November 2015

A DECADE IN
MEDICINE

UROLOGY

ENDOCRINOLOGY

CLINICAL ONCOLOGY

NEPHROLOGY

CARDIOLOGY

NEUROLOGY
GASTROENTEROLOGY &
HE PATOLOGY

RHEUMATOLOGY

© 2015 Macmillan Publishers Limited. All rights reserved

Primers in your inbox—your link
to the most recent topics covered
in Nature Reviews Disease Primers.

From mechanisms to management

Nature Reviews Disease Primers aims to cover all

diseases across all medical specialties. Modular in nature,
Primer articles describe all aspects of a condition:
epidemiology; disease mechanisms; diagnosis, screening
and prevention; management; and quality of life.

With such a comprehensive article collection for you to

read, our alerts are the perfect solution to stay updated
on the latest Primers published in the journal.

Visit go.nature.com/m72Xm5 to register for the alert

www.nature.com/nrdp
 A Decade in Medicine
The articles included in A Decade in Medicine were originally
design: laura marshall
published online and in the November 2014 or November 2015
issues of the eight clinical Nature Reviews journals. To celebrate
the 10th anniversary of the launch of these journals, the editors
commissioned international experts to write short essays
highlighting the key papers that made the biggest contribution
to their field in the past decade. Between them, the clinical
Nature Reviews journals published 47 articles, which are collated
in this eBook; if you choose to cite an article, please use the
original journal citation rather than citing the eBook.
We hope you enjoy reading A Decade in Medicine. If you would like
to find out more about the Nature Reviews series, please visit:
http://www.nature.com/reviews/

Don’t miss out on next year’s Key Advances in Medicine

The world of medicine is fast-paced and ever changing; the Key Advances in Medicine eBook is
bursting with information for medical students, physicians and clinical researchers wishing to keep
up with the most significant medical breakthroughs of the past year. Sign up for free e-alerts and
be notified when next year’s Key Advances in Medicine eBook is published. You won’t miss out on
Nature Reviews collections or Focus issues either.

http://www.nature.com/register

Nature Reviews Cardiology

1 acute coronary syndromes | Successes and future objectives in acute
COPYRIGHT © 2015 Macmillan Publishers
Limited. All rights reserved. Printed in the United coronary syndrome
Kingdom. No part of this publication may be Frans Van de Werf
reproduced, stored in a retrieval system, or
transmitted in any form (electronic or otherwise) 3 arrhythmias | Cardiac fibrillation—challenges and evolving solutions
without prior permission from permissions@
nature.com. Stanley Nattel

DISCLAIMER: Although every effort is made 5 cardiomyopathies | Cardiomyopathy on the move

by the publishers to see that no inaccurate or Magdi H. Yacoub
misleading data, opinions or statements appear
in this collection, they wish to make it clear that 6 dyslipidaemia | Resurgence of targets and compounds to treat dyslipidaemia
the data and opinions appearing in articles and
advertisements herein are the responsibility John J. P. Kastelein
of the contributor or advertiser concerned. The
journal does include the personal opinions of 8 heart failure | 10 Years of progress in HF research—what have we learned?
the authors; therefore, it is not intended to be
Henry Krum
relied on solely as a guide to good practice or
safe treatment. Accordingly, the publishers,
employees, offices and agents accept no 10 hypertension | The past decade in hypertension—facts, hopes, and hypes
liability whatsoever for the consequences of Thomas Unger
any such inaccurate or misleading data, opinion
or statement. Although every effort is made to 12 peripheral vascular disease | 10 Years of breakthroughs in peripheral
ensure that drug doses and other quantities are
vascular disease
presented accurately, readers are advised that
the new methods and techniques involving drug Mark A. Creager
usage and described within this journal should
only be followed in conjunction with the drug 14 valvular disease | Current perspectives on treatment of valvular heart disease
manufacturer’s own published literature.
Friedrich W. Mohr

© 2015 Macmillan Publishers Limited. All rights reserved

Nature Reviews Clinical Oncology 53 polycystic kidney disease | Slowing progression of autosomal
dominant polycystic kidney disease
16 clinical trials | Shifting paradigms in cancer clinical trial design Robert W. Schrier
Daniel J. Sargent and Edward L. Korn
54 renal transplantation | A spectrum of advances in renal
18 targeted therapy | Successes, toxicities and challenges in solid transplantation
tumours Bruce Kaplan
Joel W. Neal and George W. Sledge

19 haematological cancer | Advances in biology and therapy

Nature Reviews Neurology
S. Vincent Rajkumar and Philippe Moreau 57 dementia | A decade of discovery and disappointment
in dementia research
21 cancer immunotherapy | Entering the mainstream of cancer
treatment John R. Hodges
Steven A. Rosenberg 58 multiple sclerosis | New drugs and personalized medicine
for multiple sclerosis
23 genomics | A decade of discovery in cancer genomics
Paul M. Matthews
Kenneth Offit
60 epilepsy | Edging toward breakthroughs in epilepsy diagnostics
25 funding in cancer research | National Cancer Institute awards
and care
—a work in progress
Daniel H. Lowenstein
Tito Fojo and Paraskevi Giannakakou
62 movement disorders | Tracking the pathogenesis of movement
Nature Reviews Endocrinology disorders
Oksana Suchowersky
28 bone | Great strides made but still further to go
Ian R. Reid 63 stroke | Progress in acute ischaemic stroke treatment
and prevention
29 thyroid disease | The endocrinology of thyroid disease from 2005
Jose G. Romano and Ralph L. Sacco
to 2015
P. Reed Larsen 65 migraine | Incredible progress for an era of better migraine care
Peter J. Goadsby
31 type 2 diabetes mellitus | At the centre of things
Guang Ning 67 cns infections | Major advances against a moving target
of CNS infections
33 paediatric endocrinology | New genes, new therapies
Lisa F. P. Ng and Tom Solomon
Mehul T. Dattani

35 reproductive endocrinology | Understanding reproductive Nature Reviews Rheumatology

endocrine disorders
Ursula B. Kaiser 69 translational rheumatology | Ten years after: rheumatology
research from bench to bedside
Nature Reviews Gastroenterology & Hepatology Nunzio Bottini and Gary S. Firestein

37 hcv | Hepatitis C therapy—a fast and competitive race 71 paediatric rheumatology | A field on the move
Stefan Zeuzem Seza Ozen

38 hepatocellular carcinoma | HCC—subtypes, stratification 72 technology | Technological advances transforming rheumatology

and sorafenib William H. Robinson and Rong Mao
Gregory J. Gores 74 clinical rheumatology | 10 years of therapeutic advances in the
40 gut microbiota | The gut microbiota era marches on rheumatic diseases
Francisco Guarner John D. Isaacs

42 fgids | ‘Functional’ gastrointestinal disorders—a paradigm shift Nature Reviews Urology

Nicholas J. Talley
77 bladder cancer | International progress: from cytology to genomics
43 pancreatic diseases | Advances in understanding and care James C. Costello and Dan Theodorescu
of pancreatic diseases
Randall E. Brand 79 imaging | A decade in image-guided prostate biopsy
Baris Turkbey and Peter L. Choyke
45 ibd | IBD—genes, bacteria and new therapeutic strategies
Jean-Frederic Colombel 81 urinary incontinence | Advances in female urology and voiding
dysfunction
Nature Reviews Nephrology Marisa M. Clifton and Howard B. Goldman

48 glomerular disease | The glomerulus reveals some secrets 82 kidney cancer | Discoveries, therapies and opportunities
Agnes B. Fogo W. Marston Linehan and Christopher J. Ricketts

50 genetics of kidney diseases | Genetic dissection of kidney 84 sexual dysfunction | Post-RP erectile dysfunction—therapies
disorders for the next decade
Friedhelm Hildebrandt Emmanuel Weyne and Maarten Albersen

51 acute kidney injury | Acute kidney injury—a decade of progress 86 prostate cancer | A decade of progress in detection and treatment
Rinaldo Bellomo Behfar Ehdaie and Peter T. Scardino

© 2015 Macmillan Publishers Limited. All rights reserved


DECADE IN REVIEW—ACUTE CORONARY SYNDROMES
Successes and future objectives in acute
coronary syndrome
Frans Van de Werf
The past decade has seen considerable advances in the treatment of acute coronary syndromes (ACS),
particularly in the search for improved antithrombotic therapies. Despite these successes, however, renewed
efforts are needed to improve long-term outcomes after ACS by reducing recurrent ischaemic events and
lowering the risk of bleeding complications.
Van de Werf, F. Nat. Rev. Cardiol. 11, 624–625 (2014); published online 2 September 2014; doi:10.1038/nrcardio.2014.129
By 10 years ago, aspirin and clopidogrel
therapy for 1 year was already the standard
treatment for all patients with acute coro-
nary syndrome (ACS), despite the benefit
of this dual antiplatelet therapy never
having been formally studied in patients
with ST-segment elevation myocardial
infarction (STEMI). Without a doubt, the
majority of clinical studies in ACS per-
formed in the past decade have focused on
attempts to find more efficacious and safer
antithrombotic treatments than aspirin
and clopidogrel dual antiplatelet therapy.
The most successful trial in this regard
was PLATO,1 which showed significantly
improved outcomes (including recurrent
myocardial infarction, stent thrombosis,
death from vascular causes, and even all-
cause mortality) with ticagrelor, a new
oral, reversible, direct-acting­P2Y purino-
receptor 12 (P2Y12) antagonist, when com-
pared with clopidogrel; the study included
patients with either STEMI or non-STEMI
(NSTEMI). Pleiotropic actions, such as
increased adenosine plasma concentrations,
were suggested to have contributed to the
beneficial effect of ticagrelor. The overall
excess spontaneous bleeding rate with this
new, more powerful agent was small, but
the trend towards an increased risk of intra­
cranial haemorrhage observed in this study
was a concern.
One of the possible reasons why newer
P2Y12 antagonists, such as cangrelor, pra­
sugrel, and ticagrelor, are superior to
clopi­d ogrel in large clinical trials is the
existence of genetic variants of hepatic
cytochrome P450 (CYP) enzymes respon-
sible for converting clopidogrel to its active
metabolite. Several studies have shown that
A DECADE IN MEDICINE NOVEMBER 2015  |  1
© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
carriers of an allele that resulted in reduced
CYP function had lower levels of the active
metabolite and higher rates of major cardio­
vascular events. As such, carriers of CYP
loss-of-function variants who are receiving
clopidogrel are likely to be especially at risk
of stent thrombosis. The concerns about
interindividual variability in the anti­platelet
effect of clopidogrel have led to a number
of studies aimed at monitoring platelet
function and adjusting treatment accord-
ingly. Surprisingly, in one of the largest
randomized studies, which included 2,440
patients scheduled for coronary stenting
(of whom 657 had NSTEMI), no improve-
ment in clinical outcomes at 1 year was
observed with platelet-function monitor-
ing and treatment adjustments as compared
with standard treatment without monitor-
ing.2 Genetic testing and platelet-function
monitoring have not become routine tests
in patients taking clopidogrel.
‘‘
…recurrent ischaemic
events and long-term mortality
remain high
’’
Bivalirudin has replaced heparin in many
hospitals, especially for patients undergo-
ing primary percutaneous coronary inter-
vention (PCI), owing mainly to the results
of the HORIZONS‑AMI study.3 This trial
showed significantly reduced bleeding rates
and decreased 30‑day mortality with biva-
lirudin alone when compared with heparin
plus a glycoprotein (GP) IIb/IIIa (integ-
rin αIIbβ3) antagonist.3 However, the benefit
of bivalirudin for primary PCI was chal-
lenged by the results of a single-centre
trial (HEAT‑PPCI4),in which GP IIb/IIIa
antagonists were used only for bailout
(bivalirudin versus heparin): fewer ischae-
mic events occurred with heparin than with
bivalirudin and, surprisingly (in contrast to
all previous trials), no reduction in bleeding
complications was found with bivalirudin.
Whether these data should change practice
and whether a new, large multicentre trial is
needed is currently a matter of hot debate.
Despite the progress made with anti­
thrombotic therapy and revascularization
in the early phase of an ACS, recurrent
ischaemic events and long-term mortal-
ity remain high. One explanation could be
that dual antiplatelet therapy consisting of
aspirin and clopidogrel does not provide
sufficient long-term protection against
recurrent thrombotic events. In two large
studies, 5,6 an additional antithrombotic
agent was tested in combination with
aspirin and ­c lopidogrel: vorapaxar, an
antagonist of the thrombin receptor PAR‑1
(protease-activate­d receptor 1), was used
in the TRA 2P trial, 5 and rivaroxaban, a
direct oral inhibitor of fac­tor Xa, was used
in ATLAS‑2.6 Although increased bleeding
rates were observed in both studies (and
the TRA 2P study even had to be stopped
prematurely in the subgroup of patients
with a history of stroke, owing to an excess
rate of intra­cranial haemorrhage), selected
populations did benefit from adding a
third antithrombotic drug. In the group
of >17,000 patients with a previous myo-
cardial infarction, addition of vorapaxar
reduced the risk of cardiovascular death,
myo­c ardial infarction, or stroke—at the
cost of an increased risk of moderate or
severe bleeding, but without a significant
 CARDIOLOGY
PLATO1 and subgroups
of TRA 2P5 or ATLAS-26 No option
Haemorrhagic risk
Standard
antithrombotic
treatment
Range of possible results
of a simplified antithrombotic regimen
Risk of thrombosis
Figure 1 | The balance of risks in treatment
of acute coronary syndrome. Patients with
acute coronary syndrome receive therapies
aimed at reducing the risk of thrombosis,
but antithrombotic treatments carry an
increased risk of bleeding complications.
New therapeutic approaches should minimize
the possibility of thrombotic events while
keeping haemorrhagic risk at acceptable
levels, or even reduce the risk when
compared with current treatment.
increase in intra­cranial haemorrhage.5 In
ATLAS‑2,6 low-dose rivaroxaban (2.5 mg
twice daily) also significantly reduced the
risk of the composite end point of cardio­
vascular death, myocardial infarction,
and stroke and, surprisingly, also reduced
rates of stent thrombosis and death from
any cause. Whether these agents are also
beneficial and safe when, for instance,
ticagrelor or pra­sugrel is used in combina-
tion with aspirin, is unknown. Similarly,
whether aspirin can be dropped if a new
P2Y12 antagonist is given in combination
with vorapaxar or low-dose rivaroxaban
remains to be determined. New studies of
simplified antithrombotic regimens that
are underway or in the planning phase
should take into account the possible
benefits and risks of new a­ntithrombotic
combinations (Figure 1).
‘‘
…use of increasingly powerful
anti­thrombotic agents has raised
’’
concern about bleeding
A number of trials to study the optimal
timing of angiography and revascularization
in patients with NSTEMI yielded discord-
ant results. In the TIMACS trial,7 >3,000
patients with ACS were randomly assigned
to either routine early intervention or
delayed intervention (either <24 h or >36 h
after treatment assignment, respectively).
In the total population, a nonsignificant
2  |  NOVEMBER 2015  www.nature.com/reviews
15% reduction in the composite end point
of death, myocardial infarction, or stroke
was observed when comparing early versus
delayed treatment. However, in line with
other studies, early intervention in high-
risk patients led to a significant 35% reduc-
tion in the risk of death, new myocardial
infarction, or stroke when compared with
delayed intervention.7
The use of increasingly powerful anti­
thrombotic agents has raised concern
about bleeding. Major bleeding complica-
tions are indeed common, and most occur
at the vascular access site. In the RIVAL
trial,8 >7,000 patients with ACS, includ-
ing almost 2,000 patients with STEMI,
were randomly allocated to either radial
or femoral access. Overall, a lower rate of
local vascular complications was observed
with the radial approach than with femoral
access. Remarkably, patients with STEMI
had better clinical outcomes (includ-
ing reduced mortality) when treated via
radial access, suggesting that this approach
might be particularly preferable in patients
with STEMI.8
Data from all registries show that, in
patients with STEMI, door-to-balloon
times have declined significantly over the
past 10 years. The implementation of emer-
gency medical systems based on a network
of hospitals with various levels of technol-
ogy connected by an efficient ambulance
system has reduced pre-PCI delays in many
places. 9 How­e ver, in a large US study of
almost 100,000 patients with STEMI, in-
hospital mortality remained unchanged
despite significant improvement in door-
to-balloon times, indicating that other
components of the total ischaemic time
need to be targeted.9 Administration of a
fibrinolytic agent to patients who cannot
undergo timely PCI is a strategy that might
be incorporated into prehospital care. In
the STREAM trial 10 of ~2,000 patients
presenting with early STEMI, pre­hospital
administration of tenecteplase (half the
normal dose in elderly individuals) to
patients who could not undergo PCI within
1 h resulted in rates of the composite end
point of death, shock, congestive heart
failure, or recurrent infarction at 30 days
that were similar to those in patients receiv-
ing standard primary PCI. Emergency
coronary angiography on arrival in the
PCI hospital could be avoided in almost
two-thirds of patients treated with tenecte­
plase.10 This strategy needs to be further
explored in patients with long transport
times, especially elderly patients.
© 2015 Macmillan Publishers Limited. All rights reserved
In summary, 30‑day mortality after
ACS has decreased remarkably in all age
cate­gories in the past decade, thanks to
improved antithrombotic treatment and
early revascularization. As a consequence,
profound changes in the epidemiology of
ACS have also taken place: more patients
with ACS survive the initial event and go
on to develop recurrent ischaemic events,
heart failure, atrial fibrillation, or non-
cardiac diseases (such as cancer), and are
dying at an older age. In the future, more
attention will have to be paid to the long-
term care of patients after ACS, of which
antithrombotic agents will remain an
important component.
Department of Cardiovascular Sciences,
University of Leuven, Herestraat 49,
B‑3000 Leuven, Belgium.
frans.vandewerf@med.kuleuven.be
Acknowledgements
I thank all my colleagues on both sides of the
Atlantic with whom I was able to collaborate in
performing important clinical trials that have
improved the treatment of patients with acute
coronary syndrome.
Competing interests
F.V.d.W. declares that he has received research
grants and fees for participating in advisory boards,
data and safety monitoring boards, and speaking
activities from AstraZeneca, Boehringer Ingelheim,
Merck, and The Medicines Company.
1. Wallentin, L. et al. Ticagrelor versus clopidogrel
in patients with acute coronary syndromes.
N. Engl. J. Med. 361, 1045–1057 (2009).
2. Collet, J. P. et al. Bedside monitoring to adjust
antiplatelet therapy for coronary stenting.
N. Engl. J. Med. 367, 2100–2109 (2012).
3. Stone, G. W. et al. Bivalirudin during primary PCI
in acute myocardial infarction. N. Engl. J. Med.
358, 2218–2230 (2008).
4. Shahzad, A. et al. Unfractionated heparin
versus bivalirudin in primary percutaneous
coronary intervention (HEAT-PPCI): an open-
label, single centre, randomised controlled trial.
Lancet http://dx.doi.org/10.1016/S0140-
6736(14)60924-7.
5. Morrow, D. A. et al. Vorapaxar in the secondary
prevention of atherothrombotic events. N. Engl.
J. Med. 366, 1404–1413 (2012).
6. Mega, J. L. et al. Rivaroxaban in patients with a
recent acute coronary syndrome. N. Engl. J. Med.
366, 9–19 (2012).
7. Mehta, S. R. et al. Early versus delayed invasive
intervention in acute coronary syndromes.
N. Engl. J. Med. 360, 2165–2175 (2009).
8. Jolly, S. S. et al. Radial versus femoral access
for coronary angiography and intervention in
patients with acute coronary syndromes (RIVAL):
a randomised, parallel group, multicentre trial.
Lancet 377, 1409–1420 (2011).
9. Menees, D. S. et al. Door‑to‑balloon time and
mortality among patients undergoing primary
PCI. N. Engl. J. Med. 369, 901–909 (2013).
10. Armstrong, P. W. et al. Fibrinolysis or primary
PCI in ST‑segment elevation myocardial
infarction. N. Engl. J. Med. 368, 1379–1387
(2013).
 CARDIOLOGY

DECADE IN REVIEW—ARRHYTHMIAS a speci­f ic electrocardiographic variant

Cardiac fibrillation—challenges
(known as early repolarization) is much
more common in individuals resuscitated

and evolving solutions after SCD than in a matched control popu-

Stanley Nattel
Cardiac rhythm disorders, or ‘arrhythmias’, are major sources of
morbidity and mortality, and have been challenging to treat because
classic pharmacological therapies are often ineffective and sometimes
dangerous. In the past decade, groundbreaking developments have
revolutionized the management of arrhythmias and prepared the
groundwork for new advances in the future.
Nattel, S. Nat. Rev. Cardiol. 11, 626–627 (2014); published online 9 September 2014;
doi:10.1038/nrcardio.2014.133
At the onset of the 2000s, effective and
definitive therapies were available for most
types of cardiac arrhythmia. However, two
common and important rhythm disorders
remained major therapeutic challenges
—sudden cardiac death (SCD) and atrial
fibrillation (AF). SCD typically occurs
owing to a chaotic, disorganized rhythm,
‘ventricular fibrillation’, which renders
effective heart pumping impossible and
causes death if not effectively managed
within minutes of onset (Figure 1). AF
resembles ventricular fibrillation in being a
totally disorganized rhythm, but AF affects
the atria, which are not essential for car­diac
pumping (Figure 1). AF causes problems
by firing the heart at an inappropriately
rapid rate, leading to a range of symptoms
and, in extreme cases, heart muscle weak-
ening that leads to heart failure. In addi-
tion, the ineffec­t ive pump­i ng action of
the atria allows blood clots to form in the
stagnant atrial blood pool. These clots can
propagate to the brain and cause strokes.
Consequently, AF is a major cause of
strokes, particularly in the elderly. Work
dur­ing the past decade has greatly improved
our ability to prevent and manage both SCD
and AF, and enabled further progress in
the future.
Towards the end of the 20th century, the
use of implantable devices to detect and
rapidly terminate ventricular tachyarrhyth-
mias that might otherwise lead to SCD had
become widespread. However, their effec-
tiveness in reducing mortality was unclear.
Despite preventing SCD, such devices might
not actually reduce mortality, particularly
if SCD constitutes a small portion of the
overall mortality, and if device-related
com­plications can themselves increase the
likeli­hood of death. An important milestone
was a study by Bardy and colleagues, who
lation.2 This study led to extensive clinical
and basic science investigations into the
underlying mechanisms and clini­cal impor-
tance of early repolarization. The early
repolarization pattern is a common electro-
cardiographic variant in otherwise normal
individuals, particularly young, physically-
active men, so distinguishing between the
common benign early-­repolarization pat­
tern and early repolarization associated
with increased SCD risk is an important
challenge. The combination of electro­
cardio­graphic ‘J‑waves’ (an elevated junc-
compared the effects of amiodarone (the tion between the end of the QRS and the
most effective antiarrhythmic drug) with onset of the ST-segment) and a horizontal
an implantable cardioverter–­defibrillator or descending ST-segment is associated
(ICD), or placebo in patients with sys- with increased risk of SCD. How­e ver, no
tolic heart failure, who are known to be early repolarization pattern alone (in the
at high risk of SCD.1 Amiodarone did not absence of malignant arrhythmias) is pres-
reduce all-cause mortality during a median ently sufficient to warrant prophylactic
45.5‑month follow-up period compared therapy (such as an ICD) to prevent SCD.
with placebo; however, ICD-therapy signifi- A further major advance in understand-
cantly reduced all-cause mortality (by 23%; ing the pathophysiology of SCD was the
P = 0.007).1 These findings established ICDs dem­onstration by Itzhaki and colleagues
as a potentially life-saving intervention in that skin fibroblasts from patients with
patients at increased risk of SCD. an inher­ited SCD syndrome can be trans-
Whereas some patients at increased risk formed into induced pluripotent stem cells,
of SCD are easily identified, the majority of and then differentiated into cardiomyo-
these arrhythmias occur in individuals cytes that reproduce the cardiac electrical
with­out known high-risk factors. More dysfunction under­lying SCD risk. 3 This
work is, therefore, needed to understand work establishes a novel technique to aid
the factors leading to SCD and thereby the understanding of SCD pathophysio­
effectively identify at-risk individuals. logy in individual patients and might ulti-
An important contribution was provided mately lead to patient-specific medical and
by Haïssaguerre et al. who showed that ­biological therapies.
Ventricular fibrillation Atrial fibrillation
Implanted Early Rate control as good as Lenient versus strict
defibrillator repolarization: rhythm control in heart failure4 rate control5
prevents death1 a novel cause2
Prevention Drug therapy
by risk-factor complications in
reduction10 high-risk patients6
LA
RA AVN
LV Novel
RV
anticoagulants Ablation as
initial therapy7
prevent stroke9
IPSCs for pathophysiology Mechanism-directed
and therapy3 ablation8
Figure 1 | Major advances in arrhythmia research during the past decade. The rate of impulse
transmission through the AVN determines the ventricular response rate during atrial fibrillation,
and is the target of rate-control therapy. Abbreviations: AVN, atrioventricular node; IPSC, induced
pluripotent stem cell; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

A DECADE IN MEDICINE NOVEMBER 2015  |  3

© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
Two potential approaches to manag-
ing AF exist: leaving the patient in AF, but
controlling the ventricular response rate
(known as rate control); and keeping the
patient out of AF, generally with the use of
antiarrhythmic drugs (known as rhythm
control). AF is a particularly important
risk factor for death and complications
in patients with heart failure, so rhythm
control would be expected to be of maximal
value in these individuals. This hypothesis
was tested in a prospective, randomized
trial by Roy et al. who found that an effec-
tive rhythm-control strategy had no bene­
ficial effects on cardiac function, survival,
or physical function in patients with heart
failure and AF.4
‘‘
In the past decade, the
management and understanding
of cardiac arrhythmias have been
’’
greatly advanced
Despite the value of rate control in con-
trolling adverse consequences of AF, the
parameters characterizing optimum rate
control are largely unknown. Van Gelder
and collea­g ues compared a ‘lenient’
rate-control strat­e gy (resting heart rate
<110 bpm) with a ‘strict’ strategy (resting
heart rate <80 bpm; exercise heart rate
<110 bpm). 5 The lenient strategy was as
effec­tive in preventing adverse effects and
easier to achieve than strict rate control.
A further assessment of pharmaco­logical
management of patients handled by rate
control was performed by Connolly and
colleagues.6 Based on previous work sug-
gesting that the antiarrhythmia drug
dronedarone reduces cardiac mortality
and stroke rate in patients with AF, and
that this effect might be independent of
maintaining sinus rhythm, these investiga-
tors administered dronedarone or placebo
to patients with permanent AF and markers
of increased risk. Instead of improving
outcomes, dronedarone increased rates
of heart failure, stroke, and cardiovas-
cular death. 6 This study reinforced con-
cerns about the risks of antiarrhythmic
agents in patients with AF at high risk of
cardiovascul­ar complications.
Given the concerns about antiarrhyth-
mic drugs and continual improvements
in technology, ablation procedures have
assumed an increasingly important role in
the treatment of AF over the past 10 years.
Although cardiac ablations have generally
4  |  NOVEMBER 2015  www.nature.com/reviews
been reserved for patients who did not
respond to at least one antiarrhythmic
drug, ablation as first-line therapy in AF
has been given increasing consideration.
Cosedis Nielsen et al. compared antiar-
rhythmic drugs with radiofrequency
ablation as an initial therapy for paroxys-
mal AF.7 Ablation reduced the recurrence
rate of symptomatic AF (from 16% to 7%),
but the primary end point (AF burden) was
not significantly affected.7 This result sug-
gests that antiarrhythmic drugs still have
a role in the management of AF. How­ever,
encouraging developments in abla­t ion
technology have been made. Many fail-
ures of ablation techniques are suspected
to be owing to procedures that do not
target patient-specific pathophysiology.
Narayan et al. have developed a method
using intra-atrial basket-catheter arrays
and complex mathematical algorithms to
identify and target patient-specific mecha-
nisms of AF. 8 Comparing their method
with conventional ablation approaches in
a multicentre, prospective, randomized
trial, the investigators noted substantial
improvements in AF prevention: patients
undergoing mechanism-directed ablation
showed much greater maintenance of sinus
rhythm compared with those undergoing
standard procedures (77.8% versus 38.5%;
P = 0.001).8
Stroke is the most important compli-
cation of AF. Effective orally adminis-
tered anticoagulation drugs can prevent
AF-related stroke, but until the approval
of dabigatran by the European Medicines
Agency in 2008, the only medications
avail­able for this indication were vitamin‑K
antagonists, which have a very narrow toxic-
to-therapeutic win­d ow. A number of
direct-acting oral anticoagulants, including
the thrombin-inhibitor dabigatran, and the
factor‑Xa antagonists apixaban, edoxaban,
and rivaroxaban, were developed in the past
decade. The first to be studied, dabigatran,
was demonstrated in a large, multicentre,
placebo-controlled trial to be at least as
effective and safer than warfarin for stroke
prevention in patients with AF. 9 Similar
findings were subsequently published for
the other direct-acting oral anticoagulants,
which have had a major and growing effect
on AF management.
Ultimately, the best approach to AF man-
agement would be to prevent the arrhyth-
mia from occurring in the first place. An
exciting proof-of-principle was provided
by Abed et al. who compared aggressive
weight-­m anagement intervention with
© 2015 Macmillan Publishers Limited. All rights reserved
lifestyle advice (control group) in a rand-
omized study of patients with AF who were
overweight.10 In addition to a substantial
(19%) reduction in body mass, the interven-
tion group showed significant reductions
compared with the control group in scores
for AF frequency (3.4 versus 0.7; P <0.001),
duration (5.0 versus 0.8; P <0.001), and
symptoms (8.4 versus 1.7; P <0.001).10 This
study will motivate further investiga­tion
of factors and mechanisms that pro­mote
AF, as well as public-health measures for AF
prevention.
In the past decade, the management and
understanding of cardiac arrhythmias have
been greatly advanced. The groundwork
has now been laid for important further
de­velopments in the near future.
Department of Medicine and Research
Center, Montreal Heart Institute and
Université de Montréal, 5000 Belanger Street
East, Montreal, QC H1T 1C8, Canada.
stanley.nattel@icm-mhi.org
Acknowledgements
S.N. is funded by the Canadian Institutes of Health
Research (6957 and 44365) and the Heart and
Stroke Foundation of Canada.
Competing interests
S.N. declares no competing interests.
1. Bardy, G. H. et al. Amiodarone or an
implantable cardioverter-defibrillator for
congestive heart failure. N. Engl. J. Med. 352,
225–237 (2005).
2. Haïssaguerre, M. et al. Sudden cardiac arrest
associated with early repolarization. N. Engl. J.
Med. 358, 2016–2023 (2008).
3. Itzhaki, I. et al. Modelling the long QT syndrome
with induced pluripotent stem cells. Nature
471, 225–229 (2011).
4. Roy, D. et al. Rhythm control versus rate control
for atrial fibrillation and heart failure N. Engl. J.
Med. 358, 2667–2677 (2008).
5. Van Gelder, I. C. et al. Lenient versus strict rate
control in patients with atrial fibrillation. N. Engl.
J. Med. 362, 1363–1373 (2010).
6. Connolly, S. J. et al. Dronedarone in high-risk
permanent atrial fibrillation. N. Engl. J. Med.
365, 2268–2276 (2011).
7. Cosedis Nielsen, J. et al. Radiofrequency
ablation as initial therapy in paroxysmal atrial
fibrillation. N. Engl. J. Med. 367, 1587–1595
(2012).
8. Narayan, S. M. et al. Treatment of atrial
fibrillation by the ablation of localized sources:
CONFIRM (Conventional Ablation for Atrial
Fibrillation With or Without Focal Impulse and
Rotor Modulation) trial. J. Am. Coll. Cardiol. 60,
628–636 (2012).
9. Connolly, S. J. et al. Dabigatran versus warfarin
in patients with atrial fibrillation. N. Engl. J. Med.
361, 1139–1151 (2009).
10. Abed, H. S. et al. Effect of weight reduction
and cardiometabolic risk factor management
on symptom burden and severity in
patients with atrial fibrillation: a randomized
clinical trial. JAMA 310, 2050–2060
(2013).
 CARDIOLOGY

DECADE IN REVIEW—CARDIOMYOPATHIES low severity within a single family). The

Cardiomyopathy on the move development and availability of large genetic

Magdi H. Yacoub
Since Wallace Brigden first used the term ‘cardiomyopathy’ in 1952,
this group of diseases has continued to attract the interest of clinicians,
researchers, and importantly, patients. The past decade has seen a
substantial accumulation of knowledge relating to various cardiomyopathies,
which has partially lifted the mystery surrounding this topic.
Yacoub, M. H. Nat. Rev. Cardiol. 11, 628–629 (2014); published online 30 September 2014;
doi:10.1038/nrcardio.2014.157
The current definition of cardiomyo­pathy,
suggested by the ESC and now widely
accepted worldwide, is a collection of “myo-
cardial disorders in which the heart muscle is
structurally and functionally abnormal in the
absence of coronary artery disease, hyper-
tension, valvular or congenital heart disease
sufficient to cause the observed myo­cardial
abnormality”. 1 This definition excludes
channelopathies, which do not lead to
structural changes in the myocardium. The
phenotypic classification proposed by John
Goodwin in 1961 of hypertrophic cardio­
myopathy (HCM), dilated cardiomyopathy
(DCM), and restrictive cardiomyopathy has
withstood the test of time, with the addition
of arrhythmogenic, non­compaction, and
Takotsubo (also known as acute broken heart
syndrome) cardiomyopathies. Although
clinically useful, this classification does not
consider the cause or molecular mechanisms
responsible for the disease.
In the past decade, the realization that a
familial link for DCM is more frequent than
originally thought led to the discovery of a
large number of mutations in genes encod-
ing proteins involved in various functions of
the myocardium, particularly the sarcomere
(Figure 1). One example is the gene that
encodes titin (TTN), which has an impor-
tant role in cardiac function, but the sheer
length of the gene and molecule it encodes
(the titin protein is ~1 μm long and spans
half a sarcomere) has inhibited attempts to
sequence it completely. Next-generation
sequencing approaches have been used to
identify mutations in the gene that result
in a truncated molecule in 14% of patients
with DCM, which increased to 27% when
parallel-capture next-generation sequenc-
ing was used.2 These findings have sub-
stantially increased the number (up to
70%) of patients with DCM who have a
known genetic mutation, and established
TTN as the gene most-commonly associ-
ated with DCM. This development will
databases, which include phenotypically-
defined populations of different ethnic
origin, such as the 1000 Genomes Project
and the Exome Sequencing Project, has
enabled researchers to identify a discrep-
ancy between the prevalence of potential
disease-causing variants and the estimated
prevalence of the disease.3,4 One explana-
tion for this discrepancy might be that the
cardiomyopathy phenotype is affected by
modifier genes (protective or otherwise),
epigenetics, or environmental factors that
further stimulate studies to determine the have yet to be identified. 5 The increas-
worldwide distribution of d ­ isease-causing ing use of next-generation sequencing has
mutations, their possible determinants, enabled simultaneous, more efficient, and
and importantly, the u­nderlying biological cost-effective detection of mutations in a
mechanisms responsible. large number of genes.2 This development,
Inheritance of DCM or HCM can be and formal population studies to determine
autosomal dominant, X‑linked, and occa- the true prevalence of the disease in different
sionally, autosomal recessive. However, the populations, might provide some answers to
incidence of these diseases in the popula- these seemingly complex issues.
tion is grossly underestimated, has low Myocardial interstitial and focal fibro-
penetrance, and low expressivity (that is, sis are hallmarks of HCM pathology that
a HCM b DCM c ARVC d LV noncompaction
cardiomyopathy
Sarcomere
Desmosome
JAG1 MIB1
TTN DES
DSP Cytoplasm
NOTCH1
NICD
MYBPC3
TPM1 TNNC1 DSG2 DSC2
TNNT2 TNNI3 ACTC1
Sarcolemma
NICD
MYL2/3 MYH6 PKP2 JUP
MYH7
TFs
Figure 1 | Genetic mechanisms of cardiomyopathies. a | Mutations in TTN and other genes
encoding sarcomeric proteins have been shown to cause HCM (yellow stars). b | Similarly,
mutations in genes encoding either sarcomeric or desmosomal proteins have been linked with
DCM (pink stars). c | Mutations in genes encoding desmosomal proteins have also been
implicated in the pathogenesis of ARVC (green stars). d | Mutations involved in the NOTCH1
signalling pathway are thought to underlie LV noncompaction cardiomyopathy (blue stars).
Abbreviations: ARVC, arrhythmogenic right ventricular cardiomyopathy; DCM, dilated
cardiomyopathy; HCM, hypertrophic cardiomyopathy; LV, left ventricular; NICD1, NOTCH1
intracellular domain; TF, transcription factor. Permission for inset panels from parts a–c obtained
from NPG © Hershberger, R. E. et al. Nat. Rev. Cardiol. 10, 531–547 (2013).

A DECADE IN MEDICINE NOVEMBER 2015  |  5

© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
contribute to the impaired diastolic and sys-
tolic function observed in patients with this
disease. The exact molecular mechanisms
responsible for interstitial fibrosis remain
unknown, with a unifying hypothesis invok-
ing developmental biology being proposed.6
Investigators using comprehensive transcrip-
tional RNA analysis of cardio­myocytes and
nonmyocytes from two animal models of
HCM identified several changes in expres-
sion of profibrotic genes in nonmyocytes,
which preceded myocardial hypertrophy.7
The investigators then prevented patho-
logical myocardial remodelling with allele-
specifi­c silencing of mutant transcripts to
target profibrotic molecules. This study
has important mechanistic and therapeutic
implications with regard to p­reventing or
reversing fibrosis in HCM.
Studies designed to improve our under-
standing of how genetic mutations are
translated into a clinical phenotype have
employed a wide spectrum of tools in the
past decade. These have included molecu-
lar modelling,8 structural and molecular
biology, biochemistry and pharmacology, as
well as biophysics. Coppini and colleagues
used patch clamping to compare the electro­
mechanical properties of 26 myectomy
samples from patients with HCM with
control tissue.9 The HCM specimens had
abnormalities in the late sodium current,
which were corrected by administration of
ranolazine, an inhibitor of the late sodium
channels.9 Aside from its value in under-
standing the pathophysiology of the disease,
this study has stimulated the development
of a clinical trial using ranolazine for the
­treatment of HCM.9
The initial discovery that mutations in
genes encoding sarcomeric proteins, such
as myosin heavy chains, can cause DCM
and HCM (Figure 1) has raised the hope
that such diseases might be prevented or
cured with strategies to counteract the
molecular abnormality. Progress towards
this goal is slowly being made. For example,
in 2013, a study involving allele-specific,
gene-mediated­ RNA interference in the
cardio­myocytes of a mouse model of HCM
resulted in partial ‘silencing’ of the mutated
molecule.10 This approach was sufficient to
prevent the mice from developing the HCM
phenotype. The discovery of genetic cell
engineering, without the use of viral vectors,
might provide effective ways of using such
strategies in humans with cardiomyo­
pathy, in the future. The past 10 years have
seen a rapidly accelerating rate of research
into the mechanisms and treatment of
6  |  NOVEMBER 2015  www.nature.com/reviews
cardiomyopathy, which has markedly
enhanced our understanding and manage-
ment of the disease, and is poised to deliver
much more in the coming decade.
Qatar Cardiovascular Research Center,
PO Box 5825, Doha, Qatar.
m.yacoub@imperial.ac.uk
Competing interests
The author declares no competing interests.
1. Cecchi, F., Tomberli, B. & Olivotto, I. Clinical and
molecular classification of cardiomyopathies.
Glob. Cardiol. Sci. Pract. 2012, 4 (2012).
2. Herman, D. S. et al. Truncations of titin causing
dilated cardiomyopathy. N. Engl. J. Med. 366,
619–628 (2012).
3. Golbus, J. R. et al. Population-based variation in
cardiomyopathy genes. Circ. Cardiovasc. Genet.
5, 391–399 (2012).
4. Pan, S. et al. Cardiac structure and sarcomeric
genes associated with cardiomyopathy exhibit
DECADE IN REVIEW—DYSLIPIDAEMIA
Resurgence of targets and
compounds to treat dyslipidaemia
John J. P. Kastelein
Over the past decade, we have witnessed the unparalleled success of statins
to treat dyslipidaemia. Target identification by Mendelian randomization,
human monoclonal antibodies, gene therapy, RNA-based targets, and
atherogenic lipoproteins other than LDL cholesterol have fuelled intense
development efforts that might bear fruit in the very near future.
Kastelein, J. J. P. Nat. Rev. Cardiol. 11, 629–631 (2014); published online 9 September 2014;
doi:10.1038/nrcardio.2014.132
When Terje Pedersen first presented the
results of the Scandinavian Simvastatin
Survival Study at the AHA Scientific
Sessions on 16 November 1994 in Dallas,
TX, USA, he probably did not realize that
the study would initiate a global revolu-
tion in statin therapy. Today, statins are
the most widely prescribed class of drugs,
and although the results of this study were
presented 20 years ago, no other drug has
since been shown to have any additional
benefit for patients with dyslipidaemia.
Inhibitors of phospholipases, cholesteryl
ester transfer protein (CETP), and cho-
lesterol absorption, as well as agonists of
peroxisome pr­oliferator-activated receptor,
thyroxin receptor, and nicotinic acid have
either been toxic or not provided a beneficial
outcome for patients. However, not all dis-
coveries in the past decade for the therapeu-
tic control of dyslipidaemia can be claimed
by statins. In 2003, Boileau and colleagues
mapped a locus associated with familial
© 2015 Macmillan Publishers Limited. All rights reserved
marked intolerance of genetic variation.
Circ. Cardiovasc. Genet. 5, 602–610 (2012).
5. Kalozoumi, G., Tzimas, C. & Sanoudou, D.
The expanding role of epigenetics. Glob. Cardiol.
Sci. Pract. 2012, 7 (2012).
6. Olivotto, I. et al. Developmental origins of
hypertrophic cardiomyopathy phenotypes:
a unifying hypothesis. Nat. Rev. Cardiol. 6,
317–321 (2009).
7. Teekakirikul, P. et al. Cardiac fibrosis in mice
with hypertrophic cardiomyopathy is mediated
by non-myocyte proliferation and requires Tgf‑β.
J. Clin. Invest. 120, 3520–3529 (2010).
8. Gaendrarao, P. et al. Molecular modeling of
disease causing mutations in domain C1
of cMyBP‑C. PLoS ONE 8, e59206 (2013).
9. Coppini, R. et al. Late sodium current inhibition
reverses electromechanical dysfunction in
human hypertrophic cardiomyopathy. Circulation
127, 575–584 (2013).
10. Jiang, J. et al. Allele-specific silencing of
mutant Myh6 transcripts in mice suppresses
hypertrophic cardiomyopathy. Science 342,
111–114 (2013).
hyper­cholesterolaemia in the gene encod-
ing proprotein convertase s­ubtilisin/kexin
type 9 (PCSK9). As with the Scandinavian
Simvastatin Survival Study, Boileau et al.
could not have foreseen that their discov-
ery would develop into the currently most
exciting class of lipid-lowering drugs—
monoclonal antibodies against circulating
PCSK9. These two discoveries have led to an
unprecedented number of developments in
the field of dyslipidaemia in the past decade.
In 2004, the authors of two papers first
established the clinical importance of statins
in the treatment of cardio­vascular disease,
which turned the hypothesis that ‘lower
LDL cholesterol is better’ into a core prin-
ciple. In the PROVE‑IT study,1 Cannon and
co-workers convincingly demonstrated that
80 mg of atorvastatin was superior to 40 mg
of pravastatin in outcome parameters includ-
ing death, myocardial infarction, u­nstable
angina, and revascularization (a  16%
risk reduction in favour or atorvasta­t in;
 CARDIOLOGY

P = 0.005), when administered shortly after (apoC‑III), apolipoprotein(a) [apo(a)], and 1986 2013
an acute coronary syndrome (ACS).1 This PCSK9, and which have all reached phase I 450 First
study transformed care for patients with and further clinical studies. These devel- Synvinolin
ACS; 80 mg of atorvastatin straight after opments are just the beginning, and many 400 study
an ACS episode has now become routine more will come in the next decade. In the
clinical practice in coronary care units in second paper, which garnered little atten- 350

most countries. This strategy has saved tion when first published, Stroes and col- 300

LDL-C levels (mg/dl)

innumer­able lives of patients with coronary leagues performed the first proof-of-concept
RUTHERFORD
artery disease (CAD) and protected them study of gene therapy for dys­lipidaemia. 250
against recurrent events. Alipogene tiparvovec increases the removal ASAP
ENHANCE
In the TNT study,2 also published in 2004, of postprandial chylo­microns, the particles 200

LaRosa and colleagues demonstrated that
80 mg was superior to 10 mg of atorva­statin
for treating patients with stable CAD. High-
dose atorvastatin reduced the primary end
point (occurrence of a major cardiovascular
known to cause the acute and potentially 150 RADIANCE
lethal haemor­rhagic pancreatitis associ-
ated with the chylo­micro­naemia syndrome. 100
Alipogene tiparvovec (commercially known
as Glybera®; UniQure, Netherlands) became 50

event) by 22% compared with the lower dose the first gene therapy product to be approved 0
of the drug (P <0.001).2 This study has also for any indication in the Western world.5

in

in
y
enabled many secondary analyses that aid Many developments since 1994 have

ap

at

at
0
Si 80

Si 40

g + 80

10
er

st

st
4
th
our understanding of renal function during revol­v ed around the biology, pathol-

+
va

va

va

va

e
Ez
No

g
m

m
or

m
Si

At
statin therapy, the clinical safety of low LDL- ogy, and therapeutic lowering of LDL-

60

0
42
m

et
80

o
cholesterol levels, and that a reduction in cholesterol levels. Only after a number of

rc

Ev
To
va
m
adverse clinical outcomes is independent international consortia performed large

of almost any baseline patient characteris-
tic, including LDL-cholesterol level itself.
These ‘statin principles’ were extended into
the realm of primary prevention in 2008 in
the JUPITER trial.3 Ridker et al. selected
patients with an elevated C‑reactive protein
level (measured by high-sensitivity assay),
but who were free from CAD, and demon-
strated that a daily dose of 20 mg rosuvasta-
tin robustly reduced the incidence of major
adverse cardio­vascular events by 44% com-
pared with placebo (P <0.00001).3 Again,
efficacy was observed in all subgroups,
including men and women, the elderly and
young, tobacco smokers and nonsmokers,
and those with or without metabolic syn-
drome. More importantly, the JUPITER
trial3 results highlighted the link between
inflammation, dyslipidaemia, and athero-
sclerotic vascular disease, which has led to
the current development of selective anti-
inflammatory strategies in outcome trials
by the same research group.
In two papers published in 2006 and
2008, respectively, investigators applied
advanced molecular biology techniques
to the study of lipid control.4,5 In the first
study, mRNA inhibition by subcutane-
ously administered small oligonucleotides
against apolipo­protein B‑100 (apoB‑100)
lowered the levels of all atherogenic lipo-
proteins in human volunteers with mild
dyslipidaemia.4 These results have led to
the development of small inhibitory RNAs,
mRNA inhibitors, and locked nucleic acid
inhibitors that target angiopoietin-related
protein 3, apoB‑100, apolipoprotein C‑III
Mendelian randomization studies was the
link made between CAD and apo(a), tri-
glyceride-rich lipoproteins, and rem­nant
cholesterol.6,7 These observations are semi­
nal for our understanding of athero­genesis
and initiated an international effort to iden-
tify novel strategies to decrease circulat-
ing levels of these lipoproteins. More­over,
the results of these Mendelian randomi-
zation studies, such as those performed
by Nordestgaard 6 and Kathiresan, 7 have
questioned the relationship between HDL
cholesterol and CAD, which might explain
the disappointing results of strategies that
raise HDL-cholesterol levels. By contrast,
Mendelian randomization and interven-
tion studies have both validated the role of
lowering LDL-cholesterol levels to treat dys-
lipidaemia. Consequently, the dis­coveries by
investigators that CETP, t­riglyceride-rich
lipoproteins, and apo(a) are indeed involved
in atherogenesis have led to a search for
compounds to safely lower the levels of these
proteins, as well as those of apoC‑III and, of
course, PCSK9. Mendelian randomization
studies are now an integral approach in the
development of drugs to treat dyslipidaemia.
Of all the developments over the past
decade, a paper by Stein and colleagues in
2012 deserves the title of ‘game changer’. In
this study, a monoclonal antibody against
PCSK9 (alirocumab), reduced the relative
LDL-cholesterol level by >60% in both
volun­t eers with normal lipid levels and
patients with familial hypercholesterol-
aemia.8 The excellent safety profile of this
drug class has been a relief for clinicians,
Si
Therapy
Figure 1 | LDL-cholesterol levels achieved by
lipid-lowering therapies developed during the
past 3 decades in patients with familial
hypercholesterolaemia. Dosing frequencies are
per day unless otherwise stated. Abbreviations:
Atorva, atorvastatixn; Evo, evolocumab; Eze,
ezetimibe; LDL‑C, LDL-cholesterol; Simva,
simvastatin; Torcet, torcetrapib.
especially given the previous toxicity-
related failures of other compounds. Simi­
lar results were obtained in studies with
bococizumab and evolocumab, and all three
monoclonal antibodies targeted against
PCSK9 are already in advanced, phase III
out­come trials. Only 9 years have passed
between Boileau and colleagues’ initial
PCSK9 discovery and Stein and co-­workers’
2012 paper. This rapid development has
been made possible by recombinant DNA
and antibody technology. Compared with
the >25 years between cloning of the lipo-
protein lipase (LPL) gene and the LPL gene
therapy trial by our own group, the devel-
opment of PCSK9-based technologies has
been a remarkable achievement.
However, for which patients is the dis-
covery of PCSK9 monoclonal antibodies
most beneficial in the short term? To have
personally witnessed the ever-decreasing
LDL-cholesterol levels that different thera-
pies have achieved in patients with familial
hypercholesterolaemia has been immensely
gratifying. Between the first simva­statin
trial in 1986 and the RUTHERFORD
study of evolocumab just last year, a steady

A DECADE IN MEDICINE NOVEMBER 2015  |  7

© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
decrease in LDL-cholesterol levels from
9.6 mmol/l (371 mg/dl) to ~1.7 mmol/l
(65 mg/dl) has been achieved (Figure 1).
The discovery that monoclonal antibodies
against PCSK9 can lower LDL-cholesterol
levels to such an extent has essentially
cured familial hypercholesterolaemia; in
fact, the LDL-cholesterol levels achieved
in these patients are now lower than in the
general population. Who would have had
the temerity to predict that in 2003?
The final discovery that I would like to
highlight was reported in two papers in
which triglyceride-rich lipoprotein and
remnant cholesterol levels were causally
linked to apoC‑III and CAD. 9,10 ApoC‑
III was first hypothesized to reduce tri­gly­
ceride levels and contribute to the risk of
CAD, and on that basis a mRNA inhibitor
of apoC‑III was developed. These two papers
strengthen the association between apoC‑III,
triglyceride metabolism, and CAD risk.9,10
The fact that the apoC‑III inhibitor is already
in clinical development is, therefore, timely
and fortuitous. With all of these results
comes the hope that novel small-­molecule
compounds, monoclonal antibodies, and
RNA technology will transform dyslipid­
aemia treatment in the coming decade, for
a second time since 1994. We might finally
be able to eliminate dys­lipidaemia and sub-
sequent atherosclerotic vascular disease for
our patients in the very near future.
Department of Vascular Medicine,
Academic Medical Center, University
of Amsterdam, Meibergdreef 9,
1105 AZ Amsterdam, Netherlands.
j.j.kastelein@amc.uva.nl
Competing interests
J.J.P.K. declares that he has acted as a consultant
and received honouraria from the following
companies: Aegerion, Amgen, AstraZeneca,
Atheronova, Boehringer Ingelheim, Catabasis,
Cerenis, CSL Behring, Dezima Pharmaceuticals,
Eli Lilly, Esperion, Genzyme, Isis, Merck, Novartis,
Omthera, Pronova, Regeneron, Sanofi, The Medicines
Company, UniQure, and Vascular Biogenics.
1. Cannon, C. P. et al. Intensive versus moderate
lipid lowering with statins after acute coronary
syndromes. N. Engl. J. Med. 350, 1495–1504
(2004).
2. LaRosa, J. C. et al. Intensive lipid lowering with
atorvastatin in patients with stable coronary
disease. N. Engl. J. Med. 352, 1425–1435
(2005).
3. Ridker, P. M. et al. Rosuvastatin to prevent
vascular events in men and women with
elevated C‑reactive protein. N. Engl. J. Med.
359, 2195–2207 (2008).
4. Kastelein, J. J. P. et al. Potent reduction of
apolipoprotein B and low-density lipoprotein
cholesterol by short-term administration of
an antisense inhibitor of apolipoprotein B.
Circulation 114, 1729–1735 (2006).
8  |  NOVEMBER 2015  www.nature.com/reviews
5. Stroes, E. S. et al. Intramuscular administration 8. Stein, E. A. et al. Effect of a monoclonal
of AAV1-lipoprotein lipase S447X lowers antibodies to PCSK9 on LDL cholesterol.
triglycerides in lipoprotein lipase-deficient N. Engl. J. Med. 366, 1108–1118 (2012).
patients. Arterioscler. Thromb. Vasc. Biol. 28, 9. Jørgensen, A. B., Frikke-Schmidt, R.,
2303–2304 (2008). Nordestgaard, B. G. & Tybjærg-Hansen, A.
6. Kamstrup, P. R., Tybjærg-Hansen, A., Loss‑of-function mutations in APOC3 and risk of
Steffensen, R. & Nordestgaard, B. G. ischemic vascular disease. N. Engl. J. Med. 371,
Genetically elevated lipoprotein(a) and 32–41 (2014).
increased risk of myocardial infarction. 10. The TG and HDL Group of the Exome
JAMA 301, 2331–2339 (2009). Sequencing Project, National, Heart, Lung, and
7. Do, R. et al. Common variants associated with Blood Institute. Loss-of-function mutations in
plasma triglycerides and risk for coronary artery APOC3, triglycerides, and coronary heart
disease. Nat. Genet. 45, 1345–1352 (2013). disease. N. Engl. J. Med. 371, 22–31 (2014).
DECADE IN REVIEW—HEART FAILURE
10 Years of progress in HF
research—what have we learned?
Henry Krum
In this Decade in Review article, I highlight the top 10 advances in heart
failure (HF) over the past decade, including new pharmacological therapies
and expanded indications for devices in HF with reduced ejection fraction.
The poor progress in acute HF and HF with preserved ejection fraction is
emphasised. Biomarkers and devices that help prevent, detect, and guide
treatment represent the future of HF management.
Krum, H. Nat. Rev. Cardiol. 11, 631–633 (2014); published online 2 September 2014;
doi:10.1038/nrcardio.2014.127
One of the great advances in cardio­vascular shown to be unsuccessful because of study
therapy during the past 20 years has been design flaws and off-target effects; however,
the use of cardiac resynchro­nization therapy combination with an angiotensin-receptor
(CRT) in patients with heart failure (HF) blocker (ARB) seems to overcome this risk
with reduced ejection fraction (HFrEF). and might offer clinical benefits beyond
Several studies have indicated major out­ ARB monotherapy. The researchers in the
come benefits in patients with HFrEF and PARADIGM‑HF study 2 aimed to deter-
advanced NYHA class III–IV symptoms. mine whether the ARB–neprilysin inhibitor
Investigators in MADIT‑CRT1 tested the LCZ696 would be superior to the gold-
hypothesis that mildly symptomatic patients standard ACE-inhibitor enalapril in patients
with HFrEF would also benefit from CRT. with HFrEF. Patients received either the
In a cohort of >1,000 patients, the primary ARB–neprilysi­n inhibitor LCZ696 (200 mg
end point (HF events or death) was reduced twice daily) or e­nalapril (10 mg twice daily),
to 17.2% with CRT versus 25.3% without with the primary end point of cardio­vascular
CRT, driven primarily by a reduction in HF death or HF hospitalization (n = 8,442).
events.1 These findings were supported by Although the final results have not yet been
a subsequent trial, in which HF events and presented, the study was terminated pre­
mortality were reduced by CRT. CRT is now maturely for overwhelming efficacy benefit.
guideline-recommended for all patients Given the positive outcome, and assuming
with HFrEF and a prolonged QRS duration that the drug has an acceptable safety profile,
(>150 ms). LCZ696 should supplant ACE inhibitors as
Major drug therapy trials directed towards standard therapy for these patients.
novel targets have also been conducted in The clinical benefit of β‑blockers for HF
the past decade (Figure 1). Pharmacological might be partly attributable to their capac-
inhibition of neprilysin can augment the ity to lower heart rate. If channel-blocking
physiological effects of endogenous natri­ agents, which are direct sinus node inhibi-
uretic peptides, which have beneficial tors, were used to investigate this heart-rate-
vasodilatory, anti-fibrotic, and natriuretic lowering hypothesis in the SHIfT study 3
actions. Dual therapy with angiotensin- (n = 6,558; 89% receiving backgroun­d
converting enzyme (ACE) inhibitors was β‑blocker therapy), in which the If channel
© 2015 Macmillan Publishers Limited. All rights reserved
 CARDIOLOGY

blocker iva­bradine was compared with Novel drug therapies

placebo. The primary end point (cardio-
vascular death or HF hospitalization) was
reached by 24% of patients receiving ivabra-
dine compared with 29% of patients taking
placebo, driven primarily by reduced hos-
pitalization for HF. The benefits of adding
ivabradine were diminished with increasing
dose of background β‑blockers, possibly
because the additive effect of ivabradine on
lowering the heart-rate was abrogated by
full-dose β‑blocker therapy. Ivabradine is
now approved as an adjunct to background
ACE inhibitor, β‑blocker, and mineralo­
corticoid-receptor antagonists (MRAs) in
symptomatic patients with HFrEF and a
sinus rhythm heart rate >70 bpm.
MRAs increase survival in patients with
advanced HFrEF and in those who have
experienced a myocardial infarction and
have HF. However, until the EMPHASIS‑HF
trial,4 patients with mild symptoms of HFrEF
had not been formally studied. Researchers
in EMPHASIS‑HF4 evaluated the efficacy of
eplerenone compared with placebo in 2,737
patients in NYHA class II. The trial was ter-
minated prematurely (median follow-up
21 months) because only 18.3% of patients
receiving eplerenone reached the primary
end point (cardiovascular death or hospitali-
zation for HF) compared with 25.9% in the
placebo group.4 Importantly, these benefits
were additional to background ACE inhibi-
tor and β‑blocker therapy. MRAs have, there-
fore, become mandated therapy for patients
with mild HFrEF.
In contrast to HFrEF, pharmaco­logical
therapy has so far failed to reduce major
adverse events in patients with HF and pre-
served ejection fraction (HFpEF). Given the
established benefit of MRAs in patients with
HFrEF, researchers in the TOPCAT study 5
evaluated the efficacy of spirono­lactone
(15–45 mg per day) versus placebo in 3,445
patients with an ejection fraction >45% in
addition to one other HF indicator, in an
attempt to recruit a more homogenous pop-
ulation with ‘true’ HFpEF. The occurrence
of the primary end point (cardiovascular
death, cardiac arrest, or HF hospitalization)
was not significantly different between the
two groups. Patients who entered the study
because of an elevated plasma B‑type natri­
uretic peptide (BNP) concentration exhibited
a stand-alone benefit on the primary end
point with spirono­lactone. This result sug-
gests that patients who fulfilled the criteria of
‘HF hospitalization’ rather than elevated BNP
might not have had true HFpEF, and in turn
did not benefit from MRA therapy. Further
Improved understanding
of pathophysiology
Advances in HF
HF preventive strategies
Plasma biomarker and
implantable device-based
tracking of patient clinic status
Figure 1 | Key advances in the past 10 years of HF research.
efforts to recruit a more homogenous patient
population and establish the existence of
true HFpEF before study entry are required
to evaluate the role of MRAs definitively in
this setting.
Improving HF outcomes by treat-
ing comorbidities has become a concept
of great therapeutic interest. Two inter­
related strategies involve the correction of
HF‑related anaemia and iron deficiency.
Erythropoietin-stimulating agents (ESAs)
are frequently used to treat anaemia, but have
not reduced major cardiovascular events in
HF patients. By contrast, correction of iron
deficiency has produced more encouraging
results. Investigators in the FAIR‑HF study 6
enrolled 459 patients with HFrEF and iron
deficiency, based on either serum ferritin or
transferrin saturation levels. Patients were
randomly allocated to either intravenous
ferric carboxymaltose or saline. Fifty percent
of the patients who received ferric carboxy-
maltose showed improvement according to
the Patient Global Assessment, compared
with only 28% of those receiving placebo.
However, an adequately-sized trial to assess
the effect of intravenous iron on morbidity
and mortality has not yet been performed in
patients with HFrEF who are iron-deficient.
Treatment of acute HF has not changed for
≥40 years. Numerous proposed agents have
demonstrated favourable effects on natri­
uresis and diuresis, but at the expense of wors-
ened renal function and lack of benefit on
prognosis. Relaxin is an endogenous peptide
with diuretic, natriuretic, vasodilatory, and
anti-fibrotic properties. Investigators in
the RELAX‑AHF study 7 evaluated 1,061
patients with acute HF randomly allocated to
serelaxin (a recombinant form of relaxin) or
placebo. Serelaxin improved one of two dysp-
noea end points, but had no effect on cardio-
vascular/renal death or hospital readmission
at 60 days in patients with acute HF. However,
fewer deaths occurred at 180 days with sere-
laxin, and evidence existed of reduced in-
hospital worsening of HF. On the basis of
Expanding indications
for HF therapeutic devices
Gene-based therapies
these mixed findings, regulatory authorities
did not approve the agent; however, the find-
ings were sufficiently compelling to warrant
a definitive mortality trial, RELAX‑II, which
is currently ongoing.
Gene-based approaches for HF manage-
ment are beginning to emerge, after much
promise. Insertion of the sarcoplasmic reticu­
lum Ca2+-ATPase (SERCA2a) gene into the
failing heart has been considered a potential
therapeutic strategy on the basis of its pivotal
role in myocardial contractile function. The
CUPID study 8 involved 39 patients with HF
who received intracoronary adeno-associat­ed
virus type 1/SERCA2a or placebo. Initial
findings suggested that gene transfer might
improve relevant surrogate end points and
reduce cardiovascular events. The CUPID 2
trial, with an increased cohort of patients,
is ­currently ongoing.
The discovery of biomarkers that indi-
cate the likelihood of risk factors trans­
itioning to left ventricular (LV) dysfunction
might change the future of HF diagnosis.
Investigators in the STOP‑HF study 9 trialled
a BNP screening programme in which high-
risk asymptomatic patients (n = 1374) were
randomly assigned to receive care based on
knowledge of the plasma BNP level, or usual
care. Using this approach, 5.3% of the BNP
group developed LV dysfunction (with or
without HF) compared with 8.7% receiving
usual care, over a 4.2‑year period. Additional
biomarkers, including some specific to the
LV dysfunction transition phase (such as
ST2, galectin‑3), have been proposed and
are currently under assessment.
Tracking of patient clinical status has
under­g one substantial technological
advance in the past decade. In an individual
patient meta-analysis of over 2,000 individu-
als with HFpEF or HFrEF, BNP-guided HF
treatment was shown to significantly reduce
all-cause mortality and HF hospitali­zation
in those <70 years of age.10 Newly avail-
able tools for tracking patient clinical status
remotely, such as pulmonary pressure

A DECADE IN MEDICINE NOVEMBER 2015  |  9

© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
measurement using the CardioMEMS™ HF
system (St. Jude Medical, USA), can com-
plement biomarker-guided therapy, permit-
ting earlier intervention in, for example, the
acute decompensation setting.
Substantial advances have been made in
HF treatment over the past decade. Prog­
nosis is improving for patients with HFrEF,
with novel pharmacological strategies, new
devices and interventions, as well as cell-
and gene-based therapies currently being
developed (Figure 1). However, success-
ful strategies in both acute HF and HFpEF
remain elusive. Personalized medicine will
be a major research focus in the future,
integrating neurohormonal, proteomic,
metabolomic, and genomic information into
therapeutic decision-making.
Centre of Cardiovascular Research
& Education in Therapeutics, School of
Public Health & Preventive Medicine,
Monash University, 99 Commercial Road,
Melbourne, VIC 3004, Australia.
henry.krum@monash.edu
Competing interests
H.K. declares that he has received research grants
from Bayer, Medtronic, Novartis, Pfizer, and Servier.
1. Moss, A. J. et al. Cardiac-resynchronization
therapy for the prevention of heart failure events.
N. Engl. J. Med. 361, 1329–1338 (2009).
2. McMurray, J. J. et al. Baseline characteristics and
treatment of patients in Prospective comparison
of ARNI with ACEI to Determine Impact on
Global Mortality and morbidity in Heart Failure
trial (PARADIGM-HF). Eur. J. Heart Fail. 16,
817–825 (2014).
3. Swedberg, K. et al. Ivabradine and outcomes
in chronic heart failure (SHIFT): a randomised
placebo-controlled study. Lancet 376, 875–885
(2010).
4. Zannad, F. et al. Eplerenone in patients with
systolic heart failure and mild symptoms.
N. Engl. J. Med. 364, 11–21 (2011).
5. Pitt, B. et al. Spironolactone for heart failure
with preserved ejection fraction. N. Engl. J. Med.
370, 1383–1392 (2014).
6. Anker, S. D. et al. Ferric carboxymaltose in
patients with heart failure and iron deficiency.
N. Engl. J. Med. 361, 2436–2448 (2009).
7. Teerlink, J. R. et al. Serelaxin, recombinant
human relaxin‑2, for treatment of acute heart
failure (RELAX-AHF): a randomised, placebo-
controlled trial. Lancet 381, 29–39 (2013).
8. Jessup, M. et al. Calcium upregulation by
percutaneous administration of gene therapy
in cardiac disease (CUPID): a phase 2 trial
of intracoronary gene therapy of sarcoplasmic
reticulum Ca2+-ATPase in patients with
advanced heart failure. Circulation 124,
304–313 (2011).
9. Ledwidge, M. et al. Natriuretic peptide-based
screening and collaborative care for heart
failure: the STOP-HF randomized trial. JAMA
310, 66–74 (2013).
10. Troughton, R. W. et al. Effect of B‑type
natriuretic peptide-guided treatment of chronic
heart failure on total mortality and
hospitalization: an individual patient meta-
analysis. Eur. Heart J. 35, 1559–1567 (2014).
10  |  NOVEMBER 2015  www.nature.com/reviews
DECADE IN REVIEW—HYPERTENSION
The past decade in hypertension
—facts, hopes, and hypes
Thomas Unger
Hypertension research in the past decade has been a mixture of hope
and hype. In the absence of new drug developments, clinical intervention
procedures such as renal nerve ablation and baroreflex activation therapy
have dominated the research, but the results have not yet fulfilled the
great expectations.
Unger, T. Nat. Rev. Cardiol. 11, 633–635 (2014); published online 2 September 2014;
doi:10.1038/nrcardio.2014.131
“A simple, brief, catheter-based procedure to
ablate sympathetic nerves can be done safely
without long-term complications and could
result in persistent reductions of blood pres-
sure.”1 Such was the hope 5 years ago after the
publication of a study involving 45 patients
by an international team of researchers who
investigated the efficacy of catheter-based
renal sympathetic denervation for resistant
hypertension.1 Indeed, initial findings sug-
gested that resistant hypertension, which
is refractory to pharmacological therapy,
could be treated with an interventional
approach. This study fuelled an outburst
of studies on renal sympathetic nerve abla-
tions for the treatment of hypertension.
However, much to the disappointment of
cardiologists worldwide, findings from the
SYMPLICITY HTN‑3 trial2 released early in
2014 demonstrated no significant differences
in systolic blood pressure between the renal-
denervation group and the sham-controlled­
group after 6 months (Figure 1). Now that
the hype surrounding renal sympathetic
denervation has subsided, we must determine
whether this procedure can offer permanent
benefit to patients with hypertension and,
most importantly, which treatment-resistant
subgroups (age, ethnicity, sex, duration
of hypertension, state of the sympathetic
nervous system) are eligible for denervation.
Immunization against angiotensin II to
reduce blood pressure represented another
promising strategy to treat hypertension
6 years ago. The efficacy of this strategy
was evaluated in a phase IIa trial involv-
ing 72 patients.3 After a 14‑week treatment
period, a significant reduction in ambula-
tory blood pressure was observed in patients
who received a higher dose of the vaccine,
compared with placebo. However, these
findings could not be reproduced in further
clinical studies. Nevertheless, vaccinations
against a population-wide disease such as
© 2015 Macmillan Publishers Limited. All rights reserved
hyper­tension might still be effective if an
appropriate target can be identified, and the
associated immunological problems can
be overcome.
A third nonpharmacological approach for
the treatment of hypertension has emerged
for patients with resistant hypertension.
Baroreflex activation therapy (BAT) involves
the implantation of a device that per­manently
stimulates the baroreceptor reflex, which is
usually attenuated in patients with hyper-
tension. Although BAT for lowering blood
pressure is no longer a novel concept, it
has regained attention because of marked
improvements in device technology. In a ran-
domized trial published in 2011, 265 patients
with resistant hypertension were assigned to
either group A (BAT applied permanently for
up to 12 months) or group B (BAT deferred
until after 6 months).4 No differences were
apparent between the two groups after
6 months, but an ancillary analysis revealed
a significantly higher percentage of patients
(42% versus 24%, P <0.005) who achieved
systolic blood pressure ≤140 mmHg in
group A compared with group B.4 A treat-
ment regimen involving surgical implant­
ation of a device is unlikely to become
standard therapy for all hypertensive patients,
but this strategy might be suitable for selected
patients who do not respond s­ufficiently to
anti-hypertensive drug therapy.
In the past 2 decades, the development
of drugs for the treatment of hyperten-
sion and other cardiovascular diseases has
been largely dominated by inhibitors of
the renin–angiotensin system, reflected by
many comparative drug studies with large
patient cohorts. A prime example is the
ONTARGET trial,5 in which the efficacy
of the angiotensin receptor blocker (ARB),
telmisartan, was compared with the gold-
standard angiotensin-converting-enzyme
(ACE) inhibitor, ramipril, in patients with
 CARDIOLOGY

indicate a physiological role for H2S as a

Baroreflex vasodilator and blood-pressure regulator.
activation therapy
In 2011, a paper examining the association
between angiotensin II-induced hyper­
Immunisation
tension and vascular injury, and its relation-
against ship with T‑regulatory lymphocytes was
angiotensin II
published.8 The investigators convincingly
showed that T‑regulatory lymphocytes sup-
pressed angiotensin II-mediated vascular
Reduction in injury through their anti-inflammatory
blood pressure actions,8 adding support to the hypothesis
that immune mechanisms modulate the
various physiological and pathophysio­
Renal
denervation logical actions of angiotensin II. This notion
is consistent with the pathological concept
RAS inhibitors
that ‘undercurrents’, such as immune mech-
anisms, inflammation, proliferation, and
fibrosis, are important in disease processes,
Drug intervention
Nondrug intervention
forming a link between various pathologies
(such as those involving the kidneys, heart,
Figure 1 | The hits and misses in the development of blood-pressure lowering therapies and blood vessels).9
during the past decade. The ARB telmisartan was shown to be as effective as the gold- Finally, we should not forget genetics.
standard ACE-inhibitor ramipril in the ONTARGET trial. However, nondrug interventions were The findings from a study published in 2009
less successful. Renal denervation failed to reduce systolic blood pressure in patients with
offered an early insight into the genetic basis
resistant hypertension in the SYMPLICITY HTN‑3 trial. Furthermore, initial findings that
of hypertension, and the identified loci might
immunisation against angiotension II reduced ambulatory blood pressure could not be
reproduced in subsequent clinical studies. Finally, baroreflex activation therapy did not be new targets for the reduction of blood
significantly reduce systolic blood pressure after 12 months in a randomized controlled trial, pressure.10 After decades of heavy spending
but significantly increased the number of patients who achieved systolic blood pressure of and international collaborative endeavours
≤140 mmHg. Abbreviations: ACE, angiotensin-converting-enzyme; ARB, angiotensin receptor to unveil the complex genetic background
blocker; RAS, renin–angiotensin system. underlying hypertension, we might be
moving away from the general ‘one-size-fits-
coronary artery disease or diabetes mel­ such a trial. In the future, mega-trials on all’ approach and towards more personalized
litus, in addition to other cardiovascular therapeutic regimens will have to be replaced methods of diagnosis and treatment.
risk factors, including hypertension. After by smaller, better defined, and more indi- At present, the future of hypertension
a mean follow-up period of 5 years, telmi­ vidualized clinical studies. Secondly, the research is somewhat diffuse. The golden
sartan was noninferior to ramipril in terms of long-debated efficacy of ACE inhibitors age of hypertension research appears to have
the composite primary end point of cardio­ versus ARBs in treating cardiovascular come to a temporary end, as most of the
vascular death, myocardial infarction, stroke, disease can now be considered settled; each large pharmaceutical drug companies have
or hospitalization for heart failure, and was is noninferior to the other, but ARBs are stopped developing new anti-hypertensive
better tolerated than the ACE inhibitor. generally better tolerated. Thirdly, in the therapies. Despite this, new areas of research
A combination of the two drugs did not TRANSCEND trial,6 a combination of two are emerging; scientists are beginning to view
provide further benefit, but in fact increased different classes of renin–angiotensin inhibi- hypertension as a pathological interface in
the risk of hypotension, syncope, renal dys- tors was not beneficial for patients at high the development of disorders involving the
function, and hyperkalaemia. In the parallel cardiovascular risk. For these reasons, the blood vessels, heart, and kidneys, and will
TRANSCEND trial6 involving 5,926 patients ONTARGET–TRANSCEND trial5,6 was a develop strategies to target common patho-
who were unable to tolerate ACE inhibitors, hallmark study of the past decade, and will logical ‘undercurrents’ such as inflammation,
telmisartan was superior to placebo treat- probably be the last mega-trial of this kind fibrosis, and proliferative processes.
ment only when hospitalization for heart in cardiovascular therapy.
failure was removed from the combined Concerning the pathophysiology of CARIM School for Cardiovascular Diseases,
Maastricht University, PO Box 616, 6200 MD
primary end point. hypertension, two seminal publications have Maastricht, Netherlands.
At first glance, these two trials do not aroused much interest in the past decade. An t.unger@maastrichtuniversity.nl
seem to have conveyed any new knowl- article describing hydrogen sulphide (H2S)
edge; however, on further inspection, many as a physiological vasorelaxant was pub- Competing interests
The author declares no competing interests.
lessons can be gained from them. Firstly, lished in 2008.7 Genetic deletion of the H2S-
the conventional strategy of clinical trials generating enzyme cystathionine γ‑lyase 1. Krum, H. et al. Catheter-based renal sympathetic
in cardiovascular disease (that is, to extract markedly reduced H2S levels in serum and denervation for resistant hypertension:
marginally significant benefits of one drug tissue. Mutant mice lacking this enzyme a multicentre safety and proof‑of‑principle
cohort study. Lancet 373, 1275–1281 (2009).
versus another) will be questioned because showed pronounced hyper­tension, together
2. Bhatt, D. L. et al. A controlled trial of renal
the gain in clinical evidence is disproportion- with diminished ­endothelium-dependent denervation for resistant hypertension. N. Engl.
ate to the time and financial expenditure of vascular relaxation. These results strongly J. Med. 370, 1393–1401 (2014).

A DECADE IN MEDICINE NOVEMBER 2015  |  11

© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
3. Tissot, A. C. et al. Effect of immunisation
against angiotensin II with CYT006-AngQb on
ambulatory blood pressure: a double-blind,
randomised, placebo-controlled phase IIa
study. Lancet 371, 821–827 (2008).
4. Bisognano, J. D. et al. Baroreflex activation
therapy lowers blood pressure in patients with
resistant hypertension: results from the double-
blind, randomized, placebo-controlled Rheos
Pivotal Trial. J. Am. Coll. Cardiol. 58, 765–773
(2011).
5. Yusuf, S. et al. Telmisartan, ramipril, or both
in patients at high risk for vascular events.
N. Engl. J. Med. 358, 1547–1559 (2008).
6. Yusuf, S. et al. Effects of the angiotensin-receptor
blocker telmisartan on cardiovascular events in
high-risk patients intolerant to angiotensin-
DECADE IN REVIEW—PERIPHERAL VASCULAR DISEASE
10 Years of breakthroughs
in peripheral vascular disease
Mark A. Creager
Clinical trials published during the past decade have had substantial
effects on the treatment of peripheral vascular diseases. In this article,
I discuss ten important trials that have influenced treatment for common
vascular disorders, including peripheral artery disease, abdominal aortic
aneurysm, renal artery disease, extracranial carotid artery disease, and
venous thromboembolism.
Creager, M. A. Nat. Rev. Cardiol. 11, 635–636 (2014); published online 30 September 2014;
doi:10.1038/nrcardio.2014.153
Patients with peripheral artery disease
(PAD) are at increased risk of myocardial
infarction, stroke, and cardiovascular death.
Antiplatelet therapy, including aspirin, is
recommended for patients with PAD. Two
prospective studies were designed to deter-
mine whether aspirin can prevent cardio­
vascular events in asymptomatic patients
with PAD. Investigators in the Aspirin
for Asymptomatic Atherosclerosis Trial1
recruited 28,980 individuals free from
clinical cardiovascular disease, of whom
3,350 were found to have PAD on the basis
of an ankle–brachial index (ABI) ≤0.95.
Participants received either aspirin (100 mg
daily) or placebo, and were followed up for
a mean of 8.2 years. The primary end point
of fatal or nonfatal coronary events, stroke,
or revascularization was not significantly
different between the aspirin and placebo
groups (HR 1.03, 95% CI 0.84–1.27).1 In
the POPADAD trial,2 investigators exam-
ined the efficacy of aspirin in reducing
cardio­vascular events in 1,276 patients with
asymptomatic PAD (defined as an ABI
≤0.99) and either type 1 or type 2 diabetes
mellitus. Patients received aspirin (100 mg
12  |  NOVEMBER 2015  www.nature.com/reviews
converting enzyme inhibitors: a randomised
controlled trial. Lancet 372, 1174–1183 (2008).
7. Yang, G. et al. H2S as a physiologic
vasorelaxant: hypertension in mice with
deletion of cystathionine gamma-lyase. Science
322, 587–590 (2008).
8. Barhoumi, T. et al. T regulatory lymphocytes
prevent angiotensin II‑induced hypertension
and vascular injury. Hypertension 57, 469–476
(2011).
9. Hess, K., Marx, N. & Lehrke, M. Cardiovascular
disease and diabetes: the vulnerable patient.
Eur. Heart J. Suppl. 14, B4–B13 (2012).
10. Newton-Cheh, C. et al. Genome-wide
association study identifies eight loci
associated with blood pressure. Nat. Genet. 41,
666–676 (2009).
daily) or placebo, and were followed up for
a median of 6.7 years. No significant differ-
ence in the primary composite end point
of fatal coronary heart disease or stroke,
nonfatal myocardial infarction, nonfatal
stroke, or above ankle amputation for criti-
cal limb ischaemia was observed in patients
treated with or without aspirin (HR 0.98,
95% CI 0.76–1.26).2 These findings reject the
notion that aspirin therapy is effective for all
patients with PAD, and led to modifications
in ACCF/AHA guideline recommendations
regarding the use of antiplatelet therapy in
patients with asymptomatic PAD.
Although both exercise training and
endovascular revascularization improve
symptoms of intermittent claudication in
patients with PAD, the merits of each treat-
ment have been controversial. The CLEVER
study 3 involved 111 patients with PAD and
aortoiliac artery steno­sis randomly allo-
cated to receive optimal medical therapy
alone, or in combination with supervised
exercise training or stent revascularization.
After 6 months, the change in peak walking
time was greater in the exercise group than
in the stenting group (5.8 ± 4.6 min versus
© 2015 Macmillan Publishers Limited. All rights reserved
3.7 ± 4.9 min; P = 0.04), and both interven-
tions were superior to medical therapy alone
(1.2 ± 1.6 min; P <0.001 and P <0.02, respec-
tively).3 This study confirms the efficacy of
exercise training for patients with claudi­
cation and aortoiliac artery disease, but
also supports stenting for patients unable
or unwilling to participate in an exercise
t­raining programme.
Repair of large abdominal aortic aneu-
rysms (AAAs) reduces the risk of aortic
rupture and aneurysm-related mortality.
The efficacy of endovascular repair versus
open surgical repair was evaluated in
345 patients with an AAA ≥5.0 cm in the
DREAM trial.4 The primary end point, a
composite of 30‑day operative mortality
and severe complications, was numerically
greater in the open surgical group than in
the endovascular repair group (risk ratio
[RR] 2.1, 95% CI 0.9–5.4, P = 0.1), and
operative mortality alone trended higher
in the open surgical group (RR 3.9, 95% CI
0.9–32.9, P = 0.1).4 The incidence of severe
complications was higher after surgical
repair (RR 3.9, 95% CI 1.3–9.1, P <0.01). The
efficacy of endovascular repair versus open
repair of AAAs ≥5.5 cm was also assessed
in 1,082 patients in the EVAR trial 1.5 All-
cause mortality was not significantly dif-
ferent between the two groups, although
a­neurysm-related mortality was signifi-
cantly reduced after endovascular repair
(HR 0.55, 95% CI 0.31–0.96, P = 0.04). 5
Postoperative complications and the need
for re­intervention occurred more frequently
in the endovascular group. Together, these
studies support endovascular repair of
AAAs as an alternative to surgical repair,
but highlight the importance of long-term
surveillance for endograft complications.
The efficacy of endovascular repair of
AAAs among patients deemed unfit for
open surgical repair was evaluated in the
EVAR trial 2.6 Patients with an AAA ≥5.5 cm
(n = 3,338) were randomly allocated to endo-
vascular repair or no intervention and fol-
lowed up for up to 4 years. Neither all-cause
nor aneurysm-related mortality differed
significantly between the two groups.6 The
30‑day operative mortality in the endo­
vascular repair group was 9%, indicating the
high-risk nature of the study participants.6
These observations raise an important note
of caution for endovascular repair of AAAs
in patients unfit for open surgery, and suggest
that attention should be instead focused on
improving overall medical fitness.
The role of renal artery stenting for the
treatment of renal artery stenosis has been

controversial. Investigators in the CORAL
study 7 compared the efficacy of stenting
plus standard medical therapy to medical
therapy alone in 947 patients with severe
renal artery stenosis and systolic hyper­
tension or chronic kidney disease. Standard
medical therapy included an angiotensin
receptor blocker, with or without a d­iuretic,
a calcium channel blocker, and a statin. After
up to 5 years of follow-up, no signif­icant dif-
ference was found between the two groups
in the primary composite end point of death
from cardiovascular or renal causes, myo-
cardial infarction, stroke, hospitalization for
congestive heart failure, progressive renal
insufficiency, or renal replacement therapy.7
Furthermore, no significant difference
existed in all-cause mortality. Systolic blood
pressure decreased by 15.6 ± 25.8 mmHg
in the medical therapy only group and by
16.6 ± 21.2 mmHg in the stented group,
and in a longitudinal analysis, was mod-
estly less in patients from the stented group
(–2.3 mmHg; 95%  CI 0.2–4.4 mmHg,
P = 0.03).7 The observation that stenting
was no more effective than evidence-based
medical therapy to reduce cardio­vascular
or renal events should diminish referral
for stenting in the majority of patients with
renal artery stenosis. However, percutaneous
balloon angioplasty and stenting should still
be considered in patients with hypertension
associated with renal artery fibromuscular
dysplasia, and those with bilateral renal
artery sten­o sis and epi­s odic pulmonary
oedema, respectively, because these groups
were not included in the CORAL study.
Both carotid artery endarterectomy and
carotid artery stenting are used to treat
extracranial carotid artery stenosis, but
considerable debate exists about the relative
efficacy and safety of the two procedures.
Researchers in the CREST trial8 compared
2,502 patients with carotid artery stenosis
who underwent either stenting or endarter­
ectomy, and found no significant difference
in the primary end point of stroke, myo­
cardial infarction, or all-cause mortality
during the periprocedural period or any
ipsilateral stroke within 4 years (HR 1.11,
95% CI 0.81–1.51, P = 0.51). There was no
difference between treatment groups in
the 4 year rate of ipsilateral stroke in either
symptomatic or asymptomatic patients.
The peri­p rocedural rate of stroke was
higher in the stented group (4.1% versus
2.3%; P = 0.01), whereas the periprocedural
rate of myo­cardial infarction was higher
in the endarterectomy group (2.3% versus
1.1%; P = 0.03).8 No significant difference
A DECADE IN MEDICINE NOVEMBER 2015  |  13
was observed in periprocedural mortality.
These findings indicate that carotid artery
endarter­e ctomy and stenting are equally
effective for treating patients with carotid
artery stenosis. Stenting, therefore, can be
considered as a therapeutic alternative to
carotid endarterectomy. Importantly, carotid
revascularization was not compared with
optimal medical therapy in the CREST trial.
Whether medical therapy is the preferred
approach, particularly in asympto­m atic
patients with carotid s­ tenosis, remains to
be determined.
‘‘
These findings reject the
notion that aspirin is effective for
all patients with PAD...
’’
Anticoagulation therapy is indicated in
patients with venous thromboembolism
(VTE) to reduce the recurrence of deep vein
thrombosis (DVT) and pulmonary embo-
lism. Several clinical trials have examined
the efficacy of novel oral anticoagulants,
which work by inhibiting thrombin or
factor Xa. Investigators in the RE‑COVER
trial,9 a noninferiority study, compared the
efficacy and safety of dabigatran, a direct
thrombin inhibitor, with warfarin in 1,274
patients with acute DVT or pulmonary
embolism. Patients were initially treated
with a parenteral anticoagulant for at least
5 days, and subsequently received either
dabigatran or warfarin. There was no dif-
ference between the two groups in the
primary composite outcome of the inci-
dence of recurrent symptomatic venous
thromboembolism or related deaths at
6 months (HR 1.10, 95% CI 0.65–1.84),
and the difference in risk was 0.4% (95% CI
–0.8% to 1.5%; P <0.001 for noninferior-
ity). There was no significant difference
in major bleeding events between groups,
although the incidence of major or clinically
relevant nonmajor bleeding was reduced
with dabigatran.9 In the EINSTEIN-DVT
trial,10 the efficacy and safety of rivaroxa-
ban, an oral factor Xa inhibitor, were com-
pared with those of a vitamin K antagonist
in 3,449 patients with acute symptomatic
DVT. Patients received either rivaroxaban,
or enoxaparin followed by a vitamin K
antagonist. Rivaroxaban was noninferior
to vitamin K antagonism for recurrent VTE
(HR 0.68, 95% CI 0.44–1.04, P <0.001 for
noninferiority).10 No significant difference
was observed in the primary safety outcome
of major bleeding or clinically relevant non-
major bleeding. Taken together, these two
© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
studies indicate that these novel oral anti-
coagulant drugs are as effective and safe as
warfarin (or similar vitamin K antagonists),
have similar or improved safety profile, and
can thus be a safe alternative to warfarin for
the management of acute VTE.
Over the past decade, evidence from
clinical trials has guided the management of
peripheral atherosclerotic vascular diseases
and venous thrombosis. The next 10 years
of research is anticipated to uncover novel
treatments to reduce the risk of peripheral
atherosclerotic events, inhibit the growth
of aortic aneurysms, improve symptoms of
peripheral artery disease, and improve the
outcomes of patients with venous t­hrombosis
and p­ulmonary embolism.
Cardiovascular Division, Brigham and
Women’s Hospital, 75 Francis Street, Boston,
MA 02115, USA.
mcreager@partners.org
Competing interests
M.A.C. declares that he has received consulting fees
from AstraZeneca and Novartis.
1. Fowkes, F. G. et al. Aspirin for prevention of
cardiovascular events in a general population
screened for a low ankle brachial index:
a randomized controlled trial. JAMA 303,
841–848 (2010).
2. Belch, J. et al. The prevention of progression of
arterial disease and diabetes (POPADAD) trial:
factorial randomised placebo controlled trial
of aspirin and antioxidants in patients with
diabetes and asymptomatic peripheral arterial
disease. BMJ 337, a1840 (2008).
3. Murphy, T. P. et al. Supervised exercise versus
primary stenting for claudication resulting
from aortoiliac peripheral artery disease:
six-month outcomes from the claudication:
exercise versus endoluminal revascularization
(CLEVER) study. Circulation 125, 130–139
(2012).
4. Prinssen, M. et al. A randomized trial comparing
conventional and endovascular repair of
abdominal aortic aneurysms. N. Engl. J. Med.
351, 1607–1618 (2004).
5. EVAR trial participants. Endovascular aneurysm
repair versus open repair in patients with
abdominal aortic aneurysm (EVAR trial 1):
randomised controlled trial. Lancet 365,
2179–2186 (2005).
6. EVAR trial participants. Endovascular aneurysm
repair and outcome in patients unfit for open
repair of abdominal aortic aneurysm (EVAR
trial 2): randomised controlled trial. Lancet
365, 2187–2192 (2005).
7. Cooper, C. J. et al. Stenting and medical therapy
for atherosclerotic renal-artery stenosis.
N. Engl. J. Med. 370, 13–22 (2014).
8. Brott, T. G. et al. Stenting versus endarterectomy
for treatment of carotid-artery stenosis. N. Engl.
J. Med. 363, 11–23 (2010).
9. Schulman, S. et al. Dabigatran versus
warfarin in the treatment of acute venous
thromboembolism. N. Engl. J. Med. 361,
2342–2352 (2009).
10. EINSTEIN Investigators. Oral rivaroxaban
for symptomatic venous thromboembolism.
N. Engl. J. Med. 363, 2499–2510 (2010).
CARDIOLOGY

DECADE IN REVIEW—VALVULAR DISEASE technical challenges after the launch of these

Current perspectives on treatment prostheses. Meanwhile, with the availability

of a multitude of improvised valve prosthe-

of valvular heart disease ses and hardware used for implant­ation, as

well as the growing experience of cardiac
surgeons and interventional cardio­logists,
Friedrich W. Mohr
TAVI has become a routine pro­c edure
Tremendous advances in the understanding and treatment of structural and a good alternative for inoperable and
heart valve disease have been made in the past decade, including high‑risk patients (Figure 1).
Similarly in mitral valve surgery, attempts
widespread utilization of minimally-invasive surgical procedures and
to minimize patient trauma have led to
the invention and evolution of numerous interventional techniques. interesting progress in minimally-invasive
These innovations will become the norm in therapy for valvular disease surgery. Both mitral valve replacement and
in the future. reconstruction can be performed safely
Mohr, F. W. Nat. Rev. Cardiol. 11, 637–638 (2014); published online 23 September 2014; using minimally-invasive or even robot-
doi:10.1038/nrcardio.2014.138 assisted techniques in centres with sufficient
experience. Repair for mitral regurgitation
Valve repair and replacement continue to Alain Cribier,2 and its establishment as an has become an indicator of quality in many
be better treatment options than conserva- effective treatment in old, morbid, and vir- centres. We have learned that reconstruction
tive management for patients with struc- tually inoperable patients. The benefits of is feasible and desirable in most patients with
tural heart valve disease. The past decade avoiding a sternotomy, extracorporeal cir- degenerative disease. However, the results of
has seen major advances in the approaches culation, cardioplegic arrest, and (with a reconstruction for ischaemic mitral regur-
used to perform these procedures, espe- trans­femoral approach) even general anaes- gitation are mixed, particularly in the long
cially with the advent of minimally-­invasive thesia, have given this technique a boost, term. 3 As a result of the poor prognosis
surgical and interventional techniques. despite higher rates of residual aortic regur- in these patients owing to the combina-
Some of these techniques are still in their gitation, pacemaker implantations, and early tion of progressive heart failure and mitral
infancy, whereas others are being routinely
used globally, particularly in high-volume 20,000
TAVI
centres. These developments have enabled
AVR allograft
heart specialists to treat an increasingly AVR biological
18,000
wide spectrum of patients, especially those AVR mechanical
who would otherwise have been consid-
ered inoperable. Simultaneous progress in 16,000
imaging has improved delineation of the
anatomy and pathology of heart valves.
One of the most remarkable accomplish- 14,000
ments has been the establishment of the
Total number of procedures

Heart Team concept, which involves experi­ 12,000

enced cardiologists and cardiac surgeons
collaborating to devise unbiased treatment
s­trategies for patients. 10,000
In the surgical treatment of aortic valve
disease, two new trends have been observed 8,000
in the past 10 years. Firstly, a major shift
worldwide has occurred towards increased
use of biological valve prostheses, which 6,000
has been boosted by the availability of
newer-generation bioprostheses with excel-
4,000
lent haemodynamic characteristics and
long-term durability (Figure 1).1 Secondly,
minimally-invasive procedures through 2,000
small incisions have been shown to have
the same safety and efficacy as those per-
0
formed conventionally through a standard
94
95

96

97

98

99

00
01

02

03

04

05

06
07

08

09

10

11

12

median sternotomy. The development of

13
19
19

19

19

19

19

20
20

20

20

20

20

20
20

20

20

20

20

20
20

sutureless implants has further facilitated Year

these approaches. Figure 1 | Development of isolated AVR in Germany from 1994 to 2013. The graph shows
One of the most memorable develop- the rise in the total number of AVRs over this period, the increasing use of biological valves,
ments was the first transcatheter aortic and the rapid growth in TAVI since 2008. Abbreviations: AVR, aortic valve replacement; TAVI,
valve implantation (TAVI) in 2002 by transcatheter aortic valve implantation.

14  |  NOVEMBER 2015  www.nature.com/reviews

© 2015 Macmillan Publishers Limited. All rights reserved

regurgitation, the idea of interventional
treatment seems rational. Several devices
that imitate particular steps of the surgical
repair technique have been developed. The
most widely used device, the MitraClip®
(Evalve, USA), mimics the edge-to-edge
leaflet repair (Alfieri’s stitch). The device has
been shown to relieve symptoms,4 but the
long-term results of this technique have yet
to be evaluated. The next logical step is true
catheter-based mitral valve replacement,
and first-in-human experience has already
been reported. Even though the anatomical
challenges of mitral valve replacement are
much more difficult to overcome than those
with aortic valve implantation, the sheer
number of high-risk patients with mitral
valve disease seems to justify the effort of
perfecting these devices.
The tricuspid valve, which has long
been neglected, has received due attention
during the past decade. For many years,
minimal or mild tricuspid valve incompe-
tence was widely accepted to resolve, or at
least not worsen, after treatment of other
valves. How­ever, several studies published
in the past 10 years have revealed improved
long-term outcomes after concomitant tri­
cuspid valve repair, which has resulted in a
rising trend for tricuspid valve reconstruc-
tion for enlarged tricuspid annuli, even if
re­gurgitation is only minimal.5
Another development that has influenced
decision-making in valve surgery is the
concept of implanting a transcatheter valve
inside a previously implanted degenerated
bioprosthesis or ring after valve repair. 6
These valve-in-valve or valve-in-ring pro-
cedures have not only helped to prevent
high-risk patients from having to undergo
complex reoperations, but have also resulted
in an increased use of bio­prostheses during
A DECADE IN MEDICINE NOVEMBER 2015  |  15
the primary operation, often in younger
patients—a trend that will only rise in
the future.
The valve-in-valve concept has also been
applied to the pulmonary valve with the
availability of the Melody®Transcatheter
Pulmonary Valve (Medtronic, USA), which
was the first transcatheter valve available
on the market. This technology has been a
blessing for children and young adults, who
would otherwise have to undergo multiple
pulmonary valve reoperations.7
During the past decade, continued
efforts have been made to minimize surgi-
cal trauma without compromising clinical
outcomes, and to make minimally-invasi­ve
valve surgery a routine and low-risk oper­
ation. Simultaneously, the development of
catheter-based valve implantation has com-
plemented the therapeutic arsenal, thereby
expanding the indications to patients who
were formerly deemed inoperable. The rapid
development of TAVI systems and, more
importantly, the strong push to spread this
technology by the medical industry, inter-
ventional cardiologists, and some surgeons
has triggered competition between TAVI
and conventional surgical valve replace-
ment. With only two randomized studies to
compare TAVI and conventional surgery in
high-risk patients having been published to
date,8,9 further research and trials need to be
conducted in the near future, to determine
the most suitable therapy for each patient.
Unfortunately, TAVI is already being offered
to in­termediate-risk and low-risk patients
in the absence of scientific evidence. Care
needs to be exercised in advising patients
to undergo these procedures, at least until
long-term outcomes are known. Therefore,
the importance of large, independent
studies and surveys, such as the German
© 2015 Macmillan Publishers Limited. All rights reserved
CARDIOLOGY
Aortic Valve Registry (GARY),10 which will
provide 5‑year follow-up data on all aortic
valve interventions performed in Germany,
cannot be overemphasized. Nonetheless,
the fast progress in valve repair and replace-
ment technologies over the past decade is
extremely encouraging, and raises hopes for
further advances in the future.
Herzzentrum Leipzig GmbH–Universitaetsklinik,
Struempellstrasse 39, 04289 Leipzig, Germany.
friedrich.mohr@medizin.uni-leipzig.de
Competing interests
The author declares no competing interests.
1. Bavaria, J. E. et al. The St Jude Medical Trifecta
aortic pericardial valve: results from a global,
multicenter, prospective clinical study. J. Thorac.
Cardiovasc. Surg. 147, 590–597 (2014).
2. Cribier, A. et al. Percutaneous transcatheter
implantation of an aortic valve prosthesis for
calcific aortic stenosis: first human case
description. Circulation 106, 3006–3008 (2002).
3. Acker, M. A. et al. Mitral-valve repair versus
replacement for severe ischemic mitral
regurgitation. N. Engl. J. Med. 370, 23–32
(2014).
4. Glower, D. D. et al. Percutaneous mitral valve
repair for mitral regurgitation in high-risk
patients: results of the EVEREST II study.
J. Am. Coll. Cardiol. 64, 172–181 (2014).
5. Navia, J. L. et al. Moderate tricuspid
regurgitation with left-sided degenerative
heart valve disease: to repair or not to repair?
Ann. Thorac. Surg. 93, 59–67 (2012).
6. Dvir, D. et al. Transcatheter aortic valve
implantation in failed bioprosthetic surgical
valves. JAMA 312, 162–170 (2014).
7. Momenah, T. S. et al. Extended application of
percutaneous pulmonary valve implantation.
J. Am. Coll. Cardiol. 53, 1859–1863 (2009).
8. Smith, C. R. et al. Transcatheter versus surgical
aortic-valve replacement in high-risk patients.
N. Engl. J. Med. 364, 2187–2198 (2011).
9. Adams, D. H. et al. Transcatheter aortic-valve
replacement with a self-expanding prosthesis.
N. Engl. J. Med. 370, 1790–1798 (2014).
10. Mohr, F. W. et al. The German Aortic Valve
Registry: 1‑year results from 13 680 patients
with aortic valve disease. Eur. J. Cardiothorac.
Surg. http://dx.doi.org/10.1093/ejcts/ezu290.
Connect to...

Follow us on Twitter
For the latest commentary and cardiology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Cardiology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Cardiology

w w w.
Web collections
A selection of articles on a related theme, providing a panoramic view of a key area of cardiology

Posters
Visual overviews of important areas of cardiology, including dyslipidaemia and balloon angioplasty, which are free
to download

Conferences
You can meet the editors of Nature Reviews Cardiology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Cardiology!

November 2014 marked the 10th anniversary of the launch of Nature Reviews Cardiology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of cardiology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrcardio10

© 2015 Macmillan Publishers Limited. All rights reserved

 CLINICAL ONCOLOGY
DECADE IN REVIEW—CLINICAL TRIALS

Shifting paradigms in cancer clinical trial design

Daniel J. Sargent and Edward L. Korn
Over the past decade, there have been profound shifts in clinical trial design. Phase II randomized studies,
phase II/III and other adaptive designs, early surrogate end points, and prospective biomarker-based patient
selection have all increased in popularity. We discuss these shifts in clinical trial designs that have increased
efficiency in identifying which patients will benefit from specific treatments.
Sargent, D. J. & Korn, E. L. Nat. Rev. Clin. Oncol. 11, 625–626 (2014); published online 7 October 2014; doi:10.1038/nrclinonc.2014.167

The past decade has witnessed an expan- Phase II/III trials attempt to expedite
sion in knowledge and understanding of drug development by combining previ-
the biological basis of cancer, as a result ously distinct trials into a single protocol.
of ‘mega-projects’ such as the Human In the past decade, these trials have become
Gen­ome Project and The Cancer Genome both increasingly attractive and feasible.
Atlas, as well as continued scientific pursuit Key design issues face phase II/III trials:
by t­housands of individual researchers. This firstly, the choice of phase II end point; sec-
know­ledge has enabled the development ondly, whether to temporarily halt accrual
and advance­ment of rationally designed to enable maturation of the phase II data
targeted therapies in oncology, often (but before making the decision about proceed-
not always) accompanied by a predictive ing to phase III; and thirdly, the choice of
biomarker to identify the patients who phase II efficacy cut-off that informs this
iStock/Thinkstock

have the greatest likelihood of benefit. This
paradigm of targeted therapy in potentially
biomarker-­defined populations has required
evolution of the methods used to clinically
evaluate new therapies. For instance, when
a therapy demonstrates a profound and
unambiguous treatment benefit (such as
anaplastic lymphoma kinase [ALK] inhibi-
tion in ALK-positive metastatic non-small-
cell lung cancer), performing a randomized
trial versus a standard therapy might be
unnecessary.1 Unfortunately, most effec-
tive therapeutic agents are unlikely to show
such a dramatic benefit, and randomized
trials will continue to be the fundamental
approach to d­e veloping evidence-based
treatment paradigms.
most-promising of several experimental
regimens, which would subsequently be
evaluated in a controlled trial; rather than
randomized phase II studies, the activity
of a novel regimen was typically evalu-
ated in a single-arm trial with a tumour
response rate end point. Rubinstein et al.2
proposed expanding the role of randomiza-
tion in phase II trials to explicitly include
a control arm, with the intention of gener-
ating more-robust data to justify the large
resource expenditure of a phase III trial.
decision.4 Ideally, the phase II end point
should be an intermediate end point (for
example, PFS) available earlier than the
definitive phase III end point (for example,
overall survival). The stringency of the
efficacy cut-off determines the chance
of proceeding to an ultimately negative
phase III trial versus missing a potentially
positive phase III trial by stopping investi-
gation of a therapy too early based on the
phase II results.
The search for, and use of, predictive
biomarkers (that is, those that identify
the patients most likely to benefit from a
specific therapy) has become a dominant
theme in oncology drug development. In
the context of a single biomarker (such

‘‘
…a paradigm shift in
the design of phase II trials
has occurred
’’
Over the past decade, much work has
been conducted to improve randomized
clinical trial designs. For example, a para-
digm shift in the design of phase II trials has
occurred. Before 2004, randomized phase II
trials were relatively rare, and mainly used
a ‘pick the winner’ design to identify the
Randomization is particularly critical in the
setting of newly characterized, biomarker-
defined patient groups for which the natural
history of disease is poorly understood; for
new therapies with activity that might be
best assessed using a time-related end point,
such as progression-free survival (PFS); and
when novel drugs are being combined with
standard treatments known to have activity.
It has been suggested that randomization
in phase II trials should, in most circum-
stances, now become the expectation as
opposed to the exception.3
as a specific gene mutation in a tumour),
the major trial design choice is whether to
restrict eligibility to the biomarker-positive
population (enrichment design), or to ran-
domize all patients (or all patients who
have completed biomarker evaluation) and
then evaluate the treatment effect within
the biomarker-positive and biomarker-
negative subgroups separately. 5 The use
of an enrichment design potentially offers
greater trial efficiency, but at the cost of
not knowing how well the treatment would
have worked in the biomarker-­n egative

A DECADE IN MEDICINE NOVEMBER 2015  |  16

© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
subgroup, and whether the cost and incon-
venience of routine use of the biomarker
to select patients for treatment is justi-
fied. Thus, the choice of trial design (and
analysis strategy) should be dictated by the
strength of the pretrial evidence that any
treatment benefit would be restricted to
the biomarker-positive group. Continuing
improvements in biomarker assay technolo-
gies (for example, smaller sample require-
ments, the use of formalin-fixed instead
of frozen tissue, and assessments of blood
or molecular-imaging biomarkers) enable
more patients to be included in trials
i­ncorporating biomarker evaluations.
‘‘
…the next decade will be
faced … with the realization
that ‘every tumour is a
’’
rare tumour’
A number of trial designs have been
pro­posed for the situation when multiple
biomarkers are potentially associated with
therapeutic efficacy. Examples of explora­
tory trials in which the treatment is not (at
least initially) restricted by the biomarkers
are available. The phase II BATTLE trial6
randomly assigned patients with meta-
static non-small-cell lung cancer to four
treatments and subsequently evaluated
outcomes among five biomarker-defined
subgroups for each of these treatments. The
ongoing I‑SPY 2 trial is randomly assigning
patients with breast cancer between mul-
tiple investigational therapies after initial
stratification by standard FDA-approved
biomarkers, with efficacy then estimated
within patient groups with predefined bio-
marker profiles accompanied by post-hoc
analyses of a variety of additional biomark-
ers in an attempt to simultaneously speed
up biomarker and therapeutic discovery.
These trials demonstrate that obtaining
data on multiple biomarkers for patients
being treated with multiple treatments
during a trial is possible. ‘Signal-finding
trials’, in which the treatment is specifi­
cally directed by multiple biomarkers, are
also being undertaken. For example, the
phase II National Cancer Institute MATCH
trial will assign patients with disease pro-
gression after at least one standard therapy
for advanced and/or metastatic disease to
potentially any of 20–25 targeted agents
depending on the specific genomic muta-
tions detected in the patients’ tumours (if an
actionable mutation is present).8
17  |  NOVEMBER 2015  www.nature.com/reviews
‘Definitive trials’ with multiple biomark-
ers, in which patients are assigned to differ-
ent randomized phase III trials depending
on their biomarker profiles, are also being
performed. For example, the Lung Cancer
Master Protocol (Lung-MAP) trial is
screening patients with advanced-stage
squamous cell lung cancer and assigning
them to four different random­ized trials
of targeted agents depending upon the
genomic alterations in their tumours, or a
fifth randomized trial of an immunother-
apy (all versus appropriate conventional
chemotherapy). 8 Similar to many other
biomarker-based trials, Lung-MAP is
designed as a phase II/III trial. The ability
to have a uniform screening platform to
assign patients to different trials enhances
efficiency and enables more patients to be
included than if each trial screened sepa-
rately for its associated genomic alteration
(with patients without that alteration not
being included in any trial). Biomarker-
based trials are often adaptive, owing to the
ability to use the accumulating trial data to
make study modifications. Adaptive trials
are not new, as methods for formal interim
analyses date to the 1970s. The past decade,
however, has seen a rapid expansion of
the types and goals of adaption, including
defining subpopulations, adding or drop-
ping arms ‘on the fly’, altering the random-
ization ratio based on accumulating data,
and many other options. Adaptive trials
offer the possibility of obtaining clinical
answers more quickly.
Efficient clinical trial design, and in par-
ticular the ability to conduct phase II/III
or other adaptive trials, requires end points
that are rapidly accessible. Pursuit of such
early, or surrogate, end points is a major
current research focus. In oncology, the
vigor­ous debate between PFS and overall
survival as a phase  III trial end point
continues unabated, with the questions
including, but not limited to, whether PFS
is a valid surrogate for overall survival or
quality of life. The decision on which end
point to select is specific to the trial and
situ­ation. For phase II trials, an intermedi-
ate end point need not be a validated surro­
gate for the definitive end point because
the phase III evaluation will be performed
eventually using the definitive end point.
The meta-analytical approach, which
requires accurate prediction of within-trial
results for the true end point based on the
surrogate, has in the past decade been estab-
lished as the preferred method of surrogate
end point validation for assessing whether
© 2015 Macmillan Publishers Limited. All rights reserved
new therapies provide a group-level advan-
tage (as opposed to predicting how well an
i­ndividual patient will respond).9
Oncology trial design in the next decade
will be faced with the ever-increasing sub-
classification of tumours with the realiza-
tion that ‘every tumour is a rare tumour’.
One potential strategy, in addition to the
strategies mentioned earlier, is to ‘raise
the bar’ and demand greater efficacy from
new treatments.10 This strategy might be
possible owing to predictive biomarkers
and targeted therapies. Continued advances
in clinical trial design and analysis will be
required to meet these challenges and
to ensure rapid translation of biological
knowledge into clinical practice.
Department of Health Science Research,
Mayo Clinic, 200 First Street SW, Rochester,
MN 5590, USA (D.J.S.). Biometric Research
Branch, Division of Cancer Treatment and
Diagnosis, National Cancer Institute,
9609 Medical Center Drive, Room 5W112,
Bethesda, MD 20892‑9735, USA (E.L.K.).
Correspondence to: D.J.S.
sargent.daniel@mayo.edu
Competing interests
The authors declare no competing interests.
1. Sherman, R. E., Li, J., Shapley, S., Robb, M. &
Woodcock, J. Expediting drug development—the
FDA’s new “breakthrough therapy” designation.
N. Engl. J. Med. 369, 1877–1880 (2013).
2. Rubinstein, L. V. et al. Design issues of
randomized phase II trials and a proposal
for phase II screening trials. J. Clin. Oncol. 23,
7199–7206 (2005).
3. Seymour, L. et al. The design of phase II clinical
trials testing cancer therapeutics: consensus
recommendations from the clinical trial design
task force of the National Cancer Institute
Investigational Drug Steering Committee.
Clin. Cancer Res. 16, 1764–1769 (2010).
4. Korn, E. L., Freidlin, B., Abrams, J. S. &
Halabi, S. Design issues in randomized
phase II/III trials. J. Clin. Oncol. 30, 667–671
(2012).
5. Sargent, D. J., Conley, B. A., Allegra, C. &
Collette, L. Clinical trial designs for predictive
marker validation in cancer treatment trials.
J. Clin. Oncol. 23, 2020–2027 (2005).
6. Kim, E. S. et al. The BATTLE trial: personalizing
therapy for lung cancer. Cancer Discov. 1,
44–53 (2011).
7. Barker, A. D. et al. I‑SPY 2: an adaptive breast
cancer trial design in the setting of neoadjuvant
chemotherapy. Clin. Pharmacol. Ther. 86,
97–100 (2009).
8. Abrams, J. et al. National Cancer Institute’s
precision medicine initiatives for the new
National Clinical Trials Network. Am. Soc. Clin.
Oncol. Educ. Book 2014, 71–76 (2014).
9. Burzykowski, T., Molenberghs, G. & Buyse, M.
(eds) The Evaluation of Surrogate Endpoints
(Springer, 2005).
10. Ellis, L. M. et al. American Society of Clinical
Oncology perspective: raising the bar for
clinical trials by defining clinically meaningful
outcomes. J. Clin. Oncol. 32, 1277–1280
(2014).
 CLINICAL ONCOLOGY

DECADE IN REVIEW—TARGETED THERAPY

Successes, toxicities and

challenges in solid tumours
Joel W. Neal and George W. Sledge
The rise of targeted therapy for solid tumours over the past decade

Thinkstock/iStock
has yielded a cornucopia of novel agents across an array of cancers.
Amidst multiple acclaimed successes, targeted therapies are associated
with considerable toxicity, and durable responses are often thwarted by
genomic chaos driving the evolution of resistant clones; key examples
of successes in solid tumours are highlighted herein. required to maximize cell killing and, thus,
Neal, J. W. & Sledge, G. W. Nat. Rev. Clin. Oncol. 11, 627–628 (2014); published online 7 October 2014;
thwart the emergence of resistance. For
doi:10.1038/nrclinonc.2014.171 instance, in the case of melanoma, combin-
ing direct BRAF inhibition with blockade
The past decade has seen a major transi- small-molecule TKIs. One such drug, erloti­ of the downstream kinase MEK has been
tion in drug therapy for cancer, with a shift nib, was granted initial FDA approval in shown to improve the progression-free
from broad-spectrum cytotoxic agents to the second-line treatment of unselected survival of patients with metastatic dis­
highly targeted therapies. In the past few NSCLC in 2004 (Supplementary Table 1). ease (and received FDA approval in 2014;
years, FDA approvals of targeted therapies Interestingly, in early clinical trials, a sub- Supplementary Table 1), but this approach
for solid tumours have increased substan- group of patients experienced dramatic is associ­ated with increased adverse events
tially (Supplementary Table 1). Encouraged tumour shrinkage, leading to the identifica- and ‘fiscal toxicity’.4
by the success of the first FDA-approved tion of tumour-specific, sensitizing EGFR In NSCLC, the identification of an
small molecule tyrosine kinase inhibi- mutations that strongly predicted response uncom­mon tumour gene rearrangement,
tor (TKI), imati­nib, for the treatment of to erlotinib. Subsequently, the indications for EML4–ALK, led to clinical testing and rapid
chronic myeloid leukaemia, oncologists erlotinib therapy were expanded to include approval—in only 4 years—of crizotinib,
and patients hoped that targeted therapies first-line treatment—in lieu of chemo­therapy a drug designed as a hepatocyte growth
would prove to be highly effective in most —for EGFR-mutant NSCLC;2 neverthe­less, factor receptor (also known as MET) inhibi-
cancers and associ­ated with less toxi­city than as in many other tumour types, responses tor that has off-target activity against ALK.5
the cytotoxic chemo­therapeutic agents that to this targeted therapy are not durable, Repeat biopsies of tumours have revealed a
preceded them. With a decade of clinical with a simple secondary point mutation in variety of secondary resistance mutations in
research behind us, spurred on by several EGFR occurring as the dominant mecha- ALK, which could be overcome with newer-­
signal triumphs such as HER2-directed nism of acquired resistance in around half generation TKIs targeting this receptor. One
therapy in breast cancer, EGFR and ALK of these cancers. Importantly, the identi­ such second-generation ALK inhibitor, ceri-
inhibitor therapy in non-small-cell lung fication of such mutations also points tinib, received accelerated approval by the
cancer (NSCLC), antiangiogenic therapy to potential t­h erapeutic approaches to FDA in 2011 based on the demonstration
in kidney cancer, and BRAF-targeted overcome resistance. of promising efficacy in phase I/II trials.6

‘‘
therapy in metastatic melanoma, a more Other approaches to overcoming resistance
nuanced u­nderstanding of targeted therapy The past decade has seen a have been introduced in the past decade,
has emerged. and include the use of the initial targeted
major transition in drug therapy

A rapid expansion in the number of
antibody and small-molecule therapies
occurred over the past decade. These agents
were tested broadly for therapeutic efficacy,
often resulting in refinement of the bio-
logical mechanism of action after signals of
clinical benefit were observed. For example,
cetuximab, a chimeric monoclonal anti-
body that targets the extracellular portion
of EGFR, improved survival in colorectal
cancer (CRC), initially in an unselected
patient population; however, subsequent
analysis of tumour specimens from large
trials showed that the tumours with KRAS
mutations were insensitive to cetuximab.1
At the same time, a complementary strat-
egy was developed to inhibit the EGFR
—t­argeting the ATP-binding domain with
’’
for cancer…
The BRAF V600E mutation in melanoma
is another case in point; although relatively
common (approximately 50% of all mela-
nomas harbour this mutation), and readily
targetable with oral small-molecule TKIs in
the metastatic setting,3 in general, tumour
responses are of relatively brief duration and
are associated with only a modest improve-
ment in overall survival. Indeed, the rapid
emergence of drug resistance in this highly
mutated type of cancer is common, multi-
factorial, and occurs at an early time point
after treatment. Preclinical analyses have
suggested that targeting multiple drivers
of tumour growth simultaneously will be
agent as a delivery vehicle for nontargeted
drugs. Trastuzumab emtansine, a conju-
gate of the anti-HER2 monoclonal antibody
trastuzumab and the cytotoxic maytansine
derivative mertan­sine (T‑DM1), was the first
molecule using this strategy to be approved
by the FDA (Supplementary Table 1), and
has proven activity in trastuzumab-resistant
HER2‑positive breast cancer.7
A new hope for targeted therapy is in
preventing the development of disease
recurrence following curative treatment.
HER2-targeting with trastuzumab began
in the 1990s, but it was not until 2005 that
a series of adjuvant trials of this agent in
patients with breast cancer demonstrated a
dramatic reduction in the rates of recurrence
in HER2-positive early stage disease; with

A DECADE IN MEDICINE NOVEMBER 2015  |  18

© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
further follow-up, these reductions trans-
lated into durable improvements in overall
survival for this once-lethal breast cancer
subtype.8 Other attempts at translating ther-
apeutic benefit in the metastatic disease to
the adjuvant setting have been less successful,
with no benefit observed for the anti-EGFR
antibody cetuximab or anti-VEGF antibody
bevacizumab following complete resection
of CRC. However, in patients with imatinib-
sensitive gastro­intestinal stromal tumours,
both disease-free survival and overall sur-
vival were improved with imatinib treatment
after surgery.9 Intriguingly, prolonged 3‑year
treatment is superior to 1‑year regimens,
suggesting that targeted therapies might
exert a cytostatic rather than cytotoxic effect
against micrometastatic disease.9
Not all targeted therapies have yielded
such clear insight into the molecular basis
of tumour sensitivity and resistance. Anti­
angiogenic therapy with bevacizumab
found an initial role in the treatment of CRC
and NSCLC, but extensive searches in both
normal and tumour tissues for bio­markers
of therapeutic benefit have not identified
a straightforward predictive marker of
response. By contrast, renal-cell cancers,
which are generally chemoresistant, are sen-
sitive to a variety of antiangiogenic treatment
strategies, including bevacizumab and small-
molecule VEGF-receptor inhibitors such as
sorafenib, sunitinib, pazopanib, and axitinib.
Of these, sunitinib more than doubled the
median progression-free survival duration
in patients, compared with the historical
standard treatment of IFN‑α.10
With regard to toxicity, many targeted
thera­p ies are arguably associated with
important adverse events, although these
toxicities tend to be mechanistic in nature
(for example, skin and gastrointestinal toxi­
city for EGFR inhibitors and hypertension
for VEGF inhibitors) and nonmyelo­ablative.
Indeed, targeted therapies are generally
associated with increased incidence of high-
grade adverse events, treatment discon-
tinuation, and treatment-related fatalities
compared with the control interventions in
registration trials. Targeted therapies should
not, therefore, be expected to be less toxic
than the ‘nontargeted’ chemotherapeutics
that preceded them.
For tumour types with known effective
targeted therapies, the challenges over the
upcoming decade will be in both r­eactively
and proactively over­coming resistance to the
existing agents with next-­generation drugs,
and through combina­t orial therapeutic
approaches. Additionally, use of targeted
19  |  NOVEMBER 2015  www.nature.com/reviews
therapies in the adjuvant setting might delay
or prevent disease recurrence. In the next
decade, studies might also reveal additional
targetable tumour-drug dyads, but much of
the ‘low-hanging fruit’ has probably already
been identified; thus, in light of the limita-
tions of the current approaches, targeted
therapy might not be a panacea for the
majority of patients with cancer. Indeed,
geno­mic analyses point to a landscape domi-
nated by orphan diseases, with untargetable
or rare driver mutations being exceptionally
common across a broad spectrum of human
cancers, and with many cancers having
multiple drivers. For patients with these
cancers, hope lies in improving access to
approved and investigational targeted drugs
with plausible efficacy in their cancer, and in
new avenues of anticancer research, such as
immunotherapy, that might synergize with
chemotherapy and targeted therapy.
Division of Oncology, Stanford University
School of Medicine, 875 Blake Wilbur Drive,
Stanford, CA 94305‑5826, USA (J.W.N., G.W.S.).
Correspondence to: J.W.N.
jwneal@stanford.edu
Competing interests
J.W.N. has received grants for research support
from, ArQule, Boehringer-Ingelheim, Merck, and
Roche/Genentech. G.W.S. has received grants for
research support from Roche/Genentech and
honouraria for speaking from Genentech, is a
consultant for Symphogen, and is a member of
the Board of Directors for Syndax Pharmaceuticals.
DECADE IN REVIEW—HAEMATOLOGICAL CANCER
Advances in biology and therapy
S. Vincent Rajkumar and Philippe Moreau
It has been a decade of remarkable progress in the field of
haematological malignancies with the rapid translation of basic science
discoveries into effective targeted therapies. We discuss the most exciting
advances in this field, many of which have already produced meaningful
improvements in survival and quality of life of patients.
Rajkumar, S. V. & Moreau, P. Nat. Rev. Clin. Oncol. 11, 628–630 (2014); published online 14 October 2014;
doi:10.1038/nrclinonc.2014.172
In the 1990s, the median survival of mul-
tiple myeloma was 2–3 years. Nowadays, the
median survival estimates are approximately
7–10 years. In the early 2000s, the advent of
bortezomib and lenalidomide changed the
treatment of multiple myeloma in a radical
way. While bortezomib was developed with
a clear mechanistic goal in mind, lenalido-
mide had a serendipitous discovery that
promulgated its use in the treatment setting.
Bortezomib is a first-in-class proteasome
© 2015 Macmillan Publishers Limited. All rights reserved
1. Karapetis, C. S. et al. K‑Ras mutations and
benefit from cetuximab in advanced colorectal
cancer. N. Engl. J. Med. 359, 1757–1765
(2008).
2. Rosell, R. et al. Erlotinib versus standard
chemotherapy as first-line treatment for
European patients with advanced EGFR
mutation-positive non‑small‑cell lung cancer
(EURTAC): a multicentre, open-label,
randomised phase 3 trial. Lancet Oncol. 13,
239–246 (2012).
3. Chapman, P. B. et al. Improved survival with
vemurafenib in melanoma with BRAF V600E
mutation. N. Engl. J. Med. 364, 2507–2516
(2011).
4. Flaherty, K. T. et al. Combined BRAF and MEK
inhibition in melanoma with BRAF V600
mutations. N. Engl. J. Med. 367, 1694–1703
(2012).
5. Kwak, E. L. et al. Anaplastic lymphoma kinase
inhibition in non‑small‑cell lung cancer. N. Engl.
J. Med. 363, 1693–1703 (2010).
6. Shaw, A. T. et al. Ceritinib in ALK-rearranged
non‑small‑cell lung cancer. N. Engl. J. Med. 370,
1189–1197 (2014).
7. Verma, S. et al. Trastuzumab emtansine for
HER2-positive advanced breast cancer. N. Engl.
J. Med. 367, 1783–1791 (2012).
8. Romond, E. H. et al. Trastuzumab plus adjuvant
chemotherapy for operable HER2-positive
breast cancer. N. Engl. J. Med. 353,
1673–1684 (2005).
9. Joensuu, H. et al. One vs three years of
adjuvant imatinib for operable gastrointestinal
stromal tumor: a randomized trial. JAMA 307,
1265–1272 (2012).
10. Motzer, R. J. et al. Sunitinib versus interferon
alfa in metastatic renal-cell carcinoma. N. Engl.
J. Med. 356, 115–124 (2007).
Supplementary information is linked to the online
version of the paper at www.nature.com/nrclinonc.
inhibitor and functions by blocking the cel-
lular protein catabolism via the ubiquitin-
proteasome dependent pathway. Multiple
myeloma has been the first disease in which
inhibition of the proteasome was used as a
therapeutic target for the treatment of cancer.
The drug was initially tested in patients with
relapsed and refractory disease; however,
bortezomib has now become one of the
major options for the treatment of newly
diagnosed m­ultiple myeloma.1

Lenalidomide is an analogue of thalido­
mide that is more active and less toxic
than the parent drug. The original finding
that lenalidomide was remarkably effec-
tive in patients with relapsed myeloma
preceded the identification of its cellular
target, c­ereblon, by several years. Cereblon
is a component of E3 ubiquitin ligase, a
key enzyme in the ubiquitin–­proteasome
protein degradation pathway. The binding
of lenalidomide to cereblon enhances its
E3 ubiquitin ligase activity, which leads to
selective ubiquitination of two transcrip-
tion factors, lkaros and Aiolos. This causes
the degradation of these proteins by the
proteasome catalytic complex, and leads to
myeloma cell death.
As with bortezomib, the use of lenalido-
mide has now moved rapidly to the frontline
setting. In a recent landmark random­ized
trial, long-term therapy administered until
disease progression with lenalidomide
plus low-dose dexamethasone improved
progression-free survival (PFS) as well as
overall survival of patients with newly diag-
nosed multiple myeloma.2 In addition to
their direct role in prolonging the survival
of patients with myeloma, these drugs have
spawned the development of several agents
with a similar mechanism of action, includ-
ing the proteasome inhibitors carfilzomib,
ixazomib, marizomib, oprozomib and the
lenalidomide analogue pomalidomide.
‘‘ …clinical success with these
agents provides new hope for
patients…
’’
In patients with non-Hodgkin lym-
phoma, two key areas that were of great
therapeutic frustration saw important
advances. First, the need to prolong time
to relapse in patients with follicular lym-
phoma using an approach that does not
compromise quality of life was addressed.
A large randomized trial of 1,018 patients
undergoing mainten­a nce therapy with
rituximab administered every 2 months
for 2 years signifi­cantly prolonged PFS and
time to next chemotherapy; 3‑year PFS was
75% with rituximab maintenance versus
58% in the observation arm (P <0.0001).3
Nevertheless, relapse in patients with
indolent lymphomas is usually inevitable,
and eventually the disease is refractory to
available classes of agents. Thus, the second
area of interest was to identify active drugs
that are directed against novel targets to
help overcome this issue in patients with
A DECADE IN MEDICINE NOVEMBER 2015  |  20
20
200
5 04
06
20
200
200
7
2008 10
20
2012
201
3
2014
2 01
G
NP
relapsed and/or refractory disease. The
phosphatidylinositol 3‑kinase (PI3K) path­
way has an important role in cell prolif-
eration and survival. PI3Kδ is particu­larly
active in B‑cell cancers, including indolent
and non-Hodgkin lymphoma, and thus rep-
resents an attractive new target. In a phase I
trial in patients with relapsed indolent lym-
phoma, idelalisib—a specific inhibitor of
PI3Kδ—showed a striking response rate of
57% (71 out of 125 patients), including 6%
complete responses.4 These results identify
a new therapeutic target in B‑cell malignan-
cies, and led to the FDA regulatory approval
of idelalisib in the USA.
Further advances are sure to follow as we
improve our understanding of the precise
pathogenetic mechanisms involved in such
diseases, and identify ‘druggable’ targets. For
example, although translocations involving
MYC and the immunoglobulin genes have
been recognized for decades as the cyto­
genetic signature of Burkitt lymphoma, this
has not been translated into specific targeted
therapy. Of note, an interesting study has
reported new insights that might provide
new targets for Burkitt lymphoma. Using
high-throughput RNA sequencing and
RNA interference screening, Schmitz et al.5
© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
identified several new molecular abnormali-
ties in the majority of patients with sporadic
Burkitt lymphoma, including mutations of
the transcription factor TCF3 (E2A) or its
negative regulator ID3, as well as mutations
of CCND3, many of which can be targeted
for therapeutic benefit.
The treatment of Hodgkin lymphoma
has long represented one of the greatest
9 successes in oncology; however, a subset of
patients develop refractory or relapsed dis­
ease. These refractory or relapsed patients
usually fail to respond to multi-agent
combination chemotherapy and stem-cell
1 transplantation; therefore, new targets
are needed for their treatment. The CD30
antigen is overexpressed in tumour cells
in patients with Hodgkin lymphoma and
anaplastic large-cell lymphoma. However,
previous attempts to target CD30 using
monoclonal antibodies were not successful.
Brentuximab vedotin is an antibody conju-
gate in which monomethyl auristatin E, an
antitubulin agent, is attached to a CD30
monoclonal antibody. In a phase I trial,
treatment with brentuximab vedotin res­
ulted in 13 complete responders among
45 patients with relapsed and/or refractory
Hodgkin lymphoma or anaplastic large-
cell lymphoma, a remarkable result that
p­rovides hope for these patients.6
In patients with chronic lymphocytic leu-
kaemia (CLL), the main advance is ibruti-
nib. CLL, similar to indolent non-­Hodgkin
lymphoma and multiple myeloma, is
character­ized by almost inevitable relapse.
With each relapse, remissions are harder
to achieve and less durable. In fact, until
a definitive cure is found, the develop-
ments of new agents will continue to be
paramount. The hope for patients in these
situations lies in the continued identifica-
tion of such treatments. Ibrutinib is an
oral inhibitor of Bruton tyrosine kinase
(BTK), an enzyme critical for CLL cell
survival. In a phase I/II trial of 85 patients
with advanced-stage CLL, approximately
70% responded to therapy, and the PFS
rate at 2 years was over 75%. 7 Ibrutinib
is also active in mantle-cell lymphoma,
Waldenstrom m­a croglobulinaemia, and
other lymphomas.
Another major practice-changing dis-
covery in the field of haematological
malignancies came from a randomized
trial of all-trans retinoic acid (ATRA) plus
arsenic trioxide versus ATRA plus chemo-
therapy in patients with acute promyelo-
cytic leukaemia. 8 The trial was designed
to test whether a regimen without any
CLINICAL ONCOLOGY
standard chemotherapy drugs could effec-
tively treat this disease. Complete remis-
sions were achieved in all 77 patients with
the ATRA–arsenic trioxide combination
versus 75 (out of 79) patients in the ATRA–­
chemotherapy arm. The 2‑year event-free
survival rates were 97% and 86%, respec-
tively. Over­all survival was superior with
ATRA–arsenic trioxide; specifically the 2‑
year overall survival was 99% in the ATRA–
arsenic trioxide arm versus 91% in the
ATRA–c­hemotherapy arm (P = 0.02).
Chronic myeloid disorders have also had
their share of success in the past 10 years.
The first drug ever to prolong overall sur-
vival in myelodysplastic syndrome (MDS)
was reported in 2009. In a randomized trial,
azacitidine was found to improve overall
survival compared with conventional care
selected by investigators before random-
ization: 24.5 months versus 15.0 months,
respectively (P = 0.0001).9
One of the biggest advances of the past
decade in haematology, was the finding
that mutations in Janus kinase 2 (JAK2) are
associ­ated with almost all cases of poly­
cythemia vera. The BCR–ABL transloca­tion,
known as the Philadelphia transloca­tion,
separates chronic myelo­genous leukaemia
from other myeloproliferative disorders;
the ability to target this translocation using
imatinib rocked the field in 2001. However,
until the discovery of the JAK2 mutation,
the pathogenetic basis of the remain-
ing myeloprolifer­ative disorders was not
known. The JAK2 mutation is a valine-to-
phenylalanine substitution at amino acid
position 617 (JAK2 V617F) that results in
the constitutive activation of JAK2. The
JAK2 V617F substitution induced poly-
cythemia in a mouse model, and has been
reported in polycythemia vera, but not
second­ary polycythemia.10 It is, however,
not specific for polycythemia, and has been
detected in other myeloprolifer­ative dis­
orders, such as essential thrombo­cythemia
and myelofibrosis. These findings led the
way for the development of JAK2 inhibitors,
and clinical success with these agents pro-
vides new hope for patients with a variety of
myeloid disorders.
The advances discussed above have
resulted in meaningful clinical benefit to
patients affected by different lymphoid and
myeloid malignancies. The improved under-
standing of the molecular patho­genesis
gained in the past 10 years will certainly
translate into a major expansion of tar-
geted therapy for numerous h­aematological
cancers in the next decade.
21  |  NOVEMBER 2015  www.nature.com/reviews
Division of Haematology, Mayo Clinic, 200 First
Street SW, Rochester, MN 55905, USA (S.V.R.).
Department of Haematology, University Hospital
Hôtel-Dieu, 44093 Nantes, France (P.M.).
Correspondence to: S.V.R.
rajkumar.vincent@mayo.edu
Competing interests
P.M. has served on advisory boards and received
honouraria from Celgene, Janssen and Millennium.
S.V.R. declares no competing interests.
1. San Miguel, J. F. et al. Bortezomib plus
melphalan and prednisone for initial treatment
of multiple myeloma. N. Engl. J. Med. 359,
906–917 (2008).
2. Benboubker, L. et al. Lenalidomide and
dexamethasone in transplant-ineligible patients
with myeloma. N. Engl. J. Med. 371, 906–917
(2014).
3. Salles, G. et al. Rituximab maintenance for
2 years in patients with high tumour burden
follicular lymphoma responding to rituximab plus
chemotherapy (PRIMA): a phase 3, randomised
controlled trial. Lancet 377, 42–51 (2011).
DECADE IN REVIEW—CANCER IMMUNOTHERAPY
Entering the mainstream
of cancer treatment
Steven A. Rosenberg
By November 2004, when the first issue of Nature Reviews Clinical
Oncology was published, cancer immunotherapy had been successfully
applied to the treatment of selected human cancers; however, dramatic
progress in the following decade has moved immunotherapy from the
sidelines of cancer treatment into the mainstream of modern oncology.
Rosenberg, S. A. Nat. Rev. Clin. Oncol. 11, 630–632 (2014); published online 14 October 2014;
doi:10.1038/nrclinonc.2014.174
Administration of IL‑2, the first immuno­
therapy capable of mediating complete,
durable cancer responses was originally
described in 1985 and approved by the
FDA for the treatment of patients with renal
cancer and melanoma in 1992 and 1998,
respectively. In 2003, the first evidence
that the administration of an antibody
targeting the inhibitory T‑cell costimula-
tory molecule CTLA‑4, (often referred to
as a checkpoint modulator) could mediate
cancer regression in patients with metastatic
melanoma was published. Despite extensive
efforts to develop therapeutic cancer vac-
cines, prior to 2004 none had been shown
to be effective.
The first effective adoptive cell-transfer
(ACT) immunotherapy for cancer was
reported in 1988 and showed that cultured
tumour infiltrating lymphocytes (TIL) iso-
lated from resected metastatic melanomas
and administered to the autologous patients
© 2015 Macmillan Publishers Limited. All rights reserved
4. Gopal, A. K. et al. PI3Kδ Inhibition by idelalisib
in patients with relapsed indolent lymphoma.
N. Engl. J. Med. 370, 1008–1018 (2014).
5. Schmitz, R. et al. Burkitt lymphoma
pathogenesis and therapeutic targets from
structural and functional genomics. Nature
490, 116–20 (2012).
6. Younes, A. et al. Brentuximab vedotin (SGN-35)
for relapsed CD30-positive lymphomas. N. Engl.
J. Med. 363, 1812–1821 9(2010).
7. Byrd, J. C. et al. Targeting BTK with ibrutinib in
relapsed chronic lymphocytic leukemia. N. Engl.
J. Med. 369, 32–42 (2013).
8. Lo-Coco, F. et al. Retinoic acid and arsenic
trioxide for acute promyelocytic leukemia.
N. Engl. J. Med. 369, 111–121 (2013).
9. Fenaux, P. et al. Efficacy of azacitidine
compared with that of conventional care
regimens in the treatment of higher-risk
myelodysplastic syndromes: a randomised,
open-label, phase III study. Lancet Oncol. 10,
223–232 (2009).
10. James, C. et al. A unique clonal JAK2 mutation
leading to constitutive signalling causes
polycythaemia vera. Nature 434, 1144–1148
(2005).
could mediate objective cancer regres-
sions (Figure 1). The effectiveness of this
T‑cell therapy was dramatically improved
in 2002 with the discovery that lympho­
depletion prior to T‑cell administration
could improve the antitumour activity of
the transferred cells. Thus by 2004, there
were clear indications that either direct
stimulation of T cells, blockade of check-
point modulators or the administration of
antitumour T cells could mediate regression
of invasive, vascularized cancers in humans.
In the subsequent decade, increas-
ing information about the importance of
cell-surface inhibitory molecules present
on lymphocytes led to striking advances
in immunotherapy, and antibodies that
could block these inhibitory receptors
were administered to patients. The first
ran­d omized protocol to demonstrate a
survival benefit for a cancer immunother-
apy was published by Hodi et al.1 in 2010.

In this study, patients with metastatic mela-
noma were randomly assigned to receive
a peptide vaccine alone or in combination
with anti-CTLA‑4 antibody (ipilimumab)
therapy.1 This trial was controversial, as it
was already known that ipilimumab could
mediate objective (and sometimes com-
plete) responses in patients with melanoma,
whereas peptide vaccines alone had virtu-
ally no evidence of efficacy, and no cross-
over between cohorts was allowed. In the
vaccine plus ipilimumab arm, 35 of 540
patients (6.5%) experienced an objective
response compared with two of 136 patients
(1.5%) treated with peptide alone.1 Of note,
the patients who received ipilimumab
showed an increase in median survival from
6.5 months to 10 months. 1 These results
ultimately led to the approval of ipilimumab
for the treatment of patients with metastatic
melanoma by the FDA in March 2011. The
clinical value of ipilimumab in other cancer
types (with the possible e­xception of renal
cancer) is currently unknown.
‘‘
…has moved immunotherapy
from the sidelines of cancer
treatment into the mainstream
’’
of modern oncology
Clinical utility of the blockade of a
second inhibitory lymphocyte receptor
programmed cell death 1 (PD‑1) was antici-
pated in 2010 and firmly demonstrated in
2012 in a successive trial that used this agent
in 296 patients with different cancer types.2
Among 236 evaluable patients, objective
response rates were seen in 18%, 28% and
27% of patients with non-small-cell lung
cancer, melanoma and renal cell cancer,
respectively.2 No objective responses were
observed in patients with colorectal or
prostate cancers.2 Response rates seemed
to be increased when PD‑L1, the ligand for
PD‑1, was expressed on the cancer cells,
and the use of a blocking antibody against
PD‑L1 resulted in responses in nine of 52
(17%) patients with melanoma, two of 17
(12%) with renal cancer, five of 49 (10%)
with non-small-cell lung cancer and one of
17 (6%) patients with ovarian cancer.3 The
impressive responses seen with pembroli-
zumab, an anti-PD‑1 antibody, in multi-
centre phase II studies led to approval of
this agent by the FDA in September 2014
for the treatment of patients with metastatic
melanoma refractory to ipilimumab and to
a BRAF inhibitor, if the tumour was BRAF
A DECADE IN MEDICINE NOVEMBER 2015  |  22
TIL isolation
Tumour
Figure 1 | Adoptive cell transfer immunotherapy. Cells are obtained from a resected tumour
(autologous TILs) or using peripheral lymphocytes genetically engineered to express antitumour
T-cell receptors, expanded in vitro and infused to patients after they have received a preparative
lymphodepleting regimen. Abbreviation: TIL, tumour-infiltrating lymphocyte.
V600-mutation-positive. Evaluation of the
blockade of other inhibitory receptors such
as TIM‑3 or LAG‑3 is ongoing.
Very little progress has been made in the
past 10 years concerning the application
of therapeutic cancer vaccines. The only
positive randomized trial of a therapeutic
vaccine was published by Kantoff et al.4 in
July 2010. Patients with hormone-­refractory
prostate cancer received the administration
of autologous peripheral blood mono­nuclear
cells activated ex vivo with a recombinant
protein consisting of prostatic acid phos-
phatase fused to granulocyte-­macrophage
colony stimulating factor (sipuleucel‑T
vaccine) or placebo. 4 Median survival
improved from 21.7 months in the placebo
cohort to 25.8 months in the vaccine group.4
Such increase in median survival was sur-
prising as only one of the 341 patients
(0.3%) treated with the vaccine had an
objective response, and only 2.6% of patients
had a decrease in serum prostate-­specific
antigen (PSA) levels;4 the time to disease
progression was similar in the two groups
(14.4 weeks compared to 14.6 weeks).4 On
the basis of these results, the s­ipuleucel‑T
vaccine was approved by the FDA in 2010.
So far, this modest clinical result remains
the only trial to show a convincing benefit
of an active i­m munization approach to
cancer treatment.
The past 10 years have seen consider-
able progress in the development of ACT
immuno­t herapies using both naturally
occurring and genetically engineered
lympho­cytes. Long-term follow-up studies
showing that ACT immunotherapy using
autologous TIL could mediate durable
complete (and likely curative) responses in
patients with refractory metastatic mela-
noma was published in 2011.5 In a series of
© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
Cell infusion Preconditioning:
plus IL-2 chemotherapy
three sequential trials that used increasing
lymphodepleting regimens, objective cancer
regressions were seen in 52 of 93 patients
(56%), including 20 patients (22%) with
complete regressions. 19 of the 20 patients
had ongoing complete responses beyond
5 years, with the longest ongoing responses
persisting beyond 10 years.5 These results
represented the highest reported response
rates of any clinical trials in patients with
metastatic melanoma. Furthermore, Bollard
et al.6 reported responses in patients with
Epstein–Barr virus (EBV)-related Hodgkin
or non-Hodgkin lymphomas receiving
autologous cytotoxic T cells that target the
EBV latent membrane proteins LMP2 or
LMP1. Specifically, 13 of 21 patients with
measurable relapsed or resistant lympho-
mas exhibited clinical responses, including
11 complete responses.6
The difficulty in identifying cells with
in vitro antitumour activity against many
cancer types led to extensive efforts to
genetically engineer T cells for use in ACT
therapy. Retroviruses encoding either con-
ventional αβT‑cell receptors (TCR) or chi-
meric antigen receptors (CAR; composed
of a single chain antibody fused to intra-
cellular signalling chains) are capable of
inserting genes into the genome of human
lymphocytes with efficiencies exceeding
80%. The first successful use of genetically
engineered lymphocytes to mediate cancer
regression was reported in 2006. Morgan
et al.7 showed that retroviral transduction
of autologous lymphocytes with the gene
encoding a TCR against the melanocyte/
melanoma differentiation antigen, MART‑1,
mediated objective cancer regressions in
two of 15 (13%) patients with metastatic
melanoma. Objective cancer regressions
were later reported in 30% of 20 patients
CLINICAL ONCOLOGY
‘‘ The past 10 years have
seen considerable progress
in the development of ACT
’’
immunotherapies…
receiving gene-modified cells targeting the
MART‑1 antigen and in 19% of 16 patients
who received autologous cells transduced
with a mouse TCR that recognized the
gp100 antigen, demonstrating for the first
time that adoptive transfer of genetically
modified cells could mediate cancer regres-
sion. In 2008, Pule et al.8 reported that EBV-
specific autologous T cells transduced with a
CAR targeting the GD2 molecule mediated
objective tumour regressions in two of eight
patients with measurable neuro­blastoma.
In 2010, the first report of the successful
use of CAR-transduced T cells targeting
the CD19 molecule in patients with B‑cell
lymphomas was reported.9 A heavily pre-
treated patient with advanced lymphoma
exhibited a dramatic response after receiv-
ing two cycles of autologous anti-CD19
CAR-transduced T  cells and remained
progression-­free 5 years later.9 Additional
reports confirmed the ability of these anti-
CD19 CAR-transduced T cells to mediate
the regression of both indolent as well as
aggressive large B-cell lymphomas, and
acute and chronic lymphoblastic leukaemias.
These early studies using gene-­modified
cells have greatly expanded the reach of ACT
therapy to different tumour types.
The greatest obstacle to the further
development of cancer immunotherapy
is the identification of target molecules
expressed on cancer cells and not on essen-
tial human tissues. Immunogenic mutations
expressed in individual cancers could be the
ideal immunotherapy target. In 2014, Tran
et al.10 reported a general technique for the
identification of lymphocytes capable of
targeting the rare but unique immunogenic
mutations present in individual epithelial
cancers and used this ACT approach to
mediate objective regression in a patient
with metastatic cholangiocarcinoma. The
intrinsic ability of the immune system to
specifically target the unique mutations
expressed by common cancers is likely to
be the key to further progress in the field of
cancer immunotherapy.
National Cancer Institute, 10 Center Drive
MSC 1201, Bethesda, MD 20892, USA.
sar@nih.gov
Competing interests
The author declares no competing interests.
23  |  NOVEMBER 2015  www.nature.com/reviews
1.
2.
Hodi, F. S. et al. Improved survival with
ipilimumab in patients with metastatic
melanoma. N. Engl. J. Med. 363, 711–723
(2010).
Topalian, S. L. et al. Safety, activity, and immune
autologous cytotoxic T lymphocytes targeting
Epstein–Barr virus latent membrane proteins.
J. Clin. Oncol. 32, 798–808 (2012).
7. Morgan, R. A. et al. Cancer regression in
patients after transfer of genetically
correlates of anti‑PD‑1 antibody in cancer. engineered lymphocytes. Science 314,
N. Engl. J. Med. 366, 2443–2454 (2012). 126–129 (2013).
3. Brahmer, J. R. et al. Safety and activity of 8. Pule, M. A. et al. Virus-specific T cells
anti‑PD‑L1 antibody in patients with advanced engineered to coexpress tumor-specific
cancer. N. Engl. J. Med. 366, 2455–2465 receptors: persistence and antitumor activity
(2012). in individuals with neuroblastoma. Nat. Med.
4. Kantoff, P. W. et al. Sipuleucel‑T immunotherapy 14, 1264–1270 (2008).
for castration-resistant prostate cancer. N. Engl. 9. Kochenderfer, J. N. et al. Eradication of
J. Med. 363, 422 (2010). B‑lineage cells and regression of lymphoma
5. Rosenberg, S. A. et al. Durable complete in a patient treated with autologous T cells
responses in heavily pretreated patients genetically engineered to recognize CD19.
with metastatic melanoma using T‑cell transfer Blood 116, 4099–4102 (2010).
immunotherapy. Clin. Cancer Res. 17, 10. Tran, E. et al. Cancer immunotherapy based
4550–4557 (2011). on mutation-specific CD4+ T cells in a patient
6. Bollard, C. M. et al. Sustained complete with epithelial cancer. Science 344, 641–645
responses in patients with lymphoma receiving (2014).
DECADE IN REVIEW—GENOMICS
A decade of discovery in cancer
genomics
Kenneth Offit
Over the past decade, genetic testing for rare inherited mutations, such
as BRCA1 and BRCA2 mutations, has been successfully incorporated
into clinical practice. Next-generation sequencing of cancer-susceptibility
genes and entire tumour genomes has transformed cancer care and
prevention. The discoveries of new cancer syndromes have raised exciting
opportunities and potential liabilities for cancer-care providers seeking to
incorporate genomic approaches into preventive oncology practice.
Offit, K. Nat. Rev. Clin. Oncol. 11, 632–624 (2014); published online 7 October 2014;
doi:10.1038/nrclinonc.2014.170
The past decade has witnessed the incor- on the practice of preventive oncology.
poration of genetic testing for cancer- The incorporation of genetic testing for
­susceptibility syndromes into the evidence-­ BRCA mutations in breast cancer marked
based practice of oncology, and the emer- one of the first applications of ‘personal-
gence of ‘next-generation’ genome scans ized’ genomics in medicine, and enabled
for cancer-risk loci. Herein, I discuss a ‘targeted’ cancer screening, prevention
series of seminal papers published over and, in some cases, the ability to personal­
the past decade that described new cancer ize therapies according to the patient’s
syndromes, but also raised new challenges genetic lansdcape.1 The translation of BRCA
related to informed consent, incidental testing to clinical practice was highlighted
findings, and the management of genetic by Domchek and colleagues2 who showed
variants of unknown s­ignificance or that preventive surgery of the ovaries over
unproven clinical actionability. a 34-year period decreased mortality in a
In the 1980s and 1990s, rare but highly- cohort of 2,482 women with BRCA1 or
penetrant cancer-predisposition genes were BRCA2 mutations; compared with women
identified by studying cancer-prone families who did not have salpingo-oophorectomy,
that demonstrate Mendelian inheritance of women who underwent this procedure had
cancer susceptibility. These studies impli- a 60% decrease in all-cause death rates,
cated genes, such as BRCA1 and BRCA2, the driven by lower mortality associated with
DNA-mismatch-repair genes (relevant for both breast and ovarian cancer. 2 In this
colon cancer), TP53 in Li–Fraumeni syn- study, the subset of women found to have
drome, and APC in familial adenomatous occult microscopic ovarian cancer at the
polyposis. The genetic basis of these and time of ‘preventive’ surgery were excluded
other syndromes had a powerful impact from analysis.2 During subsequent years,
© 2015 Macmillan Publishers Limited. All rights reserved

risk-reducing ovarian surgery, along
with breast MRI, the option of prophy­
lactic breast surgery and hormonal chemo-
prevention, became standard practice in
p­reventive oncology.1
‘‘
…oncologists will soon be
screening the inherited genomes
’’
of all patients with cancer
In the past decade it had become obvious
that highly penetrant cancer genes (such
as BRCA1/2 and MSH2) did not account
for the bulk of familial risk of the common
hereditary cancers. A debate ensued regard-
ing whether there were many common low-
risk genetic variants or undiscovered rare
high-risk variants, which would explain the
‘missing heritability’ of cancer. A pivotal
paper tested the ‘common variant’ hypoth-
esis using the emerging techno­logy of ‘gene
chips’ to assess hundreds of thousands of
single nucleotide polymorphisms (SNPs).3
In a two-stage design, 227,876 SNPs were
assessed in 4,398  breast-cancer cases
and 4,316 controls, identifying 30 SNPs
of interest that were further analy­s ed in
21,860 cases and 22,578 controls.3 The SNP
that emerged as the best ‘hit’, which was
proximal to the gene FGFR2, had a relative
risk of around 1.2‑times the baseline risk,
compared with BRCA1 that elevated risk of
early onset breast cancer by up to 40‑fold.3
Subsequent genome-wide association
studies of other cancer types identified hun-
dreds of hits near potentially causal genes,
which were all statistically significant, but
none of a magnitude to influence preven-
tive management in the clinic.4 A possi­
ble exception to this lack of clinical utility
emerged from studies we performed as part
of an international consortium investi­gating
modifiers of risk in the carriers of BRCA
mutations. In studies involving tens of thou-
sands of BRCA1/2 mutation carriers world-
wide, panels of risk-associated SNPs could
partition breast cancer risk from 20% up to
100% in BRCA-mutation carriers.5 These
findings will likely mark the first applica-
tion of SNP-based risk profiling to inform
clinical management of individuals with
hereditary risk of a common cancer.
Over the second half of the past decade,
a shift to identifying rare genomic vari-
ants was made possible by the emergence
of next-generation sequencing (NGS)
approaches. NGS involves a series of repeat-
ing sequencing reactions, performed and
detected automatically, with the production
A DECADE IN MEDICINE NOVEMBER 2015  |  24
of thousands to millions of simultaneous
sequence reads. An immediate and obvious
application of NGS was to sequence several
genes at the same time. A technological
tour de force prefigured the current era
in ‘cancer panel’ testing. Using targeted
capture and massively parallel genomic
sequencing, a group at the University of
Washington screened 21 candidate genes in
360 women with ovarian cancer.6 Strikingly,
24% of these women carried germline loss-
of-function mutations in genes such as
BRCA1, BRCA2, BARD1, BRIP1, CHEK2,
MRE11A, MSH6, NBN, PALB2, RAD50,
RAD51C, and TP53.6 Fuelled by this techno­
logical innovation, plus the equally impact-
ful loss of patent protection for BRCA1
and BRCA2 sequence analysis, a plethora
of commercial cancer panels flooded the
oncology marketplace.
At the same time, NGS technologies were
rapidly applied to studying unexplained
familial cancer clusters. Over the past
5 years, whole-exome sequencing (WES)
and whole-genome sequencing (WGS) has
resulted in a renaissance in the discovery
of new syndromes of cancer susceptibility
(Box 1). One of the early applications of this
technology came from a group at the Johns
Hopkins University, who applied WES of
20,661 coding genes in a single case of famil-
ial pancreatic cancer.7 Of 15,461 germline
variants not found in the reference human
genome, a deletion of four base pairs within
the PALB2 gene was discovered and tested
as a pancreatic-cancer-susceptibility gene.7
Despite this early report, we and others have
failed to confirm PALB2 as a major factor in
hereditary breast–pancreas-cancer families;
however, PALB2 maintained its status as a
rare breast-cancer-susceptibility gene.
Another example of a new syndrome with
a striking phenotype was described by Testa
and colleagues in 2011,8 on the basis of their
observation of gene clustering of meso-
theliomas and melanomas. Using exome
sequencing strategies, germline muta-
tions were discovered in the gene encod-
ing BRCA1‑associated protein‑1 (BAP1) in
two families with multiple cases of meso­
thelioma, and in some cases of uveal mela-
noma.8 These findings built on the earlier
observation of inherited germline BAP1
mutations in uveal and cutaneous melano-
cytic tumours. Remark­ably, this syndrome
was extended by other groups to include
renal-cell cancers in rare families.
In some cases the ‘new’ familial cancer
types studied were not rare. For example,
we studied families with acute lymphoblastic
© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
Box 1 | Cancer susceptibility syndromes*
Familial pancreatic cancer
PALB2 identified by exome sequencing; ATM
identified by exome sequencing and WGS
Familial ovarian cancer
BRIP1 identified by WGS
Familial pheochromocytoma
MAX identified through exome sequencing
Acute myelogenous leukaemia
(with Emberger syndrome)
GATA2 identified by exome sequencing
Familial Hodgkin lymphoma
NPAT identified by exome sequencing
Familial pre‑B-cell acute lymphoblastic
leukaemia
PAX5 identified by exome sequencing
Familial melanoma
MITF identified by WGS; TERT identified
by targeted sequencing
Familial mesothelioma, melanoma
and renal-cell cancer
BAP1 identified through exome and targeted
sequencing
Hereditary mixed polyposis syndrome
(HMPS)
GREM1 identified by targeted sequencing
Colorectal adenomas and colon cancer
POLE and POLD1 identified by WGS
Familial breast cancer
XRCC2 and FAN1 identified by exome
sequencing; PPM1D (mosaic) by targeted
sequencing
*Discovered recently by NGS.4 Abbreviations: NGS,
next‑generation sequencing; WGS, whole-genome
sequencing.
leukaemia, the most-common malignancy
of childhood, and identified a mutation in
a lymphoid-associated transcription factor,
PAX5, in two such families,9 with a third
Israeli family more recently reported to
harbour the same mutation. These ‘new’
cancer syndromes have redefined our notion
of inherited cancer (Box 1).
Despite these advances over the past
decade, clinical interventions for these syn-
dromes remain relatively rudimentary, and
the ethical implications of these discover-
ies remain daunting. Risk reduction for
the adult-cancer syndromes includes organ
removal surgeries.1 True genetic preven-
tion using assisted reproductive techno­
logies is an option oncologists should
remember to discuss with their younger
patients, or patient’s families, taking into
account ethical or religious considerations.
A broader ethical debate has emerged
regarding the extent to which incidental, or
secondary genetic findings, termed the ‘inci-
dentalome’, should be disclosed to patients.
Particularly challenging for oncologists
CLINICAL ONCOLOGY
are the unexpec­ted results of NGS analysis
of tumour and normal pairs, which might
include identification of genetic predisposi-
tions to noncancer-related diseases, such as
cardiac or neurological diseases.10 A vigor-
ous discussion is in progress regarding the
potential obligations of physi­cians to inform
i­ndividuals of incidental genetic findings.
At the same time there have been recent
calls for population-based screening—for
example, BRCA testing of all 30‑year‑old
women worldwide. Although such requests
by laboratory-based scientists have the best
intentions, they overlook a more-pressing
clinical reality: oncologists will soon be
screening the inherited genomes of all
patients with cancer. In both scenarios,
population testing of healthy individuals
and tumour–normal screening in patients
with tumours, we must recognize what
has been learnt over the past decade: not
all individuals wish to know all genomic
information; risks might reflect both popu­
lation heterogeneity and differences in
penetrance; and not all genomic informa-
tion is clinically actionable. Oncology has
become the ‘ground zero’ for a tectonic shift
in paradigms regarding personalized medi-
cine, both for targeted treatment as well as
p­revention based on genomic profiles.
Department of Medicine, Clinical Genetics
Service, Program in Cancer Biology and
Genetics, Sloan-Kettering Institute, Memorial
Sloan Kettering Cancer Center, Box 192,
1275 York Avenue, New York, NY 10021, USA.
Department of Medicine and Public Health,
Weill Cornell Medical College,
Cornell University, 445 East 69th Street,
New York, NY 10021, USA.
offitk@mskcc.org
Acknowledgements
K.O. would like to acknowledge funding support from
The Robert and Kate Niehaus Clinical Cancer
Genetics Research Initiative, and The Breast Cancer
Research Foundation.
Competing interests
The author declares no competing interests.
1. Couch, F. J., Nathanson, K. L. & Offit, K.
Two decades after BRCA: setting paradigms
in personalized cancer care and prevention.
Science 343, 1466–1470 (2014).
2. Domchek, S. M. et al. Association of risk-
reducing surgery in BRCA1 or BRCA2 mutation
carriers with cancer risk and mortality. JAMA
304, 967–975 (2010).
3. Easton, D. F. et al. Genome-wide association
study identifies novel breast cancer
susceptibility loci. Nature 447, 1087–1093
(2007).
4. Stadler, K., Schrader, K. A., Vijai, J.,
Robson, M. E. & Offit, K. Cancer genomics
and inherited risk. J. Clin. Oncol. 32, 687–698
(2014).
25  |  NOVEMBER 2015  www.nature.com/reviews
5. Gaudet, M. M. et al. Identification of a BRCA2-
specific modifier locus at 6p24 related to
breast cancer risk. PLoS Genet. 9, e1003173
(2013).
6. Walsh, T. et al. Mutations in 12 genes for
inherited ovarian, fallopian tube, and peritoneal
carcinoma identified by massively parallel
sequencing. Proc. Natl Acad. Sci. USA 108,
18032–18037 (2011).
7. Jones, S. et al. Exomic sequencing identifies
PALB2 as a pancreatic cancer susceptibility
gene. Science 324, 217 (2009).
DECADE IN REVIEW—FUNDING IN CANCER RESEARCH
National Cancer Institute awards
—a work in progress
Tito Fojo and Paraskevi Giannakakou
Over the past decade, funding for cancer research by the
US government—and others—has stagnated, while the demand for
investment has grown because of the increasing cancer incidence
worldwide. We discuss how National Cancer Institute funding efforts
have developed during this period, and the contemporary and future
impact of these measures on cancer research in the USA.
Fojo, T. & Giannakakou, P. Nat. Rev. Clin. Oncol. 11, 634–636 (2014); published online 14 October 2014;
doi:10.1038/nrclinonc.2014.173
The USA has been at the forefront of
cancer research with regard to spending,
and this investment has benefited not only
US citizens, but also the wider global com-
munity. Although dwarfed by the money
invested by pharmaceutical companies and
a diverse group of foundations, govern­
ment expenditure remains critical, and
is unencumbered by bias. The National
Cancer Institute (NCI) was established as
the principal agency for cancer research
in the USA in 1937, and with their mandate
further expanded to include coordination of
the National Cancer Program in 1971, the
NCI has been responsible for the manage-
ment of the grants programme that allo-
cates research funds to investigators and
cooperative groups. For most of the past
decade, the NCI has had essentially flat
budgets, requiring it to curtail many efforts
and reduce the percentage of grants funded
to numbers that hover close to single digits.
Nobody believes that all grants deserve
support, however, there is no doubt that
the current allocation leaves many valuable
ideas unfunded, and a much higher per-
centage of grant funding would be optimal.
Indeed, given cancer’s increasing impact on
mortality statistics worldwide, the measures
taken by the NCI during the past 10 years to
maximize the value of their budgets are of
© 2015 Macmillan Publishers Limited. All rights reserved
8. Testa, J. R. et al. Germline BAP1
mutations predispose to malignant
mesothelioma. Nat. Genet. 43, 1022–1025
(2011).
9. Shah, S. et al. A recurrent germline PAX5
mutation confers susceptibility to pre‑B cell
acute lymphoblastic leukemia. Nat. Genet. 45,
1226–1231 (2013).
10. Bombard, Y., Robson, M. & Offit, K.
Revealing the incidentalome when targeting
the tumor genome. JAMA 310, 795–796
(2013).
‘‘
…clinical research has
nevertheless been severely
’’
affected by the flat NCI budget
heightened importance; herein, we provide
our personal views, and not the views of the
NCI, of these measures.
The NCI portfolio is heavily weighted
to supporting investigator-initiated proj-
ects, usually focusing on the basic science
of cancer rather than clinical trials. The
wisdom of this model can be debated, but
few can deny that support for clinical trials
has suffered greatly during the ongoing
period of fiscal austerity.1 This latter view,
held widely by clinical researchers, was rati-
fied in a 2010 Institute of Medicine (IOM)
report,2 which concluded that clinical trials
for cancer in the USA were “approaching a
state of crisis”, especially those conducted
under the auspices of the NCI Cooperative
Group Program; since that report, the
outlook has worsened. The IOM recom-
mended 12 major changes,2 many of which
the NCI has initiated as part of a compre-
hensive overhaul. An important change
was consolidation of nine Cooperative
Groups into four, with the immediate goal
of improving efficacy and avoiding dupli-
cation of effort, and a long-term goal of

hopefully enabling the conduct of studies
in patients with specific molecular tumour
subtypes who might be more difficult to
accrue in smaller groups. This programme,
now referred to as the National Clinical
Trials Network (NCTN), had an overall
budget of US$151 million in the 2014 fiscal
year, representing only 3.1% of the total
NCI budget and the same amount pro-
vided for the Cooperative Group awards
in each of the 2012 and 2013 fiscal years.3
Although the distribution of funds to
various components of the NCTN is dif-
ferent to the allocation under the former
Cooperative  Group  Program, clinical
research has nevertheless been severely
affected by the flat NCI budget. A major
consequence is low reimburse­m ents to
partici­p ating institutions for patients
enrolled in NCI-sponsored clinical trials,
which has remained largely unchanged for
nearly a decade and stands at ≈US$2000
per patient. This level of reimbursement
is increasingly inadequate, considering the
expenditures of ≈US$10,000 per patient, a
reality recognized by pharma­ceutical com-
panies that now re­imburse organisations
conducting clinical trials US$10,000–40,000
per patient.4 The shortfall in reimburse-
ment means participating institutions
must find other revenue sources to support
NCI-funded trials, discouraging involve-
ment and prevent­i ng investigators from
conducting the correlative studies that are
the essence of cutting-edge clinical trials.
The shortfall is also largely responsible for
the increasing reliance on pharmaceutical-
company sponsored trials, a trend that
some argue should be embraced, as those
with the most to gain (financially) should
cover the costs. However, most investi­gators
would counter that industry involvement
affects the choice of trials conducted and
the relevance of the questions asked, as
profits remain an important c­onsideration
for publicly held corporations.
Although the NCI predominantly funds
scientific research, valid metrics to assess
the success of any of such programmes (not
their applicants) are needed. Some might
argue that publications represent an indirect
measure of scientific output and success, but
few agree, as this metric also has its flaws.5
Although assessments of established NCI
programmes are also lacking, efforts have
been made to develop effective methodolo-
gies and quantitative indicators to ascer-
tain the success or failure of new initiatives.
Two examples include the novel ‘Physical
Science–Oncology Centers’ (PS–OC) and
A DECADE IN MEDICINE NOVEMBER 2015  |  26
the ‘Provocative Questions’ programme.
At its launch in 2009, the PS–OC funded
12 centres throughout the USA, with over
250 investigators dedicated to building
transdisciplinary teams and infrastructure
to conduct transformative research. An
emphasis on engaging indivi­duals with
backgrounds in the physical scien­ces and
limited previous involvement in cancer
iStock/Thinkstock
studies was an important goal of the PS–OC.
Initial assessments revealed rapid scien­tific
progress, such as the application of con-
cepts based on physical science principles in
clinical trials.6 Further evidence indicated a
clear increase in cross­disciplinary collabor­
ations and publications in post-award years
among the PS–OC primary investigators.
Although encouraging pro­g ress is being
made, sufficient time to reach a level of
common u­n derstanding—e­specially for
interdisci­p linary p­r ogrammes—will be
crucial before  a final assessment can
be made. Similarly, a combi­nation of subjec-
tive and objective assessments will be used
to measure the success of the Provocative
Questions programme in cancer research,
an equally novel initiative that launched
in 2012 and will be evaluated in the
coming years.7
‘‘ Initial assessments revealed
rapid scientific progress…
’’
Progress has been more difficult as the
NCI has tried to ‘think outside of the box’
in order to achieve much-needed break-
throughs. Economists typically argue that
taking risks is critical to any investment
strategy and that a minimum of 5–10% of
a portfolio should be held in ‘alternatives’,
which generally have low correlation with
the majority of the portfolio and provide
welcome diversification. One could argue
the NCI should take the same approach:
part of the budget should be devoted to such
‘high-risk/high-payoff ’ alternative invest-
ment strategies. However, pursuing such a
strategy has been difficult.
Although most agree that current NCI
funding mechanisms are far from perfect,
there seems to be a reluctance to pursue
alternatives. This reluctance is empha-
sized by the fact that peer-reviewed NIH
funding currently represents one of the
most important metrics for academic
career development. Perhaps scientists
are more risk-averse; how else could one
explain the skepticism and hesitation
expressed by so many when the Provocative
© 2015 Macmillan Publishers Limited. All rights reserved
CLINICAL ONCOLOGY
Questions initia­t ive was launched, 8 or
con­c erns regarding the well-validated
PS‑OC programme that represent only
a very small fraction of the entire NCI
budget (0.3% and 0.6%, respectively)? In
launching the Provocative Questions pro-
gramme, Varmus and Harlow 7 noted that
—p­a rticularly during periods of fiscal
austerity—grant appli­c ants and review-
ers tend to favour grants converging on
similar, popular and conservative ideas,
thus narrow­ing research portfolios; efforts
such as those taken by the NCI in the latter
half of the past decade are attempting to
challenge such thinking. Similarly, the NCI
Outstanding Investigator grants, 7‑year
‘people grants’ of up to US$600,000 a year
(plus another 50% for overhead costs) that
will replace project-based grants, has been
launched with the hope of unbridling inno-
vation. The amount of expenditure will not
be trivial, as the NCI expects to fund 50
additional Outstanding Investigator Awards
per year for 7 years, by which time the pro-
gramme will total US$317 million per year,
or about 16% of the US$2 billion that the
NCI spends on research grants currently. To
be sure, these ‘people grants’ are not uni-
formly endorsed—given concerns that they
might be unfair to younger investi­gators,
favouring established investigators based
on past track record and not based on the
scientific rigour of proposed research, as
well as the fact that their long duration will
result in less money turning over each year
and, therefore, fewer awards;9 however, if
properly executed, this grants programme
might promote much needed ‘outside of the
box t­hinking and research’.
Continued funding is critical to enable
substantive progress in the search for thera­
pies that can prolong the lives of patients
with cancer. The stakes are high and the
need for progress essential. The NCI efforts
in this direction over the past decade
remain a work in progress.
CLINICAL ONCOLOGY
Medical Oncology, Center for Cancer Research,
National Cancer Institute, Building 10,
Room 12N226, 9000 Rockville Pike, Bethesda,
MD 20892, USA (T.F.). Department of Medicine,
Division of Hematology and Medical Oncology,
Weill Cornell Medical College/Weill Cornell
Cancer Center, 520 East 70th Street, Starr 341,
New York, NY 10021, USA (P.G.).
Correspondence to: T.F.
fojot@mail.nih.gov
Competing interests
The authors declare no competing interests.
1. American Society of Clinical Oncology.
The Funding Crisis in Cancer Clinical Trials:
A Focus on Cooperative Groups [online],
27  |  NOVEMBER 2015  www.nature.com/reviews
http://www.asco.org/advocacy/funding-crisis-
cancer-clinical-trials-focus-cooperative-groups
(2014).
2. Institute of Medicine (US) Committee on Cancer
Clinical Trials and the NCI Cooperative Group
Program (eds Nass, S. J., Moses, H. L.
& Mendelsohn, J.). A National Cancer Clinical
Trials System for the 21st Century: Reinvigorating
The NCI Cooperative Group Program
(The National Academies Press, 2010).
3. National Cancer Institute at the National
Institutes of Health. Clinical Trials Programs and
Initiatives: An Overview of NCI’s National Clinical
Trials Network [online], http://www.cancer.gov/
clinicaltrials/nctn (2014).
4. Seow, H.‑Y. et al. Funding oncology clinical
trials: are cooperative group trials sustainable?
J. Clin. Oncol. 30, 1456–1461 (2012).
© 2015 Macmillan Publishers Limited. All rights reserved
5. Begley, C. G. & Ellis, L. M. Drug
development: raise standards for preclinical
cancer research. Nature 483, 531–533
(2012).
6. Basner, J. E. et al. Measuring the evolution
and output of cross-disciplinary collaborations
within the NCI Physical Sciences–Oncology
Centers Network. Res. Eval. 22, 285–297
(2013).
7. Varmus, H. & Harlow, E. Science funding:
Provocative questions in cancer research.
Nature 481, 436–437 (2012).
8. Benowitz, S. Provocative Questions initiative
to fund innovative cancers research. J. Natl
Cancer Inst. 104, 970–972 (2012).
9. Kaiser, J. Funding. NIH institute considers
broad shift to people awards. Science 345,
366–367 (2014).
Connect to...

Follow us on Twitter
For the latest commentary and oncology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Clinical Oncology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Clinical Oncology

Web focuses
Cutting-edge investigators highlight the important advances and challenges in the latest focus on paediatric cancer
research and clinical practice

w w w.
Web collections
A selection of articles on a related theme, providing a panoramic view of a key area of clinical oncology

Posters
A visual overview of intrinsic subtypes of breast cancer

Conferences
You can meet the editors of Nature Reviews Clinical Oncology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Clinical Oncology!

November 2014 marked the 10th anniversary of the launch of Nature Reviews Clinical Oncology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of clinical oncology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrco10

© 2015 Macmillan Publishers Limited. All rights reserved

ENDOCRINOLOGY
DECADE IN REVIEW—BONE

Great strides made but still further to go

Ian R. Reid
Our understanding of bone biology and the subsequent development of therapies to treat bone diseases have
both expanded greatly in the past 30 years. This article reviews some of the key advances made in these fields
during the past decade.
The recognition that osteocytes have a
central role in bone biology has been a major
develop­ment in the past decade. Elegant
studies in rodents have demon­strated the
influence of skeletal loading on osteocytic
production of the osteoblast inhibitory
protein, sclerostin. A seminal article by
Robling et al. concluded that “modulation of
sclerostin levels appears to be a finely tuned
mechanism by which osteocytes coordinate
regional and local osteogenesis in response
to increased mechanical stimulation.”1 This
finding complements the established role of
the osteocyte as a source of receptor activa­
tor of nuclear factor κB ligand (RANKL),
and enables these cells to directly influence
both bone formation and bone resorption.
Consistent with their key regulatory func­
tions, sclerostin and RANKL are both now
targets for new drug development.
The influence of osteocytes is not just
limited to bone, as they also have a central
role in phosphate and vitamin D metab­
olism via production of fibroblast growth
factor 23 (FGF23). This protein reduces
renal phosphate reabsorption by downregu­
lating the expression of sodium–phosphate
cotransporters in the apical membrane of
the proximal tubule. FGF23 also reduces
expression of 25-hydroxyvitamin  D 3
1α-hydroxylase and increases expression of
1,25-dihydroxy­vitamin D3 24-hydroxylase,
which reduces  levels of 1,25-dihyroxy­
vitamin D3. Serum levels of both phosphate
and 1,25-dihyroxy­vitamin D increase levels of
FGF23. The FGF23 receptor is a hetero­meric
complex of a canonical FGF receptor and
α‑Klotho.2 Although levels of FGF23 increase
in renal failure and are associated with worse
The past decade has seen several major
advances in therapies for osteoporosis. First,
the phase III trial for zoledronate resulted
in the availability of an annual intravenous
infusion for the management of osteo­
porosis. Zoledronate might have slightly
greater efficacy than oral bisphos­phonates
in the prevention of vertebral fractures, but
the principal benefit of zole­dronate over oral
bisphosphonates is one of convenience. This
benefit increases the likelihood of treatment
adherence, as patients only need to remem­
ber to take their medication once a year.
The study by Lyles and colleagues3 broke
new ground by demonstrating the effective­
ness of zoledronate within 3 months of a
fracture, and also the intriguing finding of
lower mortality in those allocated to active
treatment. This obser­vation is consistent, to
some extent, with the growing body of data,
in which osteoporotic fractures are associ­
ated with increased mortality; however, the
mechanism by which mortality is redu­ced
remains uncertain.
A similar study with denosumab
showed comparable efficacy
for this monoclonal anti­
body directed against
the osteoclasto­genic
protein, RANKL.4
Denosumab,
which is admin­
istered as a sub­
cutaneous injection
every 6 months, does not
cause an acute phase
response and is safe in
patients with impair­ed
renal func­tion. As
ck
to
property is theore­tically advantageous if
patients develop adverse effects resulting
from low bone turnover. Like zoledronate,
denosumab produces a sustained reduc­
tion in fracture rates over long-term follow-
up. Denosumab and zoledronate are now
also widely used in managing the skeletal
consequences of cancer.
In 2014, monoclonal antibodies directed
against the osteoblast inhibitor, sclerostin,
were reported to transiently increase
bone formation and transiently decrease
bone resorption, together with dramatic,
sustained increases in BMD at the hip and
spine.5 The antifracture efficacy of one of
these antibodies, romosozumab, is currently
being assessed.
Although new agents have been devel­
oped, the efficacy and safety of some existing
interventions have been questioned. Meta-
analyses have demonstrated an increased
risk of cardio­v ascular events with
calcium supplements.6 These risks
are sufficiently common to out­
weigh the small bene­ficial effect
of calcium supplements on frac­
tures. When combined with the
established adverse effects of calcium
(such as gastro­intestinal symp­
toms and kidney stones), cal­cium
supplements are unlikely to
produce a net benefit
in most patients. A
similar scrutiny  of
vitamin D supple­m ents
indicates a lack of effect on BMD
and fractures in nonosteomalacic
populations. Furthermore, high doses of
vitamin D supplements might, paradox­ically,

outcomes for these patients, whether FGF23 denosu­m ab has increase fractures and falls.
ks
hin

has a causative role in these consequences is a rapid  offset The accumulation of bisphosphonates
/T
ck
to

unknown. These discoveries have provided of action, strict in bone has caused concern that their
/iS
ion

insights into a range of hypophospha­taemic c ompl i anc e chronic use could result in skeletal toxi­city.
ax
Er

and hyperphospha­t aemic disorders, but is important; Osteonecrosis of the jaw and atypical femoral
many details remain to be elucidated. however, this fractures (AFF) might be mani­festations of

28  |  NOVEMBER 2015  www.nature.com/reviews

© 2015 Macmillan Publishers Limited. All rights reserved

this toxicity. Although osteo­necrosis of the
jaw is mainly a problem in oncology prac­
tice, spontaneous fractures of the femoral
shaft have been reported by many osteo­
porosis clinics and seem to have an associ­
ation with long-term oral bisphosphonate
use, as demonstrated by Schilcher and col­
leagues.7 In this study, AFFs were shown to
be transverse stress fractures originating in
the lateral femoral cortex, whereas classic,
spiral, comminuted, femoral fractures are
no more common in bisphosphonate users
then in other indivi­duals with osteoporosis.7
Risk of AFFs declines rapidly following dis­
continuation of bisphosphonate use,7 which
suggests that ‘drug holidays’ should be part of
long-term osteoporosis management.
In the FLEX study 8—a ‘drug holiday’
trial—women already on alendronate for
5 years were randomly reassigned to either
continued therapy or placebo. The results
from FLEX showed that halving the conven­
tional dose of alendronate did not reduce
its effects on BMD. Additionally, FLEX
demonstrated stable nonvertebral fracture
rates in patients who continued therapy
or crossed over to placebo, but suggested
a benefit from continued intervention in
those whose femoral neck BMD was still
in the osteo­porotic range. A similar study
with zole­dronate reached similar conclu­
sions. Together, these studies now guide the
long-term use of these drugs.
In the decades since osteoporosis treat­
ment became a reality, the definition of
osteoporosis and which patients should be
treated, has evolved. In the past 10 years,
cost-effectiveness has become an important
part of making treatment decisions, and
requires quantification of absolute fracture
risk to target interventions optimally. The
development of fracture risk estimators has
ensued, and has dramatically changed clinical
practice, with FRAX9 and the Garvan calcu­
lators being freely available via the internet.
Standard practice is now for treatment to be
provided on the basis of absolute fracture
risk, which is a convenient way of integrat­
ing a range of clinical risk factors, including
BMD. The various calculators use slightly dif­
ferent selections of risk factors, and comple­
ment one another in the assessments they
provide. The Garvan calculator is strongly
influenced by fall history, a risk factor often
ignored by endocrinologists. All calculators
recognize the importance of age in calcu­
lating fracture risk, and have probably led
to more widespread intervention in elderly
patients and a more conservative approach
to individuals in the first 1–2 decades after
A DECADE IN MEDICINE NOVEMBER 2015  |  29
menopause. Notably, these calculators do not
include certain ‘risk factors’, such as calcium
intake and exercise, that have been heavily
emphasized in the past, but which proved
to be of little utility in estimating absolute
fracture risk.
Although much less common than osteo­
porosis, Paget’s disease can be a source of sub­
stantial morbidity and, at the beginning of the
last decade, remained a significant therapeu­
tic challenge. Intravenous zoledronate infu­
sions have revolutionized the management
of Paget’s disease, normalizing serum levels of
alkaline phosphatase in ~90% of patients.10
These biochemical changes are associated
with improvements in quality of life. Perhaps
more importantly, the biochemical and
quality of life improvements are sustained
in most patients without re-dosing; <1% of
patients relapse over >6 years of follow-up
after a single dose of the drug. Zoledronate
has now become the first-line therapy for
this condition and is effectively curative in
the majority of patients.
The past decade has seen many exciting
developments, with new therapeutic and
diagnostic techniques resulting directly
from advances in our understanding of bone
biology. If the coming decade is as productive
as the last, we will be in a much stronger posi­
tion to deal with the advancing tide of bone
disease, which increasing longevity is forcing
upon us.
Faculty of Medical and Health Sciences,
University of Auckland, Private Bag 92019,
Auckland, New Zealand.
i.reid@auckland.ac.nz
doi:10.1038/nrendo.2015.143
Published online 25 August 2015
DECADE IN REVIEW—THYROID DISEASE
The endocrinology of thyroid
disease from 2005 to 2015
P. Reed Larsen
The past decade has seen exciting progress in the field of thyroid
disease, especially in the evaluation of thyroid nodules, the genomic
characterization of carcinomas and the treatment of carcinomas
after surgery. An improved understanding of hyperthyroidism during
pregnancy, as well as the causes of ‘low T3 syndrome’ and consumptive
hypothyroidism have also been achieved.
As the importance of fine needle aspiration
for evaluating thyroid nodules for potential
malignancy became apparent, a standard­
ized approach to cytologic classification of
© 2015 Macmillan Publishers Limited. All rights reserved
ENDOCRINOLOGY
Acknowledgements
I.R.R. thanks J. Cornish, N. Sims and T. J. Martin for
advice during the preparation of this manuscript.
I.R.R. is supported by the Health Research Council
of New Zealand.
Competing interests
I.R.R. has received research funding from Amgen,
Merck and Novartis, and has received honoraria
from Amgen, Lilly, Merck, Novartis and Sanofi.
1. Robling, A. G. et al. Mechanical stimulation
of bone in vivo reduces osteocyte expression
of Sost/sclerostin. J. Biol. Chem. 283,
5866–5875 (2008).
2. Urakawa, I. et al. Klotho converts canonical
FGF receptor into a specific receptor for FGF23.
Nature 444, 770–774 (2006).
3. Lyles, K. W. et al. Zoledronic acid and clinical
fractures and mortality after hip fracture.
N. Engl. J. Med. 357, 1799–809 (2007).
4. Cummings, S. R. et al. Denosumab for prevention
of fractures in postmenopausal women with
osteoporosis. N. Engl. J. Med. 361, 756–765
(2009).
5. McClung, M. R. et al. Romosozumab in
postmenopausal women with low bone
mineral density. N. Engl. J. Med. 370, 412–420
(2014).
6. Bolland, M. J., Grey, A., Avenell, A., Gamble, G. D.
& Reid, I. R. Calcium supplements with or without
vitamin D and risk of cardiovascular events:
reanalysis of the Women’s Health Initiative
limited access dataset and meta-analysis.
BMJ 342, d2040 (2011).
7. Schilcher, J., Michaelsson, K. & Aspenberg, P.
Bisphosphonate use and atypical fractures
of the femoral shaft. N. Engl. J. Med. 364,
1728–1737 (2011).
8. Black, D. M. et al. Effects of continuing or
stopping alendronate after 5 years of
treatment: the Fracture Intervention Trial
Long-term Extension (FLEX): a randomized trial.
JAMA 296, 2927–2938 (2006).
9. Kanis, J. A., Johnell, O., Oden, A., Johansson, H.
& McCloskey, E. FRAX™ and the assessment
of fracture probability in men and women from
the UK. Osteoporos. Int. 19, 385–397 (2008).
10. Reid, I. R. et al. Comparison of a single infusion of
zoledronic acid with risedronate for Paget’s
disease. N. Engl. J. Med. 353, 898–908 (2005).
this material was necessary. After several
years of preliminary collaborative studies, the
National Cancer Institute hosted a conference
of cytopathologists and endo­crinologists. The
ENDOCRINOLOGY
result of this meeting was the development of
the Bethesda System for thyroid cytopathol­
ogy.1 Six diagnostic categories were defined
and the cancer risk of each estimated from
preliminary studies. These categories were:
I, nondiagnostic (risk 1–4%); II, benign (risk
0–3%); III, atypical or follicular lesions of
un­determined significance (risk 5–15%);
IV, diagnostic of or suspicious for follicular
neoplasm (risk 15–30%); V, suspicious for
malignancy (risk 60–75%); and VI, malig­
nant (risk 97–99%). The Bethesda System has
been widely adopted. The main limitation of
this system is that 40–60% of patients with an
‘indeterminate’ classification (that is, catego­
ries III and IV) who undergo surgery actually
have benign lesions, which emphasizes the
need for more-specific preoperative criteria
to reduce the large numbers of unnecessary
diagnostic surgical procedures.
Molecular tools are now being developed
for this purpose. One approach is to rule out
the expression of mutant genes known to be
associated with malignancy while identifying
others that are associated with atypical cytol­
ogy using the Gene Expression Classifier
(GEC).2 Of 121 surgical samples classified
as suspicious by GEC, 44% were malignant.
Follow up of 174 patients with tumours classi­
fied as benign revealed that only one in 11
patients who underwent surgery had a malig­
nancy. Further ultra­sonographic and/or sur­
gical evaluations did not suggest malignancy
in any other patients in this category. These
findings indicate an improved specificity
of the GEC; however, the number of patients
in the studies was small.
A second approach that might identify
which patients with papillary thyroid carci­
nomas (PTCs) have a high risk of recurrent
disease is to screen nodules for a Val600Glu
mutation in BRAF (BRAFV600E); tumours with
these mutations have a twofold higher recur­
rence rate than lesions with wild-type BRAF.
However, this screening would not eliminate
all PTCs as only 50–60% of these tumours
express this mutation.3
In a recent landmark paper from the
The Cancer Genome Atlas consortium, in
which genomic and tumour exome, RNA-
sequencing and microRNA analyses of 496
patients has been performed, ~96% of the
causal mutations in PTCs were identified.
Overall, the main mutations associated with
tumours were in BRAF (62%), NRAS, HRAS
and KRAS (13%) and mutations in genes
encoding transmembrane receptors, mainly
RET (10%).4 Tumours expressing BRAFV600E
and those with BRAF and RET fusion had
high levels of mitogen-­activating protein
30  |  NOVEMBER 2015  www.nature.com/reviews
kinase (MAPK) signalling that was associated
with marked suppression of genes required
for thyroid hormone synthesis. By contrast,
tumours without the BRAFV600E mutation
and with mutations in the RAS genes did
not. In general, the expression profiles of
PTCs can be categorized as either BRAFV600E-
like, poorly differentiated and signalling
through MAPK, or RAS-like and signal­
ling via both phosphatidyl­inositol 3‑kinase
and MAPK, and are well differentiated.
BRAFV600E mutations were present in 78%
of tall cell variant PTCs, whereas follicu­
lar variants were RAS-like. Also correlat­
ing with aggressive behaviour of tumours
in BRAF-like lesions were levels of miR‑21
expression and mutations in TERT, which
can explain the hetero­geneity of prognoses
of BRAFV600E-expressing tumours.3,5
‘‘
...the next decade will
see widespread application
of genomic analyses…
’’
The categorization of tumours into BRAF-
like and RAS-like groups is consistent with
results of an important clinical study in
which short-term treatment with the MAPK
antagonist selumetinib was more successful
in increasing iodine uptake to therapeutically
effective levels in lesions with a mutation in
NRAS than in lesions with a mutated form
of BRAF.5 Although preliminary, this study
provides a proof-of-principle that MAPK
inhibition can redifferentiate metastatic
PTC lesions. These results also highlight the
importance of molecular characterization
of metastatic, radioiodine-resistant PTCs to
permit individualization of therapy.
Two important publications that address
therapeutic strategies after surgery in
patients with low-risk PTC lesions have
compared remnant ablation strategies and
radioiodine dosages.6,7 In prospective, ran­
domized two‑by‑two comparisons that
involved 1,105 low-risk patients, increas­
ing TSH using recombinant TSHα was
found to be as effective in ablating residual
thyroid tissue and reducing levels of thyro­
globulin as thyroid hormone withdrawal
for 3–4 weeks in patients receiving a single
dose of either 1.1 GBq (30 mCi) or 3.7 GBq
(100 mCi) radioiodine.6,7 No difference was
seen in the success rate of the two radio­
iodine doses. In fact, in many patients, the
target reduction in serum thyroglobulin
to ≤2 µg/l was achieved by surgery alone,
which demonstrates that many patients with
pT1–T3 stage PTCs might not require any
© 2015 Macmillan Publishers Limited. All rights reserved
radioiodine therapy after surgery. Follow-up
studies were performed 6–9 months after
therapy, but results after longer intervals will
be important, as will studies comparing no
radioisotope with the 1.1 Gbq dose. Perhaps
these studies will also include a molecular
­characterization of the lesions.
Another key advance in the past 10 years
relates to the treatment of pregnant patients
who have thyrotoxicosis. Treatment with
methimazole or carbimazole during the
first trimester of pregnancy is well known
to be associated with a risk of birth defects,
including aplasia cutis and choanal or
oesophageal atresia. Consequently, current
guidelines suggest that, when possible,
propyl­thiouracil (PTU) should be substi­
tuted for these agents during the first tri­
mester. In a 2014 nationwide register study
of >800,000 children born in Denmark, 11
of 564 infants exposed to PTU during early
gestation had increased risks of kidney
cyst, hydro­nephrosis or pre-auricular or
bronchial sinus abnormalities (hazard
ratio increased by threefold to fivefold).8
Although these lesions are not as serious as
those occurring with methimazole or car­
bimazole, most required surgical remedi­
ation. When antithyroid drugs cannot be
avoided during the first trimester, patients
receiving PTU should be advised of this risk.
Consumptive hypothyroidism was first
recognized in infants with large hepatic
haem­a ngiomas expressing type 3 iodo­
thyronine deiodinase (DIO3); abnormally
high levels of DIO3 results in thyroid
hormone being inactivated more rapidly
than can be replaced by the healthy thyroid.
This condition has now been identified
in adults with gastro­i ntestinal stromal
tumours, many of whom also express DIO3.9
The requirement for thyroid hormone sup­
plementation might be further exacerbated
in these patients (and others with different
malignancies) by tyro­sine kinase inhibitor
therapy (for example, imatinib or suni­
tinib), which can induce DIO3 expression
in the tumour.
For many decades, it has been known that
almost all patients with moderate to severe
illnesses develop abnormal thyroid function
tests, almost universally a reduced T3 and an
increased reverse T3 (rT3) in serum. Other
studies have now revealed that a major cause
of these abnormalities is the generation of
intracellular reactive oxygen species by the
cytokine IL‑6, which depletes the cytosolic
thiol cofactors required for activation of T4
to T3 in the liver, kidney and skeletal muscle
mediated by type 1 and type 2 iodothyronine

deiodinase (DIO1 and DIO2, respectively).10
These deiodinases are the source of 80%
of the T3 produced in iodine-sufficient
humans. Unexpectedly, IL‑6 might also
induce DIO3 expression and this T3 and T4
inactivating deiodinase, unlike DIO1 and
DIO2, has comes into contact with extra­
cellular thiols. Consequently, critically ill
patients are likely to have some degree of
consumptive hypo­thyroidism during illness,
which further reduces concentrations of
serum T3.
Hopefully, the next decade will see wide­
spread application of genomic analyses
of all indeterminate fine needle aspira­
tion samples, thus reducing the numbers
of unnecessary diagnostic surgeries. I also
anticipate the development of more potent
and specific MAPK inhibitors for thera­
peutic redifferentiation of metastatic,
r­adioiodine-resistant PTCs.
Brigham and Women’s Hospital, Division of
Endocrinology, Diabetes and Hypertension,
77 Avenue Louis Pasteur, HIM 641, Boston,
MA 02115, USA.
plarsen@partners.org
doi:10.1038/nrendo.2015.169
Published online 6 October 2015
Competing interests
The author declares no competing interests.
1. Cibas, E. S. & Ali, S. Z. The Bethesda system for
reporting thyroid cytopathology. Thyroid 19,
1159–1165 (2009).
2. Alexander, E. K. et al. Multicentre clinical
experience with the Afirma gene expression
classifier. Clin. Endocrinol. Metab. 99, 19–25
(2014).
3. Xing, M. et al. Association between BRAF V600E
mutation and recurrence of papillary thyroid
cancer. Clin. Oncol. 33, 42–50
(2015).
4. Ho, A. L. et al. Selumetinib-enhanced
radioiodine uptake in advanced thyroid
cancer. N. Engl. J. Med. 368, 623–632 (2013).
5. Cancer Genome Atlas Research Network.
Integrated genomic characterization of papillary
thyroid carcinoma. Cell 159, 676–690 (2014).
6. Schlumberger, M. et al. Strategies of radioiodine
ablation in patients with low-risk thyroid cancer.
Tumeurs de la Thyroïde Refractaires Network for
the Essai Stimulation Ablation Equivalence Trial.
N. Engl. J. Med. 366, 1663–1673 (2012).
7. Mallick, U. et al. Ablation with low-dose
radioiodine and thyrotropin alfa in thyroid
cancer. N. Engl. J. Med. 366, 1674–1685
(2012).
8. Maynard, M. A. et al. Thyroid hormone
inactivation in gastrointestinal stromal
tumors. N. Engl. J. Med. 37, 1327–1334 (2014).
9. Andersen, S. L., Olsen, J., Wu, C. S.
& Laurberg, P. Severity of birth defects after
propylthiouracil exposure in early pregnancy.
Thyroid 24, 1533–1540 (2014).
10. Wajner, S. M. et al. IL‑6 promotes nonthyroidal
illness syndrome by blocking thyroxine
activation while promoting thyroid hormone
inactivation in human cells. J. Clin. Invest. 121,
1834–1845 (2011).
A DECADE IN MEDICINE NOVEMBER 2015  |  31
DECADE IN REVIEW—TYPE 2 DIABETES MELLITUS
At the centre of things
Guang Ning
Over the past decade, the growing epidemic of type 2 diabetes mellitus
(T2DM) and its associated complications has presented both challenges
and opportunities. Progress has been made in incretin-based therapies,
bariatric surgeries and inhibiting renal glucose reabsorption; however,
long-term safety and efficacy studies are required. Advances in the
prevention of macrovascular complications of T2DM from ongoing
clinical trials are expected soon.
In the past 10 years, major advances have
been made in research into type 2 dia­betes
mellitus (T2DM). Epidemiological studies
have revealed an increasing worldwide
prevalence of T2DM and related mortality.
Two surveys published in the past 3 years on
the prevalence of T2DM and the associ­ated
disease burden, which revealed the worrying
situation we face today, might change global
and national health policies in the future.1,2
The Global Burden of Disease Study (GBD)
20131 provides the latest and most com­
prehensive assessment of causes of death,
identifying 240 specific causes of death in
188 countries; dia­betes (all forms) was iden­
tified as one of the major causes of reduced
life expectancy globally. In 2013, mortality
from diabetes reached 1.3 million worldwide,
which is almost double the mortality in 1990;
age-standardized death rates increased signi­
ficantly from 19.8 per 100,000 indivi­duals
in 1990 to 21.6 per 100,000 indivi­duals in
2013. The significant increase in mortality
was undoubtedly connected with the con­
siderably increased prevalence of diabetes
worldwide, especially in Asia.
A general population-based survey of
98,658 adults conducted in China in 20102
highlights the extent of the problem in Asia.1
This survey, which used the American
Diabetes Association 2010 criteria, estimated
the prevalence of diabetes to be 11.6% and
that of pre­diabetes to be 50.1%, making
China the country with the highest preva­
lence of diabetes in Asia and the largest
absolute disease burden for diabetes in the
world. Furthermore, poor awareness, treat­
ment and disease control among patients
with dia­betes was revealed, which highlights
an urgent need for a multi­faceted approach
in primary diabetes prevention.
Advances in our understanding of the
genetics of T2DM in the past decade can
largely be attributed to the substantial pro­
gress made by international collaborations
using genome-wide association studies
(GWAS). A 2014 study reported the largest
© 2015 Macmillan Publishers Limited. All rights reserved
ENDOCRINOLOGY
(so far) transethnic, GWAS meta-analysis
from four major ethnic groups.3 This study
identified seven novel T2DM susceptibility
loci, which had modest effects on T2DM sus­
ceptibility but were conserved across ethnic
groups, adding to the existing 69 established
T2DM susceptibility loci.3 Improvements in
the fine-mapping resolution of complex trait
loci were also demon­strated by the trans­
ethnic meta-analysis. Combining GWAS
findings from diverse ancestry groups could
facilitate discovery of common variants
for T2DM.
‘‘
Epidemiological studies have
revealed an increasing worldwide
prevalence of T2DM and related
’’
mortality
Developments in the treatment of T2DM
rely on understanding the complex patho­
physiology of the disease and, in the past
decade, breakthroughs in the genetic hetero­
geneity of T2DM have revolutionized clini­
cal practice. The first success is undoubtedly
incretin-based therapies; the multitarget
modes of action of these therapies make
them ideal for metabolic control. LEAD‑3
Mono,4 the first long-term randomized trial
evaluating the safety and efficacy of lira­
glutide (a glucagon-like peptide 1 analogue)
compared with glimepiride (a sulphonyl­
urea), is a good example of this promising
strategy. In this 52-week, phase III trial
involving 746 patients, monotherapy with
liraglutide signi­f icantly reduced HbA 1c
levels compared with glimepiride in a dose-
dependent manner.4 Although both agents
stimulated insulin secretion, liraglutide
provided more stable long-term control
of HbA1c levels, fewer hypoglycaemic epi­
sodes, more weight loss and greater insulin
sensitivity than liraglutide.
Another success is the use of bariatric
surgery to treat patients with T2DM, which
markedly improves glucose metabolism
ENDOCRINOLOGY
and remission in patients with T2DM.
However, the antidiabetic actions of bari­
atric surgery are complex; decreasing the
size of the stomach and altering levels of gut
hormones are not the only consequences,
with alterations in bile acid signalling also
being important.5 Following vertical sleeve
gastrectomy, diet-induced obese mice with
targeted deletion of Nr1h4 (encoding the
bile acid receptor; also known as farnesoid
X‑activated receptor) failed to maintain sus­
tained weight loss and had worse glycaemic
control than similarly operated wide-type
littermates. 5 Functional bile acid recep­
tors are thus essential for the therapeutic
effects of vertical sleeve gastrectomy and
represent a possible new therapeutic target
for T2DM drug develop­ment, which offers
a less invasive mode of metabolic control
than surgery.
The gut microbiota has become a promi­
nent target in controlling human health and
many studies have been conducted on its
interaction with T2DM. Qin and colleagues
performed a two-stage metagenome-wide
association study to identify changes in gut
microbiota (using deep shotgun sequen­
cing of faecal samples) in 345 patients
with T2DM.6 The researchers identified
>52,484 T2DM-associated genetic markers
and established the concept of the meta­
genomic linkage group to enable taxonomic
species-level analyses. This study 6 identifies
T2DM-related dysbiosis (associated with
a decline in butyrate-producing bacteria
and an increase in levels of opportunistic
species of pathogenic bacteria in the gut),
which might be a future therapeutic target
for treating T2DM.
Inhibiting renal glucose reabsorption
independent of the traditional glucose low­
ering approaches related to β‑cell function
or insulin sensitivity is another approach
to treat T2DM that has emerged in the past
decade. Bailey et al.7 evaluated the safety and
efficacy of dapagliflozin, a highly selective
inhibitor of sodium/glucose cotrans­porter 2
(SGLT2), in a multicentre, phase III trial of
546 adult patients with T2DM who had inad­
equate glycaemic control with met­formin.7
After 24 weeks of treatment, mean HbA1c
levels decreased significantly in each dapagli­
flozin treatment group compared with
placebo groups. Dapagliflozin treat­ment
did not significantly increase symp­toms of
hypogly­caemia but did result in significant
decreases in bodyweight in each dapagli­
flozin group.7 An increase in genital infec­
tions were reported in this study, which raises
concerns about the safety of this therapy.
32  |  NOVEMBER 2015  www.nature.com/reviews
High cholesterol levels
High blood pressure
Type 2
diabetes mellitus
Sedentary lifestyle
Overweight
Other advances in areas related to the
pathophysiology of T2DM have occurred
in the past 10 years, including those in
brown adipose tissue (BAT), pancreatic
β‑cell dediffer­entiation, inflammation and
pro­opiomelanocortin (POMC) neuronal
regulation of glucose homeostasis, all of
which might be realistic targets for treating
T2DM in the future. Among these investi­
gations, one paper on BAT aroused much
interest, as it was the first large retrospec­
tive study to identify physiologically relevant
amounts of BAT in adult individuals using
combined PET–CT.8 BAT was detected in
5.4% of the patients (n = 1,972). In addition
to significant associations with age, sex and
β‑blocker use, BMI was inversely corre­lated
with the amount of BAT, especially in elderly
patients, which indicates that BAT could
have a role in protecting against obesity and
metabolic abnormalities.
‘‘ T2DM-related dysbiosis …
might be a future therapeutic
target for treating T2DM
’’
When it comes to evidence-based infor­
mation regarding the long-term prevention
of complications related to T2DM, two hall­
mark studies in the past decade stand out:
the ADVANCE study 9 and a 10-year, post-
interventional follow-up of the UKPDS
study. 10 ADVANCE represented a series
of long-term studies comparing the effect of
intensive versus standard glycaemic control
on macrovascular outcomes in patients with
T2DM who were at high risk of vascular
events.9 This study provided support for the
well-established benefits of intensive glycae­
mic control on microvascluar complications
after a median of 5 years of treatment in
11,140 patients with T2DM. Unfortunately,
no significant reduction in macro­vascular
outcomes was found in the intensive group
compared with the control group. The
10-year follow-up of the UKPDS study
evaluated the long-term effects of intensive
© 2015 Macmillan Publishers Limited. All rights reserved
Smoking
NPG
High-fat diet
Nature Reviews | Endocrinology
glycaemic control in the early course of the
disease in patients with T2DM.10 Although
between-group differences in HbA1c levels
had disappeared by the 1‑year follow-up,
long-term reductions in myocardial infarc­
tion and all-cause mortality were observed
in the intensive treatment group compared
with the conventional therapy group. The
inconsistent findings in cardiovascular out­
comes from the two studies raise concern
about which strategy is best for managing
glycaemic control in T2DM. Anticipating
macrovasuclar complications before they
develop in the early stages of T2DM instead
of at advanced stages, integrative care for
multi­ple vascular risk factors and inten­
sive glycaemic control are imperative for
long-term intervention in these patients.
In the past decade, although epidemio­
logical studies revealed the challenges of
controlling T2DM, progress in genetics,
pathogenesis and evidence-based inter­
ventional studies raises the hope that
the harmful outcomes of T2DM can be
reduced. Even so, additional efforts are
needed to prevent and treat T2DM and its
associated complications. Future research
breakthroughs will require the integration
of methodological innovation and clinical
practice in every facet of T2DM.
Shanghai National Research Center for
Endocrine and Metabolic Diseases,
Ruijin Hospital, Shanghai Jiao Tong University
School of Medicine, 197 Ruijin 2nd Road,
Shanghai 200025, PR China.
gning@sibs.ac.cn
doi:10.1038/nrendo.2015.147
Published online 25 August 2015
Acknowledgements
The author’s work is supported by the National
Basic Research Program of China (2015CB553601),
the National Science and Technology Pillar
Program of China (2013BAI09B13) and by grants
from the National Natural Science Foundation of
China (grant numbers 81390350, 81321001
and 81261120564).
Competing interests
The author declares no competing interests.

1. GBD 2013 Mortality and Causes of
Death Collaborators. Global, regional, and
national age-sex specific all-cause and cause-
specific mortality for 240 causes of death,
1990–2013: a systematic analysis for the
Global Burden of Disease Study 2013.
Lancet 385, 117–171 (2015).
2. Xu, Y. et al. Prevalence and control of
diabetes in Chinese adults. JAMA 310,
948–959 (2013).
3. DIAbetes Genetics Replication And
Meta-analysis (DIAGRAM) Consortium et al.
Genome-wide trans-ancestry meta-analysis
provides insight into the genetic architecture
of type 2 diabetes susceptibility. Nat. Genet.
46, 234–244 (2014).
4. Garber, A. et al. Liraglutide versus
glimepiride monotherapy for type 2
diabetes (LEAD‑3 Mono): a randomised,
52-week, phase III, double-blind, parallel-
treatment trial. Lancet 373, 473–481
(2009).
DECADE IN REVIEW—PAEDIATRIC ENDOCRINOLOGY
New genes, new therapies
Mehul T. Dattani
The past 10 years have seen substantial advances in many areas
of paediatric endocrinology. Major progress has been made in our
understanding of the aetiology of many disorders with the advent of
next-generation sequencing. Furthermore, the introduction of novel
therapies has revolutionized clinical management.
Whether the increased incidence of obesity,
in both adults and children, is entirely due
to lifestyle or if a genetic element also pre­
disposes an individual to increased weight
gain has been an ongoing debate. However,
in 2007, Frayling et al. per­formed a genome-
wide associ­ation study comparing patients
with type 2 diabetes melli­tus (T2DM) and
control indivi­duals,1 and identified single
nucleotide poly­morphisms (SNP) in FTO
that were strongly associated with T2DM.
FTO encodes a nucleic acid demethy­lase
that affects food intake and might also
affect energy balance and nutrient sensing.
One SNP (rs9939609) was also studied in
other European cohorts (19,424 adults of
white European descent and 10,172 chil­
dren of white European descent); impor­
tantly the association of this genetic variant
with the risk of being overweight or obese
was found to be almost exclu­sively attribut­
able to changes in fat mass.1 Homozygosity
for the rs9939609 risk allele conferred a
1.67-fold increased odds ratio of developing
obesity and an increase in weight of ~3 kg.
Despite birth weight being unaffected in
children with rs9939609, the variant was
associ­ated with being overweight or obese
by the age of 7 years.
A DECADE IN MEDICINE NOVEMBER 2015  |  33
5. Ryan, K. K. et al. FXR is a molecular target
for the effects of vertical sleeve gastrectomy.
Nature 509, 183–188 (2014).
6. Qin, J. et al. A metagenome-wide association
study of gut microbiota in type 2 diabetes.
Nature 490, 55–60 (2012).
7. Bailey, C. J., Gross, J. L., Pieters, A.,
Bastien, A. & List, J. F. Effect of dapagliflozin
in patients with type 2 diabetes who have
inadequate glycaemic control with metformin:
a randomised, double-blind, placebo-controlled
trial. Lancet 375, 2223–2233 (2010).
8. Cypess, A. M. et al. Identification and importance
of brown adipose tissue in adult humans. N. Engl.
J. Med. 360, 1509–1517 (2009).
9. ADVANCE Collaborative Group et al. Intensive
blood glucose control and vascular outcomes
in patients with type 2 diabetes. N. Engl. J. Med.
358, 2560–2572 (2008).
10. Holman, R. R. et al. 10-year follow-up of
intensive glucose control in type 2 diabetes.
N. Engl. J. Med. 359, 1577–1589 (2008).
In 2013, Pearce et al. identified muta­
tions in KSR2 in 2.1% of 2,101 patients with
severe early-onset obesity (aged <10 years),
compared with only 1.0% of patients in the
control arm.2 In this study of individuals
with severe obesity, 44 of 45 patients were
heterozygous for this KSR2 variant, and
Table 1 | Advances in clinical paediatric endocrinology
Gene or disease Mutation phenotype
or medication
FTO ~3 kg weight gain, association
from age 7 years
KSR2 Hyperphagic early-onset
obesity, insulin insensitivity
THRA Obesity, severe chronic
constipation, growth failure,
neurocognitive delay
MKRN3 Central precocious puberty
GPR101 Pituitary gigantism
Hypophosphatasia Bone-targeted, recombinant
human TNSALP
Type 1 Islet transplantation using
diabetes mellitus Edmonton protocol
Abbreviation: GWAS, genome-wide association study.
© 2015 Macmillan Publishers Limited. All rights reserved
ENDOCRINOLOGY
one patient was homozygous for two vari­
ants. The phenotype included childhood,
but not adult, hyper­phagia, significantly
reduced basal metabolic rate, increased
respiratory quotient, reduced sleeping heart
rate and insulin insensitivity with acantho­
sis nigricans. Human KSR2 mutations are
also associated with impaired glucose and
fatty acid oxidation in a mouse C2C12 cell
line.2 Importantly, the impaired fatty acid
oxidation was completely rescued by met­
formin. Although rare loss-of-function
KSR2 variants were highly enriched in
patients with early-onset obesity, they
inconsistently cosegregate with severe
obesity, which indicates that other genes or
environ­mental factors can regulate weight
in these individuals.
In another ground-breaking study, a
dominant-negative mutation (p.E403X)
in the thyroid hormone receptor α gene
(THRA) was identified by whole-exome
sequen­cing in a single child with evidence of
hypo­thyroidism.3 This patient had growth
failure with short limbs, neurodevelop­
mental delay, hypotonia, excess weight in
relation to height and chronic severe consti­
pation. Although circulating concentrations
of thyroid hormone were near-normal, the
condition seems to be associated with a sub­
normal ratio of thyroxine to triiodothyro­
nine. Skeletal development improved with
thy­roxine treatment, as did height growth,
which suggests that thyroxine treatment
might be of some benefit. Other mutations
have also been identified that clearly define
the role of thyroid hormone receptor α in
bone, brain and the gastrointestinal tract;
Relevance
First obesity gene identified by GWAS—affects
weight in childhood as well as adults1
Mutations affect appetite, energy intake and
energy expenditure. Affects glucose and fatty
acid oxidation2
First human mutation in THRA. Aids our
understanding of bone and gastrointestinal
phenotypes in common forms of hypothyroidism3
Commonest genetic cause of early puberty.
Provides new insights into regulation of puberty5
Novel gene implicated in pituitary gigantism with
early onset of excess growth hormone secretion6
First treatment for hypophosphatasia. Improves
life expectancy and bone mineralization7
Improves blood glucose regulation and reduces
number of severe hypoglycaemic episodes8,9
ENDOCRINOLOGY
these observations are pertinent to under­
standing the pathophysiology of the con­
genital or acquired hypothyroidism seen in
both children and adults.
Our understanding of puberty-related
disorders has also dramatically increased,
and at least 26 genes have been associated
with hypogonadotropic hypogonadism.4
The discovery of heterozygous mutations
in MKRN3 in five of 15 pedigrees with
central precocious puberty revealed a
novel gene that might regulate puberty in
humans.5 MKRN3 is an imprinted, pater­
nally expressed gene located within the
Prader–Willi syndrome region on chro­
mosome 15q11‑13. MKRN3 is strongly
expressed in the arcuate nucleus before
puberty and expression decreases immedi­
ately before puberty, which suggests that
the gene product inhibits the initiation
of puberty.
‘‘
...next-generation sequencing
has enhanced our understanding
of ... paediatric endocrine
’’
disorders...
Until 2014, only AIP mutations had been
identified as a common cause of pitui­
tary gigantism, a rare condition caused
by excess secretion of growth hormone by
a pituitary adenoma. However, Trivellin
et al. identified heritable microduplica­
tions on chromosome Xq26.3 in 13 patients
(nine female) with early childhood-onset
pituitary gigantism, X‑linked acrogigan­
tism.6 One of the three genes in this region
(GPR101) encodes an orphan G‑protein
coupled receptor, which is probably respon­
sible for the pituitary gigan­t ism pheno­
type. Expression of GPR101 was greatly
increased in the pituitaries of patients with
the duplication; GPR101 mutations have
also been detected in 11 of 248 patients
with sporadic acromegaly, which sup­
ports its role in the aetiology of pituitary
gigantism and/or acromegaly.6
The identification of genetic muta­
tions by next-generation sequencing
has enhanced our understanding of the
aetiology and pathogenesis of paediatric
endocrine disorders, and can aid genetic
counselling for patients and their fami­
lies (Table 1). However, most importantly,
they will help to direct research into novel
therapies for rare genetic disorders.
The clinical management of childhood
endocrine disorders has also advanced
in the past 10 years (Table 1). Whyte and
34  |  NOVEMBER 2015  www.nature.com/reviews
coll­eagues described a major therapeutic
advance for the management of hypophos­
phatasia.7 This disorder is characterized by
the extracellular accumulation of inorganic
pyrophosphate that leads to rickets and is
associated with mutations in TNSALP,
which encodes the tissue-nonspecific
isoenzyme of alkaline phosphatase.
Children with hypophospha­t asia have
chest deformities that result in res­piratory
failure, persistent bone disease with poor
mineralization and hypercal­caemia with
nephrocalcinosis. Whyte et  al. used a
recom­binant human TNSALP (asfotase
alfa), which prevents hypophosphatasia
in Tnsalp null mutant mice, in 11 patients
aged <3 years. In nine individuals, rickets
was cured at 6 months, with an improve­
ment in motor function, neuro­cognition
and pulmonary function. Skeletal abnor­
malities also dramatically improved and
in one patient who had very little bone
visible on radiographs at baseline, some
reminerali­zation was noted at 48 weeks. At
week 48, six of nine patients were breathing
sponta­neously, compared with only one at
the beginning of the study.
Finally, one cannot summarize the past
decade’s research highlights in paediatric
endocrinology without mentioning the
inter­national trial of the Edmonton pro­
tocol for islet transplantation in type 1
dia­b etes melli­tus (T1DM).8 In this study,
Shapiro et al. used islet transplantation
in 36 indivi­d uals with unstable T1DM
(including severe recurrent hypoglycae­
mia). Of these patients, 16 (44%) achieved
insulin independence, 10 (28%) had partial
function and a further 10 (28%) had
complete graft loss 1 year after the final
transplant. All patients with residual islet
function were completely protected from
severe hypoglycaemic episodes. However,
only five individuals were insulin inde­
pendent at 2 years. Importantly, serious
adverse effects were either procedure-
related or immuno­s uppression-related.
However, a recent study suggested that
improved control (HbA1c levels <7%) and
freedom from severe hypoglycaemia can
be maintained in 60% of patients 5 years
after transplant, with up to 26% remaining
insulin independent.9
Although the best hope for a T1DM cure
is alternative sources of insulin-secreting
cells, in the meantime continued improve­
ment in T1DM control with closed-loop
insulin delivery using insulin pumps and
continuous glucose monitoring is being
made, as shown in a phase II randomized
© 2015 Macmillan Publishers Limited. All rights reserved
crossover trial by Hovorka and colleagues.10
In this study of young patients with T1DM,
a closed-loop system using real time sub­
cutaneous glucose monitoring reduced
the incidence of nocturnal hypoglycaemic
events compared with standard therapy
using an insulin pump and non-real time
continuous glucose monitoring.
Nature Reviews Endocrinology is to be
congratulated on its 10th anniversary; we
look forward to the next 10 years, which
will herald the advent of novel genomic
technologies with their inevitable impact
on clinical practice.
Genetics and Epigenetics in Health and
Disease, Genetics and Genomic Medicine
Programme, UCL Institute of Child Health,
30 Guilford Street, London WC1N 1EH, UK.
m.dattani@ucl.ac.uk
doi:10.1038/nrendo.2015.154
Published online 8 September 2015
Acknowledgements
M.T.D. is funded by the Great Ormond Street
Hospital Children’s Charity.
Competing interests
M.T.D. declares no competing interests.
1. Frayling, T. M. et al. A common variant in the
FTO gene is associated with body mass index
and predisposes to childhood and adult
obesity. Science 316, 889–894 (2007).
2. Pearce, L. R. et al. KSR2 mutations are
associated with obesity, insulin resistance,
and impaired cellular fuel oxidation. Cell
155, 765–777 (2013).
3. Bochukova, E. et al. A mutation in the thyroid
hormone receptor α gene. N. Engl. J. Med.
366, 243–249 (2012).
4. Boehm, U. et al. Expert consensus
document: European Consensus Statement
on congenital hypogonadotrophic
hypogonadism—pathogenesis, diagnosis,
and treatment. Nat. Rev. Endocrinol. 11,
547–564 (2015).
5. Abreu, A. P. et al. Central precocious puberty
caused by mutations in the imprinted gene
MKRN3. N. Engl. J. Med. 368, 2467–2475
(2013).
6. Trivellin, G. et al. Gigantism and acromegaly
due to Xq26 microduplications and GPR101
mutation. N. Engl. J. Med. 371, 2363–2374
(2014).
7. Whyte, M. P. et al. Enzyme-replacement therapy
in life-threatening hypophosphatasia. N. Engl.
J. Med. 366, 904–913 (2012).
8. Shapiro, A. M. et al. International trial of the
Edmonton protocol for islet transplantation.
N. Engl. J. Med. 355, 1318–1330 (2006).
9. Lablance, S. et al. Five-year metabolic,
functional, and safety results of patients with
type 1 diabetes transplanted with allogenic
islets within the Swiss-French GRAGIL network.
Diabetes Care http://dx.doi.org/10.2337/
dc15-0094.
10. Hovorka, R. et al. Manual closed loop insulin
delivery in children and adolescents with
type 1 diabetes: a phase 2 randomized
crossover trial. Lancet 375, 743–751 (2010).
 ENDOCRINOLOGY

DECADE IN REVIEW—REPRODUCTIVE ENDOCRINOLOGY both negative and positive feedback of estro­

Understanding reproductive gen, thereby providing new insights into the

regulation of ovulation and the menstrual
endocrine disorders or estrous cycle. Fourthly, kiss­peptin is an
important com­ponent of the link between
Ursula B. Kaiser metabo­lism and reproduction, as well as of
other disruptors of reproductive control such
The past decade has witnessed incredible advances in the field of as stress.
reproductive endocrinology. The use of new genetic and genomic tools The identification of this new regulator
has had a particular impact, leading to advances in our understanding, of GnRH secretion using human genomics
diagnosis and treatment of reproductive endocrine disorders, particularly ushered in a new era of genetic discover­
those related to the neuroendocrine control of reproduction and ies in the control of puberty and reproduc­
ovarian biology. tion, notably leading to the identification
of loss-of-function mutations in the genes
In the 10  years since Nature Reviews low levels of sex steroid hormones in associ­ encoding neurokinin  B (NKB) and its
Endocrinology was first published, many ation with inappropriately low-normal levels receptor, the neuro­kinin 3 receptor (NK3R),
important advances in the field of repro­ of LH and FSH. in patients with hypogonadotropic hypo­
ductive endocrinology have occurred. These two landmark papers reinvigorated gonadism.3 This discovery took us further up
During this decade, incredible advances the field of reproductive neuroendocrinol­ the neuro­nal ladder of GnRH control, with
have been made in science and medicine ogy. Subsequently, a multitude of studies demonstrations that both NKB and NK3R
in general as a result of new tools (such as in animal models, as well as in humans, are expressed in kisspeptin neurons and that
genetics, epi­genetics and big data) that are determined that the ligand for this receptor, this system regulates kiss­peptin secretion to
available in dis­covery research and can be kisspeptin, is a potent stimulator of GnRH control GnRH release. These discoveries are
applied to human biology and medicine. secretion, is produced in specific hypotha­ already affecting clinical care, with intense
These advances have made the past decade lamic neuronal populations and directly acti­ pharmaceutical interest in the development
an exciting time to be involved in transla­ vates kisspeptin receptors in GnRH neurons. of kisspeptin agonists and antagonists and
tional research in repro­ductive biology. These findings confirm that kiss­peptin is an a clinical study demonstrating the poten­
As the many advances in this field over upstream regulator of GnRH secretion. Four tial utility of kisspeptin to effectively and
the past decade could not all be included key roles of kisspeptin have been demon­ safely trigger oocyte maturation in women
in this article, I have limited the scope to strated by the >1,400 papers that have been under­going in vitro fertilization, with a pos­
advances involving, directly or indirectly, published on this topic over the past decade. sible reduction in the risk of ovarian hyper­
hormones, neuroendocrine factors and Firstly, the kisspeptin system is a gatekeeper stimulation syndrome.4 In the past few years,
paracrine factors. of puberty and GnRH secretion. Secondly, the application of genomic tools to study
I will start at the top of the reproductive kisspeptin is important for regulation of patients with central precocious puberty
axis, with advances in our understanding of male and female fertility. Thirdly, kisspeptin- (CPP; characterized by premature activa­
the neuroendocrine control of reproduction secreting neurons are important mediators of tion of GnRH secretion and early onset of
(Figure 1). Hypothalamic gonadotropin-
releasing hormone (GnRH) is at the top
of the axis and has effects in the pituitary Neurokinin B ?

gland, where peripheral and central signals

?
are inte­g rated to control the output of Kisspeptin MKRN3
gonado­tropins, luteinizing hormone (LH) ?
GnRH
and follicle-stimulating hormone (FSH).
Letrozole
In turn, LH and FSH act on the gonads to LH and FSH
stimu­late gametogenesis and the produc­
tion of gonadal sex steroids and peptides,
Estrogen
including estrogens, progesterone and tes­
tosterone, as well as inhibins. Although
GnRH has his­torically been considered to
be at the apex of the reproductive axis, the AMH
DENND1A
past decade has led to important advances in
our understand­ing of upstream regulators of
GnRH, heral­ded by two key publications.1,2 Figure 1 | Advances in reproductive endocrinology related to hypothalamic and ovarian
Nature Reviews control of
| Endocrinology
Two groups independently reported the reproduction. Neurokinin B, kisspeptin and MKRN3 have been identified as regulators of GnRH
identification of mutations in GPR54 (now release in both male and female individuals that modulate the timing of puberty, fertility and
feedback effects of estrogen and other sex steroids. Letrozole, an aromatase inhibitor, improves
better known as KISS1R), which encodes a
fertility induction in women with PCOS. DENND1A is involved in the pathogenesis of PCOS and
G protein-coupled receptor, in patients with might be a new target for treatment of PCOS. AMH is now recognized as a biomarker of ovarian
hypogonadotropic hypogonadism, a dis­ reserve for fertility induction therapy. Abbreviations: AMH, anti-Müllerian hormone; FSH, follicle-
order characterized by the absence of spon­ stimulating hormone; GnRH, gonadotropin-releasing hormone; LH, luteinizing hormone; PCOS,
taneous pubertal matur­ation, infertility and polycystic ovary syndrome.

A DECADE IN MEDICINE NOVEMBER 2015  |  35

© 2015 Macmillan Publishers Limited. All rights reserved
ENDOCRINOLOGY
puberty) has led to the identification of
mutations in MKRN3 as a prevalent cause
of CPP, confirming the genetic basis of this
disorder and identifying the first inhib­itor of
GnRH secretion in humans.5 This discovery
has already had clinical implications in the
diagnosis of CPP, enabling genetic counsel­
ling of family members and early treatment
of this disorder.5
These advances in our understanding
of the neuroendocrine control of repro­
duction could improve the safety of fertil­
ity induction in patients at risk of ovarian
hyperstimulation syndrome, such as those
with polycystic ovary syndrome (PCOS).
However, additional advances in the treat­
ment of these patients have also been made
in the past decade. The historical mainstay of
treatment for women with PCOS has been an
estrogen–progestin contracep­tive and clomi­
phene citrate when fertility is desired. PCOS
is also associated with metabolic distur­
bances such as obesity and insulin resistance.
Therapy with the insulin sensitizer, metfor­
min, ameliorates the meta­bolic disturbances
seen in PCOS; however, despite reports of
improved ovulation rates, metformin did not
increase the rate of live births.6 Aromatase
inhibitors (such as letro­zole) have been pro­
posed as a method to induce ovulation in
women with PCOS (Figure 1). These drugs
would also have stimulatory effects on the
hypothalamic–pituitary–gonadal axis and
might be better than clomiphene citrate, pro­
viding more physiologic hormonal stimu­
lation of the endometrium, a lower rate of
multiple pregnancies and a more favourable
adverse effect profile. Indeed, a double-blind,
multi­centre, randomized trial demonstrated
superiority of letrozole over clomiphene
citrate for fertility induction in women with
PCOS.7 Genes associated with PCOS, most
notably DENND1A, have been identified in
genome-wide association studies and could
open up opportunities for the development
36  |  NOVEMBER 2015  www.nature.com/reviews
of additional targets for the treatment of this
disorder (Figure 1).8
The importance of anti-Müllerian hor­
mone (AMH) in female reproduction has
been revealed over the past decade. AMH
was initially known as a ‘male’ hormone
because of its role in male sexual differenti­
ation; it is produced in the testes and causes
regression of the Müllerian ducts during
male fetal sexual differentiation. AMH
was subsequently shown to be produced in
granu­losa cells of the ovary and identified
as an inhibitor of primordial follicle recruit­
ment. Now recognized as a biomarker of
ovarian reserve, serum levels of AMH are
being used in the clinic as reliable predic­
tors of responsiveness to fertility induction
therapy,9 of ovarian damage and effects on
the ovarian reserve after chemotherapy or
radiotherapy and of age at the start of meno­
pause.10 Thus, AMH is now recognized as an
important ‘female’ hormone that is crucial
for maintaining the right tempo of folliculo­
genesis, and with important diagnostic
implications that enable a personalized
approach to fertility induction therapy.
As we reflect on the many advances
in reproductive endocrinology that have
occurred in the past decade, we likewise can
look forward to the discoveries of the next
decade; for example, the further develop­
ment and application of selective progester­
one receptor modulators, elucidation of the
effects of endocrine disruptors on oogenesis
and spermatogenesis and the translation of
epigenetic discoveries into clinical care. It
is with great anticipation and excitement
that we can look ahead to the next wave of
discoveries and advances to improve our
understanding of the most fundamental of
biologic processes—reproduction and the
preservation of our species.
Brigham and Women’s Hospital, 221 Longwood
Avenue, Boston, MA 02115, USA.
ukaiser@bwh.harvard.edu
© 2015 Macmillan Publishers Limited. All rights reserved
doi:10.1038/nrendo.2015.179
Published online 13 October 2015
Acknowledgements
I would like to acknowledge my many colleagues who
contributed ideas and suggestions for inclusion in
this review, including L. Halvorson, W. Kuohung,
E. Norwitz, H. Taylor and L. Heckert, among others.
I would also like to acknowledge the support of the
NIH (grant numbers: R01 HD019938, R01 HD
082314, U54 HD028138 and K12 HD051959).
Competing interests
The author declares no competing interests.
1. de Roux, N. et al. Hypogonadotropic
hypogonadism due to loss of function of the
KiSS1-derived peptide receptor GPR54.
Proc. Natl Acad. Sci. USA 100, 10972–10976
(2003).
2. Seminara, S. B. et al. The GPR54 gene as a
regulator of puberty. N. Engl. J. Med. 349,
1589–1592 (2003).
3. Topaloglu, A. K. et al. TAC3 and TACR3 mutations
in familial hypogonadotropic hypogonadism
reveal a key role for neurokinin B in the central
control of reproduction. Nat. Genet. 41,
354–358 (2009).
4. Abbara, A. et al. Efficacy of kisspeptin‑54 to
trigger oocyte maturation in women with high
risk of ovarian hyperstimulation syndrome
(OHSS) during in vitro fertilization (IVF) therapy.
J. Clin. Endocrinol. Metab. 100, 3322–3331
(2015).
5. Abreu, A. P. et al. Central precocious puberty
caused by mutations in the imprinted gene
MKRN3. N. Engl. J. Med. 368, 2467–2475
(2013).
6. Legro, R. S. et al. Clomiphene, metformin,
or both for infertility in the polycystic ovary
syndrome. N. Engl. J. Med. 356, 551–566
(2007).
7. Legro, R. S. et al. Letrozole versus clomiphene
for infertility in the polycystic ovary syndrome.
N. Engl. J. Med. 371, 119–129 (2014).
8. McAllister, J. M. et al. Overexpression of a
DENND1A isoform produces a polycystic ovary
syndrome theca phenotype. Proc. Natl Acad.
Sci. USA 111, E1519–E1527 (2014).
9. Nardo, L. G. et al. Circulating basal anti-
Müllerian hormone levels as predictor of ovarian
response in women undergoing ovarian
stimulation for in vitro fertilization. Fertil. Steril.
92, 1586–1593 (2009).
10. Broer, S. L. et al. Anti-Müllerian hormone
predicts menopause: a long-term follow-up
study in normoovulatory women. J. Clin.
Endocrinol. Metab. 96, 2532–2539 (2011).
Connect to...

Follow us on Twitter
For the latest commentary and endocrinology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Endocrinology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Endocrinology

Web focuses
A selection of articles on a related theme with content ranging from Research Highlights through to Reviews
and Perspectives

w w w.
Web collections
A selection of articles that provide a panoramic view of a key area of endocrinology

Posters
Visual overviews of the pathophysiology of type 1 diabetes mellitus and type 2 diabetes mellitus

Conferences
You can meet the editors of Nature Reviews Endocrinology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Endocrinology!

November 2015 marked the 10th anniversary of the launch of Nature Reviews Endocrinology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of endocrinology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrendo10

© 2015 Macmillan Publishers Limited. All rights reserved

GASTROENTEROLOGY & HEPATOLOGY
DECADE IN REVIEW—HCV
Hepatitis C therapy—a fast and competitive race
Stefan Zeuzem
The past 10 years have witnessed incredible developments in the treatment of hepatitis C. From an era in
which the standard of care was PEG-IFN and ribavirin back in 2004, various interferon-free regimens with
direct‑acting antivirals are now available or will be soon. Here, major milestones in the hepatitis C treatment
revolution are outlined.
Zeuzem, S. Nat. Rev. Gastroenterol. Hepatol. 11, 644–645 (2014); published online 16 September 2014; doi:10.1038/nrgastro.2014.164
Between 2004 and 2014 one of the fastest
revolutions in medicine occurred—the devel-
opment of tolerable, safe and efficacious treat-
ments for the global disease of hepatitis C.
According to the WHO, >170 million people
are infected with HCV, which causes a slowly
fibrosing hepatitis that can subsequently
lead to cirrhosis and its sequelae. Moreover,
HCV is one of the most carcinogenic viruses;
the incidence of hepatocellular carcinoma
(HCC) in patients with HCV-associated liver
c­irrhosis ranges between 2% and 5%.1
‘‘
…a race similar to Formula 1
in terms of speed and competition
started…
’’
Major basic scientific breakthroughs were
published before 2004, including the develop-
ment of an in vitro replicon system, the crys-
tallized structure of the NS3/4A protease and
numerous papers elucidating the biological
life cycle of HCV (reviewed by Lange et al.1).
Antiviral treatment of the disease at this
time, however, was still based on PEG-IFN
and ribavirin, which is poorly tolerable and
can cause substantial adverse effects. The
develop­ment and successful clinical appli-
cation of the first specific and potent direct-
acting antiviral agent (DAA), the NS3/4A
protease inhibitor ciluprevir, marked the first
major milestone of the past decade.2
Subsequent years were characterized by the
clinical development of the first-generation
NS3/4A protease inhibitors telaprevir and
boceprevir, with the publication of phase II
and III data in 2009 and 2011, respectively.
Owing to cardiotoxicity, the ciluprevir
programme was not pursued. Additional
NS3/4A protease inhibitors and DAAs of
other drug classes—mainly nucleoside
37  |  NOVEMBER 2015  www.nature.com/reviews
and non-nucleoside RNA-polymerase
inhibitors and NS5A inhibitors—were
developed and clinically investigated in
combination with PEG-IFN and ribavirin.
Inter­mittently, patients with chronic hepati-
tis C were enrol­led into hundreds of clinical
trials explor­ing several dozen DAAs. The
majority of these compounds failed either
because of i­nsufficient antiviral activity or
inappropriat­e safety, or both.
In the middle of the past decade, two
major questions dominated the field. In
2009, Ge et al.3 reported that a genetic vari-
ation in IL28B predicts HCV clearance in
patients infected with genotype 1 treated with
PEG-IFN and ribavirin. The potential role of
this genetic polymorphism in patients receiv-
ing triple therapy, comprising PEG-IFN,
ribavirin and a DAA, was heavily debated.
Step by step, data showed that the predic-
tive role of the IL28B polymorphism in triple
therapy is attenuated in previously untreated
patients and almost extinguished in previous
n­onresponders to PEG-IFN and ribavirin.
The discussion around the second ques-
tion was initially very emotional—that is,
can HCV be eradicated by interferon-free
regimens using only DAAs? Given that the
majority of leading hepatologists was con-
vinced that interferon is required for HCV
eradication, pharmaceutical com­p anies
and regulatory agencies were hesitant to
initiate the first all-oral interferon-free
treatment regimen.
Consequently, it was neither in the USA
nor in Europe, but in New Zealand that
the first such trial took place. A total of 40
previously untreated patients with HCV
genotype 2 or 3 infection were randomly
assigned to four groups; all four groups
received sofosbuvir (at a dose of 400 mg once
daily) plus ribavirin for 12 weeks. Three of
© 2015 Macmillan Publishers Limited. All rights reserved
these groups also received PEG-IFN‑α2a
for 4, 8 or 12 weeks. Of the 40 patients, all
30 (100%) who received sofosbuvir plus
ribavirin for 12 weeks and interferon for
4, 8, or 12 weeks and all 10 (100%) who
received sofosbuvir plus ribavirin without
interferon achieved a sustained virologic
response (SVR).4 When the data were first
presented by Professor Gane from Auckland
at the Annual Meeting of the American
Association for the Study of Liver Diseases
(AASLD) in November 2011 the audience
was stunned and unable for some minutes
after the presentation to ask questions and
start a scientific discussion.
The possibility of eradicating HCV with­
out interferon was confirmed in early 2012
by the combination of the NS3/4A protease
inhibitor asunaprevir and the NS5A inhibi-
tor daclatasvir in patients infected with
geno­type 1; however, major differences in
SVR rates between HCV-1a and HCV-1b
were observed, indicating differences in
the genetic barrier of HCV genotypes
or subtypes.5
Thus, from 2011 onwards a race similar to
Formula 1 in terms of speed and competition
started in the development of interferon-­
free, all-oral combination therapies for
hepatitis C. Generally, two distinct
strat­egies have been pursued:
regimens either with or
without a nucleoside NPG
polymerase inhibi-
tor. A potent and

safe nucleoside polymerase inhibitor has
resembled the Holy Grail of hepatitis C
treatment. Sofosbuvir (invented by the
small biotech company Pharmasset, which
was acquired by Gilead [Foster City, CA,
USA] for >US$11 billion in 2012) fulfilled all
major criteria including potent antiviral and
pangenotypic activity, a high barrier to resist-
ance, a good safety and tolerability profile,
and only minor issues with drug–drug
interactions. Sofosbuvir has been success­
fully combined in a step-up process with
NS3/4A protease inhibitors, NS5A inhibi-
tors, and even non-nucleoside polymerase
inhibitors plus or minus ribavirin. A fixed
drug combination consisting of sofosbuvir
and the NS5A inhibitor ledipasvir has com-
pleted phase III trials in treatment-naive and
treatment-­experienced patients.6,7 SVR rates
>95% were achieved and this combination
will be approved in 2014 both in the USA and
Europe. Other nucleoside polymerase inhibi-
tors have failed owing to either insufficient
antiviral activity (for example, valopicitabine
and mericitabine) or major safety issues
(balapiravir and BMS‑986094). Yet others
are still in very early clinical development
(ACH‑3422, IDX21437 and IDX21459).
The second strategy is to combine all drug
classes except for nucleoside polymerase
inhibitors in a step-down process—that is,
NS3/4A protease inhibitors, NS5A inhibi-
tors, non-nucleoside polymeras­e in­hibitors
and ribavirin. The combination of r­itonavir-
boosted ABT‑450, with ombitasvir, dasa­
buvir and ribavirin achieved SVR rates
>95% both in treatment-naive and treat-
ment-experienced­patients with chronic
hepatitis  C. 8,9 Additional studies have
demonstrated that ribavirin is dispensable
from this regimen in patients infected with
subtype HCV-1b. Furthermore, high SVR
rates have been reported with a two-drug
combination of second-generation NS3/4A
protease and NS5A inhibitors. Ironically,
the company that developed the first potent
and specific DAA against hepatitis C2 and
initiated one of the first large all-oral combi­
nation treatment studies10 announced in
June 2014 that they have withdrawn from
further development of this regimen owing
to sub­optimal efficacy and tolerability in the
fierce competitive environment.
Generally, both strategies might ultimately
lead to two-drug combinations without riba-
virin with treatment durations of 8–12 weeks.
The addition of a third compound might
enable a further decrease in treatment dura-
tion, at least in patients without liver cir-
rhosis. With improved antiviral potency,
A DECADE IN MEDICINE NOVEMBER 2015  |  38
GASTROENTEROLOGY & HEPATOLOGY
broad genotypic activity, and high barriers to
resistance of nucleoside polymerase inhibi-
tors, NS3/4A inhibitors, NS5A inhibitors and
even potentially non-nucleoside polymerase
inhibitors, the development of simple combi-
nation therapies suitable for most (standard)
patients is on the horizon. This development
would be one of the first examples in modern
medicine in which a strongly personalized
treatment approach might be overcome by a
simplified regimen.
Taken together, various very potent, safe,
and tolerable all-oral regimens are already
or will soon be approved in many countries.
But will the drugs be accessible to patients?
The price of these treatments might be pro-
hibitive even in the Western world. A major
debate has started in the USA about the
US$1,000 price tag for a 400 mg tablet of
sofosbuvir. The discussion is very emotional
and frequently not based on the issues that
matter: the price per cure; the cost of total
management including adverse effects; and
the reduced progression rate toward cir-
rhosis, liver failure and HCC (that is, the
overall impact on morbidity and mortality).
Fortunately, drug companies such as Gilead
have already made agreements with govern-
ments of developing countries to provide
anti-HCV drugs at low prices.
The next big question in the field to come
is whether this infectious disease could be
(globally) eradicated exclusively by preven-
tion and antiviral therapy without an avail-
able vaccine. The answer to this exciting
question, however, will take many more
years to come.
DECADE IN REVIEW—HEPATOCELLULAR CARCINOMA
HCC—subtypes, stratification
and sorafenib
Gregory J. Gores
The past 10 years have represented a whirlwind of activity with regard
to information on the risk factors, aetiopathogenesis, diagnosis and
treatment of hepatocellular carcinoma. I will describe what I consider
to be the major advances from both a tumour biology perspective and
a clinical perspective over the past 10 years.
Gores, G. J. Nat. Rev. Gastroenterol. Hepatol. 11, 645–647 (2014); published online 23 September 2014;
doi:10.1038/nrgastro.2014.157
We are beginning to obtain insights into
the initiating steps of hepatocellular carci-
noma (HCC) and to understand the tumour
biology of this lethal malignancy. We have
learnt that polymorphisms in the EGF gene
© 2015 Macmillan Publishers Limited. All rights reserved
Department of Medicine, J. W. Goethe
University Hospital, Theodor-Stern-Kai 7,
60590 Frankfurt am Main, Germany.
zeuzem@em.uni-frankfurt.de
Competing interests
The author acts as a consultant for Abbvie,
Boehringer Ingelheim, Bristol-Myers Squibb, Gilead,
Idenix, Janssen, Merck, Novartis, Roche, Santaris
and Vertex. He is a member of the speakers’ bureau
for Abbvie, Bristol-Myers Squibb, Gilead, Janssen,
Merck and Roche.
1. Lange, C. M., Jacobson, I. M., Rice, C. M.
& Zeuzem, S. Emerging therapies for the
treatment of hepatitis C. EMBO Mol. Med. 6,
4–15 (2014).
2. Lamarre, D. et al. An NS3 protease inhibitor
with antiviral effects in humans infected with
hepatitis C virus. Nature 426, 186–189
(2003).
3. Ge, D. et al. Genetic variation in IL28B predicts
hepatitis C treatment-induced viral clearance.
Nature 461, 399–401 (2009).
4. Gane, E. J. et al. Nucleotide polymerase
inhibitor sofosbuvir plus ribavirin for
hepatitis C. N. Engl. J. Med. 368, 34–44
(2013).
5. Lok, A. S. et al. Preliminary study of two antiviral
agents for hepatitis C genotype 1. N. Engl. J. Med.
366, 216–224 (2012).
6. Afdhal, N. et al. Ledipasvir and sofosbuvir for
untreated HCV genotype 1 infection. N. Engl.
J. Med. 370, 1889–1898 (2014).
7. Afdhal, N. et al. Ledipasvir and sofosbuvir for
previously treated HCV genotype 1 infection.
N. Engl. J. Med. 370, 1483–1493 (2014).
8. Feld, J. J. et al. Treatment of HCV with
ABT‑450/r-ombitasvir and dasabuvir with
ribavirin. N. Engl. J. Med. 370, 1594–1603
(2014).
9. Zeuzem, S. et al. Retreatment of HCV with
ABT‑450/r-ombitasvir and dasabuvir with
ribavirin. N. Engl. J. Med. 370, 1604–1614
(2014).
10. Zeuzem, S. Faldaprevir and deleobuvir for HCV
genotype 1 infection. N. Engl. J. Med. 369,
630–639 (2013).
(encoding epidermal growth factor), which
are associated with alterations in expres-
sion, increase the risk of HCC development
in patients with cirrhosis.1 The association
between EFG expression and HCC risk has
GASTROENTEROLOGY & HEPATOLOGY
Immune
therapies
2 conclusion. Experience from an established
Genetic
stratification
1 ing the radiographic response to determine
Figure 1 | Future strategies for HCC diagnosis and treatment. © Mayo Clinic.
resulted in ongoing clinical studies using
EGF receptor inhibitors for chemopreven-
tion of HCC in high-risk patients. Despite
the tremendous heterogeneity of the genetic
aberrations in advanced HCC, it is likely that
high-frequency somatic mutations in the
telomerase reverse-transcriptase promoter
drive early oncogenesis of this neoplasm.2 In
2008, an important observation was made
by examining gene expression in fixed tissue
samples from patients with HCC largely due
to HCV infection.3 A reproducible genetic
expression signature from nonmalignant
liver tissue in patients with HCC was identi­
fied, which predicted late development of
recurrent HCC following hepatic resection.
The nonmalignant genetic signature sup-
ports the concept that there is a hepatic field
defect (the concept that cancer results from
tissue alterations throughout the liver as
opposed to somatic mutations within a small
group of cells) in patients who develop HCC.
Although molecular profiling has yet
to be useful in selecting treatment strate-
gies for HCC, it has become increasingly
useful for prognostication of patients with
HCC. An HCC five‑gene score, identified
in 2013, was rigorously validated and asso-
ciated with survival of patients after liver
resection for HCC.4 The five‑gene score
was based on a combination expression
profile of the following genes: HN1, RAN,
RAMP3, KRT19 and TAF9. This profile
was identified empirically and the genes are
not functionally related. Disease-specific
survival was predicted by combining data
on micro­vascular invasion, the Barcelona
Clinic Liver Cancer classification and the
five‑gene score. Thus, for the first time,
39  |  NOVEMBER 2015  www.nature.com/reviews
Early diagnosis and
chemoprevention
3 diagnosis of HCC by noninvasive criteria.
we have a potential biomarker to stratify
patients for various therapeutic decisions. In
2014, work suggested that KRT19 expression
can be used to identify a subset of patients
who have HCC with a poor prognosis. 5
KRT19 is a chol­angiocyte marker, but can
be expressed by a subset of HCC cells either
due to cell differentiation, cell plasticity, or
a common progenitor cell of origin between
hepatocytes and cholangiocytes.5 Indeed, it
has been suggested that the subset of carci-
nomas that are KRT19 positive should be
considered as a separate subtype of HCC
and perhaps be managed differently.5 The
genetics involved in the pathogenesis of
HCC and those biomarkers that portend a
bad prognosis are now starting to be under-
stood. Stratification of patients with HCC
will ultimately be very important as we
begin to design p­ersonalized therapies.
A prominent paper in terms of advances
in the clinical management of HCC of the
past decade was the 2006 study demon-
strating that patients infected with HBV
and with elevated serum HBV DNA levels
were at high risk of developing HCC.6 Subse­
quently, a variety of studies have confirmed
this observation, demonstrating that reduc-
tion of HBV DNA levels reduces the risk of
HCC in this patient population; this finding
provides additional impetus for treating
selected patients who have active HBV
infection with antiviral agents. Radiographic
imaging has greatly improved over the past
decade and it is possible to identify increas-
ingly smaller nodules, which might or might
not be early liver cell cancers. A rigorous
study was published in 2008 indicating
that noninvasive techniques can be used
© 2015 Macmillan Publishers Limited. All rights reserved
to identify patients with HCC between
10–20 mm in diameter, 7 substantiating
the existing practice guidelines for early
The response to loco­regional therapy has
been assumed to relate to overall survival.
However, this concept remained more of a
supposition than it did an evidence-­based
treatment centre showed that the radio-
graphic response to loco­regional therapy
does predict patient survival.8 Memon and
colleague’s 2011 study gave further credence
for the use of loco­regional therapy as treat-
ment for patients with HCC and in monitor-
further locoregional treatment over time.
Liver transplantation for patients with
HCC also remains an area of active inter-
est. However, most issues relate to organ
shortage and how the stewardship of this
rare resource can be counterbalanced by
the potential benefit to the patient. A major
analysis was put forward in 2009 by the origi-
nal author of the Milan criteria identifying a
subgroup of patients who might do well with
liver transplantation but who have cancers
beyond the Milan criteria.9 A more precise
estimation of survival was suggested by using
the up-to-seven criteria. The total tumour
diameter of up to 7 cm in size (for example,
seven small tumours each 1 cm or a single
lesion of 7 cm) would be predicted to iden-
tify patients who would benefit s­ubstantially
from liver transplantation.
The manuscript that would probably
end up on everybody’s list for advances in
the past decade was the results of the 2008
study demonstrating a survival benefit for
patients treated with sorafenib in advanced
HCC.10 The paper by Llovet et al. was a land-
mark in so many different ways and from so
many different perspectives. First, the data
illustrated that the sole focus on tumour
response rates might not tell the entire story.
Only a few percent of the patients had classic
response rates (diminution in tumour size).
Rather, therapy worked by delaying progres-
sion of the disease and so time-to-event has
become a major end point in oncological
studies. Second, sorafenib was the first sys-
temic therapy to conclusively demonstrate a
survival benefit for HCC in a randomized,
placebo-controlled trial. Finally, no other
chemotherapeutic agent has yet matched
the results obtained with sorafenib. Thus,
sorafenib continues to stand alone as a thera-
peutic agent for HCC and we have yet to
understand precisely how and why certain
patients benefit from sorafenib therapy.

Looking back is easy—looking forward
is more challenging, but more intriguing
(Figure 1). We are beginning to understand
the heterogeneity of HCC, which clinicians
should be able to use to stratify patients for
clinical trials. I believe that approaches tar-
geting specific subtypes of HCC will prove
to be important, such as different therapies
for those who are negative or positive for
KRT19 expression. I also believe that we
need further studies using immunologi-
cal therapies for the treatment of HCC and
more attention should be paid to therapies
focused on the tumour microenvironment.
Hopefully in another 10 years I will be asked
to write a similar article as this one, which I
would hope will highlight more impressive
therapeutic advances.
Division of Gastroenterology and Hepatology,
Mayo College of Medicine, Mayo Clinic, 200
First Street SW, Rochester, MN 55905, USA.
gores.gregory@mayo.edu
Acknowledgements
This work was supported by National Institute
of Health Grant DK59427 and the Mayo Clinic.
Competing interests
The author declares no competing interests.
1. Tanabe, K. K. et al. Epidermal growth factor
gene functional polymorphism and the risk
of hepatocellular carcinoma in patients with
cirrhosis. JAMA 299, 53–60 (2008).
2. Nault, J. C. et al. High frequency of telomerase
reverse-transcriptase promoter somatic
mutations in hepatocellular carcinoma and
preneoplastic lesions. Nat. Commun. 4, 2218
(2013).
3. Hoshida, Y. et al. Gene expression in fixed
tissues and outcome in hepatocellular
carcinoma. N. Engl. J. Med. 359, 1995–2004
(2008).
4. Nault, J. C. et al. A hepatocellular carcinoma
5-gene score associated with survival of
patients after liver resection. Gastroenterology
145, 176–187 (2013).
5. Govaere, O. et al. Keratin 19: a key role player
in the invasion of human hepatocellular
carcinomas. Gut 63, 674–685 (2014).
6. Chen, C. J. et al. Risk of hepatocellular
carcinoma across a biological gradient of
serum hepatitis B virus DNA level. JAMA 295,
65–73 (2006).
7. Forner, A. et al. Diagnosis of hepatic nodules
20 mm or smaller in cirrhosis: prospective
validation of the noninvasive diagnostic criteria
for hepatocellular carcinoma. Hepatology 47,
97–104 (2008).
8. Memon, K. et al. Radiographic response
to locoregional therapy in hepatocellular
carcinoma predicts patient survival times.
Gastroenterology 141, 526–535 (2011).
9. Mazzaferro, V. et al. Predicting survival after
liver transplantation in patients with
hepatocellular carcinoma beyond the Milan
criteria: a retrospective, exploratory analysis.
Lancet Oncol. 10, 35–43 (2009).
10. Llovet, J. M. et al. Sorafenib in advanced
hepatocellular carcinoma. N. Engl. J. Med. 359,
378–390 (2008).
A DECADE IN MEDICINE NOVEMBER 2015  |  40
GASTROENTEROLOGY & HEPATOLOGY
DECADE IN REVIEW—GUT MICROBIOTA
The gut microbiota era marches on
Francisco Guarner
Research on the gut microbial communities harboured in the human gut
is progressing rapidly owing to the availability of novel and reliable tools
for analysis. Dysfunction of the gut microbiota affects human biological
fitness at multiple levels, and understanding of these effects needs to be
improved to benefit human health.
Guarner, F. Nat. Rev. Gastroenterol. Hepatol. 11, 647–649 (2014); published online 9 September 2014;
doi:10.1038/nrgastro.2014.156
Knowledge on the microbial communities was admitted that 60–80% of the microbes
that naturally inhabit the human gut has observed by microscopic examination of
previously had little effect on clinical medi- a faecal specimen were not recoverable by
cine. The fight against infectious diseases is culture.1 The gold standard for molecular
perhaps the greatest challenge of modern identification of microbial species is the
medicine, and it is not surprising that phylogenetic analysis of genes highly con-
handbooks typically refer to the gut micro- served in all bacteria and archaea: the ribo-
biota as a reservoir of germs with potential somal RNA (rRNA) genes. In particular,
pathogenicity when barrier function fails to the small subunit rRNA gene (16S rRNA)
keep them away. In contrast to this asser- has become the standard in research into
tion, Louis Pasteur, a principal promoter of prokaryotes for determination of phylo­
the germ theory of disease, wrote in 1885 genetic relationships, assessment of diver-
that animals would not be able to survive sity in the environ­ment, and detection and
when totally deprived of ‘common micro­ quantification­of specific populations.
organisms’. Experiments using germ-free In 1999, Jöel Dore’s laboratory in Paris,
facilities were successfully developed during France, examined a single faecal sample of
the 1950s and 1960s and demonstrated that an adult individual by cloning and sequenc-
germ-free animals are difficult to breed, ing 16S rRNA genes, and reported that
have high nutritional requirements in terms only 24% of the sequences could be iden-
of food variety and quantity, and do not tified in gene databases, whilst the major-
develop normal body anatomy and physio­ ity of sequences (76%) would belong to
logy. Microbial colonization of animals species unknown at that time.1 6 years later,
was shown to be critical for normal growth Eckburg et al.2 cloned and sequenced 16S
and development. As such, it is hard to rRNA genes extracted from colonic mucosal
understand why these observations were tissue and faecal samples from three healthy
ignored by medicine for decades, but a adults. A total of 13,355 rRNA gene sequen­
quick overview of leading articles in jour- ces were obtained, and again a majority
nals of most, if not all, medical disciplines of the bacterial sequences corresponded
makes it easy to predict that understanding to uncultivated species (80%) and novel
the structure and function of the human microorgan­isms (62%). They found 150–
symbiont communities might become the 300 bacterial phylotypes per participant,
first great b­reakthrough of twenty-first and 5–200 clones per phylotype. The term
century medicine.
This expectation is supported by the
availability of novel tools for analysis
G
NP
of the gut microbiota. Culture-
independent approaches are
now being used to investi­gate
microbial ecosystems, com-
bining molecular sequencing
of nucleic acids (DNA, RNA)
with powerful bioinformat-
ics for taxonomic identi-
fication and comparative
analysis of large datasets.
Before these approaches, it
© 2015 Macmillan Publishers Limited. All rights reserved
GASTROENTEROLOGY & HEPATOLOGY
phylotype is commonly used instead of
species as this method of taxonomic classi-
fication is not based on phenotypic charac-
teristics of the microbe but on phylogenetic
analysis of 16S rRNA sequences (cut-off
values of 98% identity are arbitrarily used to
delimit species). Around 3 million aligned
and annotated 16S rRNA sequences are now
available as part of the Ribosomal Database
Project. This figure is a dramatic increase
from the 79,000  and 200,000 a­v ailable
sequences in 1999 and 2004, respectively.
The molecular approach is not limited
to 16S rRNA sequencing. The decreasing
cost and increasing speed of DNA sequenc-
ing (next-generation sequencers skip the
cloning step), coupled with advances in
computational analyses of large datasets,
have made it feasible to analyse entire
genomes with reasonable coverage. The
resulting information describes the collec-
tive genetic content of the community from
which functional and meta­bolic networks
can be inferred. Importantly, whole-genome
sequencing provides information about
nonbacterial members in the community,
including viruses, yeasts and protists. The
MetaHIT project addressed this approach.
Full metagenomic analysis of faecal samples
from a cohort of European adults (n = 124)
identified a total of 3.3  million non­­­
redundant microbial genes.3 The majority
(99%) of the identifiable genes were bac­
terial, with a small proportion of virus-like
genes (bacterio­phages and viruses), isolated
DNA (plasmids) and yeasts. Each individual
carries an average of 600,000 nonredundant
microbial genes in the gastro­intestinal tract,
and around 300,000 microbial genes are
common in the sense that they are present
in about 50% of indivi­duals. By merging
data from 1,267 faecal samples from indi-
viduals from Europe, North America and
China, 4 the catalogue of the human gut
microbiome now, as of 2014, incorporates
up to 10 million non­redundant genes and,
interestingly, the core number of genes
present in >50% of the study participants
remained below 300,000, as in the original
MetaHIT catalogue.
Bacteroides, Faecalibacterium and Bifido­
bacterium are the most abundant genera, but
their relative proportion was highly variable
across individuals in the cohort.5 Network
analysis of genus abundance across different
individuals suggested that the gut microbial
ecosystem conforms to well-balanced host–
microbial symbiotic states driven by groups
of co-occurring gen­era. Published in 2011,
a cluster analysis of gut microbial sequences
41  |  NOVEMBER 2015  www.nature.com/reviews
from American, European and Japanese
individuals identi­fied three balanced states
or models, which were designated as ‘entero-
types’.5 Each of the three enterotypes is iden-
tifiable by the variation in the levels of one
of three genera: Bacteroides (enterotype 1),
Prevotella (entero­t ype  2) and Rumino­
coccus (enterotype 3). The discreteness of
these e­nterotypes is debated; some datasets
support the existence of these categories,
whereas others do not. Nonetheless, the
inverse relationship between Prevotella and
Bacteroides has been reproduced in studies
comparing the gut microbiota of residents
of agrarian societies with that of residents
of industrialized societies. Moreover, a study
exploring the associations between diet and
gut microbiota composition, based on food
frequency questionnaires collected over long
periods, indicates that diet affects the pro-
portions of Prevotella versus Bacteroides in
US populations.6 Thus, the presence of stable
gut microbial communities can be linked to
long-term dietary patterns.6
‘‘
…understanding … human
symbiont communities
might become the first great
breakthrough of twenty-first
’’
century medicine
A number of studies have associated
increased microbial richness, at either the
taxonomic or gene level, with diets high
in fruits, vegetables and fibre.7,8 Moreover,
diets high in fruits, vegetables and fibre
were linked to reduced levels of frailty in
elderly individuals (mean age 78 years, age
range 64–102 years),7 whereas low microbial
richness has been associated with obesity,9
insulin resistance, 8 dyslipidaemia, 8 low-
grade inflammation8 and IBD.3 A propor-
tion of Europeans (23%) exhibit microbial
gene counts below the previously estab-
lished median of 600,000.8 Individuals with
low microbial gene counts (below 480,000)
are characterized by more marked overall
adipos­ity, insulin resistance, leptin resist-
ance, dyslipidaemia and a more pronounced
inflammatory phenotype when compared
with individuals with high gene counts.8
These metabolic parameters are slightly
altered even in otherwise healthy individu-
als with low microbial gene counts. Obesity
is associated with a gut microbiota that is
highly efficient for harvesting energy from
the diet.9 In addition, the subset of obese
individuals with low gene counts gain more
© 2015 Macmillan Publishers Limited. All rights reserved
weight over time and have a propensity
towards severe forms of obesity than individ­
uals with high gene counts.8 Low gene rich-
ness therefore seems to be a risk factor for
the development of metabolic-­syndrome-
related complications (type 2 diabetes,
hepatic and cardiovascular patho­logies) as
well as inflammatory bowel conditions. The
effectiveness of faecal microbiota transplan-
tation in specific disease states (colitis asso-
ciated with Clostridium difficile infection,
type 2 diabetes) is proof of principle that
defects of the gut microbial e­cosystem might
be the origin of certain disorders.
Diet can alter the functional metabolism
of the gut microbiome. Food ingredients are
potential substrates for microbial metabo-
lism, which then could produce substances
or molecules that affect host physiology.
For instance, indigestible carbohydrates
in the diet are fermented by the gut micro­
biota to produce short-chain fatty acids with
a number of beneficial functions for the
host. Characterizing metabolic interactions
between host and microbes is an important
challenge for understanding human physio­
logy and predicting risk of non­communicable
diseases. An example of this approach is the
discovery that gut microbes can contribute
to the development of athero­s clerosis by
producing metabolites of dietary carnitine or
lecithin.10 Catabolism of these compounds by
the gut microbiota results in the formation
of trimethylamine, which is metabolized by
the liver into t­ rimethy­lamine-N-oxide. This
small molecule is strongly associated with
an increased risk of adverse cardiovascular
events in humans.10
The gut microbiota is unequivocally inte-
grated into the regulation of multiple host
pathways, giving rise to interactive host–
microbiota metabolic and immunological–­
inflammatory axes. Knowledge of the
struc­­ture and functions of the microbial
communities is needed for understanding
human physiology and optimizing strat­
egies to combat disease so that the impor-
tance of the gut microbiota is not ignored in
the next 10 years.
Digestive System Research Unit, University
Hospital Vall d’Hebron, CIBEREHD, Passeig Vall
d’Hebron 119‑129, Barcelona E08035, Spain.
fguarner@telefonica.net
Competing interests
The author declares no competing interests.
1. Suau, A. et al. Direct analysis of genes
encoding 16S rRNA from complex communities
reveals many novel molecular species within
the human gut. Appl. Environ. Microbiol. 65,
4799–4807 (1999).

2. Eckburg, P. B. et al. Diversity of the human
intestinal microbial flora. Science 308,
1635–1638 (2005).
3. Qin, J. et al. A human gut microbial gene
catalogue established by metagenomic
sequencing. Nature 464, 59–65 (2010).
4. Li, J. et al. An integrated catalog of reference
genes in the human gut microbiome.
Nat. Biotechnol. 32, 834–841 (2014).
5. Arumugam, M. et al. Enterotypes of the human
gut microbiome. Nature 473, 174–180 (2011).
6. Wu, G. D. et al. Linking long-term dietary
patterns with gut microbial enterotypes.
Science 334, 105–108 (2011).
DECADE IN REVIEW—FGIDS
‘Functional’ gastrointestinal
disorders—a paradigm shift
Nicholas J. Talley
In the past decade we have witnessed an explosion in the quantity and
quality of research in the functional gastrointestinal disorders. I discuss
10 top original research papers that, unless recent, have been highly
cited, published in a high-impact journal and have probably shifted
thinking in the field.
Talley, N. J. Nat. Rev. Gastroenterol. Hepatol. 11, 649–650 (2014); published online 23 September 2014;
doi:10.1038/nrgastro.2014.163
Post-dysenteric IBS was described by
Chaudhary and Truelove in 1962—the initial
clue that IBS might be an organic gut disease.
Although important work has further char-
acterized post-infectious IBS, Mearin and
colleagues’ prospective study1 from 2005
identified convincingly an increased inci-
dence of dyspepsia, IBS and an overlap syn-
drome (dyspepsia–­IBS) 6 and 12 months
after a salmonella outbreak; patients with
more severe gastro­enteritis were at increased
risk of developing one of these syndromes.1
The overlap of functional gastrointestinal
dis­orders (FGIDs) has remained mysteri-
ous but these key data suggest that infection
might explain it. Salmonella spp. induces
small-intestinal inflammation; a testable
hypothesis is that perhaps the site of inflam-
mation accounts for the symptom phenotype,
for example proximal inflammation might
be associated with dyspepsia, distal inflam-
mation with IBS and more-extensive
i­nflammation with the overlap syndrome.
How might inflammation induce the
symptoms of FGIDs? And what about
the mysterious extraintestinal symptoms
these patients often report? Liebregts et al.2
tested whether the release of proinflamma-
tory cytokines in IBS is linked to symptoms
and psychiatric comorbidity. They found
that a number of cytokines were increased
A DECADE IN MEDICINE NOVEMBER 2015  |  42
GASTROENTEROLOGY & HEPATOLOGY
7. Claesson, M. J. et al. Gut microbiota
composition correlates with diet and
health in the elderly. Nature 488, 178–184
(2012).
8. Le Chatelier, E. et al. Richness of human gut
microbiome correlates with metabolic markers.
Nature 500, 541–546 (2013).
9. Turnbaugh, P. J. et al. A core gut microbiome
in obese and lean twins. Nature 457, 480–484
(2009).
10. Tang, W. H. et al. Intestinal microbial
metabolism of phosphatidylcholine and
cardiovascular risk. N. Engl. J. Med. 368,
1575–1584 (2013).
in IBS, including TNF and IL‑6, and that
lipopolysaccharide-induced TNF levels cor-
related with levels of anxiety.2 These data
further challenge the concept that IBS is a
functional disease (a term that means no
organic, metabolic or biochemical abnormal-
ities), and suggest that gut-to-brain pathways
might be one driver of psycho­logical symp-
toms in IBS. Could psycho­logical distress in
IBS be relieved or even cured by a gut-based
intervention that down­regulates the cytokine
response? Meanwhile, the observation that
increased numbers of mast cells are present
in IBS dates back to the 1990s, but it has
been unclear what role mast cells have in the
generation of pain and other symptoms, if
any. In elegant experiments utili­zing colonic
biopsy samples from patients with IBS,
Barbara et al.,3 in 2004, observed a substan-
tial increase in the number of degranulating
mast cells releasing histamine and, strik-
ingly, mast cells observed in close proximity
to nerves correlated with abdominal pain in
IBS. The role of blocking mast cells, and
in particu­lar histamine, in IBS is now an area
of active interest.
Intense interest in the role of the gut
microbiota in gastrointestinal and non-
gastrointestinal disease continues, and
the hypothesis that the colonic and small
intesti­nal flora is altered in IBS is being
© 2015 Macmillan Publishers Limited. All rights reserved
actively explored. The study by Rajilic-
Stojanovic et al.4 is a technically excellent
example of the current state of the art, dem-
onstrating that although stool microbiota
composition is highly individual, it can be
used to discriminate IBS from health, with
reduced numbers of Bifidobacteriaceae,
Actinobacteria and Bacteroidetes but
increased levels of Firmicutes in patients
with IBS compared with healthy individu-
als. Problems remain, including limited
sampling techniques, lack of an ability to
culture most bacteria, and the fact that major
differences have been observed in different
studies. Regardless, bacterial manipulation
might be one of the keys to curing FGIDs.
If bacteria are important in symptom gen-
eration in IBS—for example by increased
gas production that distends the lumen and
induces bloating—would starving them
work? In their landmark paper from 2008,
Shepherd and colleagues5 tested dietary fruc-
tose and fructan restriction then rechallenge
in patients with IBS; 70% of patients receiv-
ing fructose, 77% receiving fructans and
79% receiving a mixture reported symptom
breakthrough versus 14% of patients ingest-
ing glucose. Other data now support the
benefit of a low-FODMAP diet in IBS, which
is being increasingly adopted although large
r­andomized blinded trials are needed.
No drug has been shown to change the
natural history of IBS (symptoms usually
rapidly reoccur) until the past decade. If
the gut microbiota is important in IBS,
manipulating it should change the natural
history of the condition at least temporarily.
In 2011, Pimentel et al.6 tested rifaximin, a
minimally absorbed antibiotic, for 2 weeks
in patients with diarrhoea-predominant and
mixed IBS with 10 weeks follow-up. Overall,
global IBS symptoms and, most notably,
bloating improved although the therapeu-
tic gain was modest (~10% of patients).
The striking observation was that symptom
improvement persisted during follow-up.6 It
is unknown why only a subgroup responded
and whether a more targeted attack on the
most relevant bacterial group would yield
better results. The response is known not to
be sustained long term and this factor is also
unexplained; is this because of microbial or
possibly mucosal factors? Probiotics have
also shown modest symptom benefits in IBS
(including a Bifidobacteria infantis strain)
but optimal probiotic therapy remains to
be defined and benefits to date are small
although this area has enormous potential.7
At last a molecular defect in IBS has been
uncovered by Coates et al.8 and these results,
GASTROENTEROLOGY & HEPATOLOGY
Gut−brain axis
(IBS)
Duodenal
eosinophilia
(functional
dyspepsia)
Intestinal
inflammation
(IBS)
Changes in
gut microbiota
(IBS)
Figure 1 | Functional’ gut disorders. Many
patients have objective gut pathology and
a gut-to-brain syndrome that explains their
gastrointestinal as well as extraintestinal
symptoms (e.g. anxiety, fatigue and
sleep disturbance).
although initially controversial, have been
confirmed. Serotonin (5-HT) is released from
enterochromaffin cells and mediates peristal-
sis and reflex responses. In this important
paper, defects in serotonin signalling (includ-
ing serotonin reuptake) in IBS and ulcerative
colitis were observed (although numbers of
enterochromaffin cells were only reduced in
colitis). Gut-specific molecular alterations
in serotonin might explain constipation
(insufficient 5HT), diarrhoea (excess 5HT)
and mixed IBS (5HT switching between
excess and insufficient levels).
Finally, two key papers indicate a para-
digm shift in adult functional dyspepsia.
The Kalixanda study 9 was a landmark
endoscopic population-based study from
northern Sweden that investigated whether
duodenal eosinophilia characterises ‘pure’
functional dyspepsia (no overlap of IBS was
permitted). Patients with increased numbers
of duodenal eosinophils had a nearly 12-fold
increased risk of functional dyspepsia, and
the pathology was linked to early satiety, a
finding subsequently confirmed by multiple
studies. A Belgian study 10 from 2014 not only
confirmed duodenal eosinophilia in func-
tional dyspepsia but also showed increased
numbers of mast cells (arguably attribut-
able to the majority of the study population
having both dyspepsia and IBS) and more
43  |  NOVEMBER 2015  www.nature.com/reviews
importantly impaired duodenal permeability.
Increased antigen presentation (from food
or bacteria) might be key in driving a subtle
upper intestinal inflammatory response in
predisposed patients, leading to duodeno­
gastric reflex responses and early satiety, a
c­ondition we label as ‘functional dyspepsia’.
Our concepts have changed over the
past decade and the short-comings of a
symptom-­b ased only classification of
FGIDs are becoming increasingly apparent
(Figure 1). IBS is ‘disappearing’ as we iden-
tify more and more subgroups (and recog-
nize organic diseases such as coeliac disease,
microscopic colitis or bile-salt-induced diar-
rhoea as clinically indistinguishable from
IBS). Accumulating data suggest people
might have either a gut-to-brain or brain-to-
gut syndrome, both with identical gut symp-
toms. One could make the case it is now time
to drop the arguably pejorative term func-
tional; on the basis of the latest information,
many patients have gut patho­logy. Identifying
serum and stool biomarkers to objectively
diagnose IBS and functional dyspepsia now
seems plausible. Prediction is dangerous but
I will stick my neck out: a purely symptom-
based classification of the FGIDs will be
extinct in the next two decades as knowledge
in this area explodes.
Faculty of Health and Medicine, University of
Newcastle, HMRI Building Lot 1 Kookaburra
Circuit, New Lambton Heights, NSW 2035,
Australia.
nicholas.talley@newcastle.edu.au
DECADE IN REVIEW—PANCREATIC DISEASES
Advances in understanding
and care of pancreatic diseases
Randall E. Brand
Clinical and basic science discoveries over the past decade have led to
considerable improvements in our understanding and care of pancreatic
diseases. Findings reported in 10 key papers highlight results that have
already substantially altered the care of patients with acute pancreatitis,
chronic pancreatitis and pancreatic cancer (or soon will).
Brand, R. E. Nat. Rev. Gastroenterol. Hepatol. 11, 650–652 (2014); published online 16 September 2014;
doi:10.1038/nrgastro.2014.159
Multicentre collaborations have enabled
investigators to perform well-powered, high-
quality studies that have led to major changes
in our approach to the care of patients with
acute pancreatitis, chronic pancreatitis or
pancreatic adenocarcinoma. Researchers
© 2015 Macmillan Publishers Limited. All rights reserved
Competing interests
The author declares no competing interests.
1. Mearin, F. et al. Dyspepsia and irritable bowel
syndrome after a Salmonella gastroenteritis
outbreak: one-year follow-up cohort study.
Gastroenterology 129, 98–104 (2005).
2. Liebregts, T. et al. Immune activation in patients
with irritable bowel syndrome. Gastroenterology
132, 913–920 (2007).
3. Barbara, G. et al. Activated mast cells in
proximity to colonic nerves correlate with
abdominal pain in irritable bowel syndrome.
Gastroenterology 126, 693–702 (2004).
4. Rajilic-Stojanovic, M. et al. Global and deep
molecular analysis of microbiota signatures in
fecal samples from patients with irritable bowel
syndrome. Gastroenterology 141, 1792–1801
(2011).
5. Shepherd, S. J., Parker, F. C., Muir, J. G. &
Gibson, P. R. Dietary triggers of abdominal
symptoms in patients with irritable bowel
syndrome: randomized placebo-controlled
evidence. Clin. Gastroenterol. Hepatol. 6,
765–771 (2008).
6. Pimentel, M. et al. Rifaximin therapy for
patients with irritable bowel syndrome without
constipation. N. Engl. J. Med. 364, 22–32
(2011).
7. O’Mahony, L. et al. Lactobacillus and
bifidobacterium in irritable bowel syndrome:
symptom responses and relationship to
cytokine profiles. Gastroenterology 128,
541–551 (2005).
8. Coates, M. D. et al. Molecular defects in
mucosal serotonin content and decreased
serotonin reuptake transporter in ulcerative
colitis and irritable bowel syndrome.
Gastroenterology 126, 1657–1664 (2004).
9. Talley, N. J. et al. Non-ulcer dyspepsia and
duodenal eosinophilia: an adult endoscopic
population-based case-control study. Clin.
Gastroenterol. Hepatol. 5, 1175–1183 (2007).
10. Vanheel, H. et al. Impaired duodenal mucosal
integrity and low-grade inflammation in
functional dyspepsia. Gut 63, 262–271 (2014).
have further improved the design of studies
through the use of standardized definitions
of diseases and their complications. Over the
past decade, findings have enabled us to move
toward individualizing preventive and treat-
ment strategies on the basis of our growing

understanding of the molecular and genetic
mechanisms of pancreatic disorders, as
shown in Figure 1 (Timeline). In my opinion,
the following 10 studies published over the
past decade best highlight these advances.
Several well-designed prospective, ran-
domized trials by the Dutch Pancreatitis
Study Group have made notable contribu-
tions to the management of patients with
acute pancreatitis. In 2010, one of their trials
changed the approach for treating patients
with necrotizing pancreatitis with infected
necrotic tissue, a major cause of death for
patients with acute pancreatitis.1 Their res­
ults demonstrated that a minimally invasive
step-up approach (percutaneous drainage
and, if required, a minimally invasive retro-
peritoneal necrosectomy) had a lower rate
of major short-term and long-term com-
plications than the traditional approach of
open necrosectomy.
Patients with severe necrotizing pancrea-
titis have also benefited from research on
the use of prophylactic antibiotics to prevent
polymicrobial pancreatic and peripancreatic
infections. In 2007, results from an inter­
national multicentre randomized, prospec-
tive, double-blind trial comparing the use of
meropenem to placebo in patients with non-
infected necrotizing pancreatitis showed no
statistically significant difference between the
two treatment groups.2 This finding began
the reversal of routine use of broad-spectrum
antibiotic prophylaxis.
In the USA, a large multicentre trial con-
firmed the effectiveness of a strategy to
prevent endoscopic retrograde cholangio-
pancreatography (ERCP)-induced pancrea-
titis.3 Prior to the publication of this study,
no pharmacological prophylaxis was broadly
accepted for the prevention of this poten-
tially life-threatening complication of ERCP.
Results from this randomized, placebo-­
controlled, double-blind study established
the benefits of 100 mg rectal indomethacin in
reducing the incidence and severity of pan-
creatitis after ERCP and led to its routine use
in selected patients at high risk of developing
the condition.
The continued identification of both
environmental and genetic risk factors for
pan­creatic diseases will enable us to focus pre-
ventive strategies in a more directed manner
than previously. For example, one rand-
omized controlled trial demonstrated that
an intervention to help patients reduce their
alcohol consumption given during their
first hospitalization for alcohol-related
acute pancreatitis and then repeated every
6 months at an outpatient gastro­intestinal
A DECADE IN MEDICINE NOVEMBER 2015  |  44
GASTROENTEROLOGY & HEPATOLOGY
Adjuvant chemotherapy High alcohol intake shown FOLFIRINOX improves
shown to improve not to be the only cause survival in patients
survival in patients with of chronic pancreatitis with pancreatic cancer
pancreatic cancer
GWAS explains sex
Repeated interventions differences in risk of
alcohol-related chronic
reduce recurrence of
acute pancreatitis pancreatitis
2004 2006 2008 2010 2012 2014
Minimally invasive First accepted
Core signalling
pathways in pancreatic step-up approach treatment to prevent
cancer revealed for treating necrotizing pancreatitis
pancreatitis induced by ERCP
Routine use of broad-spectrum
antibiotic prophylaxis in necrotizing Potential for early detection
pancreatitis begins to decline of pancreatic cancer
Figure 1 | Timeline of the advances in our understanding of pancreatic diseases and
improvements in patient care over the past decade. Abbreviations: ERCP, endoscopic retrograde
cholangiopancreatography; FOLFIRINOX, folinic acid, fluorouracil, irinotecan and oxaliplatin;
GWAS, genome-wide association study.
clinic reduced the incidence of recurrent for observed sex differences in the incidence
acute pancreatitis by two-thirds compared of alcohol‑related chronic pancreatitis.
with receiving the intervention only during The prognosis for patients with pancreatic
the initial hospitalization.4 adenocarcinoma has remained poor over
Indeed, several papers over the past decade the past decade; however, incremental pro-
have elucidated the relationship between gress has been made in the treatment of this
alcohol and chronic pancreatitis. Examining disease. In 2004, results from a multi­centre,
data from 540 patients with chronic pancrea- randomized trial by the European Study
titis who were diagnosed according to strict Group for Pancreatic Cancer that compared
entry criteria and 695 related and unrelated the use of adjuvant chemo­radiation therapy
control individuals, the North American with or without chemotherapy, chemother-
Pancreatic Study Group (20 secondary or apy alone, or no treatment led to a complete
tertiary pancreatic care centres) clarified the reassessment of the clinical utility of radiation
relationship between chronic pancreatitis therapy.7 By 2011, results from a multicentre
and both alcohol and smoking consump- trial from France had shown a consider-
tion.5 The risk of chronic pancreatitis as a able survival advantage for a combi­nation
result of alcohol consumption was dose- chemotherapy regimen consisting of folinic
dependent and occurred only in patients acid, fluorouracil, irinotecan and oxaliplatin
who drank at least five alcoholic drinks (FOLFIRINOX) compared with single-agent
per day. Cigarette smoking was found to gemcitabine.8 In addition to broadening the
be an independent risk factor. These results treatment options for patients with metastatic
changed the dogma that chronic pancreati- pancreatic adenocarcinoma who can tolerate
tis was mainly attributable to alcohol con- the increased toxicity with FOLFIRINOX,
sumption, as only a minority of patients with this study opens the door to new therapeutic
chronic pancreatitis in the study cohort were approaches beyond gemcitabine.
heavy drinkers (38% men and 11% women) With regard to more-personalized thera-
and one-quarter had never consumed alco- pies, our growing knowledge of the genetic
holic drinks. Results from an international alterations involved in carcinogenesis has
genome-wide association study that were led to the development of therapies directed
reported in 2012 might explain the long- against specific biological targets. One com-
recognized sex differences observed in the prehensive genetic analysis of 24 advanced
incidence of alcohol-related chronic pan- pancreatic adenocarcinomas showed the
creatitis.6 This study identified an X‑linked diversity and core pathway changes in these
risk allele, CLDN2 (encodes claudin‑2), pancreatic tumours, which had an average of
which conferred the greatest risk of develop- 63 genetic alterations in the proteins involved
ing alcohol-related chronic pancreatitis. As in 12 core cellular signalling pathways and
men only have one copy of the gene (hemi­ processes.9 These results indicate the impor-
zygous frequency of 0.26) whereas women tance of moving beyond single-agent thera-
have two copies (homozygous frequency of pies to target multiple signalling pathways
0.07), this finding might partially account and processes. A subsequent study, which
© 2015 Macmillan Publishers Limited. All rights reserved
GASTROENTEROLOGY & HEPATOLOGY

sequenced the genomes of pancreatic adeno- DECADE IN REVIEW—IBD

carcinoma metastases from seven individuals
obtained during rapid autopsies, compared
the somatic mutations identified in the
metastases with both the primary tumour
and anatomically distinct metastases to
determine the clonal relationships between
the primary and metastatic cancers.10 These
data were incorporated into a mathemati-
cal model to estimate the time from initial
tumour initiation to development of metasta­
ses, which pointed to an ~15-year window
during which neoplastic cells remain con-
tained within the pancreas. This finding
suggests that the prognosis of patients with
pancreatic adenocarcinoma could be vastly
improved with early detection.
These studies highlight the substantial pro-
gress achieved over the past decade in caring
for patients with pancreatic diseases, with
many new findings that can be translated into
clinical practice over the next decade.
Division of Gastroenterology, Hepatology
and Nutrition, Department of Internal
Medicine, University of Pittsburgh Medical
Center, 5200 Centre Avenue, Suite 409,
Pittsburgh, PA 15232, USA.
reb53@pitt.edu
Competing interests
The author declares no competing interests.
1. van Santvoort, H. C. et al. A step-up approach
IBD—genes, bacteria and new
therapeutic strategies
Jean-Frederic Colombel
IBD is known to be associated with an abnormal response to an
unbalanced gut microbiota in genetically predisposed individuals.
The therapeutic goal now is to control progression of the disease.
Given the heterogeneity of IBD, the two universes of basic and
clinical science must work in parallel to realize the hope of
personalized therapy.
Colombel, J.-F. Nat. Rev. Gastroenterol. Hepatol. 11, 652–654 (2014); published online 14 October 2014;
doi:10.1038/nrgastro.2014.170
In the past decade we have witnessed the prominent Crohn’s disease risk gene
major advances both in basic and clinical ATG16L1, which encodes an adaptor
aspects of IBD. Landmark discoveries in the protein central to autophagosome forma-
fields of genetics, immunology and micro- tion. This finding introduced the idea
biology have provided us with an improved of defective autophagy as a fundamental
understanding of the biological and disease mechanism, which had not previ-
genetic variations associated with disease. ously been considered by mucosal immuno­
An increased knowledge of the natural logists or gastroenterologists. Patients
history of the disease, disease predictive homozygous for the ATG16L1 allele have
factors and disease assessment has led us marked structural alterations in their secre-
to demand better therapeutic outcomes, to tory apparatus along with transcriptional
develop innovative therapeutic strategies abnormalities in Paneth cells.3 Autophagy
and to incorporate new end points in daily serves as a compensatory mechanism for
clinical practice. endoplasmic reticulum stress; when endo-
plasmic reticulum stress in the intestinal

‘‘
or open necrosectomy for necrotizing
pancreatitis. N. Engl. J. Med. 362, 1491–1502 epithelium intersects with an autophagy
(2010). The treatment paradigm has defect, a discontinuous, trans­mural inflam­
2. Dellinger, E. P. et al. Early antibiotic treatment
fundamentally shifted in IBD in mation with fissuring lesions spontan­

for severe acute necrotizing pancreatitis:
a randomized, double-blind, placebo-controlled
study. Ann. Surg. 245, 674–683 (2007).
3. Elmunzer, B. J. et al. A randomized trial of rectal
indomethacin to prevent post-ERCP
pancreatitis. N. Engl. J. Med. 366, 1414–1422
(2012).
4. Nordback, I. et al. The recurrence of acute
alcohol-associated pancreatitis can be
reduced: a randomized controlled trial.
Gastroenterology 136, 848–855 (2009).
5. Yadav, D. et al. Alcohol consumption, cigarette
smoking, and the risk of recurrent acute and
chronic pancreatitis. Arch. Intern. Med. 169,
1035–1045 (2009).
6. Whitcomb, D. C. et al. Common genetic variants
in the CLDN2 and PRSS1-PRSS2 loci alter risk
for alcohol-related and sporadic pancreatitis.
Nat. Genet. 44, 1349–1354 (2012).
7. Neoptolemos, J. P. et al. A randomized trial of
chemoradiotherapy and chemotherapy after
resection of pancreatic cancer. N. Engl. J. Med.
350, 1200–1210 (2004).
8. Conroy, T. et al. FOLFIRINOX versus gemcitabine
for metastatic pancreatic cancer. N. Engl. J. Med.
364, 1817–1825 (2011).
9. Jones, S. et al. Core signaling pathways in
human pancreatic cancers revealed by global
genomic analyses. Science 321, 1801–1806
(2008).
10. Yachida, S. et al. Distant metastasis occurs
late during the genetic evolution of pancreatic
cancer. Nature 467, 1114–1117 (2010).
’’
the past 10 years
Genetic studies culminated in a meta-
analysis of Crohn’s disease and ulcerative
colitis genome-wide association studies
by Jostins et al.1 in 2012. 163 loci associ-
ated with IBD were identified.1 Most loci
contribute to both diseases and many of
them are also implicated in other immune-
mediate­d disorders, such as ankylosing
spondylitis and psoriasis. It is worth men-
tioning that the loci identified to date rep-
resent ~20% of the potential heritable risk
of developing IBD. The ‘missing heritability’
component has led to the conclusion that
some forms of IBD—especially the familial
forms—could be due to the profound effect
of rare variants. For example, a familial
form of early-onset Crohn’s disease has been
identified as a monogenic disorder result-
ing from homozygous mutations in IL10R
(encoding the interleukin 10 receptor).2
Another major contribution from the
field of genetics was the discovery of
eously develops that mimics Crohn’s
disease.4 Auto­phagy also interacts with the
unfolded protein response, a process that
arises from endoplasmic reticulum stress
and manifests in Paneth cells. Deletion of
XPB1, which encodes a transcription factor
important in the unfolded protein response,
results in depletion of Paneth cell numbers
along with spontaneous small intestinal
inflammation that mimics human IBD with
crypt abscesses, neutrophil infiltration and
ulcerations. That these three closely con-
nected p­athways—microbial signalling
(via NOD2), autophagy and endoplas-
mic reticulum stress—affect Paneth cell
function clearly implicates altered bacte-
rial handling as having a primary role in
Crohn’s disease.4
The past decade was also marked by the
emergence of the complex and critical role
of the intestinal microbiome in health and
disease. In 2004, Rakoff-Nahoum et al. 5
were amongst the first to demonstrate the
profound role that signals from the gut
microbiota have in determining intestinal

45  |  NOVEMBER 2015  www.nature.com/reviews

© 2015 Macmillan Publishers Limited. All rights reserved
 GASTROENTEROLOGY & HEPATOLOGY

homeostasis, including protection from Early disease Late disease

injury. Conversely, bacteria-driven innate
colitis, associated with increased produc-
tion of IL‑17 and IFN‑γ in the colon, is
mediated by IL‑23-dependent innate lym-
phoid cells in mice. 6 The key role of the
IL‑23 pathway in chronic inflammation
has been established by the association of
IL23R alleles with IBD, psoriasis and anky-
losing spondylitis.1 The pathogenic and pro-
tective effects of the gut microbiota imply
that an unbalanced microbial community
(dysbiosis) is associated with a dysregulated
immune response. Compared with healthy
individuals as controls, patients with IBD
have fewer bacteria with anti-inflamma-
tory properties and/or more bacteria with
proinflammatory properties. An increased
relative abundance in Enterobacteriaceae,
mostly Escherichia coli, has been observed
in patients with Crohn’s disease and on
the basis of its patho­g enic traits a new
group of E. coli, adherent-invasive­ E. coli,
has been associated with ileal Crohn’s
disease.7 By contrast, members of the phyla
Bacteroidetes and Firmicutes are reduced
in patients with Crohn’s disease; among
Firmicutes, numbers of Faecalibacterium
prausnitzii (which have anti-inflammator­y
properties) are reduced in patients with
Crohn’s disease. Reduced counts of F. praus­
nitzii is associated with increased risk of
post-resection recurrence of ileal Crohn’s
disease. The first study in new-onset
Crohn’s disease confirmed the presence of
dysbiosis in mucosal samples, with marked
interindividual variation.7 Strikingly, the
strongest factor associated with dysbiosis
was previous exposure to antibiotics, which
is consistent with epidemiological observa-
tions that antecedent antibiotic use is a risk
factor for developing Crohn’s disease.
Combination therapy
Combination therapy
+ tight control + tight control
Deep remission
De-escalation Stabilization
Treat-to-target
Figure 1 | A treat-to-target strategy could be applied to all patients with IBD. In early disease, for
patients as defined by disease duration of <18 months, no complications and no previous use
of disease-modifying IBD drugs, the target is deep remission with minimal damage whereas in
late disease the target is to stop disease progression. In both cases, combination therapy is the
most effective and requires tight control of therapeutic efficacy and safety. In early disease and
in certain circumstances, that is when the target has been reached, de-escalation therapy could
be considered.
as endoscopy or biomarkers; the demon- particularly remarkable in patients with
stration that combination therapy using ulcerative colitis.9
an anti-TNF and an immuno­suppressor So what does the future hold? The
is the most effective therapeutic strategy;8 hope is that more clinical applications
and, finally, recognition that treatment will translate from bench to bedside. For
early in the disease course is associated instance, a huge need exists to be able to
with improved efficacy of drugs and better predict disease outcome. One success-
outcomes, such as reduced risk of sur­ ful example is the study from 2011 by Lee
gery. Figure 1 illustrates a ‘treat-to-target’ et al., 10 which reported the first robust
approach that uses the most effective com- and replicated, b­iomarker-based method
bination therapy in patients with objective for predicting severity of Crohn’s disease
inflammation and ‘tight control’ of drug and ulcerative colitis. By assessing the
levels and antibodies. The treat-to-target transcriptome of CD8+ peripheral blood
approach aims either at full control of the T cells, the authors could predict the need
disease as defined by absence of symp- for therapeutic escalation with high speci-
toms, normalization of levels of biomark- ficity and sensitivity. This approach could
ers (C-reactive protein) and endoscopic be used to identify patients at diagnosis
healing (deep remission) in early disease who might benefit most from early (and
or stabili­zation of bowel damage in patients costly) introduction of potent immuno-
with long-standing disease. Nonetheless, therapies. Finally, although most current

‘‘
The past decade was also
marked by the emergence of the
complex and critical role of the
many clinical challenges and gaps in our treatments in IBD work by dampening
knowledge remain, particularly as the the immune system, with unavoidable
current clinical approach is mainly based on adverse effects, there is great hope that
retro­spective data. Prospective studies are re-equilibrating the gut microbiota for

’’
intestinal microbiome…
The treatment paradigm has fundamen-
tally shifted in IBD in the past 10 years.
Major drivers for this evolution exist: the
recognition that IBD is not an intermittent
but rather a progressive disease, leading to
bowel damage and disability in the long
term; availability of disease-modifying IBD
drugs that are able to halt disease progres-
sion; evidence that therapeutic decisions
should be based on objective markers such
needed to validate the concept of treating
beyond symptoms; the ideal target is still
uncertain and could still evolve to become
more stringent (involving histological or
molecular healing) or perhaps more relaxed
than current approaches. Despite close
follow-up and monitoring, many patients
become refractory to the available drugs. In
addition, many promising drugs have failed
to achieve their goals. The notable excep-
tion are drugs targeting integrins, with the
first agent, vedolizumab, being approved
in 2013, for which long-term efficacy is
bene­f icial effects will be possible. For
example, specific components of the diet
have been shown to rapidly and repro-
ducibly alter the human gut microbiota.
Whether this modulation can ultimately
lead to long-term remission and even pre-
vention of disease will be a central question
moving forward.
Department of Medicine, Icahn School
of Medicine at Mount Sinai, New York,
1428 Madison Avenue, NY 10029, USA
jean-frederic.colombel@mssm.edu

A DECADE IN MEDICINE NOVEMBER 2015  |  46

© 2015 Macmillan Publishers Limited. All rights reserved
GASTROENTEROLOGY & HEPATOLOGY
Competing interests
J.-F.C. has served as consultant, advisory board
member or speaker for Abbvie, ABScience, Amgen,
Bristol–Myers Squibb, Celltrion, Danone, Ferring,
Genentech, Giuliani SPA, Given Imaging, Janssen,
Immune Pharmaceuticals, Merck, Millenium
Pharmaceuticals, Nutrition Science Partners, Pfizer,
Prometheus Laboratories, Protagonist Therapeutics,
Receptos, Sanofi, Schering Plough, Second Genome,
Takeda, Teva Pharmaceuticals, UCB Pharma, Vertex
and Dr August Wolff.
1. Jostins, L. et al. Host-microbe interactions
have shaped the genetic architecture of
inflammatory bowel disease. Nature 491,
119–124 (2012).
47  |  NOVEMBER 2015  www.nature.com/reviews
2. Glocker, E. O. et al. Inflammatory bowel disease
and mutations affecting the interleukin-10
receptor. N. Engl. J. Med. 361, 2033–2045
(2009).
3. Cadwell, K. et al. A key role for autophagy
and the autophagy gene Atg16l1 in mouse
and human intestinal Paneth cells. Nature 456,
259–263 (2008).
4. Adolph, T. E. et al. Paneth cells as a site of
origin for intestinal inflammation. Nature 503,
272–276 (2013).
5. Rakoff-Nahoum, S., Paglino, J.,
Eslami‑Varzaneh, F., Edberg, S. & Medzhitov, R.
Recognition of commensal microflora by
toll‑like receptors is required for intestinal
homeostasis. Cell 118, 229–241 (2004).
© 2015 Macmillan Publishers Limited. All rights reserved
6. Buonocore, S. et al. Innate lymphoid cells drive
interleukin‑23‑dependent innate intestinal
pathology. Nature 464, 1371–1375 (2010).
7. Gevers, D. et al. The treatment-naive
microbiome in new-onset Crohn’s disease.
Cell Host Microbe 15, 382–392 (2014).
8. Colombel, J. F. et al. Infliximab, azathioprine,
or combination therapy for Crohn’s disease.
N. Engl. J. Med. 362, 1383–1395 (2010).
9. Feagan, B. G. et al. Vedolizumab as induction
and maintenance therapy for ulcerative colitis.
N. Engl. J. Med. 369, 699–710 (2013).
10. Lee, J. C. et al. Gene expression profiling of
CD8+ T cells predicts prognosis in patients
with Crohn disease and ulcerative colitis.
J. Clin. Invest. 121, 4170–4179 (2011).
Connect to...

Follow us on Twitter
For the latest commentary and gastroenterology and hepatology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Gastroenterology
& Hepatology. Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Gastroenterology & Hepatology

Web focuses
A selection of articles on a related theme with content ranging from Research Highlights through to Reviews
and Perspectives, providing a panoramic view of a key area of medicine

Conferences
You can meet the editors of Nature Reviews Gastroenterology & Hepatology at some of the meetings listed on our
conferences page

Happy birthday Nature Reviews Gastroenterology & Hepatology!

November 2014 marked the 10th anniversary of the launch of Nature Reviews Gastroenterology
& Hepatology. To celebrate this milestone, we present an anniversary issue containing specially
commissioned articles that highlight progress in the field of gastroenterology and hepatology over
the past decade. In addition, our anniversary infographic provides a snapshot of 10 years of the
journal in numbers.
nature.com/nrgastro10

© 2015 Macmillan Publishers Limited. All rights reserved

NEPHROLOGY
DECADE IN REVIEW—GLOMERULAR DISEASE
The glomerulus reveals some secrets
Agnes B. Fogo
In the past decade, major advances have been made in defining the antigens and pathogenesis of immune
complex diseases such as membranous nephropathy and IgA nephropathy. Probing of rare genetic diseases has
revealed new pathways of injury in proteinuric conditions, including channel abnormalities in the podocyte and
complement dysregulation underlying proliferative glomerular lesions.
Novel approaches to the assessment of
human renal biopsy samples have led to
new insights into several common glo‑
merular diseases. In primary membranous
nephropathy (MN), the most common
target antigen, phospholipase A2 receptor
(PLA2R), has been identified.1,2 Moreover,
mutations in transient receptor cation
potential channel 6 (TRPC6) were dis‑
covered to cause familial focal segmental
glomerulosclerosis (FSGS),3 expanding the
paradigm beyond previous data showing
that genetic forms of this disease are caused
by mutations in structural proteins of the
podocyte. The markedly increased inci‑
dence of some chronic kidney diseases
(CKDs) in African Americans was linked
to risk allele variants of APOL1, which
encodes apolipoprotein L1. 4 The patho‑
genesis of IgA nephropathy (IgAN) was
elucidated, implicating antibody binding
to an abnormally glycosylated region of
IgA‑1—the main IgA subtype present in
the circulation.5 An u­ nbiased classification
approach to the varying biopsy appearances
of IgAN has become a model for explor‑
ing the significance of diverse patterns of
injury. 6 Aberrations in complement C3
regulation are now recognized to underlie
many proliferative glomerular injuries.7,8
Rapidly expanding translational research
is aimed at directing these new insights
into the molecular pathogenesis of disease
towards individualized therapies.
A major goal in nephrology has been to
identify the target antigen in primary MN.
Animal models were developed but the
antigens (for example megalin in Heymann
nephritis) were not present in human MN.
Using an elegant and laborious approach,
Beck et al. collected normal glomeruli from
non-transplanted deceased donor kidneys
and screened protein extracts with serum
48  |  NOVEMBER 2015  www.nature.com/reviews
samples from patients with primary or sec‑
ondary MN.1 They identified the antigen
PLA2R, which is normally expressed on
podocytes, using additional techniques,
including mass spectro­metry and elution
of IgG from the deposits. About 70% of
patients with clinically defined primary
MN had deposits containing PLA2R and
circulating anti­b odies to this receptor.
Importantly, patients with MN secondary to
lupus erythematosus did not show PLA2R
positivity, and only a very minor proportion
of patients with cancer-­associated MN had
PLA2R-positive deposits or serum PLA2R
antibodies. Using a similar approach,
another antigen, thrombospondin type‑1
domain-containing 7A, was subsequently
identified in approximately 10% of PLA2R-
negative patients with primary MN.2 These
studies indicate that extensive screening
and mass spectrometry of large numbers
of samples will likely identify additional
pathogenic antigens in patients without a
clinically evident secondary cause of MN.
Studies are currently underway to define
the utility of assessing serum levels versus
biopsy demonstration of PLA2R in delineat‑
ing specific causality and tracking responses
to intervention. Understanding inheri­
ted (for example HLA-DQ1-associated
risks9) or acquired (for example drug or
infection-­associated) mechanisms under‑
lying autoantibody formation will yield
additional insights and may provide targets
for early detection or even p ­ revention in
­susceptible populations.
Mass spectrometry was also crucial in
clarifying other types of proteins deposited
in glomeruli. For example, patients with
non-AL, non-AA amyloid can now be cor‑
rectly classified according to amyloid type,
identifying rare familial and other forms of
amyloid that require differing therapeutic
© 2015 Macmillan Publishers Limited. All rights reserved
approaches. This technique has transformed
the approach to characterizing unusual
deposits within the kidney, i­ ncluding those
in dense deposit disease (DDD).10
Primary FSGS is a common glomerulo­
pathy caused by an as yet not definitively
identified circulating factor or factors.
However, patients with familial FSGS may
offer insights into possible pathogenic
pathways that may be perturbed in the
more common sporadic FSGS. Most cases
of familial FSGS are linked to mutations in
structural proteins of the podocyte and slit
diaphragms or in components that anchor
podocytes to the glomerular basement
membrane, such as nephrin, podocin and
α‑actinin‑4. Recently, a missense mutation
in the cation channel TRPC6 was discov‑
ered in a kindred with familial adult-onset
FSGS.3 This mutation alters calcium sig‑
nalling, increasing the response to ago‑
nists such as angiotensin II. It is possible
that TRPC6 abnormalities also occur in
sporadic FSGS or other protein­u ric dis‑
eases. Thus, increased TRPC6 activity and
resulting high calcium influx may con‑
tribute to podocyte injury and play a role
Immunofluorescence image showing granular capillary wall
staining for PLA2R in primary membranous nephropathy.

in proteinuria and progressive sclerosis.
Pharmacologic targeting of a channel is an
attractive and attainable goal; manipulating
TRPC6 expression and activity thus offers
promise as a therapy for numerous pro‑
teinuric diseases ­characterized by ­podocyte
dysfunction and loss.
Genetic approaches have also shed
light on non-familial kidney diseases.
Compared with Caucasians, African
Americans have a 4–5-fold increased risk
of some CKDs, including hypertension-
associated CKD, HIV-associated nephro­
pathy and FSGS. Genome-wide association
studies found linkage of this excess risk
to the APOL1 allele variants G1 and G2.4
These variants, unlike the G0 variant
common in Caucasians, are trypanolytic
for Trypanosoma brucei rhodesiense, sug‑
gesting possible evolutionary pressures
underlying their high prevalence in popu‑
lations of African ancestry. Although the
mechanisms for increased disease risk
have not been elucidated, ongoing studies
postulate an effect on podocytes and/or
innate immunity.
‘‘
Mass spectrometry ... has
transformed the approach to
characterizing unusual deposits
’’
within the kidney...
IgAN has a highly variable clinical course
and a range of morphological expres‑
sions. A working group formed by the
International IgA Nephropathy Network
and the Renal Pathology Society undertook
a novel approach to classification of this
disease.6 In a non-biased method, all patho­
logical lesions present within a cohort of
archival IgAN biopsy samples were defined
to determine the variables that could be
reprodu­c ibly scored. The six pathologic
variables thus defined were then assessed
for correlation with clinical outcomes.
Four variables correlated independently
with increasing loss of glomerular filtration
rate: mesangial hypercellularity, endocapil‑
lary hypercellularity, segmental sclerosis
and significant tubular atrophy/­interstitial
fibrosis (>25%). These variables were
used to define the so-called MEST score.
Subsequent studies validated and refined
this scoring approach, which could form
the basis for patient stratification, analysis
of therapeutic responses and assessment of
progression risk, as well as possibly provide
insights into the pathogenesis of various
lesions. Similar to the approach commonly
A DECADE IN MEDICINE NOVEMBER 2015  |  49
used in oncology, patients with similar stage
and severity of disease may be grouped
together in clinical trials.
The pathogenesis of IgAN has also been
elucidated. The hinge region of normal
IgA‑1 has numerous serine and threonine
residues with O‑glycosylation. However,
in patients with IgAN, abnormally glyco‑
sylated IgA‑1 with deficient galactose is
bound by anti-glycan antibodies, includ‑
ing IgG, which may have a role in comple‑
ment activation and a subsequent cascade
of injury.5 This abnormal IgA‑1 is not effi‑
ciently cleared by receptors in the liver,
which may enhance its deposition within
the mesangium. These insights provide a
basis for investigation of events leading
to the generation of glycan-specific anti‑
bodies, not only as disease markers, but
also as a pathogenic ­m echanism and
therapeutic target.
Membranoproliferative glomerulo­
nephritis (MPGN) is characterized by
mesangial and subendothelial deposits,
double contours of glomerular basement
membranes due to reactions to these
deposits, endo­c apillary hypercellularity
and increased matrix. MPGN was pre‑
viously thought to represent immune
complex diseases of various types, desig‑
nated as MPGN I, MPGN II (now known
as DDD) and MPGN III. However, some
patients with an MPGN pattern of injury
have isolated C3 deposits.7 Mass spectrom‑
etry studies demonstrated that deposits
in DDD contain soluble components
of the terminal complement complex.
Aberrations in complement regulatory
proteins were identified in a group of
patients with dominant C3 deposits or
C3 deposits only, including heterozygous
mutations in factor H and CD46 (key reg‑
ulators of C3 activation) and expression
of an autoantibody that inhibits C3 con‑
vertase, C3 nephritic factor.7 Subsequent
studies in animal models and in patients
led to a new paradigm for approaching
these C3 dominant glomerulonephritides,
now called C3 glomerulopathy.8 Patients
show a spectrum of lesions from DDD to
deposits of complement components that
overlap with the ultrastructural appearance
of immune complexes. Various genetic or
acquired alterations in complement regula‑
tion have been recognized, including famil‑
ial forms with mutations in complement
factor H‑related protein 5. The therapeutic
approach in these patients may be control‑
ling complement dysregulation, rather than
searching for underlying lupus, collagen
© 2015 Macmillan Publishers Limited. All rights reserved
NEPHROLOGY
vascular diseases, chronic infections or
other triggers of MPGN.
The field of nephrology—and specifi‑
cally glomerular diseases—is catching up
with the state-of-the-art in oncology, in
which specific mutation profiles and aber‑
rant signalling pathways form the basis for
personalized therapy. We are at the dawn
of a new beginning in molecular under‑
standing of the pathogenesis of some of the
common diseases afflicting the glomerulus.
These key observations form the basis for
rapid progress in the development of new
insights into the pathogenesis of disease
and new targets for intervention.
Department of Pathology, Immunology and
Microbiology, Vanderbilt University Medical
Center, C3310 MCN, 1161 21st Avenue,
Nashville, TN 37232, USA.
agnes.fogo@vanderbilt.edu
doi:10.1038/nrneph.2015.149
Published online 15 September 2015
Competing interests
The author declares no competing interests.
1. Beck, L. H. Jr et al. M‑type phospholipase
A2 receptor as target antigen in idiopathic
membranous nephropathy. N. Engl. J. Med. 361,
11–21 (2009).
2. Tomas, N. M. et al. Thrombospondin type‑1
domain-containing 7A in idiopathic membranous
nephropathy. N. Engl. J. Med. 371, 2277–2287
(2014).
3. Winn, M. P. et al. A mutation in the TRPC6
cation channel causes familial focal
segmental glomerulosclerosis. Science 308,
1801–1804 (2005).
4. Genovese, G. et al. Association of
trypanolytic ApoL1 variants with kidney
disease in African Americans. Science 329,
841–845 (2010).
5. Suzuki, H. et al. Aberrantly glycosylated
IgA1 in IgA nephropathy patients is
recognized by IgG antibodies with restricted
heterogeneity. J. Clin. Invest. 119, 1668–1677
(2009).
6. Working Group of the International IgA
Nephropathy Network and the Renal Pathology
Society. The Oxford classification of IgA
nephropathy: pathology definitions,
correlations, and reproducibility. Kidney Int. 76,
546–556 (2009).
7. Servais, A. et al. Primary glomerulonephritis
with isolated C3 deposits: a new entity which
shares common genetic risk factors with
haemolytic uraemic syndrome. J. Med. Genet.
44, 193–199 (2007).
8. Pickering, M. C. et al. C3 glomerulopathy:
consensus report. Kidney Int. 84, 1079–1089
(2013).
9. Stanescu, H. C. et al. Risk HLA-DQA1 and
PLA2R1 alleles in idiopathic membranous
nephropathy. N. Engl. J. Med. 364, 616–626
(2011).
10. Sethi, S. et al. Laser microdissection and mass
spectrometry-based proteomics aids the
diagnosis and typing of renal amyloidosis.
Kidney Int. 82, 226–234 (2012).
NEPHROLOGY
DECADE IN REVIEW—GENETICS OF KIDNEY DISEASES
Genetic dissection of kidney
disorders
Friedhelm Hildebrandt
Advances in genome sequencing and genetic manipulation techniques
over the last decade have helped identify numerous single-gene causes
of early-onset kidney diseases and risk alleles for complex, polygenic
traits. Subsequent studies regarding the underlying disease mechanisms
will help lead to personal genetic diagnoses and unique therapeutic
interventions in the future.
The advent of next-generation sequen­cing
techniques over the past decade has led to the
discovery that a surprisingly high propor‑
tion of cases of early-onset chronic kidney
disease (CKD) are caused by mutations in
a single gene (that is, they are monogenic).
Single-gene mutations in such genes have a
very strong effect on disease causation and
are, therefore, considered to exert full pene­
trance on genetic causality. This paradigm
was exemplified in a study that performed
a mutation analysis of the 27 known causa‑
tive genes of steroid-resistant nephrotic syn‑
drome (SRNS) in an international cohort of
1,780 patients with disease that manifested
before 25 years of age.1 Strikingly, ~30%
patients were carriers of a causative muta‑
tion in one of the studied genes that could
fully explain the disease expression. The frac‑
tion of families in whom a single-gene cause
of SRNS was identified correlated inversely
with age of disease onset, with a single-gene
cause identified in 70% of affected newborns
and 12% of young adults.
Further investigations into the 27 causa‑
tive genes of SRNS have revealed that their
related gene products are part of protein–
protein interaction complexes that con‑
verge onto specific pathways critical for
glomerular podocyte function. For example,
analysis of mutations in ARHGDIA dem‑
onstrated that proper regulation by this
gene of the small GTPases Rho, Rac, and
Cdc42, is essential for the integrity of the
glomerular filter.2 This finding could open
inroads into developing Rho and/or Rac
inhibitors and activators to treat SRNS,
for which no efficient treatment currently
exists. Other opportunities for therapeutic
SilverV/iStock/Thinkstock
development have also been generated from
mutation analysis in novel disease-causing
genes. Mutations in COQ63 and ADCK4,
which regulate coenzyme Q 10 biosynth­
esis, may cause SRNS, and patients who
carry these specific variants could benefit
50  |  NOVEMBER 2015  www.nature.com/reviews
© 2015 Macmillan Publishers Limited. All rights reserved
from ­c oenzyme  Q 10 supplementation.
A high percentage of monogenic causa‑
tion has also been detected in individuals
manifesting with early-onset CKD due
to other causes, such as cystic kidney dis‑
eases (>50%), urinary stone disease (21%),
congenital anomalies of the kidneys and
urinary tract (17%), and nephritis (15%).
Overall, monogenic mutations in ~250
recessive or dominant genes are estimated to
account for ~25% of all cases of early-onset
(<25 years) CKD.
Cystic kidney diseases, also known as
renal ciliopathies, are caused by a dysfunc‑
tion of the primary cilia and centrosomes.
Recent discoveries of mutations in the
centrosomal protein CEP1644 and FAN15
have implicated DNA damage response
signalling in the pathogenesis of renal cilio­
pathies. This finding led to the proposal
that accumulation of DNA damage with
consecutive premature cellular senescence
could contribute markedly to the develop‑
ment of renal fibrosis, which is central to all
forms of CKD independent of the primary
cause. Loss-of-function mutations in reces‑
sive monogenic kidney disease genes can
be readily modelled and characterized
in animals, such as mice or zebrafish, by
knockdown and/or knockout approaches.
The first successful treatment approaches
for renal cilio­pathies were demonstrated
in mouse models, as exemplified in a study
that addressed impaired cell cycle dysregu‑
lation as a potential cause of cysto­genesis.
Administration of a cyclin-­dependent kinase
inhibitor, roscovitine, in mouse models of
juvenile and congenital poly­c ystic kidney
disease efficiently arrested cystic disease.
The mechanism of action revealed effective
cell-cycle arrest, transcriptional inhibition,
and attenuation of apoptosis.6
Studies performed over the past decade
have also focused on determining the con‑
tribution of single gene causes of adult-onset
CKD. Researchers found that the NPHS2
(podocin) gene variant R229Q, which had
been considered a variant of unknown sig‑
nificance for many years, is in fact disease-
causing when expressed in a compound
heterozygous state with one of a few other
distinct NPHS2 mutations.7 The combina‑
tion of heterozygous mutations on both
NPHS2 alleles was shown to result in adult
onset of SRNS. ‘Mild’ recessive genetic vari‑
ants are most likely to cause late-onset forms
of disease and, therefore, might explain a
considerable proportion of adult-onset CKD.
Whereas monogenic mutations are
usually discovered by whole exome sequen­
cing, variants that underlie polygenic dis­
orders are usually detected by genome-wide
association studies (GWAS). Polygenic dis‑
orders are more common than are mono‑
genic disorders, exert weak caus­a lity (or
risk) on the disease phenotype, usually
mani­fest later in life, and are often modified
by environmental effects. Risk alleles usually
account for only a small proportion of the
variance of the disease pheno­type, and the
mechanistic assignment of an associ­ated
marker allele to loss of gene function is
­difficult to ascertain.

An important discovery attained by GWAS
was the detection of a single nucleo­tide poly‑
morphism, rs12917707, located near the
UMOD gene, which if mutated, causes auto‑
somal dominant medullary cystic kidney
disease type 2.8 This finding was later con‑
firmed in a large worldwide cohort. Another
breakthrough GWAS study, published at the
turn of the decade, reported an association
between APOL1 variants and CKD. Two
variants were found to be strongly associ‑
ated with an increased risk of focal segmen‑
tal glomerulosclerosis (FSGS) and CKD
in African Americans.9 More than half of
African Americans in the USA (compared
with 4% of European Americans) carry both
recessive APOL1 risk alleles, increasing the
risk of developing FSGS by 5–30 fold. These
data likely explain the known racial dispari‑
ties in non-diabetic nephropathy seen in
African Americans.10 The two APOL1 risk
alleles always occur on different homolo‑
gous chromosomes and thereby behave like
­recessive mutations.
As this article has exemplified, techno‑
logical approaches to probe and dissect the
human genome have rapidly progressed over
the past 10 years, and the cost of perform‑
ing such investigations has decreased. The
ability to perform GWAS in large, worldwide
cohorts of patients with complex, polygenic
kidney disorders is enabling researchers to
identify new risk alleles for investigation
and to help determine disease causality and
prevalence in various populations. In the near
future we may be able to perform functional
studies of the many genetic variants of reces‑
sive and dominant causative genes for CKD,
for which there is currently not enough
evidence to indicate a deleterious effect.
Genetic manipulation techniques have also
advanced, enabling scientists to readily and
accurately model gene variants and muta‑
tions in vitro and in vivo to ascertain gene
function. The identification of single-gene
mutations in patients with kidney disorders,
by methods such as whole exome sequen­cing,
has revolutionized our practice of diagnos‑
ing and understanding disease pathogenesis.
Continued discovery of novel disease-causing
genes will enable the provision of an unequiv‑
ocal molecular genetic diagnosis to patients,
generate new insights into disease mech­an­
isms, and have consequences for personal‑
ized treatment and prevention of CKD in the
future. The functional characterization of
deleterious variants in known kidney disease
genes will be one of the most important and
rewarding contributions to understanding
renal pathogenesis in the coming decade.
A DECADE IN MEDICINE NOVEMBER 2015  |  51
Boston Children’s Hospital, Division of
Nephrology, 300 Longwood Avenue,
Enders 561, Boston, MA 02115, USA.
friedhelm.hildebrandt@childrens.harvard.edu
doi:10.1038/nrneph.2015.148
Published online 1 September 2015
Acknowledgements
F.H. has received grant support from the NIH
(grant numbers DK068306, DK076683, DK88767),
and from the Howard Hughes Medical Institute.
Competing interests
The author declares no competing interests.
1. Sadowski, C. E. et al. A single-gene cause in
29.5% of cases of steroid-resistant nephrotic
syndrome. J. Am. Soc. Nephrol. 26, 1279–1289
(2014).
2. Gee, H. Y. et al. ARHGDIA mutations cause
nephrotic syndrome via defective RHO GTPase
signaling. J. Clin. Invest. 123, 3243–3253
(2013).
3. Heeringa, S. F. et al. COQ6 mutations in human
patients produce nephrotic syndrome with
DECADE IN REVIEW—ACUTE KIDNEY INJURY
Acute kidney injury—a decade
of progress
Rinaldo Bellomo
The past decade has seen developments in several aspects of acute
kidney injury (AKI), including the discovery of an array of biomarkers,
assessment of the optimal dose intensity for renal replacement therapy,
and the impact of fluid administration. Furthermore, AKI has emerged as
an important risk factor for chronic kidney disease.
Clinical research dedicated to understand‑
ing, diagnosing, and treating acute kidney
injury (AKI)—a term that, by consensus, has
now substituted the previously used ‘acute
renal failure’—has increased markedly over
the past 10 years. The number of research
papers on this topic has doubled during the
past decade, and during this period several
trends have evolved. One key development
has been the discovery and testing of bio‑
markers for early AKI. Heralded by the first
publication of the predictive value of neutro‑
phil gelatinase associated lipocalin (NGAL)
in paediatric cardiac surgery,1 biomarker
discovery and testing for the early indica‑
tion and prognosis of AKI has undergone a
remarkable degree of exploration. More than
a dozen candidate biomarkers with variable
predictive value have now been identified in
urine and blood, with typical values for the
area under the receiver operating character‑
istic (AUROC) curve ranging between 0.65
and 0.85 (Figure 1). NGAL has been the most
investigated biomarker so far, with mixed
© 2015 Macmillan Publishers Limited. All rights reserved
NEPHROLOGY
sensorineural deafness. J. Clin. Invest. 121,
2013–2024 (2011).
4. Chaki, M. et al. Exome capture reveals
ZNF423 and CEP164 mutations, linking renal
ciliopathies to DNA damage response signaling.
Cell 150, 533–548 (2012).
5. Zhou, W. et al. FAN1 mutations cause
karyomegalic interstitial nephritis, linking chronic
kidney failure to defective DNA damage repair.
Nat. Genet. 44, 910–915 (2012).
6. Bukanov, N. O. et al. Long-lasting arrest of murine
polycystic kidney disease with CDK inhibitor
roscovitine. Nature 444, 949–952 (2006).
7. Tory, K. et al. Mutation-dependent recessive
inheritance of NPHS2-associated steroid-
resistant nephrotic syndrome. Nat. Genet. 46,
299–304 (2014).
8. Kottgen, A. et al. Multiple loci associated with
indices of renal function and chronic kidney
disease. Nat. Genet. 41, 712–717 (2009).
9. Genovese, G. et al. Association of trypanolytic
ApoL1 variants with kidney disease in African
Americans. Science 329, 841–845 (2010).
10. Freedman, B. I. et al. The apolipoprotein L1
(APOL1) gene and nondiabetic nephropathy
in African Americans. J. Am. Soc. Nephrol. 21,
1422–1426 (2010).
results and some serious concerns regard‑
ing assay reproducibility together with the
growing understanding that NGAL probably
represents a family of molecules of different
origin and with different epitopes. The most
recent and more systematically studied bio‑
marker discovery relates to the so-called ‘cell
cycle arrest biomarkers’, which have now been
assessed in several large cohorts and have
achieved consistent AUROC values of ~0.8
(Figure 1).2 Although now FDA approved, the
value of cell cycle arrest biomarkers in clinical
practice remains unclear and is hampered by
the absence of specific linked therapies.
Irrespective of the possibility of early diag‑
nosis, some patients develop severe AKI and
are deemed to require renal replacement
therapy (RRT), raising the issue of optimal
dose intensity. After some initial small and
single institution studies that produced
promising results, two large multicentre trials
(one in the USA3 and the other in Australia
and New Zealand4) tested whether increasing
the intensity of RRT from an effluent flow of
 NEPHROLOGY
u TIMP-2
× IGFBP-7
u TIMP-2
u IGFBP-7
u NGAL
Biomarker
p Cys C
u KIM-1
p NGAL
u IL-18
u pi-GST
u L-FABP
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
AUROC
Nature Reviews | Nephrology
20–25 ml/kg per hour in continuous modes
or from intermittent dialysis, to 35–40 ml/kg
per hour or daily dialysis would increase
survival. Both studies were identical in dem‑
onstrating a lack of effect on survival, thus
establishing a worldwide standard of care in
relation to RRT dose.
The epidemic of chronic kidney disease
(CKD), chronic heart failure, acute decom‑
pensated heart failure, and AKI prompted
consideration of the complex interactions
between cardiac and renal function, kidney–
heart cross talk, and the clinical syndromes
that can ensue to be conceptualized as ‘cardio­
renal syndromes’. The initial definition,
description, and proposed mechanisms by
which acute disease of each organ could affect
the other has spawned important reinvesti‑
gations and greater understanding of applied
pharmacology in this field.5 Importantly, and
together with other studies, the concept of
cardiorenal syndrome has shifted the focus
of management in patients with AKI from
the administration of fluids to ‘rescue’ the
failing kidney, to a greater understanding
and concern that fluid administration might
have little beneficial effect on renal function
and could impose a substantial contribution
to positive fluid balance. A positive fluid
balance might, in turn, cause renal oedema,
delay renal recovery, injure vital organs, and
increase mortality—an effect seen most
­strikingly in children with severe AKI.6
Faced with such varied practice patterns,
the utility of several definitions of AKI, and
uncertainty with regard to how best to manage
patients with developing or established AKI,
a large group of experts were funded to
develop a detailed framework of reference
for the definition, diagnosis, prevention,
and treatment of AKI. The Kidney Disease:
Improving Global Outcomes in Acute Kidney
Injury working group delivered an analysis
of all literature in this field and generated a
global definition for this syndrome, which
52  |  NOVEMBER 2015  www.nature.com/reviews
◀ Figure 1 | Histogram summarizing the
estimated AUROC for 12 biomarkers used to
predict early AKI. The predictive value of each
biomarker for AKI stage 2 or 3 (according to
the Kidney Disease: Improving Global
Outcomes in Acute Kidney Injury working group
guidelines) was assessed within 12 h of urine
or blood sampling. Data obtained from the
Sapphire study2 under the Creative Commons
agreement. © Kashani, K. et al. Crit. Care 17,
R25 (2013). Abbreviations: AKI, acute kidney
injury; AUROC, area under receiver operating
characteristic; p, plasma; u, urinary.
represents the most comprehensive body of
references and ­information associated with
this c­ ondition to date.7
Simultaneously, and published in the same
year, intensive care investigators tackled the
possible effect of fluid choice on renal out‑
comes in critically ill patients. Two large
multicentre double-blind randomized trials
compared the most commonly administered
artificial colloid solution in the world for fluid
resuscitation (hydroxyethyl starch [HES])
and compared it with saline. In a study of
patients with severe sepsis, the Scandinavian
Critical Care Trials Group found that fluid
resuscitation with HES increased mortality
and the incidence of AKI, including severe
AKI requiring RRT.8 Concordant with such
findings, the Australian and New Zealand
Intensive Care Society Clinical Trials Group
investigated a broad cohort of all patients in
intensive care treated with fluid bolus therapy
and also found that the administration of
HES increased the need for RRT.9 These
studies have established beyond doubt that
HES is a nephrotoxin and has led to a major
decrease in its use in developed countries.
Finally, the impact of AKI on the long-term
outcome of patients both from a general and
renal point of view has become a focus of
active investigation after a seminal study
of ~4,000 patients identified a clear associ­
ation (a threefold increase in risk) between
an episode of AKI requiring dialysis and the
subsequent need for chronic dialysis despite
initial recovery.10 Other studies have since
confirmed this associ­ation. Whether these
findings relate to the possibility that patients
who develop AKI have characteristics that
make them more likely to develop CKD
or represent a specific deleterious effect of
AKI on renal function is unclear. These data
have confirmed, however, that patients with
AKI require specific medical attention and
­represent a new at-risk population for CKD.
In summary, we have made notable strides
in understanding the pathology and improv‑
ing outcomes in AKI over the past decade.
© 2015 Macmillan Publishers Limited. All rights reserved
We have identified new biomarkers that could
prove useful in the future, defined a standard
dose for RRT, recognized the adverse conse‑
quences of overzealous fluid therapy on the
kidney as well as the patient, defined the need
to develop focused interventions to modulate
the cross-effect of simultaneous acute losses
of kidney function and cardiac function,
developed a set of practice guidelines that
provide a baseline reference point definition
and approach to AKI management for future
reference, clearly demonstrated the nephro‑
toxiciy of HES and, finally, have shown that
AKI is a major risk factor for CKD. Despite
these advances that have undoubtedly
decreased patient harm, we remain deeply
ignorant of the molecular mechanisms
responsible for the development of AKI in
humans and have not yet devised a single spe‑
cific and effective therapy to prevent, treat, or
attenuate AKI. Much remains to be done in
this field before notable improvements can be
made to patient outcomes.
Australian and New Zealand Intensive Care
Research Centre (ANZIC–RC), Monash
University, Commercial Road, Melbourne,
VIC 3181, Australia.
rinaldo.bellomo@austin.org.au
doi:10.1038/nrneph.2015.147
Published online 15 September 2015
Competing interests
R.B. has received grants from Baxter and honoraria
from Gambro and B. Braun.
1. Mishra, J. et al. Neutrophil gelatinase-
associated lipocalin (NGAL) as a biomarker for
acute renal injury after cardiac surgery. Lancet
365, 1231–1238 (2005).
2. Kashani, K. et al. Discovery and validation of
cell cycle arrest biomarkers in human acute
kidney injury. Crit. Care 17, R25 (2013).
3. VA/NIH Acute Renal Failure Trial Network.
Intensity of renal support in critically ill patients
with acute kidney injury. N. Engl. J. Med. 359,
7–20 (2005).
4. The RENAL Replacement Therapy Study
Investigators. Intensity of continuous renal
replacement therapy in critically ill patients.
N. Engl. J. Med. 361, 1627–1638 (2009).
5. Ronco, C. et al. Cardiorenal syndrome.
J. Am. Coll. Cardiol. 52, 1527–1539 (2008).
6. Sutherland, S. M. et al. Fluid overload and
mortality in children receiving continuous
renal replacement therapy: the prospective
pediatric continuous renal replacement therapy
registry. Am. J. Kidney Dis. 55, 316–325 (2010).
7. KDIGO working group. KDIGO clinical practice
guidelines for acute kidney injury. Kidney Int.
Suppl. 2, 1–136 (2012).
8. Perner, A. et al. Hydroxyethyl starch 130/0.42
versus Ringer’s acetate in severe sepsis.
N. Engl. J. Med. 367, 124–134 (2012).
9. Myburgh, J. A. et al. Hydroxyethyl starch or
saline for fluid resuscitation in intensive care.
N. Engl. J. Med. 367, 1901–1911 (2012).
10. Wald. R. et al. Chronic dialysis and death
among survivors of acute kidney injury requiring
dialysis. JAMA 302, 1179–1185 (2009).
 NEPHROLOGY

DECADE IN REVIEW—POLYCYSTIC KIDNEY DISEASE our institution receive pravastatin in addi‑

Slowing progression of autosomal tion to an ­angiotensin-converting-enzyme

(ACE) inhibitor.

dominant polycystic kidney disease Despite advances in understanding of

Robert W. Schrier
Autosomal dominant polycystic kidney disease is a challenging disorder
to diagnose and treat effectively. Promising research over the past
decade has, however, provided novel interventions, modifications to
clinical practice and new areas to investigate with the aim of identifying
approaches to slow disease progression.
Autosomal dominant polycystic kidney
disease (ADPKD) can present in children
at an early age. The disease is character‑
ized by tubular cystic expansion, kidney
enlargement and vasculature compres‑
sion of the parenchyma. Understanding
of the pathogenesis of ADPKD has rapidly
advanced since the identification of caus‑
ative mutations in PKD1 and PKD2 in
the mid-1990s. Approximately 80–85%
of patients with ADPKD are carriers of a
PKD1 mutation, with truncation mutations
in this gene known to result in faster disease
progression than mutations in PKD2. The
rapid development of sophisticated tech‑
niques to manipulate the genome over the
past 10 years has further advanced under‑
standing of the molecular and pathogenic
consequences of perturbed PKD1 func‑
tion in ADPKD. Using the Cre-lox system,
researchers have shown that inactivation
of Pkd1 in mice during renal development
causes rapid cyst growth, whereas inacti‑
vation in adulthood causes comparatively
slower cyst growth. Such studies indicate
that the type of mutation as well as the
degree and timing of inactivation of PKD1
and PKD2 might contribute to the severity
of ADPKD in humans.1
Several other risk factors for ADPKD
progression exist, such as early onset hyper‑
tension and frequent gross haematuria.
ADPKD progression seems to occur faster
in males than in females, which might be
means by which this growth can be slowed
in affected children and young adults. Early
experimental models showed that lova­statin
could decrease the severity of ADPKD by
slowing the increase in kidney volume,3
leading to the proposal that statin admin‑
istration could decrease the rate of kidney
volume growth in patients with ADPKD.
In 2014, a double blind, randomized trial in
paediatric patients with ADPKD and normal
kidney function found that pravastatin sig‑
nificantly decreased the percentage change
in height-corrected total kidney volume
over a 3‑year period compared to placebo
(23% ± 3% versus 31% ± 3%; P = 0.02).4 The
mechanisms by which this statin exerts
its effect remain to be defined, but based
on these data, patients aged 8–22 years at
Cyst compression of renal vasculature
Increased Endothelin
growth factors
Angiogenesis
Cyst proliferation Systemic vascular
and expansion resistance
the mechanisms of cyst proliferation in
ADPKD, the development of targeted
therapies has proven difficult. The major‑
ity of clinical advancement in ADPKD
over the past 10 years has primarily been
in determining the optimal management
of hypertension and cardiovascular risk
in affected patients. Hypertensive patients
with ADPKD show increased activation of
the renin–angiotensin–aldosterone system
(RAAS) compared to patients with essen‑
tial hypertension of the same age, renal
function and level of blood pressure. The
various potential pathways whereby stimu‑
lation of the RAAS can cause hypertension
and worsen kidney function include cyst
proliferation and expansion, increased sym‑
pathetic and endothelin activity, oxidant
injury and fibrosis (Figure 1). At the start of
the 21st century, no clear consensus existed
with regard to the most appropriate type of
antihypertensive therapy for patients with
ADPKD. Randomized prospective studies
of ACE inhibitors in this patient population
were lacking, although experimental and
clinical data suggested that ACE inhibition
might be an effective treatment option. In
2002, a 7‑year study in hypertensive patients
Renin
Angiotensin II
Oxidant-mediated Transforming
endothelial damage growth factor beta
Sympathetic Aldosterone
activity
Sodium retention Renal fibrosis

associated with earlier onset of more severe

hypertension in men. Among women with
ADPKD, those with a history of three or
more pregnancies exhibit more rapid loss
of kidney function than those with fewer
previous pregnancies. Overt proteinuria in
patients with ADPKD also predicts a faster
loss of kidney function. The loss of kidney
function observed in these patients strongly
correlates with rapid growth in kidney
volume.2 Research over the past decade has,
therefore, focused on identifying effective
Hypertension and renal disease
Figure 1 | Pathogenic role of the renin–angiotensin–aldosterone system in ADPKD. Angiotensin II
predominantly acts to increase blood pressure through vasoconstriction, Nature Reviews
but is | Nephrology
also known to
increase cell proliferation, angiogenesis, oxidant injury and fibrosis, which can exacerbate the
renal component of ADPKD. Potential indirect effects of angiotensin II during hypertension include
stimulation of the sympathetic nervous system, endothelin and aldosterone with sodium
retention. Angiotensin II also stimulates reactive oxygen species, which could account, at least
in part, for the observed endothelial dysfunction in ADPKD. Abbreviation: ADPKD, autosomal
dominant polycystic kidney disease. Reproduced with permission from American Society of
Nephrology © Schrier, R. W. et al. J. Am. Soc. Nephrol. 20, 1888–1893 (2009) and modified with
permission from Remedica Medical Education and Publishing.

A DECADE IN MEDICINE NOVEMBER 2015  |  53

© 2015 Macmillan Publishers Limited. All rights reserved
NEPHROLOGY
with ADPKD and left ventricular hyper‑
trophy compared the effects of lowering
blood pressure with RAAS blockade versus
a calcium channel blocker.5 Left ventricu‑
lar mass index (LVMI) decreased in both
cohorts, but the decrement was greater in
patients administered a RAAS blocker.
The results of a 5‑year randomized study
later showed that ACE inhibition in chil‑
dren and young adults with ADPKD and
blood pressures within the upper quartile
of the normal range could prevent decreases
in creatinine clearance and increases in
LVMI.6 ACE inhibition, however, did not
affect kidney volume growth. Nevertheless,
because of enhanced RAAS activation,
we have now entered an era in which
ACE inhib­ition is a fairly standard first-
line t­reatment for hypertensive patients
with ADPKD.
Data from the HALT-PKD double-blind,
placebo-controlled, randomized trial were
reported in 2014.7 This trial assessed the
effects of a standard (120/70 mmHg to
130/80 mmHg) versus a low (95/60 mmHg
to 110/75 mmHg) blood pressure target,
and of the ACE inhibitor lisinopril plus
either telmisartan (an angiotensin-­receptor
blocker) or placebo, on total kidney volume
in patients aged 15–49 years with ADPKD
and estimated glomerular filtration rate
>60 ml/min/1.73 m 2. No difference was
found in the annual percentage change in
total kidney volume associated with dual
versus single RAAS blockade; however, the
annual percentage change in total kidney
volume was significantly lower in the low
blood pressure target group compared to the
standard blood pressure target group (5.6%
versus 6.6%, P = 0.006). Moreover, LVMI
decreased to a greater extent in the former
than in the latter group (1.17 g/m2 per year
versus –0.57 g/m2 per year, P <0.001) as did
urinary albumin excretion (3.77% versus
2.43%, P <0.001).
Improved diagnosis of ADPKD has
also markedly shaped clinical practice
over the past 10 years. For example, 2009
saw the development of unified criteria for
ultra­s onographic diagnosis of ADPKD,
with at-risk patients aged 30–39  years
now requiring a minimum of three renal
cysts for diagnosis. Assessments such as
intra­cranial screening for aneurysm and
positron-­emission and computed tomo­
graphy to identify infected renal or liver
cysts are also now routinely performed.8
Improved understanding of the patho‑
genic mechanisms of cyst formation has
led to the investigation of new treatments
54  |  NOVEMBER 2015  www.nature.com/reviews
for ADPKD, such as the vasopressin
V2-receptor antagonist tolvaptan, which
has not received FDA approval.9 New con‑
cepts regarding the pathogenic mechanisms
of cyst formation have also been proposed,
such as the identification of human CD133+
progenitor cells and other specialized cells
that can promote renal cyst formation in
mouse models.10 Finally, the results of the
HALT-PKD study have helped to clarify
blood-pressure targets and antihyper­
tensive regimens for patients with ADPKD.7
These combined, cross-disciplinary efforts
will enable optimal treatment and diagno‑
sis of patients with ADPKD as well as the
identification of novel approaches to slow
disease progression.
Division of Renal Diseases and Hypertension,
University of Colorado Denver, 12700 East 19th
Avenue, Campus Box C281, Aurora, CO 80045,
USA.
robert.schrier@ucdenver.edu
doi:10.1038/nrneph.2015.164
Published online 13 October 2015
Competing interests
The author declares no competing interests.
1. Lantinga-van Leeuwen, I. S. et al. Lowering of
Pkd1 is sufficient to cause polycystic kidney
disease. Hum. Mol. Genet. 13, 3069–3077
(2004).
DECADE IN REVIEW—RENAL TRANSPLANTATION
A spectrum of advances in renal
transplantation
Bruce Kaplan
Although the first 50 years of renal transplantation were marked by great
advances in immunosuppressive therapies, the past decade has been
marked by an unprecedented increase in technology. This progress has
spurred investigators to challenge old paradigms and investigate how
best to utilize these technologies to further improve patient care.
The establishment of immunological
tolerance—­d efined as the absence of a
detectable immune response against a func‑
tional graft in the absence of immuno­
suppression—­remains a major but as yet
unachieved goal in transplantation. In 2008
Kawai et al. reported on a series of five
patients who received bone marrow condi‑
tioning with a non-myeloablative regimen
before transplantation, followed by donor-
derived bone marrow post-transplantation
in the hope of producing a mixed chimeric
tolerogenic state.1 These patients were then
weaned off immunosuppressive therapy.
© 2015 Macmillan Publishers Limited. All rights reserved
2. Cadnapaphornchai, M. A. et al. Prospective
change in renal volume and function in children
with ADPKD. Clin. J. Am. Soc. Nephrol. 4,
820–829 (2009).
3. Gile, R. D. et al. Effect of lovastatin on the
development of polycystic kidney disease in the
Han:SPRD rat. Am. J. Kidney Dis. 26, 501–507
(1995).
4. Cadnapaphornchai, M. et al. Effect of
pravastatin on total kidney volume, left
ventricular mass index, and microalbuminuria in
pediatric autosomal dominant polycystic kidney
disease. Clin. J. Am. Soc. Nephrol. 9, 889–896
(2014).
5. Schrier, R. W. et al. Cardiac and renal effects of
standard versus rigorous blood pressure control
in autosomal-dominant polycystic kidney
disease: Results of a seven-year prospective
randomized study. J. Am. Soc. Nephrol. 13,
1733–1739 (2002).
6. Cadnapaphornchai, M. et al. Increased left
ventricular mass in children with autosomal
dominant polycystic kidney disease. Kidney Int.
74, 1192–1196 (2008).
7. Schrier, R. W. et al. Blood pressure in early
autosomal dominant polycystic kidney disease.
N. Engl. J. Med. 371, 2255–2266 (2014).
8. Schrier, R. W. et al. Repeat imaging for
intracranial aneurysms in patients with
autosomal dominant polycystic kidney disease.
N. Engl. J. Med. 22, 2361–2375 (2013).
9. Torres, V. E. et al. Tolvaptan in patients with
autosomal dominant polycystic kidney disease.
N. Engl. J. Med. 367, 2407–2418 (2012).
10. Kurbegovic, A. & Trudel, M. Progressive
development of polycystic kidney disease in the
mouse model expressing Pkd1 extracellular
domain. Hum. Mol. Genet. 22, 2361–2375
(2013).
Two of the first three patients developed
antibody-mediated rejection and a decision
was made to add rituximab therapy to the
pre-transplantation regimen; however, both
subsequently transplanted patients exhib‑
ited signs of emerging HLA antibodies and
graft damage. This study was meticulously
performed and monitored, and should serve
as a template for careful assessment of the
efficacy of any future attempt to induce tol‑
erance. Interestingly, analysis of data beyond
1 year post-transplantation showed evidence
that mixed chimerism is not neces­sary to
maintain a tolerogenic state and instead
 NEPHROLOGY

Box 1 | Advances in renal transplantation in the past decade or never find a compatible donor. Over the
past decade a number of centres have per‑
■■ Only a few clinical studies of tolerance in transplantation have been performed to date,
but mechanistic insights into tolerance are increasing
formed desensitization protocols with excel‑
■■ The new US kidney allocation system aims to pair the best kidneys to patients with the lent success rates (at least in the short and
longest expected survival intermediate term of 1–5 years). However,
■■ Transplantation of patients with HIV infection has become increasingly common the outcomes of desensitization are still not
■■ Great progress has been made in molecular diagnosis of graft rejection, but the findings as good those with a fully compatible donor.
require robust validation Montgomery and colleagues reported 211
■■ Biomarker development has been slow and meticulously performed studies have not been patients who underwent desensitization
universally adopted by the transplantation community
and were subsequently transplanted.6 These
■■ The co-stimulatory blocker belatacept is the first biologic agent approved for maintenance
therapy for the prevention of graft rejection, but adoption for newly transplanted patients has
patients had a significant survival advantage
been slow over those who remained on the transplant
■■ The pathology of graft loss is coming under greater scrutiny and new paradigms of loss of waiting list. Although graft survival in the
allograft function are emerging desensitized patients was not equivalent
to graft survival associated with transplan‑
tation of an organ from a compatible live
suggested that tolerance is more dependent prognostic tools in transplantation. Although donor, patients who had been desensi­
on the emergence of T regulatory cells. This this goal has not yet been met, data from two tized still accrued great benefit. As the
study can only be called a partial success in studies have indicated the great potential of Scientific Registry of Transplant Recipients
terms of outcomes, but did reveal tremen‑ genomic analyses. In one study, researchers (SRTR)—which evaluates transplanta‑
dous mechanistic insights and the long road from Edmonton interrogated kidneys biop‑ tion outcomes across the USA—does not
that lies ahead in order to fully understand sied for cause, and found a set of molecular adjust expected survival for desensitized
how best to induce tolerance. classifiers of tissue injury that preliminarily patients, many centres that perform desen‑
Allocation of kidneys for transplantation showed high correlation with graft loss.3 This sitization find themselves at risk of out‑
has always been a daunting task that requires paradigm shift away from diagnostic cat‑ comes being rated as ‘below expected’ and
concepts of equity to be balanced with egories on the basis of histologic findings to spurring an extensive audit and large fine
approaches to achieve greatest utility. After molecular injury-based identifiers could lead from the Centres for Medicare & Medicaid
many years of public debate a new kidney to more accurate prognostic testing as well as Services. Whether the SRTR should adjust
allocation system was approved by the Board change our perception of graft loss. Taking for d
­ esensi­tization is open for debate, but
of Directors of the Organ Procurement and this finding a step further, researchers from without risk adjustment, transplant pro‑
Transplantation Network and implemented the Scripps Research Institute delineated grammes have a powerful ­d isincentive
in 2014. The goal of the new kidney alloca‑ potential noninvasive molecular classifiers towards performing desensitization.

tion system is to provide greater access to
available organs through national matching
of highly sensitized patients, while using the
Kidney Donor Risk Index (KDRI) as a metric
of kidney quality to match the ‘best’ organs
with recipients who are expected to survive
longest (as assessed using the Estimated
Post Transplant Survival score). The aim of
this revised system is to increase transplant
outcomes and maximize organ supply by
allocating kidneys with long survival poten‑
tial to recipients expected to live longest.2
The results of this new system are as yet
unknown, but many in the transplantation
community have raised questions about the
benefits of transplanting so-called compat‑
ible kidneys into highly sensitized patients
without mandatory testing for HLA-DQ and
HLA-DP alleles. Questions have also been
raised as to how well the KDRI discriminates
kidneys that are not at the extremes of the
scoring metric.
The advent of highly accurate genomics
platforms along with advances in bioinfor‑
matics analytics has ushered in a number of
genomics based studies to help understand
the underlying mechanisms of graft injury
as well as potentially serve as diagnostic and
in whole blood that could predict both acute
rejection and degree of individual immuno‑
suppression.4 Both of these studies were ele‑
gantly conceived and analysed; however, as
with all new technologies careful and robust
­validation of their f­ indings are necessary.
Early noninvasive detection of rejection
has been a goal in the setting of kidney trans‑
plantation for many years. To date no bio‑
marker has been comprehensively validated
that meets all the criteria of a fully predictive
diagnostic test. Hricik and colleagues per‑
formed a beautifully designed and executed
multi-centre study to explore the potential
of immunologic biomarkers as predictors of
acute rejection.5 They found that presence
of the chemokine CXCL9 in the urine had
excellent negative predictive value in deter‑
mining patients at low risk of acute rejection.
Unfortunately as acute rejection has become
a relatively rare phenomenon the search for a
biomarker with high positive predictive value
will be challenging.
One of the barriers to transplantation
has been sensitization of potential trans‑
plant recipients to a wide array of HLA
antigens. These highly sensitized individu‑
als often wait years to receive a transplant
‘‘
...elaborate donation ‘chains’
require elegant and quite complex
algorithms
’’
Another potential partial solution to
helping highly sensitized patients receive a
transplant is through paired kidney dona‑
tion, where donors incompatible with their
intended donors choose to donate to another
individual they are compatible with, while
their intended recipient receives an organ
from a second compatible donor.7 Simple
two-way swaps require tremendous coordi­
nation of logistics, while more elaborate
donation ‘chains’ require elegant and quite
complex algorithms. It is a great testament
to the transplantation community that such
algorithms have been not only developed,
but that extended altruistic donation chains
have been performed. Despite such inno‑
vative approaches, however, many highly
sensitized patients still cannot receive a
­compatible transplant.
The development of effective antiretroviral
therapies has dramatically improved the life
expectancy of patients with HIV infection;

A DECADE IN MEDICINE NOVEMBER 2015  |  55

© 2015 Macmillan Publishers Limited. All rights reserved
NEPHROLOGY
however, many HIV infected individuals
develop end-stage renal disease (ESRD). It
had long been held that the immunosuppres‑
sion necessitated by transplantation would
be dangerous to these individuals and not
provide an adequate risk-to-benefit profile.
In 2010 Stock et al. reported on an NIH
sponsored multi-centre trial of kidney trans‑
plantation in HIV patients with ESRD. 150
patients were enrolled with 3‑year patient
and graft survival of 88% and 73.7%, respec‑
tively.8 Although these results are slightly
inferior to transplantation outcomes in recip‑
ients without HIV infection, this study pro‑
vided hope to individuals with HIV infection
and resulted in the formation of guidelines to
help improve results in the future.
Belatacept is the first biologic agent to
be approved by the FDA for main­tenance
therapy for the prevention of acute rejec‑
tion.9 This drug offers the potential promise
of avoiding toxicities associated with calci­
neurin inhibition, while maintaining
adequate immunosuppression. Early trials
showed improved glomerular filtration
rates, but a potential signal of a greater risk
of Ebstein Barr-related lymphoma. The
role of belatacept in kidney transplantation
remains to be defined.
Although the prevention of acute rejec‑
tion has been the subject of numerous
inquiries, the aetiology of graft loss has only
been a major focus over the past decade.
56  |  NOVEMBER 2015  www.nature.com/reviews
EL‑Zoghby and colleagues looked at the
pathology of patients who suffered graft
loss and defined three major aetiologies:10
antibody-associated glomerular lesions,
interstitial fibrosis with tubular atrophy and
recurrent glomerular disease. Interestingly
calcineurin inhibitor toxicity was not a cause
of late graft loss. This study has revised our
thinking about approaches to prolong graft
survival and further studies in this area are
eagerly awaited.
The papers highlighted here are a subjec‑
tive list and are in no way meant to be all
inclusive. The advances in many areas of
transplantation, including tolerance induc‑
tion, allocation policy, molecular diagnostics,
desensitization protocols and new immuno­
suppressive agents are illustrative of the
major advances that have been made in this
field over the past decade (Box 1). These
advances will hopefully pave the way to
more individualized approaches to the care
of transplant recipients. As analytic skills
catch up with technological advances, earlier
and more accurate diagnostics will become
more commonplace. Further approaches
to decrease the toxic immunosuppressive
burden will also improve the quality of life of
transplant recipients over the coming decade.
Arizona State University, 550 North 3rd Street,
Phoenix, AZ 85004, USA.
bruce.kaplan@asu.edu
© 2015 Macmillan Publishers Limited. All rights reserved
doi:10.1038/nrneph.2015.159
Published online 22 September 2015
Competing interests
The author declares no competing interests.
1. Kawai, T. et al. HLA-mismatched renal
transplantation without maintenance
immunosuppression. N. Engl. J. Med. 358,
353–361 (2008).
2. Friedewald, J. J. et al. The kidney allocation
system. Surg. Clin. North Am. 93, 1395–1406
(2013).
3. Einecke, G. et al. A molecular classifier for
predicting future graft loss in late kidney
transplant biopsies. J Clin. Invest. 120,
1862–1872 (2010).
4. Kurian, S. M. et al. Molecular classifiers for
acute kidney transplant rejection in peripheral
blood by whole genome gene expression
profiling. Am. J. Transplant. 14, 1164–1172
(2014).
5. Hricik, D. E. et al. Multicenter validation of
urinary CXCL9 as a risk-stratifying biomarker for
kidney transplant injury. Am. J. Transplant. 13,
2634–2644 (2013).
6. Montgomery, R. A. et al. Desensitization in
HLA-incompatible kidney recipients and
survival. N. Engl. J. Med. 365, 318–326
(2011).
7. Rees, M. A. et al. A nonsimultaneous,
extended, altruistic-donor chain. N. Engl. J. Med.
360, 1096–1101 (2009).
8. Stock, P. G. et al. Outcomes of kidney
transplantation in HIV-infected recipients.
N. Engl. J. Med. 363, 2004–2014 (2010).
9. Vincenti, F. et al. Costimulation blockade with
belatacept in renal transplantation. N. Engl. J.
Med. 353, 770–781 (2005).
10. El-zoghby, Z. M. et al. Identifying specific
causes of kidney allograft loss. Am. J.
Transplant. 9, 527–535 (2009).
Connect to...

Follow us on Twitter
For the latest commentary and nephrology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Nephrology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Nephrology

Web focuses
A selection of articles on a related theme with content ranging from Research Highlights through to Reviews
and Perspectives

w w w.
Web collections
A selection of articles on a related theme, providing a panoramic view of a key area of nephrology

Conferences
You can meet the editors of Nature Reviews Nephrology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Nephrology!

November 2015 marked the 10th anniversary of the launch of Nature Reviews Nephrology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of nephrology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrneph10

© 2015 Macmillan Publishers Limited. All rights reserved

NEUROLOGY
DECADE IN REVIEW—DEMENTIA
A decade of discovery and disappointment
in dementia research
John R. Hodges
In my personal view, the past decade in dementia research has been marked by remarkable discoveries in
the field of frontotemporal dementia, accompanied by steady scientific consolidation tinged with therapeutic
disappointments related to Alzheimer disease.
A decade ago, it had already been known for
some time that up to 30% of patients with
frontotemporal dementia (FTD) have a
strong family history of early-onset demen-
tia, but the genetic basis of the majority of
cases was unknown. This situation began to
change with the discovery of the progranulin
(GRN) gene in 2006.1 GRN was the second
major FTD-associated gene to be identi-
fied and, like the microtubule-associated
protein tau (MAPT) gene, which was first
linked to FTD a decade earlier, it is located
on chromo­some 17. Mutations in GRN are
associated with tau-negative, ubiquitin-­
positive intraneuronal deposits, and they
turned out to be approximately as common
as the MAPT gene mutations, but their dis-
covery still left many apparently familial
cases of FTD unexplained.
‘‘
In 2011, the last big piece
of the FTD genetic puzzle fell
into place…
’’
Perhaps an even more impor tant
­ iscovery—­in the same year as the GRN
d
gene—was that ubiquitinated TAR DNA-
binding protein 43 (TDP‑43) is the funda-
mental pathological protein in most cases of
tau-negative, ubiquitin-positive FTD.2 This
protein was found in both sporadic and
familial FTD, as well as in the vast major-
ity of cases of amyotrophic lateral sclerosis
(ALS). These observations opened the way
to understanding the pathophysiology of
FTD, and to further genetic discoveries.
In 2011, the last big piece of the FTD
genetic puzzle fell into place with the simul-
taneous discovery, by two large research
groups, of a unique hexanucleotide repeat
expansion in the C9orf72 gene on chromo-
some 9.3,4 This expansion was found in a
57  |  NOVEMBER 2015  www.nature.com/reviews
high proportion of familial FTD and ALS
cases—especially kindreds manifesting
both disorders—as well as in a significant
minority of seemingly sporadic behavioural
variant FTD (bvFTD) cases.
The volume of research published since
2011 related to the basic science of C9orf72
and the clinical characterization of patients
with the C9orf72 syndrome has been
astounding. Interestingly, some patients with
the gene expansion have a protracted history
with prominent psychiatric symptoms and
little in the way of brain atrophy, all of which
makes diagnosis challenging.
The fact that some patients with symp-
toms of bvFTD show minimal progression
over many years has contributed to the
refinement of the diagnostic criteria for
bvFTD, with internationally agreed criteria
for possible, probable and definite bvFTD.5
It is already clear that the ‘probable’ category
is robust, and that the vast majority of such
cases have one of the pathological forms of
FTD. The utility of the ‘possible’ level is still
a topic of active research, and the prevalence
of pathological FTD among patients who do
not have one of the accepted clinical forms of
FTD remains open to question. Nevertheless,
the new criteria, which can be applied con-
sistently by researchers around the world,
represent a substantial step forward.
The other parallel development in the
FTD field has been the clear delineation of
three forms of primary progressive aphasia.6
One of the three types, semantic dementia,
stands out as a unique clinicopathological
entity with a highly characteristic clinical
presentation, a signature pattern of asym-
metric anterior temporal lobe atrophy on
MRI, and underlying TDP‑43 pathology.
Nonfluent aphasia has presented more of
a problem, in that the clinical features vary,
there is no signature MRI appearance, and
© 2015 Macmillan Publishers Limited. All rights reserved
a proportion of cases have underlying AD
pathology at postmortem. A breakthrough
was the realization that the nonfluent cases
could be split into those with agrammatism
and/or apraxia of speech in association
with inferior frontal atrophy, and those with
word-finding difficulty, phonological speech
errors and markedly reduced verbal span, in
association with atrophy around the angular
gyrus. This latter form was termed l­ ogopenic
progressive aphasia (LPA).
It has become clear that the majority of
individuals with LPA have AD pathology,
which is rare in those with true nonfluent
progressive aphasia, who typically have
FTD-tau pathology. This observation sug-
gests that the clinical presentation of AD
is quite varied, and raises questions about
the received wisdom that AD always begins
with tangle formation in the entorhinal
cortex, spreading gradually to involve the
­hippocampus and lateral cortical regions.
An important discovery that unifies
FTD and AD is that the tau pathology
found in AD, and in some cases of FTD,
spreads within the brain by a ‘prion-like’
cell-to-cell process. This phenomenon was
elegantly demonstrated using extracts from
a transgenic mouse that expressed mutant
(Pro301Ser) human tau, which were injected
NPG

into the brains of mice expressing wild-type
human tau.7 This intervention induced the
assembly of the wild-type tau into filaments,
and a progressive spread of pathology from
the site of injection to neighbouring brain
regions. A similar mechanism of prion-like
transmission has now been demonstrated
for a range of tauopathies, and is impli-
cated in other neurodegenerative disorders
associ­ated with the deposition of abnormal
protein aggregates, such as α‑synuclein in
Parkinson disease.
In the AD field, the Dominantly Inherited
Alzheimer Network (DIAN) has brought
together an amazing range of techniques in a
cohort of at-risk individuals recruited across
three continents.8 This impressive effort led
to the finding that subtle changes, such
as a decline in cerebrospinal fluid (CSF)
amyloid‑β (Aβ) 42, occur 25 years before
expected AD symptom onset. In addition,
Aβ deposition can be shown on Pittsburgh B
compound PET (PiB–PET) 15 years before
symptom onset, when tau concentrations
begin to rise in the CSF and subtle brain
atrophy can be detected.
These findings raise fundamental ques-
tions about the possible effectiveness of
disease-modifying treatments for AD, and
perhaps go some way towards explaining
the therapeutic disappointments of the past
decade. Hopes were high for the first Aβ42
(AN1792, Elan Pharmaceuticals) immuni-
zation trial, which was, sadly, stopped due
to the occurrence of severe encephalitis in
a few treated individuals. A group of eight
treated patients from the trial underwent
neuropathological examination post­mortem.
Seven patients exhibited virtually complete
amyloid plaque removal, showing that the
immunotherapy was successful in achieving
the aim of plaque removal, yet their disease
continued inexorably.9 Later analyses con-
cluded that there was no difference in sur-
vival, or time to severe dementia, between
treated and untreated individuals, suggest-
ing that such therapies need to be initiated
much earlier in the disease course, or that
other approaches, perhaps targeted at tau
pathology, are required. Trials of γ‑secretase
inhibitors have also proven a disappointment
and, consequently, effective treatments for
AD remain beyond our reach.
Another highly successful application
of PiB–PET, in combination with meta-
bolic 18F-FDG–PET and quantitative MRI,
has been to map the relationship between
amyloid deposition, hypometabolism and
atrophy early in the course of AD, and
their effects on medial brain structures that
A DECADE IN MEDICINE NOVEMBER 2015  |  58
constitute the default network.10 This work
has brought together a body of knowledge on
the brain systems that are active during quiet
contemplation and memory retrieval, so as to
understand the patterns of c­ ognitive deficits
typically found in mild cognitive impair-
ment, the prelude to full-blown dementia
in AD.
In conclusion, the past 10 years have seen
remarkable discoveries in FTD, tempered by
more-modest, incremental progress in AD
research. We can only hope that the next
decade will be the decade of AD, and that we
will see a giant step forward for mankind in
the treatment of this most p­ ernicious killer.
Neuroscience Research Australia (NeuRA) and
University of New South Wales, Barker Street,
Randwick, Sydney, NSW 2013, Australia.
j.hodges@neura.edu.au
doi:10.1038/nrneurol.2015.191
Published online 6 October 2015;
corrected online 4 November 2015
Competing interests
The author declares no competing interests.
1. Baker, M. et al. Mutations in progranulin cause
tau-negative frontotemporal dementia linked to
chromosome 17. Nature 442, 916–919 (2006).
DECADE IN REVIEW—MULTIPLE SCLEROSIS
New drugs and personalized
medicine for multiple sclerosis
Paul M. Matthews
The past decade of multiple sclerosis research has been marked by
important advances in understanding the disease, a dramatic increase
in the range of treatment options and a new spirit of data sharing in
research for patient benefit. This progress has made personalized
medicine in multiple sclerosis a realistic possibility.
Clinical research into multiple sclerosis (MS)
has evolved substantially in the past 10 years.
The clinical effectiveness of medicines for
relapsing–remitting MS has been established,
and optimism is growing that the challenge
posed by progressive disease can be met. My
personal highlights from the decade include
new evidence for a role of B-cell modula-
tion in the treatment of MS, advances in the
understanding of how immune cells traffic
between the peripheral compartment and the
CNS, and the first steps towards evidence-
based practice of personalized medicine
for MS.
A conceptual breakthrough with immedi­
ate clinical relevance came in 2008 with
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
2. Neumann, M. et al. Ubiquitinated TDP‑43
in frontotemporal lobar degeneration and
amyotrophic lateral sclerosis. Science 314,
130–133 (2006).
3. Renton, A. E. et al. A hexanucleotide repeat
expansion in C9ORF72 is the cause of
chromosome 9p21-linked ALS–FTD. Neuron 72,
257–268 (2011).
4. DeJesus-Hernandez, M. et al. Expanded
GGGGCC hexanucleotide repeat in noncoding
region of C9ORF72 causes chromosome
9p-linked FTD and ALS. Neuron 72, 245–256
(2011).
5. Rascovsky, K. et al. Sensitivity of revised
diagnostic criteria for the behavioural variant
of frontotemporal dementia. Brain 134,
2456–2477 (2011).
6. Gorno-Tempini, M. L. et al. Classification of
primary progressive aphasia and its variants.
Neurology 76, 1006–1014 (2011).
7. Clavaguera, F. et al. Transmission and
spreading of tauopathy in transgenic mouse
brain. Nat. Cell Biol. 11, 909–913 (2009).
8. Bateman, R. J. et al. Clinical and biomarker
changes in dominantly inherited Alzheimer’s
disease. N. Engl. J. Med. 367, 795–804
(2012).
9. Holmes, C. et al. Long-term effects of Aβ42
immunisation in Alzheimer’s disease: follow-up
of a randomised, placebo-controlled phase I
trial. Lancet 372, 216–223 (2008).
10. Buckner, R. L. et al. Molecular, structural, and
functional characterization of Alzheimer’s
disease: evidence for a relationship between
default activity, amyloid, and memory.
J. Neurosci. 25, 7709–7717 (2005).
clinical evidence for a central role of B cells in
the immune response in MS. A phase II trial
showed that a single course of rituximab—­
a monoclonal antibody that selectively targets
and depletes CD20+ B cells—was highly effec-
tive for suppression of new acute inflamma-
tory activity in the disease.1 The mechanisms
by which these B‑cell-directed therapies
modulate the T‑cell-associated disease pro-
cesses still need to be elucidated, but similarly
high efficacy has been observed with other
anti-CD20 monoclonal antibodies that are
now in clinical development. These antibod-
ies promise a new class of treatments for the
relapsing–remitting disease that have high
efficacy and fewer serious adverse events.
NEUROLOGY
These developments emphasize impor-
tant interactions between immune system
effectors in peripheral and CNS compart-
ments, although how CNS antigens are
presented for activation of an adaptive
immune response has been poorly under-
stood. A potentially major breakthrough
came recently in the form of a detailed
anatomical characterization of the way in
which T cells associate with the meninges.
This characterization identified functional
lymphatic vessels that line the dural sinuses
and express molecular markers of lym-
phatic endothelial cells.2 These vessels are
capable of carrying fluid and immune cells
from the cerebrospinal fluid and are con-
nected to the deep cervical lymph nodes.
Together, these characteristics suggest that
a functional lymphatic system is associated
with the meninges. The existence of such
a system in the CNS raises the possibil-
ity that u
­ nexplored mechanisms underlie
neuroinflammatory diseases. For example,
pathology that prevents activated adap-
tive immune cells from exiting the brain
could contribute to disease severity. This
discovery will drive new directions for MS
research in the coming years.
‘‘
One of the biggest clinical
impacts … has been the …
expansion in the range of
’’
approved medicines
The past decade has finally seen genet-
ics come of age as an important tool for the
molecular analysis of disease mechanisms
in MS. The collaborative development of an
MS genetic database of unprecedented size
has enabled the International MS Genetics
Consortium to make major discoveries
about susceptibility factors. A genome-
wide association study that included almost
10,000 people of European descent with MS
identified specific HLA‑DRB1 risk alleles
and confirmed a role for HLA‑A allelic
variants in protecting against MS.3 Such
genetic studies have revealed that immuno-
logically relevant genes (particularly those
involved in T helper cell differentiation)
are substantially over-represented among
those associated with susceptibility to MS,
offering further support to the hypothesis
of a primary immunopathogenesis. Work in
the next decade will address how the inter-
actions of these genes with environmental
factors (such as sunlight and vitamin D)
contribute to variation in the severity of
the disease.
59  |  NOVEMBER 2015  www.nature.com/reviews
Nevertheless, defining the molecular
mechanisms of MS remains a challenge as
long as the disease is mainly diagnosed on
the basis of characteristic features that exist
in an appropriate context, rather than by the
presence of aetiologically specific features
or biomarkers. The discovery that neuro-
myelitis optica (NMO, also known as Devic
disease) usually has a distinct, antibody-
mediated aetio-pathogenesis and is not
simply a manifestation of MS, as previously
thought, thus represented an important
step forward.4 Key to this discovery was the
identification of a serum IgG autoantibody
(AQP4‑IgG) against the membrane protein
aquaporin‑4, which is a component of the
dystroglycan protein complex in astro-
cytic foot processes at the blood–brain
barrier. Evidence suggests that AQP4‑IgG
is a specific marker of NMO and a causal
agent. Diagnostic testing for AQP4‑IgG has
enabled identification of a broader spec-
trum of NMO phenotypes5 and led to the
recognition that the condition of patients
with NMO can deteriorate if treated with
some disease-modifying drugs, making
early accurate diagnosis important in their
clinical management.
One of the biggest clinical impacts of
MS research over the past decade has been
the remarkable expansion in the range of
approved medicines. Six new treat­ments have
been approved for the t­ reatment of relapsing–­
remitting MS since 2005, including the first
orally administered treatments: fingoli-
mod, dimethyl fumarate and teriflunimide.
US and EU approval of ­alemtuzumab—­an
anti‑CD52 monoclonal antibody that leads
to immune reconstitution after induction
of panlymphocytopenia—­for treatment of
MS heralded a new generation of treatments
that are highly effective over long periods,
although alemtuzumab is also associated
with a high rate of serious autoimmunity
related adverse events (for example, thyroid
disease and immune thrombocytopenia).6
The past decade has also seen the first
important steps towards treatments that
limit or halt the worsening of disability in
people with progressive disease. An aca-
demically led phase II trial of high-dose
simvastatin for secondary progressive MS
reached pre-defined clinical and MRI end-
points,7 highlighting the opportunities for
repurposing low-cost drugs. Results of the
RENEW trial suggested that remyelination
promoted by the anti-LINGO antibody
BIIB033 has the potential for neuroprotec-
tion, as treatment slowed retinal neuro­
degeneration and improved visual function
© 2015 Macmillan Publishers Limited. All rights reserved
after acute optic neuritis.8 The next 10 years
might see the first treatment become
­available for this neglected patient group.
The wide variation in convenience,
efficacy and risk of treatments for MS is
fostering a focus on evidence-based per-
sonalization of treatment to optimize the
benefit-to-risk ratio for individual patients.
Research into this approach includes the
development of algorithms that can classify
patients according to their risk of disease
progression and enable treatment deci-
sions to be based on subclinical changes and
responses to treatment.
‘‘
The past 10 years of clinical
research in MS have consolidated
a therapeutic success story for
neurology
’’
The biggest development in personalized
treatment for MS in the past decade was
wide implementation of a risk-based model
for personalization of natalizumab treat-
ment. This quantitative model differentiates
patients according to their serum anti-JC
virus antibody status, the length of time for
which they have been using natalizumab and
their history of immuno­suppressive therapy.9
As a result, a monitoring programme that
involves regular measurements of serum
anti-JC antibody levels, as well as MRI sur-
veillance and rapid MRI evaluation of new
symptoms that are atypical of MS, is now
recommended in all countries in which
natalizumab is used. The near future could
bring insight into the autoimmune mech­
an­isms that underlie adverse responses to
alemtuzumab, enabling markers of pro-
spective risk to be identified and increasing
confidence in our u ­ nderstanding of how the
drug can best be used.
Clinical effectiveness research will be key
in accurately defining the balance between
risk and benefit of treatment for individual
patients. The UK Risk Sharing Scheme
(which ensures access to disease-modifying
treatments for MS through the National
Health Service if patients meet specific
diagnostic criteria and ensures a universal
approach to their follow-up for evaluating
the impact of treatment) has demonstrated
a practical way to couple the results of ‘real-
life’ clinical effectiveness research with
the introduction of new drugs to a health
system. A 2015 report from this large-scale,
pragmatic observational programme con-
firmed the short-term effectiveness of the
original first-line disease-modifying agents

in a population with relapsing–remitting
MS, and showed that this effectiveness
is maintained.10 Global collaborations to
address, at low cost, additional important
questions about patient natural history,
medicine effectiveness and adverse events
are being pioneered by the MSBASE
Consortium and similar organizations
through online registries.
The past 10 years of clinical research
in MS have consolidated a therapeutic
success story for neurology. MS is the
first common, chronic neurodegenerative
disease for which treatments that modify
the course of the disease have been success-
fully developed. The next 10 years are likely
to deliver the new treatments indicated spe-
cifically for progressive disease and exten-
sion of treatment paradigms developed
for MS to the increasing range of other
chronic brain disorders associated with
neuroinflammatory pathology.
Division of Brain Sciences, Department of
Medicine, Hammersmith Hospital, DuCane
Road, London WC12 0NN.
p.matthews@imperial.ac.uk
doi:10.1038/nrneurol.2015.200
Published online 27 October 2015
Acknowledgements
The author gratefully acknowledges personal
support from the Edmond Safra Foundation and
Lily Safra, the Imperial College Healthcare Trust
Biomedical Research Centre, and as an NIHR
Senior Investigator.
Competing interests
The author has received funding for his research
from GlaxoSmithKline and Biogen. The author has
received speaker’s fees or honoraria paid to his
institution from Adelphi Communications, Biogen,
IXICO, GlaxoSmithKline and Novartis.
1. Hauser, S. L. et al. B‑cell depletion with rituximab
in relapsing–remitting multiple sclerosis. N. Engl.
J. Med. 358, 676–688 (2008).
2. Louveau, A. et al. Structural and functional
features of central nervous system lymphatic
vessels. Nature 523, 337–341 (2015).
3. International Multiple Sclerosis Genetics
Consortium et al. Genetic risk and a primary
role for cell-mediated immune mechanisms
in multiple sclerosis. Nature 476, 214–219
(2011).
4. Lennon, V. A., Kryzer, T. J., Pittock, S. J.,
Verkman, A. S. & Hinson, S. R. IgG marker
of optic-spinal multiple sclerosis binds to the
aquaporin‑4 water channel. J. Exp. Med. 202,
473–477 (2005).
5. Tobin, W. O., Weinshenker, B. G. &
Lucchinetti, C. F. Longitudinally extensive
transverse myelitis. Curr. Opin. Neurol. 27,
279–289 (2014).
6. Coles, A. J. et al. Alemtuzumab for patients
with relapsing multiple sclerosis after
disease-modifying therapy: a randomised
A DECADE IN MEDICINE NOVEMBER 2015  |  60
controlled phase 3 trial. Lancet 380,
1829–1839 (2012).
7. Chataway, J. et al. Effect of high-dose
simvastatin on brain atrophy and disability
in secondary progressive multiple sclerosis
(MS‑STAT): a randomised, placebo-controlled,
phase 2 trial. Lancet 383, 2213–2221 (2014).
8. Multiple Sclerosis Discovery Forum [online]
http://www.msdiscovery.org/research-
resources/drug-pipeline/230-biib033 (2015).
DECADE IN REVIEW—EPILEPSY
Edging toward breakthroughs
in epilepsy diagnostics and care
Daniel H. Lowenstein
The past decade has yielded a host of important conceptual advances in
epilepsy, along with some promising findings related to diagnostics and
therapeutics. We are on an upswing where precise identification of the
cause of a patient’s seizure disorder can be matched to therapy that has
a high likelihood of success.
Epilepsy is a disease characterized by recur-
rent seizures, which are transient signs or
symptoms of abnormal, excessive or syn-
chronous neuronal activity in the brain.
Although the pace of translational discov-
eries in epilepsy has not yet matched some
of the dramatic achievements seen in fields
such as oncology and infectious disease, a
number of important advances have been
made in the past decade. These advances
have revised our thinking about some core
concepts related to the pathophysiological
mechanisms of epilepsy. Genetic diagnoses
are increasingly available and provide an
opportunity for patient-tailored treatment.
Moreover, on-demand neurostimulation
devices have become available for patients,
and other promising novel treatment
­strategies are on the horizon.
The conceptual advances cover the gamut
from the literal definition of epilepsy and
the range of symptoms associated with the
disorder, to the fundamental characteristics
of neuronal network dysfunction that con-
stitute a seizure. Prompted by the problems
created by the prevailing notion that the
diagnosis of epilepsy requires the patient to
have at least two unprovoked seizures, a task
force of the International League Against
Epilepsy recently proposed that the diag-
nosis can also be established when a patient
has one seizure in the setting of other
factors that make recurrence likely, such as
a family history of epilepsy, or epileptiform
abnormalities seen on EEG.1 This modifi-
cation to the diagnostic criteria of epilepsy
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
9. Bloomgren, G. et al. Risk of natalizumab-
associated progressive multifocal
leukoencephalopathy. N. Engl. J. Med. 366,
1870–1880 (2012).
10. Palace, J. et al. Effectiveness and cost-
effectiveness of interferon beta and glatiramer
acetate in the UK Multiple Sclerosis Risk
Sharing Scheme at 6 years: a clinical cohort
study with natural history comparator. Lancet
Neurol. 14, 497–505 (2015).
will have a major impact on direct patient
care in terms of making diagnostic and
treatment decisions, as well as on clinical
research, ­including epidemiological studies.
‘‘
…we are at the dawn of
’’
precision medicine in epilepsy
Another profound recent innovation in
thinking about epilepsy relates to the range of
behavioural and patho­physio­logical abnor-
malities associated with the disorder.2 By far
the dominant conceptualization of epilepsy
has been that it is a brain disease charac­
terized by recurrent seizures. However,
this view fails to recognize the existence of
numerous comorbidities in many patients,
including problems in cognition (such as
memory loss, cognitive slowing and autism)
and psychiatric diseases (such as depression,
anxiety and certain personality disorders).
Thus, it seems highly likely that many forms
of epilepsy are complex neuronal network
abnormalities that vary in their timescale
and triggers, with seizures being but one of a
variety of behavioural disturbances.
Our understanding of the basic electro-
physiological properties of a seizure has also
evolved in recent years. Until recently, the
canonical view was that a seizure arose from
the hyperexcitability and hypersynchroni-
zation of a small network of neurons, with
gradual spread depending on the degree to
which surround inhibition is maintained.
However, recent studies using single micro-
electrodes demonstrated that a seizure
NEUROLOGY
begins with ‘micro-seizures’ emanating
asynchronously from small neuronal clus-
ters, which then coalesce and constitute
the macroscale hypersynchronization that
characterizes most seizures.3 These findings
have important implications regarding our
understanding of the stochastic processes
that seem to govern the transition of a cor-
tical region from the interictal to the ictal
state and, thus, our ability to predict the
occurrence of a seizure.
The past 5–6 years have seen some clear
breakthroughs in our understanding of the
genetic basis of epilepsy. These advances
follow something of a ‘dark age’ between
2001 and 2009, when some epilepsy-related
genes were discovered among families with
rare epilepsy syndromes that followed a
Mendelian inheritance pattern, but over 100
genome-wide association studies largely pro-
duced negative or non-­reproducible results.
Arguably the most important discovery,
based on the global collaboration of several
research groups, was the recognition that
de novo mutations are the explanation for a
substantial proportion of patients with so-
called ‘epileptic encephalopathies’—severe
forms of treatment-resistant epilepsy that
arise early in life and are often associated
with profound developmental delay.4 These
findings have led to expanded use of clini-
cal genetic testing in the routine evaluation
of patients with these syndromes and, as a
consequence, a marked improvement in the
ability to establish a definitive diagnosis and
provide accurate genetic counselling.
The causes of the most common forms
of epilepsy—generalized epilepsy and
nonlesional focal epilepsy, both of which
are thought to have a strong genetic
component—­continue for the most part to
elude us despite intensive research (although
major improvements in brain imaging have
reduced the number of patients previously
diagnosed as ‘non­lesional’). However, two
areas of study have shed light on the genetic
architecture in this group of patients. First,
NPG
61  |  NOVEMBER 2015  www.nature.com/reviews
copy number variants, which have been
identified as the basis for a variety of rare
epilepsy syndromes, were shown to account
for at least a few percent of the common
epilepsies.5 Second, a recent large meta-
analysis of genome-wide association studies
of this patient population found only three
loci that reached genome-wide signifi-
cance, providing further evidence that the
common epilepsies are likely to be explained
by the combined influence of numerous
common variants of small effect, or multiple
rare variants.6 To adequately assess both
possibilities, further progress will require
the application of next-generation sequen­
cing to the analysis of genomes from tens of
thousands of patients.
Another major diagnostic advance has
been the recognition that a paraneoplastic
syndrome could be the basis of epilepsy of
particularly explosive and malignant onset
in patients who were previously healthy.7
Interestingly, in some cases, antibodies are
generated against targets that are also impli-
cated in genetic epilepsies. Thus, evaluation
in patients with this presentation increas-
ingly relies on assays for autoantibodies
directed against CNS autoantigens such as
potassium channels or glutamate receptors,
as well as a search for an underlying cancer.
Despite the introduction of well over a
dozen new antiepileptic drugs (AEDs) since
1970, there is little evidence that therapeutic
efficacy has changed. AEDs continue to fail
in approximately 30% of patients, although
some of the new medications arguably offer
better options in terms of ease of use and
adverse effect profile. The past 10 years
have, however, seen a definite improvement
in understanding the management of epi-
lepsy in women, as important studies have
demonstrated the effects of specific AEDs
on pregnancy outcome, safety of breast
milk, and the development of osteoporosis.
The past decade has also seen the intro-
duction of new and promising alter-
natives to the traditional medical and
surgical approaches for treating epilepsy.
Novel proof-of-principle experiments using
animal models have shown that closed-loop
systems incorporating EEG detection and
optogenetic control of neuronal activity can
rapidly abort seizures at onset, and implan-
tation of inhibitory neuron precursors
can reverse the process of epileptogenesis,
even in the adult animal.8,9 Pivotal clinical
trials have recently led to the approval of
new neurostimulation devices for patients,
including a closed-loop system that detects
incipient seizure activity with intra­cranial
© 2015 Macmillan Publishers Limited. All rights reserved
electrodes, which then deliver focal
­stimulation to terminate the seizures.
Finally, the genetic diagnoses described
previously are beginning to provide guid-
ance for treatment decisions. Although the
findings should be considered preliminary,
there are now examples of patients with
defined, causative mutations who respond
to specific drugs prescribed on the basis of
in vitro assays of pharmacoresponsiveness.
This step towards individualized treatment
together with the recognition that some of
the underlying mechanisms of epilepto­
genesis can be mapped onto distinct cellu-
lar pathways (such as mTOR, and pathways
related to synaptic machinery), has led to
the suggestion that we are at the dawn of
precision medicine in epilepsy.10
Department of Neurology, University of
California, San Francisco, 513 Parnassus
Avenue, Room S‑115, San Francisco,
CA 94143, USA.
lowenstein@ucsf.edu
doi:10.1038/nrneurol.2015.193
Published online 13 October 2015
Acknowledgements
D.H.L. has received research grants from the NIH
(U01NS077274 and U01NS077276).
Competing interests
The author declares no competing interests.
1. Fisher, R. S. et al. ILAE official report:
a practical clinical definition of epilepsy.
Epilepsia 55, 475–482 (2014).
2. Jensen, F. E. Epilepsy as a spectrum disorder:
implications from novel clinical and basic
neuroscience. Epilepsia 52, 1–6 (2011).
3. Truccolo, W. et al. Single-neuron dynamics
in human focal epilepsy. Nat. Neurosci. 14,
635–641 (2011).
4. Epi4K Consortium & Epilepsy Phenome/
Genome Project. De novo mutations in epileptic
encephalopathies. Nature 501, 217–221
(2013).
5. Helbig, I. et al. 15q13.3 microdeletions
increase risk of idiopathic generalized epilepsy.
Nat. Genet. 41, 160–162 (2009).
6. International League Against Epilepsy
Consortium on Complex Epilepsies. Genetic
determinants of common epilepsies: a meta-
analysis of genome-wide association studies.
Lancet Neurol. 13, 893–903 (2014).
7. Seeck, M., Zacharia, A. & Rossetti, A. O.
Autoimmune epilepsy [French]. Rev. Med.
Suisse 6, 925–929 (2010).
8. Paz, J. T. et al. Closed-loop optogenetic control of
thalamus as a tool for interrupting seizures after
cortical injury. Nat. Neurosci. 16, 64–70 (2013).
9. Hunt, R. F., Girskis, K. M., Rubenstein, J. L.,
Alvarez-Buylla, A. & Baraban, S. C. GABA
progenitors grafted into the adult epileptic
brain control seizures and abnormal behaviour.
Nat. Neurosci. 16, 692–697 (2013).
10. EpiPM Consortium. A road map for precision
medicine in the epilepsies. Lancet Neurol.
http://dx.doi.org/10.1016/S1474-
4422(15)00199-4.

DECADE IN REVIEW—MOVEMENT DISORDERS
Tracking the pathogenesis
of movement disorders
Oksana Suchowersky
Since 2005, we have made substantial progress in understanding
the pathophysiology and natural history of movement disorders such
as Parkinson disease and Huntington disease. However, disease-
modifying therapies for these conditions have proved elusive and,
as a consequence, treatments remain largely symptomatic.
Over the past few decades, the increasing
recognition of movement disorders as a sub-
speciality has led to the development of clini-
cal programmes dedicated to diagnosis and
management of disorders such as Parkinson
disease (PD), Huntington disease (HD),
dystonia, tics, and tremors. The forma-
tion of multinational collaborative research
groups, such as the Parkinson Study Group
and the Huntington Study Group, has further
enhanced awareness and fuelled expan-
sion of research. Despite improved under-
standing of the pathophysiology, aetiology
and evolution of these disorders, however,
treatment remains largely symptomatic.
‘‘
…recently recognized
autoimmune factors … result in
paroxysmal and nonparoxysmal
’’
movement disorders
PD is one of the most common neuro­
degenerative disorders of ageing, affecting 1
in 100 individuals over the age of 55 years.
Diagnosis remains clinical, based on well-
recognized motor symptoms and defined
inclusion and exclusion criteria.1 However,
the past 10 years have seen increased rec-
ognition of the fact that Lewy body for-
mation and α‑synuclein deposition in
neurons predate the development of motor
symptoms by several decades. At the pre-
symptomatic stage of PD, pathology can
be seen in the anterior olfactory nucleus,
dorsal motor nucleus of the glosso­pharyn­
geal and vagal nerves, and gastro­intestinal
nerves. The pathology subsequently spreads
to lower brainstem nuclei before involve-
ment of the substantia nigra. Thus, olfac-
tory dysfunction, anxiety, depression, REM
sleep behaviour disorder and constipation
have been identified as prodromal symp-
toms of PD, and the development of motor
symptoms indicates moderate disease
(Braak stage 3).2
A DECADE IN MEDICINE NOVEMBER 2015  |  62
With disease progression, nonmotor
symptoms (NMS) that result from involve-
ment of nondopaminergic neurotransmitter
systems become increasingly evident, result-
ing in substantial disability and poor quality
of life. As highlighted by the PRIAMO study
published in 2009,3 over 98% of patients with
PD report NMS; the number and severity
of symptoms increase with disease dura-
tion. Autonomic dysfunction (constipation,
bladder urgency and postural hypotension),
sleep disturbances and pain require prompt
recognition, as many of these symptoms can
be treated. Over 50% of individuals with PD
develop apathy, fatigue, depression and/or
anxiety;4 cognitive dysfunction and demen-
tia are present in the majority of patients
at the advanced stages. The development
of validated rating scales for both motor
symptoms and NMS has led to improved
monitoring of symptoms. However, reliable
biomarkers to accurately measure disease
progression in PD are still unavailable.
Although descriptions of PD date back
5,000 years, its aetiology remains elusive.
Various environmental factors, including
traumatic brain injury, pesticide exposure,
decreased coffee intake and lack of cigar­
ette smoking, have been associated with an
increased risk of developing PD. Over the
past decade, however, an important role of
genetic factors, even in sporadic cases, has
become increasingly evident.5 Mutations
in over 10 autosomal recessive and auto­
somal dominant genes are now known to
have a major role in at least 10% of affected
families.5 Whole-exome and whole-genome
studies, as well as carefully designed
fam­ily studies, are increasing the number of
identified genetic susceptibility loci. These
discoveries have provided valuable insights
into the pathogenesis of PD, including
common molecular pathways that involve
protein misfolding and aggregation, and
have also enabled the ­development of new
animal models.5
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
Since 1978, a variety of neuroprotective
or disease-modifying treatments have been
studied in an attempt to slow disease pro-
gression in PD. The largest and most recent
study, published in 2015, was NET-PD, a
double-blind, placebo-controlled NIH-
funded trial of creatine in 1,741 individuals.6
Like previous trials of potential disease-
modifying treatments in PD, the NET-PD
trial was unsuccessful, perhaps owing to the
fact that the pathology in the study cohort
had progressed to a point at which benefit
was precluded. Intervention in individuals
identified as being at risk of PD could lead to
improved strategies for disease modification
in the future.
HD is an autosomal dominant, slowly
progressive neurodegenerative disorder
characterized by chorea and psychiatric
and cognitive changes. In contrast to PD,
the cause is well known: an expanded CAG
triplet repeat in the huntingtin (HTT) gene
on chromosome 4. The disease typically
manifests in midlife, although the age of
onset ranges from early childhood to late
life, depending on the size of the triplet
repeat. Rate of progression also depends
on the repeat size. As in PD, pathological
changes predate symptom onset, with neuro­
nal deposition of the abnormal ­huntingtin
protein identified in the fetal brain.
‘‘
…new insights have been
gained into the premanifest and
early symptomatic stages of PD
’’
and HD…
The multinational TRACK‑HD study
was initiated to determine the natural
history of HD in the premanifest and early
symptomatic stages, and the 36-month data
were reported in 2013.7 Identification of
asymptomatic carriers of the HTT expan-
sion allows long-term evaluation in order
to objectively identify disease onset and
progression, and develop bio­m arkers.
TRACK-HD showed that atrophy of the
basal ganglia, cortex and white matter, and
loss of structural connectivity between the
basal ganglia and cortex, can be seen on
serial imaging years before symptom onset.
Disease-associated proteins identified in
the cerebrospinal fluid could serve as bio-
markers of disease progression, as well as
facilitating the exploration of potential
disease-modifying therapies. Advances in
genetic therapies to silence mutant HTT
gene and mRNA activity, involving zinc
finger proteins, antisense oligonucleotides
NEUROLOGY
and mRNA interference techniques, have
proved to be effective in animal models, and
could lead to successful disease modification
in the near future.8
With respect to other hyperkinetic dis­
orders, genotyping has led to improved
classification, as well as an appreciation of
genetic and phenotypic heterogeneity. For
example, mutations in the SLC2A1 gene
that result in GLUT1 deficiency syndrome
can lead to paroxysmal movement disorders
in addition to the classically recognized syn-
drome of developmental delay and seizures.9
The triplet repeat expansion associated
with spinocerebellar ataxia type 2 can cause
ataxia, levodopa-responsive parkinsonism
or amyotrophic lateral sclerosis. A number
of recently recognized autoimmune factors,
such as anti‑N-methyl‑d-aspartate receptor
antibodies, can result in paroxysmal and
nonparoxysmal movement disorders.10
In summary, over the past decade, the
diagnosis of movement disorders has
improved and expanded with the identifi-
cation of genetic and environmental factors,
leading to the development of new classifica-
tions, validated rating scales, and biomarkers
to measure disease progression. In particu-
lar, new insights have been gained into the
premanifest and early symptomatic stages
of PD and HD, which should facilitate the
discovery of strategies for early intervention
in these diseases. Animal models have pro-
vided a better understanding of the patho-
physiology of movement disorders, and are
enabling the development of new therapeu-
tic approaches. We are optimistic that these
advances will lead to better treatments and
disease modification in the near future.
University of Alberta, 7‑112Q Clinical Sciences
Building, 11350—83 Avenue, Edmonton,
AB T6G 2G3, Canada.
oksana.suchowersky@albertahealthservices.ca
doi:10.1038/nrneurol.2015.201
Published online 27 October 2015
Competing interests
The author has received research grants from
AbbVie and Biotie. She has received honoraria
from AbbVie, Allergan and Merz.
1. Suchowersky, O. et al. Practice parameter:
diagnosis and prognosis of new onset
Parkinson disease (an evidence-based review).
Neurology 66, 968–975 (2006).
2. Tolosa, E., Gaig, C., Santamaria, J. & Compta, Y.
Diagnosis and the premotor phase of Parkinson
disease. Neurology 72 (Suppl. 2), S12–S20
(2009).
3. Barone, P. et al. The PRIAMO study:
a multicenter assessment of nonmotor
symptoms and their impact on quality of life
in Parkinson’s disease. Mov. Disord. 24,
1641–1649 (2009).
63  |  NOVEMBER 2015  www.nature.com/reviews
4. den Brok, M. G. et al. Apathy in Parkinson’s
disease: a systematic review and meta-
analysis. Mov. Disord. 30, 759–769 (2015).
5. Van der Brug, M. P., Singleton, A, Gasser, T. &
Lewis, P. A. Parkinson’s disease: from human
genetics to clinical trials. Sci. Transl. Med. 7,
305ps20 (2015).
6. Writing Group for the NINDS Exploratory Trials in
Parkinson Disease (NET-PD) Investigators et al.
Effect of creatine monohydrate on clinical
progression in patients with Parkinson disease:
a randomized clinical trial. JAMA 313, 584–593
(2015).
7. Tabrizi, S. J. et al. Predictors of phenotypic
progression and disease onset in premanifest
and early-stage Huntington’s disease in the
TRACK-HD study: analysis of 36-month
DECADE IN REVIEW—STROKE
Progress in acute ischaemic
stroke treatment and prevention
Jose G. Romano and Ralph L. Sacco
Recent decades have seen a dramatic reduction in age-adjusted stroke-
related mortality, presumably owing to better control of vascular risk
factors, use of antithrombotic agents and improvements in acute stroke
care. Here, we highlight a few developments in stroke prevention and
acute care that have particularly influenced the care of patients.
Treatment of acute ischaemic stroke has
advanced at a remarkable pace in the past
10 years. One important advance was the
broadening of the therapeutic window for
intravenous recombinant tissue plasmino-
gen activator (rtPA) in acute ischaemic
stroke. Two important trials expanded
the original eligibility criteria (initiat-
ing treatment within 3 h from symptom
onset) for intravenous thrombolytic agents.
The European Cooperative Acute Stroke
Study III evaluated the efficacy of throm-
bolysis conducted 3–4.5 h after the onset
of symptoms.1 A favourable outcome of no
disability (modified Rankin Scale [mRS]
score 0–1) was achieved in 52.4% of patients
who were treated with rtPA, in comparison
with 45.2% in the placebo arm of the study,
yielding a significant odds ratio of 1.34 for
the primary outcome of no disability.1 The
third International Stroke Trial demon-
strated that individuals older than 80 years
derived benefits similar to those seen in
younger patients.2 Recent meta-analysis of
individual patient data from 9 randomized
controlled trials of intravenous rtPA demon-
strated better outcomes for patients treated
within 4.5 h of symptom onset than for those
given no rtTPA, irrespective of stroke sever-
ity and age. The benefit was greater with
© 2015 Macmillan Publishers Limited. All rights reserved
observational data. Lancet Neurol. 12,
637–649 (2013).
8. Stanek, L. M. et al. Silencing mutant huntingtin
by adeno-associated virus-mediated RNA
interference ameliorates disease
manifestations in the YAC128 mouse model
of Huntington’s disease. Hum. Gene Ther. 25,
461–474 (2014).
9. Leen, W. G. et al. Glucose transporter‑1
deficiency syndrome: the expanding clinical and
genetic spectrum of a treatable disorder. Brain
133, 655–670 (2010).
10. Bigi, S., Hldio, M., Twilt, M., Dalmau, J. &
Benseler, S. M. The growing spectrum of
antibody-associated inflammatory brain
diseases in children. Neurol. Neuroimmunol.
Neuroinflamm. 2, e92 (2015).
earlier thrombolysis (for mRS 0–1, OR 1.75
at <3 h versus OR 1.26 at >3 to <4.5 h from
symptom onset).3
Despite these expanded indications,
thrombolytics are of benefit for only a minor-
ity of patients with stroke, largely because
of the delay in recognizing symptoms and
the short therapeutic time window. Early
enthusi­asm for endovascular recanaliza-
tion was dampened by three trials published
in 2013 that failed to show improvements in
clinical outcomes, highlighting the need for
earlier recanalization, more-effective devices
and appropriate patient selection. Five ran-
domized controlled trials that incorporated
these lessons and were published in 2015
showed substantial benefits of mechanical
endo­vascu­lar recanalization (MER). These
studies enrolled patients with anterior cir-
culation strokes and large vessel occlusion
(most of whom were treated with intra­
venous rtPA) and employed new-generation
stent retrievers. Meta-analysis of the 2015
trials showed that MER after intravenous
rtPA was associated with an increased likeli-
hood of excellent outcomes at 90 days (mRS
0–1 OR 2.59, 95% CI 1.92–3.48) and func-
tional outcomes (mRS 0‑2 OR 2.48, 95% CI
1.95–3.15), with no difference in mortality
or symptomatic intracerebral haemorrhage.4

400
MR RESCUE*
360
Time to reperfusion (mins)
320
280
240
200
0 have not demonstrated superiority of
0 10 20
Percentage of patients with mRS score ≤2
Figure 1 | Association between time to
recanalization and clinical
Nature outcomes in trials
Reviews | Neurology
of mechanical endovascular recanalization.
Across six trials, the mean time from
symptom onset to recanalization* therapy
correlated with the likelihood of low disability
(mRS score ≤2; little or no disability) at
90 days. *In the MR RESCUE study, the time
range used was from the latest time point at
which the individual was known to be well to
the groin puncture. Abbreviation: mRS,
modified Rankin Scale.
The key differences between the new
endovascular trials and earlier studies
were a major decrease in the time to
vessel recanal­ization and tissue perfusion
(Figure 1), use of next-generation devices,
and improved selection of patients with
large vessel occlusion.5 Other imaging selec-
tion criteria have been applied in individual
trials, including estimation of the infarct
core by diffusion MRI or Alberta Stroke
Program Early CT Score (ASPECTS),
assessment of perfusion mismatch with
CT or MRI techniques, and detection of
adequate collaterals by multiphase CT
angiography. Whether implementation of
these imaging modalities results in better
outcomes or extension of the r­ ecanalization
window is being investigated.
In light of the current data, the American
Heart Association/American Stroke
Association has updated the guidelines
for acute stroke management, providing a
strong recommendation (Class I, level of
evidence A) for MER with stent retriev-
ers for patients aged ≥18 years who were
previously independent, were treated with
intravenous rtPA within 4.5 hours of stroke
onset, had an NIH Stroke Scale score >6, had
internal carotid or middle cerebral artery
M1 segment occlusion and had anASPECTS
>6, as long as endovascular treatment can
be initiated within 6 h of symp­tom onset.6
To realize the benefits of intravenous and
endovascular therapy for acute ischaemic
A DECADE IN MEDICINE NOVEMBER 2015  |  64
stroke, reductions in the time to recanal­
ization are required; these reductions could
Revascat be achieved through improvements to
IMS3
processes, including prehospital transport
policies and in-hospital workflows, facili-
tated by data-driven quality performance
SWIFT PRIME improvement programmes.
ESCAPE In contrast to the recent success of endo-
Extend IA
vascular devices in acute stroke treatment,
endovascular approaches for stroke pre-
vention have been less successful. Trials
30 40 50 60 70 80 patent foramen ovale closure or stenting of
recently symptomatic intracranial athero­
sclerotic disease over usual care, thus
emphasizing the importance of medical
management. Moreover, gaps in the evi-
dence persist regarding the choice between
extracranial carotid artery stenting, carotid
endarterectomy, and medical manage-
ment for s­ ymptomatic and ­asymptomatic
carotid stenosis.
Medical prevention of stroke that results
from atrial fibrillation has advanced dra-
matically. For many decades, warfarin was
the only effective strategy in this context,
but was under-utilized because warfarin
has a narrow therapeutic range and multi-
ple interactions with medications and foods
making frequent testing and dose adjust-
ment necessary. New non-vitamin‑K oral
anticoagulants (NOAC) do not require dose
adjustment or monitoring. Currently avail-
able agents for non-valvular atrial fibril-
lation include a direct thrombin inhibitor
(dabigatran) and three factor Xa inhibitors
(rivaroxaban, apixaban and edoxaban).
Multiple large international randomized
trials of these drugs have confirmed their
prespecified primary hypotheses. Meta-
analysis has confirmed that the available
NOACs reduce stroke or systemic embolism
by 19% when compared with warfarin (RR
0.81, 95% CI 0.73–0.91) and reduce major
haemorrhagic events by 14% (RR 0.86, 95%
CI 0.73–1.00).7 The benefits are similar for
those with and without previous cerebral
ischaemic events. Based on the available
evidence, NOACs are recommended in
guidelines for patients with non-valvular
atrial fibrillation who are at risk of stroke.8
Ongoing studies are attempting to develop
ways to rapidly detect and reverse the
­anticoagulant effects of NOACs.
Approximately 30% of ischaemic strokes
are classified as cryptogenic, and are usually
treated with antiplatelet agents. Prolonged
electrocardiographic monitoring reveals
atrial fibrillation in a proportion of crypto-
genic strokes, though the diagnostic yield
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
varies with the thoroughness of causative
evaluation and the frequency and modal-
ity of monitoring. The Cryptogenic Stroke
and Underlying Atrial Fibrillation study
addressed these concerns; patients with
cryptogenic cerebral ischaemic events
within 90 days of extensive diagnostic
testing were randomly assigned to receive
an insertable cardiac monitor or a con-
ventional Holter monitor.9 Among the 441
enrolled patients, the rate of atrial fibril-
lation detection with continuous moni-
toring was 8.9%, 12.4% and 30% at 6, 12
and 36 months, respectively, compared
with 1.4%, 2% and 3% with conventional
follow-up. Atrial fibrillation was defined
as lasting >30 s, but 92.3% of patients had
at least one episode of longer than 6 min.
These findings support the recommenda-
tion of extended cardiac rhythm monitor-
ing for appropriately selected patients with
cryptogenic stroke, because such monitor-
ing could indicate the need for a change
in therapy. Despite extensive evaluations,
including prolonged cardiac monitoring,
some patients with cryptogenic ischaemic
stroke could be reclassified as having an
embolic stroke of undetermined source.
Ongoing trials are investigating the use of
NOACs versus aspirin for secondary stroke
prevention in this subgroup.
‘‘
In the coming decade,
research and health policy for
stroke should focus on promoting
healthy lifestyles…
’’
Despite a reduction in age-adjusted mor-
tality in recent decades, the global burden of
stroke prevalence and disability is increas-
ing, with a disproportional effect on low-
income countries and minorities.10 With
an ageing population and growing preva-
lences of obesity, physical inactivity and
diabetes mellitus, the prevalence of stroke
is expected to increase, at a substantial cost
to society. In the coming decade, research
and health policy for stroke should focus
on promoting healthy lifestyles, improving
the control of hypertension and diabetes,
expanding access to health care, enhan­
cing systems of care to provide timely
and effective care, exploring better rehab­
ilitative measures, and reducing health
disparities. The many successes of the last
decade for stroke prevention and treatment
should provide a solid foundation for con-
tinued progress, but many challenges are
still ahead.
NEUROLOGY
University of Miami, Miller School of Medicine,
1120 NW 14th Street, Miami, FL 33136, USA
(J.G.R., R.L.S.).
Correspondence to: R.L.S.
rsacco@med.miami.edu
doi:10.1038/nrneurol.2015.199
Published online 20 October 2015
Acknowledgements
J.G.R. and R.L.S. have received research salary
support from the NIH/NINDS (1R01NS29993,
R01 NS40807 and and 1R01NS084288).
Competing interests
J.G.R. has received research support for MaRISS
Study and honoraria from Genentech for a steering
committee role, and honoraria and stock from Vycor
NovaVision for a scientific advisory board role.
R.L.S. has consulted for Boehringer Ingelheim as
co-chair of the RESPECT-ESUS trial with dabigatran
versus aspirin for secondary stroke prevention.
1. Hacke, W. et al. Thrombolysis with alteplase 3
to 4.5 hours after acute ischemic stroke.
N. Engl. J. Med. 359, 1317–1329 (2008).
2. IST‑3 collaborative group. et al. The benefits
and harms of intravenous thrombolysis with
recombinant tissue plasminogen activator
within 6 h of acute ischaemic stroke (the third
international stroke trial [IST‑3]): a randomized
controlled trial. Lancet 379, 2352–2363 (2012).
3. Emberson, J. et al. Effect of treatment delay,
age, and stroke severity on the effects of
intravenous thrombolysis with alteplase for
acute ischaemic stroke: a meta-analysis of
individual patient data from randomised trials.
Lancet 384, 1929–1935 (2014).
4. Falk-Delgado, A., Kuntze Söderqvist, Å.,
Fransén, J. & Falk-Delgado, A. Improved clinical
outcome 3 months after endovascular
treatment, including thrombectomy, in patients
with acute ischemic stroke: a meta-analysis.
J. Neurointerv. Surg. http://dx.doi.org/
10.1136/neurintsurg-2015-011835.
5. Prabhakaran, S., Ruff, I. & Bernstein, R. A.
Acute stroke intervention: a systematic review.
JAMA 313, 1451–1462 (2015).
6. Powers, W. J. et al. 2015 AHA/ASA focused
update of the 2013 guidelines for the early
management of patients with acute ischemic
stroke regarding endovascular treatment:
a guideline for healthcare professionals from
the American Heart Association/American
Stroke Association. Stroke http://dx.doi.org/
10.1161/STR.0000000000000074.
7. Ruff, C. T. et al. Comparison of the efficacy and
safety of new oral anticoagulants with warfarin
in patients with atrial fibrillation: a meta-
analysis of randomised trials. Lancet 383,
955–962 (2014).
8. Kirchhof, P. et al. Atrial fibrillation guidelines
across the Atlantic: a comparison of the
current recommendations of the European
Society of Cardiology/European Heart
Rhythm Association/European Association of
Cardiothoracic Surgeons, the American College
of Cardiology Foundation/American Heart
Association/Heart Rhythm Society, and the
Canadian Cardiovascular Society. Eur. Heart J.
34, 1471–1474 (2013).
9. Sanna, T. et al. Cryptogenic stroke and
underlying atrial fibrillation. N. Engl. J. Med.
370, 2478–2486 (2014).
10. Feigin, V. L. et al. Global and regional burden of
stroke during 1990–2010: findings from the
Global Burden of Disease Study 2010. Lancet
383, 245–254 (2014).
65  |  NOVEMBER 2015  www.nature.com/reviews
DECADE IN REVIEW—MIGRAINE
Incredible progress for an era
of better migraine care
Peter J. Goadsby
In the past 10 years, the realization that migraine is a brain disorder
rather than a vascular disorder has facilitated development of various
treatments, ranging from innovative immunopharmaceuticals through
to neurostimulation. Many clinical trials have been successful, and
such considerable progress holds promise for the coming decade of
migraine treatment.
Migraine has been the sixth most common
cause of disability in the world since 1990.
In the same period, medication overuse
headache—which seems only to occur in
patients with migraine—has emerged from
below 25th to rank 18th. By contrast, progress
has been made with many of the other top
causes: anxiety disorders, iron-deficiency
anaemia and diarrhoeal diseases, to name
just a few.1 This burden of migraine poses
a challenge, but increasing recognition of
the condition offers opportunities. In the
past decade, laboratory and translational
neurobiology has pushed towards a better
understanding of migraine and new thera-
pies. Here, I will highlight the therapies that
are emerging from this work.
An important theme of research in the
past decade has been the steady accumula-
tion of data that point to migraine as a brain
disorder rather than a vascular disorder.
Imaging studies in humans have focused
this discussion further; for example, one
beautifully executed study used magnetic
resonance angiography to produce clearly
negative results and further demonstrate
that no important vascular change occurs
in migraine.2 Moreover, investigation of the
premonitory phase of migraine also indi-
cates brain targets and emphasizes the need
to understand the pathogenesis of attacks in
order to develop new therapies.3
As the biology of migraine has been
unravelled, further questions with clini-
cal implications have arisen. For example,
a carefully conducted study showed that a
light challenge did not trigger a migraine
in any participants who thought their
migraines could be triggered by light.4 This
observation raises the question of whether
the fact that patients notice light in the
premonitory phase contributes to their
attribution of light as a trigger. If so, the
possibility exists that similar phenomena
explain some aspects of migraine triggered
© 2015 Macmillan Publishers Limited. All rights reserved
by food, sleep or even stress, as each could
be the result of alterations in brain func-
tion, possibly subcortical.5 Such questions
illustrate the point that a greater under-
standing of the pathogenesis of migraine
would facilitate the development of therapy
and enable clinicians to provide better basic
management advice.
An understanding of migraine as a brain
disorder has freed the development of
medicine from the shackles and concerns
of the cranial vessels. As a result, such devel-
opment in the past decade has been driven
towards targets other than vasoconstrictors
(Supplementary Table 1). The most widely
explored candidates have been treatments
that target the calcitonin gene-related
peptide (CGRP) pathway.
The CGRP pathway was first suggested as
a target by translational studies conducted
at the end of the 1980s6,7 that pointed away
from involvement of inflammatory targets.
Studies conducted within the past decade
in which CGRP has been targeted are pro-
viding substantial advances in knowledge
that even extend beyond migraine. First,
the general principle that a migraine medi-
cine can either target an acute attack or
prevent attack onset seems to be completely
incorrect: targeting the CGRP pathway is
effective in both contexts (Supplementary
Table 1). This finding suggests that further
interrogation of the biology of migraine,
for example further investigation of the
trigeminovascular system, could yield
additional targets for therapeutics that
provide both attack prevention and acute
alleviation—wonderfully unortho­dox, and
certainly encouraging. Second, targeting
the CGRP pathway drives home the fact
(borne out by the effect of aspirin and the
5‑HT1F receptor agonist lasmiditan) that
non-vasoconstrictors work: human studies
have shown that CGRP ­receptor antagonists
are not vasoconstrictors.8

Furthermore, development of the CGRP
monoclonal antibody is the first example of
immunopharmacology in which antibodies
target a molecular pathway instead of modi­
fying immune pathways. Antibodies are
highly specific and can target systems that
were previously inaccessible to therapeu-
tics, making antibody therapy a p ­ romising
­strategy for the treatment of migraine.
‘‘
Millennia after migraine was
first reported, clinicians will be
able to suggest migraine-specific
’’
preventative treatments...
The past 10 years have also taught us
lessons about other potential targets for
migraine therapy. A decade ago, nitric
oxide seemed to be a promising target, yet
still eludes. Intravenous administration of
the nonspecific neuronal nitric oxide syn-
thase (NOS) inhibitor 546C88 was effective.
However, tests of inducible NOS as a target
in both acute attack and prevention failed.
These results, together with failures of four
substance P–neurokinin 1 receptor antago-
nists, two neurogenic plasma protein extra­
vasation antagonists and a TRPV1 receptor
antagonist (Supplementary Table 1), indicate
that the role of inflammatory mechanisms
in migraine must be reconsidered and that
therapeutic ­targeting of these mechanisms
needs rethinking.
Similarly, orexinergic targets seemed
attractive 10 years ago on the basis of basic
research, but have now been shown to be
ineffective, at least when both orexin recep-
tors are targeted (Supplementary Table 1).
Testing of these targets is incomplete, as
proper targeting of just one or other orexin
receptor might have anti-migraine effects.
Certainly, dual receptor antagonists are
impractical because their sleep-inducing
effects preclude adequate target cover-
age during the day. Candesartan, which
was developed as an angiotensin recep-
tor antagonist, has been established over
the past decade as a migraine preventative.
However, a negative study with telmisartan
(Supplementary Table 1), another angio-
tensin receptor antagonist, suggests that the
benefit of candesartan is the result of more
than its action on the angiotensin receptor.
Glutamatergic mechanisms are expected
to be involved in migraine, and their
manipulation could offer a neuronal target
(Supplementary Table 1). As in other brain
disorders, development of medicines
that target glutamate receptors without
A DECADE IN MEDICINE NOVEMBER 2015  |  66
problematic adverse effects has not been
easy. Experimental and clinical studies9 have
demonstrated that the N‑methyl‑d‑aspartate
receptor is a plausible target in the manage-
ment of migraine aura, yet the practical use
will be limited unless a way around adverse
side effects is determined. Studies of the
AMPA and kainate (iGluR5) receptors
suggest that targeting the kainate receptor
will be the best way forward (Supplementary
Table 1). Targeting of the mGluR5 recep-
tor has already been established as a
­next-generation approach.
With the knowledge that migraine is a
brain disorder, neuromodulation seems
an attractive option, and has been tried
with several approaches in the past decade.
Invasive approaches, such as occipital nerve
stimulation, initially seemed promising,
but have been beset by problems with study
design and device suitability. No positive,
well-designed, placebo-controlled study
has yet been completed (Supplementary
Table 1), but newer devices and better-
designed studies could lead to a future role
for invasive techniques.
Noninvasive approaches are particularly
attractive because they carry a low risk, so
little is lost if they do not work. Transcranial
magnetic stimulation and transcutaneous
supraorbital nerve stimulation are now used
in many expert centres in Europe and the
USA. Positive studies have been published
(Supplementary Table 1), although the data
are limited. Similarly, noninvasive vagal nerve
stimulation has been studied but is yet to find
its place. Clinical experience suggests to me
that devices, when they are effective, are
truly neuromodulatory but require time and
diligent application over several months to
produce effects that can be excellent.
The future is undoubtedly bright for the
development of new treatments for migraine.
In addition to targets that have already been
tested in clinical studies, one can see other
targets emerging, such as the mGluR5 recep-
tor, pituitary adenylate-cyclase activating
peptide, acid sensing ion channels, neuro­
nal nitric oxide synthesis inhibitors and
individual orexin receptors (Supplementary
Table 1), to drive the next decade forward.
Millennia after migraine was first reported,
clinicians will be able to suggest migraine-
specific preventative treatments to patients
with migraine. Furthermore, some of the
therapies developed to date could be bene­
ficial for cluster headache and medication
overuse headache. Experimental medicine
offers a clear pathway from basic biology to
the clinic. One hopes that, as next-generation
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
genetics evolves,10 the upcoming decade will
see the development of personalized medi-
cine that will see migraine lose its position
among the top of causes of global disability.
Basic and Clinical Neurosciences, Institute of
Psychiatry, Psychology and Neuroscience, and
King’s Clinical Research Facility, Kings College
London, Wellcome Foundation Building, King’s
College Hospital, London SE5 9PJ, UK.
peter.goadsby@kcl.ac.uk
doi:10.1038/nrneurol.2015.203
Published online 27 October 2015
Competing interests
The author has received grants and personal fees
from Allergan, Amgen and eNeura, and personal
fees from Ajinomoto Pharmaceuticals, Akita
Biomedical, Alder Biopharmaceuticals, Autonomic
Technologies, Avanir Pharmaceuticals, Cipla,
CoLucid Pharmaceuticals, Dr Reddy’s Laboratories,
EMKinetics, Ethicon, Heptares, Lilly, NEJM Journal
Watch, Pfizer, Promius Pharma, Teva, UpToDate, Wells
Fargo, W. L. Gore & Associates and Zosano Pharma.
He has also received personal fees for medicolegal
work for individuals. He also has a patent pending for
magnetic stimulation for headache.
1. Global Burden of Disease Study 2013
Collaborators. Global, regional, and national
incidence, prevalence, and years lived with
disability for 301 acute and chronic diseases
and injuries in 188 countries, 1990–2013:
a systematic analysis for the Global Burden of
Disease Study 2013. Lancet 386, 743–800
(2015).
2. Amin, F. M. et al. Magnetic resonance
angiography of intracranial and extracranial
arteries in patients with spontaneous migraine
without aura: a cross-sectional study. Lancet
Neurol. 12, 454–461 (2013).
3. Maniyar, F. H., Sprenger, T., Monteith, T.,
Schankin, C. & Goadsby, P. J. Brain activations
in the premonitory phase of nitroglycerin-
triggered migraine attacks. Brain 137,
232–241 (2014).
4. Hougaard, A., Amin, F., Hauge, A. W., Ashina, M.
& Olesen, J. Provocation of migraine with aura
using natural trigger factors. Neurology 80,
428–431 (2013).
5. Akerman, S., Holland, P. & Goadsby, P. J.
Diencephalic and brainstem mechanisms in
migraine. Nat. Rev. Neurosci. 12, 570–584
(2011).
6. Goadsby, P. J., Edvinsson, L. & Ekman, R.
Vasoactive peptide release in the extracerebral
circulation of humans during migraine
headache. Ann. Neurol. 28, 183–187 (1990).
7. Goadsby, P. J., Edvinsson, L. & Ekman, R.
Release of vasoactive peptides in the
extracerebral circulation of humans and the
cat during activation of the trigeminovascular
system. Ann. Neurol. 23, 193–196 (1988).
8. Ho, T. W., Edvinsson, L. & Goadsby, P. J. CGRP
and its receptors provide new insights into
migraine pathophysiology. Nat. Rev. Neurol. 6,
573–582 (2010).
9. Ayata, C., Jin, H., Kudo, C., Dalkara, T. &
Moskowitz, M. A. Suppression of cortical
spreading depression in migraine prophylaxis.
Ann. Neurol. 59, 652–661 (2006).
10. Tolner, E. A. et al. From migraine genes to
mechanisms. Pain 156 (Suppl. 1), S64–S74
(2015).
Supplementary information is linked to the online
version of the paper at www.nature.com/nrneurol.
NEUROLOGY
DECADE IN REVIEW—CNS INFECTIONS
Major advances against a moving
target of CNS infections
Lisa F. P. Ng and Tom Solomon
CNS infections have severe manifestations, often leading to high
mortality, but the CNS is usually not the primary target of pathogens,
leaving a window of opportunity to prevent neuroinvasion. We must
prioritize development of therapies to prevent neurological sequelae
that cause long-lasting morbidity and disease burden on society.
The Nobel Prize in Physiology or Medicine
in 2015 was awarded jointly for discoveries
concerning novel therapies against malaria
and infections caused by roundworm para-
sites, highlighting the global importance of
infectious disease. The past decade has seen
some considerable advances in the diag-
nosis and treatment of CNS infections,
despite some major challenges, such as
the emergence of novel pathogens, the
spread of existing patho-
gens to new geographical
areas as a result of travel
and climate change, and
more-severe outbreaks
caused by existing patho-
gens becoming more
virulent, and increasing
numbers of immuno­
compromised people. As
organizers of the Gordon Research
Conferences on Infec­t ions of the
Nervous Sys­tem in 2013 and 2015, we
have had the opportunity to closely follow
the advances made in the prevention and
treatment of CNS infections. Despite these
advances, many unmet challenges remain,
including the rise of resistance to the anti­
malarial drug artemisin in patients with
severe cerebral malaria, which has increased
morbidity and mortality, re-emergence of
arbovirus-induced CNS disease, limited
treatment options for cryptococcal menin-
gitis, and failure of the Japanese encephalitis
vaccine, to name a few.
In the USA and developed European
countries, viral encephalitis, bacterial
meningitis and Lyme neuroborreliosis
have been studied fervently. Less attention
has been given to potentially fatal nervous
system invasion by neurotropic viruses,
parasites or mycobacteria; encephaliti-
des caused by these pathogens are more
common in developing than developed
countries, and are an important health issue
in resource-poor regions, because many of
67  |  NOVEMBER 2015  www.nature.com/reviews
the survivors are left with disability that
reduces health-related quality of life and
causes a great socioeconomic burden.
The epidemiological changes that have
occurred in recent years have placed greater
importance on recently emerged
pathogens, such as the Nipah and
Chikungunya viruses across Asia,
and West Nile virus across
the Americas. 1 Meanwhile,
enterovirus 71, has con-
tinued to cause mass out-
breaks of hand, foot and
mouth disease across Asia,
which is associated with
neurological compli-
cations that can mani-
fest as meningitis or
severe brainstem encephalitis.2
These neurological complica-
tions can, via hitherto unknown
mechanisms, lead to cardiovas-
cular collapse and death.2 Enterovirus 71
has even been described by some as the
‘new polio’, because it belongs to the same
enterovirus group IV as poliovirus, and can
cause acute flaccid paralysis that mimics
polio­myelitis. No antiviral agent or vaccine
against enterovirus 71 is currently available.
The arthropod-borne Chikungunya virus,
which was virtually unknown a decade ago,
also causes mass outbreaks of febrile disease.
Chikungunya infections are charac­terized
by fever, headache, rashes, and debilitating
arthralgia that can continue for years after
infection. Atypical manifestations have
also been described, and include neuro-
logical symptoms that range from simple
and complex febrile seizures to meningeal
syndrome, acute encephalopathy, diplopia,
aphasia, acute disseminated encephalo-
myelitis, and encephalitis.3 After its first
reported appearance in 1952, the virus
re-emerged in the Indian Ocean islands
in 2005–2006, and several outbreaks of
Chikungunya infection have occurred since
© 2015 Macmillan Publishers Limited. All rights reserved
late 2013, particularly in the Caribbean
islands and the Americas, but also in
several countries in Africa and Southern
Europe. Increased global travel, evolution
of the virus and adaptation of mosquito
vectors have all been suggested as contrib-
uting factors to the spread of the virus to
new geographical areas.3 No vaccination or
treatment for Chikungunya virus exists, and
avoiding mosquito bites is c­ urrently the only
way to prevent the disease.
Many common infections, including the
‘big three’—HIV, tuberculosis and malaria—
have important neurological manifesta-
tions. Outbreaks of new influenza strains,
such as H5N1, have also highlighted the
potential of such viruses to cause neurologi-
cal disease.4 In recent years, the number of
immuno­compromised individuals who are
particularly susceptible to CNS infection
has increased owing to the HIV epidemic
and increased use of immunosuppressive
treatments, for example to induce immune
tolerance after organ transplantation or to
treat auto­immune disorders. Neurologists
should be particularly aware of the risk of
CNS infection in patients who are immuno-
compromised, because many neurological
disorders are treated with immunosuppres-
sive therapies. Perhaps the most prominent
example of such disorders is multiple sclero-
sis (MS), for which several new drugs have
been approved in the past decade. Similarly,
progressive multifocal leukoencephalo­
pathy caused by JC virus was once rare, but
its prevalence has increased with that of
HIV, and in the past 10 years, the use of new
immunosuppressive monoclonal anti­bodies,
such as natalizumab for MS, has become
another important trigger for the disease.5
During the same period, immune-­
mediated encephalitis, such as anti-
N‑methyl-­d‑aspartate receptor encephalitis,
has been established as a cause of encepha-
litis that is as important as are infections.6
Occasionally, some infections of the NPG
nervous system, such as herpes simplex
virus, can also trigger autoimmune
encephalitis.
Improved technologies and diagnos-
tics, such as multiplex PCR,
MALDI-TOF (matrix-assisted
laser desorption/ionization
—­t ime-of-f light)

mass specto­metry, better immuno­assays
and meta­genomic next-generation sequen­
cing, have improved the speed and accuracy
of discovering pathogens and identifying
immune system correlates of protection from
infectious disease. Despite these advances,
the proportion of patients with CNS infec-
tions who are misdiagnosed remains
stubbornly high.6
Treatment of some brain infections,
including acute bacterial meningitis and
tuberculous meningitis, has improved over
the past decade owing to a better under-
standing of the clinical features and the ways
in which they contribute to poor outcomes.7
Moreover, the mechanism of corticosteroids
in the treatment of some of these conditions
has been better defined: the benefit of dexa-
methasone has been demonstrated in adults
with acute bacterial meningitis in most devel-
oped countries, as has its ability to reduce
the incidence and severity of hearing loss in
pneumococcal disease. However, the drug
is often of no benefit to adults or children
in developing, tropical countries, perhaps
because patients in these countries are more
likely to have untreated HIV.8 In Vietnamese
adults with tuberculous meningitis, dexa-
methasone reduced mortality and is now
used for most patients with this condition.
In the past 10 years, no major advances
have been made in the treatment of enceph-
alitis beyond the use of corticosteroids. In
mouse studies, monoclonal antibodies have
shown promise for the treatment of some
viral encephalitides, such as those caused
by West Nile virus, Chikungunya virus and
A DECADE IN MEDICINE NOVEMBER 2015  |  68
Japanese encephalitis virus.9 RNA interfer-
ence has been proposed as another treat-
ment strategy, but the only encephalitides
in which it has been studied to date are
Japanese encephalitis, tick-borne encephali-
tis and encephalitis caused by alphaviruses,
so it is a long way from clinical applications.
The ‘Holy Grail’ against CNS infections
—effective neuroprotective treatment—
remains elusive. Therapeutic hypothermia,
which has proved useful in hypoxic brain
injury after cardiac arrest, was found to be
of no benefit in bacterial meningitis. Novel
cocktails of antivirals and potentially neuro-
protective drugs initially looked promising,
but have, in most cases, failed to work.
Vaccination programmes have brought
some of the biggest successes in the preven-
tion of CNS infections. Advances include the
development of new vaccines for pneumo-
coccal meningitis and Japanese encephali-
tis,10 and, in the case of Japanese encephalitis,
improved administration methods and
dosage of existing vaccines. By the time
of the next Gordon Research Conference
on Infections of the Nervous System in
2017, these developments should have had
­considerable impact.
Singapore Immunology Network, Agency for
Science, Technology and Research (A*STAR),
8A Biomedical Grove, #04‑06 Immunos,
Biopolis, Singapore 138648, Singapore
(L.F.P.N.). Institute of Infection and Global
Health, University of Liverpool, Ronald Ross
Building, 8 West Derby Street, Liverpool
L69 7BE, UK (T.S.).
Correspondence to: L.F.P.N.
lisa_ng@immunol.a-star.edu.sg
© 2015 Macmillan Publishers Limited. All rights reserved
NEUROLOGY
doi:10.1038/nrneurol.2015.202
Published online 27 October 2015
Competing interests
The authors declare no competing interests.
1. Jones, K. E. et al. Global trends in emerging
infectious diseases. Nature 451, 990–993
(2008).
2. Chang, L. Y. et al. Neurodevelopment and
cognition in children after enterovirus 71
infection. N. Engl. J. Med. 356, 1226–1234
(2007).
3. Weaver, S. C. & Lecuit, M. Chikungunya virus
infections. N. Eng. J. Med. 372, 1231–1239
(2015).
4. de Jong, M. D. et al. Fatal outcome of human
influenza A (H5N1) is associated with high
viral load and hypercytokinemia. Nat. Med. 12,
1203–1207 (2006).
5. Kleinschmidt-DeMasters, B. K. & Tyler, K. L.
Progressive multifocal leukoencephalopathy
complicating treatment with natalizumab
and interferon beta‑1a for multiple sclerosis.
N. Eng. J. Med. 353, 369–374 (2005).
6. Granerod, J. et al. Causes of encephalitis
and differences in their clinical presentations
in England: a multicentre, population-based
prospective study. Lancet Infect. Dis. 10,
835–844 (2010).
7. Weisfelt, M., van de Beek, D., Spanjaard, L.,
Reitsma, J. B. & de Gans, J. Clinical features,
complications, and outcome in adults with
pneumococcal meningitis: a prospective case
series. Lancet Neurol. 5, 123–129 (2006).
8. van de Beek, D. et al. Adjunctive dexamethasone
in bacterial meningitis: a meta-analysis of
individual patient data. Lancet Neurol. 9,
254–263 (2010).
9. Oliphant, T. et al. Development of a humanized
monoclonal antibody with therapeutic potential
against West Nile virus. Nat. Med. 11, 522–530
(2005).
10. Tauber, E. et al. Safety and immunogenicity
of a Vero‑cell‑derived, inactivated Japanese
encephalitis vaccine: a non-inferiority, phase III,
randomised controlled trial. Lancet 370,
1847–1853 (2007).
Connect to...

Follow us on Twitter
For the latest commentary and neurology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Neurology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Neurology

Web focuses
In the latest focus, cutting-edge investigators highlight the important challenges in epilepsy research
and clinical practice

w w w.
Web collections
A selection of articles on a related theme, providing a panoramic view of a key area of neurology

Conferences
You can meet the editors of Nature Reviews Neurology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Neurology!

November 2015 marked the 10th anniversary of the launch of Nature Reviews Neurology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of neurology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrneurol10

© 2015 Macmillan Publishers Limited. All rights reserved


DECADE IN REVIEW—TRANSLATIONAL RHEUMATOLOGY
Ten years after: rheumatology research from
bench to bedside
Nunzio Bottini and Gary S. Firestein
Defining key advances in any medical discipline can be challenging, but is especially so in rheumatology—a
rapidly advancing field so broad that it defies traditional classifications. Here, we approach the Sisyphean
task of summarizing the translational advances in rheumatology in the past decade within several broad
categories of basic research.
Translational studies in rheumatology from
the past decade are especially prominent
in the field of cytokine biology. For instance,
a ‘bench-to-bedside and back’ loop was
spanned by the quest to understand
IL‑23–IL‑17 pathobiology. Since IL‑17-
producing T helper (TH)17 cells were first
characterized in 2005 and IL‑23 defined as
a critical cytokine for their differ­entiation,
abundant data implicate this pathway in
rheumatic diseases. For example, enthesis-
resident T cells that produce IL‑17 and
IL‑22 in response to IL‑23 stimulation
mediate inflammation and new bone for-
mation in a preclinical model of spondylo­
arthropathy (SpA).1 Clinical trials of agents
targeting this cytokine axis in SpA have
been spectacular: the anti-IL‑12/IL‑23 anti-
body ustekinumab received FDA approval
for psoriasis in 2009 and for psoriatic arthri-
tis (PsA) in 2013, changing the therapeutic
landscape of the former. The anti-IL‑17
anti­b ody secukinumab is also effective
in psoriasis and PsA. Despite high hopes
for IL‑23–IL‑17 blockade in rheuma-
toid arthritis (RA), these agents unfortu-
nately showed only modest efficacy in this
setting (Figure 1).
IL‑1 inhibition lost its lustre owing
to the disappointing results in RA, but
enthusiasm re-emerged in the past decade
with the discovery of a critical role for
IL‑1 in the  inflamma­s ome and auto­
inflammatory diseases. Ultimately, three
anti-IL‑1 agents demonstrated remarkable
efficacy in cryopyrin-associated periodic
syndromes (CAPS) and were approved for
this use in the USA. The inflammasome and
endogen­ous ‘danger signals’ also participate
in crystal-induced diseases, such as gout,
for which anti-IL‑1 agents are effective.
A DECADE IN MEDICINE NOVEMBER 2015  |  69
RHEUMATOLOGY
A newly identified anti-inflammatory difficulty of approaching a complex disease
member of the IL‑1 family, IL‑37, might such as SLE with traditional clinical
also have therapeutic potential as a target trial designs.
in autoinflammatory diseases. Antibodies against two ‘classical’ cytokine
Interferons continue to attract consider- targets that were the object of translational
able attention in rheumatology, building on efforts have entered clinical development:
the 2003 observation that leukocytes from targeting of granulocyte-macrophage
patients with systemic lupus erythematosus colony-stimulating factor (GM-CSF), first
(SLE) have a type I interferon (IFN) trans­ described in the context of RA >25 years
criptome signature. In the past decade, ago, has now been validated using block-
multiple translational studies focused on ing antibodies in RA. An open-label trial
validating type I IFN as a drug target for SLE showed that the transforming growth
and other autoimmune diseases. However, factor β (TGF-β)-blocking antibody freso-
modest efficacy of the type I IFN-blocking limumab in systemic sclerosis effectively
antibody rontalizumab reported in 2015, modulated key predictive fibrosis bio-
as well as experience with B-cell-targeted markers.2 At the pre­clinical level, IL‑27, an
approaches, exemplified the persistent IL‑12 family member highly expressed in
IL-23 IL-17 IL-6 IL-1 B-cell B-cell
survival depletion
?
Psoriasis SpA RA Autoinflammatory Systemic SLE ANCA+
PsA disease sclerosis vasculitis
JAK GM-CSF PAD TGF-β Type I IFN
Other kinases
(e.g. BTK, PI3K)
Robust clinical efficacy Modest clinical efficacy Preclinical or biomarker data
Figure 1 | Key therapeutic targets for rheumatic diseases explored in theReviews
Nature past decade.
| Rheumatology
Targets for which agents exist with robust or modest clinical efficacy are presented, as well
as targets for which preliminary or preclinical data support agents currently under development.
B‑cell depletion as a therapeutic strategy in SLE is still uncertain. Colours represent cytokines
(blue), cell-targeted therapy (green), enzymes (orange) and diseases (purple). Abbreviations:
ANCA, anti-neutrophil cytoplasmic antibody; GM‑CSF, granulocyte-macrophage colony-stimulating
factor; IFN, interferon; JAK, Janus-activated kinase; PAD, peptidylarginase deiminase;
PsA, psoriatic arthritis; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus;
SpA, spondyloarthropathy; TGF‑β, transforming growth factor β.
© 2015 Macmillan Publishers Limited. All rights reserved
RHEUMATOLOGY
the RA synovium, was effective in models
of experi­mental arthritis, and might have
potential for translation.
In addition to strategies aimed at block-
ing individual cytokines, approaches that
target cell lineages implicated in patho­
genesis have progressed in the last decade.
Studies demon­strating the critical role of
antibodies and B cells in anti-neutrophil
cytoplasmic anti­b ody (ANCA)-positive
vasculitis paved the way for FDA approval
of rituximab in 2011. B cells also received
much attention in SLE research: the TNF
superfamily member B cell stimulating
factors BAFF (TNF ligand superfamily
member 13B) and APRIL (TNF ligand
superfamily member 13), as well as their
receptors, have been explored as possible
therapeutic targets. The anti-BAFF anti­
body belimumab received FDA approval
for SLE in 2011, although its bene­f its
are limited and utilization in the clinic is
currently modest.
Neutrophils have also emerged as criti-
cal players in SLE and RA. A seminal
2011 report showed that neutrophils from
patients with SLE extrude portions of their
nuclei and cytoplasm (neutrophil extra­
cellular traps, or NETs) in a process called
NETosis. 3 This mechanism, now recog-
nized as a critical pathway for presentation
of altered antigens and self-antigens, is a
key mediator of innate immunopathol-
ogy in SLE. Importantly, neutrophils are
also a source of citrullinated proteins and
enzymes that convert arginine to citrulline
(peptidylarginine deiminases, or PADs) in
RA, and enzyme-activating anti­bodies spe-
cific for neutrophil PAD might contribute
to RA pathogenesis.4 Interest in targeting
fibroblast-like synoviocytes in inflam-
matory arthritis was re-ignited this past
decade by the identification of the trans-
membrane molecule cadherin‑11 as an
important regulator of cartilage damage
and synovial intimal lining formation.5
The success of therapies based on tyro­
sine kinase inhibitors in oncology led to
the notion that similar approaches might
be bene­f icial in inflammation. Building
upon pioneering studies showing that tyro­
sine protein kinase JAK (Janus kinase)3-
deficient individuals exhibit severe
immuno­d eficiency, the JAK inhib­itor
tofacitinib was developed and approved
for the treatment of RA in 2012. Biopsy
studies showed that tofacitinib is effective
in blocking synovial JAK1–STAT3 and
JAK–STAT1 signaling downstream of IL‑6
and type I  IFN.6
70  |  NOVEMBER 2015  www.nature.com/reviews
However, results in the signaling research
realm have been mixed. For example, fos-
tamatinib (a small-molecule inhibitor of
spleen tyrosine kinase SYK) and multiple
mitogen-activated protein kinase (MAPK)
p38 inhi­bitors were only modestly effec-
tive in RA; ‘back-to-bench’ studies later
revealed that p38 inhibition increases
activation of other MAPKs and suppresses
anti-inflammatory cytokines like IL‑10.
Despite these setbacks, kinases remain
high-profile targets in rheuma­tology: the
tyrosine kinase BTK, a regulator of B‑cell
function and innate immunity, and PI3Kδ
(phosphatidylinositol 4,5-bisphosphate
3‑kinase catalytic subunit δ isoform, which
regulates both B‑cell and synovio­cyte func-
tions), are among the kinases currently
under research.
The impressive progress in the genetics
of rheumatic diseases in the past decade
is in part due to the success of genome-
wide associ­ation studies (GWAS). GWAS
in ankylosing spondylitis and RA have
identified clear associations with IL‑23
and CD40, respectively,7,8 leading to their
explor­a tion as drug targets. However,
limitations of GWAS have emerged as
well: despite large meta-analyses that
can detect increasing numbers of loci
with ever-smaller effects on risk—over
a hundred loci have been identified in
RA to date8—GWAS have been unable to
capture most of the risk and severity influ-
ences on complex rheumatic diseases.
New epi­g enetic strategies developed
almost entirely in the past decade might
help fill the gaps left by GWAS. Genome-
wide epigenetic studies have already been
conducted in blood cells from patients
with SLE, RA or other auto­i mmune dis-
eases, in synoviocytes from patients with
RA and chondrocytes from patients with
osteoarthritis, with promising results.
The study of mucosal microbial popula-
tions and how microbial dysbiosis affects
disease is emerging as a new and important
field in rheumatology. High prevalence of
Prevotella copri was identified in the gastro­
intestinal microbiome of patients with
early RA,9 and other studies showed that
PADs produced by the gingival pathogen
Porphyromonas gingivalis could contribute
to protein citrullination.
Looking into the future, translational
studies will hopefully provide clues on the
best ways to select patients for ‘personal-
ized’ approaches to therapy. Current bio-
marker research is increasingly driven by
knowledge of disease mechanisms: one
© 2015 Macmillan Publishers Limited. All rights reserved
possible example was the identification of
C–X–C motif chemo­kine 4 (CXCL4, also
known as platelet factor 4) as a biomarker
in systemic sclerosis in 2014.10 New tech-
nology of the past decade, including the
advent of next-generation sequencing,
will enable integrative analyses of multiple
‘omics’ platforms and capture a degree
of complexity previously unattainable.
These developments should pave the way
for new algorithms to identify indivi­dual
versions of SLE, RA and other rheumatic
diseases—advances that will hopefully
occur in the next 10 years.
La Jolla Institute for Allergy and Immunology,
9420 Athena Circle, La Jolla, CA 92037, USA
(N.B.). UC San Diego School of Medicine,
9500 Gilman Drive, La Jolla, CA 92093, USA
(N.B., G.S.F.).
Correspondence to:
nunzio@lji.org
gfirestein@ucsd.edu
doi:10.1038/nrrheum.2015.126
Published online 22 September 2015
Competing interests
The authors declare no competing interests.
1. Sherlock, J. P. et al. IL‑23 induces
spondyloarthropathy by acting on ROR-γt+
CD3+CD4–CD8– entheseal resident T cells.
Nat. Med. 18, 1069–1076 (2012).
2. Rice, L. M. et al. Fresolimumab treatment
decreases biomarkers and improves
clinical symptoms in systemic sclerosis
patients. J. Clin. Invest. 125, 2795–2807
(2015).
3. Villanueva, E. et al. Netting neutrophils induce
endothelial damage, infiltrate tissues, and
expose immunostimulatory molecules in
systemic lupus erythematosus. J. Immunol.
187, 538–552 (2011).
4. Darrah, E. et al. Erosive rheumatoid arthritis
is associated with antibodies that activate
PAD4 by increasing calcium sensitivity.
Sci. Transl. Med. 5, 186ra65 (2013).
5. Lee, D. M. et al. Cadherin‑11 in synovial lining
formation and pathology in arthritis. Science
315, 1006–1010 (2007).
6. Boyle, D. L. et al. The JAK inhibitor tofacitinib
suppresses synovial JAK1–STAT signalling
in rheumatoid arthritis. Ann. Rheum. Dis. 74,
1311–1316 (2015).
7. Australo-Anglo-American Spondyloarthritis
Consortium. et al. Genome-wide association
study of ankylosing spondylitis identifies
non-MHC susceptibility loci. Nat. Genet. 42,
123–127 (2010).
8. Okada, Y. et al. Genetics of rheumatoid
arthritis contributes to biology and drug
discovery. Nature 506, 376–381 (2014).
9. Scher, J. U. et al. Expansion of intestinal
Prevotella copri correlates with enhanced
susceptibility to arthritis. eLife 2, e01202
(2013).
10. van Bon, L. et al. Proteome-wide analysis
and CXCL4 as a biomarker in systemic
sclerosis. N. Engl. J. Med. 370, 433–443
(2014).

DECADE IN REVIEW—PAEDIATRIC RHEUMATOLOGY
A field on the move
Seza Ozen
The study of rheumatic diseases that affect children has thrived in
the past 10 years. A look at several important advances in this area
illustrates how organized collaborations and advanced technologies
are contributing to the understanding and, ultimately, to improving the
treatment of these disorders.
In this past decade, research in paediatric
rheumatology has been actively revealing
the pathogenesis of the diseases particular
to this specialty. One major contribution of
paediatric research has been the definition
of monogenic diseases that have the same
phenotype as corresponding complex dis-
eases. A prominent genetic predisposition
in relevant pathways is expected to lead to
disease manifestation early in life and, thus,
such diseases often present in young children
with affected relatives.1,2 Understanding the
function of these critical genes is hoped to
offer insight into the pathogenesis of these
complex diseases as well. A leading example
is the identification of deficiency of adeno­sine
deaminase 2 (ADA2), in which muta­tions in a
single gene, CECR1, lead to a disease similar
to polyarteritis nodosa.1 As ADA2 is critical
for macrophage activation and endothelial
cell development, the defect in CECR1 results
in widespread vasculitis with a contribution
from inflammatory macro­phages.1 In many
patients, the vessels of the central nervous
system are affected, leading to stroke, whereas
others might have milder symptoms. For
further details of related publications please
refer to Supplementary Box 1 online. Another
genetic mutation that mimics a complex
disease is the LACC1 gene, which causes a
form of systemic JIA (sJIA).2
‘‘ …the spectrum of known
autoinflammatory diseases has
enlarged considerably…
’’
Gene expression profiling has enabled
the discovery of pathogenetic biomarker
signatures in paediatric rheumatic diseases.
Focusing initially on paediatric systemic
lupus erythematosus (SLE), Virginia Pascual’s
group showed that patients with this disease
almost universally have an interferon signa-
ture.3 By contrast, the inter­feron signature is
present in only about half of adult patients
with SLE, a difference that could be due
to the severity of the disease in children or to
A DECADE IN MEDICINE NOVEMBER 2015  |  71
other factors. The same group also identified
transcriptional biomarkers of disease activity
in SLE and, by analysing the transcriptional
pattern of cells from patients with sJIA,
showed that IL‑1 has a leading role in the
latter disease. In fact, many clinical and labo-
ratory features of sJIA can be explained by
increased IL‑1 production,3 and these studies
have already paved the way for targeted
therapy for sJIA, with anti-IL‑1 agents.
sJIA has many features typical of an auto­
inflammatory syndrome. In fact, most IL‑1β-
mediated human diseases have been linked to
abnormal activation of this cytokine by the
inflammasome. The IL‑1-related diseases are
now grouped as ‘autoinflammatory diseases’.
Most autoinflammatory diseases are mono-
genic, affect the innate immune system and
typically present in childhood. Given this
early presentation, it is not surprising that
paediatricians have made major contribu-
tions to this field. The leading project in this
area has been the establishment of the multi-
national, multidisciplinary Eurofever registry,
which has enabled the collection of immense
amounts of information about the clinical fea-
tures and demo­graphics of auto­inflammatory
diseases.4 This large dataset will, it is hoped,
serve as a basis for future development of
diagnostic criteria and t­ reatment protocols.
In the past 15 years, and with paedia­tricians
leading the field, the spectrum of known auto-
inflammatory diseases has enlarged consider-
ably. The newly recog­nized interferon-related
diseases have led us to reconsider the initial
‘IL‑1-centred’ concept of autoinflammation.
The term ‘interferon­opathies’ was intro­
duced in 2011 to encompass diseases such as
Aicardi–Goutières syndrome and spondylo­
enchondrodysplasias, and soon included the
PSMB8-related auto­inflammatory diseases,
including CANDLE (chronic atypical neutro­
philic dermatosis with lipodystrophy and
elevated temper­ature) syndrome.5 The latest
discovery in this growing field was SAVI
(STING-associated vasculopathy with onset
in infancy) syn­drome (see Supplementary
Box 1 online). Interferonopathies include
© 2015 Macmillan Publishers Limited. All rights reserved
RHEUMATOLOGY
at least 12 mono­genic diseases in which the
innate immune response leads to interferon
production and related autoinflammatory
features, with possible overlap with features
of auto­immunity in some cases.5 This new
perception of interferon­opathies highlights
the need for different approaches to treat-
ment, and challenges our understanding of
the intersection of the innate and adaptive
immune systems.
A number of genome-wide association
studies (GWAS) have been completed in
the most common paediatric rheumatic
disease, JIA. A 2013 GWAS genotyped
2,816 patients with JIA and 13,056 controls
using the Immunochip array.6 Patients with
oligo­articular and rheumatoid factor (RF)-
negative polyarticular JIA were studied as
these forms of the disease are specific to
children and include evidence of disorders
in the adaptive immune system. This GWAS
had certain advantages over previously pub-
lished ones: it included a large international
cohort and the Immunochip array provided
comprehensive coverage of single nucleotide
polymorphisms (SNPs) in regions implicated
in autoimmune diseases. This study con-
firmed the association with the three previ-
ously known JIA risk loci (the HLA region,
PTPN22 and PTPN2) and identified 14 new
loci associated with risk of JIA.6 Notably, the
set of SNPs included on the Immunochip
explained and estimated 18% of JIA suscepti­
bility. Thus other regions must exist that
harbour additional JIA risk loci which will
probably be the subject of future research.
‘‘
Studies of treatments for JIA
have flourished with international
collaboration
’’
Research in JIA has also concentrated
on understanding the cell biology in the
synovium. For instance, the work of two
European research groups focuses on the
autoimmune inflammatory environment of
the rheumatoid joint and the role of regula-
tory T (TREG) cells in defining the course of
arthritis (see Supplementary Box 1 online).
In a comprehensive Review on the subject,
one of the main mechanisms of local auto-
immune inflammation is suggested to be
interference with T‑cell regulation by inflam-
matory mediators, inducing T‑cell plasticity
and causing unstable TREG cells to convert to
pathogenic effector T cells.7 In fact, a skewed
T‑cell commitment occurs in the syno-
vial fluid and, upon interaction with CD4
T cells, local monocytes specifically induce
RHEUMATOLOGY
T helper type 17 (TH17) responses, which
are highly inflammatory. The inflammatory
milieu of affected joints has also been shown
to render effector T cells resistant to sup-
pression. The authors of the Review suggest
that these issues should be considered in the
development of targeted therapies.7
Studies of treatments for JIA have flour-
ished with international collaboration. In
other rheumatic diseases, however, multi­
centre treatment studies in children are
lacking, owing mainly to ethical consid-
erations and a lack of tools that enable the
use of validated classification criteria (often
used as diagnostic criteria in practice). In
the past decade, paedia­tricians have suc-
ceeded in developing, by use of sophisti-
cated approaches, classifications to guide
physicians in diagnosis and to form a basis
for future collaborative studies in two groups
of diseases: juvenile scleroderma8 and the
common childhood forms of vasculitis.9
The Ankara 2006 criteria for IgA vasculi-
tis (Henoch–Schönlein purpura), granu-
lomatous polyangiitis (formerly known as
Wegener granuloma­t osis), poly­arteritis
nodosa and Takayasu arteritis (which
were endorsed by EULAR, the Paediatric
Rheumatology European Society [PReS] and
the Paediatric Rheumatology International
Trials Organisation [PRINTO]), and the
scleroderma criteria, were validated using
online paediatric registries that included a
large cohort of international patients.
Another important contribution by pae-
diatricians was in defining S100 proteins,
which are released by activated phagocytes
and function as proinflammatory alarm-
ins, as biomarkers of inflammation.10 These
proteins have been implicated in a number
of inflammatory conditions. More impor-
tantly, however, they have been shown to
be useful biomarkers for the risk-adapted
stratification of patients with JIA in deci-
sions regarding cessation of therapy; indeed,
levels of S100 proteins have been shown to
correlate well with risk of relapse in a ran-
domized controlled drug trial, and others are
to follow.10
This overview of developments in paedi-
atric rheumatology in the past decade has
been assigned a limited space and thus many
important developments and papers could
not be cited. However, this selection is hoped
to reflect some of the leading projects in the
field. Paediatric rheuma­tology is surely pro-
gressing in the quest to better understanding
its diseases. The relevant research will enable
us to better manage our young patients, who
have a long life-expectancy.
72  |  NOVEMBER 2015  www.nature.com/reviews
Department of Pediatric Rheumatology,
Hacettepe University, Sihhiye, 06100 Ankara,
Turkey.
sezaozen@hacettepe.edu.tr
doi:10.1038/nrrheum.2015.130
Published online 22 September 2015
Acknowledgements
The author apologizes to colleagues whose work
could not be cited because of space constraints.
Competing interests
The author declares that she has acted as a
consultant for Novartis and received speaker
honoraria from Roche and Sobi.
1. Zhou, Q. et al. Early-onset stroke and
vasculopathy associated with mutations
in ADA2. N. Engl. J. Med. 370, 911–920 (2014).
2. Wakil, S. M. et al. Association of a mutation
in LACC1 with a monogenic form of systemic
juvenile idiopathic arthritis. Arthritis Rheumatol.
67, 288–295 (2015).
3. Pascual, V. et al. How the study of children
with rheumatic diseases identified
interferon-α and interleukin‑1 as novel
therapeutic targets. Immunol. Rev. 223, 39–59
(2008).
4. Toplak, N. et al. An international registry on
autoinflammatory diseases: the Eurofever
experience. Ann. Rheum. Dis. 71, 1177–1182
(2012).
DECADE IN REVIEW—TECHNOLOGY
Technological advances
transforming rheumatology
William H. Robinson and Rong Mao
Technological advances over the past decade have revolutionized many
areas of rheumatology, ranging from diagnosis, prognosis and therapeutic
development to the mechanistic understanding of rheumatic diseases.
This overview highlights key technological innovations and discusses
the major impact that these developments are having on research and
clinical practice.
The past decade has been an exciting time
of technological breakthroughs driving tre-
mendous progress in rheumatology. Many
of these innovations are high-throughput
approaches that enable robust proteomic,
genomic and epigenetic analyses ranging
from the organism level to the single-cell
level.1 Here, we describe the key innova-
tions (examples of which are described in
Box 1 and illustrated in Supplementary
Figure 1 online) and discuss how they have
shaped our understanding of rheumatic dis-
eases as well as our ability to diagnose and
treat these disorders.
Historically, many major advances in the
research and clinical practice of rheumatol-
ogy were fuelled by technological inno­vations.
© 2015 Macmillan Publishers Limited. All rights reserved
5. Crow, Y. J. & Manel, N. Aicardi-Goutières
syndrome and the type I interferonopathies.
Nat. Rev. Immunol. 15, 429–440 (2015).
6. Hinks, A. et al. Dense genotyping of
immune-related disease regions identifies
14 new susceptibility loci for juvenile
idiopathic arthritis. Nat. Genet. 45, 664–669
(2013).
7. Wehrens, E. J., Prakken, B. J. & van Wijk, F.
T cells out of control—impaired immune
regulation in the inflamed joint. Nat. Rev.
Rheumatol. 9, 34–42 (2013).
8. Zulian, F. et al. The Pediatric
Rheumatology European Society/American
College of Rheumatology/European League
Against Rheumatism provisional classification
criteria for juvenile systemic sclerosis.
Arthritis Rheum. 57, 203–212 (2007).
9. Ozen, S. et al. EULAR/PRINTO/PRES criteria
for Henoch-Schönlein purpura, childhood
polyarteritis nodosa, childhood Wegener
granulomatosis and childhood Takayasu
arteritis: Ankara 2008. Part II: final
classification criteria. Ann. Rheum. Dis. 69,
798–806 (2010).
10. Rothmund, F. Validation of relapse risk
biomarkers for routine use in patients with
juvenile idiopathic arthritis. Arthritis Care Res.
(Hoboken) 66, 949–955 (2014).
Supplementary information is linked to the online
version of the paper at www.nature.com/nrrheum.
For instance, the advent of MRI introduced a
noninvasive method for visu­alizing bone and
soft tissues in three dimen­sions that enables
improved diagnosis. The development of flow
cytometry greatly enhanced our ability to
distinguish between and characterize distinct
cell populations in tissue samples. Molecular
cloning coupled with expression profiling
using DNA micro­arrays has been pivotal in
identifying key molecules and pathways
in the pathogenesis of rheumatic diseases, and
thus in uncover­ing novel therapeutic targets.
Likewise, the past 10 years have brought us
a new raft of technological advances—both
incremen­tal and disruptive—that are enabling
us to interro­gate and manage rheumatic
diseases with increasing sophistication.

Proteomics is one notable area in which
great progress has been made over the past
decade, to far-reaching effect. Innovations
in proteomics, including advances in
mass spectrometry and the emergence of
protein-array technologies, have revolu-
tionized our ability to identify proteins and
post-translational modifications associated
with disease. Indeed, mass spectrometry
ana­lyses of proteins in cartilage, synovial
membrane, bone, synovial fluid, plasma and
serum, as well as other tissues and bodily
fluids, have uncovered molecules associated
with patho­logical changes in osteoarthritis,
rheumatoid arthritis (RA), systemic lupus
erythema­tosus (SLE) and other rheumatic
diseases. Furthermore, array-based multi­
plex pro­filing of auto­antibodies and cyto­
kines has deepened our understanding
of pre-disease and early disease states by
enabling the charac­terization of autoimmu­
nity prior to the onset of clinically apparent
symptoms.2 Several of the proteomic pro­
files gleaned with these technologies have
potential for use as actionable biomarkers
in predictive medicine. Most rheumatic
diseases are heterogeneous and only certain
subsets of patients respond to any given
therapy. Thus, proteomic profiles and other
biomarkers associated with specific disease
states or with drug responsiveness could
identify indivi­duals at high risk of develop-
ing the disease, who can then be enrolled
in primary prevention trials or treated with
preventative therapies. Proteomic profiles
and other biomarkers could also serve as
pharmaco­dynamic biomarkers to rapidly
assess patients’ responses to therapy. These
proteomic technologies are ushering in a
new era in biomarker discovery and have the
potential to revolutionize the diagnosis and
treatment of rheumatic diseases.
Large-scale sequencing is another
tech­nological tour de force that is trans-
forming rheumatology. The advent of high-
throughput DNA sequencing has made
possible sequencing of the genome to iden-
tify both common and rare genetic variants
associated with rheumatic diseases. This
method can also be applied to sequencing
the expressed genome, which includes thou-
sands of gene transcripts that reflect activa-
tion and repression of pathways involved in
rheumatic diseases. Such large-scale DNA
sequencing can now digitally define the
functional antibody and T‑cell receptor
(TCR) repertoires in autoimmune rheumatic
diseases.1,3,4 In addition, deep sequencing of
transcriptomes via RNA-seq measures the
levels of transcripts and their isoforms, and
A DECADE IN MEDICINE NOVEMBER 2015  |  73
RHEUMATOLOGY
Box 1 | Advanced technologies in rheumatology*
■■ CyTOF mass cytometry measures the binding of multiple antibodies (each tagged with a
distinct heavy-metal isotope) to cells1
■■ Single-cell antibody heavy and light chain sequencing enables bioinformatic generation
of phylogenetic trees, which reveal clonal antibody families and guide the selection of
antibodies for expression and characterization3
■■ Single-cell TCR sequencing coupled with amplification of functional genes characteristic
of T-cell subsets provides insight into the specificity and function of TCRs4
■■ ATAC-seq analyses the epigenome of cells derived from an individual6
■■ iPOP combines genomic, transcriptomic, proteomic, metabolomic, and autoantibody
profiles from an individual to reveal medical risks and dynamic molecular changes in health
and disease10
*See also Supplementary Figure 1 online. Abbreviations: ATAC-seq, transposase-accessible chromatin using
sequencing; CyTOF, cytometry by time-of-flight; iPOP, integrative personal omics profile; TCR, T-cell receptor.
can now even be done with single-cell resolu- are much better than conventional radio­
tion.5 Furthermore, epi­genetic technologies, graphy at assessing soft-tissue abnormali-
such as ATAC-seq (assay for transposase- ties and detecting bone erosions,8 and are
accessible chromatin using sequencing),6 can increasingly being used as noninvasive tools
probe disease-associated changes in DNA for detecting subclinical inflammation and
methylation and histone modification. The progression of joint damage in arthritides.
scale and efficiency of sequencing that can The growing use of these techniques in
now be achieved with these approaches is clinical practice and research is transform-
facilitating unprecedented progress in both ing diagnostic imaging. By enabling serial
basic and translational research of rheu- assessment of synovitis, interval imaging
matic diseases, and will most likely trans- via MRI or ultrasonography yields dynamic
form clinical care in the future. However, a biomarkers that are useful for monitoring
major challenge that these technologies bring the progression of disease or its response
is the need to store, retrieve and analyse to therapy.
the terabytes and petabytes of data that are Last but not least, advances in stem-cell
being generated. technologies are offering new opportunities
The past 10 years have also seen the in tissue engineering and regenerative rheu-
development of mass cytometry (known matology. Recent work on the small mol­
as cytometry by time-of-flight [CyTOF]) ecule kartogenin has shown that it is poss­ible
and considerable advances in flow cytom- to direct the differentiation of mesen­chymal
etry, enhancing our ability to analyse cellu­ stem cells into chondro­c ytes and thereby
lar markers and signalling pathways in repair damaged cartilage. 9 Moreover,
rheumatic diseases.7 Mass cytometry uses induced pluripotent stem cells (iPSCs), first
heavy metals instead of fluorophores to described in 2006, have emerged as a promis-
label cells, thereby enabling measurement ing cell source for both drug screening and
of >40 parameters per cell without spillover cell-replacement therapy.9 Patient-specific
between fluorescence spectra. Meanwhile, iPSCs are well-suited to autologous stem
flow cytometry has been improved by the cell transplantation, because they elicit only
introduction of new staining reagents, minimal immune reactions and have the
laser-excitable fluorescent compounds, potential to be used to regenerate tissues,
coupling methods and dyes that can be such as cartilage. Meanwhile, disease mod-
used to monitor cell replication and physio­ elling using patient-specific iPSCs continues
logical changes inside cells. As a result of to expand our knowledge of pathophysiology
these advances, CyTOF and polychromatic and treatment.
flow cytometry have enabled researchers to In conclusion, technological innovations
delve deeper than ever into the complexity in the past decade have opened up new
of disease-relevant cell types and molecules, horizons in our efforts to understand and
monitor their dynamics, and unravel their treat rheumatic diseases. Moving forward,
normal function as well as their contribution it is anticipated that high-throughput
to rheumatic diseases. approaches, especially those allowing for
Imaging techniques such as MRI and single-cell resolution, will move to the main-
ultrasonography, although not new, have stream of rheumatology research and form
only recently begun to be incorporated the basis for next-generation diagnostics.
into clinical trials and routine practice in The impact of high-throughput sequen­cing
rheumatology.8 MRI and ultrasonography and other data-rich technol­ogies will be
© 2015 Macmillan Publishers Limited. All rights reserved
RHEUMATOLOGY
maximized with the development of more
powerful databases, analytical methods
and analysis pipelines that can handle
the vast amount of data being generated.
Development of new biomarkers will bring
forth an era of predictive medicine, in which
individuals in pre-disease states can be
identified so that primary prevention can
be instituted and in which therapies can be
selected for use in the patient populations
most likely to benefit. Given the hetero­
geneity of most rheumatic diseases, the
diverse molecular pathways mediating their
pathogenesis and the multifaceted roles that
these pathways have in normal and patho-
logical states, advances in treatment are likely
to require approaches that integrate genomic,
transcriptomic, proteomic, metabolomic and
autoantibody profiles, such as the recently
described integrative personal omics profile,
or iPOP.10
Division of Immunology and Rheumatology,
CCSR 4135, 269 Campus Drive, Stanford,
CA 94305, USA (W.H.R., R.M.).
Correspondence to: W.H.R.
wrobins@stanford.edu
doi:10.1038/nrrheum.2015.137
Published online 6 October 2015
Acknowledgements
The authors’ work is supported by grants from the
NIH NHLBI Proteomics Center (N01-HV‑00242), NIH
NIAMS (R01 AR063676), NIH NIAID (U01 AI101981,
U01 AI057229, U19 AI110491), NIH NCI (R33
CA183659), and NIH NIAMS/NIAID/FNIH AMP
Program (UH2 AR067681), and by the Brennan
Family, the Northern California Chapter of the
Arthritis Foundation (NCCAF) Center of Excellence,
and the Bill and Melinda Gates Foundation.
DECADE IN REVIEW—CLINICAL RHEUMATOLOGY
10 years of therapeutic advances
in the rheumatic diseases
John D. Isaacs
In the past 10 years, the rheumatology community has seen an
explosion in the number of new therapies licensed for use across the
rheumatic diseases, many with outstanding clinical success. Here, the drugs
and strategies that constitute landmarks in the management of rheumatic
diseases are highlighted.
The rheumatoid arthritis (RA) investi­gational
space has remained active in the past decade
(Box 1). The major advance in the treat-
ment of this disease has been the successful
develop­ment of orally bioavail­able small-
molecule inhibitors of signalling proteins.
74  |  NOVEMBER 2015  www.nature.com/reviews
Competing interests
W.H.R. declares that he is a member of the Board
of Directors, consultant for, and owner of equity in
Atreca, Inc. R.M. declares no competing interests.
1. Maecker, H. T. et al. New tools for classification
and monitoring of autoimmune diseases.
Nat. Rev. Rheumatol. 8, 317–328 (2012).
2. Sokolove, J., Lindstrom, T. M. & Robinson, W. H.
Development and deployment of antigen arrays
for investigation of B‑cell fine specificity in
autoimmune disease. Front. Biosci. (Elite Ed.)
4, 320–330 (2012).
3. Robinson, W. H. Sequencing the functional
antibody repertoire—diagnostic and
therapeutic discovery. Nat. Rev. Rheumatol. 11,
171–182 (2015).
4. Han, A., Glanville, J., Hansmann, L.
& Davis, M. M. Linking T‑cell receptor
sequence to functional phenotype at the
single-cell level. Nat. Biotechnol. 32, 684–692
(2014).
5. Macosko, E. Z. et al. Highly parallel genome-
wide expression profiling of individual cells
using nanoliter droplets. Cell 161, 1202–1214
(2015).
6. Buenrostro, J. D., Giresi, P. G., Zaba, L. C.,
Chang, H. Y. & Greenleaf, W. J. Transposition
of native chromatin for fast and sensitive
epigenomic profiling of open chromatin,
DNA-binding proteins and nucleosome
position. Nat. Methods 10, 1213–1218
(2013).
7. Bendall, S. C., Nolan, G. P., Roederer, M.
& Chattopadhyay, P. K. A deep profiler’s guide
to cytometry. Trends Immunol. 33, 323–332
(2012).
8. Filippucci, E., Di Geso, L. & Grassi, W.
Progress in imaging in rheumatology. Nat. Rev.
Rheumatol. 10, 628–634 (2014).
9. Diekman, B. O. & Guilak, F. Stem cell-based
therapies for osteoarthritis: challenges and
opportunities. Curr. Opin. Rheumatol. 25,
119–126 (2013).
10. Chen, R. et al. Personal omics profiling reveals
dynamic molecular and medical phenotypes.
Cell 148, 1293–1307 (2012).
Tofacitinib, the first of this class of drugs to
be licensed for use in RA, inhibits the Janus-
associated kinase (JAK) signalling pathways
down­stream of various cytokine and growth-
factor receptors.1 Tofacitinib is moderately
non­selective and blocks, in descending order
© 2015 Macmillan Publishers Limited. All rights reserved
of potency, JAK3, JAK1 and JAK2. In phase 3
trials, tofacitinib had a safety and efficacy
profile similar to that of biologic DMARDs.
Several other, more-selective JAK inhibitors
are in development, although it remains to
be seen how these compare with tofacitinib.
Regardless of the results, a treatment that
offers ‘biologic efficacy in a pill’ will surely be
popular among patients with RA.
The need for novel therapies has been
more acute for patients with psoriatic
arthritis (PsA). During most of the past
decade, TNF blockade was the only alter-
native avail­able after DMARD failure, but
new oral and biologic therapies have now
been licensed. Apremilast, an oral phospho­
diesterase 4 inhibitor that indirectly reduces
pro­inflammatory cytokine production,
is now used before biologic therapies for
patients with an inadequate response to
DMARDs.2 The monoclonal antibody (mAb)
ustekinumab blocks the p40 chain common
to the cytokines IL‑12 and IL‑23, which drive
T helper (TH)1 and TH17 immune responses,
respectively.3 Both the IL‑12 and IL‑23 path­
ways have been linked to PsA in genome-
wide association studies. As the efficacy of
ustekinumab seems similar to that of TNF
blockade,3 this therapy provides an addi-
tional, potent option for patients with PsA
refractory to DMARDs. Apremilast seems
less effective than ustekinumab and TNF
blockade but is well-tolerated, and blood
monitoring is not required; therefore, apre-
milast is also a useful option for patients
with PsA.
Belimumab is the first new therapy to
be licensed for use in systemic lupus ery-
thematosus (SLE) in >50 years.4 This mAb
blocks B‑lymphocyte stimulator (BLyS,
also known as BAFF), a growth factor for
B cells. Whilst its development was proble­
matic and included a failed phase 2 clin­ical
trial, the endpoints in two phase 3 studies
of belimumab were achieved, albeit with
only modest improvements over placebo.
Important lessons were learned during this
process, including the need for new clin­ical
trial outcome criteria. Retrospective analysis
of the phase 2 trial also identified an associ­
ation between response to therapy and circu-
lating autoantibodies at baseline. Belimumab
is currently licensed as an addition to con-
ventional therapy in patients with active,
seropositive SLE.
The belimumab trials illustrated the
complexities of clinical trial design in SLE,
perhaps explaining why trials of rituximab
(a B‑cell depleting agent) did not demon­
strate efficacy in ameliorating this condition.

Rituximab has, however, gained a foothold
in the management of anti-neutrophil cyto­
plasmic antibody (ANCA)-associated vas­
culitis (AAV), in which it proved noninferior
to daily oral cyclophos­phamide in achieving
glucocorticoid-free remission in patients
with newly-diagnosed or relapsing disease.5
In a separate study, rituximab was shown
to be superior to azathioprine in sustaining
remission after treatment with cyclophos­
phamide and glucocorticoids. Whilst gluco­
corticoids remain an important component
of therapy, many patients with AAV now have
an alternative to traditional cytotoxic and
immunosuppressive medications.
Systemic sclerosis remains one of the most
refractory connective tissue diseases, with
no current treatments that slow its progres-
sion. A major aspect of the pathology of this
disease is arterial vasoconstriction, media­
ted by endothelin‑1. Whilst endothelin‑1
blockade has been available to treat pulmon­
ary hypertension for some time, only more
recently has the endothelin-receptor antag­
onist bosentan been licensed to treat severe
digital ulcers secondary to poor digital
circulation.6 These ulcers are a particularly
painful and debilitating symptom of sys-
temic sclerosis, and endothelin‑1 inhibition
has provided new hope for patients with this
unrelenting condition.
In this past decade, good news for chil-
dren with juvenile idiopathic arthritis (JIA)
emerged as IL‑6-receptor blockade with
tocilizumab was shown to be highly effec-
tive in children with active systemic or
polyarticular-course JIA. Its effectiveness
was particularly welcome in systemic JIA,
previously a condition with limited thera-
peutic options—a situation that resulted
in dependency on growth-inhibiting
glucocorticoids.7
Important advances in therapies for
more common conditions also occurred
in this period. TNF ligand superfamily
member 11, also known as RANK ligand
(RANKL), is an important differentiation,
activation and survival factor of osteoclasts.
Denosumab, an anti-RANKL mAb, has been
developed as an antiresorptive therapy for
post­menopausal women, as well as men,
with osteo­p orosis, reducing the risk of
both vertebral and nonvertebral fractures.8
Its long half-life means only twice-yearly sub­
cutaneous injections are needed for efficacy,
which will improve adherence considerably
compared with bisphosphanates. Pegloticase
(pegylated uricase), a new treatment for
patients with gout who cannot tolerate, or
have an inadequate response to, standard
A DECADE IN MEDICINE NOVEMBER 2015  |  75
treatments, is highly effective in decreasing
uric acid levels, reducing flares and resolving
tophi.9 Clinically important improvements
in quality-of-life were documented within
6 months of starting pegloticase treatment,
but anaphylaxis is common (incidence of
~7%), as are infusion reactions associated
with fortnightly dosing.9
Ankylosing spondylitis (AS) remains a
disease with few treatment options, which
currently include NSAIDs, physiotherapy and
TNF blockade. Ustekinumab trials in patients
with PsA suggested benefits for spinal inflam-
mation, and the biology of AS strongly impli-
cates the IL‑23–IL‑17 axis. In this context,
secukinumab, an anti-IL‑17 mAb, was shown
to rapidly improve the signs and symptoms of
AS, with positive changes assessed in parallel
by use of imaging techniques.10 Intriguingly,
response was associ­ated with polymorphisms
in ERAP1 and, possibly, IL23R. Several mAbs
have been developed that target the IL‑17
pathway and, notwithstanding any surprises,
patients with AS can look forward to a first
new treatment option since TNF blockade
(none of these are currently licensed for the
treatment of AS).
By reviewing these therapeutic advances,
it is satisfying to see that the majority are
targeted therapies aimed at pivotal patho­
genetic pathways. What lies ahead for the
next decade? In RA, a continuing trend
towards earlier treatment is clear, and trials
are underway in ‘pre-RA’—focused on
indivi­duals with autoantibodies but without
joint inflammation symptoms. The ‘T’
word—tolerance—is increasingly mentioned
and, in several diseases, cell-targeted strate­
gies are being developed as potential toler­
ogenic therapies. Increasingly sophisticated
methods can be used to purify and expand
T regulatory (TREG) cells, and mesenchymal
stem cell and tolerogenic dendritic-cell trials
are being reported in the literature. Methods
to generate TREG cells in vivo, such as low-dose
IL‑2 therapy, also seem promising.
Several technologies are likely to contrib-
ute to the development of rheumatic disease
therapeutics over the coming decade. In
terms of antibody engineering, we might see
the application of smaller fragments, such as
nanobodies, and of bispecific antibodies. For
example, bispecific agents that target both
TNF and IL‑17 are currently being studied
in RA. Furthermore, genetic manipulation
and epigenetic modification are already
being applied in nonrheumatic diseases (an
oral antisense oligonucleotide, mongersen,
which targets SMAD7 mRNA and leads to
a subsequent increase in the production of
© 2015 Macmillan Publishers Limited. All rights reserved
RHEUMATOLOGY
Box 1 | New drugs for rheumatic diseases
Antibodies
■■ Ustekinumab (PsA)
■■ Belimumab (SLE)
■■ Rituximab (AAV)
■■ Tocilizumab (JIA)
■■ Denosumab (osteoporosis)
Enzyme
■■ Pegloticase (gout)
Small molecule (orally bioavailable) drugs
■■ Tofacitinib (RA)
■■ Apremilast (PsA)
■■ Bosentan (SSc)
Abbreviations: AAV, anti-neutrophil cytoplasmic
antibody (ANCA)-associated vasculitis; JIA, juvenile
idiopathic arthritis; PsA, psoriatic arthritis;
RA, rheumatoid arthritis; SLE, systemic lupus
erythematosus; SSc, systemic sclerosis.
transforming growth factor β, improved
symptoms and signs in a phase 2 Crohn
disease trial). We can expect similar technol­
ogies to be tested in rheumatic diseases, not-
withstanding the challenges of extra­intestinal
delivery. There is also much interest in the
role of the microbiome in human disease, and
the capacity to modify immune and inflam-
matory processes by manipulation of the gut
flora is appealing, particularly in the context
of treating early and preclinical conditions.
If therapies become cheaper, they should
become more widely available and be used,
for example, for patients with earlier and less-
active disease—the first bio­similar infliximab
has now been approved for use in rheumatic
disease, and the next decade will see many
more biosimilars reach the market as patents
for currently licensed biologic agents expire.
Another area of inten­sive investment is strati-
fied medicine, or prescribing according to a
patient’s genetic and biologic characteristics.
This approach is likely to pay dividends over
the next decade.
Despite the numerous advances in clin­ical
management of rheumatic diseases, many
challenges lie ahead for the next decade, not
least the need for disease-modifying drugs
for the most common rheumatic disease,
osteoarthritis. Nonetheless, the new techno­
logical and strategic approaches discussed
here, including new drug classes and more
personalized treatment approaches, have the
potential to bring new solutions and algo-
rithms to rheumatology care—consequently,
the way we assess, investigate and treat
patients in 10 years’ time could look very
different from today.
Newcastle University, Institute of Cellular
Medicine, William Leech Building, Framlington
Place, Newcastle-Upon-Tyne NE2 4HH, UK.
john.isaacs@ncl.ac.uk
RHEUMATOLOGY
doi:10.31038/nrrheum.2015.138
Published online 13 October 2015
Acknowledgements
Work in the author’s laboratory is supported by
the National Institute for Health Research (NIHR)
Newcastle Biomedical Research Centre, based
at Newcastle Hospitals NHS Foundation Trust
and Newcastle University, UK. The views expressed
are those of the author and not necessarily those
of the NHS, the NIHR or the Department of Health.
The author is grateful to T. Aspray and P. Conaghan
for their insights whilst preparing thismanuscript.
Competing interests
The author has been a member of advisory
boards for Celltrion, Hospira, Janssen, Novartis,
Pfizer and Roche.
1. Norman, P. Selective JAK inhibitors in
development for rheumatoid arthritis.
76  |  NOVEMBER 2015  www.nature.com/reviews
Expert Opin. Investig. Drugs 23, 1067–1077
(2014).
2. Kavanaugh, A. et al. Treatment of psoriatic
arthritis in a phase 3 randomised, placebo-
controlled trial with apremilast, an oral
phosphodiesterase 4 inhibitor. Ann. Rheum. Dis.
73, 1020–1026 (2014).
3. McInnes, I. B. et al. Efficacy and safety of
ustekinumab in patients with active psoriatic
arthritis: 1 year results of the phase 3,
multicentre, double-blind, placebo-controlled
PSUMMIT 1 trial. Lancet 382, 780–789 (2013).
4. Stohl, W. & Hilbert, D. M. The discovery and
development of belimumab: the anti-BLyS-
lupus connection. Nat. Biotechnol. 30, 69–77
(2012).
5. Stone, J. H. et al. Rituximab versus
cyclophosphamide for ANCA-associated
vasculitis. N. Engl. J. Med. 363, 221–232 (2010).
6. Matucci-Cerinic, M. et al. Bosentan treatment of
digital ulcers related to systemic sclerosis:
© 2015 Macmillan Publishers Limited. All rights reserved
results from the RAPIDS-2 randomised,
double-blind, placebo-controlled trial.
Ann. Rheum. Dis. 70, 32–38 (2011).
7. De Benedetti, F. et al. Randomized trial of
tocilizumab in systemic juvenile idiopathic
arthritis. N. Engl. J. Med. 367, 2385–2395
(2012).
8. Cummings, S. R. et al. Denosumab for
prevention of fractures in postmenopausal
women with osteoporosis. N. Engl. J. Med. 361,
756–765 (2009).
9. Sundy, J. S. et al. Efficacy and tolerability of
pegloticase for the treatment of chronic gout
in patients refractory to conventional treatment.
Two randomized controlled trials. JAMA 306,
711–720 (2011).
10. Baeten, D. et al. Anti‑interleukin‑17A
monoclonal antibody secukinumabin treatment
of ankylosing spondylitis: a randomised, double-
blind, placebo-controlled trial. Lancet 382,
1705–1713 (2013).
Connect to...

Follow us on Twitter
For the latest commentary and rheumatology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Rheumatology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Rheumatology

Web focuses
A selection of articles on a related theme with content ranging from Research Highlights through to Reviews
and Perspectives, providing a panoramic view of a key area of medicine

w w w.
Web collections
A selection of articles on a related theme, providing a panoramic view of a key area of rheumatology

Conferences
You can meet the editors of Nature Reviews Rheumatology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Rheumatology!

November 2015 marked the 10th anniversary of the launch of Nature Reviews Rheumatology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of rheumatology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrrheum10

© 2015 Macmillan Publishers Limited. All rights reserved

UROLOGY
DECADE IN REVIEW—BLADDER CANCER
International progress: from cytology to genomics
James C. Costello and Dan Theodorescu
The past decade has seen many notable contributions to bladder cancer research. Here, we highlight the
international efforts in the field, with findings from Europe, USA and China, as well as papers resulting
from international cooperation. We anticipate the next 10 years will see even greater collaborative and
international efforts.
Costello, J. C. & Theodorescu, D. Nat. Rev. Urol. 11, 609–610 (2014); published online 2 September 2014; doi:10.1038/nrurol.2014.236
The early 2000s saw great changes in bladder
cancer research, with several very important
contributions that helped shape our molecu­
lar understanding of the disease and patient
treatment strategies. Dyrskjøt and collea­
gue­s were the first to connect tumour stage
with genome-scale molecular patterns of
the disease using microarrays.1 Dyrskjøt
et al.1 showed that gene expression patterns
sampled from patient tumours clustered
into three major groups (stages): Ta, T1 and
T2−4. They were additionally able to use the
gene expression patterns to predict whether
a tumour would recur or not, providing
insights into the molecular mechanisms
that stratify aggressive muscle‑invasive and
non-muscle-invasive tumours.
In addition to stratifying patients by
tumour stage and predicting recurrence,
gene expression profiling has been used to
develop cellular response models to drug
treatment that have been used to predict
how an untreated cell, or patient, will
respond. In a seminal 2007 paper on drug
sensitivity prediction, the NCI‑60 panel of
cancer cell lines was used to predict how
bladder cancer cells would respond to drug
treatment.2 The coexpression extrapola­
tion (COXEN) method demonstrated
Pogonici/iStock/Thinkstock
77  |  NOVEMBER 2015  www.nature.com/reviews
‘‘ …gene expression profiling
has been used to develop
cellular response models to
drug treatment…
’’
that drug sensitivities could be accurately
predicted using information derived from
other cancer cell lines. This and additional
retrospective studies showed that COXEN
could be successfully applied to patients.
The results from this study, and those of
Dyrskjøt and colleagues,1 provided early
support for the notion of personalized or
precision medicine. Indeed, a clinical trial
(Southwest Oncology Group 1314) is cur­
rently underway to validate COXEN pro­
spectively in assigning chemotherapy to
bladder cancer patients.
Urine cytology is particularly insensitive
at detecting low-grade tumours. However,
the finding that bladder tumours could be
genetically stratified, coupled with known
FGFR3 mutations that are frequently
associ­ated with low-grade disease, led to
the hypothesis that FGFR3 could serve as
a urinary biomarker of recurrent disease.
van Rhijn et al.3 reported that combined
microsatellite and FGFR3 mutation analy­
sis could detect urothelial cell carcinoma
(UCC) in voided urine. FGFR3 mutations
were found in 44% of urothelial tumours
(n = 59), but were absent in 15 G3 tumours.
The sensitivity of microsatellites to detect
cancer in voided urine was lower for
tumours harbouring FGFR3 mutations
(15 of 21 tumours; 71%) than for FGFR3
wild-type UCC (29 of 32 tumours; 91%).
By including the FGFR3 mutation, the
sensitivity of molecular cytology increased
from 71% to 89% and was superior to the
sensitivity of morphological cytology (25%)
© 2015 Macmillan Publishers Limited. All rights reserved
for every clinical subdivision. This paper
highlighted the potential power of molecu­
lar biology as an adjunct to cystoscopy and
cytology to inform follow-up care.
Minimally invasive surgery has devel­
oped into a standard outpatient pro­
cedure over the past decade in most areas
of surgery. In urology, it has considerably
changed how we manage men with prostate
cancer, primarily with the advent of robotic
technologies. In 2003, the first worldwide
attempt to perform a robot-assisted laparo­
scopic radical cystectomy and completely
intra-abdominal formation of an ortho­
topic neobladder (Hautmann neobladder)
was described by Beeken et al.4 Impressively,
operating time was 8.5 h and blood loss was
only 200 ml. Several other studies have
reported the use of the robot-assisted cys­
tectomy, with intracorporeal and extra­
corporeal formation of the diversion. Most
recently, several trials have reported com­
parisons between open and laparoscopic
approaches. Recent studies have indicated
that robotic cystectomy is carried out in
<15% of patients.
In the late 1980s, the formulation and
early trials of methotrexate, vinblastine,
doxorubicin and cisplatin (MVAC) chemo­
therapy represented an important advance
in the treatment of metastatic bladder
cancer. Despite subsequent larger trials
showing durable complete responses were
difficult to achieve, this regimen remained
the standard of care. Further developing
MVAC, a Southwest Oncology Group trial
comparing neoadjuvant chemotherapy
plus cystectomy (n = 154) with cystectomy
alone (n = 153) for locally advanced bladder
cancer showed the median survival among
patients was 77 months and 46 months,
respectively.5 Importantly, in both groups,

improved survival was associated with
the absence of residual cancer in the cys­
tectomy specimen. It was also noted that
fewer patients in the combination group
had residual disease than patients in the
cystectomy group (15% versus 38%). The
toxicity of MVAC had previously been
noted in its use in the metastatic setting,
motivating the development of alternative
less-toxic regimens. In 2005, a trial com­
paring the survival of patients with locally
advanced or metastatic UCC treated with
MVAC (n = 202) to gemcitabine and cispla­
tin (n = 203) was reported.6 Overall survival
was similar in both arms, with a median
survival of 14−15 months. The 5‑year over­
all survival rates were 13.0% and 15.3% for
the two regimens, respectively, but were
not statistically different. However, the
lower toxicity of the gemcitabine and cis­
platin regimen had a remarkable impact on
practice—most patients are now given this
treatment rather than MVAC.
‘‘
…next-generation sequencing
has been a boon to all cancer
research—and bladder cancer is
’’
no exception
Nomograms for prediction of clinical out­
comes after primary therapy gained great
popularity over the past decade in a variety
of diseases, especially in cancer. These
tools are particularly useful for patient
counselling and guidance with respect to
follow-up intensity and adjuvant therapy.
Moving into 2006, a report was published
that described a tool of great clinical
utility in this regard.7 This work was the
result of an international collaboration
that gener­ated a bladder cancer database
from 12 centres of excellence and com­
prising 9,000 patients treated with radical
cystectomy and pelvic lymphadenectomy.
A prognostic nomogram was constructed
to predict 5‑year progression-free prob­
ability; its predictive accuracy was found to
be significantly better than standard TNM
or standard pathological subgroupings.
However, its use in clinical practice is, thus
far, unreported.
In the mid-1980s, the presence and
extent of the androgen receptor (AR) in
urothelial bladder cancer tissue was recog­
nized to be higher than in normal bladder
mucosa, potentially leading to the gender
disparities in bladder cancer incidence. In
2007, investi­g ators used animal models
to show that AR signals promote bladder
A DECADE IN MEDICINE NOVEMBER 2015  |  78
cancer develop­m ent and progression. 8
They showed that 92% of wild-type male
and 42% of wild-type female mice treated
with N‑butyl‑N-(4-hydroxy­butyl)nitrosa­
mine (BBN) eventually developed bladder
cancer, whereas none of the male or female
AR‑knockout mice did. Treatment with
BBN induced bladder cancer in 25% of
AR‑knockout mice supplemented with
dihydrotestosterone and in 50% of castrated
wild-type male mice. These findings, and
that of other groups using human bladder
cancer cell lines, showed that androgens and
the AR are important promoters of bladder
tumour formation and metasta­sis. Whether
this evidence that gender differences in
bladder cancer incidence might be due
to AR will result in a change in t­reatment
p­rotocol is, as yet, unknown.
Mutations in cancer driver genes occur
at high rates; for example, in bladder cancer
TP53 is altered at a frequency of roughly
50%. Traditional single-gene cancer drivers
tend not to exceed 50% of the population,
which is why the study by Killela and col­
leagues, which showed TERT promoter
mutations occur at a rate of 66% in bladder
cancer and >80% in glioblastoma, was
so surprising.9 These findings have been
subsequently verified and although the
implications of this discovery will not be
evident for years to come, we have new
insights in the mechanism of bladder
cancer development with the potential for
therapeutic intervention.
Indeed, the advancement of genom­
ics technologies into next-generation
sequencing has been a boon to all cancer
research—and bladder cancer is no excep­
tion. Although several previous studies
had looked at whole-genome or whole-
exome sequencing, the work by Guo et al.10
reported in 2013 was revolutionary because
it sampled the genome sequences of 99
Chinese patients with bladder cancer. This
approach enabled comprehensive cata­
loguing of genomic alterations. Previously
identified bladder cancer driver mutations
were verified, and many additional candi­
date driver genes were identified, including
STAG2, ESPL1 and a group of chromatin
modifying genes. The volume of genomic
information is expanding rapidly, with
efforts such as The Cancer Genome Atlas
and the International Cancer Genome
Consortium producing petabytes of data.
We anticipate that these data will fuel much
of bladder cancer research in the future.
With advances in diagnosis, surgery,
molecular biology and medical therapy,
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
the management of patients with bladder
cancer has evolved markedly over the past
decade. The convergence of new knowledge
on the genomic landscape of bladder cancer
coupled with new targeted and immuno­
therapy approaches promises to change
bladder cancer clinical practice even more
dramatically in the next 10 years.
Department of Pharmacology, University
of Colorado, Anshutz Medical Campus,
12801 East 17th Avenue, Aurora, CO 80045,
USA (J.C.C.). University of Colorado
Comprehensive Cancer Center, MS F‑434,
13001 East 17th Place, Aurora, CO 80045, USA
(D.T.).
Correspondence to: D.T.
dan.theodorescu@ucdenver.edu
Competing interests
The authors declare no competing interests.
Acknowledgements
D.T. is supported in part by NIH grants CA075115
and CA104106. The funders had no role in study
design, data collection and analysis, decision
to publish or preparation of the manuscript.
1. Dyrskjøt, L. et al. Identifying distinct classes
of bladder carcinoma using microarrays.
Nat. Genet. 33, 90–96 (2003).
2. Lee, J. K. et al. A strategy for predicting the
chemosensitivity of human cancers and its
application to drug discovery. Proc. Natl Acad.
Sci. USA 104, 13086–13091 (2007).
3. van Rhijn, B. W. et al. Combined microsatellite
and FGFR3 mutation analysis enables a highly
sensitive detection of urothelial cell carcinoma
in voided urine. Clin. Cancer Res. 9, 257–263
(2003).
4. Beecken, W. D. et al. Robotic-assisted
laparoscopic radical cystectomy and intra-
abdominal formation of an orthotopic ileal
neobladder. Eur. Urol. 44, 337–339 (2003).
5. Grossman, H. B. et al. Neoadjuvant
chemotherapy plus cystectomy compared
with cystectomy alone for locally advanced
bladder cancer. N. Engl. J. Med. 349, 859–866
(2003).
6. von der Maase, H. et al. Long-term survival
results of a randomized trial comparing
gemcitabine plus cisplatin, with methotrexate,
vinblastine, doxorubicin, plus cisplatin in
patients with bladder cancer. J. Clin. Oncol. 23,
4602–4608 (2005).
7. International Bladder Cancer Nomogram et al.
Postoperative nomogram predicting risk of
recurrence after radical cystectomy for
bladder cancer. J. Clin. Oncol. 24, 3967–3972
(2006).
8. Miyamoto, H. et al. Promotion of bladder
cancer development and progression by
androgen receptor signals. J. Natl Cancer Inst.
99, 558–568 (2007).
9. Killela, P. J. et al. TERT promoter mutations
occur frequently in gliomas and a subset
of tumors derived from cells with low rates of
self‑renewal. Proc. Natl Acad. Sci. USA 110,
6021–6026 (2013).
10. Guo, G. et al. Whole-genome and whole-exome
sequencing of bladder cancer identifies
frequent alterations in genes involved in sister
chromatid cohesion and segregation.
Nat. Genet. 45, 1459–1463 (2013).
UROLOGY
DECADE IN REVIEW—IMAGING
A decade in image-guided
prostate biopsy
Baris Turkbey and Peter L. Choyke
The prostate is still largely assessed by random biopsy, but developments
in prostate MRI and fusion with transrectal ultrasonography (TRUS) have
made targeted biopsy of the prostate a reality. MRI/TRUS techniques
promise to address the issues of overdiagnosis and underdiagnosis
in prostate cancer.
Turkbey, B. & Choyke, P. L. Nat. Rev. Urol. 11, 611–612 (2014); published online 14 October 2014;
doi:10.1038/nrurol.2014.273
Unlike biopsies of the breast, thyroid
and colon, which are guided into lesions
of radiological concern, the prostate is
still largely assessed by random biopsy.
Transrectal ultrasonography (TRUS) has
proven to be rather poor at detecting pros­
tate cancers, so biopsies undertaken using
this method are essentially blind. In the
meantime, multi­p le technical improve­
ments have occurred in prostate MRI, which
has emerged as the imaging technique of
choice for prostate cancer. Specifically,
multiparametric MRI (mpMRI) offers high
sensitivity through a combination of high-
resolution anatomical T2-weighted (T2W)
MRI and functional pulse sequences,
such as diffusion-weighted MRI (dwMRI),
dynamic-contrast-enhanced MRI (DCE
MRI) and MR spectroscopy imaging
(MRSI). Drama­tic technological improve­
ments in dwMRI have meant that, over the
past decade, this technique has been more
heavily relied upon than the others.
The wider availability of high field
strength, 3T magnets and new coil designs
have substantially improved mpMRI tech­
niques. These methods have proven to be
the most accurate yet for identifying local­
ized prostate cancer. However, if mpMRI
simply detected lesions but did not enable
biopsy of those lesions, it would not be
clinically relevant. Thus, from the begin­
ning, attempts were made to biopsy lesions
within the MR scanner. Initial efforts
focused on in-bore MRI guided biopsies;
the main advantage of this approach is
that it enables precise lesion sampling. For
instance, Hoeks et al.1 performed in-bore
MRI guided biopsies in 265 patients
with elevated PSA and previous negative
TRUS-guided biopsies. Prostate cancer
was detected in 41% of patients and the
majority of the detected cancers (87%)
were clinically significant. Roethke et al.2
79  |  NOVEMBER 2015  www.nature.com/reviews
investigated the tumour detection rate
of in-bore MRI-guided biopsy technique
in 100 patients with previous negative
TRUS-guided biopsy with a tumour detec­
tion rate of 52.0%; 80.8% of the detected
prostate tumours were clinically signifi­
cant. However, in-bore MRI-guided biop­
sies have considerable limitations, such
as discomfort related to patient position,
increased costs related to long procedure
duration and the requirement for special
nonmagnetic equipment. Another problem
is that there is insufficient MR capacity and
expertise to handle the number of patients
requiring guided biopsy. This approach has
not proved popular with urologists either,
as the biopsy is performed in the radiology
department and interferes with normal
workflow. For these reasons, attempts
have been made to perform the biopsy
outside the MRI suite while r­etaining the
i­nformation afforded by MRI.
‘‘
…the concept of MRI/TRUS
’’
fusion has taken hold…
The first attempt to transfer MRI infor­
mation to TRUS-guided biopsy was ‘cog­
nitive fusion’. In cognitive fusion guidance
(CFG), the operator first determines the
location of the MRI-positive lesions and
then guides the needle to that location
under real-time TRUS using a ‘best guess’
approach. The main advantage of this tech­
nique is that it does not need additional
equipment; however, this method depends
strongly on the experience and training
of the operator and, therefore, results in
inconsistent outcomes. The transverse
plane on MRI and the transverse plane on
an axial TRUS are often different, and excel­
lent hand–eye coordination is required to
account for this discrepancy. Also, CFG
© 2015 Macmillan Publishers Limited. All rights reserved
iStock/Thinkstock
does not enable documentation of biopsy
locations for repeat biopsies and active
surveil­lance. Nevertheless, the impact of the
CFG technique has been positive, Haffner
et al.3 reported that CFG-targeted biopsies
had a sensitivity and specificity of 95% and
100%, respectively, compared with sensi­
tivity and specificity values for extended
systematic biopsies of 95% and 83%, respec­
tively. CFG biopsies also detected 16%
more high-grade tumours and produced
longer mean cancer core lengths (5.56 mm
compared with 4.70 mm [P = 0.002]) than
extended systematic biopsies. However,
whether this method can be effective on a
broad scale remains to be proven.
The success of CFG prompted interest
in new technologies, such as MRI/TRUS
fusion, to help guide biopsies based on
MRI findings. MRI/TRUS fusion is a col­
lection of technologies that operate under
the same principle. First the MRI image is
obtained, the prostate is segmented from
the remainder of the pelvis and the lesions
are identified. Next the patient is seen in the
ultrasonography suite, where a 3D image is
obtained, to which the segmented prostate
MRI is electronically fused. The TRUS
and the MRI are combined so that as the
TRUS probe is moved or rotated the cor­
responding MRI moves and rotates in the
same way, which allows the operator to use
an MRI obtained at a different time during
the TRUS-guided biopsy. MRI/TRUS-
fusion-guided biopsy is rapidly developing
and several commercial instruments are
already available.
In the original implementation, MRI and
TRUS images were linked together using
passive electromagnetic tracking sensors
that transmit the position of the TRUS
probe, allowing the operator to see both
MR and TRUS images moving in real-time.
Initial results reported by Pinto et al.4 dem­
onstrated that more cancers per core were

detected than standard 12-core TRUS-
guided biopsy alone. Rastinehad et al. 5
reported an overall cancer detection rate
of 62.9%, and 14.3% of cancers were only
detected using MRI/TRUS-fusion biopsy.
Also, approximately 23% of cancers deemed
clinically insignificant by 12-core biopsy
were found to be cl­inically s­ignificant by
MRI/TRUS-fusion biopsy.
‘‘
Developments in mpMRI have
been rapid and impressive over
’’
the past 10 years…
Replacing the freehand electro­magnetic
trackers with a mechanical arm that
holds the TRUS probe in place is another
approach. Once the MRI and TRUS images
are fused, the needle and probe positions are
tracked by angle-sensing encoders embed­
ded in the joints of the mechanical arm.
Using this device, Wysock et al.6 prospec­
tively compared targeted biopsy outcomes
between mechanical arm MRI/TRUS-
fusion biopsies and CGB. Mechanical arm
MRI/TRUS fusion resulted in a 32.0%
detection rate compared with 20.3% for
CGB for clinically significant cancers, and
similar results were observed by Sonn et al.7
Combining MRI and TRUS images using
spatial features alone, without GPS or
mechanical arm tracking, enables the TRUS
probe to be used freehand. Rud et al.8 evalu­
ated accuracy of this method and reported a
52% tumour detection rate. Delongchamps
et al.9 compared the detection rate of CGB
with this form of MRI/TRUS biopsy guid­
ance and found the latter to significantly
improve detection rates over that of system­
atic 12-core TRUS-guided biopsy. Targeted
biopsies decreased the number of cores
needed and the detection of microscopic
cancer, and increased the detection of high-
grade cancer. However, the accuracy of this
method has been questioned.
Transperineal biopsies, which are more
common in Europe, can also be guided
by MRI using a platform that includes a
TRUS probe mounted on a stepper fixed
to the operating table. The probe move­
ments are tracked by two encoders and the
biopsy needles are placed through a grid
mounted to the mechanical stepper, similar
to a brachytherapy seed placement setup.
Kuru et al.10 evaluated this platform in 347
patients, demon­strating that 58% had pros­
tate cancer and 73.5% of biopsy-proven
prostate cancer was clinically significant.
A DECADE IN MEDICINE NOVEMBER 2015  |  80
The tumour detection rate was 82.6% and
72% of tumours were Gleason score ≥7.
Overall, the use of targeted cores detected
significantly more cancer than systematic
biopsies (30% compared with 8.2%).
Thus,
the concept of MRI/TRUS fusion has
taken hold and multiple companies have
jumped into the market to address the
need for equipment. No direct comparison
of differ­e nt MRI/TRUS methods exists,
but it is clear that all of the methods are
superior to the current standard of care.
This technology is likely to continue to
evolve and replace traditional blind biop­
sies. However, whilst MRI-guided biopsies
are unquestionably superior to unguided
biopsies, they raise concerns over cost, as
the MRI and MRI/TRUS-fusion instru­
ments are expensive. Some costs can be
reduced by eliminating the endorectal
coil and it is also unclear whether con­
trast enhancement will be needed in the
future. Eliminating these should reduce
the cost of MRI/TRUS-fusion techniques.
As the technology evolves, market forces
should reduce the cost of fusion equipment
further, and better clinical guidelines for
the selection of candidates for biopsy will
emerge, redu­cing the number of patients
requiring this technology.
Over the past decade there has been a
distinct change in philosophy regarding
the nature of prostate cancer. Previously,
prostate cancer was thought to be a multi­
focal disease that required random samp­
ling to fully identify the extent of disease.
However, experience has shown that many
cancers are incidental and random biop­
sies can miss signifi­cant disease. Thus, the
concept that there are dominant, clinically
significant cancers and multiple inconse­
quential tumours, and that treatment
should be determined by the former, is
being increasingly accepted. The 12-core
random TRUS-guided biopsy is still the
standard of care, but there is growing
aware­ness of its limitations and the value
of image-guided biopsy. Serum PSA screen­
ing significantly increased prostate cancer
diagnosis rates; however, it brought chal­
lenges of overdiagnosis without addressing
the persistent problem of underdiagnosis.
Developments in mpMRI have been rapid
and impressive over the past 10 years and
have enabled better detection and staging
of prostate cancer. Using mpMRI in biopsy
guidance for prostate cancer diagnosis has
the potential to reduce both overdiagnosis
and underdiagnosis. As genomic assess­
ment of prostate cancer becomes more
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
important, there will be a premium on
obtaining better samples and longer cores
of cancer, which can only be provided
with imaging guidance. Nonetheless, the
ultimate acceptance of MRI/TRUS-fusion
biopsy awaits the results of large-scale
multi­centre, randomized studies, which are
needed to convince clinics to pay for this
new technology.
The Molecular Imaging Program, National
Cancer Institute, National Institutes of Health,
10 Center Drive, Building 10, Room B3B69,
Bethesda, MD 20892, USA (B.T., P.L.C.).
Correspondence to P.L.C.
pchoyke@mail.nih.gov
Competing interests
The authors declare no competing interests.
1. Hoeks, C. M. et al. Three-Tesla magnetic
resonance-guided prostate biopsy in men
with increased prostate-specific antigen
and repeated, negative, random, systematic,
transrectal ultrasound biopsies: detection
of clinically significant prostate cancers.
Eur. Urol. 62, 902–909 (2012).
2. Roethke, M. et al. MRI-guided prostate biopsy
detects clinically significant cancer: analysis
of a cohort of 100 patients after previous
negative TRUS biopsy. World J. Urol. 30,
213–218 (2012).
3. Haffner, J. et al. Role of magnetic resonance
imaging before initial biopsy: comparison of
magnetic resonance imaging-targeted and
systematic biopsy for significant prostate
cancer detection. BJU Int. 108, E171–E178
(2011).
4. Pinto, P. A. et al. Magnetic resonance imaging/
ultrasound fusion guided prostate biopsy
improves cancer detection following
transrectal ultrasound biopsy and correlates
with multiparametric magnetic resonance
imaging. J. Urol. 186, 1281–1285 (2011).
5. Rastinehad, A. R. et al. Improving Detection
of Clinically Significant Prostate Cancer:
Magnetic Resonance Imaging/Transrectal
Ultrasound Fusion Guided Prostate Biopsy.
J. Urol. 191, 1749–1754 (2014).
6. Wysock, J. S. et al. A Prospective, Blinded
Comparison of Magnetic Resonance (MR)
Imaging-Ultrasound Fusion and Visual
Estimation in the Performance of MR‑targeted
Prostate Biopsy: The PROFUS Trial. Eur. Urol.
66, 343–351 (2014).
7. Sonn, G. A. et al. Value of targeted prostate
biopsy using magnetic resonance-ultrasound
fusion in men with prior negative biopsy and
elevated prostate-specific antigen. Eur. Urol.
65, 809–815 (2014).
8. Rud, E., Baco, E. & Eggesbø, H. B. MRI and
ultrasound-guided prostate biopsy using soft
image fusion. Anticancer Res. 32, 3383–3389
(2012).
9. Delongchamps, N. B. et al. Prebiopsy
magnetic resonance imaging and prostate
cancer detection: comparison of random
and targeted biopsies. J. Urol. 189, 493–499
(2013).
10. Kuru, T. H. et al. Critical evaluation of magnetic
resonance imaging targeted, transrectal
ultrasound guided transperineal fusion biopsy
for detection of prostate cancer. J. Urol. 190,
1380–1386 (2013).
UROLOGY
DECADE IN REVIEW—URINARY INCONTINENCE
Advances in female urology
and voiding dysfunction
Marisa M. Clifton and Howard B. Goldman
Over the past decade, management of common urological problems has
metamorphosed in the field of female urology and voiding dysfunction;
from treatment of routine stress urinary incontinence (SUI) with synthetic
sling placement, to management of refractory overactive bladder (OAB)
with neuromodulation.
Clifton, M. M. & Goldman, H. B. Nat. Rev. Urol. 11, 613–614 (2014); published online 14 October 2014;
doi:10.1038/nrurol.2014.279
For most of the past decade, urodynamic
studies (UDS) were a routine part of the
evaluation for a patient with stress urinary
incontinence (SUI). UDS is an invasive
and relatively expensive procedure that can
cause anxiety in patients and, until recently,
its value in improving treatment outcomes
was unknown. However, in 2012, Nager
et al.1 published a multicentre, random­
ized noninferiority trial evaluating the
need for UDS in patients undergoing mid­
urethral sling surgery. In this study, women
with uncomplicated SUI were randomized
to undergo office evaluation and UDS or
office evaluation only. Patients had a history
of SUI symptoms for 3 months or more, a
score on the Medical, Epidemiological and
Social Aspects of Aging (MESA) inventory
for SUI that was greater than their score for
urinary urgency incontinence (UUI), evi­
dence of urethral mobility on examination
and a posi­tive stress test during physical
examination. Patients with previous anti-­
incontinence surgery, a history of pelvic
radiation, pelvic surgery within the past
3 months and women with pelvic organ pro­
lapse of stage III or greater were excluded
from the study. The primary outcome of
treatment success at 12 months was deter­
mined by a Urogenital Distress Inventory
(UDI) reduction of 70% and a Patient
Global Impression of Improvement (PGII)
Box 1 | Developments in female urology
■■ Urodynamic studies (UDS) are no longer
required in straightforward stress urinary
incontinence (SUI) evaluation
■■ Synthetic midurethral sling placement is
now the standard treatment for SUI
■■ OnabotulinumtoxinA is now a standard
treatment for overactive bladder (OAB)
■■ Percutaneous tibial nerve stimulation
(PTNS) and sacral neuromodulation are
now options for the treatment of OAB
81  |  NOVEMBER 2015  www.nature.com/reviews
© 2015 Macmillan Publishers Limited. All rights reserved
of ‘much better’. A total of 630 women were
randomized equally into the two study arms.
Successful treatment was observed in 76.9%
of patients who underwent UDS and office
evaluation, compared with 77.2% of patients
who only received office evaluation, showing
that UDS has no real effect on patient out­
comes. Eliminating the need to put patients
with straightforward SUI through this
medical examination should expedite
patient care, decrease morbidity, reduce
costs and improve the patient experience.
‘‘
…the past decade has been
an exciting one for the field
of female urology and voiding
dysfunction
’’
Throughout the previous decade, the gold
standard for the treatment of SUI involved a
retropubic suspension or autologous pubo­
vaginal sling—both of which are relatively
invasive, typically involve general or regional
anaesthesia and often require overnight
hospitalization. Tension-free vaginal tape
(TVT)—a synthetic retro­pubic midurethral
sling—was introduced in the late 1990s, and
was widely adopted by uro­logists over the
decade between 2004–2014, owing to excel­
lent 5‑year outcome results and minimal
invasiveness compared with retropubic sus­
pensions or autologous pubovaginal slings.2
Nilsson et al.3 recently published the 17 year
follow-up period results of a multicentre pro­
spective TVT study. A total of 78% of women
initially included in the study were evalu­
ated, with over 90% of these patients objec­
tively continent on follow-up after 17 years,
and 87% of patients subjectively cured or
significantly improved according to PGII,
Incontinence Impact Questionnaire-short
form (IIQ‑7), the UDI-short form (UDI‑6),
and Urinary Incontinence Severity Score
(UISS). This prospective study provides
evidence for the durability of the TVT sling
as well as excellent patient outcomes with
minimal complications. The synthetic retro­
pubic midurethral sling is now considered a
gold standard in the management of female
SUI.4 Overall, with the advent of the synthetic
retropubic midurethral sling and the transob­
turator midurethral sling (which has accrued
data demonstrating excellent results for most
patients)5 the treatment of SUI has become
less invasive, more predictable and more
acceptable to patients and physicians alike.
One of the most important advances to
have garnered attention over the past few
years has been in the treatment of over­
active bladder (OAB). Previously, OAB has
been difficult to treat, and limited thera­
peutic options have been available; but now
onabotu­linumtoxinA is considered the
standard treatment for patients with either
refractory idiopathic OAB or neurogenic
detrusor overactivity (NDO).6 Recently, Nitti
et al.7 performed a multi­centre, random­
ized, placebo-­controlled trial to evaluate
onabotulinumtoxinA for the treatment of
patients with idiopathic OAB and UUI.
Patients were ≥18 years old and experi­enced
≥3 UUI episodes over a 3‑day period. They
were randomized to receive either 100 U
of intravesical onabotulinumtoxinA or
placebo. Primary end points were a change
in the mean number of UUI episodes per
day and the proportion of patients with a
positive treatment response on the treat­
ment benefit scale. Secondary end points
were improvement in OAB symptoms, such
as a decrease in the average frequency of
micturition and urgency episodes as well as
scores on health-related quality of life (QOL)
question­naires. This study demonstrated that
onabotulinumtoxinA therapy decreased
the mean daily frequency of UUI episodes
experienced by patients by 2.65 compared
with a decrease in patients receiving placebo
of only 0.87, which was statistically signifi­
cant (P <0.001). Additionally, 60.8% of
patients receiving onabotulinumtoxinA
had a positive response according to QOL
questionnaire results, compared with only
a 29.2% positive response rate in patients
who received placebo (P <0.001). These
patients are generally difficult to treat and
have limited treatment options available to
them; onabotulinum­toxinA therapy has been
shown to improve QOL for these patients,
giving them a viable treatment option.
The efficacy of onabotulinumtoxinA in
patients with NDO has also been demon­
strated by Schurch et al. 8 using a multi

c­ entre, double-blind, randomized, placebo-­
controlled trial. The patients inclu­ded were
≥18 years old with urinary incontinence
proven by UDS, caused by NDO. Patients
had the option to continue their previously
prescribed anticholinergic therapy while
they were on the study, and all patients
performed intermittent catheterization.
The selected patients were randomized
1:1:1 to receive a single dose of 200 U or
300 U of onabotulinum­toxinA, or placebo.
Schurch and colleagues8 observed a signifi­
cant decrease in incontinence episodes in
both treatment arms (P ≤0.05), but not in
the placebo group. Additionally, patients
who received either 200 U or 300 U of ona­
botulinumtoxinA had improved bladder
function on UDS, including in factors such
as maximum cysto­metric capacity, reflex
detrusor volume and maximum detrusor
pressure as well as improvements in quality
of life scores. This study did not demonstrate
a signifi­cant difference in effect between
the different onabotulinum­toxinA doses,
owing to its small sample size; however, it
did provide significant clinical evidence that
onabotu­linumtoxinA decreases signs and
symptoms of urinary incontinence in patients
with NDO and, consequently, fundamentally
changed the management of these patients.
Neuromodulation therapy has gained
support in the literature and has become a
treatment option for patients with medi­
cally refractory OAB. Percutaneous tibial
nerve stimulation (PTNS) is an example
of a neuro­modulation treatment and is a
simple minimally invasive treatment for
OAB, which can be performed in an office
environ­ment. Peters et al.9 provided signifi­
cant support for the use of PTNS in a multi­
centre, double-blind, randomized controlled
trial of 220 patients who were ≥18 years old
and had OAB symptoms. These patients
were randomized to 12 weeks of treatment
with PTNS or a sham procedure. Participants
who underwent PTNS showed significant
improvement in bladder symptoms, with
54.5% reporting a moderate, or greater than
moderate improvement, compared with
20.9% of patients who underwent the sham
procedure (P <0.001). This study supported
the safe and effective use of PTNS in the
treatment of OAB and it is now considered a
feasible option in the treatment of medically
refractory OAB.6 Sacral neuromodulation
has also become an option in the treatment
of medically refractory OAB over the past
10 years.6 van Kerrebroeck et al.10 performed
a 5-year prospective, multicentre trial to
evaluate the use of sacral neuromodu­lation
A DECADE IN MEDICINE NOVEMBER 2015  |  82
therapy for urinary voiding dysfunction.
A total of 129 patients ≥18 years old who
had UUI, urinary frequency and/or uri­
nary retention were included in the study.
In patients with UUI, the mean number of
leak­age episodes per day had decreased from
9.6 to 4.0 at the 5-year follow-up point. In
patients with urgency frequency, the mean
number of voids per day decreased from 19
to 15 and the mean voided volume increased
from 92 ml to 165 ml. This prospective trial
demon­strated the safety and efficacy of sacral
neuromodulation in patients with refractory
OAB and has improved treatment outcomes.
Undoubtedly, the past decade has been
an exciting one for the field of female uro­
logy and voiding dysfunction. Significant
advances have been made in the evaluation
and management of patients with SUI and
a host of new treatment options for OAB
have been established (Box 1). A previously
d­ifficult-to-treat population now has excel­
lent minimally invasive options for treat­
ment of their complex problems. Hopefully,
the field will continue to evolve and develop
both better diagnostic and treatment options.
Center for Female Pelvic Medicine and
Reconstructive Surgery, Glickman Urologic
and Kidney Institute, The Cleveland Clinic,
Lerner College of Medicine, Glickman Tower
(Q Building), 9500 Euclid Avenue, Cleveland,
OH 44195, USA (M.M.C., H.B.G.).
Correspondence to: H.B.G.
goldmah@ccf.org
Competing interests
H.B.G. is a consultant and speaker for Allergan,
Medtronic and Uroplasty and a speaker for Astellas,
M.M.C. declares no competing interests.
DECADE IN REVIEW—KIDNEY CANCER
Discoveries, therapies
and opportunities
W. Marston Linehan and Christopher J. Ricketts
Several advances in kidney cancer have occurred over the past decade,
including the discovery of mutations in chromatin remodelling genes and
genomic heterogeneity in clear cell renal cell carcinoma (ccRCC). Altered
metabolic patterns in ccRCC and papillary RCC have become apparent,
and new drugs for ccRCC have been approved.
Linehan, W. M. & Ricketts, C. J. Nat. Rev. Urol. 11, 614–616 (2014); published online 7 October 2014;
doi:10.1038/nrurol.2014.262
Between 2004 and 2014, remarkable
improvements in our understanding of
the genetic basis of renal cell carcinoma
(RCC) and in the treatment of patients
with advanced kidney cancer have been
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
1. Nager, C. W. et al. A randomized trial of
urodynamic testing before stress-incontinence
surgery. N. Engl. J. Med. 366, 1987–1997
(2012).
2. Nilsson, C. G., Kuuva, N., Falconer, C.,
Retzapour, M. & Ulmsten, U. Longer-term
results of the tension-free vaginal tape (TVT)
procedure for surgical treatment of female
stress urinary incontinence. Int. Urogynecol. J.
Pelvic Floor Dysfunct. 12, 5–8 (2001).
3. Nilsson, C. G., Palva, K., Aarnio, R., Morcos, E.
& Falconer, C. Seventeen years’ follow-up of the
tension-free vaginal tape procedure for female
stress urinary incontinence. Int. Urogynecol. J.
Pelvic Floor Dysfunct. 24, 1265–1269 (2013).
4. Nager, C., Tulikangas, P., Miller, D. &
Goldman, H. Position statement on mesh
midurethral slings for stress urinary
incontinence. Female Pelvic Med. Reconstr.
Surg. 3, 123–125 (2014).
5. Richter, H. E. et al. Retropubic versus
transobturator midurethral slings for stress
incontinence. N. Engl. J. Med. 362, 2066–2076
(2010).
6. Gormley, E. A. et al. Diagnosis and treatment of
overactive bladder (non-neurogenic) in adults:
AUA/SUFU guideline. J. Urol. 188, 2455–2463
(2012).
7. Nitti, V. W. et al. Onabotulinum toxin A for the
treatment of patients with overactive bladder
and urinary incontinence: results of a phase 3,
randomized, placebo controlled trial. J. Urol.
189, 2186–2193 (2013).
8. Schurch, B. et al. Botulinum toxin type A is a
safe and effective treatment for neurogenic
urinary incontinence: results of a single
treatment, randomized, placebo controlled
6‑month study. J. Urol. 174, 196–200 (2005).
9. Peters, K. M. et al. Randomized trial of
percutaneous tibial nerve stimulation versus
sham efficacy in the treatment of overactive
bladder syndrome: results from the SUmiT Trial.
J. Urol. 183, 438–443 (2010).
10. van Kerrebroeck, P. E. et al. Results of sacral
neuromodulation therapy for urinary voiding
dysfunction: outcomes of a prospective,
worldwide clinical study. J. Urol. 178,
2029–2034 (2007).
realized. These developments include the
approval by the FDA of seven agents target­
ing the VHL/HIF pathway for patients with
advanced-stage RCC, and the identifica­
tion of gene mutations in kidney tumours.
UROLOGY
These advances built on the progress of the
previous decade, which had seen the dis­
covery of the VHL gene, the identification
of mutations of the VHL gene in clear cell
RCC (ccRCC) and the delineation of the
VHL pathway. Mutations in MET and their
involvement in hereditary type 1 papil­lary
kidney cancer were also identified, and
FLCN gene mutations were shown to be
the primary cause for inherited chromo­
phobe kidney cancer associated with
Birt–Hogg–Dubé syndrome. Mutations in
the FH gene, which encodes a Krebs cycle
enzyme, were also identified as the cause
of most cases of hereditary type 2 papillary
kidney cancer associated with hereditary
leiomyo­matosis and renal cell carcinoma
(HLRCC). The understanding that kidney
cancer is made up of a number of different
types of cancer, each having different histo­
logy, clinical course and genetic causes, has
aided p­rogress in treating this disease.
‘‘
…understanding that kidney
cancer is made up of a number
of different types of cancer …
’’
aided progress…
At the start of this decade, only IL‑2 was
approved for the treatment of patients with
advanced-stage kidney cancer. The deline­
ation of the VHL/HIF pathway and its role in
RCC in the 1990s provided the foundation
for the development of targeted t­herapeutic
agents for patients with this disease. Since
then, several randomized, double-blind
trials have identified effective agents for
the treatment of kidney cancer, targeting
vascular endothelial growth fac­tor (VEGF)
and platelet-derived growth fac­tor (PDGF)
pathways. The first targeted agent against
the VEGF and PDGF receptors, sorafenib,
was approved for patients with advanced-
stage kidney cancer in 2005. Patients
treated with sorafenib were shown to have
a median progression-­free survival (PFS)
of 5.5 months, compared with 2.8 months
for those given placebo. In 2006, suni­
tinib, which also targets the VEGF and
PDGF receptors, was approved. Sunitinib
therapy was associated with an 11-month
PFS compared with 5 months for patients
treated with IFN-α, a nonspecific immune
modulator that has been used for a number
of years in patients with advanced-stage
kidney cancer.
The mTOR pathway is an important
kidney cancer pathway that in­d irectly
83  |  NOVEMBER 2015  www.nature.com/reviews
influences VEGF and PDGF expression by
affecting HIF translation. In 2007, temsiro­
limus, an agent that targets mTOR, was
approved after patients with poor prognos­
tic factors experienced a PFS of 10.9 months
on the drug compared with 7.3 months for
those treated with IFN-α. Three agents were
approved in 2009: the first, everolimus,
also targets the mTOR pathway and was
approved for patients with RCC who had
previously received sunitinib or sorafenib
therapy. Those treated with everolimus
demonstrated PFS of 4.0 months compared
with 1.9 months for placebo. The second
drug, bevacizumab (a monoclonal anti­
body against VEGF‑A) was approved for
use in conjunction with IFN-α in patients
with advanced-stage RCC. Treatment with
bevacizumab and IFN-α increased PFS
to a median of 10.2 months, compared
with 5.4 months in patients treated with
IFN-α alone. Finally, pazopanib, a multi­
kinase inhibitor targeting the VEGF and
PDGF receptors, was also approved to treat
patients with advanced-stage RCC. Pazo­
panib therapy was found to be associ­ated
with a PFS of 9.2 months compared with
4.2 months for patients given a placebo.
In January 2012, axitinib, another multi­
kinase inhibitor, was the seventh targeted
therapeutic agent this decade approved
by the FDA for the treatment of patients
with kidney cancer. Patients treated with
axitinib were found to have a median PFS
of 6.7 months compared with 4.7 months
for patient receiving sorafenib.1
These accomplishments provided physi­
cians with a range of targeted agents to treat
patients for the first time. Although the
response rates for treatment with agents
such as sunitinib or pazopanib are reported
to be approximately 35%, and there have
been improvements in PFS and overall sur­
vival, few treatments result in a complete
response; most patients eventually progress
and many die of their cancer. This eventual­
ity could be because targeting the VHL/HIF
pathway with currently available therapies is
inadequate, or because there are other genes
associated with the initiation or progres­
sion of kidney cancer that are not currently
being targeted.
Exciting new insights into the genetic
basis of ccRCC came from reports identify­
ing mutations in several chromatin remod­
elling, histone modifying and SWItch/
SNF nucleosome remodelling complex
genes.2,3 When The Cancer Genome Atlas
Research Network performed a compre­
hensive molecu­lar characterization of
© 2015 Macmillan Publishers Limited. All rights reserved
ccRCC genomic Sorafenib
heterogeneity Sunitinib
Bevacizumab
Pazopanib
VHL–/– Axitinib
VHL–/– Everolimus
TSC2–/– Temsirolimus
VHL–/– Everolimus
PTEN–/– Temsirolimus
SETD2–/–
???
VHL–/–
BAP1–/– ???
VHL–/– ???
SETD2–/–
SMARCA4–/– ???
Figure 1 | Genomic heterogeneity in different
regions of ccRCC. The ‘truncal’ VHL gene is
mutated in all regions, and can be targeted
with sunitinib, sorafinib, bevacizumab,
pazopanib or axitinib. The ‘branch’ mutations,
in select regions of the tumour, can occur in
genes for which potential targeted therapies
do or do not exist, such as SETD2, BAP1 or
SMARCA4. The most effective strategy could
involve combination therapy targeting the
VHL/HIF pathway, or a global strategy
targeting chromatin remodelling gene
pathways. Abbreviation: ccRCC, clear cell
renal cell carcinoma.
genetic changes associated with ccRCC, 19
significantly mutated genes were identified.
After VHL, the most commonly mutated
genes were the chromosome 3 chromatin
remodel­ling genes—PBRM1, SETD2 and
BAP1. Mutations of BAP1 were found to be
correlated with poor survival. Furthermore,
mutations were identified in KDM5C, TP53
and the PI3K/Akt pathway genes (PTEN,
MTOR and PIK3CA). It was also demon­
strated that the PI3K/Akt pathway has an
important role in tumour progression.
Integrative analysis showed that high-grade,
high-stage, low-survival ccRCC tumours
demonstrated evidence of a Warburg-like
metabolic shift, involving downregula­
tion of genes involved in the Krebs cycle,
decreased AMPK and increased fatty acid
synthesis.4 These findings, which are con­
sistent with the isotopomer spectral analysis
of VHL–/– ccRCC and FH–/– type 2 papillary
RCC, reveal that a dependence on reductive
glutamine metabolism for lipid biosynthe­
sis is important in kidney cancer and could
provide the foundation for the develop­
ment of novel forms of therapy targeting
the m­etabolic basis of RCC.5,6
To further understand the genetic basis
of ccRCC, Gerlinger and colleagues 7
reported the use of multiregion exome
sequenc­ing (M-seq) to study the genomic
architecture and evolution of ccRCC, and
found extensive genomic heterogeneity.

‘Truncal’ mutations of the VHL gene were
found in every segment of each sample,
whereas driver mutations in the chro­
matin remodel­ling genes SETD2, BAP1
and KDM5C, as well as TP53 and genes
in the PI3K/mTOR pathway (MTOR,
PIK3CA, PTEN and TSC2), were found
only in some segments of the tumours;
these were labeled ‘branch’ mutations. Of all
the driver mutations, 75% were subclonal,
and intratumoral hetero­geneity increased
with the number of biopsy specimens analy­
sed.7 These findings raise profound ques­
tions about the most effective way to detect
driver mutations in ccRCC, and which
gene pathways to target. Should M‑seq be
performed on primary tumours and/or
metastatic sites to identify common driver
mutations that are associated with the initia­
tion and progression of kidney cancer? Do
both the truncal and branching driver gene
pathways need to be targeted, singularly or
together, to develop an effective form of
therapy for ccRCC? These criti­cal questions
need to be answered in the next decade in
kidney cancer research.
Although a number of novel therapies
for patients with advanced-stage ccRCC
have been developed over the past decade,
we still have no effective treatment for
patients with advanced-stage papillary
kidney cancer. Significant insight into
papil­l ary kidney cancer has come from
studies of the hereditary form of type 2
papillary renal carcinoma found in patients
with HLRCC, caused by mutation of the
FH gene. FH‑deficient kidney cancer
cells, which exhibit impaired oxidative
phosphory­lation, undergo a Warburg-like
metabolic shift to aerobic glycolysis with
decreased AMPK and increased mTOR and
HIF-1α levels. 8 Studies by Adam et al. 9
and Ooi et al.10 showed upregulation of the
antioxidant signalling pathways in both
FH‑deficient kidney cancer and sporadic
(non-inherited) papillary kidney cancer.
Elevated fumarate in FH‑deficient RCC
targets and inactivates the electrophile
sensor KEAP1, which results in NRF2 accu­
mulation and upregulation of antioxidant
response element-controlled genes, such
as NQO and the GST family genes. These
genes are critical for the survival of cancers,
which are characterized by high levels of
oxidative stress, such as hereditary and
sporadic type 2 papillary kidney cancer.9,10
This pathway provides a number of oppor­
tunities for the development of targeted
therapeutic approaches for patients with
advanced-stage papillary RCC.
A DECADE IN MEDICINE NOVEMBER 2015  |  84
Overall, this past decade has seen a
num­b er of historic advances towards the
understanding and treatment of RCC,
with the FDA approving seven drugs tar­
geting the VHL/HIF pathway for patients
with advanced-stage kid­ney cancer. Figure 1
illustrates the exten­sive genomic hetero­
geneity of ccRCC, which poses a signifi­
cant therapeutic challenge; intense efforts
are underway to translate discoveries into
new therapeutic approaches for patients.
The finding of a Warburg-like metabolic
shift and the dependence of the KEAP1/
NRF2 antioxidant pathway in both clear
cell and papillary RCC provides promis­
ing opportunities for the development of
therapies t­a rgeting the m­e tabolic basis
of kidney cancer.
Urologic Oncology Branch, National Cancer
Institute, 10 Center Drive, MSC 1107,
CRC Room 1W‑5940, Bethesda,
MD 20892‑1107, USA (W.M.L., C.J.R.).
Correspondence to: W.M.L.
wml@nih.gov
Competing interests
The authors declare no competing interests.
1. Jonasch, E. & Motzer, R. J. Ten years of progress
in renal cell carcinoma. J. Natl Compr. Canc.
Netw. 10, 690–693 (2012).
DECADE IN REVIEW—SEXUAL DYSFUNCTION
Post-RP erectile dysfunction
—therapies for the next decade
Emmanuel Weyne and Maarten Albersen
Erectile dysfunction remains a frequent sequela of radical prostatectomy,
owing to injury to the cavernous nerves that innervate the penile erectile
tissue. Over the last decade, many strategies have been proposed to
minimize the duration of denervation and prevent irreversible structural
changes from occurring in the corpora cavernosa.
Weyne, E. & Albersen, M. Nat. Rev. Urol. 11, 616–618 (2014); published online 30 September 2014;
doi:10.1038/nrurol.2014.274
Erectile dysfunction (ED) after radical pros­
tatectomy remains a major health concern
for patients, urologists and practitioners in
sexual medicine. The presence and persis­
tence of ED after radical prostatectomy has
a substantial negative impact on the quality
of daily life for men who have undergone
the procedure. Short-term oncological
results are excellent for radical prostatec­
tomy, but the preservation and rehabilita­
tion of sexual function remains a challenge.
Potency rates reported after nerve-sparing
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
2. Dalgliesh, G. L. et al. Systematic sequencing of
renal carcinoma reveals inactivation of histone
modifying genes. Nature 463, 360–363
(2010).
3. Varela, I. et al. Exome sequencing identifies
frequent mutation of the SWI/SNF complex
gene PBRM1 in renal carcinoma. Nature 469,
539–542 (2011).
4. The Cancer Genome Atlas Research Network.
Comprehensive Molecular Characterization of
Clear Cell Renal Cell Carcinoma. Nature 499,
43–49 (2013).
5. Metallo, C. M. et al. Reductive glutamine
metabolism by IDH1 mediates lipogenesis
under hypoxia. Nature 481, 380–384
(2011).
6. Mullen, A. R. et al. Reductive carboxylation
supports growth in tumour cells with
defective mitochondria. Nature 481, 385–388
(2011).
7. Gerlinger, M. et al. Genomic architecture and
evolution of clear cell renal cell carcinomas
defined by multiregion sequencing. Nat. Genet.
46, 225–233 (2014).
8. Tong, W. H. et al. The glycolytic shift in
fumarate‑hydratase‑deficient kidney cancer
lowers AMPK levels, increases anabolic
propensities and lowers cellular iron levels.
Cancer Cell 20, 315–327 (2011).
9. Adam, J. et al. Renal cyst formation in Fh1-
deficient mice is independent of the Hif/Phd
pathway: roles for fumarate in KEAP1
succination and Nrf2 signaling. Cancer Cell 20,
524–537 (2011).
10. Ooi, A. et al. An antioxidant response
phenotype shared between hereditary and
sporadic type 2 papillary renal cell carcinoma.
Cancer Cell 20, 511–523 (2011).
radical prostatectomy vary greatly between
14% and 86%, mainly owing to reporting
factors (such as the definition of ‘potency’
used in each study), surgeon experience and
patient preoperative characteristics.1
Over the past decade, researchers have
investigated many different strategies to
reduce the time that the corpora caver­
nosa lack innervation and limit irreversible
structural changes—caused by apoptosis
of smooth muscle cells and fibrosis of the
v­ascular bed—from occurring.
UROLOGY
In 1999 the concept of nerve graft inter­
position was established 2 and was sub­
sequently adopted during the past decade,
mainly in tertiary referral centres. This
technique is especially useful in men at
increased risk of extracapsular extension of
their tumour, who undergo resection of one
or both cavernous nerves resulting in a wide
gap between the nerve ends, which mini­
mizes the potential for full nerve regenera­
tion. Preclinical rodent studies and initial
reports with sural grafts in observational
prospective studies in humans were promis­
ing, reporting high potency rates after uni­
lateral nerve reconstruction.3 However, a
2009 large-scale randomized phase II trial,
which compared patients with unilateral
nerve-sparing radical prostatectomy with
or without unilateral sural nerve graft inter­
position, did not show significant differ­
ences in recovery of erectile function 2 years
after surgery.4 These disappointing results
raised questions regarding the added benefit
of nerve grafting in the setting of unilateral
nerve-sparing radical prostatectomy, and
greatly decreased the general popularity of
the procedure. The use of Schwann-cell-
seeded artificial grafts, addition of neuro­
trophic agents and combination use with
robotic procedures, which provide superior
visualisation and high manoeuvrability, are
future directions in this field.
All the postoperative strategies developed
over the past decade share the common aim
of penile rehabilitation: maintenance of
erectile function by activating the normal
physiological process of erection, with
the ultimate goal of resumption of medi­
cally unassisted sexual activity. The loss of
spontaneous and nocturnal erections after
radical prostatectomy mean that the penis
is rendered in chronic state of hypoxia,
causing structural changes to the erectile
tissue. Increasing cavernosal perfusion
would increase penile oxygenation, protect
Image courtesy of Maarten Albersen
85  |  NOVEMBER 2015  www.nature.com/reviews
endothelial function and limit muscular
apoptosis. After the initial introduction
of the concept with intracavernosal injec­
tion therapy, PDE5-inhibitors (PDE5Is)
have become the most commonly used
form of penile rehabilitation, owing to
their favourable mode of administration
and ability to increase penile oxygenation.
Preclinical studies in rodent models have
shown evidence of a benefit of PDE5Is in
rehabilitating erectile function. However,
clinical evidence remains limited to one
small randomized trial (76 patients),
in which the sildenafil group (50 mg or
100 mg) had a significantly higher propor­
tion of responders (27%) than the placebo
group (4%) after wash-out.5 The conclu­
sions drawn from this study are, however,
impaired by the small number of patients,
high withdrawal rate, the lower placebo
response than generally reported for nerve-
sparing surgery, as well as a nonvalidated
primary study outcome (responses to ques­
tion 3 and 4 of the Inter­national Index of
Erectile Function [IIEF] questionnaire and
a response to the question “were erections
good enough for satisfactory sexual activ­
ity?”, not overall IIEF-EF score). Conversely,
a larger randomized and well-performed
trial found no difference in potency rates
between groups taking 10 mg vardena­
fil nightly plus on-demand placebo, on-
demand vardenafil plus nightly placebo, or
nightly placebo plus on-demand placebo
for 9 months with a subsequent 2‑month
wash-out period.6 Furthermore, in the latest
REACTT trial,7 tadalafil—a PDE5I with an
extended half-life—failed to show an erec­
tile function benefit compared with placebo
after the wash-out interval of 6 weeks. Penile
rehabilitation with PDE5Is has a sound
conceptual basis, but hard clinical evidence
for its use is still lacking. Starting penile
rehabili­tation early—immediately after, or
even before, radical prostatectomy—and
risk stratification to select the patients
whom could most benefit from the treat­
ment, are important considerations looking
to the future of penile rehabilitation.
The absence of conclusive clinical evi­
dence in favour of PDE5Is for penile rehabi­
litation has, more recently, shifted the focus
towards the initiator of the pathophysio­
logical cascade: the nerve injury itself. One
strategy has been the use of immunophilin
proteins and ligands. Immunophilins
such as FK506 (tacrolimus), rapamycin,
and cyclosporin are multifunctional pro­
teins with roles in immunoregulation, and
have long histories in the prevention of
© 2015 Macmillan Publishers Limited. All rights reserved
allo­graft transplant rejection. The rationale is
that these agents could temper the neuro­
inflammatory reaction occurring after
nerve injury, and they have shown much
promise in preclinical studies. However,
although a 2008 study showed that tacro­
limus was highly effective in maintaining
erectile function in a bilateral cavernous
nerve crush model in rats,8 a clinical trial
(NCT00106392) failed to show a differ­
ence in recovery of erections 24 months
after bilateral nerve-sparing radical pros­
tatectomy between men who received
either tacrolimus or a placebo adminis­
tered from 4 to 10 days before and 6 months
after surgery.
‘‘
…ADSC migrate to the major
pelvic ganglion (MPG), from which
arise the axons that innervate
the penis…
’’
A similar strategy—to enhance neuro­
regeneration with neuromodulatory ther­
apy—is based on the instrinsic ability of the
peripheral nervous system to regenerate
after injury. However, endogenous regener­
ation is limited and does not usually result
in full recovery. Neuronal survival is regu­
lated by different families of endogenous
neurotrophins among which brain-derived
nerve growth factor (BDNF) is one of the
most known. In 2006, Bella et al.9 elegantly
showed that BDNF is able to increase caver­
nous nerve neurite outgrowth in vitro and
improve recovery of erectile function after
cavernous nerve crush trauma. These
results looked promising, but BDNF, like
most of these neurotrophin peptides, is not
suited for oral administration. How­e ver,
neurotrophic factors could—h­ypothetically
—be applied in slow-releasing hydrogels
placed in the proximity of the NVB at the
end of the radical prostatectomy procedure,
and this concept is something that could be
investigated in the future.
Mesenchymal stem cells (MSC) have been
shown to be a successful treatment option
in clinical trials for various cardiovascular,
inflammatory and degenerative diseases.
Of the various MSCs, adipose-derived stem
cells (ADSC) show the greatest potential
as a treatment option for ED after radical
prostatectomy. ADSC are easy to isolate and
can be administered as a stromal vascular
fraction during the same surgical session as
the prostatectomy procedure. After admin­
istration and cavernous nerve injury (CNI),
ADSC migrate to the major pelvic ganglion

‘‘ Many men remain
undertreated with the current
armoury of therapies that are
’’
available in clinical practice
(MPG), from which arise the axons that
innervate the penis10 and can, therefore, be
administered systemically by intravenous
injection. At the site of injury, they act as
a local factory of molecules that promote
neuroregeneration and temper the inflam­
matory reaction after nerve injury. In this
way, injection of ADSCs has been shown to
significantly improve erectile function after
CNI in a rat model.10 Multiple clinical trials
are currently enrolling patients, including
one at Johns Hopkins Medical Institutions
(NCT01983709), which will investigate the
safety of systemically administered allo­
genic bone-marrow-derived stem cells in
men with prostate cancer and the effect
on erectile recovery after radical prosta­
tectomy as a secondary end point. We look
forward to the results of this trial, which
have the potential to stimulate the wide­
spread use of stem cells as a new treatment
for CNI‑associated ED.
ED remains a frequent concern after
radical prostatectomy, owing to injury to
the cavernous nerves. Many men remain
undertreated with the current armoury of
therapies that are available in clinical prac­
tice. Perioperative tools to facilitate NVB
identification and cavernous nerve recon­
struction using sural nerve grafts have not
achieved the improvements in erectile func­
tion recovery that they seemed to promise
in early studies, and PDE5Is remain the
most commonly used treatment, despite
the paucity of clinical evidence in their
favour. Preclinical studies carried out
over the last decade have shown exciting
potential for stem cell and neurotropic
factor therapy to boost the endogenous
neuroregenerative response after CNI and
we eagerly await their implementation
in clinical trials. As we look to the future,
these therapies represent an auspicious new
avenue for penile rehabilitation.
Laboratory for Experimental Urology,
Department of Development and
Regeneration, KU Leuven and University
Hospitals, Campus Gasthuisberg, O&N 1,
Herestraat 49, Box 802, 3000 Leuven, Belgium
(E.W., M.A.).
Correspondence to: M.A.
maarten.albersen@uzleuven.be
Competing interests
The authors declare no competing interests.
A DECADE IN MEDICINE NOVEMBER 2015  |  86
1. Boorjian, S. A. S. et al. A critical analysis of
the long-term impact of radical prostatectomy on
cancer control and function outcomes. Eur. Urol.
61, 664–675 (2012).
2. Kim, E. D. et al. Interposition of sural nerve
restores function of cavernous nerves resected
during radical prostatectomy. J. Urol. 161,
188–192 (1999).
3. Namiki, S. et al. Impact of unilateral sural nerve
graft on recovery of potency and continence
following radical prostatectomy: 3‑year
longitudinal study. J. Urol. 178, 212–216 (2007).
4. Davis, J. W. et al. Randomized Phase II trial
evaluation of erectile function after attempted
unilateral cavernous nerve-sparing retropubic
radical prostatectomy with versus without
unilateral sural nerve grafting for clinically
localized prostate cancer. Eur. Urol. 55,
1135–1144 (2009).
5. Padma-Nathan, H. et al. Randomized, double-
blind, placebo-controlled study of postoperative
nightly sildenafil citrate for the prevention of
erectile dysfunction after bilateral nerve-sparing
radical prostatectomy. Int. J. Impot. Res. 20,
479–486 (2008).
DECADE IN REVIEW—PROSTATE CANCER
A decade of progress in detection
and treatment
Behfar Ehdaie and Peter T. Scardino
A fundamental shift in the understanding, detection and treatment of
prostate cancer has occurred over the past 10 years, especially in the
use of screening, active surveillance, novel therapies and radical surgery.
These discoveries have changed how clinicians and patients approach
prostate cancer.
Ehdaie, B. & Scardino, P. T. Nat. Rev. Urol. 11, 618–620 (2014); published online 14 October 2014;
doi:10.1038/nrurol.2014.284
The past decade has witnessed fundamen­
tal shifts in our understanding of prostate
cancer and in the detection and treatment
of all stages of this common disease. The
greatest changes in the approach to pros­
tate cancer have been in attitudes to screen­
ing, the role of patient-reported outcomes
in documenting the effects of therapy
on quality of life (QOL), the move from
radical treatment to active surveillance for
low-risk prostate cancer, the expanding
role of surgery for high-risk cancers and
the move towards multimodality treatment
for advanced cancers. Despite this progress,
prostate cancer remains the second most
lethal cancer among men, and it will con­
tinue to challenge our best judgement and
therapeutic acumen as we work to further
reduce morbidity and mortality from this
disease in the coming decades.
PSA has been widely used for the early
detection of prostate cancer in clinical prac­
tice since its discovery. However, its value in
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
6. Montorsi, F. et al. Effect of nightly versus on-
demand vardenafil on recovery of erectile
function in men following bilateral nerve-sparing
radical prostatectomy. Eur. Urol. 54, 924–931
(2008).
7. Montorsi, F. et al. Effects of tadalafil treatment
on erectile function recovery following bilateral
nerve-sparing radical prostatectomy: a
randomised placebo-controlled study (REACTT).
Eur. Urol. 65, 587–596 (2014).
8. Sezen, S. F., Lagoda, G. & Burnett, A. L. Role of
immunophilins in recovery of erectile function
after cavernous nerve injury. J. Sex. Med. 6,
340–346 (2009).
9. Bella, A. J. et al. Brain-derived neurotrophic
factor (BDNF) acts primarily via the jak/stat
pathway to promote neurite growth in the major
pelvic ganglion of the rat: part I. J. Sex. Med. 3,
815–820 (2006).
10. Fandel, T. M. et al. recruitment of
intracavernously injected adipose-derived
stem cells to the major pelvic ganglion
improves erectile function in a rat model of
cavernous nerve injury. Eur. Urol. 61, 201–210
(2011).
population-wide screening remains contro­
versial and the results of large randomized
trials have failed to resolve the debate.
The US-based Prostate Lung Colorectal
and Ovarian Cancer screening trial found
no reduction in prostate-cancer-­specific
mortal­ity (PCSM) over a median of
10 years, but the study was compromised
by pretesting for PSA in 40% of the subjects
and by contamination of the control recruits
(some 70% of the unscreened control
cohort received a PSA test during the study
period), meaning that the study was only
able to report that intense s­c reening is
no better than opportunistic screening
for reducing mortality. 1 In contrast, the
European Randomized Study of Screening
for Prostate Cancer reported a statistically
significant relative reduction of 21% in
PCSM at a follow-up duration of 13 years,
equivalent to one prostate cancer death
averted per 781 men invited for screening,
or one per 27 additional prostate cancers
UROLOGY
detected.2 Despite this substantial reduction
in PCSM with PSA screening, in 2012 the
US Preventive Services Task Force recom­
mended against PSA screening, concluding
that no net benefit is gained from the prac­
tice and that the potential harms outweigh
the benefits. Nevertheless, the longest trial
reported to date—the Göteborg random­
ized population-based prostate cancer
screening trial—showed a 44% reduction
in PCSM at a follow-up period of 14 years,
with screening of 293 men and diagno­
sis of 12 cancers preventing one death.3
The power of PSA testing was confirmed
by studies investigating the association
between PSA levels at mid-life, defined
as between 45–60 years, and the lifetime
risk of prostate cancer metastases and death,
enabling decisions to be made regarding
further screening and treatment options.4
These studies have also shown that men
with PSA levels below the reported median
concentration (1.1 ng/ml) at 60 years have
little chance of dying from prostate cancer
and can, therefore, be excluded from further
screening. This evidence provides weight to
the argument for the value of PSA testing
to risk-adjust s­creening strategies.
‘‘ Despite this progress, prostate
cancer remains the second most
lethal cancer among men…
’’
The harms of early detection as a result of
PSA screening arise mostly from the con­
sequences of unnecessary radical therapy
for low-risk cancers. Patient-reported out­
comes provide objective tools for quantify­
ing the adverse effects of therapy and its
impact on QOL. In a landmark longitudinal
observational study of the long-term effects
of surgery and radiotherapy on QOL, most
men were shown to experience significant
short-term reductions in urinary and sexual
function after surgery, with less severe
reductions in urinary and sexual func­
tion,  but greater impairment of bowel
function, after radiation. Although these
functional deficits can improve in some
men over time, long-term adverse effects
—especially in sexual f­unction—persist in
many.5 For this reason, active surveillance
for men with low-risk cancer is becoming
more common. Men followed in active
surveillance programmes have little risk of
dying from prostate cancer within 10 years,
and many cancers that do progress respond
to delayed treatment with surgery or radia­
tion. Nevertheless, longer follow-up periods
87  |  NOVEMBER 2015  www.nature.com/reviews
could reveal greater risks of progression for
men on active surveillance than for those
given radical treatment. Furthermore, active
surveillance requires frequent biopsies,
which are invasive and carry a risk of infec­
tion and bleeding, and the criteria under
which men are assessed for delayed inter­
vention are poorly defined. However, there
is a growing consensus that men with low-
iStock/Thinkstock
risk cancer do not usually need immedi­ate
definitive therapy (assuming that no high-
risk or intermediate-risk cancer is found on
careful evaluation) and can be put on active
surveillance programmes and men with
intermediate-risk and high-risk cancers
are less likely to develop metastases or to
die of cancer if they are treated with radical
p­rostatectomy or radiation therapy.
Prediction tools and nomograms have
also been improved and are now widely
available to assist physicians and patients
in the shared decision-making process
about the treatment options available
for localized prostate cancer. Together
with more accurate biopsy strategies and
improved imaging with MRI, these tools
provide more accurate risk assessments
than were previously available and a greater
confidence that active surveillance is safe
for men with low‑risk prostate cancer.
The ability to make better medical deci­
sions in collaboration with patients and
their families will continue to improve
in the next decade, as we standardize the
evaluation of candidates for active surveil­
lance. The most promising tools for better
characterization of cancer include advanced
imaging technologies such as multipara­
metric MRI, molecular profiling tools—for
example Prolaris®(Myriad Genetics, USA)
and OncotypeDX®(Genomic Health, USA)
—and tests for circulating biomarkers,
such as the 4Kscore®(OPKO Diagnostics,
USA) and the Prostate Health Index (phi).
Improved risk assessment will help to more
precisely define eligibility for active surveil­
lance programmes, and the character­istics
that indicate that delayed intervention is an
appropriate treatment option.
During most of the past century, open
radical prostatectomy (ORP) has been the
standard surgical approach in men with
localized prostate cancer. However, over
the past decade, robot-assisted laparoscopic
prostatectomy (RALP) has been widely
adopted, despite an absence of formal evalu­
ation documenting superior outcomes. In
fact, population-based observational studies
and single-institution case series continue
to show comparable outcomes for both
© 2015 Macmillan Publishers Limited. All rights reserved
approaches, except that patients receiv­
ing RALP have slightly shorter hospital
stays (for example 2 days for RALP versus
3 days for ORP) and fewer blood trans­
fusions.6 According to administrative data
RALP does not improve complication rates
or readmission rates compared with ORP.
The skill level of the surgeon, indepen­
dent of the technology employed, clearly
has the greatest impact on outcomes. The
positive association of surgical experience
with improved outcomes has been mapped
in learning curves charting surgeon experi­
ence with biochemical-recurrence rates and
patient‑reported functional outcomes.7
Landmark clinical trials have been com­
plemented by breakthroughs in translational
research, improving our understanding of
the molecular and genetic factors that drive
prostate cancer. The fusion between trans­
membrane protease serine 2 (TMPRSS2)
and ERG, a member of the erythroblast
transformation-specific transcription factor
family, was reported by Tomlins et al. 8
work­ing under Chinnaiyan, and is found in
approximately 50% of primary and meta­
static prostate cancers. Although the clini­
cal relevance of this fusion gene remains
debatable, its high prevalence provides an
important pathway for investigating new
diagnostic and prognostic markers, as well
as potential targets for novel therapeu­
tics. A comprehensive integrated genomic
profile of prostate cancer found alterations
in key molecular pathways in a surprisingly
large number of primary and metastatic
tumours, and the degree of copy-number
alterations was shown to be associated with
prognosis, independent of Gleason grade.9
Perhaps the single greatest advance in
prostate cancer biology during the past
decade was the finding by the Sawyers
group10 that upregulation of the androgen
receptor was both necessary and sufficient
for the castration-resistant phenotype. This
discovery drove the development of novel

antiandrogens (for example abiraterone
and enzalutamide), which have markedly
improved the prognosis of patients with
advanced prostate cancer some 75 years
after Huggins first identified the effects of
androgen deprivation on prostate cancer,
and after numerous studies of various regi­
mens of ADT failed to improve survival in
these patients.
In the past decade, the FDA has approved
six new therapeutic agents that prolong
overall survival in men with castration-­
resistant prostate cancer, including drugs
that target androgen-dependent and
androgen-­i ndependent pathways, bone
metasta­s es, and cell cycle regulators.
Evaluation of these therapies is occurring
in multi­modality clinical trials, raising the
hope that we might one day develop thera­
peutic programmes that can cure many
more men with advanced prostate cancer.
A DECADE IN MEDICINE NOVEMBER 2015  |  88
Memorial Sloan Kettering Cancer Center,
1275 York Avenue, New York, NY 10065, USA
(B.E., P.T.S.).
Correspondence to: P.T.S.
scardinp@mskcc.org
Competing interests
P.T.S. is a paid scientific advisor to OPKO
Diagnostics, which has licensed the 4KScore®from
P.T.S. and colleagues for commercial development.
B.E. declares no competing interests.
1. Andriole, G. L. et al. Mortality results from
a randomized prostate-cancer screening
trial. N. Engl. J. Med. 360, 1310–1319
(2009).
2. Schröder, F. H. et al. Screening and prostate
cancer mortality: results of the European
Randomised Study of Screening for Prostate
Cancer (ERSPC) at 13 years of follow-up.
Lancet http://dx.doi.org/10.1016/
S0140-6736(14)60525-0.
3. Hugosson, J. et al. Mortality results from the
Göteborg randomized population-based
prostate-cancer screening trial. Lancet Oncol.
11, 725–732 (2010).
© 2015 Macmillan Publishers Limited. All rights reserved
UROLOGY
4. Vickers, A. J. et al. Strategy for detection of
prostate cancer based on relation between
prostate specific antigen at age 40–55 and
long term risk of metastasis: case-control
study. BMJ 2013, 346–357 (2013).
5. Sanda, M. G. et al. Quality of life and
satisfaction with outcome among prostate-
cancer survivors. N. Engl. J. Med. 358,
1250–1261 (2008).
6. Gandaglia, G. et al. Comparative effectiveness
of robot-assisted and open radical
prostatectomy in the postdissemination era.
J. Clin. Oncol. 32, 1419–1426 (2014).
7. Vickers, A. et al. Cancer control and functional
outcomes after radical prostatectomy as
markers of surgical quality: analysis of
heterogeneity between surgeons at a single
cancer center. Eur. Urol. 59, 317–322 (2011).
8. Tomlins, S. A. et al. Recurrent fusion of
TMPRSS2 and ETS transcription factor genes in
prostate cancer. Science 310, 644–648 (2005).
9. Taylor, B. S. et al. Integrative genomic profiling
of human prostate cancer. Cancer Cell 18,
11–22 (2010).
10. Chen, C. D. et al. Molecular determinants of
resistance to antiandrogen therapy. Nat. Med.
10, 33–39 (2004).
Connect to...

Follow us on Twitter
For the latest commentary and urology news

Nature Reviews on Facebook

For regular updates on new content across all the Nature Reviews journals

Sign up for e-alerts

Sign up for free e-alerts and be notified when new articles are published in Nature Reviews Urology.
Research Highlights, News & Views, Reviews, and Perspectives—you won’t miss a thing!

Advance online publication

The latest content published online ahead of print from Nature Reviews Urology

Conferences
You can meet the editors of Nature Reviews Urology at some of the meetings listed on our conferences page

Happy birthday Nature Reviews Urology!

November 2014 marked the 10th anniversary of the launch of Nature Reviews Urology.
To celebrate this milestone, we present an anniversary issue containing specially commissioned
articles that highlight progress in the field of urology over the past decade. In addition,
our anniversary infographic provides a snapshot of 10 years of the journal in numbers.
nature.com/nrurol10

© 2015 Macmillan Publishers Limited. All rights reserved

To celebrate the 10th anniversaries of our
clinical review titles, we present infographics and
top cited articles charting 10 years of publication.

Follow the links to take a look.

nature.com/nrcardio10 nature.com/nrendo10

nature.com/nrco10 nature.com/nrneph10

nature.com/nrgastro10 nature.com/nrneurol10

nature.com/nrurol10 nature.com/nrrheum10