of the anastomosis is either the
efferent or a blind limb. One of the two
lumens on the other side of the anastomosis
is a blind stump or the efferent limb while
the other is the afferent limb. In either case,
the afferent limb has to cross the line of
anastomosis. B Endoscopic picture of the
two lumens seen beyond an anastomosis.
One of these two orices should lead to the
afferent limb. C Illustration of an end-to-
side jejunojejunostomy anastomosis. D Il-
lustration of a side-to-side jejunojejunostomy
anastomosis.
243
 SECTION 2 TECHNIQUES

B
A

Fig. 24.14 The double-balloon system consists of a balloon on the tip of a thin endoscope A and a balloon on an overtube B. C Both
balloons are inated. D Balloon ination device that controls air insufation, deation, pressure reading, and an alarm with a yellow-light
indicator of excessive pressure.

A B C

Fig. 24.15 A The major papilla seen with an end-viewing endoscope. Note it is a very tangential view with an awkward position for
cannulation. B After rotating the end-viewing endoscope, the major papilla appears to be optimally positioned for cannulation. The
papilla is still tangentially located and it is difcult to maintain this view for long. C A catheter has been successfully inserted into the bile
duct with this end-viewing endoscope.

244

Fig. 24.16 A total gastrectomy with a Roux-en-Y esophagojeju-
nostomy. In spite of the signicant distance, a side-viewing duode-
noscope is usually long enough to reach the papilla.
lumen of the esophagojejunostomy is a blind end, whereas the other
is the efferent jejunal limb (Fig. 24.16). In a short distance down this
limb is a side-to-side or end-to-side jejunojejunostomy to receive
pancreatic and biliary contents. Similar to the Roux-en-Y gastrec-
tomy, the endoscope has to enter the afferent limb and pass through
the proximal jejunum and most of the duodenum. But unlike Roux-
en-Y partial gastrectomy, a duodenoscope actually can reach the
major papilla on a more regular basis. Once the major papilla is
identied with the duodenoscope, the approach to ERCP cannula-
tion and therapy is quite similar to that for a Billroth II anatomy.
But if a duodenoscope is too short to reach the descending duode-
num, then an end-viewing long endoscope has to be used. In that
case, the challenge lies in cannulating and treating disease processes
without the benets of an elevator and the side-viewing capability.
UPPER GI BYPASS SURGERY
WITHOUT RESECTION
Gastrojejunostomy
The indications for gastrojejunostomy without resection of any part
of the stomach include large pancreatic head mass, benign chronic
duodenal obstruction and unresectable malignant duodenal stric-
ture. When inspecting the stomach during ERCP, the gastrojejunos-
Chapter 24 ERCP in Surgically Altered Anatomy
tomy is usually located along the dependent portion of the stomach.
However, it may be slightly off to the anterior or posterior wall along
the greater curvature (Figs 24.17A, 24.17B). While most anastomo-
ses for bypass of obstructive diseases are expected to be large, some
of these gastrojejunostomy openings appear to be quite small.
Immediately through the rim of anastomosis two jejunal orices will
be found, and either opening can be the one that leads to the afferent
limb. If it is the more distal orice, then it is an antiperistaltic con-
nection and the afferent limb is relatively short. But this distance
becomes longer if the surgery is done in the antecolic fashion
because the intestine has to drape over the transverse colon. This
limb may become even longer if the gastrojejunostomy is created in
the isoperistaltic manner. On some occasions a second anastomosis
is noted beyond the gastrojejunostomy. This Braun procedure (Fig.
24.17B) is done to add further bypass of contents between the affer-
ent and efferent limbs to reduce alkaline biliary reux into the
stomach or to provide a safety net to minimize the chance of an
afferent limb obstruction. If the scope passes through a Braun anas-
tomosis, it has a 50% chance of returning to the stomach via the
other limb of the gastrojejunostomy.
Since the major papilla is intact in this setting, a duodenoscope is
preferred for ease of inspection and cannulation unless it is proven
to be too short. In practice, reaching the descending duodenum is
often not the key issue. Instead, inspection and cannulation is a bigger
challenge because most of these cases have a highly stenotic duode-
num that has led to the gastrojejunostomy. Fortunately, duodenal
obstruction from pancreatic head cancer is frequently located proxi-
mal to the major papilla leaving sufcient room to carry out an ERCP.
In the event of an inadequate space, balloon dilation of the duodenal
stricture can be perform but the resultant mucosal trauma and hem-
orrhage may add even more obstacles to the procedure. A transhepatic
approach to biliary drainage is frequently necessary in this situation.
Alternatively a rendezvous procedure, in which a transhepatic cathe-
ter or guidewire is passed across the biliary sphincter, can be done
for endoscopic access. There are occasions when the bypass surgery
is done for gastroparesis and performing an ERCP in the usual ante-
grade fashion is preferred. In this situation the duodenoscope has to
be slightly rotated, when gliding along the greater curvature, to skip
around the gastrojejunostomy to reach the pylorus.
Duodenal bypass
Duodenal perforations are occasionally treated by a duodenojejunos-
tomy. Even though this form of operation is uncommon, the nding
of two or more intestinal lumens beyond the pylorus or at the
descending duodenum may create confusion to the endoscopist. If
there is no associated duodenal narrowing, the procedure is straight-
forward and the key is to carefully inspect each lumen until the
major papilla is found. If a mildly to moderately stenotic duodenal
lumen is found, gentle balloon dilation may be attempted to ease
the scope passage. Alternatively, a pediatric ERCP scope with a
7.5 mm outer diameter and 2.0 mm instrument channel can be
used. But the small endoscope channel allows only limited therapeu-
tic possibilities such as sphincterotomy, basket stone extraction and
placement of a 5-French stent.
BARIATRIC SURGERY
There are many forms of bariatric operations to induce weight reduc-
tion, but about 70% of them in the US are Roux-en-Y gastric bypass.
Biliopancreatic diversion (12%), vertical banded gastroplasty (7%)
245
 SECTION 2 TECHNIQUES

A B

Fig. 24.17 A Gastric bypass with an anteriorly located gastrojejunostomy. B Gastric bypass with a posteriorly located gastrojejunos-
tomy. Note a Braun procedure that connects the afferent and efferent jejunal limbs.

and gastric banding (5%) constitute the remaining practices of bar-
iatric surgery today.15 Interestingly, morbidly obese patients are most
likely to have bariatric surgery done in the northeastern United States
than the rest of the country.16 As obesity affects more Americans,
more weight-reduction operations are being performed. Since gall-
stone disease and abdominal pain are common issues in obese indi-
viduals, biliary and pancreatic evaluations are in increasing demand.
At the same time, this group of patients are the most difcult to
perform ERCP on because of the need to pass through the extremely
long intestine to get to the proximal duodenum. Variation of tech-
niques and surgeons preferences add further confusion to this dif-
cult ERCP arena.
Malabsorptive-jejunoileal bypass
This form of weight-reduction surgery is mentioned here primarily
for the historical purpose. It is neither practiced today nor does it
affect the performance of ERCP. Rarely, a consultation for ERCP
may be requested for a patient with a prior jejunoileal bypass surgery
because of jaundice. In this case, the cause of jaundice is more likely
due to hepatic failure rather than biliary tract disease. This operation,
popular prior to 1980, consists of transecting and connecting a large
jejunoileal segment to the distal colon. Alternatively, the proximal
jejunum is transected and connected to the distal ileum in an end-
to-side manner, excluding a long jejunum and ileum from contact
with intestinal nutrients (Fig. 24.18). The result of this form of
operation is a very short functioning small bowel that causes weight
loss by malabsorption and maldigestion. Chronic diarrhea, stone
diseases and fatal liver dysfunction are reasons why all patients
who have undergone this surgery should have a jejunoileal bypass
reversed.17
Biliopancreatic diversion and duodenal switch
There are two other forms of malabsorptive operations that are still
practiced today. Metabolic complications are reportedly less often
seen than in jejunoileal bypass. Both operations require resection of
most of the stomach and connection of either the duodenum or the
stomach remnant to the distal ileum (250 cm proximal to the ileoce-
cal valve) (Figs 24.19A, 24.19B). The excluded block of duodenum-
jejunum-ileum is then hooked up to the distal ileum at around
100 cm proximal to the ileocecal valve. ERCP is not a possibility
whether it is biliopancreatic bypass or duodenal switch because the
descending duodenum can only be reached by passing through most
of the small intestine.
Restrictive surgery
There are two forms of this operation that aim to restrict food intake
by creating a mere 15 ml proximal gastric pouch and a small pouch
outlet of roughly 1 cm in diameter. Vertical band gastroplasty,

246

popular in 1980s, consists of cutting off the fundus with staples and
limiting the pouch outlet with a 5 cm circumferential Marlex band17
(Fig. 24.20A). A more modern alternative is the laparoscopic adjust-
able gastric banding procedure, done with placement of a silicone
band that wraps around the gastric cardia (Fig. 24.20B). Tightness
of this band can be modied following surgery by inserting a needle
into a reservoir embedded in the abdominal wall. Restrictive gastric
surgery is easily recognized endoscopically because of the tiny gastric
pouch that leads to a small rm outlet that barely allows the passage
of an upper endoscope. The channel length is usually only 12 cm.
If this surgery is suspected in advance of the ERCP, it is best to start
the procedure with a standard upper endoscope to inspect the tight-
ness of the outlet. If difculty passing a duodenoscope is anticipated,
then dilating the ring outlet with a 13.5 mm balloon can be safely
done. Once the duodenoscope has reached the distal stomach, no
special technique is needed to accomplish an ERCP procedure. In
patients who have regained their weight after a vertical band gastro-
Chapter 24 ERCP in Surgically Altered Anatomy
Fig. 24.18 Jejunoileal bypass. This surgery does not inter-
fere with the performance of ERCP.
plasty, their stomachs may appear normal with minimal evidence of
any restriction or have two outlets from the pouch that lead to the
rest of the stomach. This loss of physical restriction is either due to
breakage of the gastric band or development of a gastric pouch-
gastric fundus stula.
Gastric bypass
This is the most commonly done surgical procedure to induce
weight reduction today. It works by both restricting food intake with
a small (<50 ml) gastric pouch and creating malabsorption with a
proximal small bowel bypass. This is the most difcult form of
Roux-en-Y surgical anatomy for performance of ERCP. In order to
reach the papilla, the endoscope has to travel through 40 cm of the
esophagus, a tiny gastric pouch, a 75150 cm Roux limb followed by
a variable-length afferent limb (Fig. 24.21). After arriving at the
descending duodenum, the tip of the endoscope has to be exible
enough to visualize the major papilla and sufciently still to perform
247
 SECTION 2 TECHNIQUES

A B

Fig. 24.19 A Biliopancreatic diversion. The long efferent and afferent limbs prevent any chance of performing ERCP in this case.
B Duodenal switch procedure. Very similar to the biliopancreatic diversion surgery, it is not possible to perform ERCP in this situation.

cannulation. Even a 250 cm-long push enteroscope is usually too
short because of looping and stretching of the intestine. Working
with long accessories to perform ERCP in this setting posts addi-
tional technical challenges that even elite endoscopists consider
overwhelming.
There are several ways of performing ERCP in this very difcult
situation (Table 24.2). The simplest, but most likely method to fail is
to use the longest endoscope available and cannulate the papilla
with extra-long accessories. A more logical alternative is to use a
double-balloon enteroscope, which is thinner, more exible and
able to travel long segments of the intestine. Indeed, there
is preliminary evidence that they can be used to carry out the proce-
dure successfully in most cases.12 If cannulation is not possible in
spite of successful scope advancement, then it is possible to perform
a percutaneous endoscopic gastrostomy (PEG) by inserting the
double-balloon enteroscope retrograde into the stomach. After
the large bore stula has matured, a duodenoscope can be inserted
through the gastric stula to perform ERCP. A similar alternative is
to have a surgical gastrostomy tube placed and then perform ERCP
after maturation of the tract.18 It is also possible to perform either

248

A B
Fig. 24.20 A Vertical banded gastroplasty, with a small aperture that allows food passage. This channel may have to be dilated in order
to pass a duodenoscope. B Laparoscopic adjustable gastric banding. The degree of constriction in the proximal stomach is adjustable
externally.
Per-oral with a long scope:
Colonoscope (therapeutic, pediatric)
Duodenoscope (not advisable)
Push enteroscope (need special long accessories)
Double-balloon enteroscope (need special long accessories)
Via a mature gastrocutaneous stula
Fistula created at surgery
Fistula created by retrograde PEG
Fistula created by percutaneous gastrectomy
Rendezvous approach with a long scope
Intra-operative via a stoma using a sterilized endoscope
Gastric stoma
Proximal jejunal stoma
Table 24.2 Methods of performing ERCP in gastric bypass
patients
an open surgical gastrostomy or a laparoscopic gastrostomy
duodenoscope insertion in one setting.19 Alternatively, a transjejunal
operative approach to advance a duodenoscope at 20 cm below the
ligament of Treitz may be attempted.14 Of course, the Rendezvous
procedure that involves placement of a percutaneous transhepatic
guidewire into the duodenum/jejunum may help passage of the
endoscope as well as gaining access into the bile duct. Finally, not
all biliary diseases require endoscopic manipulations and a solely
Chapter 24 ERCP in Surgically Altered Anatomy
transhepatic approach may be sufcient to manage these conditions
without involving the endoscopist.
PANCREATIC RESECTION
Conventional Whipple procedure
The most well-known pancreatic surgery is the Whipple operation
or pancreaticoduodenectomy, done for resection of malignant or
benign pancreatic head lesions. While the concept of resection is
simple, the multiple anastomoses in this operation create a great
deal of confusion to the non-surgeons. Since bile duct, pancreatic
duct and duodenum are interrupted, at least three separate connec-
tions are made to re-establish pancreatic, biliary and intestinal con-
tinuity. When an endoscope is passed into the mid-stomach, two
small bowel orices are seen in the conventional Whipple (Fig.
24.22A). One gastrojejunostomy ascends up the afferent limb and
joins the bile duct and eventually the pancreatic duct. The other
orice leads to the efferent limb and the rest of the gastrointestinal
tract. Since the usual challenge of cannulating an intact papilla is
absent, ERCP can be readily performed with either an end-viewing
or a side-viewing endoscope. It is possible to start this procedure
with a diagnostic upper endoscope to take advantage of its small
caliber and exibility. If a gastric loop prevents the endoscope from
reaching the biliary anastomosis, a scope-stiffening stylet may be
inserted into the gastroscope instrument channel. If a diagnostic
249
 SECTION 2 TECHNIQUES
Fig. 24.21 Gastric bypass surgery for weight reduction that com-
bined restrictive and malabsorptive principles. This increasingly
common operation is one of the most challenging surgically altered
anatomies for performance of ERCP.
upper endoscope is too short, then a pediatric colonoscope should
be able to get to the bilioenteric and even the pancreaticoenteric
anastomosis (Fig. 24.22BC).
A 120 cm, therapeutic channeled, oblique-viewing prototype
endoscope was successful in treating a post-Whipple case that had
failed other attempts with existing endoscopes.11 However, more
clinical experience with this scope is necessary to reveal the true
value of this oblique-viewing endoscope in this setting. Fluoroscopy
may occasionally be of benet to conrm if the endoscope is within
the afferent limb because it should be located in the right upper
quadrant. Fluoroscopy may also help to locate the anastomosis as it
is always coming off the most cephalad portion of the bowel gas in
the right upper quadrant. A patent hepaticojejunostomy should also
allow pneumobilia to be visualized. Endoscopically, a normal bile
duct anastomosis is readily identied as a round orice with bile
existing. The opening is frequently located eccentrically or retracted
behind an intestinal fold. It is usually a subjective opinion in diag-
nosing a mild or moderate hepaticojejunostomy narrowing. In
severe stenosis, the opening can appear as a pinhole or be covered
completely by a lm of whitish scar tissue. The pancreaticoenteric
anastomosis is notoriously difcult to identify and cannulate. It can
250
be seen at either the very proximal end of the afferent limb (invagina-
tion of the pancreatic remnant into the end lumen of the intestine)
(Figs 24.22A and 24.22C) or slightly before the surgically closed
stump is reached (end-to-side, mucosa-to-mucosa, pancreaticojeju-
nostomy) (Figs 24.22D and 24.22E).
Pylorus-preserving Whipple procedure
Aimed to minimize the gastric emptying problem and other surgery
related morbidity, a pylorus-preserving Whipple is different from the
conventional Whipple by preserving the full stomach and a tiny
segment of the proximal duodenal bulb (Fig. 24.22D). However, this
presumed advantage of keeping the antrum and pylorus has not
been proven.20 When performing an ERCP, the stomach and pylorus
appear normal. Immediately past the pylorus are two jejunal lumens,
with either one possibly leading to the biliary and pancreatic anas-
tomoses. However, the efferent limb is frequently pointing directly
away from the pylorus.
Pancreaticogastrostomy
A pancreaticoduodenectomy can be further modied to have the
main duct of the body and tail of the pancreas implanted into the
posterior wall of the stomach21 (Fig. 24.23). In this case, the bilioen-
teric anastomosis is found along the afferent limb in the usual
manner. However, the pancreatic orice is no longer seen near the
end of the proximal jejunal stump. Rather, it is located along the
posterior gastric wall as a small opening. Finding the anastomosis
among the gastric rugal folds can be difcult; but parenteral injec-
tion of secretin and spraying a color dye on the gastric mucosa may
help with the identication.
Other pancreatic resective surgery
Resection of the tail of the pancreas does not alter gastric, duodenal
or pancreaticobiliary anatomy. Injection of contrast across the pan-
creatic sphincter would obviously discover a shortened duct. Mid-
pancreatic resections for benign diseases may result in a normal
anatomy in the head of the pancreas and a very short pancreatic duct.
The tail of the pancreas is usually drained into a piece of jejunum.
Studying the pancreaticojejunostomy in this setting is difcult.
When the tail of the pancreas is connected to the posterior wall of
the stomach, the possibility of accessing the pancreatic duct becomes
much greater.22
PANCREATIC DUCT DRAINAGE PROCEDURES
Puestow procedure
This longitudinal pancreaticojejunostomy procedure is favored by
most pancreatic surgeons for decompression of a dilated pancreatic
duct for relief of pain from chronic pancreatitis. The procedure
involves opening the pancreatic duct from the head to the tail of the
pancreas and creating a side-to-side anastomosis between the open
edges of the pancreatic duct and those of a piece of jejunum.23 There
is obviously no anatomical alteration in the upper gastrointestinal
tract as a result of the surgery and ERCP can be done in the usual
manner. Contrast injection into the main pancreatic sphincter
should identify the Wirsung duct followed by immediate opacica-
tion of the jejunum. While determining if there is a stricture between
the Wirsung duct and the jejunum is relatively easy, there is no
certainty if the duct beyond that point is fully decompressed by the
surgery.
 Chapter 24 ERCP in Surgically Altered Anatomy

A
B

Fig. 24.22 A Conventional Whipple procedure. Note the pancreatic duct is found at the
upper end of the afferent limb. B A cholangiogram obtained through a hepaticojejunos-
tomy anastomosis with a pediatric colonoscope. C A pancreaticogram of a markedly
dilated pancreatic duct following a Whipple procedure. D Pylorus-preserving Whipple
procedure with an intact stomach and a very short duodenal bulb before the lumen splits
into the afferent and efferent limbs. Note the pancreaticojejunostomy is fashioned on the
side of the proximal afferent limb before the intestinal lumen ends in this case. A corre-
sponding radiograph (E) showing the relationship between the endoscope and a tortuous
and markedly dilated pancreatic duct due to stenosis of the pancreaticojejunostomy. The
afferent jejunum ends a few centimeters straight away from the tip of the endoscope.

251
 SECTION 2 TECHNIQUES
Fig. 24.23 Pancreaticogastrostomy. The arrow points to the pan-
creatic duct that is directly anastomosed to the posterior wall of the
stomach. Endoscopic pancreatography is possible if this gastric
anastomosis is identied.
Freys procedure
This is a procedure that combines a longitudinal pancreaticojejunos-
tomy (see Puestow procedure above) and duodenum-preserving,
local removal of pancreatic tissue within the head of the pancreas.24
Similar to the Puestow procedure, injection of contrast should iden-
tify a Wirsung duct that drains into the intestine. In spite of the
resection of a big portion of the head of the pancreas, the pancreatic
duct is not interrupted and therefore should show no altered anatomy
or the appearance of any anastomosis.
Duval procedure
One of the original pancreatic duct drainage procedures, this surgery
involves resection of the tail of the pancreas and the spleen and cre-
ation of an end-to-end pancreaticojejunostomy. The principal func-
tion of this surgery is to allow backward drainage of an obstructed
pancreatic duct into the jejunum rather than through the obstruc-
tion at the head of the pancreas. However, this operation has a high
incidence of treatment failure and is rarely practiced today. A
pancreatogram should reect a slightly shortened pancreatic duct
that connects to an intestinal lumen in this situation.
BILIARY SURGERY
Choledochoduodenostomy
This is a simple form of surgery to provide mid-bile duct drainage
for a benign condition such as a distal bile duct stricture or recurrent
choledocholithiasis. There is usually no ductal transection, and the
252
anastomosis is made in a side-to-side fashion between the proximal
duodenum and the mid-bile duct (Fig. 24.24A). However, this
procedure may be complicated by recurrent fever, abdominal
pain, liver abscess, pancreatitis or cholangitis. Bile duct impaction
due to trapped debris distal to a choledochoduodenostomy is referred
to as the sump syndrome (Figs 24.24B,C,F). Interestingly, symptoms
of sump syndrome usually occur 56 years after the surgery is
performed.25 Inspection of biliary complaints in this situation
should aim at studying the bile duct via the major papilla. However,
the biliary sphincter is frequently stenotic and may not allow access
through it. Therefore, contrast injection via the anastomosis may
become necessary. Identication of the orice, which is normally
about 0.5 to 1 cm in diameter, is difcult because it is often located
in the posterior aspect of the proximal descending duodenum
(Fig. 24.24D). Nonetheless, careful inspection and gentle rotation of
the duodenoscope should reveal this opening. A biliary sphincterot-
omy may provide relief from sump syndrome for up to several
years.25 Jaundice and recurrent cholangitis may also occur as a result
of stenosis of the choledochoduodenostomy; balloon dilation and
stenting through the strictured anastomosis is necessary in this
situation.
Roux-en-Y hepaticojejunostomy
The hepatic duct is surgically connected to the jejunum in a variety
of conditions, including recurrent biliary stones, benign distal biliary
strictures, cholangiocarcinoma, choledochal cyst, liver transplanta-
tion and iatrogenic bile duct injury. In stone disease and benign
biliary stasis, continuity of the entire biliary tract may be kept and
the biliojejunal anastomosis is constructed side-to-side. ERCP can
be carried out in the usual way by accessing the major papilla (Fig.
24.25A). The clue to this surgery is continuous spillage of contrast
into the jejunal lumen, located near the duodenal bulb. Quality
cholangiograms can be obtained only by injection of contrast supe-
rior to the hepaticojejunostomy or by occluding the common hepatic
duct with a stone-retrieval balloon.
Most of the time a hepaticojejunostomy is created with transec-
tion of the mid-bile duct. In this situation a cholangiogram via the
major papilla should show a complete blockage at the proximal
common bile duct (Fig. 24.25B). If this is known in advance, per-
forming a standard ERCP via the major papilla to study the biliary
tract is unnecessary and raises the risk of otherwise negligible
chance of pancreatitis. Sometimes the surgically ligated stump is
misinterpreted as tight biliary stricture and perforation may occur
due to aggressive probing of the blind end.
Studying the biliary system proximal to the transected bile duct
is a rather difcult task. It requires a long endoscope that can travel
through the entire duodenum, ligament of Treitz, proximal jejunum,
jejunojejunostomy and the afferent limb. A variable-stiffness pedi-
atric colonoscope is the ideal endoscope for this purpose, although
a push enteroscope, therapeutic colonoscope or double balloon
enteroscope may also be used. Since there is no major papilla
involved, accessing the bile duct is relatively straightforward once
the proximal afferent limb is reached. Nonetheless, searching for the
hepaticojejunostomy or choledochojejunostomy anastomosis may
take some experience. It is frequently hidden behind a sharp turn
or a recessed fold. Viewing is often difcult because it may only be
partially visible near the edge of an endoscopic image. Cannulation
may be possible with the use of a guidewire in this situation. An
immediate bifurcation may be noted if the anastomosis is created in
 Chapter 24 ERCP in Surgically Altered Anatomy

Fig. 24.24 A A choledo-

A B chojejunostomy without
interruption of the bile
duct. The bile duct can be
accessed either through
the major papilla or the
choledochoduodenosto-
my. B A pigmented
stone (thin arrow) and a
large piece of fresh vege-
table (large arrow) was
swept into the duodenum
after a sphincterotomy.
C More foreign body-like
biliary debris was extracted
from the bile duct of the
same patient. This material
appears to have been in
the bile duct for a long
duration. D A schematic
illustration of the usual
D relationship between the
choledochoduodenos-
tomy and the major
papilla. Finding the anasto-
mosis can be quite difcult
because of its usual loca-
tion in the posterior wall
of the proximal descend-
ing duodenum. E A
biliary stent was visualized
through a choledochodu-
odenostomy as expected.
F A cholangiogram typi-
cally seen in a sump syn-
C drome. The open arrow
points towards the cho-
ledochoduodenostomy
where biliary contrast
escapes laterally into the
F duodenum. Thin arrows
show lling defects
throughout the dilated
bile duct.

253
 SECTION 2 TECHNIQUES

A B

Fig. 24.25 A Roux-en-Y hepaticojejunostomy without transection of the bile duct. ERCP is done in the usual fashion and contrast is seen
owing outside of the bile duct via the biliary bypass into the jejunum. B Roux-en-Y hepaticojejunostomy without biliary continuity. In
this case, the only possible way to study the bile duct is by passing a long endoscope through the proximal jejunum and up the afferent
jejunal limb. Injection of the major papilla to evaluate the bile duct should be avoided to prevent causing pancreatitis.

the proximal hepatic duct. On rare occasions, two separate orices
may be identied when the bilioenteric anastomosis is created very
high up the bile duct. When the anastomosis is severely stenosed, a
seemingly complete closure by a lm of whitish scar tissue may be
observed. It is essential to look for a fully opacied intrahepatic
cholangiogram. If a part of the intrahepatic distribution is missing,
then a ductal stricture or an occluded separate orice should be
suspected.
Cholecystojejunostomy
This operation was once commonly employed to bypass the bile
duct obstructed by an obviously unresectable pancreatic head
cancer. However, this form of surgery is an unreliable means to
decompress the bile duct because of the potential of tumor extension
to involve the cystic duct. The surgery is a very simple one whereby
a distended gallbladder is opened up and anastomosed to a jejunal
limb (Fig. 24.26). Today, this operation is mostly reserved for the
intra-operative nding of an unresectable and obstructing cancer
that is too large to allow access to the proximal bile duct to create a
hepaticojejunostomy. This surgery does not alter the upper GI
anatomy and poses no additional difculty to the regular ERCP. In
fact, the result of this surgery is not readily recognized unless the
bile duct is overlled with contrast.
Liver transplantation
The bile duct anastomosis in liver transplantation is usually a duct-
to-duct connection or choledochocholedochostomy. There is no
special technical challenge to the performance of ERCP in this
situation. When the transplantation is done for primary sclerosing
cholangitis or other conditions in which the distal bile duct
cannot be used, then a Roux-en-Y choledochojejunostomy or
hepaticojejunostomy is constructed (Fig. 24.25b). Some living donor
liver transplantation cases also utilize Roux-en-Y hepaticojejunos-
tomy because of difculty in matching up the native bile duct with
the donor right hepatic duct. As mentioned before, a long endoscope
is mandatory for a successful ERCP study in this situation.

254

Hepaticocutaneous jejunostomy
This form of biliary surgery is rarely encountered in the US but is
occasionally done in Asia, where recurrent pyogenic cholangitis is
prevalent. This is essentially a Roux-en-Y hepaticojejunostomy with
an extension of the afferent limb to the abdominal wall as a perma-
nent stoma or to be concealed in the subcutaneous tissue. The per-
sistent nature of recurrent pyogenic cholangitis mandates an easy
access to the bile duct for periodic clearance of intrahepatic biliary
stones.26 Via this stoma a choledochoscope, bronchoscope or a pedi-
atric upper endoscope can be inserted into the intrahepatic ducts for
stricture dilation and stone extraction. This cutaneous stoma pro-
vides a highly convenient means of biliary access without having to
perform numerous difcult ERCP and per-oral choledochoscopies.
It should signicantly reduce the risks of post-ERCP cholangitis,
cumulative radiation exposure and repeated prolonged sedation.
ENDOSCOPIC TECHNIQUES COMMONLY
EMPLOYED FOR ERCP IN SURGICALLY
ALTERED ANATOMY
Performing a Rendezvous procedure
There is no single Rendezvous technique adopted by all gastroenter-
ologists for endoscopic biliary access through the use of a percutane-
Chapter 24 ERCP in Surgically Altered Anatomy
Fig. 24.26 A schematic drawing of a partially exposed
proximal afferent jejunum to illustrate biliary decompres-
sion through a patent cystic duct. This form of surgery does
not interfere with the performance of ERCP and may not
even be recognized if only a small amount of contrast is
injected above an obstructed distal bile duct.
ous transhepatic biliary guidewire or catheter. Most of these
procedures are done in two steps and possibly in two locations of a
hospital.27 An interventional radiologist would rst insert a needle
into a dilated peripheral intrahepatic duct. A guidewire, between
250 cm and 450 cm in length, is then passed through the needle into
the bile duct and eventually into the duodenum or jejunum if there
is a biliojejunal anastomosis. The external end of the guidewire is
then secured onto the abdominal wall with heavy dressing. Some
radiologists prefer to slide a thin-caliber biliary catheter over the
guidewire to protect liver and biliary tissue from slicing injuries
from a thin, tightly wound, wire.27 The patient is then transferred to
an endoscopic suite where ERCP is performed. If endoscope passage
is met with technical difculty, such as in the case of a long afferent
limb of a Whipple procedure, the guidewire can be oated down the
intestinal lumen under uoroscopy guidance until it is seen endo-
scopically. The guidewire is grasped with a snare and pulled up the
endoscope channel until it can be secured externally. With a slight
traction on the guidewire, the endoscope may be advanced along the
lumen. However, the assumption that an externally placed wire may
readily pull an endoscope up the afferent tract is inaccurate. In fact,
excessive tension on a tightly drawn wire can injure the intestine or
liver tissue and must be avoided.
For the most part, the rendezvous procedure is done for travers-
ing a difcult papilla or a tight stricture. The free end of the wire is
255
 SECTION 2 TECHNIQUES

snared and pulled up the channel of the endoscope until it can be On rare occasions, uoroscopic examination may visualize a redun-
secured externally. During the guidewire manipulation, the external dant gastric or jejunal loop that can be gently straightened before
end must be secured with a clamp to prevent it from being drawn making further advancement. Some highly angulated intestine
into the intestine. Once the guidewire has been pulled outside of the behaves like a blind pouch. The downstream lumen may be identi-
instrument channel a sphincterotome, balloon dilator or biliary stent ed only by probing with a hydrophilic wire or infusion of contrast
is readily passed over the guidewire into the bile duct to complete into the intestinal lumen. Finally, it is important to be willing to
the ERCP procedure. Alternatively, a needle-knife sphincterotomy stop a procedure if repeated attempts fail to lead to any forward
can be done over the transhepatic guidewire or catheter to create a progress.
space on the papilla for subsequent biliary cannulation. Yet another
way to complete a Rendezvous procedure is to slide a catheter directly
onto the free end of the guidewire that is barely exposed through the ERCP ACCESSORIES
papilla to perform a sphincterotomy, stricture dilation or stenting.
Finally, a two-step procedure that combines transhepatic placement Billroth II gastrectomy may be the most common surgically altered
of an 8 Fr external-to-duodenal biliary catheter with a subsequent anatomy encountered in an ERCP. All commercially available ERCP
ERCP session to cannulate the bile duct alongside the biliary drain accessories can be used; however, they may have to be modied in
has been described.28 order to gain access into the tangentially and upside-down oriented
bile duct (Fig. 24.27A). Sphincterotomes are also specially made to
accommodate the distorted anatomy (Fig. 24.27B). The techniques
Choosing an intestinal anastomotic opening of using these catheters and sphincterotomes are intuitive, but can
to enter be difcult with the rst few uses. In spite of the concern for severe
It is often confusing when the endoscope reaches an anastomosis pancreatitis,29 balloon sphincter dilation is an important therapeutic
with more than one exiting lumen. Additionally, correctly nding option if the endoscopist is unfamiliar with performing sphincter-
the afferent limb amongst several orices at an anastomosis is a otomy in this setting. In fact, a small study showed no disadvantage
difcult and time-consuming task. There is no quick solution other in employing balloon dilation to remove biliary stones in Billroth II
than examining the anatomy carefully and systematically. A thor- cases.9 When a colonoscope is used, most diagnostic and therapeutic
ough understanding of small bowel surgery may minimize the frus-
tration. Recognizing which lumen has been examined is crucial so
that there is no wasted time by entering the same lumen over and
over. Some endoscopists use India ink to mark the examined lumen,
whereas others take supercial biopsies or leave cautery marks. It is
important to know that there are either two or three lumens at any A
given anastomosis, depending on whether they are the results of a
side-to-side (3 lumens) or end-to side (2 lumens) reconstruction
(Figs 24.13AD). It is equally important to know that the afferent
limb is usually not a direct extension of the Roux limb. Therefore,
it is important to observe where the suture line is. The afferent limb
should be one of the lumens that crosses the suture line. An excep-
tion is when a Braun procedure is found but this should be sus-
pected whenever a second anastomosis in a series is noted. Since
many of these procedures have to be repeated in the future, it is B
important to construct a roadmap by carefully documenting how the
afferent limb and papilla or bilioenteric anastomosis are found.

Navigating through the small intestine

Passing through the small intestine is similar to navigating the
lumen of a tortuous colon; simply pushing an endoscope would not
get far. Whether it is using a side-viewing or end-viewing endoscope,
gentle rotation and intermittent shortening are essential maneuvers.
But handling a side-viewing duodenoscope requires very careful
manipulations to avoid causing perforation. When a duodenoscope
is used, the scope should constantly be adjusted so that the intestinal
lumen is situated at the 6 oclock position before it can be advanced
forward. Sudden rotation and forceful pulling should be avoided to
prevent trauma to the intestine. Minimizing air insufation is Fig. 24.27 A Three types of diagnostic instruments are commonly
used for cannulating the Billroth II bile duct. A tapered ERCP cathe-
equally important, as it reduces bowel tortuosity and allows stretch- ter bent into an S-shape is probably the most useful instrument.
ing room so there is less chance of traumatizing the intestinal wall. Some endoscopists prefer to use a catheter with a straight hydro-
When repeated passages result in paradoxical movements, hand philic wire. Still other endoscopists nd it helpful to start with a
compression of the abdomen may have positive effects similar to straight, uncurved cannula. B Sphincter-cutting accessories
include a push-type sphincterotome with a protruding loop when
maneuvers used in passing a colonoscope. Rotating the patient may the cutting wire is loosened, a sphincterotome with the cutting wire
straighten out a segment of the intestine and allow further passage. at the very tip, an S-shape sphincterotome, and the needle-knife.

256

ERCP accessories may be sufciently long to use. For instance, all
plastic biliary stenting can be carried out. On the other hand, most
biliary balloon dilators are too short. However, colonic balloon dila-
tors can be used instead. Some, but not all, sphincterotomes are
sufciently long for use through a colonoscope. Each endoscopy unit
should consult with its ERCP accessory manufacturers to put
together a complete set of supplies for this purpose.
It is far more challenging to nd the right accessories for use
through a push enteroscope or double-balloon enteroscope. With
lengths of 200 to 250 cm, these endoscopes are too long for most
commercially available ERCP instruments. Cook Medical Inc.
(Winston Salem, North Carolina) does make available a selection
of extra-long supplies that include diagnostic cannulas, retrieval
balloons, over-the-wire dilators and sphincterotomes. However,
exchanges of these instruments are more difcult than usual because
of added friction and inadequate lengths of guidewires. Even 450 cm-
long guidewires are too short to perform exchanges of 300 cm
length catheters. Since it is difcult to perform ERCP in a Roux-en-Y
anatomy, every effort has to be made to reduce the necessity of
REFERENCES
1. Briel JW, Tamhankar AP, Hagen JA, et al. Prevalence and risk
factors for ischemia, leak, and stricture of esophageal
anastomosis: gastric pull-up versus colon interposition. J Am Coll
Surg 2004; 198:536542.
2. Kim MH, Lee SK, Lee MH, et al. Endoscopic retrograde
cholangiopancreatography and needle-knife sphincterotomy in
patients with Billroth II gastrectomy: a comparative study of the
forward-viewing endoscope and the side-viewing
duodenoscope. Endoscopy 1997; 29:8285.
3. Faylona JM, Qadir A, Chan AC, et al. Small-bowel perforations
related to endoscopic retrograde cholangiopancreatography
(ERCP) in patients with Billroth II gastrectomy. Endoscopy 1999;
31:546549.
4. Feitoza AB, Baron TH. Endoscopy and ERCP in the setting of
previous upper GI tract surgery. Part I: Reconstruction without
alteration of pancreaticobiliary anatomy Gastrointestinal
Endoscopy 2001; 54:743749.
5. Lin LF, Siauw CP, Ho KS, et al. ERCP in post-Billroth II gastrectomy
patients: emphasis on technique. Am J Gastroenterol 1999;
94:144148.
6. Swarnkar K, Stamatakis JD, Young WT. Diagnostic and therapeutic
endoscopic retrograde cholangiopancreaticography after Billroth
II gastrectomysafe provision in a district general hospital. Ann R
Coll Surg Engl 2005; 87:274276.
7. Lee YT. Cap-assisted endoscopic retrograde
cholangiopancreatography in a patient with a Billroth II
gastrectomy. Endoscopy 2004; 36:666.
8. Hintze RE, Veltzke W, Adler A, et al. Endoscopic sphincterotomy
using an S-shaped sphincterotome in patients with a Billroth II or
Roux-en-Y gastrojejunostomy. Endoscopy 1997; 29:7478.
9. Bergman JJ, van Berkel AM, Bruno MJ, et al. A randomized trial of
endoscopic balloon dilation and endoscopic sphincterotomy for
removal of bile duct stones in patients with a prior Billroth II
gastrectomy. Gastrointest Endosc 2001; 53:1926.
10. Elton E, Hanson BL, Qaseem T, et al. Diagnostic and therapeutic
ERCP using an enteroscope and a pediatric colonoscope in long-
limb surgical bypass patients. Gastrointest Endosc 1998; 47:6267.
11. Law NM, Freeman ML. ERCP by using a prototype oblique-
viewing endoscope in patients with surgically altered anatomy.
Gastrointest Endosc 2004; 59:724728.
Chapter 24 ERCP in Surgically Altered Anatomy
repeating the procedure. Therefore, nasobiliary drainage catheters
are more likely to be used in this setting than ordinary ones.
CONCLUSION
Performing ERCP in a surgically-altered anatomy is a uniquely chal-
lenging experience. Preprocedure planning, stocking of appropriate
accessories and knowledge of surgical procedures are as important
as good endoscopic skills in achieving success. ERCP in Billroth II
predisposes to a high risk of perforation and should be done only by
those with a record of acceptable safety. Accessing the bile duct
through the afferent limb of a Roux-en-Y anatomy, once considered
an improbability, is increasingly accomplished by expert endosco-
pists. However, the bar of technical difculty has been raised by the
recent dramatic rise in bariatric surgery. Other than an occasional
case of favorably short Roux and afferent limbs, the gastric bypass
surgery makes ERCP far too challenging a procedure to perform.
Whether a double-balloon enteroscope offers a solution to this
problem remains to be seen.
12. Sato H, Tamada K, Kita H, et al. Application of double-balloon
endoscopy for afferent limb lesions of Roux-en-Y surgical
anastomoses. Gastrointest Endosc 2005; 61:AB238.
13. Byron E, Wright BE, Cass OW, et al. ERCP in patients with long-
limb Roux-en-Y gastrojejunostomy and intact papilla. Gastrointest
Endosc 2002; 56:225232.
14. Mergener K, Kozarek RA, Traverso LW. Intraoperative transjejunal
ERCP: case reports. Gastrointest Endosc2003; 58:461463.
15. Neligan PJ, Williams N. Nonsurgical and surgical treatment of
obesity. Anesthesiol Clin North America 2005; 23:501523, vii.
16. Poulose BK, Holzman MD, Zhu Y, et al. National variations in
morbid obesity and bariatric surgery use. J Am Coll Surg 2005;
201:7784.
17. DeMaria EJ, Jamal MK. Surgical options for obesity. Gastroenterol
Clin N Am 2005; 34:127142.
18. Baron TH, Vickers SM. Surgical gastrostomy placement as access
for diagnostic and therapeutic ERCP. Gastrointest Endosc 1998;
48:640641.
19. Pimentel RR, Mehran A, Szomstein S. Laparoscopy-assisted
transgastrostomy ERCP after bariatric surgery: case report of a
novel approach. Gastrointest Endosc 2004; 59:325328.
20. Stojadinovic A, Brooks A, Hoos A, Jaques DP, Conlon KC, Brennan
MF. An evidence-based approach to the surgical management of
resectable pancreatic adenocarcinoma. J Am Coll Surg 2003;
196:954964.
21. Aranha GV, Hodul, PJ, Creech S, et al. Zero Mortality after
152 Consecutive Pancreaticoduodenectomies with
Pancreaticogastrostomy. J Am Coll Surg 2003;
197:223231.
22. Sugiyama M, Abe N, Ueki H, et al. Pancreaticogastrostomy for
reconstruction after medial pancreatectomy. J Am Coll Surg 2004;
199:163165.
23. Steer ML. Townsend: Sabiston Textbook of Surgery, 17th edn,
2004 Saunders.
24. Frey CF, Amikura K. Local resection of the head of the pancreas
combined with longitudinal pancreaticojejunostomy in the
management of patients with chronic pancreatitis. Ann Surg 1994;
220:492507.
25. Caroli-Bosc FX, Demarquay JF, Peten EP, et al. Endoscopic
management of sump syndrome after
257
 SECTION 2 TECHNIQUES
choledochoduodenostomy: retrospective analysis of 30 cases.
Gastrointest Endosc 2000; 51:180183.
26. Fan ST, Mok F, Zheng SS, et al. Appraisal of hepaticocutaneous
jejunostomy in the management of hepatolithiasis. Am J Surg
1993; 165:332335.
27. Calvo MM, Bujanda L, Heras I, et al. The rendezvous technique for
the treatment of choledocholithiasis. Gastrointest Endosc 2001;
54:511513.
258
28. Dickey W. Parallel cannulation technique at ERCP rendezvous.
Gastrointest Endosc2006; 63:686687.
29. DiSario JA, Freeman ML, Bjorkrnan DJ, et al. Endoscopic balloon
dilation compared to sphincterotomy (EDES) for extraction of bile
duct stones: preliminary results. Gastrointest Endosc 1997; 45:
AB129.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Dilated Bile Duct
25 Geoffrey Spencer and Michael L. Kochman
BACKGROUND
The dilated bile duct is a commonly encountered phenomenon in
the daily practice of clinical medicine and is often generally felt to
represent a pathologic obstructive process limiting the ow of bile.
The indication and goal of any further evaluation would, therefore,
be to diagnose and treat an unrecognized obstructive lesion. Devel-
oping an approach to the evaluation of the dilated bile duct requires
denition of the normal limit of bile duct size, developing predic-
tive markers for an obstruction, and consideration of the accuracy
of the various imaging modalities for dening the etiology of an
obstruction.
Developing a denition for biliary dilation depends on the site of
measurement. Dilation may be present in the intra or extrahepatic
biliary systems or both. Assuming that this dilation is a manifesta-
tion of obstruction then a distal lesion would lead to diffuse dilation
as compared to a more proximal lesion giving rise to focal intrahe-
patic dilation. Intrahepatic ducts as small as 12 mm are seen as
scattered and non-conuent on abdominal computed tomography
or ultrasonography, but become conuent and more easily imaged
as they move centrally with diameters exceeding 2 mm in size (Fig.
25.1).1 Abnormal intrahepatic ducts are present when duct diameter
exceeds 40% of the diameter of the adjacent intrahepatic portal vein
and when they appear as parallel tubes coursing together.1
An increase in diameter of the extrahepatic bile ducts, in particu-
lar the common hepatic duct (CHD) or common bile duct (CBD), is
often referred to as biliary dilation. The normal size of the duct varies
at different levels, for example, at the conuence of the hepatic ducts
and at the union of the CBD and cystic duct, and it varies from person
to person.1,2 Numerous potential causes have been elucidated which
may affect the diameter of the extrahepatic ducts. First, the imaging
modality used to evaluate the biliary system can inuence the reported
duct diameter. Extracorporeal ultrasonography measures the inter-
nal diameter of the duct (Fig. 25.2).2,3 These measurements of the
CHD are typically obtained at the level of the hepatic artery in the
porta hepatis, anterior to the main or right portal vein or more distally
of the CBD.1,4 Most ultrasound (US) studies have placed the upper
limit of normal for the diameter of the CBD at 68 mm and that of
the CHD at 6 mm.1,47 However, on computed tomography (CT), it
is more common to accept values of 810 mm for the CBD.1,2 This
difference is in part due to measurements performed at different
locations along the duct. Unlike US, CT can more easily image the
mid to distal portions of the CBD which are often larger in diameter.1
It also more readily identies the fat around the duct and measure-
ments include the duct wall (Fig. 25.3). Evaluation of the biliary
system with cholangiogram, by endoscopic retrograde cholangiopan-
creatography (ERCP) or percutaneous transhepatic cholangiography
(PTC) may also yield results different than with other imaging tech-
niques. A study of 135 patients who underwent imaging of their
extrahepatic bile ducts with US and ERCP or PTC demonstrated duct
size of up to 4 mm by US versus 10.4 mm and 10.6 mm by ERC and
PTC respectively.8 This was likely a result of radiographic magnica-
tion of the cholangiogram and may also reect distention from con-
trast injection.
Measurements of bile duct caliber have not all been uniform even
when using the same imaging modality. For example, a study using
US to measure CBD diameters recorded normal values up to 8
10 mm in completely asymptomatic patients.5 This study noted a
trend towards increasing diameters in older-age individuals support-
ing a belief that the duct size may increase with age. This thought
has generated principles including adding 0.4 mm to the upper limit
of normal for duct size for each decade of life or 1 mm per decade
of life after age 60 years.7,9,10 However, a large US study of 1018
asymptomatic adults demonstrated a slight trend toward an increase
in duct size with age, but not as large as had been previously reported,
with a mean diameter of 3.6 mm at age 60 years and 4 mm at 85
years.7 In this study, 99% of the patients had a CBD diameter less
than 7 mm.
Patient characteristics other than age can affect the upper limit
of normal for the CBD. Since Oddi rst predicted the phenomenon
in 1887, dilation of the CBD after cholecystectomy has remained
controversial.11 Several prospective studies found no ductal dilation
on US in patients before and after cholecystectomy.12,13 However,
other studies have demonstrated a slight trend towards ductal dila-
tion post-cholecystectomy which was statistically signicant over
time.1417 In a study of 234 patients undergoing cholecystectomy, the
CBD increased after surgery from a mean of 5.9 mm before
cholecystectomy to 6.1 mm at an average of 393 days post-
cholecystectomy.17 As in other studies, the patients who had dilation
of the CBD after surgery, the increase in diameter was of the order
of 12 mm in the majority of patients and it has been suggested that
the upper limit of normal based on US be adjusted up to 8 mm in
such patients.15 In a study of 24 patients undergoing elective chole-
cystectomy, two asymptomatic patients with normal laboratory
results were noted to have an increase in ductal diameter of 4 mm
or more with duct diameters of >9 mm and >10 mm at 5 years post-
cholecystectomy.16 Though most patients have minor if any dilation
post-cholecystectomy, there are clearly those that may manifest
a more profound asymptomatic dilation of the ducts.
Variation of normal extrahepatic bile duct size due to extrinsic
factors such as time of day, respiration, or patient positioning, have
all been shown to affect the caliber of the CBD.1820 Given all the
possible circumstances that may affect the measurement of the
extrahepatic biliary system, it is difcult to dene an absolute mea-
surement that will by itself yield satisfactory predictive values for
obstruction as a cause of duct dilation. Instead, the duct diameter
should be interpreted in the context of potential causes of obstructive
biliary dilation so that any pertinent ndings from the clinical
263
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 25.1 Transabdominal ultrasound examination demonstrating
intrahepatic ductal dilation (arrowheads).
Fig. 25.2 Transabdominal ultrasound examination demonstrating
a dilated CBD (arrowheads) and dilated CHD (arrows).
presentation or biochemical tests may be considered in the decision
to pursue further diagnostic evaluation.
ETIOLOGY
Identifying the level of biliary obstruction is important in developing
a differential diagnosis. Common causes of obstruction include both
neoplastic and benign causes such as choledocholithiasis (Table
25.1).1,21,22 Less common in the United States, but more common
worldwide would be infectious etiologies such as parasitic disease.
A study on the use of EUS in the evaluation of a dilated biliary tree
in 90 patients with an unrevealing abdominal US found 40 patients
with choledocholithiasis, 13 with tumor, 8 with benign stricture, 2
choledochal cysts, 1 with infection with ascaris, and 24 dilated ducts
264
Fig. 25.3 Abdominal CT demonstrating a dilated CBD (arrow) in a
patient with choledocholithiasis.
without evidence of an obstructing lesion (Fig. 25.4).23 These
numbers are similar to other studies that have demonstrated the
general results in order of decreasing prevalence: choledocholithia-
sis, pancreatic cancer, ampullary carcinoma, and cholangiocarci-
noma as causes of biliary dilation.6 Assessing the a priori likelihood
of a particular disease given the clinical scenario should inuence
the choice of the subsequent diagnostic evaluation.
EVALUATION
Though there is a strong correlation between a dilated bile duct and
the presence of an obstruction, as previously discussed, there is a
broad interpretation of what is considered dilated. Many of the tech-
niques used for biliary imaging are based on anatomic measure-
ments and are not a functional demonstration of ow within the
ducts as may be evident on cholangiography or scintigraphy. It is
therefore important that decisions to proceed with further evalua-
tion include both a clinical and biochemical assessment for obstruc-
tion in addition to consideration of the initial imaging results
(Fig. 25.5).
Clinical evaluation
The clinical presentation should be assessed for any signs or symp-
toms relating to biliary obstruction or its cause. For example, the
history should elicit any symptoms such as abdominal pain, fever,
weight loss, jaundice, pruritus, acholic stools, dark urine, or steator-
rhea. The physical exam may be limited in its utility, but special
attention should be paid to the presence of an abdominal tenderness
or mass, hepatomegaly, jaundice, or lymphadenopathy. A positive
history or physical examination may serve to lower the threshold for
further diagnostic evaluation in those patients with an equivocal
duct size on initial imaging. In the era of laparoscopic cholecystec-
tomy, studies have attempted to develop models of these clinical

SITES OF BILIARY OBSTRUCTION
Intrahepatic Porta hepatis
Primary sclerosing cholangitis Cholangiocarcinoma
Space occupying liver lesion Primary sclerosing cholangitis
Gallbladder carcinoma
Hepatocellular carcinoma
Malignant lymph nodes
Liver metastases
Table 25.1 Common obstructive causes of biliary dilation
Fig. 25.4 Radial EUS examination demonstrating a longitudinal
view of a dilated CBD (arrows).
features in addition to biochemical values to predict choledocholi-
thiasis prior to surgery. Some of these studies have demonstrated
an increased likelihood for choledocholithiasis with jaundice or fever
at presentation, acholic stools, dark urine, or an older patient age.2427
However, it is difcult to draw a clear conclusion from these studies
as they differ in methodology and results.
Biochemical evaluation
Integral to the biochemical evaluation of obstruction are the biliru-
bin and liver associated enzymes (LAEs), commonly referred to as
liver function tests. These generally include alkaline phosphatase
(AP), alanine aminotransferase (ALT), and aspartate aminotransfer-
ase (AST).
The principal markers of cholestasis are bilirubin and AP.2830 The
total bilirubin present in the serum represents a balance between
input from production and hepatobiliary removal. In obstructive
jaundice, the serum bilirubin is principally in the conjugated form
(water soluble). Hepatobiliary AP is present on the apical membrane
of the hepatocyte and in the luminal bile duct epithelium. Increases
in AP result from increased synthesis and release into the serum. As
a result, levels may not rise until one to two days after biliary obstruc-
tion occurs. In addition, the enzyme has a half-life of a week and the
Chapter 25 Dilated Bile Duct
Suprapancreatic Intrapancreatic
Pancreatic carcinoma Pancreatic carcinoma
Cholangiocarcinoma Pancreatitis
Metastatic disease Choledocholithiasis
Direct extension of gastric/colon/ Ampullary stenosis/carcinoma
gallbladder
Carcinoma Duodenal carcinoma
Pancreatitis Cholangiocarcinoma
level may therefore remain elevated for several days even after the
resolution of biliary obstruction. Levels of AP up to three times
normal are relatively nonspecic and occur in a variety of liver dis-
eases. However, higher elevations are more specic for biliary
obstruction (intra- or extrahepatic) and inltrating liver diseases. As
AP can be produced in sources outside of the liver, it may be neces-
sary in certain instances to use other biochemical tests such as the
AP isoenzymes or the gamma-glutamyl transpeptidase or 5-nucleo-
tidase to conrm the hepatobiliary etiology of an elevated AP.
The serum aminotransferases include AST and ALT. Transient
elevations occurring rapidly (within one to two days) and with levels
into the thousands may occur in acute CBD obstruction, from
trauma or more commonly from choledocholithiasis, Subsequent
levels rapidly decline.2830 Aminotransferase levels may also rise
from other subacute or chronic obstructions but typically remain
less than 500 IU/dl.
It is difcult to interpret the predictive models using these
markers in evaluating obstruction from choledocholithiasis as the
results vary among studies.2427 In general, there is increased likeli-
hood for stones with abnormalities in bilirubin, AP, and transami-
nase levels. With these studies in mind, one can predict that it would
be unusual for a lesion to cause biliary obstruction and dilation
without clinical or biochemical abnormality. This is not universal,
however, and there have been case reports of patients with normal
LAEs despite having dilated ducts and choledocholithiasis.31
Imaging
Imaging of the biliary tract continues to evolve with the enhance-
ment of non-invasive techniques for cross-sectional evaluation and
biliary reconstruction. There are numerous radiographic and inva-
sive modalities now available to the clinician to image the anatomy
of the biliary system. These include US, CT and CT cholangiogra-
phy, magnetic resonance imaging (MRI) and magnetic resonance
cholangiopancreatography (MRCP), EUS, and ERCP. Each has
advantages and disadvantages as well as limitations in the evaluation
of the biliary system. The goal of any radiologic procedure evaluating
the dilated bile duct is to conrm the presence of obstruction and to
dene its location, extent, and cause.
Ultrasound
US is often the rst line imaging technique in the evaluation of the
bile ducts, gallbladder and right upper quadrant and is usually the
test initially demonstrating dilation of the bile ducts. US is non-
invasive, inexpensive and is a quick procedure that may be done at
the bedside. It does, however, require experience in technique and
interpretation and may be limited due to interference from gas
265
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

External bile duct diameter

Initial imaging Ultrasound MRI CT/Cholangiogram

Bile duct
measurement

<8 mm 8 mm 10 mm >10 mm

Biochemical/
clinical
assessment

() (+) () (+) () (+) () (+)

No further Indeterminate, Indeterminate, Further No further Indeterminate, Indeterminate, Further

evaluation obstruction possible. consider patient evaluation evaluation obstruction possible. consider patient evaluation
indicated. Follow labs and characteristics indicated. indicated. Follow labs and characteristics indicated.
imaging. such as age or imaging. such as age or
cholecystectomy. cholecystectomy.

Fig. 25.5 Algorithm for assessment of the external bile ducts for obstruction.

within the surrounding bowel. US has been shown to be one of the
most accurate imaging studies in the evaluation of cholelithiasis
with both a sensitivity and specicity of up to 99%.4,3234 It is also
highly sensitive for detecting dilation of the biliary tree as a whole
with a sensitivity greater than 90%.4,35 The ability of ultrasound to
dene the site and cause of biliary obstruction is slightly less reli-
able. A review of the literature with more than 700 patients demon-
strated that ultrasound has a sensitivity of 71% in delineating the
level of obstruction, a sensitivity of 57% in dening the etiology,
and a sensitivity for choledocholithiasis of 32%.32 However, there is
great variation with sensitivities ranging from 27% to 95% for the
correct level of obstruction and 2381% for the correct cause of
obstruction.3639 Some of this limitation may result from a relative
inability to image the distal CBD due to bowel gas. This area is well
visualized in only 4050% of patients.4 Clearly the ease of use,
widespread availability, and few contraindications place US early on
the algorithm for evaluation of the biliary tract or abdominal pain,
but it typically leads to further studies as it is often inconclusive and
does not provide adequate staging or surgical information in the
setting of suspected neoplasms.
Computed tomography
Similar to ultrasound, abdominal CT may be the initial test demon-
strating dilation of the bile ducts or can also be considered as a test
to further evaluate suspected biliary pathology.
Multidetector CT can obtain images at thin 1.252.5 mm inter-
vals that can be reformatted with high resolution into views to repro-
duce the biliary tree.1,40,41 This, in conjunction with careful review of
axial images, can provide a complete evaluation of the bile ducts.
Infusion of intravenous contrast agents is necessary to provide
vascular landmarks and organ opacication maximizing the vis-
ualization of the bile ducts.40 Unenhanced scans can highlight the
presence of calcications and aid in visualization of choledocholi-
thiasis. Water is often used as an oral contrast agent when biliary
tract abnormalities are suspected so as not to obscure potential
pathology at the level of the ampulla of Vater.40 CT cholangiography
has been evaluated in the United States and is used extensively in
Asia and Europe. This employs the administration of IV contrast
material to highlight the biliary tree. The fact that obstruction limits
contrast excretion into the bile ducts and the higher incidence
of adverse reactions to the contrast agents has limited the role of
CT cholangiography.42 However, with new contrast agents and
multidetector CT this may be the preferred imaging technique in
the future.
In a study evaluating the presence of biliary obstruction, dened
as extrahepatic bile duct diameter >8 mm, the sensitivity and speci-
city of CT for diagnosing dilated ducts was 96% and 91% respec-
tively.40 CT is also accurate at dening the level of obstruction in
8897% of cases as well as the cause of obstruction in 7095% of
cases.39,40,43,44
Though CT is a readily available test that can accurately identify
a dilated CBD as well as provide important details as to the level and
cause of obstruction, there are limitations as well. It requires intra-
venous contrast to optimize the images, which may lead to adverse
reactions including potential nephrotoxicity. CT also lacks sensitivity
for a common cause of obstruction, choledocholithiasis, which is
likely responsible for the lower rates of detection of the etiology of
obstruction; approximately 2025% of biliary stones are isoattenuat-
ing with bile making them difcult to detect on CT.1,40 Sensitivity of
CT for choledocholithiasis ranges from 70% to 94% depending on
the use of indirect signs of obstruction that typically coincide with
choledocholithiasis.1,45,46
Magnetic resonance imaging
Since it was rst introduced in 1991, MRCP has gained popularity
as a non-invasive imaging modality of the biliary system. MRCP

266

Fig. 25.6 MRCP image demonstrating markedly dilated bile ducts
proximal to a Klatskin-type bile duct tumor.
exploits the differences between uid lled structures in the abdomen
and adjacent soft tissues. The static or slow-moving uid within the
pancreatic and biliary system produces a different signal than solid
tissue.21,22 Images can be obtained without use of IV contrast and
are routinely performed in the axial and coronal planes.
MRCP has a high accuracy in detecting the level and cause of
biliary obstruction (Fig. 25.6). It has been shown to have a sensitivity
of 9197% and specicity of up to 99% for the diagnosis of biliary
obstruction.4749 Unlike US and CT, which can be limited in their
ability to correctly identify the level and etiology of an obstructed
biliary system, MRCP can more accurately dene these parameters,
similar to the use of direct cholangiography. The level of obstruction
can be determined with MRCP in 8798% of cases.47,48,50,51 The etiol-
ogy of obstruction was determined in roughly 84% of cases with a
malignant process and 94% in cases of a benign process.47,48 This
was corroborated by a large meta-analysis demonstrating a sensitiv-
ity and specicity of MRCP of 92% and 97% respectively for chole-
lithiasis and 88% and 95% respectively for malignant causes (Fig.
25.7).47 However, the reported accuracy in discerning benign from
malignant obstruction has varied from 30% to 98% across studies,
with a mean accuracy of 88% reported in the meta-analysis. Addition
of conventional T1 and T2 weighted images to MRCP allows for
evaluation of extraductal soft tissue increasing the diagnostic accu-
racy by demonstrating tumor extension, lymph nodes, or metastatic
disease.51,52 In one study, this increased the sensitivity, specicity,
and accuracy 1720% for differentiation of benign and malignant
causes of biliary obstruction.53 This, however, comes with increased
cost and time of exam.
The major advantages of MRCP over other imaging techniques
include the avoidance of invasive procedures, IV contrast, ionizing
radiation, and the ability to visualize the biliary system above and
below an obstruction. MRCP does have limitations. Its non-invasive
nature means an inability to perform therapeutic intervention. Tech-
Chapter 25 Dilated Bile Duct
Fig. 25.7 MRCP demonstrating choledocholithiasis (arrow).
nical and interpretive difculties can simulate or miss pathologic
conditions of the biliary system. Static images demonstrating a con-
traction on the distal CBD can simulate a stenosis for example.54
Other drawbacks include the high cost of the test, extended time to
perform an exam leading to patient intolerance, and contraindica-
tions such as magnetic objects.
Endoscopic ultrasound
EUS has emerged as a signicant advance in gastrointestinal endos-
copy since its introduction in 1987. It allows for both diagnostic
evaluation of the pancreaticobiliary system with high resolution
images as well as for therapeutic procedures. EUS is performed uti-
lizing a specialized endoscope with a radial or linear ultrasound
transducer at the tip.55,56 Images of the biliary system are obtained
from transgastric or transduodenal locations. Unlike transabdomi-
nal ultrasound, EUS allows for better visualization of the extrahe-
patic biliary tree without interference of bowel gas as the CBD passes
posterior to the duodenal bulb. Additionally, EUS offers accurate and
systematic visualization of the wall of the duodenum including the
papillary region (Fig. 25.8).
EUS has been studied extensively in a variety of disorders that
can lead to dilation of the CBD. In the setting of choledocholithiasis,
EUS has consistently demonstrated sensitivities of >90% and speci-
cities up to 100%.5759 EUS may be particularly useful for detecting
microlithiasis (stones <3 mm) though such small stones would be
less likely to lead to biliary obstruction and dilation (Fig. 25.9).60 The
close proximity of the stomach and duodenum to the pancreas
allows for sensitive evaluation of pancreatic lesions such as neo-
plasms and cyst formation which may lead to obstruction. EUS has
demonstrated greater sensitivity in the detection of pancreatic carci-
noma with sensitivities in the range of 90% and above when com-
pared to other imaging modalities (Fig. 25.10).6163 This superiority
is particularly evident with respect to smaller lesions (<3 cm). EUS
267
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 25.8 Radial EUS demonstration of ampullary adenoma
(arrows).
Fig. 25.9 Radial EUS demonstration of microlithiasis
(arrowheads).
has the further advantage of allowing for biopsy and diagnosis.
Overall, EUS has been shown to provide an accurate explanation of
biliary dilation in greater than 90% of patients.23
The limitations of EUS for the evaluation of biliary dilation
include poor visualization of obstructive lesions located more proxi-
mally in the biliary system such as in the hilum or in the right
hepatic duct. Poor visualization of the distal CBD also occurs when
the pancreas is markedly calcied, during an episode of acute pan-
creatitis, when there is altered anatomy such as previous gastric
surgery or presence of air within the biliary tract. Complications
from diagnostic EUS are uncommon with a major complication rate
of around 0.5%.64 This rate is the same as upper endoscopy with
similar risks of perforation, bleeding, and sedation.
268
Fig. 25.10 Radial EUS demonstration of a pancreatic head carci-
noma (arrowheads).
Cholangiography
Cholangiography is generally performed using ERCP, but can be
performed percutaneously (PTC). Each route provides both an
anatomic view of the bile ducts as well as a functional assessment
of whether bile can freely drain through the ducts. With the
advent of recent advances in biliary imaging, particularly MRCP,
the diagnostic indications for ERCP have been signicantly chal-
lenged. For example, detection rates for choledocholithiasis using
ERCP, MRCP, and EUS are comparable at over 90%; though
microlithiasis may be masked by contrast medium and better
imaged with EUS.22,60 In the diagnosis of pancreatic cancer, ERCP
was shown to have a sensitivity of 70% versus 84% with MRCP
(Fig. 25.11).51 Evaluation of perihilar biliary obstruction, the most
common location for cholangiocarcinoma, with both ERCP and
MRCP demonstrated to be very effective in detecting the presence
of obstruction with a sensitivity of 100%.65 However, MRCP was
superior in its investigation of the cause and anatomic extent of
disease when compared with ERCP in part because it displayed
the biliary tree cephalad to the obstruction and the character of
the intraductal lling defect. A substantial downside to perform-
ing ERCP in patients with suspected hilar cholangiocarcinoma
is the introduction of contamination into several segments of
the biliary tree which, if left undrained, may lead to severe chol-
angitis. However, ERCP remains the best means of diagnosing
ampullary cancers due to direct visualization and the ability to
biopsy.34,56
Common complications of ERCP include pancreatitis with a rate
of 17%, as well as hemorrhage, perforation, and infectious com-
plications (see Chapter 6).66 It is therefore recommended that ERCP
be reserved for patients with a reasonable likelihood of need for
therapeutic intervention.66,67 For example, retrieval of stones and
clearance of the duct is successful in over 90% of cases and in the
setting of acute suppurative cholangitis can improve the clinical
course and be life-saving.56 ERCP can also be used for palliation of
obstructive lesions in nonsurgical patients or for obtaining tissue
for diagnosis.

Fig. 25.11 ERCP demonstration of a double-duct sign in pancre-
atic carcinoma with proximal biliary dilation (arrows) and dilated
pancreatic duct (arrowheads).
Fig. 25.12 PTC demonstration of a proximal bile duct stricture in
a patient with hilar cholangiocarcinoma. The arrows demonstrate
the proximal extent of the tumor in the right and left intrahepatic
ducts and the arrowheads demonstrate the distal extent of the
tumor in the CHD.
PTC is also an invasive technique of imaging the biliary system.
It requires insertion of a needle into a dilated bile duct followed by
opacication of the bile ducts with an injection of contrast. PTC has
close to 100% sensitivity and specicity for identifying the site and
cause of biliary obstruction (Fig. 25.12).68 However, this procedure
is invasive and requires sedation. Complication rates vary with
patient status but include major complications such as sepsis, chol-
angitis, bile leak, hemorrhage, or pneumothorax at a rate of 2%.69
Chapter 25 Dilated Bile Duct
When a drainage procedure is indicated these rates increase to 2.5%
for sepsis, 2.5% for hemorrhage, 1.2% for infection, and 1.7% for
death. With the advent of enhanced imaging techniques, PTC should
be employed for patients who warrant therapeutic biliary interven-
tion, but are not candidates for ERCP or who have failed endoscopic
biliary access.
Biliary scintigraphy
Scintigraphy employs the administration of intravenous radioiso-
topes to evaluate the biliary system. The radioisotopes are taken up
by the hepatocytes and excreted into the bile with concentration in
the gallbladder. The sensitivity and specicity for biliary obstruction
in jaundiced patients are 97% and 89%, respectively, based on the
duration of time until the appearance of radiotracer in the duode-
num.70 This test may help elucidate if obstruction is present in the
setting of a dilated bile duct, but little if any information is gained
regarding the etiology of obstruction.
APPROACH TO THE PATIENT WITH A
DILATED DUCT
Developing an algorithm on the approach to the dilated bile duct
must incorporate the many variables described to this point. The
rst step in this process is to determine the clinical probability that
an underlying obstructive lesion is present that requires further
evaluation. This decision is based on the anatomic information from
the initial imaging study and the clinical and biochemical assess-
ment of the patient.
US and CT are common initial radiographic tests demonstrating
a bile duct abnormality. Dilation of intrahepatic ducts will become
evident when the diameter exceeds 40% of the adjacent intrahepatic
portal vein and the ducts become conuent throughout the liver. If
isolated dilation of all or a portion of the intrahepatic ducts cannot
be explained based on patient characteristics described above it war-
rants further investigation. The approach to dilation of the extrahe-
patic ducts is more challenging. The heterogeneity of results in
dening an upper limit of normal for duct size makes it difcult to
assign a value representing pathologic or obstructive dilation. Patient
characteristics such as age and history of cholecystectomy as well as
imaging modality must be considered when dening an abnormal
duct size. On US, extrahepatic ducts with diameters greater than
7 mm can be considered abnormal since it has been shown that 99%
of the non-obstructed population has ducts equal to or less than
7 mm. Values up to 10 mm should be considered normal for CT.
However, outliers exist with larger ducts and no pathologic obstruc-
tion. Therefore, duct size must be correlated with clinical and bio-
chemical features suggesting obstruction. In the absence of any such
abnormalities, it is unlikely that obstruction is present even when
duct diameter if greater than 7 mm. In these patients, most likely
no further evaluation is necessary. It is important to note that
obstruction may still be present with an extrahepatic duct measuring
7 mm or less. Dilation may not yet have occurred in response to
obstruction or the size represents a change from a previously smaller
duct. Again clinical and biochemical features may be helpful in
determining an underlying process.
Once it is determined that dilation of a bile duct likely represents
obstruction, the next decision is how to proceed with further evalu-
ation (Fig. 25.13). Again this approach must be individualized based
on the clinical scenario with consideration of the underlying etiol-
ogy. Generally, the algorithm should start with less invasive tests.
269
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Probable biliary obstruction
Symptoms of cholangitis
(+)
ERCP or PTC for
biliary drainage
Further therapeutic or
diagnostic procedures
Cause still unknown,
consider for imaging
* Choice of CT over MRI/MRCP based on contraindications to MR procedure,
patient characteristics and availability.
Fig. 25.13 Algorithm for the evaluation of an obstructed bile duct.
However, patients presenting with obstruction and signs and symp-
toms of cholangitis may require diagnostic and therapeutic inter-
vention acutely, in which case ERCP is an appropriate initial
procedure. Patients without emergent indication for biliary drainage
will require cross-sectional imaging with either a CT, if not per-
formed initially, or an MRI/MRCP. Availability of testing, cost,
patient characteristics, and probability of the underlying cause must
be considered when choosing between these tests. Scant useful
direct comparative data exist when comparing these tests and
attempting to control for etiology, but some trends can be observed.
Considerations in the decision process include contraindications to
IV CT contrast, potential intolerance of MRI, and presence of metal
objects which preclude MRI. CT has a high accuracy for dening
many of the common etiologies of biliary obstruction, but has a
REFERENCES
1. Baron RL, Tublin ME, Peterson MS. Imaging the spectrum of
biliary tract disease. Radiol Clin North Am 2002; 40(6):13251354.
2. Adkins RB Jr, Chapman WC, Reddy VS. Embryology, anatomy, and
surgical applications of the extrahepatic biliary system. Surg Clin
North Am 2000; 80(1):363379.
3. Sauerbrei EE, et al. The discrepancy between radiographic and
sonographic bile-duct measurements. Radiology 1980;
137(3):751755.
4. Cohen SM, Kurtz AB. Biliary sonography. Radiol Clin North Am
1991; 29(6):11711198.
5. Wu CC, Ho YH, Chen CY. Effect of aging on common bile duct
diameter: a real-time ultrasonographic study. J Clin Ultrasound
1984; 12(8):473478.
6. Niederau C, et al. Extrahepatic bile ducts in healthy subjects,
in patients with cholelithiasis, and in postcholecystectomy
patients: a prospective ultrasonic study. J Clin Ultrasound 1983;
11(1):2327.
7. Perret RS, Sloop GD, Borne JA. Common bile duct measurements
in an elderly population. J Ultrasound Med 2000; 19(11):727730;
quiz 731.
270
()
MRI/MRCP or CT*
EUS/FNA ERCP/brushing or biopsy
lower sensitivity for choledocholithiasis. MRCP with cross-sectional
imaging has the highest sensitivity for dening obstructing lesions
and should be the test of choice. Though higher in cost, a missed
diagnosis with CT would lead to further testing and more expense.
Based on the ndings from subsequent imaging, more invasive
procedures such as EUS or ERCP may be needed to obtain tissue
for denitive diagnosis.
In conclusion, the approach to the patient with a dilated bile duct
is less than straightforward. First, the clinician must predict the
possibility of an obstructing lesion that requires further evaluation
based on duct size, clinical presentation, and biochemical assess-
ment. Next further imaging modalities must be chosen based on
both test and patient characteristics. Finally, denitive diagnosis
with biopsy and treatment can be considered.
8. Niederau CA, Sonnenberg A, Mueller J. Comparison of the
extrahepatic bile duct size measured by ultrasound and by
different radiographic methods. Gastroenterology 1984;
87(3):615621.
9. Mahour GH, Wakim KG, Ferris DO. The common bile duct in
man: its diameter and circumference. Ann Surg 1967;
165(3):415419.
10. Rumack CM, Wilson SR. The gallbladder and bile ducts. 2nd edn
1998; St. Louis: Mosby-Yearbook:206207.
11. Oddi R. Dune disposition a sphincter speciale de ouverture du
canal choledoque. Arch Ital Biol 1887; 8:317322.
12. Graham MF, et al. The size of the normal common hepatic duct
following cholecystectomy: an ultrasonographic study. Radiology,
1980; 135(1):137139.
13. Mueller PR, et al. Postcholecystectomy bile duct
dilatation: myth or reality? AJR Am J Roentgenol 1981;
136(2):355358.
14. Chung SC, Leung JW, Li AK. Bile duct size after cholecystectomy:
an endoscopic retrograde cholangiopancreatographic study. Br J
Surg 1990; 77(5):534535.

15. Reinus WR, et al. Ultrasound evaluation of the common duct in
symptomatic and asymptomatic patients. Am J Gastroenterol
1992; 87(4):489492.
16. Hunt DR, Scott AJ. Changes in bile duct diameter after
cholecystectomy: a 5-year prospective study. Gastroenterology
1989; 97(6):14851488.
17. Feng B, Song Q. Does the common bile duct dilate after
cholecystectomy? Sonographic evaluation in 234 patients. AJR Am
J Roentgenol 1995; 165(4):859861.
18. Wachsberg RH. Respiratory variation of extrahepatic bile duct
diameter during ultrasonography. J Ultrasound Med 1994;
13(8):617621.
19. Lorenzon G, Corsi M. [Ultrasonic study of the dimensions of the
common bile duct in various postures]. Radiol Med (Torino) 1990;
80(4):455462.
20. Raptopoulos V, et al. Daytime constancy of bile duct diameter.
AJR Am J Roentgenol 1987; 148(3):557558.
21. Siegelman EAS, Body MRI. 1st ed. 2005. Philadelphia: Elsevier:
6377.
22. Soto JA, Barish MA, Ferrucci JT. Magnetic resonance imaging of
the bile ducts. Semin Roentgenol 1997; 32(3):188201.
23. Songur YG, Temucin G, Sahin B. Endoscopic ultrasonography in
the evaluation of dilated common bile duct. J Clin Gastroenterol,
2001; 33(4):302305.
24 Prat F, et al. Prediction of common bile duct stones by
noninvasive tests. Ann Surg 1999; 229(3):362368.
25. Barkun AN, et al. Useful predictors of bile duct stones in patients
undergoing laparoscopic cholecystectomy. McGill Gallstone
Treatment Group. Ann Surg 1994; 220(1):3239.
26. Abboud PA, et al. Predictors of common bile duct stones prior to
cholecystectomy: a meta-analysis. Gastrointest Endosc 1996;
44(4):450455.
27. Trondsen E, et al. Prediction of common bile duct stones prior to
cholecystectomy: a prospective validation of a discriminant
analysis function. Arch Surg 1998; 133(2):162166.
28. Limdi JK, Hyde GM. Evaluation of abnormal liver function tests.
Postgrad Med J 2003; 79(932):307312.
29. Green RM, Flamm S. AGA technical review on the evaluation of
liver chemistry tests. Gastroenterology 2002; 123(4):13671384.
30. Reddy KR. Hepatobiliary tract and pancreas. 1st ed. The requisites
in gastroenterology, ed. A Rustgi 2004; Philadelphia: Mosby: 117.
31. Goldman DE, Gholson CF. Choledocholithiasis in patients with
normal serum liver enzymes. Dig Dis Sci 1995; 40(5):10651068.
32. Blackbourne LH, et al. The sensitivity and role of ultrasound in
the evaluation of biliary obstruction. Am Surg 1994;
60(9):683690.
33. Harvey RT, Miller WT Jr. Acute biliary disease: initial CT and follow-
up US versus initial US and follow-up CT. Radiology 1999;
213(3):831836.
34. Stroszczynski C, Hunerbein M. Malignant biliary obstruction: value
of imaging ndings. Abdom Imaging 2005; 30(3):314323.
35. Laing FC, et al. Biliary dilatation: dening the level and cause by
real-time US. Radiology 1986; 160(1):3942.
36. Haubek A, et al. Dynamic sonography in the evaluation of
jaundice. AJR Am J Roentgenol 1981; 136(6):10711074.
37. Koenigsberg M, Wiener SN, Walzer A. The accuracy of
sonography in the differential diagnosis of obstructive jaundice: a
comparison with cholangiography. Radiology 1979;
133(1):157165.
38. Honickman SP, et al. Ultrasound in obstructive jaundice:
prospective evaluation of site and cause. Radiology 1983;
147(2):511515.
39. Baron RL, et al. A prospective comparison of the evaluation of
biliary obstruction using computed tomography and
ultrasonography. Radiology 1982; 145(1):9198.
Chapter 25 Dilated Bile Duct
40. Baron RL. Computed tomography of the bile ducts. Semin
Roentgenol 1997; 32(3):172187.
41. Kim HC, et al. Three-dimensional reconstructed images using
multidetector computed tomography in evaluation of the biliary
tract. Abdom Imaging 2004; 29(4):472478.
42. Stockberger SM, Sherman S, Kopecky KK. Helical CT
cholangiography. Abdom Imaging 1996; 21(2):98104.
43. Pedrosa CS, et al. Computed tomography in obstructive jaundice.
Part II: The cause of obstruction. Radiology 1981; 139(3):635645.
44. Pedrosa CS, Casanova R, Rodriguez R. Computed tomography in
obstructive jaundice. Part I: The level of obstruction. Radiology
1981; 139(3):627634.
45. Neitlich JD, et al. Detection of choledocholithiasis: comparison of
unenhanced helical CT and endoscopic retrograde
cholangiopancreatography. Radiology 1997; 203(3):753757.
46. Baron RL. Common bile duct stones: reassessment of criteria for
CT diagnosis. Radiology 1987; 162(2):419424.
47. Romagnuolo J, et al. Magnetic resonance
cholangiopancreatography: a meta-analysis of test performance
in suspected biliary disease. Ann Intern Med 2003;
139(7):547557.
48. Magnuson TH, et al. Utility of magnetic resonance
cholangiography in the evaluation of biliary obstruction. J Am
Coll Surg 1999; 189(1):6371; discussion 7172.
49. Guibaud L, et al. Bile duct obstruction and choledocholithiasis:
diagnosis with MR cholangiography. Radiology 1995;
197(1):109115.
50. Georgopoulos SK, et al. Comparison of magnetic resonance
and endoscopic retrograde cholangiopancreatography in
malignant pancreaticobiliary obstruction. Arch Surg 1999;
134(9):10021007.
51. Adamek HE, et al. Pancreatic cancer detection with magnetic
resonance cholangiopancreatography and endoscopic
retrograde cholangiopancreatography: a prospective controlled
study. Lancet 2000; 356(9225):190193.
52. Barish MA, Soto JA. MR cholangiopancreatography: techniques
and clinical applications. AJR Am J Roentgenol 1997;
169(5):12951303.
53. Kim MJ, et al. Biliary dilatation: differentiation of benign from
malignant causesvalue of adding conventional MR imaging
to MR cholangiopancreatography. Radiology 2000;
214(1):173181.
54. David V, et al. Pitfalls in the interpretation of MR
cholangiopancreatography. AJR Am J Roentgenol, 1998; 170(4):
10551059.
55. Ingram M, Arregui ME. Endoscopic ultrasonography. Surg Clin
North Am 2004; 84(4):10351059, vi.
56. Ahmad NA, Shah JN, Kochman ML. Endoscopic ultrasonography
and endoscopic retrograde cholangiopancreatography imaging
for pancreaticobiliary pathology: the gastroenterologists
perspective. Radiol Clin North Am 2002; 40(6):13771395.
57. Canto MI, et al. Endoscopic ultrasonography versus
cholangiography for the diagnosis of choledocholithiasis.
Gastrointest Endosc 1998; 47(6):439448.
58. Palazzo L, Levy P, Bernades P. Usefulness of endoscopic
ultrasonography in the diagnosis of choledocholithiasis. Abdom
Imaging 1996; 21(2):9397.
59. Amouyal P, et al. Diagnosis of choledocholithiasis by endoscopic
ultrasonography. Gastroenterology 1994; 106(4):10621067.
60. Liu CL, et al. EUS for detection of occult cholelithiasis in patients
with idiopathic pancreatitis. Gastrointest Endosc 2000; 51(1):2832.
61. Akahoshi K, et al. Diagnosis and staging of pancreatic cancer by
endoscopic ultrasound. Br J Radiol 1998; 71(845):492496.
62. Muller MF, et al. Pancreatic tumors: evaluation with endoscopic
US, CT, and MR imaging. Radiology 1994; 190(3):745751.
271
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
63. Snady H, Cooperman A, Siegel J. Endoscopic ultrasonography
compared with computed tomography with ERCP in patients with
obstructive jaundice or small peri-pancreatic mass. Gastrointest
Endosc 1992; 38(1):2734.
64. Yusuf TE, Bhutani MS. Role of endoscopic ultrasonography in
diseases of the extrahepatic biliary system. J Gastroenterol
Hepatol 2004; 19(3):243250.
65. Yeh TS, et al. Malignant perihilar biliary obstruction: magnetic
resonance cholangiopancreatographic ndings. Am J
Gastroenterol 2000; 95(2):432440.
66. Mallery JS, et al. Complications of ERCP. Gastrointest Endosc 2003;
57(6):633638.
272
67. NIH state-of-the-science statement on endoscopic retrograde
cholangiopancreatography (ERCP) for diagnosis and therapy. NIH
Consens State Sci Statements 2002; 19(1):126.
68. Gold R, Casarella WJ, Stern G. Transhepatic cholangiography: The
radiologic method of choice in suspected obstructive jaundice.
Radiology 1979; 133:3944.
69. Burke DR, et al. Quality improvement guidelines for percutaneous
transhepatic cholangiography and biliary drainage. J Vasc Interv
Radiol 2003; 14(9 Pt 2):S243246.
70. Lee AW, et al. Technetium-99m BIDA biliary scintigraphy in the
evaluation of the jaundiced patient. J Nucl Med 1986;
27(9):14071412.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Ampullary Neoplasm
26 Ellert J. van Soest and Paul Fockens
INTRODUCTION
Malignant tumors of the ampulla of Vater are rare. One person per
80.000 inhabitants older than 40 years will be diagnosed with ampul-
lary cancer each year in the United States.1 Fifteen times more
prevalent than ampullary cancer is the heterogeneous group of
malignancies of the periampullary region, which include mainly
tumors of the pancreas, but also less prevalent tumors of the distal
bile duct, the periampullary duodenum, and the ampulla of Vater.
Prior to surgical resection, the origin of these tumors is often dif-
cult to differentiate. However, survival rates after Whipples resec-
tion differ signicantly between the different periampullary cancers,
in favour of ampullary tumors (Fig. 26.1).2 This chapter will focus
on neoplasms arising from the ampulla of Vater itself.
Adenoma is the most common benign ampullary tumor. Since
an adenoma-carcinoma sequence has been documented at this level,
all adenomas must be resected upon their detection. In short, many
adenomas are suitable for endoscopic therapy (see also Chapter 19),
whereas adenocarcinomas should primarily be resected surgically.
SYMPTOMS AND SIGNS
Patients with ampullary tumors frequently develop symptoms early
in the course of the disease, when the tumor is still relatively small.
These early symptoms are caused by its typical anatomical location
at the junction of the bile duct and pancreatic duct, impeding biliary
and/or pancreatic outow. Jaundice is the presenting symptom in
the majority of patients. Unlike the jaundice observed with cancer
of the pancreas, jaundice produced by ampullary cancers may uctu-
ate, especially early in the course of the obstructive process. In
addition, nonspecic symptoms such as weight loss, abdominal
discomfort, nausea and vomiting are often seen. Obstructive jaun-
dice, anemia due to gastrointestinal blood loss, and a palpable, non-
tender gallbladder (Courvoisier sign) make up the classic triad of an
ampullary carcinoma. Acute pancreatitis and cholangitis can occur
as the result of obstruction of the pancreatic or bile duct.
Benign ampullary tumors present less often with jaundice,
reecting the smaller tumor size and the absence of tissue invasion.
Nondescript abdominal pain is common. An increased incidence of
common duct calculi, observed in both benign and malignant
tumors of the ampulla, due to bile stasis, may account for some of
the abdominal pain and jaundice observed with these lesions. A
signicant subset of ampullary adenomas are found during surveil-
lance for adenomas in patients with Familial Adenomatous Polypo-
sis (FAP), or as an incidental nding at upper gastrointestinal
endoscopy in patients with symptoms that are not attributable to the
ampullary lesion. In a relatively large series of 55 ampullary adeno-
mas, 45% of patients were asymptomatic, abdominal pain was seen
in 16%, jaundice in 15%, pancreatitis in 9%, and miscellaneous
symptoms were present in 15%.3
Presentation with severe acute gastrointestinal bleeding is rare,
but has been described with adenomas, carcinomas and mesenchy-
mal tumors of the ampulla, when the overlying mucosa has become
eroded.4
There are no specic laboratory ndings that accompany ampul-
lary tumors. Usually, because of disturbed biliary outow, alkaline
phosphatase and bilirubin are elevated; sometimes elevated amino-
transferases and iron deciency anemia are seen. Abnormal liver
function tests can be the only presenting feature. Tumor markers
are used mostly as a prognostic rather than a diagnostic tool. In a
study of 56 patients with ampullary carcinoma, CA 19-9 had a sen-
sitivity of 78% in detecting cancer, and CEA only 33%. Specicity of
both CA 19-9 and CEA is low.5 Gene expression analysis of ampul-
lary adenocarcinoma has identied other tumor markers, such as
serum osteopontin, which in the future may also aid in the early
detection and differential diagnosis of patients with periampullary
lesions.6
DIAGNOSTIC WORK-UP AND STAGING
Endoscopy
Endoscopy for the assessment of an ampullary tumor should be done
with a high quality side-viewing duodenoscope. Regular forward-
viewing endoscopy will miss abnormalities of the papilla in a signi-
cant number of cases. Endoscopy visualizes the site of the lesion, its
size, and macroscopic appearance, as well as extension beyond the
limits of the ampulla. More important, the value of endoscopy lies
in its ability to obtain tissue for histology. A wide variation exists in
the endoscopic appearance of the normal major duodenal papilla.
Recognition of adenomas can therefore be difcult.
Grossly, ampullary tumors may manifest in one of four well-
described appearances (Figs 26.2 and 26.3):7
Macroscopically normal papilla: suspicion of an ampullary process
arises when unexplained common bile duct (and/or pancreatic
duct) dilatation to the level of the ampulla is seen on CT scan.
Completely intra-ampullary tumors may only become apparent
after endoscopic sphincterotomy, but can often be visualized with
endoscopic ultrasound (EUS) nowadays.
Intramural protrusion: A bulge underneath a normal-appearing
papilla (prominent infundibulum).
Exposed protrusion: Neoplastic-appearing tissue extending out from
an otherwise normal or abnormal-appearing papilla.
Ulcerating tumor: This situation is very suspicious for malignancy
(Fig. 26.5a).
A characteristic appearance distinguishing benign from malig-
nant disease is the presence of an ulcerating tumor mass. Failure to
achieve a cleavage plane with submucosal injection during ampul-
273
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B
100

90

80

70

60
Survival (%)

50

40
C D
30
Ampulla
20 Duodenum
Distal bile duct
10 Pancreas

0
0 6 12 18 24 30 36 42
Time after diagnosis (months)

No. at risk
Ampulla 88 78 67 57 49 45 37 32
Distal bile duct 41 30 17 13
Duodenum 13 7 5 3
Pancreas 419 207 95
Fig. 26.2 Different appearances of ampullary adenomas.
A Macroscopic almost normal papilla, containing adenoma at
biopsy in a FAP-patient. B Adenoma with granular surface
C Exposed protrusion of villous adenoma. D Large polypoid
mass at the papilla.
Fig. 26.1 Survival curves for 561 patients with periampullary
tumors, stratied by site of origin (Redrawn from reference no. 2
with the permission of John Wiley & Sons Ltd on behalf of the
British Journal of Surgery Society Ltd.)

A B C

Fig. 26.3 Adenocarcinoma of the ampulla of Vater. Although multiple biopsies yielded only adenoma, the ampullectomy specimen
showed adenocarcinoma. Patient underwent a pancreaticoduodenectomy afterwards.

lectomy is also a strong predictor of malignancy.8 Other signs that
indicate malignancy are induration and friability with easy bleeding.
Ulceration located on the roof of the ampulla, separated from the
papilla by normal mucosa, indicates a lesion invading the duodenal
submucosa.9
In a series of 52 patients with biopsy-proven ampullary adenomas
or carcinomas, Ponchon et al. noted a normal endoscopic appearance
of the papilla in 37% of patients. In these patients the intra-ampullary
tumors only became apparent after endoscopic sphincterotomy.7
Most sporadic lesions are solitary, while there is almost always
macroscopic evidence of adenomatous transformation elsewhere in
the duodenum in FAP-patients.
ERCP
Since the introduction of MRCP and EUS, ERCP has limited
application as a diagnostic test because of its associated mor-
bidity and possible mortality. However, in case of an ampullary
tumor, ERCP is performed when biliary drainage is required.

274

Sphincterotomy can be performed to biopsy deeper within the
ampulla. Furthermore, ERCP is helpful when performing an ampul-
lectomy, to establish the presence or absence of intraductal tumor
in the bile- or pancreatic duct. Some centers use cholangioscopy for
this purpose.10 Other uoroscopic ndings of ampullary cancer at
ERCP include bile- and pancreatic duct dilatation in most cases, a
defect in the ampulla, delayed drainage of contrast, and common
bile duct stones in a minority of patients.
Forceps biopsy
Endoscopic forceps biopsy specimens, taken from the surface of a
tumor, can be falsely negative for malignancy in a considerable
number of patients. Carcinomatous changes inside adenomas are
found in 2530%, and percentages up to 56% have been reported
in older series, depending on the size of the adenoma.11,12 Despite
these limitations, recent endoscopic series show an acceptable per-
centage of only 68% carcinomas in ampullectomy specimens in
patients who were treated for presumed benign ampullary tumors.10,13
This is probably the result of improved patient selection before per-
forming an ampullectomy. To improve diagnostic yield, it is essen-
tial to obtain a large number of biopsies (at least six) from an
ampullary neoplasm to minimize sampling error. Even then, nega-
tive biopsies do not rule out the presence of cancer. Larger biopsies
with a diathermic snare improve the quality of samples, but also
carry more risk of complications. In some patients, sphincterotomy
with deeper biopsies from within the incised papilla may be consid-
ered. Biopsies are more reliable when obtained 10 days or more after
sphincterotomy, as sphincterotomy can cause coagulation effects
that can be misinterpreted.14 Endoscopic ultrasound-guided ne-
needle aspiration biopsy (EUS-guided FNAB) seems to be accurate
in the assessment of suspected primary ampullary masses. In the
only published report on this subject, the reported sensitivity and
specicity were 82% and 100%, respectively. However, adenomas
were underrepresented in this study (2 out of 35 patients), so the
role of EUS-guided FNAB for ampullary adenomas needs to be
determined.15
Transabdominal ultrasound, CT and MRI
Obstructive jaundice is usually initially analysed with transabdomi-
nal ultrasound (US) or a CT scan. If biliary obstruction is attributable
to an ampullary neoplasm, dilatation of the entire biliary tree is
typically demonstrated. The tumor itself is not seen in the majority
of cases, however. Sensitivity of CT for detecting ampullary tumors
is only around 2030%, but increases with larger tumor size.1618 If
no mass is depicted with CT, endoscopic ultrasound should be con-
sidered (see below). In the setting of ampullary carcinoma with a
clear mass, CT plays an important role in predicting resectability, by
depicting both vessel inltration and distant metastasis.19 MRI with
MRCP has not been studied extensively for ampullary tumors, but
its role in periampullary tumors is promising.20
Endoscopic ultrasonography (EUS)
When evaluating patients with ampullary pathology, EUS can be
used for different purposes.
Diagnosing an ampullary tumor
EUS can be used in the primary detection of an ampullary tumor in
patients with obstructive jaundice, without a mass on US or CT.
Studies in patients with ampullary cancer have clearly shown a
higher sensitivity of EUS when compared to US and CT for the
Chapter 26 Ampullary Neoplasm
detection of ampullary lesions (respectively 95100% versus 15%
and 2030%).1618 However, lesions of 10 mm or smaller can be dif-
cult to demonstrate with EUS.17
Secondly, EUS can be used in the assessment of the cause of an
endoscopically abnormal looking ampulla. EUS should ideally be
able to differentiate an ampulla with a (pre-) malignant lesion in it,
from other conditions, such as a an enlarged inammatory papilla
or normal-variant protruding papilla. From the sparse literature on
this subject it seems that the specicity of EUS in the diagnosis of
an ampullary mass is less impressive than its sensitivity, and lies
arround 75%.21 A false positive result is primarily caused by a swollen
papilla as a consequence of stone migration (Fig. 26.4). In summary,
EUS should be considered an important tool in clarifying the cause
of an abnormal papilla, but only biopsies can reliably conrm the
diagnosis.
Staging advanced ampullary cancer
EUS is helpful with staging of ampullary cancers, according to the
TNM system (Table 26.1). Pretherapeutic staging is essential for
making therapeutic decisions and to determine prognosis of the
patient. EUS has been proven to be the most reliable modality for
preoperative T-staging, compared to CT and MRI.18,22 In the largest
series of 50 consecutive patients with ampullary neoplasms, EUS
was found to be more accurate (78%) than CT (24%) and MRI (46%)
compared with surgical-pathologic staging in the assessment of the
T stage. The main limitation of EUS in T-staging is its relative inac-
curacy in differentiating stage T2 from T3. Overstaging of true T2
carcinomas can be caused by peritumoral pancreatitis. The presence
A B
C D
Fig. 26.4 Enlarged inammatory papilla (odditis) due to
gallstones. A Swollen papilla (with a stent in situ), suspicious for
a tumor with intramural protrusion. B Therefore a sphincterot-
omy was performed to biopsy deeper. Biopsies showed only
inammation. C Gallstones were removed. D Normalization of
the papilla two months later.
275
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
T1 Tumor limited to the ampulla of Vatera
T2 Tumor invades duodenal wall (muscularis propria)
T3 Tumor invasion into the pancreas <2 cm
T4 Tumor invasion into the pancreas >2 cm or
into adjacent organs or vessels
N0 No regional lymph node metastasis
N1 Regional lymph node metastasis
M0 No distant metastasis
M1 Distant metastasis
Table 26.1 TNM staging system for ampullary carcinomas
a
Some divide T1 tumors in d0 (limited to Oddis sphincter) and d1 (tumor invading
the duodenal submucosa).
Sobin LH, Wittekind Ch (eds) International union against cancer (UICC): The TNM
classication of malignant tumors. 6th ed. New York: Wiley; 2002.
A B
C Ingrowth in the common bile- and pancreatic duct is regarded as
D
Fig. 26.5 T2N1 adenocarcinoma of the ampulla of Vater.
A Endoscopic appearance. B EUS showing the polypoid
tumor, C a dilated common bile duct and pancreatic duct, D and
a lymph node metastasis.
of endobiliary stents also may decrease the accuracy of EUS in
staging ampullary tumors, mainly by understaging T2/T3 carcino-
mas. Nevertheless, this differentiation is not mandatory as the same
surgical treatment is applied in T2 and T3 tumors.
No signicant difference in N-stage accuracy has been noted
between CT, MRI and EUS. Metastatic regional lymph nodes are
detected by EUS with an accuracy of around 65% (Fig. 26.5).22,23
Locoregional lymph node involvement usually does not inuence
the surgical treatment plan. Currently, suspicion of lymph node
metastasis should always be conrmed by EUS-guided FNAB when
treatment decisions are dependent on it.
Resectability of ampullary tumors depends mainly on vessel
involvement, and on distant metastasis. Since ampullary cancer
causes symptoms early in the course of the disease and originates a
276
relatively long distance from the mesenteric and other major vessels,
vascular involvement is uncommon. EUS can obtain adequate
visualization of the portal venous system from the the duodenal
bulb, and celiac vessels can be scanned from the gastric fundus. EUS
is a very sensitive modality in demonstrating vascular tumor
involvement.24,25
Liver metastasis and ascites can be evaluated with EUS, but
distant metastasis should be exluded with CT scan as well, before
considering a surgical resection.
Staging early ampullary lesions
When local endoscopic or surgical resection is considered, EUS T-
staging becomes of vital importance. Endoscopic resection is mainly
reserved for benign lesions since early cancers already have a 10%
risk of lymph node metastases and should be treated with radical
surgery (pancreaticoduodenectomy). EUS cannot distinguish
between benign adenoma and ampullary cancer, unless inltrative
growth exists, such as inltration in the duodenal muscularis
propria. In case of benign biopsies, exclusion of inltrating tumors
deeper in the lesion is the primary goal of EUS. The accuracy of EUS
to assess that the T stage > T1 is high, and has been reported to be
between 87 and 94%.17,22,24,26 Subdivision of stage T1 in d0 and d1
further renes staging. T1d0 tumors are limited to Oddis sphincter
and T1d1 tumors invade the duodenal submucosa. Local resection
of ampullary tumors is appropriate when there are no signs of
malignancy macro- and microscopically, and staging is T1d0. For
this subdivision intraductal ultrasonography (IDUS) seems to be the
best imaging modality (see below).
a (relative) contraindication for endoscopic treatment of ampullary
adenomas, a condition that can be visualized by either EUS or
IDUS.
Intraductal ultrasound (IDUS)
IDUS is a promising imaging modality in staging ampullary tumors.
Ultrasound catheter probes are inserted during ERCP and advanced
over a guidewire into the bile duct and/or pancreatic duct. Intraductal
catheter probes have a higher frequency (20 MHz) than standard
EUS, resulting in better resolution. In the largest prospective study
on this subject, IDUS was signicantly superior to EUS and CT in
terms of tumor visualization, especially in depicting small lesions.17
IDUS is the only modality that distinguishes the sphincter of Oddi
as a distinct layer.26 Potentially, IDUS could be performed in every
patient with seemingly benign tumors of the ampulla. The decision
to perform endoscopic resection would then be taken only if no
submucosal ingrowth or intraductal extension was detected with
IDUS. Larger prospective series are needed to validate this policy.
Colonoscopy
There are reports in the literature that the risk of colonic adenoma
and carcinoma is increased in patients with adenomas of the ampulla,
not only as part of hereditary polyposis syndromes.27,28 In one study
of 16 patients with adenomas of the ampulla, 10 colonoscopies were
performed, and colonic adenomas were found in four cases, and one
patient turned out to have FAP.29 The group of Saurin and Ponchon
has also described positive ndings in approximately half the colo-
noscopies that they performed in 13 patients with sporadic ampul-
lary adenomas, including one patient who had eight colonic
adenomas, and one patient found to have a sigmoid carcinoma.30
Although these studies are small, the signicant percentage of

detected colon adenomas/carcinomas mandate a colonoscopy in all
patients with sporadic adenomas of the ampulla.
PATHOLOGY
Only 5% of all tumors of the ampulla are not adenomas or adeno-
carcinomas. These 5% of cases comprise a wide range of benign and
malignant disease. Most ampullary tumors develop sporadically, and
some occur in connection with cancer syndromes, as summarized
in Table 26.2.
Adenoma
The ampulla of Vater is a complex anatomic site, and is composed
of the common channel, the intraduodenal portion of the common
bile duct and pancreatic duct, and duodenal mucosa. The common
channel is probably the site from which most ampullary tumors
derive, as shown in histologic studies and autopsy series.31 A patho-
physiological role of bile and pancreatic juice in its carcinogenesis
has been hypothesised, but, to date, has not been convincingly
demonstrated.
Ampullary adenomas are histologically very similar to their
colonic counterparts and are classied as tubular, tubulovillous, or
villous, in order of increasing malignant potential. As in the colon,
progressive dysplastic changes may occur, including eventual trans-
formation to cancer. This adenoma-carcinoma sequence is sup-
ported by the following observations:
In patients who had been operated on for benign ampullary tumors
in the past, a signicant percentage presented later with ampullary
cancer.7
Adenomatous remnants, with histological transitions of adenoma
into carcinoma, have been found in the vicinity of ampullary
adenocarcinoma.7,32,33
Histological progression has been demonstrated during follow-up
of ampullary adenomas in FAP patients.34
A genetic alteration model delineating ampullary tumor develop-
ment has been described, and is similar to that seen in colonic
carcinogenesis.35
Some cancers have no detectable adenomatous component.
Either tumor extension has already destroyed the adenomatous
tissue, or certain cancers develop without going through an adenoma
stage. The nding in autopsy series that atypical epithelium occurs
without adenomatous growth supports the latter theory.31
Carcinoma
Carcinoma of the ampulla can be classied histopathologically and
immunohistochemically as either an intestinal or pancreaticobiliary
FAP
HNPCC
Neurobromatosis type 1
Muir-Torre syndrome
Table 26.2 Cancer syndromes associated with ampullary
carcinoma
Hereditary nonpolyposis colorectal carcinoma (HNPCC) is weakly associated with
ampullary carcinoma.
Neurobromatosis seems to be a predisposition of both somatostatinomas and
carcinomas of the ampulla of Vater.
Muir-Torre syndrome is a condition characterized by the association of multiple
sebaceous tumors and kerato-acanthomas with internal malignancies, including
ampullary carcinomas.
Chapter 26 Ampullary Neoplasm
type.36 The intestinal type resembles tubular adenocarcinoma of
the colon and originates from the ampulloduodenal mucosa; the
pancreaticobiliary type originates from the common channel, or
from bile- or pancreatic duct epithelium. Carcinomas of the intestinal
and pancreaticobiliary types may develop via different mechanisms,
and demonstrate different molecular biological characteristics.37
There are conicting reports whether survival differs between
types; some suggest survival is better in the intestinal type.32,36
Neuroendocrine tumors (NET)
For many years, disseminated neuroendocrine neoplasms of the
gastrointestinal tract have been subsumed as carcinoids. Cur-
rently, the World Health Organization (WHO) recommends using
the term neuroendocrine tumor. The WHO distinguishes (i) well-
differentiated neuroendocrine tumors (carcinoids); (ii) well-differen-
tiated neuroendocrine carcinomas (malignant carcinoids); and (iii)
poorly differentiated neuroendocrine carcinomas.
These tumors are morphologically and biologically diverse, with
a broad spectrum of subtypes. Ampullary NETs are rare. Between
100 and 200 cases have been reported in the international literature.
NETs that arise at the ampulla of Vater are mainly somatostatino-
mas, but gangliocytic paragangliomas and other NETs can occur also
(Figs 26.6 and 26.7). The majority of ampullary NETs are non-
functional: there is production of peptides when examined immu-
nocytochemically, but plasma hormone levels are not elevated and
they do not produce a clinical syndrome.38
There is a strong relationship with neurobromatosis (von
Recklinghausen disease): approximately one quarter of all reported
patients with ampullary NETs had neurobromatosis.39
Jaundice is the predominant presenting symptom. The correct
diagnosis is frequently made only after operation, because the sub-
mucosal location hinders adequate tissue sampling with forceps
biopsies.40 In contrast to carcinoid tumors at other gastrointestinal
sites, the tumor size of ampullary carcinoids does not correlate with
either the metastatic potential or prognosis.39,40 For assessment of
tumor extension, the same imaging modalities are used as described
above for ampullary adenocarcinomas. In addition, somatostatin
receptor (octreotide) scintigraphy can be helpful.38 Pancreati-
coduodenectomy is the standard treatment for ampullary, well-
differentiated carcinoids >2.0 cm and for ampullary neuroendocrine
Fig. 26.6 Carcinoid at the papilla.
277
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B A B

C D Fig. 26.8 Metastasis of a melanoma to Vaters ampulla. A Pa-

tient with a history of a melanoma was admitted because of chol-
angitis. ERCP showed a bulging papilla. B Easy bleeding during
cannulation. Endoscopic ultrasound-guided ne-needle aspiration
biopsy conrmed the suspicion of a metastasis of melanoma.

Benign disease
Adenoma
Carcinoid
GIST
Lipoma
Fig. 26.7 Periampullary gangliocytic paraganglioma. A Large, Leiomyoma
stalked polyp. B Indentication of the orice of the papilla. Hamartoma (Peutz-Jeghers polyp)
C Cannulation. D Snare resection just under the papilla, after
Schwannoma
placing an endoloop.
Lymphangioma
Hemangioma
Fibroma
Neurobroma
Granular cell tumor
Adenomyoma
carcinomas.40 Local surgical, or endoscopic resection can be consid-
Eosinophilic gastroenteritis
ered for smaller carcinoids without evidence of metastases. Gener- Duodenal duplication
ally, the prognosis of ampullary carcinoids after resection is good, Choledochocele
with a ve-year survival period of 90%.40 Poorly differentiated neuro- Heterotopic pancreas
endocrine carcinomas have an aggressive clinical behaviour, however, Heterotopic gastric mucosa
with early metastases, and often, a fatal outcome. Brunners gland hyperplasia
Inammatory nonneoplastic lesions: odditis/papillitis (e.g. due to
Lymphoma lithiasis)
Lymphomas of the biliary system are rare, but should be considered Large, but normal variant ampulla
in the differential diagnosis of malignant strictures of the common
bile duct. The ampulla is even more rarely the site of origin of these Table 26.3 Differential diagnosis of tumors and pseudotumors of
the ampulla of Vater
tumors. Four types of primary lymphomas of the ampulla of Vater
have been described:
Primary ampullary MALT lymphoma arise from the mucosa-
associated lymphoid tissue of the duodenum. Similar to MALT Gastrointestinal stromal tumor (GIST)
lymphomas in the stomach, ampullary MALT lymphomas are There are few reports of gastrointestinal stromal tumors of the
sometimes controlled by eradication of Helicobacter pylori. In other ampulla. Most patients have been treated with a pancreaticoduode-
cases, tumor regression is accomplished only with more aggressive nectomy. Treatment with imatinib mesylate has shown to improve
forms of therapy, such as radio- or chemotherapy.41 outcome in unresectable, metastatic or recurrent disease.45,46
Diffuse largeB-cell lymphoma. Together with MALT lymphomas,
diffuse large B-cell lymphomas seem to be the most prevalent Metastasis
lymphomas in the periampullary region. Their absolute incidence Metastasis to the ampulla is extremely unusual. It may arise from a
is low, however.42 variety of primary lesions including renal cell carcinomas, melano-
Follicular lymphomas are infrequently limited to the ampulla. mas, and ovarian-, breast-, esophagus- and larynx cancer (Fig. 26.8).47
Surgery can be avoided, if a proper diagnosis is made before opera-
tion, as chemotherapy is reported to be effective.43 Miscellaneous
T-cell lymphomas very rarely inltrate the sphincter of Oddi and The ampulla of Vater can harbour a variety of other neoplasms, and
cause jaundice.44 Association with celiac sprue has been described. an extended overview is given in Tables 26.3 and 26.4.

278
 Chapter 26 Ampullary Neoplasm

Malignant disease effectively prevents cancer, it is questionable whether all stage IV

patients should be exposed to this procedure in view of its consider-
(Adeno)carcinomaa able morbidity, and mortality.52 An alternative would be to follow
Neuroendocrine carcinoma
these patients regularly with upper endoscopy. Unfortunately, endo-
Malignant GIST
scopic surveillance seems not to detect all carcinomas, as has been
Lymphoma
Pancreatoblastoma shown in the study by Bjrk et al.51 This was potentially related to
Leiomyosarcoma the use of forward-viewing, rather than forward- and side-viewing
Neurobrosarcoma endoscopes. Alternatively, endoscopic biopsies may not have been
Kaposi sarcoma sensitive enough, or disease may have progressed rapidly between
Angiosarcoma screening intervals.
Malignant schwannoma In spite of these limitations, upper gastrointestinal endoscopic
Rhabdomyosarcoma (only described in children) surveillance is recommended for all FAP patients, with special atten-
Metastasis to the ampulla tion for the periampullary region. Adenomas of the major duodenal
papilla are more likely to undergo malignant transformation than
Table 26.4 Differential diagnosis of malignant tumors and an adenoma arising elsewhere in the duodenum, reecting the fact
pseudotumors of the ampulla of Vater
a
Mixed cellular populations of carcinoma have been described, such as:
that the occurrence of dyplasia in the periampullary region has been
Sarcomatoid carcinomas, an intermixture of carcinomatous and sarcomatous
reported to be as high as 6674%.27,34,49 Surveillance intervals of 3
elements. years are recommended by experts for less severe disease; those with
Adenocarcinoid tumors, an intermixture of adenocarcinoma and carcinoid stage IV disease should be examined every 6 months to 1 year.53
tumor
Ampullary carcinomas with unusual patterns of differentiation, such as papillary Biopsies should be taken routinely, also from normal appearing
carcinomas, Paneth cell carcinomas and signet-ring cell carcinomas are also papillae, as a normal appearing papilla can harbour an adenoma in
included. more than 50% of cases (Fig. 26.2A).34 The role of endoscopic treat-
ment of ampullary lesions in FAP is not yet clearly dened. Random-
ized trials of the different surgical and endoscopic modalities are
lacking. Endoscopic therapy has not been studied in large prospec-
Points 1 2 3
tive studies yet, but the available literature on this subject seems
Polyp number 14 520 >20 promising.13,54 Endoscopic treatment of (peri) ampullary lesions in
Polyp size (mm) 14 510 >10 FAP does not differ from therapy for sporadic ampullary lesions,
Histology Tubular Tubulovillous Villous although one should take into account that recurrence of adenomas
Dysplasia Mild Moderate Severe
is more common in FAP patients.13
Table 26.5 Staging system for the severity of duodenal polyposis
in FAP: the Spigelman classication (Reproduced from reference
no. 49 with permission.) TREATMENT
0 points = Stage 0; 14 = Stage I; 56 = Stage II; 78 = Stage III; 912 = Stage IV
Optimal therapy results in a radical excision of all neoplastic tissue,
minimizes the chance of recurrence, and has an acceptable proce-
Familial adenomatous polyposis (FAP) dure-related morbidity and mortality.

FAP syndrome and its variants (Gardners syndrome and Turcot
syndrome) afict approximately 1 in 20 000 people (data from
Denmark).48 Duodenal cancer, mainly in a (peri) ampullary location,
is the leading cause of cancer death in patients with FAP who have
undergone prophylactic colectomy.48 Patients with FAP have a
cumulative lifetime risk of over 90% for developing duodenal adeno-
mas, and a lifetime risk of 410% for developing periampullary or
duodenal adenocarcinoma.4851 The risk of adenocarcinoma of the
major duodenal papilla has been estimated to be greater than 100
times that of the general population.50 In an attempt to prevent
cancer, screening programmes have been developed using a well-
dened staging system to detect those patients most at risk, as
shown in Table 26.5.
Patients with stage IV disease have a 1030 times higher chance
of developing carcinoma, as compared to patients at a lower stage.51,52
How to treat patients with Spigelman stage IV remains controver-
sial. Around 25% of patients with stage IV periampullary adenomas
developed cancer in a large Scandinavian study.51 It is our impres-
sion however, that the use of current high resolution endoscopes
and chromoendoscopy leads to upstaging of patients in the Spigel-
man classication. This upstaging will probably diminish the cancer
risk in stage Spigelman IV. Although (pancreatico) duodenectomy
Adenomas
Pancreaticoduodenotomy has been the traditional treatment
of ampullary adenomas for years. This procedure effectively
eliminates all adenomatous tissue, but has been associated with
considerable morbidity (2565%), and with mortality (010%). Two
strategies have been developed to diminish complications: transduo-
denal local surgical resection (surgical ampullectomy) and endo-
scopic therapy. Postoperative morbidity is denitely lower with
local surgical resection (025%), and mortality is low to absent
(05%), but recurrence of adenomas occurs in 530% of operated
patients. Endoscopic surveillance is therefore indicated after this
type of surgery.12,55,56
Since 1983, reports have been published on the endoscopic treat-
ment of ampullary adenomas, using snare resection and laser pho-
tocoagulation.5759 Later, with the results of snare removal of the
entire ampulla as a single specimen, described by Binmoeller, it
became clear that endoscopic snare resection in selected patients
provides an excellent alternative to traditional surgery (Figs 26.9
and 26.10) (for indications, see Chapter 19).60 Controlled studies
comparing endoscopic and surgical strategies have not (yet) been
conducted. The published reports on this subject are case series
from surgical and endoscopic centers, and vary considerably in

279
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B A B

C D C D

Fig. 26.9 En bloc resection of a tubulovillous adenoma. A Peri- Fig. 26.10 Piecemeal resection of a large tubulovillous adenoma.
ampullary polypoid mass. B Indentication and cannulation of A Large polyp with orice of the papilla on top. B Twisted stalk.
the papilla. C Snare resection. D Result after ampullectomy, C Piecemeal resection. D Result after two months.
and sphincterotomy of the common bile duct.

100
patient selection and demographics, thus making comparison dif-
cult. Table 26.6 summarizes the literature on endoscopic therapy. 90
The choice on how to treat ampullary adenomas is highly dependent 80
on local expertise. In our opinion, because of its favourable outcome,
its low morbidity and absent mortality in most case series, endo- 70
scopic snare resection as the initial treatment is justied in centers
60
Survival (%)

with experience in this type of endoscopic interventions. An addi-

tional advantage is that this treatment can be performed on an out- 50
patient basis.
40
Carcinomas 30
Standard surgical management of invasive ampullary carcinoma is
resection by pancreaticoduodenectomy. Local surgical therapy for 20 T1
pT1 carcinoma is questionable, as the resection is often limited,61 T2
10 T3
and a reduced survival has been reported compared to pancreatico-
duodenectomy.62 Conversion to pancreaticoduodenectomy is indi- 0
0 6 12 18 24 30 36 42
cated when pathological examination of an ampullectomy specimen
Time after resection (months)
identies invasive carcinoma. The subsequent operation does not
seem to increase morbidity when compared to initial pancreatico- No. at risk
duodenectomy.63 For high grade dyplasia/carcinoma in situ, radi- T1 18 17 17 15 13 12 11 10
T2 32 30 28 24 23 18 16 18
cally removed endoscopically or with local surgical excision, there is
T3 20 17 12 11 11 8 5
insufcient evidence that a subsequent operation (i.e. lymphadenec-
tomy) is benecial. Lymph node metastasis in such cases seems to
be absent.64 A wait and see policy may be justied, but opinions
are divided on this issue.
Survival after resection in a relatively recent, large, single center Fig. 26.11 Survival curves for 70 patients after resection of ampul-
study is summarized in Figure 26.11. Operative mortality has dimin- lary carcinoma, stratied by tumor stage. Patients with T3 tumors
had poorer survival than those with T1 or T2 disease (P = 0.05).
ished over time, with no postoperative death in this and in other (Redrawn from reference no. 2 with the permission of John Wiley
studies.2 Factors predictive of improved survival in ampullary carci- & Sons Ltd on behalf of the British Journal of Surgery Society Ltd.)

280
 Inclusion: Studies published from 1990, and with more than 5 patients included, on endoscopic treatment of ampullary tumors with benign features.
Complication rate (percentage per total number of patients): Only clinically evident bleeding that occurred after completion of the procedure was regarded as a complication.
Pancreatitis and perforation were managed conservatively in the majority of patients.
Complete endoscopic resection (percentage per total number of patients): total clearance of adenoma in one or more treatment sessions, without the need for surgery (this includes
recurrences that could be treated endoscopically).
Surgery (percentage per total number of patients) includes the need of surgery for malignant disease, for persistent adenoma, for complications, and for recurrences that could not be
treated endoscopically.
Recurrence: recurrence of adenoma (or adenocarcinoma) after complete endoscopic resection (patients lost to follow-up are excluded).
NR: Not reported
Number of Complete
First Year of patients Retrospective/ Intraductal Follow- endoscopic
author publication included Prospective Technique growth Complications Histology up resection Surgery Recurrence
Katsinelos 2006 n = 14 Retrospective Snare excision 0 Pancreatitis 7%, bleeding 7% 11 adenoma, 28 79% 21% 18%
3 carcinoma months
Harewood 2005 n = 19 Prospective Snare excision 0 Pancreatitis 33% without pd NR NR NR NR NR
(RCT of stent, vs 0% with pd stent.
prophylactic Cholestasis 5%
pancreatic stent
placement)
Bohnacker 2005 n = 106 Prospective Snare excision 31 Pancreatitis 12%, bleeding 3% 92 adenoma 43 73% 19% 18%
(109 and/or (18 with months
lesionsb) electrocoagulation HGD), 4
carcinoma, 12
inammatory
lesion, 1
lymphangioma
Han 2005 n = 22 Retrospective Snare excision 0 Bleeding 5%, perforation 5%, 15 adenoma 8 86% NR 5%
papillary stenosis 5%, (3 with months
cholangitis 5%, cholestasis 5% HGD), 2
carcinoma, 1
carcinoid, 3
inammatory
lesion, 1
lymphangioma

Table 26.6 Summary of the literature on endoscopic resection of ampullary tumors

a
Median number of 3 therapeutic sessions to achieve complete destruction of adenomatous tissue.
b
Synchronous tumors of the major and minor papillae.
Chapter 26 Ampullary Neoplasm

281
282
Moon 2005 n=6 Prospective Snare excision 0 Late onset pancreatitis 17%, 6 adenoma (1 7 100% 0% 0%
(wire-guided cholangitis 17% with HGD) months
endoscopic
papillectomy).
Cheng 2004 n = 55 Retrospective 6 Pancreatitis 9%, bleeding 7%, 45 adenoma 30 74% 13% 33%
perforation 2% (7 with months
HGD), 5
carcinoma, 2
carcinoid, 1
gastric
heterotopia, 2
normal
histology
Catalano 2004 n = 103 Retrospective Snare excision 0 Pancreatitis 5%, bleeding 2%, 97 adenoma 36 80% 16% 20%
papillary stenosis 3% (14 with months
HGD), 6
SECTION 3 APPROACH TO CLINICAL PROBLEMS

carcinoma
Saurin 2003 n = 24 Retrospective Mainly laser 0 Pancreatitis 17%, bleeding Forcepsbiopsies: 81 67% 17% 6%
photodestructiona 13%, perforation 4% 24 months
adenoma (10
with HGD)
Norton 2002 n = 26 Retrospective Snare excision 0 Pancreatitis 15%, perforation 25 adenoma, 9 96% 4% 10%
(28 4%, pancreatic duct stenosis 1 carcinoma, months
lesionsb) 8% 1
inammatory
lesion, 1
normal
histology
Desilets 2001 n = 13 Retrospective Snare excision 0 Pancreatitis 8% 13 adenoma 19 92% 8% 0%
(piecemeal)a (1 with HGD) months
Zdarov 2001 n = 16 Retrospective Snare excision NR Pancreatitis 13%, bleeding 13% 16 adenoma NR 100% 6% 19%
Vogt 2000 n = 18 Retrospective Snare excision NR Pancreatitis 11%, bleeding 11%, 18 adenoma 75 100% 17% 44%
stent dysfunction 6% months
Park 2000 n=6 Retrospective Snare excision NR Pancreatitis 33% 4 adenoma, 2 21 67% 17% 0%
carcinoma months
Binmoeller 1993 n = 25 Prospective Snare excision 2 Pancreatitis 12%, bleeding 8% 25 adenoma 37 92% 12% 26%
(1 with HGD) months

Table 26.6 Continued


noma vary per study, and include resectability, negative margins (R0
resection), negative lymph nodes, differentiation of the tumor, and
absence of pancreatic, vascular and perineural invasion.2,61
Palliative treatment for ampullary carcinoma can consist of a local
resection in very frail patients with a small, pT1 tumor.64 Literature
on endoscopic resection of this kind of tumor is only anecdotal.65
Furthermore, biliary- or pancreatic drainage can be achieved with
ERCP by placing an endoprosthesis or by papillotomy. In case
of duodenal obstruction an expandable metal stent can be placed
endoscopically to palliate the gastric outlet obstruction.
CONCLUSIONS
Ampullary tumors are rare lesions, but of interest. Their prognosis
is generally better than that of pancreatic tumors and their treatment
REFERENCES
1. Urbach DR, Swanstrom LL, Khajanchee YS, et al. Incidence of
cancer of the pancreas, extrahepatic bile duct and ampulla
of Vater in the United States, before and after the introduction
of laparoscopic cholecystectomy. Am J Surg 2001; 181(6):
526528.
2. Bettschart V, Rahman MQ, Engelken FJ, et al. Presentation,
treatment and outcome in patients with ampullary tumours. Br J
Surg 2004; 91(12):16001607.
3. Cheng CL, Sherman S, Fogel EL, et al. Endoscopic snare
papillectomy for tumors of the duodenal papillae. Gastrointest
Endosc 2004; 60(5):757764.
4. Nowak A, Botdys H, Marek T, et al. Unusual cause of severe
upper gastrointestinal hemorrhage treated using a simple endo-
loop technique. Endoscopy 2001; 33(8):733.
5. Kau SY, Shyr YM, Su CH, et al. Diagnostic and prognostic values of
CA 19-9 and CEA in periampullary cancers. J Am Coll Surg 1999;
188(4):415420.
6. Van Heek NT, Maitra A, Koopmann J, et al. Gene expression
proling identies markers of ampullary adenocarcinoma. Cancer
Biol Ther 2004; 3(7):651656.
7. Ponchon T, Berger F, Chavaillon A, et al. Contribution of
endoscopy to diagnosis and treatment of tumors of the ampulla
of Vater. Cancer 1989; 64(1):161167.
8. Kahaleh M, Shami VM, Brock A, et al. Factors predictive of
malignancy and endoscopic resectability in ampullary neoplasia.
Am J Gastroenterol 2004; 99(12):23352339.
9. Napolon B, Pialat J, Saurin JC, et al. Adenomas and
adenocarcinomas of the ampulla of Vater: endoscopic therapy.
Gastroenterol Clin Biol 2004; 28(4):385392.
10. Bohnacker S, Seitz U, Nguyen D, et al. Endoscopic resection of
benign tumors of the duodenal papilla without and with
intraductal growth. Gastrointest Endosc 2005; 62(4):551560.
11. Yamaguchi K, Enjoji M, Kitamura K. Endoscopic biopsy has limited
accuracy in diagnosis of ampullary tumors. Gastrointest Endosc
1990; 36(6):588592.
12. Cahen DL, Fockens P, de Wit LT, et al. Local resection or
pancreaticoduodenectomy for villous adenoma of the ampulla of
Vater diagnosed before operation. Br J Surg 1997; 84(7):948951.
Chapter 26 Ampullary Neoplasm
varies between a local endoscopic resection in case of a benign
lesion, and a radical surgical pancreaticoduodenectomy for malig-
nant disease. The diagnostic work-up of ampullary lesions is chal-
lenging. A high index of suspicion together with careful endoscopic
examination of the papilla are crucial for early detection. CT-scan
and endoscopic ultrasonography are usually necessary to decide on
the optimal treatment for the individual patient. Ampullary tumors
are typically lesions that require close cooperation of a multidisci-
plinary team consisting of a gastroenterologist, radiologist, surgeon
and pathologist.
13. Catalano MF, Linder JD, Chak A, et al. Endoscopic management of
adenoma of the major duodenal papilla. Gastrointest Endosc
2004; 59(2):225232.
14. Bourgeois N, Dunham F, Verhest A, et al. Endoscopic biopsies of
the papilla of Vater at the time of endoscopic sphincterotomy:
difculties in interpretation. Gastrointest Endosc 1984;
30(3):163166.
15. Defrain C, Chang CY, Srikureja W, et al. Cytologic features and
diagnostic pitfalls of primary ampullary tumors by endoscopic
ultrasound-guided ne-needle aspiration biopsy. Cancer 2005;
105(5):289297.
16. Skordilis P, Mouzas IA, Dimoulios PD, et al. Is endosonography
an effective method for detection and local staging of the
ampullary carcinoma? A prospective study. BMC Surg 2002; 25;
2:1.
17. Menzel J, Hoepffner N, Sulkowski U, et al. Polypoid tumors of the
major duodenal papilla: preoperative staging with intraductal US,
EUS, and CTa prospective, histopathologically controlled study.
Gastrointest Endosc 1999; 49(3):349357.
18. Chen CH, Tseng LJ, Yang CC, et al. The accuracy of endoscopic
ultrasound, endoscopic retrograde cholangiopancreatography,
computed tomography, and transabdominal ultrasound in the
detection and staging of primary ampullary tumors.
Hepatogastroenterology 2001; 48(42):17501753.
19. Schwarz M, Pauls S, Sokiranski R, et al. Is a preoperative
multidiagnostic approach to predict surgical resectability of
periampullary tumors still effective? Am J Surg 2001;
182(3):243249.
20. Andersson M, Kostic S, Johansson M, et al. MRI combined
with MR cholangiopancreatography versus helical CT in the
evaluation of patients with suspected periampullary tumors:
a prospective comparative study. Acta Radiol 2005; 46(1):
1627.
21. Will U, Meyer F, Erhardt C, et al. Correlation of differential
diagnosis between inammatory and malignant lesions of the
papilla of Vater using endosonography with results of histologic
investigation [Abstract]. Gastroenterology 2003; 124 (suppl 1)
A 440.
283
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
22. Cannon ME, Carpenter SL, Elta GH, et al. EUS compared with CT,
magnetic resonance imaging, and angiography and the inuence
of biliary stenting on staging accuracy of ampullary neoplasms.
Gastrointest Endosc 1999; 50(1):2733.
23. Kubo H, Chijiiwa Y, Akahoshi K, et al. Pre-operative staging of
ampullary tumours by endoscopic ultrasound. Br J Radiol;
72(857):443447.
24. Mukai H, Nakajima M, Yasuda K, et al. Evaluation of endoscopic
ultrasonography in the pre-operative staging of carcinoma of the
ampulla of Vater and common bile duct. Gastrointest Endosc
1992; 38(6):676683.
25. Tierney WM, Francis IR, Eckhauser F, et al. The accuracy of EUS
and helical CT in the assessment of vascular invasion by
peripapillary malignancy. Gastrointest Endosc 2001; 53(2):
182188.
26. Itoh A, Goto H, Naitoh Y, et al. Intraductal ultrasonography in
diagnosing tumor extension of cancer of the papilla of Vater.
Gastrointest Endosc 1997; 45(3):251260.
27. Seifert E, Schulte F, Stolte M. Adenoma and carcinoma of the
duodenum and papilla of Vater: a clinicopathologic study. Am J
Gastroenterol 1992; 87(1):3742.
28. Schneider AR, Seifert H, Trojan J, et al. Frequency of colorectal
polyps in patients with sporadic adenomas or adenocarcinomas
of the papilla of vateran age-matched, controlled study. Z
Gastroenterol 2005; 43(10):11231127.
29. Zdorov Z, Dvork M, Hajer J. Endoscopic therapy of
benign tumors of the papilla of Vater. Endoscopy 2001;
33(4):345347.
30. Saurin JC, Chavaillon A, Napolon B, et al. Long-term follow-up of
patients with endoscopic treatment of sporadic adenomas of the
papilla of Vater. Endoscopy 2003; 35(5):402406.
31. Kimura W, Futakawa N, Zhao B. Neoplastic diseases of the papilla
of Vater. J Hepatobiliary Pancreat Surg 2004; 11(4):223231.
32. Fischer HP, Zhou H. Pathogenesis of carcinoma of the papilla of
Vater. J Hepatobiliary Pancreat Surg 2004; 11(5):301309.
33. Spigelman AD, Talbot IC, Penna C, et al. Evidence for adenoma-
carcinoma sequence in the duodenum of patients with familial
adenomatous polyposis. The Leeds Castle Polyposis Group
(Upper Gastrointestinal Committee). J Clin Pathol 1994;
47(8):709710.
34. Burke CA, Beck GJ, Church JM, et al. The natural history of
untreated duodenal and ampullary adenomas in patients with
familial adenomatous polyposis followed in an endoscopic
surveillance program. Gastrointest Endosc 1999; 49(3):358364.
35. Takashima M, Ueki T, Nagai E, et al. Carcinoma of the ampulla of
Vater associated with or without adenoma: a clinicopathologic
analysis of 198 cases with reference to p53 and Ki-67
immunohistochemical expressions. Mod Pathol 2000;
13(12):13001307.
36. Kimura W, Futakawa N, Yamagata S, et al. Different
clinicopathologic ndings in two histologic types of carcinoma of
papilla of Vater. Jpn J Cancer Res 1994; 85(2):161166.
37. Zhao B, Kimura W, Futakawa N, et al. p53 and p21/Waf1 protein
expression and K-ras codon 12 mutation in carcinoma of the
papilla of Vater. Am J Gastroenterol 1999; 94(8):21282134.
38. Hoffmann KM, Furukawa M, Jensen RT. Duodenal neuroendocrine
tumors: Classication, functional syndromes, diagnosis and
medical treatment. Best Pract Res Clin Gastroenterol 2005;
19(5):675697.
39. Makhlouf HR, Burke AP, Sobin LH. Carcinoid tumors of the
ampulla of Vater: a comparison with duodenal carcinoid tumors.
Cancer 1999; 85(6):12411249.
40. Hatzitheoklitos E, Buchler MW, Friess H, et al. Carcinoid of the
ampulla of Vater. Clinical characteristics and morphologic
features. Cancer 1994; 73(6):15801588.
284
41. Toyoda H, Yamaguchi M, Nakamura S, et al. Regression
of primary lymphoma of the ampulla of Vater after
eradication of Helicobacter pylori. Gastrointest Endosc 2001;
54(1):9296.
42. Yildirim N, Oksuzoglu B, Budakoglu B, et al. Primary duodenal
diffuse large cell non-hodgkin lymphoma with involvement of
ampulla of Vater: report of 3 cases. Hematology 2005;
10(5):371374.
43. Nadal E, Martinez A, Jimenez M, et al. Primary follicular
lymphoma arising in the ampulla of Vater. Ann Hematol 2002;
81(4):228231.
44. Weinstock LB, Swanson PE, Bennett KJ, et al. Jaundice caused by a
clinically undetectable T-cell lymphoma inltrating the sphincter
of Oddi. Am J Gastroenterol 2001; 96(11):31863189.
45. Matsushita M, Kobayashi Y, Kobayashi H, et al. A case of
gastrointestinal stromal tumour of the ampulla of Vater. Dig Liver
Dis 2005; 37(4):275277.
46. Schubert ML, Moghimi R. Gastrointestinal Stromal Tumor (GIST).
Curr Treat Options Gastroenterol 2006; 9(2):181188.
47. Le Borgne J, Partensky C, Glemain P, et al.
Pancreaticoduodenectomy for metastatic ampullary and
pancreatic tumors. Hepatogastroenterology 2000;
47(32):540544.
48. Blow S. Results of national registration of familial adenomatous
polyposis. Gut 2003; 52(5):742746.
49. Spigelman AD, Williams CB, Talbot IC, et al. Upper gastrointestinal
cancer in patients with familial adenomatous polyposis. Lancet
1989; 2(8666):783785.
50. Offerhaus GJ, Giardiello FM, Krush AJ, et al. The risk of upper
gastrointestinal cancer in familial adenomatous polyposis.
Gastroenterology 1992; 102(6):19801982.
51. Bjrk J, kerbrant H, Iselius L, et al. Periampullary adenomas and
adenocarcinomas in familial adenomatous polyposis: cumulative
risks and APC gene mutations. Gastroenterology 2001;
121(5):11271135.
52. Blow S, Bjrk J, Christensen IJ, et al. Duodenal adenomatosis in
familial adenomatous polyposis. Gut 2004; 53(3):381386.
53. Burke C. Risk stratication for periampullary carcinoma in patients
with familial adenomatous polyposis: does theodore know what
to do now? Gastroenterology 2001; 121(5):12461248.
54. Norton ID, Gostout CJ. Management of periampullary adenoma.
Dig Dis 1998; 16(5):266273.
55. Martin JA, Haber GB. Ampullary adenoma: clinical manifestations,
diagnosis, and treatment. Gastrointest Endosc Clin N Am 2003;
13(4):649669.
56. Farnell MB, Sakorafas GH, Sarr MG, et al. Villous tumors of the
duodenum: reappraisal of local vs. extended resection.
J Gastrointest Surg 2000; 4(1):1323.
57. Shemesh E, Nass S, Czerniak A. Endoscopic sphincterotomy
and endoscopic fulguration in the management of adenoma
of the papilla of Vater. Surg Gynecol Obstet 1989; 169(5):
445448.
58. Lambert R, Ponchon T, Chavaillon A, et al. Laser treatment of
tumors of the papilla of Vater. Endoscopy 1988; 20 Suppl
1:227231.
59. Suzuki K, Kantou U, Murakami Y. Two cases with ampullary cancer
who underwent endoscopic excision. Prog Dig Endosc 1983;
23:236239.
60. Binmoeller KF, Boaventura S, Ramsperger K, et al. Endoscopic
snare excision of benign adenomas of the papilla of Vater.
Gastrointest Endosc 1993; 39(2):127131.
61. Beger HG, Treitschke F, Gansauge F, et al. Tumor of the
ampulla of Vater: experience with local or radical resection
in 171 consecutively treated patients. Arch Surg 1999;
134(5):526532.

62. Roggin KK, Yeh JJ, Ferrone CR, et al. Limitations of ampullectomy
in the treatment of nonfamilial ampullary neoplasms. Ann Surg
Oncol 2005; 12(12):971980.
63. de Castro SM, van Heek NT, Kuhlmann KF, et al. Surgical
management of neoplasms of the ampulla of Vater: local
resection or pancreatoduodenectomy and prognostic factors for
survival. Surgery 2004; 136(5):9941002.
Chapter 26 Ampullary Neoplasm
64. Yoon YS, Kim SW, Park SJ, et al. Clinicopathologic analysis of early
ampullary cancers with a focus on the feasibility of ampullectomy.
Ann Surg 2005; 242(1):92100.
65. Jung S, Kim MH, Seo DW, et al. Endoscopic snare papillectomy of
adenocarcinoma of the major duodenal papilla. Gastrointest
Endosc 2001; 54(5):622.
285
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Malignant Biliary Obstruction: Distal
27 Raed M. Alsulaiman and Alan Barkun
INTRODUCTION
Malignant biliary obstruction is most frequently encountered in
the setting of pancreatic adenocarcinoma, but can also be due to
cholangiocarcinomas, gallbladder, and ampullary neoplasms. The
approach and management of distal biliary obstruction, including
the role of endoscopic retrograde cholangiopancreatography (ERCP),
will be discussed in this chapter. For discussions on the approach
to managing patients with proximal biliary obstruction and the prac-
tical aspects of biliary stenting, readers are referred to Chapters 17
and 28.
EPIDEMIOLOGY
Malignancies of the biliary and pancreatic systems are not uncom-
mon; together they represent two of the 10 most incident cancers in
North America and Europe. While the incidence of pancreatic cancer
has remained stable over the last 25 years, the epidemiology of chol-
angiocarcinoma has changed. The incidence of intrahepatic cholan-
giocarcinomas seems to be rising while that of extrahepatic biliary
tumors is decreasing, even though the reasons for such a change in
pattern are not known.1 Because these cancers are usually diagnosed
at advanced stages, when the probability of cure is very low, the
mortality rate is very high. Consequently pancreatic cancer ranks as
the fth most lethal cancer in the US, and second as a cause of
digestive cancer-related deaths, behind colon cancer. The incidence
of pancreatic and biliary malignancies increases with age and, in
fact, these tumors are rarely seen before the age of 45. Epidemi-
ological surveys have shown that the median age of diagnosis
approximates 70 years. Exceptions include genetically predisposed
individuals, and those with chronic pre-malignant conditions such
as primary sclerosing cholangitis.2 Pancreatic cancer is more
common in males, people of Jewish and Afro-American descent.
Diabetes, chronic pancreatitis, pernicious anemia, inherited disor-
ders such as familial adenomatous polyposis, and high fat and meat
intake have been cited as risk factors for pancreatic cancer.3 Although
rare and conned to clusters of families, genetic disorders such as
hereditary pancreatitis and familial pancreatic cancer have also been
linked to pancreatic cancer; individuals with these conditions appear
to have up to a 40% lifetime risk of malignant transformation.4
The majority of cases of cholangiocarcinoma have no identiable
underlying etiology. However, a number of risk factors have been
implicated in its development; most factors exhibit long-standing
inammation and chronic injury of the biliary epithelium. Primary
sclerosing cholangitis is an uncommon disease, more commonly
seen in middle-aged males. It is characterized by stricturing, brosis
and inammation of the biliary tree, and is closely associated with
inammatory bowel disease, particularly ulcerative colitis. Approxi-
mately 1020% of patients with primary sclerosing cholangitis will
develop cholangiocarcinoma. The rare congenital bropolycystic dis-
eases of the biliary system are associated with increased risks of
cholangiocarcinoma, particularly choledochal cysts and Carolis
disease. Choledochal cysts are associated with a 10% lifetime inci-
dence of cholangiocarcinoma: there is a 1% per year risk which pla-
teaus after 1520 years.5 In the Far East, other forms of chronic
inammation associated with cholangiocarcinoma include infesta-
tion with the liver ukes Clonorchis sinensis and Opisthorchis viv-
erinni. Cholangiocarcinoma is also rarely seen in association with
cirrhosis and has been weakly linked to hepatitis C infection.
Among neoplasms involving the biliary tree, carcinoma of the
gallbladder has the poorest prognosis with a 5-year survival ranging
between 0% and 10% in most reported series.6 Gallbladder cancer is
the most incident malignant lesion of the biliary tract, and the fth
most common among malignant neoplasms of the digestive tract. It
affects women two to six times more often than men, and the inci-
dence increases with age. Although its etiology is unknown, choleli-
thiasis is thought to be an important risk factor for gallbladder cancer.
Other risks factors such as the presence of a porcelain gallbladder,
gallbladder polyps, an anomalous pancreaticobiliary junction and
obesity have also been suggested in epidemiological studies.7
NATURAL HISTORY
Although not always present, obstruction of the distal common bile
duct (CBD) occurs during the natural evolution of most of these
tumors, depending on their location and behavior. The most common
malignancy causing distal biliary malignant obstruction is pancre-
atic cancer accounting for more than 90% of cases (2), followed by
gallbladder cancer, malignant lymphadenopathy and cholangiocar-
cinoma, the latter being relatively uncommon in Western countries.
Carcinoma of the ampulla of Vater can also obstruct the distal CBD
and although rarely seen in otherwise healthy individuals, it is par-
ticularly common in patients with familial adenomatous polyposis.
In fact, it is a leading cause of death in this population. Gallbladder
cancer and cholangiocarcinoma involving the distal CBD may also
obstruct, but represent just a small fraction of all such patients. The
overall prognosis of malignancies that cause biliary obstruction is
dismal. Except for extrinsic compressions caused by enlarged lymph
nodes in the case of hematological malignancies such as non-
Hodgkins lymphomas and for ampullary tumors, the majority of
patients found with unresectable disease have a median survival
of 35 months.8
CLINICAL FEATURES
The most common presenting symptoms of pancreaticobiliary
malignancies are painless jaundice, anorexia and weight loss, and
are seen in most patients. If pain occurs it is often located in the
287
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Palliated (N = 256) Resected (N = 512) Ampullary adenocarcinoma

Demographic Pancreatic adenocarcinoma
Age 64.0 0.7 years 65.8 0.5 years Head
Gender 57% male 55% male Body/tail
Race 91% white 91% white Gallbladder adenocarcinoma
Symptoms and signs Cholangiocarcinoma
Abdominal pain 64% 36%a Non-hilar
Jaundice 57% 72%a Metastatic disease
Weight loss 48% 43%
Nausea/vomiting 30% 18%a Table 27.2 Most prevalent distal pancreaticobiliary malignancies
Back pain 26% 2%

Table 27.1 Presenting symptoms and signs in patients with

resectable and unresectable pancreatic cancer
a
P < 0.001 vs. Palliated group
With permission from Sohn TA, Lillemoe KD, Cameron JL, et al. Surgical palliation of
unresectable periampullary adenocarcinoma in the 1990, J Am Coll Surg 1999;
188:658666. 1999 The American College of Surgeons.

epigastric region or right upper quadrant, and may radiate to the

back. Back pain usually indicates retroperitoneal inltration by the
tumor, and therefore, probable unresectability (Table 27.1). Other
symptoms may include dark urine, pale stools and pruritus. As
many as 80% of patients with pancreatic cancer will present with
impaired glucose tolerance or frank diabetes mellitus. Carcinoma of
the body and tail of the pancreas presents with similar features,
although jaundice is usually absent or develops very late in the
course of the disease. A history of inammatory bowel disease or
previously diagnosed malignancies should be sought. A complete
physical examination, including assessment for abnormal lymph
nodes, jaundice, hepatomegaly, palpable gallbladder, or mass should
be performed. A chest radiograph may be appropriate to exclude
pulmonary metastases. Obtaining serum tumor markers such as
CA19-9 and CEA may be appropriate. Once there is a clinical suspi-
cion of a pancreaticobiliary malignancy, further investigation with
abdominal imaging studies is appropriate.
DIFFERENTIAL DIAGNOSIS OF DISTAL BILIARY
MALIGNANCIES AND IMAGING TECHNIQUES
The differential diagnosis of distal biliary malignancies is shown in
Table 27.2.
Ampullary carcinoma
Ampullary carcinoma is suspected based upon the demonstration of
obstructive jaundice, often with dilation of the pancreatic and biliary
ducts seen on abdominal imaging studies. A discrete mass may or
may not be identiable by using standard transabdominal US (TUS)
or helical computerized tomography (CT) scanning. ERCP allows for
direct identication and biopsy conrmation, although the diagnos-
tic accuracy of biopsy is not 100%. MRCP may allow identication
of the lesion and obviate diagnostic ERCP. Endoscopic ultrasound
(EUS) allows for more accurate diagnosis and staging of these
lesions than CT, and also allows for ne needle aspirate (FNA) tissue
sampling.9 EUS also may facilitate the selection of patients who
can undergo local resection instead of pancreaticoduodenectomy
(Whipple operation). Once the lesion is identied and staged, the
choice of operative resection for cure or some form of jaundice
palliation are similar to treatment options for carcinoma of the
pancreatic head.
Fig. 27.1 CT demonstrating a pancreatic carcinoma causing biliary
obstruction: enlarged head of the pancreas with an irregular area
of low attenuation of the tumor.
Pancreatic malignancies
The approach to the patient with pancreatic carcinoma involving the
pancreatic head is different than in the patient with body/tail lesions
in terms of accessibility, curative potential, and palliation.
Most patients with cancer of the pancreatic head present with
obstructive jaundice. Radiological imaging studies are performed
allowing for (a) detection of the tumor, (b) determination of tumor
resectability, and (c) tissue acquisition under imaging guidance.
TUS will suggest biliary obstruction by the demonstration of
biliary ductal dilation. It may also identify the presence of obvious
liver metastases. However, standard TUS is operator dependent and
has a poor sensitivity for detecting small neoplasms of the pancreatic
head. Recent advances in TUS, such as color-power Doppler US,
US angiography, harmonic imaging (tissue harmonic imaging
and contrast harmonic imaging), and three-dimensional US, may
improve the usefulness of this modality in the staging of pancreatic
cancer. Nonetheless, more information regarding staging and extent
of disease, and possible nodal or vascular involvement is obtainable
with other imaging modalities.
Helical CT of the abdomen with ne cuts through the pancreas
during the arterial and portal phases of contrast enhancement has
a high sensitivity and specicity for the detection of pancreatic car-
cinoma (Fig. 27.1). It allows for the determination of tumor exten-
sion, liver metastases, and invasion of vascular structures, and thus,
resectability. Multislice (multidetector) CT has been introduced and

288
 Chapter 27 Malignant Biliary Obstruction: Distal

may improve on the accuracy of helical CT. If the CT ndings are

found to be highly suggestive of a resectable pancreatic carcinoma
in the appropriate clinical setting, and the patient is felt to be an
operative candidate, a reasonable approach is to then refer the patient
directly for an attempt at surgical resection (pancreaticoduodenec-
tomy) with or without further imaging (depending on local avail-
ability and expertise) or diagnostic testing. Transabdominal or
CT-guided biopsy of the pancreatic mass rarely may result in tumor
seeding at the needle track or within the peritoneum and has been
reported to increase the risk of postoperative recurrence.10 If the CT
scan reveals overt evidence of unresectable pancreatic cancer or the
patient is a not an operative candidate because of co-morbid medical
conditions, non-operative palliation of obstructive jaundice should
be performed at ERCP. If a denitive tissue diagnosis is required
for the administration of chemotherapy and/or radiation therapy,
tissue acquisition can be performed at the time of the palliative
ERCP. If a tissue diagnosis cannot be made at that time, then trans-
abdominal biopsy (CT-guided or US) of the mass or metastatic
disease sites (i.e. liver lesions), or EUS-guided FNA of the mass or
metastatic sites should be performed.
Magnetic resonance imaging (MRI) of the pancreas may include
MRI, MR cholangiopancreatography (MRCP), or magnetic reso-
nance angiography. Standard abdominal MRI appears to be an accu-
rate modality for staging pancreatic carcinoma, though it does not
appear to be more specic or sensitive than helical CT. In addition,
it is more expensive and more time consuming to perform than CT11
(Fig. 27.2).
If expertise in EUS is readily available, it should be used as a
preoperative staging modality in patients with suspected pancreatic
cancer (Fig. 27.3). This is particularly important in patients with
equivocal ndings on CT or those with co-morbidities and, there-
Fig. 27.2 Cholangiogram at MRCP showing a pancreatic carci-
fore, at higher risk for intra-operative or postoperative complica- noma with upstream dilation of pancreatic and common bile ducts.
tions. EUS allows identication of vascular invasion as well as A plastic stent is shown within the common bile duct.
sampling of suspicious-appearing lymph nodes, which, if positive,
may change the treatment approach as it alters prognosis. EUS
appears to be complementary to helical CT, with EUS better at
detecting small (<3 cm) masses, staging the portal vein, and detect-
ing lymph node metastases, while helical CT is superior for staging
arterial involvement and distant metastases.12 An EUS-guided FNA
biopsy specimen allows for a denitive tissue diagnosis of a pancre-
atic mass when results on other biopsy methods are negative but
pancreatic cancer is suspected. If EUS suggests resectability, EUS-
guided biopsy of the mass is not necessary before proceeding with
operative resection, although this point remains controversial.
Advantages of needle biopsy of the mass include identication of
alternative diagnoses to primary pancreatic adenocarcinoma (lym-
phoma, islet cell tumors, and metastatic disease). It also allows for
preoperative patient counseling. Potential disadvantages of preop-
erative EUS-guided FNA include the risks of pancreatitis, bleeding,
and, theoretically, tumor seeding.13 The latter has never been reported
and appears to be inconsequential in most cases since the needle
path usually lies within the resected specimen. Ideally, EUS should
be performed before ERCP and stent placement since the latter may
interfere with the accuracy of EUS staging and EUS ndings of
unresectable carcinoma allow improved patient selection for place-
ment of a self-expanding metallic stent.14 In patients with unresect-
able cancer, EUS-guided celiac plexus neurolysis has been shown to
control disabling abdominal pain.
Fig. 27.3 EUS appearance of a hypoechoic lesion measuring
The near-pathognomonic ndings on ERCP of a pancreatic head 1.5 cm in the head of the pancreas (Grey arrow). CBD = common
cancer are strictures of the bile and pancreatic ducts with proximal Bile Duct, PD = Pancreatic Duct.

289
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
dilation (the double-duct sign) (Fig. 27.4). While ductal abnormali-
ties are almost invariably present in patients with adenocarcinoma,
other imaging modalities (CT, MR, and EUS) have supplanted ERCP
in the diagnosis of pancreatic cancer. Preoperative ERCP does not
add further staging information and may result in complications15
that may make operative intervention more difcult and/or may
considerably delay operative intervention, resulting in a decreased
potential for curative resection. Furthermore, several studies suggest
postoperative complications after pancreaticoduodenectomy may be
higher when a preoperative ERCP is done.16 However, if the patient
suffers from cholangitis or severe pruritus, or if there is a substantial
delay in operative resection, preoperative ERCP with biliary drainage
should be performed. A more extensive discussion on biliary stent-
ing follows in this chapter.
Cholangiocarcinoma
A primary tumor of the bile duct should be suspected based on clini-
cal and imaging ndings. Abdominal CT scans will show biliary
Fig. 27.4 Cholangiogram at ERCP showing a stenosis of both pan-
creatic and common ducts also called a double-duct sign as seen
in pancreatic adenocarcinoma.
Type I Type II
Fig. 27.5 The Bismuth classication for cholangiocarcinomas.
290
dilation without an associated pancreatic mass or pancreatic ductal
dilation, and the level of obstruction usually can be localized to a
level above that of the pancreatic duct. The differentiation of hilar vs
non-hilar tumors is important because of implications for both
operative resection and endoscopic palliation. The Bismuth classi-
cation of cholangiocarcinoma is useful for determining surgical
resectability and type of surgery (Fig. 27.5). If imaging studies map
the level of obstruction below the bifurcation (Bismuth type I lesions,
Fig. 27.5), operative resection should be considered in t patients
without metastatic disease. If the patient is a poor operative candi-
date, palliation with plastic or metal stents, as with pancreatic carci-
noma, should be undertaken.17 The management of patients with
hilar cholangiocarcinomas is detailed in the next chapter.
Metastatic disease
A variety of malignant diseases may metastasize to and around the
biliary tree, resulting in obstruction. These may lead to biliary
obstruction either intrinsically or extrinsically (porta hepatis involve-
ment) anywhere from the level of the bifurcation to the ampulla. The
diagnosis may be obvious, based upon known widespread malig-
nancy, or more occult and discovered at the time of endoscopic
evaluation or surgical resection. CT scan ndings may mimic
primary malignant disease of the bile ducts or pancreas.18 An MR
examination may be useful in dening the presence of perihilar
obstructive disease. If disease is widespread, palliation of obstruc-
tion is indicated as for primary malignancies. Surgical resection may
be indicated in selected cases.
AN APPROACH TO THE MANAGEMENT OF
PATIENTS WITH DISTAL BILIARY MALIGNANCIES
If a pancreaticobiliary malignancy is suspected based on clinical
and US ndings, further imaging must be performed to obtain a
diagnosis, stage the extent of the malignant process for resectabil-
ity, and evaluate the appropriateness of possible palliative treat-
ment. Identication of the level of obstruction is of great importance
since the differential diagnosis and therapeutic implications differ
accordingly. Conceptually, management may be stratied according
to whether the biliary obstruction is proximal or distal.19 Patients
with a distal CBD obstruction may be amenable to endoscopic or
surgical drainage, whereas a more proximal blockage of the biliary
tree may require a more complex intrahepatic anastomosis or per-
cutaneous drainage. The optimal approach to patients with malig-
nant biliary obstruction must take into account the performance
characteristics of the different imaging modalities, the level and
cause of the obstruction, the risk of cholangitis when opacifying an
obstructed biliary tree, and the potential for curative versus pallia-
tive therapy. Recent data suggest that non-invasive biliary imaging
Type IIIa Type IIIb Type IV

may greatly assist endoscopic drainage and diminish septic compli-
cations that occur when there is a failed attempt at unilateral or
bilateral drainage.20 A proposed algorithm for the management of
patients with a suspected pancreaticobiliary malignancy is shown
in Figure 27.6. The roles of radiotherapy and nutritional support
are not addressed in this chapter.
Curative surgery
Operable patients with a distal pancreaticobiliary neoplasm and no
evidence of metastatic disease or local vascular invasion should be
offered curative surgical resection. Unfortunately these patients
account for only 1020% of all presenting cases. Many elderly
patients are not referred for consideration of surgery as they are
judged unt for surgery due to advanced age or the presence of
unrelated co-morbidities. The rst step towards potential resection
should be laparoscopy to determine resectability and to prevent a
lengthy hospital stay and prolonged convalescence associated with
an unnecessary laparotomy. Laparoscopy is used mostly to detect
peritoneal carcinomatosis, liver metastases, malignant ascites, and
unexpected cirrhosis. Despite an extensive preoperative work-up,
11%-53% of patients are found to be unresectable at the time of
laparotomy. Most patients thus end up undergoing palliative treat-
ment tailored to the symptomatology, i.e. either a surgical bypass
(biliary or biliary and gastric), or placement of a biliary stent.21
Palliation
The three most important conditions requiring treatment in patients
with unresectable biliary and pancreatic cancers are cholestasis,
pain, and gastrointestinal obstruction. These may be consequences
of local tumor invasion into adjacent structures including the bile
ducts, duodenum, and neural celiac plexus.
Endoscopic stenting
The following emphasizes issues relating to an approach to the man-
agement of patients with distal malignant biliary obstruction; practical
aspects of biliary endoscopic stenting and a more detailed description
of the equipment are discussed in Chapters 4, 16, 17, and 28.
Clinical suspicion of primary
pancreaticobiliary malignancy OR
Transabdominal US suggestive
Helical CT or MS-CT
Suspected resectable Unresectable primary
pancreatic head, or metastatic disease
ampullary, or distal
cholangio CA
Resectable Resectable
EUS Surgery
Unresectable
Chapter 27 Malignant Biliary Obstruction: Distal
Background
Endoscopic placement of plastic biliary stents were rst described
by Soehendra et al. as an alternative to surgical biliary bypass in
high-risk and inoperable cancer patients.22 Self-expandable metal
stents (SEMS) for use in the biliary system were introduced into
clinical practice over a decade later. The ability to place a larger-
diameter plastic stent is limited by the size of the endoscope acces-
sory channel. SEMS were developed to overcome this limitation.
They have the advantage of larger diameter stenting (up to 10 mm)
but are more costly than plastic stents.
Indications
Biliary decompression is indicated if there is cholangitis or pruritus
in the face of advanced malignant biliary obstruction. Biliary stenting
has also been shown to improve symptoms of anorexia and quality
of life.23,24 When a patient has advanced malignant disease, drainage
of the biliary system for palliation is not routinely indicated because
the risk of complications related to the procedure may outweigh the
potential benet. Indeed, the best treatment for a patient with asymp-
tomatic obstructive jaundice and liver metastases may be supportive
care alone. In the preoperative setting, as jaundice itself is thought
to be associated with multiple adverse systemic effects (e.g. renal
failure, sepsis, and impaired wound healing), it has been postulated
to improve surgical outcomes if performed before pancreaticoduo-
denectomy. Routine preoperative drainage of an obstructed biliary
system, however, has not been shown to benet patients who will
soon undergo a surgical procedure, and may in fact be deleterious
in some.15,25,26 If preoperative drainage is indicated because of chol-
angitis or an anticipated delay to surgery in the face of clinically
signicant symptoms, such as pruritus, drainage has traditionally
been performed using plastic stents. A recent cost-minimization
study favors the preoperative placement of a short SEMS.27
Plastic endoprostheses
Plastic stents are easy to insert, and can be removed if necessary
(Fig. 27.7). Their biggest advantage compared to metal stents is that
their upfront cost is signicantly lower (tenfold in many markets).
Fig. 27.6 An algorithm suggested by the
American Society for Gastrointestinal Endos-
copy for the diagnosis and management of
patients with suspected pancreaticobiliary
malignancies.
Suspected hilar
tumor
MRCP
Unresectable
Palliation
291
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 27.7 Cholangiogram at ERCP showing a 10 French, 9 cm
plastic biliary endoprosthesis inserted across a distal common bile
duct stricture.
A large variety of biliary plastic stents are available with internal
diameters ranging from 5 to 11.5 French (Fr) gauge with lengths
varying from 5 to 15 centimeters (cm). Straight plastic stents with
aps at both ends and side-holes are the most common type of stent
used. The presence of aps minimizes the risk of stent migration
which is less likely to occur in pigtail stents due to their physical
characteristics that allow greater anchoring inside the CBD and
duodenum. Although no study has compared the occlusion and
migration rates between straight and pigtail stents, animal studies
suggest that straight stents may provide better bile drainage than
pigtail stents.28 The main limitation of plastic stents is their shorter
duration of patency and the potential for occlusion.
Many studies have assessed the mechanisms of plastic stent
occlusion. Although recent investigations have focused on the
importance of ingested ber matter,29 most studies have historically
emphasized bacterial colonization of the prosthesis interior leading
to production of a bacterial biolm and subsequent stent clogging.30
These ndings have resulted in modied stent designs or the use of
adjuvant medications in the hope of prolonging stent patency.
(a) Size of the internal diameter: The duration of patency for
stents with an internal diameter of 10 Fr or greater is 2132 weeks
compared to 1012 weeks for 7 or 8.5 Fr plastic stents. Additionally,
there may be a lower associated incidence of cholangitis with larger
gauge stents. The superior performance of large-caliber stents is
attributed to improved internal ow dynamics.31 There is no con-
clusive evidence favoring 11.5 Fr compared to 10 Fr stents,32 and
the insertion of 10 Fr stents does not require the routine perfor-
292
mance of a sphincterotomy. Stent length does not seem to affect
patency duration.
(b) Plastic-stent design: Evidence linking the presence of side holes
as a contributing factor for stent occlusion prompted the develop-
ment of a stent without side holes. As materials that allow a lower
coefcient of friction may theoretically reduce stent clogging, the
stent was made of Teon. However these changes have not resulted
in proven clinical benets.33,34 More recently, a double layer plastic
stent, combining a very smooth inner surface with an absence of
sideholes, has been found in one randomized trial to display a longer
patency duration than the standard polyethylene stents,40 while a
pilot study has assessed a stent without a lumen, that may result in
prolonged stent patency.41
(c) Position of the distal tip of the stent: Placement of plastic stent
above an intact sphincter of Oddi appeared useful in animals by
preventing colonization by bacteria, but did not result in demon-
strable patency improvement in a human, randomized clinical trial.
Moreover, stents placed above the papilla had higher stent migration
rates.35
(d) Administration of choleretic agents and/or antibiotics: The role
of pharmacological agents to lengthen the patency period fo plastic
stents has been assessed using choleretic agents that may enhance
bile ow and reduce stent occlusion. Antibiotics may also be useful
by inhibiting bacterial colonization of the stent. However, both
classes of drugs, alone or in combination, have failed to demonstrate
improvement in the duration of stent patency.36,37 A series of meta-
analyses have conrmed the negative results for all the aforemen-
tioned attempts in improving the duration of patency with plastic
stents.38 In addition, no improvement in survival has been noted. A
recent randomized trial of patients with malignant biliary obstruc-
tion treated by plastic stent insertion demonstrated that long-term
ciprooxacin administration initiated prior to the ERCP signicantly
decreased the incidence of cholangitis, and resulted in improvement
of certain domains of quality of life.39
Self-expandable metal stents
The idea of inserting an expandable stent has been applied to stric-
tures of the biliary tree comparable to their use in vascular stenoses.8
SEMS are braided in the form of a tubular mesh using a surgical
grade steel alloy. The elastic properties of the material allow the stent
to adopt different congurations according to the site and intensity
of forces applied to it. SEMS are delivered into the bile duct while
completely constrained by a sheath, allowing insertion as a small-
circumference delivery system. As the constraining sheath is pro-
gressively retracted from its more distal end, the intrinsic expansile
forces of the stent make it regain its original conguration. After the
sheath is completely withdrawn, the end result is an expanded stent
which accommodates to the shape of normal and strictured bile
ducts by maintaining constant radial pressure against its bile duct
wall (Figs 27.8 and 27.9). Since the initial application of Wallstents
in patients with biliary malignancies, a multitude of SEMS types
have been released. SEMS differ in regard to the type of delivery
system, structural composition, design, length and diameter.
However, all achieve a much larger internal diameter and subse-
quent longer patency rate compared to plastic stents (see below). The
mechanisms of SEMS blockage include stent ingrowth and over-
growth by tumor, as well as mucosal hyperplasia.42
There are limited published data directly comparing different
models of uncovered SEMS.43,44,44a Generally, most biliary endosco-
pists consider SEMS models to be equivalent. Familiarity with the

Fig. 27.8 Cholangiogram at ERCP showing a distal common bile
duct stricture caused by a pancreatic adenocarcinoma. A self-
expanding metallic stent has been inserted for palliation.
mechanical characteristics of the stent and its delivery system is
important to consider when choosing a SEMS.
More recently, polyurethane-covered SEMS have been developed
in the hope of prolonging stent patency by presenting a physical
barrier to tumor ingrowth. In the sole randomized comparative trial
to date, the covered SEMS technology was associated with a signi-
cant increase in patency duration as compared to the uncovered
SEMS.45 More recent non-randomized studies have questioned this
nding.45a,45b However the covered SEMS may be more prone to
migration, and may occlude ductal branches, leading to complica-
tions such as cholecystitis, cholangitis, and pancreatitis.45,46 Risk
factors for the development of cholecystitis in this setting include
gallstones and cystic duct invasion by tumor.46a
SEMS are difcult to remove so they are generally reserved for
patients with established, unresectable malignant disease, although
recently, an increasing number of endoscopists are describing
removal of covered SEMS.47 Additional complications related to
biliary stenting (common to both plastic and metal stents) are much
more infrequent and include upper gastrointestinal bleeding, duo-
denal perforation, and intestinal luminal obstruction.
Stent choices for palliation of malignant biliary obstruction
The optimal stent choice for the palliation of malignant biliary
obstruction is dependent on multiple factors and varies from patient
to patient. The major decision that needs to be made is the type of
stent to be placed (plastic or metal). Important measures for this
decision include several stent-related factors, such as stent efcacy
(relief of jaundice), stent patency, need for reinterventions, and
Chapter 27 Malignant Biliary Obstruction: Distal
Fig. 27.9 Endoscopic picture of metallic stent after full deploy-
ment with good bile ow from the stent.
costs. Patient-related issues, such as the extent of disease and
expected survival time, also need to be considered and inuence the
optimal and cost-effective choice of stent.
Several trials have directly compared plastic stents to SEMS for
the palliation of malignant biliary obstruction. Plastic stents and
SEMS both provide palliation of jaundice and improve liver tests
after placement in over 95% of patients; they are equivalent in this
regard. These two stent types, however, signicantly differ in dura-
tion of patency. Median stent patency ranges from 2 to 5 months for
plastic stent, whereas it is 4 to 10 months or more for SEMS.48,49,49a
The decreased patency of plastic stents seems to inuence the need
for additional interventions. In comparative trials, the use of plastic
stents led to signicantly increased repeat endoscopic interventions
to re-establish biliary drainage as compared with either covered or
uncovered SEMS.48,49,49b In several of these studies there were sig-
nicantly increased numbers of hospitalization days associated with
the use of plastic stents versus SEMS. Median patient survival ranges
from 4 to 6 months after plastic or metallic stenting. There have
been no associated survival benets demonstrable that are attribut-
able to any stent type in individual trials, although a recent meta-
analysis presented in abstract form demonstrated an improvement
in survival attributable to the use of SEMS compared to plastic stents
in patients with distal malignant biliary obstruction. Self-expandable
metal stents confer a survival advantage in palliation of distal malig-
nant biliary obstruction.49c The recent Cochrane systematic review,
however, did not conclude on any survival benets attributable to
metal versus plastic stents.49a
Cost-effectiveness analyses have shown that the optimal choice
of stent (plastic versus SEMS) is inuenced by the ratio of the
cost of stent to the cost of the ERCP, and the anticipated life expec-
tancy of the patient.50,51 The greater the cost of the ERCP, the more
likely the SEMS will be a cost-effective choice. Unfortunately, it is
difcult to estimate survival in a patient with malignant biliary
293
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

obstruction. Plastic stents may be preferred to SEMS in patients with

large tumors (>3 cm) or liver metastases, both of which are poor
predictors of survival, as plastic stents are cost-saving in patients
surviving less than 34 months while SEMS are more cost-effective
in patients expected to survive longer than 6 months.50,51 Direct cost
measurements from a randomized controlled trial demonstrated
similar results.52
Despite these considerations, there are no strict criteria for
selecting between plastic stents and SEMS for the palliation of
unresectable malignant biliary obstruction. Indeed, the decision
must be individualized as patient factors must also be considered. For
example, SEMS may be preferred in a patient who is non-compliant
or resides in a remote area without medical access, despite an antici-
pated short life expectancy. Patients with difcult endoscopic biliary
access from associated duodenal stenosis may benet from early
SEMS placement, and may even be candidates for endoscopic double
Fig. 27.10 Plain abdominal radiograph showing a plastic stent
bypass where a biliary and an enteric metallic stents are placed.52a inserted within a blocked SEMS.

The optimal stenting strategy

In addition to deciding on the optimal stent technology (plastic
versus SEMS), an endoscopist must also consider the optimal stent-
ing strategy. For example, if a plastic stent is initially inserted, should
it be replaced at regular intervals to prevent stent blockage, or is it
best to change the plastic stent on demand during the life of the
patient? In a randomized trial, routine exchanges every 3 months
were associated with longer symptom-free intervals for patients than
exchanges at signs of stent occlusion, but there was no difference in
overall survival.52 An extensive cost-effectiveness model suggested,
in order of the most to the least cost-effective approaches, the fol-
lowing approaches: initial insertion of a covered SEMS, or of an
uncovered metal stent, and a plastic stent with its subsequent on
demand replacement (when the patient developed symptoms sug-
gesting stent occlusion). Least cost-effective was the routine three-
monthly replacement of a plastic stent.53 As discussed previously, a
cost-minimization analysis suggested that the preoperative insertion
of a short covered SEMS is cost saving.27
Occluded SEMS are managed by a variety of methods. The most
commonly used techniques include insertion of a plastic stent within
the occluded SEMS (Fig. 27.10), insertion of a second SEMS (Fig
27.11) and mechanical cleaning of the occluded stent lumen. Overall
success rates for re-establishing biliary drainage are over 80%.
Mechanical cleaning methods, such as catheter irrigation or the use
of stone-extraction balloons, may be less successful and are associ-
ated with decreased duration of patency than repeat stenting. Given
the typical short median survival at the time of the rst SEMS occlu-
sion, treatment with a plastic prosthesis seems to be the most cost-
effective method. Other techniques for SEMS recanalization include
application of thermal energy and intaductal brachytherapy but are
not widely used.
Fig. 27.11 Plain abdominal radiograph showing a second SEMS
Percutaneous approach inserted within an initially placed, now occluded SEMS.
Percutaneous insertion of plastic stents and SEMS is an acceptable
alternative for management of distal biliary malignant obstruction
not successfully treated by an endoscopic approach. Percutaneous allows both internal and external bile ow. The technique has evolved
drainage was the preferred palliative method in patients with malig- over the years and currently insertion of an indwelling catheter
nant obstruction until several years ago. This procedure entails without external drainage is possible; furthermore, the advent of
sterile catheterization of a peripheral biliary radical after percutane- SEMS has obviated in many cases the need for serial tract dilation.
ous puncture. External drainage is accomplished by percutaneous Prior to the era of metallic expandable stents, the percutaneous
transhepatic insertion of a catheter, manipulation of a guidewire and approach had permitted the insertion of plastic stents with larger
insertion of a drainage catheter through the obstructing lesion that diameters when compared to endoscopic drainage. The consequent

294

benet of a longer stent patency represented a signicant advantage
over the prosthesis inserted by ERCP which was historically limited
by the size of the accessory channel of duodenoscopes. Also, percu-
taneous drainage appeared to be as effective as biliary bypass and
had other inherent advantages. Bornman et al. found the overall
survival to be similar in both surgical and percutaneous groups,
whereas percutaneous drainage was associated with a lower proce-
dure-related complication and 30-day mortality rate.54 The disadvan-
tages of external biliary drainage include the risk of spontaneous
catheter dislodgment, inammation and pain around the puncture
site, leak of ascitic uid and bile around the catheter, and loss of
uid and electrolytes. The complication rate for transhepatic biliary
drainage can be substantial and varies with the patient status prior
to the procedure as well as the diagnosis. The presence of coagu-
lopathy, cholangitis, stone, malignant obstruction or intrahepatic
lesions are all associated with high complication rates.
The advent of self-expandable metal endoprostheses, larger size
accessory channels in duodenoscopes, and the complication rate
observed with percutaneous drainage have changed the standard of
practice. Speer and colleagues conducted a prospective, randomised
study comparing percutaneous and endoscopic drainage. While
overall survival was not different between either arm, 30-day mortal-
ity, both by intention-to-treat and per-protocol analysis, was signi-
cantly lower in the endoscopy group and justied the early
termination of the study.55 The authors found that complications
associated with the percutaneous procedure accounted for the dif-
ference in mortality and that endoscopic insertion of a stent was
safer and more likely to succeed.55 In contradistinction, a recent RCT
showed that patients undergoing percutaneous drainage had a
longer survival than those in the endoscopy group.56
However, a number of possible confounders of outcome need to
be examined: First, the authors selected not only patients with unre-
sectable distal biliary obstruction but also subjects with more proxi-
mal obstruction including hilar tumors. The latter may be more
amenable to percutaneous management. These inclusion criteria
could explain the low success rate of PE stent insertion by endoscopy
(58%) which, in turn, accounted for the suboptimal efcacy observed
in this group. Secondly, the authors used SEMS in the percutaneous
group and PE stents in the endoscopic drainage group, which biases
against the endoscopic approach even though it may represent that
institutions practice. The economic evaluation of this study is also
difcult to interpret as it did not include procedural costs, and may
thereby have further disadvantaged the endoscopic treatment arm.
At present, there is insufcient evidence in the literature to advocate
the routine use of percutaneous drainage as the preferred approach
in the palliation of patients with distal biliary obstruction other than
for reasons of institutional expertise or availability.
Surgical palliation
Historically, surgery was the favored method of palliation, but has
been replaced by percutaneous and endoscopic insertion of stents.57
The 30-day mortality after surgical palliation for pancreatic cancer
and cholangiocarcinoma is signicant, especially in the face of
advancing age and metastatic disease. Surgical biliary and gastroin-
testinal bypass have been advocated for patients who also suffer from
chronic pain, since celiac nerve block can also be performed at the
time of surgery. Whether prophylactic gastrointestinal bypass should
be offered to patients with malignant obstructive jaundice, and if so,
when, remains unclear. Recent studies have shown that gastro-
jejunostomy, in addition to biliary bypass may decrease the inci-
Chapter 27 Malignant Biliary Obstruction: Distal
dence of late gastric outlet obstruction without higher morbidity or
mortality.58 Surgery has the advantage of precluding multiple re-
interventions, associated with less invasive procedures, namely
endoscopic stenting. Three prospective randomized trials have com-
pared open surgery to endoscopic stenting.5961 Smith and colleagues
randomized 203 patients to 10 French (Fr) plastic stents or biliary
bypass (choledochoduodenostomy and choledochojejunostomy).
Patients who underwent stent placement had lower procedure-
related and major complication rates as well as a shorter hospital
stay than the surgical group. Most complications occurred in the rst
30 days in the surgical group. In contrast to the endoscopy group,
therefore numerically fewer late complications due to cholangitis or
gastric obstruction. Shepherd and Andersen conducted smaller
studies that have shown similar results. While overall survival did
not differ between treatments, they demonstrated that endoscopic
stenting had a lower rate of short-term complications than surgical
treatment. Although patients in the endoscopy group had more
obstructions and needed more re-interventions, the total number of
days in hospital was higher in the surgical group. A meta-analysis
performed with these three studies conrmed a higher likelihood of
intervention in the stent group.62 The recent Cochrane systematic
review by Moss et al. suggested a lower recurrent biliary obstruction
rate attributable to surgery, with a higher overall complication rate,
and no difference in mortality when compared to plastic biliary
stenting.49a Although performed more than 10 years ago, before the
advent of newer stent technologies and less invasive surgical proce-
dures, these studies suggest that endoscopic prostheses are effective
and less costly than surgery. A recent, single-center, retrospective
cost-analysis in the US also revealed a striking difference between
endoscopic palliation and surgery despite the need for repetitive
interventions and readmissions in the endoscopic group.63 However,
surgical bypass remains an excellent alternative and may be favored
in patients with unresectable disease at the time of laparotomy, and
for those requiring concomitant gastrointestinal bypass and/or
celiac nerve block for management of chronic pain.
Adjuvant therapy
While the cases of biliary obstruction due to lymphomas can be
managed with stent insertion or surgical bypass, cure can only be
achieved with remission of the underlying disease. Responsiveness
to chemotherapy is the main predictor for outcome in these patients.
In contrast, cure of tumors of epithelial origin can only be achieved
with surgical resection, even though adjuvant chemotherapy has
been shown to improve 1- and 5-year survival after resection of
pancreatic adenocarcinoma.64 The role of chemotherapy in patients
with unresectable disease is still limited. Studies have shown that
5-Fluoracil (5-FU) based regimens are superior to observation or
supportive treatment in patients with unresectable adenocarcinoma
of the pancreas. Unfortunately, the combination of other chemo-
therapeutic agents such as cisplatin with 5-FU is no better than
5-FU alone.65
In fact, this combination is associated with an increased rate
of systemic toxicity, which seems to be unrelated to the biliary
obstruction and inability to excrete the drug metabolites. An impor-
tant breakthrough in the management of advanced pancreatic
cancer occurred with the introduction of gemcitabine and other
cytotoxic drugs which have been shown to improve major symp-
toms such as pain and weight loss, clinical benet response, time
to progression, and length of survival, but maintain an acceptable
toxicity prole.66
295
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
The effect of chemotherapy in the management of malignant
biliary obstruction is unknown. Because tumor invasion into the
biliary tree is unlikely to be relieved by chemotherapy alone, a pro-
cedure to palliate the obstruction is still necessary regardless of the
administration and response to adjuvant therapy, and may in fact
be required to improve liver tests and function prior to the initia-
tion of this treatment. In contrast, addition of a chemotherapeutic
regimen for the treatment of patients with unresectable disease
could potentially result in an improvement in survival and inu-
ence the choice of palliation. There are no studies evaluating the
effect of chemotherapy on the patency of stents. While chemothera-
peutic agents are unlikely to affect the mechanisms involved in
plastic stent occlusion, reduction of the tumor mass could diminish
the probability of tumor ingrowth and prolong patency of SEMS.
It is unknown if adjuvant chemotherapy can increase the risk of
stent migration and malfunctioning as has been suggested for
esophageal malignancies.67
REFERENCES
1. Shaib Y, El-Serag HB. The epidemiology of cholangiocarcinoma.
Semin Liver Dis 2004; 24:115125.
2. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA
Cancer J Clin 2004; 54:829.
3. Ghadirian P, Lynch HT, Krewski D. Epidemiology of pancreatic
cancer: an overview. Cancer Detect Prev 2003; 27:8793.
4. Villeneuve PJ, Johnson KC, Hanley AJ, et al. Alcohol, tobacco and
coffee consumption and the risk of pancreatic cancer: results
from the Canadian Enhanced Surveillance System case-control
project. Canadian Cancer Registries Epidemiology Research
Group. Eur J Cancer Prev 2000; 9:4958.
5. Okuda K, Nakanuma Y, Miyazaki M. Cholangiocarcinoma: recent
progress. Part 1: epidemiology and etiology. J Gastroenterol
Hepatol 2002; 17:10491055.
6. Piehler JM, Crichlow RW. Primary carcinoma of the gallbladder.
Surg Gynecol Obstet 1978; 147:929942.
7. Chijiiwa K, Kimura H, Tanaka M. Malignant potential of the
gallbladder in patients with anomalous pancreaticobiliary ductal
junction. The difference in risk between patients with and without
choledochal cyst. Int Surg 1995; 80:6164.
8. Nix GA, Dubbelman C, Wilson JH, et al. Prognostic implications of
tumor diameter in carcinoma of the head of the pancreas. Cancer
1991; 67:529535.
9. Irie H, Honda H, Shinozaki K, et al. MR imaging of ampullary
carcinomas. J Comput Assist Tomogr 2002; 26:711717.
10. Mortele KJ, Ji H, Ros PR. CT and magnetic resonance imaging in
pancreatic and biliary tract malignancies. Gastrointest Endosc
2002; 56:S206S212.
11. Lopez Hanninen E, Amthauer H, Hosten N, et al. Prospective
evaluation of pancreatic tumors: accuracy of MR imaging with MR
cholangiopancreatography and MR angiography. Radiology
2002; 224:3441.
12. Mallery JS, Centeno BA, Hahn PF, et al. Pancreatic tissue sampling
guided by EUS, CT/US, and surgery: a comparison of sensitivity
and specicity. Gastrointest Endosc 2002; 56:218224.
13. Hunt GC, Faigel DO. Assessment of EUS for diagnosing, staging,
and determining resectability of pancreatic cancer: a review.
Gastrointest Endosc 2002; 55:232237.
14. Cannon ME, Carpenter SL, Elta GH, et al. EUS compared with CT,
magnetic resonance imaging, and angiography and the inuence
296
SUMMARY
Distal malignant biliary obstruction is a commonly
encountered problem facing endoscopists that requires a
multidisciplinary approach involving surgeons, radiologists and
gastroenterologists. The optimal approach to manage this
complex medical problem depends more on available
expertise and resources than the evidence in the literature.
Endoscopic techniques have advanced signicantly over the
past two decades and maintain a central role in the palliation
of distal pancreaticobiliary malignancy.
of biliary stenting on staging accuracy of ampullary neoplasms.
Gastrointest Endosc 1999; 50:2733.
15. Sewnath ME, Karsten TM, Prins MH, et al. A meta-analysis on the
efcacy of preoperative biliary drainage for tumors causing
obstructive jaundice. Ann Surg 2002; 236:1727.
16. Isenberg G, Gouma DJ, Pisters PW. The on-going debate about
perioperative biliary drainage in jaundiced patients undergoing
pancreaticoduodenectomy. Gastrointest Endosc 2002;
56:310315.
17. Hawes RH. Diagnostic and therapeutic uses of ERCP in pancreatic
and biliary tract malignancies. Gastrointest Endosc 2002; 56:
S201S2015.
18. Moon SG, Han JK, Kim TK, et al. Biliary obstruction in metastatic
disease: thin-section helical CT ndings. Abdom Imaging 2003;
28:4552.
19. Barkun J, Barkun A. Chapter 6: Jaundice. In: Wilomre D, Cheung L,
Harken A, et al. (eds) American College of Surgeons ACS
Surgeryprinciples and practice New York: Web MD,
www.acssurgery.com 2003.
20. Freeman ML, Overby C. Selective MRCP and CT-targeted drainage
of malignant hilar biliary obstruction with self-expanding metallic
stents. Gastrointest Endosc 2003; 58:4149.
21. Kuhlmann KF, de Castro SM, Wesseling JG, et al. Surgical
treatment of pancreatic adenocarcinoma; actual survival and
prognostic factors in 343 patients. Eur J Cancer 2004; 40:549558.
22. Soehendra N, Reynders-Frederix V. [Palliative biliary duct
drainage. A new method for endoscopic introduction of a new
drain]. Dtsch Med Wochenschr 1979; 104:206207.
23. Ballinger AB, McHugh M, Catnach SM, et al. Symptom relief and
quality of life after stenting for malignant bile duct obstruction.
Gut 1994; 35:467470.
24. Abraham NS, Barkun JS, Barkun AN. Palliation of malignant biliary
obstruction: a prospective trial examining impact on quality of life.
Gastrointest Endosc 2002; 56:835841.
25. Lai EC, Mok FP, Fan ST, et al. Preoperative endoscopic drainage
for malignant obstructive jaundice. Br J Surg 1994; 81:11951198.
26. Martignoni ME, Wagner M, Krahenbuhl L, et al. Effect of
preoperative biliary drainage on surgical outcome after
pancreatoduodenectomy. Am J Surg 2001; 181:5259;
discussion 87.

27. Chen VK, Arguedas MR, Baron TH. Expandable metal biliary stents
before pancreaticoduodenectomy for pancreatic cancer: a
Monte-Carlo decision analysis. Clin Gastroenterol Hepatol 2005;
3:12291237.
28. Scheeres D, OBrien W, Ponsky L, et al. Endoscopic stent
conguration and bile ow rates in a variable diameter bile duct
model. Surg Endosc 1990; 4:9193.
29. van Berkel AM, van Marle J, Groen AK, et al. Mechanisms of biliary
stent clogging: confocal laser scanning and scanning electron
microscopy. Endoscopy 2005; 37:729734.
30. Libby ED, Leung JW. Prevention of biliary stent clogging: a clinical
review. Am J Gastroenterol 1996; 91:13011308.
31. Speer AG, Cotton PB, MacRae KD. Endoscopic management of
malignant biliary obstruction: stents of 10 French gauge are
preferable to stents of 8 French gauge. Gastrointest Endosc 1988;
34:412417.
32. Kadakia SC, Starnes E. Comparison of 10 French gauge stent with
11.5 French gauge stent in patients with biliary tract diseases.
Gastrointest Endosc 1992; 38:454459.
33. Catalano MF, Geenen JE, Lehman GA, et al. Tannenbaum Teon
stents versus traditional polyethylene stents for treatment of
malignant biliary stricture. Gastrointest Endosc 2002; 55:
354358.
34. van Berkel AM, Huibregtse IL, Bergman JJ, et al. A prospective
randomized trial of Tannenbaum-type Teon-coated stents versus
polyethylene stents for distal malignant biliary obstruction. Eur J
Gastroenterol Hepatol 2004; 16:213217.
35. Pedersen FM, Lassen AT. Response. Gastrointest Endosc 2000;
51:117.
36. De Ledinghen V, Person B, Legoux JL, et al. Prevention of
biliary stent occlusion by ursodeoxycholic acid plus
noroxacin: a multicenter randomized trial. Dig Dis Sci 2000;
45:145150.
37. Halm U, Schiefke, Fleig WE, et al. Ooxacin and ursodeoxycholic
acid versus ursodeoxycholic acid alone to prevent occlusion of
biliary stents: a prospective, randomized trial. Endoscopy 2001;
33:491494.
38. Waschke K, da Silveira E, Toubouti Y, et al. The role of plastic
stents, adjuvant therapy and metal stents in distal malignant biliary
obstruction. a systematic review and series of metal-analyses.
American Journal of Gastroenterology 2004; 99:AB 74.
39. Chan G, Barkun J, Barkun AN, et al. The role of ciprooxacin in
prolonging polyethylene biliary stent patency: a multicenter,
double-blinded effectiveness study. J Gastrointest Surg 2005;
9:481488.
40. Tringali A, Mutignani M, Perri V, et al. A prospective, randomized
multicenter trial comparing DoubleLayer and polyethylene stents
for malignant distal common bile duct strictures. Endoscopy
2003; 35:992997.
41. Raju GS, Sud R, Elfert AA, et al. Biliary drainage by using stents
without a central lumen: a pilot study. Gastrointest Endosc 2006;
63:317320.
42. Huibregtse K, Carr-Locke DL, Cremer M, et al. Biliary stent
occlusiona problem solved with self-expanding metal
stents? European Wallstent Study Group. Endoscopy 1992;
24:391394.
43. Dumonceau JM, Cremer M, Auroux J, et al. A comparison of
Ultraex Diamond stents and Wallstents for palliation of distal
malignant biliary strictures. Am J Gastroenterol 2000; 95:
670676.
44. Shim CS, Lee YH, Cho YD, et al. Preliminary results of a new
covered biliary metal stent for malignant biliary obstruction.
Endoscopy 1998; 30:345350.
44a. Shah RJ, Howell DA, Desilets DJ, Sheth SG, Parsons WG, Okolo P
3rd, Lehman GA, Sherman S, Baillie J, Branch MS, Pleskow D,
Chapter 27 Malignant Biliary Obstruction: Distal
Chuttani R, Bosco JJ. Multicenter randomized trial of the spiral
Z-stent compared with the wallstent for malignant biliary
obstruction. Gastrointest Endosc 2003; 57(7):830836.
45. Isayama H, Komatsu Y, Tsujino T, et al. A prospective randomised
study of covered versus uncovered diamond stents for the
management of distal malignant biliary obstruction. Gut 2004;
53:729734.
45a. Park do H, Kim MH, Choi JS, Lee SS, Seo DW, Kim JH, Han J, Kim
JC, Choi EK, Lee SK. Covered versus uncovered wallstent for
malignant extrahepatic biliary obstruction: a cohort comparative
analysis. Clin Gastroenterol Hepatol. 2006; 4(6):790796.
45b. Yoon WJ, Lee JK, Lee KH, Lee WJ, Ryu JK, Kim YT, Yoon YB.
A comparison of covered and uncovered wallstents for the
management of distal malignant biliary obstruction. Gastrointest
Endosc 2006; 63(7):9961000.
46. Nakai Y, Isayama H, Komatsu Y, et al. Efcacy and safety of the
covered Wallstent in patients with distal malignant biliary
obstruction. Gastrointest Endosc 2005; 62:742748.
46a. Suk KT, Kim HS, Kim JW, Baik SK, Kwon SO, Kim HG, Lee DH,
Yoo BM, Kim JH, Moon YS, Lee DK. Risk factors for cholecystitis
after metal stent placement in malignant biliary obstruction.
Gastrointest Endosc 2006; 64(4):522529.
47. Baron T, Poterucha J. Insertion and removal of covered
expandable metal stents for closure of complex biliary leaks.
Clinical gastroenterology and hepatology 2006; 4:381386.
48. Davids PH, Groen AK, Rauws EA, et al. Randomised trial of self-
expanding metal stents versus polyethylene stents for distal
malignant biliary obstruction. Lancet 1992; 340:14881492.
49. Kaassis M, Boyer J, Dumas R, et al. Plastic or metal stents for
malignant stricture of the common bile duct? Results of a
randomized prospective study. Gastrointest Endosc 2003;
57:178182.
49a. Moss AC, Morris E, Leyden J, MacMathuna P. Malignant distal
biliary obstruction: a systematic review and meta-analysis of
endoscopic and surgical bypass results. Cancer Treat Rev 2007;
33(2):213221.
49b. Soderlund C, Linder S. Covered metal versus plastic stents
for malignant common bile duct stenosis: a prospective,
randomized, controlled trial. Gastrointest Endosc 2006;
63(7):986895.
49c. Waschke K, da Silveira E, Toubouti Y, Rahme E, Barkun A. Self-
expandable metal stents confer a survival advantage in palliation
of distal malignant biliary obstruction. Gastrointest Endosc 2005;
61(5):T1322.
50. Arguedas MR, Heudebert GH, Stinnett AA, et al. Biliary stents
in malignant obstructive jaundice due to pancreatic carcinoma:
a cost-effectiveness analysis. Am J Gastroenterol 2002;
97:898904.
51. Yeoh KG, Zimmerman MJ, Cunningham JT, et al. Comparative
costs of metal versus plastic biliary stent strategies for malignant
obstructive jaundice by decision analysis. Gastrointest Endosc
1999; 49:466471.
52. Prat F, Chapat O, Ducot B, et al. A randomized trial of endoscopic
drainage methods for inoperable malignant strictures of the
common bile duct. Gastrointest Endosc 1998; 47:17.
52a. Maire F, Hammel P, Ponsot P, Aubert A, OToole D, Hentic O,
Levy P, Ruszniewski P. Long-term outcome of biliary and
duodenal stents in palliative treatment of patients with
unresectable adenocarcinoma of the head of pancreas. Am J
Gastroenterol 2006; 101(4):735742.
53. da Silveira E, Waschke K, Barkun A, et al. Cost-effectiveness
decision analysis comparing covered to uncovered self-
expandable metal stents to elective or on-demand polyethylene
stent changes in patients with distal biliary malignant obstruction.
Gastrointest Endosc 2005; 61:AB 203.
297
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
54. Bornman PC, Harries-Jones EP, Tobias R, et al. Prospective
controlled trial of transhepatic biliary endoprosthesis versus
bypass surgery for incurable carcinoma of head of pancreas.
Lancet 1986; 1:6971.
55. Speer AG, Cotton PB, Russell RC, et al. Randomised trial of
endoscopic versus percutaneous stent insertion in malignant
obstructive jaundice. Lancet 1987; 2:5762.
56. Pinol V, Castells A, Bordas JM, et al. Percutaneous self-expanding
metal stents versus endoscopic polyethylene endoprostheses for
treating malignant biliary obstruction: randomized clinical trial.
Radiology 2002; 225:2734.
57. Sharma D, Bhansali M, Raina VK. Surgical bypass is still relevant in
the palliation of malignant obstructive jaundice. Trop Doct 2002;
32:216219.
58. Lillemoe KD, Grosfeld JL. Addition of prophylactic
gastrojejunostomy to hepaticojejunostomy signicantly
reduces gastric outlet obstruction in people with
unresectable periampullary cancer. Cancer Treat Rev 2004;
30:389393.
59. Andersen JR, Sorensen SM, Kruse A, et al. Randomised trial of
endoscopic endoprosthesis versus operative bypass in malignant
obstructive jaundice. Gut 1989; 30:11321135.
60. Shepherd HA, Royle G, Ross AP, et al. Endoscopic biliary
endoprosthesis in the palliation of malignant obstruction of the
298
distal common bile duct: a randomized trial. Br J Surg 1988;
75:11661168.
61. Smith AC, Dowsett JF, Russell RC, et al. Randomised trial of
endoscopic stenting versus surgical bypass in malignant low
bileduct obstruction. Lancet 1994; 344:16551660.
62. Taylor MC, McLeod RS, Langer B. Biliary stenting versus bypass
surgery for the palliation of malignant distal bile duct obstruction:
a meta-analysis. Liver Transpl 2000; 6:302308.
63. Martin RC, 2nd, Vitale GC, Reed DN, et al. Cost comparison of
endoscopic stenting vs surgical treatment for unresectable
cholangiocarcinoma. Surg Endosc 2002; 16:667670.
64. Shore S, Raraty MG, Ghaneh P, et al. Review article: chemotherapy
for pancreatic cancer. Aliment Pharmacol Ther 2003; 18:10491069.
65. Ducreux M, Rougier P, Pignon JP, et al. A randomised trial
comparing 5-FU with 5-FU plus cisplatin in advanced pancreatic
carcinoma. Ann Oncol 2002; 13:11851191.
66. Heinemann V. Gemcitabine in the treatment of advanced
pancreatic cancer: a comparative analysis of randomized trials.
Semin Oncol 2002; 29:916.
67. Siersema PD, Hop WC, Dees J, et al. Coated self-expanding metal
stents versus latex prostheses for esophagogastric cancer with
special reference to prior radiation and chemotherapy: a
controlled, prospective study. Gastrointest Endosc 1998;
47:113120.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Malignant Biliary Obstruction: Hilar
28 Giovanni D. De Palma
INTRODUCTION
BOX 28.1 KEY POINTS
1. Review of epidemiology of malignant bile duct obstruction
at the hepatic hilum.
2. Review of anatomical classication.
Epidemiology of malignant bile duct obstruction
Malignant biliary obstruction at the liver hilum is caused by a het-
erogeneous group of tumors that includes primary bile duct cancer
(the so-called Klatskin tumor), cancers that involve the conuence
by direct extension (e.g., gallbladder and liver cancer), and metastatic
cancer to hilar lymphatic nodes or to the liver (Table 28.1).
Primary cholangiocarcinoma of the hepatic hilus affecting either
the right or left main hepatic ducts or the biliary conuence was rst
described in 1957 by Altemeier et al. Other investigators have also
documented the existence of this tumor, but it was not until 1965
that Klatskin1 reported the rst comprehensive examination of the
pathology, diagnosis, and management of the tumor that now bears
his name.
Cholangiocarcinoma is an uncommon malignancy comprising
less than 2% of all cancer diagnoses. The overall rate of occurrence
of cholangiocarcinoma is 1.2/100 000 individuals, with two-thirds of
all cases occurring in patients more than 65 years old, and a near
tenfold increased rate of occurrence in patients more than 80 years
of age.2
In recent years, there has been a worldwide trend towards
decreased mortality from extrahepatic tumors, particularly for
females. In contrast to the observations in intrahepatic cholangio-
carcinoma, the estimated annual percentage change in mortality for
extrahepatic biliary tract cancers decreased in most countries, with
the exception of the United Kingdom.3
Chronic biliary tract inammation represents a major risk factor
for the development of cholangiocarcinoma. The association between
chronic parasitic infection of the biliary tract and cholangiocarci-
noma is obvious in regions of high endemicity, such as in certain
Far Eastern nations. In Western nations, primary sclerosing cholan-
gitis is the most common risk factor identied with the development
of cholangiocarcinoma.4
Extrahepatic cholangiocarcinoma has traditionally been separated
into three groups, based on anatomical location. Upper third or hilar
tumors are those located in the common hepatic duct and/or the
right and left hepatic ducts including their conuence. Middle third
tumors occur in the region bounded by the upper border of the
duodenum and extending to the common bile duct. Lower third or
distal bile duct tumors arise between the ampulla of Vater and the
upper border of the duodenum.
Tumors at the biliary conuence of the liver are the most common
and comprise 4060% of the total. Middle third and distal third
tumors comprise 1720%, and 1827%, respectively. A small per-
centage of patients (<10%) have diffuse tumors involving the entire
extrahepatic bile duct.2.4,5
Anatomical classication
The extent of duct involvement by perihilar tumors may be classied
as suggested by Bismuth and Corlette6 (Fig. 28.1):
A. type I: tumors below the conuence of the left and right hepatic
ducts (ceiling of the biliary conuence is intact; right and left
ductal systems communicate);
B. type II: tumors reaching the conuence but not involving the left
or right hepatic ducts (ceiling of the conuence is destroyed; bile
ducts are separated);
C. type III: tumors occluding the common hepatic duct and either
the right (IIIa) or left (IIIb) hepatic duct;
D. type IV: multicentric tumors or tumors involving the conuence
and both hepatic ducts, the right one and the left one.
MANAGEMENT STRATEGIES FOR HILAR TUMORS
BOX 28.2 KEY POINTS
1. Understanding the role of imaging studies.
2. Review of the management strategies for resectable
tumors.
3. Review of the pros and cons of alternative palliative
techniques.
Hilar tumors have proven to be a challenge to treat and manage
because of their poor sensitivity to conventional therapies and our
inability to prevent or to detect early tumor formation. Untreated
patients usually die within six months to a year of diagnosis.
The range of therapeutic modalities varies from a curative
approach by performing extensive liver resectionsin some
299
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
cases even total hepatectomy and liver transplantationto a more
palliative approach in which a surgical bypass or even percutaneous
or endoscopic stent insertion is undertaken, with or without
radiotherapy.
All patients should be fully evaluated for resectability before any
type of intervention is performed because stent-associated inam-
mation or infection often makes assessment more difcult.
Patients being evaluated for resectability must at rst be physio-
logically suitable for a potential operative resection that may include
a partial hepatectomy. The patients nutritional status and risk of
postoperative liver failure are important factors to consider before
proceeding to exploration for resection. A retrospective review of
resected hilar cholangiocarcinoma cases demonstrated that a preop-
erative serum albumin level <3 g/dl and a total bilirubin level
>10 mg/dl were both associated with poorer survival.7
All available data must be used to distinguish resectability (local-
ized cancers) from unresectability (Table 28.2). The preoperative
evaluation of a patient with suspected hilar tumors is directed toward
the following three primary objectives: (1) an assessment of the
extent and level of biliary tract and portal vein involvement; (2) an
assessment of the liver status to detect evidence of lobar atrophy or
concomitant liver pathology; (3) an evaluation of the presence and
extent of nodal disease and/or distant metastases.
Radiological studies
Radiographic studies are essential in therapeutic planning of patients
with hilar cholangiocarcinoma.811
Most jaundiced patients undergo initial transabdominal ultra-
sound (US) before being referred to a hepatobiliary specialist. In
patients with hilar lesions US typically demonstrates intrahepatic
bile duct dilation and normal diameter of extrahepatic ducts, or dila-
Malignant Non-malignant
Cholangiocarcinoma Sclerosing cholangitis
Gallbladder carcinoma Inammatory strictures
Nodal mets at porta hepatis Post operative
Hepatocellular carcinoma Post radiation chemotherapy
Hepatic metastases Carolis disease
Metastase to biliary tree Retroperitoneal brosis
AIDS
Table 28.1 Hilar strictures
Type I Type II Type IIIa
Fig. 28.1 Schematic representation of Bismuth classication of hilar cholangiocarcinoma. Type I: tumors below the conuence of the left
and right hepatic ducts (ceiling of the biliary conuence is intact; right and left ductal systems communicate); type II: tumors reaching the
conuence but not involving the left or right hepatic ducts (ceiling of the conuence is destroyed; bile ducts are separated); type III: tumors
occluding the common hepatic duct and either the right (IIIa) or left (IIIb) hepatic duct; type IV: multicentric tumors or tumors involving
the conuence and both hepatic ducts, the right one and the left one.
300
tion of both intrahepatic and extrahepatic ducts in more distal
lesions. Centers with expertise in duplex ultrasound have found that
this method is an accurate predictor of vascular involvement and
resectability. Duplex ultrasonography can identify the site of biliary
obstruction, as well as the presence or absence of portal venous
involvement, with 93% sensitivity and 99% specicity.
Contrast CT is sensitive to detect bile duct tumors, the level of
biliary obstruction, and the presence of liver atrophy. In addition,
CT may also permit visualization of the pertinent nodal basins.
Performance of a triple-phase helical CT will detect essentially all
cholangiocarcinomas greater than 1 cm. However, CT may only be
able to establish resectability in approximately 60% of patients. Dila-
tion of the intrahepatic bile ducts in a single, small hepatic lobe with
hypertrophy of the contralateral lobe suggests the atrophy-hypertro-
phy complex, as seen with tumors chronically obstructing a single
lobe and invading the ipsilateral portal vein. Bilobate dilated intra-
hepatic ducts and a normal or collapsed gallbladder and common
bile duct suggest a perihilar tumor.
At the present time magnetic resonance (MR) is the optimal
initial investigation for suspected hilar cholangiocarcinoma. MR
permits excellent visualization of hepatic parenchymal abnormali-
ties, as well as the visualization of the biliary tree and vascular
structures. MR with the use of ferrous oxide and gadolinium yields
information similar to that yielded by CT, cholangiography, and
angiography combined (Fig. 28.2). MR cholangiopancreatography
(MRCP), has the capability to evaluate the bile ducts both above and
Medical co-morbidities limiting the patients ability to undergo
major surgery
Signicant underlying liver disease prohibiting liver resection
necessary for curative surgery based on preoperative imaging
Bilateral tumor extension to secondary biliary radicals
Encasement or occlusion of the main portal vein
Lobar atrophy with contralateral portal vein involvement
Contralateral tumor extension to secondary biliary radicals
Evidence of metastases to N2 level lymph nodesa
Presence of distant metastases
Table 28.2 Criteria for unresectability in patients with hilar
cholangiocarcinoma
a
N2 lymph nodes, metastasis in the peripancreatic (head only), paraduodenal,
periportal, celiac, superior mesenteric, and/or posterior pancreaticoduodenal lymph
nodes.
Type IIIb Type IV Type IV

A B
Fig. 28.2 Contrast enhanced MR A and MRCP B showing tumor
involving the conuence and both hepatic ducts, the right one and
the left one. Courtesy of Mainenti P., MD, and Maurea S., MD. Dipar-
timento di Scienze Biomorfologiche e Funzionali; University of
Naples Federico II, School of Medicine.
below a stricture. The efcacy of MRCP as a non-invasive means of
acquiring reliable and precise information about the anatomy of
both the intrahepatic and the extrahepatic biliary tree, has been well
documented and has almost totally replaced percutaneous and endo-
scopic cholangiography.
EUS has not been proven to offer more than using other imaging
modalities in patients with suspected hilar cholangiocarcinoma. One
small series suggests that EUS may allow a denitive histological
diagnosis to be made in patients with hilar tumors.12
Intraductal US (IDUS), at the time of ERCP, may add useful
information in the patient with a suspected pancreaticobiliary malig-
nancy, especially cholangiocarcinoma. However, there are limited
data to date, the exact role has not been dened yet, and the avail-
ability of this technology is limited to specialized centers.
Positron emission tomography (PET) using the radionucleotide
tracer 18-uorodeoxyglucose (FDG) can reliably detect cholangio-
carcinomas as small as 1 cm. A recent study has demonstrated
that preoperative staging using FDG PET detected distant meta-
static disease that was not suspected based on other radiological
studies in 30% of patients.13
PET may be useful to detect primary cholangiocarcinoma in
patients with PSC.
Preoperative histological conrmation
The diagnosis of malignant biliary strictures depends on the identi-
cation of tumor cells obtained by ultrasound or CT-guided percu-
taneous ne needle aspiration, bile sampling, endobiliary brushings,
or bile duct biopsies. Percutaneous needle biopsies are reliable only
if US or CT identify a malignancy (sensitivity of 50%).
Bile samples, obtained through a percutaneous or endoscopic
stent, contain cancerous cells in 3040% of cases of cholangiocarci-
noma. The use of brush biopsy and cytologic examination may
increase the yield to 4070%. Unfortunately, even percutaneous or
endoscopic biopsy not infrequently yields non-diagnostic tissue
because of the desmoplastic nature of the lesion.
There are currently many problems with tissue sampling during
ERCP. Despite combination sampling with brush cytology, FNA,
and biopsy, the sensitivity for all three combined is only 62% with
a negative predictive value of 39%.14,15 Additionally, multiple sam-
pling requires considerable time and technical expertise and there
is a risk of losing guidewire access across the biliary stricture. Finally,
Chapter 28 Malignant Biliary Obstruction: Hilar
forceps biopsy generally requires a sphincterotomy, which adds a
small risk of bleeding and perforation and is not required for sub-
sequent stent placement. As a result of time and technical consider-
ations, most practitioners perform brush cytology alone, which has
sensitivity as low as 30%.
Whereas the vast majority of extrahepatic strictures, particularly
hilar strictures, are the result of a cholangiocarcinoma, histological
diagnosis is not mandatory before exploration. At the present time,
particularly if preoperative biliary stents are placed, distinguishing
between alternative diagnoses can be difcult, if not impossible,
even intraoperatively. In the absence of clear evidence of unresect-
ability, all suspected lesions should be considered for resection.
Surgery for localized cancer
Surgery remains the only intervention offering the possibility of a
cure. The main treatment goal should be complete excision with
negative margins. Several studies have concluded that radical exci-
sion of the lesion offers the best treatment option with respect to
long-term survival.16,17 However, curative resections are difcult to
achieve owing to the unfavorable location of the tumor, its tendency
to grow into the perineural tissue, and to inltrate proximally into
the biliary tree and the liver. A signicant number of patients have
peritoneal implants or locoregional lymph node involvement that is
not easily detected on preoperative imaging studies.
The surgical treatment of perihilar cholangiocarcinoma depends
on the Bismuth class. Most authors report that about one-third of
patients can undergo curative resection, but some suggest that up
to two thirds of patients should undergo resection with curative
intent. En bloc resection of the extrahepatic bile ducts and gallblad-
der, regional lymphadenectomy, and Roux-en-Y hepaticojejunos-
tomy are recommended for type I and II tumors, and that treatment
plus hepatic lobectomy is recommended for type III tumors. The
intent is to achieve a tumor-free proximal margin of at least
5 mm. Because type II and III tumors often involve the ducts of
the caudate lobe, caudate lobectomy is recommended to improve
local control and survival for patients with type II or III tumors.
However, extended resection has shown an increased surgical
risk with higher morbidity (overall up to 45%) and mortality rates
of 333%.13,18
Liver transplantation
Orthotopic liver transplantation (OLT) has been performed for both
resectable and unresectable cholangiocarcinomas. Liver transplanta-
tion is currently contraindicated because it is usually associated with
rapid recurrence of disease and death within three years.19 In pilot
studies, OLT following preoperative chemoirradiation for unresect-
able cholangiocarcinoma has resulted in long-term survival of
carefully selected patients and may be appropriate within clinical
trials.20
Surgery for unresectable cancer
Palliative options include operatively placed trans-tumoral stents, or
the performance of an operative bilioenteric bypass. Several surgical
techniques have been described for intrahepatic biliary bypass. The
two most common approaches are bypass to the segment III duct
and the right sectoral hepatic ducts. The rst one, because of the
more constant anatomy and long extrahepatic course of the left
hepatic duct, is technically easier and preferred. However, the type
of bypass is usually dictated by the location of the tumor. In general,
segment III bypass is performed unless the left lobe is atrophic or
301
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 28.3 Percutaneous drainage. A X-ray showing internal-external drainage of Bismuth type-IV hilar stricture secondary to colonic
hepatic metastases (recurrence after liver resection) (courtesy of Iaccarino V, MD, UOC of Radiology. University of Naples Federico II,
School of Medicine). B Endoscopic view of duodenal end of internal-external prosthesis.
heavily involved with tumor or if the primary lesion extends to the
umbilical ssure of the liver.
Percutaneous approach (see Fig. 28.3)
Distinction should be made between three transhepatic procedures:
(1) external drainage; (2) external-internal drainagea long catheter
is placed through the obstruction into the duodenum, so that bile
can ow internally, but the proximal end still exits through the skin
for ushing, cholangiography and exchange; (3) internal drainage
a prosthesis is placed by the percutaneous route without subsequent
external access.
Adjuvant therapy
At this time, there is no effective adjuvant therapy for cholangio-
carcinoma. No prospective randomized trials that demonstrate a
therapeutic role for chemotherapy, brachytherapy, or external beam
radiation have been published. However, given the poor prognosis
associated with this diagnosis, patients should be encouraged to
enroll in phase I or II clinical trials.
Recently, photodynamic therapy has been used in the palliative
management of extrahepatic cholangiocarcinoma with promising
results.2123
ENDOSCOPIC APPROACH TO MALIGNANT
STRICTURES AT THE HEPATIC HILUM
BOX 28.3 KEY POINTS
1. Understanding the role of endoscopic management.
2. Review of the current techniques for stent implantation.
302
3. Identing the criteria for stent selection (plastic versus metal
stent).
Endoscopic drainage of malignant
hilar obstruction
Endoscopic stent drainage has been proposed as an alternative to
biliary-enteric bypass surgery and percutaneous drainage to palliate
malignant biliary obstruction. In addition, alternative approaches to
biliary stent placement have been compared with particular interest
in determining optimal stent material, design, and placement
strategies.2429
Prosthetic palliation of patients with malignant hilar stenoses
poses particular difculties, especially in advanced lesions (type II
lesions or higher). The risk of cholangitis after contrast injection
into the biliary tree in cases where incomplete drainage is achieved
is well known. Retention of contrast and subsequent segmental
cholangitis is a risk associated with endoscopic attempts to treat
advanced hilar lesions and this has prompted some to question the
role of endoscopic drainage in this situation.30
Some studies suggest that patients undergoing stent placement
for malignant low bile duct obstruction had signicant improve-
ment in abdominal discomfort, weight loss, or anorexia and
sleep patterns, in addition to the expected improvement in pruritus
and jaundice.31 Similar studies would be needed to conrm that
endoscopic stent placement of hilar obstruction is associated with
an improved quality of life and can be justied by economic
considerations.
Although, metabolic and immune parameters appear improved
with biliary drainage, there has been no evidence that endoscopic
stent placement translates into prolonged survival.
 Chapter 28 Malignant Biliary Obstruction: Hilar

Shepherd (1988) Andersen (1989) Smith (1994)

surgery stent surgery stent surgery stent
Patients (n) 19 25 25 23 103 101
Successful drainage (%) 84 96 92 91 91 94
Complications (%) na na 40 22 28 10
30-day mortality (%) na na 20 9 17 7
Duodenal bypass (%) 0 0 0 0 1 6
Recurrent jaundice (%) 16 28 2 17 3 18
Median survival (days) 100 84 125 152 150 150

Table 28.3 Treatment outcomes: comparison of endoscopic stent placement versus surgical bypass (randomized controlled trials)
na = not available

The success rate of plastic stent insertion is around 80% in
patients with proximal tumors. Relief of symptoms can be achieved
in nearly all patients successfully stented. Failure of endoscopic
drainage may be due to previous gastric surgery, presence of peri-
ampullar diverticulum, duodenal obstruction, failure of cannulation
of the common bile duct, or inability to pass a guidewire or push a
stent through the stricture.
Endoscopic drainage compared with
alternative techniques
(a) Endoscopic stenting versus surgical bypass
Prospective comparisons between surgery and stents do not exist for
proximal bile duct obstruction. There have been three prospective
studies comparing biliary stents with surgical bypass for distal bile
duct obstruction (Table 28.3). These studies demonstrate that surgery
is associated with greater early morbidity and mortality but greater
long-term patency and a lower incidence of recurrent jaundice.
There was no long-term survival advantage for either technique.
Neither of the randomized trials looked at cost. Two retrospective
studies both found lower cost for the endoscopic approach.32,33 A
meta-analysis suggested that patients who live less than six months
would be best served with endoscopic stent placement, whereas
those who live longer would benet most by surgical bypass. Unfor-
tunately, there were insufcient data in existing trials to draw any
rm conclusions.34 There have also been no randomized compara-
tive studies of endoscopic stent placement with expandable metal
stents and surgery.
(b) Endoscopic versus percutaneous stenting
No direct comparative studies exist between endoscopic and percu-
taneous placement of biliary stents for proximal bile duct obstruc-
tion. Percutaneous transhepatic stent placement has been compared
with endoscopic stent placement showing signicant advantages for
the endoscopic approach (higher success in relieving jaundice and
lower complication rate).35 These studies were carried out in the era
of large, plastic stents, which are clearly disadvantageous for the
transhepatic route. With the introduction and widespread use of
smaller expandable metal stents, percutaneous placement is now
associated with fewer risks and complications including hemor-
rhage, cholangitis, and pneumothorax. Studies with metal stent
technology have not been reported primarily because it is now
accepted that endoscopic access is less invasive than the percutane-
ous transhepatic route, offers the advantage of favorable uid and
electrolyte homeostasis, more satisfactory cosmetic results, and
greater psychological acceptance.
Preoperative stenting
The role of preoperative stenting is controversial. It is generally
accepted that preoperative biliary instrumentation is associated with
an increase in the bile colonization rate and an increase in periopera-
tive infectious complications, although mortality is not increased.
Most surgeons will agree that those patients with renal impairment,
cholangitis, or signicant pruritis (where an operation cannot be
scheduled in reasonable time) should undergo preoperative
drainage.
In cases of preoperative stenting, use of metal stents should be
avoided because initial insertion of an expandable metal stent makes
subsequent surgery more difcult as these stents cannot readily be
removed surgically.
Technique of stent implantion
The options include draining only the left hepatic system, draining
only the right hepatic system, or draining both systems.
The decision whether to place a single biliary stent or multiple
stents depends initially on the location of the stricture in the
biliary tract.
In patients who have strictures that do not involve the conuence
of right and left hepatic ducts (Bismuth type I hilar strictures), jaun-
dice can be palliated completely with a single biliary stent because
both the right and left intrahepatic ductal systems are in communi-
cation (Fig. 28.4).
In patients who have more complex strictures (Bismuth type II
to IV strictures) the central question is whether adequate palliative
relief of obstruction requires the placement of two endoprostheses
(Fig. 28.5), one to drain the left system and one to drain the right,
or if one prosthesis placed in either system will sufce.
Palliation of jaundice generally requires drainage of 1/4 to 1/3 of
a healthy liver, or proportionally more in those with underlying
dysfunction. Hence unilateral drainage is usually adequate, and
many studies have reported good results using a single stent in
about 80% of patients with type II and III tumors. No difference in
efcacy has been shown between single stent placement in the left
or the right system.
Really, the necessity to ensure the drainage of both systems,
including additional endoscopic or percutaneous stent, if necessary,
pertains more to the prevention of procedure-induced cholangitis
caused by contrast injection in undrained biliary branches than to
effective palliation. Generally, if both lobes are imaged with contrast

303
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B C

Fig. 28.4 Unilateral stent implantation. A ERCP showing Bismuth type I hilar stricture. B A guidewire and a guide-catheter advanced
through the stricture in the left hepatic duct. C Post-ERCP x-ray image showing single 10 Fr plastic stent in place.

A B

C E
D

Fig. 28.5 Bilateral stent implantation. (a) ERCP showing Bismuth type-II hilar stricture. (b) Endoscopic view of two guidewires respectively
inserted in the right and left hepatic duct. (c) Endoscopic view of second stent insertion. (d) Two stents have been selectively placed into
each major hepatic duct. (e) Endoscopic view of the two stents in place.

during cholangiography bilateral stenting reduces the potential Patient preparation

sequelae of cholangitis in contaminated but undrained areas. If
contrast does not contaminate both sides then unilateral stenting
should be sufcient.30,3639
Patients with multiple intrahepatic strictures may not benet
from any type of drainage procedure if several segments (>1/4)
always remain undrained (Fig. 28.6). In the absence of intract-
able symptoms, these patients should not undergo further
endoscopic measures, as the risk of inducing cholangitis outweighs
any benets realized from the establishment of endoscopic
drainage.
The patients should have an intravenous line for administration of
sedatives, antibiotics and hydration. Antibiotic coverage is manda-
tory, particularly in those patients with more complex strictures
(Bismuth type III and IV). Prophylaxis can be given as a single,
adequate dose shortly before the procedure and should be contin-
ued for 45 days after the procedure. Escherichia coli, and to a lesser
extent, Klebsiella spp. (gram-negative bacteria) and gram-positive
Enterococcus spp. are the most common organisms in bile. There-
fore, antibiotics should be aimed mainly at gram-negative bacteria
with good penetration in liver tissue and bile. Ciprooxacin is

304

Fig. 28.6 ERCP showing patient with multiple intrahepatic
strictures
currently the rst choice of antibiotic. In case of cholangitis, the
addition of amoxicillin or a switch to piperacillin/tazobactam is
advisable, Patients should be routinely sedated with diazepam or
midazolam, sometimes combined with fentanyl or pethidine. The
patients should be monitored by an assistant and by mechanical
methods including pulse oximetry. Supervision by an anesthetist
may be required.
MRCP and CT-guided stent implantation (Fig. 28.7)
Recent reports describe the utility of MRCP or CT imaging to guide
selection of the target lobe for subsequent endoscopic stenting, often
without use of contrast.37,39
MRCP or CT images are used to conrm the diagnosis of Klatskin
tumor to exclude other biliary diseases (Fig. 28.1) and to demon-
strate the stenoses as well as dilation of proximal liver segments.
The left or right main hepatic duct is chosen for stent insertion,
depending on the number of drainable liver segments. Subsequent
to MRCP selective endoscopic retrograde contrast injection is
deliberately limited to the distal end of the malignant tumor
stenosis.
Thereafter, sphincterotomy is generally performed, the papillo-
tome or a catheter is advanced to the distal margin of the stricture
and a guidewire (hybrid or hydrophilic, with a torquable angle-tip if
necessary) is advanced, under uoroscopic guidance, in the direction
of the duct preselected for drainage based on prior imaging. Once
Chapter 28 Malignant Biliary Obstruction: Hilar
the guidewire passes through the stricture in the desired direction,
it is advanced as deeply as possible into that lobe. Then a catheter is
advanced over the guidewire and through the stricture as far as pos-
sible, the guidewire is removed, and as much bile as possible is
aspirated to decompress the accessed duct. Contrast is injected with
the catheter and the unilateral cholangiogram is completed. Subse-
quently, a stiff guidewire is substituted for the initial guidewire and
the catheter removed, leaving the guidewire in that duct for the
remainder of the procedure until nal stent deployment. Thereafter,
if necessary, dilation of the malignant stenosis is performed using
either balloon catheters or bouginages. If histological diagnosis is
not already established, sampling is performed with a biopsy forceps
and cytology brush. Finally a plastic or a metal stent is inserted to
decompress the proximal ductal system.
If bilateral stent placement was planned, immediately after inser-
tion of the rst guidewire a second guidewire is inserted into the
contralateral side, stents are placed sequentially into the left and then
the right hepatic ducts over dual guidewires.
Unilateral random stent implantation (Figs 28.828.9)
MRCP images are used to conrm the diagnosis of Klatskins
tumor, to exclude other biliary diseases, and to demonstrate the
stenoses, as well as dilation of proximal liver segments. Contrast
injection at ERCP is deliberately limited to the extrahepatic bile
duct distal to the tumor. Then, sphincterotomy is performed in
all cases, and a guidewire is subsequently advanced through the
malignant stenosis into the duct that is technically easiest to access.
A catheter is then passed over the guidewire and through the
stenosis, and, after removal of the guidewire, a unilateral cholangio-
gram is completed. Finally, a single plastic or metallic stent is
deployed.36,38
Contrast-free stent implantation (Fig. 28.10)
Stents are placed in these patients under uoroscopic guidance as
follows: the stent assembly is passed over the guidewire above the
suspected site of stricture (Fig. 28.10A) and the stent is deployed at
the desired site.40
Rendezvous technique
An interventional radiologist passes a guidewire transhepatically
down the bile duct and into the duodenum; this wire is then grasped
by the endoscopist to place stents in the bile duct.
The combined percutaneous-endoscopic approach has been
reported by many groups. The rationale is that the complications
should be lower than those with a purely percutaneous approach,
since only small catheters are passed through the liver, and rather
briey. However, the complication rates are not negligible. The
overall morbidity, including the initial failed endoscopic attempts,
proved to be 62% and 27% in one series of 74 patients.
Plastic versus metal stents
Theoretically, a metal stent should result in better drainage than
plastic stent in hilar strictures.
Metal stents have two advantages over the plastic stents: they
do not occlude side branches because of the mesh; furthermore,
because most hilar tumors are rm and scirrhous, tumor ingrowth
probably occurs less frequently.
Metallic stents offer longer but still limited stent patency duration
of about 46 months compared with a patency duration of 24
months for plastic stents (Table 28.4).
305
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B

C D

Fig. 28.7 MRCP-targeted unilateral implantation of SEMS. A Contrast enhanced MR showing hilar cholangiocarcinoma causing marked
dilation of the bile ducts in the lateral segments of the right lobe. B MRCP image showing tumor occluding the common hepatic duct
and the right (IIIa) hepatic duct (same patients as in Fig. 28.5A). (A and B courtesy of Mainenti P, MD, and Maurea S, MD. Dipartimento di
Scienze Biomorfologiche e Funzionali; University of Naples Federico II, School of Medicine.) C The stent is advanced over the guidewire
into dilated right hepatic duct as planned. D The SEMS has been deployed. E Post-ERCP x-ray image showing SEMS in place providing
drainage.

306
 Chapter 28 Malignant Biliary Obstruction: Hilar

A C
B

Fig. 28.8 Retrograde cholangiograms illustrating unilateral contrast injection and unilateral stent deployment. A Contrast injected into
distal common duct distal to stenosis. Stenosis itself and the more proximal bile duct regions are deliberately not opacied. B Guidewire
is advanced through tumor stenosis toward left hepatic duct under uoroscopic control. A minimal contrast injection into dilated left
hepatic ductal system shows correct position of the guidewire and guiding catheter. Injection of contrast into right hepatic duct system
has been avoided. C Stent advanced through stenosis into dilated left hepatic duct.

B C

Fig. 28.9 Retrograde cholangiograms illustrating unilateral stent deployment in a patient with type IV Bismuth hilar stricture. A ERCP
showing dilated biliary tree with high-grade stricture affecting the distal common bile duct, caused by carcinoma of the pancreas.
Narrowed common hepatic duct and dilation of proximal biliary tree are compatible with metastatic disease producing obstruction
at the bifurcation. B A guidewire and a guide-catheter advanced through the stricture in the left hepatic duct. C Single 10 Fr plastic
stent advanced through stenosis.

In contrast to plastic stents, metallic stents are not removable
after the rst few days of deployment, as the stent becomes embed-
ded in the tumor tissue which may grow into each individual mesh
opening. Thus, metallic stents should be used in patients with
proven unresectable malignancies, because initial insertion of an
expandable metal stent makes subsequent surgery more difcult as
these stents cannot readily be removed surgically.
The main disadvantage is the cost of the metallic stent (US$
9001200), and identication of patients who are likely to outlive
their rst plastic stent, and warrant a metal stent, is a major chal-
lenge for the managing clinician. Cost analysis showed that metallic
stents were advantageous versus plastic stents in patients surviving
more than six months and very costly when patients survived less
than three months. Therefore, the use of metal stents should be
restricted to those patients with unresectable tumors who will, in all
probability, live longer than three months. Unfortunately there is no
a good way to predict life expectancy at this time. Tumor size (>3 cm),
evidence of diffuse liver metastases, and general condition of the
patient could guide the choice of stent.
An additional indication for the use of metal stents is the small
group of patients who suffer rapid and repeated obstruction of
plastic stents. These patients have not been well studied and pres-
ently cannot be identied at the start. This group constitutes patients
who will also benet from a metal stent. All patients, who need a
stent exchange because of clogging of a plastic stent within 1 month
after insertion are good candidates for metal stent insertion.

307
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A 30-days
Patients Drainage mortality Occlusion Patency
(N) (%) (%) (%) (days)
Davids (1992)
Metal 49 96 14 33 372
Plastic 56 95 4 54 126
Carr-Locke (1993)
Metal 86 98 5 13 111
Plastic 78 95 5 13 62
Knyrim (1993)
Metal 31 100 13 22
Plastic 31 100 9 43

Table 28.4 Treatment outcomes: comparison of plastic versus

metal stents for low bile duct obstruction (randomized controlled
trials)

Fig. 28.11 Duonenal perforation by guidewire in a patient with

high-grade stricture affecting common hepatic duct and dilation of
proximal biliary tree.

Fig. 28.10 Contrast-free deployment of stent. A A guidewire is

passed above the suspected site of stricture, deeply in the left
hepatic duct. B Stent is passed over the guidewire above the
stricture.
COMPLICATIONS
BOX 28.4 KEY POINTS
1. Review of the complications of the endoscopic approach.
2. Identifying strategies to reduce complications and their
management.
Early and late complications of stent insertion
Immediate complications of attempted and successful stent place-
ment are similar to those of other ERCP procedures. Pancreatitis
can be provoked. A small sphincterotomy, when performed,
rarely results in direct complications, such as bleeding or perforation
(Fig. 28.11). In contrast, post-ERC bacterial cholangitis in patients
with Klatskin tumors occurs in 1749%.5,23
Bacterial cholangitis is caused by infected bile and inammation
of ductal epithelium. In animal studies bacterial reux from bile to
blood is enhanced by increased intrabiliary pressure.16 This suggests
that increased intrabiliary pressure during ERC is the main reason
for increased bacterial access to the blood.
The major late complication is clogging of the prosthesis,
occurring in 2136% of cases. Much higher rates of 2154%

308

are reported in prospective randomized studies, with an overall
incidence of 42%.
Stents placed for hilar obstruction appear to occlude faster than
stents placed for more distal obstruction.41
The problem with stent occlusion has been studied intensively
but attempts at altering bile composition using choleretic agents,
reducing bacterial load with antibiotics, or inuencing mucin pro-
duction with aspirin have failed to prolong stent patency.
Prophylactic stent changes have been advocated by many authori-
ties: However, nearly 50% of patients undergoing stenting with 10
or 11.5 French plastic stents die prior to stent occlusion. Thus, not
all patients will require stent changes and a watchful waiting remains
a reasonable option if a good follow-up system is in place.
Other late complications are unusual and include migration into
the more proximal bile duct or bowel, duodenal or bile duct perfora-
tion and acute cholecystitis.
Complications specic to metal stents
Incomplete removal of the covering membrane, failure of stent
expansion and inability to remove the inner catheter after stent
release are rare technical problems.
About 1015% of metal stents eventually get occluded because
of tumor ingrowth through the wire mesh or tumor overgrowth
above or below the stent. The problem can be solved by inserting
another stent, either plastic or metal, through the occluded stent
(Figs 28.1228.13).
B
A The cost-effectiveness of palliation with endoscopic stent is
Fig. 28.12 Tumor ingrowth: endoscopic treatment. A ERCP
showing a case of tumor ingrowth in a SEMS. B A plastic stent
inserted through the SEMS.
A B
Fig. 28.13 Stent clogging: endoscopic treatment. A Endoscopic
view showing the distal end of stent completely occluded. B In-
sertion of new SEMS through the occluded stent.
Chapter 28 Malignant Biliary Obstruction: Hilar
How to reduce the risk of acute cholangitis
All patients undergoing evaluation and therapy should receive anti-
biotics before and after ERCP. The antibiotic chosen should pene-
trate an obstructed biliary tree (see patient preparation).
For Bismuth type II and III strictures, good endoscopic
techniques are paramount. Minimal contrast medium should
be injected only into the duct to be drained. Once access is
obtained to the obstructed segment, the pressure in the system
should be reduced before more complete lling, by aspirating
bile. If possible, manipulation of ducts that will not be drained
should be avoided. A single stent into an obstructed segment that
drains at least 25% of the liver should be placed. There does not
seem to be any advantage to choosing one lobe of the liver over
the other. If cholangitis occurs, ERCP or a percutaneous approach
to drain the obstructed lobe of the liver should be performed
promptly.36,38,42
COSTS
There are no economic analyses which compare various palliative
options in patients with malignant biliary obstruction of the hepatic
hilum. A cost-effectiveness study previous reported a tabulation of
hospital charges comparing surgical bypass with endoscopic stenting
in patients with distal malignant biliary obstruction. The study
reported total costs per patients of 25 000 US$ for patients treated
with surgical bypass, and 5000 US$ per patient treated with endo-
scopic stent.43
enhanced when metal stents are used. Studies have demonstrated
that the cost for both treatment strategies are different only because
of stent price.
Cost analysis has concentrated on patency rates, repeat ERCPs,
and total length of hospital stays. Of critical importance in determin-
ing cost-efciency is the length of survival of each patient. Because
occlusion rates are dependent on survival, the longer a patient sur-
vives, the more likely a metal stent will be benecial (as plastic stents
occlude earlier than metal). Despite the initial heavy cost of expand-
able metal stents (3040 times that of the plastic stents), in patients
surviving over than 46 months, metal stents ultimately (because of
longer patency periods) prove to be the least expensive method of
relieving biliary obstruction.44,45
SUMMARY
1. The evaluation of patients with suspected malignancy of the
hepatic hilum should include helical or multislice CT of the
abdomen. An MRCP should be obtained to assess for
resectability.
2. If the disease is resectable and the patient is t, surgical resection
of the lesion should be performed.
3. Preoperative ERCP should be avoided unless there is
cholangitis or signicant delay in surgery and the patient is
symptomatic.
4. If the lesion is unresectable or the patient is unt for surgery,
then endoscopic palliation of jaundice should be performed by
using MRCP as a guide for unilateral drainage to minimize
cholangitis.
309
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
5. If cholangitis occurs, ERCP or a percutaneous approach to drain
the obstructed lobe of the liver should be performed promptly.
6. The use of metal stents should be restricted to those patients who
will, in all probability, live longer than 3 months.
REFERENCES
1. Klatskin G. Adenocarcinoma of the hepatic duct at its bifurcation
within the porta hepatis: an unusual tumor with distinctive clinical
and pathological features. Am J Med 1965; 38:241256.
2. De Groen PC, Gores GJ, LaRusso NF, et al. Biliary tract cancers.
N Engl J Med 1998, 341; 18:13681378.
3. Patel T. Worldwide trends in mortality from biliary tract
malignancies. BMC Cancer 2002; 2(10):15.
4. Chamberlain SS, Blumgart RH. Hilar cholangiocarcinoma: a
review and commentary. Annals of Surgical Oncology, 2000;
7:5566.
5. Michaud DS. The epidemiology of pancreatic, gallbladder, and
other biliary tract cancers. Gastrointest Endosc 2002; 56(6):
S195S200.
6. Bismuth H, Corlette MB. Intrahepatic cholangioenteric
anastomosis in carcinoma of the hilus of the liver. Surg. Gynecol.
Obstet. 1975; 140:170176.
7. Gerhards MF, van Gulik TM, de Wit LT, et al. Evaluation of
morbidity and mortality after resection for hilar
cholangiocarcinoma-a single center experience. Surgery 2000;
127:395404.
8. Han JK, Choi BJ, Kim AY, et al. Cholangiocarcinoma: pictorial essay
of ct and cholangiographic ndings. RadioGraphics 2002;
22:173187.
9. Yeh T, Jan YY, Tseng JH, et al. Malignant perihilar biliary
obstruction: magnetic resonance cholangiopancreatographic
ndings. Am J Gastroent 2000; 95:432440.
10. Mortele KJ, Ji H, Ros PR. CT and magnetic resonance imaging in
pancreatic and biliary tract malignancies. Gastrointest Endosc
2002; 56:S206S212.
11. Lee WJ, Lim HK, Jang KM, et al. Radiologic spectrum of
cholangiocarcinoma: emphasis on unusual manifestations and
differential diagnoses RadioGraphics 2001; 21:S97S116.
12. Fritscher-Ravens A, Broering DC, Sriram PVJ, et al.
EUS-guided ne-needle aspiration cytodiagnosis of hilar
cholangiocarcinoma: a case series. Gastrointest Endosc 2000;
52:534540.
13. Anderson CS, Pinson CV, Berlin J, et al. Diagnosis and treatment of
cholangiocarcinoma The Oncologist 2004; 9:4357.
14. Gerhards MF, van Gulik TM, de Wit LT, et al. Evaluation of
morbidity and mortality after resection for hilar
cholangiocarcinomaa single center experience. Surgery 2000;
127:395404.
15. De Bellis M, Sherman S, Fogel EL, et al. Tissue sampling at ERCP in
suspected malignant biliary strictures (Part 1). Gastrointestinal
Endosc 2002; 56:552561.
16. De Bellis M, Sherman S, Fogel EL, et al. Tissue sampling at ERCP in
suspected malignant biliary strictures (Part 2). Gastrointestinal
Endosc 2002; 56:720730.
17. Sugiura Y, Nakamura S, Iida S, et al. Extensive resection of the bile
ducts combined with liver resection for cancer of the main
hepatic duct junction: a cooperative study of the Keio Bile Duct
Cancer Study Group. Surgery 1994; 115:445451.
310
Acknowledgments
I thank Francesca Salvatori, MD, for the English text revision and
Pietro Addeo, MD, for the bibliographic research.
18. Smimada K, Sano T, Sakamoto Y, et al. Safety and effectiveness of
left hepatic trisegmentectomy for hilar cholangiocarcinoma.
World J. Surg. 2005; 29:723727.
19. Capussotti L, Muratore A, Polastri R, et al. Liver resection for hilar
cholangiocarcinoma: in-hospital mortality and longterm survival.
J Am Coll Surg 2002; 195:641647.
20. Meyer CG, Penn I, James L. Liver transplantation for
cholangiocarcinoma: results in 207 patients. Transplantation 2000;
69:16331637.
21. De Vreede I, Steers JL, Burch PA, et al. Prolonged disease-free
survival after orthotopic liver transplantation plus adjuvant
chemoirradiation for cholangiocarcinoma. Liver Transpl 2000;
6:309316.
22. Zoepf T, Jakobs R, Rosenbaum A, et al. Photodynamic therapy
with 5-aminolevulinic acid is not effective in bile duct cancer
Gatrointest Endosc 2001; 54:763766.
23. Ortner MEJ, Caca K, Berr F, et al. Successful photodynamic therapy
for nonresectable cholangiocarcinoma: a randomized prospective
study. Gastroenterology 2003; 125:13551363.
24. Wiedmann M, Berr F, Schiefke I, et al. Photodynamic therapy in
patients with non-resectable hilar cholangiocarcinoma: 5-year
follow-up of a prospective phase II study Gastrointest Endosc
2004; 60:6875.
25. Hawes RH. Diagnostic and therapeutic uses of ERCP in pancreatic
and biliary tract malignancies. Gastrointest Endosc 2002; 56:
S201S205.
26. Strasberg SM. ERCP and surgical intervention in pancreatic and
biliary malignancies Gastrointest Endosc 2002; 56:S213S217.
27. Flamm CR, Mark DH, Aronson N. Evidence-based assessment of
ERCP approaches to managing pancreaticobiliary malignancies
Gastrointest Endosc 2002; 56:S218S225.
28. Rey JF, Dumas R, Canard JM, et al. Guidelines of the French
Society of Digestive Endoscopy: Biliary Stenting. Endoscopy 2002;
34:169173.
29. ASGE Guidelines. The role of endoscopy in the evaluation and
treatment of patients with pancreaticobiliary malignancy.
Gastrointest Endosc 2003; 58:643649.
30. ASGE guideline: the role of ERCP in diseases of the biliary tract
and the pancreas. Gastrointest Endosc 2005; 62:18.
31. Chang W, Kortan P, Haber G. Outcome in patients with
bifurcation tumors who undergo unilateral versus bilateral hepatic
duct drainage. Gastrointest Endosc 1998; 47:354362.
32. Abraham NS, Barkun JS, Barkun A. Palliation of malignant biliary
obstruction: a prospective trial examining impact on quality of life
Gastrointest Endosc 2002; 56:835841.
33. Raiker GV, Melin MM, Ress A, et al. Cost-effective analysis of
surgical palliation versus endoscopic stenting in the management
of unresectable pancreatic cancer. Ann Surg Oncol 1996;
3:470475.
34. Brandabur JJ, Kozarek RA, Ball TJ, et al. Non-operative versus
operative treatment of obstructive jaundice in pancreatic cancer:
cost and survival analysis. Am J Gastroenterol 1988; 83:11321139.

35. Taylor MC, McLeod RS, Langer B. Biliary stenting versus bypass
surgery for the palliation of malignant distal bile duct obstruction;
a meta-analysis. Liver Transpl 2000; 6:302308.
36. Speer AG, Cotton PB, Russell RCG. Randomized trial of
endoscopic versus percutaneous stent insertion in malignant
obstructive jaundice. Lancet 1987; 2:5762.
37. De Palma GD, Galloro G, Iovino P, et al. Unilateral versus bilateral
endoscopic hepatic duct drainage in patients with malignant hilar
biliary obstruction. Results of a prospective, randomized, and
controlled study. Gastrointest Endosc 2001; 53:547553.
38. Hintze RE, Abou-Rebyeh H, Adler A, et al. Magnetic resonance
cholangiopancreatographyguided unilateral endoscopic stent
placement for Klatskin tumors. Gastrointest Endosc 2001;
53:4046.
39. De Palma GD, Pezzullo A, Rega M, et al. Unilateral placement of
metallic stents for malignant hilar obstruction: a prospective study.
Gastrointest Endosc 2003; 58:5053.
40. Freeman ML, Overby C. Selective MRCP and CT-targeted drainage
of malignant hilar biliary obstruction with self-expanding metallic
stents Gastrointest Endosc 2003; 58:4149.
Chapter 28 Malignant Biliary Obstruction: Hilar
41. Singh V, Sigh G, Verma GR, et al. Contrast-free unilateral
endoscopic palliation in malignant hilar biliary obstruction: New
method Journal of Gastroenterology and Hepatology 2004;
19:589592.
42. Shermann S, Lehman G, Earle D. Are the patency rates for 10-
French, 11.5-French stents different for common duct obstruction
and hilar obstruction? Randomized, prospective study.
Gastrointest Endosc 1996; 43:396.
43. Shermann S. Endoscopic drainage of malignant hilar obstruction:
Is one biliary stent enough or should we work to place two?
(Editorial) Gastrointest Endosc 2001; 53:681684.
44. Raikar GV, Melin MM, Ress A, et al. cost-effective analysis of
surgical palliation versus endoscopic stenting in the management
of unresectable pancreatic cancer. Ann Surg Onc 1996;
3:470475.
45. Schmassmann A, Von Gunten E, Knuchel J, et al. Wallstents
versus plastic stents in malignant biliary obstruction: effects
of stent patency of the rst and second stent on patient
compliance and survival. Am J Gastroenterol 1996;
91:654659.
311
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Indeterminate Biliary Stricture
29 Bret T. Petersen
Biliary obstruction results from diverse benign and malignant
processes and patients can present acutely or chronically with signs
and symptoms ranging in severity. The nature of an obstruction is
often immediately clear at the time of initial investigation, while at
other times obstruction is readily apparent but the nature of the
pathologic process remains uncertain. No single denition exists
for the term indeterminate stricture, but it commonly refers to
biliary strictures in patients in whom cross-sectional imaging is
unrevealing, i.e. without an associated mass lesion, and without
pathologic conrmation. Recent experience with inammatory pan-
creatic masses might prompt expansion to include all strictures,
including those with associated mass lesions, prior to histological
characterization.
When biliary obstruction is identied, an efcient approach to
early diagnostic testing and management is important for reduction
of morbidity and guidance of denitive therapy. Untreated obstruc-
tive cholestasis of even moderate degree can culminate in secondary
biliary cirrhosis within several months.1,2 Patients with inadequately
treated strictures also risk development of acute or chronic cholan-
gitis, particularly following invasive testing.
Key steps in the assessment and management of patients
with indeterminate biliary strictures include characterization of the
pathogenesis of the stricture, relief of biliary obstruction and/or
denitive treatment of the pathologic processemploying medical,
endoscopic, percutaneous, or surgical means. Stricture characteriza-
tion and relief of obstruction are not independent pursuits but are
typically accomplished in unison. Stricture characterization is based
upon historical features, laboratory testing, non-invasive and inva-
sive imaging, and by the use of various tissue sampling methods
(Fig. 29.1).3
HISTORICAL FEATURES
Historical features may contribute to both the correct diagnosis and
the management strategy for newly identied biliary strictures
(Table 29.1). Prior history of ulcerative colitis, complicated biliary
surgery, or chronic pancreatitis suggest PSC, postoperative stric-
tures, and pancreatic compression of the CBD, respectively. An
acute presentation in the early postoperative period or during an
episode of pancreatitis suggests signicant operative injury or stone-
related obstruction whereas sub-acute but early (<3 months) presen-
tations suggest inammatory processes that may resolve with
timehence minimally invasive and temporizing approaches may
sufce. Presentation more than three months after a prior insult
suggests a more brotic and rigid stricture which may require more
aggressive or prolonged therapy. Strictures that present in an occult
or delayed fashion, and those presenting without known predispos-
ing factors all raise the specter of a malignant etiology. A waxing and
waning presentation is suggestive of benignity whereas inexorable
progression of symptoms associated with weight loss suggests
malignant etiologies.
LABORATORY FEATURES
Laboratory features obtained at the time of presentation with a stricture
may provide an assessment of the strictures severity and chronicity as
well as the etiology. Isolated mild to moderate elevations of alkaline
phosphatase, without transaminase or bilirubin elevations, imply
modest impairment to bile ow due to intra or extrahepatic etiologies.
Enzyme fractionation should conrm the hepato-biliary source of the
elevation and cross-sectional imaging should identify when obstruc-
tion involves larger central or extrahepatic ducts. Concurrent eleva-
tions in the transaminases imply either a hepatitic process or relative
acuity of onset to the obstruction. Total bilirubin values are not highly
indicative of obstruction, but in the setting of complete obstruction
with an otherwise healthy liver, bilirubin is said to generally peak
under 20 mg/dl, whereas values beyond this imply hepatocellular
injury, with or without obstruction. Chronic obstruction with deep
jaundice can induce malabsorption of the fat soluble vitamins, includ-
ing vitamin K, thus leading to elevated prothrombin times. Hence an
INR should be checked prior to interventional techniques in these
patients. Elevated pancreatic enzymes imply concurrent pancreatitis
or pancreatic duct obstruction, commonly due to biliary stone disease,
pancreatic carcinoma, or advanced chronic pancreatitis.
Very few serologic markers contribute to characterization of the
benign or malignant nature of indeterminate biliary strictures. Car-
bohydrate antigen 19-9 (CA19-9) is a serum moiety that is elevated
in the settings of pancreatic and biliary carcinoma, cholangitis, and
to a lesser degree, pancreatitis.4 Marked CA19-9 elevations above
1000 IU are seen only with cancer or orid cholangitis. Elevations
above 100 IU are strongly suggestive of cancer in the absence of
known pancreatitis or cholangitis. When an elevated CA19-9 is
detected in the setting of cholangitis it should be reassessed follow-
ing appropriate therapy of the infectious process.
Immunoglobulin G, sub-fraction 4 levels are often, but not invari-
ably, elevated in autoimmune pancreatitis, which can cause biliary
strictures that mimic those that occur with chronic pancreatitis of
other etiologies, pancreatic cancer, or even PSC.5 Such strictures are
often amenable to therapy using corticosteroids.6
NON-INVASIVE CROSS-SECTIONAL IMAGING
Ultrasonography (US), computerized tomography (CT), and mag-
netic resonance imaging (MR) play a primary role in conrmation
of obstruction (based upon ndings of duct dilation or mass lesions),
identication of associated complications such as abscess or bowel
obstruction, and initial characterization of the pathologic process.
Settings in which ductal dilation proximal to a stricture may not be
313
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Fig. 29.1 Algorithm for Evaluation of Jaun-

Obstructive jaundice dice and Suspected Biliary Obstruction.
See text for discussion. (Redrawn with per-
mission from Gastrointestinal Endoscopy
2002.)
Ultrasound or CT scan

Mass? Yes Biliary dilation? No Mass? No

ERCP with tissue No Yes Surgical candidate and Yes

potentially resectable Hepatitis screen
sampling and stenting
Drugs screen
Liver biopsy
MRCP

EUS Yes No ERCP with tissue

with FNA sampling and stenting

Resectable Not resectable

Tissue diagnosis
Surgery not established

EUS, FNA
or CT FNA

ent resectability of the lesion based on initial studies. Ultrasono-

A. Historical features suggestive of benign etiologies
History of right upper quadrant surgery
Trauma
Ulcerative colitis or Crohns disease
Chronic pancreatitis
Difcult biliary stone disease
Stable weight
Fluctuating labs
B. Historical features suggestive of malignant etiologies
Never operated abdomen
Absent history of abdominal illness
Weight loss
Short course without antecedent illness
Decompensation of known PSC
Table 29.1 Historical features and character of biliary strictures
present include: early or uctuating processes that are inadequately
advanced to cause obstruction, and diseases in which the ducts
and/or the liver are brosed and cannot dilate easily, as in sclerosing
cholangitis.
Transabdominal ultrasonography (TUS) is usually the rst study
employed in jaundiced patients to identify the presence and level of
duct dilation and to look for bile duct or gallbladder stones or masses.
Ultrasonography is the most sensitive of the non-invasive tests for
biliary stones. US is extremely sensitive for duct dilation but less so
for identifying the specic etiology of a stricture.
Once a stricture is localized by cross-sectional imaging with US
(or CT scanning) the next step in evaluation is highly dependent
upon the clinical judgment as to whether the setting favors a benign
or malignant process, the patients tness for surgery, and the appar-
graphic evidence of a stricture, without evidence of advanced cancer,
is usually followed by abdominal CT scanning to dene whether a
mass exists and to provide initial staging information. If ultrasonog-
raphy demonstrates a distal unresectable mass, based on local-regional
extension, hepatic metastasis, or associated ascites, then ERCP is
usually performed for both tissue acquisition and palliation of
obstructive jaundice. If US demonstrates a hilar mass, with or without
evidence of unresectability, then magnetic resonance cholangio-
pancreatography (MRCP) is helpful to better dene the level of the
obstruction, assist with assessment of resectability, and guide the
subsequent approach to invasive cholangiography, tissue sampling,
and palliative stenting.7 An extrahepatic stricture without a mass in
the setting of fever, apparent biliary pancreatitis, or gallbladder
stones can often be evaluated directly with ERCP in anticipation of
identifying an obstructing duct stone.
Abdominal CT scanning is commonly employed in patients with
associated weight loss, fever, or signicant pain, as it is particularly
useful for identication and staging of extra-ductal mass lesions,
inammatory processes, and bile collections or leaks (Fig. 29.2). CT
is also preferred over ultrasonography in obese subjects. CT images
are very familiar to most clinicians. Recently the inclusion of coronal
renderings has provided clinicians with even greater understanding
of the anatomic relationships between neighboring structures in the
right upper quadrant and their involvement by pathology. CT chol-
angiography is also available but seems to be less utilized currently,
in the era of MRCP.
Abdominal magnetic resonance imaging yields cross-sectional
information analogous to CT scanning and provides relatively
sensitive cholangiographic images that usually allow determination
of stricture location. MRCP is the most sensitive non-invasive

314
 Chapter 29 Indeterminate Biliary Stricture

A B

Fig. 29.2 Abdominal CT scan demonstrates distal extrahepatic obstruction, based upon traditional cross-sectional views showing dilation
of the intrahepatic ducts A and proximal extrahepatic ducts B, and similar dilation plus pancreatic duct dilation and a distal mass seen on
the coronal view C.

imaging test for biliary obstruction and it approaches the sensitivity
of ERCP for identication of biliary strictures.8 Studies differ,
however, as to whether it is inferior or equivalent to ERCP for the
differentiation of benign from malignant lesions.8,9 When acquired
and displayed by standard cross-sectional MR scanning information
regarding extra-ductal pathology or extent of disease tends to be less
readily interpreted by the non-radiologist than by CT scanning.
MRCP has largely replaced diagnostic endoscopic cholangiography
when there is not a need for tissue acquisition, therapy, or dynamic
measurements of motility. The primary benet of MRCP is the
avoidance of intubation, sedation, and the risk of pancreatitis. Other
advantages of MRI over ERCP include the ability to display the
anatomy of the ducts and the liver above a stricture even when com-
plete obstruction is present, and the ability to generate multiple
perspectives or angles of view for the same lesion. A disadvantage
of MRI is that the cholangiographic display includes all ducts,

315
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

without the ability to localize images to the region of interest around
a stricture, as is done with early lms acquired during initial con-
trast instillation during ERCP. This sometimes makes interpreta-
tion of a central or a complex stricture difcult due to overlapping
peripheral ducts that are of little consequence (Fig. 29.3). Several
studies have demonstrated the utility of MRCP as a guide to
Fig. 29.3 Abdominal MRCP demonstrating a distal extrahepatic
stricture, analogous to that seen on CT in Figure 29.2.
subsequent ERCP and palliative stent placement for hilar lesions7
(Fig. 29.4) as discussed in Chapter 28.
INVASIVE IMAGING TECHNIQUES
Evaluation of the biliary tree can be achieved using endoscopic ultra-
sonography, as well as by traditional contrast-based cholangiography
via percutaneous transhepatic (PTC) or endoscopic retrograde
(ERCP) routes. Intraductal ultrasound (IDUS) and cholangioscopy
are specialized techniques employed during performance of ERCP
with a duodenoscope which will be discussed in subsequent sec-
tions. Cholangioscopy is also discussed in Chapter 21.
Endoscopic ultrasonography (EUS) is useful for both diagnosis and
staging of malignant biliary strictures. EUS is accomplished from
the duodenal bulb, and/or the antrum, depending upon the patients
anatomy. Radial or linear technology can be employed, but the
frequent use of ne needle aspiration is driving an evolution
toward predominantly linear imaging. Malignancies are identied
as hypoechoic masses or thickening of the bile duct wall. In one
study of 40 biliary strictures (24 malignant, 16 benign) EUS ndings
of a pancreatic head mass and/or an irregular bile duct were more
sensitive than concurrent FNA sampling. EUS imaging alone was
88% sensitive and 100% specic for malignancy. Wall thickness
>3 mm was 79% sensitive and 79% specic for malignancy. The
sensitivity of FNA was 47%, with 100% specicity and positive pre-
dictive value but only 50% negative predictive value.10 In a compara-
tive study of several modalities, EUS sensitivity and specicity (79%
and 62%) were less than ERCP or MRCP but complementary to
them.8 In contrast, a study evaluating EUS with FNA in 28 patients
with non-diagnostic sampling of biliary strictures obtained during
ERCP, PTC, or CT demonstrated 86% sensitivity, 100% specicity,
100% positive predictive value, 57% negative predictive value, and
88% accuracy for malignant lesions.11 Importantly, management

A B

Fig. 29.4 Cross-sectional imaging with MRCP or CT provides guidance to the preferred lobe for palliative biliary drainage during ERCP
in patients with proximal biliary obstruction. A MRCP suggests access should be pursued toward the dominant right lobe. B CT in the
same patient demonstrates left lobe atrophy, also suggesting access should be to the right lobe.

316

was inuenced in 84% of patients. Some but not all studies note a
greater sensitivity of EUS/FNA for pancreatic lesions than for extra-
hepatic cholangiocarcinoma.12 EUS and CT are complementary
studies for staging and determination of resectability for distal biliary
strictures due to pancreatic mass lesions.13 Hence, while not a
primary imaging modality for biliary strictures, EUS with FNA is an
important ancillary technique when diagnosis remains elusive and
when staging for determination of resectability is sought.
Cholangiography is the mainstay for diagnosis and characteriza-
tion of extrahepatic biliary lesions of all types. Endoscopic and
percutaneous approaches to cholangiography are complementary
studies and, on occasion, both will be necessary to characterize and
treat difcult biliary lesions (Table 29.2). In general, proximal lesions
that appear to involve the hilar region are best investigated initially
with non-invasive MRCP, as this study provides directional guidance
for subsequent invasive imaging and palliation7 and avoids the risk
of cholangitis that occurs with ERCP when contrast is injected into
areas which may not be drainable. However, preoperative planning
for hilar lesions may still require the clarity of contrast-based chol-
angiography (ERCP or PTC).
The cholangiographic appearance of the stricture is generally
inadequate for interpretation of malignancy and many strictures that
are interpreted as benign prove to be malignant.14 Features sugges-
tive of malignancy include progressive focal stricturing over time,
abrupt shelf-like borders, length greater than 14 mm, intrahepatic
duct dilation, and presence of intraductal polypoid or nodular
areas.14,15 In the setting of background sclerosing cholangitis with
dominant strictures, malignant lesions are more likely to exceed
1 cm in length, be located at the bifurcation as opposed to the
common bile duct, and have irregular margins.16 Despite these
criteria, cholangiography alone correctly identied only 8 of 12
(66% sensitivity) malignant lesions and 21 of 41 (51% specicity)
benign lesions.16
Endoscopic retrograde cholangio-pancreatography (ERCP) has
become the primary non-operative modality for both investigation
and palliation of biliary strictures because it provides high quality
contrast-based images of the ductal systems, access for tissue sam-
pling, and means of therapy via internal drainage. ERCP is preferred
if there is a likely need for stone extraction or stent placement in the
extrahepatic ducts, when coagulopathy or ascites is present, when
Settings favoring endoscopic approach (ERCP)
Preferred in most settings
Intact upper gut anatomy
Anticipated need for therapeutic stenting or stone removal
Need for luminal exam
Ascites
Coagulopathy
Small caliber ducts
Failed percutaneous approaches
Settings favoring percutaneous approach (PTC)
Altered upper gut anatomy, especially Roux-en-Y gastric bypass,
+/ Whipple anatomy
Complete biliary obstruction
Failed endoscopic access for cholangiography or stent placement
Need to better image and stage proximal end of stricture for
surgical planning
Table 29.2 Indications and favored settings for endoscopic vs
percutaneous cholangiographic approach to biliary strictures
Chapter 29 Indeterminate Biliary Stricture
the bile ducts are not dilated, and when percutaneous approaches
fail. It should be undertaken only by endoscopists with the experi-
ence and ability to proceed with appropriate imaging, tissue sam-
pling, and therapies. In inexperienced hands initial studies often
yield poor stricture denition, inadequate drainage, or procedural
complications.
Endoscopic cholangiography in the setting of obstructive jaun-
dice should be performed with peri-procedural antibiotic coverage
and anticipation of continuing antibiotics for a brief interval, par-
ticularly if complete drainage cannot be achieved. Stent placement
should be performed whenever signicant contrast is instilled above
a lesion that prevents spontaneous drainage. In the setting of hilar
strictures intrahepatic lling of contrast should be avoided until wire
access is accomplished, to ensure the ability to provide subsequent
palliative drainage of the imaged segments. As noted, prior high-
quality CT or MRCP imaging can guide selection of optimal intra-
hepatic systems for wire access and stent placement.
Optimal characterization and successful access for sampling and
treatment require attention to imaging principles that are often not
appreciated by non-radiologists. The following points apply equally
to imaging of benign or malignant strictures (Table 29.3): (1) Stric-
tures, unlike large dilated systems, are best imaged with full-strength
contrast. (2) Multiple early lms should be taken as contrast is rst
crossing the lesion, continuing the injection until the image has
been obtained (Fig. 29.5). This allows later reference back to the
minute details of angles or bifurcations, which may not be evident
when the ducts are completely lled. (3) Coning down on an area of
interest will sharpen the detail seen on hard copy. This requires at
least some larger views to maintain anatomic reference. (4) Tissue
sampling and therapy should be performed with the least contrast
lling required to adequately demonstrate the anatomy. Excessive
lling of intrahepatic ducts often obscures the bifurcation. (5) The
proximal extent of duct involvement must be well demonstrated for
surgical or endoscopic decision making. Following wire access,
1. Use pre-procedure antibiotics and full strength contrast.
2. Obtain multiple early images during contrast injection.
3. Obtain selected broad views to maintain anatomic reference.
4. Cone down on the area of interest to sharpen radiographic
detail.
5. Use the least contrast lling required to adequately
demonstrate anatomy.
6. Employ prior CT or MRCP for hilar lesions to guide intrahepatic
duct selection for contrast lling and decompression. When
wire access is conrmed the proximal extent of duct
involvement must be well demonstrated for surgical or
endoscopic decision making.
7. Head-up and head-down positions on a tilt table can facilitate
imaging of stricture extent by inducing contrast ow to the area
of interest.
8. An open hilar view is best obtained with a 20 degree oblique
position using a C-arm or by rolling the prone patient
rightward toward the endoscopist.
8. After demonstrating a distal stricture, the bifurcation and central
intrahepatic ducts should be lled to exclude secondary
proximal obstruction (e.g. adenopathy).
9. Some lesions can only be well characterized by percutaneous
cholangiography.
Table 29.3 Pointers for stricture characterization during
cholangiography
317
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
Fig. 29.5 Early A vs later B images acquired during ERCP. Note detail evident on early image is obscured by intrahepatic lling overlap-
ping the area of interest in the subsequent view. Note improved view C with oblique imaging.
A B
Fig. 29.6 A and B Overlapping and open views of the ducts at the
hilum. B is obtained by rolling the patient rightward 1520
degrees or by leftward rotation of a C-arm.
marked lling may be required to delineate the proximal end of the
stricture. (6) Head-up and head-down positions on a tilt table can
facilitate imaging of stricture extent by employing gravity to shift
contrast to the area of interest. (7) An open view of the hilum,
without overlapping ducts, is best seen in an oblique position, which
is achieved with use of a C-arm or by rolling the prone patient right-
ward toward the endoscopist (Fig. 29.6). (8) After demonstrating a
distal extrahepatic stricture, the central intrahepatic ducts and
hepatic conuence should be lled to exclude additional proximal
strictures (e.g. adenopathy) within reach of endoscopic manage-
ment. (9) Some lesions can only be well characterized by percutane-
ous cholangiography (Fig. 29.7). (10) Limited pancreatography may
aid in demonstrating or excluding a pancreatic primary lesion when
biliary strictures involve the distal third of the duct.
Stricture access with guidewires and subsequently other over-the-
wire accessories is required to accomplish cytology brushing, dila-
tion, and palliative stenting or denitive endoscopic management.
Access is usually gained with multipurpose plastic-coated guide-
wires or specialty hydrophilic wires that are extremely slippery,
exible, and torqueable.17 Manipulation of angled wires using simul-
taneous torque and advancement can be performed by the assistant,
or by the endoscopist using any of several recently designed short-
318
C
A
Fig. 29.7 Benet of percutaneous trans-hepatic cholangiography
(PTC) for proximal evaluation of selected duct lesions. A Partial
endoscopic cholangiogram demonstrating complete duct obstruc-
tion following laparoscopic cholecystectomy. B PTC-placed
hepatic drain demonstrating intrahepatic ducts, conrming com-
plete duct disruption.
wire systems. This is best done with two hands to facilitate ne wire
control and maintenance of position (Fig. 29.8).
Stricture dilation should be performed prior to passage of larger
caliber devices. In the case of benign lesions this constitutes the
rst step in their therapy. For the tightest strictures that will not
accept anything beyond 0.035 guidewires, initial dilation can be
accomplished with angioplasty balloons that traverse 0.018 wires
and expand up to 4 mm from their deated 0.035 caliber (Fig.
29.9). Rigid 457 F dilators can be passed over a 0.025 guidewire.
Standard balloon dilators can then be used to expand to larger cali-
bers. Balloon selection is based upon the size of the non-obstructed
duct just distal to the stricture. Most often this calls for 4, 6, or
8 mm diameter balloons. Tight chronic strictures carry some risk
for rupture or tear during dilation. Should this occur, adequate

stenting for drainage is mandatory and addition of a nasobiliary
drain may be useful during a several-day hospital stay for parenteral
antibiotics.
Percutaneous transhepatic cholangiography (PTC) has similar
capabilities of duct imaging, access, and palliative drainage as does
ERCP; however, it is performed through a sterile prepped cutaneous
eld and hence cholangitis risks are lessened and drainage of lled
Fig. 29.8 A view of the endoscopists hands during manipulation
of a slippery guidewire through difcult strictures. Note that two
hands are used to hold and move the wire, while the base of one
hand holds the control section of the endoscope.
A B
Fig. 29.9 Dilation of a web-like anastomotic stricture with an angioplasty balloon passed
over a 0.018 guidewire. This balloon dilates from an outer diameter of 0.035 to 4 mm,
allowing subsequent passage of standard 5 Fr balloon catheters for dilation to 6 or
8 mm.
Chapter 29 Indeterminate Biliary Stricture
segments is less critical. PTC is only indicated when the proximal
end of a stricture has not been adequately characterized by MRCP
or retrograde methods (if this information will change manage-
ment), when endoscopy fails to access and decompress an obstructed
system, or altered anatomy dictates that percutaneous routes be
used. When an extrahepatic stricture cannot be accessed from
below, a guidewire can be advanced via PTC for subsequent retro-
grade access. Use of this so-called combined procedure to enable
stent placement and decompression should have less morbidity
than conversion of the entire management plan to a percutaneous
approach.
TISSUE ACQUISITION AND PATHOLOGIC
INVESTIGATIONS
Methods of tissue acquisition and analysis for cholangiography
include performance of ne needle aspiration and mucosal brushing
for thin preparation cytologic exam, mucosal biopsy for standard his-
tological analysis, and evaluation of both cytology and biopsy speci-
mens using a variety of specialized tests for nucleic abnormalities or
by-products of neoplasia. Tissue acquisition is a key element of each
of these techniques. In a recent review, all available sampling tech-
niques during ERCP were evaluated.18 Overall, the results for patho-
logic examination of tissues acquired at ERCP remain frustratingly
low. This is due to several factors, including the scirrhous nature of
many tumors, the small tissue samples acquired, difculty in target-
ing the abnormality in question, and time constraints in procedures
geared primarily toward palliation and less so to tissue acquisition.
During cholangiography, spot lm radiographic documentation
should be obtained of all sampling techniques and locations.
C
319
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 29.10 Various cytology brush designs include: A metal tip
brush; B brush with exible wire leader in single lumen catheter;
C brush with leader and guidewire in dual lumen catheter; and (d)
large caliber brush. (With permission, from Clinical Gastrointestinal
Endoscopy, Elsevier Saunders.)
Brush cytology
The yield of brush cytology for the diagnosis of strictures varies
widely, with conrmation of malignancy in 1565% of biliary stric-
tures secondary to pancreatic cancer and in 4480% of strictures due
to cholangiocarcinoma.18,19 Combined results in over 800 patients
reported sensitivity of 42%, specicity of 98% and positive predictive
value (PPV) of 98% among patients with conrmed cancer.18 Studies
pertaining to sampling technique note that cellular yield is improved
by using a minimum of ve brush passes through the stricture,
removal of the catheter and brushing together while avoiding brush
withdrawal through the length of the catheter, and ushing residual
cells from within the catheter into the sample vial after removal of
the brush.20 It is unclear whether stricture dilation improves sample
cellularity. Inclusion of washings from the barb or lumen of removed
plastic stents may also enhance cytology yield. A variety of brushes
are available but few comparative data exist among them. One major
design variant is incorporation of a leading exible wire guide for
maintenance of position both within the stricture and within the
duct (Fig. 29.10). One study showed no difference between tradi-
tional biliary brushes and a new design with longer, stiffer, angled
bristles.19 Both brushes yielded adequate specimens in over 80% of
cases, suggesting cytologic technology itself is suboptimal for all
varieties of malignant biliary lesions.
Today biliary cytology brushing is most commonly employed
using wire-guided devices (Fig. 29.11). The technique involves rst
establishing wire-access through the stricture, advancement of the
cytology device over the guidewire until through the stricture,
advancement of the brush beyond the end of the sheath, and with-
drawal of the two together until the brush is within the stricture.
The brush is then passed up and down through the stricture at
least ve times, using either combined movement of the sheath
and the brush by the endoscopist, or movement of the brush itself
by the assistant as the sheath is held in place. Those devices with
a exible wire leader ahead of the brush can be withdrawn through
most of the length of the stricture and safely advanced without
risking loss of access or perforation. Some tight or angled strictures
can only be brushed with a downward movement or brush with-
drawal, requiring repeated access with the entire assembly for each
brush pass.
320
A B
Fig. 29.11 A Cholangiogram demonstrating an indeterminate
biliary stricture. B Wire-guided brush cytology device within the
stricture.
A B
Fig. 29.12 A An indeterminate biliary stricture with a neighboring
pancreatic stricture representing a double duct sign, suspicious for
pancreatic carcinoma. B Forceps biopsy being performed paral-
lel to a guidewire.
Intraductal transmucosal ne needle aspiration
This FNA method was reported to yield positive or suspicious cytol-
ogy in 67% of cancers in the hands of one proponent21 but cumula-
tive data from over 220 patients in ve series yielded a sensitivity of
only 34%, with 100% specicity and 100% PPV.18 The technique has
not gained favor as it is difcult and is optimally performed with a
cytopathologist in the room.
Intraductal forceps biopsies
Intraductal biopsies provide the greatest yield for detection of malig-
nancy among the ERCP-based modalities, with a cumulative sensi-
tivity of 56%, specicity of 97% and positive predictive value of 97%
based upon 500 patients in ve cumulative studies.18,20 A variety of
straight, angled, and malleable forceps are available in adult (7 F)
and pediatric (56 French) calibers for intraductal use. Passage of
these devices may require performance of a biliary sphincterotomy.
However, passing the forceps alongside of a guidewire without
sphincterotomy is possible.22 Trocars or sheaths for transpapillary
passage of biopsy cables are also available. There are limited data
comparing the different biopsy devices.
The technique of passing a biopsy forceps into the bile duct
involves impacting the rigid leading end of the biopsy cable into the
papillary os or the sphincterotomy opening from a short scope posi-
tion, then advancing the endoscope several centimeters while simul-
taneously exing the large ratchet backward to look upward from
below the papilla, followed by upward advancement of the cable
(Fig. 29.12). Alternatively one can occasionally advance the biopsy

Fig. 29.13 Fluorescent in-situ hybridization demonstrates a single microscopic eld with uorescent probes of different colors attached
to specic chromosomal loci. Two copies of each probe should be present. Presence of more than two copies represents aneuploidy
in a cell. A a normal cell. B a cell from a malignant stricture in a patient with PSC.
cable directly into the papilla from a slightly longer exed position,
looking upward from below the papilla.
The highest diagnostic yield for tissue sampling during ERCP is
obtained when two or more of the standard modalities are combined
at the same procedure. Ponchon increased the cumulative yield to
63% by combining brush cytology (43% sensitivity) and intraductal
biopsy (30%).23 Combined biopsy, brushing, FNA, and stent cytology
yielded positive diagnosis in 82% of patients in one study.23a
Given the suboptimal diagnostic yield from standard analyses
of brush cytology and tissue biopsy samples, a variety of advanced
analytic techniques have been investigated. They include ow
cytometry, digital image analysis (DIA), and uorescent in-situ hybrid-
ization (FISH). In a limited number of studies, ow cytometry for
DNA assessment of large cellular populations yielded improved
sensitivity at the expense of signicantly reduced specicity.24
Digital image analysis uses a computerized assessment of
cellular DNA ploidy within a smaller number of individual cells
identied on a cytology slide to estimate the relative proportion
with aneuploidy, which serves as a marker of malignancy. In a
recent prospective study of 100 patients with mixed benign and
malignant strictures the sensitivity, specicity, and accuracy of DIA
of biliary brush cytology samples was 39.3%, 77.3%, and 56%,
compared to 17.9%, 97.7%, and 53% for standard cytology.25 False
positive results from DIA (10 of 44, 22.7%) occurred only in
patients with primary sclerosing cholangitis (PSC). The only false
positive for routine cytology (1/44, 2.3%) was also in a patient with
PSC.
Fluorescent in-situ hybridization employs uorescent probes that
label specic portions of selected chromosomes, allowing for deter-
mination of cellular ploidy via uorescent microscopy of specic
cellular samples (Fig. 29.13). Recent studies have employed chromo-
somal probes that were originally designed for identication of
urothelial cancers (centromeres to chromosomes 3, 7, and 17 plus
chromosomal band 9p21).26 Detection of more than ve cells with
polysomy is considered evidence for malignancy. In preliminary
Chapter 29 Indeterminate Biliary Stricture
studies, the FISH technique increased the sensitivity of brush sam-
pling for detection of malignancy from 15% to 34% (p < 0.01), with
corollary non-signicant reduction in specicity from 98% to 91%
(p = 0.06).26 Concerns about specicity persist and conrmatory
series are needed. Studies designed to identify products of other
genetic mutations (p-53, k-ras) in bile or tissues have not yielded
adequate sensitivity and specicity to be of clinical use for
diagnosis.
ANCILLARY TECHNIQUES
Intraductal ultrasonography (IDUS) employs a 20 mHz radial ultra-
sound probe on the leading end of a 7 French catheter that can be
passed over a guidewire into the biliary and pancreatic ducts during
ERCP (Fig. 29.14). IDUS has been employed for identication of
residual duct stones, characterization of strictures, and staging of
local cancer involvement. Following performance of cholangiogra-
phy, a 0.035 guidewire is left in the duct and the ultrasound probe
is advanced over the wire with the radial crystal stationary. Imaging
is then performed primarily during catheter withdrawal, to limit
mechanical trauma to the mechanical drive of the probe. Acquisition
of IDUS skills is less involved than acquisition of EUS skills and
most experienced endoscopists should be able to adopt IDUS for the
management of stone disease with limited training, and for stricture
assessment with slightly greater experience.
Ultrasonographic features of malignant strictures include
hypoechoic asymmetric wall thickening, poorly demarcated borders
and abrupt shoulders (Fig. 29.15). Benign lesions tend to be hyper-
echoic, have less asymmetry, sharper demarcation with surrounding
tissues, preserved tissue planes, and smooth edges. IDUS interpreta-
tion is more difcult in the setting of primary sclerosing cholangitis,
where widespread background inammation and duct thickening
are present. Similarly, prolonged stenting can induce more wide-
spread duct abnormalities than were originally present at the level
of the index stricture.
321
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
Fig. 29.14 A Leading end of 5 F wire-guided intraductal ultrasound (IDUS) probe. The
mechanical radial ultrasound crystal is noted with an arrow. B IDUS catheter exiting from
the duodenoscope. C IDUS probe within stricture in Figure 29.11.
Fig. 29.15 IDUS images from an extrahepatic cholangiocarcinoma.
Note the asymmetry of wall thickening and its irregular outer
boundary.
Several studies have demonstrated the superior sensitivity and
accuracy of IDUS for characterization of strictures as malignant
(sensitivity >90%; accuracy 8892%) compared to standard cholan-
giography, with or without cytology and biopsy sampling (sensitivity
4857%; accuracy 7378%).2729 In a recent prospective study in 87
patients, elevated serum CA 19-9 (>100 units), routine cytology and
322
intraductal biopsy were compared to advanced cytologic assessment
with DIA, FISH and stricture assessment with IDUS. IDUS showed
the greatest sensitivity (87%) and accuracy (90%), and the combina-
tion of IDUS, DIA, and FISH allowed diagnosis of malignancy in
87% of those with falsely negative routine cytology.30 Studies of
IDUS for tumor staging also report utility for dening longitudinal
extent as well as the extent of invasion to the pancreatic parenchyma,
portal vein, and right hepatic artery.31 Periductal, nodal, and distant
spread is not adequately assessed by IDUS.
Cholangioscopy, or direct visual evaluation of the biliary tree, is
increasingly used for visually directed sampling of indeterminate
strictures and for electrohydraulic lithotripsy therapy of intractable
stones. Cholangioscopy is also discussed in Chapter 21 (Fig. 29.16).
It usually employs 810 French mini-endoscopes that are passed
through the working channel of a therapeutic duodenoscope, with
or without guidewire assistance. Fiberoptic cholangioscopes are
commercially available; digital video-chip versions are in use but not
widely available. Cholangioscopy is usually performed with the assis-
tance of a second endoscopist to manipulate the cholangioscope
controls and biopsy cables while the primary endoscopist controls
the duodenoscope and the insertion of the cholangioscope. Some
centers employ custom endoscope holders for the cholangioscope to
enable studies by a single operating physician.32 During cholangios-
copy attention must be paid to frequent or continuous ushing to
clear the eld of bile or debris. Fluid run-off to the stomach requires
frequent aspiration, use of a nasogastric tube, or endotracheal intu-
bation. Cholangioscopes are relatively fragile and care must be taken
to avoid excessive angulation or force, particularly at the level of the
elevator on the duodenoscope.
Recent series demonstrate improved sensitivity and accuracy
for diagnosis of malignant obstruction when cholangioscopy is
employed33,34 In one study, ERCP with uoroscopically guided tissue
sampling had a sensitivity of 58% and accuracy of 78% for malig-
nancy while addition of cholangioscopy raised these values to 100%
 Chapter 29 Indeterminate Biliary Stricture

A B

Fig. 29.16 A Radiograph of a cholangioscope advanced to the level of the proximal

extrahepatic duct. B Cholangioscopic view of hepatic conuence with open left hepatic
duct and tumor occluding right hepatic duct.

A
and 94% respectively.33 Cholangioscopy may be particularly useful
in differentiating benign from malignant strictures in the setting of
primary sclerosing cholangitis. One study demonstrated improved
sensitivity (92% vs 66%, p = 0.25), specicity (93% vs 51%, p <
0.001), and accuracy (93% vs 55%, p < 0.001) for cholangioscopic B
characterization compared to radiographic characterization.16 The
criteria for suspicion of malignancy in this study were the presence
of an associated polypoid or villous mass or irregularly shaped
ulceration.
A recently developed technology employs a single-use 10 Fr
multi-channeled sheath (SpyglassTM Direct Visualization System;
Boston Scientic, Marlboro, MA) that attaches to the head of the
duodenoscope just below the biopsy port and advances through
the accessory channel for insertion into the duct (Fig. 29.17).
The sheath provides 4-way steering of the tip, illumination, water
ushing, a channel for passage of a 0.035 caliber beroptic
probe for visualization of the duct (SpyScope), and another for C D
either wire guidance or passage of therapeutic or sampling devices
such as the electrohydraulic probe, biopsy cables (SpyBite), or
cytology brushes. Preliminary data demonstrate utility in 18 of 20
cases for directed biopsy, stone therapy, or clarication of anatomy
or pathology.35
Stent placement for biliary decompression is the major therapeu-
tic modality used for patients with indeterminate strictures. Stent-
ing as denitive therapy for benign lesions or as palliative therapy
for malignant obstruction is discussed in Chapters 16 and 17.
Plastic stents are usually employed for palliation of indeterminate
strictures to ensure the ability to remove them at a later endoscopic
or surgical procedure and to limit the cost of potentially brief dura-
tion stenting. When diagnosis remains indeterminate, serial proce-
dures and repeated tissue sampling are often performed; hence
shorter intervals of drainage and smaller caliber 7 and 8.5 French
stents may sufce. If subsequent diagnostic investigations are
chosen that employ EUS rather than ERCP or if the patient is not Fig. 29.17 The SpyGlass Direct Visualization System attached
a surgical candidate regardless of the diagnosis, it is preferable to to the duodenoscope and advanced into the accessory channel.
Close-up demonstrates ratchets for steering and ports for passage
palliate the indeterminate lesion with larger caliber 10 French of guidewire, biopsy forceps, and 0.035 SpyGlass beroptic
stents that remain patent longer and may minimize the number of probe.

323
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
subsequent procedures. In the patient with an indeterminate stric-
ture self-expandable metal stents (SEMS) are typically avoided due
to both their permanence and their expense. Partially coated SEMS
are generally removable when they are left extending into the duo-
denum, and fully coated variants designed specically for removal
are under investigation.36,37 Their use can be entertained in the
indeterminate lesion if prolonged stenting for either treatment of
benign stricture or palliation of cancer is desirable. Plastic stents
REFERENCES
1. Afroudakis A, Kaplowitz N. Liver histopathology in chronic
common bile duct stenosis due to chronic alcoholic pancreatitis.
Hepatology 1981; 1:6572.
2. Lesur G, Levy P, Flejou JF, et al. Factors predictive of liver
histopathological appearance in chronic alcoholic pancreatitis
with CBD stenosis and increased serum alkaline phosphatase.
Hepatology 1993; 18:107811081.
3. ASGE Standards of Practice Committee. An annotated algorithmic
approach to malignant biliary obstruction. Gastrointest Endosc
2001; 53:849852.
4. Torok N, Gores GJ. Cholangiocarcinoma. Seminars in
Gastrointestinal Disease 2001; 12:125132.
5. Klimstra DS, Adsay NV. Lymphoplasmacytic sclerosing
(autoimmune) pancreatitis. Seminars in Diagnostic Pathology
2004; 21:237246.
6. Chari ST, Smyrk TC, Levy MJ, et al. Diagnosis of autoimmune
pancreatitis: the Mayo Clinic experience. Clin Gastroenterol
Hepatol 2006 Aug; 4(8):10101016.
7. De Palma GD, Pezzullo A, Rega M, et al. Unilateral placement of
metallic stents for malignant hilar obstruction: A prospective
study. Gastrointestinal Endoscopy 2003; 58:5053.
8. Rosch T, Meining A, Fruhmorgen S, et al. A prospective
comparison of the diagnostic accuracy of ERCP, MRCP, CT, and
EUS in biliary strictures. Gastrointestinal Endoscopy 2002;
55:870876.
9. Domagk D, Wessling J, Reimer P, et al. Endoscopic retrograde
cholangio-pancreatography, intraductal ultrasonography, and
magnetic resonance cholangio-pancreatography in bile duct
strictures: A prospective comparison of imaging diagnostics with
histopathological correlation. American Journal of
Gastroenterology 2004; 99:16841689.
10. Lee JH, Salem R, Aslanian H, et al. Endoscopic ultrasound and
ne-needle aspiration of unexplained bile duct strictures.
American Journal of Gastroenterology 2004; 99:10691073.
11. Eloubeidi MA, Chen VK, Jhala NC, et al. Endoscopic ultrasound-
guided aspiration biopsy of suspected cholangiocarcinoma.
Clinical Gastroenterology and Hepatology 2004; 2:209213.
12. Rosch T, Hofrichter K, Fringberger E, et al. ERCP or EUS for tissue
diagnosis of biliary strictures? A prospective comparative study.
Gastrointestinal Endoscopy 2004; 60:390396.
13. Wiersema MJ, Fletcher JG, Jondal ML, et al. Prospective evaluation
of triple phase multidetector computed tomography, dynamic
gadolinium-enhanced magnetic resonance imaging and endo-
sonography in potentially resectable pancreatic adenocarcinoma.
Gastrointestinal Endoscopy 2002 Oct; 56(4 Suppl):S123.
14. Bain VG, Abraham N, Jhangri GS, et al. Prospective study of biliary
strictures to determine the predictors of malignancy. Endoscopy
2000; 14:397402.
15. MacCarty RL, LaRusso NF, May GR, et al. Cholangiocarcinoma
complicating primary sclerosing cholangitis: cholangiographic
appearances. Radiology 1985; 156:4346.
324
remain preferable for the patient with a mass lesion and high likeli-
hood of resectability.
Modern imaging and new analytical techniques have advanced
our ability to characterize indeterminate biliary strictures. Neverthe-
less, in some patients with indeterminate strictures a denitive diag-
nosis cannot be made via minimally invasive approaches. In these
instances, surgical exploration with a goal of diagnosis and resection
should be considered in operable patients.
16. Tischendorf JJW, Kruger M, Trautwein C, et al. Cholangioscopic
characterization of dominant bile duct stenosis in patients with
sprimary sclerosing cholangitis. Endoscopy 2006; 38:665669.
17. ASGE Technology Committee. Technology assessment status
evaluation: guidewires in gastrointestinal endoscopy.
Gastrointestinal Endoscopy 1998; 47:579583.
18. De Bellis M, Sherman S, Fogel EL, et al. Tissue sampling at ERCP in
suspected malignant biliary strictures (Part 2). Gastrointestinal
Endoscopy 2002; 56:720730.
19. Fogel EL, deBellis M, McHenry L. Effectiveness of a new long
cytology brush in the evaluation of malignant biliary obstruction:
a prospective study. Gastrointest Endosc 2006; 63:7177.
20. Barkun A, Liu J, Carpenter S, et al. Update on endoscopic tissue
sampling devices. Gastrointestinal Endoscopy 2006; 63:741745.
21. Howell DA, Beveridge RP, Bosco J, et al. Endoscopic needle
aspiration biopsy at ERCP in the diagnosis of biliary strictures.
Gastrointestinal Endoscopy 1992; 38:531535.
22. Lin LF, Siauw CP, Ho KS, et al. Guidewire technique for
endoscopic transpapillary procurement of bile duct biopsy
specimens without endoscopic sphincterotomy.Gastrointest
Endosc 2003 Aug; 58(2):272274.
23. Ponchon T, Gagnon P, Berger F, et al. Value of endobiliary brush
cytology and biopsies for the diagnosis of malignant bile duct
stenosis: results of a prospective study, Gastrointestinal
Endoscopy 1995; 42:565572.
23a. Jailwala J, Fogel EL, Sherman S, et al. Triple-tissue sampling at
ERCP in malignant biliary obstruction. Gastrointest Endosc 2000;
51:383390.
24. Ryan ME, Baldauf MC. Comparison of ow cytometry for DNA
content and brush cytology for detection of malignancy in
pancreaticobiliary strictures. Gastrointest Endosc 1994; 40:
133139.
25. Baron TH, Harewood GC, Rumalla A, et al. A prospective
comparison of digital image analysis and routine cytology for the
identication of malignancy in biliary tract strictures. Clinical
Gastroenterology and Hepatology 2004; 2:214219.
26. Kipp BR, Stadheim LM, Halling SA, et al. A comparison of routine
cytology and uorescence in situ hybridization for the detection
of malignant bile duct strictures. Am Journal of Gastroenterology
2004; 99:16751681.
27. Farrell RJ, Agarwal B, Brandwein SL, et al. Intraductal US is a useful
adjunct to ERCP for distinguishing malignant from benign biliary
strictures. Gastrointest Endoscopy 2002; 56:681687.
28. Vazquez-Sequeiros E, Baron TH, Clain JE, et al. Evaluation of
indeterminate bile duct strictures by intraductal US.
Gastrointestinal Endoscopy 2002; 56:372379.
29. Inui K, Miyoshi H. Cholangiocarcinoma and intraductal sonography.
Gastrointestinal Endoscopy Clinics of NA 2005; 15:143155.
30. Levy MJ, Rumalla A, Baron TH, et al. Prospective Evaluation of
Intraductal Ultrasound (IDUS), Digital Image Analysis (DIA), Fluores-
cence in Situ Hybridization (FISH), CA 19-9, and ERCP with routine

cytology and intraductal biopsy in the evaluation of indeterminate
bile duct strictures. Gastrointestinal Endoscopy 2006; 63:AB88.
31. Tamada K, Ido K, Ueno N, et al. Preoperative staging of
extrahepatic bile duct cancer with intraductal ultrasonography.
American Journal of Gastroenterology 1995; 90:239246.
32. Farrell JJ, Bounds BC, Al-Shalabi S, et al. Single-operator
duodenoscope-assisted cholangioscopy is an effective alternative
in the management of choledocholithiasis not removed by
conventional methods, including mechanical lithotripsy.
Endoscopy 2005; 37:542547.
33. Fukuda Y, Tsuyuguchi T, Sakai Y, et al. Diagnostic utility of peroral
cholangioscopy for various bile-duct lesions. Gastrointestinal
Endoscopy 2005; 62:374382.
Chapter 29 Indeterminate Biliary Stricture
34. Shah RJ, Langer DA, Antillon MR, et al. Cholangioscopy and
cholangioscopic forcepts biopspy in patients with indeterminate
pancreaticobiliary pathology. Clinical Gastroenterology and
Hepatology 2006; 4:219225.
35. Chen YK. Results from the rst human use clinical series utilizing
a new peroral cholangiopancreatoscopy system (Spyglass(tm)
Direct Visualization System). Gastrointestinal Endoscopy 2006; 63:
AB86.
36. Familiari P, Bulajic M, Mutignani M, et al. Endoscopic removal of
malfunctioning biliary self-expandable metallic stents.
Gastrointestinal Endoscopy 2005; 62:903910.
37. Petersen BT. SEMS removal: salvage technique or new paradigms.
Gastrointestinal Endoscopy 2005; 62:911913.
325
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Benign Biliary Strictures
30 Guido Costamagna, Syed G. Shah and Lalit Shimpi
BOX 30.1 KEY POINTS
There are a variety of causes of benign biliary strictures
Postoperative biliary strictures are the most common benign
biliary stricture
Endoscopic therapy for benign biliary strictures consists of
dilation and stent placement using multiple, large-bore
plastic stents
Chronic pancreatitis strictures are less responsive to
endoscopic therapy
The short-term outcome following endoscopic treatment of
benign strictures is excellent
Successful long-term outcome following endoscopic
therapy of benign biliary strictures is comparable to surgery,
and does not preclude subsequent surgical therapy in cases
that fail or recur
The majority of benign bile duct strictures occur as a result of post-
operative iatrogenic injury, most commonly following cholecystec-
tomy, or occur at the site of biliary anastomosis after hepatic resection
or liver transplantation. Benign strictures may also result from a
variety of other causes (Table 30.1). The endoscopic management of
benign biliary strictures is also covered in Chapters 31, 35, and 43.
Bile duct injuries are reported to be higher during laparoscopic
cholecystectomy than open surgery.1 The estimated overall incidence
of biliary injuries following laparoscopic cholecystectomy has been
reported to be between 0.2% and 1.7%.2,3 Misidentication of ana-
tomic structures during dissection, presence of acute inammation
or brous adhesions in the gallbladder fossa, excessive use of elec-
trocautery to control bleeding, inaccurate placement of clips or
sutures, and excessive traction on the gallbladder neck are major
causes.2 Bergman et al.4 described four types of postoperative bile
duct injuries: Type A-cystic duct leaks or leakage from aberrant or
peripheral hepatic radicles, Type B-major bile duct leaks with or
without concomitant biliary strictures, Type C-bile duct strictures
without bile leakage, and Type D-complete transection of the duct
with or without excision of some portion of the biliary tree. Postop-
erative biliary strictures occur in 0.20.5% of patients and usually
occur as a result of partial or complete transection by clipping or
ligation of the bile duct, or less frequently as a result of vascular
injury during dissection or cauterization. Injury to sectorial or
segmental branches may occur in patients with anatomic anomalies
of the biliary tree.
Approximately 1030% of patients with chronic pancreatitis
develop symptomatic biliary stenosis.5 Biliary obstruction due to
compression by an edematous pancreatic head or pseudocyst usually
resolves when the inammation subsides or after resolution of the
pseudocyst. However, obstruction caused by a brotic stricture does
not resolve spontaneously and requires therapeutic intervention.
CLINICAL FEATURES
Approximately 10% of postoperative bile duct strictures present
within 1 week of surgery. These usually occur as a result of inad-
vertent clipping or ligation of the common bile duct and may or
may not be associated with biliary leaks. Patients may present with
abdominal pain, fever, pruritus, jaundice or biliary stula. However,
in the majority of cases presentation is delayed and 7080% present
within 612 months of surgery.6 The presentation is symptomatic
or asymptomatic cholestasis, recurrent cholangitis, stone formation,
or secondary biliary cirrhosis. Bismuth7 classied benign strictures
into ve types: Type 1-Low common hepatic duct (CHD) or bile
duct (CHD >2 cm), Type 2-Mid common hepatic duct (CHD <2 cm),
Type 3-Hilar stricture, Type 4-Destruction of hilar conuence (right
and left hepatic ducts separated) and Type 5-Involvement of right
hepatic branch alone or with common duct. Bismuth types 1 and 2
strictures are most frequent in reported series.810 The clinical
presentation of biliary strictures is somewhat different in patients
with chronic pancreatitis.11 In a retrospective survey of 78 patients
with chronic pancreatitis, overt jaundice was found in only a
minority of patients.12 No relationship was found between features
of the common bile duct and severity of pancreatitis, or disease
duration.
DIAGNOSIS
The clinical diagnosis of postoperative biliary stricture is usually
suspected by the onset of either symptomatic and/or biochemical
cholestasis in the early or late postoperative period. In the rst
instance, ultrasound examination is performed to conrm biliary
dilation and may suggest the level of biliary obstruction. MRCP is a
useful, non-invasive diagnostic modality for accurately delineating
the biliary anatomy and site of stenosis, and for planning denitive
therapy.13
MANAGEMENT
Traditionally, postoperative bile duct strictures were managed
surgically and the role of ERCP was limited to the diagnosis
and denition of the level and extent of the stricture.14 With the
327
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Postoperative
Anastomotic
Non-anastomotic
Ischemia (including polyarteritis nodosa)
Primary or secondary sclerosing cholangitis
Post-sphincterotomy
Chronic pancreatitis
Radiation therapy
Portal bilopathy
Post-treatment of biliary lymphoma
Tuberculosis
Abdominal trauma
Post-radiofrequency ablation
Endoscopic injection sclerotherapy of duodenal ulcer bleeding
Table 30.1 Causes of benign biliary strictures
increasing use of ERCP for the treatment of acute complications
of cholecystectomy, therapeutic ERCP has been extensively adopted
to manage postoperative strictures and other benign biliary stric-
tures. Percutaneous transhepatic therapy with balloon dilation of
the stricture is limited by low success rates, high stricture recur-
rence rates, and high complication rates.15,16 The high stricture
recurrence rate following percutaneous transhepatic pneumatic
dilation is most likely due to forceful disruption of the scar, which
can add further traumatic damage to the tissue and consequential
development of a new brotic reaction. Endoscopic treatment of a
postoperative bile duct stricture is often preferred over percutane-
ous techniques because it avoids the need for liver puncture and
access to non-dilated intrahepatic ducts is easier. Also, the endo-
scopic approach is more comfortable for the patient and is safer in
the presence of cirrhosis, ascites, or coagulopathy. Percutaneous
transhepatic techniques are now usually reserved for failed endo-
scopic procedures.
ENDOSCOPIC TECHNIQUE
Endoscopic treatment of benign biliary strictures involves two
technical steps: (1) negotiating the stricture and (2) dilation of
the stricture.
Negotiating the stricture
Negotiating the stricture requires continuity of the common bile
duct (CBD). In cases of complete transection or ligation of the CBD,
a guidewire cannot be passed across the lesion and thus endother-
apy alone is not feasible. In such cases, surgical reconstruction is
indicated, though a combined percutaneous endoscopic approach
has been described.17 Negotiating benign biliary strictures is often
much more difcult than neoplastic strictures because the stenosis,
even if short, may be asymmetric. Furthermore, associated brosis
makes them thin and tighter. It is, therefore, often necessary to use
thin hydrophilic guidewires (0.021 or 0.018-inch diameter) with a
straight or curved (J-shaped) tip to get across. Guidewire manipula-
tion requires patience, skill, and optimal uoroscopic imaging.
Forceful maneuvers may create false passages and should be
avoided. Pulling on an inated stone retrieval balloon positioned
just distal to the stricture results in stretching of the bile duct and
modication of the axis of the guidewire. Steerable catheters or
papillotomes may also be used to achieve the same result. Once the
stricture is traversed, the hydrophilic guidewire is exchanged for a
328
stiffer one to facilitate dilation. Sphincterotomy is usually per-
formed due to the necessity for repeated stent exchanges and
to allow for side-by-side placement of large-bore stents. When
a stricture cannot be traversed at ERCP, a combined endoscopic
percutaneous approach (rendezvous) can be used.
Dilation of the stricture
Stricture dilation has two objectives: (1) to reopen the CBD to achieve
bile drainage and (2) to keep the stenosis open and avoid re-strictur-
ing. Insertion of the guidewire through the stricture is followed by
placement of a 6 Fr catheter over the guidewire, and by mechanical
dilation with a 9.5 Fr Cunningham-Cotton sleeve (Cook Endoscopy,
Winston-Salem, NC, USA) to test the caliber of the stricture prior to
attempting stent insertion. Hydrostatic balloon dilation with 4, 6,
and 8 mm balloons (e.g. Max Force, Boston Scientic, MA, USA)
may be necessary in cases where the stricture is not amenable to
mechanical dilation. Balloon dilation is usually performed to a size
12 mm larger than the downstream bile duct diameter. Although
immediately effective, endoscopic and percutaneous balloon dilation
alone, whether in single or multiple sessions, is considered inade-
quate and associated with a high restenosis rate (up to 47%).8,18,19
Stent placement, on the other hand, keeps the stricture open for a
prolonged period to allow scar remodeling and consolidation.20
When mechanical and/or balloon dilation is unsuccessful, leaving a
5 or 6 Fr nasobiliary drainage tube in situ for 24 to 48 hours may
increase the chances of subsequent endoscopic stent placement.
Typically, 10 to 12 Fr polyethylene stents are placed and exchanged
every 34 months to prevent cholangitis from stent occlusion.
However, the optimal number of stents and duration of stent place-
ment for stricture resolution has not been established.
OUTCOMES OF BILIARY STENTING
Several variations in endoscopic stent protocols have been described
for the treatment of postoperative benign biliary strictures with suc-
cessful stricture resolution reported in 7490% of patients.8,21
Bergman and colleagues reported on the results of endoscopic stent
therapy in 74 patients with postoperative biliary strictures.22 Two
10 Fr stents were placed whenever possible, and exchanged every 3
months for a period of 1 year. Technical success of initial stenting
was reported to be 80%. However, due to a variety of reasons only
59% (44 patients) of the original cohort completed the 12-month
stenting period. Amongst these, endoscopic stent placement failed
in 11 patients due to complete bile duct obstruction. Recurrent ste-
nosis occurred in 9 of 44 patients (20%) after a median follow-up of
9.1 years with the majority of patients presenting within six months
of stent removal (median 2.6 months, range 2 months to 15 years).
Of the original cohort, 35 of 74 patients (47%) were stricture free at
the end of the follow-up period. Major complications, including
cholangitis, pancreatitis, bleeding and stent migration, were
more common in patients who were non-compliant with the stent
exchange protocol.
A more aggressive biliary stent approach has been reported by
our own institution.9 In 45 patients with postoperative biliary stric-
tures, we placed as many stents as possible during each ERCP
session with elective stent exchange performed every 3 months. At
each exchange, all previously placed stents were removed and the
maximum number of large-diameter stents inserted, as permitted
by the diameter of the duct distal to the stricture and the diameter
of the stricture itself. The treatment protocol was stopped when
 Chapter 30 Benign Biliary Strictures

A B C D

E F G H

Fig. 30.1 A Bismuth type 2 stricture after laparoscopic cholecystectomy. B Guidewire negotiated across stricture. C A single large-
bore plastic stent has been placed. Cholangiogram shows co-existent bile leak at the site of the biliary stricture. D At repeat ERCP, two
large-bore plastic stents have been placed. E Persistence of biliary stricture after removal of stents. The stricture has been traversed
using a guidewire and biliary balloon dilator. F Biliary balloon dilation performed. G Post-dilation, three large-bore plastic stents have
been placed. H Disappearance of the stricture after removal of the stents.

complete morphologic disappearance of the stricture was demon-
strated by occlusion cholangiography after stent removal, and con-
rmed by subsequent cholangiography performed through a
nasobiliary drain 24 to 48 hours later (Fig. 30.1).
Complete disappearance of the stricture was dened as absence
of any signicant indentation at the site of previous narrowing. Of
the 42 patients who completed the stenting protocol, a percutaneous
endoscopic rendezvous technique was necessary for endoscopic
biliary access in 3 patients, and balloon dilation was required in 40%
of patients prior to stent placement. The mean number of stents
placed at the rst session was 1.7 0.7 (range 14), while that at the
last ERCP session was 3.2 1.3 (range 16). Mean duration of treat-
ment was 12.1 5.3 (range 224) months with a mean number of
4.1 1.3 (range 28) procedures per patient. Stents were ultimately
removed from all patients. The mean follow-up was 48 months
(range 24 months11.3 years). On an intention-to-treat basis, endo-
scopic treatment was successful in 89% of patients. Complications
occurred in 9% of patients (including three patients with cholangitis,
and one with pancreatitis) and all were managed conservatively.
There was no mortality. Even though the follow-up period in our
study was shorter than the Amsterdam group, it is longer than the
typical period during which all the recurrences after endoscopic
therapy have been described (within 2 years).
These promising results with endoscopic therapy have also been
replicated in a recent prospective trial involving 43 patients with
post-laparoscopic cholecystectomy bile duct strictures in whom a
similar protocol for endoscopic stenting was followed.23 A mean of
3.4 0.6 (range 35) stents were placed for a period of 1 year. The
authors reported a 100% success rate for stricture dilation with no
stricture recurrence at a mean follow-up of 16.0 11.1 (range 142)
months after stent removal.
Previously published data have shown that endoscopic therapy is
at least as effective as surgical treatment for benign biliary strictures
with reported success rates of approximately 80%.24 Although there
are no randomized controlled trials comparing the two modalities,
retrospective and case controlled studies comparing both treatment
modalities have shown similar long-term results and re-stenosis
rates.2426 Surgery may, however, be associated with higher early
morbidity and mortality.25,27,28 Endoscopic therapy, on the other
hand, offers several advantages including relative simplicity,

329
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
reversibility, and above all minimal invasiveness. An additional
advantage of endotherapy is that treatment failure does not preclude
subsequent surgery, whereas hepaticojejunostomy, which is the
classical surgical procedure, makes future endotherapy difcult, if
not impossible. The major disadvantage of endotherapy, however, is
the need for multiple procedures. Most studies have employed a
protocol of elective stent exchange to avoid cholestasis and/or chol-
angitis, secondary to stent occlusion. In case of patient non-compli-
ance, late morbidity with endoscopic therapy may increase. Assurance
of patient compliance is therefore of paramount importance to
ensure successful endoscopic therapy.
As with surgery, strictures at or above the main hepatic conu-
ence are often more challenging than strictures below the hepatic
conuence (Fig. 30.2). In the study by Draganov et al. a high success
rate was achieved in patients with Bismuth type 1 or 2 strictures
(80%), and the lowest in type 3 strictures (25%).8
The favorable results of endoscopic therapy for post-cholecystec-
tomy biliary strictures are also reected in the treatment of anasto-
motic strictures following orthotopic liver transplantation. Though
poor long-term outcome has been reported with endoscopic balloon
dilation alone,29 good to excellent results (7490% clinical success
rates at 35 years follow-up) have been reported following multiple
stent insertion in patients with post-transplant anastomotic biliary
strictures.30,31 Endoscopic therapy is associated with a low complica-
tion rate and may avoid the need for surgical intervention.30,32
However, surgery should be considered for those patients who are
non-responsive to endoscopic therapy.31
Unlike postoperative benign biliary strictures, long-term results
of endoscopic therapy for bile duct strictures associated with chronic
pancreatitis have not been encouraging. Although short-term results
are excellent with immediate relief of cholestasis in almost all cases,
it remains unclear whether endoscopic plastic stent placement can
achieve denite stricture resolution, especially in tight, brotic stric-
tures.33 No prospective studies have been performed, and a number
of small, retrospective studies have reported success rates ranging
between 10 and 95%. Cahen et al. recently reported a success rate
of only 38% for endoscopic plastic stent placement in 58 patients
with distal bile duct strictures due to chronic pancreatitis.33 In this
retrospective study single, large-bore (10 Fr) plastic stents were
placed in the majority of patients and exchanged every 3 months, or
when signs of stent dysfunction were present, for a period of 1 year.
In those patients in whom stricture dilation was successful there was
no evidence of recurrent biliary stricture after a median follow-up of
85 months. Concomitant acute pancreatitis was the only factor iden-
tied as predictive of a successful outcome (92% stricture resolution
with concomitant pancreatitis vs 24% without pancreatitis).33 Stric-
ture dilation beyond 1 year had no additional benets. A more
aggressive approach using multiple plastic stents has also been tried,
but with little improvement in success rates.8
The poor results of endoscopic therapy for chronic pancreatitis
induced biliary strictures following plastic stent placement have
encouraged the use of self-expandable metal stents (SEMS). However,
the majority of studies reporting on the use of uncovered SEMS have
been small with variable results.34,35 Uncovered SEMS may be useful
and considered in those patients who either refuse surgery, or in
whom surgery is contra-indicated due to concomitant portal hyper-
tension or patient co-morbidity.
Overall, the long-term results of stenting using uncovered SEMS
for all types of benign biliary strictures has been disappointing
because patency is usually short term, and most stents eventually
330
obstruct due to mucosal hyperplasia, biliary sludge or calculi.26,36,37
Placing SEMS in such patients may therefore add to morbidity,
require repeated endoscopic sessions to re-establish stent patency,
and may preclude surgical therapy if placed too proximally. Current
data does not support the routine placement of SEMS for benign
biliary strictures.
As opposed to uncovered SEMS, covered SEMS may have added
applicability because of potential removability. One trial, published
in abstract form is encouraging.38 Wallstents were placed in 55
patients for benign biliary strictures (chronic pancreatitis, n = 25;
stones, n = 16; liver transplantation, n = 11; surgical bile duct injury,
n = 3) and the stents could be easily removed. At a median follow-up
of 3 months (range 111), relapse occurred in only 10% of patients.
The advantage of this approach is the large dilation diameter
which can be achieved after one endoscopic procedure. Larger
studies with longer follow-up are needed to conrm these results.
In addition, when the Wallstent is placed, one must be sure that the
distal end is well into the duodenum so that it can be subsequently
removed.39
Post-sphincterotomy distal bile duct strictures are a distinct entity
that deserve mention. It is estimated that post-sphincterotomy distal
biliary strictures occur in about 2% of patients when the indication
is for choledocholithiasis. A recent study in 20 patients with post-
sphincterotomy stricture using sequential insertion of multiple
plastic biliary stents was published.40 After a median of nine months
of follow-up with 18 patients the success rate was 90%. Two patients
developed stricture recurrences at 10 and 24 months.
Complications of biliary stenting may occur at any time during
the rst or subsequent treatment sessions. Early complications
relate to sphincterotomy, such as acute pancreatitis, perforation or
bleeding, and occur with similar frequency to those encountered
during therapeutic ERCP for other indications, such as removal of
CBD stones. Complications arising during the course of stent
therapy are mostly due to stent dysfunction, including obstruction,
BOX 30.2 INDICATIONS AND
CONTRAINDICATIONS
The indications for treatment of a benign biliary stricture are
symptoms and/or cholestasis.
Contraindications to endoscopic therapy are inability to
traverse the stricture and severe, uncorrectable
coagulopathy.
BOX 30.3 COMPLICATIONS AND
CONTROVERSIES
Complications of endoscopic therapy include cholangitis,
ductal or luminal perforation, pancreatitis, bleeding, and
stent migration.
The use of self-expanding metal stents for treatment of
benign biliary strictures is controversial.
 Chapter 30 Benign Biliary Strictures

Fig. 30.2 A Bismuth type 4 bile duct stricture with two guidewires
A B in place in the paramedian branch of the right hepatic and left
hepatic ducts. B Three large-bore plastic stents have been placed.
C Persistence of stricture after removal of stents. D Five large-
bore plastic stents have been placed. E At subsequent ERCP, eight
large-bore plastic stents have been placed. F Cholangiogram
showing resolution of biliary stricture after stent removal.

331
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
migration, and stent impaction. Stent dysfunction usually manifests
as acute cholangitis, and requires prompt endoscopic assessment
and stent replacement or repositioning to establish biliary drainage.
Complications of long-term stenting include the formation of stones
or biliary sludge above the stricture, which may be asymptomatic or
present with cholangitis. Regular stent exchange every three months
is essential in minimizing the risk of stone formation. Removal of
all stones and sludge by basket and/or balloon extraction is manda-
tory before replacement of new stent(s) to minimize the risk of early
reocclusion.
In the future, novel endoscopic stents may become available
which will improve the success rates of endoscopic therapy in
patients with benign biliary strictures. The development of bioab-
sorbable stents, dedicated removable expandable stents, or drug-
eluting stents that may be coated with steroids or chemotherapeutic
agents that inhibit endothelial growth or brosis (similar to coronary
stents),41 could offer new possibilities for treatment.
The management of primary sclerosing cholangitis (PSC) biliary
strictures is discussed in Chapter 35. In general, the approach differs
from that of postoperative strictures in several ways. The intra and
extrahepatic bile ducts are usually small and do not lend themselves
REFERENCES
1. Deziel DJ, Millikan KW, Economou SG. Complications of
laparoscopic cholecystectomy: a national survey of 4292
hospitals and an analysis of 77,604 cases. Am J Surg 1992;
165:L9L14.
2. Archer SB, Brown DW, Smith CD, et al. Bile duct injury during
laparoscopic cholecystectomy: results of a national survey. Ann
Surg 2001; 234:549558.
3. Nuzzo G, Giuliante F, Giovannini I, et al. Bile duct injury during
laparoscopic cholecystectomy: results of an Italian national survey
on 56 591 cholecystectomies. Arch Surg 2005; 140:986992.
4. Bergman JJGHM, van den Brink GR, Rauws EA. Treatment of bile
duct lesions after laparoscopic cholecystectomy. Gut 1996;
38:141147.
5. Jakobs R, Riemann JF. The role of endoscopy in acute recurrent
and chronic pancreatitis and pancreatic cancer. Gastroenterol Clin
North Am 1999; 28:783800.
6. Pitt HA, Miyamoto T, Parapatis SK. Factors inuencing outcome in
patients with postoperative biliary strictures. Am J Surg 1982;
144:1421.
7. Bismuth H. Postoperative strictures of the bile duct. In: Blumgart
LH, ed. The biliary tract. Edinburgh: Churchill Livingstone
1982:209218.
8. Draganov P, Hoffman B, Marsh W, et al. Long-term outcome in
patients with benign biliary strictures treated endoscopically with
multiple stents. Gastrointest Endosc 2002; 55:680686.
9. Costamagna G, Pandol M, Mutignani M, et al. Long-term results
of endoscopic management of postoperative bile duct strictures
with increasing numbers of stents. Gastrointest Endosc 2001;
54:162168.
10. Bauer TW, Morris JB, Lowenstein A, et al. The consequences of a
major bile duct injury during laparoscopic cholecystectomy.
J Gastrointest Surg 1998; 2:6166.
11. Mutignani M, Tringali A, Costamagna G. Therapeutic biliary
endoscopy. Endoscopy 2004; 36:147159.
12. Angelini G, Sgarbi D, Castagnini A, et al. Common bile duct
involvement in chronic pancreatitis. Ital J Gastroenterol Hepatol
1994; 26:7982.
332
to placement of multiple large-bore stents. The occlusion rate of
large-bore stents is higher, such that shorter duration (4 weeks) stent
placement has been advocated. Finally, the possibility that an under-
lying cholangiocarcinoma is present must always be considered.
Therefore, tissue sampling is crucial during the management of PSC
patients.
At the present time there are no studies that have addressed
cost-effectiveness of endoscopic therapy compared to surgical
therapy.
In conclusion, an endoscopic attempt at management of post-
operative bile duct strictures should be undertaken as a rst-line
treatment in most instances. An aggressive approach placing mul-
tiple stents improves the results of endoscopic therapy. Surgical
reconstruction should be considered for complete transection of the
bile duct, when endotherapy fails, and following stricture recur-
rence. The role of SEMS in benign biliary strictures is not yet
clearly dened due to variable results and small numbers, and is
currently not recommended. In patients with biliary strictures
secondary to chronic pancreatitis, SEMS may be considered as an
alternative to surgery in patients who are considered to be poor
surgical candidates.
13. Khalid TR, Casillas VJ, Montalvo BM. Using MR
cholangiopancreatography to evaluate iatrogenic bile duct injury.
AJR 2001; 177:13471352.
14. Vallon AG, Mason RR, Laurence BH, et al. Endoscopic retrograde
cholangiography in post-operative bile duct strictures. Br J Radiol
1982; 55:3235.
15. Mueller PR, van Sonnenberg E, Ferrucci Jr JT. Biliary stricture
dilatation: multicenter review of clinical management in 73
patients. Radiology 1986; 160:1722.
16. Williams HJ, Jr., Bender CE, May GR. Benign postoperative biliary
strictures: dilation with uoroscopic guidance. Radiology 1987;
163:629634.
17. Dumonceau JM, Baize M, Deviere J. Endoscopic transhepatic
repair of the common hepatic duct after excision during
cholecystectomy. Gastrointest Endosc 2000; 52:540543.
18. Smith MT, Sherman S, Lehman G. Endoscopic management
of benign strictures of the biliary tree. Endoscopy 1995;
27:253266.
19. Foutch PG, Sivak MJ, Jr. Therapeutic endoscopic balloon dilation
of the extra-hepatic biliary ducts. Gastroenterology 1985;
80:575580.
20. Berkelhammer C, Kortan P, Haber GB. Endoscopic biliary
prostheses as treatment for benign postoperative bile duct
strictures. Gastrointest Endosc 1989; 35:95101.
21. De Masi E, Fiori E, Lamazza A, et al. Endoscopy in the treatment of
benign biliary strictures. Ital J Gastroenterol Hepatol 1998;
30:9195.
22. Bergman JJ, Burgemeister L, Bruno MJ. Long-term follow-up after
biliary stent placement for postoperative bile duct stenosis.
Gastrointest Endosc 2001; 54:154.
23. Kuzela L, Oltman M, Sutka J, et al. Prospective follow-up of
patients with bile duct strictures secondary to laparoscopic
cholecystectomy, treated endoscopically with multiple stents.
Hepatogastroenterology 2005; 52:13571361.
24. Tocchi A, Mazzoni G, Liotta G, et al. Management of benign biliary
strictures: biliary enteric anastomosis vs endoscopic stenting. Arch
Surg 2000; 135:153157.

25. Davids PHP, Tanka AKF, Rauws EAJ. Benign biliary strictures.
Surgery or Endoscopy? Ann Surg 1993; 217:237243.
26. Dumonceau JM, Deviere J, Delhaye M, et al. Plastic and metal
stents for postoperative benign bile duct strictures: the best and
the worst. Gastrointest Endosc 1998; 47:817.
27. Csendes A, Diaz C, Burdiles P, et al. Indications and results of
hepaticojejunostomy in benign strictures of the biliary tract.
Hepatogastroenterology 1992; 39:333336.
28. Frattaroli FM, Reggio D, Guadalaxara A, et al. Benign biliary
strictures: a review of 21 years of experience. J Am Coll Surg
1996; 183:506513.
29. Schwartz DA, Petersen BT, Poterucha JJ, et al. Endoscopic therapy
of anastomotic bile duct strictures occurring after liver
transplantation. Gastrointest Endosc 2000; 51:169174.
30. Morelli J, Mulcahy HE, Willner IR, et al. Long-term outcomes for
patients with post-liver transplant anastomotic biliary strictures
treated by endoscopic stent placement. Gastrointest Endosc 2003;
58:374379.
31. Rossi A, Grosso C, Zanasi G, et al. Long-term efcacy of
endoscopic stenting in patients with stricture of biliary
anastomosis after orthotopic liver transplantation. Endoscopy
1998; 30:360366.
32. Mahajani RV, Cotler SJ, Ulzer MF. Efcacy of endoscopic
management of anastomotic biliary strictures after hepatic
transplantation. Endoscopy 2000; 32:943949.
33. Cahen DL, van Berkel AM, Oskam D, et al. Long-term results of
endoscopic drainage of common bile duct strictures in chronic
pancreatitis. Eur J Gastroenterol Hepatol 2005; 17:103108.
Chapter 30 Benign Biliary Strictures
34. Kahl S, Zimmermann S, Glasbrenner B, et al. Treatment of benign
biliary strictures in chronic pancreatitis by self-expandable metal
stents. Dig Dis 2002; 20:199203.
35. van Berkel AM, Cahen DC, Van Westerloo DJ, et al. Self-expanding
metal stents in benign biliary strictures due to chronic pancreatitis.
Endoscopy 2004; 36:381384.
36. OBrien SM, Hateld AR, Craig PI, et al. A 5-year follow-up of self-
expanding metal stents in the endoscopic management of
patients with benign bile duct strictures. Eur J Gastroenterol
Hepatol 1998; 10:141145.
37. Siriwardana HP, Siriwardena AK. Systematic appraisal of the role
of metallic endobiliary stents in the treatment of benign bile duct
stricture. Ann Surg. 2005 Jul; 242(1):1019.
38. Dumonceau JM. Systematic appraisal of the role of metallic
endobiliary stents in the treatment of benign bile duct stricture.
Ann Surg. 2006 Jul; 244(1):164165.
39. Baron TH, Poterucha JJ. Insertion and removal of covered
expandable metal stents for closure of complex biliary leaks. Clin
Gastroenterol Hepatol. 2006 Mar; 4(3):381386.
40. Pozsar J, Sahin P, Laszlo F, et al. Endoscopic treatment of
sphincterotomy-associated distal common bile duct strictures by
using sequential insertion of multiple plastic stents. Gastrointest
Endosc. 2005 Jul; 62(1):8591.
41. Degertekin M, Serruys PW, Foley DP. Persistent inhibition
of neointimal hyperplasia after sirolimus-eluting stent
implantation: long-term (up to 2 years) clinical, angiographic,
and intravascular ultrasound follow-up. Circulation 2002;
106:16101613.
333
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Biliary Surgery Complications
31 including Transplantation
Claudio Navarette, Eduardo Valdivieso and Jaquelina M. Gobelet
Endoscopic management of iatrogenic biliary complications contin-
ues to be technically challenging for the therapeutic endoscopist. In
this area, some progress has been reported and ERCP constitutes a
valuable tool.
This chapter deals with the endoscopic treatment of biliary leaks
after laparoscopic cholecystectomy (LC) or liver resection. Consider-
ation of ERCP in the treatment of complications following liver
transplantation is also described.
The treatment of late strictures will not be discussed in detail as
this particular topic is described in Chapter 30 of this textbook.
Currently, medical literature does not provide much information
from experimental trials, and a systematic approach to guide deci-
sions about the use of ERCP as a treatment for biliary surgery com-
plications has not been clearly established. Indications and
contraindications of ERCP in the treatment of a biliary injury are
based only on empirical evidence provided by highly experienced
groups. This chapter will provide a tool to determine when and how
endoscopic management could be the treatment of choice for an
individual patient with a complicated biliary surgery.
PHYSIOLOGICAL BASIS OF ERCP TECHNIQUES IN
TREATING BILIARY SURGERY COMPLICATIONS
ERCP is useful in the treatment of biliary surgery complications by
means of different physiological mechanisms:
Decreasing the pressure inside the biliary ducts
The rst aim of endoscopic therapy is to reduce sphincter of Oddi
tone. Endoscopic sphincterotomy (ES) can achieve this, favoring
transpapillary bile ow and controlling the drainage trough the leak
site. As ES usually does not cut the sphincter completely, a stent is
often used concomitantly, passing through the papilla in order to
bypass the muscular activity of the remnant sphincter (Fig. 31.1).
Bypassing bile ow trough the leak site
In addition to reducing intraductal pressure, the stent bypasses the
bile toward the papilla and reduces the ow through the leak site.
This fact, contributes indirectly to the closure of the leak (Fig. 31.2).
Considering the low morbidity associated with stents, we recom-
mend the combined use of ES and stenting to minimize intrabiliary
pressure.
Traditionally, plastic stents have been used for this purpose.
However, in a recent report, bioabsorbable self-expanding stenting
appears to be a promising option both because of larger diameter
and elimination of need for removal.1
Nasobiliary tubes can be an alternative to stenting in the
treatment of biliary leaks after laproscopic cholecystectomy (LC).
Advantages include precluding need for sphincterotomy, easy
removal, and interval cholangiograms to assess leak resolution.2
However, because of patient discomfort, we do not routinely
use nasobiliary tubes as an alternative to stents in the treatment of
biliary leaks.
Sealing the leak
In case of bilio-pleural stula, where negative pressure sucks bile
into the thoracic cavity, application of sealants is often necessary to
facilitate control of the abnormal ow. ERCP can be an expeditious
way to apply sealants to a leak as described by Seewald, who reported
his preliminary experience using endoscopically injected glue to
control stula output.3 Endoscopic cyanoacrylate glue occlusion is
an effective method to occlude the lumen of a disrupted duct. As
total lumenal closure is not feasible for lesions of major ducts, the
use of this resource is limited to peripheral injuries of the biliary
tree and may be potentially useful in peripheral pancreatic stulas.4
However, because of potential risk of pulmonary embolism, we have
used sealants to occlude biliary stulas only for highly selected
peripheral cases of bilio-pleural stulas, or bilio-peritoneal stulas
that have been refractory to combined treatment with ES and plastic
prostheses (Fig. 31.3).
Dilating the stricture and maintaining
intraductal ow
Benign biliary strictures require dilatation followed by multiple
stent placements and exchanges. This therapy, which is
described in Chapter 30, offers a minimally invasive alternative to
hepaticojejunostomy in the management of postoperative biliary
strictures.5
Biolm accumulation in the prosthesis and secondary loss of
patency require repeated exchanges. This constitutes a problem not
only because of infection risk related to bile ow obstruction, but
also because of patient discomfort and the morbidity related to repet-
itive procedures. In this setting, bio-absorbable biliary stents, as well
as self-expanding metal prostheses, have been recently described
with promising, but preliminary results.6
As noted above, there are several physiological mechanisms
that are potentially efcacious for treatment of a biliary injury by
means of an ERCP. Unfortunately, there are no randomized clinical
trials to dene the ideal mechanism or combined strategy for the
successful endoscopic resolution of an injury of the biliary tree. In
general, we suggest that pressure inside the biliary ducts should be
335
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 31.1 Endoscopic sphincterotomy and transpapillary stenting
in a patient with a tear of the common bile duct. Post-stenting, the
biliary drainage is improved and the pressure inside the biliary
ducts is reduced.
A B
Fig. 31.2 A Common bile duct tear in a 52-year-old woman after
a laparoscopic cholecystectomy. B Final position of the stent
bridging the leak site.
A B
Fig. 31.3 68-year-old man after segmental liver resection.
A Central biliary leakage treated with Hystoacryl injection.
B After emptying of contrast, radiopaque cyanoacrylate is noted to
occlude the injured duct.
336
BOX 31.1 PHYSIOLOGICAL BASIS FOR
ENDOSCOPIC MANAGEMENT OF BILIARY
SURGERY COMPLICATIONS
Decreasing pressure inside the biliary ducts
Bypassing bile ow through the leak site
Sealing the leak
Dilatation of a stricture
Maintaining bile ow
reduced as much as possible, using not only ES but also stent
placement. In selected cases of peripheral leak, endoscopic
cyanoacrylate application may be considered as adjunctive therapy
(Box 31.1).
ERCP FOR TREATMENT OF COMPLICATIONS
FOLLOWING LAPAROSCOPIC
CHOLECYSTECTOMY
Laparoscopic cholecystectomy has become the rst choice for the
treatment of symptomatic cholecystolithiasis. A different assortment
of surgical complications such as leaks from the cystic duct or inju-
ries to aberrant ducts have been described as surgical teams gain
experience and classical contraindications to LC have been progres-
sively eliminated.
Strasberg classied biliary injuries during LC according to
anatomical considerations and their relationship to treatment.7 We
recommend this classication as it identies lesions morphologi-
cally amenable to endoscopic management and it has been widely
adopted by surgical groups (Fig. 31.4).
The majority of biliary injuries are suitable for endoscopic man-
agement. They correspond to small tears and leaks that are easily
misidentied during surgery. Rarely, transections of the Common
Bile Duct (CBD) have been reported and are generally not amenable
to endoscope therapy.8
An evidence-based approach cannot be used to determine what
type of bile duct lesion is going to improve by means of an ERCP.
Currently, indications and contraindications for endoscopic proce-
dures are determined by experts, supported by factors such as:
nature and magnitude of the injury, ow trough the leak, timing of
the diagnosis postoperatively, presence of sterile or infected bilomas
and surgical risk associated with operative repair.
Nature and magnitude of the
biliary injury
Continuity of the injured bile duct is the most important factor
concerning the nature and magnitude of the injury and its relation
to endoscopic management.
If there is continuity of the injured bile duct (Fig. 31.4 types A,
C and D), ERCP is considered as primary therapy. Otherwise,
 Chapter 31 Biliary Surgery Complications including Transplantation

A B C D

>2cm <2 cm

E1 E2 E3 E4 E5

Fig. 31.4 Classication of biliary injury following cholecystectomy. Type A is a transection of small bile ducts that enter the liver bed or
the cystic duct. Type B and C almost always involve aberrant right hepatic ducts. Type D is a tear or burn of the CBD. Type E implies tran-
section or resection of the CBD. Type A, C, D and some E injuries frequently cause bilomas of stulas requiring external drainage. Type B
may or may not require subsequent treatment. (Redrawn with permission from Strasberg SM, Hertl M, Soper NJ. An analysis of the problem
of biliary injury during laparoscopic cholecystectomy. J Am Coll Surg 1995 Jan; 180(1):10125 1995 The American College of
Surgeons.)

endoscopic therapy is generally precluded for injuries with complete
bile duct transection and the presence of clips at the distal stump or
lack of continuity between segments (Fig. 31.4 type E).9,10
If the CBD has been resected, surgery is always necessary to
establish the lost continuity of the duct. ERCP is helpful only to
determine the type and extent of the injury and a complementary
magnetic resonance cholangiogram is usually required to dene
proximal biliary anatomy.
In case of complete transection without resection of the CBD,
surgical reconstruction is also almost always required. Even though,
we have successfully treated some patients with complete transec-
tion of the CBD by means of ERCP without the use of a simultane-
ous percutaneous transhepatic approach, after successful passage of
a wire guide, these cases should be considered anecdotal. Complete
transection of a major bile duct is a strong indication for operative
repair. Only if it is possible to adequately traverse the lesion and
decompress the entire biliary system with stents, should exclusive
endoscopic therapy be considered.11
Aberrant ducts usually drain bile from areas of the liver directly
to the gallbladder or the CBD and they uncommonly have commu-
nication to the left or right biliary system. If an aberrant damaged
duct is misidentied during surgery (Fig. 31.4 type C), clinical sus-
picion of an iatrogenic injury will often require an ERCP. If the
injured duct is visualized, it means that communication with the
biliary system exists. In this case, an ERCP could be therapeutic
if procedures such as ES, stenting and, occasionally, glue injection
are undertaken. If the injury is not detected, other studies such
as scintigraphy or magnetic resonance cholangiogram are required
and resective surgery is mandatory.
Flow through the leak
Clinically signicant bile leak is a potentially serious complication
that occurs after 0.10.5% of conventional open cholecystectomy.
Either because of intrinsic risk or increased tendency to report com-
plications, the shift to LC has been associated with up to a 2% inci-
dence of bile leak. Given the large number of LC performed, the
incidence of postoperative bile leaks is substantial.
The most common site of leakage is the cystic duct (78%), fol-
lowed by peripheral right hepatic ducts of Luschka (13%) and other
sites (9%) including the common hepatic duct, CBD and T-tube
insertion points.12
High output leaks were traditionally considered as an indication
for surgery. However, more and more successful endoscopic
treatments have been reported (Table 31.1). Shanda et al. dened a
low-grade leak as a leak identied only after opacication of the
intrahepatic biliary radicals with contrast and a high-grade leak as
a leak observed uoroscopically before intrahepatic opacication.12
As ERCP is generally used for diagnosis when a leak is suspected,
endoscopic therapy should be attempted even for high-grade ductal
leaks.
Time to diagnosis
If a patient presents with a bile duct injury early or late after
surgery, the treatment of choice is an endoscopic procedure. If
the injury is identied during surgery, the complication should be
solved during the same procedure that caused the iatrogenic
lesion.

337
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Series n Stones ES Stent ES&Stent NT Efcacy Factor ECRP Surgery

Kozarek 11 18% 2 7 1 82% Injury Cystic duct (Strasberg A) X
Foutch 23 30% 4 6 12 1 100% Luschka duct (Strasberg A) X
Barkum 52 22% 27 1 27 8 88% Ligated sectorial duct Xa
Ryan 50 22% 6 13 31 88% (Strasberg B)
Davids 48 31% 20 25 3 90% Non-ligated sectorial duct Xb X
Prat 26 31% 15 3 8 70% (Strasberg C)
Himal 12 6 6 100% CBD small; tearing/burning X
DePalma 64 33% 25 18 21 96.9% (Strasberg D)
Chow 19 16% 19 19 95% CBD transection (Strasberg E) Xc X
Shanda 204 20% 75 99% CBD resection (Strasberg E) X
Flow through Low-grade leak X
Table 31.1 Published series using an endoscopic approach to the leak High-grade leak Xd X
treat bile leaks
n = number of patients; ES = endoscopic sphincterotomy; NT = nasobiliary tube.
Timing Intraoperative X
Early postoperative X
Late postoperative (strictures) Xd X
Associated No associated or small collection X
biloma Aseptic no septated collection X
However, biliary endoprostheses can be used to facilitate a Infected or septated collection Xe X
correct cicatrisation of a sutured bile duct and as an alternative Surgical risk High-risk X
method of CBD drainage after a choledochorraphy.13 Non-high-risk Xd X

Collections and related septic
complications
The rst step in the treatment of iatrogenic biliary injury is sepsis
control and patient stabilization. Septic complications are mainly
associated with loculated collections and their drainage is
mandatory.
Currently, most collections and bilomas are drained percutane-
ously by placing an indwelling catheter, and systemic antibiotics are
administered. If percutaneous drainage is considered insufcient,
minimally invasive techniques can be used to laparoscopically
resolve the collection.
It is important to note that the role of ERCP in the modern mini-
mally invasive treatment of bilomas and septic collections related to
biliary surgery is to diagnose and control the bile leak, while sepsis
control must be the initial treatment.
Surgical risk
Signicant surgical risk has historically been considered to be a
contraindication for conventional surgery. As a consequence, high-
risk patients have helped engender the initial development of mini-
mally invasive procedures. ERCP has not been the exception, but
currently, enough evidence and experience support the safety and
effectiveness of surgical endoscopy to recommend it for almost all
patients in whom conventional surgery was historically considered
the only alternative.
Considering this complex scenario, iatrogenic lesions of the
biliary tree represent serious injuries with common complications,
high costs and uncertain results at late follow-up. The management
of these patients is a difcult medical problem which requires the
cooperation and skill-set of hepatobiliary surgeons, interventional
radiologists and biliary endoscopists.
It is desirable but impossible to establish a rigorous evidence-
based list of indications and contraindications for ERCP in the treat-
ment of LC complications. Instead, we propose a useful guide to
identify the most appropriate clinical setting that supports the prin-
ciples of endoscopic therapy and surgery in the management of iat-
rogenic injuries of the biliary tree. (Table 31.2).
Table 31.2 Factors inuencing endoscopic vs. surgical treatment
of biliary tree injuries
a
Absent communication to CBD branches leads to late liver resection.
b
Only if aberrant system communicates with CBD branches.
c
Stenting only if continuity can be established.
d
ERCP seems as effective as surgery; other factors should be considered.
e
Drainage is priority, post-drainage, endoscopic therapy should be used.
BOX 31.2 CONTRAINDICATIONS FOR
ENDOSCOPIC TREATMENT OF
COMPLICATIONS FOLLOWING LC
Injuries identied during surgery
Injuries without continuity to the CBD
1. Strasberg B
2. Strasberg C
In summary, most patients are candidates for a therapeutic ERCP
if a complication from LC is suspected. As shown by our proposed
algorithm (Fig. 31.5), ERCP is a widely accepted therapy for high
output stulas and low risk patients. Only type B and E lesions
without continuity to the CBD, and injuries identied during
surgery limit ERCP to adjuvant therapy to the classical surgical
repair (Box 31.2).
ERCP AND BILIARY LEAKS AFTER
HEPATIC RESECTION
The true incidence of bile leaks after hepatic resection is unknown.
Many series report an incidence approximating 11% but most
cases correspond to minor bile leaks that often seal spontaneously.
Major bile leaks are often central in origin and occur as a

338
 Chapter 31 Biliary Surgery Complications including Transplantation

Fig. 31.5 ERCP for treatment of complica-

Biliary complication tions after LC.
after LC * In selected cases of septic collections and/
or biloma, open surgery is indicated as
primary treatment.
** Consider open surgery only if Lapara-
scopic management is not feasible.
Identified during Misidentified during LC = Laparoscopic Cholecystectomy.
surgery surgery

**

* Septic No collection
collection/biloma

Percutaneous/
laparoscopic drainage

Failure Success

Therapeutic ERCP

Open surgery Failure Success

Follow up

consequence of failure to ensure the integrity of the bile ducts at the

liver hilum (Fig. 31.6). B
A
Coagulopathy is common in patients following hepatic resection.
For this reason ES is sometimes avoided and cyanoacrylate is occa-
sionally used.14 For correct occlusion with sealants, the duct must
be selectively visualized by ERCP and its distal part embolized with
approximately 1.5 ml of cyanoacrylate. This procedure usually effec-
tively stops further bile leakage (Fig. 31.3). Although supported only
by case reports, selective embolization is a useful adjuvant therapy
to stenting procedures and its use as an exclusive treatment for
stopping major bile leaks has also been described.
C D
THE RETAINED COMMON BILE STONE
A retained bile duct stone after cholecystectomy is present in up to
2.5% of LC. The effectiveness of ERCP to treat CBD stones has
improved due to the increased experience of endoscopic surgeons,
and the introduction of new lithotripsy devices which allow the
extraction of giant and complicated stones (Fig. 31.7). To support
the use of ERCP for retained CBD stones, Anward conducted a ret-
rospective study of patients who presented with symptomatic
retained stones after LC.15 The study included 25 patients who Fig. 31.6 65-year-old male patient with a biliobronchial stula after
returned to the hospital with symptoms and signs suggestive of CBD hydatid cyst resection treated with ES and stenting. A Biliary
leakage from a left liver duct. B Leakage draining to an organized
stones. All 25 patients underwent ERCP and there was successful biliobronchial stula. C Wire placed into the injured duct.
removal of stones in 16 cases, failure in ve cases, and no stones in D Final position of the stent.

339
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
Fig. 31.7 A Giant choledocholithiasis. B Multiple common bile
duct stones.
four cases. A second ERCP was successful in removing retained
stones in four of the ve failed patients. For most patients with bile
duct stones, ES continues to be the treatment of choice to facilitate
extraction. We have previously described 8204 patients treated in
three surgical endoscopy centers (Chile, Germany, and India); 86
91% of all CBD stones could be extracted subsequently after ES
using a Dormia basket, 47% required mechanical lithotripsy before
removal, and 310% of the patients required different techniques,
such as electrohydraulic lithotripsy, laser-induced shock-wave litho-
tripsy, or extracorporeal shock-wave lithotripsy. We concluded that
endoscopic techniques can resolve the vast majority of retained
stones and that less than 1% of all patients with bile duct stones still
require surgical intervention.16
In summary, ERCP is an effective technique for diagnosis and
treatment of retained post-LC stones. Its morbidity is acceptably low,
and should be considered as an alternative to surgical CBD explora-
tion even in cases in which choledocholithiasis is diagnosed
intraoperatively.
BILIARY COMPLICATIONS ASSOCIATED WITH
KEHRS TUBE
Historically, postoperative T-tube drainage has been used by biliary
surgeons to prevent stasis, to decompress the biliary tree, to mini-
mize the risk of leakage through a choledochotomy, and to prevent
strictures with cicatrisation of a sutured choledochus. T-tubes also
offer easy percutaneous access for cholangiography and extraction
of retained stones, but this resource is only useful 4 or more weeks
following surgery, when a stulous track has been established.
Despite these advantages, T-tube morbidity rates as high as 16%
have been reported.17 Accidental T-tube displacement leading to
CBD obstruction, bile leakage, duodenal erosion, persistent biliary
stula and excoriation of the skin, dehydration, saline depletion,
cholangitis and CBD stenosis have been reported as postoperative
complications secondary to T-tube use (Fig. 31.8). Such complica-
tions and the demonstrated benet of minimally invasive methods
have been associated with a tendency among laparoscopic surgeons
to avoid T-tube usage. Both laparoscopic endobiliary stent placement
under direct or uoroscopic guidance with primary closure of the
common bile duct (CBD), and primary closure of the CBD without
drainage, have been proposed as safe and effective alternatives to T-
tube placement. Isla prospectively collected 53 consecutive patients
who underwent laparoscopic CBD exploration through a choledo-
chotomy and compared the results of a group treated with T-tube
placement versus a group treated with biliary stent placement and
primary CBD closure. His results suggest that the use of a biliary
340
Fig. 31.8 Displaced T-tube.
endoprosthesis leads to lower morbidity, shorter hospital stay, less
postoperative discomfort and earlier return to full activities.13
If a T-tube is placed, a postoperative cholangiogram will show the
presence of residual retained stones, but it is always necessary to
wait at least four weeks before percutaneous extraction or tube
withdrawal.
If accidental tube dislodgement occurs, we recommend attempt-
ing guidewire passage through the early stulous track in order to
reach the CBD and place a new tube. If this is not possible, an ES
with stent placement will be necessary for treatment of the remnant
bile leak.
Otherwise, if an endoprosthesis has been placed as an alternative
to a T-tube, early ERCP is possible, and with the use of a small
catheter passed through the stent, a cholangiogram is readily
obtained. If stones are not found the prosthesis can be removed. In
the case of retained stones, a precut ES is performed over the pros-
thesis which minimizes pancreatic duct injury and leads to an inher-
ently safer stone extraction.
Even though prosthesis insertion during surgery mandates a
postoperative ERCP, we prefer this alternative because of our ability
to remove retained stones more safely and the poor quality of life
related to the T-tube.
Wani studied the use of an endonasobiliary drainage tube as an
alternative to the T-tube for postoperative, intraductal drainage. His
paper reviewed 20 patients who underwent closure of the common
duct over an endonasobiliary tube without any difculty.18 None of
the patients experienced biliary complications such as bile leak,
biliary peritonitis, biliary stula, pancreatitis or cholangitis. The
tubes were removed between days 6 and 8 and the length of the
postoperative hospital stay varied from 7 to 15 days. We do not use
this alternative because of possible patient discomfort and need for
prolonged hospitalization.
SUMP SYNDROME
Even though side-to-side choledochoduodenostomy is a commonly
performed operation, postoperative biliary sump syndrome is
an unusual complication. Caroli-Bosc analyzed 30 cases of sump

Fig. 31.9 Sump syndrome. Food debris and calculi in the distal
choledochus; (arrow) note Dormias basket passed through a
choledochoduodenostomy.
syndrome and described the clinical presentation and the outcome
of endoscopic sphincterotomy.19 The median clinical latency was ve
years and the median delay between surgery and diagnosis six years.
Abdominal pain with fever was the main symptom and hepatic
abscesses, and acute pancreatitis were also noted. Liver function
tests were abnormal in 79% of cases.
Endoscopic sphincterotomy appeared to be a safe, reliable treat-
ment and symptom recurrence post-ES has not been reported. Fol-
lowing an ES, food debris and calculi can be removed from the
biliary sump (Fig. 31.9).
As a consequence of debris and stones impacted within the sump,
catheterization of the papilla is sometimes difcult. In those
instances, a wire should be advanced through the choledochoduode-
nostomy to reach the sump and cross the papilla in an antegrade
way. Once the wire is placed in the duodenal lumen, the duodeno-
scope is withdrawn and the wire is pushed at the same time to
release it from the endoscopes channel. With the wire entering
through the patients mouth and crossing the choledochocholedos-
tomy and the papilla, the duodenoscope is introduced again with a
snare in order to grip the guidewire and pull it out through the
accessory channel of the duodenoscope. Finally, the papillotome is
inserted endoscopically over the guidewire allowing ES in almost all
cases. Alternatively, the wire can be grasped with a polyp snare and
pulled through the endoscope channel, a so-called internal
rendezvous procedure.
POST-CHOLECYSTECTOMY SYNDROME
Endoscopy plays an important role in diagnosis and treatment of the
post-cholecystectomy syndrome (PCS). In our experience, the vast
majority of PCS can be attributed to organic causes and can be
effectively treated. This experience is mirrored in a retrospective
Chapter 31 Biliary Surgery Complications including Transplantation
Fig. 31.10 Dilatation of anastomotic stricture through a sub-
cutaneously placed afferent limb in a patient with choledocho-
jejunostomy.
study, in which Zhou described ERCP in the management of 371
patients with PCS.20
ERCP is a useful tool for both diagnosis and treatment of PCS.
Patients found to have choledocholithiasis can be subjected to
endoscopic stone extraction, those with papillary inammatory
stricture or sphincter of Oddi dysfunction to ES, and those
with papillary or hepatobiliary tumor to endoscopic biliary endo-
prosthesis placement.
DILATATION OF BILIARY STRICTURES THROUGH
A SUBCUTANEOUSLY PLACED AFFERENT LIMB IN
PATIENTS WITH A CHOLEDOCHOJEJUNOSTOMY
Until the development of minimally invasive procedures to access
hepaticojejunostomy anastomoses, reoperation was needed for anas-
tomotic strictures or retained stones. Endoscopic techniques have
been described which can replace the radiological approach using a
combined technique that includes endoscopy to reach the stricture
both to guide balloon dilatations but also to allow directed stricture
biopsy, strictureplasty, and injection of depot corticosteroids such as
triamcinolone.
Using local anesthesia, a small incision is performed over the
subcutaneous limb and a minimally invasive access to the lumen is
established. With a front-viewing endoscope, the hepaticojejunos-
tomy is easily reached through a subcutaneously placed afferent
limb. If malignancy is suspected, biopsies can be taken to conrm
the diagnosis or in case of a benign stricture, different sized balloons
can be used to effect dilatation. By means of a precut ES, several
radial cuts can facilitate dilatation (Fig. 31.10).
Local injection of depot corticosteroids into benign esophageal
strictures has been proven to maintain the effects of bougienage or
balloon dilatation. This resource has not been proven for biliary
strictures and a single pilot trial which included eight patients with
CBD strictures has been published.21
Given the absence of strong evidence demonstrating efcacy, we
do not use corticosteroids for the treatment of CBD strictures as an
adjuvant to multiple stent therapy. However, we have used injection
of two 10 mg doses of triamcinolone into the stricture site with a
341
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

sclerotherapy needle to improve late results of balloon dilatation and
strictureplasty when treating selected benign anastomotic stenoses
managed endoscopically through a subcutaneous limb.
SPECIAL CONSIDERATIONS FOR LIVER
TRANSPLANTATION
Biliary complications are important causes of early and late postop-
erative morbidity and mortality following orthotopic liver transplan-
tation (OLT) and are present in up to 20% of the patients.22 The most
common complications in OLT are strictures, bile leaks, stones and
sphincter of Oddi dysfunction (Table 31.3).
Reconstruction following OLT has an important relationship to
the treatment of biliary complications. The continuity of donor bile
ducts with the intestinal tract is performed as either an end-to-end
choledochocholedochostomy or as a Roux-en-Y reconstruction with
an end-to-side choledochojejunostomy. Choledochocholedochos-
tomy anastomosis is the technique of choice because it is simpler,
has a lower complication rate, preserves sphincter of Oddi function
and allows retrograde evaluation by means of an ERCP. Roux-en-Y
reconstruction with an end-to-side choledochojejunostomy is the
reconstruction of choice in children, and in adults with biliary duct
conditions such as sclerosing cholangitis, or in cases of different bile
duct diameters between the donor and recipient. As previously
described, a subcutaneously placed afferent limb of a choledochoje-
junostomy can be used to perform a minimally invasive access to
the transplanted biliary ducts and recent data suggest that double-
balloon enteroscopes or Shape-LockTM overtubes facilitate ERCP in
patients with Roux anatomy.
The clinical presentation of post-liver transplant bile duct com-
plications is often subtle. The classical right upper quadrant abdomi-
nal pain is absent due to hepatic denervation and the only clue for
Bile leaksa Anastomotic siteb
EARLY COMPLICATIONS
the diagnosis is commonly an asymptomatic increase in the baseline
serum transaminases or bilirubin levels.
Even in the presence of severe obstruction, bile duct dilatation
may not be present early in the transplanted liver. Moreover,
common, non-invasive examinations often lack the sensitivity to
detect mild but clinically signicant sources of biliary obstruction.
If there is a clinical suspicion of biliary obstruction or leak fol-
lowing OLT, an early cholangiogram is needed for denitive diagno-
sis. The use of T-tube splinting of the anastomosis allows a prompt
evaluation of the biliary anatomy during the early postoperative
period, but has been associated to a higher rate of complications
related to biliary sludge, migration of the tube, and a higher inci-
dence of bile leak.23 Moreover, there is an increased tendency to
avoid T-tubes, not only because of complications, but also because
it doesnt help in the treatment of identied abnormalities, found in
up to 80% of suspicious cases. ERCPs have been reported to be
normal in only 15% of symptomatic liver transplant patients.24
The early recognition and prompt treatment of most biliary com-
plications after OLT, has resulted in a clear recognition that endo-
therapy is an effective tool to prevent unnecessary surgical re
interventions and improve long-term graft and patient survival. The
use of endoscopy for biliary complications after OLT is inuenced
by the type of biliary reconstruction, the use of T tubes and the
availability of subcutaneously placed limbs (Fig. 31.11).
Bile leaks and stulas associated with liver
transplantation
OTL has been associated with numerous leakage sites, including
the anastomosis, cystic duct, the T tube tract, and accessory con-
duits. Endoscopic management of bile leaks after transplantation
follows the same principles described for the treatment of bile
leaks secondary to LC or biliary surgery. As a general rule, drainage
of a related uid collection, ES, and stenting must be used for
all cases.
It is important to note that an early nonanastomotic leak usually
means vascular insufciency, and therefore thrombosis of the
hepatic artery must be ruled out. Furthermore, in live-donor trans-

Cystic duct plantation, leaks from the cut surface of the liver are common, but
Accessory bile ducts most cases close spontaneously and rarely need endoscopic
T tube tract
treatment.
Incidental intrahepatic injury
If a T Tube cholangiogram identies a minor leak, it can be
Cut surface of the liverc
Migrated T-tube conservatively managed leaving the tube in place. Only refractory
cases need endoscopic treatment (Fig. 31.11).
Early strictures Mismatched bile duct diameters
Technical errors
Late strictures Anastomotic Strictures associated with liver
Nonanastomotic
transplantation
Cholangitis
LATE COMPLICATIONS

Strictures are the most common complication associated with OTL.

Filling defects Choledocholithiasis They can be divided into early (within 60 days of transplantation),
Sludge subacute (between 60 days and 1 year of transplantation) and late.
Biliary cast syndrome Early strictures are mainly associated with technical errors at the
Ampullary obstruction Spinchter of Oddi dysfunction anastomosis, while subacute and late strictures are often the conse-
Stenosis quence of vascular insufciency. This classication is signicant
Recurrent biliary disease Recurrent sclerosing cholangitis because if transient narrowing in a duct-to-duct connection appears
Recurrent malignant neoplasms within the rst 3060 days after transplantation, it is likely a conse-
quence of postoperative edema and inammation. These patients
Table 31.3 Biliary complications after liver transplantation respond very well to balloon dilatation and temporary stent place-
a
Septic related collections must be appropriately managed.
b
Thrombosis of the hepatic artery should be ruled out. ment (3 months). Repeated procedures are required uncommonly.22
c
Only in live-donor liver transplantation. Anastomotic strictures identied during the rst postoperative year

342
 Chapter 31 Biliary Surgery Complications including Transplantation

Biliary complication
after OLT

Choledochocholedochostomy Choledochojejunostomy

T tube in situ No T tube in situ Subcutaneous limb No subcutaneous limb*

T tube Endoscopic Percutaneous

ERCP
cholangiogram treatment procedure

Normal Abnormal**

Failure Success Success Failure Success Failure

Open Open Open

surgery surgery surgery

Follow up

Fig. 31.11 Role of endoscopy in the treatment of biliary complications after orthotopic liver transplant (OLT).
* Endoscopic therapy can be attempted when trans-oral access to hepaticojejunostomy is feasible.
**Abnormal T tube cholangiogram usually leads to ERCP. Minor leaks can be conservatively managed by leaving the tube in place.

show excellent response to short-term stent placement (36 months),
but these patients require long-term and repeated endoscopic treat-
ment as strictures may recur years after the treatment.
Late strictures usually respond well to initial balloon dilatation
and temporary stent placement (3 months), but relapse rates can
reach 40% leading to a more complex treatment which includes
repetitive balloon dilatation and long-term (1224 months) therapy
using multiple (up to three) large diameter stents.
Except when the stricture compromises the hilum, nonanasto-
motic strictures are difcult to distinguish on the basis of cholangi-
ography. A more difcult endoscopic treatment should be expected
as nonanastomotic strictures are multiple and ischemic in nature.
In fact, up to 50% of patients with nonanastomotic strictures require
a new transplant or die despite multiple attempts using balloon dila-
tation, debris removal, and stenting.
It is important to note that any ischemic injury that involves a
large portion of intrahepatic bile ducts is associated with poor graft
survival and is best managed by repeated transplantation.22
Filling defects and biliary cast syndrome
The differential diagnosis of lling defects after OLT includes stones,
blood clots, debris, sludge, and casts. Filling defects can appear as
an isolated nding, but frequently stones, sludge and casts manage-
ment is complementary to stricture therapy. An unusual form of
massive cast lling of the proximal bile ducts is the biliary cast syn-
drome. This incompletely explained disease has been associated
with ischemic events and is often accompanied by strictures which
can include diffuse stricturing of the hilum.
ERCP plays an important role in the treatment of this entity.
Baron et al. reported a severe case which was endoscopically treated,
suggesting that even when massive compromise is present, com-
bined therapy of stricture treatment and biliary toilette can yield
good results.25 (Fig. 31.12)
Endoscopic management of recurrent biliary
disease after liver transplantation
Sclerosing cholangitis is a benign disease that can recur in a trans-
planted liver. In fact, there is a high risk of anastomotic stricture
after OLT for this entity. As Roux-en-Y reconstruction with an end-
to-side choledochojejunostomy is the reconstruction of choice in
case of sclerosing cholangitis, for the most part, an endoscopic
approach can be done only if an afferent limb has been placed sub-
cutaneously. With this anatomy, balloon dilatation, strictureplasty,

343
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
B
A
C D
Fig. 31.12 Biliary cast syndrome: Successful endoscopic treatment.
A A large lling defect was found in the common hepatic duct
proximal to the anastomosis with extension into the intrahepatic
system. B Endoscopic biliary sphincterotomy was performed fol-
lowed by balloon retrieval of a large biliary cast. C Immediate
post extraction cholangiogram was markedly improved. D The
cholangitis resolved, but the patient developed recurrent biliary
obstruction 3 months later. ERCP demonstrated nonanastomotic
narrowing at the bifurcation requiring repeated endoscopic and
percutaneous therapy. (Reproduced with permission from Baron
TH, Yates RM, III, Morgan DE, et al. Biliary cast syndrome: successful
endoscopic treatment. Gastrointest Endosc 2000 Jul; 52(1):7879.)
stenting and even injection of depot glucocorticoids can be under-
taken as described previously (Fig. 31.13).
Liver transplantation for malignancies is still a controversial issue.
Many centers worldwide perform OLT for hepatocellular carcinoma
associated with liver cirrhosis, liver metastases from neuroendocrine
tumors and epithelioid hemangio-endothelioma. This setting is
potentially associated with bile duct obstruction due to local recur-
rence. Endotherapy is a well-recognized entity for the palliation of
malignancies that obstruct bile ow and these techniques, to include
REFERENCES
1. Ginsberg G, Cope C, Shah J, et al. In vivo evaluation of a new
bioabsorbable self-expanding biliary stent. Gastrointest Endosc
2003 Nov; 58(5):777784.
2. Elmi F, Silverman WB. Nasobiliary tube management of
postcholecystectomy bile leaks. J Clin Gastroenterol 2005 May;
39(5):441444.
344
Fig. 31.13 Strictureplasty of an hepaticojejunostomy undertaken
through a subcutaneously placed limb. Note mucosectomy cap
on tip of the endoscope.
self-expandable metal stent (SEMS) placement, can be expeditiously
applied in the treatment of malignant obstruction after OLT.
In view of the minimal morbidity and the proven efcacy, we
believe that for biliary complications after OLT, endoscopic manage-
ment should be considered as initial management before either
percutaneous or surgical treatment. If endoscopic therapy fails, if
patients have multiple intrahepatic strictures, or if a Roux-en-Y
hepaticojejunostomy has been done without a subcutaneously placed
afferent limb, re-operation should be considered.
SUMMARY
The morbidity of endoscopic treatment of biliary surgery
complications is well documented. However, these have
approximated 1.5% in one large series, an incidence lower
than operative morbidity. Complications are mainly composed
of post-ERCP pancreatitis and duodenal perforations which
can generally be managed by conservative therapy; surgery is
rarely required. Mortality has not been reported, in the largest
series to date.12
As mentioned throughout the chapter, the safety and effec-
tiveness of endoscopy in treating the complications of biliary
surgery and liver transplantation have become rmly estab-
lished. These minimally invasive techniques have complemented
and supplemented classic surgery, but with lower morbidity,
mortality and metabolic impact.
3. Seewald S, Groth S, Sriram PV, et al. Endoscopic treatment of
biliary leakage with n-butyl-2 cyanoacrylate. Gastrointest Endosc
2002 Dec; 56(6):916919.
4. Seewald S, Brand B, Groth S, et al. Endoscopic sealing of
pancreatic stula by using N-butyl-2-cyanoacrylate. Gastrointest
Endosc 2004 Apr; 59(4):463470.

5. Fogel EL, Sherman S, Park SH, et al. Therapeutic biliary
endoscopy. Endoscopy 2003 Feb; 35(2):156163.
6. van Berkel AM, Cahen DL, van Westerloo DJ, et al. Self-expanding
metal stents in benign biliary strictures due to chronic pancreatitis.
Endoscopy 2004 May; 36(5):381384.
7. Strasberg SM, Hertl M, Soper NJ. An analysis of the problem of
biliary injury during laparoscopic cholecystectomy. J Am Coll Surg
1995 Jan; 180(1):101125.
8. Csendes A, Navarrete C, Burdiles P, et al. Treatment of common
bile duct injuries during laparoscopic cholecystectomy:
endoscopic and surgical management. World J Surg 2001 Oct;
25(10):13461351.
9. Parlak E, Cicek B, Disibeyaz S, et al. Treatment of biliary leakages
after cholecystectomy and importance of stricture development
in the main bile duct injury. Turk J Gastroenterol 2005 Mar;
16(1):2128.
10. Kaffes AJ, Hourigan L, De LN, Impact of endoscopic intervention
in 100 patients with suspected postcholecystectomy bile leak.
Gastrointest Endosc 2005 Feb; 61(2):269275.
11. Baron TH, Feitoza AB, Nagorney DM. Successful endoscopic
treatment of a complete transection of the bile duct complicating
laparoscopic cholecystectomy. Gastrointest Endosc 2003 May;
57(6):765769.
12. Sandha GS, Bourke MJ, Haber GB, et al. Endoscopic therapy for
bile leak based on a new classication: results in 207 patients.
Gastrointest Endosc 2004 Oct; 60(4):567574.
13. Isla AM, Griniatsos J, Karvounis E, et al. Advantages of laparoscopic
stented choledochorrhaphy over T-tube placement. Br J Surg
2004 Jul; 91(7):862866.
14. Kiltz U, Baier J, Adamek RJ. [Selective embolization of a bile leak
after operative resection of an echinococcal cyst]. Dtsch Med
Wochenschr 1999 May 28; 124(21):650652.
15. Anwar S, Rahim R, Agwunobi A, et al. The role of ERCP
in management of retained bile duct stones after
Chapter 31 Biliary Surgery Complications including Transplantation
laparoscopic cholecystectomy. N Z Med J 2004 Oct 8; 117(1203):
U1102.
16. Seitz U, Bapaye A, Bohnacker S, et al. Advances in therapeutic
endoscopic treatment of common bile duct stones. World J Surg
1998 Nov; 22(11):11331144.
17. Martin IJ, Bailey IS, Rhodes M, et al. Towards T-tube free
laparoscopic bile duct exploration: a methodologic evolution
during 300 consecutive procedures. Ann Surg 1998 Jul;
228(1):2934.
18. Wani MA, Chowdri NA, Naqash SH, et al. Primary closure of the
common duct over endonasobiliary drainage tubes. World J Surg
2005 Jul; 29(7):865868.
19. Caroli-Bosc FX, Demarquay JF, Peten EP, et al. Endoscopic
management of sump syndrome after
choledochoduodenostomy: retrospective analysis of 30 cases.
Gastrointest Endosc 2000 Feb; 51(2):180183.
20. Zhou PH, Liu FL, Yao LQ, et al. Endoscopic diagnosis and
treatment of post-cholecystectomy syndrome. Hepatobiliary
Pancreat Dis Int 2003 Feb; 2(1):117120.
21. Wehrmann T, Schmitt T, Caspary WF, et al. [Local injection of
depot corticosteroids in endoscopic therapy of benign bile duct
strictures]. Z Gastroenterol 2000 Mar; 38(3):235241.
22. Thuluvath PJ, Pfau PR, Kimmey MB, et al. Biliary complications
after liver transplantation: the role of endoscopy. Endoscopy
2005 Sep; 37(9):857863.
23. Qian YB, Liu CL, Lo CM, et al. Risk factors for biliary
complications after liver transplantation. Arch Surg 2004 Oct;
139(10):11011105.
24. Thuluvath PJ, Atassi T, Lee J. An endoscopic approach to biliary
complications following orthotopic liver transplantation. Liver Int
2003 June; 23(3):156162.
25. Baron TH, Yates RM, III, Morgan DE, et al. Biliary cast syndrome:
successful endoscopic treatment. Gastrointest Endosc 2000 Jul;
52(1):7879.
345
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
ERCP for the Acute and Chronic
32 Complications of Pancreatic Surgery
James J. Farrell and David L. Carr-Locke
Types of pancreatic surgery performed for benign and malignant
pancreatic diseases include pancreaticoduodenectomy (Whipple,
classic and pylorus preserving), distal pancreatectomy (tail resec-
tion), lateral pancreaticojejunostomy (Puestow and related proce-
dures), and duodenum preserving local resection of the pancreatic
head (Beger and Frey procedures).16 This review will focus predomi-
nantly on the role of ERCP in patients who have undergone pan-
creaticoduodenectomy. The role of ERCP following other types of
surgery are discussed in detail in Chapter 24.
Although most often performed for malignant disease (including
pancreatic cancer, ampullary cancer and duodenal cancer), pancre-
aticoduodenectomy is increasingly being performed for manage-
ment of benign disease (e.g. chronic pancreatitis, premalignant
pancreatic cystic neoplasms) (Table 32.1).7 For example, in one
study, there was an increasing proportion of patients undergoing
pancreaticoduodenectomy for intraductal papillary mucinous
neoplasia (IPMN) compared with pancreatic ductal adenocarcin-
oma over a 10 year period.8 Due to this shift in indications and the
improved postoperative survival of patients with cancer undergoing
this operation, a greater proportion of these patients are surviving
longer. Consequently, there has been an increase in both short-term
and long-term postoperative complications requiring intervention.
Distal pancreatectomy is performed for resection of both malig-
nant and premalignant pancreatic diseases. It can be complicated by
the development of pancreatic duct leak, which is most often treated
conservatively by external management of the leak. However, pan-
creatic duct stenting may help in healing pancreatic duct leaks that
fail medical treatment. Prophylatic placement of pancreatic duct
stent to prevent development of a postoperative pancreatic duct leak
after a distal pancreatectomy has also been suggested, but is not
practiced widely (Fig. 32.1).9
Lateral pancreaticojejunostomy (Puestow procedure) is per-
formed for management of symptomatic chronic pancreatitis in
setting a dilated main pancreatic duct. Endoscopists are frequently
asked to perform ERCP to assess patency of the pancreaticojejunal
anastomosis when symptoms recur after surgery. If the pancreati-
cojejunal anastomosis is found to be stenosed, endoscopic dilatation
and stent placement can be performed (Fig. 32.2).
ROLE OF ERCP IN PREVENTING POST
PANCREATIC SURGERY COMPLICATIONS
Specic avoidance of preoperative ERCP may be instrumental in
preventing postoperative pancreatic surgical complications. Because
obstructive jaundice can cause defects in hepatic, renal, and
immune function, it was long believed that preoperative relief of
biliary obstruction would correct these defects and decrease post-
operative morbidity and mortality rates. However, several random-
ized, controlled studies have proven that routine preoperative bile
duct stent placement does not improve postoperative outcomes
and, in fact, may increase postoperative infectious complications
(Table 32.2).1019 Though it is recommended that when operative
intervention is delayed by several weeks (including in the setting
of neoadjuvant therapy), or if cholangitis or pruritus occurs, a
biliary stent (at least 10 French diameter) should be placed endo-
scopically, recent data suggests that placement of short-length
expandable metal biliary stents does not interfere with pancreatico-
duodenectomy and decreases the rate of cholangitis as compared
to plastic biliary stents in patients receiving preoperative neo-
adjuvant therapy.20,21 In addition, patients found to be unresectable
at the time of surgery do not require biliary bypass.22
Preoperative pancreatic duct stent placement prior to distal pan-
createctomy has been shown in a small study to reduce the incidence
of postoperative pancreatic duct leaks.9
SHORT-TERM COMPLICATIONS ASSOCIATED
WITH PANCREATIC SURGERY
Short-term complications following pancreaticoduodenectomy
include anastomotic leakage (either a bile leak or a pancreatic
stula), hemorrhage, uid or abscess collection, afferent limb
obstruction, and delayed gastric emptying. Magnetic resonance chol-
angiopancreatography (MRCP) is becoming increasingly important
in the diagnostic evaluation of patients with suspected bile or pan-
creatic duct leaks. Anastomotic pancreaticobiliary leaks may require
intervention, including ERCP therapy. However, the potential risks
associated with endoscopy in the acute post surgical setting have led
to the use of interventional radiologic techniques for the manage-
ment of these post surgical complications. This applies especially
to biliary complications including abscess formation, bilomas,
bile leaks and biliary obstruction.23,24 In a recent study, up to 44% of
patient who had undergone a pancreaticoduodenectomy required
interventional radiologic procedures.23
Post pancreatic surgery pancreatic duct leaks occur in 330% of
cases. Patients present with pancreatico-cutaneous stula, intra-
abdominal uid collection, ileus or delayed gastric emptying. Expect-
ant management may be undertaken by placement of operative drains
to monitor output. Conservative management involves nutritional
support with nil per os and administration of octreotide. Occasionally,
re-operation or placement of percutaneous drains is required. ERCP
plays a very limited role in the management of post pancreatic surgery
pancreatic leaks in the early postoperative period.
Before contemplating endoscopic evaluation and management of
post pancreatic surgery pancreatic duct leaks, the risk of infecting a
347
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

UCLA JHH pancreatic uid collection should be addressed. A secretinMRCP

Pathology (19902004) (19901999) prior to consideration of endoscopic treatment may assist in plan-
ning endoscopic treatment of the pancreatic duct leak25 and may
Pancreatic adenocarcinoma 38 43
allow determination of stula location, degree of pancreatic ductal
Ampullary cancer 21 11
disruption (incomplete or complete) and the cause of obstruction
Bile duct cancer 6 10
Duodenal cancer 7 4 (stone, main pancreatic duct stricture, anastomotic stricture), if
Ampullary adenoma 3 present.
Pancreatic cystic neoplasm 11 6 Endoscopic pancreatic duct stent placement has been reported
Islet cell tumor 3 5 for the management of pancreatic duct leaks following pancreatic
Chronic pancreatitis 8 11 surgery.2631 It may be used to bypass a distal obstruction of the MPD
as well as for bridging a MPD leakage. Telford et al. suggest that
Table 32.1 Pathologies associated with panceaticoduodenectomy bridging a leak is a associated with a better long-term outcome (92%
(%) vs 50%).31 The endoscopic application of N-butyl-2-cyanoacrylate
UCLA, University of California, Los Angeles, Los Angeles, CA
glue has been successfully used in the management of distal pan-
JHH, Johns Hopkins Hospital, Baltimore, MD
createctomy associated leaks.32

A B
Fig. 32.1 A ERCP showing a pancreatic duct leak with cyst forma-
tion after a distal pancreatectomy for chronic pancreatitis. (Image
courtesy of Todd Barron, MD.) B Pancreatic duct stenting with a
7 fr pancreatic duct stent for management of pancreatic duct leak
after distal pancreatectomy. (Image courtesy of Todd Barron, MD.)
A B
LONG-TERM COMPLICATIONS
OF PANCREATIC SURGERY
Long-term complications associated with pancreatic surgery include
the development of diabetes mellitus and obstruction at either the
choledochojejunostomy or the pancreaticojejunostomy. Complica-
tions of either the choledochojejunostomy or pancreaticojejunos-
tomy may be related to either persistence or recurrence of the
original disease including adenocarcinoma, IPMN and chronic pan-
creatitis. Occasionally, ERCP may be required for the evaluation and
management of these complications.
Patients with problems related to the choledochojejunostomy
typically present with obstructive jaundice or recurrent cholangitis.
Possible causes include benign anastomotic strictures, tumor recur-
rence or biliary stones. Complications of the pancreaticojejunostomy
also include development of benign or malignant strictures. These
may present with abdominal pain or recurrent pancreatitis. Retained,
surgically placed anastomotic pancreatic stents may cause steator-
rhoea or recurrent pancreatitis. Endoscopic retrieval is feasible and
can result in resolution of symptoms.33
A decision to pursue ERCP in the post pancreatic surgery patient
should be based on the decision to provide therapy (e.g. stent place-
ment, removal of stones). This is best achieved by preproce-
dure evaluation with laboratory studies and non-invasive imaging
including CT and MRCP. Although there is an increasing role for
MRCP in the evaluation of such patients, therapeutic intervention
is often necessary.7

Fig. 32.2 A ERCP showing an anastomotic stricturing of a lateral UNDERSTANDING THE ANATOMY
pancreaticojejunostomy (Peustow Procedure). (Image courtesy of
Todd Barron, MD.) B Pancreatic duct stenting of the anastomotic
structuring of a lateral pancreaticojejunostomy (Puestow Proce- Prior to performing endoscopy in a patient who has undergone
dure). (Image courtesy of Todd Barron, MD.) pancreaticoduodenectomy it is important to understand the exact
anatomy and reconstruction in the individual patient. There are two
main types of pancreaticoduodenectomy: classic and pylorus
Study n Stented Unstented preserving.
Povoski 240 41 25
Sohn 567 32 22
Classic pancreaticoduodenectomy (Fig. 32.3)
In the classic pancreaticoduodenectomy, the gastric antrum, head of
Hochwald 71 66 38
Heslin 74 46 11 pancreas, distal bile duct and duodenum are resected. Although
Pisters 265 37 11 there are several different techniques described, one well-accepted
Hodul 212 28 20 reconstruction creates all necessary anastomoses with a single limb
of bowel. A side-to-side gastroenteroanastomosis is usually encoun-
Table 32.2 Studies of preoperative biliary stenting in pancreatic tered endoscopically with a relatively large gastric remnant. The
cancer: infectious complications (%) opening to the afferent limb which should lead to the biliary and

348
 Chapter 32 ERCP for the Acute and Chronic Complications of Pancreatic Surgery

Fig. 32.3 Diagram of a classic pancreaticoduodenec-

tomy. (Reproduced with permission: Feitoza AB, Baron
TH. Gastrointest Endosc 2002; 55(1):7579.)

pancreatic anastomoses is often found along the lesser curvature of

the stomach.
The length of the afferent limb varies according to the position
of the jejunal loop relative to the mesocolon and surgeon preference.
Shorter limbs (about 40 cm) may be found within antecolic anasto-
moses, whereas longer limbs (60 cm or more) may be found with
retrocolic/antecolic combinations. Several sharp angulations are
found along the afferent limb due to xation of the limb to adjacent
organs. The afferent limb ends blindly, where a pancreatic duct
anastomosis may be found if an end-to-end pancreaticojejunostomy
has been created.
The choledochojejunostomy tends to be located approximately
10 cm proximal (endoscopically) to the pancreaticojejunostomy,
along the antimesenteric border of the afferent limb. It is an end-to-
side anastomosis that may be difcult to nd, often hidden behind
a fold.
The pancreaticojejunostomy may be an end-to-end or an end-to-
side anastomosis. The anastomosis may be performed by suturing
the pancreatic ductal mucosa to the jejunal mucosa (mucosal to
mucosal anastomosis), thereby creating a small opening; or by
invaginating the pancreatic stump into the jejunum, with either an
end-to-end or end-to-side pancreaticojejunostomy.34,35 This is known Fig. 32.4 Diagram of pylorus-preserving pancreaticoduodenec-
tomy. (Reproduced with permission: Feitoza AB, Baron TH. Gastro-
as the dunking anastomosis. Due to the various locations and type intest Endosc 2002; 55(1):7579.)
of anastomoses (at with mucosa-to-mucosa; protuberant with
lateral dunking) the identication and subsequent cannulation of
the pancreatic duct is sometimes difcult, and may be facilitated by
administration of intravenous secretin. traversing the pylorus endoscopically, a short segment of duodenal
bulb is present and two openings are identied which correspond
The pylorus-preserving to the end-side duodenojejunostomy. Typically the opening visual-
pancreaticoduodenectomy (Fig. 32.4) ized to the right in the endoscopic eld leads to the afferent limb
In this variation of the Whipple procedure the antrum of the with its biliary and pancreatico anastomoses. Issues relating to
stomach, is not resected. A short proximal duodenal stump remains length of afferent limb and the locations of the anastomoses are the
and is anastomosed to the jejunum as a duodenojejunostomy. After same as with the classic Whipple procedure.

349
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
ENDOSCOPES AND ACCESSORIES
Following pancreaticoduodenectomy the pancreatic and biliary anas-
tomoses may be reached using a standard forward-viewing or side-
viewing endoscope. In some patients it may be necessary to use
a longer length endoscope such as a colonoscope, or even a push-
enteroscope. The use of variable stiffness colonoscopes is useful in
avoiding loop formation during passage of the colonoscope to the
distal aspect of the limb after initial intubation. There have been
reports of using oblique viewing endoscopes with elevators in the
post pancreatic surgery setting.36 There is limited data on the role of
double-balloon enteroscopy in accessing the afferent jejunal limb in
patients with pancreaticoduodenectomy.
The use of a colonoscope does not limit the range of endoscopic
accessories that may be used for ERCP therapy, including standard
catheters and sphincterotomes. However, most balloon catheters
and stent introducing systems are not long enough. When the stan-
dard ERCP balloon catheter is not of sufcient length, using non-
ERCP wire-guided balloons may be an alternative. Similarly the use
of a nasobiliary tube as a pusher tube may be required for stent
placement. However, the absence of an elevator makes diagnostic
and therapeutic manipulation more difcult. The number of avail-
able accessories for use with an enteroscope is limited.
NEGOTIATING THE ANATOMY
Several technical, often time-consuming challenges are involved in
performing ERCP in patients with post pancreatic surgery anatomy,
especially in patients who have undergone pancreaticoduodenec-
tomy. These include accessing and traversing the afferent jejunal
limb, and the identication of the choledochojejunostomy and
pancreaticojejunostomy.
Accessing and traversing the afferent limb
Having detailed operative information about the surgical proce-
dure is helpful. In addition to differentiating between a non-pylorus
sparing versus a pylorus sparing operation, critical information
about the length of the afferent limb and the relative location of the
choledochojejunostomy and the pancreaticojejunostomy is helpful.
For most patients with a pancreaticoduodenectomy, the initial
endoscope of choice is a duodenoscope, as this allows the greater
exibility for the use of therapeutic instruments. Should there be
failure to access the afferent limb or advance the appropriate length
of the limb, a standard therapeutic channel gastroscope or colono-
scope can be used. Both of these instruments allow passage of thera-
peutic diameter accessories.
The primary goal is to identify and access the afferent limb. For
patients with a standard pancreaticoduodenectomy, the afferent
limb is typically off the lesser curve of the stomach. For patients with
a pylorus preserving pancreaticoduodenectomy, the afferent limb is
typically distal to the pylorus and in the right of the endoscopic eld.
Unfortunately however, the distinction between afferent and effer-
ent limb is not always possible using these guidelines. Invariably,
the more easily accessed limb is the efferent limb.
Using a combination of endoscopic and uoroscopic guidance,
the chosen limb is traversed for up to 20 to 40 cm before a decision
can be made about whether the correct limb has been intubated.
Using uoroscopy and demonstrating the endoscope moving in the
350
direction of the liver and the possible surgical clips associated with
the choledochojejunostomy anastomosis can be enough evidence to
continue forward.
Often due to the acute angulation at the site of the afferent jejunal
limb anastomosis initial intubation of the limb can be difcult. This
is due to the need for 180 degree exion of the endoscope to enter
the limb. This tends to put excessive pressure on bowel wall by the
bend of the endoscope and increases the risk of perforation. This
can be corrected by backing into the afferent limb, whereby the
endoscope is guided almost blindly into the afferent limb without
excessive exion of the endoscope, similar to the maneuver required
to enter the ileum during colonoscopy.
The length of the afferent limb is variable and is often surgeon-
dependent. Typically it ranges from 20 cm up to 80 cm. Therefore a
considerable length of jejunal limb needs to be traversed before
deciding whether one is in the correct limb. If the efferent limb has
been intubated, careful and slow withdrawal of the endoscope should
be performed in order to identify and intubate the afferent limb at
the level of the jejunal anastomosis. An alternative strategy to prevent
repeated intubation of the efferent limb includes taking endoscopic
biopsies at 10 cm and 20 cm distal to the jejunal anastomosis. The
limited bleeding associated with these biopsies serves as an indicator
of limb status.
When the afferent limb is nally identied and intubated, care
should be taken to carefully advance the endoscope. In the post sur-
gical patient the presence of adhesions and extraluminal recurrence
of malignancy increase the risk for perforation. Intermittent endo-
scope reduction should be performed to avoid excessive looping.
Often abdominal pressure and changing the position of the patient
(including to a left lateral or supine position) may facilitate passage
of the endoscope to the end of the afferent limb where the hepati-
cojejunostomy and pancreaticojejunostomy are located.
As these are post operative patients with a high risk of endo-
scope induced perforation, it is important to know when to stop
advancing the endoscope. Occasionally information about the
hepaticojejunostomy and pancreaticojejunostomy can be obtained
without reaching the end of the afferent limb. The presence of a
non-dilated appearing air cholangiogram on uoroscopy, although
not positive proof, favors against complete biliary obstruction
(Fig. 32.5). It may also be possible to perform an occlusion contrast
study by inating an extraction balloon in the afferent jejunal limb,
and allowing contrast to ow into the bile duct and pancreatic
duct outlining the anatomy.
Identication of the hepaticojejunostomy
The hepaticojejunostomy is typically an end-to-side anastomosis
located within the last 5 to 10 cm of the afferent limb, proximal to
the pancreaticojejunostomy site. When not strictured, it should be
widely patent (Figs 32.6, 32.7). If not clearly visible, the endoscopist
should search carefully behind jejunal folds on the antimesenteric
side of the jejunum for a possible opening, and insure that the end
of the afferent limb has truly been identied. The use of uoros-
copy can sometimes help locate where the anastomosis should
be, either by tracing an air cholangiogram or recognizing the pre-
sence of surgical clips at the site of the hepaticojejunostomy.
Often the only endoscopic evidence of a hepaticojejunostomy is a
small dimple.
When faced with a strictured hepaticojejunostomy anastomosis,
probing the anastomosis with a wire-guided system is preferred.
 Chapter 32 ERCP for the Acute and Chronic Complications of Pancreatic Surgery

Fig. 32.5 Fluoroscopic pancreatogram in a patient with a pancre-

aticoduodenectomy for chronic pancreatitis. Note the air cholan- B
giogram (arrow).

Blind injection of a presumed anastomosis can lead to submucosal

injection of contrast and even perforation. After access to the biliary
tree has been conrmed by contrast injection, standard ERCP acces-
sories may be used including stents, balloons and sphincterotomes
(for orientation).

Identication of the pancreaticojejunostomy

The pancreaticojejunostomy is more difcult to identify than the
hepaticojejunostomy (Figs 32.8, 32.9). Knowledge of the operation,
including the type and location of the anastomosis is important. All
that may be present is a focal dimple in the region of the end of the
afferent limb. The use of methylene blue spray (0.5%) may enhance
the endoscopic image to locate the pancreaticojejunostomy as the
pancreatic secretions create a halo of clearing of the dye. Alterna-
tively, the use of secretin to stimulate pancreatic juice ow may help
locate the opening. As with stenotic biliary orices, wire-guided
probing of the presumed pancreatic duct is preferred to blind con-
trast injection. Fig. 32.6 A Endoscopic image of a widely patent hepaticojeju-
nostomy. B Endoscopic image of biliary tree through the heapti-
cojejunostomy orice.
ENDOTHERAPY FOR THE POST-PANCREATIC
SURGERY PATIENT
The reason for obstruction at the hepaticojejunostomy or pancreati- stricture re-evaluated after 25 months. If stenting of a pancreatic
cojejunostomy anastomoses needs to be dened prior to attempted anastomotic stricture is required, a single 5 Fr plastic pancreatic
therapeutic intervention. If there is evidence of tumor recurrence, stent is often inserted and the stricture re-evaluated after 24
then endoscopic stent placement is adequate. If there is evidence of weeks.
benign stricturing, then short-term stenting or balloon dilatation
may be adequate. NOVEL APPROACHES TO ERCP IN THE
In all patients requiring anastomotic dilatation (e.g. hepaticoje- POST PANCREATICODUODENECTOMY
junostomy anastomosis, pancreaticojejunostomy anastomosis), PATIENTS
endoscopic balloon dilatation with standard biliary dilating cathe-
ters is preferable. If stenting of a biliary anastomotic stricture is The use of combined interventional radiology ERCP rendezvous
required, single 10 Fr plastic biliary stents are employed and the procedures has been described, when the endoscopist is able to

351
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 32.7 Balloon occlusion cholangiogram of the biliary tree
in a patient with a hepaticojejunostomy after pancreaticoduo-
denectomy.
Fig. 32.8 Endoscopic pancreaticojejunostomy stenting for pan-
creatic duct obstruction secondary to pancreaticojejunostomy
stenosis.
reach the end of the afferent limb, but cannot obtain access to the
biliary system. The percutaneous transhepatic placement of an inter-
nal-external biliary catheter can serve as a guide for the endoscopist
to access the biliary tree. A combined EUS-ERCP rendezvous proce-
dure has been described whereby EUS guided transgastric access to
the pancreatic duct with placement of a wire antegrade through the
pancreaticojejunostomy is followed by ERCP access to the pancreatic
352
Fig. 32.9 Fluoroscopic normal pancreatogram in patient with pan-
creaticoduodenectomy for pancreatic neoplasia. Evidence of ste-
notic choledochojejunostomy requiring endoscopic stenting.
duct.37 Although it provides the potential for improved access, there
is limited experience with the use of double-balloon enteroscopy on
post pancreaticoduodenectomy afferent jejunal limbs. The develop-
ment of this technology for the management of patients after pan-
creatic surgery will require the development of compatible ERCP
accessories.
BRIGHAM AND WOMENS
HOSPITAL EXPERIENCE
Because of the difculties encountered in negotiating the proximal
portion of the afferent limb with an endoscope, as well as identica-
tion of the pancreatic duct opening and sometimes the bile duct
opening, the overall success rate of cannulation is variable. There
are limited specic data available from either the endoscopic or
interventional radiology literature about the indications and out-
comes (either technical or clinical) of non-operative intervention in
patients undergoing biliary or pancreatic therapy after pancreatic
surgery.35
One hundred and thirty-ve ERCPs were performed on 61
patients after pancreatic surgery at the Brigham and Womens Hos-
pital between 1990 and 2000. Four endoscopists performed all the
procedures, although one (DCL) was involved in the majority of
cases. Surgical operations included pancreaticoduodenectomy,32
distal pancreatectomy (tail resection)11 and lateral pancreaticojeju-
nostomy.21 Indications for surgery included neoplasia (pancreatic
ductal adenocarcinoma, intraductal papillary mucinous neoplasia
(IPMN), neuroendocrine tumors), chronic pancreatitis and trauma.
No procedure-related limb perforations occurred.
Twenty-nine patients with a pancreaticoduodenectomy
underwent 56 ERCPs. Ten of these patients had a classic
 Chapter 32 ERCP for the Acute and Chronic Complications of Pancreatic Surgery
29 Patients
ERCP indication Jaundice (16) Pain (13)
Afferent limb access 15/16 (94%) 12/13 (92%)
Cholangiogram 11/15 (73%) 7/12 (58%)a
Pancreatogram N/A 5/10 (50%)b
Table 32.3 ERCP in pancreaticoduodenectomy patients: Brigham
and Womens experience
a
Choledochojejunostomy seen in all 12, but cholangiogram performed in only 7
b
Two of these patients had a satisfactory biliary explanation for their symptoms,
hence not necessitating a pancreatic evaluation
pancreaticoduodenectomy, while 19 had a pylorus-preserving
pancreaticoduodenectomy. Indications for surgery included
neoplasia,17 chronic pancreatitis11 and trauma.1
The mean time from surgery to ERCP was 15 months (range: 1
120 months). The mean number of ERCPs performed per patient
was 1.9 (range: 111). Only three patients had more than three
ERCP procedures performed. These were all due to recurrent
stricturing of the pancreaticojejunostomy anastomosis, requiring
repeated dilatation and stenting, in patients with a history of
chronic pancreatitis.
The indications for ERCP in patients with pancreaticoduodenec-
tomy were jaundice16 and pain.13 Patients undergoing ERCP for
evaluation of jaundice were more likely to have undergone surgery
for neoplasia (14/16, 88%), compared with patients undergoing
ERCP for evaluation of pain who were more likely to have undergone
surgery for chronic pancreatitis (10/13, 77%) (p < 0.05). Pain in the
immediate postoperative period in one patient was attributed to a
bile leak (Table 32.3).
Of the 16 patients undergoing ERCP for jaundice, 14 had had
surgery for neoplasia, and one each had surgery for trauma or
chronic pancreatitis. The afferent limb was accessed in 15/16 (94%).
The reason for failure in one patient was the presence of a metal
enteral stent traversing the gastrojejunostomy into the efferent limb
(previously placed for malignant gastric outlet obstruction). The cho-
ledochojejunostomy was identied in 11/15 patients (73%) whose
afferent limb was accessed for evaluation of jaundice. Inability to
identify the choledochojejunostomy was related to an excessively
long afferent limb (n = 2) or poor visualization even when the end
of the afferent limb was reached (n = 2). Inability to identify the
choledochojejunostomy was equally distributed amongst both types
of surgical procedures (pylorus-preserving pancreaticoduodenec-
tomy (n = 2) and classic pancreaticoduodenectomy (n = 2)).
For patients being evaluated for jaundice and in whom the cho-
ledochojejunostomy was identied, all had successful cholangio-
graphy (often balloon occlusion cholangiography). Only four
cholangiograms provided an explanation for the jaundice (bile duct
stricture2 and choledochojejunostomy anastomotic stricture2) and all
required endoscopic therapy. Patients with the choledochojejunos-
tomy strictures were treated by dilatation and temporary stenting.
All four patients had long-term resolution of their jaundice. The
remaining seven cholangiograms performed in this subgroup were
normal. Thus, of the 16 patients undergoing ERCP for jaundice,
cholangiography was successful in 11/16 (69%), excluding a biliary
structural issue in 7/11. An air cholangiogram was identied in two
patients, suggesting the absence of a signicant high-grade obstruc-
tion and the remaining three underwent a subsequent percutaneous
cholangiogram.
Thirteen patients underwent ERCP for evaluation of pain. Three
had surgery for neoplasia, and 10 had surgery for chronic pancreati-
tis. The afferent limb was accessed in 12/13 (92%). The choledocho-
jejunostomy was identied in all 12 patients whose afferent limbs
were accessed. A cholangiogram was performed in seven of these
12 patients, of which only one was abnormal, showing a choledo-
chojejunostomy anastomotic stricture which was dilated and stented
successfully. The remaining ve patients did not have a cholangio-
gram performed in the evaluation of their pain, because a pancreatic
etiology was suspected based on the nature of the pain, laboratory
tests, or a non-invasive imaging study.
Of the 12 patients undergoing ERCP for evaluation of pain whose
afferent limbs were accessed, the pancreaticojejunostomy was iden-
tied in ve, aided by the use of secretin or dye-spraying in three,
and no pancreaticojejunostomy was identied in the remaining
seven. Three of these ve patients had a pancreaticojejunostomy
stricture which required dilatation and stenting, with long-term
improvement in their pain symptoms. The remaining two had no
evidence of a pancreaticojejunostomy stricture.
In the remaining seven patients who were undergoing ERCP
evaluation for pain and in whom the afferent limb was accessed,
reasons for not visualizing the pancreaticojejunostomy included:
very long afferent limbs,3 inability to identify the pancreaticojejunos-
tomy despite the use of dye-spraying or secretin,2 and an alternative
satisfactory non-pancreatic explanation (e.g. choledochojejunostomy
anastomotic stricture1 or exclusion of a bile duct leak1).
Overall, patients with a previous history of cancer were more
likely to undergo ERCP for the evaluation of jaundice (14/17, 82%),
and those with a previous history of chronic pancreatitis were more
likely to undergo an ERCP for evaluation of pain (10/11, 91%),
although technical success was slightly better in patients being eval-
uated for jaundice than for pain (69% vs 54%).
This experience with ERCP in the post pancreaticoduodenectomy
setting has several limitations. Its retrospective nature affects the
ability to clearly dene the indications for the procedures and long-
term clinical outcome. Indications for the procedure were obtained
from medical records and procedure reports. Long-term clinical
outcome (including resolution of symptoms) was not sought and
limits the value of this data.
The use of other non-invasive imaging studies such as MRCP in
the evaluation of these patients was not evaluated. The evolution of
non-invasive imaging technology during the study period has altered
the diagnostic role of ERCP in this patient population. There are two
main benets of MRCP imaging in this patient population, The rst
is to identify patients who would benet from either biliary or pan-
creatic endotherapy, and those who could be spared an endoscopic
procedure. The second main benet of MRCP is the evaluation of
the anatomy in cases where endoscopy has not been successful due
to inability to access the relevant limb or inability to identify the
choledochojejunostomy or pancreaticojejunostomy. With respect to
the rst benet, we estimate that for indications relating to jaundice
in our study population, it is likely that an adequate MRCP would
have precluded need for an ERCP in 9 of 16 patients. For indications
of pain, it is likely that adequate MRCP would have resulted in
endoscopic cholangiograms not being performed in at least 6 of 7
353
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
cases and endoscopic pancreatograms not being performed in 2 of
5 cases. This is based on the assumptions that MRCP could easily
diagnose or rule out pancreatic lesions requiring endotherapy such
as pancreaticojejunostomy anastomotic stricturing, and that all
biliary duct abnormalities on MRCP would result in an ERCP. For
the second benet, it is much more speculative to determine how
MRCP might have impacted on the diagnosis and management of
patients in whom minimal endoscopic evaluation was possible in
our study. Based on these data it seems prudent to undertake an
initial non-invasive MRCP rather than an ERCP, except perhaps in
cases where there is a very high suspicion of the need for a thera-
peutic intervention, e.g. recurrent anastomotic stricture, occlusion
of a biliary stent.
REFERENCES
1. Bolman RM, 3rd. Surgical management of chronic relapsing
pancreatitis. Ann Surg 1981; 193(2):125131.
2. Traverso LW, Tompkins RK, Urrea PT, Longmire WP, Jr. Surgical
treatment of chronic pancreatitis. Twenty-two years experience.
Ann Surg 1979; 190(3):312319.
3. Whipple A. Treatment of carcinoma of the ampulla of Vater. Ann
Surg 1935; 102:763769.
4. Puestow C. Retrograde surgical drainage of pancreas for chronic
relapsing pancreatitis. Arch Surg 1958; 76:698707.
5. DuVall M. Caudal pancreatico-jejunostomy for chronic relapsing
pancreatitis. Ann Surg 1954; 140:775785.
6. Feitoza AB, Baron TH. Endoscopy and ERCP in the setting of
previous upper GI tract surgery. Part II: postsurgical anatomy with
alteration of the pancreaticobiliary tree. Gastrointest Endosc 2002;
55(1):7579.
7. Yeo CJ, Cameron JL, Sohn TA, et al. Six hundred fty consecutive
pancreaticoduodenectomies in the 1990s: pathology,
complications, and outcomes. Ann Surg 1997; 226(3):248257;
discussion 257260.
8. Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous
neoplasms of the pancreas: an increasingly recognized
clinicopathologic entity. Ann Surg 2001; 234(3):313321; discussion
321322.
9. Abe N, Sugiyama M, Suzuki Y, et al. Preoperative endoscopic
pancreatic stenting for prophylaxis of pancreatic stula
development after distal pancreatectomy. Am J Surg 2005;
191(2):198200.
10 Hateld AR, Tobias R, Terblanche J, et al. Preoperative external
biliary drainage in obstructive jaundice. A prospective controlled
clinical trial. Lancet 1982; 2(8304):896899.
11. McPherson GA, Benjamin IS, Hodgson HJ, et al. Pre-operative
percutaneous transhepatic biliary drainage: the results of
a controlled trial. Br J Surg 1984; 71(5):371375.
12. Pitt HA, Gomes AS, Lois JF, et al. Does preoperative percutaneous
biliary drainage reduce operative risk or increase hospital cost?
Ann Surg 1985; 201(5):545553.
13. Lai EC, Mok FP, Fan ST, et al. Preoperative endoscopic drainage
for malignant obstructive jaundice. Br J Surg 1994;
81(8):11951198.
14. Povoski SP, Karpeh MS, Jr., Conlon KC, et al. Association of
preoperative biliary drainage with postoperative outcome
following pancreaticoduodenectomy. Ann Surg 1999;
230(2):131142.
15. Sohn TA, Yeo CJ, Cameron JL, et al. Do preoperative biliary stents
increase postpancreaticoduodenectomy complications? J
Gastrointest Surg 2000; 4(3):258267; discussion 267268.
354
CONCLUSION
ERCP in the post pancreatic surgery setting is useful for both biliary
and pancreatic indications. With the hope of improved survival with
adjuvant therapy for pancreatic neoplasia, increased surgical rates
for preneoplastic conditions of the pancreas (e.g. intraductal papil-
lary mucinous tumor, benign cysts of the pancreas), and continued
surgical management of patients with chronic pancreatitis, the need
for therapeutic ERCP in these patient populations will continue.
Improved cooperation between the pancreatic surgeon and endos-
copist (especially in patients with benign disease) and the develop-
ment of more specialized endoscopes will improve the success of
ERCP in these patients.
16. Pisters PW, Hudec WA, Hess KR, et al. Effect of preoperative
biliary decompression on pancreaticoduodenectomy-associated
morbidity in 300 consecutive patients. Ann Surg 2001;
234(1):4755.
17. Hochwald SN, Burke EC, Jarnagin WR, et al. Association of
preoperative biliary stenting with increased postoperative
infectious complications in proximal cholangiocarcinoma. Arch
Surg 1999; 134(3):261266.
18. Heslin MJ, Brooks AD, Hochwald SN, et al. A preoperative biliary
stent is associated with increased complications after
pancreatoduodenectomy. Arch Surg 1998; 133(2):149154.
19. Hodul P, Creech S, Pickleman J, et al. The effect of preoperative
biliary stenting on postoperative complications after
pancreaticoduodenectomy. Am J Surg 2003; 186(5):420425.
20. Mullen JT, Lee JH, Gomez HF, et al. Pancreaticoduodenectomy
after placement of endobiliary metal stents. J Gastrointest Surg.
2005 Nov; 9(8):10941104; discussion 11041105.
21. Wasan SM, Ross WA, Staerkel GA, et al. Use of expandable
metallic biliary stents in resectable pancreatic cancer. Am J
Gastroenterol. 2005 Sep; 100(9):20562061.
22. Lawrence C, Howell DA, Conklin DE, et al. Delayed
pancreaticoduodenectomy for cancer patients with prior
ERCP-placed, nonforeshortening, self-expanding metal
stents: a positive outcome. Gastrointest Endosc 2006 May;
63(6):804807.
23. Sohn TA, Yeo CJ, Cameron JL, et al. Pancreaticoduodenectomy:
role of interventional radiologists in managing patients and
complications. J Gastrointest Surg 2003; 7(2):209219.
24. Gervais DA, Fernandez-del Castillo C, ONeill MJ, et al.
Complications after pancreatoduodenectomy: imaging and
imaging-guided interventional procedures. Radiographics 2001;
21(3):673690.
25. Gillams AR, Kurzawinski T, Lees WR. Diagnosis of duct disruption
and assessment of pancreatic leak with dynamic secretin-
stimulated MR cholangiopancreatography.AJR Am J Roentgenol.
2006 Feb; 186(2):499506.
26. Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic
transpapillary therapy for disrupted pancreatic duct and
peripancreatic uid collections. Gastroenterology 1991;
100(5 Pt 1):13621370.
27. Saeed ZA, Ramirez FC, Hepps KS. Endoscopic stent placement for
internal and external pancreatic stulas. Gastroenterology 1993;
105(4):12131217.
28. Bracher GA, Manocha AP, DeBanto JR, et al. Endoscopic pancreatic
duct stenting to treat pancreatic ascites. Gastrointest Endosc 1999;
49(6):710715.
 Chapter 32 ERCP for the Acute and Chronic Complications of Pancreatic Surgery
29. Boerma D, Rauws EA, van Gulik TM, et al. Endoscopic stent
placement for pancreaticocutaneous stula after surgical drainage
of the pancreas. Br J Surg 2000; 87(11):15061509.
30. Costamagna G, Mutignani M, Ingrosso M, et al. Endoscopic
treatment of postsurgical external pancreatic stulas. Endoscopy
2001; 33(4):317322.
31. Telford JJ, Farrell JJ, Saltzman JR, et al. Pancreatic stent placement
for duct disruption. Gastrointest Endosc 2002; 56(1):1824.
32. Seewald S, Brand B, Groth S, et al. Endoscopic sealing of
pancreatic stula by using N-butyl-2-cyanoacrylate. Gastrointest
Endosc 2004; 59(4):463470.
33. Levy MJ, Chari S, Adler DG, et al. Complications of temporary
pancreatic stent insertion for pancreaticojejunal anastomosis
during pancreaticoduodenectomy. Gastrointest Endosc 2004
May; 59(6):719724.
34. Toledo-Pereyra L. The pancreasprincipals of medical and
surgical practice. New York: John Wiley & Sons; 1985.
35. Braasch JG, Ganger M. Pylorus-preserving
pancreaticoduodenctomy. Langenbecks Arch Chirg 1991;
376:5058.
36. Law NM, Freeman ML. ERCP by using a prototype
oblique-viewing endoscope in patients with
surgically altered anatomy. Gastrointest Endosc 2004;
59(6):724728.
37. Mallery S, Matlock J, Freeman ML. EUS-guided rendezvous
drainage of obstructed biliary and pancreatic ducts: Report of
6 cases. Gastrointest Endosc 2004; 59(1):100107.
38. Kugelberg C, Wehlin L, Arnesjo B, et al. Endoscopic
pancreatography in evaluating results of pancreatico-jejunostomy.
Gut 1976; 17(4):267272.
355
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Choledocholithiasis
33 Yuk Tong Lee and Joseph Sung
INTRODUCTION
Choledocholithiasis, dened as the presence of stone inside the
common bile duct (CBD), is a common condition. At least 15% of
patients with cholelithiasis have choledocholithiasis. Conversely,
95% of patients with CBD stones also have gallstones.1 It is esti-
mated that over 20 million Americans have gallstone disease and up
to 25% of elderly patients have calculi in the CBD at the time of
cholecystectomy.2
In the West, the majority of CBD stones are composed of
cholesterol stones that originated from the gallbladder. Less than
10% of CBD stones are formed de novo within the CBD. On the
other hand, in the East, because of a higher incidence of chronic
biliary tree infection and infestation, the occurrence of pigment
stones is much more common. This condition, which begins
with inammation in the bile ducts, is referred to as recurrent pyo-
genic cholangitis (RPC). It is characterized by recurrent cholangitis
due to the presence of multiple pigment stones formed inside the
intrahepatic ducts.
This chapter focuses on the current management of choledocho-
lithiasis, and the related problem of cholelithiasis. Readers should
refer to other chapters for the technique of bile duct cannulation
(Chapter 8), sphincterotomy (Chapter 12), CBD stone extraction
(Chapter 13), management of RPC (Chapter 38) and acute biliary
pancreatitis (Chapter 39).
CLINICAL MANIFESTATION
The natural history of choledocholithiasis is not fully recognized.
The clinical manifestation largely depends on the location of the
stones and the presence of bacterial invasion. Stones formed in the
intrahepatic ducts could be asymptomatic and remain in the bile
ducts for a long period of time without being recognized. Small CBD
stones may pass through the papilla spontaneously causing few
symptoms.3 From a recent study up to one-fth of CBD stones less
than 5 mm in diameter pass into the duodenum spontaneously
without causing signicant symptoms.4 On the other hand, sponta-
neous passage of small calculi through the papilla may be compli-
cated by biliary pancreatitis. In patients with mild biliary pancreatitis,
spontaneous stone passage may occur in 7080% of patients. Stones
impacted in the distal bile duct often lead to biliary colic, jaundice,
cholangitis or pancreatitis. When bacterial infection develops, stone
impaction will be complicated by suppurative cholangitis. If the bile
duct obstruction is gradual and insidious, patients may present with
progressive obstructive jaundice and rarely, secondary biliary cir-
rhosis. If the stone is impacted in the cystic duct causing obstruction
of bile ow in the gallbladder and/or the biliary tract, acute chole-
cystitis or Mirizzi syndrome may arise.
DIAGNOSIS OF CHOLEDOCHOLITHIASIS
Trans-abdominal ultrasound (US) is commonly used for diagnosis
of choledocholithiasis. However, limited by various technical factors
and compounded by variable experience of operators, the sensitivity
of US in detecting choledocholithiasis ranges from 22% to 80%.1 As
extrahepatic biliary dilation is present in 7094% of patients with
choledocholithiasis, it serves as a useful surrogate marker for the
diagnosis. However, it is important to note that a normal US study
does not exclude the possibility of choledocholithiasis.
Although ERCP is a time-honored gold standard for the diagnosis
of choledocholithiasis, its accuracy for the detection of small stones
and biliary sludge is being questioned.5,6 Both under-lling and over-
lling of the bile duct during contrast injection may cause CBD and
intrahepatic ductal stones to be missed. Occasionally, stones cannot
be differentiated from air bubbles. A recent study in which ERCP
alone was compared to ERCP combined with miniprobe intraductal
ultrasound (IDUS) showed the latter approach had a higher accuracy
(87% versus 97%) in diagnosing bile duct stones and sludge.5 Fur-
thermore, because of its invasive nature and the possible complica-
tions associated with cannulation of the CBD, the use of diagnostic
ERCP has been gradually replaced with other modalities for the
detection of choledocholithiasis.
Helical CT has a sensitivity of detecting choledocholithiasis of
65% to 88%. With the use of contrast-enhanced CT cholangiography,
the sensitivity increases to 90%.7 However, the use of CT cholangi-
ography is limited in patients with signicant hyperbilirubinemia or
advanced liver disease. Up to 10% of patients have adverse events
after administration of contrast.
Magnetic resonance cholangiopancreatography (MRCP) is proven
to be an accurate non-invasive imaging for choledocholithiasis. In a
recent meta-analysis of 67 studies, the pooled sensitivity and speci-
city for MRCP in diagnosing biliary obstruction were 95% and
97%, respectively.8 The sensitivity for detecting choledocholithiasis
is slightly lower (92%) as detection of small CBD stones can still be
a challenge. For CBD stones less than 5 mm in size, the sensitivity
of detection by MRCP drops to below 65%. MRCP is also poor in
detecting biliary sludge. Despite these limitations, using MRCP to
select patients for ERCP is shown to be substantially cost-saving and
improves the quality of life in patients with low to intermediate risk
of choledocholithiasis.9 If available, MRCP should replace diagnostic
ERCP.
In the last decade, the advent of EUS has changed the manage-
ment of CBD stones. The sensitivity of EUS in detecting choledo-
cholithiasis reported in various studies ranges from 93 to 100%.
Unlike the other imaging modalities, performance of EUS is less
affected by the size of CBD stones and the diameter of the bile duct
(Figs 33.133.3). Most studies comparing MRCP against EUS
357
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Fig. 33.1 EUS examination demonstrating a small stone (arrow) in Fig. 33.2 EUS examination demonstrates sludge (arrow) in the
a non-dilated bile duct. Acoustic shadowing is seen behind the distal CBD. The sludge appears as homogeneous amorphous mate-
stone. rial without casting any acoustic shadowing.

showed superiority of the latter in accuracy. In a recent study com-

paring MRCP with EUS in detecting choledocholithiasis, MRCP
(sensitivity 87.5%, specicity 96.6%) was outperformed by EUS (sen-
sitivity 93.8%, specicity 96.6%).10 The high negative predictive value
of EUS helps to reduce unnecessary diagnostic ERCP. Based on a
decision analysis model, using EUS to triage patients for therapeutic
ERCP was shown to be safe and cost-effective in patients with biliary
pancreatitis.11 The nding is in accord with a prospective study in
which 485 patients who were suspected (based on clinical, biochemi-
cal and US parameters) to have choledocholithiasis were triaged by
EUS to receive ERCP and stone extraction.12 EUS showed a superb
result of sensitivity, specicity, positive and negative predictive value
of 98%, 99%, 99% and 98% respectively. In 214 (46%) patients,
invasive ERCP was avoided. The mean cost for patients managed by
the EUS-based strategy was signicantly lower than the theoretical
mean cost for patients managed by the ERCP-based strategy.
However, EUS causes more discomfort than MRCP and patient
preference may aid in the choice of one modality over another.
Depending on clinical suspicion and serum laboratory tests,
patients can be divided into low risk (e.g. atypical pain with normal
liver function test), intermediate risk (pain and biochemical evidence
of biliary obstruction) and high risk (fever, jaundice and pain) of
having choledocholithiasis. The following algorithm is suggested to
investigate these patients according to the risk stratication (Fig. Fig. 33.3 Endoscopic appearance of biliary sludge extraction using
an occlusion balloon after a sphincterotomy was performed.
33.4).
For low-risk patients, a normal US examination is adequate to
exclude the presence of choledocholithiasis and no further testing is sidering invasive ERCP examination. There may be a small benet
needed. If a CBD stone is detected the patient is referred for ERCP. by adding EUS to a patient who has a dilated duct but otherwise
If a dilated bile duct without a stone is detected, the patient should normal MRCP examination.
undergo an additional non-invasive study such as CT or MRCP, or For the intermediate risk patient, if US detects any CBD stone,
less invasive study (EUS) to look for underlying causes before con- the patient should be referred for ERCP. However, if the US

358

Suspected choledocholithiasis
High risk Intermediate risk
US
CBD Normal Dilated duct CBD stone Normal Dilated duct
stone with no stone with no stone
MRCP
(contrast CT)
ERCP (+ IDUS)
+ sphincterotomy Normal
EUS
examination is normal, additional testing such as CT, MRCP or EUS
is still needed. If US examination shows a dilated duct with no stone,
MRCP or EUS would be needed to rule out underlying pathology.
The patient may also be considered for ERCP.
For the high-risk patient, especially in a patient with features of
acute cholangitis, the patient may go directly for ERCP without US
examination. If the patients condition is stable, an US can be per-
formed to exclude other intra-abdominal pathology before an ERCP
is performed. Whether a normal EUS examination spares the patient
from invasive ERCP deserves further study.
MANAGEMENT OF CHOLANGITIS SECONDARY
TO CHOLEDOCHOLITHIASIS
When stones become impacted in the distal CBD, biliary stasis
ensues. The normal pressure in the ductal system is 1015 cm H2O.
When the pressure exceeds 30 cm H2O, bile ow stops. Due to the
raised intraductal pressure, secretion of IgA, the predominant
immunoglobulin in bile, is suppressed and hence antibacterial func-
tion of bile weakened. Bacteria ascending through the transpapillary
route or translocating through the portal venous circulation may
invade into the biliary system resulting in cholangitis.13 The most
commonly found bacteria are Escherichia coli, enterococci, Klebsiella,
Proteus species and Pseudomonas (after surgical or endoscopic
manipulations). Anaerobic organisms such as Bacteroides fragilis or
Clostridium perfringens are found in about 15% of cases. With the
raised intraductal pressure in the setting of choledocholithiasis, bac-
teria in the bile ducts may reux back into the hepatic venous system
leading to the development of septicemia. The elevated intraductal
pressure also suppresses active and passive secretion of antibiotics
into the biliary system thus hampering their antimicrobial activity.
Among commonly used antibiotics, only ciprooxacin is excreted in
detectable levels in the presence of biliary obstruction.14 As elevated
intrabiliary pressure is the main pathophysiological mechanism of
biliary sepsis, decompression of the obstructed system is a critical
step in reversing these processes and controlling endotoxaemia,15
and is the mainstay of treatment for cholangitis.
Chapter 33 Choledocholithiasis
Fig. 33.4 Algorithm in investigating patients
with suspected choledocholithiasis. More than
one option can be followed in each clinical sit-
uation (solid arrow). The dotted line represents
Low risk an alternative EUS-based approach.
CBD stone Normal
ERCP Observe
CBD stone
The symptoms of cholangitis include right upper quadrant pain,
fever and jaundice which represent the classical Charcots triad.
However, these classical symptoms may occur in only 70% of cases.
When more severe septicemia occurs, the patient may present with
hypotension, mental confusion and sign of peritonitis (when occur-
ring with Charcots triad is known as Reynolds pentad). Laboratory
examination in patients with cholangitis demonstrates an obstruc-
tive liver enzyme pattern and leukocytosis.
Acute cholangitis carries a high morbidity and mortality,
especially in elderly patients. Conservative treatment with antibiot-
ics for patients with cholangitis is effective in only 80% of cases.
Before the development of ERCP and endoscopic drainage, the
surgical mortality rate of acute suppurative cholangitis was
around 40%. ERCP has established its role in the treatment of
cholangitis in the 1980s. In patients suffering from suppurative
cholangitis, emergency endoscopic drainage through the insertion
of a nasobiliary catheter was shown to be effective in controlling
sepsis.16 A randomized controlled trial showed endoscopic decom-
pression of the bile duct reduces morbidity and mortality rates
when compared with emergency open surgery and CBD explora-
tion.17 With successful drainage of the biliary system, abdominal
pain and fever subside, and the patients hemodynamic status sta-
bilizes. The mortality rate of cholangitis has been brought down to
58% in recent years.
When performing ERCP for patients with acute suppurative chol-
angitis, one should exercise extra caution when introducing radio-
logical contrast into the obstructed biliary system since this will
further increase the intraductal pressure precipitating septicemia.
The use of a guidewire instead of contrast during the cannulation is
recommended. After deep cannulation is achieved, the CBD should
be decompressed by aspiration of the infected bile prior to contrast
injection. Only a minimal amount of contrast should be used to
outline the bile ducts.
There are several potential strategies in the treatment of choledo-
cholithiasis after successful cannulation and decompression of the
bile duct is achieved. A one-step approach with endoscopic sphinc-
terotomy and stone extraction is the treatment of choice for patients
359
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 33.5 Endoscopic view of basket extraction of a large mixed
stone through a biliary sphincterotomy.
with stable clinical conditions (Fig. 33.5). In patients who are
critically ill or with coagulopathy due to prolonged cholestasis, the
complication rate of biliary sphincterotomy is substantial. Complete
ductal clearance at rst ERCP may only be achieved in 5785%.18
Prolongation of the procedure and increasing sedation poses unnec-
essary challenges to the cardiopulmonary function of these patients.
A B
Fig. 33.6 A Cholangitis due to impacted distal CBD stone. B A typical nasobiliary drain (NBD). C Radiographic image after successful
NBD insertion in same patient.
360
For patients that are not suitable for prolonged endoscopic interven-
tion, urgent decompression followed by elective ERCP to remove the
stone is recommended.
After sphincterotomy and stone extraction, it is sometimes useful
to leave a nasobiliary drain (NBD) or biliary stent for temporary
drainage. This procedure is particularly useful after edema develops
due to manipulation of the papilla. Incomplete clearance of the CBD
after basket mechanical lithotripsy (BML) may leave small stone
fragments behind blocking bile ow. A NBD will secure drainage of
the CBD. After the sepsis subsides a follow-up cholangiogram can
be obtained by injecting contrast through the NBD without need for
passage of the endoscope. If a stone is not detected, the drain can
safely be removed.
NBD is widely used for the initial decompression in patients with
acute cholangitis (Fig. 33.6). The insertion of a NBD is an easy pro-
cedure once deep cannulation of bile duct is achieved. Bile sample
can be obtained through NBD for bacteriological culture; bile output
can be continuously monitored. In case of clogging of the NBD, the
drain can be gently ushed with water. However, the insertion and
re-routing of the drain through the nose may be difcult in septic
patients with confusion. The drain may be pulled out inadvertently
by the patients. In patients that require prolonged drainage, it is
uncomfortable to the patients and the uid and electrolyte loss
through the external drainage could be signicant.
The alternative to NBD is the insertion of a biliary stent. Unlike
NBD, stent placement does not require re-routing through the nose
so that it can be done even in uncooperative or confused patients.
An indwelling biliary stent cannot be removed by the patient, does
not cause discomfort, and does not disturb uid and electrolyte
balance.
The benet of combining sphincterotomy to NBD or stent place-
ment to improve biliary drainage has been debated. In a retrospec-
C

tive study, 166 patients who underwent ERCP and NBD insertion
with and without concomitant sphincterotomy for treatment of chol-
angitis were compared.19 The success rates of NBD insertion were
95% and 96% and efcient drainage was achieved in 92% and 94%
in patients with and without sphincterotomy, respectively. Eleven
percent of the patients who underwent sphincterotomy developed
complications (including hemorrhage, cholecystitis and pancreati-
tis) as compared to 2% in the non-sphincterotomy group. In another
retrospective study, 74 patients received ERCP and biliary stent
placement for decompression in the setting of acute cholangitis.20
Half of the patients had a concomitant sphincterotomy. Three
patients in the sphincterotomy group had bleeding that required
endoscopic treatment and one had acute pancreatitis. One patient
who received stent placement alone developed pancreatitis. Clinical
outcomes of the two groups were similar but procedure time was
signicantly shorter in those without a sphincterotomy. Both studies
suggest that endoscopic sphincterotomy may not be necessary in
acute cholangitis when biliary decompression can be achieved by
either NBD or biliary stenting.
Two randomized, controlled studies compared initial NBD
insertion to biliary stent placement in the treatment of acute chol-
angitis. Lee et al. recruited 79 patients presenting with severe acute
cholangitis to receive either a 6.5 Fr NBD or insertion of a 10-Fr
stent.21 Endoscopic sphincterotomy was not performed in either
group. NBD were successfully inserted in all cases whereas there
was one failed stent insertion. Four patients pulled out their NBD
inadvertently and in one patient the NBD kinked. There was one
early occlusion in the stent group. Patients in the stent group expe-
rienced less discomfort than patients in the NBD group. The overall
mortality rate was 6.8% (2.5% NBD group; 12% stent group). The
higher mortality rate in the stent group was attributed to more severe
cholangitis in this group. Sharma et al. randomized 150 patients
with severe cholangitis to receive either a 7 Fr NBD or a 7 Fr biliary
stent insertion for biliary drainage.22 Biliary drainage was achieved
in 74/75 (99%) patients in the NBD group and in 73/75 (98%)
patients in the stent group. There were no episodes of displacement,
kinking, or occlusion of the NBD, and no episodes of stent occlusion
or migration. Two patients died in each group. Based on these two
randomized trials, biliary drainage can be achieved equally well by
NBD or stenting in the treatment of severe cholangitis.
In patients who are debilitated by chronic illness and those with
limited life expectancy, endoscopic stenting can be used as a deni-
tive treatment for cholangitis. However, for patients treated with
biliary stents alone the rate of recurrent cholangitis is expected to be
higher than those who have achieved complete stone clearance in
long-term follow-up.23 In one non-randomized study, 36 patients
with difcult CBD stones were treated with either electrohydraulic
lithotripsy (EHL) for ductal clearance or stent placement only.
Patients treated with stents experienced a signicantly higher rate
of recurrent cholangitis (63.2% versus 7.7%) and mortality (73.7%
versus 41.2%) compared to patients treated with EHL for duct clear-
ance.24 Therefore in patients who are medically t, one should aim
for complete ductal clearance of the common bile duct.
Age alone should not be a determining factor limiting patients to
receive endoscopic treatment of choledocholithiasis. From a retro-
spective study, 126 elderly patients (median age 92.2 years) under-
went a total of 159 ERCPs.25 Most of the patients (92.4%) tolerated
the procedures well. In the 68 patients with choledocholithiasis, 59
(86.8%) achieved bile duct stone clearance. There were 3 mild com-
plications (2 cholangitis, 1 hemorrhage) and one fatal complication
Chapter 33 Choledocholithiasis
(post-sphincterotomy bleeding). The mean survival for patients with
non-malignant disease after ERCP was 22.5 months.
Occasionally in patients with suspected cholangitis, no stone is
found at ERCP. It is because a small stone or sludge can be missed
by the cholangiogram, which was reported in 25.5% of the patients
in one study.6 It raises the question of performing empirical endo-
scopic sphincterotomy to ensure biliary drainage, realizing the
potential complications. In a prospective randomized study of 111
patients with cholangitis but without a stone found at cholangiogra-
phy, they were randomized to undergo endoscopic biliary sphincter-
otomy or no endoscopic therapy. In the sphincterotomy group both
duration of fever and length of hospital stay were signicantly
shorter than patients who did not receive sphincterotomy. However,
during follow-up no difference in occurrence of recurrent cholangi-
tis was seen.26 Therefore, routine sphincterotomy does not appear to
be justied in these patients because of potential complications.6
Alternatively, NBD insertion is effective in providing temporary
drainage.19 If follow-up NBD cholangiography conrms absence of
a stone, the patient is then referred for cholecystectomy.
Box 33.1 summarizes the endoscopic management of cholangitis
due to choledocholithiasis.
BOX 33.1 PRACTICAL TIPS IN MANAGING
PATIENTS WITH CHOLANGITIS CAUSED BY
CHOLEDOCHOLITHIASIS
Adequate resuscitation before performing ERCP
Avoid over-sedation of the patient
Avoid over-injection of contrast during bile duct cannulation
Aspirate infected bile once deep cannulation is achieved to
reduce biliary pressure before contrast injection
Avoid performing biliary sphincterotomy and stone
extraction in an unstable patient
Insert a nasobiliary drain or a biliary stent for decompression
in unstable patients
Avoid unnecessary sphincterotomy when NBD or biliary
stent is placed
Delay second stage ERCP to remove stones until stabilization
of the patient
In the stable patient always attempt to achieve complete
stone clearance
Consider surveillance for recurrent choledocholithiasis after
endoscopic clearance in high-risk patients or those who
would poorly tolerate recurrence
Recommend cholecystectomy in healthy patients with
residual gallstones after endoscopic clearance of bile duct
stones
361
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
MANAGEMENT OF DIFFICULT CBD STONES
The standard endoscopic technique for stone removal consists of
biliary sphincterotomy followed by stone extraction with Dormia
basket or balloon. Ductal clearance is achieved in 95% of cases.
However, there are situations where ductal clearance is difcult.
These include stones larger than 2 cm in size, stones impacted in
relatively small or non-dilated common bile duct, and small sphinc-
terotomy because of difcult anatomic position.
For large CBD stone (>2 cm), biliary mechanical lithotripsy (BML)
allows successful ductal clearance in 8090% of cases. Stone impac-
tion in the bile duct is a signicant predictive factor of endoscopic
failure.27 Electrohydraulic lithotripsy (EHL) via peroral endoscopic
choledochoscopy is a safe technique that allows successful manage-
ment of difcult CBD and intrahepatic ductal stones in a high per-
centage of patients (Figs 33.733.8). The fragmentation rate is as
high as 96% and complete stone clearance as high as 90%.28 In a
cohort of 313 patients with difcult CBD stones, 90% achieved
ductal clearance with the use of high-energy extracorporeal shock
wave lithotripsy (ESWL). The success of ESWL was not inuenced
by stone location, size, or presence of bile duct stricture.29 The selec-
tion of specic treatment methods depends on the availability of the
equipment and local expertise. If this equipment is not available the
temporary insertion of a biliary stent may fragment the stone and
aid in future endoscopic treatment (Fig. 33.9).30
SURVEILLANCE AFTER TREATMENT
OF CHOLEDOCHOLITHIASIS
Recurrence of biliary complications after endoscopic and surgical
treatment of choledocholithiasis ranges from 3.7% to 32%. Higher
rates of recurrent choledocholithiasis are seen in patients with an
A B
362
intact gallbladder, dilated CBD, periampullary diverticula, history of
primary bile duct stone, and use of T-tube drainage at the time of
cholecystectomy.31 Patients with a history of recurrent cholangitis
may have a shortened period of remission.32 Therefore, in high-risk
patients it is logical to perform surveillance. In one study of patients
who underwent a follow-up program of serum liver enzymes testing
and US examination every 36 months, recurrent stones were
detected when asymptomatic.33 Therefore, in patients with risk
factors of recurrent choledocholithiasis, multiple episodes of chol-
angitis, and presence of major systemic diseases, a regular surveil-
lance program should be considered. Repeat ERCP and endoscopic
sphincterotomy in patients with prior sphincterotomy may be safe
in expert hands but does not eliminate the risk of recurrent
choledocholithiasis.32
PERI-OPERATIVE MANAGEMENT
OF CHOLEDOCHOLITHIASIS
The management of bile duct stone in the era of laparoscopic cho-
lecystectomy (LC) has been a subject of much debate. The current
options available include preoperative ERCP, intraoperative ERCP,
postoperative ERCP, laparoscopic exploration of the CBD (LECBD)
through a transcystic route or laparoscopic choledochotomy, and
open CBD exploration.
In patients who are at high risk of choledocholithiasis (history of
cholangitis, pancreatitis, deranged liver function, dilated CBD) pre-
operative ERCP should be performed prior to LC. From a retrospec-
tive audit of 1139 patients, 227 (20%) patients were selected for
ERCP examinations based on the above criteria. 53% of the patients
had choledocholithiasis; among them, 97% had successful endo-
scopic stone extraction.34 In a prospective study, 427 patients were
assessed for ERCP before surgery based on the same criteria. Forty-
Fig. 33.7 A Multiple intrahepatic CBD
stones. B Photograph of a mother and
baby scope (choledochoscope) seen
passing through the duodenoscope. A
forceps is pictured exiting the baby scope
working channel which allows extraction of
stone fragments after intraductal lithotripsy.

A baby scope shows the detail of the bile duct
one patients (9.6%) met the criteria and among them 22 patients
(53.7%) were found to have choledocholithiasis. The strongest pre-
dictive factor for CBD stones on ERCP was a dilated CBD in associa-
tion with abnormal serum liver chemistries. During follow-up, 28
patients (6.6%) were found to have recurrent or residual CBD
stones.35 These studies demonstrated that selective use of ERCP
before LC reduces the routine use of time-consuming laparoscopic
cholangiography or LCBDE. However, there are drawbacks in
employing these selection criteria since only half of the patients
selected were conrmed to have choledocholithiasis and hence up
to 50% of patients were subjected to an unnecessary invasive proce-
dure. Thus better methods of selecting patients for ERCP are needed.
Studies have shown that MRCP36 and EUS37 are both useful in select-
ing patients for ERCP before LC. Recently, a decision analysis model
suggested that when the risk of bile duct stones is less than 10%
(low risk), expectant management with postoperative ERCP for
recurrent symptoms is less costly. When the risk of stones is greater
than approximately 55% (high risk), preoperative ERCP is most cost-
effective. When the risk is intermediate (1055%), preoperative
testing with either EUS or MRCP, or intraoperative cholangiography
is the best approach.38
Chapter 33 Choledocholithiasis
Fig. 33.8 A Endoscopic view from the
and the presence of pigment stones. B The
extracted pigment stone.
Occasionally bile duct stones are discovered during LC. A recent
study shows that by using intraoperative ERCP, one can avoid
LCBDE which is technically more demanding. In a two-year prospec-
tive study of 674 patients, 34 (5.7%) patients were found to have a
CBD stone detected during laparoscopic cholangiography and
required intraoperative ERCP.39 Cannulation of the CBD was aided
by surgical passage of a guidewire through the cystic duct and across
the papilla. Successful cannulation was reported in 100% and no
pancreatitis occurred. CBD stones were successfully extracted in
93.5% of patients. The operating time was prolonged but the length
of hospitalization was not increased. This combined procedure may
actually reduce the cost and hospital stay for patients by preventing
two separate procedures. If the stones cannot be cleared during the
intraoperative ERCP the operation can be converted to LECBD or
even open bile duct exploration.40 Postoperative ERCP and stone
removal allows successful clearance in 93100% of patients.3,41
Should the stone be removed during the operation or postoperatively
at ERCP? In a prospective randomized trial, 471 patents underwent
LC and 80 (17%) patients were found to have CBD stones by chol-
angiogram.41 Half of the patients were randomized to receive LECBD
and the other half randomized to postoperative ERCP and stone
363
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
C
D
Fig. 33.9 The use of biliary stent in the treatment of difcult CBD
stone. A Fluoroscopic image of large CBD stone impacted in the
CBD; there was no room in which to open the biliary mechanical
lithotripter (BML). B A biliary stent was inserted. C Three months
later the stone had reduced in size to allow entrapment and crush-
ing using the BML. D Only small fragments remained in the CBD
which were subsequently removed using standard retrieval
techniques.
extraction. The ductal clearance after rst intervention was 75% in
both groups. There was no difference in clinical outcome but the
postoperative hospital stay was signicantly shorter in the laparo-
scopic group than the ERCP group (median 1 day versus 35 days).
The advantage of one-stage laparoscopic approach is also supported
364
by a decision analysis model.42 It showed LCBDE is a cost-effective
method of managing CBD stones found on LC. If expertise
in LECBD is unavailable, then postoperative ERCP should be
offered.
These conclusions from the above studies should be considered
with caution. Despite the reported high success rate of postoperative
ERCP, there are still about 5% of cases in whom complete stone
clearance is not achieved. The patient may then need to undergo
another operation for stone removal. Therefore, patient selection for
preoperative stone extraction is very important. If endoscopic stone
extraction fails, CBD exploration is deemed inevitable. Findings of
large stone size (>25 mm diameter), presence of intrahepatic stones
or multiple tightly packed CBD stones, CBD stricture, duodenal
diverticulum, and history of Bilroth II or Roux-en-Y operation during
LC predicts failure of postoperative ERCP.3 In these situations, open
or laparoscopic CBDE for stone removal seems to be a better
choice.
An algorithm for the management of choledocholithiasis
incorporating endoscopic and laparoscopic treatment had been
proposed.43
CHOLECYSTECTOMY AFTER
ENDOSCOPIC SPHINCTEROTOMY
Endoscopic sphincterotomy and stone extraction have gained wide
acceptance in the management of choledocholithiasis. However, fol-
lowing the endoscopic removal of bile duct stones, the need for
cholecystectomy in patients with concomitant gallstone is less clear.
Various studies had shown that further biliary complications occur
in 424% of patients after varying periods of follow-up and the rate
of subsequent cholecystectomy ranges from 6% to 18%.44,45 The age
and presence of comorbidity of the patients also factors into the
equation.
A randomized, prospective study of 120 patients (mean age
62 years) who underwent ERCP with CBD stone removal followed
by either LC or expectant management showed that recurrent
biliary symptoms, mainly biliary pain and acute cholecystitis,
occurred in 2% of patients in the LC group as compared with 47%
of patients in the expectant group.46 In the expectant group, 37%
of the patients subsequently required cholecystectomy and in over
half of them conversion to open surgery was required. In contrast,
in an Oriental non-randomized study of 140 patients (mean age 67
years) there were no signicant differences in the incidence of
recurrent bile duct stones, biliary symptoms and complications
between those who received elective LC or were managed expec-
tantly.47 The discrepancy in outcome arises from the origin of cho-
ledocholithiasis which is gallstone predominantly from the
gallbladder in the West and pigment stone from the bile ducts in
the East. In a recent study from Hong Kong 178 patients (mean
age 71 years) were randomized to LC or expectant management
after endoscopic sphincterotomy and stone extraction.48 During 5
years of follow-up, 6 patients in the cholecystectomy group returned
with biliary events (cholangitis 5, epigastric pain 1); whereas in
those with gallbladders in situ, 21 patients developed further biliary
events including recurrent choledocholithiasis with cholangitis in
13, epigastric pain in 2, obstructive jaundice in 1, and acute chole-
cystitis in 5. The cumulative probability of recurrent biliary events
in the cholecystectomy group and gallbladders in situ group was

5.8% and 25.4%. It is concluded that after endoscopic sphincterot-
omy and removal of bile duct stones, even in Asian patients, cho-
lecystectomy should be performed to reduce the risk of recurrent
biliary events.
CONCLUSION
Advances in recent imaging techniques have replaced diagnostic
ERCP for choledocholithiasis. ERCP remains the mainstay for
REFERENCES
1. Ralls PW, Jeffrey Jr. RB, Kane RA, et al. Ultrasonography. In:
Yamada T, Alpers DH, eds. Textbook of gastroenterology, 4th
edn. Philadelphia: Lippincott Williams & Wilkins; 2003:3124.
2. Greenberger NJ, Paumgartner G. Diseases of the gallbladder and
bile duct. In: Braunwald E, Fauci AS, Kasper DL, eds. Harrisons
principles of internal medicine, 15th edn. New York: McGraw-Hill;
2001:1785.
3. Park AE, Mastrangelo MJ Jr. Endoscopic retrograde cholangio-
pancreatography in the management of choledocholithiasis. Surg
Endosc 2000; 14(3):219226.
4. Frossard JL, Hadengue A, Amouyal G, et al. Choledocholithiasis: a
prospective study of spontaneous common bile duct stone
migration. Gastrointest Endosc 2000; 51(2):175179.
5. Das A, Isenberg G, Wong RC, et al. Wire-guided intraductal US: an
adjunct to ERCP in the management of bile duct stones.
Gastrointest Endosc. 2001; 54(2):3136.
6. Siddique I, Mohan K, Khajah A, et al. Sphincterotomy in patients
with gallstones, elevated LFTs and a normal CBD on ERCP.
Hepatogastroenterology 2003; 50(53):12421245.
7. Soto JA, Alvarez O, Munera F, et al. Diagnosing bile duct stones:
comparison of unenhanced helical CT, oral contrast-enhanced CT
cholangiography, and MR cholangiography. AJR Am J Roentgenol
2000; 175(4):11271134.
8. Romagnuolo J, Bardou M, Rahme E, et al. Magnetic resonance
cholangiopancreatography: a meta-analysis of test performance
in suspected biliary disease. Ann Intern Med 2003;
139(7):547557.
9. Kaltenthaler E, Vergel YB, Chilcott J, et al. A systematic review and
economic evaluation of magnetic resonance
cholangiopancreatography compared with diagnostic endoscopic
retrograde cholangiopancreatography. Health Technol Assess
2004; 8(10):iii,189.
10. Aube C, Delorme B, Yzet T, et al. MR cholangiopancreatography
versus endoscopic sonography in suspected common bile duct
lithiasis: a prospective, comparative study. AJR Am J Roentgenol
2005; 184(1):5562.
11. Romagnuolo J, Currie G. Calgary Advanced Therapeutic
Endoscopy Center study group. Noninvasive vs. selective invasive
biliary imaging for acute biliary pancreatitis: an economic
evaluation by using decision tree analysis. Gastrointest Endosc
2005; 61(1):8697.
12. Buscarini E, Tansini P, Vallisa D, et al. EUS for suspected
choledocholithiasis: do benets outweigh costs? A prospective,
controlled study. Gastrointest Endosc 2003; 57(4):510518.
13. Sung JY, Shaffer EA, Olson ME, et al. Bacterial invasion of the
biliary system by way of the portal-venous system. Hepatology
1991; 14(2):313317.
14. Sung JJ, Lyon DJ, Suen R, et al. Intravenous ciprooxacin as
treatment for patients with acute suppurative cholangitis: a
Chapter 33 Choledocholithiasis
treatment of choledocholithiasis. Advances in therapeutic
ERCP have resulted in successful clearance rates of choledo-
cholithiasis and improved patient outcome. Endoscopic and
laparoscopic treatments of bile duct stones should be con-
sidered complementary approaches to the management of
choledocholithiasis.
randomized, controlled clinical trial. J Antimicrob Chemother
1995; 35(6):855864.
15. Lau JY, Chung SC, Leung JW, et al. Endoscopic drainage
aborts endotoxaemia in acute cholangitis. Br J Surg 1996;
83(2):181184.
16. Leung JW, Chung SC, Sung JJ, et al. Urgent endoscopic drainage
for acute suppurative cholangitis. Lancet 1989; 1(8650):13071309.
17. Lai EC, Mok FP, Tan ES, et al. Endoscopic biliary drainage for
severe acute cholangitis. N Engl J Med 1992; 326(24):15821586.
18. Garca-Cano J. Success rate for complete choledocholithiasis
extraction by means of endoscopic retrograde
cholangiopancreatography. Surg Endosc 2004; 18(11):16811682.
19. Sugiyama M, Atomi Y. The benets of endoscopic nasobiliary
drainage without sphincterotomy for acute cholangitis. Am J
Gastroenterol 1998; 93(11):20652068.
20. Hui CK, Lai KC, Yuen MF, et al. Does the addition of endoscopic
sphincterotomy to stent insertion improve drainage of the bile
duct in acute suppurative cholangitis? Gastrointest Endosc 2003;
58(4):500504.
21. Lee DW, Chan AC, Lam YH, et al. Biliary decompression by
nasobiliary catheter or biliary stent in acute suppurative
cholangitis: a prospective randomized trial. Gastrointest Endosc
2002; 56(3):361365.
22. Sharma BC, Kumar R, Agarwal N, et al. Endoscopic biliary
drainage by nasobiliary drain or by stent placement in patients
with acute cholangitis. Endoscopy 2005; 37(5):439443.
23. Chopra KB, Peters RA, OToole PA, et al. Randomised study of
endoscopic biliary endoprosthesis versus duct clearance for
bileduct stones in high-risk patients. Lancet 1996;
348(9030):791793.
24. Hui CK, Lai KC, Ng M, et al. Retained common bile duct stones: a
comparison between biliary stenting and complete clearance of
stones by electrohydraulic lithotripsy. Aliment Pharmacol Ther
2003; 17(2):289296.
25. Rodriguez-Gonzalez FJ, Naranjo-Rodriguez A, Mata-Tapia I, et al.
ERCP in patients 90 years of age and older. Gastrointest Endosc
2003; 58(2):220225.
26. Hui CK, Lai KC, Wong WM, et al. A randomised controlled trial of
endoscopic sphincterotomy in acute cholangitis without common
bile duct stones. Gut 2002; 51(2):245247.
27. Garg PK, Tandon RK, Ahuja V, et al. Predictors of unsuccessful
mechanical lithotripsy and endoscopic clearance of large bile
duct stones. Gastrointest Endosc 2004; 59(6):601605.
28. Arya N, Nelles SE, Haber GB, et al. Electrohydraulic lithotripsy in
111 patients: a safe and effective therapy for difcult bile duct
stones. Am J Gastroenterol 2004; 99(12):23302334.
29. Sackmann M, Holl J, Sauter GH, et al. Extracorporeal shock wave
lithotripsy for clearance of bile duct stones resistant to
endoscopic extraction. Gastrointest Endosc 2001; 53(1):2732.
365
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
30. Chan AC, Ng EK, Chung SC, et al. Common bile duct stones
become smaller after endoscopic biliary stenting. Endoscopy
1998; 30(4):356359.
31. Uchiyama K, Onishi H, Tani M, et al. Long-term prognosis after
treatment of patients with choledocholithiasis. Ann Surg 2003;
238(1):97102.
32. Sugiyama M, Suzuki Y, Abe N, et al. Endoscopic retreatment of
recurrent choledocholithiasis after sphincterotomy. Gut 2004;
53(12):18561859.
33. Lai KH, Lo GH, Lin CK, et al. Do patients with recurrent
choledocholithiasis after endoscopic sphincterotomy benet from
regular follow-up? Gastrointest Endosc 2002; 55(4):523526.
34. Coppola R, Riccioni ME, Ciletti S, et al. Selective use of endoscopic
retrograde cholangiopancreatography to facilitate laparoscopic
cholecystectomy without cholangiography. A review of 1139
consecutive cases. Surg Endosc 2001; 15(10):12131216.
35. Katz D, Nikfarjam M, Sfakiotaki A, et al. Selective endoscopic
cholangiography for the detection of common bile duct
stones in patients with cholelithiasis. Endoscopy 2004;
36(12):10451049.
36. Laokpessi A, Bouillet P, Sautereau D, et al. Value of magnetic
resonance cholangiography in the preoperative diagnosis of
common bile duct stones. Am J Gastroenterol 2001;
96(8):23542359.
37. Palazzo L, OToole D. EUS in common bile duct stones.
Gastrointest Endosc 2002; 56(4 Suppl):S49S57.
38. Sahai AV, Mauldin PD, Marsi V, et al. Bile duct stones and
laparoscopic cholecystectomy: a decision analysis to assess the
roles of intraoperative cholangiography, EUS, and ERCP.
Gastrointest Endosc 1999; 49(3 Pt 1):334343.
39. Enochsson L, Lindberg B, Swahn F, et al. Intraoperative
endoscopic retrograde cholangiopancreatography (ERCP) to
remove common bile duct stones during routine laparoscopic
366
cholecystectomy does not prolong hospitalization: a 2-year
experience. Surg Endosc 2004; 18(3):367371.
40. Saccomani G, Durante V, Magnolia MR, et al. Combined
endoscopic treatment for cholelithiasis associated with
choledocholithiasis. Surg Endosc 2005; 19(7):910914.
41. Rhodes M, Sussman L, Cohen L, et al. Randomised trial of
laparoscopic exploration of common bile duct versus
postoperative endoscopic retrograde cholangiography for
common bile duct stones. Lancet 1998; 351(9097):159161.
42. Urbach DR, Khajanchee YS, Jobe BA, et al. Cost-effective
management of common bile duct stones: a decision analysis of
the use of endoscopic retrograde cholangiopancreatography
(ERCP), intraoperative cholangiography, and laparoscopic bile
duct exploration. Surg Endosc 2001; 15(1):413.
43. Lilly MC, Arregui ME. A balanced approach to choledocholithiasis.
Surg Endosc 2001; 15(5):467472.
44. Hill J, Martin DF, Tweedle DE. Risks of leaving the gallbladder in
situ after endoscopic sphincterotomy for bile duct stones. Br J
Surg 1991; 78(5):554557.
45. Hansell DT, Millar MA, Murray WR, et al. Endoscopic
sphincterotomy for bile duct stones in patients with intact
gallbladders. Br J Surg 1989; 76(8):856858.
46. Boerma D, Rauws EA, Keulemans YC, et al. Wait-and-see policy or
laparoscopic cholecystectomy after endoscopic sphincterotomy
for bile-duct stones: a randomised trial. Lancet 2002;
360(9335):761765.
47. Lai KH, Lin LF, Lo GH, et al. Does cholecystectomy after
endoscopic sphincterotomy prevent the recurrence of biliary
complications? Gastrointest Endosc 1999; 49(4 Pt 1):483487.
48. Lau JY, Leow CK, Fung TM, et al. Cholecystectomy or Gallbladder
in situ after endoscopic sphincterotomy and bile duct stone
removal in Chinese patients. Gastroenterology 2006; 130(1):
96103.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Pancreaticobiliary Pain and
34 Suspected SOD
Paul R. Tarnasky and Robert H. Hawes
INTRODUCTION
The diagnosis and treatment of suspected sphincter of Oddi dys-
function (SOD) presents a signicant challenge for physicians who
care for patients with digestive diseases. This chapter is intended to
provide readers with a practical guide to the evaluation and manage-
ment of patients with pancreaticobiliary type pain and suspected
SOD. The overall goals of this chapter include identifying the chal-
lenges which these patients present and offering a pragmatic
approach to the clinical evaluation and decisions regarding treat-
ment. The specic goals are: (1) describe pain patterns that are con-
sistent and not consistent with SOD; (2) dene SOD and the clinical
scenarios where SOD might be considered; (3) describe a rational
initial evaluation for patients with suspected SOD; (4) provide guid-
ance for patient and physician decisions regarding management of
SOD; (5) describe techniques of sphincter of Oddi manometry
(SOM) and endoscopic treatment of SOD; and (6) reinforce the risks
inherent to the endoscopic evaluation of SOD and how they can be
minimized. It should be emphasized that there is a paucity of good
data to guide clinicians in this arena. When data is available, recom-
mendations will be evidence based but much of the following infor-
mation is derived from anecdotal experience of which the authors
have a considerable amount.
Clinical syndromes which may be attributed to SOD range from
functional disorders with purely subjective symptomatology to struc-
tural disorders having objective pathologic features. Functional and
structural SOD diagnoses are widely divergent with regard to their
presentation and management. Unexplained upper abdominal pain
and acute pancreatitis represent the two most important examples
at each end of this spectrum and will be the focus of this review.
Other clinical scenarios that may be associated with SOD include
chronic acalculous cholecystitis, early chronic pancreatitis, biliary
pancreatitis, postoperative bile leak, and pancreatic stula.
DEFINITIONS
Confusing terminology and varied clinical presentations explain part
of the complexity regarding SOD. Biliary dyskinesia is the encom-
passing term for a group of disorders with acalculous biliary-type
pain. Subgroup diagnoses include chronic acalculous cholecystitis,
gallbladder dyskinesia, cystic duct syndrome, and SOD. Sphincter
of Oddi dysfunction may occur in patients with or without a gall-
bladder but is most commonly diagnosed in patients with post-
cholecystectomy symptoms.
Attempts have been made to develop consensus on dening the
signs and symptoms of SOD culminating in what are called the
Rome criteria.1 Denitions established for post cholecystectomy
patients and those with gallbladder in situ are listed in Tables 34.1
and 34.2. Revisions in the Rome criteria for SOD were recently
published.2 The Rome criteria are meant to provide a general frame-
work for clinicians but obviously do not describe all patients. A uni-
fying symptom, present in all patients with SOD, is pain. There may
be associated symptoms such as nausea with or without vomiting
but the hallmark symptom is painlocated in the epigastrium and/
or right upper quadrant (RUQ). When evaluating a patient with pos-
sible SOD, the most important aspect of the evaluation is the history.
It is imperative that the clinician gain a clear understanding of the
nature, location and timing of pain. The Rome criteria specify that
the pain should be intermittent with pain-free intervals. This is a
very controversial point. While biliary pain is typically intermittent,
in some cases, patients will have a constant, low-grade discomfort
with exacerbations. This can be seen particularly in those with pan-
creatic sphincter hypertension who typically have exacerbations after
eating. These patients should undergo careful review and extensive
evaluation for other causes of pain (Table 34.3) but should not be
excluded from evaluation for SOD based solely on there being a
constant component to their pain. However, if associated symptoms
such as nausea, vomiting, abdominal distention or bowel dysfunc-
tion are dominant, then the patient likely does not have SOD as the
predominant explanation for their symptoms.
Based on observations and after developing correlations be-
tween patients presentation and outcomes after endoscopic sphinc-
terotomy, Joseph Geenen, Walter Hogan, and Wylie Dodds
published what have become to be known as The Geenen-Hogan
Criteria (Table 34.4).3 These have been modied over the years
but still serve as a very good compass to clinicians to direct them
in their evaluation and therapeutic decision making. The original
criteria were applied to patients who had previously undergone
cholecystectomy and were based on three factors which could be
assessed without ERCP-presence of typical pancreatic or biliary
type pain, the presence or absence of elevated liver or pancreatic tests
during or shortly following an episode of pain, and the presence
or absence of bile and/or pancreatic duct dilation. The original cri-
teria also included measurement of pancreatic and biliary drainage
times. Drainage times are very imprecise, require instillation of
contrast into the respective duct and in the case of biliary drainage
times, the endoscope must be withdrawn, the patient placed in the
supine position and an abdominal lm obtained at 45 minutes.
Studies have shown that drainage times do not correlate with
SOM4 and delayed drainage is common in asymptomatic post-
cholecystectomy volunteers.5 As a result, drainage times are no
longer performed and are not part of the current Geenen-Hogan
(G-H) criteria.
367
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Post cholecystectomy
Episodes of severe steady pain located in the epigastrium and
right upper quadrant and all of the following:
1. Episodes lasting 30 minutes or longer
2. Recurrent symptoms occurring at different intervals and not
daily
3. The pain is steady, interrupts daily activity and/or leads to
medical encounter
4. The pain is not relieved by bowel movements, postural
change, or antacids
5. Structural diseases that could explain symptoms are excluded
Table 34.1 Rome criteria for sphincter of Oddi dysfunction
Gallbladder in situ
Episodes of severe steady pain located in the epigastrium and
right upper quadrant and all of above criteria plus:
Normal liver and pancreas chemistries
Absence of gallbladder stones, sludge, or microlithiasis
Abnormal gallbladder emptying
Table 34.2 Rome criteria for sphincter of Oddi dysfunction
The G-H criteria are important because they represent a frame-
work around which a clinician can plan patient evaluation. If one
obtains an appropriate history of pain, bile duct imaging should be
obtained and the patient should be given a prescription directing
health care providers (in an emergency room, hospital lab or clinic)
to obtain liver and pancreatic tests (amylase and lipase) during or
shortly after a pain episode. These data then can be used to stratify
patients as to their likelihood of having SOD.
CLINICAL EVALUATION
The rst step is a detailed review of prior health care encounters
pertinent to the clinical presentation with a focus on questions of
when, where, and what (Table 34.5). A complete history and thor-
ough review of records will dene the clinical symptoms, reveal what
tests have been done, what treatments (surgical, endoscopic, medical)
have been tried, and what the impact has been on the patient.
Patients with unexplained symptoms that may be attributed to SOD
often end up undergoing a massive assault with both diagnostic
and therapeutic fronts. It can be helpful to organize objective data
with regards to prior laboratory testing, imaging and treatments
(Table 34.6).

Esophageal When did the attacks begin?

Spasm or other motility disorder When do the attacks occur?
Esophagitis Where is the pain?
Gastric Where does the pain radiate?
Gastroparesis What is associated with the attacks?
Ulcer What has been done to investigate the cause?
Hiatal hernia What has been done to treat the attacks?
Volvulus What are the consequences of the attacks?
Pyloric stenosis
Duodenal Table 34.5 Important history questions for suspected SOD
Stricture
Ulcer
Diverticulitis
Ampullary neoplasm Laboratory and pathology
Biliary Serum liver and pancreas chemistries
Stone Serum fasting triglyceride
Benign stricture Gallbladder pathology
Sump syndrome Imaging
Neoplasm Transabdominal ultrasound
Pancreatic Computed tomography
Chronic pancreatitis Magnetic resonance with MRCP
Neoplasm Biliary scintigraphy
Abdominal Wall Endoscopic ultrasound
Neuroma Intraoperative cholangiography
Myopathy/myositis Previous Treatment
Irritable bowel syndrome Surgical
Cholecystectomy
Table 34.3 Diagnoses to consider (other than SOD) for Biliary bypass
unexplained upper abdominal pain Pseudocyst drainage
Pancreatic bypass or resection
Partial gastrectomy
Typical pain LFT > 2X Normal 2 BD diam > 10 mm Gastric bypass
Type I + + + Endoscopic
Type II + + or + Biliary sphincterotomy
Type III + Pancreatic sphincterotomy
Stenting
Table 34.4 Geenen-Hogan classication for SOD
LFT = Liver function tests. Table 34.6 Clinical details pertinent to sphincter of Oddi
BD = Bile duct. dysfunction

368

Type I II
Denition Pain + 3 criteria*
Baseline pressure >40 mmHg 70100%
Benet from sphincterotomy 5591%
Table 34.7 Correlation between Geenen/Hogan Criteria, results of SOM and outcome with sphincterotomy
The criteria used were the following:
1. ALT and alkaline phosphatase over twice upper limit of normal.
2. Dilated bile duct on sonography.
3. Delayed drainage of contrast material at ERCP.
Some historical details may indicate that SOD is likely. It is not
uncommon for SOD patients to have undergone cholecystectomy
because of a diseased or dysfunctional gallbladder. Patients with a
history of chronic narcotic analgesic use who then develop pancre-
aticobiliary pain often have SOD. Symptomatic patients who have a
history of common bile duct exploration, postoperative bile leak
and/or post-ERCP pancreatitis are often discovered to have SOD.
Pain is a subjective complaint. Nevertheless, considerable infor-
mation can be obtained. There are a number of classic descriptors
that can help guide whether or not SOD is a likely cause for pain.
Typical pancreatic/biliary pain occurs intermittently, begins after
meals, and lasts minutes to hours. It is located in the epigastric or
right upper quadrant areas and may radiate to the back, chest or
right shoulder. Occasionally, the pain is perceived rst in the back
or chest. Daily pain that is constant is not typical for SOD unless
associated with chronic pancreatitis. Patients may be awakened from
sleep because of pain. It is not uncommon for patients to describe
their symptoms as my gallbladder pain and even describe symp-
toms that are worse than my gallbladder attack. Transient eleva-
tions of serum liver and/or pancreas enzymes drawn hours after
pain onset may suggest SOD.
The possibility of more common and potentially more treatable
diagnoses should be considered before proceeding with an evalua-
tion for possible SOD. Symptom history and diagnostic testing
should be directed at evaluation for the potential diagnoses listed in
Table 34.3. For example, bile duct dilation should raise a suspicion
for neoplasia or bile duct stones if associated with abnormal liver
tests. Alternatively, a dilated bile duct with normal liver tests in a
patient with intermittent pain should raise suspicion for SOD. Eval-
uation for possible common bile duct stones deserves careful con-
sideration. Bile duct stones are very rarely found when routine
imaging tests such as transabdominal ultrasound and laboratory
testing are normal. Therefore, unless there are objective indicators
to suggest bile duct pathology, ERCP should be avoided when purely
used to rule out bile duct stones. Additional imaging such as
MRCP or EUS can be helpful in this setting. It is most reasonable
to consider ERCP when SOM and/or denitive endoscopic therapy
is planned.
Ideally, patients with unexplained upper abdominal pain can be
categorized as to the likelihood for SOD and a favorable response to
endoscopic treatment. The Geenen-Hogan classication (Table 34.4)
is the standard in this regard. Type I SOD patients have objective
evidence of impaired drainage and are more likely to have structural
obstruction (papillary stenosis). In addition to characteristic pain,
they have dilated duct(s) and abnormal liver tests during episodes
of pain.
Patients with Type II SOD will have characteristic pain and either
a dilated duct or abnormal laboratory tests with pain. Type III SOD
Chapter 34 Pancreaticobiliary Pain and Suspected SOD
III
Pain + 1 or 2 criteria* Pain only
4086% 2055%
p > 40 mmHg: 8090% p > 40 mmHg: 856%
p < 40 mmHg: 3035%
patients have typical biliary or pancreatic pain but no objective evi-
dence of impaired drainage. Such patients likely have a purely func-
tional disorder. The reason that this categorization of patients is
important is that it predicts, to a certain extent, the chance of nding
an abnormal SOM and having a favorable outcome following sphinc-
terotomy (Table 34.7).6
Remarkably, disease and/or interventions of such a small struc-
ture may lead to a tremendous burden on behalf of the patient and
their physician. From the patient standpoint, SOD may present a
wide spectrum of physical and emotional symptoms ranging from
nuisance to total disability. Much of the emotional burden is derived
from uncertainty. Patients become desperate when not knowing the
cause of their symptoms; whether and when they will have future
attacks, and if there are safe and effective treatments. Diverse chal-
lenges which face physicians include substantial time-requirements,
potential legal ramications, and the broad range of necessary skills
such as history taking, record-keeping, radiology interpretation, and
psychological assessment. Moreover, physicians who decide on
doing ERCP in this setting need to possess appropriate technical
skills such as sphincter manometry, selective cannulation, sphinc-
terotomy (perhaps precut), and pancreatic therapy. Compassion and
judgment are the intangible physician qualities that are more impor-
tant than knowing how to cut a sphincter or place a stent. These
qualities are tested when faced with the often asked question: What
would you do if I was your mother or daughter or . . .?
Once a clinical impression of SOD is established, ideally a non-
invasive test would be available to conrm ones clinical impression
before proceeding to ERCP. Several tests have been studied and
individual centers have reported good correlation with SOM and/or
sphincterotomy. The problem is that when these tests are evaluated
on a broader scale, their accuracy does not match previous, single
center reports. The Hopkins group rst reported on the accuracy of
dynamic (quantitative) biliary scintigraphy.78 The test was designed
to measure delayed bile ow through the ampulla by assessing the
time it takes for the radionuclide to reach the duodenum. These
authors found a good correlation with SOM. Their results were sup-
ported by Corazziari, et al.9 This prompted the Hopkins group to
suggest that this test could substitute for SOM.10 However, when
this test was evaluated in normal volunteers, we found it had very
poor specicity and it had little value in excluding SOD in patients
suspected to suffer from this disorder.11
Another test hypothesized to detect SOD is fatty meal ultrasound
(FMS). An abnormal test is dened by a >2 mm dilation of the bile
duct 45 minutes after ingestion of a standardized fatty meal.
Rosenblatt et al. compared SOM, FMS and hepatobiliary scintigra-
phy (HBS) in a retrospective comparative study.12 Poor correlation
was observed between FMS and HBS with SOM. However, of the
patients with abnormal SOM who had a good long-term response to
369
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
sphincterotomy, 85% (11/13) had an abnormal FMS and HBS. This
raises an interesting point; perhaps non-invasive tests should be
evaluated as to whether they predict response to sphincterotomy
rather than whether they correlate with SOM. What a clinician really
wants to know from a non-invasive test is whether or not the patient
will respond to endoscopic sphincterotomy.
UPPER ABDOMINAL PAIN WITH GALLBLADDER
IN SITU
Management of patients with biliary-type pain without evidence of
gallstones on standard imaging represents a challenge. Physicians
(including surgeons) and patients usually prefer to identify some
proof of gallbladder pathology before considering cholecystectomy.
Biliary crystal analysis can be performed on bile collected from the
duodenum or bile duct after cholecystokinin (CCK) stimulation.
Endoscopic ultrasound is more sensitive for discovering biliary
sludge1314 and can also be used to assess for evidence of pancreatitis.
If EUS and CCK stimulated biliary drainage are performed and
biliary crystals or gallbladder sludge are found, >90% of patients will
have resolution of pain with cholecystectomy.15 Biliary scintigraphy
may reveal evidence of chronic acalculous cholecystitis (gallbladder
ejection fraction <35%).16 Empiric cholecystectomy, however, will
benet about three-fourths of those patients with classic biliary pain,
independent of other testing.1721
The exact role for SOM in this setting is not established. There
has been limited study of the prevalence of sphincter of Oddi dys-
function (SOD) in patients with gallbladder in situ. Guelrud reported
on 121 patients with biliary pain and a nding of gallstones but a
normal caliber bile duct by ultrasound.22 ERCP and sphincter of
Oddi manometry was performed and he found elevated basal sphinc-
ter pressures in 14 patients (11.6%). Interestingly, 4% of patients in
this group with a normal alkaline phosphatase had elevated basal
sphincter pressures while 40% with an elevated alkaline phospha-
tase were found to have SOD. Ruffolo et al. investigated 81 patients
with typical biliary-type pain and a normal gallbladder ultrasound.23
When ERCP and sphincter of Oddi manometry was performed, 53%
of these patients had sphincter of Oddi dysfunction as diagnosed by
elevated basal sphincter pressures. For the whole group, 49% had
an abnormal ejection fraction on gallbladder scintigraphy but the
nding of sphincter of Oddi dysfunction did not correlate with ejec-
tion fraction. All patients in this group with elevated sphincter pres-
sures underwent biliary sphincterotomy and the short-term results
of pain relief (one year) were quite good. However, with longer-term
follow-up, most patients ultimately required cholecystectomy.24
Our approach is to avoid SOM in patients with gallbladder in situ
because laparoscopic cholecystectomy is safer than ERCP. Also,
SOM testing may be misleading in this setting. While sphincter of
Oddi dyskinesia (tachyoddia) may be detected, basal sphincter pres-
sure may still be normal in patients without prior upper abdominal
surgery because inhibitory neurons to the sphincter remain intact.
However, ERCP with SOM may be reasonable where typical biliary
pain is accompanied by transient elevations of liver enzymes.
UNEXPLAINED ACUTE PANCREATITIS
The diagnosis of acute pancreatitis is usually straightforward. Pan-
creatic pain is most often epigastric and radiates through to the back.
Serum amylase and/or lipase should be three times normal before
370
making a diagnosis of acute pancreatitis. In situations where there
are equivocal laboratory results or when labs were drawn many
hours after the onset of pain, a diagnosis can be made if there is
radiographic evidence of pancreatitis.
A thorough review on the approach to acute pancreatitis is covered
elsewhere is this book. Sphincter of Oddi dysfunction is considered
along with other structural causes (microlithiasis, neoplasia,
pancreas divisum, duodenal diverticulum, choledochocele) when
more common causes (alcohol, gallstones, medications) have been
excluded.
Evaluation with ERCP and SOM should be considered when a
patient has had either recurrent attacks or at least one attack that
was considered to be severe. Patients of 50 years and older should
also undergo endoscopic evaluation in order to exclude neoplasia.
INFORMED CONSENT FOR ERCP FOR
SUSPECTED SOD
Most important to any discussion on suspected SOD is informed
consent. Proper informed consent before ERCP in a patient with
suspected SOD is in itself a complex venture. Informed means
that both the physician and patient have a thorough understanding
of the clinical situation before determining the potential for risk and
benet. The physicians role is largely limited to acquiring and
sharing information. Information that is relayed to the patient
includes review of relevant data (if any) regarding efcacy and safety
of endoscopic treatment. It may be just as important to disclose the
fact that there is very little efcacy data. Fortunately, we do have
reasonable data regarding complications of ERCP in suspected SOD
(see below). A physician should share their own complication data
that is specic for that situation. For example it is inappropriate to
either state complication data from other endoscopists or that from
other clinical situations e.g. bile duct stones. At the present time, it
remains particularly important that patients with purely functional
symptoms who have no objective evidence of digestive duct obstruc-
tion understand that they are making a benet/risk decision that
pertains to a quality of life problem. Patients should understand that
they are ultimately responsible for giving their consent and that they
should not solely depend on physician advice.
SPHINCTER OF ODDI MANOMETRY
Equipment
Traditionally, sphincter of Oddi manometry has been performed
using a water perfused, low-compliance pneumohydraulic system.
This is the same system that was originally used for esophageal
manometry. However, unlike esophageal manometry, which is now
undertaken primarily with electronic systems, almost all centers
performing sphincter of Oddi manometry, continue to use the water
perfused system because:
1. All of the normal data has been generated with a water perfused
system.
2. The electronic manometry catheters are expensive and fragile.
Advances have been made in water perfused systems, particularly in
the software, which makes set-up, recording and interpreting
the manometry much easier. These systems are available through
Sandhill (Sandhill Scientic, Inc., Highlands Ranch, Colorado) and
Medtronic (Minneapolis, Minnesota), as well as other manufac-

Fig. 34.1 Distal tip of SOM catheter. Top arrow points to guide-
wire. Bottom arrow points to distal side hole.
turers. The entire system consisting of the computer and the water
perfusion system, can be placed on a small cart and is readily
mobile.
The original catheter used for sphincter of Oddi manometry was
manufactured by Arndorfer (Arndorfer Inc., Greendale, Wisconsin)
and some practitioners still use these catheters. However, the major-
ity of catheters used in sphincter of Oddi manometry are manufac-
tured by Cook GI Endoscopy (Winston-Salem, NC). The catheter
consists of three lumens, two of which terminate in a side hole of
the catheter while the third lumen has both a side port as well as an
end port (Fig. 34.1). The lumen with the end port does accommodate
a 0.018 guidewire. All three channels can be used for the manom-
etry recording but a randomized study showed that sacricing the
third lumen with the side and end port, and using that for aspiration
during a pancreatic manometry, signicantly reduced the post-
manometry pancreatitis rate.25 It was found that aspiration during
manometry of the biliary sphincter was not necessary.26 Water is
perfused at 0.25 ml per minute through each port. The triple lumen
manometry catheter made by Cook GI Endoscopy consists of Teon
and is tapered at the end. In the distal end of the catheter, there are
black rings spaced 1 mm apart. Moving proximally to distally, there
are seven black rings followed by a red ring, a black ring and another
red ring in sequence. The rings allow communication between the
endoscopists and the manometry assistant to record the position of
the catheter relative to the papilla orice. The proximal end of the
catheter (that portion that is outside the scope channel) is bolstered
by an additional plastic coating that helps stiffen the catheter and
prevents kinking as the catheter is inserted and withdrawn. The
catheter is supplied in two typesthe so-called short nose and
long nose. The short-nose catheter has 5 mm between the last
black ring and the tip of the catheter. The length of the distal tip on
the long nose catheter is 20 mm. The main advantage of the long
nose catheter is that the manometry can be completed (withdrawn
to the last ring) while maintaining the cannulation. The downside
of this catheter, in the opinion of these authors, is that the long nose
catheter is harder to cannulate with.
In summary, the equipment to perform sphincter of Oddi
manometry is reasonably uncomplicated. The computer and water
Chapter 34 Pancreaticobiliary Pain and Suspected SOD
perfused systems are readily available as are the catheters used to
perform SOM. The equipment can be contained on a small mobile
cart which can be wheeled into and out of the ERCP suite easily. It
takes a matter of only a few minutes to set up the system and thus
is not an onerous addition to an ERCP case.
Technique
The technique of sphincter of Oddi manometry is relatively straight-
forward. It consists of deep cannulation followed by a slow sta-
tioned withdrawal.
There are two basic cannulation techniques employed when
performing sphincter of Oddi manometry. The rst is called
the kissing technique (video). Here, the manometry catheter is
advanced a short distance into the visual eld and the elevator is
moved maximally up. The endoscopists left thumb is positioned
on the up/down knob with the right hand grasping the shaft of the
scope. Using the right hand, the scope is inserted or withdrawn as
necessary to allow the manometry catheter to be inserted into the
papillary orice using the up dial of the endoscope. Once the tip
of the catheter is seated in the papillary orice, the manometry
catheter is slowly advanced until slight resistance is met. Now, with
a greater amount of catheter beyond the elevator, the shaft of the
endoscope is then regrasped with the right hand and then deep can-
nulation is accomplished by varying the trajectory of the catheter
using the up/down dial and driving it forward into the duct by with-
drawing the scope.
The second method of cannulation relies on the more standard
approach of advancing the catheter into the papillary orice using
the elevator (video). However, once the tip of the catheter is seated
into the ampullary orice (a step that we call insinuation), the most
effective way to advance the catheter deeply is to withdraw the scope.
The most common mistake in trying to achieve deep cannulation is
to simply advance the catheter. The natural curve of the catheter will
cause the tip to be driven into the roof of the papilla with this tech-
nique and one usually fails to get deep cannulation. One must
remember that whatever technique is used to advance the catheter,
it must be advanced directly in line with the trajectory of the duct.
As mentioned earlier, this is usually best achieved by withdrawing
the scope and adjusting the up/down of the scope tip. Another
common problem is the natural mucosal folds that are usually
present within the intraduodenal segment of the ampulla. The cath-
eter tip is commonly driven into these folds and trying to forcibly
advance the catheter against these folds usually fails and can result
in trauma and even tearing of the ampullary mucosa. The most
effective way to get around these folds is to advance the scope very
slightly which results in backing the catheter out of the papilla. Once
the catheter has been withdrawn, the tip should be redirected and
then the scope withdrawn which will drive the catheter into the
papilla hopefully this time avoiding the folds. One should remember
that the ampulla is never so tight that brute force is needed to can-
nulate. It is virtually always an issue of achieving the proper tip
trajectory.
There are two general approaches to cannulation when one is
planning to perform sphincter of Oddi manometry. One rst
achieves cannulation using a standard accessory such as a diagnostic
cannula or a sphincterotome. Once cannulation is achieved, a chol-
angiogram or pancreatogram is performed and if no pathology is
seen (stone or stricture) a 0.018 inch guidewire is passed and then
the manometry catheter is advanced over the guidewire. This is
usually employed at centers that do not do a lot of sphincter of Oddi
371
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
manometry and who wish to perform manometry of the biliary
sphincter only. The technique that we advocate is to perform free
cannulation as the initial step in the ERCP with the manometry
catheter alone. Our reasoning for this approach is:
1. With good technique and experience, free cannulation with a
manometry catheter in one or both ducts can be achieved in over
90% of the time. Attempted free cannulation of course does not
prohibit the use of a standard catheter if this technique fails.
2. Cannulation with a standard catheter is no easier than with
a manometry catheter and adds additional steps to the proce-
dure. A guidewire exchange then must be performed. The best
guidewire to use in conjunction with sphincter of Oddi man-
ometry is the Nitinol based 0.018 wire. This particular wire
does not aid in cannulation as its very thin, soft and oppy tip
easily gets caught on the mucosal fold within the ampullary
segment.
3. The manometry catheter can be used to obtain a cholangiogram
by infusing contrast through the aspiration port.
Before recording the sphincter pressures, one must obtain a
duodenal baseline pressure. This can be achieved two ways. The
technique that we utilize (and is utilized in most centers) is to
advance the entire tip of the manometry catheter into the duodenal
lumen. The catheter is then perfused and a baseline pressure is
established. During this baseline, the catheter should not come in
contact with the duodenal wall. A second method for obtaining
duodenal baseline achieves a continuous recording of the duode-
num. With this technique, a separate manometry catheter is taped
to the side of the duodenoscope and the duodenal baseline is
recorded continuously throughout the manometry.
We do not make any attempt to select a particular duct as we
begin sphincter of Oddi manometry. We perfuse only two of the
three channels and use the third channel for aspiration. Once the
manometry catheter is deeply in the duct, gentle aspiration is applied
to the auxiliary channel (video). If a clear uid is withdrawn, then
it is known that we are in the pancreatic duct. If yellow uid is seen,
then this identies the bile duct. We then perform a standard sta-
tioned pull-back. In a stationed pull-back, we withdraw the catheter
until the rst black ring is identied and at this point, we notify the
manometry assistant that we are at one black. The manometry
assistant then informs the endoscopists as to whether or not any
phasic contractions are seen. If not, then the endoscopist withdraws
the catheter to two black. Again, this position is held to determine
if any phasic waves are identied. This process of withdrawal of one
black ring at a time continues until typical phasic contractions are
identied. Once phasic contractions are identied, it is determined
whether the nadir of the phasic contractions dips below 40 mm Hg
or not. If so, then the catheter is withdrawn to the next station. If
however, a station is achieved in which the nadir of the phasic con-
tractions remains above 40 mm Hg, then the manometry assistant
informs the endoscopists of this and instructs the endoscopists to
hold that position for at least 30 seconds. If the nadir of the phasic
contractions remains above 40 mm Hg throughout that 30 second
time period, then this would be interpreted as an abnormal record-
ing for that lead. Once the 30 second span has been achieved, then
the manometry assistant instructs the endoscopists to withdraw
another station. Again, the nadir of the phasic waves is viewed to
determine if it remains above 40 mm Hg. An abnormal sphincter of
Oddi manometry is determined when the nadir of the phasic waves
remains above 40 mm Hg for a 30 second span in both leads. As the
catheter is withdrawn, the more proximal of the two side ports
372
should come into the sphincter zone and one should obtain abnor-
mal 30 second recording. Then as the catheter is withdrawn, the
more proximal port will eventually come back to normal while the
distal port then enters the abnormal sphincter zone. With further
withdrawal, both leads are withdrawn out of the sphincter zone and
the nadir of the contractions returns to baseline.
One weakness in sphincter of Oddi manometry is that interpre-
tation of the recordings is not standardized. Most people accept
40 mm or greater as being abnormal, yet the largest study looking
at normals suggests that 35 mm Hg is a better gure.27 Various
systems are used to obtain an actual value for the basal sphincter
pressure. The Indiana group advocates taking the nadir value for the
four lowest nadirs over a 30 second run and then averaging those
three values.28 The most important factor however is that most agree
that there should be a sustained time (most agree on 30 seconds)
where the nadir of the phasic waves does not dip below 40 mm Hg.
During this time, it can be extremely important to keep the position
of the catheter steady. This can be challenging if there is active duo-
denal motility or respiratory movements are transmitted to the
abdominal cavity. During the critical zone of recording, it is extremely
important that there be immediate communication between the
endoscopists, and the manometry nurse to indicate the exact posi-
tion of the manometry catheter relative to the ampullary orice. If
the basal pressure is changed, it is important to know that the cath-
eter is in the same position.
A very controversial part of sphincter of Oddi manometry has
been the type of sedation and adjunctive medications which are
acceptable while performing sphincter of Oddi manometry. Tradi-
tionally, conscious sedation for endoscopy has been accomplished
with the combination of narcotic and benzodiazepine. It is known
that narcotics do affect intestinal motility and sphincter recordings.
Thus for many years, conscious sedation for ERCP and sphincter of
Oddi manometry was accomplished with a benzodiazepine alone.
However, benzodiazepines alone frequently provide inadequate
sedation for patients and many patients develop paradoxical agita-
tion when high doses of benzodiazepines are used. Credit should
be given to Grace Elta and her colleagues at the University of Michi-
gan who were the rst to question the validity of avoiding narcotics
for conscious sedation for sphincter of Oddi manometry.29 In a small
limited study, they found that basal sphincter pressures (the values
used to determine if a manometry is normal or abnormal) were not
affected by Demerol in a dose of 1 mg/kg. This nding was
conrmed by a larger, better designed study done at Indiana
University.30 This initiated an era where meperidine was routinely
used in conjunction with benzodiazepines for sphincter of Oddi
manometry. More recently, propofol has been advocated for use
during ERCP to achieve even better sedation. Several animal studies
have demonstrated that it does not affect SOM in dogs and sheep.31
32
However, only one human study has been reported and involves
only 11 patients. This study concluded that propofol did not alter
basal sphincter of Oddi pressures.33 Today, propofol is routinely
used in many centers to achieve deep sedation for the performance
of sphincter of Oddi manometry. In some cases, duodenal motility
can make cannulation very difcult. While glucagon (Eli Lilly,
Indianapolis, Indiana) is used routinely in standard ERCP to control
duodenal movement, it cannot be used during sphincter of Oddi
manometry as it does affect sphincter pressures. If it is impossible
to cannulate without the aid of glucagon, it is recommended that 5
minutes pass between a dose of glucagon and manometry
recording.

TREATMENT
Medical
Medical therapy has not been widely studied as a treatment for
sphincter of Oddi dysfunction. Because the sphincter of Oddi is a
smooth muscle structure, it makes some sense that pharmacologic
therapy might be of benet. If the theory is correct that sphincter of
Oddi dysfunction falls into functional or structural categories, medi-
cations would then be of benet only for those with functional
disease. Empiric pharmacologic trials are most reasonable in
Geenen-Hogan Type III patients with relatively mild and infrequent
episodes of pain. The drugs most studied in sphincter of Oddi dys-
function are calcium channel blockers and nitrates. Khuroo et al.
investigated nifedipine in a placebo-controlled crossover trial in 28
patients.34 End points for the study included reduction in pain scores,
emergency room visits and use of oral pain medication. Seventy-ve
percent of 28 patients responded to nifedipine. Sand et al. looked at
the effects of three calcium channel blockers with differing smooth
muscle selectivity (verapamil, nifedipine and felodipine) on human
sphincter of Oddi contractions.35 Results showed that all 3 calcium
channel blockers are potent inhibitors of contraction and it was
concluded that this category of drugs might be helpful in SOD. Sand
et al. performed a 16-week double-blind crossover study using nife-
dipine in Geenen-Hogan type II patients and showed that it decreased
the number of days in which patients experienced pain.36 A slow-
release form of nifedipine was tested in a small pilot study in patients
with SOD with encouraging early results.37 Nitrates have been
studied experimentally but there has not been a large literature using
this class of medications in humans. Gocer et al. found that isosor-
bide dinitrate decreased rhythmic and tonic contraction in guinea
pig-isolated sphincter of Oddi muscle.38 Bar-Meir described the dis-
appearance of pain as well as a decrease in both basal and phasic
sphincter activity on repeat manometry after nitrate therapy in a
woman with manometry-proven papillary dysfunction.39 Finally,
Wehrmann et al. looked at topical application of nitrates onto the
papilla of Vater and found that topically applied nitrates had a pro-
found inhibition of sphincter of Oddi motility.40
The new drug on the block may be nitric oxide (NO). Nitric
oxide plays an important role in the regulation of intestinal and
pancreaticobiliary motility. Inhibition of nitric oxide synthetase
(NOS) increases intraluminal pressure within the GI tract. We
looked at the effect of an inhibitor of NOS, NG-Nitro-L-arginine
methyl ester (L-NAME) on the mean basal pressure of the sphincter
of Oddi in the anesthetized pig. We found that L-NAME signicantly
increases mean sphincter of Oddi pressure in this animal model and
the physiologic affect was sustained for the duration of the experi-
ment (3 hours).41 Noting that topical administration of a NO donor
induces SO relaxation in humans, Niiyama et al. looked at the effect
of intrasphincteric injection of sodium nitroprusside (SNP) on the
pig SO. They found that intrasphincteric injection of SNP signi-
cantly reduced the mean basal pressures which lasted up to 45
minutes without inducing side effects or signicantly lowering
blood pressure.42 Research is ongoing to develop pharmaceutical
agents that will generate nitric oxide and these may serve as potential
treatment for sphincter of Oddi dysfunction.43
Despite some promising developments, medical therapy for
sphincter of Oddi is still problematic for a number of reasons:
1. Current therapy, particularly nitrates, have a signicant side effect
prole (especially headache).
Chapter 34 Pancreaticobiliary Pain and Suspected SOD
2. There is lack of long-term data.
3. The variability of response may be due to our inability to differ-
entiate between xed stenosis and functional spasm.
To move forward with medical therapy, well-conducted, placebo-
controlled, randomized trials with long-term follow-up need to be
performed. The main drawbacks of medical therapy at this point are
the lack of specicity for the sphincter of Oddi and the lack of a
long-acting medication with a low side effect prole.44
Endoscopic
Endoscopic therapy has been the most widely employed treatment
for sphincter of Oddi dysfunction. When assessing outcomes after
endoscopic sphincterotomy, specics of the patient population (G-H
Type I, II or III) and the exact nature of the intervention (biliary or
dual sphincterotomy) must be taken into account. There is only one
study that focuses on sphincterotomy solely in G-H type I patients.45
This was a relatively small study population of 17 patients with
biliary type pain, dilated bile duct and abnormal liver tests during
episodes of pain. At ERCP, only 65% had abnormal basal sphincter
pressures with SOM but all beneted from biliary sphincterotomy
with a mean of 2.3 years follow-up.
The strongest data supporting the efcacy of endoscopic interven-
tion is contained in three randomized trials,4648 two of which46,48
included only G-H type II patients who underwent either sham
therapy or biliary sphincterotomy alone. In the landmark study by
Geenen et al. all patients underwent ERCP with SOM and all patients
were then randomized to either sham therapy or biliary sphincter-
otomy (the endoscopist was blinded to the results of the manome-
try).46 The results helped validate both the predictive capability of
SOM as well as the benet of sphincterotomy. The patients with
normal SOM did not benet from sphincterotomy but those with
abnormal basal sphincter pressures did. The important take home
points of this trial are:
1. Only Type II patients were included
2. Patients underwent biliary sphincterotomy alone
3. Those patients with abnormal basal sphincter pressures treated
with sphincterotomy beneted signicantly more than those with
normal SO pressures treated by sham or sphincterotomy.
The Toouli et al. study48 was designed somewhat similar to the
Geenen study. It included only G-H type II patients and involved
randomization of all patients to sham therapy or biliary sphincter-
otomy. In this trial, however, if the manometry was initially normal,
the patients were provoked with cholecystokinin in an effort to detect
a subset of patients with functional SOD. The outcomes were
similar to the Geenen study, 85% (11/13) patients with elevated basal
sphincter pressures beneted from sphincterotomy whereas only
38% (5/13) patients beneted from sham therapy (P = 0.041). The
outcomes were similar for both the sphincterotomy and the sham
group who had normal SOM.
The Indiana trial, however, was distinctive for several reasons:
it had a 3 group randomizationsham, endoscopic biliary sphinc-
terotomy, surgical (dual) sphincteroplasty, and the trial involved
G-H Type II and III patients.47 The latter characteristic of the Indiana
trial is particularly important because most centers that have exper-
tise in SOD see a predominance of G-H type III patients. However,
this trial did not randomize all patients; only those with abnormal
SOM. The results with 3-year follow-up revealed that 69% of patients
in the endoscopic sphincterotomy and surgical sphinctero-
plasty groups beneted compared to only 24% of the sham group
(P = 0.009).
373
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
The most serious void in our data is the lack of well-designed
trials that address whether SOM predicts outcome and whether
sphincterotomy is benecial in G-H type III patients. The Indiana
data did contain type II patients but did not address the predictability
of SOM. Such a trial is currently being designed and will be under-
taken as part of a multi-center protocol.
The other issue regarding endoscopic intervention that has not
been adequately addressed in properly designed studies is the ef-
cacy of biliary sphincterotomy alone versus combined biliary and
pancreatic sphincterotomy. There is, however, a growing body of
evidence that suggests that some patients may benet from dual
sphincterotomy. It is known from reports in which dual manometry
has been performed that there is generally a concordance between
the pancreatic and biliary sphincter pressures. In the majority of
cases, they are either both normal or both abnormal. There is
however, some discordance which suggests that about 10% of the
time there is isolated biliary sphincter hypertension and about 20%
of the time there is isolated pancreatic sphincter hypertension.4950
In a previous report by Guelrud51 biliary sphincterotomy alone
versus biliary and pancreatic sphincterotomy was evaluated in a
group of patients with Type II pancreatic sphincter of Oddi dysfunc-
tion and recurrent pancreatitis. Twenty-eight percent of patients
undergoing biliary sphincterotomy alone showed improvement
whereas 86% (12/14) showed improvement if dual sphincterotomy
was performed. There was also a subset of 13 patients who under-
went biliary sphincterotomy followed by pancreatic sphincterotomy
in a later session, and ultimately 77% of those patients improved. In
both cases (biliary and pancreatic sphincterotomy at the same session
or pancreatic sphincterotomy at a later session), results were signi-
cantly improved over patients with biliary sphincterotomy alone.
Another study from the University of Iowa52 looked at a group of 26
patients who had not responded to biliary sphincterotomy despite
abnormal sphincter of Oddi motility. 25/26 underwent a repeat
ERCP and pancreatic sphincterotomy and 16 of them (64%)
responded. A further study by Kaw et al.53 looked at biliary versus
dual sphincterotomy and related it to which sphincter had abnormal
manometry. For those with an abnormal biliary manometry alone,
80% responded to biliary sphincterotomy. However, if isolated pan-
creatic sphincter hypertension or combined biliary and pancreatic
sphincter hypertension were found, only 7/23 (30%) responded if
patients underwent a biliary sphincterotomy alone. Alternatively, if
patients had isolated pancreatic sphincter hypertension or combined
biliary and pancreatic sphincter abnormalities, 11/16 (69%)
responded to dual sphincterotomy. Although there is increasing
enthusiasm for performing a dual sphincterotomy in those patients
in which the pancreatic sphincter has been shown to be abnormal,
a denitive recommendation must await a properly designed, ran-
domized clinical trial.
Some investigators have tried botulinum toxin (BoTox, Allergan
Inc., Irving, California) injections directly into the sphincter as a
substitute for sphincter of Oddi manometry,54 as a permanent treat-
ment for sphincter of Oddi dysfunction,5556 or to prevent post-ERCP
pancreatitis.57 The presumption with BoTox is that the pain of
sphincter of Oddi dysfunction comes from the sphincter itself and
by preventing tonic contraction, symptoms can be relieved. Sand et
al.58 demonstrated that botulinum toxin inhibits pig sphincter of
Oddi smooth muscle contractions and Wang et al.59 showed that in
dogs, it reduced contractile activity for a prolonged time. The rst
clinical report of its use in SOD was by Wehrmann.55 They injected
100 international units of botulinum toxin in 22 patients with
374
manometrically documented SOD (all were Geenen-Hogan type III
patients). Six weeks later, 55% (12 patients) were symptom-free and
45% (10 patients) were not. The 10 non-responders underwent
ERCP and biliary sphincterotomy. Five of ten patients had normal-
ized their sphincter pressure and did not respond to biliary sphinc-
terotomy with longer-term follow-up. Eleven of twelve initial
responders relapsed at a median of six months. Repeat manometry
revealed sphincter hypertension in 11/23 and 11 responded to endo-
scopic sphincterotomy. This initial report has not been followed by
a prospective randomized study. Goerlick et al.57 found it to be effec-
tive in decreasing the risk of post-ERCP pancreatitis in manometri-
cally positive SOD patients; however, the incidence of pancreatitis
was 25%, which in the era of pancreatic stenting, is unacceptable.60
There are several drawbacks to this approach.
1. Botulinum toxin has not been subjected to a randomized study
in the way that manometry hasat least in Geenen-Hogan type
II patients.
2. When botulinum toxin is used as a treatment of sphincter of Oddi
dysfunction, it is logical that its effect will be transient as we have
seen in botulinum toxin in achalasia patients.
3. When used as a predictor of response to sphincterotomy, it requires
a second procedure, which exposes patients with suspected SOD
to yet another procedure that could cause pancreatitis.
In summary, there are inherent drawbacks to botulinum toxin
use in sphincter of Oddi dysfunction. The most important factors
are that overall experience is very small and botulinum toxin has not
been subjected to well-designed, randomized studies with sufcient
follow-up to determine efcacy. Though data is scant, BoTox does
not appear to be as effective as short-term pancreatic stenting in
preventing post-ERCP pancreatitis.
PREVENTION OF POST ERCP PANCREATITIS
It was originally thought that it was the actual performance of SOM
that caused post ERCP pancreatitis (PEP). We have come to learn
however that the risk is in fact inherent in the patients themselves.
This was outlined in a study published by Freeman et al. This land-
mark paper clearly outlines risk factors for post ERCP pancreatitis
and most are related to the patients themselves (Table 34.8).61 Per-
formance of ERCP, with or without manometry, in a young female
with suspected SOD carries a very high risk of post ERCP pancre-
atitis. To date, the best study done which has looked at the role of
manometry itself as a causative factor in PEP reviewed 76 patients
with suspected SOD undergoing SOM.25 The group was randomized
to manometry in the standard fashion with all three ports perfused
with 0.25 cc of water per minute versus perfusion through two
Multivariate analysis p. value
Suspected SOD <0.001
Younger age <0.001
Univariate analysis
History of ERCP induced pancreatitis <0.001
Female sex <0.001
History of pancreatitis <0.001
Distal bile duct diameter 0.02
Table 34.8 Patient factors correlated with increased risk of
pancreatitis
From Freeman et al., NEJM 1996 [61].

leads with simultaneous aspiration through the third channel. A
previous study proved that aspiration during SOM did not affect the
manometry results.62 This study was important because the proce-
dure consisted only of SOM and the patients did not undergo ERCP
after the manometry was complete. Thus, the study isolated SOM
and recorded the incidence of PEP. The results showed that in the
group being perfused, the pancreatitis rate was 23.5% whereas in
the group undergoing aspiration, the pancreatitis rate was 3% (p =
0.01). The recorded rate of pancreatitis in the aspiration group is an
acceptable rate for PEP, in general, and well below the rate generally
quoted for patients with suspected SOD.
The most important factor in reducing the risk of PEP in patients
with suspected SOD reported to date is stenting of the pancreatic
duct. The hypothesis is that manipulation of the ampulla in the
course of performing ERCP (with or without sphincterotomy) may
cause swelling and compromise pancreatic uid drainage leading
to pancreatitis. In a landmark study, patients with manometri-
cally documented SOD were randomized to short-term pancreatic
stenting versus no stent following biliary sphincterotomy.63 The
results showed that the stented group had a PEP rate of 7% com-
pared to a rate of 26% in the non-stented group (p = 0.03). A number
of other studies have also suggested that there was benet to pan-
creatic stenting to prevent PEP.6467 Subsequently, Singh et al. per-
formed a meta-analysis of published studies and concluded
that pancreatic stenting was effective in reducing the incidence of
post ERCP pancreatitis in patients with suspected or proven
SOD.68
Pancreatic stenting is not without potential problems. The origi-
nal prospective trials used short 5 Fr stents that required endoscopic
removal. In an effort to avoid a second procedure to remove stents,
we began utilizing stents that did not have a ap on the pancreatic
side. The latter typically migrate within 12 weeks and are equally
effective in preventing PEP.69 Many experts favor the use of smaller
caliber (3 Fr), longer (8, 10 or 12 cm) pancreatic stents for this indica-
tion for the same reason (video). These stents have a pigtail on the
duodenal side and no retention ap on the pancreatic side (Lehman
pancreatic stent, Cook GI Endoscopy, Winston-Salem, NC). They
must be placed over a 0.018 guidewire and because of the length
of the stent, the guidewire must be passed into the tail of the pan-
creas. These stents typically fall out within 2 weeks and can be
checked by obtaining a plain lm of the abdomen to include the
diaphragms 23 weeks after the procedure. If the stent is still in
place, it will be seen as a radio-opaque thread crossing the spine in
the upper abdomen underneath the diaphragm. If the stent is
present at 3 weeks, one can wait an additional week and repeat the
x-ray. If the stent remains after 4 weeks, then it is probably best to
remove it with a side-viewing endoscope and a mini-snare. The main
reason for using this stent is that it is very soft and exible, and when
left in place for 34 weeks, is associated with less iatrogenic ductal
damage than that seen with larger stents.7071 Three French stents
also predictably stay in place for at least 72 hours which is probably
the timeframe necessary to prevent PEP. The only drawback to this
stent is that it requires passing the 0.018 guidewire to the tail of the
pancreas. This can be difcult if there are multiple acute turns as
the duct courses through the head of the pancreas. Occasionally we
also encounter patients with an ansa pancreaticus in which the
main pancreatic duct makes a 360 turn as it courses through the
head of the pancreas. In cases with severe sigmoid bends or ansa
pancreaticus, we favor placement of a short 5 Fr stent with the inter-
nal (pancreatic side) ap removed.
Chapter 34 Pancreaticobiliary Pain and Suspected SOD
Though controversial and not yet subjected to a randomized trial,
current retrospective data suggests that pancreatic stenting is helpful
in preventing PEP in those patients found to have a normal SOM.72
The reason for this is unknown but the risk of PEP is inherent in
patients with suspected SOD not just those proven to have it with
abnormal manometry. Also, our data shows that 42% of patients
with suspected SOD who have a normal manometry at their index
ERCP, have an abnormal manometry if they return for a repeat
examination suggesting that the initial manometry was falsely nega-
tive.73 Though current evidence is not as strong as it could be, our
recommendation is that a short-term pancreatic stent be placed even
if SOM is found to be normal.
EVALUATION OF PATIENTS WITH RECURRENT
PAIN AFTER ENDOSCOPIC INTERVENTION FOR
SPHINCTER OF ODDI DYSFUNCTION
Despite an abnormal manometry, patients may not respond to endo-
scopic intervention or may demonstrate a transient response fol-
lowed by a relapse. The results of a reinvestigation can lead to several
potential ndings:
1. incomplete prior biliary sphincterotomy
2. residual pancreatic sphincter hypertension
3. restenosis of the pancreatic sphincter
4. completely normal examination
5. evidence of early chronic pancreatitis
If patients re-present with typical pancreaticobiliary pain after
endoscopic therapy for sphincter of Oddi dysfunction and the recur-
rent symptoms warrant the risks of ERCP, this examination should
be repeated. There are no comprehensive reports that systematically
detail the ndings at a second examination. In most centers that see
a signicant number of sphincter of Oddi patients, as complete a
biliary sphincterotomy as possible is usually performed if biliary
sphincterotomy is indicated. The exact source of pain in sphincter
of Oddi patients is still not known but it is likely that in those with
functional stenosis (as opposed to structural obstruction), the pain
emanates from the sphincter itself. This is the reason to perform a
complete biliary sphincterotomy. Under these circumstances, it is
unusual to nd recurrent stenosis of the biliary sphincter on follow-
up examination, but rather residual pancreatic sphincter hyper-
tension. In this setting, available data suggests that patients
symptomatically improve following pancreatic sphincterotomy. In a
study by Elton,74 pancreatic manometry was performed following
biliary sphincterotomy in patients with sphincter of Oddi dysfunc-
tion. If pancreatic sphincter hypertension was found, pancreatic
sphincterotomy was performed. The results showed that 73% of
patients had complete resolution of symptoms after the index
ERCP and an additional 18% showed partial or transient change.
Only 8% of these Type I and Type II SOD patients had no change
in symptoms. These results are somewhat better than other reports
in the literature. Additionally, Eversman et al.75 looked at long-term
follow-up after biliary sphincterotomy and correlated it with the
sphincter of Oddi manometry results. In this study, 37 patients
had isolated biliary sphincter hypertension and only 16% required
reintervention. In a group of 62 patients who had elevated biliary
and pancreatic basal pressures, 29% required reintervention. For
the 33 patients who had isolated pancreatic sphincter hypertension
(and underwent a biliary sphincterotomy alone), 39% required
reintervention.
375
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
This concept of improved outcomes with dual sphincterotomy
when pancreatic sphincter hypertension is present was further
investigated by Park et al.76 In this report, there was no signicant
difference in outcome when comparing dual sphincterotomy versus
biliary sphincterotomy alone in patients with isolated biliary sphinc-
ter hypertension. Interestingly, there was also no difference between
dual and biliary sphincterotomy alone in patients who had both
abnormal biliary as well as abnormal pancreatic sphincter pressures.
However, there was a signicant difference in the rate of reinterven-
tion in patients with isolated abnormal pancreatic sphincter pres-
sures (21% vs 39%, p value < .05).
In summary, it does appear that outcomes can be improved if
pancreatic sphincterotomy is performed in patients with docu-
mented pancreatic sphincter hypertension. However, denitive rec-
ommendations await appropriate randomized trials.
CONCLUDING REMARKS
The evaluation and treatment of patients with suspected sphincter
of Oddi dysfunction remains a challenge for gastroenterologists.
Obtaining a detailed history is a critical step and the evaluation for
other causes of upper abdominal pain should be undertaken and
empiric medical trials for endoscopically and radiologically negative
diseases (GERD, IBS) should be tried. However, with an appropriate
history and a failure to respond to empiric interventions, one should
consider sphincter of Oddi dysfunction. In these cases, evaluation
for ductal dilation and hepatic and pancreatic chemistry panels
REFERENCES
1. Corazziari E, Shaffer EA, Hogan WJ, et al. Functional disorders of
the biliary tract and pancreas. Gut 1999; 45:4854.
2. Behar J, Corazziari E, Guelrud M, et al. Functional gallbladder and
sphincter of Oddi disorders. Gastroenterology 2006;
130(5):1498509.
3. Hogan WJ, Geenen JE, Dodds WJ. Dysmotility disturbances of the
biliary tract: classication, diagnosis and treatment. Semin Liver Dis
1987; 7(4):302310.
4. Troiano F, Hawes R, Schultze E, et al. Sphincter of Oddi (SO)
assessment: correlation between ductal drainage time (DT) and
manometry. Gastrointest Endosc 1988; 34(2):192(Abst).
5. Elta GH, Barnett JL, Ellis JH, Ackermann R, Wahl R. Delayed biliary
drainage is common in asymptomatic post-cholecystectomy
volunteers. Gastrointest Endosoc 1992; 38:435439.
6. Sherman S, Troiano FP, Hawes RH, et al. Frequency of abnormal
sphincter of Oddi manometry compared with the clinical
suspicion of sphincter of Oddi dysfunction. Am J Gastrointerol
1991; 86(5):586590.
7. Sostre S, Kalloo AN, Spiegler EJ, et al. A noninvasive test of
sphincter of Oddi dysfunction in post-cholecystectomy patients:
the scintigraphic score. J Nucl Med 1992; 33(6):12231124.
8. Kalloo AN, Sostre S, Pasricha PJ. The Hopkins scintigraphic score:
a noninvasive, highly accurate screening test for sphincter of
Oddi dysfunction. Gastroenterology 1994; 106:342(Abst).
9. Corazziari E, Cicala M, Habib FI, et al. Hepatoduodenal bile transit
in cholecystectomized subjects. Relationship with sphincter of
Oddi function and diagnostic value. Dig Dis Sci 1994;
39(9):19851993.
10. Jagannath S, Kalloo AN. Efcacy of biliary scintigraphy in
suspected sphincter of Oddi dysfunction Curr Gastroenterol 2001;
3(2):160165.
376
during or shortly after an episode of signicant pain should be
obtained. With this information, the patient can be categorized into
Geenen-Hogan criteria (I, II or III). Because of a 90% response to
sphincterotomy and the fact that they appear to have a lower inci-
dence of post-ERCP pancreatitis, those patients who fall into cate-
gory I can undergo ERCP and sphincterotomy without the need for
manometry. Geenen-Hogan category II and III patients, should be
referred to a gastroenterologist or pancreaticobiliary center with
expertise in sphincter of Oddi manometry. Category II patients
should undergo a manometry because randomized trails have proven
that manometry is an accurate discriminator of those who will
respond to sphincterotomy. We clearly need randomized trials in
category III patients. In the absence of those trials, current data
suggests that manometry is helpful in selecting those who will
respond to sphincterotomy. Moreover, emerging data suggests that
both sphincters should be studied, and in selected patients with
pancreatic sphincter hypertension, one should consider dual sphinc-
terotomy. Additionally, category III patients are at high risk for
post-ERCP pancreatitis and should undergo short-term pancreatic
stenting, even if their manometry is normal. The need for dual
manometry combined with a potential need for pancreatic sphinc-
terotomy and the mandatory requirement for pancreatic stenting all
dictate that type III patients be managed by endoscopists who have
signicant experience and interest in patients with sphincter of Oddi
dysfunction. Finally, patients with sphincter of Oddi dysfunction
who relapse after initial endoscopic intervention should be re-
evaluated if the severity of their recurrent symptoms warrants either
endoscopical or surgical intervention.
11. Pineau BC, Knapple WL, Spicer KM, et al. Cholecystokinin-
stimulated mebrofenin (99mTc-choletec) hepatobiliary
scintigraphy in asymptomatic postcholecystectomy individuals:
assessment of specicity, interobserver reliability, and
reproducibility. Am J Gastroenterol 2001; 96(11):31063109.
12. Rosenblatt ML, Catalano MF, Alcocer E, et al. Comparison of
sphincter of Oddi manometry, fatty meal sonography, and
hepatobiliary scintigraphy in the diagnosis of sphincter of Oddi
dysfunction. Gastrointest Endosc 2001; 54(6):697704.
13. Mirbagheri SA, Mohamadnejad M, Nasiri J, et al. Prospective
evaluation of endoscopic ultrasonography in the diagnosis of
biliary microlithiasis in patients with normal transabdominal
ultrasonography. J Gastrointest Surg 2005; 9(7):961964.
14. Dill JE, Hill S, Callis J, et al. Combined endoscopic ultrasound
and stimulated biliary drainage in cholecystitis and
microlithiasisdiagnoses and outcomes. Endoscopy 1995;
27(6):424427.
15. Dill JE. Symptom resolution of relief after cholecystectomy
correlates strongly with positive combined endoscopic
ultrasound and stimulated biliary drainage. Endoscopy 1997;
29(7):646648.
16. Bingener J, Richards ML, Schwesinger WH, et al. Laparoscopic
cholecystectomy for biliary dyskinesia: correlation of preoperative
cholecystokinin cholescintigraphy results with postoperative
outcome. Surg Endosc 2004; 18(5):802806.
17. Adams DB, Tarnasky PR, Hawes RH, et al. Am Surg 1998;
64(1):15.
18. DiBaise JK, Oleynikov D. Does gallbladder ejection fraction predict
outcome after cholecystectomy for suspected chronic acalculous
gallbladder dysfunction? A systematic review. Am J Gastroenterol
2003; 98(12):26052611.

19. Yap L, Wycherley AG, Morphett AD, et al. Acalculous biliary pain:
cholecystectomy alleviates symptoms in patients with abnormal
cholescintigraphy. Gastroenterology 1991; 101:786793.
20. Patel NA, et al. Therapeutic efcacy of laparoscopic
cholecystectomy in the treatment of biliary dyskinesia. Am J
Gastroenterol 2004; 187:209212.
21. Yost F, Margenthaler J, Presti M, et al. Cholecystectomy is an
effective treatment for biliary dyskinesia. Am J Surg 1999;
178:462465.
22. Guelrud M, Mendoza S, Mujica V, et al. Sphincter of Oddi (SO)
motor function in patients with symptomatic gallstones.
Gastroenterology 1993; 104(A361).
23. Ruffolo TA, Sherman S, Lehman GA, et al. Gallbladder ejection
fraction and its relationship to sphincter of Oddi dysfunction. Dig
Dis Sci 1994; 39:289292.
24. Choudhry U, Ruffolo T, Jamidar P, et al. Sphincter of Oddi
dysfunction in patients with intact gallbladder: Therapeutic
response to endoscopic sphincterotomy. Gastrointest Endoscopy
1993; 39:492495.
25. Sherman S, Troiano FP, Hawes RH, et al. Sphincter of Oddi
manometry: decreased risk of clinical pancreatitis with use of a
modied aspirating catheter. Gastrointest Endosc 1990;
36(5):462466.
26. Sherman S, Hawes RH, Troiano FP, et al. Pancreatitis following
bile duct sphincter of Oddi manometry: utility of the aspirating
catheter. Gastrointest Endosc 1992; 38(3):347350.
27. Guelrud M, Mendoza S, Rossiter G, et al. Sphincter of Oddi
manometry in healthy volunteers. Dig Dis Sci 1990; 35(1):3846.
28. Sherman S, Lehman GA. Sphincter of Oddi dysfunction: diagnosis
and treatment. JOP 2001; 2(6):382400.
29. Elta GH, Barnett JL. Meperidine need not be proscribed during
sphincter of Oddi manometry. Gastrointest Endosc 1994;
40(1):79.
30. Sherman S, Gottleib K, Uzer MF, et al. Effects of Meperidine on
the pancreatic and biliary sphincter. Gastrointest Endosc 1996;
44(3):239242.
31. Baron TH, Dalton CB, Cotton PB, et al. The effect of propofol on
the canine sphincter of Oddi. HPB Surg 1994; 7(4):297304.
32. Turan M, Bagcivan I, Gursoy S, et al. In-vitro effects of intravenous
anesthetics on the sphincter of Oddi strips of sheep.
Pancreatology 2005; 5(23):215219.
33. Goff JS. Effect of propofol on human sphincter of Oddi. Dig Dis
Sci 1995; 40(110):23642367.
34. Khuroo MS, Zargar SA, Yattoo GN. Efcacy of nifedipine therapy
in patients with sphincter of Oddi dysfunction: a prospective-
double-blind-randomized-placebo-controlled-cross over trial. Br J
Clin Pharmacol 1992; 33:477485.
35. Sand J, Arvola P, Nordback I. Calcium channel antagnosits and
inhibition of human sphincter of Oddi contractions. Scand J
Gastrroenterol 2005; 40(12):13941397.
36. Sand J, Nordback I, Koskinen M, et al. Nifedipine for suspected
type II sphincter of Oddi dyskinesia. American Journal of
Gastroenterology 1993; 88:530535.
37. Craig AG, Toouli J. Slow-release nifedipine for patients with
sphincter of Oddi dyskinesia: results of a pilot study. Int Med J
2002; 32(3):119120.
38. Gocer G, Yearis E, Tuncer M, et al. Affect of nitrovasodilators on
the rhythmic contractions of guinea-pig isolated sphincter of
Oddi. Arzneimittelforschung 1995; 45(7):809812.
39. Bar-Meir S, Halperin Z, Bardan E. Nitrate therapy in a patient
with papillary dysfunction. Am J Gastroenterol 1983;
78(20):9495.
40. Wehrmann T, Schmitt T, Stergiou N, et al. Topical application
nitrates onto the papilla of Vater: Manometric and clinical results.
Endoscopy 2001; 33(4):323328.
Chapter 34 Pancreaticobiliary Pain and Suspected SOD
41. Bak AW, Perini RF, Muscara M, et al. An endoscopic animal model
of sphincter of Oddi dysfunction. Gastroenterology 2001;
120(5)suppl 1:A390.
42. Niiyama H, Jagannath S, Kantsevoy S, et al. Intrasphincteric nitric
oxide reduces sphincter of Oddi motility in an endoscopic
porcine model. Dig Dis Sci 2003; 48(11):21872190.
43. Herrmann BW, Cullen JJ, Ledlow A, et al. The affect of
peroxynitrite on sphincter of Oddi motility. J Surg Res 1999;
8191:5558.
44. Craig A, Toouli J. Sphincter of Oddi dysfunction: is there a
role for medical therapy? Curr Gastrenterol Rep 2002;
4(2):172176.
45. Rolny P, Geenen JE, Hogan WJ. Post-cholecystectomy patients
with objective signs of partial bile outow obstruction: clinical
characteristics, sphincter of Oddi manometry ndings, and results
of therapy. Gastrointest Endosc 1993; 39:778781.
46. Geenen JE, Hogan WJ, Dodds WJ, et al. The efcacy of endoscopic
sphincterotomy after cholecystectomy in patients with sphincter
of Oddi dysfunction. N Engl J Med 1989; 320:8287.
47. Sherman S, Lehman GA, Jamidar P, et al. Efcacy of endoscopic
sphincterotomy and surgical sphincteroplasty for patients with
sphincter of Oddi dysfunction (SOD): randomized, controlled
study. Gastrointest Endosc 1994; 40:A125.
48. Toouli J, Roberts-Thomson IC, Kellow J, et al. Manometry based
randomized trial of endoscopic sphincterotomy fro sphincter of
Oddi dysfunction. Gut 2000; 46(1):98102.
49. Knapple W, Tarnasky P, Coyle W, et al. Sphincter of Oddi
manometry (SOM) of both ducts after conscious sedation with
meperidine. Gastrointest Endosc 1996; 43(4):385(Abst).
50. Eversman D, Sherman S, Bucksot L, et al. Frequency of abnormal
biliary and pancreatic basal sphincter pressure at sphincter of
Oddi manometry (SOM) in 463 patients. Gastrointest Endosc
1996; 43(4):381(Abst).
51. Guelrud M, Plaz J, Mendoza S, et al. Endoscopic treatment in Type
II pancreatic sphincter dysfunction. Gastrointest Endosc 1995;
41(40):A398.
52. Soffer EE, Johlin FC. Intestinal dysmotility in patients with sphincter
of Oddi dysfunction: a reason for failed response to
sphincterotomy. Dig Dis Sci 1994; 39(9):19421946.
53. Kaw M, Verma R, Brodmerkel Jr GJ. Biliary and/or pancreatic
sphincter of Oddi dysfunction (SOD): response to endoscopic
sphincterotomy (ES). Gastrointest Endosc 1996; 43(4):A384.
54. Banerjee B, Miedema B, Saifuddin T, Marshall JB. Intrasphincteric
botulinum toxin type A for the diagnosis of sphincter of Oddi
dysfunction: a case report. Surg Laparosc Endosc Percutan Tech
1999; 9(3):194196.
55. Wehrmann T, Seifert H, Seipp M, et al. Endoscopic injection of
botulinum toxin for biliary sphincter of Oddi dysfunction.
Endoscopy 1998; 30(8):702707.
56. Pasricha PJ, Miskovsky EP, Kalloo AN. Intrasphincteric injection of
botulinum toxin for suspected sphincter of Oddi dysfunction. Gut
1994; 35:13191321.
57. Goerlick A, Barnett J, Chey W, et al. Botulinum toxin injection after
biliary sphincterotomy. Endoscopy 2004; 36(2):170173.
58. Sand J, Nordback I, Arvola P, et al. Effects of botulinum toxin A on
the sphincter of Oddi: an in vivo and in vitro study. Gut 1998;
42(4):507510.
59. Wang HJ, Tanaka M, Konomi H, et al. Effect of local injection of
botulinum toxin on sphincter of Oddi cyclic motility in dogs. Dig
Dis Sci 1998; 43(4):694701.
60. Cotton PB, Hawes RH. Botulinum toxin injection after biliary
sphincterotomy. Endoscopy 2004; 36(8):744.
61. Freeman ML, Nelson DB, Sherman S, et al. Complications of
endoscopic biliary sphincterotomy. N Engl J Med 1996;
335(13):909918.
377
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
62. Sherman S, Troiano FP, Hawes RH, et al. Does continuous
aspiration from an end and side port in a sphincter of Oddi
manometry catheter alter recorded pressures? Gastrointest
Endosc 1990; 36(5):500503.
63. Tarnasky PR, Palesch YY, Cunningham JT, et al. Pancreatic stenting
prevents pancreatitis after biliary sphincterotomy in patients with
sphincter of Oddi dysfunction. Gastroenterology 1998;
115(6):15181524.
64. Aizawa T, Ueno N. Stent placement in the pancreatic duct
prevents pancreatitis after endoscopic sphincter dilation for
removal of bile duct stones. Gastrointest Endosc 2001;
54(2):209213.
65. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent
prevent post-ERCP pancreatitis? A prospective randomized study.
Gastrointest Endosc 2003; 57(3):291294.
66. Smithline A, Silverman W, Rogers D, et al. Effect of prophylactic
main pancreatic duct stenting on the incidence of biliary
endoscopic sphincterotomy-induced pancreatitis in high-risk
patients. Gastrointest Endosc 1993; 39(5):652657.
67. Sherman S, Bucksot EL, Esber E, et al. Does leaving a main
pancreatic duct stent in place reduce the incidence of precut
biliary sphincterotomy-induced pancreatitis? Randomized
prospective study. Am J Gastroenterol 1995; 90:241 (Abstr).
68. Singh P, Das A, Isenberg G, et al. Does prophylactic pancreatic
stent placement reduce the risk of post-ERCP acute pancreatitis?
A meta-analysis of controlled trials. Gastrointest Endosc 2004;
60(4):544550.
378
69. Rashdan A, Fogel EL, McHenry L Jr, et al. Improved stent
characteristics for prophylaxis of post-ERCP pancreatitis. Clin
Gastroenterol Hepatol 2004; 2(4):322329.
70. Sherman S, Hawes RH, Savides TJ, et al. Stent-induced pancreatic
ductal and parenchymal changes: correlation of endoscopic
ultrasound with ERCP. Gastrointest Endosc 1996; 44(3):276282.
71. Smith MT, Sherman S, Ikenberry SO, et al. Alterations in pancreatic
ductal morphology following polyethylene pancreatic stent
therapy. Gastrointest Endosc 1996; 44(3):268275.
72. Fogel EL, Varadarajulu S, Sherman S, et al. Prophylactic
pancreatic duct stenting in patients with suspected sphincter of
Oddi dysfunction but normal sphincter of Oddi manometry.
Gastrointest Endosc 2003; 57(5):AB88.
73. Varadarajulu S, Hawes RH, Cotton PB. Determination of sphincter
of Oddi dysfunction in patients with prior normal manometry.
Gastrointest Endosc 2003; 58(3):341344.
74. Elton E, Howell DA, Parsons WG, et al. Endoscopic pancreatic
sphincterotomy: indications, outcome and a safe stentless
technique. Gastrointest Endosc 1998; 47(3):240249.
75. Eversman D, Fogel E, Phillips S, et al. Sphincter of Oddi
dysfunction (SOD): long-term outcome of biliary sphincterotomy
(BES) correlated with abnormal biliary and pancreatic sphincters.
Gastrointest Endosc 1999; 49(4):AB78.
76. Park S-H, Watkins JL, Fogel EL, et al. Long-term outcome of the
endoscopic dual pancreatobiliary sphincterotomy in patients with
manometry-documented sphincter of Oddi dysfunction and
normal pancreatogram. Gastrointest Endosc 2003; 57(4):483491.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Sclerosing Cholangitis
35 Jawad Ahmad and Adam Slivka
BACKGROUND
Primary sclerosing cholangitis (PSC) is a chronic inammatory
disease of the biliary tree. It is characterized by stricturing and dila-
tion of the intra- and/or extrahepatic bile ducts, with concentric
obliterative brosis of intrahepatic biliary radicles. PSC is closely
associated with inammatory bowel disease (IBD), particularly
ulcerative colitis (UC), which is found in approximately two-thirds
of northern European PSC patients.1,2 The disease leads to chronic
cholestasis but patients can be asymptomatic at presentation, diag-
nosed by abnormal liver enzymes, particularly elevation of the alka-
line phosphatase, or they can present with pruritus, fatigue, right
upper quadrant pain, and jaundice. As the disease progresses, symp-
toms of cirrhosis can be manifested. PSC is associated with an
unpredictable risk of developing cholangiocarcinoma in up to 30%
of patients.3 The etiology and pathogenesis of PSC are unclear but
it is likely an immune-mediated disease involving an exaggerated
cell-mediated immune response leading to chronic inammation of
the biliary epithelium.
PSC is diagnosed by radiographic imaging of the biliary tree (Fig.
35.1).4 This has traditionally been performed using endoscopic ret-
rograde cholangiopancreatography (ERCP) but more recently mag-
netic resonance cholangiopancreatography (MRCP) is thought to be
as sensitive as ERCP in the diagnosis of PSC if the best equipment
and operator are available (Fig. 35.2).
Liver biopsy has a limited role in diagnosis but is a useful adjunct
to determine the stage of the disease. Histology can range from
normal to frank biliary cirrhosis with the typical appearances being
of portal inammation, concentric onion skin periductal brosis a diagnosis of PSC can be condently made. Biliary surgery, calculi
and periportal brosis developing into septal and bridging necrosis.
The endoscopists role in PSC involves diagnostic cholangiogra-
phy, therapeutic intervention of strictures in the bile duct including
dilation and stenting, and differentiating between benign and malig-
nant strictures.
DIAGNOSIS AND NATURAL HISTORY
Introduction and scientic basis
The role of ERCP in the diagnosis of PSC has become more contro-
versial with the availability of high-quality MRCP. The latter has the
benet of being non-invasive but is operator and machine dependent
and does not allow therapeutic intervention or cytological sampling.
ERCP is still considered to be the gold standard and allows sampling
and intervention, although it is still operator dependent. In addition,
ERCP provides endoscopic staging of portal hypertension.
Several studies have compared ERCP and MRCP in patients with
clinical or biochemical evidence of cholestasis and MRCP appears
to have a comparable diagnostic accuracy. Textor et al. examined 150
patients with cholestatic liver enzymes and obtained a diagnostic
MRCP in 146 of these patients (97%). PSC was found in 34 of the
150 patients (23%) at ERCP and MRCP was able to correctly identify
88% (29 of 33) and had a specicity of 99% (108 of 109).5 A similar
study by Angulo and colleagues determined that MRCP had a diag-
nostic accuracy of 90% compared to 97% with invasive cholangiog-
raphy in 73 patients with a variety of cholestatic diseases, including
23 patients with PSC. However, three-quarters of the PSC patients
required a therapeutic intervention.6
Description of technique
The technique for ERCP in PSC does not differ from the standard
approach to biliary cannulation and is described elsewhere (see
Chapter 8). In certain cases an occlusion cholangiogram is required
using a stone-extraction balloon to prevent drainage of contrast from
the biliary tree or lling of the gallbladder. However, care should be
taken to avoid lling of segments of the intrahepatic ducts that sub-
sequently cannot be drained thus increasing the risk of infection.
We treat all PSC patients with antibiotics immediately before and
for several days after ERCP.
Indications/contraindications
Any patient with a clinical picture consistent with cholestasis is a
candidate for imaging of the biliary tree. This is especially true in
patients with underlying inammatory bowel disease. The use of
ERCP or MRCP will be affected by several factors as described above.
If therapy is potentially indicated, then ERCP has the advantage of
treating a stricture without the need for an additional test, although
MRCP may help to plan a therapeutic intervention.
Secondary causes of biliary sclerosis need to be excluded before
and neoplasms, hepatic artery injury, hepatic arterial chemotherapy
and AIDS can lead to strictures in the biliary tree. Figure 35.3 illus-
trates the cholangiogram of a patient several months following intra-
arterial chemotherapy with oxuridine (FUDR) with resultant toxic
cholangiopathy.
Several processes can mimic PSC on a cholangiogram. Hepatic
malignancies polycystic liver disease, inltrative liver disease and
inammatory pseudotumors need to be considered. Abdominal
CAT scan or ultrasound can differentiate many of these disease enti-
ties from PSC.
Another potential role for ERCP in PSC patients may be in pre-
dicting prognosis. A recent study suggests that using a scoring
system for severity of disease based on the initial cholangiogram has
prognostic value, and in combination with the age at rst ERCP, is
strongly predictive of survival.7
Complications
The complications of ERCP in the setting of PSC are typical of those
for any other indication and are described in Chapter 6. There may
379
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

BOX 35.1 KEY POINTSDIAGNOSIS

OF PSC

ERCP MRCP
Invasive Non-invasive
Operator dependent Operator dependent
Gold standard Accuracy < 100%
Therapeutic Non-therapeutic
Tissue sampling No sampling
Stage portal hypertension Less expensive
No complications

Fig. 35.1 Typical endoscopic retrograde cholangiogram of a

patient with PSC. The disease is in an early stage with areas of stric-
turing and beading of the intrahepatic biliary tree but little attenua-
tion. Note that an occlusion cholangiogram has been performed
with the balloon inated proximal to the take off of the cystic duct,
so that contrast lls the intrahepatic bile ducts and not the gallblad-
der. This technique is useful to minimize possible post-procedure
cholecystitis.

Fig. 35.3 Cholangiogram demonstrating the effects of intra-

arterial chemotherapy using FUDR. Note the discrete area of nar-
rowing in the extrahepatic duct which otherwise looks normal,
and the areas of diffuse structuring in the intrahepatics.

Fig. 35.2 Magnetic resonance cholangiogram of a patient with
PSC. The disease is relatively early without much attenuation of the
intrahepatic biliary tree. The biliary radicles appear to be greater in
diameter more peripherally and several discrete strictures are seen
centrally. Both the left and right systems are involved but the extra-
hepatic duct is not well seen. The gallbladder is seen in the center
of the image.
be an increased risk of cholangitis but the use of prophylactic anti-
biotics has become standard care in these patients. A prospective
study from Europe assessed the complications occurring within a
week after ERCP on 83 patients with PSC who underwent 103 pro-
cedures.8 Of the 47 diagnostic procedures the only complication that
occurred was worsening of symptoms in one patient. The vast major-
ity of these patients received prophylactic antibiotics. A similar study
in 104 patients with PSC who underwent diagnostic and therapeutic
ERCP found an 18% complication rate in those with a diagnostic
procedure but these were all mild and not life-threatening.9
Relative cost
A recent study examined the cost of MRCP versus ERCP in the
diagnosis of PSC.10 The average cost per correct diagnosis by MRCP
or ERCP, as the initial testing strategy in 73 patients with clinically
suspected biliary disease, was $724 and $793, respectively. MRCP
had a sensitivity of 82% and a specicity of 98%. MRCP thus resulted
in cost savings when used as the initial test strategy for diagnosing
PSC, particularly as there are essentially no procedure-related com-
plications. However, this was in a cohort of patients with a 32%

380

BOX 35.2 KEY POINTSENDOSCOPIC
THERAPY IN PSC
Dominant strictures in PSC can be treated at ERC.
More important than the stricture is the state of the
pre-stenotic biliary tree.
Tissue sampling and liberal antibiotics are mandatory.
Reserve treatment for patients with symptomatic jaundice.
Concomitant dilation with stenting may improve results.
Balloon dilation and short-term (1014 day) stenting
preferable.
Avoid sphincterotomy if possible.
More complications compared to diagnostic ERCP.
No convincing data we are altering long-term natural
history.
prevalence of PSC and with a very high specicity of MRCP. With a
lower MRCP specicity (<85%) and a higher prevalence of PSC
(>45%), ERCP becomes more cost-effective, suggesting that ERCP
should be used when the suspicion of PSC is high or if local MRCP
facilities are sub-optimal.
The same study illustrated the high cost of dealing with complica-
tions of PSC. The average cost of managing post-ERCP-related com-
plications was $2902 with a range of $1915$5032.
ENDOSCOPIC TREATMENT
Introduction and scientic basis
Interpreting the results of endoscopic therapy trials for PSC is
limited by the small numbers of cases tested and the variety of
endoscopic techniques used. In addition, most series reporting
therapy involve dilation or stenting of a dominant stricturea term
for which there is no consensus, although a stenosis of less
than 1.5 mm in the extrahepatic bile duct, or less than 1 mm in
the right or left common hepatic duct is commonly used. The
status of the upstream bile duct is not considered in this denition
and is critically important in determining the impact of an
intervention.
Repeated endoscopy to maintain biliary patency may improve the
survival of patients with PSC.11 By comparing the survival of 63
patients with PSC who underwent therapeutic ERCP (primarily
repeated balloon dilation of dominant biliary strictures) over a 6-year
period with the predicted survival using the Mayo Clinic survival
model for PSC, Baluyut and colleagues demonstrated an 83% 5-year
survival compared to an expected survival of 65% (p = 0.027).
However, it should be remembered that the Mayo risk score uses
bilirubin in its formula and therefore will be profoundly affected by
stenting of a stricture with resultant rapid decrease in serum biliru-
Chapter 35 Sclerosing Cholangitis
bin. This raises the question whether studies looking at outcome of
therapy in PSC can use this model, which was designed to assess
slow longitudinal decompensation, as a comparison for a control
group following acute endoscopic interventions in highly selected
patients.
A recent study by Bjornsson et al. suggested that cholestasis in
PSC patients did not appear to be related to the presence of domi-
nant strictures.12 They found no difference in the change in choles-
tatic laboratory values in patients with and without dominant
strictures after therapeutic ERCP and hence concluded that endo-
scopic therapy of dominant strictures should not be undertaken
routinely. However, there was very little difference between serum
bilirubin and alkaline phosphatase levels between patients with and
without dominant strictures prior to therapeutic ERCP casting doubt
as to the validity of their assessment of what constituted a dominant
stricture.
The paucity of controlled data indicates that it is unclear whether
endoscopic therapy alters the long-term natural history of PSC.
Description of technique
Once biliary cannulation has been achieved there are a variety of
instruments that can be used to perform stricture dilation (see
Chapter 30). Wire access across strictures is the rst step in therapy
and soft tip wires with diameters of 0.0180.035 must be used to
avoid perforation of the biliary tree. Push catheters have a tapered tip
and the dilating part is typically 710 Fr in diameter. They are wire-
guided but their limited diameter and limited radial force mean that
they are seldom used in PSC. More commonly inatable balloons are
employed. These are also wire-guided but come in a variety of diam-
eters (up to 12 mm) and have a greater radial force. They are difcult
to use if the stricture is tortuous. A temporary plastic stent can be
placed after dilation or in some cases can be placed without dilation.
Any dominant stricture should undergo sampling for cholangiocar-
cinoma. We perform brush cytology and/or intraductal forceps
biopsy. In cases of refractory strictures, a screw catheter can be used
over a wire although there is no controlled data on its efcacy.
Several studies have performed biliary sphincterotomy prior to
dilation or stent insertion but we do not advocate this as there is no
reliable data that this is required and the complication rate is
undoubtedly higher.
Initial studies in PSC patients with dominant strictures did not
use a standardized technique. Van Milligen de Wit and colleagues
demonstrated technical success in 21 of 25 PSC patients with a
dominant stricture who underwent endoscopic stent therapy.13 Of
these 25 patients, 18 had a biliary sphincterotomy and 9 underwent
dilation prior to stent insertion with either a balloon or dilating
catheter. Stents were exchanged electively every 23 months or if
they became occluded. After a median follow-up of 29 months, 16
of the 21 patients had improved or stable liver tests.
More recent data indicates that short-term stenting may be effec-
tive with the benet extending for several years.14 A European study
described 32 PSC patients with dominant strictures treated with
plastic stent placement for a mean of only 11 days. Again the tech-
nique was heterogeneous with some patients undergoing biliary
sphincterotomy and push dilation, and both 7 Fr and 10 Fr stents
were used. Improvements in symptoms and cholestasis were seen
in all patients and these improvements were maintained for several
years, with 80% of patients intervention free at 1 year, and 60% at 3
years. There were 7 transient procedure-related complications out of
45 procedures but all but one was managed conservatively.
381
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

The addition of ursodeoxycholic acid (UDCA) to endoscopic

A
therapy has been examined in a prospective trial by Stiehl et al., in
106 PSC patients followed for up to 13 years.15 All patients received
UDCA along with balloon dilation of dominant strictures or place-
ment of short-term biliary stents whenever necessary. Ten patients
had dominant strictures at the beginning of the trial and a further
43 patients developed dominant strictures during the follow-up
period. This combined approach appeared to improve overall
survival rates compared with the Mayo PSC survival model but as
the study was not controlled it was unclear whether the UDCA,
the endoscopic therapy or the combination of the two improved
outcome.
Our approach in PSC patients is to use therapeutic ERCP in
patients with a dominant stricture with pre-stenotic dilation who B
have an elevated serum bilirubin or severe symptomatic cholestasis.
All our PSC patients are on UDCA at a dose of 2025 mg/kg. We
use balloon dilation over a guidewire ensuring that the diameter of
the balloon is no greater than the smallest diameter of the duct either
proximal or distal to the stricture. The balloon is inated until there
is no waist to the balloon or to a pressure of 12 atmospheres. We
then leave a 10 Fr plastic stent across the stricture for 23 weeks and
then repeat an ERCP and provide further therapy if indicated. All
procedures are covered with prophylactic antibiotics and we will use
several days of oral antibiotics after the procedure to minimize the
risk of cholangitis. Figures 35.4 and 35.5 illustrate dilation and stent
therapy in patients with a dominant stricture.

Indications/contraindications Fig. 35.4 A Balloon dilation of a dominant stricture in a patient

Endoscopic therapy is indicated in PSC patients if there is clinical
or biochemical evidence of cholangitis or if a dominant stricture is
suspected. However, due to lack of controlled data it is unclear if
treatment alters the natural history of the disease. Occasionally bile
duct calculi are encountered in patients with PSC and can be
removed using standard techniques (see Chapter 13), although
stones proximal to a stricture can be challenging. We typically do
not perform therapy in patients who are not clinically jaundiced
as the complication rate is higher and the potential benet is
questionable.
with PSC. The cholangiogram on the left demonstrates intrahepatic
and extrahepatic PSC. The extrahepatic duct has a dominant stric-
ture (just below the scope) with pre-stenotic dilation. Retained con-
trast in the pancreatic duct is seen. The cholangiogram on the right
shows a deated balloon introduced into the extrahepatic duct
over a wire. Note the radiopaque markers on the proximal and
distal ends of the balloon. B Balloon dilation of a dominant stric-
ture in a patient with PSC. The cholangiogram on the left demon-
strates the balloon inated across the stricture. There is no waist to
the balloon. Typically the balloon is inated to 12 atmospheres for
3045 seconds. This can lead to pain and additional sedation may
need to be given. The post-dilation appearance is shown on the
right. Note the marked improvement.

Complications and their management

Complications after therapeutic procedures in PSC patients are
more frequent than after diagnostic ERCP. Of the 10 complications
after 106 procedures studied by van den Hazel et al., all but one was
in therapeutic cases.8 These included two episodes of cholangitis,
three episodes of pancreatitis, a post-sphincterotomy bleed, a perfo-
ration, an upper extremity venous thrombosis, and a single patient
with worsening of symptoms.
Of the two patients with cholangitis, one had not received prepro-
cedure antibiotics, but both responded to treatment. The pancreatitis
patients were treated conservatively and all recovered. The post-
sphincterotomy bleed occurred three days after a needle-knife
sphincterotomy and was mild. It was treated endoscopically. The
perforation was asymptomatic but resulted from a guidewire making
a false route in the cystic duct. The patient with worsening symp-
toms had had a stent placed and this needed a second ERCP to
remove the stent.
Factors that appeared to increase the risk of complications
included a therapeutic indication such as jaundice or recurrent chol-
angitis, and also if the patient had undergone ERCP previously.
These patients had a 14% complication rate. A similar number of
early complications have been noted by other investigators with
cholangitis usually the most common.913 These are usually easily
treated with antibiotics but liver abscesses and septic shock have
been reported. Prolonged stent therapy is associated with cholangitis
or jaundice due to stent occlusion that can be successfully treated
by stent exchange or removal. Studies using short-term stent therapy
indicate a similar early complication rate but cholangitis is less
frequent.14
Relative cost
There is no cost-effectiveness data available in studies of therapeutic
endoscopy in PSC patients. Although the equipment used in balloon
dilation is more expensive than push dilators the likely increased
efcacy of the former may translate into fewer follow-up procedures
and hence reduce the cost. We typically always deploy a stent after
dilation and this necessitates a repeat procedure in a few weeks that
adds to the cost of treatment. Due to the lack of controlled trials in
PSC patients undergoing therapeutic ERCP, it is unclear if dilation
therapy alone is sufcient. Reducing the number of subsequent

382
 Chapter 35 Sclerosing Cholangitis

A B C D

E F G H

Fig. 35.5 Diffusely irregular distal bile and common hepatic ducts in deeply jaundiced patient with
PSC, pigment stones in the biliary tree A. Note intrahepatic duct changes. Following biliary sphincter-
otomy and stone extraction B,C,D the strictures are balloon dilated E, brushed for cytology F, and
stented with a 10 Fr biliary stent G. Note persistent but improved stenosis at time of stent retrieval 4
weeks later H,I.

procedures without affecting long-term outcome would improve the
cost-effectiveness of endoscopic treatment.
CHOLANGIOCARCINOMA
Introduction and scientic basis
Cholangiocarcinoma (CCA) will develop in up to 1030% of patients
with PSC, with a lifetime risk of 1015%. A recent study determined
the incidence and risk factors for CCA in 161 patients with PSC,
monitored for a median of 11.5 years. CCA developed in 11 patients
(6.8%) at a rate of approximately 0.6% per year.16 This equated to a
relative risk of CCA compared with that in the general population
of 1560. No association was found between the duration of PSC and
the incidence of CCA. Similarly, in a Swedish cohort of 604 PSC
patients followed for many years, the frequency of CCA was 13%
and the incidence rate of CCA after the rst year of diagnosis was
1.5% per year.17
Early diagnosis of CCA may improve patient survival as it may
permit curative surgical resection, but it is hampered by the absence
of sufciently accurate and non-invasive diagnostic tests. Diagnos-
ing CCA in PSC patients is even more challenging because of the
presence of multiple non-neoplastic strictures.
Description of technique
Diagnosis
Several endoscopic methods have been employed to try and diagnose
CCA in PSC patients. Brush cytology, ne needle aspiration and
forceps biopsy have been used but all have low sensitivity and high
specicity. In addition, the tumor markers carbohydrate antigen
19-9 (CA19-9) and carcinoembryonic antigen (CEA) have been
examined alone or in various combinations.
Brush cytology involves gaining access to the biliary tree and
then inserting a wire into the intrahepatic ducts. The closed
cytology brush is then advanced over the wire into the area of
the stricture to be brushed and then opened and vigorously pushed
in and out of the stricture to try and increase the cellular yield.
Occasionally, push or balloon dilation of the stricture is required
to enable passage of the cytology brush. The brush is then closed
while still inside the duct and removed through the endoscope
and the cellular specimen sent to cytology on preprepared slides.
It is important for the nurse or endoscopy technician to prepare
the cytology slides quickly to prevent excessive drying which can
cause artifacts that affect interpretation. The overall sensitivity of
cytology in the diagnosis of CCA in PSC is around 50%. Adding
K-ras or p53 mutational analysis of brush samples does not

383
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 35.6 Intraductal biopsy in a patient with PSC and a dominant
stricture. The scope is in a short position and the forceps have been
placed into the bile duct. The bile duct proximal to the stricture is
irregular and dilated. This technique enables a biopsy of the stric-
ture such that tissue can be obtained for pathology rather than
cytology, improving the diagnostic yield.
appear to increase the sensitivity although repeated brushing on
two or three occasions appears to increase the sensitivity
signicantly.18,19
The increased cellular load obtained by an intraductal biopsy also
appears to increase the sensitivity. The technique involves cannulat-
ing directly with the biopsy forceps. Figure 35.6 illustrates a biopsy
being taken in a patient with a suspicious stricture.
Using a formula with the sum of the CA19-9 and 40 times the
CEA, Ramage et al. concluded that a diagnostic accuracy of 86%
could be achieved in making a diagnosis of CCA in a small cohort
of PSC patients.20
We have shown that a CEA level of >5.2 ng/ml or a CA19-9 of
>180 U/ml leads to 70% sensitivity and combined with brush cytol-
ogy a sensitivity of close to 90% can be achieved.21 Similarly, combin-
ing brush cytology, DNA analysis by ow cytometry, CA19-9 and
CEA, a diagnostic sensitivity of 88% and specicity of 80% can be
expected. Interestingly, measurement of CA19-9 or CEA in bile has
no diagnostic signicance.22
Recently, several newer techniques have shown some promise in
differentiating benign from malignant disease in PSC. Quantifying
nuclear DNA content from a stricture using digital image analysis
(DIA) provides a diagnostic accuracy equivalent to brush cytology.23
Analysis of tumor suppressor gene linked microsatellite marker loss
of heterozygosity and detection of k-ras mutation in brush cytology
samples also yielded excellent sensitivity and specicity in a small
group of patients with biliary strictures.24
Screening patients with PSC for CCA with CA19-9 and CEA
would appear to be reasonable, but the ideal interval at which
to obtain these tests and the cost-effectiveness remain to be
determined.
384
Treatment and palliation
Until the emergence of ERCP, patients who presented with jaundice
in the setting of malignant biliary obstruction required surgical
biliary bypass if t enough for surgery, or percutaneous drainage.
Smith and colleagues demonstrated the efcacy of endoscopic stent
insertion compared to surgical biliary bypass in a randomized pro-
spective controlled trial of 204 patients with malignant low bile duct
obstruction.25 Technical success was achieved in 94 surgical and 95
stented patients, with functional biliary decompression obtained in
92 patients in both groups. The overall survival between the two
groups did not differ (median survival: surgical 26 weeks; stented 21
weeks). The authors concluded that endoscopic stenting and surgery
were both effective palliative treatments with the former having
fewer early treatment-related complications and the latter fewer late
complications.
For patients with unresectable disease an expandable metal stent
can be deployed for palliation using a standard technique (see
Chapter 17). In a prospective randomized trial, Davids et al. demon-
strated that metal stents resulted in signicantly prolonged patency
compared to polyethylene stents in 105 patients with unresectable
distal bile-duct malignancy.26 Median patency of the metal stent was
273 days compared to 126 days with a plastic stent. Tumor ingrowth
typically led to occlusion in metal stents whereas sludge deposition
caused occlusion with plastic stents. However, the overall median
survival was 149 days and did not differ signicantly between patients
with metal or plastic stents.
The addition of chemotherapy to stenting has been tried without
much success but a recent study demonstrated the benet of photo-
dynamic therapy (PDT) and stenting in non-resectable proximal
CCA (involving the hilum) compared to stenting alone.(27) The
technique requires initial placement of plastic stents (see Chapter
16) into both the left and right intrahepatics so that biliary drainage
is achieved. Patients receiving PDT then are treated with Photofrin
at a dose of 2 mg/kg intravenously 48 hours before laser activation.
The plastic stents are then removed during a repeat ERCP and intra-
luminal photoactivation is performed, following which the plastic
stents are redeployed. Patients are kept in a darkened room for 34
days following the procedure. In this randomized study, PDT and
stenting resulted in a median survival of 493 days compared to 98
days with stenting alone. In addition, jaundice and quality of life
were also signicantly improved. The only complication seen in the
PDT group was photosensitivity in 10% of the patients. This proce-
dure should only be performed by experienced endoscopists and in
facilities with PDT capability. The costs may also be prohibitive.
Indications/contraindications
Any patient with PSC who has an unexpected rise in cholestatic
enzymes or bilirubin should be investigated for the development of
CCA. In addition, a sudden rise in CA19-9 or CEA should prompt a
cholangiogram. We do not advocate routine surveillance ERCP in
asymptomatic patients with PSC. Diagnosing cholangiocarcinoma
at an early stage is not usually helpful in PSC patients as resection
is usually contraindicated because of the presence of cirrhosis, and
the results of liver transplant for early CCA are not encouraging. The
diagnosis should ideally be made in a premalignant stage but as yet
this is not possible. The timing of liver transplant in patients with
PSC is therefore difcult as many patients will have preserved liver
synthetic function and are early for transplant but are still at risk of
developing a tumor that will then likely preclude transplant.

Complications and their management
The complications and management of metal or plastic stent inser-
tion for patients with CCA in the setting of PSC are as described
elsewhere (see Chapters 16 and 17).
Comparing stent insertion to surgical bypass in CCA, Smith et
al. demonstrated a lower procedure-related mortality (3% vs 14%),
major complication rate (11% vs 29%), and median total hospital
stay (20 vs 26 days) in stented patients compared to surgery. However,
late complications including recurrent jaundice and late gastric
outlet obstruction were more frequent in stented patients.25
Relative cost
For diagnostic purposes, CEA and CA19-9 measurement are rela-
tively inexpensive but their low sensitivity means cytology is required.
The cost-effectiveness of these tumor markers and the newer DNA-
based tests remains to be seen.
In terms of providing palliation, metal stents are more cost-
effective than plastic stents because the longer patency compared to
plastic stents translates into fewer follow-up procedures. Although
there are no formal studies looking at risks and benets of plastic
versus metal stents in PSC patients who develop CCA, Davids et al.
using incremental cost-effectiveness analysis, showed that intial
placement of a metal stent resulted in a 28% decrease in subsequent
endoscopic procedures in patients with distal malignant obstructive
jaundice.26 However, in patients with a life expectancy of less than
3 months a plastic stent may be adequate palliation as a follow-up
procedure is unlikely to be required.
REFERENCES
1. Aadland E, Schrumpf E, Fausa O, et al. Primary sclerosing
cholangitis: a long-term follow-up study. Scand J Gastroenterol
1987; 22:655664.
2. Quigley EM, LaRusso NF, Ludwig J, et al. Familial occurrence of
primary sclerosing cholangitis and ulcerative colitis. Gastroenterol
1983; 85:11601165.
3. Boberg KM, Bergquist A, Mitchell S, et al. Cholangiocarcinoma in
primary sclerosing cholangitis: risk factors and clinical
presentation. Scand J Gastroenterol 2002; 37:12051211.
4. Cullen SN, Chapman RW. Review article: current management of
primary sclerosing cholangitis. Alimen Pharmacol Therap. 2005;
21:933948.
5. Textor HJ, Flacke S, Pauleit D, et al. Three-dimensional magnetic
resonance cholangiography with respiratory triggering in the
diagnosis of primary sclerosing cholangitis: comparison with
endoscopic retrograde cholangiography. Endoscopy 2002;
34:984990.
6. Angulo P, Pearce DH, Johnson CD, et al. Magnetic resonance
cholangiography in patients with biliary disease: its role in
primary sclerosing cholangitis. J Hepatol 2000; 33:520527.
7. Ponsioen CY, Vrouenraets SM, Prawirodirdjo W, et al. Natural
history of primary sclerosing cholangitis and prognostic value
of cholangiography in a Dutch population. Gut 2002;
51:562566.
8. Van den Hazel SJ, Wolfhagen EH, van Buuren HR, et al.
Prospective risk assessment of endoscopic retrograde
cholangiography in patients with primary sclerosing cholangitis.
Endoscopy 2000; 32:779782.
9. Enns R, Eloubeidi MA, Mergener K, et al. Predictors of successful
clinical and laboratory outcomes in patients with primary
sclerosing cholangitis undergoing endoscopic retrograde
Chapter 35 Sclerosing Cholangitis
The high initial cost of surgical palliation compared to endo-
scopic therapy means that the former has very limited application
(Box 35.3).
BOX 35.3 KEY POINTSDIAGNOSING
CHOLANGIOCARCINOMA IN PSC
Cholangiocarcinoma may develop in 1030% patients with
PSC.
No proven screening method.
Brush cytology has only 50% sensitivity.
Combining tumor markers with cytology may increase
sensitivity.
Newer DNA and molecular techniques may increase
sensitivity.
Pre-malignant diagnosis would allow liver transplant prior to
development of cholangiocarcinomanot currently
possible.
cholangiopancreatography. Canadian Journal of Gastroenterology
2003; 17:243248.
10. Talwakar JA, Angulo P, Johnson CD, et al. Cost-minimization
analysis of MRC versus ERCP for the diagnosis of primary
sclerosing cholangitis. Hepatology 2004; 40:3945.
11. Baluyut AR, Sherman S, Lehman GA, et al. Impact of endoscopic
therapy on the survival of patients with primary sclerosing
cholangitis. Gastrointest Endosc 2001; 53:308312.
12. Bjornsson E, Lindqvist-Ottosson J, Asztely M, et al. Dominant
strictures in patients with primary sclerosing cholangitis. Am J
Gastroenterol 2004; 99:502508.
13. Van Milligen de Wit AWM, van Bracht J, Rauws EAJ, et al.
Endoscopic stent therapy for dominant extrahepatic bile duct
strictures in primary sclerosing cholangitis. Gastrointest Endosc
1996; 44:293299.
14. Ponsioen CY, Lam K, van Milligen de Wit AWM, et al. Four years
experience with short term stenting in primary sclerosing
cholangitis. Am J Gastroenterol 1999; 94:24032407.
15. Stiehl A, Rudolph G, Kloters-Plachky P, et al. Development of
dominant bile duct stenoses in patients with primary sclerosing
cholangitis treated with ursodeoxycholic acid: outcome after
endoscopic treatment. J Hepatol 2002; 36:151156.
16. Burak K, Angulo P, Pasha TM, et al. Incidence and risk factors for
cholangiocarcinoma in primary sclerosing cholangitis. Am J
Gastroenterol 2004; 99:523526.
17. Bergquist A, Ekbom A, Olsson R, et al. Hepatic and extrahepatic
malignancies in primary sclerosing cholangitis. J Hepatol 2002;
36:321327.
18. Ponsioen CY, Vrouenraets SME, van Milligen de Wit AWM, et al.
Value of brush cytology for dominant strictures in primary
sclerosing cholangitis. Endoscopy 1999; 31:305309.
385
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
19. de Bellis M, Fogel EL, Sherman S, et al. Inuence of stricture
dilation and repeat brushing on the cancer detection rate of
brush cytology in the evaluation of malignant biliary obstruction.
Gastrointest Endosc 2003; 58(2):176182.
20. Ramage JK, Donaghy A, Farrant JM, et al. Serum tumor markers
for the diagnosis of cholangiocarcinoma in primary sclerosing
cholangitis. Gastroenterol 1995; 108:865869.
21. Siqueira E, Schoen RE, Silverman W, et al. Detecting
cholangiocarcinoma in patients with primary sclerosing
cholangitis. Gastrointest Endosc. 2002; 56:4047.
22. Lindberg B, Arnelo U, Bergquist A, et al. Diagnosis of biliary
strictures in conjunction with endoscopic retrograde
cholangiopancreaticography, with special reference to patients
with primary sclerosing cholangitis. Endoscopy 2002; 34:909916.
23. Baron TH, Harewood GC, Rumalla A, et al. A prospective
comparison of digital image analysis and routine cytology for the
386
identication of malignancy in biliary tract strictures. Clin
Gastroenterol Hepatology. 2004; 2:214219.
24. Khalid A, Pal R, Sasatomi E, et al. Use of microsatellite marker loss
of heterozygosity in accurate diagnosis of pancreaticobiliary
malignancy from brush cytology samples. Gut 2004;
53:18601865.
25. Smith AC, Dowse JF, Russell RC, et al. Randomized trial of
endoscopic stenting versus surgical bypass in malignant low bile
duct obstruction. Lancet 1994; 344:16551660.
26. Davids PH, Groen AK, Rauws EA, et al. Randomized trial of
self-expanding metal stents versus polyethylene stents for
distal malignant biliary obstruction. Lancet 1992; 34:
14881492.
27. Ortner MEJ, Caca K, Berr F, et al. Successful photodynamic therapy
for nonresectable cholangiocarcinoma: A randomized
prospective study. Gastroenterol 2003; 125:13551363.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Choledochal Cysts
36 Annie On On Chan, Chi Leung Liu and Benjamin Chun-Yu Wong
INTRODUCTION AND SCIENTIFIC BASIS
Choledochal cysts are congenital anomalies characterized by cystic
dilatation of the biliary tree, extrahepatic or intrahepatic, or both
extrahepatic and intrahepatic. Choledochal cysts are classied into
ve subtypes under the Todani classication, based on the anatomi-
cal site of the cystic dilatation1 (Fig. 36.1). They are more common
in women than in men, and their prevalence is higher in Asian and
Japanese descendants than in Caucasians. For additional images
and discussion see Chapters 22 and 42.
BOX 36.1 INDICATIONS/CONTRA-
INDICATIONS OF ERCP FOR
CHOLEDOCHAL CYST
Indications
Diagnosis
Preoperative assessment
Therapeutics
Management of complications
Contraindications:
Coagulopathy
Unt to undergo sedation
Clinical perforation
Usual contraindications to ERCP
Choledochal cysts are thought to be the result of an anomalous
junction of the common bile duct with the pancreatic duct (anoma-
lous pancreatobiliary junction (APBJ)) (Figs. 36.2A, 36.2B). An APBJ
is characterized when the pancreatic duct enters the common bile
duct 1 cm or more proximal to where the common bile duct reaches
the ampulla of Vater. Miyano and Yamataka2 have demonstrated
such APBJs in more than 90% of their patients with choledochal
cysts. The APBJ allows pancreatic secretions and enzymes to reux
into the common bile duct. In the relatively alkaline conditions
found in the common bile duct, pancreatic proenzymes can be acti-
vated, resulting in inammation and weakening of the bile duct wall.
Severe damage leads to complete denuding of the common bile duct
mucosa. Another proposed theory for the formation of choledochal
cysts is the presence of defects in epithelialization and recanalization
of the developing bile ducts during organogenesis resulting in con-
genital weakness of the duct wall.
DESCRIPTION OF TECHNIQUE
Choledochal cysts can present at any age. With the increasing use
of imaging studies in clinical management, more cases are diag-
nosed incidentally. Presenting symptoms are age-dependent with
jaundice prevailing in children and abdominal pain in adults. In
view of the high risk of cholangiocarcinoma, early resection and not
internal drainage is the appropriate treatment of extrahepatic cysts.
Patients who underwent internal drainage in the past still should
undergo resection of the cyst. Neonates and children usually present
with abdominal pain and/or abdominal mass3 and jaundice.4 Initial
presentation in adults is rare, but, when present, may cause non-
specic right upper quadrant abdominal pain, jaundice, acute chol-
angitis, or acute pancreatitis.4,5 Hence, prior to the advent of MRCP,
endoscopic retrograde cholangiopancreatography (ERCP) was one of
the most commonly used methods for the diagnosis of choledochal
cyst, especially in the initial management of acute cholangitis, acute
pancreatitis and biliary malignancies.
ERCP can be performed with a standard diagnostic or therapeutic
channel duodenoscope. After identication of the papilla, cannula-
tion of the common bile duct (CBD) should be performed using 30%
water-soluble contrast solution. Immediate and delayed spot lm
radiographs should be obtained (Figs 36.3A, 36.3B).
Sphincterotomy and biliary stent placement are performed as
needed. Bile duct stones, if present, are extracted with a Dormia
basket and/or balloon catheters as in patients with non-variant
anatomy as described in Chapters 13 and 33. Biopsy specimens are
obtained if a tumor mass is suspected with conventional or specially
designed biopsy forceps (Fig. 36.4).
Choledochal cyst is mainly a disease seen in children (see Chapter
22). The technique of ERCP is no more difcult in children than in
adults, and there is no need for a special pediatric endoscope, except
in infants younger than 12 months of age. The rate of successful
cannulation of the desired duct in children is as good as in adults,
and is reported to be in the range of 9296%.6 Occasionally, ERCP
via the accessory papilla is an effective method for visualization of
the detailed structure of the entire pancreatic ductal system and
junction of the pancreatic and biliary ducts when ERCP via the major
duodenal papilla is unsuccessful.7 Most endoscopists use general
anesthesia with an endotracheal tube in situ in children, particularly
in infants, to prevent airway obstruction and dyspnea caused by
abdominal distension. There are also reports of conducting ERCP
under ketamine sedation. It has been shown to be safe and effective
and does not require an extensive monitoring system.8 The compli-
cation rate of performing ERCP in children is reported to be less
than that in adults.9
387
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

I II III Fig. 36.1 Todani classication of choledochal

cysts. (Redrawn with permission from de
Vries JS et al. J Pediatr Surg. 2002:37(11);1568
1573.)

IVa V

Fig. 36.2 A Initial injection during ERCP

A showing a long common channel in a
patient with type IVA choledochal cyst and
B long bile duct stricture. B Same patient.
The long stricture is seen with cystic dilata-
tion above the stricture.

Fig. 36.3 A Initial injection into a massive

A type I choledochal cyst (arrow). B Late
lling clearly demonstrates cyst.

388

Fig. 36.4 Biopsy of stricture depicted in Figure 36.2B.
Cystic dilatation is known to occur in all parts of the biliary
system (see Chapter 42). Todanis classication (Fig. 36.1) comprises
ve types of cysts:
Type I Common type: (a) choledochal cyst in a narrow sense; (b)
segmental choledochal dilatation; and (c) diffuse or cylin-
drical dilatation.
Type II Diverticulum type in the extrahepatic duct
Type III Choledochocele
Type IVa Multiple cysts in the intra- and extrahepatic ducts
Type IVb Multiple cysts in the extrahepatic duct only
Type V Intrahepatic bile duct cyst only (single or multiple)
For the technique on diagnosing choledochocele (type III chole-
dochal cyst), it has been recommended that patients fulll these
criteria: (a) a radiolucent halo around the distal end of the CBD, (b)
bulbous end of the distal CBD, and (c) dynamic sequential morpho-
logic changes of the distal CBD. Patients who meet two or more of
these criteria are diagnosed as having choledochocele on cholangi-
ography. In addition, the presence of a waist, a pseudoweb, and
wrinkling of the distal CBD may also be observed.10 Characteristic
duodenoscopic features of choledochocele include a hemispherical
or pear-shaped bulge protruding into the duodenal lumen, soft and
smooth overlying mucosa, easy compressibility with a cannulating
catheter, and visible ballooning of the ampulla during injection of
contrast (Figs 36.5A, 36.5B).11 Park et al.12 recommended that the
duodenal lumen should be lled with an optimal amount of contrast
material. The patient should be positioned so that the dilated sac
Chapter 36 Choledochal Cysts
does not overlap with the second duodenal portion or the endoscope
in order to identify the dilated end of the distal CBD in the intramu-
ral portion (Fig. 36.5C). A radiolucent halo can be seen between the
contrast-lled dilated sac and the contrast-lled duodenal lumen.
Cholangiography should be done in prole. When adequate cholan-
giographic images are combined with a careful ERCP examination,
it is possible to diagnose small choledochoceles that may otherwise
be overlooked by inspection during ERCP alone. Optimal radio-
graphic technique during ERCP is necessary and it has been pro-
posed that a cannulating catheter should be withdrawn from the
duodenum after contrast material has been instilled into the bile
duct. This is necessary to avoid masking the dilated sac by the
balloon and the catheter. In patients with a choledochocele sphinc-
terotomy is performed by cannulating both the cyst and the common
bile duct with a papillotome (Fig. 36.5D). The latter allows simulta-
neous unroong of the cyst and sphincterotomy. In cases where a
case of a large choledochocele is suspected needle knife sphincter-
otomy appears to be safe.13
DIAGNOSIS
As mentioned above, the presentation of choledochal cyst in adults
is usually atypical, and individuals seldom present with an abdomi-
nal mass. Instead, they usually present with biliary complications.
Historically, the nal diagnosis of choledochal cyst has been estab-
lished by ERCP as it was considered the only method to precisely
dene the extent of the cyst as well as conrm the presence or
absence of an anomalous pancreaticobiliary junction.
PREOPERATIVE ASSESSMENT
ERCP not only provides an accurate diagnosis, but also provides
useful anatomical details that help to plan appropriate surgical inter-
vention and obviates an intraoperative cholangiogram. Traditionally,
preoperative ERCP is used to determine the upper and lower margin
of a cyst and provide detailed anatomy of the biliary tract and pan-
creaticobiliary junction before is contemplated. More recently,
MRCP has been shown to be an effective imaging modality for
diagnosis and preoperative evaluation of choledochal cysts (Figs
36.6A, 36.6B).14,15
THERAPEUTIC MEASURES
Due to cystic formation, stones and strictures may form within the
biliary tree. Therapeutic ERCP should be carried out for biliary de-
compression, stone removal, stricture dilatation, stent insertion, and
sphincterotomy. In children with choledochal cysts, surgery has tra-
ditionally been performed soon after ERCP because of the risk of
cholangitis. Malignancy may arise in long-standing choledochal cysts.
Hence brush cytology or biopsy from a suspected tumor mass should
be undertaken during ERCP. The current treatment of choledochal
cysts is essentially surgical, irrespective of the age of the patient. Endo-
scopic treatment is the method of choice for uncomplicated choled-
ochocele (type III) (Fig. 36.5),16,17 after which long-term follow-up is
unnecessary.18 Choledochoceles are difcult to distinguish from dupli-
cation cysts of the duodenum unless the mucosal lining is examined
histologically. Endoscopic sphincterotomy or unroong the cyst to
permit adequate drainage are common methods of treatment.19
However, carcinoma may rarely co-exist with choledochocele, and
ERCP with biopsies may be needed to exclude malignancy.
389
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B

C D

Fig. 36.5 Choledochocele. A The choledochocele (arrow) can be seen above the
papilla (arrowhead). B The choledochocele is seen to enlarge as contrast is injected into
the bile duct. C Characteristic cholangiographic appearance of choledochocele. D En-
doscopic appearance immediately after biliary sphincterotomy.

MANAGEMENT OF COMPLICATIONS FROM
CHOLECHODAL CYSTS
Usually, total choledochal cyst resection followed by Roux-en-Y
hepaticojejunostomy is the preferred treatment in most types of
choledochal cysts (Fig. 36.7). Internal cyst drainage is associated with
an increased risk of cholangitis,20 pancreatitis,21 and biliary tract
cancer.22,23 Patients with choledochal cysts who have undergone
surgery remain at increased risk for recurrent cholangitis, pancre-
atitis, intrahepatic strictures, stones, and malignancy. ERCP will be
indicated if these complications arise, assuming the hepaticojeju-
nostomy can be reached (Fig. 36.8). Long-term management of
patients with Carolis disease, however, remains controversial.
Usually this involves treatment of recurrent cholangitis with even-
tual surgical intervention, including external abscess drainage,
partial resection, and even orthotopic liver transplantation.24,25 The
time interval between diagnosis and surgery is, however, highly
variable. Alternatively, bile stasis can be prevented by internal drain-
age via ERCP with endoscopic sphincterotomy, biliary stent place-
ment, and stone extraction. Gold et al.25 described placement of an
endobiliary stent in the right liver lobe alone for two years in a
patient with type V choledochal cyst, with a good clinical response.
Ciambotti et al.26 treated a patient with monolobar Carolis disease
and multiple intrahepatic stones by stent placement for one year,

390
 Chapter 36 Choledochal Cysts

Fig. 36.6 MRCP of choledochal cyst.

A B A Long common channel corresponds to
Figure 36.2a. B Intrahepatic cystic dilata-
tion and extrahepatic stricture corresponds
to Figure 36.2b.

A B

Fig. 36.7 A Roux-en-Y hepaticojejunostomy after resection of

extrahepatic cyst. B Temporary percutaneous transhepatic stents
may be placed across the hepaticojejunostomy.

together with administration of ursodeoxycholic acid (UDCA), until

the stone burden was eliminated.

COMPLICATIONS AND THEIR MANAGEMENT

Choledochal cysts should be completely resected if possible because
of the long-term consequences of cholangitis, liver cirrhosis, pancre-
atitis, and cancer.27 These problems can be exacerbated if internal
drainage procedures rather than cyst resection are performed.
Pancreatitis is a common presentation of choledochal cysts which
may be due to the activation of pancreatic enzymes by bile reux. In
addition, an APBD not only predisposes to biliary tract anomalies
but also causes pancreatic duct abnormalities such as dilatation,
protein plugs and stones.28 The association between pancreatic
calculi and choledochal cysts has been recognized. Pancreatic calculi
can cause recurrent or chronic pancreatitis after cyst excision.29
Guelrud et al.30 showed that sphincter of Oddi dysfunction may be
associated with APBD and choledochal cysts and could be a cause
of acute recurrent pancreatitis. In addition, patients with choledochal
cysts may be more prone to developing post-ERCP pancreatitis
because of the presence of the long common channel (Fig. 36.2A).
Similarly, some have postulated that endoscopic placement of a stent
Fig. 36.8 PTC from patient illustrated in Figure 36.2b after hepati-
cojejunostomy for management of cholangitis. ERCP failed to reach
the hepaticojejunostomy.
through the long common channel may increase the high risk of
pancreatitis, though there are no data to support this observation.
Cholangitis is a common complication of choledochal cysts and
may be the presenting feature, as mentioned earlier. It is also a
commonly reported complication after surgical management.
Strictures of the hepatic ducts have been attributed to recurrent
infection, especially in patients with type V cysts, a group in which
strictures occur in up to 50% of patients.31 Matsumoto et al.32 and

391
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Ando et al.33 suggest that concomitant congenital strictures contrib-
ute to dilatation of the intrahepatic bile ducts. These investigators
recommend more extensive evaluation using choledochoscopy, as
the strictures may be characteristic of choledochal cyst disease.
REFERENCES
1. Todani T, Urushihara N, Morotomi Y, et al: Characteristics of
choledochal cysts in neonates and early infants. Eur J Pediatr Surg
1995; 5:143145.
2. Miyano T, Yamataka A. Choledochal cysts. Curr Opin Pediatr 1997;
9:283288.
3. Stringer MD, Dhawan A, Davenport M, et al. Choledochal cysts:
lessons from a 20 year experience. Arch Dis Child. 1995;
73:528531.
4. de Vries JS, de Vries S, Aronson DC, et al. Choledochal cysts: age
of presentation, symptoms, and late complications related to
Todanis classication. J Pediatr Surg. 2002 Nov; 37(11):15681573.
5. Lipsett PA, Pitt HA, Colombani PM, et al. Choledochal cyst disease.
A changing pattern of presentation. Ann Surg. 1994; 220:644652.
6. Hsu RK, Draganov P, Leung JW, et al. Therapeutic ERCP in the
management of pancreatitis in children. Gastrointest Endosc.
2000 Apr; 51(4 Pt 1):396400.
7. Kouchi K, Yoshida H, Matsunaga T, et al. Efcacy of ERCP via the
accessory papilla in children with choledochal cysts. Gastrointest
Endosc. 2004; 59:119123.
8. Aggarwal A, Ganguly S, Anand VK, et al. Efcacy and safety of
intravenous ketamine for sedation and analgesia during pediatric
endoscopic procedures. Indian Pediatr. 1998 Dec;
35(12):12111214.
9. Shirai Z, Toriya H, Maeshiro K, et al. The usefulness of endoscopic
retrograde cholangiopancreatography in infants and small
children. Am. J. Gastroenterol. 1993; 88:536541.
10. Park KB, Auh YH, Kim JH, et al. Diagnostic pitfalls in the
cholangiographic diagnosis of choledochoceles:
cholangiographic quality and its effect on visualization. Abdom
Imaging 2001 JanFeb; 26(1):4854.
11. Kim MH, Myung SJ, Lee SK, et al. Ballooning of the papilla during
contrast injection: the semaphore of a choledochocele.
Gastrointest Endosc 1998; 48:258262.
12. Park KB, Auh YH, Kim JH, et al. Diagnostic pitfalls in the
cholangiographic diagnosis of choledochoceles:
cholangiographic quality and its effect on visualization. Abdom
Imaging 2001; 26:4854.
13. Katsinelos P, Dimiropoulos S, Galanis I, et al. Needle-knife
sphincterotomy. Surg Endosc 2003 Jan; 17(1):158.
14. Shaffer E. Can MRCP replace ERCP in the diagnosis of congenital
bile-duct cysts? Nat Clin Pract Gastroenterol Hepatol. 2006 Feb;
3(2):7677.
15. Park DH, Kim MH, Lee SK, et al. Can MRCP replace the diagnostic
role of ERCP for patients with choledochal cysts? Gastrointest
Endosc 2005 Sep; 62(3):360366.
16. Geenen JE. Choledochocele: endoscopic diagnosis and treatment.
In: GNJ Tytgat and K Huibregtse (eds), Bile and bile duct
abnormalities, Thieme, New York (1989):7278.
17. Martin RF, Biber BP, Bosco JJ, et al. Symptomatic choledochoceles
in adults: endoscopic retrograde cholangiopancreatography
recognition and management. Arch Surg 1992; 127:536539.
18. Ladas SD, Katsogridakis I, Tassios P, et al. Choledochocele, an
overlooked diagnosis: report of 15 cases and review of 56
392
The incidence of malignancy in choledochal cysts is reported at
between 10% and 30%.34,35 Anomalous pancreaticobiliary ductal
junction is strongly associated with gallbladder cancer among
Chinese patients.36
published reports from 1984 to 1992. Endoscopy 1995;
27:233239.
19. Chatila R, Anderson DK, Topazean M. Endoscopic resection of a
choledochocele. Gastrointest Endosc 1999; 50:578.
20. Scudamore CH, Himming AW, Teare JP, et al. Surgical
management of choledochal cysts. Am J Surg 1994; 167:497500.
21. Sugivama M, Atomi Y. Anomalous pancreaticobiliary junction
without choledochal cyst. Br J Surg 1998; 85:911916.
22. Ishibashi T, Kasakara K, Yasuda Y, et al. Malignant change in the
biliary tract after excision of choledochal cyst. Br J Surg 1997;
84:16871691.
23. Watanabe Y, Toki A, Todani T. Bile duct cancer developed after
cyst excision for choledochal cyst. J Hepatobil Pancreat Surg
1999; 6:207212.
24. Gold DM, Stark B, Pettei MJ, et al. Successful use of an internal
biliary stent in Carolis disease. Gastrointest Endosc 1995;
42:589592.
25. Nagasue N. Successful treatment of Carolis disease by
hepatic resection: report of six patients. Ann Surg 1984;
200:718734.
26. Ciambotti GF, Ravi J, Abrol RP, et al. Right-sided monolobar
Carolis disease with intrahepatic stones: Nonsurgical
management with ERCP. Gastrointest Endosc 1994; 40:761764.
27. Kobayashi S, Asano T, Yamasaki M, et al., Risk of bile duct
carcinogenesis after excision of extrahepatic bile ducts in
pancreaticobiliary maljunction. Surgery 1999; 126:939944.
28. Kato O, Hattori K, Suzuta T, et al. Clinical signicance of
anomalous pancreaticobiliary union. Gastrointest Endosc 1983;
29:9498.
29. Yamataka A, Ohshiro K, Okada K, et al. Complication after cyst
excision with hepatico enterostomy for choledochal cysts and
their surgical management in children versus adults. J Pediatr Surg
1997; 32:10971102.
30. Guelrud M, Morera C, Rodriguez M, et al. Sphincter of Oddi
dysfunction in children with recurrent pancreatitis and anomalous
pancreatiocobiliary union: an etiological concept. Gastrointest
Endosc 1999; 50:194199.
31. Benhamou JP, Congenital hepatic brosis and Carolis syndrome.
In: ER Schiff, L Schiff (eds), Diseases of the liver (7th edn),
Lippincott, Philadelphia (1993): 12041209.
32. Matsumoto Y, Fujii H, Yoshioka M, et al. Biliary strictures as a
cause of primary bile duct stones. World J Surg 1986; 10:867875.
33. Ando H, Ito T, Kaneko K, et al. Congenital stenosis of the
intrahepatic bile duct associated with choledochal cysts. J Am Coll
Surg 1995; 181:426430.
34. Stain SC, Guthrie CR, Yellin AE, et al. Choledochal cyst in the adult.
Ann Surg. 1995; 222:128133.
35. Ohtsuka T, Inoue K, Ohuchida J, et al. Carcinoma arising in
choledochocele. Endoscopy 2001; 33:614619.
36. Hu B, Gong B, Zhou DY. Association of anomalous
pancreaticobiliary ductal junction with gallbladder carcinoma in
Chinese patients: an ERCP study. Gastrointest Endosc 2003;
57:541545.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Parasitic Disease: Endoscopic Diagnosis
37 and Management of Tropical
Parasitic Infestations
Nageshwar Reddy, G. Venkat Rao, Wai Choung Ong and Banerjee Rupa
INTRODUCTION
Parasitic infestation of the biliary tract is a common cause of hepa-
tobiliary disease in developing countries and in rural areas of devel-
oped countries. In developing countries biliary parasitoses often
mimic biliary stone disease. With the advent of international travel
and immigration, clinicians in developed countries will encounter
these conditions with increasing frequency. Ascariasis, hydatid liver
disease, clonorchiasis, opisthorchiasis and fascioliasis are the com-
monly encountered parasitic infestations of the biliary tract. They
may present with cholestasis, obstructive jaundice, biliary colic,
acute cholangitis and less commonly pancreatitis. An abdominal
ultrasound facilitates the diagnosis in most cases. Although medical
therapy remains the mainstay of treatment, endoscopic retrograde
cholangiopancreatography (ERCP), and endoscopic sphincterotomy
with bile duct clearance is essential when biliary complications
occur.1 In contrast to ascariasis and hydatid disease in which the
radiological assessment may suggest the diagnosis clonorchiasis,
opisthorchiasis and fascioliasis require astute clinical suspicion in
non-endemic areas.
ASCARIS LUMBRICOIDES
The roundworm, Ascaris lumbricoides, is the most common hel-
minthic infestation the world over, infecting an estimated one billion
people. Cases have been reported from non-endemic areas in both
developing and developed countries.25 Often the infestation is
asymptomatic. When present, symptoms can vary from a non-
specic abdominal pain to intestinal obstruction.
Symptomatic patients present with a variety of hepatobiliary and
pancreatic complications due to migration of worms into the
common bile duct. Biliary-pancreatic ascariasis is commonly
reported from high endemic regions such as the Kashmir valley in
India. In a study of 500 patients with hepatobiliary and pancreatic
ascariasis, Khuroo et al. reported biliary colic in 56%, acute cholan-
gitis in 24%, acute cholecystitis in 13%, acute pancreatitis in 6% and
hepatic abscess in less than 1%.6 Biliary-pancreatic ascariasis should
be suspected in patients from an endemic area presenting with
biliary symptoms.6 In this setting, identication of eggs, larva or the
adult worm from bile or feces is strongly suggestive of the disease.
Diagnosis is conrmed by ultrasound, abdominal CT, or ERCP.
Abdominal ultrasonographic features highly suggestive of biliary
ascariasis include the presence of long, linear, parallel echogenic
structures without acoustic shadowing and the four lines sign of
non-shadowing echogenic strips with a central anechoic tube repre-
senting the digestive tract of the parasite.7
Endotherapy
During endosocpy, worms can be seen in the duodenum and are
often seen protruding from the ampulla of Vater. During ERCP,
radiographic features of the Ascaris worm include the presence of
long, smooth, linear lling defects with tapering ends (Fig. 37.1);
smooth, parallel lling defects; curves, and loops crossing the hepatic
ducts transversely; and dilatation of bile ducts (usually the common
bile duct).
Endoscopy is the mainstay of treatment for biliary ascariasis.1,811
Worm extraction is easy when it protrudes out of the ampulla of
Vater (Fig. 37.2). Held with a grasping forceps, the worm can be
brought out by withdrawing the endoscope out of the patient. A
Dormia basket can also be used: the outer end of the worm should
be maneuvered into the strings of the basket and gently held before
extraction.9 It is best to avoid using a polypectomy snare as it tends
to cut the worm. As remnant worms lead to stone formation, worms
in the biliary tree should be extracted completely.1
Worms within the common bile duct occasionally protrude out
of the papilla after contrast injection. Alternatively, they can be
extracted using a Dormia basket or a biliary occlusion balloon.1
Endoscopic sphincterotomy should be avoided in endemic areas in
view of the high reinfestation rates and easy entry of worms into
post-sphincterotomy bile ducts. Ascariasis may co-exist with biliary
calculi or strictures. In these situations, endoscopic balloon dilata-
tion of the biliary sphincter (sphincteroplasty) is an alternative to
sphincterotomy to retrieve the parasite and associated calculi.12
In endemic areas, pregnant women are prone to develop biliary
ascariasis. Endoscopic intervention in such cases requires special
precautions including lead shielding of the foetus and limitation of
total uoroscopic exposure. Failure of endoscopic extraction may
require surgical extraction which has increased risks of fetal wastage
and premature labor.13
Extraction of the culprit worm is usually associated with rapid
symptom relief. Following endoscopic therapy, all patients should
receive antihelminthic therapy to eradicate remaining worms.
A single dose of albendazole (400 mg) is highly effective against
393
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 37.1 Linear lling defect within the opacied common bile
duct in a case of biliary ascariasis. The worm was eventually
removed with a Dormia basket following endoscopic biliary
sphincterotomy.
ascariasis.14 In endemic areas periodic deworming may have a
signicant role in preventing recurrences.
ECHINOCOCCUS GRANULOSUS
The domestic strain of Echinococcus granulosus is the main cause
of human hydatid disease. Infections are found worldwide and
remain endemic in sheep raising areas. The life cycle involves two
hosts; the adult tapeworm is usually found in dogs (denitive host)
whereas sheep (intermediate host) are the usual host for the larval
stages. Human exposure is via the oral-fecal route with food or water
contaminated by the feces of the infected denitive host, usually
dogs.15 Embryonated eggs hatch in the small intestines and liberate
oncospheres that migrate to distant sites. The right lobe of the liver
is the most common site for hydatid cyst formation. The majority of
patients remain asymptomatic. In symptomatic patients, abdominal
ultrasound and serologic studies usually establish the diagnosis.
In approximately one-fourth of cases, hydatid cysts rupture into
the biliary tree causing obstructive jaundice.16 Contents of the cyst
(the scolices and daughter cysts) draining into the biliary ducts cause
intermittent or complete obstruction of bile duct, resulting in
obstructive jaundice, cholangitis, and sometimes cholangiolytic
abscesses. Rarely, acute pancreatitis complicates intrabiliary rupture
of hydatid cyst.17
Cysto-biliary communication is common, occurring in 1042%
of patients. Cysto-biliary communications are often recognized
at surgery when cysts are stained with bile.18,19 Unrecognized
cysto-biliary communications may present in the postoperative
period as a persistent biliary stula resulting in prolonged hospital-
ization and increased morbidity.
Hydatid cyst involving the pancreatic head has been rarely
reported.20 These cysts enlarge manifesting as acute pancreatitis,
394
Fig. 37.2 Ascaris protruding from the ampulla of Vater. Held with a
grasping forceps, the worm can be brought out by withdrawing
the endoscope out of the patient.
chronic pancreatitis or obstructive jaundice and are easily confused
with pancreatic pseudocyst, tumor or other congenital pancreatic
cyst. Surgical intervention is generally required for management.
Endotherapy
Treatment of hydatid disease involves antihelminthic therapy
(Albendazole) combined with surgical resection of the cyst.
Endoscopic intervention plays an important role (1) when intra-
biliary rupture of the hydatid cyst occurs and (2) in the management
of biliary complications following surgery.2135
1. Intrabiliary rupture of a hydatid cyst
Intrabiliary rupture is a common and serious complication of hepatic
hydatid cyst. The incidence varies from 1% to 25%.36 ERCP is indi-
cated when intrabiliary rupture is suspected clinically (jaundice),
biochemically (cholestasis) or sonographically (a dilated biliary
ductal system in association with hydatid cysts in the liver).23,25 Duo-
denoscopy sometimes shows whitish, glistening membranes lying
in the duodenum, or protruding from the papilla of Vater. On
cholangiography, the hydatid cyst remnants appear as: (a) liform,
linear wavy material in the common bile duct representing the lami-
nated hydatid membranes; (b) round or oval lucent lling defects
representing daughter cysts oating in the common bile duct; or (c)
brown, thick, amorphous debris.25,35 Cholangiography often reveals
minor communications, particularly with peripheral ducts. The sig-
nicance of these communications remains unclear.
In patients presenting with obstructive jaundice or cholangitis,
endoscopic biliary sphincterotomy facilitates extraction of the cysts
and membranes with use of a Dormia basket (Fig. 37.3) or a biliary
occlusion balloon.27,28 Saline irrigation of the bile duct may be neces-
sary to ush out the hydatid sand and small daughter cysts. Life-
threatening episodes of acute cholangitis can be managed by initial
nasobiliary drainage as a temporizing method, followed by extrac-
tion of hydatid cysts and membranes with or without sphincterot-
omy. The nasobiliary drain output can be examined for hydatid
hooklets or membranes. Endoscopic management of acute biliary
 Chapter 37 Parasitic Disease: Endoscopic Diagnosis and Management of Tropical Parasitic Infestations
Fig. 37.3 Hydatid membranes protruding from the ampulla of
Vater in a case of hydatid disease of the liver with intrabiliary
rupture. Endoscopic biliary sphincterotomy facilitates extraction of
the cysts and membranes using a Dormia basket.
complications allows for denitive surgery to be performed elec-
tively. Rarely, rupture with complete drainage may be treated endo-
scopically alone.37
In the presence of a hydatid cyst communicating with the biliary
ductal system, a hydrophilic guidewire can usually be negotiated into
the cyst followed by a placement of a nasobiliary catheter to facilitate
cyst evacuation. Irrigating the cyst with hypertonic saline solution
through the nasobiliary catheter ensures sterilization of the germi-
nal layers plus the remaining daughter cysts.22 In the presence of
extensive disease with multiple cysto-biliary communications, this
method should not be used because of the potential to elicit biliary
strictures from seepage of the hypertonic saline solution into the
biliary tree.38,39
2. Biliary complications following surgery
Biliary complications occur in up to 1416% of patients following
surgery for Echinococcus complications.18,40 Early postoperative com-
plications include persistent biliary stula and obstructive jaundice.
Sclerosing cholangitis and sphincter of Oddi stenosis are late post-
operative complications.
Persistent biliary stula is the most common postoperative com-
plication and occurs in 5063% of patients following surgery.19,40
Unrecognized cysto-biliary communications manifest as persistent
biliary drainage through a T-tube or development of an external
biliary stula in the postoperative period. Low output stulas (less
than 300 mL/day) usually close spontaneously after a mean duration
of four weeks. Patients with high output stulae require endoscopic
intervention.18 Endoscopic biliary sphincterotomy and ductal clear-
ance, followed by internal biliary stenting for approximately 48
weeks, is usually sufcient to achieve stula closure. Sphincterot-
omy alone may also be effective.24 Occasionally, stulous communi-
cation may develop between the hepatic hydatid cyst and the bronchi,
leading to the development of a broncho-biliary stula either de novo
or following surgery. Endoscopic biliary sphincterotomy and naso-
biliary drainage or stenting is effective in closing these stulae
non-operatively.33
Obstructive jaundice occurs in up to 2% of patients following
surgical resection of a hydatid cyst. This typically presents within
two to four weeks of surgery.2327 Obstructive jaundice results from
common bile duct obstruction by echinococcal remnants in the pres-
ence of cysto-biliary communications. As such, endoscopic biliary
sphincterotomy and ductal clearance followed by internal stenting
may be required for approximately four to eight weeks to achieve
stula closure.
Sclerosing cholangitis and Sphincter of Oddi stenosis are seen in
patients in whom formalin is used to sterilize the cysts during
surgery. Seepage of formalin into bile ducts through minor com-
munications may result in inammatory changes and stricture for-
mation in the long term. Almost all scolicidal agents are associated
clinically or experimentally with this complication. Amongst the
various scolicidal agents currently available, the use of hypertonic
saline (20%) may be preferable.41,42 These complications can be
treated endoscopically by sphincterotomy and stenting with or
without dilatation of the stricture using biliary balloons.
CLONORCHIS SINENSIS
Clonorchis sinensis, also known as the Chinese liver uke, is a
trematode commonly found in South East and Far East Asian coun-
tries, mainly China, Japan, Korea, Taiwan, and Vietnam. It is esti-
mated that about 35 million people are infected globally, of whom
approximately 15 million are in China.43 It is harbored in the biliary
tract of man and other sh-eating animals. Liver ukes have a life-
span of 1030 yrs; this creates a problem for Asian immigrants who
develop clinical symptoms many years after leaving the endemic
area.44 Opisthorchis felineus and Opisthorchis viverrini also cause
similar clinical manifestations.
Clonorchiasis is acquired from eating infested raw fresh water
sh (carp and salmon group). The infective metacercariae adhere to
the common bile duct and migrate along the epithelial lining of the
duct into the intrahepatic ducts, where they mature into at, elon-
gated, 1023 mm long adult worms. The smaller branches of the left
lobe of liver are commonly involved where the adult form attains
maturity at about one month and begins to lay eggs. The migration
of the immature ukes causes trauma, ulceration and desquamation
of the bile duct epithelium. Adenomatous hyperplasia and goblet cell
metaplasia develop as a result of epithelial injury and may lead to
encapsulating brosis of the bile duct. While a single exposure to
the parasite is of little signicance, repeated exposures provoke
diffuse involvement of the biliary tree, including the large bile ducts
and gallbladder. The average infection leads to harboring of about
20200 adult ukes, which which can increase to 20 000 ukes
during a heavy infection. Dilated sub-capsular bile ducts, adenoma-
tous ductal hyperplasia with or without peri-ductal brosis, and
eosinophilic inltration are seen in early infections. Cirrhosis may
develop in patients with repeated infections in later phases. The
endemic areas of Clonorchiasis and Opisthorchiasis coincide with
the geographical distribution of liver tumors in Southeast Asia, par-
ticularly cholangiocarcinoma.45
The clinical presentation of biliary clonorchiasis is protean. Most
patients with low parasite loads remain asymptomatic. Patients with
large parasite loads present with cholangitis, cholangiohepatitis or
intrahepatic calculi. The liver uke causes mechanical obstruction
of bile ow; subsequent bile stasis predisposes to cholangitis that
results in the death of the uke. Paroxysms of colicky upper abdomi-
nal pain due to cholangitis may be confused with gallstone disease.
395
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Biliary calculi may coexist as the eggs can act as a nidus for stone
formation. Chronic infection is associated with the development of
cholangiocarcinoma.
Clonorchiasis should be suspected in any patient who has lived
in or has traveled to an endemic region, consumed raw fresh water
sh and subsequently developed clinical signs consistent with a
biliary or hepatic disease.
Endotherapy
In patients presenting with acute cholangitis, emergency biliary
decompression following sphincterotomy is the treatment of choice.46
Aspirated bile may show adult worms and ova. On cholangiography,
characteristic features include mulberry-like appearance due to mul-
tiple saccular or cystic dilatations of the intrahepatic bile ducts; the
arrow head sign due to rapid tapering of the intrahepatic bile ducts
towards the periphery; and decrease in the number of intrahepatic
radicles due to portal and periportal brosis. Ductal irregularities are
due to adenomatous hyperplasia, which vary from small indenta-
tions to hemispherical lling defects. A scalloped appearance is
seen, sometimes visualized as lamentous, wavy, and elliptical
shaped lling defects.
Endoscopic biopsy or brush cytology is indicated whenever chol-
angiocarcinoma is suspected. Surgical intervention is indicated
in patients with hepatolithiasis complicated by multiple biliary
strictures.
All patients with biliary clonorchiasis should receive praziquantel
(75 mg/kg per day in three divided doses for 2 days) to eradicate the
infection. Biliary ductal abnormalities usually persist even after suc-
cessful drug therapy.47
FASCIOLA HEPATICA
Fascioliasis is caused by Fasciola hepatica, the sheep liver uke. The
most important denitive hosts are sheep and fasciola infestation
remains an important veterinary disease. A wide variety of mam-
malian ruminants, particularly goats, cattle, horses, camels, hogs,
rabbits, and deer, are commonly infected. Intermediate hosts include
numerous species of snail, both amphibious and aquatic. Because
of the wide range of denitive and intermediate hosts, the disease
is geographically widespread and occurs worldwide. Physicians
should therefore be aware of the possibility of infection in all geo-
graphical areas. Peru and Bolivia (La Paz, Lake Titicaca) are areas of
highest endemicity.48
Fascioliasis occurs where watercress (water plants) is eaten; its
epidemiology is related to the distribution of the intermediate snail
host populations in freshwater areas. Human infection occurs fol-
lowing ingestion of watercress that is infested with metacercariae,
the infective form of the uke. These larvae pass through the duo-
denal wall into the abdominal cavity and migrate toward the liver.
The disease occurs in two stages. The acute or hepatic phase of
the illness occurs when the organism penetrates the liver capsule
and migrates through the liver parenchyma toward the biliary
system. Patients in the acute phase usually present with dyspepsia,
followed by an acute onset of fever and abdominal pain, especially
in the right hypochondrium or in the right upper quadrant. Urticaria
and eosinophilia may be present. These symptoms result from the
destruction and inammatory response caused by the migrating
larvae. In approximately 50% of such cases, the infection remains
subclinical. The acute phase usually lasts for 3 months following
ingestion of the metacercariae.
396
The second chronic or biliary phase occurs when the parasite
enters the bile canaliculi 3 to 4 months after the contaminated meal
is ingested. Typical manifestations are jaundice, fever, right upper
quadrant pain, and rarely acalculus cholecystitis, severe hemobilia,
and acute pancreatitis.49 During the chronic stage, motile ukes may
be visualized in the gallbladder.50 Liver function tests reect a cho-
lestatic picture. Serological tests (FAST-ELISA/Falcon assay screen-
ing test or dot blot ELISA) are highly sensitive (95100%) and specic
(97%), thus aiding in diagnosis.51
Inammation due to toxic metabolites and mechanical effects of
the larvae in the bile ducts leads to epithelial necrosis and adenoma-
tous changes, eventually leading to biliary brosis. These changes
further evolve into cystic dilatation, total or partial obstruction of bile
ducts, and periportal brosis and cirrhosis. Although the brotic
changes are likely to persist despite successful therapy, some of the
ductal changes are reversible.49 The adult form has a life span of
approximately 913 years. Eggs or the dead parasites can form a
nidus for calculus formation, potentially leading to intra or extra
hepatic biliary calculi.
Endotherapy
Oral drug therapy is the standard treatment for hepatic fascioliasis.
Triclabendazole (10 mg/kg as a single dose) is the drug of choice.
In severe or persistent infections, two doses of 10 mg/kg adminis-
tered 1224 hours apart are recommended.51 An alternative includes
Bithionol (3050 mg/kg on alternate days for 1015 doses). Chloro-
quine, mebendazole, albendazole, and praziquantel have also been
used with variable success. Patients should be advised regarding
expected biliay colic caused by expulsion of parasites or parasite
fragments which usually occurs 27 days after commencing drug
therapy. Endoscopic therapy is required (1) when biliary com-
plications occur or medical therapy fails and (2) in management of
severe infection with multiple worms. During ERCP, Fasciola appear
as small, radiolucent linear or crescent-like shadows, with jagged,
irregular margins in the gallbladder or in dilated bile ducts. The
worms can be extracted by using a balloon catheter or Dormia basket
following biliary sphincterotomy (Fig. 37.4). Patients usually harbor
a single Fasciola worm in the bile duct with an occasional one in the
Fig. 37.4 Fasciola hepatica extracted using a balloon catheter fol-
lowing biliary sphincterotomy.
 Chapter 37 Parasitic Disease: Endoscopic Diagnosis and Management of Tropical Parasitic Infestations
gallbladder. When worms are present in the gallbladder or in the
intrahepatic biliary radicles where they are not amenable to mechan-
ical extraction, irrigating the biliary system with 20 ml of 2.5% povi-
done iodine solution (5 ml of 10% povidone iodine plus 15 ml of
contrast material) during ERCP is helpful.52 Aspirated bile may be
examined for parasite eggs.
The management of massive forms of biliary fascioliasis where
dozens or hundreds of mature parasites reside in the intrahepatic
and extrahepatic ducts has been successfully described.53 Initial
extraction of parasites with a basket or balloon catheter is performed.
This is followed by a 10 minute instillation of 20 ml of 2.5% povi-
done iodine solution (5 ml of 10% povidone iodine plus 15 ml of
contrast material) with balloon occlusion of the common hepatic
duct. The ducts are then washed with saline solution and the dead
parasites are removed instrumentally. Repeat treatment sessions
REFERENCES
1. Reddy DN, Sriram PVJ, Rao GV. Endoscopic diagnosis and
management of tropical parasitic infestations. Gastrointest Endosc
Clin N. Am. 2003; 13:765773,xxi.
2. Reddy DN. Endoscopic diagnosis and management of biliary
parasitosis. http://patients.uptodate.com.
3. Botembe N, Cabrera-Alvarez G, LeMoine O, et al. A rare cause of
biliary pain in Belgium. Acta Gastroenterol Belg 1999; 62:443445.
4. Misra SP, Dwivedi M. Clinical features and management of
biliary ascariasis in a non-endemic area. Postgrad Med J 2000;
76:2932.
5. Valentine CC, Hoffner RJ, Henderson SO. Three common
presentations of ascariasis infection in an urban emergency
department. J Emerg Med 2001; 20:135139.
6. Khuroo MS, Zargar SA, Mahajan R, Hepatobiliay and pancreatic
ascariasis in India. Lancet 1990; 335:15031506.
7, Khuroo MS, Zargar SA, Mahajan R, et al. Sonogaphic appearances
in biliary Ascariasis. Gastoenterology 1987; 93:276272.
8. Al Karawi M, Sanai FM, Yasawy MI, et al. Biliary strictures and
cholangitis secondary to ascariasis: Endoscopic management.
Gastrointest Endosc 1999; 50:695697.
9. Beckingham IJ, Cullis SN, Krige JE, et al. Management of
hepatobiliary and pancreatic ascarisis infestation in adults after
failed medical treatment Br J Surg 1998; 85:907910.
10. Misra SP, Dwivedi M. Endoscopyassisted emergency treatment
of gastroduodenal and pancreatobiliary ascariasis. Endoscopy
1996; 28:629632.
11. Sandouk F, Haffar S, Zada MM, et al. Pancreatic-biliary
ascariasis: experience of 300 cases. Am J Gastroenterol 1997;
92:22642267.
12. Misra SP, Dwivedi M. Removal of Ascaris lumbricoides from the
bile duct using balloon sphincteroplasty. Endoscopy 1998; 30:
S6S7.
13. Shah OJ, Robbani I, Zargar SA, et al. Management of biliary
ascariasis in pregnancy. World J Surg 2005; 29:12941298.
14. Albonico M, Smith PG, Hall A, et al. A randomized controlled trial
comparing mebendazole and albendazole against Ascaris,
Trichuris and hookworm infections. Trans R Soc Trop Med Hyg
1994; 88:585589.
15. Reather W, Hanel H. Epidemiology, clinical manifestations and
diagnosis of zoonotic cestode infections: an update. Parasitol Res
2003; 91:412438.
16. Dadoukis J, Gamvros O, Aletras H. Intrabiliary rupture of the
hydatid cyst of the liver. World J Surg 1984; 8:786790.
may be required for complete parasite clearance. In cases with chol-
angitis and liver abscesses, nasobiliary drainage with iodine treat-
ment repeated three times under direct uoroscopic control may be
used.
SUMMARY
Biliary parasitosis will remain a problem infrequently
encountered by the practicing endoscopist in non-endemic
areas. Ascariasis and hydatid disease are clinically and
radiologically evident: fasciola, clonorchis and opisthorchis
infections require astute clinical suspicion and awareness for
early diagnosis and appropriate management.
17. A1 Toma AA, Vermeijden RJ, Van De WA. Acute pancreatitis
complicating intrabiliary rupture of liver hydatid cyst. Eur J Intern
Med 2004; 15:6567.
18. Agarwal S, Sikora SS, Kumar A, et al. Bile leaks following surgery
for hepatic hydatid disease. Indian J Gastreenterol 2005;
24:5558.
19. Simsek H, OZaslan E, Sayek I, et al. Diagnostic and therapeutic
ERCP in hepatic hypatic hydatid disease. Gastrointest Endosc
2003; 58:384389.
20. Lemmer ER, Krige JE, Price SK, et al. Hydatid cyst in the head of
the pancreas with obstructive jaundice. J Clin Gastroenterol 1995;
20:136138.
21. A1 Karawi MA, Mohamed AR, Yasawy I, et al. Non-surgical
endoscopic trans-papillary treatment of ruptured echinococcus
liver cyst obstructing the biliary tree. Endoscopy 1987; 19:8183.
22. A1 Karawi MA, Yasawy MI, Shiek Mohamed AR. Endoscopic
management of biliary hydatid disease: report on six cases.
Endoscopy 1991; 23:278281.
23. Magistrelli P, Masetti R, Coppola R, et al. Value of ERCP in the
diagnosis and management of pre- and postoperative biliary
complications in hydatid desease of the liver. Gastrointest Radiol
1989; 14:315320.
24. Iscan M, Duren M. Endoscopic sphincterotomy in the
management of postoperative complications of hepatic hydatid
disease. Endoscopy 1991; 23:282283.
25. Aargar SA, Khuroo MS, Khan BA, et al. Intrabiliary rupture of
hepatic hydatid cyst: sonographic and cholangiographic
appearances. Gastrointest Radiol 1992; 17:4145.
26. Spiliadis C, Georgopoulos S, Dailianas A, et al. The use of ERCP in
the study of patients with hepatic echinococcosis before and after
surgical intervention. Gastrointest Endosc 1996; 43:575579.
27. Tekant Y, Bile O, Acarli K, et al. Endoscopic sphincterotomy in the
treatment of postoperative biliary stulas of hepatic hydatid
disease. Surg Endosc 1996; 10:909911.
28. Rodriguez AN, Sanchez del Rio AL, Alguacil LV, et al. Effectiveness
of endoscopic sphincterotomy in complicated hepatic hydatid
disease. Gastrointest Endosc 1998; 48:593597.
29. Yilmaz U, Sakin B, Boyaxioglu S, et al. Management of
postoperative biliary strictures secondary to hepatic hydatid
disease by endoscopic stenting. Hepatogastroenterology 1998;
45:6569.
30. De A, X, Perez OL. The use of endoprostheses in biliary stula of
hydatid cyst. Gastrointest Endosc 1999; 49:797799.
397
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
31. Dumas R, Le Gall P, Hastier P, et al. The role of endoscopic
retrograde cholangiopancreatography in the management of
hepatic hydatid disease. Endoscopy 1999; 31:242247.
32. Giouleme O, Nikolaidis N, Zezos P, et al. Treatment of
complications of hepatic hydatid disease by ERCP. Gastrointest
Endosc 2001; 54:508510.
33. Partrinou V, Dougenis D, Kritikos N, et al. Treatment of
postoperative bronchobiliary stula by nasobiliary drainage. Surg
Endosc 2001; 15:758.
34. Saritas U, Parlak E, Akoglu M, et al. Effectiveness of endoscopic
treatment modalities in complicated hepatic hydatid disease after
surgical intervention. Endoscopy 2001; 33:858863.
35. Busic Z, Amic E, Servis D, et al. Common bile duct obstruction
caused by the hydatid daughter cysts. Coll Antropol 2004;
28:325329.
36. Erzurumlu K, Dervisoglu A, Polat C, et al. Intrabiliary rupture: an
algorithm in the treatment of controversial complication if hepatic
hydatidosis. World J Gastroenterol 2005; 11:24722476.
37. Hilmioglu F, Karincaoglu M, Yilmaz S, et al. Complete treatment of
ruptured hepatic cyst into biliary tree by ERCP. Dig Dis Sci 2001;
46:463467.
38. Belghiti J, Benhamou JP, Houry S, et al. Caustic sclerosing
cholangitis. A complication of the surgical treatment of hydatid
disease of the liver. Arch Surg 1986; 121:11621165.
39. Belghiti J, Perniceni T, Kabbej M, et al. Complication of
preoperative sterilization of hydatid cysts of the liver. Apropos of
6 cases. Chirurgie 1991; 117:343346.
40. Bilse Y, Bulut T, Yamaner S, et al. ERCP in the diagnosis and
management of complications after surgery for hepatic
echinococcosis. Gastrointest Endosc 2003; 57:210213.
41. Sahin M, Eryilmaz R, Bulbuloglu E. The effect of scolicidal agents
on liver and biliary tree (experimental study). J Invest Surg 2004;
17:323326.
398
42. Tozar E, Topeu O, Karayalcin K, et al. The effects of cetrimide-
chlorhexidine combination on the hepato-pancreatico-biliary
system. World J Surg 2005; 29:754758.
43. Lun ZR, Gasser RB, Lai DH, et al. Clonorchiasis: a key foodborne
zoonosis in China. Lancet Infect Dis 2005; 5:3141.
44. Stauffer WM, Sellman JS, Walker PR. Biliary liver ukes
(Opisthorchiasis and Clonorchiasis) in immigrants in the United
States: often subtle and diagnosed years after arrival. J Travel Med
2004; 11:157159.
45. Srivatanakul P, Sriplung H, Deerasamee S. Epidemiology of
liver cancer: an overview. Asian Pac J Cancer Prev 2004;
5:118125.
46. Navab F, Diner WC, Westbrook KC, et al. Endoscopic biliary
lavage in a case of Clonorchis sinensis. Gastrointest Endosc 1984;
30:292294.
47. Leung JW, Sung JY, Banez VP, et al. Endoscopic
cholangiopancreatography in hepatic clonorchiasis: a follow-up
study. Gastrointest Endosc 1990; 36:360363.
48. Mas-Coma MS, Esteban JG, Bargues MD. Epidemiology of human
fascioliasis: a review and proposed new classication. Bull World
Health Organ 1999; 77:340346.
49. Dias LM, Silva R, Viana HL, et al. Biliary fascioliasis: diagnosis,
treatment and follow-up by ERCP. Gastrointest Endosc 1996;
43:616620.
50. Bassily S, Iskander M, Youssef FG, et al. Sonography in diagnosis
of fascioliasis. Lancet 1989; 1:12701271.
51. Saba R, Korkmaz M. Human fascioliasis. Clinical Microbiology
Newsletter 2005; 27:2734.
52. Dowidar B, EI Sayad M, Osman M, et al. Endoscopic therapy of
fascioliasis resistant to oral therapy. Gastrointest Endosc 1999;
50:345351.
53. Roig GV. Hepatic fascioliasis in the Americas: a new challenge for
therapeutic endoscopy. Gastrointest Endosc 2002; 56:315317.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Recurrent Pyogenic Cholangitis
38 Khean-Lee Goh and Dong Wan Seo
INTRODUCTION AND SCIENTIFIC BASIS
Recurrent pyogenic cholangitis (RPC) is a condition that is charac-
terized by repeated attacks of bacterial infection of the biliary tract.
It is believed that the initiating event is the entry of enteric ora into
the biliary tree causing infection and inammation and through
bacterial deconjugation of bilirubin diglucoronide, the formation of
primary biliary stones.1 Persistent inammation results in biliary
strictures and stasis of bile in the biliary tree which encourages
further formation of stones leading to a vicious cycle of repeated or
persistent inammation and infection. There have been reports
linking helminthiasis to RPCAscaris lumbricoides and Clonorchis
sinensis worms have been identied in the biliary tract of patients
with RPC.2 RPC has been most commonly reported in countries in
the Asian Pacic region including China, Taiwan, Japan, Korea and
the South East Asian countries and is distinctly uncommon in the
western world.3
BOX 38.1 KEY POINTS
Recurrent pyogenic cholangitis (RPC) is characterized by
repeated attacks of cholangitis and presence of intrahepatic
strictures and stones.
Modalities of treatment include: endoscopic retrograde
cholangiography techniques, percutaneous transhepatic
cholangioscopy (PTCS) and surgery.
PTCS requires skill and patience and involves the passage of
a cholangioscope through a percutaneous transhepatic tract.
Repeated procedures are usually needed.
Dilation of strictures with catheters and balloons and
fragmentation of stones with electrohydraulic lithotripsy and
laser may be necessary.
The hallmark of RPC is the presence of stones and strictures,
which can be located in both intrahepatic and extrahepatic ducts (Fig
38.1). The treatment of RPC is difcult and requires a multimodality
approach encompassing endoscopy, radiological techniques and
surgery. Successful treatment of RPC depends on the success in
clearing stones, dilating strictures and maintaining the patency of
the stenosed ducts. Specic management aims at accurate localiza-
tion of the pathology, application of specic techniques to removing
stones and dilating strictures. It serves to eliminate bile stasis and
achieve control of cholangitis. Amongst the endoscopic techniques
used are standard endoscopic retrograde cholangiopancreato-graphy
(ERCP) with or without cholangioscopy, percutaneous transhepatic
cholangioscopy (PTCS) and postoperative cholangioscopy (POCS)
though a T-tube tract.
Surgery is an important treatment modality in RPC and will be
discussed briey at the end of the chapter to place it in the overall
scheme of management.
INITIAL MANAGEMENT OF THE PATIENT
WITH CHOLANGITIS
Patients with RPC often present with acute ascending cholangitis.
Acute cholangitis may be the rst attack or a recurrent episode.
These patients may develop septic shock quite rapidly. Initial man-
agement includes intravenous uid replacement and the institution
of intravenous potent, broad-spectrum antibiotics. Emergency surgi-
cal decompression may be necessary in some patients but carries
with it a high postoperative morbidity and mortality rate.4
Non-surgical biliary drainage procedures provide an important
alternative treatment option in these patients, especially in those
with concomitant common bile duct stones. Urgent ERCP with
placement of an indwelling stent or a nasobiliary catheter has sig-
nicantly reduced the mortality rate5 (Fig, 38.2). Percutaneous trans-
hepatic biliary drainage (PTBD) may be necessary in patients with
cholangitis associated with intrahepatic stones.6
SPECIFIC TREATMENT OF INTRAHEPATIC STONES
Description of techniques
Standard ERCP and peroral cholangioscopy
Stones that are found in the extrahepatic bile duct and the main
intrahepatic ducts can often be dealt with using conventional ERCP
techniques of sphincterotomy and passing a basket and balloon into
the appropriate ducts and extracting the stones. Strictures in the CBD
or in the main intrahepatic ducts can be dilated with biliary dilating
balloons of inated diameters of 410 mm (Quantum TTC balloon,
Cook Endoscopy Inc, Winston Salem, USA, MaxForce Hurricane
Boston Scientic, Natrick, USA) to facilitate passage of a retrieval
basket for stone extraction. The use of a through the scope mechan-
ical lithotripter (Olympus Optical Company, Tokyo, Japan or MTW,
Wessel, Germany) is helpful in crushing large stones. The usefulness
of this approach is limited by the difcult access into the intrahepatic
ducts. This is often due to angulated bile ducts, tight strictures and
impacted stones making it difcult to pass and open a Dormia
retrieval basket and sometimes even to pass a guidewire.7
Lithotripsy using an electrohydraulic lithotripsy (EHL) or a laser
probe through a mother-baby scope set-up can be attempted to
break-up main intrahepatic duct stones and allow further passage
399
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 38.1 Endoscopic retrograde cholangiogram from a patient
presenting with acute ascending cholangitis showing intra- and
extrahepatic strictures with stones.
Fig. 38.2 Intra- and extrahepatic stones and strictures. A long
plastic stent has been placed via ERCP to relieve the cholangitis.
of Dormia basket and balloons. Eventual success is limited by loca-
tion and severity of strictures and duct angulation.
Percutaneous transhepatic cholangioscopy (PTCS)
This technique requires a percutaneous transhepatic tract and three
steps are required for PTCS; percutaneous transhepatic biliary drain-
age, dilation of the tract, and cholangioscopic examination.
1. Percutaneous Transhepatic Biliary Drainage (PTBD)
The technique, PTBD, has been used to relieve obstructive jaundice,
to drain infected bile, and to prevent or control cholangitis and sepsis.8
It is an initial step for creating a percutaneous tract and can be per-
formed under uoroscopic or ultrasonographic guidance.9 For PTCS,
the site of PTBD is very important. If the puncture site is misplaced,
there can be an acute angulation during the course to the target lesion
(Fig. 38.3). Acute angulation is an important factor causing PTCS
failure. Before selection of a PTBD site, the cholangioscopist or inter-
ventional radiologist should be familiar with the anatomy of the
biliary tree. The ideal PTBD site is selected after meticulous review
400
Fig. 38.3 Acute angulation caused by inappropriate selection of
PTBD site.
Fig. 38.4 PTBD kit. A complete kit includes Chiba needles for
puncture, guidewires, a dilator, and pigtail drainage catheters.
of several imaging studies such as ultrasonography, CT scan, endo-
scopic retrograde cholangiography or magnetic resonance cholangi-
ography (MRCP). A PTBD kit is composed of puncture needles, a
guidewire, dilators and pigtail drainage tubes (Fig. 38.4). The usual
diameter of an initial PTBD tube is around 68.5 Fr. For the initial
puncture, the selection of a peripheral duct is important because the
direct insertion of a PTBD tube into the central duct carries a signi-
cant risk of bleeding. Ultrasonography-guided puncture or uoros-
copy-guided puncture technique is commonly adopted for initial
peripheral duct selection. After selective puncture of a peripheral
duct, a guidewire is inserted and a bougienage dilator is pushed
over the guidewire. After dilation of the tract, a pigtail catheter is
introduced into the biliary tree (Fig. 38.5). In cases with a non-
dilated intrahepatic duct, PTBD carries a high risk of bleeding
and biliary leakage. To prevent these complications, the insertion of

an endoscopic nasobiliary drainage tube before PTBD and simultane-
ous cholangiography using this tube during PTBD is helpful in the
accurate targeting of the desired intrahepatic duct.10
2. Tract dilation
For cholangioscopic examination of the biliary tree, the diameter of
the percutaneous transhepatic tract should be larger than that of the
cholangioscope (e.g. CHF P 20, Olympus Optical Company, Tokyo,
Japan). The cholangioscope diameter varies from 3.0 mm to 5.2 mm.
Therefore the diameter of the percutaneous tract should be dilated to
at least 1112 Fr when an 11 Fr cholangioscope is to be used. In most
centers, the tract is dilated to 1618 Fr because a cholangioscope with
an outer diameter of 16 Fr is commonly used for therapeutic pur-
poses. This dilation procedure can be performed by the aid of a spe-
cialized dilation Kit (Nipro, Tokyo, Japan) (Fig. 38.6). The dilation
process can be accomplished by multi-stage dilation or by one-
stage dilation.11 Multi-stage dilation means reaching the fully dilated
diameter of percutaneous tract through several repeated dilations.
The diameter of PTBD tube is around 68.5 Fr at the rst attempt.
The tract can be dilated in stages every two to four days; to 1012 Fr,
then up to 1416 Fr, and, nally, up to more than 18 Fr. In the one-
A B
C D
Fig. 38.5 Fluoroscopic images of PTBD procedures. A A guide-
wire is inserted into the left intrahepatic duct after selection of S3
segmental duct by puncture needle. B The tract is dilated by a
dilator up to 8 Fr. C A 7.5 Fr pigtail drainage catheter is inserted
into the dilated tract. D Previously injected contrast material is
well drained through the pigtail catheter.
A B
Fig. 38.7 Percutaneous transhepatic cholangioscopic examination. A Dressing set for PTCS. B Draping was done before cholangio-
scopic examination. An 18 Fr PTCS catheter is visible and the tube is tied to the skin to prevent migration. C Checking of light source
and saline ow before insertion of the cholangioscope into the body. D Insertion of cholangioscope. After insertion of the tip of the
cholangioscope into the tract, the cholangioscopist monitors the videoscopic view and guides the cholangioscope tip to maintain the
view of the bile duct lumen.
Chapter 38 Recurrent Pyogenic Cholangitis
stage dilation protocol, however, the PTBD tract is dilated up to 16
or 18 Fr in a single session 24 days after the initial PTBD.
There are advantages and disadvantages of each method. The
main advantage of multi-stage dilation is that the dilation process
is less painful and less traumatic to the patient. Gradual dilation with
repeated procedures can reduce the risk of severe pain or the chance
of signicant bleeding after dilation. However, multi-stage dilation
is time consuming, requires several procedures until satisfactory
dilation is achieved, and is costlier. The one-stage dilation protocol
saves time and money compared to the multi-stage dilation proto-
col but can cause signicant pain and bleeding during the proce-
dure. For the one-stage dilation protocol, it is mandatory to provide
adequate analgesia in addition to antibiotics.
3. Percutaneous transhepatic cholangioscopic examination
Following full dilation of the percutaneous transhepatic tract, 1014
days are usually required for maturation of the sinus tract at which
time the cholangioscopic examination can be safely performed (Fig.
38.7). The patient is positioned in the supine position on the uo-
roscopy table. The position of the cholangioscopist can be changed
according to the site of PTBD. For example, when PTBD is per-
formed on a right intrahepatic duct, the right side of the patient is
the preferred position and when PTBD is done on a left intrahepatic
duct, the left side of the patient is the preferred position. The video
monitor and uoroscopic monitor should be located at a favorable
Fig. 38.6 Nipro set for tract dilation. A complete set for tract dila-
tion is composed of a guidewire, a tapered tip catheter, variable
sized bougienage tubes, and a PTCS catheter.
C D
401
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

angle for the cholangioscopist. Premedication is required to relieve
pain and anxiety using a combination of meperidine and midazolam
or diazepam. The insertion of a cholangioscope into a fully dilated
tract is not difcult. However, if the waiting period between the tract
dilation and cholangioscopic examination is short, the insertion of
the cholangioscope can be difcult and sometimes traumatic. The
tract may collapse after removal of the dilation tube, especially
during the rst cholangioscopic examination. A guidewire, which is
inserted before removal of the dilation tube, is used to smoothly
guide the cholangioscope into the biliary tree.
PTCS offers several advantages. The percutaneous approach
allows evaluation of both the intrahepatic ducts and common bile
duct and it is the shortest distance to the biliary tree. The handling
and angulation maneuvers of the tip of the percutaneous cholangio-
scope are easier compared to those of peroral cholangioscopy. The
application of various techniques such as biopsy, suction, and dye
spraying during cholangioscopic examination is also not difcult.
Insertion of a balloon or a catheter under cholangioscopic guidance
and application of biopsy forceps or electrohydraulic lithotripsy are
much easier than the peroral approach. The only drawback com-
pared to the peroral route is the necessity of creating a percutaneous
tract, which is an invasive process.
During cholangioscopic evaluation of the biliary tree, irrigation
is required to obtain an optimal view of the bile duct. In the biliary
tree, pus, sludge and blood can cause blurring of the view. Thick bile
may also coat the bile duct wall. To obtain a clear view of the bile
duct continuous saline irrigation is recommended and is usually
achieved by suspending a bottle of normal saline and letting the
saline ow continuously into the instrumental channel of the chol-
angioscope to wash the bile duct (Fig. 38.8).
When compared to a percutaneous transhepatic tract, the T-tube
tract is a relatively long tract and traverses a longer length of free
peritoneal space before reaching the common bile duct. When per-
cutaneous transhepatic cholangioscopy is employed, the distance of
free space between parietal peritoneum and liver capsule is usually
less than 1 cm whereas the T-tube tract has a free space between
parietal peritoneum and common bile duct which is usually longer
than 45 cm. Because of this difference, maturation of the T-tube
tract usually takes a longer period of time, normally at least 4 weeks
after T-tube insertion.
An ideal T-tube tract should run a straight course from the skin
to the insertion point in the bile duct. If this tract meets the common
bile duct at a right angle, cholangioscopic examination of the
common bile duct and the intrahepatic ducts is not difcult (Fig.
38.9). The technique of bile duct examination is basically similar to
that of percutaneous transhepatic cholangioscopy. However, there
are also some limitations of postoperative cholangioscopy. In addi-
A B

Postoperative Cholangioscopy (POCS)

Although the practice of performing ERCP and laparoscopic chole-
cystectomy is commonly carried out in patients with gallbladder and
common bile duct stones, many patients still undergo open chole-
cystectomy with common bile duct exploration and with a T-tube left
in situ in the common bile duct. POCS can be performed in these
patients following maturation of the T-tube tract, when there are
concomitant common bile duct and/or intrahepatic stones.12
Fig. 38.8 Normal saline irrigation by gravity. A A bottle of
normal saline is connected to the cholangioscope. Gravity provides
a continuous ow of normal saline into the bile duct during chol-
angioscopic examination. B The ow can be controlled by the
on and off switching of a two-way stopcock.

A B C

Fig. 38.9 Ideal course of T-tube tract. A A successful cholangioscopic examination is dependent on the straightness of the tract. The
ideal T-tube tract for POCS should be straight and the insertion angle into CBD is a right angle. B,C In the setting of a right angle, the
insertion of the cholangioscope into intrahepatic duct or distal CBD is possible with the aid of exion or extension movement of tip of
cholangioscope.

402

tion to having to wait for a longer period until full maturation of the
tract, insertion of the cholangioscope may not be easy because the
T-tube tract may contain angulations (Fig. 38.10). An acute angula-
tion makes cholangioscopic examination very difcult and perfora-
tion of T-tube tract during insertion of a cholangioscope can occur.
Techniques for cholangioscopic stone removal
1. Removal with a basket
Biliary stones can be found in a straight duct and/or in an angulated
duct. Cholangioscopic removal of stones requires an experienced
operator in order to make an otherwise laborious procedure less
time-consuming and safer. The basic steps for cholangioscopic
removal of stones are as follows: rst, the basket is inserted into the
biliary tract; second, the basket is opened just beyond the stones;
third, the stone is engaged by withdrawing the opened basket; fourth,
stones are grasped rmly by withdrawing the basket further; nally
removal of the stone by removal of the cholangioscope and basket
simultaneously (Fig. 38.11).
A B
Fig. 38.10 Various angulations of a T-tube tract. A,B During
operative insertion of T-tube into the bile duct, various angulations
can be made at the sinus tract and are important factors that limit
successful cholangioscopy. The surgeon should be cautious to
prevent these types of angulations.
A B
Fig. 38.12 Cholangiography during basket stone removal. A The cholangioscope is just in front of an intrahepatic stone (arrow) and
a basket tip is located just beyond the stone before opening. B The basket is opened beyond the stone. C By withdrawing the opened
basket, a small stone is engaged within the basket.
Chapter 38 Recurrent Pyogenic Cholangitis
For removal of stones, a Dormia-type basket is used. The ideal
position of the basket tip before opening is just beyond the stones
and they can be captured only by gentle withdrawal of the opened
basket (Fig. 38.12). If a basket is opened in front of a stone, there is
a possibility of pushing the stone into a peripheral duct. The basket
may be opened just before the stone or at the exact site of the stone;
the stones can be captured by pushing the opened basket or by the
back and forth movement of the basket. However, this pushing
maneuver or back and forth movement is not recommended because
these movements can cause deformity of the basket. Gentle with-
A B
C D
Fig. 38.11 Illustration of the four basic steps of cholangioscopic
stone removal. A A basket is inserted through the cholangio-
scope and the basket tip is positioned beyond the stone. B The
basket is opened just beyond the stone to grasp the stone easily.
C Gentle withdrawing of the opened basket is usually enough
to engage intraluminal stones. To facilitate this step, the basket
should keep its original shape when it is opened within the lumen.
D Further closure of the basket in which the stone is engaged
allows the stone to be grasped tightly; removal of stone is usually
achieved by gentle, simultaneous withdrawal of the cholangio-
scope and stone basket.
C
403
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

drawal of the basket and closure of the basket to engage the stones
during this withdrawal movement is recommended.
2. Stone fragmentation
Large stones cannot be extracted directly from the biliary tree, and
need to be broken into smaller fragments. The easiest way to frag-
ment large stones is to crush them by tightening or closing the
basket. Some of the softer brown pigment stones can be broken into
several pieces by this simple maneuver. Electrohydraulic lithotripsy
(EHL) or laser lithotripsy are needed to fragment harder stones. The
mechanism of EHL fragmentation is compression waves (shock
waves). The shock waves are generated at the tip of the probe directly
in front of the stones by an explosive spark discharge in a liquid
medium (Fig. 38.13). The different power adjustment and different
frequency modes permit adaptation to various elds of application
and sites. Higher energy with high frequency is very effective to
fragment hard stones, but it can easily damage the bile duct wall. To
avoid tissue damage, lower energy levels are preferable. For the best
results, the probe must be placed directly on the stone. The EHL
shock wave can only be generated in an aqueous media and there-
fore an electrolyte solution such as physiologic saline (0.9% NaCl)
is used to irrigate the area.
This technique can be applied not only to the fragmentation of
large and hard stones, but also to the fragmentation of smaller
stones that are impacted in the peripheral duct or in a strictured
segment.13 It may not be possible to remove impacted stones using
only a basket because there is no space to open the basket and
grasp the stones. When EHL lithotripsy is used, only a small aper-
ture is needed and the tip of probe can then come into contact with
the stone. Once the tip of the electrohydraulic lithotripsy probe is
directed at the stones, the electrohydraulic shock wave can be gen-
erated in an aqueous media and the force applied to the stone (Fig.
38.14). After fragmentation of large stones into small pieces, a
basket can be introduced past the strictured segment and small
fragmented stones can then be extracted or washed away by forceful
saline irrigation.
3. Stricture dilation
Intrahepatic stones frequently occur upstream to a strictured
segment of the bile duct. For cholangioscopic removal of these

A B C

Fig. 38.13 Stone fragmentation using EHL. A The tip of the EHL probe is placed on the surface of a pigmented stone. B During EHL
ring, an electric spark and accompanying shock wave is generated in the area of the probe tip. C Exposed inner core. Layered structure
of inner core is visible after destruction of outer shell.

A B
C

Fig. 38.14 Application of EHL to an impacted stone. A This cholangiogram shows a large stone impacted in the left intrahepatic duct.
There is no space to permit the introduction of a stone basket. B Cholangioscopic view reveals only small anterior surface of the
impacted large stone. The tip of the EHL probe is positioned at the surface of the stone. C Shockwave is generated to the impacted
stone.

404
 Chapter 38 Recurrent Pyogenic Cholangitis

stones, stricture dilation is required. A balloon dilator or a dilating
catheter can be used for this purpose.14
When balloon dilation is performed, a balloon and pressure
gauge are required. There are several types of balloons which can be
used for dilation of strictures (Cook, Bloomington, USA). The ideal
balloon should have a small shaft diameter and the balloon should
be strong enough to endure the pressure applied during dilation.
The balloon must be long enough to cover the stricture segment.
For balloon dilation, a guidewire is inserted across the strictured
segment under cholangioscopic and uoroscopic guidance. After
insertion of a guidewire, the balloon is advanced though the stric-
tured segment over a guidewire and positioned under uoroscopy.
It is important to keep the guidewire straight during the advance-
ment of the balloon catheter. The optimal location of the balloon is
usually obtained when the mid-portion of the balloon (where the
expansile force is greatest) is located at the stricture. If the balloon
is not placed centrally over the stricture as described, there is a ten-
dency for the balloon to slip in or out of the stricture during ination
of the balloon. Contrast media or distilled water mixed with radio-
opaque contrast material is used to inate the balloon and allow
visualization of the dilating balloon under uoroscopy. Radio-opaque
contrast material acts as an indicator of the degree of dilation (Fig.
38.15) by showing up waisting of the balloon at the site of the stric-
ture and obliteration of the waisting with successful dilation (Fig
38.16). The dilation pressure should be monitored to achieve and
maintain the optimum ination pressure of 610 atmospheres.
Dilation of the bile duct can also be achieved with catheters.13
There are two types of catheters according to the shape of the tips;
a tapered tip catheter (Akita Sumitomo, Bakelite Company, Akita,
Japan) and a straight tip catheter (Cook, Bloomington, USA) (Fig
38.17). The main advantage of the tapered tip catheter is that it is
easier to pass into and across the strictured segment. However, a
catheter with tapered tip has a tendency to slip away from the lesion
when the strictured segment is tight, while a straight tip catheter
does not. The main disadvantage of a straight tip catheter is that the
insertion of this catheter can be traumatic and may cause severe pain
or bleeding because of its blunt end.
For the insertion of a catheter into the strictured segment, a
guidewire is passed through the strictured portion. The catheter is
pushed though the strictured segment over the guidewire. If there
is an acute angulation during the course of passage, catheter inser-
tion frequently fails. To overcome difcult angulations, two or more

A B C

Fig. 38.15 Balloon dilation of left main duct. A A tight stricture in the left main duct with upstream duct dilation is seen. Under cholan-
gioscopic and uoroscopic guidance, a guidewire is passed into the stricture segment. B A balloon is inserted over the guidewire and
centered across the strictured segment (the two radio-opaque markers of the balloon are seen). The balloon is inated with contrast
material. C After successful dilation, the cholangioscope is able to pass the stricture segment.

Fig. 38.16 Waist formation and disappear-

A B ance during balloon dilation. A The tight
stricture causes a waist of the balloon during
pressure exertion. B When the pressure
exceeds the strength of stricture, the waist
disappears. Once this occurs the ination
pressure should not be increased.

405
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Fig. 38.17 Two types of catheters. According to the shape of the
tip, catheters are classied into two types; tapered (lower two cathe-
ters) and straight (uppermost catheter).
guidewires can be used simultaneously to guide the tip of the dilat-
ing catheter (Fig. 38.18). Balloon dilation and catheter dilation can
be used in combination. In case of tight strictures, balloon dilation
followed by catheter placement into the stricture segment allows for
an effective dilation.
To allow efcient bile drainage from other intrahepatic ducts at
the same time, the catheters should have side-holes. The side-holes
can be fashioned just before catheter insertion (Fig. 38.19). The
number, location, and size of side-holes are important to ensure
adequate drainage through the catheter and to prevent bile leakage.
Before making side-holes, the cholangioscopist should measure the
length of catheter which will be inserted into the bile duct. This
measurement can be done by measuring the length of the cholan-
gioscope inserted into the target lesion or the length of a guidewire
after insertion into the stricture segment. Side-holes should not be
located at the portion of the catheter at the sinus tract because this
can cause bile leakage.
Results of cholangioscopic stone removal
The cholangioscopic approach is a good therapeutic option for treat-
ment of intrahepatic stones, especially for multiple and bilateral
intrahepatic stones. Complete stone removal can usually be achieved
after several cholangioscopic sessions (Fig. 38.20). However, there
are some difcult cases with multiple strictures and angulations.
From studies of long-term results and risk factors for stone recur-
rence, complete stone clearance can be achieved in 80% of patients.15
The rate of complete stone clearance by the cholangioscopic approach
is signicantly lower in patients with severe intrahepatic strictures
than those without strictures. Patients with severe intrahepatic stric-
tures also show a higher recurrence rate than those with mild stric-
tures or no strictures. In addition, the stone recurrence rate is
different according to the hepatic functional reserve. The recurrence
rate is signicantly higher in patients with advanced biliary cirrhosis
such as Child class B or C than those with mild cirrhosis such as
Child class A or no cirrhotic change.16
406
INDICATIONS AND CONTRAINDICATIONS
Standard ERCP is important and indicated in the initial evaluation
of a patient presenting with acute cholangitis. A good quality chol-
angiogram can be obtained with ERCP. However, with tight stric-
tures and impacted stones the biliary anatomy upstream to the
pathology may not be seen. Extrahepatic stones and stones with
strictures in the main intrahepatic ducts can be treated using con-
ventional ERCP techniques with retrieval basket and balloons and
dilating balloons. This is indicated in providing acute, albeit tempo-
rary, relief of biliary obstruction and cholangitis with the placement
of plastic stents or nasobiliary drain.
PTCS is indicated in patients with peripheral intrahepatic stones,
multiple bilateral intrahepatic stones, or stones located upstream to
tight intrahepatic strictures. It allows better access to the affected
intrahepatic ducts with a cholangioscope. It is also indicated in pro-
viding biliary drainage with a PTBD.
Apart from the general contraindications to ERCP there are no
specic contraindications in the setting of RPC. PTCS is contrain-
dicated in patients with bleeding diathesis. The risk of bleeding is
high in patients with concomitant advanced cirrhosis. Ascites makes
establishment of a mature percutaneous tract difcult and special
precautions are needed, including the use of sheaths. For patients
with a history of allergy to contrast media, prophylactic steroids
should be given or, alternatively, non-ionic contrast media is given.
Patients must be cooperative and well sedated for both ERCP and
PTCS. These procedures are relatively high risk and sophisticated
and the full concentration of endoscopists and assistants is
needed.
COMPLICATIONS AND THEIR MANAGEMENT
Complications of ERCP are as for other indications. Cholangitis may
be a particularly difcult problem especially when contrast is injected
into intrahepatic ducts that cannot be drained and/or following the
insertion of catheters and manipulation in the biliary tree.17 In these
instances percutaneous drainage of the appropriate duct should be
carried out. Intravenous antibiotics are administered prior to and
continued post-procedurally.
The main complications from PTCS are related to transhepatic
catheter placement and dilation of the cutaneous hepatic stula.
Hemobilia due to biliovenous stula is a commonly encountered
problem. Bleeding may not be apparent when the percutaneous
catheter is in place as the catheter provides a tamponading effect and
occurs when the catheters are removed.
Complications with the use of the cholangioscope are usually
minor. Bleeding is reported in about 10% of cases but major bleed-
ing requiring transfusion or therapeutic intervention occurs in 1
2%. Perforation of the intrahepatic bile ducts is reported in 1.7%.18
Cholangitis is a problem with PTCS and occurs following vigor-
ous manipulation of the biliary tree and the inability to completely
drain the bile ducts where lakes of contrast-lled intrahepatic
ducts may be present.
If a sinus or T-tube tract is not mature enough, partial or com-
plete migration of the catheter may cause bile leakage and bile peri-
tonitis, which is a serious complication. The risk of a percutaneous
tube dislocation is reduced if the distal end of the tube is place in
the common bile duct or through the papilla into the duodenum.
When dislocation of the tube occurs, immediate replacement of the
 Chapter 38 Recurrent Pyogenic Cholangitis

A B C

D E

Fig. 38.18 Catheter insertion using double guidewires. A Two guidewires are inserted into the distal common bile duct. B The
cholangioscope was removed while two guidewires are kept in the bile duct. C Insertion of a 16 Fr PTCS catheter is tried and would
not pass the angulated area. D After several negotiations, it was possible for the catheter to pass the angulated area. E Catheter chol-
angiogram shows the tip of PTCS catheter located in the distal common bile duct.

tube along the same tract is required, although this may not be
possible.

LONG-TERM MANAGEMENT OF RPC

As complete dilation of intrahepatic strictures is rarely successful,
one of the biggest problems in RPC is persistent infection and recur-
rent stone formation. When a percutaneous access has already been
created for cholangioscopy, access can be maintained with the place-
ment of a Yamakawa-type transhepatic tube, which can be occluded
at the skin-level. Repeat cholangioscopy with extraction of newly
formed stones or further dilation of strictures can then be carried
out.

SURGERY

Fig. 38.19 PTCS catheters before (above) and after making side- Non-surgical approaches to treatment of RPC have often proven to
holes (below). be difcult with persisting or recurrent problems. Another concern

407
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

B C

Fig. 38.20 Sequential dilations of multiple strictures and stone removal. A Upon catheter cholangiogram, many branches of right
intrahepatic ducts are missing. Several right intrahepatic stones are faintly delineated. B After several sessions of stricture dilation using
balloons and PTCS catheters, this cholangiogram reveals multiple stones in the right superior and inferior branch ducts. C After full
dilation of multiple segmental ducts and cholangioscopic stone removal, many right intrahepatic ducts are visualized.

in RPC is the development of cholangiocarcinoma. Hepatectomy,
where feasible, can provide denitive treatment for hepatolithiasis
as it removes not only the stones, but the strictured bile ducts and
abolishes the possibility of recurrent stone formation and risks of
cholangiocarcinoma. Intrahepatic stones that are conned to one
lobe allow removal of the affected lobe with cure of the disease. In
most cases there is predilection for stones to be conned to the left
lobe of the liver. Bilobar stone disease poses a therapeutic dilemma.
At aggressive surgical centers the more severely affected lobe is
removed and a hepaticocutaneous jejunostomy is fashioned in the
remaining lobe for access to allow further percutaneous cholangio-
scopic treatment.19
RELATIVE COST
Surgery, although the highest initial cost, offers the best denitive
treatment when feasible and is probably the most cost-effective treat-
ment. PTCS requires repeated procedures and uses both endoscopy
and radiology time. Cost of treatment will therefore escalate over
time.

REFERENCES
1. Maki T. Pathogenesis of calcium bilirubinate gallstone: role of E.
coli, b-glucuronidase and coagulation by inorganic ions,
polyelectrolytes and agitation. Ann. Surg 1966; 164:90100.
2. Leung JW, Yu AS. Hepatolithiasis and biliary parasites. Ballires
Clinical Gastroenterology 1997; 11:681706.
3. Cheung KL, Lai EC. The management of intrahepatic stones.
Advances in Surgery 1996; 29:111129.
4. Fan ST, Lai EC, Mok FP, et al. Acute cholangitis secondary to
hepatolithiasis. Arch Surg 1991; 126:10271031.
5. Lai EC, Mok FP, Tan ES, et al. Endoscopic biliary drainage for
severe acute cholangitis. N Engl J Med 1992; 326:15821586.
6. Huang MH, Ker CG. Ultrasonic guided percutaneous transhepatic
bile drainage for cholangitis due to intrahepatic stones. Arch Surg
1988; 123:106109.
7. Mahadeva S, Prabakharan R, Goh KL. Endoscopic intervention for
hepatolithiasis associated with sharp angulation of right
intrahepatic ducts. Gastrointest Endosc 2003; 58:279282.
8. Lukes P, Ceder S, Wihed A, et al. Evaluation of percutaneous
cholangiography and percutaneous biliary drainage in obstructive
jaundice. Eur J Radiol 1985; 5:267270.
9. Makuuchi M, Yamazaki S, Hasegawa H, et al. Ultrasonically guided
cholangiography and bile drainage. Ultrasound Med Biol 1984;
10:617623.
10. Seo DW, Kim MH, Lee SK, et al. Usefulness of cholangioscopy in
patients with focal stricture of the intrahepatic duct unrelated to
intrahepatic stones. Gastrointest Endosc 1999; 49:204209.
11. Nimura Y, Shionoya S, Hayakawa N, et al. Value of percutaneous
transhepatic cholangioscopy (PTCS). Surg Endosc 1988; 2:213219.
12. Cheng YF, Chen TY, Ko SF, et al. Treatment of postoperative
residual hepatolithiasis after progressive stenting of associated
bile duct strictures through the T-tube tract. Cardiovasc Intervent
Radiol 1995; 18:7781.
13. Sheen-Chen SM, Cheng YF, Chen FC, et al. Ductal dilatation and
stenting for residual hepatolithiasis: a promising treatment
strategy. Gut 1998; 42:708710.
14. Yoshida J, Chijiiwa K, Shimizu S, et al. Hepatolithiasis: outcome of
cholangioscopic lithotomy and dilation of bile duct stricture.
Surgery 1998; 123:421426.
15. Cheng YF, Lee TY, Sheen-Chen SM, et al. Treatment of
complicated hepatolithiasis with intrahepatic biliary stricture by

408

ductal dilatation and stenting: long-term results. World J Surg
2000; 24:712716.
16. Lee SK, Seo DW, Myung SJ, et al. Percutaneous transhepatic
cholangioscopic treatment for hepatolithiasis: an evaluation of
long-term results and risk factors for recurrence. Gastrointest
Endosc 2001; 53:318323.
17. Audisio RA, Bozzetti F, Severini A, et al. The occurrence
of cholangitis after percutaneous biliary drainage:
Chapter 38 Recurrent Pyogenic Cholangitis
evaluation of some risk factors. Surgery 1988; 103:
507512.
18. Seo DW. Complications of cholangioscopy. In: Seo DW, Lee SK,
Kim MH, et al. (eds) Cholangioscopy. Seoul: Koonja/Lippincott
Williams & Wilkins; 2002:160167.
19. Chen DW, Poon RTP, Liu CL, et al. Immediate and long-term
outcomes of hepatectomy for hepatolithiasis. Surgery 1993;
135:386393.
409
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Biliary Intervention in Acute
39 Gallstone Pancreatitis
Kanul Jajoo and David L. Carr-Locke
BACKGROUND
Gallstones are the most common cause of pancreatitis, accounting
for approximately 35% of cases of acute pancreatitis (AP) in the
United States and Europe1,2 and up to 65% of cases in Asia.3 The
majority of patients with acute gallstone pancreatitis (AGP) will
follow a benign clinical course. However, up to 25% will progress to
severe acute pancreatitis (SAP), which confers a signicant increase
in morbidity and mortality.4 Although the exact mechanism by
which gallstones cause AP remains elusive, the correlation between
gallstones and AP is well documented. Gallstones were found in the
stool of approximately 90% of patients with recent ABP, whereas
they are found in only 10% of patients with cholelithiasis without
AP.5 In addition, persistent obstruction of the ampulla by a common
bile duct (CBD) stone is believed to result in more severe pancreatic
injury. Endoscopic retrograde cholangiopancreatography (ERCP)
and endoscopic sphincterotomy (ES) are effective tools for removal
of such an obstructing stone and re-establishment of biliary drain-
age,3,6,7 with success rates exceeding 90%.
BOX 39.1
Four controlled studies, one in abstract form only, have
addressed the role of ERCP in acute gallstone pancreatitis
(AGP)
Three studies demonstrated a benet in morbidity for those
patients who underwent early ERCP and one also
demonstrated a mortality benet
Systematic review of these studies conrmed that the
morbidity benet of early ERCP is seen in those patients with
severe AGP
DIAGNOSIS OF ACUTE GALLSTONE
PANCREATITIS
The effective utilization of ERCP intervention for the management
of AGP necessitates differentiating AGP from other causes of AP.
As with all examples of diagnostic investigation, this requires a
combination of keen history taking, physical examination and inter-
pretation of laboratory values and imaging. A history of cholelithiasis
or symptoms consistent with biliary colic is suggestive, but not
diagnostic, of a biliary etiology. The physical examination is not
specic for distinguishing AGP from other causes. However, con-
current cholecystitis producing Murphys sign or signs of cholangi-
tis are ndings that can increase the likelihood that gallstones are
the etiology of a patients pancreatitis. Much of the published litera-
ture involves the use of biochemical values and imaging studies for
predicting a biliary etiology of AP.
Serum amylase levels have been shown to be higher in patients
with AGP in comparison to those with alcohol-related AP and
authors have postulated that a serum amylase level of greater than
1000 indicates a biliary cause of AP.8,9 The presence of elevated liver
chemistries has been evaluated and meta-analysis of these studies
demonstrated that elevations of alanine aminotransferase (ALT)
levels greater than threefold are suggestive of AGP.10 This study also
found that total bilirubin level and alkaline phosphatase level were
not useful and aspartate aminotransferase (AST) was no more useful
than ALT in diagnosing AGP. In addition, once AGP is established,
patients with rising serum pancreatic enzymes or liver tests carry a
fourfold risk of persistent CBD stones and approximately threefold
risk of complications when compared to patients with stable or
declining laboratory values.11
Demonstration of cholelithiasis by imaging can further support
the diagnosis of AGP. Abdominal ultrasound is the preferred initial
imaging study given its high sensitivity and specicity (>95%) for
gallstones.12 In the setting of AP, however, this sensitivity can be
reduced.13 A more recent study found that abdominal ultrasound in
the setting of AP remains a very sensitive test (86%) and, when
combined with an elevation of ALT greater than 80 IU/L, is 98%
sensitive and 100% specic for a biliary etiology.14 The clinician
should be aware that the lack of biliary dilatation on ultrasound does
not exclude choledocholithiasis as the cause of AP, especially in the
rst 48 hours of an attack.
The attendant risks of ERCP, the gold standard for detecting
choledocholithiasis, have prompted the study of magnetic resonance
cholangiopancreatography (MRCP) and endoscopic ultrasound
(EUS) as alternative diagnostic modalities. MRCP has been shown
to have high sensitivity (8495%) and high specicity (96100%) for
the diagnosis of common duct stones.1517 The most common cause
of a false negative MRCP was gallstone size of less than 5 mm.16
EUS demonstrates equivalent accuracy to MRCP for the detection
of choledocholithiasis.17 EUS can detect choledocholithiasis at a sen-
sitivity of 98% with 99% specicity,18 and may safely replace diag-
nostic ERCP.19 The exact role of each of these modalities in the
diagnosis of AGP and the patient group to which they should be
individually applied must be claried in future studies and is inevi-
tably dependent on local availability and logistics.
411
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

System Complication At admission

a
Pulmonary Mechanical ventilation; pneumonia with Age > 70 years
a
hypoxemia (PaO2 60 mm Hg); and WBC > 18,000/mm3
hypoxemia (PaO2 60 mm Hg) or dyspnea Serum glucose > 200 mg/dl (11.1 mmol/L)
a
requiring frequent assessment of need for Serum LDH > 400 IU/L
intubation Serum AST > 250 IU/L
Cardiovasoular Hypotension requiring pressor support; Within 48 hours of hospital admission
ischemia or acute myocardial infarction noted Hematocrit fall > 10 percentage points
a
on electrocardiogram or cardiac enzymes; BUN rise > 2 mg/dl (0.7 mmol/L)
a
and new onset arrhythmia other than sinus Base decit > 5 meq/L (5 mol/L)
a
tachycardia Fluid sequestration > 4 L
Infectious Sepsis of any origin Serum calcium < 8 mg/dl (2 mmol/L)
Renal New onset oliguric or nonoliguric renal failure Arterial PaO2 < 60 mm Hg
or new onset dialysis
Hematologic Disseminated intravascular coagulation and Table 39.3 Ransons criteria of pancreatitis severity for biliary
platelet counts <50 109/L pancreatitis
Modied from Table 4, Reference 24.
Neurologic Glasgow Coma Scale score 9 and diminished a
Denotes changes from original Ranson criteria for all other causes of acute
responsiveness or agitation (requiring pancreatitis.
signicant sedation) with need for frequent
airway monitoring
Gastrointestinal Stress ulcer with hematemesis or melena
tract (requiring >2 U of blood per 24 hours) the presence of local complications such as pancreatic necrosis,
abscess or uid collection by cross-sectional imaging. With these
Table 39.1 Denition of severe complications requiring intensive parameters, the Atlanta Classication of 1992 standardized the de-
care unit monitoring and treatment (Reproduced with permission nition of SAP as the presence of local complications and/or organ
from Meek K et al. Arch Surg 2000;135(9):10481052.) failure (Table 39.1), or an Acute Physiology and Chronic Health
Evaluation II (APACHE II) (Table 39.2) score greater than 8 or
greater than three Ransons criteria.21 It is important to note that a
Physiological variable Reference range modication of Ransons original criteria is used for biliary pancre-
Rectal temperature, C 3638.4 atitis (Table 39.3).22
Mean arterial pressure, mm Hg 70109 Organ failure, particularly persistent or worsening organ failure,
Heart rate (ventricular response), 70109 is a strong determinant of mortality in patients with SAP.23,24 Though
beats/min many denitions of organ failure have been used, more recent
Respiratory rate, breaths/min 1224 studies utilize the multiple organ dysfunction syndrome (MODS)
Oxygenation, mm Hg PAO2 PaO2 < 200 or
score or the systemic inammatory response syndrome (SIRS) score
PO2 > 70
to ensure that ndings can be generalized. Mortality in the setting
Arterial pH 7.337.49
Serum sodium level, mmol/L 130149 of AP with organ failure can range from 20% to as high as 50% and
Serum potassium level, mmol/L 3.55.4 is dependent upon the duration, severity and number of organ
Serum creatinine level, mol/L (mg/dL) 0.61.4 (53123) systems in failure.20,21,25 Prognostic indices have been formulated to
(double point score for acute renal predict which patients are more likely to develop severe AGP and to
failure) direct appropriate care toward that group. These include Ransons
Hematocrit 0.300.46 criteria (biliary version), modied Glasgow criteria and APACHE II
Leukocyte count, 109/L 0.0030.015 score. Radiologic scores such as the Balthazar score and the modi-
Glasgow Coma Scale score (GCS) 15actual GCS score ed CT severity index which are based on the extent of pancreatic
necrosis and uid collections, have been shown to correlate with
Table 39.2 The APACHE II scoring systema (Reproduced with mortality.26,27 Several biochemical markers of inammation have
permission from Meek K et al Arch Surg 2000;135(9):104852.)
a
To calculate the Acute Physiology and Chronic Health Evaluation (APACHE) II score,
been studied to predict SAP, but serum C-reactive protein level of
the 12 physiological variables are assigned points between 0 and 4, with 0 being greater than 150 mg/L at 4872 hours after symptom onset remains
normal and 4 being the most abnormal. The sum of these values is added to a point the standard.28 Recent data suggest that a genetic polymorphism that
weighting for patient age (44 years = 0; 4554 years = 2; 5564 years = 3; 6574
confers an enhanced chemokine response to an inammatory stim-
years = 5; 75 years = 6) and a point weighting for chronic health problems. PAO2
PaO2 indicates alveolar-arterial difference in partial pressure of oxygen. ulus is a risk factor for progression to SAP.29 The search continues
for a biochemical marker that can be easily measured in the rst 24
hours of AP and reliably predicts progression to severe disease.
ASSESSMENT OF SEVERITY OF
ACUTE PANCREATITIS ENDOSCOPIC THERAPY FOR ACUTE
GALLSTONE PANCREATITIS
Early recognition of patients with severe acute pancreatitis (SAP) is
crucial, as those patients will require intensive care management The mainstay of therapy for all forms of SAP remains supportive
and will likely benet from endoscopic intervention.3,7,20 Several care including aggressive hydration, adequate nutritional support,
clinical and radiographic parameters have been used to evaluate the pain control and often an intensive care unit (ICU).30 With regard
severity of AP: the presence of organ failure; prognostic indices; and to severe AGP in particular, early endoscopic therapy has become

412
 Chapter 39 Biliary Intervention in Acute Gallstone Pancreatitis

an integral management strategy, supported by anecdotal reports31
and evidence from randomized clinical trials (RCTs).3,7,20 An addi-
tional RCT from Germany brought into question the benet of early
ERCP with ES in a subgroup of patients without signs of biliary
obstruction.6 These studies differ on the assessment of pancreatitis
severity, timing to ERCP, exclusion criteria and possibly endoscopic
expertise. The four RCTs designed to assess the safety and benet
of early ERCP in AGP are described below and summarized in Table
39.4.
Neoptolemos et al. 1988
This landmark study comparing ERCP and ES against conservative
management of AGP was performed by Neoptolemos and colleagues
from 1983 through 1987 and published in 1998.7 The investigators
randomized 121 of 146 consecutive patients who presented to a
single institution with suspected AGP to receive either conservative
management or ERCP within 72 hours of admission. The diagnosis
of AGP was established by ultrasound and laboratory data. The
severity of pancreatitis was predicted within 48 hours of admission
using the modied Glasgow criteria.32 If choledocholithiasis was
found on ERCP, an ES with stone extraction was performed.
Outcome measures included mortality, length of stay, local compli-
cations and organ failure. Predicted severe AP was present in 44%
of all patients enrolled (25 of 59 in the ERCP group and 28 of 62 in
the conservative management group). ERCP was successful in 94%
of mild disease and 80% of severe disease. One ERCP-related com-
plication was cited, a case of vertebral osteomyelitis. There were no
cases of ERCP-related hemorrhage, cholangitis or perforation.
The overall mortality was not signicantly different in the two
patient groups (ERCP group: 2% vs conservative management
group: 8%; p = 0.23). However, the overall morbidity was signi-
cantly lower in the group that underwent ERCP within 72 hours of
admission (17% vs 34%; p = 0.03). Sub-group analysis demonstrated
that the morbidity difference was limited to the group of patients
with predicted SAP. In patients with predicted SAP who were ran-
domized to urgent ERCP, the complication rate was 24%, in com-
parison to 61% in patients with predicted SAP managed conservatively
(p < 0.01). Accordingly, the length of hospitalization was shorter in
the patients with SAP who underwent urgent ERCP (9.5 days vs 17
days; p < 0.035).
The investigators acknowledged the concern that the benet of
early ERCP +/ ES might be a result of treating cholangitis and not
pancreatitis. They controlled for this possible confounding factor by
excluding the patients who presented with cholangitis and analyzing
the remaining patients separately. The complication rate remained
signicantly lower in the group of patients without cholangitis who
underwent urgent ERCP (11% vs 33%; p = 0.02). Again, the majority
of this difference occurred in the sub-group of patients with pre-
dicted SAP.
In summary, this study by Neoptolemos and colleagues demon-
strated that it is safe to perform ERCP in patients with AGP admitted
to an expert center and early ERCP is associated with signicantly
decreased morbidity and hospital stay in patients with predicted
severe AGP in comparison to conservative management.
Fan et al. 1993
The investigators of this trial from Hong Kong randomized 195
patients with acute pancreatitis of all etiologies to undergo urgent
ERCP within 24 hours of hospital admission or conservative man-
agement followed by selective ERCP for clinical deterioration. The
authors utilized this approach of selecting all patients with pan-
creatitis in order to minimize selection bias. Analysis of the sub-
group of patients with AGP, revealed that 127 of the 195 randomized
patients (65%) had biliary stones. Sixty-four of the 97 patients ran-
domized to early ERCP were found to have biliary stones and 38 of
these required ES for CBD or ampullary stones. Of the 98 patients
in the conservative therapy group, 63 had biliary stones and 27 of
these patients required ERCP for clinical deterioration. Ten of these
patients were found to have CBD or ampullary stones.
The severity of pancreatitis was graded by serum urea concentra-
tion and plasma glucose concentration and Ransons score. Patients
were categorized as having SAP if serum urea concentration was
greater than 45 mg/dL or if plasma glucose concentration was
greater than 198 mg/dL at admission. Predicted SAP was diagnosed
in 41.5% of the patient population, distributed evenly between the
treatment groups. The overall morbidity (urgent ERCP group: 18%

No. treated No. control

Study patients patients Study design Outcomes
Neoptolemos 59 62 Single center Signicant morbidity reduction in severe AGP
Consecutive patients with suspected Signicant length of stay reduction in severe
AGP AGP
Fan 97 98 Single center Signicant morbidity reduction in AGP
Consecutive patients with AP, regardless Signicant reduction in biliary sepsis in severe
of etiology AGP
AGP analyzed separately
Flsch 126 112 Multi-center Similar morbidity rates between study groups
Patients with suspected Signicantly higher incidence of respiratory
AGP, excluded those with bilirubin insufciency in ERCP group
> 5 mg/dL
Nowak 178 102 Single center Signicant reduction in both morbidity and
Consecutive patients with suspected AGP mortality in the early ERCP + ES group
All underwent duodenoscopy, immediate
ES if obstructed, randomized if not

Table 39.4 Summary of randomized controlled trials

413
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
vs conservative management group: 29%; p = 0.07) and mortality
(5% vs 9%; p = 0.4) were not signicantly different in the two patient
groups. When considering only those patients with biliary stones,
the morbidity rate in the urgent ERCP group was signicantly lower
than in the conservative management group (16% vs 33%, p = 0.03)
and there was a trend toward lower mortality (2% vs 8%; p = 0.09).
These ndings were driven by the signicant morbidity advantage
of urgent ERCP in the sub-group of patients with predicted SAP. In
particular, the incidence of biliary sepsis among those patients pre-
dicted to have SAP was signicantly lower in the urgent ERCP group
than in the conservative management group (0% vs 20%; p = 0.008).
In contrast, among patients with mild pancreatitis, there was no
difference in the incidence of biliary sepsis between the two study
groups.
In summary, this trial demonstrated a morbidity benet in
patients with predicted severe AGP who underwent urgent ERCP
+/ ES as compared to those managed conservatively. Despite the
high prevalence of cholelithiasis in the study population, this trial
corroborates the ndings of the earlier study from the UK.
Flsch et al. 1997
In this German multi-center study, 126 patients with AGP were
randomly assigned to early ERCP within 72 hours of the onset of
symptoms and 112 patients with AGP were assigned to conservative
management. The inclusion criteria in this study were distinct from
the previous studies in that patients with obstructive jaundice (total
bilirubin >5 mg/dL) were excluded. In doing so, the investigators
sought to determine the effect of early ERCP upon AGP independent
of its known benet in patients with cholangitis.33 In these patients
with acute pancreatitis, the diagnosis of AGP was made if gallstones
were seen on imaging or if two of three serum liver chemistry values
(ALT, alkaline phosphatase and/or total bilirubin) were abnormal.
The severity of pancreatitis was predicted by the modied Glasgow
criteria. Early ERCP was successful in 96% of the treatment group
and 46% of patients in this group were found to have choledocholi-
thiasis. Elective ERCP was required in 20% of the conservative treat-
ment group and 59% of those patients were found to have bile duct
stones.
Predicted SAP was seen in 19.3% of patients overall and similarly
distributed between the treatment groups. Complications directly
attributable to ERCP were minimal, with post-sphincterotomy hem-
orrhage seen in 2.8% and no duodenal wall perforations reported.
Overall complications were similar in the early ERCP and control
groups (46% vs 51%) and mortality rates were also similar (11% vs
6%; p = 0.10). Stratication of patients by predicted severity of pan-
creatitis did not alter these ndings. Though systemic complications
overall were not signicantly different, the patients in the early ERCP
group had a higher rate of respiratory insufciency, as dened by pO2
<60 mm Hg despite use of an oxygen mask (12% vs 4%; p = 0.03).
Several critiques of this study have been put forth in the litera-
ture. In this multi-center trial involving 22 institutions, the majority
of patients were enrolled by three centers. This brings into question
the level of experience and frequency of patients with AGP at a
number of the study centers. Also, the excessive rate of respiratory
insufciency in the treatment group was not seen in any of the other
trials addressing this subject. The investigators concluded that early
ERCP in patients with AGP, without biliary obstruction or sepsis,
does not confer a mortality or morbidity benet and may result in a
higher rate of respiratory insufciency as compared to conservative
management.
414
Nowak et al. 1998
The fourth and largest prospective study to assess the role of early
ERCP in acute gallstone pancreatitis included 280 patients and was
conducted at a single center in Poland. The predicted severity of
AGP was assessed using Ransons criteria. All patients underwent
an ERCP within 24 hours of admission and 75 patients underwent
an endoscopic sphincterotomy due to evidence of an impacted stone
at the papilla. Patients with a normal appearing papilla were ran-
domized to immediate ES (n = 103) or conservative management (n
= 102). The investigators found a signicant reduction in morbidity
(17% vs 36%; p < 0.001) and mortality (2% vs 13%; p < 0.001) in the
group of patients randomized to ERCP + ES. After stratifying for
severity of pancreatitis, the morbidity and mortality benet of early
ERCP + ES remained signicant in the group with predicted mild
pancreatitis.
This study is discussed in brief, as it has only been published in
abstract form. Particulars of the study design, patient groups and
criteria for enrollment are not provided. Comprehensive critical
analysis is not possible until a complete manuscript is made
available.
SYSTEMATIC REVIEWS
Any attempt to create a unied recommendation for the care of
patients with AGP based on these trials is hindered by their distinct
study methods. A meta-analysis by Sharma and Howden34 sought to
estimate the overall effect of ERCP for AGP. They performed a
pooled analysis of all four trials, assessing 460 treated patients and
374 controls. In analyzing complications and mortality, they found
that the number of patients with AGP needed to treat (NNT) with
ERCP + ES for avoidance of complications was 7.6 and the NNT for
avoidance of death was 25.6. Sub-group analysis by severity of AGP
was not possible due to unavailable data. The authors concluded that
ERCP + ES reduces mortality and morbidity in patients with AGP.
These results must be viewed with caution as this was pooled data
and the largest contribution to the pool of patients came from the
Nowak study, which is only available in abstract form.
The Cochrane Database systematic review of this subject by Ayub
and colleagues only included the studies by Neoptolemos, Fan and
Flsch. The authors sought to assess the value of ERCP +/ ES
versus conservative therapy in patients with AGP. In particular, this
review sought to address the effect of confounding by indication by
controlling for associated acute cholangitis and by stratifying accord-
ing to disease severity. To this end, the investigators of the Flsch
trial provided additional data regarding the severity of pancreatitis
in each patient group. The authors concluded that ERCP +/ ES was
associated with a signicant reduction in morbidity in predicted
severe AGP (OR = 0.27, 95% CI = 0.14 to 0.53). However, there was
no signicant difference in morbidity in patients with predicted mild
AGP. In addition, no signicant difference in mortality was found,
regardless of predicted disease severity. Figure 39.1 demonstrates
their ndings, comparing ERCP +/ ES versus conservative man-
agement and stratied by severity of AGP.
Acute gallstone pancreatitis with biliary obstruction
Acosta et al. 2006
The four studies above and their inherent differences in patient
inclusion, interval time to ERCP and classication of severe acute
pancreatitis prompted Acosta and colleagues to perform a random-
 Chapter 39 Biliary Intervention in Acute Gallstone Pancreatitis

Review: Endoscopic retrograde cholangiopancreatography in gallstone-associated acute pancreatitis Fig. 39.1 Cochrane Database systematic
Comparison: 01 Early ERCP+/ES versus Conservative Mx review, K Ayub et al.
Outcome: 02 Complications stratified by severity of GAP

Study Early Conservative Odds ratio (fixed) Weight Odds ratio (fixed)
ERCP+/ES Mx 95% CI (%) 95% CI
n/N n/N
01 Mild GAP
Fan 1993 8/56 6/58 7.7 1.44 [0.47, 4.47]
Flsch 1997 35/84 36/76 33.5 0.79 [0.42, 1.48]
Neoptolemos 1988 3/33 4/32 5.6 0.70 [0.14, 3.41]

Subtotal (95% CI) 46/173 46/166 46.8 0.89 [0.53, 1.49]

Test for heterogeneity chi-square = 0.92 df = 0.6299
Test for overall effect = 0.45 p = 0.7
02 Severe GAP
Fan 1993 9/41 23/40 27.6 0.21 [0.08, 0.55]
Flsch 1997 17/26 14/20 8.3 0.81 [0.23, 2.83]
Neoptolemos 1988 3/20 15/25 17.2 0.12 [0.03, 0.51]

Subtotal (95% CI) 29/87 52/85 53.2 0.27 [0.14, 0.53]

Test for heterogeneity chi-square = 4.47 df = 2 p = 0.1071
Test for overall effect = 3.86 p = 0.0001

Total (95% CI) 75/260 98/251 100.0 0.56 [0.38, 0.83]

Test for heterogeneity chi-square = 12.68 df = 5 p = 0.0266
Test for overall effect = 2.86 p = 0.004

.1 .2 1 5 10
Favors ERCP +/ ES Favors cons Mx

ized trial of early ERCP + ES in a more narrowly dened group of
patients. The investigators randomized 61 consecutive patients with
AGP and persistent ampullary obstruction to undergo ERCP +/ ES
between 24 and 48 hours of the onset of symptoms (study group, n
= 30) or conservative management followed by selective ERCP +/
ES if jaundice or cholangitis were present 48 hours after the onset
of symptoms (control group, n = 31). Persistent ampullary obstruc-
tion was dened by a previously validated method.35 This method
utilized three clinical ndings to detect ampullary obstruction:
severe and continuous epigastric pain, bile-free gastric aspirate and
elevated serum bilirubin (followed serially every 6 hours). Ranson
or Acosta criteria were utilized to predict severity of pancreatitis.
The majority of patients experienced spontaneous relief of biliary
obstruction within 48 hours of the onset of symptoms (71% of
control group and 53% of study group). Fourteen patients in the
study group underwent ERCP within 48 hours of symptom onset;
impacted stones were found in 11 (79%) of these patients. There
were no deaths in either group and no complications attributable to
ERCP or ES. The study group had a signicantly lower incidence of
immediate complications (3% vs 26%; p = 0.026) and overall com-
plications (7% vs 29%; p = 0.043). The incidence of severe AGP
(10%) was relatively low in this study. The two groups did not differ
in length of hospitalization or time to cholecystectomy. Collective
analysis of both the study and control groups demonstrated that
ampullary obstruction of less than 48 hours duration was associated
with fewer complications (p < 0.001), shorter time interval to chole-
cystectomy (p = 0.018), and shorter hospitalization (p = 0.003).
Biliary microlithiasis
Small diameter biliary stones, measuring less than 5 mm and known
as microlithiasis, biliary sludge or biliary sand, have been implicated
as a cause of recurrent acute pancreatitis and other biliary com-
plications.36 In practice, microlithiasis can be diagnosed by trans-
abdominal ultrasound and is seen as mobile, echogenic material that
layers with gravity and does not produce shadows.37 EUS has also
been shown to be effective in detecting microlithiasis, particularly
in the setting of typical biliary pain and normal abdominal ultra-
sound.38 The gold standard for the detection of biliary microlithiasis
is microscopic analysis, which documents cholesterol monohydrate
crystals or calcium bilirubinate granules in up to 80% of patients
with AP of presumed biliary origin in whom gallstones could not be
documented on imaging.39 Though no prospective, randomized con-
trolled trials have been performed to establish the role of ERCP in
patients with AGP due to microlithiasis, uncontrolled studies have
suggested that these patients benet from intervention.40,41
Cholecystectomy after AGP
Once a patient stabilizes from an episode of mild AGP, laparoscopic
cholecystectomy should be performed prior to discharge from the
hospital.42 Delay in cholecystectomy is associated with a 20% risk of
recurrent biliary complications including AGP, cholangitis and cho-
lecystitis43 and a near 50% recurrence rate of any biliary symptoms.44
A recent prospective study of 178 Chinese patients over the age of
60 demonstrated that those patients randomized to early cholecys-
tectomy after ES and bile duct clearance had a signicantly lower
rate of biliary events compared to those randomized to conservative
management after ES (7% vs 24%; p = .001). In patients who are
unable to undergo surgery, ES does confer some degree of protec-
tion from subsequent biliary events.
In the case of severe AGP, cholecystectomy should be delayed
until the systemic inammatory response has subsided. In cases
with signicant pancreatic necrosis or pancreatic uid collection,

415
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Fig. 39.2 Algorithm for the management

Acute gallstone pancreatitis (AGP) of acute gallstone pancreatitis.

Severe AGP Mild AGP

Deterioration, jaundice,
ERCP +/ ES Recovery
cholangitis

Cholecystectomy ERCP + ES Cholecystectomy

(after recovery) (non-surgical candidate) (same admission)

cholecystectomy should be delayed 36 weeks due to an increased
risk of infection and surgical complications.45,46 If necessary and
indicated, cholecystectomy can be combined with drainage proce-
dures for pancreatic uid collection or debridement of pancreatic
necrosis.
ALGORITHM FOR THE MANAGEMENT OF ACUTE
GALLSTONE PANCREATITIS
The studies presented above provide a framework within which to
manage patients with AGP (Fig. 39.2). We advocate ERCP +/ ES in
patients with severe AGP, dened by Ransons criteria or the modi-
ed CT severity index when available, as soon as that diagnosis is
made. Additional indications for ERCP include concurrent cholan-
gitis or jaundice, persistent ampullary obstruction or clinical deterio-
ration in a patient who initially presented with mild disease. Once
selection criteria are met for ERCP, ES should be performed in those
patients with conrmed choledocholithiasis or ampullary edema
causing obstruction. In patients who cannot undergo cholecystec-
tomy due to medical co-morbidities, ES is protective against further
bouts of AGP, but may not protect against other biliary complica-
tions. The fear that ERCP with or without ES can exacerbate existing
AP is not borne out by the literature nor our clinical experience.

REFERENCES
1. Go V, Everhart J. Pancreatitis. In: Everhart J (ed.) Digestive diseases
in the United States: epidemiology and impact. Washington, DC:
US Government Printing Ofce, 1994; 941447:693712.
2. Gullo L, Migliori M, Olah A, et al. Acute pancreatitis in ve
European countries: etiology and mortality. Pancreas 2002;
24:223227.
3. Fan S-T, Lai E, Mok F, et al. Early treatment of acute biliary
pancreatitis by endoscopic papillotomy. New England Journal of
Medicine 1993; 328:228232.
4. Bhatia M, Brady M, Shokuhi S, et al. Inammatory mediators in
acute pancreatitis. Journal of Pathology 2000; 190:117125.
5. Acosta JM, Ledesma CL. Gallstone migration as a cause of acute
pancreatitis. New England Journal of Medicine 1974;
290:484487.
6. Flsch U, Nitsche R, Ludtke R, et al. Early ERCP and papillotomy
compared with conservative treatment for acute biliary
pancreatitis. New England Journal of Medicine 1997;
336:237242.
7. Neoptolemos J, Carr-Locke D, London N, et al. Controlled trial of
urgent endoscopic retrograde cholangiopancreatography and
endoscopic sphincterotomy versus conservative treatment for
acute pancreatitis due to gallstones. Lancet 1988; 2:979983.
8. Nordestgaard AG, Wilson SE, Williams RA. Correlation of serum
amylase levels with pancreatic pathology and pancreatitis
etiology. Pancreas 1988; 3:159162.
9. Frakes JT. Biliary pancreatitis: a review. Emphasizing appropriate
endoscopic intervention. Journal of Clinical Gastroenterology
1999; 28:97109.
10. Tenner S, Dubner H, Steinberg W. Predicting gallstone
pancreatitis with laboratory parameters: a meta-analysis. American
Journal of Gastroenterology 1994; 89:18631866.
11. Cohen ME, Slezak L, Wells CK, et al. Prediction of bile duct stones
and complications in gallstone pancreatitis using early laboratory
trends. American Journal of Gastroenterology 2001;
96:33053311.
12. Cooperberg PL, Burhenne HJ. Real-time ultrasonography:
diagnostic technique of choice in calculous gallbladder disease.
New England Journal of Medicine 1980; 302:12771279.
13. Neoptolemos JP, Hall AW, Finlay DF, et al. The urgent diagnosis
of gallstones in acute pancreatitis: a prospective study of three
methods. British Journal of Surgery 1984; 71:230233.
14. Ammori BJ, Boreham B, Lewis P, et al. The biochemical detection
of biliary etiology of acute pancreatitis on admission: a revisit in
the modern era of biliary imaging. Pancreas 2003; 26:e32e35.
15. Topal B, Van de Moortel M, Fieuws S, et al. The value of magnetic
resonance cholangiopancreatography in predicting common bile
duct stones in patients with gallstone disease. British Journal of
Surgery 2003; 90:4247.
16. Grifn N, Wastle ML, Dunn WK, et al. Magnetic resonance
cholangiopancreatography versus endoscopic retrograde

416

cholangiopancreatography in the diagnosis of
choledocholithiasis. European Journal of Gastroenterology &
Hepatology 2003; 15:809813.
17. Aube C, Delorme B, Yzet T, et al. MR cholangiopancreatography
versus endoscopic sonography in suspected common bile duct
lithiasis: a prospective, comparative study. American Journal of
Roentgenology 2005; 184:5562.
18. Buscarini E, Tansini P, Vallisa D, et al. EUS for suspected
choledocholithiasis: do benets outweigh costs? A prospective,
controlled study. Gastrointestinal Endoscopy 2003; 57:510518.
19. Liu CL, Fan ST, Lo CM, et al. Comparison of early endoscopic
ultrasonography and endoscopic retrograde
cholangiopancreatography in the management of acute biliary
pancreatitis: a prospective randomized study. Clinical
Gastroenterology and Hepatology 2005; 3(12):12381244.
20. Nowak A, Marek TA, Nowakowska-Dulawa E, et al. Biliary
pancreatitis needs endoscopic retrograde
cholangiopancreatography with endoscopic sphincterotomy for
cure. Endoscopy 1998; 30:A256A259.
21. Bradley EL, 3rd. A clinically based classication system for acute
pancreatitis. Summary of the International Symposium on Acute
Pancreatitis. Archives of Surgery 1993; 128:586590.
22. Meek K, Toosie K, Stabile BE, et al. Simplied admission criterion
for predicting severe complications of gallstone pancreatitis. Arch
Surg 2000; 135(9):10481052.
23. Buter A, Imrie CW, Carter CR, et al. Dynamic nature of early organ
dysfunction determines outcome in acute pancreatitis. British
Journal of Surgery 2002; 89:298302.
24. Johnson CD, Abu-Hilal M. Persistent organ failure during the rst
week as a marker of fatal outcome in acute pancreatitis. Gut
2004; 53:13401344.
25. Perez A, Whang EE, Brooks DC, et al. Is severity of necrotizing
pancreatitis increased in extended necrosis and infected necrosis?
Pancreas 2002; 25:229233.
26. Balthazar EJ. Acute pancreatitis: assessment of severity with clinical
and CT evaluation. Radiology 2002; 223:603613.
27. Mortele KJ, Wiesner W, Intriere L, et al. A modied CT severity
index for evaluating acute pancreatitis: improved correlation with
patient outcome. American Journal of Roentgenology 2004;
183:12611265.
28. Papachristou GI, Whitcomb DC. Predictors of severity and
necrosis in acute pancreatitis. Gastroenterology Clinics of North
America. 2004; 33:871890.
29. Papachristou GI, Sass DA, Avula H, et al. Is the monocyte
chemotactic protein-12518 G allele a risk factor for severe acute
pancreatitis? Clinical Gastroenterology and Hepatology 2005;
3:475481.
30. Nathens AB, Curtis JR, Beale RJ, et al. Management of the critically
ill patient with severe acute pancreatitis. Critical Care Medicine
2004; 32:25242536.
Chapter 39 Biliary Intervention in Acute Gallstone Pancreatitis
31. van der Spuy S. Endoscopic sphincterotomy in the management
of gallstone pancreatitis. Endoscopy 1981; 13:2526.
32. Blamey SL, Imrie CW, ONeill J, et al. Prognostic factors in acute
pancreatitis. Gut 1984; 25:13401346.
33. Leese T, Neoptolemos JP, Baker AR, et al. Management of acute
cholangitis and the impact of endoscopic sphincterotomy. British
Journal of Surgery 1986; 73:988992.
34. Sharma VK, Howden CW. Metaanalysis of randomized controlled
trials of endoscopic retrograde cholangiography and endoscopic
sphincterotomy for the treatment of acute biliary pancreatitis.
American Journal of Gastroenterology 1999; 94:32113214.
35. Acosta JM, Ronzano GD, Pellegrini CA. Ampullary obstruction
monitoring in acute gallstone pancreatitis: a safe, accurate, and
reliable method to detect pancreatic ductal obstruction. American
Journal of Gastroenterology 2000; 95:122127.
36. Ko CW, Sekijima JH, Lee SP. Biliary sludge. Annals of Internal
Medicine 1999; 130:301311.
37. Chen EY, Nguyen TD. Images in clinical medicine. Gallbladder
sludge. New England Journal of Medicine 2001; 345:e2.
38. Thorboll J, Vilmann P, Jacobsen B, Hassan H. Endoscopic
ultrasonography in detection of cholelithiasis in patients with
biliary pain and negative transabdominal ultrasonography.
Scandanavian Journal of Gastroenterology 2004; 39:267269.
39. Kohut M, Nowak A, Nowakowska-Dulawa E, et al. The frequency
of bile duct crystals in patients with presumed biliary pancreatitis.
Gastrointestinal Endoscopy 2001; 54:3741.
40. Lee SP, Nicholls JF, Park HZ. Biliary sludge as a cause of acute
pancreatitis. New England Journal of Medicine 1992; 326:589593.
41. Ros E, Navarro S, Bru C, et al. Occult microlithiasis in idiopathic
acute pancreatitis: prevention of relapses by cholecystectomy or
ursodeoxycholic acid therapy. Gastroenterology 1991;
101:17011709.
42. Uhl W, Warshaw A, Imrie C, et al. IAP Guidelines for the surgical
management of acute pancreatitis. Pancreatology. 2002;
2:565573.
43. Hernandez V, Pascual I, Almela P, et al. Recurrence of acute
gallstone pancreatitis and relationship with cholecystectomy or
endoscopic sphincterotomy. American Journal of
Gastroenterology 2004; 99:24172423.
44. Boerma D, Rauws EA, Keulemans YC, et al. Wait-and-see policy or
laparoscopic cholecystectomy after endoscopic sphincterotomy
for bile-duct stones: a randomised trial. Lancet 2002;
360:761765.
45. Uhl W, Muller CA, Krahenbuhl L, et al. Acute gallstone pancreatitis:
timing of laparoscopic cholecystectomy in mild and severe
disease. Surgical Endoscopy 1999; 13:10701076.
46. Nealon WH, Bawduniak J, Walser EM. Appropriate timing of
cholecystectomy in patients who present with moderate to
severe gallstone-associated acute pancreatitis with peripancreatic
uid collections. Annals of Surgery 2004; 239:741749.
417
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Pancreatic Interventions in Acute
40 Pancreatitis: Ascites, Fistulae, Leaks and
Other Disruptions
Richard A. Kozarek
INTRODUCTION
BOX 40.1
1. With the exception of sphincter of Oddi dysfunction, the
use of ERCP to diagnose the etiology of relapsing attacks of
pancreatitis has been supplanted by pancreas protocol CT,
MRI/MRCP, and endoscopic ultrasound.
2. The most common therapeutic interaction between ERCP
and acute pancreatitis is in the setting of biliary pancreatitis.
3. Biliary intervention in pancreatitis may be related not only
to biliary calculi but also biliary obstruction from pancreatic
edema and uid collections.
4. Therapeutic pancreatography in acute pancreatitis includes
bypass of obstruction and treatment of leaks and their
consequences and should be undertaken as one aspect of
a multidisciplinary approach.
Background
The role of ERCP in acute pancreatitis has primarily been twofold.1
On the one hand, it has been used after resolution of an acute attack,
or more commonly, multiple attacks, in an attempt to dene etiol-
ogy. As such, congenital variants to include duodenal duplication,
anomalous pancreaticobiliary union, annular pancreas, or pancreas
divisum can be diagnosed as can other anatomic causes of pancre-
atitis to include ampullary adenoma or surreptitious stone disease.
For the most part, endoscopic ultrasound (EUS) and magnetic reso-
nance imaging (MRI/MRCP) have supplanted the need for an inva-
sive study that can actually cause the disease for which it is being
applied.2 There remains, however, a major role for ERCP in conjunc-
tion with sphincter of Oddi manometry and most series actually
suggest that sphincter of Oddi dysfunction (SOD) is the most
common cause of idiopathic pancreatitis when other diagnostic
studies have been exhausted.
The second role that ERCP has played is in the treatment of acute
biliary pancreatitis.3,4 This subject is covered, in detail, in the preced-
ing chapter. Sufce it to say, however, most endoscopists use ERCP
selectively, as opposed to universally, in patients with presumptive
biliary pancreatitis, particularly in patients with an intact gallbladder
or those with mild disease.
In addition to its application in conjunction with SOM in patients
with ARP and its selective application in biliary pancreatitis, ERCP
is being evaluated as a means to provide pancreatic endotherapy in
the setting of severe or smoldering pancreatitis related to ductal
disruption, or spasm or edema of the sphincter of Oddi.57 This
chapter will focus on these latter indications, although admittedly,
many of these ductal disruptions can also occur in a background of
chronic as well as acute pancreatitis. Moreover, in contrast to con-
trolled observations about the timing and appropriateness of ERCP
in acute relapsing or severe biliary pancreatitis, respectively, most
publications related to pancreatic endotherapy during an attack of
pancreatitis have been uncontrolled, anecdotal, but also, often
dramatic.
Epidemiology of ductal disruption
If the underlying pathophysiology of acute pancreatitis is colocaliza-
tion of zymogen granules with cell membranes, setting off an
inammatory cascade with local effects related to cytokine release
and recruitment of inammatory cells, it seems reasonable that this
sequence antedates disruption of ductular epithelial cells and sub-
sequent pancreatic juice leak in most cases of acute pancreatitis.5,8
However, acute sphincter obstruction in the setting of a common
bile duct stone may increase intraductal pressure leading to side-
branch or acinar leak with resultant pancreatitis. Likewise, any other
downstream obstruction may increase upstream duct pressure
leading to PD blowout and perpetuation or exacerbation of pancre-
atitis.9,10 In acute pancreatitis, this is most commonly seen with
severe edema, whereas in chronic pancreatitis, disruptions are
usually the consequence of a downstream stricture or stone. In the
setting of severe pancreatic necrosis, ductal disruption is almost
invariable, although whether the ductal disruption is the cause or
the consequence of the necrosis remains ill-dened.10,11 Nor does
the presence of a peripancreatic uid collection imply a signicant
ongoing leak in all instances. In fact, up to 40% of patients with
acute pancreatitis develop a uid collection, yet less than 5% of these
patients go on to develop a pseudocyst.12
Anatomic classication
Although this chapter will focus on pancreatic duct leaks and their
endoscopic treatment, Table 40.1 summarizes some of the other
endoscopically amenable lesions that endoscopists see not infre-
quently in a busy ERCP practice. They include bile duct obstruction
from stones, edema within the head of the pancreas, and neoplasms
419
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

that occasionally present with pancreatitis. From a pancreatic stand- or absence of concomitant necrosis, often determines the manifesta-
point, they include neoplastic obstruction of the papilla or duct, tions8,12,13 (Fig. 40.1). As such, a major blowout of the PD tail may
edema or spasm of the sphincter mechanism, and an inammatory cause an acute perisplenic uid collection with or without a high
PD stenosis. amylase, left pleural effusion. Alternatively, pancreatic juice may
Pancreatic duct leaks can be dened anatomically by site of dis- follow anatomic pathways around the left kidney and even into the
ruption, which in conjunction with the size of the leak, and presence pelvis with resultant scrotal or labial edema. Blowouts in the head
of the pancreas may be associated with C Loop edema and gastric
outlet obstruction, biliary compression or even pancreaticobiliary
stulization, right perinephric uid accumulation, and dissection
Biliary obstruction: jaundice, cholangitis into the pelvis or perihilar area. A central disruption may result in
Common bile duct stones uid collection within the lesser sac, dissection into the mediasti-
Biliary stenosis from edema, head of pancreas num or pericardium, and, when associated with signicant central
Extrinsic obstruction from pseudocyst pancreatic necrosis, result in a permanently disconnected duct/
Pancreatic duct obstruction: exacerbation/perpetuation of gland syndrome.13
pancreatitis Anatomic classications based upon the presence of an acute or
Sphincter spasm/stenosis/edema chronic pancreatic duct leak are outlined in Table 40.2. Classically,
Stenosis pancreatic duct leaks (stulas) are classied as internal or external,
Acute, inammatory the latter almost always a consequence of trauma, surgery, or inter-
Fixed, brotica
ventional radiologic drainage procedures.1418 Internal stulas, in
Neoplastic
turn, classically have included pseudocysts, pancreatic ascites, high
PD stonea
Pancreatic duct leak: see Table 40.2 amylase pleural effusions, and erosion of pancreatic uid collections
into contiguous organs, resulting in pancreaticoenteric, gastric, colic
Table 40.1 Endoscopically amenable anatomic lesions seen in or biliary stulae.1923 They also include evolving pancreatic necrosis
acute pancreatitis in which variable amounts of high amylase uid collect, usually in
a
Implies concomitant chronic pancreatitis. the context of central pancreatic necrosis.

2 Pancreaticoenteric/biliary fistula
5 Pancreatic 6 Pancreatic
pleural effusion necrosis

Surgical or
percutaneous
drain

7 External
pancreatic
fistula

3 Pseudocyst

1 Bile duct compression by 4 Pancreatic ascites

fluid collection/edema

Fig. 40.1 Consequences of pancreatic duct leak. (1) Bile duct compression by uid collection/edema. (2) Pancreaticoenteric/biliary stula.
(3) Pseudocyst. (4) Pancreatic ascites. (5) Pancreatic pleural effusion. (6) Pancreatic necrosis. (7) External pancreatic stula.

420
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions
Internal Fistula
Peripancreatic uid collection
Pseudocyst
Pancreatic ascites
High amylase pleural effusion
Pancreaticoenteric/biliary/bronchial stula
Evolving pancreatic necrosis
smoldering pancreatitis
External stula
Pancreaticocutaneous stula
Table 40.2 Manifestations of pancreatic duct leak
MANAGEMENT STRATEGIES
BOX 40.2 KEY POINTS
1. The diagnosis of external pancreatic stulas is usually
self-evident.
2. Pancreas protocol CT scan is most often the best way to
dene the consequences (uid collection, necrosis) of an
internal pancreatic stula.
3. Unless endotherapy can be done at the time of an ERCP,
Secretin-MRCP (S-MRCP) may be a better test to dene the
location or persistence of an internal pancreatic stula.
Diagnosis
The management of pancreatic stulas presupposes their diagnosis.
The diagnosis of external stulas is usually self-evident and
consists of variable output of clear pancreatic juice following
percutaneous drainage of a pseudocyst or peripancreatic uid
collection (Table 40.3).1 Alternatively, persistent output from a JP
drain following pancreatic resection, decompression, or peripancre-
atic surgery (e.g. splenectomy, left nephrectomy or gastrectomy) is
not usually a subtle manifestation of an external leak. More trouble-
some, however, may be the patient who sustains a penetrating
abdominal surgery such as a knife or gunshot wound in whom the
external stula is overlooked because of concerns of more signicant
injury.
The diagnosis of internal stulas is outlined in Table 40.3. In
essence, non-invasive imaging, particularly pancreas protocol CT,
remains the best initial diagnostic test in patients with smoldering
or severe pancreatitis or those patients with underlying chronic pan-
creatitis and an acute exacerbation of symptoms.13 Not only will CT
dene the consequences of pancreatitis (uid collections, necrosis,
effusions, ascites . . .)24 but can also be used to dene the potential
etiology (e.g. stones or strictures) as well as following the subsequent
evolution of pancreatitis. CT remains an imperfect tool, however, in
that biliary stone disease is underestimated, the uid component
associated with evolving pancreatic necrosis is overestimated, and
leaks are implied rather than dened.25 Further conrmation of a
ductal disruption may require sequential scans demonstrating an
External Fistulas
High amylase output through a surgically or percutaneously
placed drain
demonstrable pancreatogram through JP drain
Internal Fistula
Pseudocyst/evolving pancreatic necrosis
CT
MRI/MRCP
US/
EUS
ERCP
Pancreatic Ascites
High amylase uid with aspiration
Flat lm (ground glass appearance)
CT/MRCP
ERCP } Concomitant pseudocyst 1/31/2
Ductal disruption vs obstruction/upstream leak
High Amylase Pleural Effusion
ERCP
S-MRCP
Table 40.3 Diagnostic studies in pancreatic duct leaks
JP = Jackson-Pratt; CT = computerized tomography; MRI = magnetic resonance
imaging; S-MRCP = secretin magnetic resonance cholangiopancreatography; US =
ultrasound; EUS = endoscopic ultrasound; ERCP = endoscopic retrograde
cholangiopancreatography.
enlarging uid collection, aspiration of that uid collection with
measurement of amylase or lipase, an ERCP demonstrating pres-
ence and location of the leak, or S-MRCP. The latter study has been
shown by Deviere and his colleagues to be predictive of ongoing
ductal disruption and clearly minimizes potential ERCP complica-
tions such as exacerbation of pancreatitis and iatrogenic infection of
an undrained uid collection.26 It may also demonstrate patients
with a complete ductal disruption and a disconnected gland syn-
drome who may be better managed with non-endoscopic forms of
therapy. Finally, use of S-MRCP prior to ERCP may help in dening
subsequent endoscopic management comparable to its utilization in
hilar neoplasms of the liver. ERCP, in turn, is usually denitive
in showing not only the site of the ductal disruption (if persistent)
but also the proximate cause or reason for persistence (PD stones,
inammatory or brotic structure).13,27 For the most part, however,
ERCP adds an unnecessary risk to the care of acutely or chronically
ill patients with presumptive leak, unless endoscopic percutaneous,
or surgical therapy is contemplated.28
MANAGEMENT
BOX 40.3 KEY POINTS
1. The appropriate management of pancreatic stulas should
include a multidisciplinary team.
2. Transpapillary stent placement has a markedly higher
success rate in treating internal stulas if the disruption is
bridged.
421
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
The presence of a presumptive pancreatic stula is not, per se, a
demand to undertake endotherapy. Important considerations include
whether the patient has acute or underlying chronic pancreatitis,
whether pancreatic necrosis is present or absent, whether there is
superinfection of a uid collection, and whether the leak is con-
trolled or uncontrolled. For instance, the vast majority of low-volume
leaks following pancreatic resection are low grade, controlled by a
surgically placed JP drain, and spontaneously close with or without
concomitant octreotide over days or several weeks.29,30 On the other
hand, a patient may have rapidly increasing ascites or pleural effu-
sion or concomitant jaundice or cholestasis that demand urgent
attention.
From a personal perspective, I use the following indications to
dene the necessity of approaching a patient with pancreatic duct
leak endoscopically:
1. an enlarging pancreatic uid collection (pseudocyst, pancreatic
ascites, high amylase pleural effusion) despite conservative
management;
2. a symptomatic uid collection;
3. persistence of an external stula; and
4. inability to refeed a patient without developing recurring pain or
pancreatitis.13
A fth indication may be the question of concomitant biliary tract
disease. While the latter may occasionally be a concern for a retained
stone in the setting of biliary pancreatitis, it is more commonly seen
in patients with jaundice or cholangitis from pancreatic head edema
or pseudocyst.
Perhaps as important as indications for study are contraindica-
tions. The foremost is inability to render therapy if a ductal disrup-
tion is demonstrated. As such, diagnosis of a leak may result in
iatrogenic infection of necrosis or pseudocyst as the consequences
of internal stulization. The latter may require endoscopic, percuta-
neous, or even surgical drainage.
The endoscopic and non-endoscopic management of pancreatic
pseudocysts3135 and evolving pancreatic necrosis3641 are covered by
Dr Baron in Chapter 45. Sufce it to say that therapy usually requires
treatment of the underlying ductal disruption, if anatomically feasi-
ble, as well as the consequences of that disruption. Thus, surgical,
percutaneous, and endoscopic decompression of uid collections
have all been variously described.
Likewise, treatment of other types of internal and external pan-
creatic stulas does not occur in a vacuum and a multidisciplinary
approach between an endoscopist, pancreaticobiliary surgeon, and
interventional radiologist is often required to diagnose and properly
treat these complex and acutely ill patients.
Pancreatic ascites and high amylase
pleural effusions
Historically, pancreatic ascites and pleural effusions were treated
with gut rest and total parenteral nutrition (TPN) to minimize pan-
creatic juice stimulation. Diuretics, large volume thoracentesis and
paracentesis, and octreotide have all been used for weeks or months
in an attempt to preclude need for surgical resection or bypass. Suc-
cessful in <50% of these patients, salvage type surgery, usually
dened by ERCP preoperatively, consisted of partial pancreatectomy
or Roux-en-Y cyst-jejunostomy in the subset of patients with con-
comitant pancreatic pseudocysts. These attempts were associated
with high morbidity, an 815% peri procedural mortality, and recur-
rence of 1520%.13
422
Our group initially published a small series in which transpapil-
lary stent placement beyond the site of ductal disruption, in con-
junction with large volume paracentesis, was successful in treating
patients with pancreatic ascites42 (Fig. 40.2.) This therapy was deriv-
ative and a consequence of previous experience using transpapillary
stents with or without concomitant cyst-gastrostomy or cyst-duode-
nostomy in patients with pseudocysts. Since our initial publication,
a number of other authors including Bracher et al. have conrmed
our ndings.43 Combining the two series, over 90% of the patients
resolved their ascites without complication, and there were no
recurrences in the two series at 5 years and 14 months, respectively.
Transpapillary stenting appears to work by changing the ductal
drainage gradient and making the duodenum the area with the
least resistance to ow. Potential areas of downstream obstruction
that are bypassed include the sphincter, possible stones, and the
inammatory or brotic stricture frequently associated with a leak
(Fig. 40.3). Transpapillary stenting will not work in the setting of a
disconnected gland syndrome in which the bulk of the pancreatic
juice that enters the thoracic or abdominal cavity comes from a
disconnected pancreatic duct tail. This situation is also noted in
other forms of internal as well as external pancreatic stulas, all of
which are better handled with resective surgery unless there is con-
comitant portal vein thrombosis and prohibitive surgical risk (Figs
40.4, 40.5).13 A nice review of the endoscopic and surgical treat-
ments of pancreatic ascites has recently been published by Gomez-
Cerezo et al.44
An additional retrospective series by Telford et al. reported 43
patients with pancreatic duct disruption and a variety of clinical
manifestations.45 The etiology was acute pancreatitis in 24, chronic
pancreatitis in 9, operative injury in 7, and trauma in 3 patients.
Stent placement was successful in resolution of the disruption in 25,
unsuccessful in 16, and indeterminate in 2 patients. On univariate
analyses, bridging of the ductal disruption and duration of stenting
were associated with a statistically signicant successful outcome
whereas female gender and acute pancreatitis were negative predic-
tive factors of success. With multivariant analysis, only bridging of
the disruption remained statistically signicant as a dener of
success (Figs 40.6, 40.7, 40.8).
Pancreaticoenteric stulae and acute
pancreatic trauma
To date, our group has treated over 30 patients with pancreaticoen-
teric or biliary stulae. Although these patients may present with
spontaneous and rapid resolution of a uid collection for which no
treatment is required, a stenosis at the site of ductal disruption may
result in relapsing attacks of pancreatitis. Alternatively, stulization
into the bile duct or splenic exure (Fig. 40.5) of the colon may
result in cholestasis or cholangitis, or recurrent sepsis, respectively.
In our initial series of 8 patients with pancreaticoenteric stulae, 3
healed their stulas after downsizing or removal of an external
drain that had eroded into a contiguous loop of bowel, 3 healed
with transpapillary stent placement, and 2 patients ultimately
required surgical resection.46 Fistulization into the bile duct, in
turn, is almost invariably treated successfully by concomitant biliary
and pancreatic duct stenting assuming that the stula is not from
the upstream disconnected portion of the pancreas47 (Figs 40.9 and
40.10).
In addition to pancreaticoenteric or biliary stulae that usually
occur in the setting of pancreatic necrosis or chronic pancreatitis,
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions

A B

D
C

Fig. 40.2 Abdominal CT demonstrates bilateral pleural effusions

A and ascites B,C in patient with pancreatic ascites. ERCP demon-
strates site of disruption D treated with PD stent E. Ascites was
tapped and did not reaccumulate. Stent was retrieved after 6
weeks.

423
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
D
Fig. 40.3 ERCP demonstrates ductal disruption in pancreatic head A in patient with huge, high amylase effusion of right lung B. Patient
treated with transpapillary pigtail stent into uid collection C as well as stenting of minor papilla D,E to decompress upstream pancreatic
duct.
ERCP has also been used to treat internal stulae as a consequence
of acute pancreatic trauma. By way of example, Kim et al. diagnosed
normal pancreatograms in 14 of 23 patients with acute abdominal
trauma.18 Eight of these patients had a leak from the main PD into
the parenchyma and resolved spontaneously where 3 had an MPD
leak that could be bridged and responded to transpapillary stenting.
Although early ERCP was felt to be advantageous to project the need
for medical, surgical, or endoscopic therapy, it is possible that S-
MRCP may evolve to play a diagnostic role, selecting patients who
have the greatest potential for benet from therapeutic ERCP.17
External stulas
As previously discussed, with the exception of penetrating abdomi-
nal trauma, the vast majority of external stulas are iatrogenic. They
may occasionally follow partial pancreatic resection or bypass in the
setting of a downstream stricture. Most, however, are a consequence
of disconnected gland following percutaneous or surgical drainage
of a pancreatic uid collection.13 Our group initially reported a series
of patients undergoing transpapillary stenting for amenable external
424
C
F
stulas almost a decade ago.48 Since that time, multiple additional
series have been published or abstracted.13,49,50 By way of summary
of the available series, 86% of patients (50/58) could be successfully
stented and 46 of the latter (92%) had resolution of their stulas.
Procedural complications were limited to mild are of pancreatitis
in several series although there were two deaths in the series by
Costamagna et al., neither related to the stula or its endoscopic
treatment. No recurrences were reported in patients undergoing
successful stula closure at follow-up ranging from 12 to 36
months.
In our practice, endotherapy was initially reserved for postopera-
tive or percutaneously drained patients whose external stula did not
respond to several weeks of clear liquids, total parenteral nutrition,
and octreotide. Currently, we have become considerably more
aggressive, studying patients with high-volume stulas with S-
MRCP if they have marginal change in stula volume after several
days of somatostatin analogue, and undertake ERCP and transpapil-
lary stent placement unless imaging documents a disconnected
gland syndrome (Figs 40.4, 40.5). Historically, the latter patients
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions

A B C D

E F H

I J K

Fig. 40.4 JP drain demonstrates disconnected pancreatic duct A in patient with a history of pancreatic necrosis, portal hypertension. ERCP
demonstrates cut off main pancreatic duct below the genu B. Transduodenal contrast injection lls residual pancreatic duct through small
upstream pseudocyst C,D,E. Following needle-knife stulization through the bulb F,G,H, parallel 7 Fr pigtail and straight stents were placed
I,J,K. Fistula closed within 24 hours, allowing JP drain removal 3 years after original placement.

have usually required distal pancreatectomy,13 although our inter-
ventional radiologists have treated a subset of these patients with
cyanoacrylate injection. The latter therapy requires guidewire place-
ment to the pancreatic duct tail through the stulous tract, place-
ment of a microcatheter over the wire, and injection of the entire
disconnected portion of the duct to include side branches. Mild post-
procedural pancreatitis has been noted in approximately 50% of
patients and recurrent stulas may occur unless the entire duct and
its side branches are sealed. This procedure works best with 34 cm
of disconnected gland and is less likely to be successful when the
glandular disconnection is at the genu requiring the major portion
of the gland to be glued shut.
In addition to the percutaneous approach to the disconnected
gland, as well as the surgical approach using glue injection to mini-
mize post-pancreatectomy leak,51 Soehendras group has utilized
this technique in internal stulas by transpapillary injection of
methyl-butyl cyanoacrylate into the distal duct at the site of glandular
disruption.52 Eight of 11 patients in their series healed their disrup-
tion without recurrence although all had concomitant pancreatic
duct stents or drains as well as endoscopic drainage of associated
uid collections. This group has also used glue injection as an
adjunct in patients with severe pancreatic necrosis who are undergo-
ing aggressive endoscopic drainage using EUS-directed lavage or
debridement.53

425
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B

C D

Fig. 40.5 ERCP demonstrates transgastric pigtail stent (arrows) A,B, in patient with severe necrotizing pancreatitis, portal vein thrombosis,
and disconnected pancreatic duct (arrow) C. Note embolized coils for previous splenic artery aneurysm C (arrows), JP drain for persistent
colonic stula into pancreatic head D.

426
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions

A B C

D E F

Fig. 40.6 CT demonstrates perisplenic uid collection A and markedly thickened gastric
wall (arrow) in patient with hereditary pancreatitis. ERCP demonstrates severe chronic pan-
creatitis B with dilated tail and ductal disruption C requiring percutaneous drainage. Stricture
is dilated with Soehendra stent extractor D followed by balloon dilation E and 7 Fr stent
placement beyond ductal disruption F. Note improvement in CT scan, persistent dilation of
the pancreatic duct tail G.

427
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B

C D

Fig. 40.7 ERCP demonstrates PD disruption at junction of body and tail A,B in patient with severe LUQ/ank pain, splenic vein
thrombosis. Pain, ductal disruption resolved with prosthesis placement C,D. Disruption recurred 1 year later, requiring distal
pancreatectomy/splenectomy.

428
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions

A B C D

E F G H

Fig. 40.8 Patient with chronic pancreatitis, pancreas divisum, and ductal disruption in pancreatic head who presented with jaundice,
intractable pain, and 50 lb weight loss. Note pseudocyst in head A, leak from VPD B, which was stented C, dorsal papillary sphincterotomy
and, 2 dorsal pancreatic duct stents into uid collection in the head D,E. Note stent retrieval progressive pseudocyst resolution and recon-
stitution of the DPD following stent exchange after 6 weeks F,G,H.

It is the authors opinion that endoscopic glue injection into

the pancreatic duct needs further critical evaluation before
widespread adoption and its application must be weighed against
the risks of long-term, indwelling drain placement or surgical
resection.

COMPLICATIONS

BOX 40.4 KEY POINTS

1. The risk of procedural or post-procedural complications in

treating PD stulas endoscopically should be weighed
against the risk of persistent stula or alternative treatments.

2. Procedural pancreatitis and iatrogenic infection of a

concomitant uid collection are the major risks in
endoscopically treating internal PD stulae.

Fig. 40.9 Arrow demonstrates stula between CBD (stented) and 3. Iatrogenic ductitis to include irreversible stricture formation
PD. Fistula resolved at 4 weeks following pancreaticobiliary stent can follow PD stenting.
placement.

429
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A
C
B

D E F

Fig. 40.10 High grade bile duct stricture in patient with pancreatic necrosis, intramural pancreatic ductal disruption. Note contrast in
colon A, deformed edematous papilla B, biliary stent placement C. Small arrows demonstrate intraduodenal abscess and large arrow,
dilated pancreatic duct D. Note dual stent placement E,F which resolved jaundice, concomitant PD disruption and obstruction.

Immediate complications sphincter.55 Pancreatitis is more common in the setting of a poorly

The immediate complications of transpapillary stent placement are
those of diagnostic ERCP and include drug reaction, aspiration,
cardiopulmonary events, pancreatitis from contrast injection or
sphincter manipulation, and cholangitis if there is a concomitant
biliary stenosis that is not endoscopically treated. Bleeding and
iatrogenic perforation can occasionally be seen if sphincterotomy is
done to facilitate stent placement.54 Pancreatitis are, in my experi-
ence, approximates 10% in normal ducts and is uncommon in
patients who already have ductal changes of chronic pancreatitis. It
may approach 50% in the setting of unsuccessful stenting when
multiple accessories and guidewires have been placed into the PD
to facilitate bridging the area of leak. This pancreatitis is usually
attenuated, however, if a short transpapillary stent is left to preclude
ductal obstruction by an edematous papilla or traumatized
chosen prosthesis even if the disruption can be bridged. As such,
placement of a 7 Fr stent into a 5 Fr diameter duct should be
discouraged. Likewise, placement of a 12 cm prosthesis to bridge
a ductal leak 3 cm from the papilla is inappropriate.
Subacute complications
Subacute complications are usually infectious and result from iatro-
genic introduction of bacteria into a uid collection or necrotic
debris at the time of ERCP. As such, all patients with a presumptive
internal stula deserve antibiotic coverage with a broad spectrum
antibiotic prior to a diagnostic ERCP and may require more pro-
longed treatment afterwards, particularly in the setting of necrosis.
Moreover, clearly contaminated uid collections should be consid-
ered for concomitant endoscopic or percutaneous drainage of necro-

430
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions
sis as described by Dr. Baron later in this text. Note that our group
has previously demonstrated that bacterial contamination within the
pancreatic duct is invariable in patients with indwelling stents and
that stent occlusion is a necessary, but not sufcient, cause of pan-
creatic sepsis.56 Stent occlusion may also be associated with obstruc-
tive pancreatitis and it is for this reason as well as fear of iatrogenic
duct injury,57 that indwelling stents should be retrieved within
1 week of external stula closure and after 46 weeks of treating
an internal PD stula.13
Chronic complications
Although iatrogenic ductal injury is listed under chronic complica-
tions, prostheses never belong in the pancreas, particularly in those
with otherwise normal pancreatic ducts. A number of procedural
and stent modications have decreased trauma to the major pancre-
atic duct and minimized side branch occlusion over the past several
years. These modications include utilization of 34 Fr diameter
stents, elimination of internal stent anges, and recognition that
stents which apply signicant pressure proximally may cause duct
ulceration and subsequent brosis.57,58 Despite this, 3 Fr, unanged,
pigtail stents almost always spontaneously migrate within a week or
two and are probably appropriate only in the patient in whom bridg-
ing of the disruption was unsuccessful, and then only to prevent or
ameliorate post-ERCP pancreatitis. Iatrogenic ductitis should be
anticipated and minimized by selecting a prosthesis with a smaller
diameter than the PD downstream from the leak, one that is the
appropriate length to bridge the disruption without an inordinate
stent length beyond the site of leak, and one that avoids upstream
impaction or angulation on the ductal wall.
REFERENCES
1. Kozarek R. Role of ERCP in acute pancreatitis. Gastrointest Endosc.
2002; 56:S231S236.
2. Mariani A. Is secretin magnetic resonance cholangio-
pancreatography an effective guide for a diagnostic and/or
therapeutic ow-chart in acute recurrent pancreatitis? JOP. 2001;
2:414421.
3. Enns R, Baillie J. Review article: the treatment of acute biliary
pancreatitis. Aliment Pharmacol Ther. 1999; 1:13791389.
4. Frakes JT. Biliary pancreatitis: a review. Emphasizing appropriate
endoscopic intervention. J Clin Gastroenterol. 1999; 28:97109.
5. Kozarek RA, Attia FM, Traverso LW, et al. Pancreatic duct leak in
necrotizing pancreatitis. Role of diagnostic and therapeutic ERCP
as part of a multi-disciplinary approach. J Gastrointest Endosc.
2000; 51:AB138.
6. Delhaye M, Matos C, Deviere J. Endoscopic technique for the
management of pancreatitis and its complications. Best Pract Res
Clin Gastroenterol. 2004; 18:155181.
7. Varadarajulu S, Noone T, Hawes RH, et al. Pancreatic duct stent
insertion for functional smoldering pancreatitis. Gastrointest
Endosc. 2003; 58:438441.
SUMMARY
1. Pancreatic duct leaks are the consequence of acute
inammation with duct disruption, downstream obstruction,
or both.
2. Minor leaks in the setting of acute pancreatitis/necrosis
are probably common and respond to conservative
therapy.
3. The consequences of major ductal disruptions include
internal stulas (pseudocysts, necrosis, pancreatic ascites,
pancreatic pleural effusions, pancreaticoenteric or biliary
communication), and external stulas.
4. Treatment of internal stulas requires treatment of the leak
and/or the sequelae/consequences of the leak.
5. Bridging the site of ductal disruption with a transpapillary
prosthesis is more likely to result in resolution of the
disruption unless there is a disconnected gland
syndrome.
6. External pancreatic stulas that are the consequence of a
disconnected pancreatic duct may ultimately close if low
volume but are traditionally treated with resection of the
disconnected gland.
8. Kozarek RA, Traverso LW. Pancreatic stulas and ascites. In: Brandt
JL (ed.) Textbook of clinical gastroenterology. Philadelphia:
Current Medicine 1998:11751181.
9. Chebli JM, Gaburri PD, de Souza AF, et al. Internal
pancreatic stulas: proposal of a management algorithm
based on a case series analysis. J Clin Gastroenterol. 2004;
38:795800.
10. Lau ST, Simchuk EJ, Kozarek RA, et al. A pancreatic ductal leak
should be sought to direct treatment in patients with acute
pancreatitis. Am J Surg. 2001; 181:411415.
11. Uomo G, Molino D, Visconti M, et al. The incidence of main
pancreatic duct disruption in severe biliary pancreatitis. Am J Surg.
1998; 176:4952.
12. Andrn-Sandberg , Dervenis C. Pancreatic pseudocysts in the
21st century. Part I: classication, pathophysiology, anatomic
considerations and treatment. JOP. 2004; 5:824.
13. Kozarek RA. Pancreatic duct leaks and pseudocysts. In: Ginsberg
G, Kochman M, Norton I, et al. (eds) Clinical gastrointestinal
endoscopy; Textbook with DVD. Philadelphia: Saunders,
2005:807820.
431
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
14. Poon RT, Lo SH, Fong D, et al. Prevention of pancreatic
anastomotic leakage after pancreaticoduodenectomy. Am J Surg.
2002; 183:4252.
15. Sheehan MK, Beck K, Creech S, et al. Distal pancreatectomy: does
the method of closure inuence stula formation? Am Surg.
2002; 68:264267.
16. Freeny PC, Hauptmann E, Althaus SJ, et al. Percutaneous CT-
guided catheter drainage of infected acute necrotizing
pancreatitis: techniques and results. AJR Am J Roentgenol. 1998;
170:969975.
17. Memis A, Parildar M. Interventional radiological treatment
in complications of pancreatitis. Eur J Radiol. 2002; 43:
219228.
18. Kim HS, Lee DK, Kim IW, et al. The role of endoscopic retrograde
pancreatography in the treatment of traumatic pancreatic duct
injury. Gastrointest Endosc. 2001; 54:4955.
19. Oksuz MO, Altehoefer C, Winterer JT, et al. Pancreatico-
mediastinal stula with a mediastinal mass lesion
demonstrated by MR imaging. J Magn Reson Imaging. 2002;
16:746750.
20. Sakorafas GH, Sarr MG, Farnell MB. Pancreaticobiliary stula: an
unusual complication of necrotising pancreatitis. Eur J Surg. 2001;
167:151153.
21. De Backer AI, Mortele KJ, Vaneerdeweg W, et al.
Pancreatocolonic stula due to severe acute pancreatitis: imaging
ndings. JBR-BTR. 2001; 84:4547.
22. Salih A. Massive pleural effusion. Postgrad Med J. 2001; 77:536,
546536, 547.
23. Ito H, Matsubara N, Sakai T, et al. Two cases of thoracopancreatic
stula in alcoholic pancreatitis: clinical and CT ndings. Radiat Med.
2002; 20:207211.
24. Balthazar EJ, Robinson DL, Megibow AJ, et al. Acute pancreatitis:
value of CT in establishing prognosis. Radiology. 1990;
174:331336.
25. Baron TH, Morgan DE. Acute necrotizing pancreatitis. N Engl J
Med. 1999; 340:14121417.
26. Matos C, Bali MA, Delhaye M, et al. Magnetic resonance imaging
in the detection of pancreatitis and pancreatic neoplasms. Best
Pract Res Clin Gastroenterol. 2006; 20:157178.
27. Kozarek RA. Endoscopic therapy of complete and partial
pancreatic duct disruptions. Gastrointest Endosc Clin N Am. 1998;
8:3953.
28. Szentes MJ, Traverso LW, Kozarek RA, et al. Invasive treatment of
pancreatic uid collections with surgical and nonsurgical
methods. Am J Surg. 1991; 161:600605.
29. Kaman L, Behera A, Singh R, et al. Internal pancreatic stulas with
pancreatic ascites and pancreatic pleural effusions: recognition
and management. ANZ J Surg. 2001; 71:221225.
30. Li-Ling J, Irving M. Somatostatin and octreotide in the
prevention of postoperative pancreatic complications and the
treatment of enterocutaneous pancreatic stulas: a systematic
review of randomized controlled trials. Br J Surg. 2001;
88:190199.
31. Pitchumoni CS, Agarwal N. Pancreatic pseudocysts. When and
how should drainage be performed? Gastroenterol Clin North
Am. 1999; 28:615639.
32. Kozarek RA, Brayko CM, Harlan J, et al. Endoscopic drainage
of pancreatic pseudocysts. Gastrointest Endosc. 1985;
31:322327.
33. De Palma GD, Galloro G, Puzziello A, et al. Endoscopic drainage
of pancreatic pseudocysts: a long-term follow-up study of 49
patients. Hepatogastroenterology. 2002; 49:11131115.
34. Sanchez Cortes E, Maalak A, Le Moine O, et al. Endoscopic
cystenterostomy of nonbulging pancreatic uid collections.
Gastrointest Endosc. 2002; 56:380386.
432
35. Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic transpapillary
therapy for disrupted pancreatic duct and peripancreatic uid
collections. Gastroenterology. 1991; 100:13621370.
36. Seifert H, Wehrmann T, Schmitt T, et al. Retroperitoneal
endoscopic debridement for infected peripancreatic necrosis.
Lancet. 2000; 356:653655.
37. Takeda K, Matsuno S, Sunamura M, et al. Surgical aspects and
management of acute necrotizing pancreatitis: recent results
of a cooperative national survey in Japan. Pancreas. 1998;
16:316322.
38. Bchler P, Reber HA. Surgical approach in patients with acute
pancreatitis. Is infected or sterile necrosis an indicationin whom
should this be done, when, and why? Gastroenterol Clin North
Am. 1999; 28:661671.
39. Bchler MW, Gloor B, Muller CA, et al. Acute necrotizing
pancreatitis: treatment strategy according to the status of
infection. Ann Surg. 2000; 232:619626.
40. Baron TH, Thaggard WG, Morgan DE, et al. Endoscopic therapy
for organized pancreatic necrosis. Gastroenterology. 1996;
111:755764.
41. Baron TH, Harewood GC, Morgan DE, et al. Outcome differences
after endoscopic drainage of pancreatic necrosis, acute pancreatic
pseudocysts, and chronic pancreatic pseudocysts. Gastrointest
Endosc. 2002; 56:717.
42. Kozarek RA, Jiranek GC, Traverso LW. Endoscopic treatment of
pancreatic ascites. Am J Surg. 1994; 168:223226.
43. Bracher GA, Manocha AP, DeBanto JR, et al. Endoscopic pancreatic
duct stenting to treat pancreatic ascites. Gastrointest Endosc. 1999;
49:710715.
44. Gomez-Cerezo J, Barbado CA, Suarez I, et al. Pancreatic ascites:
study of therapeutic options by analysis of case reports and case
series between the years 1975 and 2000. Am J Gastroenterol.
2003; 98:568577.
45. Telford JJ, Farrell JJ, Saltzman JR, et al. Pancreatic stent placement
for duct disruption. Gastrointest Endosc. 2002; 56:1824.
46. Wolfsen HC, Kozarek RA, Ball TJ, et al. Pancreaticoenteric stula:
no longer a surgical disease? J Clin Gastroenterol. 1992;
14:117121.
47. Carrere C, Heyries L, Barthet M, et al. Biliopancreatic stulas
complicating pancreatic pseudocysts: a report of three cases
demonstrated by endoscopic retrograde
cholangiopancreatography. Endoscopy. 2001; 33:9194.
48. Kozarek RA, Ball TJ, Patterson DJ, et al. Transpapillary Stenting
for Pancreaticocutaneous Fistulas. J Gastrointest Surg. 1997;
1:357361.
49. Costamagna G, Mutignani M, Ingrosso M, et al. Endoscopic
treatment of postsurgical external pancreatic stulas. Endoscopy.
2001; 33:317322.
50. Saeed ZA, Ramirez FC, Hepps KS. Endoscopic stent placement for
internal and external pancreatic stulas. Gastroenterology. 1993;
105:12131217.
51. Suc B, Msika S, Fingerhut A, et al. Temporary brin glue occlusion
of the main pancreatic duct in the prevention of intra-abdominal
complications after pancreatic resection: prospective randomized
trial. Ann Surg. 2003; 237:5765.
52. Seewald S, Brand B, Groth S, et al. Endoscopic sealing of
pancreatic stula by using N-butyl-2-cyanoacrylate. Gastrointest
Endosc. 2004; 59:463470.
53. Seewald S, Groth S, Omar S, et al. Aggressive endoscopic therapy
for pancreatic necrosis and pancreatic abscess: a new safe and
effective treatment algorithm. Gastrointest Endosc. 2005;
62:92100.
54. Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic pancreatic duct
sphincterotomy: indications, technique, and analysis of results.
Gastrointest Endosc. 1994; 40:592598.
 Chapter 40 Pancreatic Interventions in Acute Pancreatitis: Ascites, Fistulae, Leaks and Other Disruptions

55. Fazel A, Quadri A, Catalano MF, et al. Does a pancreatic duct stent
prevent post-ERCP pancreatitis? A prospective randomized study.
Gastrointest Endosc. 2003; 57:291294.
56. Kozarek R, Hovde O, Attia F, et al. Do pancreatic duct stents cause
or prevent pancreatic sepsis? Gastrointest Endosc. 2003;
58:505509.
57. Smith MT, Sherman S, Ikenberry SO, et al. Alterations in
pancreatic ductal morphology following polyethylene
pancreatic stent therapy. Gastrointest Endosc. 1996; 44:
268275.
58. Raijman I. Biliary and pancreatic stents. Gastrointest Endosc Clin N
Am. 2003; 13:561592.

433
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Idiopathic Acute Pancreatitis: Role of
41 ERCP in Diagnosis and Therapy
Stuart Sherman
Determining the cause of acute pancreatitis is not usually difcult.
Alcohol and gallstones are the two most common etiologies and
account for 6090% of the cases (Table 41.1). Alcoholism is diag-
nosed by history, and gallstones, by a combination of demographic
characteristics, laboratory ndings, and radiographic imaging
studies. In patients in whom acute pancreatitis is due to hypertri-
glyceridemia, hypercalcemia, drug reactions, trauma, surgery, ERCP,
etc, the relationship of the episode of pancreatitis to the cause is
usually clear. Nevertheless, a cause for acute pancreatitis will not be
identied in 1030% of patients after a careful history, physical
examination, laboratory testing, and radiological evaluation.1 These
patients are conventionally classied as having idiopathic acute pan-
creatitis (IAP).2,3 Patients with recurrent episodes of IAP are diag-
nosed with idiopathic acute recurrent pancreatitis (IARP). This
chapter will review the role of ERCP and ancillary endoscopic tech-
niques in the evaluation and therapy of patients with IAP and IARP.
The reader should be aware that signicant controversy exists as to
the appropriate use of ERCP in IAP because the available evidence
is often sparse and largely uncontrolled.4
RATIONALE FOR AN ENDOSCOPIC APPROACH
The literature is now replete with evidence to suggest that pancreatic
duct obstruction can lead to acute pancreatitis. The pathophysiologic
mechanisms are based on two key points: (1) pancreatic ductal
obstruction leads to ductal hypertension that is exacerbated by pan-
creatic secretion and (2) Ductal hypertension causes inhibition of
enzyme secretion resulting in colocalization of inactive pancreatic
enzymes and lysosomal hydrolases with subsequent acinar cell
injury and the clinical sequelae of acute pancreatitis.5 Advanced
endoscopic evaluation of IAP and IARP focuses on identifying
causes of pancreatic ductal obstruction with the therapeutic goal of
alleviating the obstruction. It is assumed that relief of the obstruc-
tion will prevent further episodes of pancreatitis. The obstructive
theory of acute pancreatitis assumes that ductal obstruction is
intermittent or that a second risk factor predisposes patients with
impaired ductal drainage.6
DIAGNOSTIC FINDINGS AND TIMING OF ERCP
There are two major concerns that prompt the physician to do a
more intensive evaluation of the patient with acute pancreatitis in
whom no obvious cause is determined. The rst is that the patient
may have an underlying disease which will predispose to further
attacks of acute pancreatitis unless the cause is identied and ade-
quately treated. Acute pancreatitis is likely to recur in 3367% of
patients with biliary tract disease when not diagnosed and treated.7
Similarly, other anatomical or functional disorders of the pancreati-
cobiliary tree may predispose patients to recurrent episodes of acute
pancreatitis. The second concern is that the pancreatitis may be
related to a tumor. As a result, ERCP now plays a central role in the
evaluation and therapy of patients with IAP.
There are a number of potential causes of IAP that can be diag-
nosed by ERCP and ancillary techniques. These include occult gall-
stones, abnormalities and anomalies of the pancreatic duct and bile
duct, sphincter of Oddi dysfunction, and ampullary and pancreatic
neoplasms. The techniques applied at ERCP to diagnose the cause
of IAP are shown on Table 41.2.
Although there are potential gains by performing ERCP (identify-
ing and treating the cause and preventing another episode of acute
pancreatitis), there are potential downsides for the patient and the
health care system as a whole (the inappropriate performance of the
procedure and its complications).8 The timing of ERCP in patients
with IAP is controversial. Ballinger and colleagues2 reported only
one of 27 patients with one unexplained episode of acute pancreatitis
and the gallbladder in situ had a second episode of pancreatitis
during a 3-year follow-up period. They felt that the risks of ERCP
were greater than the risks of a second episode of acute pancreatitis
and advised against its use. On the other hand, Trapnell and Duncan
reported that 35 of 148 patients (24%) with IAP suffered a recur-
rence, but that if gallstones were present, the rate increased to 38%.9
Using an actuarial method, the authors found that 10% of patients
with IAP were likely to have a rst recurrence within one year of the
initial attack, 17% within two years and 25% within six years. Patients
who had one recurrence were likely to have a second. In a cost-utility
analysis, Gregor and colleagues found that performing an ERCP on
all patients after a rst episode of idiopathic pancreatitis was neither
of great benet nor particularly costly.8 However, it is of substantial
benet and cost-effective in a subgroup of patients with greater prob-
ability of having an occult gallstone.
Our approach is (usually) to evaluate patients 40 years and older
by ERCP (usually with sphincter of Oddi manometry and bile
microscopy) after their rst episode. This is based on the ndings
of Choudari et al.10 In this study, 21% of patients aged 4060 years
and 25% of patients older than 60 years had a neoplastic process as
the cause for their pancreatitis in contrast to only 3% younger than
40 years (Table 41.3). Because of the low incidence of a neoplastic
process, patients younger than 40 years undergo ERCP after their
second episode unless the rst episode was considered severe. In a
series of 1108 patients with idiopathic pancreatitis, Fischer and
colleagues reported a progressive increase in the incidence of
mucinous tumors (as a cause for the pancreatitis) with age ranging
from 1.3% in patients <25 years to 11.4% in patients older than
70 years.11
435
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Alcohol Idiopathic <20 yr 2040 yr 4060 yr >60 yr

Autoimmune pancreatitis Infection Diagnosis n = 15 n = 53 n = 95 n = 62
Biliary calculous disease Bacterial Pancreatic cancer 0% 0% 2% 2%
Macrolithiasis (bile duct stone) Parasites/worms Ampullary Ca/Adenoma 0% 2% 2% 0%
Microlithiasis (biliary crystals) Viral Mucinous tumor 0% 2% 17% 23%
Biliary cystic disease Metbolic SOD 47% 43% 35% 26%
Choledochal cyst Hypercalcemia Pancreas divisum 13% 15% 19% 23%
Choledochocele/duplication Hyperlipidemia Chronic pancreatitis 27% 11% 9% 13%
cyst Inborn errors of metabolism Miscellaneous 7% 9% 9% 3%
Congenital anomaly Neoplasm Normal 7% 21% 6% 11%
Annular pancreas Duodenal
Anomalous pancreaticobiliary Ampullary
Table 41.3 Idiopathic acute pancreatitis: yield of ERCP sphincter
junction Pancreatic
of Oddi manometry bile microscopy correlated with age (n =
Pancreas divisum Biliary
225)
Chronic pancreatitis Renal disease SOD = Sphincter of Oddi dysfunction; Ca = cancer.
Duodenal obstruction Chronic renal failure From reference 10.
Afferent limb obstructed Dialysis related
(Billroth II) Sphincter of Oddi dysfunction
Atresia Toxin
Crohns disease Organophosphate ultrasound may help identify patients with underlying microlithiasis
Diverticulum insecticides when conventional transcutaneous ultrasound is normal.1315
Drugs Scorpion bite Microlithiasis has been implicated as a common cause of IAP.
Genetic Trauma
Two prospective studies found that approximately two-thirds to
Alpha 1-antitrypsin deciency Tropical
three-fourths of patients with IAP harbored occult gallstones in the
Cystic brosis Vasculitis
Hereditary pancreatitis Polyarteritis nodosa gallbladder.16,17 The diagnosis was based on microscopic examina-
Iatrogenic Systemic lupus tion of bile for crystals and usually conrmed on evaluation of the
ERCP erythematosus resected gallbladder or follow-up gallbladder ultrasound showing
Abdominal surgery gallstones and/or sludge. On multivariate analysis, the nding of
biliary crystals in bile is a strong predictor of small stones or sludge
Table 41.1 Etiologies of acute pancreatitis in the gallbladder (p < 0.001).17 Moreover, the nding of crystals in
bile had a sensitivity of 86%, specicity of 86% and a positive predic-
tive value of 92% for the diagnosis of gallstone disease as the missed
cause of IAP. In contrast to the results of Lee and colleagues, and
Screening endoscopy Ros and colleagues, some investigators have detected microlithiasis
Ampullary and mucinous tumor
in fewer than 10% of patients with IAP.1619 Bile may be collected at
Ductography
Bile duct stones
the time of ERCP from the duodenum or bile duct after gallbladder
Anomalies/abnormalities of the pancreatic duct and bile duct stimulation with cholecystokinin or by direct cannulation of the
Chronic pancreatitis gallbladder.
Tumors Treatment of microlithiasis can signicantly reduce the incidence
Sphincter of Oddi manometry of recurrent pancreatitis.16,17 There are several therapeutic options
Sphincter of Oddi dysfunction for managing patients with pancreatitis due to microlithiasis. Lapa-
Aspiration of bile for crystals roscopic cholecystectomy should be considered the procedure of
Microlithiasis choice as it is almost always curative. Ros and colleagues reported
no further episodes of pancreatitis in 17 of 18 patients followed for
Table 41.2 Techniques applied at ERCP to diagnose cause of 3 years after cholecystectomy.17 Endoscopic biliary sphincterotomy
idiopathic acute pancreatitis
is an excellent alternative for the elderly and high-risk surgical
patient.7 Dissolution therapy with ursodeoxycholic acid has also been
shown to prevent recurrent pancreatitis.17 However, maintenance
The ndings of ERCP and the outcomes of therapy for each therapy appears necessary to prevent recurrent stone formation.
disease identied will be discussed individually. Given the high prevalence of occult microlithiasis in some series,
some authorities advocate empiric cholecystectomy as a rst line of
OCCULT GALLSTONE DISEASE therapy, particularly in patients with recurrent attacks.20,21

Although microlithiasis and biliary sludge are technically different,
the terms are often used interchangeably. Microlithiasis most com-
monly refers to stones <3 mm in diameter and biliary sludge is
considered to be a suspension of crystals, mucin, glycoproteins, cel-
lular debris, and proteinaceous material within bile.12 The crystals
are made up of cholesterol monohydrate, calcium bilirubinate, or
calcium carbonate. Microlithiasis and biliary sludge may not be
detected by standard gallbladder imaging techniques. Endoscopic
SPHINCTER OF ODDI DYSFUNCTION
The sphincter of Oddi (SO) is a small complex of smooth muscles
surrounding the terminal common bile duct, main pancreatic duct,
and the common channel. It is important to appreciate that there
are separate biliary and pancreatic sphincters and that treatment
aimed at the biliary sphincter may leave a diseased pancreatic
sphincter intact (see below). The main function of the SO is to

436
 Chapter 41 Idiopathic Acute Pancreatitis: Role of ERCP in Diagnosis and Therapy

regulate bile and pancreatic exocrine juice ow and prevent reux
of duodenal contents into the ducts.
SO dysfunction (SOD) refers to an abnormality of SO contractility
that is manifested clinically by pancreaticobiliary pain, pancreatitis,
or deranged liver function tests. Sphincter of Oddi manometry
(SOM) is considered by most authorities to be the gold standard test
for diagnosing SOD.2226 SOM can be performed percutaneously or
intraoperatively, but is most commonly performed at the time of
ERCP. SOM utilizes a water-perfused catheter which is inserted into
the common bile duct, pancreatic duct, or both to measure sphincter
pressures. The diagnosis of SOD is established when the basal
sphincter pressure is equal to or greater than 40 mm Hg.24 Because
SOM is difcult to perform, invasive, not widely available, and asso-
ciated with a high complication rate, several non-invasive and pro-
vocative tests have been designed in an attempt to identify patients
with SOD. The currently available data suggest that these tests lack
the sensitivity and specicity to replace SOM.23,27 However, Mariani
and colleagues recently reported that secretin-stimulated magnetic
resonance cholangiopancreatography (s-MRCP) and SOM were con-
cordant in 13 of 15 patients (86.7%).28
SOD is a frequent cause of recurrent pancreatitis previously
labeled as IARP. It has been manometrically documented in 1572%
of such patients (Table 41.4).10,11,18,2933 Pancreatic sphincter manom-
etry should be done in patients with IARP, particularly those with
normal biliary manometry and in those who have recurrent attacks
after a biliary sphincterotomy. It is not surprising that isolated
pancreatic sphincter hypertension is common among patients with
IARP found to have SOD.6,34 Also pancreatic sphincter dysfunction
may explain recurrent pancreatitis despite biliary sphincterotomy or
surgical biliary sphincteroplasty.6
Sphincter ablation is the recommended therapy for patients with
recurrent pancreatitis due to SOD. Historically, this has been done
surgically.35,36 However, with increasing experience, endoscopic
sphincterotomy has become the treatment of choice.
The value of ERCP, SOM and sphincter ablation therapy was
studied in 51 patients with idiopathic pancreatitis.37 Twenty-four
patients (47.1%) had an elevated basal sphincter pressure. Thirty
were treated by biliary sphincterotomy (n = 20), or surgical sphinc-
teroplasty with septoplasty (n = 10). Fifteen of 18 patents (83%) with
an elevated basal sphincter pressure had long-term benet (mean
follow-up, 38 months) from sphincter ablation therapy (including 10
of 11 treated by biliary sphincterotomy) in contrast to only 4 of 12
(33.3%, p < 0.05) with a normal basal sphincter pressure (including
4 of 9 treated by biliary sphincterotomy). However, Guelrud and
colleagues found that severance of the pancreatic sphincter was
necessary to resolve the pancreatitis (Fig. 41.1).38
In this series, 69 patients with idiopathic pancreatitis due to SOD
underwent treatment by standard biliary sphincterotomy (n = 18),
biliary sphincterotomy with pancreatic sphincter balloon dilation (n
= 24), biliary sphincterotomy followed by pancreatic sphincterotomy
in separate sessions (n = 13), or combined pancreatic and biliary
sphincterotomy in the same session (n = 14). Eighty-one percent of
patients undergoing pancreatic and biliary sphincterotomy had reso-
lution of their pancreatitis compared to 28% of patients undergoing
biliary sphincterotomy alone (p < 0.005). Sherman and colleagues
reported that only 44% of SOD patients with IARP had no further
attacks during a 5-year follow-up interval after biliary sphincterot-
omy alone.32 These data are consistent with the theory that many
such patients who benet from biliary sphincterotomy alone may
have subtle gallstone pancreatitis or perhaps the follow-up has not
been long enough to detect another attack of pancreatitis. The results
of Guelrud and colleagues also support the anatomic ndings of
separate biliary and pancreatic sphincters, and the manometry nd-
ings of residual pancreatic sphincter hypertension in more than 50%
of persistently symptomatic patients who undergo biliary sphincter-
otomy alone.38 Kaw and Brodmerkel reported that among patients
with idiopathic pancreatitis secondary to SOD, 78% had persistent
manometric evidence of pancreatic sphincter hypertension despite
a biliary sphincterotomy.33 Toouli et al. also demonstrated the impor-
tance of pancreatic and biliary sphincter ablation in patients with
idiopathic pancreatitis.39
In this series, 23 of 26 patients (88%) undergoing surgical
ablation of both the biliary and pancreatic sphincter were either

Author (yr) Frequency

Toouli (1985) 16/26 (57%)
Guelrud (1986) 17/42 (40%)
Gregg (1989) 38/125 (30%)
Venu (1989) 17/116 (15%)
Sherman (1993) 18/55 (33%)
Choudari (1998) 79/225 (35%)
Kaw (2002) 67/126 (53%)
Fischer (2005) 445/1108 (40%)
Total 697/1823 (38%)
Fig. 41.1 Patient with idiopathic acute recurrent pancreatitis that
Table 41.4 Manometrically documented sphincter of Oddi reoccurred after biliary sphincterotomy. Top left shows pancreatic
dysfunction causing idiopathic acute recurrent pancreatitis manometry being performed. Top right, bottom left and bottom
From references 10, 11, 18, 2933. right show pancreatic sphincterotomy being performed.

437
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

asymptomatic or had minimal symptoms at a median follow-up of ducts, the major portion of the pancreatic exocrine juice drains into
24 months (range 9105 months). Okolo and colleagues retrospec- the duodenum via the dorsal duct and minor papilla. It has been
tively evaluated the long-term results of endoscopic pancreatic proposed that a relative obstruction to pancreatic exocrine juice ow
sphincterotomy in 55 patients with manometrically documented or through the minor papilla with resultant pancreatic duct hyperten-
presumed pancreatic sphincter hypertension (presumption based on sion could precipitate recurrent pancreatitis in a subpopulation
recurrent pancreatitis with pancreatic duct dilation and contrast of pancreas divisum patients.5052 Whereas a few epidemiological
medium drainage time from the pancreatic duct greater than 10 studies dispute the relation of pancreas divisum and pancreatitis,
minutes).40 During a median follow-up of 16 months (range 352 three lines of evidence favor this association: (1) histological studies
months), 34 patients (62%) reported signicant pain improvement. and pancreatograms have demonstrated features of chronic pancre-
Patients with normal pancreatograms were more likely to respond atitis isolated to the dorsal pancreas; (2) numerous studies have
to therapy than those with pancreatographic evidence of chronic shown a statistically signicant higher prevalence of pancreas
pancreatitis (73% vs 58%). Jacob et al. postulated that SOD might divisum in this patient population; and (3) numerous studies indi-
cause recurrent episodes of pancreatitis even though SOM was cated symptom resolution by facilitating dorsal duct decompression
normal, and pancreatic stent placement might prevent further endoscopically or surgically.49,53,54
attacks.41 In a randomized study, 34 patients with IARP, normal The diagnosis of pancreas divisum is suspected at ERCP when
pancreatic duct SOM, ERCP, secretin testing, and no biliary crystals only a small ventral ductal system is visualized following contrast
(probably best considered true IARP) were treated with pancreatic injection of the major papilla. It is conrmed when the remainder
stents (n = 19, 57 Fr, with stents exchanged 3 times over a 1-year of the pancreatic ductal system (dorsal duct) is visualized by inject-
period) or conservative therapy (n = 15). During a three-year follow- ing contrast into the minor papilla and there is no communication
up, pancreatitis recurred in 53% of the patients in the control group between the two ductal systems. The clinical presentation and
and only 11% of the stented patients (p < .02). This study suggests response to therapy for incomplete pancreas divisum where there is
that SOM may be an imperfect test, as patients may have SOD but a small lamentous communication between the ventral and dorsal
not be detected at the time of SOM. However, long-term studies are ducts appears to be similar to complete pancreas divisum.3,55
needed to evaluate the outcome after removal of stents, and concern The aim of endoscopic therapy in symptomatic pancreas divisum
remains regarding stent-induced ductal and parenchymal changes.42,43 patients is to alleviate the outow obstruction at the level of the
Because of the concern of stent-induced injury to the pancreas, trial minor papilla. The available endoscopic options include dilation,
pancreatic duct stenting to predict outcome from pancreatic sphinc- long-term dorsal duct stenting, minor papilla sphincterotomy, or a
terotomy is not recommended.44 combination of therapies. Table 41.5 shows the outcome of endo-
Wehrmann and colleagues evaluated the feasibility and effective- scopic therapy in 8 selected series in the literature.5663 Overall, 81%
ness of botulinum toxin injection in patients with recurrent pancre- of 127 patients had no further episodes of pancreatitis during a
atitis due to pancreatic sphincter hypertension.45 No side effects of mean follow-up interval of 27 months.
the injection were noted in any of the 15 treated patients. Twelve It must be appreciated that acute pancreatitis is an episodic
patients (80%) remained asymptomatic at 3-month follow-up, but 11 illness. A follow-up duration as short as 20 months after intervention
developed a relapse at a follow-up period of 6 2 months. These 11 would not be long enough to conclude that a patient is cured.46
patients underwent pancreatic or combined pancreatobiliary sphinc- Moreover we lack the necessary randomized trials proving the
terotomy with subsequent remission after a median follow-up of 15 efcacy of endoscopic intervention.
months. This study showed that injection of botulinum toxin is safe, There is only one randomized study evaluating the role of endo-
may be effective short term, and predict the outcome from pancre- scopic therapy of pancreas divisum in the setting of IARP. Lans and
atic sphincter ablation in patients having frequent episodes of pan- colleagues reported the results of a randomized controlled trial of
creatitis, but the need for denitive sphincter ablation in the majority long-term (12 months) stenting of the minor papilla in patients with
of patients limits its clinical use. at least 2 prior episodes of unexplained pancreatitis (n = 19).59 The
In summary, these data show that SOD is the most common mean follow-up interval was 2.5 years and all stented patients were
cause of IARP when detailed endoscopic evaluation is performed.
SOM should be considered the gold standard for diagnosing SOD.
Complete sphincter evaluation requires manometric assessment of
Endoscopic Follow-up
both the biliary and pancreatic sphincters. Although the best endo-
Author, yr N Therapy (mos) % Improved
scopic therapy of SOD warrants further investigation, there is mount-
ing evidence that pancreatic sphincter ablation will be necessary in Liguory 86 8 MiES 24 63
the majority of patients to achieve the best long-term results. However, McCarthy 88 19 Stent 21 89
Lans 92 10 Stent 30 90
at present, controversy persists as to the appropriateness of perform-
Lehman 93 17 MiES 20 76
ing SOM in patients with IAP.46 In the absence of a well-conducted
Coleman 94 9 MiES/Stent 23 78
randomized controlled trial and long-term patient follow-up, many Kozarek 95 15 MiES/Stent 26 73
authorities consider sphincter ablation therapy experimental and Ertan 00 25 Stent 24 76
should not be performed in routine clinical practice.47,48 Heyries 02 24 MiES/Stent 39 92
Total 127 27 81
PANCREAS DIVISUM
Table 41.5 Endoscopic therapy of acute recurrent pancreatitis
Pancreas divisum, the most common congenital variant of pancre-
due to pancreas divisum
atic duct anatomy, occurs when the dorsal and ventral ducts fail to MiES = Minor papilla sphincterotomy.
fuse during the second month of gestation.49 With non-union of the From references 5663.

438
 Chapter 41 Idiopathic Acute Pancreatitis: Role of ERCP in Diagnosis and Therapy
followed for at least 12 months after stent removal. Stented patients
had statistically signicant fewer hospitalizations and episodes
of pancreatitis (p < 0.05) and were more frequently judged to be
improved (90% vs 11% for controls, p < 0.05). These promising
results certainly support the role of endoscopic therapy in patients
with pancreas divisum presenting with IARP. However, long-term
stenting requires repeated procedures for stent change, each with
an associated risk. Moreover pancreatic stenting is associated with
pancreatic ductal and parenchymal changes which may be perma-
nent.42,43 Finally, an unanswered question is whether stenting for 1
year permanently alleviates the obstruction at the level of the minor
papilla. We prefer performing a minor papilla sphincterotomy which
results in a more permanent enlargement of the minor papilla
orice (Fig. 41.2).49
In summary, patients with IARP found to have pancreas divisum
are good candidates for minor papilla therapy. However long-term
outcome studies (at least 5- to 10-year follow-up), preferably as ran-
domized trials, are necessary to prove the safety and efcacy of
endoscopic therapy not only in the setting of pancreas divisum but
in other settings where a cause for the IARP has been uncovered.
CHOLEDOCHOCELE
Choledochal cysts are uncommon anomalies of the biliary tree mani-
fested by cystic dilation of the intrahepatic or extrahepatic ducts. A
choledochocele, or Type III choledochal cyst using the Todani et al.
classication system, most commonly refers to a cystic dilation of
Fig. 41.2 Minor papilla sphincterotomy. Top left: A normal minor
papilla. A highly tapered catheter and guidewire used to cannulate
the dorsal duct are in view. Top right: A sphincterotome is in the
minor papilla. Bottom left: Completed minor papilla sphincterot-
omy. Bottom right: A dorsal pancreatic duct stent was placed for
pancreatitis prophylaxis.
the terminal common bile duct usually involving the intramural
segment.64 Although pancreatitis has been reported to occur in asso-
ciation with all types of extrahepatic choledochal cysts, it occurs most
commonly with the choledochocele. Moreover, choledochoceles are
rarely identied by standard radiologic imaging (and therefore meet
the denition of IAP) while other extrahepatic choledochal cysts are
typically suspected or recognized by abdominal ultrasound or CT
scan. IAP has been reported in 3070% of patients found to have a
choledochocele.65 Although choledochoceles commonly present with
pancreatitis they are an uncommon cause of IAP because of their
low prevalence.
Choledochoceles are most commonly diagnosed at the time of
ERCP. Endoscopically, the papilla has a bulging appearance but is
soft (pillow sign) when probed with a catheter tip. A rounded cystic
structure can be demonstrated at the terminal end of the common
bile duct following contrast injection into the biliary tree with associ-
ated progressive enlargement or ballooning of the papilla.66,67
Surgical therapy, either by excision or sphincteroplasty, has been
the traditional approach to choledochoceles.65 There is limited data
to suggest that endoscopic therapy is a safe and effective alternative
to surgery. The endoscopic approach is to unroof the cyst and
perform a biliary sphincterotomy (Fig. 41.3). Table 41.6 presents the
results of three selected series reporting the outcome of endoscopic
therapy. Ten of 11 patients treated had no further episodes of
pancreatitis during the follow-up period.66,68,69
In summary, choledochoceles are an uncommon cause of IARP
but commonly present with pancreatitis. The diagnosis is made on
endoscopic views of the papilla and contrast injection of the biliary
tree. Endoscopic therapy appears to be an effective treatment in the
large majority of patients.
Fig. 41.3 Left: Endoscopic view of a choledochocele. Right: Six
weeks after endoscopic unroong of the choledochocele and
biliary sphincterotomy.
# Improved/# with IARP
Author (yr) N # Pancreatitis Treated at ERCP
Venu (1984) 8 5 2/3
Martin (1992) 10 7 7/7
Ladas (1995) 15 1 1/1
Total 33 13 (39%) 10/11 (91%)
Table 41.6 Endoscopic therapy of choledochoceles in patients
presenting with idiopathic acute recurrent pancreatitis
IARP = idiopathic acute recurrent pancreatitis.
From references 66, 68, 69.
439
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

TUMORS A B
Five to seven percent of patients with benign or malignant pancre-
aticobiliary and ampullary tumors present with IAP.3 These tumors
should be considered in patients 40 years or older who present with
their rst episode of acute pancreatitis.10,11 Patients with hereditary
conditions such as Familial Adenomatous Polyposis (FAP) may have
ampullary involvement and present with IAP at a youger age. The
most common tumors reported in IARP series are: intraductal
papillary mucinous tumors (IPMT) and cystic tumors, ampullary
(papillary) tumors, pancreatic adenocarcinoma, and islet cell
tumors. Ampullary tumors and IPMT deserve special mention since
they are commonly missed on standard abdominal imaging tests
C D
and identied at the time of ERCP.6
There are a wide array of benign tumors that arise at the major
papilla including adenoma, lipoma, broma, lymphangioma, leio-
myoma and hamartoma.70 All have the potential to cause pancreatitis
by obstructing pancreatic juice ow. Adenoma is the most common
benign tumor of the major papilla. Endoscopy is the most sensitive
and specic method for diagnosis of papillary tumors, as it accu-
rately localizes the lesion and provides biopsy conrmation. Although
there is uniform agreement that papillary adenomas should be
resected, there is controversy as to the optimal method of excision.
Regardless of the method of resection, complete removal is manda-
tory.70 The trend in management of papillary adenomas has been
Fig. 41.4 A Endoscopic view of a tubulovillous adenoma involv-
toward an increased use of endoscopic therapy perhaps because of ing the papilla and extending caudually down the duodenal
more widespread use and experience with endoscopic mucosal wall. B The papilla was snared and electrocautery applied. C Ap-
resection in other parts of the GI tract. Evidence is accumulating to pearance of the ampullary segment following snare resection of
indicate that endoscopic resection (snare papillectomy), thermal the papilla. D The tumor involving the duodenal wall was resected
and a biliary sphincterotomy was done (patient has pancreas
ablation or a combination of the two are the treatment of choice divisum, so a pancreatic stent was not placed in the ventral pancre-
for most papillary adenomas (Fig. 41.4).7175 The techniques of atic duct). This endoscopic picture shows the area of the resected
endoscopic management and surveillance are discussed in detail tumor.
by Kim and colleagues.70
There are a variety of primary malignant tumors of the major
papilla including carcinoma, lymphoma, and neuroendocrine
tumors. Metastatic tumors include malignant melanoma, hyperne-
phroma, and lymphoma.70 Although most patients with malignant
tumors of the papilla present with obstructive jaundice, there are
occasional patients who develop pancreatitis as their rst sign of the
disease. ERCP is used to conrm the diagnosis and offer palliative
stenting in non-resectable patients.
IPMTs are clearly premalignant lesions, with up to 50% of
patients having invasive carcinoma at operation.76 It is not uncom-
mon to nd patients with IPMT present with recurrent pancreatitis
for many years before the diagnosis is made.77 Endoscopy and ERCP
are essential to the diagnosis. ERCP was previously considered the
standard for evaluation and diagnosis.76,78 Pancreatography typically
reveals a dilated main pancreatic duct with intraductal cast-like Fig. 41.5 Intraductal papillary mucinous tumor (IPMT). Endoscopic
lling defects representing mucin (Fig. 41.6).76 A patulous pancre- retrograde pancreatogram showing a markedly dilated pancreatic
atic orice exuding mucin is seen in up to 80% of patients (Fig. duct containing lling defects in the body and tail consistent with
mucus.
41.7).76 However, pancreatographic ndings may be much more
subtle and be misinterpreted as normal when a small cast-like lling
defect is missed or a lling defect is labeled a pancreatic stone or air
bubble. A missed diagnosis often comes in the setting of a normal obtain cytologic specimens by aspiration or brushing, guided forceps
diameter pancreatic duct (Fig. 41.7). Finally, IPMT may be misinter- biopsy specimens for histology and tissue and pancreatic juice for
preted as chronic pancreatitis. A high index of suspicion of the tumor markers. Pancreatoscopy and intraductal ultrasound are
endoscopist is therefore critical particularly in patients older than 40 ancillary techniques utilized at the time of ERCP to help localize the
years. During pancreatography, early x-ray lms should be obtained tumor, differentiate benign from malignant disease and aid in the
and the uoroscopic image should be carefully observed so that differential diagnosis of amorphous lling defects in the pancreatic
small lling defects are not missed. At ERCP, it is also possible to duct.79

440
 Chapter 41 Idiopathic Acute Pancreatitis: Role of ERCP in Diagnosis and Therapy
Fig. 41.6 Endoscopic views of an intraductal papillary mucinous
tumor (IPMT). The patulous pancreatic orice is exuding mucin.
Fig. 41.7 Intraductal papillary mucinous tumor (IPMT). In this case,
the cast-like lling defect (arrows) is present in a normal diameter
pancreatic duct.
There appears to be no role or very little role of endoscopic
therapy in IPMT (except when biliary obstruction occurs). However,
the value of pancreatic sphincterotomy to assist in passage of mucin
in high-risk surgical patients has not been evaluated. Given the fact
that many of these patients already have gaping pancreatic orices,
it is unlikely to be of signicant long-term benet.
EUS is assuming an increasingly important role in the assess-
ment of patients with IAP and IARP (see below).80 Many studies
have found EUS to have the highest sensitivity for identifying pan-
creatic neoplasms relative to other imaging modalities, especially for
tumors smaller than 23 cm in diameter.3,81,82 This seems to be a
reasonable test to perform particularly in patients older than 40,
when other radiological imaging tests are negative, and perhaps
should be considered prior to ERCP.
OTHER ANATOMICAL CAUSES
There are a number of other non-neoplastic structural lesions that
can cause acute pancreatitis. Pancreatic ductal strictures, which
result in upstream ductal hypertension, are usually the result of
prior trauma, or develop after healing of a pseudocyst or necrosis.6
Duodenal diverticula are rarely associated with pancreatitis.83
Anomalous pancreaticobiliary duct junction is a rare congenital
malformation in which the union of the pancreatic duct and bile
duct occurs outside the duodenal wall. As a result, the sphincter of
Oddi is unable to prevent the reciprocal regurgitation of pancreatic
enzymes and bile into the alternate biliary and pancreatic ducts.
Such unions may occur in isolation or be associated with chole-
dochal cyst disease. Pancreatitis may be a complication of this
anomaly. Samavedy and colleagues suggested that endoscopic
therapy eliminates or reduces the frequency of recurrent pancreatitis
and would be a logical rst step in the management of most symp-
tomatic patients.84 Annular pancreas is another congenital anomaly
associated with pancreatitis that manifests as a band of pancreatic
tissue partially or completely encircling the duodenum. ERCP typi-
cally identies the duct of the annulus. Pancreas divisum is present
in about one-third of patients.49,85 Duodenal duplication cyst, also a
distinctly rare congenital anomaly, may present with pancreatitis. A
potential role of endoscopic therapy in the treatment of such cysts
has been described.86 Finally, chronic pancreatitis can be diagnosed
at ERCP when main duct and/or side branch changes are
present. In a series of 90 patients with IAP and IARP, 44% of
patients had chronic pancreatitis by EUS and ERCP criteria.87 An
obvious structural cause for the clinical episodes of acute pancreati-
tis may not be apparent after detailed endoscopic evaluation has
been done.
OTHER INVESTIGATIONS
The literature is now replete with evidence showing that patients
with IAP and IARP may have mutations in the genes encoding cat-
ionic trypsinogen (PRSS1), pancreatic secretory trypsin inhibitor
(Serine Protease Inhibitor Kazal Type I or SPINK-1), and cystic
brosis transmembrane conductance regulator (CFTR).20,8892 Testing
for these genetic mutations is commercially available although many
of the mutations causing pancreatitis cannot be detected by the cur-
rently available techniques. The role of genetic testing in IAP is
controversial.20,88 However, detection of a genetic cause for the pan-
creatitis may obviate further testing, assist with family planning, and
identify the patient for surveillance for complications of their pan-
creatitis including pancreatic cancer.93 In most situations, endo-
scopic therapy does not differ in patients with acute pancreatitis,
chronic pancreatitis or pancreaititis complications regardless of
whether a genetic mutation is present.
Autoimmune pancreatitis is a relatively newly described entity
characterized radiographically by diffuse or segmental irregular nar-
rowing of the main pancreatic duct and diffuse enlargement of the
pancreas, laboratory evidence of elevated levels of serum IgG (par-
ticularly the IgG 4 subtype) and the presence of autoantibodies, and
histopathologically by brotic changes with lymphoplasmacytic inl-
tration of the pancreas.94 In contrast to other forms of chronic pan-
creatitis, it responds dramatically to steroids. Thus, laboratory
screening with antinuclear antibody and serum immunoglobulins
with IgG subtypes can be considered in selected patients with
IAP.20
Endoscopic Ultrasound (EUS) has assumed a central role in the
evaluation of patients with IAP and IARP.3,4 The ndings on EUS
may direct an alternative therapy (e.g. cholecstectomy for microli-
thiasis) and obviate a more invasive ERCP. Clearly EUS can identify
small tumors of the pancreas not seen on other radiologic imaging
studies and that may also be missed at ERCP.82 Frossard and col-
leagues reported that EUS identied a cause for IAP in 131 of 168
patients (78%) including 103 (61%) with biliary tract disease, 16
(10%) with chronic pancreatitis, 4 (2%) with IPMT, 4 (2%) with
pancreatic cancer, 3 (2%) with amullary tumors, and 1 (1%) with a
choledochal cyst.80 The denitive diagnosis was made at surgery in
101 patients (60%), at ERCP in 49 (29%) and by bile crystal analysis
441
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Diagnosis Number abnormal

Sphincter of Oddi Dysfunction 179 (34%)
Pancreas Divisum 70 (13%)
Pancreatic or papillary tumor 46 (9%)
Gallbladder or duct stones 37 (7%)
Pancreatic duct stricture/chronic pancreatitis 37 (7%)
Choledochocele 12 (2%)
Total abnormal 381/522 (73%)

Table 41.7 Idiopathic acute recurrent pancreatitisdiagnostic

yield of ERCP, sphincter of Oddi manometry and bile microscopy
(4 selected series of 522 patients)
From references 10, 18, 32, 33.

No.
No. treated Follow-
Diagnosis patients at ERCP up (mos) Asymptomatic
Sphincter of Oddi 67 67 33 79%
Dysfunction
Gallbladder or duct 18 16a 31 89%
stones
Pancreas Divisum 9 8b 24 89%
Tumor 2 0c 28 50%
Fig. 41.8 Secretin-stimulated MRCP demonstrating pancreas Choledochocele 2 2 18 100%
divisum. Note the normal dorsal duct. Pancreatic duct 2 2 31 50%
stricture
Total 100 95 30 81%

Table 41.8 Outcome of endoscopic therapy in 100 patients

or clinical follow-up in 18 (11%). Canadian researchers compared
the diagnostic yield of EUS in 201 patients with a single episode of
IAP with 169 patients with recurrent episodes.95 When chronic pan-
creatitis was not considered a positive nding, 29.2% of patients
were found to have a cause for their pancreatitis. Biliary tract disease
(18.7%) was the most common positive nding in the patients with
the gallbladder in situ, whereas pancreas divisum (11.3%) was the
most common nding in the post-cholecystectomy group. The fre-
quency of positive ndings did not differ between those with a single
attack vs those with recurrent attacks. EUS identied an unsus-
pected cystic or solid neoplasm in 3.5% of patients. Chronic
pancreatitis, the most common abnormality found (29.5%), was
identied in about twice as many patients with recurrent episodes
as in those with single episodes.
MRI/MRCP has also assumed a vital role in the evaluation of
patients with IAP and IARP because of its high-quality imaging of
parenchymal and ductal structures. Secretin stimulaton can add to
the diagnostic accuracy of this test (Fig. 41.8).9699 Like EUS, MRI/
MRCP has a 90100% sensitivity and specicity for detecting bile
duct stones.
Both EUS and MRI/MRCP are operator-dependent and their
sensitivity and specicity in identifying a cause for the pancreatitis
will depend on local expertise and availability.
YIELD OF ERCP FOR IAP AND IARP
Table 41.7 summarizes the results of 4 selected series that utilized
ERCP, SOM and bile microscopy in the assessment of 522 patients
with IAP and IARP.10,18,32,33 Overall, 73% of patients were found to
have a cause for their pancreatitis with SOD being the most common
diagnosis.
found to have a cause for idiopathic acute recurrent pancreatitis
From reference 33.
a
Two treated with cholecystectomy.
b
One had unsuccessful minor papilla sphincterotomy.
c
One treated with Whipple; 1 with unresectable adenocarcinoma.
OUTCOMES OF ENDOSCOPIC THERAPY IN IARP
Unfortunately there is a paucity of controlled data regarding out-
comes in patients undergoing endoscopic therapy for previously
diagnosed IARP. In fact, only two randomized controlled trials have
been reported.41,59 It is also difcult to interpret what data are
available due to loosely dened outcome measures, inhomogeneous
clinical characteristics, generally short-term follow-up, and varied
treatment techniques.6 Also, as emphasized above, since recurrent
pancreatitis is an episodic illness, long-term follow-up (at least 510
years) after therapy is necessary before concluding that the therapy
was effective.46,100
The outcomes of endoscopic therapy are detailed in the sections
above. One recent study should be highlighted.33 One hundred and
twenty-six patients with IARP underwent ERCP, SOM, and biliary
crystal analysis. Patients had a mean of 3.2 episodes of pancreatitis
(mean interval between recurrent attacks was 3.8 months). A cause
for the pancreatitis was identied in 100 patients (79%) and included
sphincter of Oddi dysfunction or papillary stenosis with or without
crystals in 67 (53%), microcrystals alone in 12 (9.5%), pancreas
divisum in 9 (7.1%), common duct stones in 6 (4.8%), malignancy
in 2 (1.6%), chronic pancreatitis with a pancreatic stricture in 2
(1.6%) and choledochocele in 2 (1.6%). Endoscopic therapy was
performed in 95 patients and 3 patients underwent surgery. The
outcome of therapy is shown on Table 41.8. During a mean follow-

442
 Chapter 41 Idiopathic Acute Pancreatitis: Role of ERCP in Diagnosis and Therapy
up of 30 months, 81% of patients were asymptomatic. Twenty-four
patients had procedure-related complications; 20 had pancreatitis.
CONCLUSIONS
IAP and IARP are a challenging clinical problem for the physician
and often a frustrating one for the patient. ERCP with ancillary
techniques can identify a probable cause for the pancreatitis in
about 75% of patients. The majority of the diseases uncovered
appear to be treatable by endoscopic or surgical techniques. EUS
and MRI/MRCP are assuming a more central role in the evaluation
REFERENCES
1. Grendell JH. Idiopathic acute pancreatitis. Gastroenterol Clin
North Am 1990; 19:843848.
2. Ballinger AB, Barnes E, Alstead EM, et al. Is intervention
necessary after a rst episode of acute idiopathic pancreatitis?
Gut 1996; 38:293295.
3. Levy MJ, Geenen JE. Idiopathic acute recurrent pancreatitis. Am J
Gastroenterol 2001; 96:25402555.
4. Kozarek R. Role of ERCP in acute pancreatitis. Gastrointest
Endosc 2002; 56(supp):S231S236.
5. Steer ML, Saluja AK. Pathogeneis and pathophysiology of acute
pancreatitis. In: Beger HG, Warshaw AL, Buchler MW et al. (eds),
The Pancreas. Oxford: Blackwell Science, 1998:383391.
6. Tarnasky PR, Hawes RH. Endoscopic diagnosis and therapy of
unexplained (idiopathic) acute pancreatitis. Gastrointest Endosc
Clin N Am 1998; 8:1337.
7. Frakes JT. Biliary pancreatitis: a review emphasizing appropriate
endoscopic intervention. J Clin Gastroenterol 1999; 28:97109.
8. Gregor JC, Ponich TP, Detsky AS. Should ERCP be routine after
an episode of idiopathic pancreatitis? A cost-utility analysis.
Gastrointest Endosc 1996; 44:118123.
9. Tranpnell JE, Duncan EHL. Patterns of incidence in acute
pancreatitis. Brit Med J 1975; 2:179183.
10. Choudari CP, Fogel EL, Sherman S, et al. Idiopathic pancreatitis:
yield of ERCP correlated with patient age. Am J Gastroenterol
1998; 93:1654A.
11. Fischer M, Fogel EL, McHenry L, et al. ERCP/manometry ndings
in 1108 idiopathic pancreatitis patients. Gastrointest Endosc
2005; 61:AB190.
12. Levy MJ. The hunt for microlithiasis in idiopathic acute recurrent
pancreatitis: should we abandon the search or intensify our
efforts. Gastrointest Endosc 2002; 55:286293.
13. Dahan P, Andant C, Levy P, et al. Prospective evaluation of
endoscopic ultrasonography and microscopic examination of
duodenal bile in the diagnosis of cholecystolithiasis in 45
patients with normal conventional ultrasonography. Gut 1996;
38:277281.
14. Dill JE, Hill S, Callis J, et al. Combined endoscopic ultrasound and
stimulated biliary drainage in cholecystitis and microlithiasis
diagnosis and outcomes. Endoscopy 1995; 27:424427.
15. Liu CL, Lo CM, Chan JK, et al. EUS for detection of occult
cholelithiasis in patients with idiopathic pancreatitis. Gastrointest
Endosc 2000; 51:2832.
16. Lee SP, Nicholls JF, Park HZ. Biliary sludge as a cause of acute
pancreatitis. N Engl J Med 1992; 326:589593.
17. Ros E, Navarro S, Bru C, et al. Occult microlithiasis in idiopathic
acute pancreatitis: Prevention of relapses by cholecystectomy or
ursodeoxycholic acid therapy. Gastroenterology 1991;
101:17011709.
of patients with IAP and IARP. When these studies identify the
cause of the pancreatitis, appropriate targeted treatment should be
recommended. In this setting, ERCP should be utilized for therapy
when appropriate. ERCP with ancillary endoscopic techniques is
clearly indicated when EUS and/or MRI/MRCP fail to identify a
cause for the recurrent pancreatitis. Further investigation is neces-
sary to develop an algorithm which will provide the most cost-effec-
tive approach for these patients. Patients with IAP and IARP are
best evaluated in centers where specialized expertise and equip-
ment are available when advanced endoscopic methods are
required.
18. Venu RP, Geenen JE, Hogan W, et al. Idiopathic recurrent
pancreatitis: An approach to diagnosis and treatment. Dig Dis Sci
1989; 34:5660.
19. Sherman S, Gottlieb K, Earle D, et al. The role of microlithiasis in
idiopathic pancreatitis. Gastrointest Endosc 1997; 45:165A.
20. Draganov P, Forsmark CE. Idiopathic pancreatitis.
Gastroenterology 2005; 128:756763.
21. Steinberg WM, Geenen JE, Bradley EL, et al. Controversies in
clinical pancreatology. Recurrent idiopathic acute pancreatitis.
Should a laparoscopic cholecystectomy be the rst procedure
of choice? Pancreas 1996:329334.
22. Sherman S. What is the role of ERCP in the setting of abdominal
pain of pancreatic or biliary origin (suspected sphincter of Oddi
dysfunction)? Gastrointest Endosc 2002; 56(Supp):S258S266.
23. Lehman G, Sherman S. Sphincter of Oddi dysfunction: a review.
International J Pancreatology 1996; 20:1125.
24. Hogan WJ, Sherman S, Pasricha P, et al. Sphincter of Oddi
manometry. Gastrointest Endosc 1997; 45:342348.
25. Corazziari E, Hogan W, Shffer EA, et al. Functional disorders of
the biliary tract and pancreas. Gut 1999; 45:II48II54.
26. Kuo WH, Pasricha PJ, Kalloo AN. The role of sphincter of Oddi
manometry in the diagnosis and therapy of pancreatic disease.
Gastrointest Endosc Clin N Am 1998; 8:7985.
27. Sherman S, Fogel EL, Watkins JL, et al. Sphincter of Oddi
dysfunction. In: Ginsberg GG, Kochman ML, Norton I, et al. (eds).
Clinical Gastrointestinal Endoscopy. Philadelphia: Elsevier Health
Sciences, 2005:733750.
28. Mariani A, Curioni S, Zanello A, et al. Secretin MRCP and
endoscopic pancreatic manometry in the evaluation of
sphincter of Oddi function: a comparative pilot study in patients
with idiopathic recurrent pancreatitis. Gastrointest Endosc 2003;
58:847852.
29. Toouli J, Roberts-Thomson IC, Dent J, et al. Sphincter of Oddi
Motility disorders in patients with idiopathic recurrent
pancreatitis. Br J Surg 1985; 72:859863.
30. Guelrud M, Mendoz S, Viera L. Idiopathic recurrent pancreatitis
and hypercontractile sphincter of Oddi. Treatment with
endoscopic sphincterotomy and pancreatic duct dilation.
Gastroenterology 1986; 90:1443.
31. Gregg JA. Function and dysfunction of the sphincter of Oddi. In:
Jacobson IM (ed.) ERCP: Diagnostic and therapeutic applications.
New York: Elsevier, 1989:137170.
32. Sherman S, Jamidar P, Reber H. Idiopathic acute pancreatitis
(IAP): endoscopic diagnosis and therapy. Am J Gastroenterol
1993; 88:1541A.
33. Kaw M, Brodmerkel GJ. ERCP, biliary crystal analysis, and
sphincter of Oddi manometry in idiopathic pancreatitis.
Gastrointest Endosc 2002; 55:157162.
443
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
34. Raddawi HM, Geenen JE, Hogan WJ, et al. Pressure measurement
from biliary and pancreatic segments of sphincter of Oddi.
Comparison between patients with functional abdominal pain,
biliary, or pancreatic disease. Dig Dis Sci 1991; 36:7174.
35. Sherman S, Hawes RH, Madura JA, et al. Comparison of
intraoperative and endoscopic manometry of the sphincter of
Oddi. Surg Gynecol Obstet 1992; 175:410418.
36. Madura JA, Madura JA II, Sherman S, et al. Surgical
sphincteroplasty in 446 patients. Archives of Surgery 2005;
140:504513.
37. Lans JL, Parikh NP, Geenen JE. Applications of sphincter of Oddi
manometry in routine clinical investigations. Endoscopy 1991;
23:139143.
38. Guelrud M, Plaz J, Mendoza S, et al. Endoscopic treatment in
Type II pancreatic sphincter dysfunction. Gastrointest Endosc
1995; 41:A398.
39. Toouli J, Di Francesco V, Saccone G, et al. Division of the
sphincter of Oddi for treatment of dysfunction associated with
recurrent pancreatitis. Br J Surg 1996; 83:12051210.
40. Okolo PI, Pasricha PJ, Kalloo AN. What are the long-term results
of endoscopic pancreatic sphincterotomy. Gastrointest Endosc
2000; 52:1519.
41. Jacob L, Geenen JE, Catalano MF, et al. Prevention of pancreatitis
in patients with idiopathic recurrent pancreatitis: a prospective
non-blinded randomized study using endoscopic stents.
Endoscopy 2001; 33:559562.
42. Kozarek RA. Pancreatic stents can induce ductal changes
consistent with chronic pancreatitis. Gastrointest Endosc 1990;
36:9395.
43. Smith MT, Sherman S, Ikenberry SO, et al. Alterations in
pancreatic duct morphology following polyethylene pancreatic
stent therapy. Gastrointest Endosc 1996; 44:268275.
44. Testoni PA, Caporuscio S, Bagnolo F, et al. Idiopathic recurrent
pancreatitis: Long-term results after ERCP, endoscopic
sphincterotomy, or ursodeoxycholic acid treatment. Am J
Gastroenterol 2000; 95:17021707.
45. Wehrmann T, Schmitt TH, Arndt A, et al. Endoscopic injection of
botulinum toxin in patients with recurrent acute pancreatitis due
to pancreatic sphincter of Oddi dysfunction. Aliment Pharmacol
Ther 2000; 14:14691477.
46. Steinberg WM. Should the sphincter of Oddi be measured in
patients with idiopathic recurrent acute pancreatitis, and should
sphincterotomy be performed if the pressure is high. Pancreas
2003; 27:118121.
47. Petersen BT. Sphincter of Oddi dysfunction, part 2 Evidence-
based review of the presentations, with objective pancreatic
ndings (types I and II) and of presumptive type III. Gastrointest
Endosc 2004; 59:670687.
48. Mark DH, Lefevre F, Flamm CR, et al. Evidence-based assessment
of ERCP in the treatment of pancreatitis. Gastrointest Endosc
2002; 56(Supp):S249S254.
49. Lehman GA, Sherman S. Diagnosis and therapy of pancreas
divisum. Gastrointest Endosc Clin N Am 1998; 8:5577.
50. Cotton PB, Kizu M. Malfusion of dorsal and ventral pancreas: a
cause of pancreatitis. Gut 1977; 18:A400.
51. Gregg JA. Pancreas divisum: its association with pancreatitis. Am
J Surg 1977; 134:539543.
52. Cotton PB. Congenital anomaly of pancreas divisum as a
cause of obstructive pain and pancreatitis. Gut 1980;
21:105114.
53. Warshaw AL, Simeone JF, Schapiro RH, et al. Evaluation and
treatment of the dominant dorsal duct syndrome (pancreas
divisum redened). Am J Surg 1990; 159:5966.
54. Somogyi L, Martin SP, Venkatesan T, et al. Recurrent acute
pancreatitis: an algorithmic approach to identication and
444
elimination of inciting factors. Gastroenterology 2001;
120:708717.
55. Jacob L, Geenen JE, Catalano MF, et al. Clinical presentation and
short-term outcome of endoscopic therapy of patients with
symptomatic incomplete pancreas divisum. Gastrointest Endosc
1999; 49:5357.
56. Liguory C, Lefebvre JF, Canard JM, et al. Pancreas divisum: clinical
and therapeutic study in man: Apropos 87 cases. Gastroenterol
Clin Biol 1986; 10:820825.
57. McCarthy J, Geenen JE, Hogan WJ. Preliminary experience with
endoscopic stent placement in benign pancreatic disease.
Gastrointest Endosc 1988; 34:1618.
58. Lehman GA, Sherman S, Nisi R, et al. Pancreas divisum: Results
of minor papilla sphincterotomy. Gastrointest Endosc 1993;
39:18.
59. Lans JI, Geenen GE, Johanson JF, et al. Endoscopic therapy in
patients with pancreas divisum and acute pancreatitis: A
prospective, randomized, controlled clinical trail. Gastrointest
Endosc 1992; 38:430434.
60. Kozarek RA, Ball TJ, Patterson DJ, et al. Endoscopic approaches
to pancreas divisum. Dig Dis Sci 1995; 40:19741981.
61. Coleman SD, Eisen GM, Troughton AB, et al. Endoscopic
treatment in pancreas divisum. Am J Gastroenterol 1994;
89:11521155.
62. Ertan A. Long-term results after endoscopic pancreatic stent
placement without pancreatic papillotomy in acute recurrent
pancreatitis due to pancreas divisum. Gastrointest Endosc 2000;
52:915.
63. Heyries L, Barthet M, Delvasto C, et al. Long-term results of
endoscopic management of pancreas divisum with acute
recurrent pancreatitis. Gastrointest Endosc 2002; 55:376381.
64. Todani T, Watanabe Y, Narusue M, et al. Congenital bile duct
cysts: Classication, operative procedures, and review of thirty-
seven cases including cancer arising from choledochal cyst. Am
J Surg 1977; 134:263269.
65. Sherman S. Choledochal cysts. In: Snape WJ (ed.) Consultations
in Gastroenterology. Philadelphia: WB Saunders Co,
1996:814828.
66. Venu RP, Geenen JE, Hogan WJ, et al. Role of endoscopic
retrograde cholangiopancreatography in the diagnosis and
treatment of choledochocele. Gastroenterology 1984;
87:11441149.
67. Kim MH, Myung SJ, Lee SK, et al. Ballooning of the papilla
during contrast injection: the semaphore of a choledochocele.
Gastrointest Endosc 1998; 48:258262.
68. Ladas SD, Katsorgridakis I, Tassios P, et al. Choledochocele, an
overlooked diagnosis: report of 15 cases and review of 56
published reports from 1984 to 1992. Endoscopy 1995;
27:233239.
69. Martin RF, Biber BP, Bosco JJ, et al. Symptomatic
choledochoceles in adults. Arch Surg 192; 127:536539.
70. Kim MH, Lee SK, Seo DW, et al. Tumors of the major duodenal
papilla. Gastrointest Endosc 2001; 54:609620.
71. Norton ID, Geller A, Petersen BT, et al. Endoscopic surveillance
and ablative therapy for peripapillary adenomas. Am J
Gastroenterol 2001; 96:101106.
72. Binmoeller KF, Boaventura S, Ramsperger K, et al. Endoscopic
snare excision of benign adenomas of the papilla of Vater.
Gastrointest Endosc 1993; 39:127131.
73. Desilets DJ, Dy RM, Ku PK, et al. Endoscopic management of
ampullary tumors: rened techniques to improve outcome and
avoid complications. Gastrointest Endosc 2001; 54:202208.
74. Catalano MF, Linder JD, Chak A, et al. Endoscopic management
of adenoma of the major duodenal papilla. Gastrointest Endosc
2004; 59:225232.
 Chapter 41 Idiopathic Acute Pancreatitis: Role of ERCP in Diagnosis and Therapy
75. Cheng CL, Sherman S, Fogel EL, et al. Endoscopic snare
papillectomy for tumors of the duodenal papillae. Gastrointest
Endosc 2004; 60:757764.
76. Farrell JJ, Brugge WR. Intraductal papillary mucinous tumor of
the pancreas. Gastrointest Endosc 2002; 55:701714.
77. Enner S, Carr-Locke DL, Banks PA, et al. Intraductal mucin-
hypersecreting neoplasm mucinous duct ectasia: endoscopic
recognition and management. Am J Gastroenterol 1996;
91:25482553.
78. Loftus EV, Olivares-Pakzad BA, Batts KP, et al. Intraductal papillary
mucinous tumors of the pancreas: clinicopathologic features,
outcome, and nomenclature. Gastroenterol 1996;
110:19091918.
79. Hara T, Yamaguchi T, Ishihara T, et al. Diagnosis and patient
management of intraductal papillary mucinous tumor of the
pancreas by using peroral pancreatoscopy and intraductal
ultrasonography. Gastroenterology 2002; 122:3443.
80. Frossard JL, Sosa-Valencia L, Amouyal G, et al. Usefulness of
endoscopic ultrasonography in patients with idiopathic acute
pancreatitis. Am J Med 2000; 109:196200.
81. Rosch T. Staging of pancreatic cancer: Analysis of literature
results. Gastrointest Endosc Clin North Am 1995; 5:735739.
82. DeWitt J, Devereaux B, Chriswell M, et al. Comparison of
endoscopic ultrasonography and multidetector computed
tomography for detecting and staging pancreatic cancer. Ann
Intern Med 2004; 141:753763.
83. Uomo G, Manes G, Ragozzino A, et al. Periampullary
extraluminal duodenal diverticula and acute pancreatitis: An
underestimated etiological association. Am J Gastroenterol 1996;
91:11861190.
84. Samavedy R, Sherman S, Lehman GA. Endoscopic therapy in
anomalous pancreatobiliary duct junction. Gastrointest Endosc
1999; 50:623627.
85. Tagge EP, Smith SD, Raschbaum GR, et al. Pancreatic ductal
abnormalities in children. Surgery 1991; 110:709717.
86. Johanson JF, Geenen JE, Hogan WJ, et al. Endoscopic therapy of
a duodenal duplication cyst. Gastrointest Endosc 1992;
38:6064.
87. Coyle WJ, Pineau BC, Tarnasky PR, et al. Evaluation of
unexplained acute and acute recurrent pancreatitis using
endoscopic retrograde cholangiopancreatography, spincter of
Oddi manometry and endoscopic ultrasound. Endoscopy 2002;
34:617623.
88. Gorry MC, Gabbaizedeh D, Furey W, et al. Mutations in
the cationic trypsinogen gene are associated with recurrent
acute and chronic pancreatitis. Gastroenterol 1997;
113:10631068.
89. Cohn JA, Friedman KJ, Noone PG, et al. Relation between
mutations of the cystic brosis gene and idiopathic pancreatitis.
N Engl J Med 1998; 339:653658.
90. Choudari CP, Imperiale TF, Sherman S, et al. Risk of pancreatitis
with mutation of the cystic brosis gene. Am J Gastroenterol
2004; 99:13581363.
91. Cohn JA, Neoptolemos JP, Feng J, et al. Increased risk of
idiopathic chronic pancreatitis in cystic brosis carriers. Hum
Mutat 2005; 26:303307.
92. Morinville V, Whitcomb D. Recurrent acute and chronic
pancreatitis: Complex disorders with a genetic basis.
Gastroenterol & Hepatol 2005; 1:195205.
93. Ellis I. Genetic counseling for hereditary pancreatitis-the role of
molecular genetics testing for the cationic trypsinogen gene,
cystic brosis and serine protease inhibitor Kazal type 1.
Gastroenterol Clin NAm 2004; 33:839854.
94. Kim KP, Kim MH, Song MH, et al. Automimmune chronic
pancreatitis. Am J Gastroenterol 2004; 99:16051616.
95. Yusoff IF, Raymond G, Sahai AV. A prospective comparison of
the yield of EUS in primary vs. recurrent idiopathic acute
pancreatitis. Gastrointest Endosc 2004; 60:673678.
96. Moon JH, Cho YD, Cha SW, et al. The detection of bile duct
stones in suspected biliary pancreatitis: comparison of MRCP,
ERCP, and intraductal US. Am J Gastroenterol 2005;
100:10511057.
97. Arvanitakis M, Delhaye M, De Maertelaere V, et al. Computed
tomography and magnetic resonance imaging in the
assessment of acute pancreatitis. Gastroenterology 2004;
126:715723.
98. Matos C, Metens T, Deviere J, et al. Pancreatic duct:
morphologic and functional evaluation with dynamic MR
pancreatography after secretin stimulation. Radiology 1997;
203:435441.
99. Hellerhoff KJ, Helmberger, H, 3rd, Rosch T, et al. Dynamic MR
pancreatography after secretin administration: image quality and
diagnostic accuracy. AJR 2002; 179:121129.
100. Steinberg WM, Chari ST, Forsmark CE, et al. Controversies in
clinical pancreatology: management of acute idiopathic
recurrent pancreatitis. Pancreas. 2003; 27:103117.
445
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Pancreas Divisum and Other
42 Pancreaticobiliary Anomalies
Mark Topazian
INTRODUCTION
Anomalies of the biliary and pancreatic ducts are commonly encoun-
tered during ERCP, and are important to both surgeons and gastro-
enterologists. This chapter reviews the diagnosis, clinical relevance,
and therapy of these variants.
AMPULLARY ANOMALIES
Ectopic major papilla
The major papilla, which is typically located in the mid or distal
second duodenum, is occasionally located in the third duodenum.1
Ectopic distal location of the ampulla is associated with anomalous
pancreaticobiliary junction, congenital biliary dilatation, and biliary
cysts.2 The distal displacement of the papilla may correspond to the
length of an abnormally long common channel,3 and may reect
failure of the ducts to migrate normally into the duodenum during
embryologic development. Rarely the major papilla may be located
in the duodenal bulb.4 Double papilla of Vater has been described.5
When the papilla is in an anomalous location, the oblique intramu-
ral course of the bile duct is often absent,6 leaving less room for
endoscopic biliary sphincterotomy.
Anomalous pancreaticobiliary junction
The bile duct and pancreatic duct typically form a short common
channel in the papilla of Vater. In about 1% of persons there is
no common channel, and separate orices of the bile duct and
pancreatic duct are found on the major papilla. Less commonly
the ducts join each other proximal to the sphincter apparatus,
and a long common channel (>10 mm, and often >2 cm) is present
(Fig. 42.1). A long common channel is often referred to as an
anomalous pancreaticobiliary junction (APBJ); synonyms include
pancreaticobiliary malunion. APBJ can be sub-classied according
to the presence or absence of pancreas divisum, a dilated common
channel, and an acute angle between the bile duct and pancreatic
duct.7 These ndings may inuence choice of management
and surgical strategy in symptomatic patients. For instance, symp-
tomatic obstruction of the common channel is often treated
endoscopically.
APBJ is present in the majority of patients with biliary cysts, and
appears to be a risk factor for the development of malignancy in a
biliary cyst as discussed below. Patients with APBJ and no biliary
cyst have increased risk of gallbladder cancer, and develop gallblad-
der cancer at an earlier age (Fig. 42.2).8,9 The nding of an isolated
APBJ should prompt consideration of elective cholecystectomy, even
when asymptomatic.
BILIARY ANOMALIES
Bile duct anatomy
Coiunaud10 described the liver as being composed of 4 sectors,
dened by the 3 hepatic veins.11 The 4 sectors can be further subdi-
vided into 8 segments, which are drained by segmental bile ducts
(Fig. 42.3). Ducts from segments II, III, and IV form the left hepatic
duct, and ducts from segments V, VI, VII, and VIII form the right
hepatic duct. The right and left hepatic ducts drain into the common
duct at the biliary conuence. The caudate lobe (segment 1) is typi-
cally drained by several short, small ducts into both the right and
left hepatic ducts, and the caudate branches are generally not well
seen during ERCP.
The right hepatic duct is typically formed by a right anterior sec-
toral duct (draining segments V and VIII) and a right posterior sec-
toral duct (draining segments VI and VII). The right anterior duct
has a relatively vertical and medial course, and the right posterior
duct has a more horizontal and lateral course (Fig. 42.4, see also
Figures 42.6, 42.7).11 Moving away from the conuence along the
left hepatic duct, the rst visualized branches generally drain
segment IV, which may be drained by I to III segmental ducts.
Moving further to the left, the left hepatic duct bifurcates into the
segment II and III ducts (Fig. 42.4).11
Variants of conuence anatomy are common, and the normal
conuence anatomy described above is seen in only 57% of persons.
The commonest variations involve the right anterior and posterior
ducts, and are shown in Figure 42.5. These include low drainage of
one of the right sectoral ducts into the common duct (seen in 20%)
(Fig. 42.6), a triple conuence in which the two right sectoral ducts
drain separately into the conuence (12%), and drainage of a right
sectoral duct into the left duct (6%). In about 2% a right sectoral duct
drains into the cystic duct as shown in Figure 42.7. These variants
of right duct anatomy may increase the risk of a bile duct injury
during cholecystectomy, and if identied on a preoperative ERCP,
should be conveyed to the operating surgeon. They are also impor-
tant to the endoscopist when managing hilar malignant obstruction
and evaluating postoperative biliary strictures and leaks. When a
right sectoral duct draining into the cystic duct is divided and clipped
during laparoscopic cholecystectomy, resulting in a Bismuth V
ductal injury, cholangiographic diagnosis is difcult and requires a
447
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A junction (APBJ) (arrows) and pancreas
B creatitis. A A long, dilated common channel
Fig. 42.2 Anomalous pancreaticobiliary junction (APBJ) (arrow)
with a bile duct stricture (arrowhead) caused by gallbladder cancer
in a 32-year-old with obstructive jaundice. APBJ is a risk factor for
development of gallbladder cancer. Note the absence of a bile duct
cyst.
high degree of suspicion to recognize that a right sectoral duct is not
visualized.
Variants of segmental duct anatomy also commonly occur (par-
ticularly segments IV, V, VI, and VIII) and are shown in Figure 42.8.
In one common variant, a segment IV duct drains into one of the
right sectoral ducts rather than the left duct. Ectopic drainage of the
gallbladder and cystic duct may occur, and is discussed elsewhere.11
448
Fig. 42.1 Anomalous pancreaticobiliary
divisum in a 14-year-old with recurrent pan-
is present, which contains a stone. B The
common bile duct is dilated, suggestive of a
Type I choledochal cyst.
The cystic duct may drain into the ampulla separately from the
common bile duct.12
Biliary cysts
Biliary cysts, also called choledochal cysts, are cystic dilatations of
the biliary tree. One widely adopted classication of biliary cysts was
described by Alonso-Lej and modied by Todani, and is shown in
Figure 42.9.13,14 Type I cysts, which are commonest, are dilatations
of the common bile duct, as illustrated in Figure 42.10. They can
be subdivided into Types 1a, 1b, and 1c based on the presence or
absence of an APBJ and fusiform or segmental dilatation, as shown
in Figure 42.9. Type II cysts are diverticula of the common duct.
Type III cysts involve the major papilla and may also be termed
choledochoceles or duodenal duplications (as they are often lined by
duodenal rather than biliary mucosa). Type III cysts may be further
subdivided into Type IIIA (in which the bile duct and pancreatic duct
enter the cyst proximally, and the cyst drains through a separate
distal opening into the duodenum) and Type IIIB (a diverticulum
of the intra-ampullary common channel). Endoscopically Type IIIA
cysts present as an enlargement of the ampulla and intramural
ducts, although this may not be apparent until they bulge during
contrast injection (Fig. 42.11). Type IV cysts are multiple cysts
located in both the intra- and extrahepatic ducts (IVA) or in the
extrahepatic biliary tree (IVB). Type V cysts, also called Carolis
disease, are located in the intrahepatic ducts.
Various mechanisms probably lead to formation of biliary cysts,
as reviewed elsewhere.15 In the majority of patients with extrahepatic
cysts an anomalous pancreaticobiliary junction (APBJ) is present,
and it is likely that chronic reux of pancreatic juice into the bile
duct leads to ductal dilatation and biliary mucosal inammation.
Sequelae may include pancreatitis, ductal stones, biliary dysplasia,
and cholangiocarcinoma. The risk of cholangiocarcinoma increases
with age, and may be as high as 14% in young adults16 and 50% in
older adults17 presenting with symptomatic complications of their
cysts. In one report, the increased risk of cholangiocarcinoma in
extrahepatic biliary cysts appeared to be conned to those patients
with APBJ.18 The possibility of cholangiocarcinoma should always

VIII
VII
I III
IV
VI
Fig. 42.4 Normal intrahepatic ductal anatomy. RA = right anterior
duct, RP = right posterior duct, L = left duct, numbers indicate
drained hepatic segments. Note that the right anterior duct has
a relatively vertical and medial course, and the right posterior
duct has a more horizontal and lateral course. See also Figures
42.6 and 42.7.
be considered in an adult patient with a newly diagnosed biliary cyst,
especially when an APBJ is also present.
Diagnosis of a biliary cyst requires a high degree of clinical sus-
picion, particularly for Type I cysts, which may have a similar chol-
angiographic appearance to a chronically obstructed bile duct. The
absence of obstruction distal to the dilated segment is a key diag-
nostic nding, although patients with biliary cysts may become
symptomatic only when an obstruction (due to stone or malignancy)
occurs. An APBJ, when present, is an important clue to diagnosis.
Chapter 42 Pancreas Divisum and Other Pancreaticobiliary Anomalies
Fig. 42.3 Functional division of the liver
into segments, according to Couinauds
nomenclature. From LH Blumgart and Y
Fong (eds) Surgery of the liver and biliary
tract, Saunders, 2002, Philadelphia, PA, Fig
1.7B, p. 7. Reproduced with permission.11
II
Type I cysts with associated dilatation of central intrahepatic ducts
can be distinguished from Type IVA (combined intrahepatic and
extrahepatic cysts) by the presence of a distinct change in ductal
caliber at the distal end of a true intrahepatic cyst.13 Diagnosis
of early cholangiocarcinoma in a dilated segment of bile duct is
difcult, and it is likely that intraductal ultrasound is more sensitive
than cholangiography for detection of early cholangiocarcinoma in
this setting.
Extrahepatic biliary cysts are best treated with surgical resection,
which manages local complications of the disease and decreases the
risk of subsequent malignancy. Exceptions include Type III cysts or
choledochoceles, which can be treated with endoscopic sphincterot-
omy19 (Type IIIA) or endoscopic resection20 (Type IIIB). It appears
that cancer is uncommon in choledochoceles, but has been reported.19
In some patients with Type I cysts, obstruction of a long common
channel may be best treated endoscopically.
PANCREATIC ANOMALIES
Pancreas divisum
Embryology and terminology
The embryological development of the pancreatic ductal system is
shown in Figure 42.12. The pancreas develops from dorsal and ventral
pancreatic buds that appear in the dorsal mesentery during the fth
week of embryologic development. The dorsal bud is larger, and
eventually forms the pancreatic tail, body, neck, and portions of head
including the uncinate process. The ventral bud arises together with
the bile duct, and eventually forms part of the periampullary pancre-
atic head. Growth and rotation of the duodenum brings the ventral
bud around the posterior aspect of the duodenum towards the dorsal
bud. The two buds fuse as shown in Figure 42.12. Typically the ducts
449
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B C Fig. 42.5 Common variations of biliary

ra
ra conuence anatomy. (a) typical anatomy,
ra ra lh (b) triple conuence, (c) ectopic drainage
of a right sectoral duct into common
rp lh rp lh hepatic duct, (d) ectopic drainage of a right
rp
sectoral duct into left hepatic duct, (e)
absence of a conuence, (f) ectopic drain-
rp age of the right posterior sectoral duct into
57% 12% 20% 16% 4% the cystic duct. rp = right posterior, ra =
right anterior, lh = left hepatic duct. From
C1 C2 LH Blumgart and Y Fong (eds) Surgery of
D the liver and biliary tract, Saunders, 2002,
Philadelphia, PA, Fig 1.25, p. 19. Repro-
ra duced with permission.11
ra

rp
rp lh
lh

6% 5% 1%

D1 D2

E F
ra
IV III
IV rp
ra
III lh
rp ra

II II
rp
3% 2% I 1% I 2%

E1 E2 F1

Fig. 42.6 Ectopic drainage of the right posterior duct into the
common hepatic duct. RA = right anterior duct, RP = right posterior
duct, L = left duct, CD = cystic duct.
Fig. 42.7 Ectopic drainage of the right anterior duct into the cystic
duct. RA = right anterior duct, RP = right posterior duct, L = left duct,
CD = cystic duct.

450
 Chapter 42 Pancreas Divisum and Other Pancreaticobiliary Anomalies

VII
A B
VII VIII
seg V seg VI
86%
91% VIII
VI V
VII VII VII VII VII
VIII VIII

VIII VIII VIII

5% VI 4% 10% 2% 2%
VI V V V

C D

seg VIII seg IV

VII 67% 1% 1%
VII II II II

VI III III III

VI
V 80%
V 20%

25% II 1% II 1%
II

III III III

4% II

III

Fig. 42.8 Common variations of the segmental intrahepatic ducts. A segment V, B segment VI, C segment VIII, D segment IV. From LH
Blumgart and Y Fong (eds) Surgery of the liver and biliary tract, Saunders, 2002, Philadelphia, PA, Fig 1.26, p. 19. Reproduced with
permission.11

I II III Fig. 42.9 Classication of biliary cysts.

From Todani, Watanabe, Toki, et al.
Classication of congenital biliary cystic
disease: special reference to type Ic
and IVA cysts with primary ductal
stricture. J Hepatobiliary Pancreat Surg.
2003; 10(5):340344, Figure 1. Repro-
duced with kind permission of Springer
Science and Business Media.13

IV-A IV-B V

451
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Fig. 42.10 Type I biliary cyst. An anomalous pancreaticobiliary

junction was also present.

A B C

Fig. 42.11 Choledochocele, or Type IIIA biliary cyst. A The major papilla appears normal prior to cannulation. B The intramural duct bal-
loons outward after cholangiography. C Cholangiography demonstrates a cystic dilatation of the intra-ampullary common channel.

452
 Chapter 42 Pancreas Divisum and Other Pancreaticobiliary Anomalies

Stomach Dorsal mesentery

Ventral mesentery
Dorsal pancreatic bud
Liver

Gall bladder
Foregut part of
duodenum
Ventral pancreatic bud Midgut part of
duodenum

Dorsal pancreatic bud E

Dorsal mesentery
Ventral
Level of section E pancreatic bud

Bile duct Dorsal

pancreatic bud
Gall bladder Duodenum
Duodenum
Bile duct
Dorsal mesentery

C
Dorsal mesentery

F
Level of section F Ventral pancreatic bud

Free edge of Fusion of

lesser omentum dorsal and ventral
pancreatic buds

Dorsal pancreatic bud

Duodenum

D
G
Head of Main pancreatic
Bile duct pancreas duct
Spleen
Duodenum

Level of section G
Tail of
pancreas

Head of Tail of pancreas

Opening of bile and main Accessory Body of
pancreas pancreatic ducts pancreatic duct pancreas

Fig. 42.12 Embryological development of the pancreas and pancreatic ductal system. From KL Moore and TVN Persaud, The developing
human: clinically oriented embryology, 7th edn, Saunders, 2003, Philadelphia, PA. Fig. 12.10, p. 267. Reproduced with permission.47

453
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A A
B
Fig. 42.13 Complete pancreas divisum. A Cannulation of the
major papilla demonstrates an arborizing ventral pancreatic duct
that does not cross the midline or give rise to an uncinate branch.
B Cannulation of the minor papilla demonstrates a dominant
dorsal pancreatic duct, with no communication to the ventral
pancreas. The uncinate branch arises from the dorsal pancreatic
duct.
within the buds also fuse, and the main pancreatic duct drains both
dorsal and ventral pancreas into the duodenum via the major papilla.
In pancreas divisum the ducts do not fuse or fuse incompletely, and
the dorsal pancreas duct drains the bulk of pancreatic exocrine secre-
tions into the duodenum via the minor papilla.
Pancreas divisum can be subdivided into cases in which there is
no communication between ventral and dorsal pancreas (complete
pancreas divisum) (Fig. 42.13), and cases in which there is a com-
munication between dorsal and ventral pancreatic ducts (incom-
plete pancreas divisum) (Fig. 42.14). Patients with complete divisum
usually have a small ventral pancreatic duct communicating with the
major papilla, but this cannot always be demonstrated. Incomplete
pancreas divisum can be further subdivided into cases in which a
narrow connecting duct joins ventral and dorsal pancreas, but most
pancreatic exocrine secretion must drain through the minor papilla
(dominant dorsal duct drainage), and cases in which the connect-
ing duct is large and most pancreatic drainage is likely to occur via
the major papilla.
454
B
Fig. 42.14 Incomplete pancreas divisum. A Cannulation of the
major papilla demonstrates a small ventral pancreatic duct without
lling of the dorsal duct. B Cannulation of the minor papilla dem-
onstrates a dominant dorsal pancreatic duct, with communication
to the ventral pancreatic duct and major papilla. The uncinate
branch arises from the dorsal pancreatic duct.
Diagnosis
Diagnosis of pancreas divisum can typically be made by CT, MR,
or EUS, although this may require review of diagnostic images
with attention to the possibility of divisum. EUS demonstrates the
main pancreatic duct coursing through the dorsal pancreas to the
expected location of the minor papilla, entering the duodenum
separately from the bile duct, with or without visualization of a
diminutive ventral pancreatic duct.21 When viewed from the duo-
denal bulb, the bile duct and pancreatic duct may appear to cross
each other rather than converging at the duodenal wall.22 While the
accuracy of EUS is as high as 97% using a linear echoendoscope,21
CT, MR, or EUS do not fully exclude a diminutive or collapsed duct
connecting ventral and dorsal pancreas, and they do not prove
patency of the minor papilla. ERCP thus remains the gold standard
for diagnosis.

A of the major papilla demonstrates an arbo-
B and further injection a stricture of the main
Major papilla injections in complete divisum usually demonstrate
a diminutive, arborizing ventral pancreatic duct, which does not
cross the midline (Fig. 42.13). The uncinate branch is not visualized
because the uncinate process is part of the dorsal pancreas; this is
another clue to diagnosis.23 When divisum is diagnosed on the basis
of a ventral pancreatogram alone, the possibility of pseudodivisum
(in which an obstruction of the main pancreatic duct in the head
simulates divisum) should be considered (Fig. 42.15). Pseudodivi-
sum typically causes an eccentric, rapidly tapering, or abrupt termi-
nus of the ventral pancreatic duct, while in true divisum the ventral
duct arborizes. In some clinical situations, cannulation of the minor
papilla to conrm the diagnosis of divisum may be necessary to
exclude pseudodivisum. Techniques of minor papilla cannulation
are discussed in Chapter 8.
Association with pancreatitis and role of
endoscopic treatment
The relationship of pancreas divisum to pancreatic disease is con-
troversial. An early study found pancreas divisum in up to 25% of
patients with idiopathic pancreatitis presenting for ERCP,24 but in
a subsequent larger study divisum was not seen more commonly
in ERCP patients with a history of acute, chronic, or idiopathic
pancreatitis compared to those with no history of pancreatic
disease.25 Referral bias may explanation the high incidence of
divisum seen at some centers. Pancreas divisum is a common
anomaly, with a prevalence of 5% to 10% at autopsy.26,27 Since, in
the United States, less than 0.1% of the population is admitted to
hospital yearly with pancreatitis of any cause,28 the vast majority of
persons with divisum must be asymptomatic. If pancreas divisum
causes pancreatic disease it does so only in a very small percentage
of persons with this anomaly. Factors proposed to trigger pancre-
atitis in persons with divisum include obstruction to outthe minor
papilla, and the presence of genetic abnormalities linked to
pancreatitis.
A stenotic minor papilla orice or spasm of a minor papilla
sphincter29 might result in divisum-related pancreatitis due to
relative obstruction of dorsal duct drainage. Several lines of evi-
Chapter 42 Pancreas Divisum and Other Pancreaticobiliary Anomalies
Fig. 42.15 Pseudodivisum. A Cannulation
rizing ventral pancreatic duct, suggestive of
pancreas divisum, but with some lling of an
irregular duct superiorly. B With deep can-
nulation of the ventral pancreatic duct
pancreatic duct is demonstrated. Biopsies
showed adenocarcinoma.
dence lend some support to this hypothesis. In a surgical series,
minor papilla sphincteroplasty was of most benet in divisum
patients whose minor papilla was stenotic by intra-operative assess-
ment with lacrimal probes.30 In a small randomized, prospective
study, endoscopic stenting of the minor papilla resulted in signi-
cantly better outcomes than sham therapy in patients with pancreas
divisum and at least 2 attacks of unexplained acute pancreatitis.31
The occasional nding of a Santorinicele, or a dilated terminus of
the dorsal pancreatic duct in the duodenal wall, may also be taken
as evidence of pancreatic outow obstruction in some patients with
pancreas divisum. Secretin-stimulated MRI in patients with San-
torinicele demonstrates larger pancreatic duct diameters and
delayed drainage into the duodenum compared to divisum patients
without a Santorinicele.32 Manometry of the minor papilla and
dorsal pancreatic duct has been performed, showing high dorsal
duct pressures, although normal control data is not available.29
Botulinum toxin injections into the minor papilla predicted subse-
quent response to minor papilla sphincterotomy in one small
series, presumably by treating sphincter or duodenal wall
spasm.33
The obstruction theory has led to widespread adoption of endo-
scopic therapy for idiopathic pancreatitis associated with pancreas
divisum. In addition to the randomized, controlled series described
above, uncontrolled surgical and endoscopic reports indicate that
about 70% of persons with divisum and idiopathic recurrent acute
pancreatitis will improve following sphincterotomy or stenting of the
minor papilla. These relatively good outcomes are reported in
patients with recurrent acute pancreatitis who have no evidence of
chronic pancreatitis; patients with chronic pancreatitis and divisum
are less likely to respond, and patients with chronic upper abdominal
pain and no history of pancreatitis will improve about 30% of the
time.34 A randomized trial of minor papilla sphincterotomy versus
sham therapy in divisum patients with pain alone, reported only in
abstract form, did not show a signicant benet to endoscopic
treatment.35
Difculties with the obstruction theory persist. Those patients
who respond best to endoscopic therapy do not have pancreato-
455
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
graphic evidence of chronic ductal obstruction. Endoscopists have
not developed a reliable method of assessing the minor papilla for
stenosis or spasm at the time of ERCP: normal values for minor
papilla manometry are not known, and delayed drainage has not
been studied as a predictor of outcome following endoscopic treat-
ment. The secretin ultrasound test has not proven to reliably dem-
onstrate pancreatic duct obstruction in divisum patients with
recurrent acute pancreatitis. In addition, a recent series reports
poorer response to minor papilla treatment than previously seen,
with immediate improvement in the majority but later recurrence
in most patients.36
These difculties have led investigators to look for other physio-
logic or genetic factors responsible for pancreatitis in divisum
patients. Two groups of investigators have studied cystic brosis
transmembrane receptor (CFTR) abnormalities in patients with
divisum and recurrent pancreatitis. One group assayed for 13
common CFTR mutations, and found a mutated allele signicantly
more often in divisum patients with pancreatitis (22%) than in
divisum patients without pancreatitis (0%).37 Since about 1000
CFTR mutations have been described, the authors pointed out that
the true prevalence of mutations in their subjects may well be
higher. A second group assayed CFTR function by nasal transepi-
thelial potential difference testing. They found that patients with
divisum and recurrent pancreatitis had intermediate results between
normal controls and patients with classic cystic brosis, a result that
parallels ndings in other adult onset single organ diseases linked
to CFTR dysfunction.38 These ndings suggest that an underlying
physiologic abnormality likely contributes to the development of
pancreatitis in some divisum patients. In the case of CFTR, dimin-
ished function may result in more viscous pancreatic secretions,
potentially contributing to ductal obstruction. Only 2 of 12 divisum
patients with diminished CFTR function improved after endoscopic
or surgical therapy.38
A non-invasive test that reliably identies obstruction to pancre-
atic duct outow and predicts response to minor papilla therapy is
needed. The secretin ultrasound test has been used for this purpose.
In this test intravenous secretin is administered and changes in
pancreatic duct diameter are measured using either transabdominal
ultrasound or endoscopic ultrasound. While transient dilatation of
the pancreatic duct occurs after secretin administration in normal
subjects, ductal dilatation of more than 1 mm, persisting for at
least 15 minutes after secretin administration, may indicate ductal
obstruction. In some series this technique has been a strong
predictor of clinical response to minor papilla therapy in pancreas
divisum.30,39 Other investigators, however, have not reproduced
these ndings.40 In one study, abnormal secretin ultrasound results
were seen after episodes of acute pancreatitis due to a variety
of causes.41 The major concern with this test is the potential for
false positive results; false negative results are likely to be seen
only in patients with chronic pancreatitis and some exocrine
insufciency. Secretin MRI has theoretical advantages over secretin
ultrasound, since the entire pancreatic duct is visualized, and the
timing and volume of pancreatic juice secretion into the duodenum
can be estimated.42,43 The value of secretin MRI for predicting
response to endoscopic therapy in pancreas divisum has not been
reported.
In summary, while endoscopic minor papilla therapy probably
helps a minority of patients with pancreas divisum, its role is
limited, and optimal patient selection requires further clarication.
456
A practical approach to divisum patients based on current evidence
is to discourage endoscopic therapy in patients with pain alone or
one episode of pancreatitis, and to consider endoscopic therapy in
those who have had at least two episodes of otherwise unexplained
acute pancreatitis. Even these patients face a substantial chance of
persistent or recurrent disease after endoscopic treatment. Prior to
endoscopic treatment, both secretin-stimulated MR and tests of
CFTR status may be useful, although more data is needed regard-
ing the predictive value of these tests. If testing is performed for
CFTR mutations it should include mutations associated with adult
onset, single organ disease. If the patient is having frequent epi-
sodes of pancreatitis, botulinum toxin injection of the minor papilla
may also be considered as a therapeutic trial, again based on limited
data. Patients considering minor papilla sphincterotomy or stenting
should be told about the variable outcomes reported with these
interventions, the chance of recurrent symptoms, and the possibil-
ity of complications such as post-sphincterotomy stenosis. Endo-
scopic techniques of minor papilla therapy are discussed in detail
in Chapter 15.
Incomplete pancreas divisum
As discussed earlier in this chapter, incomplete pancreas divisum is
characterized by communicating ventral and dorsal pancreatic ducts,
with a patent minor papilla and pancreatic duct drainage at both the
major papilla and the minor papilla. Despite having two routes of
pancreatic drainage, some patients with incomplete divisum may
nevertheless have symptoms related to pancreatic outow obstruc-
tion. Some have dominant dorsal duct drainage, in which the duct
connecting ventral and dorsal pancreas is diminutive, and most
pancreatic exocrine secretion must drain through the minor papilla
(Fig. 42.14). Others may have stenosis at both the major and minor
papillae.
As with complete pancreas divisum, incomplete divisum is a
common anatomic variant, and most persons with incomplete
divisum are asymptomatic. It seems likely that incomplete divisum
occasionally causes or contributes to disease, as in the reported case
of an adult with acute relapsing pancreatitis, incomplete divisum,
and a carcinoid tumor of the minor papilla causing partial obstruc-
tion.44 However the caveats discussed above regarding pancreas
divisum are germane to incomplete divisum as well. There is no
randomized, controlled trial of endoscopic therapy for incomplete
divisum, and convincing evidence of benet from endoscopic treat-
ment is lacking. Two series of endoscopic treatment for incomplete
divisum report short-term improvement in 5060%, without long-
term follow-up.45,46 Therapy was most successful in patients with
otherwise unexplained recurrent acute pancreatitis. The decision to
direct endoscopic treatment at the minor papilla, major papilla, or
both is affected by the presence or absence of dominant dorsal duct
drainage.
Annular pancreas
Annulus is the Latin word for ring, and an annular pancreas is a
ringed pancreas partially or completely encircling the duodenum. As
shown in Figure 42.16, the likely embryologic basis of annular pan-
creas is encirclement of the duodenum by the ventral pancreas
during duodenal rotation in the 5th to 8th week of fetal development.
The encircling ring of pancreas typically involves the second duode-

A B
Stomach
Bile duct
Bile duct
Bifid ventral
pancreatic bud
Fig. 42.16 Embryological basis of annular pancreas. From KL Moore and TVN Persaud, The developing human: clinically oriented embry-
ology, 7th edn, Saunders, 2003, Philadelphia, PA. Fig. 12.11, p. 268. Reproduced with permission.47
Fig. 42.17 Annular pancreas. The ventral pancreatic duct encircles
the second duodenum (arrow). This patient also has a stricture of
the main pancreatic duct, caused by pancreatic adenocarcinoma.
REFERENCES
1. Schwartz A, Birnbaum D. Roentgenologic study of the
topography of the choledocho-duodenal junction. Am J
Roentgenol Radium Ther Nucl Med 1962; 87:772776.
2. Li L, Yamataka A, Wang Y, et al. Anomalous pancreatic duct
anatomy, ectopic distal location of the papilla of Vater and
congenital biliary dilatation: a new developmental triad?
Pediatr Surg Int 2003; 19:180185.
3. Li L, Yamataka A, Yian-Xia W, et al. Ectopic distal location
of the papilla of vater in congenital biliary dilatation:
Implications for pathogenesis. J Pediatr Surg 2001; 36:
16171622.
4. Kubota T, Fujioka T, Honda S, et al. The papilla of Vater emptying
into the duodenal bulb. Report of two cases. Jpn J Med 1988;
27:7982.
5. Rajnakova A, Tan W, Goh P. Double papilla of Vater: a rare
anatomic anomaly observed in endoscopic retrograde
cholangiopancreatography. Surg Laparosc Endosc 1998;
8:345348.
6. Pereira-Lima J, Pereira-Lima L, Nestrowski M, et al.
Anomalous location of the papilla of Vater. Am J Surg 1974;
128:7174.
Chapter 42 Pancreas Divisum and Other Pancreaticobiliary Anomalies
C
Bile duct (passing dorsal to
duodenum and pancreas)
Duodenum
Annular
pancreas
Dorsal Site of
pancreatic bud duodenal
obstruction
num or apex of duodenal bulb. The commonest clinical presentation
of annular pancreas is duodenal obstruction, most often diagnosed
in infancy but occasionally presenting later in life. Cases of annular
pancreas associated with pancreatitis and pancreatic cancer have
been reported, but it is unclear if annular pancreas predisposes to
these diseases.
In adults annular pancreas is typically diagnosed by CT. EUS is
also useful for diagnosis, showing a ring of ventral pancreatic tissue
encircling the duodenum proximal to the papilla. At pancreatogra-
phy a ventral pancreatic duct is seen encircling the duodenum, as
shown in Figure 42.17.
7. Komi N, Takehara H, Kunitomo K, et al. Does the type of
anomalous arrangement of pancreaticobiliary ducts inuence the
surgery and prognosis of choledochal cyst? J Pediatr Surg 1992;
27:728731.
8. Elnemr A, Ohta T, Kayahara M, et al. Anomalous pancreaticobiliary
ductal junction without bile duct dilatation in gallbladder cancer.
Hepatogastroenterology 2001; 48:382386.
9. Hu B, Gong B, Zhou D. Association of anomalous pancreaticobiliary
ductal junction with gallbladder carcinoma in Chinese patients: an
ERCP study. Gastrointest Endosc 2003; 57:541545.
10. Couinaud C, Le Foi. Etudes anatomogiques et chirurgicales.
Masson, 1957.
11. Blumgart L, Fong Y. Surgery of the liver and biliary tract. 3rd ed.
New York: WB Saunders, 2000.
12. Watanabe A, Ohashi Y, Nagashima H. Anomalies of the bile ducts:
a case report of direct drain of the cystic duct into the papilla of
Vater. Acta Med Okayama 1983; 37:409415.
13. Todani T, Watanabe Y, Toki A, et al. Classication of congenital
biliary cystic disease: special reference to type Ic and IVA cysts
with primary ductal stricture. J Hepatobiliary Pancreat Surg 2003;
10:340344.
457
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
14. Alonso-Lej F, Rever WJ, Pessagno D. Congenital choledochal cyst,
with a report of 2, and an analysis of 94, cases. Int Abstr Surg
1959; 108:1.
15. Corso M, Topazian M. Biliary cysts. UpToDate 2005. [http://
patients.uptodate.com]
16. Voyles C, Smadja C, Shands W, et al. Carcinoma in choledochal
cysts. Age-related incidence. Arch Surg 1983; 118:986988.
17. Todani T, Watanabe Y, Toki A, et al. Carcinoma related to
choledochal cysts with internal drainage operations. Surg
Gynecol Obstet 1987; 164:6164.
18. Song H, Kim M, Myung S, et al. Choledochal cyst associated the
with anomalous union of pancreaticobiliary duct (AUPBD) has a
more grave clinical course than choledochal cyst alone. Korean J
Intern Med 1999; 14:18.
19. Ladas S, Katsogridakis I, Tassios P, et al. Choledochocele, an
overlooked diagnosis: report of 15 cases and review of 56
published reports from 1984 to 1992. Endoscopy 1995; 27:
233239.
20. Chatila R, Andersen D, Topazian M. Endoscopic resection of a
choledochocele. Gastrointest Endosc 1999; 50:578580.
21. Lai R, Freeman M, Cass O, Mallery S. Accurate diagnosis of
pancreas divisum by linear-array endoscopic ultrasonography.
Endoscopy 2004; 36:705709.
22. Bhutani MS, Hoffman BJ, Hawes RH. Diagnosis of pancreas
divisum by endoscopic ultrasonography. Endoscopy 1999;
31:167169.
23. Morgan D, Logan K, Baron T, et al. Pancreas divisum: implications
for diagnostic and therapeutic pancreatography. AJR Am J
Roentgenol 1999; 173:193198.
24. Cotton P. Congenital anomaly of pancreas divisum as cause of
obstructive pain and pancreatitis. Gut 1980; 21:105114.
25. Burtin P, Person B, Charneau J, et al. Pancreas divisum and
pancreatitis: a coincidental association? Endoscopy 1991;
23:5558.
26. Dawson W, Langman J. An anatomical-radiological study on
pancreatic duct pattern in man. Anat Rec 1961; 139:5968.
27. Berman L, Prior J, Abramow S, et al. A study of the pancreatic
duct system in man by the use of vinyl acetate casts of
postmortem preparations. Surg Gynecol Obstet 1960;
110:391403.
28. Russo M, Wei J, Thiny M, et al. Digestive and liver diseases
statistics, 2004. Gastroenterology 2004; 126:14481453.
29. Sattereld S, McCarthy J, Geenen J, et al. Clinical experience in 82
patients with pancreas divisum: preliminary results of manometry
and endoscopic therapy. Pancreas 1988; 3:248253.
30. Warshaw A, Simeone J, Schapiro R, et al. Evaluation and treatment
of the dominant dorsal duct syndrome (pancreas divisum
redened). Am J Surg 1990; 159:5964.
31. Lans J, Geenen J, Johanson J, et al. Endoscopic therapy in patients
with pancreas divisum and acute pancreatitis: a prospective,
randomized, controlled clinical trial. Gastrointest Endosc 1992;
38:430434.
458
32. Manfredi R, Brizi M, Costamagna G, et al. Pancreas divisum and
santorinicele: assessment by dynamic magnetic resonance
cholangiopancreatography during secretin stimulation. Radiol
Med (Torino) 2002; 103:5564.
33. Wehrmann T, Schmitt T, Seifert H. Endoscopic botulinum toxin
injection into the minor papilla for treatment of idiopathic
recurrent pancreatitis in patients with pancreas divisum.
Gastrointest Endosc 1999; 50:545548.
34. Lehman G, Sherman S, Nisi R, et al. Pancreas divisum: results of
minor papilla sphincterotomy. Gastrointest Endosc 1993; 39:18.
35. Sherman S, Hawes R, Nisi R, et al. Randomized controlled trial of
minor papilla sphincterotomy in patients with pancreas divisum
and pain only. Gastrointestinal Endoscopy 1994; 40:P125.
36. Gerke H, Byrne M, Stifer H, et al. Outcome of endoscopic minor
papillotomy in patients with symptomatic pancreas divisum. JOP
2004; 5:122131.
37. Choudari C, Imperiale T, Sherman S, et al. Risk of pancreatitis with
mutation of the cystic brosis gene. Am J Gastroenterol 2004;
99:13581363.
38. Gelrud A, Sheth S, Banerjee S, et al. Analysis of cystic brosis
gener product (CFTR) function in patients with pancreas divisum
and recurrent acute pancreatitis. Am J Gastroenterol 2004;
99:15571562.
39. Catalano M, Lahoti S, Alcocer E, et al. Dynamic imaging of the
pancreas using real-time endoscopic ultrasonography with
secretin stimulation. Gastrointest Endosc 1998; 48:580587.
40. Lowes J, Lees W, Cotton P. Pancreatic duct dilatation after secretin
stimulation in patients with pancreas divisum. Pancreas 1989;
4:14.
41. Cavallini G, Rigo L, Bovo P, et al. Abnormal US response of main
pancreatic duct after secretin stimulation in patients with acute
pancreatitis of different etiology. J Clin Gastroenterol 1994;
18:298303.
42. Khalid A, Peterson M, Slivka A. Secretin-stimulated magnetic
resonance pancreaticogram to assess pancreatic duct outow
obstruction in evaluation of idiopathic acute recurrent
pancreatitis: a pilot study. Dig Dis Sci 2003; 48:14751481.
43. Punwani S, Gillams A, Lees W. Non-invasive quantication of
pancreatic exocrine function using secretin-stimulated MRCP. Eur
Radiol 2002; 13:273276.
44. Singh V, Bhutani M, Draganov P. Carcinoid of the minor papilla in
incomplete pancreas divisum presenting as acute relapsing
pancreatitis. Pancreas 2003; 27:9697.
45. Kim M, Lee S, Kim C, et al. Incomplete pancreas divisum: is it
merely a normal anatomic variant without clinical implications?
Endoscopy 2001; 33:778785.
46. Jacob L, Geenen JE, Catalano M, et al. Clinical presentation and
short-term outcome of endoscopic therapy of patients with
symptomatic incomplete pancreas divisum. Gastrointestinal
Endoscopy 1999; 49:5357.
47. Moore KL, Persaud TVN. The developing human: clinically
oriented embryology, 7th edn. Saunders, Philadelphia, PA, 2003.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Chronic Pancreatitis: Stones
43 and Strictures
Jacques Deviere
INTRODUCTION
Chronic pancreatitis (CP) is a rare disease in Western countries
(incidence 210/100000/ year). It ultimately leads to irreversible
damage of the pancreas with exocrine and endocrine insufciency.
Pain is the major clinical symptom and is present early in the
course of the disease, in the majority of the cases.1,2 With the excep-
tion of the rare hereditary chronic pancreatitis, the etiology of CP
has not yet been demonstrated. Chronic alcoholism is a precipitat-
ing factor and dramatically increases the probability of CP develop-
ment but the disease can also develop in non-alcoholic subjects
without any obvious genetic background and is then dened as
idiopathic CP.
The pathophysiology of CP is still debated. Adherents of the
stone theory believe that the initiating event is protein plug forma-
tion due to a congenital lack of lithostatin.3 Proponents of the necro-
sis brosis theory, in turn, ascribe brosis and ductal stricture as
the consequence of focal inammation and necrosis.4
Pain associated with chronic pancreatitis is obviously multifacto-
rial and includes increased interstitial and intraductal pressures,
closed compartment syndrome, neural inltration, ongoing acute
pancreatitis, pseudocyst(s) and/or biliary obstruction. Elevated
intraductal pressure due to the presence of stones and/or stricture
is one of the major phenomena leading to pain in CP.58 Due to
the lack of compliance of the pancreatic gland, which is already
present in the early stages of CP, elevated intraductal pressure is
quickly associated with increased parenchymal pressure that
impairs blood ow, leading to hypoxia, release of oxygen-derived
free radicals and further stimulation of inammation with subse-
quent brosis development.8 Surgical decompression of the main
pancreatic duct has been shown to be associated with pain relief in
many patients and is associated with a decrease in intraductal and
interstitial pressures.9
Another characteristic of pain in CP is its heterogeneous pattern,
from relapsing episode to persistent pain of varying intensity which
cannot be predicted by pancreatic morphology. The initial episodes
of acute recurrent abdominal pain or acute pancreatitis often increase
and may evolve into a continuous pain syndrome requiring narcot-
ics. During the natural history of chronic pancreatitis, pain may
disappear after several years, often associated with the development
of endocrine and/or exocrine insufciency.10 This heterogeneous
pain pattern is one of the difculties we have in interpreting the
clinical results of studies reporting on the effectiveness of surgical
or endoscopic drainage for pain relief in CP.
ENDOSCOPIC TREATMENT: DUCTAL
DECOMPRESSION BY MANAGING STONES
AND STRICTURES
The goal of endotherapy in severe CP is to decompress the main
pancreatic duct (MPD) by removing stones and bypassing strictures.
Another clinical goal for MPD drainage might be to reduce the
incidence, or delay the development, of steatorrhea as a consequence
of increased ow of pancreatic juice to the duodenum. This remains
controversial.11,12
Therapeutic endoscopy offers several modalities for pancreatic
duct drainage including endoscopic pancreatic sphincterotomy,
stone removal, extracorporeal shock wave lithotripsy (ESWL), stric-
ture dilation and insertion of pancreatic stents.
Pretherapeutic planning
In addition to standard laboratory testing and plain lms of the
pancreatic area or a non-contrast CT scan for detection of pancreatic
calcications, magnetic resonance imaging is currently the best
modality for the selection of patients who may benet from endo-
scopic treatment and for pretherapeutic planning of the sequence of
interventions which may be required (Fig. 43.1).
Where preformed in conjunction with secretin stimulation, it
adequately depicts the pancreatic ductal anatomy, the presence of
peripancreatic uid collections, and possible obstruction of the bile
duct. Moreover, it can be used to quantify pancreatic exocrine func-
tion and to evaluate the short- and long-term effect of a pancreatic
ductal drainage procedure (Fig. 43.2).13,14
MPD cannulation and endoscopic
pancreatic sphincterotomy
This is the rst step of pancreatic endoscopy and provides improved
access to the MPD. Minor papilla sphincterotomy may be needed
in up to 20% of patients in cases of dominant dorsal duct anatomy
(complete or incomplete pancreas divisum, ansa pancreatica).
In a small subset of patients, pancreatic sphincterotomy (EPS)
itself may resolve papillary stenosis and allow the removal of
small oating stones. However, especially in Western countries,
when a calcication in the head of the pancreas is visible on
routine radiographs or unenhanced CT scan, and when MPD dila-
tion is recognized at MRCP, the stones are most often deeply
impacted in the ductal wall and very difcult to remove. In this
case, ESWL should be performed prior to attempted endoscopic
intervention.
459
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

B
A C

Fig. 43.1 Example of pretherapeutic planning in a patient with CP and severe pain. The plain lm A shows a dense calcication, which is
even more clearly visible on an unenhanced CT Scan B, while the dynamic MRCP after secretin injection shows an impacted stone at the
level of the genu of the pancreas, with upstream dilation, while the distal MPD is normal size C. This patient will rst undergo ESWL before
any endoscopic intervention.

B A
A

D
Position 1

B
C

Fig. 43.2 Patient with a residual prepapillary stricture after extrac-
tion of stone fragments A, treated by placement of two 8.5 Fr stents
side by side B. The comparison of S-MRCP performed before C and
after D drainage shows the decrease in MPD diameter and earlier
duodenal lling of pancreatic secretions (arrow).
Position 2
Fig. 43.3 The use of a minitome (Cook, Winston Salem, NC) with
a 0.018 J tip Terumo guidewire manipulated by the assistant using
a torque device offers the best performance for cannulating difcult
and tortuous strictures under uoroscopic control.

Biliary sphincterotomy may be performed before endoscopic pan-
creatic sphincterotomy in case of cholangitis or obstructive jaundice,
associated cholestasis, or when it is technically necessary to facilitate
access to the pancreatic duct. If such a biliary sphincterotomy is
performed, the orice of the pancreas is always located between 3
and 6 oclock on the right margin of the sphincterotomy. After pan-
creatic opacication, a hydrophilic guidewire (Terumo Inc, Japan)
can be maneuvered through the stricture or alongside the stones,
using a torque device under precise radiological control (Fig. 43.3).
Pancreatic sphincterotomy is performed over the guidewire after
deep cannulation. Section of the pancreatic sphincter is performed
under direct vision with a standard or tapered pull-type sphinctero-
tome, using, in our experience, only pure cutting current, extending
the incision to the duodenal wall. The same technique can be used
for minor papilla sphincterotomy.
Alternatively, when feasible, a small stent can be inserted into the
pancreatic duct and sphincterotomy performed with a needle knife
over the stent.
Extracorporeal shock wave lithotripsy
In multidisciplinary referral centers which use endoscopic manage-
ment to treat severe chronic pancreatitis, ESWL is performed as the
initial procedure before any endoscopic access to the pancreas.
Good-quality plain lms of the pancreatic area are taken in left and
right oblique positions to help improve uoroscopic control during
lithotripsy.
Technically, it is of major importance to use a lithotriptor with a
bidimensional x-ray focusing system and a high-power generator.
Ultrasonic localization of pancreatic stones lacks precision and ef-
cacy. When performed under general anesthesia or deep sedation,

460
 Chapter 43 Chronic Pancreatitis: Stones and Strictures

3 to 6000 shock waves can be applied at an intensity of 0.33 to
0.54 mJ/mm2, which provides a complete fragmentation of the
stones after a median of 1 session (with a maximum of 5 sessions
in our experience).15,16 The patient is in a prone position and the
shock waves generator is placed to his right when stones are in the
head of the pancreas and to his left for stones located in the body or
tail of the pancreas. The effectiveness of ESWL on calcium carbonate
pancreatic stones is much better than that obtained for biliary stones
and provides fragmentation into millimeter fragments which can
usually be easily removed by ERCP (Fig. 43.4). When the lithotriptor
is available within or close to the endoscopy unit, ESWL and thera-
peutic ERCP may be performed consecutively, even during the same
general anesthesia.
Intraductal lithotripsy
Intraductal mechanical lithotripsy has been claimed to facilitate pan-
creatic stone fragmentation before extraction.17 However, mechani-
cal lithotripsy requires encircling the stone with a Dormia basket,
which is most often impossible for impacted calcied stones
and much less successful than when utilized for stones in the bile
duct. A pulsed dye-laser lithotriptor has also been used to fragment
pancreatic stones, the laser ber inserted under visual control with
a pancreatoscope or directly applied under uoroscopic control.18,19
Laser lithotripsy has proven useful in a minority of patients. However,
intraductal fragmentation remains anecdotal and does not really
challenge ESWL which, because of efcacy, simplicity, and lack
of complications, make it the gold standard for pancreatic stone
fragmentation.
Stones extraction and dilation
After ESWL, minute stone fragments are visible within the pancre-
atic duct. If located above a stricture, dilation of this stricture using
a 4 to maximum 6 mm dilating balloon (Maxforce, Boston Scien-
tic, Boston, MA) should be performed before stone removal. Most
commonly, we initially use a small Dormia basket to remove the
fragments (Fig. 43.4). When these are visible on uoroscopy, a good
trick is to rst introduce a guidewire into the main pancreatic duct
and then the Dormia basket with either no or minimal contrast
injection. The localization of residual fragments is easier and the
basket can be manipulated at their level to trap them. Most often,
the basket is left opened in the duct, turning on its axis while gently
perfusing the duct with saline. A slightly inated balloon catheter
may be used in some cases but is of limited use in the pancreas
because sharp stone fragments frequently break the balloons. Tight
strictures may be present. Although balloon dilation is used most
frequently, bougies may be necessary and, in case of strictures in
which any catheter passage proves impossible, a Soehendra
stent retriever (8.5 Fr) can usually be rotated through the stricture
to create the necessary passage for placement of a dilation balloon
(Fig. 43.5).
If multiple sessions of endoscopy are necessary for stone frag-
ment removal or fragmentation of multiple stones, a nasopancreatic
catheter (NPC) is left in place for drainage between the sessions.
This may decrease the risk of acute pancreatitis due to fragment
impaction.20 In addition, placement of a NPC can also be used as a
clinical predictor for the requirement of pancreatic stenting when
the presence of a stricture is not obvious. Indeed, if a patient toler-

A B
C

D
F
E

Fig. 43.4 Same patient as Figure 43.1. Successful fragmentation (A versus B) after ESWL is illustrated by a decrease in radiological density,
an increase of the stone surface area and a heterogeneity of the stones (powder-like material). After fragmentation, a pancreatic sphinc-
terotomy is performed C, a guidewire is inserted in the pancreatic duct D and a small Dormia basket E is maneuvered alongside the
guidewire to remove the stone fragments. At the end of the procedure, a nasopancreatic catheter is left in place F.

461
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

A B C

Fig. 43.5 Usual techniques of dilation: A 4 cm 6 mm Maxforce balloon is inated through the stricture of which the imprint is visible
at the beginning of ination. B A biliary bougie is passed over a guidewire. C In extremely tight strictures, an 8.5 F Soehendra stent retriever
is rotated through the stricture in order to create the room necessary for insertion of a balloon. In this case, after dilation, the stent retriever
should be removed while turning it counterclockwise in order to avoid the risk of displacing the guidewire.

ates perfusion of a NPC without pain, the absence of a signicant

A B
stricture of the pancreatic duct is likely and further stenting may be
avoided or postponed. In contrast, if perfusion of NPC is painful,
the catheter should be placed to gravity drainage and further stone
extraction or stenting considered.

Stenting
When a stricture is present, adequate outow from the pancreas to
the duodenum must be achieved by placement of a stent. In contrast
to stenting to prevent acute pancreatitis in high-risk patients, in
severe chronic pancreatitis only large (8.5 or 10 Fr) stents are used. C
D
Stent lengths are adapted to the morphology of the pancreas and are
placed through the stricture after complete removal of stone frag-
ments. Stents are usually replaced every 6 months for a period of 2
years, or on demand in case of relapsing symptoms. The tendency
now is to place multiple stents side by side. Two 8.5 Fr stents can
usually be easily placed after a dilation to 6 mm. The number of
stents can be further increased over successive exchanges if the
morphology of the pancreas allows it (Fig. 43.2). This policy has been
recently recommended as a means to reduce the duration of stenting
needed to provide effective calibration and prolonged symptom Fig. 43.6 Patient with CP and daily pain. Left (A,C): plain lm and
relief.21 The technique of multiple stent placement is easier when MRCP at admission, showing a single impacted stone in the head
of pancreas with upstream dilation of the MPD. Right (B,D): after a
two guidewires are placed initially through the stricture and the two single session of ESWL, without any endoscopic therapy, the major
stents are implanted successively over the two guidewires. This tech- part of the stone has disappeared B, the size of the MPD has
nique avoids the need to recannulate the pancreas and bypass a tight decreased D and duodenal lling, at the same time-point after
stricture with a guidewire when the rst stent is already in place. secretin injection, is much more obvious.
The use of the new Fusion systemTM (Cook Endoscopy, Winston
Salem, NC) allows intraductal exchange and placement of multiple
stents side by side without losing access with a single guidewire. cases, and both stones and stricture in 32%.11 Numerous reports
This system has become our preferred technique for placement of have shown that ESWL is a low-risk and technically successful
multiple plastic stents in the pancreas (Fig. 43.6). method with fragmentation rate of up to 100%. Nevertheless, com-
plete clearance of the main pancreatic duct is only achieved in 44 to
Technical results 75% of the cases (Table 43.1). A recent meta-analysis showed that
The majority of patients with severe painful chronic pancreatitis the use of ESWL was clearly associated with ductal clearance and
have stones which require shock wave lithotripsy. In our experience, pain relief.22 Technical success of endoscopic ductal drainage,
ESWL is required in 2/3 of the patients who are referred for treat- however, is generally dened as a decrease in the diameter of main
ment, whereas in a multicentric study of more than 1000 patients pancreatic duct with or without complete ductal stone clearance.12
pancreatic obstruction was due to the presence of obstructive stones Using this denition, technical successes have been noted in 54
alone in 17%, stricture of the main pancreatic duct in 47% of the 99% of the cases in the largest series published to date.11,15,16,2331

462
 Chapter 43 Chronic Pancreatitis: Stones and Strictures

Complete
Complete or partial Need for Mean
No. of Fragmentation clearance pain relief surgery follow-up
Study (ref) Year patients (%) (%) (%) (%) (months)
ESWL and endotherapy
Delhaye (15) 1992 123 99 59 85 8 14
Schneider (23) 1994 50 86 60 62 12 20
Costamagna, et al. (27) 1997 35 100 74 72 3 27
Adamek, et al. (28) 1999 80 54 ND 76 10 40
Brand, et al. (29) 2000 48 60 44 82 4 7
Farnbacher, et al. (26) 2002 125 85 64 48a 13 29
Kozarek, et al. (30) 2002 40 100 ND 80 20 30
Inui, et al. (16) 2005 470 92 73 69a 4 44
ESWL alone
Ohara (31) 1996 32 100 75 86 3 44

Table 43.1 Results of extracorporeal shock wave lithotripsy (ESWL) and endotherapy for chronic calcic pancreatitis
a
Patients with complete pain relief during follow-up.

Number of Technical Clinical

Series (ref) patients success (%) Associated factor success (%) Associated factor (ref)
Short-term follow-up <2 y (15) 123 90 None 85 Decrease in MPD diameter
Medium-term follow-up 2y5y 53996 5499 Availability of ESWL (24) 4884 Short duration of disease (24,33)
(11,2426,28,3335) Single stone (28) Low frequency of pain (24)
Absence of MPD stricture (24)
Total 1557 86 65
Long-term follow-up >5 y (12) 56 86 None 66 Short duration of disease
No current smoking

Table 43.2 Predictive factors of technical and clinical success in published series of more than 50 chronic pancreatitis patients treated
by ESWL and endoscopic pancreatic ductal drainage

Dumonceau et al. identied ESWL as the only independent factor
associated with technical success. In most reports, successful frag-
mentation and stone clearance have not been correlated with the
initial size or the number of main pancreatic duct stones.24 In one
recent study, stone clearance was negatively correlated with the pres-
ence of a stricture.32
Clinical results
Early pain relief after endoscopic pancreatic duct drainage is experi-
enced by 8294% of patients and can be expected when drainage of
the main pancreatic duct is adequate.23,24,26 Medium-term clinical
improvement has been observed in 4884% of the patients after a
mean follow-up period of 25 years and an independent predictor
for pain relapse during follow-up has been reported as a high fre-
quency of pain attacks.24,36 Other factors include a long duration of
disease before treatment and the presence of a stenosis of the main
pancreatic duct (Table 43.2).24,32,33 In our series, reporting the longest
follow-up to date (14,4 years), a good clinical outcome was recorded
in 2/3 of the patients and was associated with a short duration of
disease before treatment and cessation of smoking.12
This suggests that ESWL and/or endoscopic therapy should be
initiated as early as possible in the course of CP since it increases
the probability of long-term benet.
In most series, recurrent pain attacks during the follow-up period
were related to stone fragment migration or recurrent, progressive
stricture of the main pancreatic duct, or pancreatic stent obstruction
or dislodgement. Interestingly, re-treatments are usually easier than
initial treatment and remain very effective in controlling pain.26 The
latter is dramatically different from surgery which is associated with
increased morbidity when it has to be repeated.
Dominant strictures of the main pancreatic duct are often the
indication for insertion of a pancreatic stent, required in 5060% of
the patients with severe chronic pancreatitis.11,26,27,33,34 The problem
with stenting is that even large stents may become occluded and
result in relapsing symptoms and even infectious pancreatitis. They
can be exchanged on a regular basis or on demand (in patients with
recurrence of pain and recurrent dilation of the main pancreatic
duct).
This latter strategy results in stent replacement at a mean period
of 812 months, likely because even an occluded stent may serve as
a wick to allow pancreatic juice to ow into the duodenum.33,36 Stent-
ing for a short period of time (6 months) is not enough to provide
adequate calibration of a stricture and long-term pain relief.37
However, since the presence of a stent may be associated with the
need for repeated endoscopy, two large studies have assessed long-
term outcomes in patients following pancreatic stent removal. In
one study, stents could be removed from 49 of 93 patients after a
mean stenting period of 16 months and 73% of these patients
remained pain-free without a stent during a mean follow-up of 3.8
years.33 In the second study, following a median stenting duration

463
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
of 23 months, 62% of the patients maintained satisfactory pain
control without pancreatic stent replacement during a median
follow-up of 27 months.36 The only signicant predictive factor of
the need of pancreatic restenting within one year of stent removal
was the presence of pancreas divisum. Interestingly, the majority of
patients with pain recurrence requiring restenting relapsed during
the rst year after removal of the stent and almost all of them
relapsed within two years. Therefore, if a patient remains stable
during the rst year after stent removal, subsequent relapse and
need for restenting is unlikely. It is interesting to note that if a
patient has frequent pain relapses due to stent clogging but good
control of pain when a patent stent is in place, the decision for elec-
tive pancreaticojejunostomy can be considered as there is a high
chance for good outcome after surgical pancreatic decompression.
This further reinforces the idea that endotherapy should be per-
formed within multidisciplinary teams. From the latter perspective,
data are clear that pancreatic stenting does not complicate subse-
quent surgical procedures when they are indicated.38
In an attempt to decrease the duration of stenting, Costamagnas
group was the rst to propose placement of multiple stents into the
pancreas for a period of 612 months before removal.21 These endos-
copists inserted as many stents as possible, contingent upon the
tightness of the stricture and the pancreatic duct diameter (a median
of 3). After removal of the stents, 84% of 19 patients remained
asymptomatic for a mean follow-up of 38 months. Additional infor-
mation gleaned from these latter patients is that the majority of them
have all stents occluded at time of removal, yet no pain relapse. This
suggests that pancreatic juice can still ow through the space located
between the plastic tubes and that a more aggressive approach of
dominant pancreatic strictures might decrease the duration of stent-
ing required to provide long-term pain relief.
Another area needing further investigation in patients with distal
strictures, is the long-term efcacy of an iatrogenic pancreaticoduo-
denal stula. The latter can then be maintained by placement of one
or two stents.39 This technique, can be also applied to patients with
a complete obstruction of the main pancreatic duct and has the
potential advantage of creating a true pancreaticoduodenal stula
that might not be dependent on stent patency. The stula would act
more as a wick comparable to the communication created after
drainage of a pancreatic pseudocyst associated with a disconnected
main pancreatic duct.
Impact of pancreatic duct drainage on pancreatic
endocrine and exocrine functions
In contrast to the multicentric study published by Rsch et al., our
long-term outcome study suggests that endoscopic ductal drainage
including ESWL can delay the development of clinical steatorrhea
by about 10 years when compared to the natural history of chronic
pancreatitis patients.1012
The risk of new onset steatorrhea was higher in alcoholic patients
and also signicantly associated with long duration of symptomatic
ductal obstruction before treatment, suggesting that early ductal
decompression in the course of the disease may be benecial.
However, in the setting of painless steatorrhea endotherapy does not
appear warranted at this time.
Moreover, previous studies have conrmed that the development
of diabetes mellitus was not prevented by pancreatic duct drainage.
Rather it appeared to be a consequence of ongoing alcohol abuse
suggesting that only the exocrine pancreatic function may be depen-
dent on early relief of ductal obstruction.11,28,40
464
DISCUSSION
There are now multiple large series with long follow-up which dem-
onstrate that pancreatic endotherapy for painful chronic pancreatitis
is effective and should be rst-line interventional management.
There are still criticisms, however, and papers which suggest that
this treatment is experimental and should be conducted only in the
setting of clinical trials.41
One of the criticisms is the absence of sham-controlled trial,
something very difcult to undertake in a referral center where
patients with severe pain are referred for treatment. Over a 3-year
period, we were able to include only 8 patients in such a trial cur-
rently ongoing in our institution. This number represents only 5%
of the new patients treated. A recent publication has reported the
results of pancreatic duct drainage in a subgroup of patients with
continuous pain and a dilated duct.42 Authors observed 100% com-
plete pain relief and analgesic discontinuation after endotherapy,
suggesting that in this particular group, MPD drainage is at least
better than a placebo.
Another discussion centers around whether endotherapy chal-
lenges surgery in the treatment of pain. The rst published, ran-
domized trial comparing endotherapy with surgery disclosed a
similar efcacy for short-term pain relief while surgery was better
for mid-term pain control.43 However, surgery included resection
in 80% of patients, a procedure difcult to compare with ductal
drainage. Moreover, ESWL was not available for endoscopic treat-
ment (which, in our experience, would render treatment not feasi-
ble in 44% of cases) and repeated endotherapy was not available in
case of symptom recurrence.15 Multiple previous studies have dem-
onstrated the need for such retreatment during the early period
following initially directed endoscopic decompression.15,26 Another
trial, recently published,44 reported better efcacy at 2 years follow-
ing surgery (75% vs 32%) for pain relief. However, in this study,
stenting in case of stricture was only maintained for a median of
6 months, a policy previously shown to be ineffective for long-term
pain relief 37 and the vast majority of patients had tight strictures.
Randomized studies will be helpful to better establish the efcacy
of this therapy.45 Moreover, standardization of the endoscopic
approach (ESWL, duration of stenting in case of strictures, multi-
disciplinary units with the ability to perform additional interven-
tions in the early phase of endotherapy) is required. Measurement
of outcomes is also important as most pain relapses following
endotherapy occur within one year after initial treatment.12 After
surgery, in turn, relapse commonly occurs after a median of 67
years.10
A nal issue is to dene whether or not we are doing too many
procedures in these patients. We have learned from Japanese series
that ESWL alone, without any endoscopic therapy, is able to relieve
pain in many patients.16,31 This is a likely consequence of minute
fragmentation of calcium carbonate stones, especially in patients
with a single stone located in the head of the pancreas.
We have recently conducted a RCT in this particular group of
patients, comparing ESWL alone as an initial treatment vs ESWL
plus endotherapy.46 We observed similar clinical benets and only
31% of the patients from the ESWL alone required endotherapy over
a median 4 years follow-up (Fig. 43.6).
This is another argument to support the proposition that ESWL
remains the cornerstone technology to manage these patients and it
has proven to be an incentive to critically look at even less invasive
procedures.

REFERENCES
1. Amman RW, Akovbiantz A, Largiader F, et al. Course and
outcome of chronic pancreatitis. Gastroenterology 1984;
86:820828.
2. Sahel J, Sarles H. Chronic calcifying pancreatitis and obstructive
pancreatitistwo entities. In: Gyr KE, Singer MV, Sarles H (eds)
Pancreatitis, concepts and classication. Excepta Medica,
Amsterdam; 1984:4749.
3. Multigner L, Sarles H, Lombardo D, et al. Pancreatic stone protein
II. Implication in stone formation during the course of chronic
alcoholic pancreatitis. Gastroenterology 1985; 89:387391.
4. Amman RW, Heigz PU, Kloppel G. Course of alcoholic chronic
pancreatitis. A prospective clinicomorphological long-term study.
Gastroenterology 1996; 111:224231.
5. Madsen P, Winkler K. The intraductal pancreatic pressure in
chronic obstructive pancreatitis. Scand J Gastroenterol 1982;
17:553554.
6. Bradley EL, 3rd. Pancreatic duct pressure in chronic pancreatitis.
Am J Surg 1982; 144:313316.
7. Ebbehoj N, Borly L, Bulow J, et al. Pancreatic tissue uid pressure
in chronic pancreatitis. Relation to pain, morphology, and
function. Scand J Gastroenterol 1990; 25:10461051.
8. Karanjia ND, Widdison AL, Leung F, et al. Compartment syndrome
in experimental chronic obstructive pancreatitis: effect of
decompressing the main pancreatic duct. Br J Surg 1994; 81:
259264.
9. Ebbehoj N, Borly L, Bulow J, et al. Evaluation of pancreatic tissue
uid pressure and pain in chronic pancreatitis. A longitudinal
study. Scand J Gastroenterol 1990; 25:462466.
10. Ammann RW, Muellhaupt B. The natural history of pain in
alcoholic chronic pancreatitis. Gastroenterology 1999; 116:
11321140.
11. Rosch T, Daniel S, Scholz M, et al. Endoscopic treatment of
chronic pancreatitis: a multicenter study of 1000 patients with
long-term follow-up. Endoscopy 2002; 34:765771.
12. Delhaye M, Arvanitakis M, Verset G, et al. Long-term clinical
outcome after endoscopic pancreatic ductal drainage for patient
with painful chronic pancreatitis. Clin Gastroenterol Hepatol 2004;
2:10961106.
13. Matos C, Metens T, Devire J, et al. Pancreatic duct: morphology
and functional evaluation with dynamic MR pancreatography after
secretin stimulation. Radiology 1997; 203:435441.
14. Bali MA, Sztanticz A, Metens T, et al. Quantication of pancreatic
exocrine function with secretin and magnetic resonance
cholangiopancreatography: normal values and short term effects
of pancreatic duct drainage procedures in chronic pancreatitis:
initial results. Eur Radiol 2005; 15:21102121.
15. Delhaye M, Vandermeeren A, Baize M, et al. Extracorporeal shock
wave lithotripsy of pancreatic calculi. Gastroenterology 1992;
102:610620.
16. Inui K, Tazuma S, Yamaguchi T, et al. Treatment of pancreatic
stones with extracorporeal shock wave lithotripsy: results of a
multicenter survey. Pancreas 2005; 30:2630.
17. Freeman ML. Mechanical lithotripsy of pancreatic duct stones.
Gastrointest Endosc 1996; 44:833835.
18. Neuhaus H, Hoffman W, Classen M. Laser lithotripsy of pancreatic
and biliary stones via 34 and 37 miniscopes: rst clinical results.
Endoscopy 1992; 24:208214.
19. Hiray T, Goto H, Hirooka Y, et al. Pilot study of pancreatoscopic
lithotripsy using a 5-fr instrument: selected patients may benet.
Endoscopy 2004; 36:212216.
20. Hookey LC, Tinto RR, Delhaye M, et al. Risk factors for pancreatitis
after pancreatic sphincterotomy: a review of 572 cases.
Endoscopy 2006; 38:670676.
Chapter 43 Chronic Pancreatitis: Stones and Strictures
21. Costamagna G, Bulajic M, Tringali A, et al. Multiple stenting of
refractory pancreatic duct strictures in severe chronic pancreatitis:
long-term results. Endoscopy 2006; 38:254259.
22. Guda NM, Partington S, Freeman ML. Extracorporeal shock wave
lithotripsy in the management of chronic calcic pancreatitis: a
meta-analysis. JOP 2005; 6:612.
23. Schneider HT, May A, Benninger J, et al. Piezoelectric shock wave
lithotripsy of pancreatic duct stones. Am J Gastroenterol 1994; 89:
20422048.
24. Dumonceau JM, Devire J, Le Moine O, et al. Endoscopic
pancreatic drainage in chronic pancreatitis associated with ductal
stones: long-term results. Gastrointest Endosc 1996; 43:547555.
25. Smits ME, Rauws EAJ, Tytgat GNJ, et al. Endoscopic treatment of
pancreatic stones in patients with chronic pancreatitis. Gastrointest
Endosc 1996; 43:556560.
26. Farnbacher MJ, Schoen C, Rabenstein T, et al. Pancreatic duct
stones in chronic pancreatitis: criteria for treatment intensity and
success. Gastrointest Endosc 2002; 56:501506.
27. Costamagna G, Gabbrielli A, Multignani M, et al. Extracorporeal
shock wave lithotripsy of pancreatic stones in chronic pancreatitis:
immediate and medium-term results. Gastrointest Endosc 1997;
46:231236.
28. Adamek HE, Jakobs R, Buttmann A, et al. Long-term follow-up
of patients with chronic pancreatitis and pancreatic stones
treated with extracorporeal shock wave lithotripsy. Gut 1999;
45:402405.
29. Brand B, Kahl M, Sidhu S, et al. Prospective evaluation of
morphology, function and quality of life after extracorporeal
shockwave lithotripsy and endoscopic treatment of chronic
calcic pancreatitis. Am J Gastroenterol 2000; 95:34283438.
30. Kozarek RA, Brandabur JJ, Ball TJ, et al. Clinical outcomes in
patients who undergo extracorporeal shock wave lithotripsy for
chronic calcic pancreatitis. Gastrointest Endosc 2002; 56:
496500.
31. Ohara H, Hoshino M, Hayakawa T, et al. Single application
extracorporeal shock wave lithotripsy is the rst choice for
patients with pancreatic duct stones. Am J Gastroenterol 1996;
91:13881394.
32. Tadenuma H, Ishihara T, Yamaguchi T, et al. Long-term results
of extracorporeal shock wave lithotripsy and endoscopic
therapy for pancreatic stones. Clin Gastroenterol Hepatol 2005;
3:11281135.
33. Binmoeller KF, Jue P, Seifert H, et al. Endoscopic pancreatic stent
drainage in chronic pancreatitis and a dominant stricture: long-
term results. Endoscopy 1995; 27:638644.
34. Cremer M, Devire J, Delhaye M, et al. Stenting in severe chronic
pancreatitis: results of medium-term follow-up in 76 patients.
Endoscopy 1991; 23:171176.
35. Sherman S, Lehman GA, Hawes RH, et al. Pancreatic ductal stones:
frequency of successful endoscopic removal and improvement in
symptoms. Gastrointest Endosc 1991; 37:511517.
36. Eleftheriadis N, Dinu F, Delhaye M, et al. Long-term outcome after
pancreatic stenting in severe chronic pancreatitis. Endoscopy
2005; 37:223230.
37. Ponchon T, Bory RM, Hedelius F, et al. Endoscopic stenting for
pain relief in chronic pancreatitis: results of a standardized
protocol. Gastrointest Endosc 1995; 42:452456.
38. Boerma D, van Gulik TM, Rauws EAJ, et al. Outcome of
pancreaticojejunostomy after previous endoscopic stenting in
patients with chronic pancreatitis. Eur J Surg 2002; 168:223228.
39. Tessier G, Bories E, Arvanitakis M, et al. EUS guided
pancreaticogastrostomy and pancreaticobulbostomy for the
treatment of pain in patients with pancreatic ductal dilatation
465
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
inaccessible for transpapillary endoscopic therapy. Gastrointest
Endosc 2007; 65:233241.
40. Malka D, Hammel P, Sauvanet A, et al. Risk factors for diabetes
mellitus in chroinc pancreatitis. Gastroenterology 2000; 119:
13241332.
41. Mel Wilcox C. Endoscopic therapy for pain in chronic pancreatitis:
is it time for the naysayers to throw in the towel? Gastrointest
Endosc 2005; 61:582586.
42. Gabbrielli A, Pandol M, Mutignani M, et al. Efcacy of main
pancreatic-duct endoscopic drainage in patients with chronic
pancreatitis, continuous pain, and dilated duct. Gastrointest
Endosc 2005; 61:576581.
466
43. Dite P, Ruzicka M, Zboril V, et al. A prospective, randomized trial
comparing endoscopic and surgical therapy for chronic
pancreatitis. Endoscopy 2003; 35:553558.
44. Cahen D, Gouma DJ, Nio Y, et al. Endoscopic versus surgical
drainage of the pancreatic duct in chronic pancreatitis. N Engl J
Med 2007; 356:676684.
45. Elta GH. Is there a role for the endoscopic treatment of pain in
chronic pancreatitis? N Engl J Med 2007; 356:727729.
46. Dumonceau JM, Costamagna G, Tringali A, et al. Treatment for
painful calcied chronic pancreatitis: extracorporeal shock wave
lithotripsy versus endoscopic treatment: a randomized controlled
trial. Gut 2007; 56:545552.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Suspected IPMN and Cystic Lesions
44 of the Pancreas
William R. Brugge
INTRODUCTION
Cystic lesions of the pancreas consist of a spectrum of benign, pre-
malignant, and variably invasive malignancies. In the past, cystic
neoplasms of the pancreas were thought to be relatively rare, but the
widespread use of cross-sectional imaging has dramatically increased
the frequency of diagnosis. Although the vast majority of the lesions
are discovered incidentally, large or invasive lesions may produce
sufcient symptoms to cause the patient to seek medical attention.
Cystic neoplasms are often confused or misdiagnosed as pseudo-
cysts. Alternatively, peripancreatic collections of inammatory uid
may morphologically mimic cystic neoplasms. Furthermore, the
presenting symptoms of pseudocysts may be identical to the symp-
toms associated with cystic neoplasms.
Cystic neoplasms of the pancreas are traditionally organized by
the type of lining epithelium since this feature dominates the risk
of malignancy and management (Table 44.1).1 There are three types
of mucinous lesions, benign mucinous cystadenomas, malignant
mucinous cystic lesions, and intraductal papillary mucinous neo-
plasms (IPMNs). The non-mucinous lesions include serous cystad-
enomas, cystic endocrine tumors and other rare lesions.
PREVALENCE
The prevalence of pancreatic cysts has been examined with autopsy
examinations of the pancreas in adults without known pancreatic
disease. The prevalence of pancreatic cysts found at autopsies in
Japan was approximately 73 of 300 (24.3%) cases.2 The prevalence
of cysts increased with increasing age of the patient. The cysts were
located throughout the pancreatic parenchyma and were not related
to chronic pancreatitis. The epithelium of the cysts displayed a
spectrum of neoplastic change, including atypical hyperplasia
(16.4%) to carcinoma in situ (3.4%). The malignancy simulated
pancreatic adenocarcinoma and arose from the epithelium lining
the cyst.
The prevalence of pancreatic cysts in the United States has been
estimated in patients undergoing MRI for a variety of medical prob-
lems.3 This study revealed that approximately 1520% of 1444
patients had at least one pancreatic cyst. Older patients are more
likely to have a cyst than younger patients. Screening abdominal
ultrasound in a younger population revealed that (0.21%) of 130,951
adults had a pancreatic cystic lesion.4
CLINICAL EPIDEMIOLOGY
Mucinous cystic neoplasms account for approximately 25% of
all exocrine pancreatic tumors and are the more common type of
cystadenoma. Women are affected far more commonly than men
(9 : 1 ratio), with a mean age at diagnosis in the fth decade.
Intraductal papillary mucinous neoplasms share many of the
features of mucinous cystic neoplasms. Their true incidence is
uncertain, but estimates range from 1% to 8% of all pancreatic
tumors. IPMNs affect men and women equally or men predomi-
nantly, depending on the reported series, and they tend to occur in
an older age group than MCNs.
Serous cystadenomas have been estimated to account for about
25% of all cystic neoplasms of the pancreas.5 Estimates of the inci-
dence and prevalence vary. Using surgical pathology studies, it has
been estimated that serous cystadenomas account for about 12%
of all exocrine pancreatic neoplasms.
Serous cystadenomas occur only in adults with a median age in
the sixth or seventh decade. The vast majority of patients with serous
cystadenomas are female.6 Traditionally about 1/2 of tumors are
discovered as incidental ndings during abdominal imaging or
surgery or at autopsy.
RISK FACTORS FOR CYSTIC LESIONS
In the vast majority of patients with a cystic lesion, no risk factor is
apparent. Von HippelLindau (VHL) syndrome is the best-described
inherited disorder associated with cystic lesions.7 In the largest
series to date, pancreatic involvement was observed in 122/158
patients (77.2%) and included true cysts (91.1%), serous cystadeno-
mas (12.3%), neuroendocrine tumors (12.3%), or combined lesions
(11.5%).
PATHOGENESIS
The pathogenesis of cystic neoplasms of the pancreas is poorly
understood. Serous cystadenomas are strongly associated with muta-
tions of the VHL gene, located on chromosome 3p25.8 The VHL
gene is likely to play an important role in the pathogenesis of spo-
radic serous cystadenomas. In one study, 70% of the sporadic serous
cystadenomas studied demonstrated loss of heterozygosity (LOH) at
3p25 with a VHL gene mutation in the remaining allele.9 The muta-
tions in the VHL gene probably affect most commonly the centro-
acinar cell and result in hamartomatous proliferation of these small
cuboidal cells. The expression of keratin in clear epithelial cells
resembles that in ductal and/or centroacinar cells and is most likely
responsible for the bro-collagenous stroma.10
The pathogenesis of mucinous cystic neoplasms and intraductal
papillary mucinous neoplasms (IPMN) is likely very different com-
pared to serous cystadenomas. K-ras mutations are present only in
mucinous cystic neoplasms but not in serous microcystic adenomas.
467
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Tumor type Gender Age Morphology Type of epithelium Risk of malignancy

Mucinous cystadenoma Female Middle-aged Unilocular Mucinous Moderate
Mucinous cystic neoplasm Female Middle-aged Associated mass Malignant mucinous High
Intraductal papillary mucinous Mixed Elderly Unilocular, septated, Papillary mucinous Moderate
tumor associated dilated ducts
Serous cystadenoma Female Middle-aged Microcystic Serous (PAS positive Low
for glycogen)
Cystic endocrine tumor Mixed Middle-aged Associated mass Endocrine Low
Solid cystic pseudopapillary tumor Female Young Mixed solid and cystic Endocrine-like Low

Table 44.1

In addition, LOH at 3p25, the chromosomal location of VHL gene,

was present in 57% (8/14) of serous microcystic adenomas com-
pared with 17% (2/12) of mucinous cystic neoplasms in one study.11
Mucinous cystic neoplasms frequently contain mutations of the K-
ras oncogene and p53 tumor suppressor gene, and the frequency of
these mutations increases with increasing degrees of dysplasia in
the neoplasm.12 The frequency of k-ras mutation in mucinous cystic
neoplasms is linearly related to the grades of atypia.13 However, the
degree of atypia in IPMT does not seem to correlate with the pres-
ence of k-ras mutations. LOH of the p16 gene was observed with
increasing degrees of histological atypia in IPMN, whereas LOH of
the p53 gene was seen only in invasive carcinomas. The distribution
of loss of heterozygosity in 9p21(p16) and 17p13(p53) of IPMT
lesions is mostly clonal, without the presence of genetic alterations.
The identical genetic statuses in the precursor lesions are consistent
with the presence of clonal progression during the development of
this tumor.14

PATHOLOGY

Serous cystadenomas (Fig. 44.1)

Serous cystadenomas are benign, solitary, cystic tumors that arise
from centroacinar cells. Although the majority of serous cystadeno-
mas are microcystic, there are two other variants based on growth
pattern: macrocystic and solid. Microcystic serous cystadenomas are
composed of multiple small thin-walled cysts with a honeycomb-like
appearance on cross section. Microcystic serous cystadenomas may
grow to a large diameter over the long term and the large lesions
often have a brotic or calcied central scar. Macrocystic serous
cystadenomas are composed of far fewer cysts, and the diameter of
each cyst varies from microcystic to large cavities.15 The presence of
discrete, large cystic cavities mimics the appearance of mucinous
lesions. However, the cyst uid from serous cystadenomas is non-
viscous and may contain blood as a result of the vascular nature of
the lesions.
All serous lesions contain a prominent brous stroma, glycogen-
rich epithelial cells, endothelial and smooth muscle cells.16 Ultra-
structurally, brocollagenous stroma is composed of myobroblasts
and endothelial cells embedded in thick collagen bundles.10 Estrogen
and progesterone receptors are not present.
Mucinous cystic neoplasms
Mucinous cystic neoplasms (MCNs) are composed of discrete indi-
vidual locules that vary in diameter. MCNs are lined by mucin-
producing cells in a columnar epithelium. The World Health
Organization classication catalogues MCNs into three types, based
Fig. 44.1 Gross photograph of a serous cystadenoma.
on the degree of epithelial dysplasia: benign, borderline, and malig-
nant. The degree of atypia of the tumor is classied according to the
most advanced degree of dysplasia/carcinoma present.
Mucinous cystic neoplasms of the pancreas often contain a
unique, highly cellular (so-called ovarian) stroma that often con-
tains estrogen and progesterone receptors. It occurs almost exclu-
sively in female patients, although rare cases of MCNs with ovarian
stroma in male patients have been encountered. Many authorities
have restricted the very denition of MCNs to include only those
cystic mucinous tumors that contain ovarian stroma. The cyst uid
from MCNs is often viscous and clear.
IPMNs are similar to MCNs in that they are cystic tumors that
secrete mucin. However, IPMNs are characterized by a unique papil-
lary epithelium and arise from ductal epithelium. The presence of a
papillary neoplasm and obstructing mucus causes the pancreatic
duct to dilate. The degree of ductal ectasia produced varies with the
degree of mucin production, but duct dilation great enough to be
seen on imaging studies or gross pathologic examination is a diag-
nostic feature of the diagnosis. Mucin production may be so exces-
sive that mucin will be spontaneously secreted out of the ampulla.
The degree of dysplasia exhibited by the epithelium may range from

468

mild to moderate to severe (carcinoma in situ), and the foci of early
malignancy may be evident by the presence of mural nodules.17 The
solid malignancies that arise from IPMN are more likely to have
papillary features, as compared to typical pancreatic malignancies
that arise from the main pancreatic duct.18
Cystic endocrine neoplasms
Cystic neoplasms are composed of neuroendocrine tissue. The cystic
lesions are not true cysts since they are composed of centralized
tumor necrosis. The lesions are rare and comprise 0.54% of all
primary pancreatic neoplasms. The classic neuroendocrine cystic
tumor is populated with a characteristic small, granular population
of cells that are stainable for immunoreactive hormones, chromo-
granin and synaptophysin.19 It is rare for the cystic endocrine tumors
to produce sufcient hormones to be clinically active. Cystic endo-
crine tumors are seen in association with Von HippelLindau syn-
drome.20 A related cystic lesion, the solid pseudopapillary tumor,
contains cells with neuroendocrine characteristics.21 The tumor
often contains areas of hemorrhage and necrosis as well as cystic
components. These neoplasms exhibit low-grade malignant behav-
ior and have an excellent prognosis when resected, but a small
percent metastasize.
CLINICAL PRESENTATION
Most patients with a pancreatic cystic lesion have non-specic symp-
toms.22 The cystic lesion is usually found with CT or US imaging
Fig. 44.2 Gross photograph of a mucinous cystic neoplasm.
Type of lesion Morphology Location
Serous Microcystic Evenly distributed
Mucinous Macrocystic Tail
IPMT Mixed Head
Pseudocyst Unilocular Evenly distributed
Table 44.2
Chapter 44 Suspected IPMN and Cystic Lesions of the Pancreas
performed for the evaluation of another condition. When symptoms
are present, the most common presentation is recurrent abdominal
pain, nausea, and vomiting as result of mild pancreatitis.22 Cystic
lesions that cause duct compression or involvement of the main
pancreatic duct are prone to cause pancreatitis. Chronic abdominal
pain and jaundice are rare presentations of a cystic lesion and
suggest a malignancy or a pseudocyst. Patients with a cystic malig-
nancy will present with symptoms and signs similar to pancreatic
cancer, i.e. pain, weight loss, and jaundice.23 Pseudocysts may arise
after an episode of acute pancreatitis or insidiously in the setting of
chronic pancreatitis and are associated with chronic abdominal pain.
It is common for cystic lesions associated with pancreatitis to be
diagnosed as pseudocysts and be confused with cystic neoplasms
that also cause pancreatitis.24
DIFFERENTIAL DIAGNOSIS
The differential diagnosis of a cystic lesion of the pancreas is
very wide and often causes confusion. Since the treatment of a pseu-
docyst and cystic neoplasm are so different, it is incumbent on the
clinician to rst differentiate between these major categories of
lesions. Although it is unusual for a patient with a pseudocyst to
present without preceding symptoms, it may occur in mild chronic
pancreatitis. Evidence of inammatory changes or calcications in
the pancreas is suggestive of a pancreatic pseudocyst. However, in
the initial setting of mild pancreatitis it may be difcult to differenti-
ate between a cystic neoplasm that has caused pancreatitis and a small
pseudocyst that has formed as a result of pancreatitis. If a cystic lesion
has been present for many years, it is highly likely that the lesion
represents a cystic neoplasm. Congenital cysts of the pancreas are
rare.25
Once a pancreatic pseudocyst has been excluded, attention
should be focused on the differential between the types of cystic
neoplasms. The principal differentiation is between mucinous
and serous lesions because the fundamental difference in manage-
ment is based on the neoplastic potential of mucinous lesions.
The non-neoplastic serous cystadenomas may be diagnosed on the
initial imaging test because of their typical microcystic morphology.
Once a serous lesion has been condently diagnosed, the lesion
may be followed by serial imaging, looking for evidence of growth
of organ impingement. In contrast, the approach to mucinous
lesions is quite different. The underlying risk of malignancy or the
development of malignancy often results in resection of the lesion.
Under some clinical circumstances, such as patients who are at high
risk for complications of pancreatectomy, differentiation between
benign and grossly malignant mucinous lesions is important. The
risk of surgery must be weighed against the risk of malignancy in
the decision-making process. The risk of surgery must also take into
account the variation in risk that is inherent in the location of the
lesion.
Cytology CEA (ng/ml)
Bland PAS+ <0.5
Mucinous >200
Mucinous >200
Pigmented histiocytes <200
469
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

DIAGNOSTIC METHODS solid mass or invasive tumor (Fig. 44.5).38 EUS is also very sensitive
for detecting IPMN lesions, but imaging alone may not be sufcient
CT is an excellent test for cystic lesions of the pancreas because of for differentiating between benign and malignant lesions.39,40 The
its widespread availability and ability to detect cysts (Fig. 44.3).26 MR strength of EUS is its ability to detect and aspirate small cystic
imaging is used increasingly because of its ability to determine if lesions with a high level of safety.41 EUS is more accurate in the
there is involvement of the main pancreatic duct.27 Ultrasonography evaluation of lesions less than 3 cm in diameter.42 The macrocystic
whether performed transabdominally or intraoperatively is generally variant of serous cystadenomas can be diagnosed with EUS using a
not helpful.28 Recently PET scanning has been shown to be positive combination of a thick cyst wall, the presence of microcysts, and a
in a high percentage of malignant cystic lesions.29 low cyst uid CEA.43
Although seen in less than 20 percent of lesions, demonstration The uid contents of cystadenomas are often analyzed for cytol-
of a central scar by CT or MR is a highly diagnostic feature of a ogy.44 However, the low cellular content of cyst uid has hampered
serous cystadenoma.30 The honey-combed or microcystic appearance the use of the cytologic analysis of cyst uid. Small, cuboidal cells
of the lesion is commonly used to provide a diagnosis. However, in cytologic specimens are diagnostic of serous cystadenomas.
macrocystic serous cystadenomas are difcult to diagnose with
cross-sectional imaging because of the morphologic similarities with
mucinous lesions.15,31 The presence of multiple small thin-walled
cysts is suggestive of Von HippelLindau syndrome.32 Mucinous
cystic neoplasms, in contrast, are commonly diagnosed with CT
based on the unilocular or macrocystic characteristics.33 Although
not frequently seen, the nding of peripheral calcication by CT is
specic for a mucinous cystic neoplasm. Intraductal papillary muci-
nous neoplasms (IPMN) may involve the main pancreatic duct
exclusively, a side branch or both. MRCP can demonstrate the diag-
nostic ndings of pancreatic duct dilation, mural nodules, and ductal
connection better than ERCP.34 However, ERCP can demonstrate
the intraductal lling defects, mucin extrusion, and cystic side
branches that are associated with IPMN in 7090% of patients.35,36
Despite these imaging features, the ability to accurately diagnose
a specic cystic lesion and to determine whether malignancy is
present by CT and MR remains uncertain (Fig. 44.4). The diagnosis
of a pancreatic pseudocyst is more dependent upon the clinical
history and the associated ndings of chronic pancreatitis. Pancre-
atic pseudocysts appear as unilocular uid-lled cavities associated
with parenchymal changes such as calcications and atrophy.
Recently endoscopic ultrasound (EUS) has been used to diagnose
cystic lesions of the pancreas and guide ne needle aspiration
(FNA).37 The detailed imaging features of cystic neoplasms by EUS
Fig. 44.4 CT scan of a mucinous cystadenocarcinoma.
do not appear to be sufciently accurate to differentiate between
benign and malignant cystadenomas unless there is evidence of a

Fig. 44.3 CT scan of a mucinous cystic neoplasm in the tail of the Fig. 44.5 EUS image of a mucinous cystadenocarcinoma: note the
pancreas. mass in the wall of the cyst.

470
 Chapter 44 Suspected IPMN and Cystic Lesions of the Pancreas

In contrast, mucinous cystadenoma may have large secretory epi-
thelial cells with evidence of mucin secretion or atypia.45 Only
inammatory cells should be present in the uid aspirated from
pseudocysts.
A variety of cyst uid tumor markers have been studied to help
differentiate between the major types of cystic neoplasms. Serum
19-9 is elevated in frank cystic malignancies.46 A number of gly-
coproteins are present in the epithelium of mucinous neoplasms
and are secreted into the cyst uid.47 The presence of extracellular
mucin in aspirated cyst uid is moderately predictive of a muci-
nous neoplasm.48 Several studies suggest that carcinoembryonic
antigen (CEA) or CA 72-4 are useful for identifying mucinous
lesions.49,50 These carbohydrate antigens are secreted by the epi-
thelium lining mucinous lesions and may be present in high con-
centrations. Cyst uid concentrations of CEA and CA 72-4 are very
low in serous cystadenomas.51 Despite considerable overlap
between mucinous and non-mucinous cysts, cyst uid CEA is the
most accurate marker.50,52 Cyst uid CEA of less than 5 ng/ml is
highly diagnostic of serous cystadenomas and values greater than
800 are predictive of mucinous lesions.52 Recently molecular
studies of cyst uid DNA have revealed k-ras, tumor suppressor
gene mutations, and altered telomerase activity in mucinous cystic
lesions.12,53
Intraductal papillary mucinous neoplasms can be evaluated with
endoscopic retrograde cholangiopancreatography (ERCP) or EUS
(Fig. 44.6).54 Prior to ERCP or EUS, the endoscopic nding of a
patulent ampulla lled with mucin is diagnostic of an intraductal
papillary mucinous neoplasm. Contrast retrograde pancreatography
will demonstrate the characteristic ndings of mucinous lling
defects within the duct, diffuse ductal dilation, and cystic dilation of
side branches.55 MRCP may be more sensitive for detecting the side
branch lesions of IPMN.56 EUS may assist in the detection of malig-
nancy arising from intraductal papillary mucinous neoplasms by
demonstrating focal nodules, invasive lesions and guiding ne
needle aspiration of suspicious lesions.57 EUS can also be used to
monitor IPMN lesions, looking for increases in the size of cysts and
the diameter of the main pancreatic duct.58
DIAGNOSTIC EVALUATION
Patients suspected of having a cystic neoplasm of the pancreas
should undergo a CT scan with contrast as the initial test. If no
lesion is seen in the pancreas, it is very unlikely that a clinically
signicant neoplasm is present. MR imaging, particularly MRCP,
may be substituted for CT scanning (Fig. 44.7). If the CT scan
demonstrates a diagnostic nding, such as a classic microcystic
serous cystadenoma, a malignant cystic mass, or pancreatitis with
a uid collection, no further evaluation is necessary. If the patient
is a young woman and there is a solitary, unilocular cystic lesion in
the tail of the pancreas, the patient should undergo surgical resec-
tion. Indeterminate lesions should undergo EUS with FNA. Aspi-
rated cyst uid should be analyzed for CEA, cytology, and amylase
(Table 44.3). Preference should be given to sending the uid for
CEA as opposed to cytology in small mucinous lesions. Cytology
should be used preferentially in malignant-appearing cystic lesions.
IPMN lesions should have FNA of cystic lesions, an enlarged
pancreatic duct, and focal mass lesions. Cytology should be used
preferentially.

Fig. 44.7 MRCP of side branch IPMT of the head of the

pancreas.

Type of First Second

lesion priority priority Third priority Experimental
Serous CEA Imaging Fluid VHL gene
cytology testing
Mucinous CEA Cytology Subjective k-ras
assessment LOH analysis
of viscosity
Malignant Tissue Fluid CEA k-ras
cytology cytology LOH analysis
Fig. 44.6 ERCP of IPMT: note the dilated and tortuous main pan-
creatic duct. Table 44.3 Use of FNA samples: prioritizing the use of samples

471
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
TREATMENT
Surgical resection is the treatment of choice for pre-malignant cystic
neoplasms. The decision to resect a lesion, however, is based on the
presence or absence of symptoms, the risk of malignancy, and the
surgical risk of the patient. High-risk patients with low-grade cystic
neoplasms may be monitored with periodic CT/MRI scanning or
EUS-FNA.59 Experimentally, high-risk patients have been treated
with EUS-guided ethanol lavage which safely produces variable abla-
tion of the cyst epithelium.60 Small cystic lesions in the elderly can
be safely monitored.61
The increasing safety of surgical resection has prompted the use
of surgery for a wider range of lesions.62 However, serous cystade-
nomas do not require resection except for relief of symptoms.63
Since most mucinous cystic neoplasms are located in the tail of the
pancreas, a distal pancreatectomy is sufcient for these pre-malig-
nant lesions. Since intraductal papillary mucinous neoplasms
invade the pancreas along ductal structures, it is important that
REFERENCES
1. Brugge WR, Lauwers GY, Sahani D, et al. Cystic neoplasms of the
pancreas. N Engl J Med 2004; 351:12181226.
2. Kimura W, Nagai H, Kuroda A, et al. Analysis of small cystic
lesions of the pancreas. Int J Pancreatol 1995; 18:197206.
3. Zhang XM, Mitchell DG, Dohke M, et al. Pancreatic cysts: depiction
on single-shot fast spin-echo MR images. Radiology 2002;
223:547553.
4. Ikeda M, Sato T, Morozumi A, et al. Morphologic changes in the
pancreas detected by screening ultrasonography in a mass
survey, with special reference to main duct dilatation, cyst
formation, and calcication. Pancreas 1994; 9:508512.
5. Compton CC. Serous cystic tumors of the pancreas. Semin Diagn
Pathol 2000; 17:4355.
6. Bassi C, Salvia R, Molinari E, et al. Management of 100 consecutive
cases of pancreatic serous cystadenoma: wait for symptoms and
see at imaging or vice versa? World J Surg 2003; 27:319323.
7. Hammel P. Diagnostic value of cyst uid analysis in cystic lesions
of the pancreas: current data, limitations, and perspectives. J
Radiol 2000; 81:487490.
8. Moore PS, Zamboni G, Brighenti A, et al. Molecular
characterization of pancreatic serous microcystic adenomas:
evidence for a tumor suppressor gene on chromosome 10q. Am
J Pathol 2001; 158:317321.
9. Vortmeyer AO, Lubensky IA, Fogt F, et al. Allelic deletion and
mutation of the von Hippel-Lindau (VHL) tumor suppressor gene
in pancreatic microcystic adenomas. Am J Pathol 1997; 151:
951956.
10. Yasuhara Y, Sakaida N, Uemura Y, et al. Serous microcystic
adenoma (glycogen-rich cystadenoma) of the pancreas: Study of
11 cases showing clinicopathological and immunohistochemical
correlations. Pathol Int 2002; 52:307312.
11. Ahmad NA, Kochman ML, Brensinger C, et al. Interobserver
agreement among endosonographers for the diagnosis of
neoplastic versus non-neoplastic pancreatic cystic lesions.
Gastrointest Endosc 2003; 58:5964.
12. Khalid A, McGrath KM, Zahid M, et al. The role of pancreatic cyst
uid molecular analysis in predicting cyst pathology. Clin
Gastroenterol Hepatol 2005; 3:967973.
472
frozen section histology be used during surgery to assure negative
margins.64,65
PROGNOSIS
Cystic neoplasms are slow growing and 19% will demonstrate an
increase in diameter at 16 months.66 Surgical resection has been
associated with a morbidity of 27.9%, with a reoperation rate of 7.3%
and a very low mortality rate.6 The overall 5-year survival for patients
having IPMNs without invasive cancer was 77%, compared with 43%
in those patients with an invasive component.67 The overall postop-
erative recurrence rate varies from 7% to 43%.68 Approximately 50%
of patients will have evidence of malignancy in the resected speci-
men.69 Similar survival rates are seen in patients with mucinous
cystic neoplasms.70 Side branch lesions arising from IPMN have a
better prognosis than main duct IPMN.71 The worst prognosis is for
advanced, transmural adenocarcinomas arising from mucinous
lesions; the 5-year survival is only 30% for resected lesions.
13. Yoshizawa K, Nagai H, Sakurai S, et al. Clonality and K-ras
mutation analyses of epithelia in intraductal papillary mucinous
tumor and mucinous cystic tumor of the pancreas. Virchows
Arch 2002; 441:437443.
14. Wada K, Takada T, Yasuda H, et al. Does clonal progression
relate to the development of intraductal papillary mucinous
tumors of the pancreas? J Gastrointest Surg 2004; 8:289296.
15. Khurana B, Mortele KJ, Glickman J, et al. Macrocystic serous
adenoma of the pancreas: radiologic-pathologic correlation. AJR
Am J Roentgenol 2003; 181:119123.
16. Mohr VH, Vortmeyer AO, Zhuang Z, et al. Histopathology and
molecular genetics of multiple cysts and microcystic (serous)
adenomas of the pancreas in von Hippel-Lindau patients. Am J
Pathol 2000; 157:16151621.
17. Tanaka M. Intraductal papillary mucinous neoplasm of the
pancreas: diagnosis and treatment. Pancreas 2004;
28:282288.
18. Furuta K, Watanabe H, Ikeda S. Differences between solid and
duct-ectatic types of pancreatic ductal carcinomas. Cancer 1992;
69:13271333.
19. Kann P, Bittinger F, Engelbach M, et al. Endosonography of
insulin-secreting and clinically non-functioning neuroendocrine
tumors of the pancreas: criteria for benignancy and malignancy.
Eur J Med Res 2001; 6:385390.
20. Marcos HB, Libutti SK, Alexander HR, et al. Neuroendocrine
tumors of the pancreas in von Hippel-Lindau disease: spectrum
of appearances at CT and MR imaging with histopathologic
comparison. Radiology 2002; 225:751758.
21. Notohara K, Hamazaki S, Tsukayama C, et al. Solid-
pseudopapillary tumor of the pancreas: immunohistochemical
localization of neuroendocrine markers and CD10. Am J Surg
Pathol 2000; 24:13611371.
22. Wiesenauer CA, Schmidt CM, Cummings OW, et al. Preoperative
predictors of malignancy in pancreatic intraductal papillary
mucinous neoplasms. Arch Surg 2003; 138:610617; discussion
617618.
23. Holly EA, Chaliha I, Bracci PM, et al. Signs and symptoms of
pancreatic cancer: a population-based case-control study in the

San Francisco Bay area. Clin Gastroenterol Hepatol 2004;
2:510517.
24. Sand J, Nordback I. The differentiation between pancreatic
neoplastic cysts and pancreatic pseudocyst. Scand J Surg 2005;
94:161164.
25. Caillot JL, Rongieras F, Voiglio E, et al. A new case of congenital
cyst of the pancreas. Hepatogastroenterology 2000; 47:916918.
26. Curry CA, Eng J, Horton KM, et al.. CT of primary cystic pancreatic
neoplasms: can CT be used for patient triage and treatment? AJR
Am J Roentgenol 2000; 175:99103.
27. Fukukura Y, Fujiyoshi F, Hamada H, et al. Intraductal papillary
mucinous tumors of the pancreas. Comparison of helical CT and
MR imaging. Acta Radiol 2003; 44:464471.
28. Kubota K, Noie T, Sano K, et al. Impact of intraoperative
ultrasonography on surgery for cystic lesions of the pancreas.
World J Surg 1997; 21:7276; discussion 77.
29. Sperti C, Pasquali C, Decet G, et al. F-18-uorodeoxyglucose
positron emission tomography in differentiating malignant from
benign pancreatic cysts: a prospective study. J Gastrointest Surg
2005; 9:2228; discussion 2829.
30. Torresan F, Casadei R, Solmi L, et al. The role of ultrasound in the
differential diagnosis of serous and mucinous cystic tumours of
the pancreas. Eur J Gastroenterol Hepatol 1997; 9:169172.
31. Chatelain D, Hammel P, OToole D, et al. Macrocystic form of
serous pancreatic cystadenoma. Am J Gastroenterol 2002; 97:
25662571.
32. Cheng TY, Su CH, Shyr YM, et al. Management of pancreatic
lesions in von Hippel-Lindau disease. World J Surg 1997; 21:
307312.
33. Sahani D, Prasad S, Saini S, et al. Cystic pancreatic neoplasms
evaluation by CT and magnetic resonance
cholangiopancreatography. Gastrointest Endosc Clin N Am 2002;
12:657672.
34. Sugiyama M, Atomi Y, Hachiya J. Intraductal papillary tumors of
the pancreas: evaluation with magnetic resonance
cholangiopancreatography. Am J Gastroenterol 1998; 93:156159.
35. Madura JA, Wiebke EA, Howard TJ, et al. Mucin-hypersecreting
intraductal neoplasms of the pancreas: a precursor to cystic
pancreatic malignancies. Surgery 1997; 122:786792; discussion
792793.
36. Azar C, Van de Stadt J, Rickaert F, et al. Intraductal papillary
mucinous tumours of the pancreas. Clinical and therapeutic issues
in 32 patients. Gut 1996; 39:457464.
37. Brugge WR. Evaluation of pancreatic cystic lesions with EUS.
Gastrointest Endosc 2004; 59:698707.
38. Ahmad NA, Kochman ML, Lewis JD, et al. Can EUS alone
differentiate between malignant and benign cystic lesions of the
pancreas? Am J Gastroenterol 2001; 96:32953300.
39. Brandwein SL, Farrell JJ, Centeno BA, et al. Detection and tumor
staging of malignancy in cystic, intraductal, and solid tumors of
the pancreas by EUS. Gastrointest Endosc 2001; 53:722727.
40. Terris B, Ponsot P, Paye F, et al. Intraductal papillary mucinous
tumors of the pancreas conned to secondary ducts show
less aggressive pathologic features as compared with those
involving the main pancreatic duct. Am J Surg Pathol 2000;
24:13721377.
41. Lee LS, Saltzman JR, Bounds BC, et al. EUS-guided ne needle
aspiration of pancreatic cysts: a retrospective analysis of
complications and their predictors. Clin Gastroenterol Hepatol
2005; 3:231236.
42. Volmar KE, Vollmer RT, Jowell PS, et al. Pancreatic FNA in 1000
cases: a comparison of imaging modalities. Gastrointest Endosc
2005; 61:854861.
43. OToole D, Palazzo L, Hammel P, et al. Macrocystic pancreatic
cystadenoma: The role of EUS and cyst uid analysis in
Chapter 44 Suspected IPMN and Cystic Lesions of the Pancreas
distinguishing mucinous and serous lesions. Gastrointest Endosc
2004; 59:823829.
44. Centeno BA, Warshaw AL, Mayo-Smith W, et al. Cytologic
diagnosis of pancreatic cystic lesions. A prospective study of 28
percutaneous aspirates. Acta Cytol 1997; 41:972980.
45. Recine M, Kaw M, Evans DB, et al. Fine-needle aspiration
cytology of mucinous tumors of the pancreas. Cancer 2004;
102:9299.
46. Sperti C, Pasquali C, Guolo P, et al. Serum tumor markers and cyst
uid analysis are useful for the diagnosis of pancreatic cystic
tumors. Cancer 1996; 78:237243.
47. Yamaguchi K, Enjoji M. Cystic neoplasms of the pancreas.
Gastroenterology 1987; 92:19341943.
48. Walsh RM, Henderson JM, Vogt DP, et al. Prospective
preoperative determination of mucinous pancreatic cystic
neoplasms. Surgery 2002; 132:628633; discussion 633634.
49. Frossard JL, Amouyal P, Amouyal G, et al. Performance of
endosonography-guided ne needle aspiration and biopsy in the
diagnosis of pancreatic cystic lesions. Am J Gastroenterol 2003;
98:15161524.
50. Brugge WR, Lewandrowski K, Lee-Lewandrowski E, et al.
Diagnosis of pancreatic cystic neoplasms: a report of the
cooperative pancreatic cyst study. Gastroenterology 2004; 126:
13301336.
51. Hammel P, Voitot H, Vilgrain V, et al. Diagnostic value of CA 72-4
and carcinoembryonic antigen determination in the uid of
pancreatic cystic lesions. Eur J Gastroenterol Hepatol 1998; 10:
345348.
52. van der Waaij LA, van Dullemen HM, Porte RJ. Cyst uid analysis
in the differential diagnosis of pancreatic cystic lesions: a pooled
analysis. Gastrointest Endosc 2005; 62:383389.
53. Zhou GX, Huang JF, Li ZS, et al. Detection of K-ras point mutation
and telomerase activity during endoscopic retrograde
cholangiopancreatography in diagnosis of pancreatic cancer.
World J Gastroenterol 2004; 10:13371340.
54. Telford JJ, Carr-Locke DL. The role of ERCP and pancreatoscopy in
cystic and intraductal tumors. Gastrointest Endosc Clin N Am
2002; 12:747757.
55. Dachman AH, Namieno T, Ichimura T, et al. Mucin-producing
pancreatic tumors: comparison of MR cholangiopancreatography
with endoscopic retrograde cholangiopancreatography.
Radiology 1998; 208:231237.
56. Koito K, Namieno T, Ichimura T, et al. Mucin-producing pancreatic
tumors: comparison of MR cholangiopancreatography with
endoscopic retrograde cholangiopancreatography. Radiology
1998; 208:231237.
57. Sugiyama M, Atomi Y, Saito M. Intraductal papillary tumors of the
pancreas: evaluation with endoscopic ultrasonography.
Gastrointest Endosc 1998; 48:164171.
58. Kobayashi G, Fujita N, Noda Y, et al. Mode of progression of
intraductal papillary-mucinous tumor of the pancreas: analysis of
patients with follow-up by EUS. J Gastroenterol 2005; 40:744751.
59. Irie H, Yoshimitsu K, Aibe H, et al. Natural history of pancreatic
intraductal papillary mucinous tumor of branch duct type: follow-
up study by magnetic resonance cholangiopancreatography. J
Comput Assist Tomogr 2004; 28:117122.
60. Gan SI, Thompson CC, Lauwers GY, et al. Ethanol lavage of
pancreatic cystic lesions: initial pilot study. Gastrointest Endosc
2005; 61:746752.
61. Kimura W, Makuuchi M. Operative indications for cystic lesions of
the pancreas with malignant potentialour experience.
Hepatogastroenterology 1999; 46:483491.
62. Fernandez del Castillo CF, Targarona J, Thayer SP, et al. Incidental
pancreatic cysts: clinicopathologic characteristics and comparison
with symptomatic patients. Arch Surg 2003; 138:427434.
473
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
63. Le Borgne J, de Calan L, Partensky C. Cystadenomas and
cystadenocarcinomas of the pancreas: a multiinstitutional
retrospective study of 398 cases. French Surgical Association. Ann
Surg 1999; 230:152161.
64. Gigot JF, Deprez P, Sempoux C, et al. Surgical management of
intraductal papillary mucinous tumors of the pancreas: the role
of routine frozen section of the surgical margin, intraoperative
endoscopic staged biopsies of the Wirsung duct, and
pancreaticogastric anastomosis. Arch Surg 2001; 136:
12561262.
65. Chari ST, Yadav D, Smyrk TC, et al. Study of recurrence after
surgical resection of intraductal papillary mucinous neoplasm of
the pancreas. Gastroenterology 2002; 123:15001507.
66. Spinelli KS, Fromwiller TE, Daniel RA, et al. Cystic pancreatic
neoplasms: observe or operate. Ann Surg 2004; 239:651657;
discussion 657659.
474
67. Sohn TA, Yeo CJ, Cameron JL, et al. Intraductal papillary mucinous
neoplasms of the pancreas: an updated experience. Ann Surg
2004; 239:788797; discussion 797799.
68. Lai EC, Lau WY. Intraductal papillary mucinous neoplasms of the
pancreas. Surgeon 2005; 3:317324.
69. Falconi M, Salvia R, Bassi C, et al. Clinicopathological features and
treatment of intraductal papillary mucinous tumour of the
pancreas. Br J Surg 2001; 88:376381.
70. Suzuki Y, Atomi Y, Sugiyama M, et al. Cystic neoplasm of the
pancreas: a Japanese multiinstitutional study of intraductal
papillary mucinous tumor and mucinous cystic tumor. Pancreas
2004; 28:241246.
71. Kobari M, Egawa S, Shibuya K, et al. Intraductal papillary mucinous
tumors of the pancreas comprise 2 clinical subtypes: differences
in clinical characteristics and surgical management. Arch Surg
1999; 134:11311136.
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
Chapter
Endoscopic Drainage of Pancreatic
45 Pseudocysts, Abscesses and Organized
(Walled-Off ) Necrosis
Todd H. Baron
BOX 45.1 KEY POINTS
Pancreatic uid collections comprise pancreatic pseudocysts,
abscesses, and organized pancreatic necrosis
Pancreatic pseudocysts may be acute or chronic
Endoscopic drainage of pancreatic uid collections can be
achieved using transmural (transgastric or transduodenal)
and transpapillary drainage techniques, alone or in
combination
The short-term outcome following endoscopic drainage of
pancreatic uid collections depends on the degree of liquid
or solid components
The long-term outcome following endoscopic drainage of
pancreatic uid collections depends on the presence or
absence of pancreatic ductal damage
Transmural drainage is usually required for drainage of
pancreatic uid collections
INTRODUCTION AND SCIENTIFIC BASIS
Pancreatic pseudocysts, abscesses, and pancreatic necrosis are types
of pancreatic uid collections (PFCs) that arise as a consequence of
pancreatic injury.1 At the basis of this pancreatic injury is disruption
of the pancreatic duct or side branches. Ductal disruption can be due
to acute pancreatic injury (acute pancreatitis, trauma, surgical resec-
tion or injury to the pancreas during abdominal surgery) or chronic
injury (chronic pancreatitis, autoimmune pancreatitis). The sequela
to this ductal injury is the formation of a collection of uid with or
without solid debris.
Thus the basis of endoscopic therapy is directed at either the
drainage of the uid and solid components using a transmural
approach and/or correction of ductal abnormalities using a trans-
papillary approach. Correction of pancreatic abnormalities may
decrease the long-term recurrence rate and improve the outcome
following successful resolution of a collection and can be assessed
and treated endoscopically. This chapter will discuss the endoscopic
approaches to PFCs.
SPECIFIC TYPES OF FLUID COLLECTIONS
Classication systems exist for dening types of PFCs1 which are
useful for understanding mechanisms of formation and making
meaningful comparisons of therapies between or amongst disci-
plines. At the time of this writing, the classication and nomen-
clature for PFCs is being revised. When undertaking endoscopic
therapy of a PFC, however, it is more important to make two
main distinctions: (1) Is the collection composed primarily of
liquid or does it contain signicant solid debris? and (2) What is
the status and etiology of the disruption of the main pancreatic
duct? Using these two basic questions, an approach to drainage
can be formulated for both the short-term and long-term approach
of patients with PFCs. The approach to collections that are com-
posed primarily of uid is different than that of those containing
signicant solid debris, since liqueed collections can be managed
with either placement of modest sized diameter stents via trans-
papillary or transmural approaches alone, whereas those with
solid debris usually require aggressive transmural dilation to allow
egress of solid material and placement of large-bore stents and
irrigation catheters, and are associated with higher complication
rates.
COLLECTIONS COMPOSED ENTIRELY OR
PREDOMINANTLY OF LIQUID
1. Acute uid collections:
Acute uid collections arise early in the course of acute
pancreatitis, are usually peripancreatic in location, and usually
resolve without sequelae but may evolve into pancreatic
pseudocysts.1
2. Pancreatic pseudocysts
(a) Acute pancreatic pseudocysts: Acute pseudocysts arise as a
sequela of acute pancreatitis, require at least four weeks to
form, and are devoid of signicant solid debris. The mecha-
nism of formation of an acute pancreatic pseudocyst is usually
as a result of limited pancreatic necrosis that produces a pan-
creatic ductal leak (Fig. 45.1). Alternatively, areas of pancreatic
and peripancreatic fat necrosis may completely liquefy over
time and become a pseudocyst.2 Despite the requirement of
at least 4 weeks for a pseudocyst to form, it is important to
note that this time period in and of itself does not dene the
collection as a pancreatic pseudocyst. Patients with signicant
pancreatic necrosis (>30%) may evolve the early acute pancre-
atic necrosis and peripancreatic necrosis into a collection that
475
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
resembles a pseudocyst radiographically, but has been present
for more than 4 more weeks (see organized pancreatic necro-
sis below). By denition, if these collections contain signi-
cant solid debris, they are not pseudocysts and endoscopic
treatment of these collections by typical pseudocyst drainage
methods may result in infectious complications because of
inadequate removal of solid debris.1,3,4
(b) Chronic pseudocysts: A chronic pseudocyst arises as a sequela
of chronic pancreatitis due to downstream pancreatic obstruc-
tion from brotic strictures and/or stones.5 This results in a
pancreatic ductal blowout (leak) and accumulation of pancre-
atic uid. These collections do not contain solid debris and
usually do not arise as a result of acute inammatory pro-
cesses (Fig. 45.2).
3. Infected pancreatic pseudocysts (in some nomenclatures also known
as pancreatic abscesses1)
4. Pancreatic abscesses. True pancreatic abscesses have formerly been
considered rare and not synonymous with infected pancreatic
pseudocysts.1 However, for the purposes of this chapter and based
476
upon pending revisions of the existing nomenclature, an abscess
will be considered as an infected PFC that contains little to no
solid debris (as opposed to infected pancreatic necrosis which will
be described later). It is my belief that using this denition,
abscesses can be drained through modest sized catheters without
need for irrigation or debridement.
INDICATIONS FOR DRAINAGE OF
LIQUEFIED COLLECTIONS
In general, the indications for drainage of a liqueed PFC are
symptom driven and infection driven, not merely the presence of
a collection by imaging studies. For many years, a size cut-off of
approximately 6 cm or persistence of a collection for more than a
certain amount of time was used a criterion for drainage. It has
been established, however, that patients may remain asymptomatic
with collections of 6 cm or more with little risk of complication
such as rupture, infection, or bleeding,6 whereas endoscopic inter-
vention is associated with a nite (and presumably higher) risk of
Fig. 45.1 Illustration of the mechanism of formation
of an acute pancreatic pseudocyst. Limited necrosis
of the main pancreatic duct produces a ductal leak
with accumulation of amylase-rich uid.
Fig. 45.2 Illustration of the mechanism of formation
of a chronic pancreatic pseudocyst. Obstruction of the
main pancreatic duct by stones and/or stricture pro-
duces a duct blowout with accumulation of amylase-
rich uid.
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis
BOX 45.2 INDICATIONS AND
CONTRAINDICATIONS
In general, the indications for drainage of pancreatic uid
collections are the presence of symptoms and/or infection
Pancreatic pseudocysts produce symptoms because of mass
compression of surrounding structures, pancreatic leaks, and
pancreatic stulae
Prior to drainage it should be established that the collection
is an inammatory collection and not a true pancreatic cyst
or cystic neoplasm
Prior to drainage it should be established whether the
collection contains signicant solid debris
Contraindication to any type of endoscopic drainage is free
luminal perforation. Contraindications to transmural drainage
include a distance of free space between the gastric or
duodenal wall of more than 1 cm from the and presence of
underlying coagulopathy.
Cystic pancreatic neoplasm
Duplication cyst
True pancreatic cyst
Pseudoaneurysm
Solid necrotic neoplasm (e.g. retroperitoneal sarcoma)
Lymphocele
Gallbladder
Table 45.1 Masqueraders of pancreatic uid collections
complications. Progressive enlargement of a collection is one excep-
tion to symptoms that is cited as an indication for drainage1,
although even then a patient can be followed until symptoms
occur.
Symptoms related to liqueed pancreatic collections include
abdominal painoften exacerbated by eating, weight loss, gastric
outlet obstruction, obstructive jaundice, and leakage. The leakage
results in pancreatic ascites and pancreatic stulae and has been
discussed in Chapter 40. Infection is considered an absolute indica-
tion for drainage.
PRE-DRAINAGE EVALUATION
Prior to undertaking drainage of a liqueed pancreatic collection, a
pre-drainage evaluation should be performed. The goals of the pre-
drainage evaluation include the following:
1. Establish whether the collection represents a PFC or a masquer-
ader of a PFC such as a cystic neoplasm7 or other entity (Table
45.1, Fig. 45.3). If the patient does not have a well-documented
history of acute or chronic pancreatitis, the endoscopist should
be wary that something other than a pseudocyst is present.8
With the development of endoscopic ultrasonography (EUS),
Fig. 45.3 Metastatic malignant brous histiocytoma detected in
a woman with abdominal pain and pancreatitis mimicking a
pancreatic pseudocyst. She had a previous known primary lesion.
The lesion was diagnosed by EUS-FNA and resected for
palliation.
cystic neoplasms have been better recognized and dened.9 Cystic
neoplasms are discussed in more detail in Chapter 44.
2. Establish whether the collection is predominately liquid or con-
tains signicant solid debris.
3. Establish the relationship of the collection to surrounding lumenal
and vascular structures.
4. Consider underlying etiologies of true pancreatic pseudocyst
which have implications for alternative/adjuvant therapies such
as pancreatic cancer,10 autoimmune pancreatitis,11 and intraductal
pancreatic mucinous neoplasms.
In addition to a complete history and physical examination, the
following evaluation should be undertaken:
1. Coagulation prole in patients with a suspicion of coagulopathy
and/or liver disease, especially when transmural drainage is
considered.
2. Contrast-enhanced abdominal CT. This allows assessment of
the precise location of the collection in relation to the stomach
and duodenum in anticipation of possible transmural drainage.
Additionally, the relationship of the collection to potential
intervening vascular structures can be assessed. Surrounding
varices from splenic vein or portal vein thrombosis may also be
visualized. The nding of inhomogeneity within the collection
suggests the presence of underlying solid debris and/or blood
(Fig. 45.4).
Consideration should be given to additional studies:
1. Endoscopic ultrasonography (EUS). EUS can be used prior to
drainage to allow assessment for the presence of signicant solid
debris that may alter the management strategy. In addition, if
there is uncertainty as to whether the collection represents a true
pseudocyst or other non-inammatory cystic lesion, EUS allows
one to obtain a denitive diagnosis by using ultrasonographic
features, aspiration and analysis of cyst contents, and biopsy of
the cyst wall.12 Once the endoscopist is certain that the lesion in
question is a PFC and the decision has been made to proceed with
477
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
endoscopic drainage, EUS may be used to guide transmural
drainage as discussed in the next section under endoscopic drain-
age methods.
2. MRI with or without MRCP. Magnetic resonance imaging (MRI)
allows detection of the presence of solid debris,13 so that plans for
removal and/or alternative drainage strategies can be chosen
depending on local expertise and necrosis drainage preferences.
Fig. 45.4 CT scan obtained 4 weeks after clinically severe gallstone
pancreatitis. Note large homogenous uid collection posterior to
the stomach with inhomogenous density (arrowheads) suggesting
solid debris.
A B
D
E
Fig. 45.5 Patient with pseudocyst due to chronic pancreatitis. A pseudocyst (PC) is seen compressing duodenum; calcications present
near tail (arrows). B Lower cuts same patient. A large stone (arrow) is obstructing the main pancreatic duct. C Follow-up CT after transmural
drainage and ESWL. Transduodenal stents can be seen (arrow). The prior stone has been fragmented (arrowhead). D At the time of duo-
denal stent removal pancreatography shows a stricture in the head (arrow). E Stone fragments were removed, the stricture was balloon
dilated, and a pancreatic duct stent placed.
478
DRAINAGE TECHNIQUES
Liqueed PFCs as described above may be drained transpapillary,
transmurally, or using a combination of the two. The decision to use
one approach over the other depends on the size of the collection,
its proximity to the stomach or duodenum, and the ability to enter
the pancreatic duct and/or reach the area of disruption. For example,
although the intended approach to draining a pseudocyst that formed
from an obstructing pancreatic duct stone may be transpapillary
(Figs 45.5AB), failure to negotiate a guidewire beyond the obstruct-
ing ductal stone may require transmural drainage. Assessment and
treatment of the ductal stone at a later date by other techniques such
as ESWL can then be performed (Figs 45.5CE).
Transpapillary drainage
If the collection communicates with the main pancreatic duct, place-
ment of a pancreatic endoprosthesis with or without pancreatic
sphincterotomy is an approach that is useful, especially for collec-
tions measuring <6 cm that are not otherwise approachable trans-
murally.1 The proximal end of the stent (toward the pancreatic tail)
may directly enter the collection or bridge the area of leak into the
pancreatic duct upstream from the leak (Figs 45.6AD). The latter
is the preferred approach (Fig. 45.7), since it restores ductal continu-
ity.14 In patients with chronic pseudocysts, it is important that the
stent bridges any obstructive process (stricture or stone) between the
duodenum and the leak. The diameter of pancreatic stent used is
dependent on the pancreatic ductal diameter, although 7 Fr stents
are most frequently used. In patients with chronic pancreatitis,
endoscopic therapy of underlying pancreatic duct strictures and
pancreatic stones may reduce the recurrence rate of pancreatic
pseudocysts.1
The advantage of the transpapillary approach over the transmural
approach is the avoidance of bleeding or perforation that may occur
C
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis

A B

D
C

Fig. 45.6 Transpapillary drainage of pancreatic pseudocyst. A CT scan. Collections are seen (arrows). B Pancreatogram shows leakage at
the tail. C Transpapillary stent placed to tail. D Follow-up pancreatogram shows no leak.

with transmural drainage. The disadvantage of transpapillary drain-
age is that pancreatic stents may induce scarring of the main pan-
creatic duct in patients whose pancreatic duct is otherwise normal
(i.e., patients with acute pseudocysts and small side branch disrup-
tion or those patients with leakage after surgical resection of the
pancreatic tail, see Figure 16.22).
TRANSMURAL DRAINAGE
Transmural entry devices
The type of devices used to puncture through the gastric or duodenal
wall into the collection can be divided into cautery and non-cautery
devices. Cautery devices include standard diathermy wires (needle
knife) and specialized stulotomy devices (Cystotome, CST-10, Cook
Endoscopy, Winston-Salem, NC). Non-cautery devices include EUS
ne needle aspiration (FNA) needles and aspiration needles (BaronTM
Aspiration Needle, BAN-18, Cook Medical).
Transmural entry techniques
There is no standardized approach to this method of drainage and
some endoscopists feel that EUS evaluation is mandatory prior to
performing endoscopic transmural drainage of PFCs, although the
superiority of EUS-guided versus non EUS-guided drainage has not
been demonstrated.15 Approximately 50% of all respondents to a
recent survey of transmural drainage practices claimed to use EUS-
guided transmural drainage.16 EUS-guided and non-EUS guided
drainage will be discussed separately.
EUS-GUIDED TRANSMURAL DRAINAGE
EUS imaging may theoretically reduce complications related to
transmural entry of PFCs, although this has not been proven.16,17
There are two ways EUS can be used for transmural drainage. The
rst way, using either a radial or linear echoendoscope, is to local-
ize the collection in relationship to surrounding structures and

479
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A
Fig. 45.7 A Schematic of pancreatic duct leak B Illustration of ideal position of pancreatic duct stent placement across a leak site.
endoscopic landmarks. The optimal location of entry into the gastric
or duodenal wall may be marked with contrast injection or
mucosal biopsy; the echoendoscope is then removed and a thera-
peutic endoscope or duodenoscope is then used to perform trans-
mural drainage by puncturing into the collection as described under
non-EUS guided drainage below. The second way EUS may be used
is to perform the evaluation and entry into the collection using
direct EUS guidance with a linear echoendoscope similar to EUS-
guided FNA (Figs 45.8AD). Using this approach either with a stan-
dard or Doppler-equipped echoendoscope, successful entry has
been reported in up to 94% of patients with low complication rates,
including those without an endoscopically visible extrinsic com-
pression.1820 Therefore, if EUS is readily available, consideration
should be given to EUS assistance during transmural drainage.
The lack of EUS availability, however, does not preclude trans-
mural drainage except in the following instances: small window
of entry based upon CT, especially in the absence of an endo-
scopically dened area of extrinsic compression or unusual loca-
tion;1 coagulopathy or thrombocytopenia; documented intervening
varices;21 and failed transmural entry using non-EUS-guided
techniques.
NON-EUS-GUIDED TRANSMURAL DRAINAGE
The collection is entered at a point of endoscopically visible extrin-
sic compression using electrocautery with or without prelocaliza-
tion using a needle (Fig. 45.9ac). An alternative is localization and
entry into the collection using a needle which accepts an 0.035
guidewire (Fig. 45.10) without the use of electrocautery using the
480
B
Seldinger technique. Entry is conrmed by aspiration of uid and/
or injection of radiopaque contrast (Figs 45.11AC). This method
may be safer than a cautery puncture for non-EUS guided entry
since if unsuccessful entry occurs, the needle is simply withdrawn
without adverse sequelae. Similarly, if bleeding occurs upon needle
entry, if gross blood is aspirated, or if a visible hematoma develops,
the needle is withdrawn to allow the vessel to tamponade. Another
transmural entry site may be chosen during the same endoscopic
session. Using this technique, successful non-EUS-guided trans-
mural entry has been reported in 41/43 patients (95%) in lesions
as small as 3 cm and without endoscopically visible extrinsic com-
pression.22 Other authors have described successful drainage of
non-bulging collections with or without EUS guidance.23
Stent placement
Transmural drainage of liqueed pancreatic collections is achieved
by placing one or more stents through the gastric or duodenal wall.
After entry into the collection, the transmural tract is balloon dilated
with a standard ERCP dilating balloon of 810 mm diameter (Fig.
45.11d) to allow placement of one or two 10 Fr stents (Fig. 45.11e).
It is important to secure an ample length of guidewire into the col-
lection with at least one 360 degree loop of wire (Figs 45.8d and
45.11d). The practice of enlarging the transmural tract using cautery
and a sphincterotome has been abandoned because it is associated
with an increased risk of bleeding.
The type of stents used for transmural drainage may be straight
or pigtail. Double pigtail stents are recommended for at least two
reasons. First, they are less prone to migrate into or out of the col-
lection. Secondly, recent data suggests that straight stents are associ-
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis

A B

C D

Fig. 45.8 EUS-guided drainage. A Illustration of EUS-guided

transmural drainage of pancreatic uid collection. B Ultraso-
nographic image taken during EUS-guided drainage of pan-
creatic pseudocyst showing needle entering the collection
(arrowheads). C Illustration of guidewire passed through EUS
needle. D Same patient as B, contrast has been injected
through the FNA needle and a guidewire coiled into the col-
lection. Subsequent stents were placed with successful
drainage.

A B C

Fig. 45.9 Transmural drainage without EUS using the Cystotome (Cook Endoscopy). A Extrinsic compression with Cystotome in view.
B Initial entry is made with the inner, smaller-sized cautery device. C The outer, larger (10 Fr) cautery portion is passed over the smaller
one and through the wall of the collection.

481
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

ated with delayed bleeding complications resulting from impaction

of the stent against the wall of the collection as it contracts around
the stent.24 We routinely place one to two short-length (35 cm) 10 Fr
double pigtail stents during transmural drainage. Stents are avail-
able from several companies. I use standard 10 Fr double pigtail
stents (Cook Endoscopy, Winston-Salem, NC), although the tapered
tip will not allow an inner guiding catheter to pass unless it is
trimmed off. One must be careful when placing double pigtail stents
to not push the entire stent into the collection. This can be avoided
by passing no more than 50% of the stent. An indelible marker can
be placed at the midpoint of the stent if radiopaque markers are not
already present. When the midpoint of the stricture is at the gastric
or duodenal wall, the endoscope is withdrawn away from the collec-
tion while simultaneously pushing the stent out of the endoscope
channel. Recently, softer stents (Hobbs Medical Inc., Stafford
Springs, CT and the SolusTM stent, Cook Endoscopy) have become
available. The SolusTM stent has an inner guiding catheter as well as
endoscopic and radiopaque markers (Fig. 45.12).

PANCREATIC NECROSIS
Fig. 45.10 Baron needle with guidewire in place (Cook Pancreatic necrosis is dened as non-viable pancreatic parenchyma
Endoscopy). usually with associated peripancreatic fat necrosis. In the earliest
form, this is detected radiographically on contrast enhanced CT by
the presence of non-enhancing pancreatic parenchyma (Fig. 45.13).
Pancreatic necrosis is frequently accompanied by major pancreatic

A B C

D E

Fig. 45.11 Transmural drainage using the Seldinger technique (same patient as in Figure 45.5). A A needle is passed transduodenally
through the duodenal wall. B Contrast is injected and lls the pseudocyst. C A guidewire is coiled within the collection. D The duodenal
wall is balloon dilated using a 10 mm biliary dilating balloon. E Two double pigtail stents are placed.

482
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis
Fig. 45.12 SolusTM double pigtail stent. The markers are both
endoscopically and radiographically visible (arrows).
Fig. 45.13 Early acute necrotizing pancreatitis. Lack of glandular
perfusion of non-viable pancreatic parenchyma (NV) is seen in the
neck of the pancreas. The viable parenchyma (V) is seen in the body
and tail.
ductal disruptions. Over the course of several weeks, the collection
may continue to evolve and expand the initial area of necrosis and
contains both liquid and solid debris (Fig. 45.14). The term orga-
nized pancreatic necrosis has been used to differentiate this process
from the early (acute phase) of pancreatic necrosis, although many
investigators now refer to this entity as walled-off necrosis (WON)
(Fig. 45.15). The CT appearance of organized pancreatic necrosis
may be similar to that of an acute pseudocyst. Because the underly-
ing solid debris is frequently not discernible by CT, its homogeneous
appearance may lead one to embark on standard pseudocyst drain-
Fig. 45.14 Same patient as in Figure 45.13, ve weeks later. There
is now a large collection occupying the area of the pancreatic bed
consistent with walled-off necrosis (organized pancreatic
necrosis).
age methods which do not adequately remove the underlying solid
material. This may result in serious infectious complications.1
The distinction between an acute pseudocyst and organized
necrosis may be made on clinical, radiologic, or endoscopic nd-
ings at the time of drainage. Clinically, if the patient suffered a
severe or complicated course of acute pancreatitis, it is likely that
pancreatic necrosis occurred and is present within the collection.
Radiographically, several features indicate the presence of underly-
ing solid material within the collection. Firstly, if an initial contrast-
enhanced computerized tomographic scan obtained at the time
ofor soon afterthe initial bout of pancreatitis demonstrated
signicant glandular necrosis, the collection likely contains solid
debris. Secondly, the evolution of changes on serial CT scans can
be traced from the original pancreatic glandular necrosis to the
present collection. Thirdly, magnetic resonance imaging (MRI)
prior to attempted drainage can delineate the solid debris within
the collection. Lastly, a repeat abdominal CT scan after endoscopic
drainage will depict solid material once some of liquid has been
evacuated (Fig. 45.16). Endoscopic ndings at the time of drainage
may alert the endoscopist to the presence of necrotic debris within
the collection. If the collection is drained transmurally, solid mate-
rial may be seen to ow from the collection; the presence of choco-
late-brown or extremely turbid uid (in the absence of clinical
infection) also suggests underlying necrosis is present. During pan-
creatography, the nding of complete main pancreatic duct disrup-
tion (Figs 45.17a and b) suggests that pancreatic necrosis occurred
during the initial course of pancreatitis and may be present in the
collection. During contrast injection, either through the main pan-
creatic duct or transmurally, the nding of large lling defects
within the collection denotes the presence of solid material. If any
or all of the above are recognized, then appropriate steps must be
taken to evacuate the underlying solid debris to prevent secondary
infection. Overall, one should consider the evolution of a pancreatic
collection from the early phase of acute pancreatic necrosis toward
a pseudocyst as a spectrum, with organized pancreatic necrosis as
an intermediate stage, realizing that some collections will never
become completely liqueed.
483
 SECTION 3 APPROACH TO CLINICAL PROBLEMS

Fig. 45.15 Illustration depicting organized pancre-

atic necrosis (walled-off necrosis). Note viable pan-
creatic head and tail typically occurs and is the
mechanism of pancreatic duct disconnection.

Fig. 45.16 Typical appearance after intervention in necrosis. The

collection (arrows) contains non-dependent air and debris.

Fig. 45.17 Severe pancreatic ductal disrup-

A tion identied at the time of endoscopic
B
drainage of necrosis. A,B Initial injection
shows short normal pancreatic duct that
then disrupts into necrotic cavity.

484
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis
AP CP
Successful resolution 23/31 (74%) 59/64 (92%)
Complications 6/31 (19%) 11/64 (17%)
Recurrence 2/23 (9%) 7/59 (12%)
Hospital days 9 3
Table 45.2 Outcomes after attempted endoscopic drainage of patient uid collections
AP, Acute pseudocyst; CP, chronic pseudocyst; PN, pancreatic necrosis.
Source: Baron TH, Harewood GC, Morgan DE, Yates MR.
Gastrointest Endosc. 2002 Jul;56(1):717 (with permission)
The indications forand the timing ofdrainage of sterile
pancreatic necrosis are controversial. Pancreatic necrosis is not
amenable to endoscopic drainage until the process becomes orga-
nized, which usually occurs several weeks after onset of pancreatitis.
If the process remains sterile, the general indications for drainage
are refractory abdominal pain, gastric outlet obstruction or failure to
thrive (continued systemic illness, anorexia, and weight loss) at 4 or
more weeks after the onset of acute pancreatitis. The severity of CT
scan ndings alone is not an indication for drainage. Since endo-
scopic drainage of these collections is more technically difcult,
carries a higher rate of complications, and tends to involve a more
severely ill patient group (Table 45.2), the decision to endoscopically
intervene in patients with sterile pancreatic necrosis must be care-
fully considered.25 Alternative management options to endoscopic
drainage include nutritional support with parenteral or enteral
jejunal feeding and non-endoscopic drainage methods such as per-
cutaneous or surgical drainage. The nal management option is
usually based upon local expertise and severity of comorbid medical
illnesses. Ideally, these patients are best managed by a multidisci-
plinary approach in tertiary centers.
Infected pancreatic necrosis is considered an indication for
drainage. Infected necrosis may not be distinguishable clinically
from sterile necrosis because of leukocytosis and fever. Percutane-
ous ne needle aspiration may be required to determine the bacte-
riologic status of the necrosis, especially when the decision to
intervene is predicated on infection. Surgical therapy is still consid-
ered the gold standard when a debridement is felt to be necessary
in the management of pancreatic necrosis. In some centers, place-
ment of large-bore percutaneous drains is undertaken as an alterna-
tive to surgery and in preference to endoscopic therapy as described
below.
Endoscopic drainage of organized
pancreatic necrosis
Because of the need to evacuate solid material, the endoscopic
approach to drainage of organized pancreatic necrosis differs from
drainage of other PFCs. In general, the transpapillary approach is
not adequate to allow removal of solid debris. Therefore, the trans-
mural drainage approach is recommended for these collections.
After entry into the collection, the gastric or duodenal wall is balloon
dilated to a diameter 15 mm (Fig. 45.18a). This allows egress of
solid material around the endoprostheses. Several approaches, alone
or in combination, can be used to evacuate solid debris. The rst is
to place an irrigation system to lavage the solid debris. This is
achieved by placing a 7 Fr naso-irrigation tube (standard nasobiliary
tube) into the collection alongside the transmural stents (Figs 45.18b
and c). Up to 200 cc of normal saline is forcefully and rapidly infused
via the tube every two to four hours initially. In patients who are
PN AP vs. CP AP vs. PN CP vs. PN
31/43 (72%) p = 0.02 NS p = 0.006
16/43 (37%) NS NS p = 0.02
9/31 (29%) NS NS p = 0.047
20 p = 0.0003 NS p = 0.0001
intolerant to nasocystic irrigation tubes and/or it is anticipated that
irrigation may be required for many weeks, an alternative to naso-
cystic lavage is the placement of a percutaneous endoscopic gastros-
tomy tube (PEG) with placement of a jejunal extension tube into
the collection26 (Fig. 45.19). The gastric port may then be used for
supplementing nutritional needs. More recently, is the description
of placing two PEG tubes. One is placed for irrigation, as described
above. Over the course of endoscopic therapy, the authors described
upsizing the PEG debridement tube to 20 Fr for even greater irriga-
tion. The second PEG was placed to allow for jejunal feeding through
an extension tube.27
Another approach for removal of necrotic debris is to perform
endoscopic debridement. Indirect endoscopic debridement can be
performed under uoroscopic guidance by passing catheters trans-
murally into the collection to irrigate and remove debris with stone
retrieval baskets, the Roth Basket, and/or balloons. Direct endo-
scopic debridement can be performed by passing forward-viewing
endoscopes through the balloon-dilated transmural tracts into the
collection (Fig. 45.20).28 Baskets and/or grasping forceps are then
used to remove solid debris (Figs 45.21ac).
Regardless of the approach, repeat procedures are needed to re-
dilate the transmural tract as it closes, exchange transmural cathe-
ters, debride the collection, and to treat pancreatic ductal disruptions.
These procedures may be scheduled or on-demand based upon
clinical status and/or CT scan ndings.
Patients undergoing endoscopic drainage sterile organized pan-
creatic necrosis should receive preprocedural antibiotics. We recom-
mend intravenous carbapenem agents such as imipenem or
meropenem or an extended penicillin agent such as piperacillin.
Those with infected necrosis should continue to receive antibiotics,
either empirically or based upon culture data. Outpatients should be
admitted to the hospital post-procedurally for observation and insti-
tution of irrigation. Patients are discharged home after they are able
to tolerate oral intake and care for irrigation tubes. Post-procedurally
oral antibiotics and antifungal agents (e.g. ciprooxacin and ucon-
azole) are administered, and irrigation is continued until the collec-
tion has resolved as documented by follow-up CT. CT scans are
obtained weekly or every other week to follow the progress of drain-
age and to guide need for further debridement. The internal drains
are endoscopically removed several weeks after complete resolution
of the collection.
Recently, we have changed our strategy for the management of
organized (walled-off) pancreatic necrosis. After successful entry
and dilation of the transmural tract to at least 15 mm we now rou-
tinely advance a forward-viewing endoscope into the necrotic collec-
tion at the time of the rst procedure. Schedule debridements are
performed on a weekly basis with follow-up CT performed weekly
as well. Naso-irrigation tubes are placed into the collection and
485
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
removed when all of the necrotic material has been evacuated
endoscopically.
RESULTS OF ENDOSCOPIC THERAPY OF
PANCREATIC FLUID COLLECTIONS
It must be emphasized that there are no prospective studies compar-
ing endoscopic drainage to conservative (medical) therapy, percuta-
neous drainage or surgical drainage.
Pancreatic pseudocysts
The success rates, recurrence rates and complication rates following
endoscopic drainage of pancreatic pseudocysts are variable, likely
because the patient populations and interventions in most of these
486
Fig. 45.18 Endoscopic transmural drain-
age of necrosis. Patient depicted in Figure
45.14. A The transmural tract is dilated with
a 16 mm balloon. B Double pigtail stents
and a naso-irrigation tube (arrow at tip) are
placed. C Illustration of transmural drainage
and naso-irrigation catheter.
series are heterogeneous, comprised of acute and chronic pseudo-
cysts, pancreatic abscesses, and some patients have undergone
transpapillary drainage while others transmural drainage methods.
Nonetheless, cumulatively successful drainage is achieved in approx-
imately 7590% with complication rates of about 510% and recur-
rence rates of 520%.1,15,29 The results of endoscopic therapy compare
favorably to surgical therapy. Although percutaneous drainage of
pancreatic pseudocysts has a high success rate for resolution, it is
can create external stula formation if the pseudocyst communicates
with the main pancreatic duct. In a recent study in which a national
database was used to compare percutaneous drainage and surgical
drainage of pancreatic pseudocysts, there was a signicantly lower
length of stay and inpatient mortality (5.9% vs 2.8%) in the surgical
group.30
Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis

Fig. 45.19 Illustration of PEG with jejunal extension placed

through the posterior gastric wall for irrigation.

Fig. 45.20 Illustration of direct endo-

scopic debridement using a forward-
viewing endoscope.

487
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
A B
Fig. 45.21 Direct endoscopic debridement. A The endoscope is just within the cavity through a large transmural tract. A pelican forceps
can be seen grasping the solid material. B The necrotic material is withdrawn and deposited into the antrum. C Large amount of necrotic
debris is seen in the stomach at the end of the procedure.
In terms of transmural drainage results, there may be a slightly
lower pseudocyst recurrence rate following transduodenal drainage
than transgastric drainage due to the sustained patency of a trans-
duodenal stula that allows long-term drainage of the main pancre-
atic duct, although this has not been proven. Therefore, in a patient
with severe underlying and irreversible pancreatic ductal abnormali-
ties, I recommend transduodenal drainage, whenever possible.
Pancreatic abscesses
When a broad denition of pancreatic abscess is taken to include
infected pseudocysts and other infected liqueed collections without
necrosis, success rates following endoscopic drainage are high,
although there are few series and small patient numbers.19,3133
Organized pancreatic necrosis (walled-off necrosis)
There are limited series of endoscopic treatment of pancreatic
necrosis. Successful non-surgical resolution of pancreatic necrosis
(symptomatic sterile and infected) was published by our group with
a non-surgical success rate of 72% of 43 patients.34 At that time
smaller-sized transmural dilating balloons of 810 mm were used.
Outcome after changing the drainage strategy to the use of large-
diameter balloon dilation of the transmural entry tract to >15 mm
at the initial procedure has not yet been published, although in a
series of 53 patients who underwent transoral/transmural endo-
scopic therapy treated from 1998 to 2006, our non-surgical success
rate was 81% (in press). Predictors of failure included diabetes mel-
litus and a collection size = 15 cm. The presence of paracolic gutter
involvement was also associated with failure as well as a need for
adjuvant percutaneous drain placement. Other authors have
reported results of small numbers of patients who underwent endo-
scopic therapy for necrosis. Success rates of only 4/8 (50%)18 were
reported using standard irrigation methods. Using aggressive
direct debridement, the results have been better with complete res-
olution in 5/5 patients29 and 2/2 patients,28 although many endo-
scopic procedures per patient were required to achieve this result in
the former group.26
488
C
OUTCOME DIFFERENCES FOLLOWING
ENDOSCOPIC DRAINAGE OF PANCREATIC
FLUID COLLECTIONS
Based upon published results18,34 signicant differences in success-
ful endoscopic drainage have been noted between patients with
pseudocysts and pancreatic necrosis (Table 45.2). Complications
occur more commonly in patients with pancreatic necrosis than for
pseudocysts. Likewise, hospital stay is shorter in patients with pseu-
docysts, while pancreatic necrosis is predictive of signicantly longer
hospital stays. The recurrence rates are signicantly higher for
patients with pancreatic necrosis than chronic pseudocysts.
The differences in success rates, complication rates, recurrences
and hospital stay may be explained by the differences in pathology,
pathophysiology, and severity of illness between the groups. Patients
with pancreatic necrosis tend to be more severely ill patients and
endoscopic evacuation of solid debris is less efcient than evacuation
of liquid. In terms of recurrence rates, acute pancreatic ductal dis-
ruptions occurring in patients with necrosis frequently lead to
either severe stricturing or a completely disconnected duct whereby
the head and tail of the pancreas are not in communication (Fig.
45.22). Recurrent collections may arise from the undrained viable
pancreatic tail. Patients with acute pseudocysts tend to have less
severe ductal abnormalities and less recurrence, while patients with
chronic pancreatitis have underlying ductal abnormalities, such as
strictures and stones, that may lead to recurrences, especially if left
unrecognized and untreated.34,35 I recommend aggressive endo-
scopic intervention to correct underlying ductal abnormalities, if
possible, in all types of PFCs so that recurrent collections or symp-
toms may be avoided (Figs 45.5d and e).
ROLE OF ENDOSCOPIC EXPERIENCE
Endoscopic therapy of PFCs requires a high skill level. Operator
experience may play a role in the outcome of these patients as there
appears to be a learning curve associated with drainage of PFCs.36
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis

Animal training models for learning pseudocyst drainage techniques
have been described37 and may be helpful for acquiring these
techniques.
COMPLICATIONS OF ENDOSCOPIC THERAPY OF
PANCREATIC FLUID COLLECTIONS
Life-threatening complications may arise following attempted endo-
scopic drainage of PFCs and are listed in Table 45.3. It is recom-
mended that endoscopic drainage of PFCs is performed with the
availability of surgical and interventional radiology support. The
most feared complications of transmural drainage are bleeding and
perforation. Bleeding after transmural drainage may be managed
supportively, endoscopically, surgically, or with angiographic embo-
lization. If perforation occurs during attempted transgastric drain-
age and is limited to the gastric wall (does not involve the collection),
it may be successfully managed non-surgically, assuming a stent has
not been placed through the perforation; the gastric wall rapidly
closes with conservative treatment consisting of nasogastric suction
and antibiotics. Some authors feel that transduodenal perforation
may be managed conservatively, since the perforation is retroduode-
nal,38 although this is not proven. Infectious complications usually
occur from inadequate drainage of uid and/or solid debris. If trans-
papillary drainage was performed on a liqueed collection, stent
exchange and/or upsizing of the stent may resolve the infection. If
solid material was unrecognized during the initial procedure, place-
ment of irrigation tubes or changing to transmural drainage may
resolve the infection. Occasionally, some patients will require adju-
vant placement of percutaneous drainage and/or irrigation catheters
to manage infectious complications, particularly when the necrosis
extends to the paracolic gutters. Stent migration into the collection

BOX 45.3 COMPLICATIONS AND

CONTROVERSIES

Fig. 45.22 Indirect endoscopic debridement of patient in Figure Complications of endoscopic drainage of pancreatic uid
45.14. Stone retrieval balloon was inated inside the cavity and solid collections include perforation, bleeding, infection,
debris swept out of the transmural site. respiratory complications, pancreatic ductal damage, and
stent migration

Endoscopic drainage of pancreatic necrosis is controversial

and associated with a higher complication rate than drainage
of pancreatic pseudocysts

Bleeding
Perforation
Infection
Pancreatitis
Sedation complications
Aspiration
Fig. 45.23 Final pancreatogram of patient in Figure 45.14. The Stent migration/occlusion
necrotic cavity resolved as documented by CT. Extravasation of
contrast is seen from the pancreatic duct back into the duodenum Pancreatic ductal damage
at the transmural entry site. There is no communication with the
tail. Table 45.3 Complications of endoscopic therapy of PFCs

AP CP PN AP vs. CP AP vs. PN CP vs. PN

Successful resolution 23/31 (74%) 59/64 (92%) 31/43 (72%) p = 0.02 NS p = 0.006
Complications 6/31 (19%) 11/64 (17%) 16/43 (37%) NS NS p = 0.02
Recurrence 2/23 (9%) 7/59 (12%) 9/31 (29%) NS NS p = 0.047
Hospital days 9 3 20 p = 0.0003 NS p = 0.0001

Table 45.4 Outcomes after attempted endoscopic drainage of patient uid collections
AP, Acute pseudocyst; CP, chronic pseudocyst; PN, pancreatic necrosis.

489
 SECTION 3 APPROACH TO CLINICAL PROBLEMS
through the gastric or duodenal wall may occur during or after
endoscopic stent placement. Endoscopic retrieval is possible if the
collection has not completely collapsed and the transmural tract is
still patent.
REFERENCES
1. Baron TH. Endoscopic drainage of pancreatic uid collections and
pancreatic necrosis. Gastrointest Endosc Clin N Am. 2003;
13(4):743764.
2. Kloppel G. Pathology of severe acute pancreatitis. In Bradley EL,
III (ed.) Acute pancreatitis: diagnosis and therapy. New York:
Raven Press, 1994:3546.
3. Hariri M, Slivka A, Carr-Locke DL, et al. Pseudocyst drainage
predisposes to infection when pancreatic necrosis is
unrecognized. Am J Gastroenterol 1994; 89:17811784.
4. Soliani P, Franzini C, Ziegler S, et al. Pancreatic pseudocysts
following acute pancreatitis: risk factors inuencing therapeutic
outcomes. JOP. 2004; 5(5):338347.
5. Andren-Sandberg A, Dervenis C. Pancreatic pseudocysts in the
21st century. Part I: classication, pathophysiology, anatomic
considerations and treatment. JOP. 2004; 5(1):824.
6. Andren-Sandberg A, Dervenis C. Pancreatic pseudocysts in the
21st century. Part II: natural history. JOP 2004; 5(2):6470.
7. Kim YH, Saini S, Sahani D, et al. Imaging diagnosis of cystic
pancreatic lesions: pseudocyst versus nonpseudocyst.
Radiographics 2005; 25(3):671685.
8. Beckingham IJ, Krige JE, Bornman PC, et al. Endoscopic
management of pancreatic pseudocysts. Br J Surg 1997; 84:
16381645.
9. Brugge WR. Approach to cystic pancreatic lesions. Gastrointest
Endosc Clin N Am. 2005; 15(3):485496, viii.
10. Balthazar EJ. Pancreatitis associated with pancreatic carcinoma.
Preoperative diagnosis: role of CT imaging in detection and
evaluation. Pancreatology. 2005; 5(45):330344.
11. Nishimura T, Masaoka T, Suzuki H, et al. Autoimmune pancreatitis
with pseudocysts. J Gastroenterol 2004; 39(10):10051010.
12. Levy MJ, Smyrk TC, Reddy RP, et al. Endoscopic ultrasound-
guided trucut biopsy of the cyst wall for diagnosing cystic
pancreatic tumors. Clin Gastroenterol Hepatol 2005; 3(10):
974979.
13. Morgan DE, Baron TH, Smith JK, et al. Pancreatic uid collections
prior to intervention: evaluation with MR imaging compared with
CT and US. Radiology 1997; 203(3):773778.
14. Telford JJ, Farrell JJ, Saltzman JR, et al. Pancreatic stent
placement for duct disruption. Gastrointest Endosc 2002; 56(1):
1824.
15. Kahaleh M, Shami VM, Conaway MR, et al. Endoscopic ultrasound
drainage of pancreatic pseudocyst: a prospective comparison
with conventional endoscopic drainage. Endoscopy 2006;
38(4):355359.
16. Yusuf TE, Baron TH. Endoscopic transmural drainage of
pancreatic pseudocysts: results of a national and an international
survey of ASGE members. Gastrointest Endosc. 2006; 63(2):
223227.
17. Hookey LC, Debroux S, Delhaye M, et al. Endoscopic drainage of
pancreatic-uid collections in 116 patients: a comparison of
etiologies, drainage techniques, and outcomes. Gastrointest
Endosc 2006; 63(4):635643.
18. Giovannini M, Pesenti C, Rolland AL, et al. Endoscopic ultrasound-
guided drainage of pancreatic pseudocysts or pancreatic
490
Endoscopic therapy may be associated with complications and/or
failures that require surgical management. It is possible that the
outcome of surgical therapy may be adversely altered when com-
pared to those patients undergoing primary surgical therapy.39
abscesses using a therapeutic echo endoscope. Endoscopy 2001;
33(6):473477.
19. Azar RR, Oh YS, Janec EM, et al. Wire-guided pancreatic
pseudocyst drainage by using a modied needle knife and
therapeutic echoendoscope. Gastrointest Endosc 2006; 63(4):
688692.
20. Kruger M, Schneider AS, Manns MP, et al. Endoscopic
management of pancreatic pseudocysts or abscesses after an
EUS-guided 1-step procedure for initial access. Gastrointest
Endosc 2006; 63(3):409416.
21. Sriram PV, Kaffes AJ, Rao GV, et al. Endoscopic ultrasound-guided
drainage of pancreatic pseudocysts complicated by portal
hypertension or by intervening vessels. Endoscopy 2005;
37(3):231235.
22. Monkemuller KE, Baron TH, Morgan DE. Transmural drainage of
pancreatic uid collections without electrocautery using the
Seldinger technique. Gastrointest Endosc 1998; 48(2):195200.
23. Sanchez Cortes E, Maalak A, Le Moine O, et al. Endoscopic
cystenterostomy of nonbulging pancreatic uid collections.
Gastrointest Endosc 2002; 56(3):380386.
24. Cahen D, Rauws E, Fockens P, et al. Endoscopic drainage of
pancreatic pseudocysts: long-term outcome and procedural
factors associated with safe and successful treatment. Endoscopy
2005; 37(10):977983.
25. Kozarek RA. Endoscopic management of pancreatic necrosis: not
for the uncommitted. Gastrointest Endosc 2005; 62(1):101104.
26. Baron TH, Morgan DE. Endoscopic transgastric irrigation tube
placement via PEG for debridement of organized pancreatic
necrosis. Gastrointest Endosc 1999; 50(4):574577.
27. Raczynski S, Teich N, Borte G, et al. Percutaneous transgastric
irrigation drainage in combination with endoscopic
necrosectomy in necrotizing pancreatitis (with videos).
Gastrointest Endosc 2006; 64(3):420424.
28. Seewald S, Groth S, Omar S, et al. Aggressive endoscopic therapy
for pancreatic necrosis and pancreatic abscess: a new safe and
effective treatment algorithm (videos). Gastrointest Endosc 2005;
62(1):92100.
29. Weckman L, Kylanpaa ML, Puolakkainen P, et al. Endoscopic
treatment of pancreatic pseudocysts. Surg Endosc 2006;
20(4):603607.
30. Morton JM, Brown A, Galanko JA, et al. A national comparison of
surgical versus percutaneous drainage of pancreatic pseudocysts:
19972001. J Gastrointest Surg 2005; 9(1):1520; discussion
2021.
31. Park JJ, Kim SS, Koo YS, et al. Denitive treatment of pancreatic
abscess by endoscopic transmural drainage. Gastrointest Endosc
2002; 55(2):256262.
32. Venu RP, Brown RD, Marrero JA, et al. Endoscopic transpapillary
drainage of pancreatic abscess: technique and results. Gastrointest
Endosc 2000; 51(4 Pt 1):391395.
33. Baron TH, Harewood GC, Morgan DE, et al. Outcome differences
after endoscopic drainage of pancreatic necrosis, acute pancreatic
pseudocysts, and chronic pancreatic pseudocysts. Gastrointest
Endosc 2002; 56(1):717.
 Chapter 45 Endoscopic Drainage of Pancreatic Pseudocysts, Abscesses and Organized (Walled-Off) Necrosis
34. Zhang AB, Zheng SS. Treatment of pancreatic pseudocysts in line
with DEgidios classication. World J Gastroenterol 2005;
11(5):729732.
35. Nealon WH, Walser E. Main pancreatic ductal anatomy can direct
choice of modality for treating pancreatic pseudocysts (surgery
versus percutaneous drainage). Ann Surg 2002; 235(6):
751758.
36. Harewood GC, Wright CA, Baron TH. Impact on patient
outcomes of experience in the performance of endoscopic
pancreatic uid collection drainage. Gastrointest Endosc 2003;
58(2):230235.
37. Scho R, Buchmeier B, Hauder G. Adaptation of the Erlangen
Active Simulator for Interventional Endoscopy (EASIE) model for
transmural pancreatic pseudocyst drainage. Endoscopy 2006;
38(1):100.
38. Beckingham IJ, Krige JE, Bornman PC, et al. Endoscopic
management of pancreatic pseudocysts. Br J Surg 1997;
84(12):16381645.
39. Nealon WH, Walser E. Surgical management of complications
associated with percutaneous and/or endoscopic management of
pseudocyst of the pancreas. Ann Surg 2005; 241(6):948957;
discussion 957960.
491
 Index
A alkaline phosphatase (AP) 265
accessories for the ERCP room 17, 3141
disposable 40
evolution of 41
single use vs reusable 40
storage 16, 40
for use in altered anatomy patients 40
see also individual accessories
access (precut) sphincterotomes 323
access (precut) sphincterotomy 8790, 110
techniques 77
see also precut accessotomy
AccuTouch computer simulator 63
acinarization 26
Acosta criteria 415
acute gallstone pancreatitis (AGP)
background 411
biliary intervention 41117
with biliary obstruction 41415
cholecystectomy after 41516
diagnosis 41112
endoscopic therapy 41214
management algorithm 416
randomized controlled trials (RCTs) 41314
severity assessment 412
systematic reviews 41416
acute pancreatic trauma 424
acute pancreatitis
causes 435, 441
in children 2278
concomitant 330
endoscopically amenable lesions 41920
episodic illness 438
ERCP in 419
idiopathic see idiopathic acute pancreatitis
unexplained 370
see also acute gallstone pancreatitis
adenomas
ampullary see ampullary adenomas
cystadenomas see cystadenomas; serous
cystadenomas; mucinous cystic
neoplasms
major papilla 440
administrators 5
air, retroperitoneal/intraperitoneal 21
air bubbles
after sphincterotomy 26
vs gallstones 26
airway problems, risk factors 45
alanine aminotransferase (ALT) 265, 411
albendazole 3934, 394, 396
alcoholism
causing acute pancreatitis 435
and chronic pancreatitis (CP) 459, 464
transabdominal ultrasound 275
allergies see contrast allergies treatment 27983
allopurinol 56 see also carcinoid/neuroendocrine tumors
altered anatomy, accessories for use with 40 ampullary tumors 189, 440
alveolar hypoventilation 46 literature on endoscopic resection of 2812
American Society for Gastrointestinal Endoscopy surgical treatment 189
(ASGE) see also ampullary carcinoma; ampullary
guidewire review 35 neoplasm/ampullary adinoma
pregnancy guidelines 231, 234 ampullectomy 18997
training guidelines 7, 61, 62 complications 194, 195
American Society of Anesthesiologists (ASA) contraindications 194
patients physical classication system 43, 44 controversies 1945
practice guidelines 43, 448 costs 196
amoxicillin 305 en bloc 191
ampicillin 220 endoscopic evaluations 18990
ampulla of Vater 87, 88 indications 194
evaluation 18990 postampullectomy cholangitis 194
lymphomas 278 postampullectomy pancreatitis 1934, 195
neoplasms 2789 postampullectomy sphincterotomy 1934
ampullary adenomas 273, 277 postampullectomy surveillance 194
colonoscopy 2767 preresection sphincterotomy 1923
treatment 27980 snare excision 1912
ampullary anomalies 447 specimen retrieval 192
ampullary carcinoma 288 stent placement 1934
classication 277 success rates 1956
colonoscopy 2767 techniques 18994
survival rates 2803 thermal ablation 192
symptoms and signs 273 see also papillectomy
transabdominal ultrasound (TUS) 288 analgesia see sedation and analgesia
treatment 2803 anastomosis
ampullary diverticula 79 dunking anastomosis 349
ampullary neoplasm 27385 with more than one exiting lumen 256
adenomas see ampullary adenomas anastomotic strictures, in children 226
appearances 273 anatomy, difcult, EST and 11213
benign vs malignant 2734 anesthesia
carcinoma see ampullary carcinoma induction agents 478
colonoscopy 2767 mortality rates 434
computed tomography (CT) scan 275 animals, training using 636
diagnosis 275 annular pancreas 441, 4567
differential diagnosis 278, 279 anomalous pancreaticobiliary junction (APBJ)
endoscopic ultrasonography (EUS) 2756 387, 441, 447
endoscopy 2734 biliary cysts and 448, 449
ERCP and 2745 gallbladder cancer and 392, 447
familial adenomatous polyposis (FAP) 189, anomalous pancreatobiliary union (APBU), in
273, 274, 277, 279 children 223
forceps biopsy 275 ansa pancreaticus 92, 375
gastrointestinal stromal tumors (GIST) 278 antegrade sphincterotomy 113
intraductal ultrasound (IDUS) 276 antibiotics
lymphomas 278 in cholangitis 359
magnetic resonance imaging (MRI) 275 in hilar malignant biliary obstruction 304,
pathology 2779 309
staging advanced cancer 2756 pancreatic sphincterotomy and 129
staging early lesions 276 in primary sclerosing cholangitis 379, 380,
symptoms and signs 273 382
493
 INDEX

anticoagulants, pancreatic sphincterotomy and in infants 220 pancreatic duct stent placement to facilitate
129 lithotripsy baskets 39, 1237, 126 767
antiplatelet drugs 114 spiral 122 with a sphincterotome 756
Apache II scoring system 412 wire 39, 1212 standard techniques 756
aprons, lead 16, 19 basket stone extraction 1213, 128 biliary cast syndrome 343
articial tissue models 66 assessing stone size 119 biliary cirrhosis 313
ASA see American Society of Anesthesiologists complications 121, 123 biliary cysts see choledochal (biliary) cysts
ascariasis contraindications 121, 123 biliary decompression, stent placement 323
biliary 393 indications 121, 123 biliary dilation, denition 263
biliary-pancreatic 393 key points 121 biliary disorders, in pregnancy 234
complications 393 technique 1223 biliary dyskinesia, denition 367
infestation in children 226 battery (criminal charge) 6 biliary lling defects 21314
in pregnancy 231, 234, 393 B-cell lymphomas 278 biliary infections, in children 226
Ascaris 231 benign biliary strictures 32733 biliary leaks 234, 26, 3357
Ascaris lumbricoides 228, 3934, 399 causes 327, 328 after hepatic resection 3389
endotherapy 3934 chronic pancreatitis and 327, 330 after laparoscopic cholecystectomy (LC) 337
worm extraction 393 classication 327 associated with liver transplantation 342
ascites, pancreatic see pancreatic ascites clinical features 327 bypassing bile ow through leak site 335
ASGE see American Society for Gastrointestinal complications 3302 in children 226
Endoscopy contraindications 330 ERCP treatment techniques 3356
aspartate aminotransferase (AST) 265 controversies 3302 maintaining intraductal ow 3356
aspirin 56, 129 diagnosis 327 sealing 335
atropine 16 endoscopic technique 328 biliary lymphomas 278
autoimmune pancreatitis 441 indications 330 biliary malignancies see cholangiocarcinoma;
management 3278 malignant biliary obstruction, distal;
B postoperative 327 malignant biliary obstruction, hilar
bacteria, in cholangitis 359 post-sphincterotomy distal 330 biliary mechanical lithotripsy (BML) 362
Bacteroides fragilis, in cholangitis 359 SEMS, uncovered vs covered in 330 biliary microlithiasis 415
balloon dilation stenting 32832 biliary obstruction 313
after biliary sphincterotomy 55 stricture dilation 328 benign 160
of the biliary sphincter 55 stricture negotiation 328 distal 1537
in Billroth II gastrectomy 2567 surgery 32930 hilar 158
endoscopic 98100, 113 benzodiazepines 44, 91 malignant 160, 165
informed consent 97 in pregnancy 233 causes 165
of the papilla 97107, 231 in sphincter of Oddi manometry (SOM) palliation 165
see also endoscopic papillary balloon dilation; 372 see also malignant biliary obstruction,
large balloon dilation after minimal bile duct distal; malignant biliary obstruction,
biliary sphincterotomy; balloon adenocarcinoma 214 hilar
sphincteroplasty anomalous anatomy 4478 MRCP and 267
balloon dilators 38 decreasing pressure inside 335 perihilar, detection rates 268
balloons dilated see dilated bile duct see also indeterminate biliary strictures
extraction balloons 11920, 121, 123 extrahepatic 105, 269 biliary papillomatosis 214, 216
over-the-guidewire type 105 extravasation from 26 biliary scintigraphy, dilated bile duct 269
balloon sphincter dilation, Billroth II gastric leaks see biliary leaks biliary sludge 436
resection 2412 lesion evaluation 21315 biliary sphincter, balloon dilation 55
balloon sphincteroplasty lesion therapy 216 biliary sphincterotomy 460
in children 222 opacication 224 balloon dilation after 55
in pregnancy 231 perforation 116 Billroth II gastric resection 2412
balloon stone extraction 11921 postoperative injuries 327 immediately before pancreatic
assessing stone size 119 stones see bile duct stones sphincterotomy 134
complications 119, 1212 tumors 214 as indication of EST 114
contraindications 119, 120 bile duct stones 357 precut 779
extraction balloons 11920, 121, 123 common bile duct (CBD) 357, 362, 363 in sphincter of Oddi dysfunction (SOD) 114,
indications 119, 120 extrahepatic, EST for 105 374, 3756
key points 119 found during laparoscopic cholecystectomy see also endoscopic biliary sphincterotomy
technique 1201 (LC) 3634 biliary stenting, complications 3302
balloon traction, stent removal 17980 irretrievable, stents for 157 biliary stents
bariatric surgery 2459 post-EST/EPBD 103 in acute cholangitis 360
biliopancreatic diversion 245, 246 retained after cholecystectomy 33940 categories 165
duodenal switch 246 bile microscopy 442 complications 162
gastric banding 246, 247 biliary 38 xed-diameter plastic (FDPS) see xed-
gastric bypass 2479 biliary anomalies 4479 diameter plastic stents
malabsorptive-jejunoileal bypass 246 biliary atresia (BA), diagnosis in children 222, palliative insertion 153
restrictive surgery 2467 226 plastic 1538
vertical banded gastroplasty 245, 2467 biliary biopsy forceps 38 with antireux valve 153, 185
basket mechanical lithotripsy (BML) 360 biliary cannulation 756 indications 160, 177
baskets double-wire technique 77 technique 153
cost 124 guidewire-assisted 76 self-expandable metallic (SEMS) see self-
ower baskets 122 methods 75 expandable metallic stents

494
 INDEX

biliary strictures cannulas 312 parasitic disease and 396

benign see benign biliary strictures cannulation risk factors 287, 299
in children 2236 of common bile duct see biliary cannulation stenting 384
evaluation of 21315 of major papilla 7381 technique 3834
indeterminate see indeterminate biliary of minor papilla see minor papilla treatment 384
strictures of pancreatic duct 801 cholangiograms 119
biliary surgery 2525 parallel 80 cholangiography
cholecystojejunostomy 254 patients with periampullary duodenal dilated bile duct 2689
choledochoduodenostomy 252 diverticula 7980 extrahepatic biliary lesions 317
complications see biliary surgery capnography 46 identifying malignant lesions 317
complications carbapenem agents 485 percutaneous transhepatic (PTC) 268, 269,
hepaticocutaneous jejunostomy 255 carbohydrate antigen 19-9 (CA 19-9), in 319
liver transplantation 2545 indeterminate biliary strictures 313 cholangiopancreatoscopy, duodenoscope-assisted
Roux-en-Y choledochojejunostomy 254 carbon dioxide (CO@2) monitoring 46 39
Roux-en-Y hepaticojejunostomy 2524 carcinoids see neuroendocrine tumors cholangiopathy, HIV associated, in children
biliary surgery complications, management of carcinoma 226
33545 ampullary see ampullary carcinoma cholangioscopes 39
collections 338 gallbladder, prognosis and risk factors 287 maneuvering 213
diagnosis 3378 neuroendocrine 2778 cholangioscopic stone removal
endoscopic vs surgical treatment 338 see also cholangiocarcinoma; pancreatic basket removal 4034
ERCP treatment techniques 3356 cancer results 406
following laparoscopic cholecystectomy 336 cardiopulmonary complications of ERCP 512 stone fragmentation 404
injury, nature and magnitude of 3367 cardiovascular complications of ERCP 9 stricture dilation 4046
Kehrs tube 340 care techniques 4036
septic complications 338 duty of 4 cholangioscopy 21117
summary 344 standards of 45 accessories used 211
surgical risk 338 C-arm uoroscopy units bile duct preparation 213
see also biliary leaks benets and disadvantages 19 biopsy and 21415
biliary tract disease 442 scatter radiation 19 brush cytology 215
biliopancreatic diversion 245, 246 space in ERCP room for 13 cannulation 21213
bilirubin 265 Carolis disease 287, 390, 448 complications 21617
Billroth I gastrectomy 237 catheters contraindications 216
Billroth II anatomy manometry 92, 94 cost 217
endoscopic biliary sphincterotomy (EST) and with microtransducers 94 diagnostic technique 21112
110, 112 needle tip 31, 82 with uoroscopy 213
endoscopic papillary balloon dilation (EPBD) pancreaticobiliary manometry catheters 31 without uoroscopy 21516
and 100 sleeve 95 indeterminate biliary strictures 3223
Billroth II gastrectomy 23742, 2567 in sphincter of Oddi manometry (SOM) indications 21316
EPBD/EST and 98100 956, 371 malignancy criteria 214
ERCP accessories 2567 swing-tip 31 nomenclature evolvement 211
bilobar stone disease 412 tapered-tip 31, 95 peroral 399400
biopsy types 312 postoperative (POCS) 4023
in children 226 cefazolin 220 preprocedure room set-up 211
cholangioscopy 21415 cephalosporin 129, 220 stone fragmentation 215
image of 27 Charcots triad 359 stone removal see cholangioscopic stone
malignant biliary obstruction, distal 289 chemotherapy removal
pancreatoscopy and 2034 in cholangiocarcinoma (CA) 384 techniques 211
see also forceps biopsies in malignant biliary obstruction 2956 transhepatic 216
bipolar electrocautery, in pregnancy 233 Child-Pugh cirrhosis 101 see also percutaneous transhepatic
bispectral index (BIS) monitoring 46 children cholangioscopy
Bithionol 396 choledochal cysts 387, 389 cholangitis
bleeding see hemorrhage ERCP in see pediatric ERCP acute, morbidity and mortality 359
botulinum toxin injection 455, 456 Chinese liver uke see Clonorchis sinensis acute suppurative, radiological contrast 359
in recurrent pancreatitis 438 chloroquine 396 as complication of choledochal cysts 391
in sphincter of Oddi dysfunction (SOD) 374 cholangiocarcinoma (CA) 214, 268, 290, 299, as complication of ERCP 52, 58
bougies 38 3835 due to choledocholithiasis/ductal stenosis
bowel perforation 57 biliary cysts and 4489 114
Braun procedure 2401, 245, 256 complications 385 endoscopic stenting 361
breach of duty of care 4 contraindications 384 EST-related 116
brush cytology cost 385 postampullectomy 194
in children 226 diagnosis 3834 post-EST/EPBD 101, 103
cholangioscopy 215 extrahepatic 299 recurrent pyogenic (RPC) see recurrent
devices 38 hilar see hilar cholangiocarcinoma pyogenic cholangitis
indeterminate biliary strictures 320 incidence 287 reducing risk 309
pancreatoscopy 2034 indications 384 risk factors 58
mucin-hypersecreting 214 sclerosing see primary sclerosing cholangitis
C occurrence rate 299 secondary to choledocholithiasis 35961
CA 19-9 see carbohydrate antigen 19-9 (CA 19-9) palliation 384 symptoms 359
calcium channel blockers 373

495
 INDEX

cholecystectomy idiopathic 459 in diagnosis of pancreatic stulas 421

after acute gallstone pancreatitis (AGP) pain in 459, 464 dilated bile duct 266
41516 pancreatic sphincterotomy as primary hilar cholangiocarcinoma 3001
after endoscopic sphincterotomy 3645 treatment 13840 indeterminate biliary strictures 314
common bile duct dilation after 267 pathophysiology 459 pancreatic malignancies 28990
laparoscopic see laparoscopic cholecystectomy pretherapeutic planning 459 computer simulators 63
cholecystitis stenting 463 conscious sedation 43
as complication of ERCP 52, 58 see also severe chronic pancreatitis consent see informed consent
post-EST/EPBD 101, 103 ciprooxacin 220, 292, 3045, 359, 486 contraindications to ERCP 114
risk factors 293 cirrhosis, Child-Pugh 101 contrast
cholecystojejunostomy 254 cisplatin 2956 non-ionic 56
cholecystokinin 91 clonorchiasis 3956 strength of 20, 26
choledochal anomalies, in children 223 biliary 396 contrast allergies, policy 9
choledochal (biliary) cysts 287, 38792 endotherapy 396 contrast-enhanced abdominal CT, and pancreatic
anomalous pancreaticobiliary junction Clonorchis 231 uid collection 477
(APBJ) and 448, 449 Clonorchis sinensis 3956, 399 corticosteroids 55, 313, 3412
in children 223, 387, 389 clopidogrel 114 Crohns disease, in children 224
cholangiocarcinoma and 4489 Clostridium perfringens, in cholangitis 359 Cryptosporidium parvum 226
complications 3902 coagulation prole, and pancreatic uid cyanoacrylate
contraindications of ERCP for 387 collection 477 after hepatic resection 339
diagnosis 389, 449 coagulopathy, after hepatic resection 339 for external pancreatic stulas 425
extrahepatic 449 colonoscopes 31 sealing biliary/pancreatic leaks 335, 336,
indications of ERCP for 387 colonoscopy, ampullary neoplasm 2767 348
in pregnancy 234 common bile duct (CBD) 74 cystadenomas 471
preoperative assessment 389 cannulation see biliary cannulation serous see serous cystadenomas
technique 3878 dilation 263 cyst uid tumor markers 471
therapeutic measures 389 laparoscopic exploration of (LECBD) 3624 cystic endocrine neoplasms, pathology 469
Todani classication 387, 389, 448 size 2634 cystic brosis transmembrane receptor (CFTR)
see also choledochoceles stones 357, 362, 363 abnormalities 456
choledochoceles 389, 448 common hepatic duct (CHD) cystic lesions
cancer in 449 dilation 263 of the pancreas see pancreatic cystic lesions
IAP and 439 size 263 risk factors 467
choledochoduodenostomy 252 communication cysts see choledochal (biliary) cysts; hydatid cysts;
choledochojejunostomy 349 of endoscopists thoughts 26, 27 pancreatic pseudocysts
complications 348 with patient and family 10 cytology brush systems see brush cytology
dilatation of biliary strictures 3412 radiologic reporting 28 cytology sampling
choledocholithiasis 35766 CompactEASIE simulator 64, 656 cholangioscopy 21415
after cannulation and decompression of bile competence pancreatoscopy 2034
duct 35960 continued 7 cytomegalovirus 226
age of patients 361 denition 7 cytotoxic drugs 295
in children 2223 number of procedures to achieve/maintain
cholangitis secondary to 35961 7, 612 D
clinical manifestation 357 of patient, and consent 6, 9 damages 4
detection rates 268 privileging and 8 punitive 4
diagnosis 3579 trainees 7 diazepam 44, 305
endoscopic ultrasound and 267 complications of ERCP 518, 268 in pregnancy 233
as indication for biliary sphincterotomy 114 adverse events 51 diffuse large B-cell lymphomas 278
natural history 357 avoidance strategies 910 digital image analysis (DIA), indeterminate
patients cardiopulmonary 512 biliary strictures 321
high-risk 359, 362, 3623 cardiovascular 9 dilated bile duct 26372
intermediate-risk 3589, 363 denitions 51 approach to patient 26970
low-risk 358, 363 endoscopists experience and 54, 58 background 2634
risk stratication 3589 legal issues 89 biliary scintigraphy 269
peri-operative management of 3624 long-term 58 biochemical evaluation 265
in pregnancy 231, 234 managing 10 cholangiography 2689
surveillance after treatment 362 rates 512 clinical evaluation 2645
see also common bile duct, stones risk factors 8, 513 computed tomography (CT) 266
cholelithiasis risky clinical situations 10 endoscopic ultrasound (EUS) 2678
in children 2223 severity 51 etiology 264
in pregnancy 231 strategies for reduction of 58 evaluation 2649, 26970
see also gallstones unplanned events 51 imaging 2659
cholestasis computed tomography (CT) magnetic resonance imaging (MRI) 2667
markers of 265 ampullary carcinoma 288 ultrasound (US) 2656
neonatal 222 ampullary tumors 275 dilation
obstructive see obstructive cholestasis in choledocholithiasis diagnosis 357 pancreaticobiliary 38
chronic pancreatitis (CP) 208, 441, 45966 contrast-enhanced abdominal CT, and see also balloon dilation; large balloon
biliary strictures and 327, 330 pancreatic uid collection 477 dilation after minimal biliary
in children 2289 in diagnosis of pancreatic cystic lesions 470, sphincterotomy
endotherapy in 459 471 dilation devices 38

496

diverticula
ampullary 79
duodenal 113
periampullary 79
periampullary duodenal 7980
documentation 910
of informed consent 910
dominant dorsal duct drainage 454
double-duct sign 290
in children 224
drainage
dominant dorsal duct drainage 454
drainage times, pancreatic/biliary 367
in hilar malignant biliary obstruction 3023
pancreatic duct see pancreatic duct
drainage
transmural 47881
transpapillary 4780
see also pancreatic uid collections
drainage devices 36
see also nasobiliary drainage catheters;
nasopancreatic drainage catheters;
stents
drugs, emergency, availability of 16
dual sphincterotomy 376
ductal disruption 475, 483, 488
epidemiology 419
ductal hypertension 435
duct of Wirsung see pancreatic duct
duct size, calculation 212
dunking anastomosis 349
duodenal bypass 245
duodenal diverticula 7980, 113
juxtapapillary 112
duodenal duplication cysts 441
see also choledochoceles
duodenal obstruction, SEMS placement 1723
duodenal papilla, dilatation 202
duodenal polyposis, staging system 279
duodenal switch 246
duodenojejunostomy 245
duodenoscopes 31, 39
pediatric 31
in self-expandable metallic stents (SEMS)
placement 16970
therapeutic 105
used in children 220
duty of care 4
breach of 4
Duval procedure 252
E success rates 97, 100
EASIE (Erlangen Active Training Simulator for
Interventional Endoscopy) 64
CompactEASIE simulator 64, 656
EBD see endoscopic balloon dilation
Echinococcus granulosus 3945
postoperative complications 395
electrocautery, in pregnancy 233
electrocoagulation 116
electrohydraulic lithotripsy (EHL) 215, 362,
399400, 402, 404
success rates 215
see also intraductal electrohydraulic
lithotripsy; lithotripsy
electrosurgical current
for EST 111, 115
minor papilla sphincterotomy 1434, 150
during pancreatic sphincterotomy 12930
pure-cut electrocautery vs blended current
111
electrosurgical generators 1415
ENDO CUT mode 111
employer liability 5
endoscope processor, in ERCP room 14
endoscopes 31
forward-viewing 31
oblique-viewing 31
side-viewing 31
storage 16
upper 31
see also pancreatoscopes; cholangioscopes
endoscopic balloon dilation (EBD) 113
indications for 98100
see also endoscopic papillary balloon dilation
endoscopic biliary sphincterotomy (EST)
97107, 10918
alternatives to 113
complications 1013, 11417
contraindications 114
cost 117
electrosurgical current for 111, 115
vs endoscopic papillary balloon dilation
(EPBD) 97, 98103
indications 97, 11314
instruments 10910
long-term consequences 11617
pancreatic stent placement 115
for pancreatitis due to microlithiasis 436
patients with difcult anatomy 11213
procedure 11012
success rates 97
technique 10913
endoscopic mechanical lithotripsy (EML) 105
endoscopic pancreatic ductal drainage
impact on endocrine and exocrine functions
464
pain relief after 463
in severe chronic pancreatitis 462
endoscopic pancreatic sphincterotomy (EPS)
45960
see also pancreatic sphincterotomy
endoscopic papillary balloon dilation (EPBD)
97107
complications 1013
vs endoscopic biliary sphincterotomy (EST)
97, 98103
indications for 98100
informed consent 97
limitations 1001
endoscopic resection (snare papillectomy) 440
endoscopic retrograde cholangiography,
indications 26
endoscopic sphincterotomy, indications 26
endoscopic transpancreatic papillary septotomy
79
endoscopic ultrasound (EUS)
in acute gallstone pancreatitis 411
ampullary neoplasm 2756
in assessment of ampullary tumors 18990
in choledocholithiasis diagnosis 3578
complications 268
in diagnosis of pancreas divisum 454
in diagnosis of pancreatic cystic lesions 470
dilated bile duct 2678
EUS-guided rendezvous techniques 80, 83
hilar cholangiocarcinoma 301
INDEX
in IAP and IARP 441, 4412
indeterminate biliary strictures 31617
limitations 268
for minor papilla cannulation 83
and pancreatic uid collection 4778
pancreatic malignancies 289
endoscopists
communication of thoughts 26, 27
experience of, and complications of ERCP
51, 54, 58
levels of training 66
number of procedures to achieve/maintain
competence 7, 612
trainees 7
see also training
endotherapy
and pain, in chronic pancreatitis 464
palliation of malignancies 344
post-pancreatic surgery 351
EndoTrainer simulator 64, 65
enterococci, in cholangitis 359
Enterococcus spp. 304
enteroscopes 31
EPBD see endoscopic papillary balloon dilation
epinephrine 56, 88, 96, 116, 150, 195, 229
equipment for ERCP room
radiologic imaging 1516
see also accessories for the ERCP room; and
individual accessories and pieces of
equipment
ERCP, past, present and future ix-x
ERCP room 1318
conguration 1315
emergency drugs 16
location 13
size 13
see also accessories; equipment for ERCP
room; and individual accessories and
pieces of equipment
Erlangen Active Training Simulator for
Interventional Endoscopy (EASIE) 64
Escherichia coli 304
in cholangitis 359
esophageal intubation 73
esophageal resection 237
EST see endoscopic biliary sphincterotomy
ESWL see extracorporeal shock wave lithotripsy
EUS see endoscopic ultrasound
exchange assistance devices 35
experience see competence; endoscopists
expert testimony 4, 45
extracorporeal shock wave lithotripsy (ESWL) 119
common bile duct stones 362
in severe chronic pancreatitis 4601
clinical results 4634
stone extraction and dilation after 4612
technical results 4623
extraction balloons 11920, 123
cost 121
extrahepatic bile duct
size 269
stones, EST for 105
F
familial adenomatous polyposis (FAP) 189, 273,
274, 277, 279, 440
screening programs 279
see also ampullary adenomas; Gardners
syndrome
497
 INDEX

Fasciola 231 gastric banding 246, 247 surgery for localized cancer 301
Fasciola hepatica 3967 gastric bypass 2479 surgery for unresectable cancer 3012
fascioliasis 3967 gastric resection 23745 see also cholangiocarcinoma
acute/hepatic phase 396 Billroth I 237 hilar obstruction, SEMS placement 172
chronic/biliary phase 396 Billroth II 23742, 2567 hilar strictures 300, 3028
endotherapy 3967 Roux-en-Y gastrectomy 2423 hilar tumors 23
massive forms 397 total gastrectomy 2435 HIV-associated cholangiopathy, in children 226
stages of 396 gastroenteroanastomosis 348 hospital liability 5
tests 396 gastroenterology, medicolegal issues 3 hydatid cysts 394
fatty meal ultrasound (FMS), in SOD 36970 gastrointestinal endoscopy, medicolegal issues cysto-biliary communication 394, 395
felodipine 373 3, 9 intrabiliary rupture 3945
fentanyl 305 gastrointestinal stromal tumors (GIST) 278 and obstructive jaundice 394, 395
brin glue 116 gastrojejunostomy 245 postoperative complications 395
lling defects, liver transplantation 343 gastroplasty, vertical banded 245, 2467 hydatid disease 3945
stulas, associated with liver transplantation 342 Geenen-Hogan (G-H) criteria, sphincter of endotherapy 394
stulotomy Oddi dysfunction (SOD) 3678, 369, hyoscyamine 40
precut 77 376 hyperamylasemia, post-EPBD 1023
suprapapillary 89 gemcitabine 295 hypoventilation, alveolar 46
xed-diameter plastic stents (FDPS) 165 gene analysis, pancreatitis in children 228
cost 174 generator currents 32
I
patency 1656 idiopathic acute pancreatitis (IAP)
generators 1415
vs self-expandable metallic stents (SEMS) causes 435
genetic mutations/abnormalities 441, 455
1656, 174 controversy about SOM 438
gentamicin 129
stent occlusion 165, 166 diagnosis 4356
glucagon 40, 143, 192, 220, 372
ow cytometry, indeterminate biliary strictures diagnostic yield of ERCP 442
glue injection, into the pancreatic duct 429
321 due to SOD 4378
guidelines, professional 4
ower baskets 122 endoscopic approach 435
guidewires 335
oxuridine 379 ERCP timing 435
coated (sheathed) 34
uconazole 485 genetic testing 441
coiled 34
umazenil 16 investigations 4412
hydrophilic 345
5-Fluoracil (5-FU) 295 recurrent see idiopathic acute recurrent
monolament 34
uorescent in-situ hybridization (FISH), pancreatitis
types 345
indeterminate biliary strictures 321 idiopathic acute recurrent pancreatitis (IARP)
wire safety 35
uoroquinolone 220 435, 455
uoroscopy H choledochoceles causing 439
during cholangioscopy 213 Helicobacter pylori 278 diagnostic yield of ERCP 442
for pediatric ERCP 21920 hemorrhage (bleeding) investigations 4412
views of images 20 as complication of ERCP 52, 53 outcomes of endoscopy therapy 4423
uoroscopy rooms 19 as complication of EST/EPBD 1012 pancreas divisum and 4389
follicular lymphomas 278 as complication of large balloon dilation after SOD as a cause of 4378
forceps, biliary biopsy forceps 38 minimal biliary sphincterotomy 105 stenting 438
forceps biopsies as complication of sphincterotomy 956 idiopathic chronic pancreatitis 459
ampullary neoplasm 275 EST-related 11516 idiopathic pancreatitis 419
biliary 38 prevention 56 IDL see intraductal lithotripsy
indeterminate biliary strictures 3201 risk factors for 8 IDUS see intraductal ultrasound
formalin, in hydatid cyst surgery 395 after sphincterotomy 56 image intensier 15
Freys procedure 252 during sphincterotomy 56 imaging, digital vs conventional 15
Fusion system 36, 462 treatment 56 imatinib mesylate 278
heparin 56 imipenem 485
G hepatectomy 412 immunoglobulin G, in indeterminate biliary
gabexate 55 hepatic ducts 447 strictures 313
gabexate mesylate 103, 115, 209 hepaticocutaneous jejunostomy 255 see also autoimmune pancreatitis
gallbladder, cholecystitis after sphincterotomy 58 hepaticojejunostomy identication 3501 indemnity insurance 10
gallbladder cancer, anomalous pancreaticobiliary hepatobiliary disease, parasitic infestation 393 indemnity payments 3
junction (ABPJ) and 392, 447 hepatobiliary scintigraphy (HBS), for SOD indeterminate biliary strictures 31325
gallbladder carcinoma, prognosis and risk 36970 ancillary techniques 3214
factors 287 hepatolithiasis 412 brush cytology 320
gallstones 357 hilar cholangiocarcinoma characterization 313, 317
vs air bubbles 26 adjuvant therapy 302 denition 313
causing acute pancreatitis 435 Bismuth classication 299 ERCP in 314, 317, 319
nonradiopaque 20 criteria for unresectability 300 historical features 313
occult gallstone disease 435, 436 imaging 3001 imaging
origin 364 orthotopic liver transplantation (OLT) for invasive 31619
see also acute gallstone pancreatitis 301 non-invasive 31316
Gardners syndrome 279, see also familial palliation 3012 principles 31719
adenomatous polyposis percutaneous approach 302 inadequate treatment 313
gastrectomy preoperative histological conrmation 301 intraductal forceps biopsies 3201
Roux-en-Y see Roux-en-Y gastrectomy radiological studies 3001 intraductal transmucosal ne needle
total 2435 risk of acute cholangitis 309 aspiration 320

498

laboratory features 313
malignant 317
pathologic investigations 31921
stent placement 323
stricture dilation 31819
surgical exploration 324
tissue acquisition 31921
indications for ERCP 26
marginal 9, 10
in pancreaticoduodenectomy 353
in recurrent pyogenic cholangitis (RPC) 406
indigocarmine staining 191
infection, post-EST/EPBD 101, 103
inammatory bowel disease, in children 2234
informed consent 9
disclosure of physicians experience 6, 9
documentation of 910
elements of risk in disclosure 6
endoscopic pancreatic sphincterotomy 129
endoscopic papillary balloon dilation (EPBD)
97
for ERCP for suspected sphincter of Oddi
dysfunction (SOD) 370
exceptions 6
failure to obtain, legal consequences 6
informed refusal 6
legal principles 56
material risks 6
patients competence and 6, 9
risk management and 3, 6
theory of 56
informed refusal 6
insurance, indemnity insurance 10
insurance companies
information for 10
information from 3
interleukin 10 55
intraductal electrohydraulic lithotripsy 1278
complications 1278
contraindications 127
indications 127
key points 127
technique 1278
intraductal lithotripsy (IDL)
for pancreatic stone fragmentation 461
see also intraductal electrohydraulic
lithotripsy; laser lithotripsy
intraductal papillary mucinous neoplasia (IPMN)
46774
diagnosis 470, 471
incidence 467
pancreaticoduodenectomy for 347
pathogenesis 467
pathology 4689
prognosis 472
intraductal papillary mucinous tumors (IPMTs)
199, 440, 4401
endoscopic therapy 441
as indication for pancreatoscopy 2067
intraductal transmucosal ne needle aspiration,
indeterminate biliary strictures 320
intraductal ultrasound (IDUS) 440
ampullary neoplasm 276
biliary cysts (choledochal) 449
in choledocholithiasis diagnosis 357
hilar cholangiocarcinoma 301
indeterminate biliary strictures 3212
intrahepatic stones, treatment in RCP,
techniques 399406
IPMN see intraductal papillary mucinous
neoplasia
IPMTs see intraductal papillary mucinous
tumors
isosorbide dinitrate 103, 202, 373
J (PSC) 379
jaundice
in biliary malignancies 287, 288
obstructive see obstructive jaundice
post pancreatic surgery
Brigham and Womens Hospital
experience 353
choledochojejunostomy and 348
as symptom of ampullary tumors 273
as symptom of neuroendocrine tumors 277
juxtapapillary duodenal diverticula 112
K for malignancies, controversy 344
Kasai procedure (portoenterostomy) 222
Kehrs tube, associated biliary complications
340
Klatskin tumors 299, 305, 306, 308
Klebsiella, in cholangitis 359
Klebsiella spp. 304
L follicular 278
laparoscopic cholecystectomy (LC) 97
bile duct stones found during 3634
bile duct stones retained after 33940
biliary leaks after 335, 337
classication of biliary injuries during 336
complications after 336
contraindications for endoscopic treatment
following 338
for pancreatitis due to microlithiasis 436
laparoscopic exploration of the common bile
duct (LECBD) 3624
laparotomy 116
large balloon dilation after minimal biliary
sphincterotomy 1035
complications 105
laser lithotripsy 127, 399400, 404
lasso technique, stent removal 185
lateral pancreaticojejunostomy (Puestow
procedure) 347
lawsuits
avoidance strategies 910
if led 1011
limiting costs of 1011
reasons for 89
risky clinical situations 10
lead aprons 16, 19
legal principles
in medical practice 4
see also medicolegal issues
liability
employers 5
hospitals 5
vicarious 5
lithotripsy
basket mechanical lithotripsy (BML) 360
electrohydraulic (EHL) 215, 362, 399400,
402, 404
endoscopic mechanical lithotripsy (EML) 105
intraductal see intraductal lithotripsy
laser lithotripsy 127, 399400, 404
see also extracorporeal shock wave lithotripsy;
intraductal electrohydraulic lithotripsy;
mechanical lithotripsy
INDEX
lithotripsy baskets 39, 1237, 127
lithotripters 1245
electrohydraulic 127
mechanical 39
through-the-scope (TTS) 125
liver biopsy, in primary sclerosing cholangitis
liver enzymes, abnormal 379
liver function tests 265
liver transplantation 2545
bile leaks associated with 342
biliary cast syndrome 343
biliary complications 3424
complications 3424
lling defects 343
stulas associated with 342
for hilar cholangiocarcinoma 301
primary sclerosing cholangitis (PSC) and 384
recurrent biliary disease after 343
stent removal 185
strictures associated with 3423
lymphomas
biliary 278
diffuse large B-cell lymphomas 278
MALT (mucosa-associated lymphoid tissue)
lymphomas 278
T-cell lymphoma 278
M
macrocystic serous cystadenomas 468, 470
magnetic resonance cholangiopancreatography
(MRCP)
in acute gallstone pancreatitis 411
advantages and limitations 267
benets in post-pancreatic surgery 353
in biliary obstruction 267
in choledochal cyst diagnosis 389
in choledocholithiasis diagnosis 357
in diagnosis of bile/pancreatic duct leaks 347
dilated bile duct 2667
disadvantages 31516
hilar cholangiocarcinoma 3001
in IAP and IARP 442
indeterminate biliary strictures 314, 31415,
315
in primary sclerosing cholangitis (PSC) 379
secretin-MRCP (S-MRCP) 421, 424, 437, 442
magnetic resonance imaging (MRI)
ampullary neoplasm 275
in chronic pancreatitis (CP) 459
dilated bile duct 2667
hilar cholangiocarcinoma 3001
in IAP and IARP 442
indeterminate biliary strictures 31416
pancreas 289
and pancreatic uid collection 478
periampullary tumors 275
magnication, calculation of 21
main pancreatic duct (MPD), cannulation 45960
major papilla
cannulation of 7381
ectopic 447
endoscope positioning 745
evaluation 745
malignant tumors 440
malabsorptive-jejunoileal bypass 246
malabsorptive operations 246
499
 INDEX
malignant biliary obstruction, distal 28798
adjuvant therapy 2956
ampullary carcinoma 288
biopsies 289
chemotherapy 2956
cholangiocarcinoma 290
clinical features 2878
differential diagnosis 28890
endoscopic stenting 2914
epidemiology 287
imaging techniques 28890
metastatic disease 290
natural history 287
palliation 2915
pancreatic malignancies 287, 28890
patient management 2906
percutaneous stenting 2945
presenting symptoms 2878
stent choices 2934
stenting 2915
surgery, curative 291
surgery, palliative 295
tumors causing 287
malignant biliary obstruction, hilar 299311
classication 299
costs 303, 307, 30910
diagnosis 301
endoscopic drainage 3023
epidemiology 299
management strategies 299302
patient preparation 3045
percutaneous stenting 303
preoperative histological conrmation 301
preoperative stenting 303
radiological studies 3001
risk of acute cholangitis 309
stent implantation technique 3036
stent insertion complications 3089
summary 30910
surgical bypass 3012, 303
tissue sampling 301
see also hilar cholangiocarcinoma
malpractice 4
insurance 10
new technology and 7
MALT (mucosa-associated lymphoid tissue)
lymphomas 278
manometry, pancreatobiliary 95; see also
sphincter of oddi
manometry catheters 92
interpretation of recordings 94
triple lumen 94
mask ventilation, difculties with 45
MCNs see mucinous cystic neoplasms
mebendazole 396
mechanical lithotripsy 1247, 128
basket (BML) 360
complications 124, 1267
contraindications 124, 126
endoscopic (EML) 105
indications 124, 126
key points 1234
technique 1246
through-the-scope (TTS) 125
medical practice, legal principles 4, 7
medicolegal environment 3
medicolegal issues
in gastroenterology 3
in gastrointestinal endoscopy 3, 9
500
in medical practice 4 nasobiliary drain (NBD)
summary 11 in acute cholangitis 3601
meperidine 91, 372 temporary 360
meropenem 485 nasobiliary drainage catheters 36, 38, 260
metastatic disease 290 nasobiliary tubes, in treatment of biliary leaks 335
methylene blue 143, 191, 193, 351 nasopancreatic drainage catheters 36, 38
microcystic serous cystadenomas 468, 470 N-butyl-2-cyanoacrylate 178
microlithiasis 436 needle-knife access sphincterotomy, precautions
biliary 415 against complications and lawsuits 9
endoscopic ultrasound and 267, 268 needle-knife papillotomy 89
midazolam 44, 91, 305 needle-knife precut accessotomy 8890
in pregnancy 233 needle-knife sphincterotomy 779
minor papilla complications 779
cannulation 814 minor papillary 143, 145
advanced techniques 834 pancreatic 134, 136
indications for 81 needle tip catheters 31, 82
standard techniques 813 negligence 4, 5, 8
locating 812 neonatal cholestasis 222
pancreatitis and 455 neonates, duodenoscopes for examination in 31
sphincterotomy see minor papilla neo-papilla 65
sphincterotomy neuroendocrine carcinomas 2778
minor papilla sphincterotomy 14351, 460 neuroendocrine tumors (NET) 2778
accessories 143 neurobromatosis (von Recklinghausen
cautery unit 1434 disease) and 277
complications 14950 neurobromatosis, neuroendocrine tumors and
contraindications 143 277
current type 1434 new onset steatorrhea 464
guidewires 143 nifedipine 373
indications 143 nitrates 373
needle-knife sphincterotomy 143, 145 nitric oxide (NO) 373
over pancreatic duct stent 147 non-steroidal anti-inammatory drugs (NSAIDs)
outcomes 14950 55, 56, 129
precut sphincterotomy technique
14750 O
orice seen 1478 obstructive cholestasis, untreated 313
no orice seen 148 obstructive jaundice, hydatid cysts and 394, 395
re-do cases 149 occult gallstone disease 435, 436
with Santorinicele 1489 octreotide 55, 96
pull-type sphincterotomy 143, 145 operator experience, and complications of ERCP
technique 1457 54, 58
sedation 143 Opisthorchis felineus 395
sphincterotomes, choice of 145 Opisthorchis viverrini 395
stents 1445, 150 organ failure 412
summary 151 organized (walled-off ) pancreatic necrosis 483
supplemental drugs 143 antibiotics 485
Mirizzi syndrome 357 endoscopic debridement 485
monitors, placement in ERCP room 1314 endoscopic drainage 4856
MRCP see magnetic resonance endoscopic therapy results 488
cholangiopancreatography orthotopic liver transplantation (OLT) see liver
MRI see magnetic resonance imaging transplantation
mucinous cystic neoplasms (MCNs) oxygen administration 46
clinical epidemiology 467 P
diagnosis 470 PACS (picture archiving and computerized
k-ras mutation 468 storage) system 16, 19
ovarian stroma 468 pain
pathogenesis 467, 468 in chronic pancreatitis (CP) 459, 464
pathology 4689 diagnoses other than SOD 367, 368, 369
types 468 Geenen-Hogan (G-H) criteria 3678, 369
mucinous ductal ectasia 140 pancreatic 370
see also IPMN, IPMT pancreaticobiliary 367, 369
mucosa-associated lymphoid tissue (MALT) in sphincter of Oddi dysfunction (SOD)
lymphomas 278 3678, 370
multiple organ dysfunction syndrome with gallbladder in situ 370
(MODS) score 412 pain evaluation, post pancreatic surgery, Brigham
Murphys sign 411 and Womens Hospital experience 353
pancreas
N
annular 4567
nafamostat mesilate 209
imaging techniques 199
naloxone 16
 INDEX

pancreas divisum 27, 4389, 442, 44956 operator experience 4889 indications for 13540
association with pancreatitis 4556 results 4868 needle-knife sphincterotomy 134, 136
complete 454 liqueed 4756 precut pancreatic sphincterotomy 134
diagnosis 82, 438, 4545 contraindications to drainage 477 preparation 129
embryology 44954 drainage techniques 478 as primary therapy 13640
endoscopic therapy 4389, 4556 indications for drainage 4767 pull-type sphincterotomy 1314
idiopathic acute recurrent pancreatitis (IARP) pre-drainage evaluation 4778 as secondary therapy 140
and 4389 stent placement 4802 in sphincter of Oddi dysfunction (SOD)
incomplete 82, 454, 456 symptoms 477 1368
minor papilla therapy 143 masqueraders of 477 pancreatic stenosis, differentiation 2078
minor papilla ductography and 84 outcome differences following endoscopic pancreatic stents 37
minor papilla therapy 143 drainage 488 complications 162
pseudodivisum 455 types 475 for EST patients 115
stenting 4389 pancreatic malignancies 287, 28890 plastic 153, 155, 1589
terminology 454 presenting symptoms 2878 indications 1601, 177
pancreatectomy, distal 347 risk factors 287 precautions against complications and
pancreatic abscesses 476 see also pancreatic cancer lawsuits 9
endoscopy therapy results 488 pancreatic necrosis 4756, 4825 reducing risk of post-ERCP pancreatitis 545
pancreatic anomalies 44957 organized see organized (walled-off ) in treatment of post-ERCP pancreatitis 56
pancreatic ascites 422 pancreatic necrosis pancreatic surgery, complications 34755
pancreatic cancer walled-off (WON) see organized (walled-off ) endotherapy 351
differentiation 2078 pancreatic necrosis ERCP for 34755
early detection 208 pancreaticobiliary dilation 38 accessing and traversing the afferent limb
see also pancreatic malignancies pancreaticobiliary manometry 95 350
pancreatic cystic lesions catheters 31 Brigham and Womens Hospital
clinical presentation 469 see also sphincter of Oddi manometry experience 3524
diagnosis, methods 4701 pancreaticoduodenal stula, iatrogenic 464 endoscopes and accessories 350
differential diagnosis 469 pancreaticoduodenectomy 347 endotherapy 351
loss of heterozygosity (LOH) and 467, 468 for ampullary adenomas 279 hepaticojejunostomy identication 3501
pathogenesis 467 for ampullary carcinomas 280 negotiating the anatomy 3501
prevalence 467 Brigham and Womens Hospital experience pancreaticojejunostomy identication 351
prognosis 472 353 long-term 348
treatment 472 classic 3489 prevention of 347
types 467 ERCP approaches after 3512 short-term 3478
pancreatic duct (PD) 74 in familial adenomatous polyposis (FAP) 279 pancreatitis
cannulation of 801 indications for ERCP 353 acute see acute pancreatitis
drainage see pancreatic duct drainage postoperative complications 290 association of pancreas divisum with 4556
extravasation from 26 pylorus-preserving 349 biliary see acute gallstone pancreatitis (AGP)
leaks see pancreatic stulas types 3489 in children 2279
obstruction 1601, 435 see also Whipple procedure acute 2278
opacication 246 pancreaticoenteric stulas 4224 as complication of ERCP 229
perforation 116 pancreaticogastrostomy 250 gene analysis 228
stents see pancreatic duct stents pancreaticojejunostomy persistent/recurrent/chronic 2289
pancreatic duct drainage 2502, 463 complications 348 chronic see chronic pancreatitis
Duval procedure 252 identication of 351 as complication of ERCP 51, 526
Freys procedure 252 lateral (Puestow procedure) 347 in children 229
Puestow procedure 250 pancreatic pseudocysts 422, 467, 469 as complication of EST/EPBD 101, 1023
pancreatic duct stents acute 4756, 483 as complication of pancreatic sphincterotomy
to facilitate biliary cannulation 767 chronic 476 141
insertion 1589 diagnosis 470 as complication of pancreatobiliary
in sphincter of Oddi manometry 94, 95 endoscopic therapy results 4868 manometry 95
pancreatic stulas 4201 infected 476 ERCP-related 115
classication 420 pancreatic abscesses 476, 488 EST-related 114, 115
complications 42931 pancreatic resection 24950 idiopathic 419
chronic 431 mid-pancreatic 250 lawsuits involving 8
immediate 430 pancreaticogastrostomy 250 pharmacologic prophylaxis 115
subacute 4301 tail of the pancreas 250 postampullectomy 1934, 195
contraindications for endoscopic approach 422 Whipple procedure, conventional 24950 risk factors, EST-related 11516
diagnosis 421 Whipple procedure, pylorus-preserving 250 pancreatitis, post-ERCP (PEP)
external 420, 4249 see also pancreaticoduodenectomy patient-related risk factors 534
indications for endoscopic approach 422 pancreatic sphincter hypertension 376, 438 pharmacological agents 556
internal 420 pancreatic sphincterotomy 12942, 45960 prevention of 56, 3745
management 4219 antibiotics and anticoagulants 129 risk factors 8
pancreatic uid collections (PFCs) 475 biliary sphincterotomy immediately before 134 technique-related 54
acute uid collections 475 complications 1401 risk reduction techniques 545
drainage 159 costs 141 treatment 56
endoscopic therapy current type 12930 pancreatobiliary malunion see anomalous
complications 48990 endoscopic technique 1301 pancreaticobiliary junction; anomalous
controversies 489 equipment 12930 pancreatobiliary union

501
 INDEX

pancreatoscopes 199 percutaneous transhepatic cholangiography indications 87

baby scopes 199200, 204 (PTC) 268, 269, 319 needle-knife 89
cost 209 percutaneous transhepatic cholangioscopy technique 878
beroptic 199, 2002, 204 (PTCS) 4002, 4012 precut biliary sphincterotomy 779
image converters 2002 advantages 402 precut stulotomy 77
insertion procedure 2023 complications 411 precut pancreatic sphincterotomy 134
mother scopes 199200 cost 409 precut papillotomy 87
peroral electronic (PEPS) 199, 204, 207; see drawback 402 precut (access) sphincterotomes 323
also pancreatoscopy examination 4012 precut sphincterotomy technique 14750
ultrathin 199, 2023, 203, 204, 209 percutaneous transhepatic biliary drainage pregnancy 19, 41, 2315
video 199, 200 (PTBD) 4001 ascariasis in 231, 234, 393
pancreatoscopy 199210, 440 tract dilation 401 avoidance of uoroscopic images 2323
biopsy 2034 percutaneous transhepatic pneumatic dilation biliary disorders 234
complications 209 328 choledocholithiasis 231, 234
cytology sampling 2034 percutaneous transhepatic therapy 328 cholelithiasis 231
endoscopic procedure 2025 perforation complications, ERCP-related 231
equipment 199202 as complication of ERCP 52, 568 uoroscopy time 2323
ndings in pancreatic diseases 205 ERCP-related rates 116 indication for ERCP 231
indications 2069 EST-related 116 outcomes 234
procedure timing 202 post-EST/EPBD 101, 103 physiologic changes of pregnancy 233
scope insertion 2023 risk factors during ERCP 8, 57 positioning 233
success rates 2045 periampullary diverticula 79 radiation protection 2313
technique 199202 periampullary duodenal diverticula sedation 2334
visualization 204 cannulation in patients with 7980 primary sclerosing cholangitis (PSC) 287, 37986
papillae terminology 79 background 379
balloon dilation 97107 periampullary tumors, magnetic resonance biliary strictures 332
see also major papilla; minor papilla imaging (MRI) 275 cholangiocarcinoma in see
papilla of Vater, double 447 peroral electronic pancreatoscope (PEPS) 199, cholangiocarcinoma (CA)
papillary roof incision 33 204, 207; see also pancreatoscopy cost 3801
papillary spasm see sphincter of Oddi pethidine 305 diagnosis 379
dysfunction photodynamic therapy (PDT) 384 complications of ERCP 37980
papillary stenosis 96, 140 Photofrin 384 contraindications 379
re-stenosis 150 Physician Insurers Association of America indications 379
see also sphincter of Oddi dysfunction (PIAA) 3 technique 379
papillectomy 189 physicianpatient relationship 4, 8, 9 endoscopic treatment 3813
papillitis see sphincter of Oddi dysfunction physicianphysician interaction 19 complications 382
papillomatosis, biliary 214, 216 picture archiving and computerized storage contraindications 382
papillotomes 54 (PACS) system 16, 19 cost 3823
papillotomy piperacillin 129, 305, 485 indications 382
needle-knife 89 plastic stents 367 technique 3812
precut 87 biliary 1538 natural history 379
parallel cannulation 80 congurations/sizes 153 symptoms 379
parasites 231 cost 162, 2912, 293 privileging 8
parasitic disease 3938 in distal malignant biliary obstruction 2912 proctors 5
patients xed-diameter see xed-diameter plastic propofol 91
monitoring 467 stents (FDPS) case studies 44
physicianpatient relationship 4, 8, 9 in hilar malignant biliary obstruction 3058 guidelines for administration of 478
positioning during ERCP 19, 23, 25 indications for 1601 in pregnancy 233, 2334
pregnant 19, 41 pancreatic 153, 1589 in sphincter of Oddi manometry (SOM) 372
radiation exposure 19, 401 small-caliber 1589 Proteus spp., in cholangitis 359
patient table 13 stent patency 37, 153, 1656 pseudocysts see pancreatic pseudocysts
pediatric duodenoscopes 31 stent removal 177, 1789, 183, 185 pseudodivisum 455
pediatric ERCP 21930 platelet-activating factor inhibitors 56 Pseudomonas, in cholangitis 359
complications 229 Plavix 56 Puestow procedure (lateral
contraindications 2229 pleural effusions 422 pancreaticojejunostomy) 250, 347
cost 229 polyposis syndromes pull-type sphincterotomes 32
endoscopists 219 hereditary 189 pull-type sphincterotomy
equipment 220 see also familial adenomatous polyposis; minor papillary 143, 145, 1457
uoroscopy 21920 Gardners syndrome pancreatic 1314
indications 2229 portoenterostomy (Kasai procedure) 222 pulse oximetry 46
procedure setting 219 positron emission tomography (PET), hilar
sedation 219 cholangiocarcinoma 301 R
supplemental medications 220 post-cholecystectomy syndrome (PCS) 341 radiation
technique 2201 see also sphincter of Oddi dysfunction ERCP method to avoid exposure 233
peer review, adverse events 8 practice parameters 4 exposure types 231
percutaneous endoscopic gastrostomy (PEG) 248 praziquantel 396 monitoring 19
percutaneous transhepatic biliary drainage preceptors 5 patient exposure to 19, 401
(PTBD) 4001 precut accessotomy 87 protection, in pregnancy 2313
equipment 400 complications 90 staff exposure to 15, 16, 19, 401, 231

502
 INDEX

radiation dosimeters 16 sources 82 setting sun 73

radiographic images, white-on-black vs black-on- ultrasound tests 456 severe acute pancreatitis (SAP)
white 20 secretin-MRCP (s-MRCP) 421, 424, 437, 442 organ failure 412
radiologic imaging equipment 1516 sedation recognition of 412
comparison of 15 minor papilla sphincterotomy 143 scoring systems 412
components 15 for pediatric ERCP 219 severe chronic pancreatitis
controls 1516 in pregnancy 2334 clinical results 4634
digital vs conventional 15 in sphincter of Oddi manometry (SOM) 372 stenting 462
xed vs portable 15 summary 48 technical results 4623
image capture and storage 16 sedation and analgesia 439 see also extracorporeal shock wave lithotripsy
imaging components 15 adverse effects 44 sheep liver uke see Fasciola hepatica
radiographic hardware 16 conscious sedation 43 shouldered strictures 23
radiologic issues, in ERCP 1930 endoscopists and 44 Simbionix GI-Mentor computer simulation
Ransons criteria, biliary pancreatitis 412, 416 levels of 43, 44 models 63, 68
Rapid Exchange (RX) Biliary System 35 patient evaluation guidelines 45 simulators 667
recurrent cholangitis, post pancreatic surgery patient monitoring 467 skills
348 physiology of 43 interpersonal 8
recurrent pyogenic cholangitis (RPC) 357, practice guidelines 445 maintaining 68
399410 pre-procedure evaluation 456 major vs minor 7
complications 4068 safety of 434 see also competence
contraindications for ERCP/PTCS 406 self-expandable metallic stents (SEMS) 378, sleeve catheters 95
cost 412 165, 177 small intestine, navigating through 256
indications for ERCP/PTCS 406 benign biliary strictures 330 s-MRCP (secretin-MRCP) 421, 424, 437, 442
initial management 399 complications 1734 snare papillectomy (endoscopic resection) 440,
key points 399 contraindications 166 see also ampullary adenomas;
long-term management 408 cost-effectiveness 174, 293, 294 ampullary tumors
surgery 4089 covered 166, 1689, 174, 177, 186 SO see sphincter of Oddi
see also intrahepatic stones Diamond Ultraex stent 1678 SOD see sphincter of Oddi dysfunction (SOD)
rendezvous procedure/technique 80, 110, 113, in distal malignant biliary obstruction 291, sodium nitroprusside (SNP) 373
3512 2924 solid pseudopapillary tumor 469
EUS-guided 80, 83 vs xed-diameter plastic stents (FDPS) SOM see sphincter of Oddi manometry (SOM)
internal rendezvous procedure 341 1656, 174 somatostatin 55, 115, 424
malignant biliary obstruction, hilar 305 Flexxus stent 168 sphincter ablation therapy 437, 438
SEMS placement 172 in hilar malignant biliary obstruction 3058 sphincter of Oddi (SO) 91
in surgically altered anatomy 249, 2556 complications 3089 function 91, 4367
repeat ERCP 80 indications 1656 sphincter of Oddi dysfunction (SOD) 91, 4368
reprivileging 8 patency 37, 1656, 177 biliary sphincterotomy in 114, 374, 3756
respondeat superior 5 placement techniques 16973 in children 228
Reynolds pentad 359 cholangiogram 170 clinical evaluation 36870
risk management strategies 68 deployment 1712 clinical syndromes associated with 367
avoiding complications and lawsuits 910 dilation 170 complication rate 51
informed consent and 3, 6 duodenal obstruction 1723 denitions 3678
summary 10 duodenoscope 16970 diagnosis 437, 438
technique-related 9 endoscopic guides 171 emotional burden on patient 369
Robotics Interactive Endoscopy Simulation uoroscopic guides 171 Geenen-Hogan (G-H) criteria 3678, 369, 376
(RIES) System 63 guidewire use 170 informed consent for ERCP 370
Rome criteria, sphincter of Oddi dysfunction hilar stricture 172 pain with gallbladder in situ 370
(SOD) 367 rendezvous technique 172 pain in 367, 369
roundworms see Ascaris lumbricoides SEMS positioning 1701 pancreatic 1368
Roux-en-Y anatomy, EST and 113 sphincterotomy 170 diagnostic criteria 137
Roux-en-Y choledochojejunostomy 254 stent selection 170 post-endoscopic recurrent pain 3756
Roux-en-Y gastrectomy 2423 polyurethane-covered 293 as risk factor for pancreatitis 53, 419
Roux-en-Y gastric bypass 245, 247 removal 177, 178, 1856 Rome criteria 367, 368
Roux-en-Y hepaticojejunostomy 2524, 255 stent occlusion 165, 174 symptoms 369
RPC see recurrent pyogenic cholangitis types 1679 tests for 36970
RX see Rapid Exchange uncovered 177, 186, 187 tests used for detection of 91
Viabil 169 sphincter of Oddi dysmotility, in children 2267
S Wallstent 1656, 167, 168, 169 sphincter of Oddi manometry (SOM) 32, 916,
Santoriniceles 1489, 455 Y stent 169 1367, 3702
scintigraphy Zilver stent 168 abnormalities in 94
biliary, dilated bile duct 269 Z stent 168 cannulation 3712
hepatobiliary (HBS), for SOD 36970 septicemia 359 catheters 956
sclerosing cholangitis 343 serous cystadenomas 467, 468, 470 complications 956
pediatric 223 macrocystic 468, 470 for diagnosing SOD 437, 438, 442
see also primary sclerosing cholangitis microcystic 468, 470 duodenal baseline pressure 372
secretin 55, 81, 82, 91, 193, 220, 349, 351 pathology 468 equipment 956, 3701
injection 40 solid 468 equipment preparation 915
in minor papilla sphincterotomy 143 serum amylase levels 411 indications 95
in pancreatoscopy 203 serum C-reactive protein level 412 interpretation of recordings 372

503
 INDEX

sphincter of Oddi manometry (SOM) (cont.) stent patency stones

kissing technique 371 plastic stents 37, 153, 1656 assessing size 119
pancreatic duct stent 94, 95 self-expandable metal stents (SEMS) 37, see also bile duct stones; gallstones; stone
patient monitoring 96 1656, 177 extraction
patient preparation 91 stent placement strictures
scientic basis 91 EST and 115 dilation devices 38
sedation 372 to facilitate biliary cannulation 767 shouldered 23
sphincterotomy 91, 94 image of 27 sulbactam 220
technique 915, 3712 indeterminate biliary strictures 323 sump syndrome 80, 252, 3401
treatment, endoscopic 3734 PFCs and 4802 supine hypotensive syndrome 233
treatment, medical 373 postampullectomy 1934 suprapapillary stulotomy 89
type of current 94 self-expandable metallic stents (SEMS) surgery
sphincterotomes techniques 16973 ampullary tumors 189
access (precut) 323 stent removal 37, 17787 ERCP and 237
biliary cannulation with 756 accessories used 1789 hilar cholangiocarcinoma 31314
in Billroth II gastrectomy 256 balloon traction 17980 malignant biliary obstruction, distal 291, 295
for minor papilla sphincterotomy 143, 145 complications 186 malignant biliary obstruction, hilar 3012,
needle-knife 32 contraindications 1778 303
for pancreatic sphincterotomy 130 controversies 186 see also bariatric surgery; biliary surgery;
pull-type 32 costs 187 biliary surgery complications;
reverse 40 direct grasping 179 pancreaticoduodenectomy; Whipple
rotatable 32 duct access and 177 procedure
single vs reusable 40 indications 1778 surgically altered anatomy 23758
types of 32, 10910 lasso technique 185 bariatric surgery 2459
sphincterotomy liver transplantation and 185 biliary surgery 2525
access techniques 77 plastic stents 177, 1789, 185 endoscopic techniques 2556
antegrade 113 self-expandable metallic stents (SEMS) 177, ERCP accessories 2567
biliary see biliary sphincterotomy; endoscopic 178, 1856 esophageal resection 237
biliary sphincterotomy stent cannulation 1803 gastric resection 23745
in children 222, 228, 229 techniques 177, 17885 pancreatic duct drainage 2502
complication rates 512 stent retrievers 37 pancreatic resection 24950
dual sphincterotomy 376 as dilators 38 upper GI bypass surgery without resection
indications 26 stents 245
minor papilla see minor papilla antireux 153, 185 swing-tip catheters 31
sphincterotomy associated with minor papilla sphincterotomy
and nasobiliary drain (NBD) in cholangitis 1445 T
360 biliary see biliary stents tachyoddia see sphincter of Oddi dysfunction
needle-knife 779 bioabsorbable self-expanding 335 tapered-tip catheters 31, 95
pancreatic see pancreatic sphincterotomy double pigtail 159 tazobactam 129, 305
postampullectomy 1934 endoscope requirements 153 T-cell lymphoma 278
precut, hemorrhage after 56 migrated 178 teaching skills, train-the-trainer sessions 678
preresection 1923 occluded 178 techniques, risk management strategies 9
prophylactic 193 pancreatic see pancreatic stents technology 78
as risk factor for pancreatitis 54 patency see stent patency and malpractice 7
sphincter of Oddi manometry 91, 94 pigtail 37, 157 and training 7
thermal injury after 55, 56 placement see stent placement and vicarious liability 78
trans-pancreatic precut 90 plastic see xed-diameter plastic stents thyroid collars, storage 16
see also endoscopic biliary sphincterotomy (FDPS); plastic stents ticlopidin 114
sphincterotomy perforation 57 removal see stent removal tissue sampling devices 38
spiral baskets 122 self-expandable metallic see self-expandable TNM staging system 2756
staff metallic stents (SEMS) Todani classication, choledochal (biliary) cysts
experience, and complications of ERCP 54, temporary 178 387, 389, 448
58 Viaduct 185 tort law 4
radiation exposure 15, 16, 19, 401, 231 steroids, for autoimmune pancreatitis 441 tort of negligence 5
standards of care 45 stomach, endoscope passage through 73 torus pylorus 64
majority vs minority standards 4 stone extraction 11928 traction papillotome techniques 90
steatorrhea, new onset 464 accessories 389 training 7, 6170
stenosis, papillary 96, 140 assessing stone size 119 articial tissue models 66
re-stenosis 150 balloon stone extraction 11921 ASGE guidelines 61, 62
stent cannulation, in stent removal 1803 basket stone extraction 1213, 128 clinical training 613
stenting and dilation, after extracorporeal shock wave guidelines 7
affecting bile leakage 335 lithotripsy (ESWL) 4612 hands-on workshops 678
benign biliary strictures 32832 image of 27 levels of 66
biliary, complication 3302 stone fragmentation with live animals 634
in cholangiocarcinoma (CA) 384 methods 119 models 63, 66
cholangitis 361 using electrohydraulic lithotripsy (EHL) new technology and 7
chronic pancreatitis (CP) 463 215 porcine tissue models 646
pancreatic 375 stone removal see cholangioscopic stone simulators 63, 667
severe chronic pancreatitis 462 removal survey of 68

504

transabdominal ultrasound (TUS)
advances in 288
ampullary carcinoma 288
ampullary neoplasm 275
in choledocholithiasis diagnosis 357
hilar cholangiocarcinoma 300
indeterminate biliary strictures 314
jaundiced patients 314
pancreatic malignancies 288
transmural drainage
contraindications 477, 478
entry devices 479
entry techniques 479
EUS-guided 47980
non-EUS-guided 4802
transpancreatic papillary septotomy 79
transpancreatic precut sphincterotomy 90
transpapillary drainage 4789, 485
triamcinolone 341
triclabendazole 396
T-tubes 340, 342, 362, 4023, 411
tumors
benign, as cause of pancreatitis 440
IAP and 4401
INDEX
mucinous 435 vital dye staining 191
see also mucinous cystic neoplasms Von HippelLindau (VHL) syndrome 467, 469,
Turcot syndrome 279 470
TUS see transabdominal ultrasound von Recklinghausen disease, neuroendocrine
tumors and 277
U V-system 36
ulcerative colitis, in children 2234
ulinastatin 209 W
Ulistatin 55 walled-off necrosis (WON) see organized (walled-
ultrasound (US) off ) pancreatic necrosis
dilated bile duct 2656 warfarin 129
endoscopic see endoscopic ultrasound (EUS) weight-reduction surgery see bariatric surgery
intraductal see intraductal ultrasound (IDUS) Whipple procedure
transabdominal see transabdominal conventional 24950
ultrasound (TUS) pylorus-preserving 250, 349
ultrasound probes 3940 wire baskets 1212
ursodeoxycholic acid (UDCA) 382, 391, 436 lithotripsy 39
US see ultrasound for stone extraction 39
V X
vancomycin 129 x-ray imaging units 15, 19
verapamil 373
vertical banded gastroplasty 245, 2467 Z
vicarious liability 5 zipper cutting 94, 110, 111, 116
new technology and 78
505