CASE REPORTS
Retroperitoneal Fibrosis
Secondary to Actinomycosis
With No Intrauterine Device
Michael R. Milam, MD,
Stephen J. Schultenover, MD,
Marta Crispens, MD, and Lynn Parker, MD
Departments of Obstetrics and Gynecology and Pathology, Vanderbilt University
Medical Center, Nashville, Tennessee
BACKGROUND: Actinomycotic pelvic infection usually oc-
curs in the presence of an intrauterine device. It can result
in pelvic inflammatory disease, tubo-ovarian abscess, and
retroperitoneal fibrosis.
CASE: A 35-year-old multipara who had never used an
intrauterine device presented with a 5-month history of
progressively worsening, colicky, right-sided abdominal
pain, dysuria, weight loss, and constipation. She was found
to have retroperitoneal fibrosis. The diagnosis of actinomy-
cotic pelvic infection was made at laparotomy.
CONCLUSION: Actinomycosis may be considered in the dif-
ferential diagnosis of women with retroperitoneal fibrosis,
even when there is no history of an intrauterine device.
(Obstet Gynecol 2004;104:1134 – 6. © 2004 by The Amer-
ican College of Obstetricians and Gynecologists.)
Actinomycotic pelvic infection usually occurs in associa-
tion with the use of an intrauterine device (IUD). It can
result in pelvic inflammatory disease, tubo-ovarian ab-
scess, and retroperitoneal fibrosis.1 We present a case in
which actinomycotic pelvic infection was diagnosed in a
patient without a history of IUD use.
CASE
A 35-year-old, G2P2002 woman with no prior history of
IUD use presented with a 5-month history of progres-
sively worsening, colicky, right-sided abdominal and
flank pain, dysuria, weight loss, and constipation. Her
past medical history was unremarkable. The patient had
normal Pap tests and denied any history of sexually
transmitted diseases.
Approximately 4 months after delivery, the patient
had a single episode of right pelvic and flank pain for
which she was treated for a presumed pelvic infection.
Address reprint requests to: Lynn P. Parker, MD, 7506 Clipping Cross
Road, Louisville, KY 40241; e-mail: lynn.parker@vanderbilt.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1134 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
The patient then underwent laparoscopic bilateral Fa-
lope ring tubal sterilization 7 months postpartum; find-
ings at the laparoscopy were normal. However, 10 days
later, the patient developed recurrence of pain. Of note,
the patient also admitted to a 16-lb weight loss over the
preceding 3 months. On examination, her uterus was
fixed in the pelvis. A computed tomographic (CT) scan
of the abdomen and pelvis revealed right hydronephro-
sis and changes in the right retroperitoneum consistent
with retroperitoneal fibrosis. The patient was referred to
a gynecologic oncologist and a urologist (Table 1).
An examination under anesthesia, right ureteral stent
placement, and true-cut biopsies of the bilateral parame-
tria were performed. The cervix was normal. The uterus
was fixed with smooth, thickened cardinal and uterosa-
cral ligaments. The fibrous tissue was thought to origi-
nate from the pelvic sidewalls. Pathology revealed dense
fibrosis with a mixed inflammatory cell infiltrate. There
was no evidence of malignancy. The patient was diag-
nosed with idiopathic retroperitoneal fibrosis and treated
with methylprednisolone, tapered 24 mg to 0 mg, by
mouth over 7 days. She initially improved. However, a
month later, she presented to the Emergency Depart-
ment with increasing abdominal pain, nausea, constipa-
tion, and 14-lb weight loss. The patient was febrile and
had a white blood cell count of 20,000. She was admitted
to the hospital and treated with ampicillin, gentamicin,
and metronidazole. A pelvic ultrasonography and CT
scan of the abdomen and pelvis revealed a new pelvic
mass consistent with an abscess (Fig. 1). When she had
persistent fever, worsening pain, and abdominal radio-
graphs consistent with small bowel obstruction, the de-
cision was made to proceed with surgery.
The patient underwent exploratory laparotomy, right
salpingo-oophorectomy, left salpingectomy, appendec-
tomy, cystoscopy, and bilateral ureteral stent placement.
Findings at laparotomy included a 10-cm right tubo-
ovarian abscess adherent to the cecal and sigmoid mes-
enteries, the colon, the posterior aspect of the uterus, and
the bladder. The left fallopian tube was normal, with a
Falope ring in place. The left ovary and upper abdomen
were normal. The appendix was involved with, but did
not appear to be the source of, the inflammatory process.
The small intestine was dilated above a site of narrowing
in the posterior cul-de-sac. This resolved with lysis of
adhesions.
Postoperatively, the fever defervesced and her pain
improved on antibiotics. The left ureteral stent was
removed on postoperative day one. Culture of the pelvic
abscess grew Bacteroides fragilis. The pathology report
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000144118.90772.a1
Table 1. Timeline of Symptom Presentation and Treatment
Time: 0–19 mo Patient complaint or action
0 Vaginal delivery
4 mo Patient requested information about tubal ligation
4.5–5 mo Patient complained of pain
7 mo Laparoscopy for tubal ligation (Falope ring)
7 mo, 10 d Patient complained of pain
9 mo Diagnosed with fibrosis; exam with true-cut
biopsies and stent
10 mo Laparotomy; abscess discovered; treated with
long-term antibiotics
16 mo Computed tomographic scan; resolution of
hydronephrosis
19 mo Stent removal
described a right ovarian abscess with acute inflamma-
tion and colonies of filamentous organisms consistent
with Actinomyces sp (Fig. 2).
An infectious disease consultation recommended that
the patient be treated with intravenous penicillin V, 2.5
million units every 6 hours for 6 weeks, followed by oral
penicillin V, 1 g twice daily, to complete a year of
treatment. For the Bacteroides, she received metronida-
zole, 500 mg orally 3 times daily, to complete a 2-week
course.
At a follow-up visit 2 months later, the patient had
gained 10 pounds and had lessening of her pain, as well
as decreased parametrial thickening on examination.
Computed tomographic scans done 7 months postoper-
atively revealed improvement of the bilateral hydrone-
phrosis. The right ureteral stent was removed 10 months
after surgery.
Fig. 1. Computed tomography scan demonstrating complex
fluid and soft tissue mass involving the right adnexa with
foci of air consistent with abscess. Abscess is contained
within the arrows.
Milam. Pelvic Actinomycosis. Obstet Gynecol 2004.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Fig. 2. Hematoxylin-eosin stained section of ovarian ab-
scess with classic sulfur granules surrounded by eosino-
philic material (! 200, original magnification).
Milam. Pelvic Actinomycosis. Obstet Gynecol 2004.
COMMENT
Actinomycosis can be diagnosed by identification of the
classic sulfur granule on light microscopy. Diagnosis can
also be made by identification of the organism in anaer-
obic culture.1 Blood or tissue cultures from individuals
with actinomycosis will often grow other pathogens.
However, an 86% success rate in culturing Actinomyces has
been reported when cultures are pretreated with metro-
nidazole to inhibit the growth of other, faster growing,
anaerobes.2
Treatment with penicillin for 6 –12 months is usually
recommended for actinomycotic infections. Metronida-
zole is commonly added for treatment of associated
anaerobic infection.3 Atad et al4 recommend a shorter
course of penicillin for pelvic actinomycosis following
complete surgical drainage of a pelvic abscess. In their
study, 5 patients with actinomycotic pelvic infections
who had complete surgical drainage were treated with
penicillin for 6 months or less. After at least 2 years of
follow-up, all 5 patients were without evidence of recur-
rent infection.
Actinomyces sp are normal inhabitants of the human
gastrointestinal tract, being found in both the orophar-
ynx and small bowel. Ascending actinomycotic infection
is thought to originate from the perineum or by oro- and
anogenital contact.3 The most common presenting
symptoms are abdominal pain, weight loss, and foul-
smelling vaginal discharge. The diagnosis of pelvic acti-
nomycosis is difficult. According to one review, only
17% of actinomycotic pelvic infections are diagnosed
preoperatively, despite the use of Pap test, percutaneous
biopsy, and cultures.3
Milam et al Pelvic Actinomycosis 1135
 Actinomycotic pelvic infections are commonly associ-
ated with the use of an IUD. Our patient had no history
of current or prior IUD use. Retroperitoneal fibrosis
secondary to Actinomyces sp without a history of an intra-
uterine device has been reported only rarely.5 Hibbard et
al6 reported 10 cases of pelvic and retroperitoneal infec-
tions following paracervical or pudendal blocks. A po-
tential etiology for this patient’s infection is pudendal
anesthesia administered at the time of her vaginal deliv-
ery. Unfortunately, this patient’s delivery records are
unavailable.
In addition to actinomycosis, the differential diagnosis
of retroperitoneal fibrosis includes malignancy, inflam-
matory abdominal aortic aneurysm, giant cell arteritis,
prior surgery, endometriosis, inflammatory bowel dis-
ease, and other infections. The majority of cases of
retroperitoneal fibrosis are idiopathic. Treatment op-
tions for idiopathic retroperitoneal fibrosis include glu-
cocorticoids, immunosuppressive agents, tamoxifen,
and surgery.7 Response rates for these treatments vary.
The differential diagnosis of a woman with retroperi-
toneal fibrosis should include actinomycotic pelvic infec-
tion, even in a patient with no history of IUD use. Our
patient’s course illustrates the difficulty in diagnosing
this condition.
Massive Fetal Ascites Causing
Increased Middle Cerebral
Artery Systolic Velocity
Gloria Chiang, Deborah Levine, MD,
Philip Hess, MD, and Kee-Hak Lim, MD
Harvard Medical School, Boston, Massachusetts; and Departments of Radiology,
Obstetrics and Gynecology, and Anesthesiology, Beth Israel Deaconess Medical
Center, Boston, Massachusetts
BACKGROUND: An elevated peak systolic velocity in the
middle cerebral artery, assessed by Doppler ultrasonogra-
phy, is commonly associated with fetal anemia. Other fetal
abnormalities associated with a high middle cerebral ar-
tery velocity have rarely been reported.
CASE: A fetus with increasing ascites was found to have an
elevated middle cerebral artery peak systolic velocity. Fol-
lowing paracentesis, the peak systolic velocity normalized.
Peak systolic velocity continued to correlate with the level
of ascites, falling to normal ranges when large-volume
Address reprint requests to: Deborah Levine, MD, Department of
Radiology, Beth Israel Deaconess Medical Center, 330 Brookline
Avenue, Boston, MA 02215; e-mail: dlevine@caregroup.harvard.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1136 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
REFERENCES
1. Binford CH, Conner DH, editors. Pathology of tropical and
extraordinary diseases. Washington, DC: The Armed
Forces Institute of Pathology; 1976.
2. Traynor RM, Parratt D, Duguid Helen LD, Duncan ID.
Isolation of actinomycetes from cervical specimens. J Clin
Pathol 1981;34:914 –16.
3. Fiorino AS. Intrauterine contraceptive device-associated
actinomycotic abscess and Actinomyces detection on cervi-
cal smear. Obstet Gynecol 1996;87:142–9.
4. Atad J, Hallak M, Sharon A, Kitzes R, Kelner Y,
Abramovici H. Pelvic actinomycosis. Is long-term antibiotic
therapy necessary? J Reprod Med 1999;44:939 – 44.
5. Willscher MK, Mozden PJ, Olsson CA. Retroperitoneal
fibrosis with ureteral obstruction secondary to Actinomyces
israeli. Urology 1978;12:569 –71.
6. Hibbard LT, Snyder EN, McVann RM. Subgluteal and
retropsoal infection in obstetrics practice. Obstet Gynecol
1972;39:137–50.
7. Katz R, Golijanin D, Pode D, Shapiro A. Primary and
postoperative retroperitoneal fibrosis: experience with 18
cases. Urology 2002;60:780 –3.
Received May 1, 2004. Received in revised form July 2, 2004. Accepted
July 22, 2004.
amniocentesis and paracentesis were performed. At birth,
the infant was found to have a normal hematocrit.
CONCLUSION: An elevated middle cerebral artery peak
systolic velocity may result from massive fetal ascites with-
out anemia. We hypothesize that the massive ascites led to
increased afterload of the heart, with relatively preserved
preload, leading to an increased systolic blood pressure and
an elevated middle cerebral artery peak systolic velocity.
(Obstet Gynecol 2004;104:1136 – 40. © 2004 by The
American College of Obstetricians and Gynecologists.)
The diagnostic value of Doppler ultrasonography of the
middle cerebral artery in evaluating nonhydropic fetuses
at high risk of anemia has been well reported. Several
studies have demonstrated that an elevated peak systolic
velocity in the middle cerebral artery accurately predicts
the low fetal hemoglobin level that results from maternal
red blood cell alloimmunization,1– 4 parvovirus B19 in-
fection,3 and !-thalassemia.5
One recent study has suggested that an elevated time-
averaged mean velocity in the middle cerebral artery is
associated with low fetal hemoglobin levels in fetuses
with hydrops, with ascites being the defining feature.6
Although this particular study did not find ascites to be a
complicating factor in the middle cerebral artery exami-
nation, minimal literature has examined the impact that
other fetal anomalies may have on the observed velocity
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000121830.94197.70
in the middle cerebral artery. We performed an OVID
MEDLINE search, from 1980 through October 2003,
for the terms “middle cerebral artery,” “Doppler,” “ul-
trasound,” and “fetus,” and found no reported cases of
anomalies other than anemia causing elevated middle
cerebral artery Doppler. Since that time, there has been a
report7 of 2 fetuses with hydrops without anemia associ-
ated with elevated middle cerebral artery Doppler. In
that report, both fetuses died soon after the study was
performed. We report a case in which massive fetal
ascites, presumably causing increased intraabdominal
pressure, dramatically raised the peak systolic velocity in
the middle cerebral artery. The velocity then returned to
the normal range once large-volume amniocentesis and
paracentesis were performed. The hematocrit level was
normal at birth. This suggests that middle cerebral artery
Doppler may not be reliable as an indicator of anemia in
fetuses with concomitant conditions.
CASE
A 33-year-old woman (gravida 3, para 2) presented at 21
weeks of gestation, as determined by last menstrual period,
with a diagnosis of fetal ascites by ultrasonography. Fetal
biometry revealed biparietal diameter and head circumfer-
ence corresponding to 24 weeks, femur length correspond-
ing to 21 weeks, and abdominal circumference correspond-
ing to 27 weeks. Moderate ascites caused enlargement of
the abdomen. There was normal amniotic fluid volume,
with an amniotic fluid index of 16 cm, and no other signs of
hydrops. Amniocentesis revealed normal 46XY karyotype.
Fetal echocardiography showed mild cardiomegaly but,
otherwise, was normal.
At 24 weeks of gestation, fetal biometric data contin-
ued to show head size 2 weeks ahead of gestational age,
with abdominal circumference corresponding to 36
weeks. The ascites had increased dramatically (Fig. 1).
Hydramnios with an amniotic fluid index of 26 and
placentomegaly with placental thickness of 5.6 cm were
also seen. Magnetic resonance imaging (MRI) of the
abdomen was conducted at the time and was remarkable
for isolated fetal ascites with elevated hemidiaphragms
and compression of the fetal lungs.
Because of concern about lung development, a plan
was made to perform large-volume amniocentesis and
paracentesis. Because of concern about atypical presen-
tation of anemia, it was also planned to perform percu-
taneous umbilical cord blood sampling after the afore-
mentioned procedures. Before the amniocentesis,
Doppler examination of the middle cerebral artery re-
vealed an elevated peak systolic velocity of 56 cm/s,
which is in the range suggestive of severe anemia (Fig.
2).1 A total of 1,600 mL of amniotic fluid was removed.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Fig. 1. Transverse view of the fetus at 24 weeks of gestation
shows a large amount of ascites distending the fetal
abdomen.
Chiang. Massive Ascites and Middle Cerebral Artery Doppler. Obstet Gynecol
2004.
Fetal paracentesis was performed with removal of 650
mL of serosanguinous fluid. After these procedures, the
middle cerebral artery peak systolic velocity fell to 30.0
cm/s, within the normal range.1 Because the middle
cerebral artery peak systolic velocity normalized, there
were no other signs of hydrops, and the fetal position
made approaching the umbilical cord insertion site diffi-
cult, percutaneous umbilical cord blood sampling was
not performed. Examination of the ascitic fluid showed a
white blood cell count (WBC) of 1,250/mL, red blood
cell count (RBC) of 1,650/mL, total protein of 2.4 g/dL,
glucose of 85 mg/dL, lactate dehydrogenase (LDH) of
107 IU/L, and amylase of 2 IU/L. Low protein and low
LDH with a high white cell count suggested that this was
a transudate with lymphatic etiology.
A week later, at 25 weeks of gestation, it was noted
that there was a reaccumulation of the massive ascites
and hydramnios, with an amniotic fluid index of 31. The
middle cerebral artery peak systolic velocity was 48
cm/s. Paracentesis was repeated (400 mL removed), and
chemistries showed similar results. At 28 weeks of gesta-
tion, a third large-volume amniocentesis, with the re-
moval of 1,500 mL over 20 minutes, and paracentesis
(650 mL) were performed without complication. The
middle cerebral artery peak systolic velocity was 44
cm/s, both before and 1 minute after the amniocentesis,
and decreased to 29 cm/s after the ascites was tapped.
By 32 weeks of gestation, the fetus again had severe
ascites and moderate-to-severe hydramnios. Fetal biom-
Chiang et al Massive Ascites and Middle Cerebral Artery Doppler 1137

etry showed a head circumference 4 weeks greater than
that appropriate for gestational age and an abdominal
circumference greater than that at 42 weeks. The ascites
also elevated the diaphragm and made the chest seem
relatively small. There was an umbilical vein varix mea-
suring 9 mm. The middle cerebral artery peak systolic
velocity was 79 cm/s. No additional paracentesis or
therapeutic amniocentesis was performed before birth,
because the ascites had rapidly reaccumulated after the 2
prior procedures, and the short-term benefit of these
palliative measures was not considered worth the poten-
tial risk. At 37 weeks of gestation, the baby was born via
cesarean delivery with a markedly large abdomen. The
hematocrit was 43% (normal) and remained normal
throughout his hospitalization, so no blood products
were given. He was immediately intubated for respira-
tory distress. On the first day of life, he underwent 2
more paracenteses of 500 mL of brownish fluid. Analysis
of the ascites fluid showed WBC of 3300/mL, RBC of
1520/mL, total protein of 3.3 g/dL, glucose of 53 mg/dL,
LDH of 350 IU/L, and amylase of 1 IU/L. Gram stain
and culture results were negative. The following day,
abdominal ultrasonography revealed normal abdominal
organs, normal Doppler hepatic flow, and ascites. Liver
function tests were remarkable for an aspartate amino-
transferase of 87 U/L (normal 0 – 40 U/L) and "-
glutamyltranspeptidase of 139 U/L (normal 11– 63 U/L).
The infant remained intubated until the fourth day
after delivery. Analysis of a peritoneal tap of 100 mL of
1138 Chiang et al Massive Ascites and Middle Cerebral Artery Doppler
Fig. 2. Middle cerebral artery
Doppler at 24 weeks of gestation
shows an elevated peak systolic
velocity before the amniocentesis
and paracentesis were performed.
Chiang. Massive Ascites and Middle Cerebral
Artery Doppler. Obstet Gynecol 2004.
fluid revealed a triglyceride count of 309 mg/dL, LDH of
181 IU/L, total protein of 3.5 g/dL, albumin of 2.2 g/dL,
WBC of 11,534/mL, and RBC of 1,078/mL, with a
differential of 1 neutrophil and 90% lymphocytes. A
presumptive diagnosis was made of chylous ascites. An
MRI of the chest and abdomen revealed no pleural or
pericardial effusion. An MRI/magnetic resonance an-
giography of his head was performed because of a hem-
angioma on his right cheek and was found to be normal.
No syndrome was consistent with the findings. He was
started on total parenteral nutrition without oral feeds to
prevent further accumulation of ascites.
At 3 months of age, he underwent an exploratory
laparotomy to localize the chylous leakage site. Contrast
studies during surgery documented a lymphatic leak,
and a possible site was ablated. However, the source of
leakage was not definitively found.
As an outpatient, his ascites remained stable.
COMMENT
Chylous ascites is a rare cause of ascites that results from
congenital abnormalities, lymphatic obstruction, or trau-
matic injury of the intraabdominal portion of the tho-
racic duct. Lymphatic leakage leads to abdominal disten-
tion and, in severe cases, scrotal edema, inguinal and
umbilical herniation, and respiratory problems in neo-
nates.8 It is diagnosed by paracentesis of milky fluid after
OBSTETRICS & GYNECOLOGY
a fat-containing feeding that contains high protein levels,
elevated triglycerides, and lymphocytosis.8
The use of middle cerebral artery blood flow velocity
has been studied as a noninvasive diagnostic tool for
evaluating fetal anemia to reduce the risks associated
with amniocentesis and cordocentesis. Vyas et al9 first
described the association between an elevated time-aver-
aged mean blood velocity in the middle cerebral artery
and low levels of hemoglobin in nonhydropic fetuses.
Because the brain is particularly sensitive to hypoxia, it
has been hypothesized that the body compensates by
increasing cardiac output and decreasing blood viscosity
to maintain cerebral perfusion.
The largest study in the literature analyzing the asso-
ciation between fetal anemia and high blood velocity
(using elevated peak systolic velocity rather than time-
averaged mean blood velocity) was conducted by Mari
et al2 Based on data from 35 fetuses with moderate-to-
severe anemia, on which cordocentesis would normally
be performed, the sensitivity of middle cerebral artery
peak systolic velocity was 100%, with a false-positive rate
of 12%.
A study of 17 hydropic fetuses with ascites supported
the association between a high velocity in the middle
cerebral artery and low fetal hemoglobin levels.6 This
study found that the sensitivity of elevated middle cere-
bral artery time-averaged velocity in predicting fetal
anemia was 91%, while the specificity was 100%. Be-
cause there were no false positives, this study suggests
that the level of ascites seen in the fetuses did not distort
the middle cerebral artery findings in hydropic fetuses
without anemia.
Four factors in the literature have been noted to affect
the observed middle cerebral artery peak systolic veloc-
ity. The first is in utero transfusion of the fetus, which
normalizes the hemoglobin level and thus normalizes the
middle cerebral artery peak systolic velocity.3 The sec-
ond, described by Vyas et al,10 is an increased maternal
abdominal pressure that increases fetal intracranial pres-
sure and thus reduces fetal cerebral perfusion through
the middle cerebral artery. The third is the observation
that small-for-gestational age fetuses exhibit a higher
blood velocity in the middle cerebral artery than appro-
priate-for-gestational age fetuses.11 The fourth factor,
described by Zimmerman et al,12 is that the 7 fetuses in
their study that developed an elevated middle cerebral
artery peak systolic velocity after 35 weeks of gestation
were not found to be anemic at birth, which suggests that
middle cerebral artery peak systolic velocity may not be
reliable after 33 weeks of gestation.
The elevated middle cerebral artery peak systolic ve-
locity seen in our case is of interest because the fetus had
a normal hematocrit and no evidence of anemia, and the
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
level of the observed middle cerebral artery peak systolic
velocity appeared to correlate with the level of fetal ascites.
We hypothesize that the massive ascites led to increased
afterload of the heart, with relatively preserved preload
because most of the venous return was from the umbilical
vein, leading to an increased systolic blood pressure and
thus an elevated middle cerebral artery peak systolic veloc-
ity. Because large-volume amniocentesis and paracentesis
were performed sequentially without allowing time for
equilibration, we cannot be entirely confident of the role
that decreased amniotic fluid played in decreasing the mid-
dle cerebral artery peak systolic velocity.
In summary, fetal structural anomalies can cause ele-
vated peak systolic velocity.
REFERENCES
1. Mari G, Adrignolo A, Abuhamad A, Pirhonen J, Jones D,
Ludomirsky A, et al. Diagnosis of fetal anemia with Dopp-
ler ultrasound in the pregnancy complicated by maternal
blood group immunization. Ultrasound Obstet Gynecol
1995;5:400 –5.
2. Mari G, Deter R, Carpenter R, Rahman F, Zimmerman R,
Moise K, et al. Noninvasive diagnosis by Doppler ultra-
sonography of fetal anemia due to maternal red-cell allo-
immunization. N Engl J Med 2000;342:9 –14.
3. Delle Chiaie L, Buck G, Grab D, Terinde R. Prediction of
fetal anemia with Doppler measurement of the middle
cerebral artery peak systolic velocity in pregnancies com-
plicated by maternal blood group alloimmunization or
parvovirus B19 infection. Ultrasound Obstet Gynecol
2001;18:232– 6.
4. Dukler D, Oepkes D, Seaward G, Windrim R, Ryan G.
Noninvasive tests to predict fetal anemia: a study compar-
ing Doppler and ultrasound parameters. Am J Obstet
Gynecol 2003;188:1310 – 4.
5. Lam Y, Tang M. Middle cerebral artery Doppler study in
fetuses with homozygous !-thalassaemia-1 at 12–13 weeks
of gestation. Prenat Diagn 2002;22:56 – 8.
6. Abdel-Fattah S, Soothill P, Carroll S, Kyle P. Noninvasive
diagnosis of anemia in hydrops fetalis with the use of
middle cerebral artery Doppler velocity. Am J Obstet
Gynecol 2001;185:1411–5.
7. Shah NK, Martin WL, Whittle MJ. Middle cerebral artery
Doppler velocimetric assessment in two cases of hydrops
fetalis with fetal anaemia. Prenat Diagn 2004;24:17– 8.
8. Hyams JS. Malformations. In: Behrman R, Kliegman R,
Jenson H, editors. Nelson textbook of pediatrics. 16th ed.
Philadelphia (PA): W. B. Saunders; 2000. p. 1229.
9. Vyas S, Nicolaides H, Campbell S. Doppler Examination
of the middle cerebral artery in anemic fetuses. Am J
Obstet Gynecol 1990;162:1066 – 8.
10. Vyas S, Campbell S, Bower S, Nicolaides K. Maternal
abdominal pressure alters fetal cerebral blood flow. Br J
Obstet Gynaecol 1990;97:740 –2.
Chiang et al Massive Ascites and Middle Cerebral Artery Doppler 1139
11. Vyas S, Nicolaides K, Bower S, Campbell S. Middle cere-
bral artery flow velocity waveforms in fetal hypoxaemia.
Br J Obstet Gynaecol 1990;97:797– 803.
12. Zimmerman R, Carpenter RJ Jr, Durig P, Mari G. Longi-
tudinal measurement of peak systolic velocity in the fetal
middle cerebral artery for monitoring pregnancies compli-
Pregnancy and Delivery After
Right Common Carotid Artery
Endarterectomy
Filiberto M. Severi, MD, Erika Ignacchiti, MD,
Francesco Setacci, MD,
Giancarlo Palasciano, MD, Carlo Setacci, MD,
and Felice Petraglia, MD
Obstetrics and Gynecology, Department of Pediatrics, Obstetrics and Reproductive
Medicine, and Vascular Surgery, Department of General Surgery, University of
Siena, Siena, Italy
BACKGROUND: Carotid artery atherosclerosis and essential
hypercholesterolemia can add a predisposing risk factor
for coagulation in pregnancy. Careful management of an-
ticoagulation during labor, delivery, and puerperium is
called for in such a case.
CASE: A 41-year-old woman, gravida 2, para 1, with a
previous endarterectomy at the right common carotid ar-
tery because of atherosclerotic plaques, underwent antico-
agulation studies and prophylactic antithrombotic ther-
apy. Low-molecular-weight heparin was administrated
during pregnancy and puerperium. She successfully deliv-
ered by cesarean at 36 weeks of gestation.
CONCLUSION: Low-molecular-weight heparin treatment is
an effective and safe therapy in pregnancy. The healthy
course of therapy, delivery, and puerperium reported here
is a reference that may support women with a similar history.
(Obstet Gynecol 2004;104:1140 –2. © 2004 by The Ameri-
can College of Obstetricians and Gynecologists.)
Vascular complications during pregnancy are a major
cause of maternal and fetal morbidity. Pregnancy is
considered a hypercoagulable state affecting both coag-
ulation and fibrinolytic systems, and any exacerbation of
the predisposing factors for coagulation may lead to a
thrombotic event more often in pregnant women than in
the general population.1 The incidence of ischemic
stroke in women of child-bearing age is estimated be-
tween 3.5 and 18 per 100,000 per year in Western
Address reprint requests to: Felice Petraglia, MD, Chair of Obstetrics
and Gynecology, Department of Pediatrics, Obstetrics and Reproduc-
tive Medicine, University of Siena, Policlinico “Le Scotte” Viale Bracci,
53100 Siena, Italy; e-mail: petraglia@unisi.it.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1140 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
cated by red cell alloimmunisation: a prospective multicenter
trial with intention-to-treat. BJOG 2002;109:746–52.
Received July 15, 2003. Received in revised form September 12, 2003.
Accepted September 26, 2003.
countries.2,3 We describe the course of pregnancy and
delivery in a woman who became pregnant 5 years after
an endarterectomy in the right common carotid artery
because of the presence of atherosclerotic plaques.
CASE
A 41-year-old woman, gravida 2, para 1, at the eighth
week of gestation presented with the diagnosis of preg-
nancy under treatment with indobufen (200 mg/d) as
anticoagulation therapy. She was referred to our depart-
ment for anticoagulation studies, including antiphospho-
lipid, hyperhomocystinemia, and gene polymorphism
assessment. No evidence of acquired or genetic throm-
bophilias was found. The indobufen administration was
stopped and low-molecular-weight heparin introduced
(Clexane; Gruppo Lepetit, Milan, Italy, 0.4 mL/d), as a
safer method of anticoagulation therapy in pregnancy.4
When she was aged 36 years, the patient underwent
duplex scan investigation and right carotid angiogram,
resulting in a diagnosis of severe obstruction of the
common carotid artery extending from the origin to the
bifurcation on the right side, as a result of atherosclerotic
plaques in the vessel and producing a 70% stenosis of the
lumen. The investigation was prompted by her reports
of frequent headaches and worsening episodes of ver-
tigo, without loss of consciousness, for a period of 3– 4
months. She was nondiabetic, her body mass index was
21 kg/m2, she was not hypertensive, and she had no
history of smoking or alcohol consumption. Her past
obstetric history consisted of 1 spontaneous vaginal de-
livery at term at age 32 years , and a miscarriage in the
first trimester of pregnancy. Postload blood glucose lev-
els and postload insulin levels were normal. Total cho-
lesterol levels were markedly increased, with lower high-
density lipoprotein levels and higher total/high-density
lipoprotein cholesterol ratios; therefore, essential hyper-
cholesterolemia was diagnosed. With the aim of prevent-
ing a future stroke and in light of the young age of the
patient, corrective surgery was indicated, and an endar-
terectomy was performed, with application of an intralu-
minal shunt and coverage of the arterectomy with a
Dacron (E. I. du Pont de Nemours and Company,
Wilmington, DE) patch. The lumen of the carotid bifur-
cation was successfully restored, and extended anticoag-
ulation therapy and a hypolipidic diet were prescribed.
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000125552.00171.9a
 We saw her during the present gestation at 15 weeks of
pregnancy. Complete blood count, serum urea and creati-
nine, electrolytes, total cholesterol, low-density lipoprotein
cholesterol, and triglyceride levels were within the normal
range, whereas the coagulation screen showed increased
levels of D-dimer (706 ng/mL, normal range less than 500
ng/mL) and serum fibrinogen (819 mg/dL, normal range
150–350 mg/dL), generally considered predictors of pro-
gression of carotid atheromatous lesions.5
During an ultrasound examination a viable pregnancy
with a singleton fetus was detected, with an anterior
normally inserted placenta. Pelvic examination demon-
strated mild bleeding from a closed cervix. A duplex
ultrasound scan of carotid arteries revealed the presence
of irregular plaques in the right common carotid artery
(stenosis less than 50%) and focal intimal and fibromus-
cular thickening of 1.8 cm extending from the bifurcation
to the origin of the internal carotid, without flow accel-
eration, and normal vertebral artery patency. She was
discharged 5 days later.
At 21 weeks of gestation the patient was admitted to
the hospital for symptomatic cramping and mild vaginal
bleeding. The transvaginal ultrasound examination
showed a cervical length of 35 mm and the absence of
funneling. Laboratory values were all normal except for
elevated fibrinogen and D-dimer levels. A new duplex
ultrasound scan of the carotid arteries confirmed the
previous findings.
At 25 weeks of pregnancy she once more experienced
painful and regular uterine contractions and was success-
fully treated with atosiban (Tractocile; Ferring SA,
Malmö, Sweden, 0.9 mL intravenous loading dose plus
20-mL infusion/24 hours), an oxytocin antagonist, con-
sidered the choice of treatments in this case for the
limited cardiovascular adverse effects compared with the
other tocolytic agents. After 5 days the patient was
discharged with the recommendation of bed rest at home
and continuation of low-molecular-weight heparin. Ac-
cording to the cardiologist team, she was followed up
with monitoring of blood and coagulation parameters
every 4 weeks, and fibrinogen and D-dimer levels re-
mained consistently above the normal range, without a
significant increment. In addition, a duplex scan of the
carotid arteries was performed every 8 weeks, confirm-
ing the previous findings. Biometric parameters detected
by ultrasonography showed a regular fetal growth.
At 36 weeks of pregnancy the patient had an elective
cesarean delivery for breech presentation of a healthy
male fetus of 2,580 g, with Apgar scores of 9 and 10 at 1
and 5 minutes, respectively. During the postpartum pe-
riod she remained on long-term anticoagulation therapy,
all the laboratory values were normal, and a duplex scan
of the carotid arteries confirmed the previous flow pa-
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
rameters. She was discharged on the seventh postdeliv-
ery day and followed up uneventfully for a period of 6
weeks.
COMMENT
The present case is remarkable as a pregnancy in a
patient affected by atherosclerotic stenosis of the carotid
arteries and essential hypercholesterolemia who had un-
dergone surgery 5 years before. Pregnancy and delivery
in women who are diagnosed with a carotid artery
atherosclerosis are strongly associated with disease-re-
lated complications such as ischemic attacks6 besides
gestational vascular complications.
To our knowledge, no precise data are available on
the influence of pregnancy on the risk of recurrent
stroke, thereby making it difficult to counsel women with
a history of ischemic stroke regarding future pregnan-
cies. In addition, we are not aware of data on or guide-
lines for the management of labor and delivery in
women with a history of ischemic stroke or any consen-
sus on the risk-benefit ratio of antithrombotic treatment
to prevent recurrent stroke during pregnancy.
The use of low-molecular-weight heparin in the
present case is a novel report. Low-molecular-weight
heparin has been used for more than a decade in the
prophylaxis and treatment of venous thromboembolism
and recently in coronary artery disease; however, con-
troversies still exist regarding anticoagulation therapy
during pregnancy.7
With respect to stroke prophylaxis, aspirin and vita-
min K antagonists are alternatives that have not been
adequately studied in the management of these patients
during pregnancy after arterial surgical procedures.
Low-molecular-weight heparin treatment seems to be an
effective and safe therapy in pregnancy and offers impor-
tant advantages over unfractionated heparin or warfa-
rin.8 Oral anticoagulants, such as warfarin, can transfer
easily to the placenta, causing significant fetal wastage
and an increased incidence of fetal mortality and mor-
bidity. They are also associated with various congenital
malformations caused by teratogenic effects. For all these
reasons they should be avoided throughout pregnancy.9
In addition, heparin-induced thrombocytopenia and os-
teopenia are occasionally seen, making such therapy less
desirable than fractionated heparins.
Low-molecular-weight heparin treatment was contin-
ued for 6 weeks after delivery for our patient because the
postpartum period is associated with an increased relative
risk of stroke recurrence (risk ratio 9.7; 95% confidence
interval 1.2%, 78.9%10). The higher risk ratio of first or
recurrent ischemic stroke during the postpartum period
suggests a causal role for the large decrease in blood volume
Severi et al Atherosclerosis and Pregnancy 1141
or the rapid changes in hormonal status that follow deliv-
ery, perhaps because of hemodynamic, coagulative, or ves-
sel wall changes.10,11 For this reason if prophylactic antico-
agulation therapy is beneficial, it is even more so during the
late pregnancy period and the puerperium.
Thus, the successful course of pregnancy, delivery,
and puerperium in our report may support women with
a similar history who desire a pregnancy. The outcome
of pregnancy in these women seems to be similar to that
expected in the general population. A previous ischemic
stroke is not a contraindication to a subsequent pregnan-
cy.12 Although women with a history of stroke often
may be encouraged by their physicians to avoid preg-
nancy or to undergo an induced abortion solely because
they have had a stroke, such an advice is not appropriate.
In our opinion, the final decision depends on several
factors, including the cause of the initial stroke, the desire
for pregnancy, the residual deficit, and the possibility of
preventing a future episode.
REFERENCES
1. Hague WM, Dekker GA. Risk factors for thrombosis in
pregnancy. Best Pact Res Clin Haematol 2003;16:
197–210.
2. Petitti DB, Sidney S, Quesenberry CP Jr, Bernstein A.
Incidence of stroke and myocardial infarction in women of
reproductive age. Stroke 1997;28:280 –3.
3. Williams GR, Jiang JG, Matchar DB, Samsa GP. Inci-
dence and occurrence of total (first-ever and recurrent)
stroke. Stroke 1999;30:2523– 8.
A Fatal Case of Clostridium
sordellii Septic Shock Syndrome
Associated With Medical Abortion
Ellen Wiebe, MD, Edith Guilbert, MD,
Francis Jacot, MD, Caitlin Shannon, MPH, and
Beverly Winikoff, MD, MPH
Department of Family Practice, University of British Columbia, Vancouver, British
Columbia, Canada; University of Laval, Laval, Quebec, Canada; University of
Sherbrooke, Sherbrooke, Quebec, Canada; and Gynuity Health Projects, New
York, New York
BACKGROUND: Clostridia bacteria are infrequent human
pathogens. In the obstetric and gynecologic literature, Clos-
tridium sordellii infections have been very rarely reported.
Address reprint requests to: Caitlin Shannon, Gynuity Health Projects,
15 East 26th Street, Suite 1609, New York, NY 10010; e-mail:
caittespencer@yahoo.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1142 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
4. Kher A. Low-molecular-weight heparins: weeks or months
instead of days of treatment. Clin Appl Thromb Hemost
2001;7:314 –20.
5. Colas JL, Montalescot G, Tribouilloy C. Fibrinogen. A
cardiovascular risk factor "in French#. Presse Med 2000;
29:1862– 6.
6. Lewis CE, Funkhouser E, Raczynski JM, Sidney S, Bild
DE, Howard BV. Adverse effect of pregnancy on high den-
sity lipoprotein (HDL) cholesterol in young adult women.
The CARDIA Study. Coronary Artery Risk Development
in Young Adults. Am J Epidemiol 1996;144:247–54.
7. Schlman S. Unresolved issues in anticoagulant therapy.
J Thromb Haemost 2003;1:1464 –70.
8. Bowles L, Cohen H. Inherited thrombophilias and antico-
agulation in pregnancy. Best Pract Res Clin Obstet Gynae-
col 2003;17:471– 89.
9. Ginsberg JS, Hirsh J. Use of antithrombotic agents during
pregnancy. Chest 1995;108(suppl):305–11.
10. Leys D, Lamy C, Lucas C, Henon H, Pruvo JP, Codaccioni
X, et al. Arterial ischemic strokes associated with pregnancy
and puerperium. Acta Neurol Belg 1997;97:5–16.
11. Grosset DG, Ebrahim S, Bone I, Warlow C. Stroke in
pregnancy and puerperium: what magnitude of risk?
J Neurol Neurosurg Psychiatry 1995;58:129 –31.
12. Lamy C, Hamon JB, Coste J, Mas JL. Ischemic stroke in
young women. Risk of recurrence during subsequent preg-
nancies. French Study Group on Stroke in Pregnancy.
Neurology 2000;55:269 –74.
Received September 30, 2003. Received in revised form December 16,
2003. Accepted December 30, 2003.
This is a case of infection following medical termination of
early pregnancy with mifepristone and misoprostol.
CASE: A 27-year-old woman presented for termination of
pregnancy at 5.5 weeks from her last menstrual period. She
received mifepristone 200 mg orally followed by 800 !g
vaginal misoprostol. Three days after administration of
misoprostol, she complained of dizziness, pelvic pain, and
bleeding. The next day, she experienced worsening of
symptoms and was hospitalized. She developed pulmo-
nary edema, ascites, and heart failure. Despite supportive
measures, antibiotics, and hysterectomy, she died 3 days
later. The post mortem examinations indicated that death
was caused by shock secondary to C sordellii infection.
CONCLUSION: The frequency of infection following medi-
cal abortion is low. The rapid and fatal course of this
infection is similar to other obstetric and gynecologic cases
reported in the literature. Although providers should re-
main vigilant to the possibility of infection following med-
ical abortion, the overall proven safety of medical abortion
remains the same. (Obstet Gynecol 2004;104:1142– 4.
© 2004 by The American College of Obstetricians and
Gynecologists.)
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000142738.68439.9e
A 27-year-old woman presented August 16, 2001, for ter-
mination of early pregnancy. Her reproductive history
included one vaginal delivery and a surgical abortion with
no complications. Other than childhood asthma, there was
no significant medical history. A transvaginal ultrasound
examination revealed an intrauterine pregnancy consistent
with 5.5 weeks after last menstrual period. On August 20,
she was seen again. The hemoglobin was 13.6 g/dL, the
hematocrit was 40.3%, and the white blood cell (WBC)
count was 6,100/mm3. A Chlamydia trachomatis screen was
negative, and the blood pressure was 95/60 mm Hg. The
patient was given 200 mg mifepristone and 800 #g of
misoprostol to take vaginally 2 days later.
On August 23, she took 200 mg mifepristone orally,
and 2 days later, she inserted 800 #g misoprostol vagi-
nally. She telephoned the clinic on August 27 while at
work complaining of persistent bleeding heavier than
menstruation accompanied by painful cramping. She
had not taken any analgesics. She was advised to take an
additional dose of misoprostol (800 #g, vaginally) and
200 #g of methylergonovine by mouth and to come to
the clinic the following day.
On August 28, she presented at the clinic complaining
of severe pelvic pain and dizziness. She was noted to
have good color, a blood pressure of 90/60 mm Hg, and
a pulse rate of 125 beats per minute (bpm). On bimanual
examination, the uterus was tender but difficult to pal-
pate. A small amount of serous discharge was seen
coming from the cervical os. A transvaginal ultrasound
examination revealed a completed medical abortion. A
complete blood count was drawn, and the patient was
sent home with analgesics.
On August 29, she called the clinic complaining of
nausea and palpitations. The hemoglobin, measured the
previous day, was 17.3 g/dL, the hematocrit was 51.1%,
the platelet count was 281,000/mm3, and the WBC
count was 21,000/mm3 with a left shift. Later that
evening, she presented to hospital with complaints of
abdominal pain, nausea, vomiting, subjective sensations
of fever and chills, decreased frequency of urination,
slight vaginal bleeding, and foul-smelling discharge. Her
temperature was 35.3°C, blood pressure was 90/60 mm
Hg, pulse was 120 bpm, and O2 saturation was 94% on
room air. On examination, there was abdominal re-
bound tenderness with guarding and cervical motion
tenderness. Pertinent laboratory results were as follows:
WBC count 55,100/mm3, hemoglobin 19.9 g/dL, hemat-
ocrit 58.2%, platelet count 202,000/mm3, blood urea
nitrogen 7.1 mmol/L (20 mg/dL), creatinine 90 #mol/L
(1.0 mg/dL), sodium 128 mmol/L, potassium 4.2
mmol/L. Blood gas analysis revealed pH 7.34, PCO2 31.3
mm Hg, HCO3 16.4 mmol/L. The diagnosis was endo-
metritis, “preshock,” and dehydration.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
On August 30, the patient was treated with ampicillin,
gentamicin, and clindamycin, as well as oxygen, analge-
sics, and antiemetics. She developed hypotension with
decreased oxygen saturation. Additionally, urine output
decreased, complete blood count remained abnormal,
and serum levels of sodium, potassium, calcium, and
albumen were all low. She was treated with vigorous
crystalloid therapy and potassium replacement and
transferred to the intensive care unit.
Ultrasonography revealed no retained tissue in the
uterus. A computed tomography scan revealed pleural
effusion, ascites, and 2 small air bubbles in the vagina
and cervix. The antibiotics were changed to piperacillin,
tazobactam, and ciprofloxacin. The patient’s condition
continued to deteriorate, with increasing hemodynamic
instability. An endometrial biopsy sample revealed mod-
erate-to-severe necrosis and gram-positive rods.
On August 31, a hysterectomy was performed. Oper-
ative findings included ascites, edema of the pelvic tissue,
and bluish discoloration of the uterus without crepitation.
On September 1, the patient suffered refractory multiple
organ system failure, and, despite intraaortic balloon ther-
apy, experienced cardiac arrest and died. Final cultures,
both from the endometrial biopsy and the surgical speci-
men, were positive for C sordellii, Streptococcus milleri, and
Peptostreptococcus species. Autopsy revealed ascites, bilateral
pleural effusions, and hemorrhagic gastritis.
COMMENT
Clostridia are ubiquitous gram-positive, spore-forming,
obligate anaerobes that are generally found in soil and in
the alimentary tract of humans and other animals. Clos-
tridial species have been isolated from the vaginas of
4 –18% (average 11%) of normal healthy women.1 The
most common isolates are C perfringens.2
Infections with C sordellii are exceedingly rare, with
only 22 cases reported in the literature. Eight cases were
in obstetric or gynecologic patients, 7 after vaginal or
cesarean deliveries, and 1 case of endometritis was re-
ported in a woman with no known risk factors.3– 8 These
women all developed septic shock that was uniformly
fatal. Factors predisposing to endometritis or myometri-
tis in postpartum women include trauma and necrosis of
retained decidua. In the 7 reported cases of C sordellii
infections in postpartum women, 2 had had cesarean
deliveries, 2 had had episiotomy infections, and 1 had
had a retained vaginal pack.
In the case reported here, in association with medical
abortion, there was no trauma, no retained tissue in the
uterus or vagina, and no other known predisposing
factors. Additionally, the medications used to induce
abortion in this case—mifepristone and misoprostol—
Wiebe et al Case of Clostridium sordellii Infection 1143
have no known effects that could be considered contrib-
uting factors for this event. The rate of infection follow-
ing medical abortion is reported to be between 0.09%
and 0.5% and is generally lower than the rate of infection
following surgical abortion.9
Because Clostridia are part of normal vaginal flora, it
is unlikely the patient introduced the bacteria when she
inserted the misoprostol into her vagina. Vaginal miso-
prostol has been used by thousands of women as part of
a medical abortion regimen, with either methotrexate or
mifepristone, with very low infection rates. Moreover,
many other vaginal medications are used routinely, for
example, antifungal agents for vaginal candidiasis, with-
out an increase in endometritis or myometritis.
This case followed the clinical course reported in other
cases of C sordellii sepsis, including absence of fever,
absence of purulent discharge, high WBC count and
hemoglobin level, minimal uterine tenderness and flu-
like symptoms until the onset of symptoms and signs of
shock.6 The standard of care for postabortal endometritis is
oral antibiotics not adequate to treat clostridial infection.
Additionally, because of the nature of this infection, early
recognition of its severity is uncommon. The majority of C
sordellii infections are fulminant, and in reported obstetric
and gynecologic cases, uniformly fatal.
CONCLUSION
This patient developed fatal septic shock secondary to
infection with C sordellii after medical abortion with mife-
pristone and misoprostol. This syndrome is extremely
rare, with only 8 obstetric or gynecologic cases previ-
ously reported. The source of this infection was likely
the resident flora in the patient’s vagina. The infecting
organisms gained access to the upper genital tract, as can
occur during menstruation, delivery, and spontaneous
Amyloidosis of the Endometrium:
An Asymptomatic Presentation
Danielle D. Winkler, MD,
Jacqueline A. Emery, MD, and
Carol B. Alan, MD
Department of Obstetrics and Gynecology and Department of Pathology, University
of South Carolina and Palmetto Richland Memorial Hospital, Columbia, South
Carolina
Address reprint requests to: Danielle D. Winkler, MD, Department of
Obstetrics and Gynecology, University of South Carolina, 2 Medical
Park, Suite 208, Columbia, SC 29203; e-mail: ddwinkler@att.net.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1144 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
or induced abortion. Given the rarity of this infection
and its virulence, it is unlikely that there was any way to
prevent this death. Although physicians should remain
vigilant for the symptoms and signs of infection follow-
ing medical abortion, this case does not alter the overall
proven safety of medical abortions induced with mife-
pristone and misoprostol.
REFERENCES
1. Hammill HA. Normal vaginal flora in relation to vaginitis.
Obstet Gynecol Clin North Am 1989;16:329 –36.
2. Sweet RL, Gibbs RS. Clinical microbiology of the female
genital tract. In: Infectious diseases of the female genital
tract. Baltimore (MD): Williams & Wilkins; 1995. p. 3–15.
3. Hogan SF, Ireland K. Fatal acute spontaneous endometritis
resulting from Clostridium sordellii. Am J Clin Pathol 1989;91:
104 – 6.
4. Golde S, Ledger WJ. Necrotizing fasciitis in postpartum pa-
tients: a report of four cases. Obstet Gynecol 1977;50:670–3.
5. Soper DE. Clostridial myonecrosis arising from an episiot-
omy. Obstet Gynecol 1986;68:26S– 8.
6. McGregor JA, Soper DE, Lovell G. Todd JK. Maternal
deaths associated with Clostridium sordellii infection. Am J
Obstet Gynecol 1989;161:987–95.
7. Bitti A, Mastrantonio P, Spigaglia P, Urru G, Spano AI,
Moretti G, et al. A fatal postpartum Clostridium sordellii associ-
ated toxic shock syndrome. J Clin Pathol 1997;50:259–60.
8. Rorbye C, Petersen IS, Nilas L. Postpartum Clostridium
sordellii infection associated with fatal toxic shock syndrome.
Acta Obstet Gynecol Scand 2000;79:1134 –5.
9. Kruse B, Poppema S, Creinin MD, Paul M. Management of
side effects and complications in medical abortion. Am J
Obstet Gynecol 2000;183:S65–75.
Received December 3, 2003. Received in revised form February 6, 2004.
Accepted March 18, 2004.
BACKGROUND: Amyloidosis of the endometrium is a rare
occurrence according to current literature. Previously re-
ported cases have presented with menorrhagia or post-
menopausal bleeding.
CASE: A postmenopausal woman with multiple medical
problems presented with fatigue and weight loss. During
the evaluation for her 18-kg weight loss, a computed to-
mography scan revealed an enlarged uterus and liver le-
sions. Endometrial and liver biopsies were performed sec-
ondary to concern over metastatic cancer, given an
enlarged uterus in a postmenopausal woman with liver
masses. She was found to have systemic primary amyloid-
osis in multiple organs, including her endometrium.
CONCLUSION: This patient represents an interesting case of
systemic amyloidosis involving the endometrium that is
not associated with vaginal bleeding. The presence of amy-
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000124985.48138.7e
loid in the endometrium may be more common than
currently recognized, because patients without vaginal
bleeding are not routinely evaluated for amyloid deposi-
tion in their reproductive organs. (Obstet Gynecol 2004;
104:1144 –7. © 2004 by The American College of Obste-
tricians and Gynecologists.)
Amyloidosis is a group of diseases involving the deposi-
tion of similar-appearing proteins in the extracellular
space of various tissues and organs.1 The disease, which
was first described in 1856, primarily affects the heart,
kidneys, liver, gastrointestinal tract, respiratory system,
and nervous system.2 It can be either systemic or local-
ized to a particular organ or organ system. Cases involv-
ing the reproductive tract are rare, with only 12 cases of
cervical amyloid and 4 cases of uterine amyloid in the
literature.3– 6
Taylor et al3,5,6 described uterine amyloidosis in the
presence of systemic disease, whereas Jongen4 noted a case
of amyloidosis localized to the uterus. All of the aforemen-
tioned cases presented with vaginal bleeding, with either
postmenopausal bleeding or menorrhagia as the chief com-
plaint. We report a case of systemic amyloidosis with
uterine involvement as an incidental finding.
CASE
A 74-year-old African-American woman, gravida 4 para
3, presented with a report of a 18-kg weight loss during
the past 2 months. She had had diarrhea for a few days
and was having difficulty moving from her bed and
walking. She was admitted to the nephrology service
secondary to her required hemodialysis.
The patient’s medical history was significant for non–
insulin-dependent diabetes mellitus, hypertension, end-
stage renal disease requiring hemodialysis, congestive
heart failure, retinopathy, degenerative joint disease, and
bilateral carpal tunnel syndrome. Her renal function had
declined significantly within the past year, and she was
placed on hemodialysis 8 months before her current
presentation. Her serum creatinine level on admission
was 1.9 mg/dL. Her renal failure was attributed to dia-
betic nephropathy and hypertension given her 20-year
history of both conditions. She stopped smoking 9 years
ago and denied alcohol usage. She further denied vaginal
bleeding, chest pain, shortness of breath, fevers, or chills.
Physical examination was remarkable for a 4-cm nod-
ule in her left breast, which was mobile and nontender, a
small, reducible umbilical hernia, and mild lower-ex-
tremity edema. She had no palpable abdominal masses.
At pelvic examination, the patient had a slightly en-
larged, nontender, mobile, midplane uterus. The cervix
was grossly normal, and the ovaries were not palpable.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Fig. 1. Endometrial biopsy. Trichrome stain. Fragmented
portion of endometrium with lightly staining amorphous
amyloid protein in stroma. Original magnification: ! 4.
Winkler. Asymptomatic Amyloidosis of the Endometrium. Obstet Gynecol 2004.
Computed tomography scans of the chest, abdomen,
and pelvis were done to rule out cancer as a cause of the
weight loss. Abdominal computed tomography revealed
diffuse inhomogeneous attenuation of the liver compat-
ible with parenchymal disease or neoplasm. Her uterus
was noted to be enlarged, lobulated, and inhomoge-
neous. Transvaginal ultrasonography revealed an irreg-
ular uterus measuring 4.7 cm ! 4.5 cm ! 5.3 cm and
containing hyperechoic areas with calcifications compat-
ible with leiomyomata. The endometrial stripe was dis-
torted by the presence of presumed leiomyomata. The
ovaries were normal in size. Ultrasonography of the
patient’s breasts revealed 2 left breast lesions. Both had
ultrasound appearance suggestive of fibroadenomas.
Breast biopsy was declined by the patient and her family.
An endometrial biopsy was performed because of
concern about increased uterine size in a postmeno-
pausal woman. The endometrial biopsy sample con-
tained superficial endometrium admixed with acellular
proteinaceous and eosinophilic debris, often globular
(Fig. 1). No perivascular accumulation was noted. No
inflammation, hyperplasia, or neoplasia was noted.
Congo red staining with polarization showed only ques-
tionable apple-green birefringence. Immunohistochemi-
cal staining revealed positivity for amyloid P. No stain-
ing for $ or % light chains, amyloid A, or &-amyloid was
noted. Because the amyloid P component is not respon-
sible for the classic staining characteristics, it is not
surprising that little birefringence was identified on
Congo red staining.
Fine needle aspiration of the liver revealed marked
amounts of acellular, eosinophilic, hyaline material with
a panlobular, perisinusoidal distribution (Fig. 2). Scat-
tered areas of portal triad involvement were seen, with
Winkler et al Asymptomatic Amyloidosis of the Endometrium 1145

Fig. 2. Liver (fine needle aspiration-cell block). Trichrome
stain. Perisinusoidal distribution of amyloid protein. Orig-
inal magnification ! 40.
Winkler. Asymptomatic Amyloidosis of the Endometrium. Obstet Gynecol 2004.
sparing of the arterioles. Congo red stain with subse-
quent polarization displayed scant apple-green birefrin-
gence with a weak control. Immunohistochemical eval-
uation showed uniform positivity for amyloid P protein.
No staining for $ or % light chains, Amyloid A or
&-amyloid was noted.
An extensive workup was then begun for the source of
the patient’s amyloidosis, as well as for additional sites of
protein deposition. A transthoracic echocardiogram
showed severe left ventricular hypertrophy with globally
depressed function, as well as biatrial enlargement and
mild mitral insufficiency. A ground-glass appearance was
noted in the left ventricular myocardial mass, which is
suggestive, but not diagnostic, of amyloid involvement.
The workup for multiple myeloma had negative results,
and the patient was thought to have systemic primary
amyloidosis. No other cause was found for the patient’s
weight loss, and she was then transferred to a nursing
home for continued care.
COMMENT
Amyloidosis refers to the pathologic deposition of a
proteinaceous substance in the extracellular spaces of
various tissues and organs.1 Amyloid is made up largely
of nonbranching fibrils, which form a characteristic
&-pleated sheet. This configuration is responsible for the
diagnostic staining and apple-green birefringence seen
with Congo red stain.1 Other minor components are also
present: serum amyloid P component, proteoglycans,
and highly sulfated glycosaminoglycans.1
The incidence of amyloidosis is approximately 8 in 1
million, with an average age of 73.5 years.2 Approxi-
mately 60% to 65% of amyloidosis patients are male.2
1146 Winkler et al Asymptomatic Amyloidosis of the Endometrium
Fatigue and weight loss are the most common presenting
symptoms.2 The disease primarily affects the heart, kid-
neys, liver, gastrointestinal tract, respiratory system, and
nervous system. Although it is commonly classified on
the basis of the major fibril component, such as amyloid
associated, amyloid light-chain, or &-amyloid protein,
there is also a clinical classification (primary versus sec-
ondary). Amyloidosis can be associated with immuno-
cyte dyscrasias, such as multiple myeloma, (primary,
light-chain) and chronic inflammatory conditions (sec-
ondary, amyloid associated), or it can be a familial,
heterogenous disease.
Localized disease occurs when the amyloid deposition
is confined to a single organ, most commonly the kidney
(33–50%) and the heart, with symptoms of nephrotic
syndrome and restrictive cardiomyopathy, respective-
ly.2 There is 1 case in the literature in which the amyloid
deposition was confined to the uterus and was discov-
ered after a hysterectomy for postmenopausal bleeding.4
Systemic disease involves multiple organ systems and
is commonly associated with immunocyte dyscrasias,
most notably multiple myeloma, and chronic inflamma-
tory conditions, such as rheumatoid arthritis and inflam-
matory bowel disease. Patients undergoing long-term
hemodialysis for renal failure also may develop systemic
amyloidosis from the deposition of &2-microglobulin.
Once the histologic diagnosis of amyloidosis is made, the
patient is routinely tested for possible underlying causes,
if not already known, such as multiple myeloma and
other immunocyte dyscrasias, as well as the presence of
amyloid in other organ systems.
In our patient, amyloid deposition was noted in the
liver as well as the endometrium. Echocardiography
suggested a presence of amyloid in the heart as well. The
patient also had a history of bilateral carpal tunnel syn-
drome, a common sign of neurologic system amyloid
involvement. While a specific cause for this patient’s
amyloidosis was not found, the disease itself had likely
been present for many years and may have been a
contributing factor to her congestive heart failure and
end-stage renal disease.
Although this patient’s presentation with weight loss is
not uncommon, her endometrial involvement is unusual
in that she did not experience symptoms of menorrhagia
or postmenopausal bleeding. Had her radiologic exami-
nations not revealed an enlarged, lobulated uterus, the
involvement of the uterus most likely would never have
been reported. This suggests that the female reproductive
organs may have a greater incidence of amyloid deposition
than reported in the literature and that women with sys-
temic amyloidosis should be evaluated for genital tract
involvement, even if symptoms of bleeding are not present.
OBSTETRICS & GYNECOLOGY
REFERENCES
1. Robbins SL, Cotran RS, Kumar V. Diseases of immunity.
In: Cotran RS, Kumar V, Collins T, editors. Pathologic
basis of disease. 6th ed. Philadelphia (PA): WB Saunders;
1999. p. 251– 6.
2. Gertz MA, Lacy MQ, Dispenzieri A. Amyloidosis. Hematol
Oncol Clin North Am 1999;13:1211–33.
3. Taylor E, Gilks B, Lanvin. Amyloidosis of the uterine
cervix presenting as postmenopausal bleeding. Obstet
Gynecol 2001;98:966 – 8.
4. Jongen VHWM, Grond AJK, van Veelen H, Santema JG.
Low-Grade Endometrial Stromal
Sarcoma Arising From Sciatic
Nerve Endometriosis
Magali Lacroix-Triki, MD, Laurent Beyris, MD,
Pierre Martel, MD, and Bernard Marques, MD
Departments of Pathology, Radiology, and Gynecologic Surgery, Institut Claudius
Regaud, Toulouse, France
BACKGROUND: Endometrial stromal sarcoma arising from
extrauterine endometriosis is a rare and not easily diag-
nosed tumor. We present a case arising from sciatic nerve
endometriosis in a 50-year-old woman.
CASE: The patient had a long history of endometriosis and
presented with a left buttock mass and motor deficit. Mag-
netic resonance imaging showed a large tumor of the sciatic
nerve with pelvic extension. She underwent total hysterec-
tomy, bilateral salpingo-oophorectomy, and excision of
pelvic endometriotic pockets, allowing the diagnosis of
low-grade endometrial stromal sarcoma arising from en-
dometriosis. She received systemic chemotherapy, gonado-
tropin-releasing hormone agonist therapy, and palliative
radiation therapy, but her disease progressed.
CONCLUSION: Aggressive endometriosis raises important di-
agnostic and therapeutic difficulties and may correspond to
misdiagnosed rare malignant neoplasms, which should be
treated. (Obstet Gynecol 2004;104:1147–9. © 2004 by The
American College of Obstetricians and Gynecologists.)
Endometriosis is a well-known disease, usually easily
diagnosed, and is associated with a benign clinical
course. However, the significant occurrence of unrecog-
nized atypical forms is still responsible for important and
unexpected diagnostic and therapeutic difficulties. We
present a very unusual case of aggressive endometriosis
Address reprint requests to: Dr. Magali Lacroix-Triki, Laboratoire
d’Anatomie et Cytologie Pathologiques, Institut Claudius Regaud,
20--24 rue du pont Saint Pierre, 31052 Toulouse Cedex, France;
e-mail: lacroix_m@icr.fnclcc.fr.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
Uterine amyloidosis in menopause. Br J Obstet Gynaecol
1998;105:362– 4.
5. Copeland W, Hawley PC, Teteris NJ. Gynecologic amy-
loidosis. Am J Obstet Gynecol 1985;153:555– 6.
6. Lee JA, Angus B. Amyloidosis of the uterine vessels: an
unusual case of menorrhagia. Br J Obstet Gynaecol 1993;
100:1056 –7.
Received October 2, 2003. Received in revised form December 15, 2003.
Accepted December18, 2003
located to the sciatic nerve, which exemplifies such diffi-
culties. The observed clinical and pathological features
are consistent with a low-grade endometrial stromal
sarcoma arising from endometriosis of the sciatic nerve.
CASE
A 50-year-old woman was admitted for treatment of a
mass in the left sciatic nerve that was causing disabling
sciatica associated with leg motor deficit. Medical history
included an endometriosis that had been developing for
10 years, was presenting as a cyclic chronic pelvic pain,
and for which she successively received danazol, proges-
tins, and more recently, gonadotropin-releasing hor-
mone (GnRH) agonist therapy. Furthermore, the patient
had a periodic left-sided sciatica that first appeared post-
partum at age 24 and had gradually increased over the
past 2 years, leading to disabling constant pain and
motor deficit causing foot drop. Clinical examination
revealed a mass in the left buttock. Magnetic resonance
imaging showed a 10 ! 5 cm tumor of the left sciatic
nerve with a “dumbbell” extension to the left pelvis and
intratumoral hemorrhage foci (Fig. 1). Hysterectomy
and bilateral salpingo-oophorectomy for progressive and
recurrent endometriosis were determined to be the ap-
propriate treatments. At laparotomy, left-sided perito-
neal pelvic soft masses were excised. Pathologic exami-
nation showed extensive adenomyosis in the uterus and
multiple foci of pelvic endometriosis (Fig. 2). Image-
guided biopsy of the left buttock mass showed a poorly
differentiated sarcomatous proliferation that could not
be further characterized because of material insuffi-
ciency. Given the large size of the mass, which could not
be widely excised, systemic chemotherapy, combining
mesna, doxorubicin, ifosfamide, and dacarbazine, was
performed.
Midtreatment evaluation showed a clinical and radio-
logical improvement with stabilization of the tumor
growth. A second and larger image-guided biopsy al-
lowed a more accurate diagnosis of the left sciatic nerve
tumor. It showed a diffuse proliferation of spindle cells
0029-7844/04/$30.00 1147
doi:10.1097/01.AOG.0000128114.97877.33

Fig. 1. Magnetic resonance imaging (A. T1-weighted se-
quences; B. T2-weighted sequences) showing a large tumor
of the left sciatic nerve with a “dumbbell” extension to the
left pelvis through the foramen ischiadicum (arrow) and
intratumoral hemorrhage foci (arrowheads).
Lacroix-Triki. Aggressive Endometriosis. Obstet Gynecol 2004.
with scanty cytoplasm, ill-defined borders, and mild
nuclear atypia (Fig. 2). Mitotic activity was low, with 4
mitoses per 10 high-power fields. Sparse endometrioid
glands were observed. Immunostains for vimentin, es-
trogen, and progesterone receptors were strongly posi-
tive. The monoclonal antibody MIB-1 stained less than
10% of tumor cells. These pathologic findings were
consistent with a low-grade endometrial stromal sar-
coma developing from endometriosis.
An assessment laparotomy with first-stage tumor ex-
cision was decided upon next, but significant fibrous
adhesions and pelvic tumor extension to the bladder and
the sigmoid colon prevented surgical tumor debulking.
The omentum grossly appeared normal, and there was
no ascites nor any lymphadenopathy. The patient under-
went GnRH agonist therapy postoperatively. Clinical
outcome was marked by a disease progression with
Fig. 2. A. Pelvic peritoneum: endometriotic focus, charac-
terized by endometrial epithelium (arrow) surrounded by a
cellular stroma with hemosiderin deposits (asterisk) (he-
matoxylin-eosin, ! 150, original magnification). B. and C.
Sciatic nerve: diffuse proliferation of short spindle cells
with mild atypia and few mitoses (arrow) (hematoxylin-
eosin, ! 150 and ! 400, original magnifications, respec-
tively).
Lacroix-Triki. Aggressive Endometriosis. Obstet Gynecol 2004.
1148 Lacroix-Triki et al Aggressive Endometriosis
intense recurrence of sciatica and emergence of multiple
and quickly growing pelvic masses. The left sciatic nerve
tumor was measured at 24 ! 12 cm, with involvement of
second and third foramina sacralia pelvina. The patient
presented with significant deterioration of her general
condition, associated with an increase of inflammatory
parameters. She received palliative pelvic radiation ther-
apy for a total of 30 Gy with a partial improvement of
abdominal pains. She died not long afterward.
COMMENT
Considering the involvement site, but also the strikingly
aggressive clinical course, the case that we present is
particularly unusual and raised important diagnostic and
therapeutic difficulties. With about 20 cases described in
the literature, endometriosis of the sciatic nerve is a rare
event, typically revealed by catamenial sciatica in
women with a history of endometriosis.1,2 Malignant
sarcomatous transformation of endometriosis is an ex-
tremely rare complication, which is probably underesti-
mated because of tumor sampling that seldom retrieves
the preexistent endometriosis.1 Furthermore, incomplete
tumor sampling by image-guided biopsies was responsi-
ble for important diagnostic problems. Indeed, extrauter-
ine mixed epithelial and stromal tumor– expressing hor-
mone receptors may raise difficult differential diagnoses,
ie, diagnoses of an endometriosis, a müllerian adenosar-
coma, or an endometrial stromal sarcoma.1 Partly be-
cause of the large size of the tumor and the aggressive
clinical behavior associated with hormone therapy resis-
tance, we made the diagnosis of low-grade endometrial
stromal sarcoma with endometrioid glandular differenti-
ation. The occurrence of endometrioid glandular differ-
entiation has been previously described and is a source
of misdiagnosis as endometriosis.3,4 The same authors
even suggested that the cases of “aggressive endometri-
osis” reported in the literature may be endometrial stro-
mal sarcoma with glandular differentiation.3 Indeed, our
case typically corresponds to the “aggressive endometri-
osis” entity described in the literature and, particularly,
to one case of aggressive endometriosis in bone reported
by Oei et al in 1992.5 Over the past 10 years, many
authors have highlighted the disparity in the clinical
course of endometriosis, and some of them have made
attempts to classify the disease,6 but without any success.
Therefore, the term of “aggressive endometriosis” was
used to describe unusual cases presenting poor clinical
behavior often associated with hormone therapy resis-
tance.5,7 Considering the progression of such a disease,
sprinkled with local recurrences, it has been recently
proposed that aggressive cases actually correspond to
sarcomas that are misdiagnosed, partly because of their
OBSTETRICS & GYNECOLOGY
rarity and partly because of inadequate tumor sam-
pling.3
As expected, such cases raise important therapeutic
difficulties. Lack of hormonal responsiveness, ie, resis-
tance to progestins, GnRH analogs, and tamoxifen, is a
main characteristic.5–7 Interestingly, successful antiaro-
matase treatment of an unusually aggressive case of
endometriosis has been recently reported.8 Further stud-
ies will certainly explore this promising therapeutic ap-
proach. As in our case, a surgical approach may be very
difficult but generally consists in total hysterectomy and
bilateral salpingo-oophorectomy associated, if possible,
with a tumor debulking.5–7 Because it is very seldom
used, the responsiveness to radiotherapy at an anti-
inflammatory dose is not known. In our case, a partial,
but indisputable, clinical improvement was obtained
with a total dose of 30 Gy.
REFERENCES
1. Kurman RJ. Blaustein’s pathology of the female genital
tract. 5th ed. New York (NY): Springer-Verlag; 2002.
2. Vilos GA, Vilos AW, Haebe JJ. Laparoscopic findings, man-
agement, histopathology, and outcome of 25 women with
cyclic leg pain. J Am Assoc Gynecol Laparosc 2002;9:145–51.
Symptomatic Diaphragmatic
Endometriosis Ten Years After
Total Abdominal Hysterectomy
Barak Nahir, Talia Eldar-Geva, MD, PhD,
Joseph Alberton, MD, and Uzi Beller, MD
Department of Obstetrics and Gynecology, Shaare Zedek Medical Center, Jerusa-
lem, and the Faculty of Health Sciences, Ben-Gurion University of The Negev,
Be’er-Sheva, Israel
BACKGROUND: Endometriosis is a disease that affects
women, mostly in the age range of 25–35 years, and in most
cases pelvic organs are involved. Involvement of the dia-
phragm after hysterectomy is extremely uncommon.
CASE: A 50-year-old woman presented to our department
with right upper-quadrant abdominal pain. Ten years
before her admission, she underwent total hysterectomy
and right salpingo-oophorectomy for a large leiomyoma-
tous uterus. On evaluation, a right diaphragmatic lesion
was identified by computed tomography. An explorative
laparotomy was then performed, which revealed a 4-cm
Address reprint requests to: Uzi Beller, MD, Shaare Zedek Medical
Center, Obstetrics and Gynecology, Faculty of Health Sciences, Ben-
Gurion University of The Negev, Be’er-Sheva, Jerusalem, Israel;
e-mail: beller@szmc.org.il.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
3. Clement PB, Scully RE. Endometrial stromal sarcomas of
the uterus with extensive endometrioid glandular differen-
tiation: a report of three cases that caused problems in
differential diagnosis. Int J Gynecol Pathol 1992;11:163–73.
4. Levine PH, Abou-Nassar S, Mittal K. Extrauterine low-
grade endometrial stromal sarcoma with florid endometri-
oid glandular differentiation. Int J Gynecol Pathol 2001;20:
395– 8.
5. Oei SG, Peters AA, Welvaart K, Bode PJ, Fleuren GJ. Aggres-
sive endometriosis in bone. Lancet 1992;339:1477–8.
6. Abrao MS, Podgaec S, Carvalho FM, Pinotti JA. Endome-
triosis in the presacral nerve. Int J Gynaecol Obstet 1999;
64:173–5.
7. Metzger DA, Lessey BA, Soper JT, McCarty KS Jr, Haney
AF. Hormone-resistant endometriosis following total
abdominal hysterectomy and bilateral salpingo-oophorec-
tomy: correlation with histology and steroid receptor con-
tent. Obstet Gynecol 1991;78:946 –50.
8. Takayama K, Zeitoun K, Gunby RT, Sasano H, Carr BR,
Bulun SE. Treatment of severe postmenopausal endometri-
osis with an aromatase inhibitor. Fertil Steril 1998;69:
709 –13.
Received November 13, 2003. Received in revised form January 9,
2004. Accepted January 15, 2004.
diaphragmatic cyst compressing the liver surface and con-
taining thick chocolate-colored material. The lesion was
totally excised. Pathological examination confirmed the
diagnosis of endometriotic cyst.
CONCLUSION: The diagnosis of endometriosis involving the
diaphragm with no evidence of disease in the pelvis 10
years after hysterectomy, although a rare situation, should
be considered in the differential diagnosis of a symptom-
atic diaphragmatic lesion in a woman with a single func-
tioning ovary. (Obstet Gynecol 2004;104:1149 –51.
© 2004 by The American College of Obstetricians and
Gynecologists.)
Endometriosis developing after hysterectomy is a rare
condition and in most cases is attributed to recurrent
disease. We found in the English literature only 1 de-
scription of endometriosis as a primary disease with no
pelvic involvement.1 The location of endometriosis in
the diaphragm is also very unusual. Here we report the
case of the oldest woman with symptomatic diaphrag-
matic endometriosis with no evidence of pelvic disease
and the longest interval after hysterectomy.
CASE
A 50-year-old woman was referred initially to general
surgery because of right upper-quadrant abdominal
pain. Her complaint did not include any periodic pelvic
0029-7844/04/$30.00 1149
doi:10.1097/01.AOG.0000125000.20582.64

Fig. 1. Abdominal ultrasound showing a 4-cm cyst on the
right lobe of the liver adjacent to the diaphragm. The arrow
indicates the cyst’s wall.
Nahir. Diaphragmatic Endometriosis De Novo. Obstet Gynecol 2004.
or lower abdominal pain, dyspareunia, or catamenial
pain. Total abdominal hysterectomy with right salpingo-
oophorectomy had been performed 10 years before to
treat symptomatic uterine leiomyomata, which were
confirmed by histological examination. At the patient’s
request, the left ovary was not removed to avoid early
menopause. Evaluation on admission ruled out choleli-
thiasis by normal liver enzymes and by normal abdom-
inal ultrasonography. A gastric endoscopy demonstrated
esophageal hernia with esophagitis, and she was treated
with ranitidine. Despite a short period of relief, the
symptoms continued, and the patient was referred to our
hospital 6 months later.
Abdominal ultrasonography (Fig. 1) and computed
tomography demonstrated a 4-cm cystic mass on the
surface of the right lobe of the liver adjacent to the
diaphragm. A small multilobular left adnexal cystic mass
was also seen. The level of CA 125 was 34.5 mg/dL on
admission. Because of the continuation of symptoms and
the unexplained lesion in the abdomen, an explorative
laparotomy was performed. During the operation the left
ovary, the small simple cysts adjacent to it, and the
diaphragmatic lesion were all completely excised. A lac-
eration of the diaphragm during dissection was sutured.
Pathological examination from the right diaphragm
showed a 4-cm cyst with chocolate-colored material and
a simple ovarian cyst. Microscopy revealed a benign endo-
metriotic cyst, a normal ovary, and 2 paraovarian cysts.
The abdominal symptoms disappeared on follow-up.
1150 Nahir et al Diaphragmatic Endometriosis De Novo
COMMENT
Endometriosis is defined as the presence of ectopic en-
dometrial tissue outside the endometrium. This condition
affects 10% to 15% of premenopausal women, usually
those between the ages of 25 and 35 years, but it can occur
rarely in younger and older women.2 The locations of
the endometrial lesions are usually within the pelvis.
A MEDLINE search of English-language publications
on the search term “diaphragmatic endometriosis” dur-
ing the past 3 decades was performed. Only 22 cases of
symptomatic endometriosis on the diaphragm3 were
found. All reported cases, but one, had disease in the
pelvis as well. Only one reported case of symptomatic
diaphragmatic endometriosis had disease located also in
the right upper pole of the kidney. This latter case also
appeared after total abdominal hysterectomy, and it was
assumed to be related to the use of hormone therapy.1
Various theories have been proposed to explain the
occurrence of late endometriosis. Considerable evidence
supports the migratory theory, which suggests that en-
dometrial tissue elements originate in the uterine mucosa
and reach an ectopic position by lymphatic or hematog-
enous spread.4 Other evidence suggests that endometri-
osis is caused by transplantation of viable endometrial
fragments shed during menses that are regurgitated di-
rectly through the fallopian tubes under the influence of
prostaglandin-mediated uterine contractions.5 The the-
ory that could explain the occurrence of diaphragmatic
endometriosis, as in our case, is hematogenous spread.
However, that would be very unlikely for 2 reasons.
First, menses stopped 10 years before the symptomatic
disease started, and second, there was no evidence of
disease on ultrasonography 6 months before the finding
of the diaphragmatic lesion. It may be that the endome-
trial cells that were spread to the diaphragm 10 years
before did not cause any lesion or symptoms and sud-
denly developed a 4-cm cyst within 6 months. It is also
possible to explain this case with past migration of endo-
metrial cells in the peritoneal cavity with the peristalsis of
the ascending colon because the lesion was found on the
right diaphragm and most cases occur on this side.3
However, this theory is also unlikely because there was
no evidence of disease on the tract from the uterus to the
diaphragm.
The other theory by which we may explain the patho-
genesis of endometriosis, especially its occurrence at
extragenital sites, is the coelomic metaplastic theory.6
According to this theory, endometriosis develops from
metaplasia of coelomic cells under the influence of a local
and hormonal inducing substance. It is possible that in
our case endometriosis arose because of coelomic meta-
plasia of the peritoneum under the diaphragm under the
OBSTETRICS & GYNECOLOGY
influence of estrogen, which was produced by the func-
tioning left ovary. The combination of peritoneum that
had gone through metaplastic changes to become endo-
metrial tissue with estrogen as an endometrial growth
hormone could give rise to an extrapelvic endometriosis
lesion that developed in a relatively short time.
REFERENCES
1. Chinegwundoh F, Ryan P, Lusley T, Chan S. Renal and
diaphragmatic endometriosis de novo associated with hor-
mone replacement therapy, J Urol 1995;153:380 –1.
2. Case records of the Massachusetts General Hospital.
Weekly clinicopathological exercises. Case 33-1992. A
34-year-old woman with endometriosis and bilateral hydro-
nephrosis. N Engl J Med 1992;327:481–5.
3. Redwine DB. Diaphragmatic endometriosis: diagnosis, sur-
Hepatic Ischemia Associated
With Coarctation of the Aorta in
Pregnancy: Key Issues in
Differential Diagnosis
George J. M. Webster, Jelica Kurtovic,
Sandra A. Lowe, and
Stephen M. Riordan, MD, FRACP
Gastrointestinal and Liver Unit, The Prince of Wales Hospital, and Royal
Hospital for Women, Sydney, Australia.
BACKGROUND: Hepatic ischemia associated with coarcta-
tion of the aorta has not previously been reported in an
adult; pregnancy increases the pressure gradient across a
coarctation.
CASE: A young woman with known coarctation of the
aorta developed severe hepatic ischemia in pregnancy. A
pregnancy-induced increase in the mean pressure gradient
across the coarctation, from 18 mm Hg before pregnancy to
40 mm Hg in the third trimester, predisposed to critical
hepatic hypoperfusion in the setting of dehydration.
CONCLUSION: This case documents an association between
coarctation of the aorta and hepatic ischemia, precipitated
by pregnancy and dehydration in combination. It empha-
sizes the need in the assessment of patients with liver disease
in pregnancy to consider not only “traditional” pregnancy-
related conditions such as acute fatty liver and the hemolysis,
elevated liver enzymes, low platelets syndrome, in which
Address reprint requests to: Stephen Riordan, MD, FRACP, Director,
Gastrointestinal and Liver Unit, The Prince of Wales Hospital, Barker
Street, Randwick 2031, New South Wales, Australia; e-mail:
sriordan@ozemail.com.au.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
gical management, and long-term results of treatment. Fertil
Steril 2002;77:288 –96.
4. Bergqvist A. Extragenital endometriosis: a review. Eur
J Surg 1992;158:7–12.
5. Haney A. Etiology and histogenesis of endometriosis. Prog
Clin Biol 1990;323:1–14.
6. Suginami H. A reappraisal of the coelomic metaplasia the-
ory by reviewing endometriosis occurring in unusual sites
and instances. Am J Obstet Gynecol 1991;165:214 – 8.
Address reprint requests to: Uzi Beller, MD, Shaare Zedek
Medical Center, Obstetrics and Gynecology, Faculty of Health
Sciences, Ben-Gurion University of The Negev, Be’er-Sheva,
Jerusalem, Israel; e-mail: beller@szmc.org.il.
Received September 11, 2003. Received in revised form January 21,
2004. Accepted January 28, 2004.
delivery may be necessary as a clinical emergency, but also
those in which the circulatory and metabolic demands of
pregnancy may precipitate liver injury. (Obstet Gynecol
2004;104:1151–4. © 2004 by The American College of Ob-
stetricians and Gynecologists.)
Hepatic ischemia associated with coarctation of the aorta
has not previously been reported in an adult. This case
documents an association between coarctation and he-
patic ischemia, precipitated by pregnancy and dehydra-
tion in combination. Perhaps even more importantly, it
reiterates the need in the assessment of patients with liver
disease in pregnancy to consider not only those “tradi-
tional” pregnancy-specific conditions such as acute fatty
liver of pregnancy and the hemolysis, elevated liver
enzymes, low platelets (HELLP) syndrome, in which
delivery may be necessary as a clinical emergency, but
also those in which the circulatory and metabolic de-
mands of pregnancy may precipitate clinically apparent
liver injury.
CASE
A young woman, 30 weeks into her second pregnancy,
presented to the emergency department with a 36-hour
history of nausea, vomiting, colicky periumbilical ab-
dominal pain, and watery nonbloody diarrhea. These
symptoms were preceded by a 5-day history of fever,
rhinorrhea, myalgia, and headache. She had taken less
than 2 g per day of acetaminophen on 2 occasions during
this time and no other prescribed or nonprescription
medications. She had undergone cardiothoracic surgery
at the age of 18 months for correction of a coarctation of
the arch of the aorta, which was located just proximal to
the origin of the left subclavian artery and had resulted in
0029-7844/04/$30.00 1151
doi:10.1097/01.AOG.0000128108.68908.73
the blood pressure in the left arm being lower than in the
right arm as a consequence of obstruction to blood flow.
Surgical intervention had substantially reduced the de-
gree of aortic obstruction due to the coarctation, al-
though a minor pressure gradient across it had persisted.
Before pregnancy her blood pressure in the right arm
was 140/80 mm Hg and the pressure gradient across the
coarctation, measured by ultrasound doppler, was 18
mm Hg. Her first pregnancy, 2 years earlier, had been
complicated by an exacerbation of hypertension but was
otherwise uneventful.
On examination she was icteric, with no signs of
chronic liver disease (such as Dupuytren’s contractures
or leukonychia) or hepatic encephalopathy (such as poor
concentration, dyspraxia, confusion, asterixis, or im-
paired consciousness). There was a left thoracotomy
scar. Her temperature was 37.6°C. She was clinically
dehydrated, with dry mucosae and a jugular venous
impulse that was not visible while lying flat. Her blood
pressure, as measured in the right arm in the semirecum-
bent position, was 90/60 mm Hg, and her heart rate was
90 beats per minute and regular. Her blood pressure in
the left arm was 60/35 mm Hg. A loud systolic murmur
was heard over the precordium and the posterior chest
wall. There were no peripheral stigmata of infective
endocarditis (such as splinter hemorrhages, Janeway le-
sions, Osler’s nodes, or Roth’s spots). Respiratory exam-
ination was unremarkable. The arterial blood oxygen
saturation while breathing room air was 99%. Abdomi-
nal examination revealed a uterine fundal height com-
patible with dates. No peripheral edema was demon-
strated. There were no abnormal neurologic signs. In
particular, hyperreflexia and clonus were not demon-
strated.
Results of initial blood tests, reported as value (normal
range), included bilirubin 57 #mol/L (0 –25), alkaline
phosphatase 171 U/L (38 – 400), "-glutamyl transpepti-
dase 23 U/L (0 –30), alanine aminotransferase 738 U/L
($30), aspartate aminotransferase 1,858 U/L ($30), to-
tal protein 61g/L (58 –70), albumin 34 g/L (30 –35),
international normalized ratio 2.5 (0.8 –1.1), glucose 6.8
mmol/L (3.9 – 6.1), creatinine 0.08 mmol/L (0.03– 0.08),
and uric acid 0.19 mmol/L (0.10 – 0.27). The arterial
lactate level was 4.0 mmol/L (0.7–1.6). The hemoglobin
was 131 g/L (115–165), and white cell count was 11 !
109/L (3.5–11), with normal differential counts and plate-
lets 143 ! 109/L (150 – 450). The D-dimer level was
% 1.6 mg/L ($0.19mg/L). Hemolytic screen results were
negative. Results of tests for infective, metabolic, and
immune-mediated liver disorders (hepatitis A virus IgM,
hepatitis B surface antigen, hepatitis B core IgM, hepati-
tis C virus antibody, hepatitis E antibody, antibodies to
cytomegalovirus, Epstein-Barr virus, adenovirus and
1152 Webster et al Hepatic Ischemia and Coarctation in Pregnancy
toxoplasmosis, antinuclear, anti–smooth muscle and an-
timitochondrial antibodies, serum immunoglobulin lev-
els, iron studies, alpha1-antitrypsin, and ceruloplasmin
levels) were all negative. Antibody profiles for herpes
simplex virus and varicella zoster virus were consistent
with previous infection. A mononucleosis spot test gave
negative results. Blood cultures proved sterile. Urinary
drug screen results, including for acetaminophen, was
negative. Liver ultrasound demonstrated normal size,
echogenicity, and outline and no biliary dilatation. He-
patic and portal veins were patent, with normal venous
waveforms on doppler examination. The spleen and
kidneys were of normal size and there was no ascites.
There was an intrauterine pregnancy, dated at 29 –30
weeks of gestation, with normal umbilical vein wave-
form on doppler examination. Urinalysis showed pro-
tein &&. A subsequent 24-hour urine collection yielded
0.54 g of protein (normal $ 0.3). Testing for viral agents
responsible for gastroenteritis demonstrated an elevated
enterovirus complement fixation titer (1:64), in keeping
with recent infection.
A diagnosis was made of severe hepatic ischemia
associated with coarctation of the aorta, with critical
hepatic hypoperfusion precipitated by both the hemody-
namic changes of pregnancy, which act to increase the
pressure gradient across a coarctation, and hypovolemia
associated with viral gastroenteritis. The patient received
aggressive fluid resuscitation with colloid and crystal-
loid. The blood pressure quickly increased to 140/80 mm
Hg in the right arm. Echocardiography with ultrasound
doppler, performed after fluid resuscitation, confirmed
coarctation just proximal to the origin of the left subcla-
vian artery, with a mean pressure gradient across it of 40
mm Hg, and associated left ventricular hypertrophy.
After a peak aspartate aminotransferase level of 5,226
U/L on day 3 of her admission, the liver enzyme, biliru-
bin, and international normalized ratio levels rapidly
improved and normalized over the subsequent 2 weeks.
The serum creatinine level fell to 0.04 mmol/L, and
proteinuria resolved. Nausea, vomiting, and diarrhea
resolved. The remainder of her pregnancy was unevent-
ful, and the patient delivered a live baby by normal
vaginal delivery at 38 weeks of gestation. The mean
pressure gradient across the coarctation was 19 mm Hg
when reassessed 8 weeks postpartum.
COMMENT
The very high peak aspartate aminotransferase level of
5,226 IU/L in our patient limited the differential diagno-
sis to acetaminophen toxicity, of which there was no
evidence, and hepatic ischemia. Such markedly elevated
aminotransferase levels are rarely if ever seen in infec-
tive, metabolic, or immune-mediated pathologies, inves-
OBSTETRICS & GYNECOLOGY
tigation for which, in any regard, proved negative.
Severe hepatic ischemia due to systemic arterial hypoten-
sion is rarely seen in the absence of at least some degree
of associated renal impairment, and it is notable that,
although the serum creatinine level characteristically
falls in uncomplicated pregnancy, our patient’s serum
creatinine level on admission was double that recorded
both before pregnancy and after subsequent fluid resus-
citation. The diagnosis of hepatic ischemia was further
supported by the markedly elevated blood lactate level at
presentation, which, in the setting of normal oxygen-
ation of arterial blood, was indicative of a critical reduc-
tion in tissue perfusion, sufficient to result in anaerobic
metabolism.
A pregnancy-induced increase in the mean pressure
gradient across the patient’s known coarctation of the
aorta, from 18 mm Hg before pregnancy to 40 mm Hg in
the third trimester, was central to the development of
hepatic ischemia in our patient’s case, as a result of the
increased degree of aortic obstruction predisposing to
impaired perfusion distal to it. Pregnancy increases the
pressure gradient across a coarctation as a result of
associated increases in blood and stroke volumes and
reduction in systemic vascular resistance,1 such that an
increased quantity of blood is presented to a relatively
tighter, fixed stricture per unit time. However, coarcta-
tion alone could not be held accountable for the devel-
opment of hepatic ischemia in our patient, because the
severe liver injury resolved despite this pressure gradient
persisting throughout the remainder of the pregnancy.
Rather, the increased functional severity of the coarcta-
tion induced by pregnancy predisposed our patient to a
critical reduction in hepatic perfusion in the setting of
hypovolemia associated with enterovirus-related gastro-
enteritis. It might be argued that hepatic ischemia was
solely due to hypovolemia and that the coarctation was
irrelevant. However, such a severe liver insult in our
patient could not be explained by hypovolemia alone,
because the mean arterial blood pressure, measured
proximal to the coarctation, remained above 70 mm Hg
throughout. Hepatic perfusion is not compromised at
this level. Conversely, the mean arterial pressure distal
to the coarctation at admission was only 44 mm Hg, a
level at which organ perfusion is known to be critically
impaired. The possibility that hepatic ischemia was
caused by Budd-Chiari syndrome, which results from
obstruction to hepatic venous outflow and which may
occur in pregnancy as a result of the associated hyperco-
agulable state, rather than by coarctation and dehydra-
tion in combination, was excluded by the normal hepatic
venous doppler examination.
Despite a systematic search of the literature (including
MEDLINE and PubMed from 1966 and 1965, respec-
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
tively, to the present, encompassing all languages and
using the search terms “coarctation and liver,” “hepatic,”
“ischemia,” and “ischemia or adult”), we could find no
other report of hepatic ischemia due to coarctation of the
arch of the aorta in the adult. However, the potentially
adverse effect of coarctation on hepatic perfusion has
been reported in infants. Hepatic necrosis has been ob-
served as a particularly common finding in infants who
died from coarctation.2,3 In another report, more than
70% of 137 infants with congenital heart disease who had
hepatic necrosis on autopsy were found to have coarcta-
tion or the related hypoplastic left heart syndrome. Man-
ifestations of liver failure were commonly seen as part of
the terminal illness.4
The estimated incidence of coarctation of the aorta is
on the order of 1:12,000 live births. There is a particular
association with certain genetic conditions, including
Turner’s syndrome. In the presence of uncorrected co-
arctation, historical data suggest that pregnancy carries a
high risk to both mother and fetus, with complications
related to placental insufficiency and maternal hyperten-
sion.5 After advances in surgical repair and balloon
angioplasty, the great majority of pregnancies are now
successful, even though the risk of complicating hyper-
tension remains increased, especially when the residual
gradient across the coarctation is in excess of 20 mm
Hg.6,7
Hepatic ischemia occurring in association with a preg-
nancy-induced increase in the functional severity of co-
arctation and hypovolemia in our patient needed to be
differentiated from other, more widely recognized, preg-
nancy-related liver disorders in which hepatocellular
injury is a feature, such as the HELLP syndrome and
acute fatty liver of pregnancy. Proteinuria and hyperten-
sion, which were seen after resuscitation in our patient,
raised the possibility of preeclampsia, the initial insult to
the liver in which is presumed to result in the HELLP
syndrome.8 The latter, which typically presents in the
third trimester, often with malaise, nausea, headache,
and abdominal discomfort, was an early consideration in
our patient. However, subsequent investigation ex-
cluded hemolysis, whereas the platelet count, which is
usually $ 100 ! 109/L in HELLP syndrome, was only
minimally reduced. Although marked elevations in ala-
nine aminotransferase and aspartate aminotransferase
levels may be seen in conjunction with complicating
subcapsular hematoma or hepatic infarction, no imaging
evidence of which was apparent in our patient, these
values are classically elevated to less than 500 U/L in
HELLP syndrome.8 Moreover, it turned out that each of
the findings that raised the possibility of preeclampsia
could be explained on other grounds. Hypertension was
of long-standing origin, as evidenced by the finding of
Webster et al Hepatic Ischemia and Coarctation in Pregnancy 1153
left ventricular hypertrophy on echocardiography, and
accounted for by the coarctation, whereas proteinuria
resolved after fluid resuscitation and restoration of ade-
quate renal perfusion. Other features of preeclampsia,
such as hyperreflexia and clonus, were not present, and
the uric acid level was normal.
Acute fatty liver of pregnancy also required consider-
ation in our patient, in view of both the characteristic
presentation in the third trimester, often with nonspe-
cific, viral-like symptoms, and the marked degree of
coagulopathy. The latter, along with hypoglycemia, is
typically disproportionately severe in acute fatty liver of
pregnancy when compared with the degree of jaundice
and liver enzyme abnormality.8 Hence, the markedly
elevated alanine transaminase and aspartate transami-
nase levels and normal blood glucose level were impor-
tant early pointers against this diagnosis, and the rapid
overall clinical improvement in response to fluid resusci-
tation, without the necessity of delivery, finally excluded
acute fatty liver of pregnancy as the cause of our patient’s
severe liver disturbance.
Other causes of liver damage whose prevalence or
natural history can be influenced by pregnancy could
similarly be discounted in our patient. Although chronic
liver disorders such as autoimmune hepatitis, chronic
viral hepatitis, and primary biliary cirrhosis may first
become clinically evident during pregnancy, the more
common pattern is of improvement in liver enzyme
levels during this time, sometimes followed by a postpar-
tum flare of disease activity. Reduction in liver damage
during pregnancy has been attributed to maternal immu-
nosuppression during the third trimester, with an exac-
erbation of liver injury after delivery being the conse-
quence of immune reconstitution. In any case, serology
results for these various disorders proved negative in our
Fetal Femur Fracture and
External Cephalic Version
Steven Papp, MD, Gupreet Dhaliwal, MD,
Greg Davies, MD, and Dan Borschneck, MD
Department of Orthopedics, Queen’s University, Kingston, Ontario, Canada; and
Department of Obstetrics and Gynecology, Queen’s University, Kingston, Ontario,
Canada
BACKGROUND: Femoral fractures due to birth trauma are
extremely rare. External cephalic version is considered a
Address reprint requests to: Dr. Steven Papp, Victory 3 — Room 308,
Kingston General Hospital, 76 Stuart Street, Kingston, Ontario, Can-
ada, K7L 2V7; e-mail: srpapp21@yahoo.ca
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1154 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
patient. Biliary sludge and gallstones are a common
ultrasound finding in pregnancy, and although symp-
tomatic choledocholithiasis is unusual, this complication
seems to occur more frequently than in the nonpregnant
female population. No imaging evidence of gallbladder
or biliary pathology was apparent in our patient, whose
clinical features could not, in any case, be explained on
the basis of biliary obstruction.
REFERENCES
1. Oakley C. Acyanotic congenital heart disease. In: Oakley
C, editor. Heart disease in pregnancy. London: BMJ Pub-
lishing Group; 1997. p. 70 – 82.
2. Shiraki K. Hepatic cell necrosis in the newborn. Am J Dis
Child 1970;119:395– 400.
3. Coen R, McAdams AJ. Visceral manifestation of shock in
congenital heart disease. Am J Dis Child 1970;119:383–9.
4. Weinberg AG, Bolande RP. The liver in congenital heart
disease. Am J Dis Child 1970;119:390 – 4.
5. Goodwin JF. Pregnancy and coarctation of the aorta. Lancet
1958;1:16 –20.
6. Saidi AS, Bezold LI, Altman CA, Ayres NA, Bricker JT.
Outcome of pregnancy following intervention for coarcta-
tion of the aorta. Am J Cardiol 1998;82:786 – 8.
7. Beauchesne LM, Connolly HM, Ammash NM, Warnes
CA. Coarctation of the aorta: outcome of pregnancy. J Am
Coll Cardiol 2001;38:1728 –33.
8. Bacq Y, Riely CA. The liver in pregnancy. In: Schiff ER,
Sorrell MF, Maddrey WC, editors. Schiff’s diseases of the
liver. 9th ed. Philadelphia: Lippincott Williams & Wilkins;
2003. p. 1435–57.
Received December 17, 2003. Received in revised form February 23,
2004. Accepted March 3, 2004.
relatively safe alternative when dealing with breech pre-
sentation, but it can be associated with complications.
CASE: This patient underwent elective cephalic version for
breech presentation at 36 weeks of gestation. Due to con-
cerns of fetal distress after the version, a cesarean delivery
was performed. The newborn had a bruised leg at delivery,
and X-ray studies confirmed a distal femoral corner
fracture.
CONCLUSION: This case demonstrates another risk of exter-
nal cephalic version. Physicians and patients should be
aware of this potential complication. (Obstet Gynecol
2004;104:1154 – 6. © 2004 by The American College of
Obstetricians and Gynecologists.)
Infant fractures at birth are uncommon injuries and
usually involve the clavicle, humerus, and femur. Al-
though the incidence of femoral fractures at birth is not
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000128112.33398.31
Fig. 1. Initial radiograph showing that a distal femoral
fracture can often be subtle in neonates, as the distal
femur is mostly cartilaginous at this point.
Papp. Fracture After Cephalic Version. Obstet Gynecol 2004.
well known, it is thought to be extremely rare.1 Femoral
fractures at birth can occur as a result of vaginal breech
deliveries and cesarean deliveries.1– 4
External cephalic version is considered a relatively safe
alternative when dealing with a breech presentation. Com-
plications of external cephalic version include fetal heart
rate changes most commonly, but can include preterm
labor, rupture of membranes, placental abruption, fetoma-
ternal hemorrhage, and fetal demise.5 We present a case of
a femur fracture as a result of an external cephalic version.
CASE
A healthy 29-year-old woman gravida 3 para 0 presented
to the hospital at 36 weeks of gestation to undergo
external cephalic version for a frank breech presentation.
The pregnancy up to the point of the procedure had been
unremarkable. Successful external cephalic version was
accomplished by using a single attempt at forward roll
after elevation of the breech out of the maternal pelvis. No
analgesia or tocolysis was used. Fetal heart rate monitoring
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Fig. 2. Radiographs taken 3 weeks after birth showing
abundant callous formation around the entire distal femur,
more suggestive of extent of injury.
Papp. Fracture After Cephalic Version. Obstet Gynecol 2004.
immediately after the procedure identified a bradycardia to
60 beats per minute (bpm) lasting 10 minutes. Over the
next 4 to 5 hours, the fetus continued to have a baseline
tachycardia of 170 bpm, minimal variability, and intermit-
tent decelerations. The decision was made to proceed to
cesarean delivery. There were no difficulties in the delivery,
and no resuscitation was required after the delivery. The
Apgar scores were 8 and 9 at 1 and 5 minutes, respectively,
and the infant weighed 3,085 g.
Immediately after birth, the infant was noticed to have
a floppy and blue left leg. The remainder of the exami-
nation was within normal limits. The orthopedics service
was notified within 4 hours after the birth. Ecchymosis
was present from the knee to the ankle, and crepitus was
palpable at the knee. There was generalized laxity at the
left knee as compared with the right side. A Doppler
revealed normal pedal pulses bilaterally. X-rays ob-
tained revealed a distal metaphyseal corner fracture to
the lateral aspect of the left femur (Fig. 1).
The infant’s course in the hospital was complicated by
respiratory distress syndrome and a short intensive care
Papp et al Fracture After Cephalic Version 1155
unit stay. No further complications were noted. The
fracture itself was treated with cast immobilization. At 3
weeks, an X-ray was repeated and showed abundant
callous formation typical of the quick healing response to
fracture at this age (Fig. 2). The cast was removed.
COMMENT
Femoral fractures occur after vaginal breech deliveries or
difficult deliveries in which the infant has engaged in the
pelvis in the breech position and a forceful cesarean
extraction is required.6 The fractures themselves are
typically found in the shaft of the femur and are spiral in
nature.1–3
Metaphyseal fractures of the long bones (corner frac-
tures) in neonates are considered pathognomonic of
nonaccidental injury due to the force required to gener-
ate this injury. In our patient, leg bruising was noted at
birth suggesting an injury a few hours old. The prompt
discovery of the bruising combined with the atraumatic
cesarean delivery suggests that the fracture occurred in
utero, as a result of the attempted version. X-rays con-
firmed a corner fracture of the distal femur. These me-
taphyseal fractures extend to the growth plate centrally.7
In neonates, this fracture pattern has a low probability of
a growth arrest, but the results of an arrest at this age
would be disastrous. The distal femur has the fastest
growing physeal plate of all the long bones. A growth
arrest at an early age would leave a child with a large
leg-length discrepancy. Regalia et al5 report a femur
fracture in an infant delivered by cesarean 1 week after a
failed version attempt. The authors state that the physi-
cian felt something strange under his hands while at-
tempting the external cephalic version. However, be-
cause the child also had a difficult cesarean delivery, the
authors were not able to say when the fracture occurred.
The type and location of the fracture were not described.
This case report illustrates that a distal metaphyseal
corner type femoral fracture can occur as a result of an
Prenatal Ultrasound Diagnosis
of Poland Syndrome
Dario Paladini, MD,
Maria Rosaria D’Armiento, MD, and
Pasquale Martinelli, MD
Fetal Cardiology Unit, Department of Gynecology and Obstetrics, and Department
of Pathology, University Federico II of Naples, Italy
Address reprint requests to: Dario Paladini, MD, Via Petrarca, 72,
80122 - Naples, Italy; e-mail: paladini@unina.it.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1156 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
external cephalic version attempt. Even in the hands of
an experienced obstetrician, this procedure requires
some controlled force. Obviously, it is difficult to control
the position of the extremities during this procedure,
putting them at risk. We found no other case reports or
discussion on this complication using a MEDLINE
search (1966 to June 2003; keywords: “complication
fracture,” “external cephalic version”) and review of
recent obstetric and orthopedic reference texts. We feel it
is likely that similar injuries have occurred as a result of
external cephalic versions in a small number of cases. It
should be recognized as a potential complication associ-
ated with this procedure.
REFERENCES
1. Morris S, Cassidy N, Stephens M, McCormack D. Birth-
associated femoral fractures: incidence and outcome.
J Pediat Orthopedics. 2002;22:27–30.
2. Awwad JT, Nahhas DE, Karam KS. Femur fracture during
cesarean breech delivery. Int J Gynecol Obstet 1993;43:
324 – 6.
3. Barnes AD, Van Geem T. A. Fractured femur of the new-
born at cesarean section: a case report. J Reprod Med
1985;30:203–5.
4. Curran JS. Birth-associated injury. Clin Perinatol 1981;8:
111–29.
5. Regalia AL, Curiel P, Natale N, Galluzzi A, Spinelli G,
Ghezzi GV, et al. Routine use of external cephalic version in
three hospitals. Birth 2000;27:19 –24.
6. Alexander J, Gregg JEM, Quinn MW. Femoral fractures at
caesarean section: case reports. Br J Obstet Gynaecol 1987;
94:273.
7. Kleiman PK, Marks SC, Blackbourne B. The metaphyseal
lesion in abused infants: a radiologic- histopathologic study.
Am J Roentgenol 1986;146:895–905.
Received December 1, 2003. Received in revised form February 27,
2003. Accepted March 11, 2003.
BACKGROUND: Poland syndrome is a congenital nongenetic
anomaly characterized by unilateral chest wall hypoplasia,
ipsilateral hand abnormalities, and hemivertebrae. It has
not ben described so far in the fetus.
CASE: The patient was referred for suspected left-arm hy-
pomelia at 22 weeks of gestation. On ultrasonography, we
confirmed the presence of severe left-sided hypomelia and
detected an asymmetry of the rib cage and 3 thoracic hemi-
vertebrae. The absence of heart defects led us to make the
putative diagnosis of Poland syndrome. After termination of
pregnancy, the diagnosis was confirmed by the pathologist.
CONCLUSION: The possibility of diagnosing Poland syn-
drome in utero is important for proper management
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000128115.55051.90
and counseling. If the syndrome is suspected in a fetus,
counselors may refer to specific postnatal data to provide
the couple with survival rates, treatment options, and
results and morbidity figures. If the pregnancy is termi-
nated, a detailed necropsy is warranted to confirm the
diagnosis because familial transmission has been re-
ported. (Obstet Gynecol 2004;104:1156 –9. © 2004 by
The American College of Obstetricians and
Gynecologists.)

Fig. 1. A. Oblique scan of the fetal thorax showing the

Poland syndrome is a rare congenital nongenetic anom-
aly with an incidence of 1:7,000 to 1:100,000 live births.1
It represents a sporadic condition, but familial transmis-
sion has been reported in some cases.2 Poland syndrome
is characterized by unilateral chest wall hypoplasia, ipsi-
lateral hand abnormalities, and hemivertebrae and af-
fects the right side of the body in 60 –75% of the cases.1
According to the prevailing pathogenetic theory, it is the
interruption of the embryological blood supply to the
subclavian artery at the end of the 6th week of gestation
that causes the syndrome, with the severity of the anom-
alies depending on the site and the degree of flow impair-
ment.3 In fact, Poland syndrome covers a spectrum of
anomalies ranging from complete absence of pectoralis
major and minor muscles, amastia, severe rib anomalies,
and absence of the hand to minor hand abnormalities
and isolated pectoralis minor muscle aplasia.4 We report
here the prenatal ultrasound diagnosis of Poland syn-
drome made at 22 weeks of gestation in a fetus with
severe left hypomelia and rib cage asymmetry.
CASE
A gravida 2 was referred to our attention at 22 weeks of
gestation for suspicion of left hypomelia in the fetus. By
ultrasonography we confirmed the presence of severe
hypoplasia of the whole left arm and hand (Fig. 1A). In
particular, the clavicle, humerus, ulna, and metacarpals
were highly hypoplastic, and the radius was completely
absent. In addition, we detected an evident asymmetry
of the rib cage (Fig. 1B), with at least 4 hypoplastic ribs
on the right side and the presence of 3 thoracic hemiver-
tebrae (Fig. 1C). The absence of heart defects and the
detection of at least 3 hemivertebrae led us to make the
putative diagnosis of Poland syndrome, considering
the presence of CHILD (Congenital Hemidysplasia with
Ichthyosiform erythroderma and Limb Defects) syn-
drome less likely, although not completely excludible.
An attempt was made to visualize the left subclavian
artery by color Doppler, but the extent of the lesion and
the fixed position of the left hypoplastic arm on the
severe hypomelia. The arm is flexed at the elbow. Note that
no radial and ulnar ossification nuclei are visible in the
forearm and that the single digits cannot be distinguished
because of the oligo-brachy-syndactyly (arrowhead). B. Trans-
verse view of the fetal upper abdomen, showing the hypopla-
sia of the ribs on the right side (arrows). The sonolucent area
regularly positioned in the left hemi-abdomen is the stomach.
C. Coronal view of the upper thoracic spine demonstrating one
of the hemivertebrae (arrowhead).
Paladini. Prenatal Diagnosis of Poland Syndrome. Obstet Gynecol 2004.
thorax impaired proper visualization. After counseling,
the couple opted for termination of pregnancy.
The necropsy of the 426-g female fetus confirmed the
diagnosis of Poland syndrome (Fig. 2). In particular, the
following abnormalities were detected on the babygram:
the left arm was highly hypoplastic, with the hypoplasia
involving the scapulo-humeral joint, clavicle, humerus,
and ulna. The radius was absent. The hand presented
oligo-brachy-syndactyly, with 2 metacarpal rays, agene-
sis of the 5th digit, and syndactyly of the 2nd and the 3rd
ones. The ipsilateral foot had 3 metatarsal rays. As for the
spine, 4 thoracic hemivertebrae were present. At dissection
of the specimen, a pterygium-like attachment of the mal-
formed left arm to the thorax was noted, with athelia
(complete absence of the nipple and areola) and amastia.
The sterno-costal head of the pectoralis major muscle and
Fig. 2. The specimen at birth. Note the hypomelia, the
deformation of the chest, and the oligo-brachy-syndactyly
(magnification of the hand in the lower right-end panel).
Paladini. Prenatal Diagnosis of Poland Syndrome. Obstet Gynecol 2004.

VOL. 104, NO. 5, PART 2, NOVEMBER 2004 Paladini et al Prenatal Diagnosis of Poland Syndrome 1157
the whole of the pectoralis minor muscle were absent. The
rib cage was asymmetrical, with rib abnormalities on the
right side. The heart was unremarkable.
COMMENT
Poland syndrome is an extremely rare condition charac-
terized by hypoplasia of the breast and nipple, absence of
the costo-sternal head of the pectoralis major muscle,
absence of the pectoralis minor muscle, rib abnormali-
ties, and unilateral brachysyndactyly.3 Although these
abnormalities, when severe, are potentially detectable by
ultrasonography in utero, this condition has not been
described in the fetus before. In particular, we could not
find any article in the English literature describing the
occurrence of Poland syndrome in a human fetus in a
MEDLINE search (http://www3.ncbi.nlm.nih.gov/Entrez)
conducted with the terms “Poland syndrome” or “Poland
sequence” and “fetus,” with no time limits.
Differential diagnosis in the fetus is limited to those
conditions characterized by body asymmetry, such as
CHILD syndrome, and to the cluster of anomalies
caused by the occurrence of thoracic hemivertebrae,
such as chest asymmetry and rib hypoplasia.5 However,
in cases of isolated thoracic hemivertebrae, upper limb
malformations are usually absent. The real problem is
represented by CHILD syndrome. This rare X-linked
condition, recognized for the first time in 1968,6 has been
demonstrated to be caused by mutations in the glutathi-
one reductase (NAD"P#H) steroid dehydrogenase-like
protein gene.7 It is characterized by unilateral hypome-
lia, skin hypoplasia, and, often but not always, major
heart defects. Because the erythroderma of CHILD syn-
drome and abnormalities of the chest muscles of Poland
syndrome are not detectable by ultrasonography in
utero, the sonographic appearance of the 2 entities is
similar, both featuring a unilateral limb hypomelia with a
variable degree of arm and hand involvement. The
differential diagnosis is made even more difficult by the
possible occurrence in CHILD syndrome of vertebral
anomalies and rib hypoplasia. A possible hint in the
differential diagnosis issue is the common occurrence in
CHILD syndrome of major heart defects, which are
indeed detectable in utero.8 However, the final diagnosis
can be reached only at autopsy, when anomalies of the
pectoralis muscles and the athelia/amastia are found in
Poland syndrome (as in the index case), and the skin
defects (erythroderma, regional alopecia, etc) become
evident in CHILD syndrome. In conclusion, if unilateral
rib, chest, and arm anomalies are found at routine midtri-
mester ultrasonography, both Poland and CHILD syn-
dromes represent possible diagnoses. In such an occur-
rence, the detection of congenital heart disease might
1158 Paladini et al Prenatal Diagnosis of Poland Syndrome
indicate the likely presence of CHILD syndrome,
whereas the disclosure of thoracic hemivertebrae would
support a diagnosis of Poland syndrome.
The possibility of diagnosing Poland syndrome in
utero is important for proper management and counsel-
ing. Should the syndrome be suspected in a fetus, the
counselors may refer to specific postnatal data to provide
the couple with survival rates, treatment options, and
results and morbidity figures. In case of termination of
pregnancy, the prenatal recognition of the syndrome
strengthens the need for a detailed necropsy to confirm
the diagnosis because familial transmission has been
reported in some instances.2 In these families, the ability
to reassure the couple that the reoccurrence of Poland
syndrome can be excluded during the second trimester
of pregnancy is of the utmost importance. However,
taking into consideration the variable expression of the
syndrome and the ultrasound findings in the index case,
2 further points need be addressed: 1) It is likely that, in
high-risk families, full-blown forms of the syndrome
could be detected by ultrasonography as early as the
12th–13th gestational weeks, whereas mild forms, fea-
turing only pectoralis minor muscle hypoplasia/aplasia
and subtle digit anomalies, may escape prenatal ultra-
sound diagnosis completely. 2) Ultrasound differential
diagnosis between Poland and CHILD syndromes may
be impossible in some cases, with the final diagnosis
based only upon postmortem findings.
REFERENCES
1. Freire-Maia N, Chautard EA, Opitz JM, Freire-Maia A,
Queice-Salgado A. The Poland syndrome: clinical and
genealogical data, dermatoglyphic analysis, and incidence.
Hum Hered 1973;23:97–104.
2. Sujansky E, Riccardi VM, Mathew AL. The familial occur-
rence of Poland syndrome. Birth Defects Orig Artic Ser
1977;13:117–21.
3. Bavinck JN, Weaver DD. Subclavian artery supply disrup-
tion sequence: hypothesis of a vascular etiology for Poland,
Klippel-Feil, and Möbius anomalies. Am J Med Genet 1986;
23:903–18.
4. Fokin AA, Robicsek F. Poland’s syndrome revisited. Ann
Thorac Surg 2002;74:2218 –25.
5. Fraser FC, Teebi AS, Walsh S, Pinsky L. Poland sequence
with dextrocardia: which comes first? Am J Med Genet
1997;73:194 – 6.
6. Falek A, Heath CW Jr, Ebbin AJ, McLean WR. Unilateral
limb and skin deformities with congenital heart disease in
two siblings: a lethal syndrome. J Pediatr 1968;73:910 –3.
7. Konig A, Happle R, Fink-Puches R, Soyer HP, Bornholdt
D, Engel H, et al. A novel missense mutation of NSDHL in
an unusual case of CHILD syndrome showing bilateral,
OBSTETRICS & GYNECOLOGY
 almost symmetric involvement. J Am Acad Dermatol 2002;
46:594 – 6.
8. Paladini D, Russo MA, Teodoro A, Pacileo G, Capozzi G,
Martinelli P, et al. Prenatal diagnosis of congenital heart
disease in the Naples area during the years 1994 –1999: the
Buttock Necrosis After Uterine
Artery Embolization
Danielle M. Dietz, MD, Kurt R. Stahlfeld, MD,
Surendra K. Bansal, MD, and
Wayne A. Christopherson, MD
Departments of Surgery, Radiology, and Obstetrics and Gynecology, The Mercy
Hospital of Pittsburgh, Pittsburgh, Pennsylvania
BACKGROUND: Uterine artery embolization is an increas-
ingly popular alternative to hysterectomy or myomectomy
for treatment of symptomatic uterine leiomyomata.
CASE: A woman with a symptomatic uterine fibroid devel-
oped 2 areas of full-thickness necrosis on her right buttock
following uterine artery embolization. After surgical de-
bridement, healing occurred over 14 weeks.
CONCLUSION: Buttock necrosis is a possible complication of
nontarget embolization during uterine artery emboliza-
tion. (Obstet Gynecol 2004;104:1159 – 61. © 2004 by
The American College of Obstetricians and
Gynecologists.)
Uterine artery embolization is an increasingly used alter-
native to myomectomy or hysterectomy for the treat-
ment of symptomatic uterine leiomyomata. Reported
complications of nontarget embolization during uterine
artery embolization include ovarian failure,1 labial ne-
crosis,2 and unilateral buttock pain.3 We present a case
of nontarget embolization resulting in full-thickness but-
tock necrosis requiring surgical debridement.
CASE
A 46-year-old multiparous woman with a history of mor-
bid obesity (body mass index 56 kg/m2), hypertension,
asthma, and a known uterine leiomyoma presented for
uterine artery embolization for increasing leiomyoma size.
Pelvic examination was limited because of the patient’s size,
and the cervix was not visible on speculum examination.
An anterior, subserosal myoma, measuring 19 ! 10.2 ! 12
cm, was diagnosed by ultrasonography.
Address reprint requests to: Danielle M. Dietz, MD, Department of
Surgery, The Mercy Hospital of Pittsburgh, 1400 Locust Street, Pitts-
burgh, PA 15219; e-mail: danielledietz@hotmail.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
experience of a joint fetal-pediatric cardiology unit. Prenat
Diagn 2002;22:545–52.
Received November 3, 2003. Received in revised form January 20,
2004. Accepted February 5, 2004.
Bilateral selective uterine artery embolization with
500 –700 #m Microspheres (Boston Medical, Boston,
MA) was performed in accordance with the Society of
Cardiovascular and Interventional Radiology protocol
by an experienced interventional radiologist. Angiogra-
phy revealed stasis of flow, with no evidence of embolic
reflux out of either uterine artery (Fig. 1).
The patient was admitted to the hospital for observa-
tion and pain control. Postprocedure, the patient had
pelvic pain requiring ketorolac and parenteral narcotics.
She was able to tolerate oral pain medication and was
discharged on postprocedure day 2.
Two weeks postprocedure, the patient was complain-
ing of buttock pain when sitting or lying supine. On
examination, 2 large ecchymotic areas on the right but-
tock were noted. These patches had intact skin and
appeared to be superficial. At follow-up 2 weeks later,
however, the skin had broken down revealing significant
areas of underlying necrosis (Fig. 2).
The patient was admitted to the hospital for intrave-
nous broad-spectrum antibiotics and operative debride-
ment. The necrotic lesions extended 6 cm deep to super-
ficially involve the gluteus maximus muscle. She was
discharged home on postoperative day 2 with wound
irrigation 3 times daily and wet-to-dry gauze dressings.
Over the next 14 weeks, the lesions healed with con-
tinued conservative wound management. The wounds
re-epithelialized, and the patient has no functional limi-
tations. She requires only acetaminophen for pain con-
trol. Postembolization, she has had one episode of light
menses. The uterine leiomyoma has decreased in vol-
ume approximately 65% and now measures 9.2 ! 9.4 !
9.4 cm on transvaginal ultrasonography.
COMMENT
Because of the time relation to the uterine artery embo-
lization and clinical appearance of the lesions (2 spatially
separate full-thickness areas of necrosis), we believe the
patient’s buttock injuries were caused by nontarget em-
bolization during therapeutic uterine artery emboliza-
tion. The main blood supply to the gluteus maximus
muscle and overlying soft tissues is the superior gluteal
artery, the terminal branch of the posterior division of
the internal iliac artery. The inferior gluteal artery typi-
0029-7844/04/$30.00 1159
doi:10.1097/01.AOG.0000141567.25541.26

Fig. 1. Angiogram of left internal iliac arterial system
during uterine artery embolization. Pictured are the inter-
nal iliac artery (A), posterior division of the internal iliac
artery (B), anterior division of the internal iliac artery (C),
branches of the inferior gluteal artery (D), branches of the
superior gluteal artery (E), and uterine artery (arrows).
Dietz. Nontarget Buttock Embolization. Obstet Gynecol 2004.
cally originates from the anterior division of the inter-
nal iliac artery, in proximity to the uterine artery. The
region supplied by the inferior gluteal artery is vari-
able, often including the skin and muscles dorsal to the
Fig. 2. Buttock lesions appearing after uterine artery
embolization. The lateral lesion measures 5.5 ! 5.0 cm,
and the gluteal fold lesion measures 12.0 ! 6.5 cm.
Dietz. Nontarget Buttock Embolization. Obstet Gynecol 2004.
1160 Dietz et al Nontarget Buttock Embolization
sacrum and extending down the buttock to the poste-
rior thigh.
Buttock ischemia from nontarget embolization most
likely would occur as a direct result of reflux of embolic
material from the uterine artery into the inferior gluteal
artery. It is also possible that these spheres could migrate
through anastomotic channels that exist between the
superior or muscular branches of the inferior gluteal
artery and the superficial branch of the superior gluteal
artery. This is supported by the documentation by
Payne et al4 of polyvinyl alcohol particles passing unde-
tected through anastomotic channels between the ovar-
ian and uterine vessels.
To minimize nontarget embolization during injec-
tion, specific techniques are recommended by the Society
of Cardiovascular and Interventional Radiology (Siskin
GP. Gynecologic interventions. SCVIR 2002 Work-
shop Book, 27th Annual Scientific Meeting, Baltimore,
Maryland, April 6 –11, 2002. Fairfax [VA]: The Society
of Cardiovascular & Interventional Radiology, 2002).
The embolization catheter should be placed in the trans-
verse segment of the uterine artery, distal to the cervico-
vaginal branch (if visualized). Recommended embolic
agents are polyvinyl alcohol particles, microspheres, or
gelatin sponge. Power injection is associated with ex-
treme pressures and should be avoided. Hand injection
of embolic particles is performed under direct fluoro-
scopic visualization until complete stasis of flow to the
level of the catheter is seen.
Before our patient, there has been one reported case of
presumed gluteal muscle nontarget embolization (Ovid
MEDLINE search, 1966 to March 6, 2004, all languages,
keywords: “uterine artery embolization” or “therapeutic
embolization” and “uterine hemorrhage” or “leiomy-
oma”) that manifested as deep buttock pain.3 If embolic
spheres occlude the terminal branches of the superior
or inferior gluteal arteries, where no collateral blood flow
is present, significant skin and subcutaneous necrosis
may occur. Our case demonstrates that this is potentially
a significant complication. Initial conservative manage-
ment failed, and the patient required surgical debride-
ment. The affected areas were eventually reperfused
from the extensive pelvic collateral blood supply, and
the wounds then healed secondarily over an extended
period of time.
REFERENCES
1. Amato P, Roberts AC. Transient ovarian failure: a compli-
cation of uterine artery embolization. Fertil Steril 2001;75:
438 –9.
2. Yeagley TJ, Goldberg J, Klein TA, Bonn J. Labial necrosis
OBSTETRICS & GYNECOLOGY
 after uterine artery embolization for leiomyomata. Obstet
Gynecol 2002;100:881–2.
3. Hutchins FL Jr, Worthington-Kirsch R. Embolotherapy for
myoma-induced menorrhagia. Obstet Gynecol Clin North
Am 2000;27:397– 405.
4. Payne JF, Robboy SJ, Haney AF. Embolic microspheres
Necrotic Leiomyoma and Gram-
Negative Sepsis Eight Weeks
After Uterine Artery
Embolization
Matthew Aungst, MD, Mark Wilson, MD,
Karen Vournas, MD, and
Sarah McCarthy, MD
Department of Obstetrics and Gynecology, David Grant Medical Center, Travis
Air Force Base, California; Department of Interventional Radiology, University
of California, San Francisco, California; Department of Obstetrics and Gynecology,
Kaiser Permanente, Vallejo, California.
BACKGROUND: Uterine artery embolization for symptom-
atic leiomyomata is generally safe, but rare life-threatening
complications, including sepsis, can result.
CASE: A 39-year-old woman with primary antiphospho-
lipid syndrome, who was on chronic warfarin therapy,
underwent uterine artery embolization for severe menor-
rhagia and a 12-cm intracavitary leiomyoma. Eight weeks
postembolization, the patient, who had been essentially
asymptomatic, presented in septic shock from gram-nega-
tive anaerobic bacteria. She underwent hysterectomy and
bilateral salpingo-oophorectomy for a large infarcted ne-
crotic leiomyoma and partial uterine necrosis. The pa-
tient’s 8-day hospitalization required extended care in the
intensive care unit and blood transfusion and resulted in
surgical menopause in a patient who is not a candidate for
hormone therapy.
CONCLUSION: Uterine artery embolization is a procedure
not without significant risks. From published case reports,
it appears that patients most at risk for severe infection of
an infarcted leiomyoma after this procedure are those with
a large dominant leiomyoma. (Obstet Gynecol 2004;104:
1161– 4. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
Address reprint requests to: Matthew Aungst, Captain, 60 MSGS/
SGCG, 101 Bodin Circle, Travis AFB, CA 94535; e-mail:
Mathew.aungst@60mdg.travis.af.mil.
No funding was received for this project.
The views expressed in this material are those of the authors and do not reflect the
official policy or position of the U.S. Government, the Department of Defense, or the
Department of the Air Force.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
within ovarian arterial vasculature after uterine artery
embolization. Obstet Gynecol 2002;5:883– 6.
Received March 12, 2004. Received in revised form April 12, 2004.
Accepted April 29, 2004.
Embolization of the uterine arteries is being performed at
increasing rates as women seek conservative therapy for
symptomatic leiomyomata. Improvements in menorrha-
gia, dysmenorrhea, and pressure symptoms are reported
to reach or exceed 80%.1 Common complications of the
procedure include expulsion of submucosal leiomyo-
mata, allergic reaction to intravenous contrast dye, and
prolonged pain. Approximately 90% of patients experi-
ence no complications, and serious complications are
reportedly rare.2
CASE
A 39-year-old married multigravida presented with 4
days of vaginal bleeding and cramping and 1 day of fever
and chills. Eight weeks before this, the patient had un-
dergone uterine artery embolization of a large 12 ! 11 !
9 – cm intracavitary fundal leiomyoma. Before emboliza-
tion the patient’s menorrhagia had resulted in chronic
anemia requiring transfusion on 1 occasion and severe
fatigue. Her fatigue had markedly improved in the weeks
after the embolization procedure, and she felt well at her
postoperative follow-up visit, which occurred 5 days
before admission. The patient’s medical history was
significant for primary antiphospholipid syndrome,
which was diagnosed after a cerebrovascular accident at
age 37 years. The patient had no residual deficits from
this right-parietal-lobe infarction and was maintained on
chronic warfarin therapy with an international normal-
ized ratio (INR) of 3.0. She was considered a poor
surgical candidate because of a significant risk of periop-
erative thrombotic events.
The uterine artery embolization procedure was per-
formed with a microcatheter (Cordis, Miami, FL) manip-
ulated into each uterine artery. The embolic agent con-
sisted of 500 –710 #m polyvinyl alcohol particles (Target
Theraputics, Fremont, CA), which were suspended in
nonionic contrast medium (Omnipaque 300; Amersham
Health, Princeton, NJ). This suspension was subse-
quently injected through the microcatheter until stasis of
flow was achieved radiographically in both uterine arter-
ies. Five 2-mL vials of polyvinyl alcohol particles were
used for the entire procedure. The patient received doxy-
cycline before the procedure and for a week postproce-
dure. Five days before the embolization procedure, the
0029-7844/04/$30.00 1161
doi:10.1097/01.AOG.0000128107.58898.e1
Fig. 1. Computed tomography of pelvis showing air pock-
ets (arrow) in necrotic leiomyoma.
Aungst. Sepsis After Uterine Artery Embolization. Obstet Gynecol 2004.
patient was taken off warfarin and placed on enoxaparin.
The enoxaparin therapy was stopped the morning of the
procedure but restarted later that evening. The day after
the procedure, warfarin was restarted and the patient
was discharged. She continued the enoxaparin until her
INR was therapeutic.
The patient was well after the embolization procedure
until the 4 days before her admission, when she devel-
oped pelvic cramps and heavy vaginal bleeding. The
initial evaluation in the emergency department found the
patient to be in septic shock, with a temperature of
39.78°C, heart rate of 163 beats per minute, blood pres-
sure of 99/60 mm Hg, and anuric. Her examination was
notable for a moderately tender uterus and intermit-
tently waning consciousness. The patient’s white blood
cell count was 1,400 with 37% bands; her hemoglobin
was 8.5 mg/dL and INR was 3.0. The peripheral smear
demonstrated findings consistent with disseminated in-
travascular coagulation. Liver function studies were ele-
vated with aspartate transaminase of 129 IU/L and ala-
nine transaminase of 101 IU/L. Computed tomography
of the abdomen and pelvis (Fig. 1) showed a thick-walled
uterus with a 10-cm intracavitary mass containing air
pockets.
Intravenous antibiotics were started, as was aggressive
fluid resuscitation. Her condition deteriorated despite
1162 Aungst et al Sepsis After Uterine Artery Embolization
Fig. 2. Gross photo of bivalve uterine specimen containing
the infarcted necrotic leiomyoma.
Aungst. Sepsis After Uterine Artery Embolization. Obstet Gynecol 2004.
crystalloids and vasopressors, and she required emer-
gent transfusion of O Rh-negative packed red cells while
awaiting type-specific blood. The patient’s homody-
namic status stabilized after receiving 5 units of packed
red blood cells. After reversal of her anticoagulation with
vitamin K and 5 units of fresh frozen plasma, the patient
underwent total abdominal hysterectomy and bilateral
salpingo-ophorectomy. The 18-week uterus was re-
moved intact and grossly showed a 10-cm infarcted,
necrotic leiomyoma (Fig. 2). The patient’s sepsis syn-
drome began to improve after removal of the specimen,
and the vasopressors were discontinued before conclu-
sion of the procedure. Full heparin anticoagulation was
started 12 hours after her hysterectomy, and she was
continued on antibiotics. Blood cultures obtained upon
initial admission grew the anaerobic gram-negative rods
Fusobacterium necrophorum and Corynebacterium, and cultures
of the purulent ascites obtained at surgery grew the
anaerobic gram-negative rods Bacteroides. The patient
was transferred from the intensive care unit on postop-
OBSTETRICS & GYNECOLOGY
Table 1. Reported Cases of Uterine and Pelvic Infection After Uterine Artery Embolization
Time from
Largest embolization
Report myoma (cm) to infection Therapy Complications
Aungst et al 12 8 wk TAH/BSO, intravenous Sepsis, transfusion,
antibiotics hospitalization (8 d)
Vashisht et al10 14 7d TAH/BSO Sepsis, death
Mehta et al4 9 29 wk Intravenous antibiotics
16 5 wk Intravenous antibiotics, manual Hospitalization (20 d)
extraction of myoma
10 1 wk Intravenous antibiotics Ovarian failure
10 3 wk Intravenous antibiotics Hospitalization (21 d)
7 20 wk Failed intravenous antibiotics TAH
Not stated 1 wk Intravenous antibiotics, manual Hospitalization (16 d)
extraction of myoma
10 3 wk Intravenous antibiotics
Pron et al5 7 12 h Failed intravenous antibiotics TAH/RSO
10 18 d Failed intravenous antibiotics TAH
Walker and Pelage1 11 10 d Failed intravenous antibiotics TAH/BSO
14 3 mo TAH
11 5d Failed intravenous antibiotics TAH
Spies et al2 Not stated 10 wk Intravenous antibiotics
Not stated Not stated Hysteroscopic myomectomy
Payne and Hayney6 14 4d TAH Transfusion
Robson et al8 6 24 d Failed intravenous antibiotics Exploratory laparotomy,
BSO, sepsis
Al-Fozan and Tulandi9 Not stated 12 d Intravenous antibiotics, TAH/ Sepsis
BSO
Godfrey and Zbella7 17 3 mo TAH/LSO
TAH ' total abdominal hysterectomy; BSO ' bilateral salpingo-oophorectomy; RSO ' right salpingo-oophorectomy; LSO ' left salpingo-
oophorectomy.

erative day 4, and her further postoperative course was
uneventful until her discharge on postoperative day 8.
COMMENT
The high risk of a perioperative thrombotic event, 40 –
80% in patients with antiphopholipid syndrome who are
undergoing major surgery without anticoagulation,3 was
the primary motivation for this patient and her providers
to seek a nonsurgical therapy for her symptomatic uter-
ine leiomyomata. The embolization procedure itself was
associated with a briefly elevated risk of a thrombotic
event resulting from arterial catheterization and a short
period of anticoagulation reversal. However, this risk
was perceived to be considerably less than that associ-
ated with pelvic surgery and the prolonged period of
postoperative inactivity. Indeed the vast majority of pa-
tients who undergo uterine artery embolization have
no complications.2 Emergency hysterectomy for infec-
tion is estimated to complicate only approximately 1%
of embolization procedures.1
However, it appears from the reviewed cases that
serious infection after uterine artery embolization is
more common after embolization of uteri with large
myomata (Table 1). Most reported cases have involved
endometritis and have resolved with intravenous antibi-
otics alone. Mehta et al4 described 7 patients requiring
readmission for infection during a period of 1 to 29
weeks after uterine artery embolization of large leiomy-
omata. Three patients improved with antibiotic therapy
alone, 3 patients required vaginal removal of an ex-
truded leiomyoma in addition to antibiotics, and 1 pa-
tient failed to improve with antibiotics and required
hysterectomy. The Canadian Multicenter Clinical Trial5
of 555 women undergoing embolization found 8 cases
requiring hysterectomy for postembolization complica-
tions at 3 months of follow-up. Two of the 8 hysterecto-
mies were performed for infection-related morbidities.
Both of theses cases required hysterectomy after intrave-
nous antibiotics failed to resolve endometritis after em-
bolization of 7- and 10-cm leiomyomata. In Walker and
Pelage’s1 series of 400 uterine artery embolizations, 3
women required hysterectomy for infection refractory to
antibiotic therapy. A series of 400 consecutive emboliza-
tions by Spies et al2 lists 2 cases of postembolization
endometritis, neither necessitating hysterectomy. Payne
and Haney6 and Godfrey and Zbella7 each described
single cases of endometritis after embolization necessitat-
ing hysterectomy. There are no guidelines indicating

VOL. 104, NO. 5, PART 2, NOVEMBER 2004 Aungst et al Sepsis After Uterine Artery Embolization 1163
when medical therapy should be abandoned and hyster-
ectomy performed. However, it seems reasonable to
continue medical therapy in hemodynamically stable
patients who show a response to therapy and who re-
main motivated to avoid hysterectomy.
Sepsis after uterine artery embolization has been less
frequently reported than endometritis. In Australia, a
case of “pelvic sepsis” occurred 24 days after the original
embolization procedure.8 A 6-cm submucosal leiomy-
oma was expelled from the cervix 21 days postproce-
dure. Three days later the patient presented with fever
and abdominal pain and purulent peritoneal fluid was
found at laparotomy. The patient recovered after bilat-
eral salpingectomy and antibiotic therapy. The patient’s
cervical swabs grew Escherichia coli and Trichomonas. Al-
Fozan and Tulandi9 also reported a case of sepsis after
uterine artery embolization resulting in hysterectomy. A
devastating case of sepsis and disseminated intravascular
coagulation is described occurring in a 51-year-old par-
ous women 10 days after embolization of a 14 ! 12 !
11-cm submucosal fibroid.10 The woman underwent
hysterectomy but died as a result of the sepsis 15 days
later.
With respect to myoma size and time from emboliza-
tion to presentation, our case is very similar to the
previously reported cases complicated by infection. Of
the cases listed in Table 1, the average diameter of the
largest myoma was 11 cm (range 6 –14 cm), and the
average time from embolization to presentation was 6
weeks (range 12 hours to 29 weeks). Limiting emboliza-
tion to uteri with a myoma less than 8 cm in the largest
diameter has been suggested by some authors to reduce
the risk of infection.9,11 It has also been suggested that
submucosal myomata are more likely to cause infection
and should not be embolized.11
Although uterine artery embolization for symptom-
atic leiomyomata appears to have a low risk of compli-
cations for most women, significant infectious morbidity
is reported. Because women with large dominant myo-
mata may be at increased risk of postprocedure endome-
tritis, careful patient selection and counseling, followed
1164 Aungst et al Sepsis After Uterine Artery Embolization
by close surveillance, is warranted in women undergoing
embolization of large leiomyomata.
REFERENCES
1. Walker WJ, Pelage JP. Uterine artery embolization for
symptomatic fibroids: clinical results in 400 women with
imaging follow up. BJOG 2002;109:1262–72.
2. Spies JB, Spector A, Roth AR, Baker CM, Mauro L,
Murphy-Skrynarz K. Complications after uterine artery
embolization for leiomyomas. Obstet Gynecol 2002;100:
873– 80.
3. Geerts WH, Heit JA, Clagett GP, Pineo GF, Colwell CW,
Anderson FA, et al. Prevention of venous thromboembo-
lism. Chest 2001;119:132S–75S.
4. Mehta H, Sandhu C, Matson M, Belli AM. Review of
readmissions due to complications from uterine fibroid
embolization. Clin Radiol 2002;57:1122– 4.
5. Pron G, Mocarski E, Cohen M, Colgan T, Bennett J,
Common A, et al. Hysterectomy for complications after
uterine artery embolization: results of a Canadian multi-
center clinical trial. J Am Assoc Gynecol Laparosc 2003;
10:99 –106.
6. Payne JF, Haney AF. Serious complications of uterine
artery embolization for conservative treatment of fibroids.
Fertil Steril 2003;79:128 –31.
7. Godfrey CD, Zbella EA. Uterine necrosis after uterine
artery embolization for leiomyoma. Obstet Gynecol 2001;
98:950 –2.
8. Robson S, Wilson K, Munday D, Sebben R. Pelvic sepsis
complicating embolization of a uterine fibroid. Aust N Z J
Obstet Gynaecol 1999;39:516 –7.
9. Al-Fozan H, Tulandi T. Factors affecting early surgical
intervention after uterine artery embolization. Obstet
Gynecol Surv 2002;57:810 –5.
10. Vashisht A, Studd J, Carey A, Burn P. Fatal septicemia
after fibroid embolisation. Lancet 1999;354:307– 8.
11. McLucas B, Adler L, Perrella R. Uterine fibroid emboliza-
tion: nonsurgical treatment for symptomatic fibroids. J Am
Coll Surg 2001;192:95–105.
Received December 19, 2003. Received in revised form February 5,
2004. Accepted February 18, 2004.
OBSTETRICS & GYNECOLOGY
Rectovaginal Fistula Repair
Using a Porcine Dermal Graft
Robert D. Moore, DO, John R. Miklos, MD,
and Neeraj Kohli, MD
Atlanta Urogynecology Associates, Atlanta, Georgia, and Brigham and Women’s
Hospital, Harvard Medical School, Boston, Massachusetts
BACKGROUND: Rectovaginal fistula repair is commonly per-
formed through a vaginal route. In many cases, healthy
tissue such as an autologous fat pad may be interposed
between the suture lines and the vaginal epithelium to
facilitate healing and prevent recurrence. We present a
simple alternative to autologous flaps with the use of por-
cine dermal grafts in the repair of rectovaginal fistulae.
CASES: Two patients are presented with rectovaginal fistu-
lae. In both cases the patients were found to have insuffi-
cient native tissue to achieve an adequate traditional mul-
tilayered closure, and therefore an acellular collagen
porcine dermal graft was used as an interposition graft
between the rectum and the vaginal epithelium in the
repair.
CONCLUSION: Porcine dermal grafts may be a viable alter-
native to traditional autologous flaps or human dermal
grafts for the repair of rectovaginal fistula (Obstet Gy-
necol 2004;104:1165–7. © 2004 by The American College
of Obstetricians and Gynecologists.)
Despite the use of preoperative antibiotics and meticu-
lous surgical techniques, rectovaginal fistula repair con-
tinues to be a challenge for the gynecologic or colorectal
surgeon. Repair can be accomplished through the va-
gina, perineum, sphincter, rectum, or abdomen (with the
abdomen usually the choice of approach for high fistu-
lae). Regardless of approach, standard principles should
be used to repair fistulae, including excision of the epi-
thelialized tract, complete closure of the rectal opening
with clean healthy tissue margins, adequate tissue mobi-
lization, tension-free multilayer closure, and hemosta-
sis.1 Recurrent or complicated fistulae, such as those
associated with irradiation, often require a more compli-
cated operation. This may include interposing healthy
tissue between layers (such as a pedicle graft from omen-
tum, peritoneum, muscle, or fat pad) to facilitate healing.
Additionally, patients found to have very poor tissue
quality or lacking sufficient native tissue to accomplish
an adequate multilayer repair may also benefit from the
use of an interposition graft. The use of human dermal
allografts in this type of repair of rectovaginal fistula has
Address reprint requests to: Robert D. Moore, DO, Atlanta Urogyne-
cology Associates, 3400-C Old Milton Parkway, Suite 330, Alpharetta,
GA 30005; moorer33@hotmail.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
been reported.2 We report the successful repair of a
rectovaginal fistula using porcine dermal graft as an
interposition material in 2 patients with rectovaginal
fistulae.
CASE 1
A 39-year-old multigravida presented after being in-
formed of the presence of a possible rectovaginal fistula.
She stated that she had been passing gas vaginally and
having brownish/green vaginal discharge for years since
the vaginal birth of her first child in southeast Asia 10
years prior. This was a vaginal delivery of a 8-lb, 8-oz
child, complicated by what she recalled as being a very
large vaginal laceration. Physical examination, as well as
barium enema, confirmed a rectovaginal fistula, approx-
imately 5 mm in size, 3 cm inside of vaginal introitus in
the lower third of the vagina. The anal sphincter was
intact, and there was no evidence of any other fistulous
tracts. The patient was placed on a clear liquid diet for 24
hours before surgery, completed a sodium phosphate
enema the night before surgery, and a laxative supposi-
tory the morning of surgery. Antibiotic prophylaxis was
given preoperatively. The patient opted for general an-
esthesia and was placed in the dorsal lithotomy position,
and the rectovaginal fistula was identified. A vaginal
incision was made, circumscribing the fistula and ensur-
ing healthy tissue margins around it. Vertical incisions
were then made cranially and caudally in the vaginal
epithelium, extending away from the circular incision
and increasing access to the subepithelial plane. The
vaginal epithelium was then mobilized in all directions
from the underlying rectovaginal fascia. The fistulous
tract was then excised all the way down through the
rectum, incorporating the vaginal epithelium, fistulous
tract, and rectal mucosa, as well as scar tissue surround-
ing the tract. Because necrotic tissue was encountered in
the dissection and healthy viable tissue margins were
needed, a large defect (approximately 2 cm) was left to
close. A 2-layer closure of the rectum was attempted with
interrupted 3/0 absorbable suture in a tension-free clo-
sure. The second layer imbricated the first and was
placed in the same plane, but it was very difficult to
obtain healthy tissue for the second layer because the
patient was found to have very tenuous and poor quality
tissue. We felt she was at high risk of breakdown of the
repair. Therefore, before closing the vaginal epithelium, an
acellular 3 ! 3 cm porcine dermal graft was placed over the
incision line and anchored laterally to the rectovaginal
septum with 2/0 absorbable suture. The vaginal epithelium
was then closed over the graft with interrupted 2/0 absorb-
able suture. The patient was discharged home in 24 hours
and had an uncomplicated postoperative course. She was
0029-7844/04/$30.00 1165
doi:10.1097/01.AOG.0000128116.52051.10
placed on a clear liquid diet for 3 days, followed by a low
residue diet for 3 weeks and bowel regimen including
softeners for 6 weeks. Intercourse was discouraged for 12
weeks. The patient has been followed for 18 months with-
out recurrence of the fistula or complication from place-
ment of the porcine graft. She has resumed sexual inter-
course and reports no dyspareunia.
CASE 2
A 35-year-old woman presented with complaints of fecal
incontinence approximately 3 months after a normal
spontaneous vaginal delivery complicated by a third-
degree laceration. Physical examination, including pelvic
examination, was normal except for a 3-mm rectovaginal
fistula slightly to the right of midline and 1 cm proximal
to the hymenal ring. The perineal body was shortened,
and rectal examination revealed decreased external anal
sphincter tone. Dye instillation into the rectum excluded
coexisting fistula tracts. The patient underwent uncom-
plicated fistula repair in a fashion similar to that de-
scribed in Case 1, with concurrent external anal sphinc-
teroplasty. An interpositional porcine dermal graft was
placed after very poor tissue quality was noted during
dissection. There was essentially no rectovaginal fascia
found that could be used for a second-layer closure of the
rectum, and attempts to go out laterally farther to the
side walls would have caused significant narrowing of
her introitus. Therefore, a porcine graft was placed as an
additional layer for the repair and to interrupt the suture
lines. She was discharged on postoperative day 1 with
conservative bowel management. On postoperative fol-
low-up, there was complete resolution of the rectovagi-
nal fistula, with good wound healing. At the 6-month
postoperative visit, the examination was normal, and the
patient is without complaints of fecal incontinence or
urgency. She has resumed sexual intercourse and denies
dyspareunia.
COMMENT
Surgical correction of the rectovaginal fistula can be a
complicated procedure, even for the advanced pelvic
surgeon. Simple fistula of the mid and lower vagina are
most often treated with transrectal advancement flap or
transvaginal layered closure techniques and are associ-
ated with good surgical results. However, complicated or
recurrent fistulae often require additional procedures,
with meticulous surgical technique and tissue handling.
The cause of recurrent fistulae is controversial, but su-
ture line breakdown may be due to infection caused by
contamination from stool or vaginal bacteria, not having
a tension-free closure with healthy tissue margins, or
1166 Moore et al Fistula Repair With Porcine Dermal Graft
having overlying suture lines in the same direction on
top of each other. Patients who are found to have very
poor tissue quality or inadequate “native” tissue to
achieve a multilayer closure may also be at risk of failure
or recurrence. Various methods of interpositioning of
healthy tissue between the suture lines to help reduce the
risk of failure or recurrence have been described. The
theory is that the use of a bulbocavernosus fat pad
(Martius flap) or gracilis muscle flap may result in en-
hanced blood supply to devascularized epithelium, oblit-
eration of dead space, and the interruption of suture lines
along the length of multilayer closure. These techniques,
however, involve additional surgery, advanced surgical
skills, and can be associated with increased morbidity.
We have previously described the use of a cadaveric
dermal allograft in the surgical repair of rectovaginal
fistula as a simplified alternative to autologous interposi-
tional flaps.2 This approach minimizes the need for
additional surgery, is based on common anatomical prin-
ciples well known to the gynecologic surgeon, and offers
the advantage of an interpositional graft between the
suture lines to help prevent recurrence. This may be
especially helpful in patients who are found to have poor
tissue quality or an absence of healthy rectovaginal fascia
to use in a multilayer closure. Compared with traditional
autologous flaps, a cadaveric allograft is readily avail-
able, cost-effective, and associated with minimal compli-
cations or morbidity. Their use has also been recently
described in complicated pelvic floor repairs,3,4 as well in
augmenting rectocele repairs to help improve cure rates
and achieve a more anatomical repair.5 A very similar
anatomic technique is used when placing the graft during
a fistula repair.
The use of porcine dermis in the repair of rectovaginal
fistulae has not been previously reported (MEDLINE
and PubMed searches with the terms “rectovaginal fis-
tula,” “dermal graft,” and “allograft,” for the years
1940 –2004, in all languages). Pelvicol (Bard Urology,
Covington, GA) porcine dermis is a natural acellular,
nonallergenic matrix that has emerged as an alternative
to cadaveric dermis for pelvic floor repairs. Porcine
dermis offers the same advantages as cadaveric dermal
grafts in repairs, as well as these additional advantages:
offers an essentially unlimited supply, is easily stored
(does not require refrigeration), handles very easily, has
excellent strength properties, is readily incorporated into
host tissue, and has decreased risk of bacteria or viral
transmission. Pelvicol has been used for permanent im-
plantation in humans since 1998. It has been used exten-
sively in dermal grafting for burn victims and other areas
of plastic, ear, nose, and throat, and general surgery. A
recent report described its use in 60 different surgical
procedures in 140 patients.6 There is limited data regard-
OBSTETRICS & GYNECOLOGY
ing porcine graft use in gynecologic surgery, but initial
studies have reported its use in anterior and posterior
repair, as well as in pubovaginal slings (Graul ES, Hurst
B. Porcine allograft in the repair of anterior and posterior
vaginal defects [abstract]. Int Urogynecol J Pelvic Floor
Dysfunct 2002;13 suppl:S36).7,8 Our initial experience
with porcine dermis in prolapse repair has been encour-
aging, and the indications for its use continue to increase.
We have found it to be an excellent material for pelvic
floor repairs and have had no significant intraoperative
or postoperative complications (Moore RD, Kohli N,
Miklos JR. Laparoscopic uterosacral colpoperineopexy
utilizing dermal graft for vaginal vault prolapse [ab-
stract]. Int Urogynecol J Pelvic Floor Dysfunct 2002;13
suppl:S39).
The use of a porcine dermal graft in repair of recto-
vaginal fistula is a method that offers a simplified ap-
proach for placing an interpositional graft with minimal
risk or morbidity, giving the surgeon the advantages of
interruption of the suture lines without the extra surgical
risks of bringing in an autologous tissue flap. Consider-
ation of the use of a dermal graft in the repair of recto-
vaginal fistula should be given to patients who have a
higher risk of failure, including patients found to have
poor tissue quality that inhibits the surgeon in achieving
an adequate multilayer closure. Their use should also be
considered in recurrent or complicated fistulae or in a
patient where a previous autologous tissue flap has
failed. A porcine dermal graft should be avoided in
patients known to have a hypersensitivity to porcine
products or noted to have an active infection. A greater
number of cases using porcine dermal grafts in rectovagi-
nal fistula repairs is needed before these grafts can be
recommended in primary repairs to reduce risk of failure
or in more complicated repairs to avoid having to use
Port-Site Implantation After
Laparoscopic Treatment of
Borderline Ovarian Tumors
Philippe Morice, MD, Sophie Camatte, MD,
Dominique Larregain-Fournier, MD,
Anne Thoury, MD, Pierre Duvillard, MD, and
Damienne Castaigne, MD
Institut Gustave Roussy, Villejuif, France; and Centre Hospitalier de la Côte
Basque, Bayonne, France
Address reprint requests to: Docteur Philippe Morice, Service de
Chirurgie Gynécologique, Institut Gustave Roussy, 39 rue Camille
Desmoulins, 94805 Villejuif Cedex, France; e-mail: morice@igr.fr.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
traditional autologous flaps, but initial results are encour-
aging.
REFERENCES
1. Wiskind AK, Thompson JD. Fecal incontinence. In: Rock
JA, Thompson JD, editors. Telinde’s operative gynecology.
9th ed. Philadelphia (PA): Lippincott Williams & Wilkins;
1997. p. 1223–36.
2. Miklos JR, Kohli N. Rectovaginal fistula repair utilizing a
cadaveric dermal allograft. Int Urogynecol J Pelvic Floor
Dysfunct 1999;10:405– 6.
3. Moore RD, Miklos JR. Repair of a vaginal evisceration
following colpocleisis utilizing an allogenic dermal graft. Int
Urogynecol J Pelvic Floor Dysfunct 2001;12:215–7.
4. Miklos JR, Kohli N, Moore RD. Levator plasty release and
reconstruction of rectovaginal septum using allogenic der-
mal graft. Int Urogynecol J Pelvic Floor Dysfunct 2002;13:
44 – 6.
5. Kohli N, Miklos JR. Dermal augmented rectocele repair. Int
Urogynecol J Pelvic Floor Dysfunct 2003;14:146 –9.
6. Harper C. Permacol: clinical experience with a new bioma-
terial. Hosp Med 2001;62:90 –5.
7. Arunkalaivanan AS, Barrington JW. Randomized trial of
porcine dermal sling (Pelvicol implant) vs tension free vag-
inal tape (TVT) in surgical treatment of stress incontinence.
Int Urogynecol J Pelvic Floor Dysfunct 2003;14:17–23.
8. Gomelsky A, Rudy DC, Dmochowski RR. Porcine dermis
interposition graft for repair of high grade anterior compart-
ment defects with or without concomitant pelvic organ
prolapse procedures. J Urol 2004;171:1581– 4.
Received October 26, 2003. Received in revised form February 5, 2004.
Accepted February 11, 2004.
BACKGROUND: The aim of this article is to report 3 cases of
port-site implantation after laparoscopic treatment of a
borderline ovarian tumor.
CASES: Three patients underwent a laparoscopic proce-
dure for a serous (2 patients) or mucinous (1 patient)
borderline ovarian tumor. In 2 patients, the port-site im-
plantation was discovered during a later surgical proce-
dure, and one was discovered clinically 11 months after the
initial laparoscopic oophorectomy. Surgical resection of the
port-site was the only treatment in all cases. These women
are currently alive and disease-free 11, 23, and 51 months
after the treatment of the scar metastasis.
CONCLUSIONS: These results suggest that, unlike port-site
metastasis in other gynecologic malignancies, the prognosis
in patients with a port-site implantation after laparoscopic
management of borderline ovarian tumor is excellent. The
treatment of this complication is surgical resection.
0029-7844/04/$30.00 1167
doi:10.1097/01.AOG.0000124988.46203.f2
(Obstet Gynecol 2004;104:1167–70. © 2004 by The Amer-
ican College of Obstetricians and Gynecologists.)

Tumor recurrence at port sites after laparoscopic treatment

of malignancies has been largely reported in the literature.
Nevertheless, the exact impact of the laparoscopic proce-
dure on the occurrence of this complication continues to be
discussed. Whatever the mechanisms of the onset of trocar-
site lesions, they seem to arise particularly after laparo-
scopic management of advanced stage disease (peritoneal
disease) in patients with an ovarian tumor. Port-site “metas-
tases” after laparoscopic treatment of a borderline ovarian
tumor are exceptional. Here, we report 3 cases of port-site
metastasis after laparoscopic surgery for borderline ovarian
tumor, focusing on the clinical presentation, histologic find-
ings, management, and outcomes.
Fig. 1. Case 1: Implants of the serous borderline tumor in
the scar of the lateral port site (thin arrows). Striated
muscle is located at the top of the figure (thick arrows).
CASE 1
Hematoxylin-eosin stained; ! 25, original magnification.
A 28-year-old woman with a history of a left borderline Morice. Port-Site Metastasis in Borderline Tumor. Obstet Gynecol 2004.
ovarian tumor 5 years prior, and a negative second-look
surgical procedure 6 months later, underwent a laparo-
referred to our institution. The histologic examination
scopic procedure for a suspicious lesion on the left ovary. A
revealed bilateral serous borderline ovarian tumor. A
laparoscopic oophorectomy of the left ovary and a biopsy
laparotomy was performed 3 months later, including a
of a small excrescence on the surface of the right ovary were
bilateral salpingo-oophorectomy (with cryopreservation
performed. The specimens were removed by using an
of part of 1 ovary), an omentectomy, and excision of the
endoscopic bag. There was no evidence of suspicious le-
peritoneal implants. During the surgical procedure, a
sions on the surface of the peritoneum. The histologic
3-mm nodule was removed from a lateral laparoscopic
analysis revealed a recurrent serous borderline tumor on
port site. The histologic examination revealed bilateral
the left ovary associated with a borderline lesion on the
serous borderline ovarian tumor with noninvasive peri-
right ovary with negative peritoneal cytology. Eleven
toneal implants. A port-site borderline ovarian tumor
months later, the clinical examination, blood markers (CA
was found on the peritoneum and aponeurosis (Fig. 2).
125), and abdominopelvic ultrasonography results were
normal, but the patient reported a 3-cm nodule on the
5-mm scar at the right-lateral port site. A port-site resection
was associated with laparoscopic exploration of the ab-
dominopelvic cavity and systematic peritoneal and right
ovarian biopsies. The histologic examination revealed a
port-site borderline ovarian tumor but no intraabdominal
disease (Fig. 1). Two months later the patient had a spon-
taneous pregnancy. This patient gave birth by cesarean
delivery to a healthy full-term baby weighing 4,000 g. The
peritoneum and remaining ovary were macroscopically
normal, and she is disease-free 11 months after surgical
excision of the port-site tumor.

CASE 2
A 28-year-old woman, with a history of infertility under-
went a laparoscopic procedure for bilateral ovarian tu-
mors associated with peritoneal implants. She under-
went simple laparoscopic biopsies of both ovarian
tumors (without the use of an endoscopic bag) and was
Fig. 2. Case 2: Noninvasive implants (arrows) on the perito-
neum and inside the abdominal wall at a lateral port site.
Hematoxylin-eosin stained; ! 200, original magnification.
Morice. Port-Site Metastasis in Borderline Tumor. Obstet Gynecol 2004.

1168 Morice et al Port-Site Metastasis in Borderline Tumor OBSTETRICS & GYNECOLOGY


Fig. 3. Case 3: Presence of mucinous product (arrows)
inside the abdominal wall at the suprapubic trocar. Hema-
toxylin-eosin stained; ! 25, original magnification.
Morice. Port-Site Metastasis in Borderline Tumor. Obstet Gynecol 2004.
The patient is currently disease-free 20 months after this
surgical procedure.
CASE 3
A 55-year-old woman underwent a laparoscopic procedure
for a right 7-cm cyst with a few intracystic septa but no solid
component at ultrasonography, and CA 125 was normal.
During the surgical procedure, no ascites or any external
excrescences were observed. A right salpingo-oophorec-
tomy (without use of an endoscopic bag) was carried out.
Frozen section analysis of the tumor was performed during
the laparoscopic procedure, but it was not possible to dif-
ferentiate between a benign cyst and a borderline mucinous
tumor. The final histologic analysis demonstrated a muci-
nous borderline tumor. We planned laparoscopic comple-
tion surgery (contralateral salpingo-oophorectomy and
hysterectomy). Just before the surgical procedure (per-
formed 1 month later), we discovered a small 20-mm nod-
ule at the 12-mm suprapubic trocar site. The patient under-
went laparoscopic radical surgery with resection of the
suprapubic scar. The histologic examination revealed an
isolated mucinous implant in the aponeurosis of the medi-
ally positioned suprapubic trocar site (Fig. 3). The patient is
currently disease-free 48 months after this surgical proce-
dure.
COMMENT
The estimated incidence of port-site metastasis in pa-
tients undergoing laparoscopic management of gyneco-
logical disease is between 1% and 2%.1 A number of
mechanisms have been discussed in the literature to
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
explain this complication: the effect of the pneumoperi-
toneum (related to the pressure and/or use of carbon
dioxide), exfoliation or spillage of tumor cells along the
trocar (caused by a leakage of gas), inoculation of the
trocar site through contact of the laparoscopic instru-
ments with the tumor, contamination of the port sites as
resected specimens are extracted through an excessively
small incision, and modification of local immune reac-
tions related to the use of a laparoscopic procedure.1
The 3 cases reported here are exceptional because very
few cases of port-site metastasis after treatment of border-
line ovarian tumor have been reported in the literature.2–5
Furthermore, we have correlative clinical and histologic
findings, and our cases raise important questions about the
treatment and prognosis of this rare event. These 3 cases
were observed in a series of 54 patients who underwent
laparoscopic management of a borderline ovarian tumor in
our institution (Morice P, Camatte S, Rouzier R, Atallah D,
Pautier P, Pomel C, et al. Laparoscopic treatment of bor-
derline tumors [abstract]. Proc ASCO 2003;22:471). Port-
site metastasis is a well-established complication after lapa-
roscopic treatment of invasive ovarian cancer. In the
literature, 8 patients, including our 3, had tumor implanta-
tion at the trocar site after treatment of a borderline ovarian
tumor. In 2 cases reported by Hsiu et al,2 this event oc-
curred after simple laparoscopic biopsies. One of our cases
is similar. In the other reported cases, an adnexectomy or a
cystectomy was performed. It is noteworthy that in the 2
cases reported by Hsiu et al,2 a case reported by Crouet and
Heron,3 and a case by Gleeson et al,4 port-site implantation
was associated with peritoneal implants. These peritoneal
implants were not spotted during the initial laparoscopic
procedure in the 2 cases reported by Hsiu et al.2 In 2 of our
patients and in the case reported by Shepherd et al,5 there
was no peritoneal disease at the time of completion surgery.
None of the patients in the 9 reported cases with port-site
dissemination died of their tumors (follow-up of between 6
and 72 months). The prognosis in patients with a port-site
implantation was excellent, as in the case of borderline
ovarian tumor and noninvasive peritoneal implants, which
also have an excellent prognosis.6 Treatment of this com-
plication is based on local excision.
Such events could be prevented by using an endo-
scopic bag when extracting the operative specimens (and
even when simple biopsies are performed). This bag was
not used in 6 cases reported in the literature. However, in
case 1 in our study, an endoscopic bag was used, and this
patient developed a lateral port-site implantation. This
complication was not related to direct contamination by
the tumor when it was removed, because the metastasis
did not arise at the site where the ovarian tumor was
extracted (medial port). It is probably related to contam-
ination of the surgical instruments (forceps) used to
Morice et al Port-Site Metastasis in Borderline Tumor 1169
grasp and maneuver the ovary, which were introduced
into the 5-mm lateral ports. Using an endoscopic bag
does not totally rule out the occurrence of port-site
implantations. Several surgical procedures aimed at low-
ering the risk of port-site metastasis from ovarian tumors
have been discussed in the literature. Van Dam et al7
reported a reduction of the risk of this complication in
patients who underwent a laparoscopic procedure with
closure of all layers. But this type of measure could not
be used in 5-mm port sites. Another measure aimed at
reducing port-site metastasis consists of irrigating the
trocar site with sterile water or 5% povidone iodine.1
This procedure was not used in our cases, but because it
is very simple, perhaps it should be done systematically
once the trocar has been removed.1
Should systematic excision of trocar sites be performed if
restaging surgery is planned? In both of the cases of Hsiu et
al,2 in that of Gleeson et al,4 and in 3 of our patients,
dissemination to the abdominal wall was diagnosed during
clinical or surgical exploration. In the case reported by
Shepherd et al,5 this tumor was diagnosed fortuitously
during the histologic examination. In our opinion, system-
atic removal of the port site should not be performed in
patients without clinical abnormalities.
These results seem to suggest that, unlike port-site
metastasis in other gynecologic malignancies, the prog-
nosis in patients with a port-site implantation after lapa-
roscopic management of borderline ovarian tumor is
excellent. The treatment of this complication is based on
surgical resection. An endoscopic bag should be used to
remove specimens as a preventive measure against this
complication.
Early Abdominal Incision
Recurrence in a Patient With
Stage I Adenocarcinoma of the
Endometrium
Chi Chiung Grace Chen, MD,
J. Michael Straughn Jr, MD, and
Larry C. Kilgore, MD
Division of Gynecologic Oncology, University of Alabama, Birmingham, Alabama
BACKGROUND: Although most patients with a surgical stage
I endometrial cancer have an excellent prognosis, some
patients will experience a recurrence. Endometrial cancer
Address reprint requests to: J. Michael Straughn Jr, MD, 619 19th
Street South, OHB Room 538, Division of Gynecologic Oncology,
Birmingham, AL 35249; e-mail: jmstraughn@yahoo.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1170 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
REFERENCES
1. Ramirez PT, Wolf JK, Levenback C. Laparoscopic port-site
metastases: etiology and prevention. Gynecol Oncol 2003;
91:179 – 89.
2. Hsiu JG, Given FT, Kemp GM. Tumor implantation after
diagnostic laparoscopic biopsy of serous ovarian tumors of
low malignant potential. Obstet Gynecol 1986;68:90S–3S.
3. Crouet H, Heron JF. Dissémination du cancer de l’ovaire
lors de la chirurgie coelioscopique: un danger réel. Presse
Med 1991;20:1738 –9.
4. Gleeson NC, Nicosia SV, Mark JE, Hoffman MS,
Cavanagh D. Abdominal wall metastases from ovarian
cancer after laparoscopy. Am J Obstet Gynecol 1993;169:
522–3.
5. Shepherd JH, Carter PG, Lowe DG. Wound recurrence by
implantation of a borderline ovarian tumour following lapa-
roscopic removal. Br J Obstet Gynaecol 1994;101:265– 6.
6. Morice P, Camatte S, Rey A, Atallah D, Lhommé C,
Pautier P, et al. Prognostic factors of patients with advanced
stage serous borderline tumor of the ovary. Ann Oncol
2003;14:592– 8.
7. Van Dam PA, De Cloedt J, Tjalma WA, Buytaert P, Bec-
quart D, Vergote I. Trocar implantation metastasis after
laparoscopy in advanced ovarian cancer: can the risk be
reduced. Am J Obstet Gynecol 1999;181:536 – 41.
Address reprint requests to: Docteur Philippe Morice, Service
de Chirurgie Gynécologique, Institut Gustave Roussy, 39 rue
Camille Desmoulins, 94805 Villejuif Cedex, France; e-mail:
morice@igr.fr.
Received November 14, 2003. Received in revised form December 17,
2003. Accepted January 15, 2004.
typically recurs at the vaginal cuff or in the pelvis; however,
it can recur distantly in the abdomen or lung. Although
recurrences have been reported at laparoscopic trochar
sites, it is unusual to have a recurrence in the abdominal
incision after laparotomy.
CASE: A 51-year-old woman was diagnosed with stage Ib
grade 2 adenocarcinoma of the endometrium and stage IV
endometriosis. Six months after surgery, she presented
with a mass in the lateral aspect of her Maylard incision.
Computed tomography scans of the chest, abdomen, and
pelvis showed no evidence of recurrent disease. The mass
was resected and confirmed to be an adenocarcinoma sim-
ilar to the endometrial primary.
CONCLUSION: This case is interesting because of the rapid
recurrence of the endometrial primary at an unusual site—
the abdominal incision. It illustrates the need to carefully
evaluate all suspicious masses, even as early as 6 months
after diagnosis and surgical staging. (Obstet Gynecol
2004;104:1170 –2. © 2004 by The American College of
Obstetricians and Gynecologists.)
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000123268.57857.0d
Endometrial cancer is the fourth most common cancer in
women, with approximately 36,100 new cases per year
and 6,500 deaths.1 Overall survival is excellent because
the majority of patients are diagnosed with disease con-
fined to the uterine corpus. Pathologically, the tumor
spreads from the endometrial stroma into the uterine
myometrium and other adjacent pelvic structures such
as the cervix, fallopian tubes, and ovaries. Most recur-
rences occur in the pelvis or at the vaginal cuff, although
distant metastases to the chest, abdomen, and brain are
not uncommon. Recurrences have been reported at lapa-
roscopic trochar sites2; however, it is unusual to have a
recurrence in the abdominal incision after laparotomy.
In a series of patients with grade 3 endometrial carci-
noma, only 1 patient experienced a recurrence in the
abdominal incision.3
CASE
A 51-year-old postmenopausal woman presented with
abnormal uterine bleeding for 3 months. Medical history
was significant for menorrhagia unresponsive to hor-
mone therapy. Surgical history included hysteroscopy
and dilation and curettage. Social and family histories
were noncontributory. Pelvic examination revealed a
slightly enlarged uterus, but no adnexal masses were felt.
Nodal survey revealed no lymphadenopathy. An endo-
metrial biopsy revealed grade 2 papillary serous adeno-
carcinoma of the endometrium.
The patient underwent an exploratory laparotomy
with total abdominal hysterectomy, bilateral salpingo-
oophorectomy, omentectomy, and pelvic and paraaortic
lymphadenectomy. A Maylard incision was performed,
and peritoneal cytology was obtained after entering the
abdominal cavity. Intraoperative findings revealed
dense adhesions and stage IV endometriosis with bilat-
eral endometriomas and extensive peritoneal and cul-de-
sac implants. All gross endometriosis was removed, and
surgical pathology confirmed no evidence of metastatic
adenocarcinoma. Final pathology was consistent with
stage Ib grade 2 endometrial adenocarcinoma with su-
perficial myometrial invasion. The peritoneal implants
were consistent with endometriosis. Final pathology did
not demonstrate a papillary serous component of the
tumor, despite the initial endometrial biopsy. Postoper-
atively, the patient did well and no further therapy was
recommended. Six months after her initial laparotomy,
she presented with a mass in the lateral aspect of her
Maylard incision. Computed tomography scans of the
chest, abdomen, and pelvis were performed after the
mass was noted and showed no evidence of recurrent
disease. Needle aspiration of this mass revealed glandu-
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
lar cells with mild cytologic atypia. The pathologist could
not determine whether this represented endometriosis or
endometrial adenocarcinoma.
The patient underwent a wide local excision of the
mass and surrounding subcutaneous fat. Intraopera-
tively, there was no involvement of the anterior rectus
fascia or intraabdominal disease. Final pathology re-
vealed a 3.5 cm ! 3.9 cm ! 1.8 cm nodule that was a
grade 2 endometrial adenocarcinoma, similar to the en-
dometrial primary. The surgical margins were free of
neoplasm. Treatment options considered were observa-
tion, electron beam radiation to the incision, whole pel-
vic radiation, or hormonal therapy. The patient received
5,400 cGy of whole pelvic radiation without significant
acute complications. The patient remained disease-free
for 1 year after radiation therapy; however, she subse-
quently developed intraabdominal disease and is cur-
rently receiving chemotherapy.
COMMENT
It is estimated that approximately 17% of endometrial
carcinoma will recur. The prognostic factors that contribute
to recurrence are depth of myometrial invasion, tumor
grade, and extrauterine disease.4 This patient demon-
strated intermediate-risk factors for recurrence, specifically
superficial myometrium invasion and grade 2 histology.
In 1 series, the recurrence rate for similar intermediate-risk
tumors was estimated at 8% without adjuvant therapy.5
This case is interesting because this patient was at relatively
low risk for recurrence; nevertheless, the tumor rapidly
recurred at an unusual site—the abdominal incision. Final
pathology confirmed that this mass was consistent with the
patient’s primary tumor.
There have been other case reports in the literature
involving abdominal wall recurrences in endometrial
cancer patients. For example, 1 case described a recur-
rence in the abdominal incision of a patient with a stage
Ib grade 2 endometrial adenocarcinoma 7 years after
diagnosis. The authors postulated that this occurred as a
result of tumor cells extruding from the cervix and
implanting in the abdominal incision at the time of
surgery.6 In our case, it is also likely that tumor cells were
deposited in the incision and subsequently grew rapidly
until this nodule was palpable. However, it is extremely
rare to experience recurrence in the abdominal incision
this soon after a staging laparotomy for endometrial
cancer.
This case is further complicated by the presence of
stage IV endometriosis at the time of staging laparotomy.
Similar to a neoplastic process, endometriosis can spread
either locally with direct extension or distantly through a
hematogenous route. Furthermore, it is not uncommon
Chen et al Abdominal Incision Recurrence 1171
to find endometriosis implants in surgical scars, transab-
dominal needle tracts, and laparoscopic trochar sites.
There are numerous reports describing such cases after
cesarean deliveries or after surgical removal of endome-
triosis. It is hypothesized that during these procedures
the endometrial cells become implanted in the various
locations.7 Therefore, endometriosis was in the differen-
tial diagnosis of this patient with an abdominal incision
mass. One could have easily attributed this mass to
endometriosis, especially in light of the needle biopsy
that showed glandular cells with mild cytological atypia.
Unfortunately, the lesion was consistent with recurrent
endometrial cancer and not endometriosis.
Although isolated recurrences in the abdominal inci-
sion occur in patients with endometrial cancer, they are
uncommon and usually occur several years after the
initial diagnosis. Furthermore, this case was confounded
by the presence of stage IV endometriosis at the staging
laparotomy. Because of the short time interval and the
inconclusive needle biopsy, this abdominal incision re-
currence could have been mistaken for endometriosis,
an incisional hernia, a hematoma, or scar tissue. This
case demonstrates that recurrent disease must be aggres-
sively ruled out, even in patients with intermediate-risk
factors for recurrence.
Trophoblastic Tissue Spread to
the Sigmoid Colon After Uterine
Perforation
Ishai Levin, MD, Dan Grisaru, MD, PhD,
David Pauzner, MD, and Benny Almog, MD
The Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel
BACKGROUND: Trophoblastic tissue spread following uter-
ine perforation during dilation and curettage is rare. We
present a case of trophoblastic spread to the sigmoid colon
following uterine perforation, which was treated by surgi-
cal removal of the implants and intramuscular administra-
tion of methotrexate.
CASE: A woman presented 3 weeks after curettage for a
blighted ovum. Laparotomy performed for suspected in-
tra-abdominal bleeding revealed bleeding trophoblastic
implants in a perforation tract and the anterior uterine
wall and on the appendix epiploica of the sigmoid colon.
Address reprint requests to: Ishai Levin, MD, Department of Gynecol-
ogy, Lis Maternity Hospital, Tel Aviv Sourasky Medical Center, 6
Weizmann Street, Tel Aviv, 64239, Israel; e-mail: i-levin@barak-
online.net.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1172 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
REFERENCES
1. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer
statistics, 2000. CA Cancer J Clin 2000;50:7–33.
2. Dargent DF. Laparoscopic techniques for gynecologic can-
cer: description and indications. Hematol Oncol Clin North
Am 1999;13:1–19.
3. Chambers SK, Kapp DS, Peschel RE, Lawrence R, Merino
M, Kohorn EI, et al. Prognostic factors and sites of failure in
FIGO Stage 1, Grade 3 endometrial carcinoma. Gynecol
Oncol 1987;27:180 – 8.
4. Morrow CP, Bundy BN, Kurman RJ, Creasman WT,
Heller P, Homesley HD, et al. Relationships between surgi-
cal-pathologic risk factors and outcome in clinical stage I
and II carcinoma of the endometrium: a Gynecologic
Oncology Group study. Gynecol Oncol 1991;40:55– 65.
5. Straughn JM, Huh WK, Kelly FJ, Leath CA, Kleinberg MJ,
Hyde J, et al. Conservative management of stage I endome-
trial carcinoma after surgical staging. Gynecol Oncol 2002;
84:194 –200.
6. Kotwall CA, Kirkbride P, Zerafa AE, Murray D. Endome-
trial cancer and abdominal wound recurrence. Gynecol
Oncol 1994;53:357– 60.
7. Fair KP, Patterson JW, Murphy RJ, Rudd RJ. Cutaneous
deciduosis. J Am Acad Dermatol 2000;43:102–7.
Received August 25, 2003. Received in revised form October 21, 2003.
Accepted November 6, 2003.
The implants were surgically removed and methotrexate
was administered for persistently high "-hCG levels. The
patient fully recovered.
CONCLUSION: Extrauterine trophoblastic implants should
be considered in women evaluated for abdominal pain
whose pregnancy test is positive after uterine perforation.
Conservative treatment with methotrexate in nonacute pa-
tients may be considered. (Obstet Gynecol 2004;104:
1172– 4. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
Trophoblastic tissue spread to the abdominal cavity
following perforation of the uterus during dilation and
curettage is a rare event. The vast majority of intra-
abdominal trophoblastic tissue spread described previ-
ously1 were complications of ectopic rather than intra-
uterine pregnancies. Unusual cases of trophoblastic
tissue spread to the myometrium, the uterine serosa,
and the omentum,2– 4 subsequent to intrauterine preg-
nancies, have been reported. We report a case of tropho-
blastic tissue spread to the sigmoid colon following dila-
tion and curettage for a blighted ovum pregnancy. A
MEDLINE search, performed from January 1, 1966 to
December 31, 2003, using the terms “colon,” “tropho-
blast,” “sigmoid,” “implants,” “uterine perforation,” “in-
tra-abdominal,” and “complications,” alone and in com-
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000128113.82312.34
bination, revealed that this is the first such case to be
reported in the English literature.
CASE
A 30-year-old woman presented to our gynecological
emergency room for evaluation of abdominal pain.
Three weeks earlier she had undergone dilation and
curettage for a blighted ovum pregnancy of 8 weeks. She
was in general good health with no history of previous
operations. Her obstetric history included 2 induced
abortions and 2 spontaneous abortions with dilation and
curettage, and her menstrual cycles were regular with a
normal amount of bleeding.
The patient reported slight bleeding and episodes of
abdominal pain since undergoing the latest obstetrical pro-
cedure. Her current complaints of severe abdominal pain
started 24 hours before she presented to our hospital. Vital
signs were blood pressure 110/70 mm Hg, pulse 90 beats
per minute, and oral temperature 37.1°C. Her abdominal
examination revealed signs of peritoneal irritation with
guarding and rebound tenderness. She experienced severe
pain from cervical manipulation during a vaginal examina-
tion. Her hemoglobin level was 10.6 g/dL. A qualitative
urine analysis &-hCG was positive. Three days before her
admission, the quantitative assessment of &-hCG was 6,500
mIU/mL. A transvaginal ultrasound scan revealed a small
amount of amorphous material in the uterine cavity, and a
moderate amount of fluid, suspected of being blood, was
visible in the cul-de-sac.
A review of the sonographic scan, which had been
performed before the procedure to evacuate the uterus,
revealed an empty gestational sac with no fetal echo or
yolk sac, a diagnosis compatible with a blighted ovum.
We telephoned the surgeon who had performed the last
dilation and curettage and learned that he had difficulty
while dilating the cervix, possibly creating a false route,
but that there was no indication or suspicion of a perfo-
ration of the uterus. He stated that the pathological
examination of the material extracted from the uterus
described it as “products of conception.”
At this point, the patient’s differential diagnosis con-
sisted of an ectopic pregnancy, a heterotopic pregnancy,
another hemorrhagic abdominal event, or perforation of
the uterus. We believed that perforation of the uterus
manifesting as late as 3 weeks after a dilation and curet-
tage was unlikely.
The patient underwent a diagnostic laparoscopy, and
exploration of the abdomen revealed 500 mL of blood in
the pelvic cavity. There was a 2-cm diameter mass
resembling trophoblastic tissue protruding from the an-
terior uterine wall (Fig. 1), and bleeding could be seen
from the area of attachment of this mass to the uterine
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Fig. 1. Laparoscopic view of the anterior uterine wall
harboring a bleeding trophoblastic implant.
Levin. Trophoblastic Colon Implants. Obstet Gynecol 2004.
serosa. A laparotomy was performed under the consid-
eration that the hemorrhage could not be controlled
under laparoscopy. Further inspection of the abdomen
revealed a similar lesion on the appendix epiploica of the
proximal sigmoid colon.
The two lesions were excised and removed. The
black-blue uterine wall mass was contiguous with the
endometrial cavity, and there was no other source of
bleeding. A repeat uterine curettage was performed un-
der direct vision through the abdominal incision.
The patient’s recovery was uneventful. Quantitative
&-hCG testing was repeated once a week postoperatively
and revealed a plateau of (800 mIU/mL throughout 3
consecutive weeks. The patient was readmitted and
treated with alternate intramuscular methotrexate 1
mg/kg and folinic acid 0.1 mg/kg for 5 days. Repeated
&-hCG testing demonstrated a decline in its levels to
undetectable ones at 1 month after the operation. The
patient has been asymptomatic since then.
Pathological examination of the material obtained
during surgery revealed gestational trophoblastic tissue.
COMMENT
We report a case of trophoblastic tissue spread to the
sigmoid colon following dilation and curettage for a
blighted ovum pregnancy. Many cases of trophoblastic
spread to peritoneal surfaces have been described follow-
ing surgical treatment for ectopic pregnancy,1 but contig-
uous spread of trophoblastic tissue through a uterine
perforation is a rare event. Cases of trophoblastic spread
to the myometrium2 and to the uterine serosa3 have been
described, as was a case of trophoblastic spread to the
Levin et al Trophoblastic Colon Implants 1173
omentum following perforation during a first-trimester
termination of pregnancy.4 Giuliani et al5 reported a case
of trophoblastic spread with large macroscopic lesions
on the sigmoid colon after salpingostomy for ectopic
pregnancy.
We report the first case in the English literature of
trophoblastic spread to the sigmoid colon after termina-
tion of an intrauterine pregnancy by dilation and curet-
tage. Awareness of the possibility of this adverse event
following dilation and curettage is important because it
may aid the clinician in prompt diagnosis and manage-
ment of such patients. Trophoblastic spread in this case
was probably due to the proximity of the appendix
epiploica of the sigmoid colon to the route of perforation
in the anterior uterine wall. Exfoliating trophoblastic
cells probably have the ability to implant and grow on
peritoneal surfaces. This case also illustrates the effective
use of methotrexate to treat small or microscopic tropho-
blastic implants that were not visible and therefore not
removed during laparotomy.
We conclude that a diagnosis of extrauterine tropho-
blastic implants should be considered in women with
uterine perforation who are being evaluated for abdom-
inal pain and whose pregnancy test is positive. Conser-
Paraneoplastic Limbic
Encephalitis: Ovarian Cancer
Presenting as an Amnesic
Syndrome
Michael H. Bloch, BA, Wei Chan Hwang, BA,
Joachim M. Baehring, MD, and
Setsuko K. Chambers, MD
Departments of Neurology and Neurosurgery, Department of Obstetrics and
Gynecology, Yale University School of Medicine, New Haven, Connecticut
BACKGROUND: Paraneoplastic limbic encephalitis (PLE) is a
rare neurologic manifestation of malignancy. Paraneoplas-
tic limbic encephalitis typically presents with short-term
memory loss, seizures, or other limbic system abnormali-
ties. The majority of PLE cases are associated with lung and
testicular cancer.
CASE: We present the first case of PLE attributable to an
epithelial ovarian cancer. The 58-year-old woman pre-
sented with rapidly progressing short-term memory loss
and amnesia. The diagnosis of ovarian cancer was sus-
pected on the basis of computed tomography scan findings.
Address reprint requests to: Michael Bloch, 196 Crown Steet, Apart-
ment 206, New Haven, CT 06510; e-mail: michael.bloch@yale.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1174 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
vative treatment with methotrexate may be considered
in nonacute patients.
REFERENCES
1. Cartwright PS. Peritoneal trophoblastic implants after sur-
gical management of tubal pregnancy. J Reprod Med 1991;
36:523– 4.
2. Pascual MA, Tresserra F, Dexeus D, Grases PJ, Dexeus S.
Myometrial trophoblastic implant as a complication of sur-
gically induced first-trimester termination of pregnancy.
Ultrasound Obstet Gynecol 2003;22:194 –5.
3. Lam PM, Yim SF, Leung TN. Entrapment of viable tropho-
blastic tissue in a uterine hematoma after surgical evacua-
tion: a case report. J Reprod Med 2002;47:170 –2.
4. Dessouky DA. Ectopic trophoblast as a complication of
first-trimester induced abortion. Am J Obstet Gynecol 1980;
136:407– 8.
5. Giuliani A, Panzitt T, Schoell W, Urdl W. Severe bleeding
from peritoneal implants of trophoblastic tissue after lapa-
roscopic salpingostomy for ectopic pregnancy. Fertil Steril
1998;70:369 –70.
Received November 5, 2003. Received in revised form December 15,
2003. Accepted January 27, 2004.
The patient showed dramatic clinical improvement within
2 weeks of debulking surgery and after the initiation of
plasmapheresis and chemotherapy.
CONCLUSION: Paraneoplastic limbic encephalitis is a rare
complication of ovarian tumors that is potentially revers-
ible with prompt surgical management of the primary
tumor, plasmapheresis, and chemotherapy. (Obstet Gy-
necol 2004;104:1174 –7. © 2004 by The American College
of Obstetricians and Gynecologists.)
Paraneoplastic limbic encephalitis (PLE) is an autoim-
mune manifestation of malignancy in which antibodies
or cytotoxic T cells, which combat the primary progress-
ing neoplasm, cross-react with neuronal tissue in the
temporal lobe.1 Neurologic symptoms of PLE include
memory loss or seizures, but also it can involve other
limbic system abnormalities such as hypersomnia, hy-
perphagia, confusion, irritability, or depression. PLE is
typically associated with lung and testicular cancer.
Because PLE is a rare syndrome and causative anti-
bodies are identified in only a minority of cases, PLE
remains a diagnosis of exclusion. Diagnostic criteria
include neuropathological examination (biopsy or au-
topsy) demonstrating neuronal loss and perivascular
lymphocytic infiltrate or all 4 of the following: 1) a
clinical picture of short-term memory loss, seizures, or
psychiatric symptoms suggesting limbic system involve-
ment; 2) an interval of less than 4 years between the onset
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000128110.31784.c8
of neurological symptoms and the cancer diagnosis; 3)
exclusion of other cancer-related complications (metastasis,
infection, metabolic and nutritional deficits, cerebrovas-
cular disorder, or side-effects of therapy) that may cause
symptoms of limbic dysfunction; and 4) at least 1 of the
following: cerebrospinal fluid (CSF) with inflammatory
changes (pleocytosis, oligoclonal bands, increased immu-
noglobulin, or total protein content in the absence of mea-
sured immunoglobulin); magnetic resonance imaging
showing unilateral or bilateral temporal lobe abnormalities
on T2-weighted images or atrophy on T1-weighted images;
or an electroencephalogram showing slow or sharp wave
activity in 1 or both temporal lobes.2
Treatment for PLE involves aggressive treatment of the
primary tumor and immunomodulatory therapy (intrave-
nous immunoglobulin or plasmapheresis) to remove the
offending humor. Even with proper therapy, the majority
of patients given the diagnosis of PLE fail to recover. We
present the first reported case of PLE associated with pap-
illary serous adenocarcinoma of the ovary.
CASE
A 58-year-old woman presented with a 1-week history of
rapid-onset memory loss and confusion. She reported a
1- to 2-day history of nausea preceding the onset of her
memory loss and a 23- to 27-kg weight-loss over the last
year since joining Overeaters Anonymous.
On admission to a local hospital the patient had a
Mini-Mental Status Examination (MMSE) score of 22
with the ability to register the 3 objects immediately but
was unable to recall any of the objects at 5 minutes. A
comprehensive neurological, psychiatric, and physical
examination was otherwise unremarkable. Results of an
initial workup for metabolic, infectious, and toxicologic
causes of memory loss were negative. Additionally, a
chest X-ray, electroencephalogram, head computed to-
mography (CT), and brain magnetic resonance imagine
pre- and postgadolinium contrast were negative.
A lumbar puncture revealed a mild pleocytosis with
lymphocytic predominance. Cerebrospinal fluid results
were 10 white blood cells/mm3 (89% lymphocytes, 11%
monocytes), 22 red blood cells/mm,3 glucose 69 mg/dL,
and protein 25 mg/dL. Screening test results for herpes
simplex virus (HSV), syphilis, tuberculosis, and crypto-
coccal meningitis as well as bacterial and viral cultures in
the CSF returned negative. The patient was initially
treated with acyclovir pending the CSF HSV results and
failed to show clinical improvement on this regimen.
An abdominal and pelvic CT revealed an 8-cm mul-
tilocular solid and cystic mass emanating from the pelvis
just superior to the uterus extending into the right lower
abdomen. Nodularity of the greater omentum, infiltra-
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
tion into the peritoneal fat pad anterior to the mass,
enlarged pelvic nodes, and nodularity in the small bowel
mesentery were also appreciated on CT scan. A trans-
vaginal ultrasound examination confirmed the presence
of a large right-sided, multiloculated cyst and omental
studding. Serologic tumor marker panel revealed a CA
125 level that was elevated at 2,508.2 U/L. A serological
panel for paraneoplastic autoantibodies including an-
ti-Hu and anti-Yo antibodies was negative. The patient
was transferred to Yale-New Haven Hospital for treat-
ment of her primary gynecologic malignancy.
Upon transfer (hospital day 10) the patient scored
16/30 on the MMSE. The patient was completely disori-
ented to time and place, unable to even register any of 3
objects nor recall 5 minutes later that she had even been
asked to remember objects. When the examiner left and
immediately returned to the room, the patient was un-
able to recall having previously met him.
A pelvic examination revealed the cul-de-sac to be free
and the presence of a large mass indistinguishable from the
uterus. Abdominal examination was remarkable only for
an enlarged, hard, midline, subumbilical mass. Magnetic
resonance imagine of the pelvis revealed a 6.9-cm !
12.0-cm ! 10.3-cm heterogeneous lobulated mass invading
the superior, posterior uterus with no evidence of perito-
neal implants, lymphadenopathy or ascites.
On hospital day 17, an exploratory laparotomy with a
total abdominal hysterectomy, bilateral salpingo-oopho-
rectomy, right pelvic lymphadenectomy, omentectomy,
and extensive debulking with the cavitation ultrasound
aspirator was performed. The tumor mass extended
from the right ovary, infiltrated the adjoining uterus, and
studded the pelvic cul-de-sac, left gutter, mesocolon,
small bowel, and omentum. The largest implant was on
the sigmoid mesocolon measuring approximately 3 cm.
Intraoperative frozen section of the right ovary revealed
a poorly differentiated adenocarcinoma consistent with
an ovarian primary. Complete surgical resection of all
visible tumor mass was achieved. Final pathologic anal-
ysis revealed a stage IIIc papillary serous adenocarci-
noma emanating from the right ovary. Disease was
confirmed in the omentum, mesentery of the sigmoid
colon, cul-de-sac, and both ovaries. The sigmoid colon
implant was a benign peritoneal cyst.
In the first postoperative week, the patient’s MMSE
score remained at 16/30 with the same performance
difficulties previously described. On postoperative day 4
she recalled having “had surgery to remove her female
parts because of cancer.” On postoperative day 7, the
patient began a course of 3 plasmapheresis sessions,
followed by chemotherapy consisting of 6 cycles with
carboplatin and paclitaxel. She was discharged home on
postoperative day 12, still disoriented and confused.
Bloch et al Ovarian Cancer With PLE 1175
 After discharge, the patient’s memory deficits began to
improve dramatically. Three weeks after discharge, her
short-term memory abilities had returned to near prema-
lignancy levels although she remained completely amne-
sic of the events that took place during her hospitaliza-
tion as well as many other events from the previous year.
Her MMSE score had now improved to 26/27 (figure
copying, reading, and writing of a sentence were omit-
ted). She was able to register all 3 objects and recall 2 of
3 objects at 5 minutes and the third with slight prompt-
ing. The patient reported feeling much more paranoid,
anxious, and depressed since the illness began.
At five months postoperation, the patient had com-
pleted chemotherapy. A repeat abdominal and pelvic
CT showed no evidence of residual disease. Her CA 125
level was 23 U/mL. Repeat and more extensive paraneo-
plastic antibody panels that was drawn 7 weeks postop-
eratively and 6 weeks after initial plasmapheresis were
negative. The patient has returned to work part-time and
reports no current memory deficits, but she began taking
sertraline for depression that is currently in remission.
COMMENT
Since the initial description of PLE in 1968, approximately
150 cases have been identified in the English medical liter-
ature. A study reporting all PLE cases described before
2000 that met strict criteria for PLE (N ' 132) found that
the vast majority of cases have been associated with lung
cancer (n ' 68, 52%), (usually small cell), testicular neo-
plasm (n ' 14, 11%) or thymoma (n ' 6, 5%).2 Addition-
ally, case reports have implicated Hodgkin’s disease, neu-
roblastoma, colorectal, ovarian, esophageal, breast,
prostate, and renal cancers as other possible primary
tumor sites that produce PLE symptoms.
An extensive MEDLINE search that included the
search patterns 1) paraneoplastic limbic encephalitis, 2)
memory loss or amnesia and ovarian cancer, and 3)
paraneoplastic syndromes and ovarian cancer yielded
only 5 previous cases in the English medical literature
after 1966 that linked PLE with ovarian primary tumors.
Four of these cases were teratomas; none was epithelial
ovarian tumors.4 – 8 In all 5 cases as well as ours, patients
with PLE associated with ovarian tumors tested negative
for anti-Yo and anti-Hu antibodies. This case represents
the first reported link between papillary serous adenocar-
cinoma of the ovary and PLE.
Patients with PLE caused by primary tumors of the
ovary tend to have better outcomes than most PLE
patients. In 4 of 5 previously reported cases as well as
ours, neurological symptoms significantly improved or
completely resolved after a combination of surgery with
or without chemotherapy and immunomodulatory ther-
1176 Bloch et al Ovarian Cancer With PLE
apy. In the 1 case in which PLE symptoms failed to
improve after surgical resection of the primary tumor,
plasmapheresis was never attempted. These outcomes
compare favorably to the general population of PLE pa-
tients of whom only approximately 44% of those afflicted
ever experience neurological improvement after resection
of their tumor.2 Possible explanations for the improved
outcome for reported PLE cases with ovarian primary
neoplasms include the relative responsiveness of ovarian
immature teratomas to medical therapy compared with the
other neoplasms responsible for PLE (eg small cell lung
cancer) and the relatively young age of patients with PLE-
associated ovarian primary neoplasms.
Treatment for PLE should involve aggressive treat-
ment of the underlying neoplasm. Approximately 44%
of PLE patients show neurologic improvement after
treatment of their primary tumor.2 Attributes associated
with a greater likelihood of neurologic improvement
included decreased tumor burden at diagnosis, absence
of residual tumor after multimodal therapy, minimal
neurologic damage at time of treatment, and absence of
identifiable antineuronal antibodies.
Immunomodulatory therapy has been used in addi-
tion to traditional multimodal therapy against the
primary tumor to ameliorate the neurologic symptoms
in patients who have PLE. Due to the time-course of
interventions it cannot be ascertained whether it was
surgical treatment of the primary tumor, plasmapheresis,
chemotherapy, or the combination of all three which was
responsible for this patient’s dramatic improvement.
Nevertheless, given the effectiveness of plasmaphere-
sis in treating other autoimmune disorders of the central
nervous system and the anecdotal evidence suggest-
ing its effectiveness in paraneoplastic syndromes, plas-
mapheresis remains a prudent treatment option in pa-
tients who have PLE. The current case literature
suggests that this combination of therapies may be par-
ticularly effective in PLE patients with ovarian primary
neoplasms.
REFERENCES
1. Posner JB. Paraneoplastic syndromes. In: Posner JB, editor.
Neurologic complications of cancer. Philadelphia (PA): F.A.
Davis; 1995. p. 353– 85.
2. Gultekin SH, Rosenfeld MR, Voltz R, Eichen J, Posner JB,
Dalmau J. Paraneoplastic limbic encephalitis: neurological
symptoms, immunological findings and tumour association
in 50 patients. Brain 2000;123:1481–94.
3. Cao Y, Abbas J, Wu X, van Amburg AL. Anti-Yo paraneo-
plastic cerebellar degeneration associated with ovarian car-
cinoma: case report and review of the literature. Gynecol
Oncol 1999;75:178 – 83.
OBSTETRICS & GYNECOLOGY
4. Aydiner A, Gurvit H, Baral I. Paraneoplastic limbic enceph-
alitis with immature ovarian teratoma: a case report. J Neu-
rooncol 1998;37:63– 6.
5. Nokura K, Yamamoto H, Okawara Y, Koga H. Osawa H.
Sakai K. Reversible limbic encephalitis caused by ovarian
teratoma. Acta Neurol Scand 1997;95:367–73.
6. Okamura H, Oomori N, Uchitomi Y. An acutely confused
15-year-old girl. Lancet 1997;350:488.
7. Taylor RB, Mason W, Kong K, Wennberg R. Reversible
Coccidioidomycosis Mimicking
Ovarian Cancer
Michael W. Ellis, MAJ, MC, USA,
David P. Dooley, COL, MC, USA,
Michael J. Sundborg, LTC, MC, USA,
Laura L. Joiner, CPT, MC, USA, and
Edward R. Kost, LTC, MC, USA
Departments of Medicine and Obstetrics and Gynecology, Brooke Army Medical
Center, Fort Sam Houston, Texas
BACKGROUND: Dissemination of coccidioidomycosis to the
abdominal cavity is rare. No previous case of peritoneal
coccidioidomycosis has presented as an adnexal mass.
CASE: We report a case of peritoneal coccidioidomycosis
mimicking ovarian carcinoma. The patient presented with
a complex ovarian mass, ascites, omental caking, and an
elevated CA 125. The ultimate diagnosis was not made
until frozen section histopathology was performed at stag-
ing laparotomy.
CONCLUSION: Peritoneal coccidioidomycosis can present
with the clinical, radiographic, and serologic features of
ovarian cancer. Although essential for diagnosis and stag-
ing, radiographic studies and tumor markers have limited
specificity. Coccidioidomycosis now joins other benign con-
ditions that comprise the differential diagnosis of patients
who present with what seems to be advanced ovarian
carcinoma. Infectious diseases consultation is recom-
mended for the management of peritoneal coccidioidomy-
cosis. (Obstet Gynecol 2004;104:1177–9. © 2004 by The
American College of Obstetricians and Gynecologists.)
The noninvasive diagnosis of ovarian cancer may be
difficult. Multiple disease processes may present with
adnexal masses and ascites. Ultimately, operative biopsy
establishes the definitive diagnosis. The finding of an
elevated CA 125 increases the specificity of a preopera-
Address reprint requests to: Michael West Ellis, MD, MAJ, MC,
Infectious Disease Service MCHE-MDI, Brooke Army Medical Cen-
ter, 3851 Roger Brooke Drive, Fort Sam Houston, TX; e-mail:
Michael.Ellis2@cen.amedd.army.mil.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
paraneoplastic encephalomyelitis associated with a benign
ovarian teratoma. Can J Neurol Sci 1999;26:317–20.
8. Rosenbaum T, Gartner J, Korholz D, JanBen G, Schnei-
der D, Engelbrecht V, et al. Paraneoplastic limbic
encephalitis in two teenage girls. Neuropediatrics 1998;
29:159 – 62.
Received December 11, 2003. Received in revised form February 7,
2004. Accepted February 11, 2004.
tive diagnosis of malignancy. We present a case of
peritoneal coccidioidomycosis masquerading as ovarian
malignancy that possessed each of these clinical features.
CASE
A 42-year-old Hispanic woman presented in June 2003
with a 2-month history of slowly increasing abdominal
girth, occasional abdominal discomfort, and several epi-
sodes of fever and night sweats. The patient also noted
ribbon-like stools and irregular menses. She denied neu-
rologic, pulmonary, urinary, skin, or musculoskeletal
symptoms. The patient was a gravida 3 para 3 with her
past medical and surgical history being remarkable only
for left oophorectomy 25 years before presentation for
benign disease. A routine pelvic examination in Febru-
ary 2003 was normal. She took no medications and had
no known drug allergies. The patient grew up in San
Antonio, Texas, was currently living in Laredo, and had
had no recent out-of-state travel. There was no family
history of neoplastic disease.
The patient seemed well on examination. She was
afebrile. Her abdomen was distended with an evident
fluid wave, but had normal bowel sounds, was non-
tender, and had no palpable masses. Results of the
remainder of the examination, including the pelvic ex-
amination, were normal. Pelvic ultrasonography demon-
strated a normal-appearing uterus, ascites, and a right
ovarian complex hypoechogenic mass. Abdominal com-
puted tomography (CT) with contrast showed massive
ascites, a 3-cm right ovarian complex mass, and irregular
anterior abdominal peritoneum suggestive of peritoneal
studding (Fig. 1). The CA 125 was 365.1 U/mL (normal
range 0.6 –35.0). Urinalysis results were normal and
human immunodeficiency virus antibody testing was
negative.
The patient underwent laparotomy (total abdominal
hysterectomy, right salpingo-oophorectomy, and omen-
tectomy) for debulking and staging for the presumed
diagnosis of ovarian cancer. The peritoneum was stud-
ded with malignant-appearing nodules on all serosal
surfaces, including the uterus, fallopian tube, and ovary.
0029-7844/04/$30.00 1177
doi:10.1097/01.AOG.0000128111.90367.f2

Fig. 1. Computed tomography scan with contrast demon-
strating right adnexal mass (arrow).
Ellis. Coccidioidomycosis. Obstet Gynecol 2004.
Frozen sections of omental tissue, however, demon-
strated Coccidioides immitis spherules within noncaseating
granulomas that were diagnostic of coccidioidomycosis
(Fig. 2). Calcofluor white examination of both centri-
fuged peritoneal fluid and urine also revealed C immitis
spherules. There was no histopathologic evidence of
invasion into the ovary, fallopian tube, or uterus. The
patient’s serum complement fixation (CF) titer was 1:2,
and the cerebrospinal fluid CF titer was negative (testing
done by Dr. Demosthenes Pappagianis, University of
California, Davis). Further evaluation showed the ab-
sence of pulmonary disease or dissemination to the
central nervous system or bone.
On the first postoperative day, the patient developed
fever, tachycardia, hypotension, and decreased urine
Fig. 2. Photomicrograph of peritoneal serosa demonstrating
granuloma with spherule (arrow) of Coccidioides immitis
(Hematoxylin-eosin, ! 400, original magnification).
Ellis. Coccidioidomycosis. Obstet Gynecol 2004.
1178 Ellis et al Coccidioidomycosis
output. The patient was treated with volume resuscita-
tion, liposomal amphotericin B, and empiric antimicro-
bials for bacterial sepsis. She rapidly improved and re-
ceived the liposomal amphotericin B for a total of 9 days
before switching to oral fluconazole at 800 mg/d alone. A
dose reduction to 400 mg/d was necessary due to nausea.
After 8 months of therapy, the patient continued to do
well, with a normalized CA 125 and a stable CF. At least
1 year of fluconazole therapy was planned for the patient
contingent upon her symptoms, CA 125, CF, and non-
specific inflammatory markers.
COMMENT
Coccidioidomycosis is a fungal infection that is acquired
by the inhalation of airborne Coccidioides immitis arthro-
conidia (spores). Endemic to the southwestern United
States, C immitis is a desert soil fungus. Most commonly,
coccidioidomycosis manifests itself as a lower respira-
tory tract illness, with influenza-like symptoms, that re-
solves without therapy. Perhaps 1 patient per 100 in-
fected will develop extrapulmonary or disseminated
disease. Sites of disseminated disease typically include
the skin, bones, joints, and meninges. Many other sites of
disseminated infection have been reported but are rare.
Risk factors for disseminated disease are male sex; Fili-
pino, African-American, or Hispanic race; pregnancy;
human immunodeficiency virus infection; and other im-
mune suppression.1
Phillips and Ford2 recently reviewed the 26 previously
reported cases of peritoneal coccidioidomycosis. The
disease has generally presented with abdominal pain,
ascites, and radiographic or operative findings com-
monly mistaken for tuberculosis. The diagnosis of peri-
toneal coccidioidomycosis has been most often estab-
lished by demonstrating spherules in biopsy specimens.
The majority of reported peritoneal coccidioidomycosis
cases were in men, and little involvement of the female
genital tract has been noted. Unlike the prognosis of
disseminated coccidioidomycosis involving other sites,
mortality is rare with isolated peritoneal disease.
Coccidioidomycosis of the upper female genital tract
has been separately described.3 Two general patterns of
disease have been recognized: pelvic inflammatory dis-
ease with abdominal pain and endometritis with men-
strual abnormalities. The diagnosis has been made with
biopsies obtained at laparotomy or endometrial curet-
tage. Most patients were racially predisposed (Mexican-
American or African-American) or had recently been preg-
nant. Ascites has been unusual. Three deaths occurred in
the 11 reported patients, only 1 of whom had received
antifungal therapy. After an English-language MEDLINE
search (1966 to the present with search terms “coccidioid-
OBSTETRICS & GYNECOLOGY
omycosis,” “ovary,” “ovarian cancer,” “CA 125,” “perito-
neal carcinomatosis,” and “peritonitis”) and a search of the
bibliographies of the pertinent articles, no similar case was
found that presented with a syndrome of afebrile ascites, a
suspicious adnexal mass, and an elevated tumor marker as
in the current patient’s presentation.
Whereas disseminated coccidioidomycosis is a rare
disease, ovarian cancer is the fourth leading cause of
cancer death in American women.4 Detection of CA 125
is a valuable tumor marker for the evaluation of ovarian
cancer and monitoring response to therapy for the dis-
ease; however, nonspecificity can plague its usefulness.
In adults, many tissues may express CA 125, and conse-
quently, many conditions, both benign and malignant,
are associated with elevated levels. Essentially, any pro-
cess that causes serosal inflammation may elevate CA
125.5 In this patient the combination of ascites, occa-
sional abdominal pain, the radiographic appearance of a
complex ovarian mass, and an elevated CA 125 all
contributed to a high probability of finding ovarian
cancer with peritoneal spread at laparotomy.
Computed tomographic findings typical for peritoneal
involvement with ovarian carcinoma may also be non-
specific. Other diseases can also mimic this patient’s
radiographic findings that suggest peritoneal carcinoma-
tosis, especially tuberculous peritonitis.6 Besides ovarian
carcinoma, carcinomas of the colon, pancreas, stomach,
appendix, kidney, uterus, and biliary tract may metasta-
size to the greater omentum.7 A variety of nontubercular,
nonmalignant conditions can also present with similar ra-
diographic peritoneal patterns, such as granulomatous peri-
tonitis (inflammatory or infectious), mesenteric panniculi-
tis, and leiomyomatosis peritonealis disseminata.6
The diagnosis of coccidioidomycosis can be made in 3
ways: by directly demonstrating spherules in tissue, ex-
udates, sputum, or bronchoalveolar lavage specimens;
by recovering Cs immitis from cultured specimens; and by
serologic testing.8 Culture-positive patients are not them-
selves contagious and do not need to be isolated. How-
ever, culture plates that grow C immitis are highly infectious
and require special microbiologic handling. Complement
fixation is the most useful serologic test employed, because
it is a quantitative assay for coccidioidal immunoglobulin G
and adds useful prognostic information in following up
treated patients.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
The Infectious Diseases Society of America has re-
cently published guidelines for the treatment of coccid-
ioidomycosis.9 Disseminated disease may be treated with a
variety of agents, including the azoles (fluconazole or itra-
conazole) and the amphotericin B preparations. Effective
management of coccidioidomycosis can be difficult, and
infectious diseases consultation is recommended.
This case illustrates the ability of coccidioidomycosis to
present in an uncommon manner, in this case mimicking
ovarian carcinoma. Likewise, it is a reminder that many
diseases other than ovarian carcinoma may elevate CA 125
levels and produce radiographic abnormalities suggesting
peritoneal carcinomatosis. Finally, this case emphasizes the
importance of integrating medical and surgical specialties
to effectively manage patients with complicated infectious
diseases.
REFERENCES
1. Chiller TM, Galgiani JN, Stevens DA. Coccidioidomycosis.
Infect Dis Clin North Am 2003;17:41–57.
2. Phillips P, Ford B. Peritoneal coccidioidomycosis: case
report and review. Clin Infect Dis 2000;30:971– 6.
3. Bylund DJ, Nanfro JJ, Marsh WL. Coccidioidomycosis of
the female genital tract. Arch Pathol Lab Med 1986;110:
232–5.
4. DiSaia PJ, Creasman WT. Epithelial ovarian cancer. In:
DiSaia PJ, Creasman WT, editors. Clinical gynecologic oncol-
ogy. 6th ed. St. Louis (MO): Mosby; 2002. p. 289–315.
5. Blast RC, Xu FJ, Yu YH, Barnhill S, Zhang Z, Mills GB. CA
125: the past and the future. Int J Biol Markers 1998;13:
179 – 87.
6. Coakley FV. Staging ovarian cancer: role of imaging.
Radiol Clin North Am 2002;40:609 –36.
7. Sompayrac SW, Mindelzun RE, Silverman PM, Sze R. The
greater omentum. AJR Am J Roentgenol 1997;168:683–7.
8. Pappagianis D. Serologic studies in coccidioidomycosis.
Semin Respir Infect 2001;16:242–50.
9. Galgiani JN, Ampel NM, Catanzaro A, Johnson RH,
Stevens DA, Williams PL. Practice guideline for the treat-
ment of coccidioidomycosis: Infectious Diseases Society of
America. Clin Infect Dis 2000;30:658 – 61.
Received December 5, 2003. Received in revised form February 5, 2004.
Accepted February 11, 2004.
Ellis et al Coccidioidomycosis 1179
Abdominal Compartment
Syndrome Secondary to Ovarian
Mucinous Cystadenoma
Anne Chao, MD, Angel Chao, MD,
Yu Shiuan Yen, MD, and
Chi-Hsiang Huang, MD
Department of Anesthesiology, National Taiwan University Hospital and National
Taiwan University College of Medicine, Taipei, Taiwan; and Department of
Obstetrics and Gynecology, Graduate Institute of Clinical Medical Sciences, Chang
Gung Memorial Hospital and University, Taipei, Taiwan
BACKGROUND: Abdominal compartment syndrome second-
ary to a very large benign ovarian tumor has been rarely
reported in gynecology. With the increase of intraabdomi-
nal pressure in abdominal compartment syndrome, all
major organ systems are adversely affected, causing a po-
tentially fatal condition.
CASE: A 43-year-old woman presenting with a tensely
distended abdomen developed hypotension, difficulty in
ventilation, and anuria. An ovarian tumor complicated by
abdominal compartment syndrome was diagnosed, along
with hemodynamic decompensation. Prompt resuscitation
with immediate surgical removal of the tumor reversed the
life-threatening situation.
CONCLUSION: Timely aggressive resuscitation, prompt sur-
gical decompression, and intensive perioperative hemody-
namic management are required for patients with ovarian
mucinous cystadenoma complicated by abdominal com-
partment syndrome. (Obstet Gynecol 2004;104:1180 –2.
© 2004 by The American College of Obstetricians and
Gynecologists.)
Very large ovarian tumors occasionally occur in the
practice of modern medicine and, when complicated by
abdominal compartment syndrome, represent a surgical
emergency. Most are diagnosed and removed before the
association of marked tumor mass and, therefore, rarely
develops into abdominal compartment syndrome. The
management of a large benign ovarian tumor with ab-
dominal compartment syndrome poses a great challenge
to clinicians, because of impairment of all major organ
functions secondary to the increased intraabdominal
pressure.1 In this report, we share our experience in
managing a critically ill patient with abdominal compart-
ment syndrome resulting from a very large ovarian
mucinous cystadenoma.
Address reprint requests to: Chi-Hsiang Huang, MD, Department of
Anesthesiology, National Taiwan University Hospital and National
Taiwan University College of Medicine, 7, Chung-Shan South Road,
Taipei, Taiwan: e-mail: chuang@ha.mc.ntu.edu.tw.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1180 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
CASE
A 43-year-old woman had been admitted to the ward as
an emergency. Although a gradual abdominal distension
was noted in the past 2 years, the patient developed
difficulty in food intake and decreased urine amount
only a few days before admission. Physical examination
of the patient disclosed a tensely distended abdomen.
Abdominal ultrasonography and computed tomography
scan showed a large cystic tumor with little ascites in the
entire peritoneal cavity (Fig. 1). Chest X-ray showed
decreased lung fields without effusion. During the pre-
operative workup, the patient suddenly developed hypo-
tension (blood pressure dropped to 70/50 mm Hg, heart
rate 140 beats per minute) and severe dyspnea. Resusci-
tation was immediately initiated with endotracheal intu-
bation. The tidal volume of the ventilator was 500 mL
with a flow rate of 60 L/min, producing a high airway
pressure of 40 cm H2O (normally 35 cm H2O or less).
The PaO2/FIO2 ratio was 195 (normal 470 or more). The
dynamic lung compliance fell to 14 mL/cm H2O (40 – 80
mL/cm H2O in an intubated patient). Dopamine and
intravenous fluid infusion were administered. Because
her hemoglobin decreased from 13 to 10 g/dL, internal
bleeding was suspected, and blood components were
transfused. The patient remained anuric despite diuret-
ics infusion. Her serum urea nitrogen rose to 66.3 mg/dL
and serum creatinine to 3.9 mg/dL. The patient was
taken to the operating room under the diagnosis of
ovarian tumor complicated by abdominal compartment
syndrome. During laparotomy, a multilocular cystic tu-
mor, measuring 35 ! 25 ! 25 cm in size and weighing 17
kg, was completely excised from the left adnexa. Dark-
brownish fluid, 8,000 mL, was removed from the perito-
neal cavity. Despite the continuous intravascular fluid
support, blood pressure dropped to 80/50 mm Hg, and
central venous pressure fell to 7 mm Hg when the
peritoneum was opened. Hemodynamics were stabilized
by the administration of sodium bicarbonate, mannitol,
and dopamine and fluid support. The urine output re-
sumed shortly after the abdomen was decompressed.
The postoperative peak inspiratory pressure decreased
to 25 cm H2O, dynamic lung compliance increased to 31
mL/cm H2O, and oxygenation returned to normal. The
patient began tube feeding on the second postoperative
day, and the tracheal tube was extubated on the fourth
postoperative day. She was hospitalized for 13 more
days to ensure a complete recovery from preoperative
poor nutrition status and to guard against the refeeding
syndrome. The patient was discharged in good condi-
tion, and the pathology showed an ovarian mucinous
cystadenoma.
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000128106.96563.8b

Fig. 1. Computed tomography shows an intraabdominal
mass (margin identified by arrows) occupying the entire
peritoneal cavity.
Chao. Abdominal Compartment Syndrome. Obstet Gynecol 2004.
COMMENT
Abdominal compartment syndrome has rarely been re-
ported as a complication in very large benign ovarian
tumors.2 Abdominal compartment syndrome can result
from the accumulation of blood in the abdominal cavity,
bowel edema after direct bowel injury or compromise to
the mesenteric vessels, excessive crystalloid fluid resus-
citation, and inappropriate packing of the retroperitoneal
cavity. The diagnosis of abdominal compartment syn-
drome is made based on symptoms and signs caused by
an increase in intraabdominal pressure. Persistent intra-
abdominal pressure of 20 cm H2O or greater (normal
range ' 10 mm Hg) is a strong indication for decom-
pressing intervention.1 If uncorrected, abdominal com-
partment syndrome may lead to hypotension, hypoxia,
difficulty in ventilation, oliguria or anuria, and intestinal
ischemia. Reported etiologies of abdominal compart-
ment syndrome in gynecology (Table 1) include large
ovarian tumors,2 hemorrhage after gynecologic sur-
gery,3 ascites and bowel edema after debulking of malig-
nant mesodermal tumors and drainage of intraabdomi-
nal abscesses,3 and ovarian hyperstimulation syndrome
complicated by ovarian bleeding.4 Most of these cases
were treated successfully by surgical decompression, and
treatment resulted in good outcome.
Although ovarian mucinous cystadenomas are benign
in nature, excessive enlargement can result in dysfunc-
tion of adjacent intraabdominal organs by direct com-
pression, as in our patient. The abdominal cavity, unlike
the intracranial cavity, can accumulate a large volume
before the pressure rises rapidly. However, when a
critical pressure is reached, intraabdominal organs de-
compensate and lead to life-threatening sequelae.
Treatment for abdominal compartment syndrome in-
cludes immediate laparotomy to relieve the increase in
pressure. Drainage alone might be inappropriate for this
patient because internal bleeding was suspected. Great
vigilance must be exercised when the abdomen is opened
because severe hypotension and cardiac asystole could
occur.5,6 Therefore, sufficient fluid infusion and the use
of vasopressor are mandatory. Metabolic acidosis, re-
sulting partly from the washout of products of anaerobic
metabolism after reperfusion, can be corrected by bicar-
bonate. Decompression in abdominal compartment syn-

Table 1. Patient Characteristics of Abdominal Compartment Syndrome in Gynecology

Case Age (y) Etiology Management Outcome
1* 88 Ovarian tumor Stable hemodynamics, oliguria; decompressive Alive
laparotomy, 30 ! 25-cm tumor, 1,500 mL
fluid
2† 42 Hemorrhage after abdominal total hysterectomy Unstable hemodynamics; decompressive Dead
and BSO; alcoholic cirrhosis laparotomy with open abdomen technique,
repaired active bleeders
3† 54 Ascites and bowel edema; malignant Unstable hemodynamics; decompressive Alive
mesodermal tumor with intra-abdominal laparotomy, abscess drainage and
abscesses externalization of abdominal contents
4‡ 35 Ovarian bleeding after heparinization in ovarian Stable hemodynamics, acute renal failure; Alive
hyperstimulation syndrome with DVT placement of IVC filter; discontinuation of
Crohn’s disease with ileotomy heparin; short-term hemodialysis
5 43 Ovarian tumor Unstable hemodynamics, anuria; Alive
decompressive laparotomy, 35 ! 25 ! 25-
cm tumor, 8,000 mL fluid
BSO, bilateral salpingo-oophorectomy; DVT, deep vein thrombosis; IVC, Inferior vena cava.
* Celoria et al.2
†
von Gruenigen et al.3
‡
Cil et al.4

VOL. 104, NO. 5, PART 2, NOVEMBER 2004 Chao et al Abdominal Compartment Syndrome 1181
drome is the cornerstone for reversal of renal function.
Without decompression, anuria is unresponsive to the
diuretics.1,2 Patients should be guarded against pulmo-
nary edema because of rapid reexpansion of collapsed
lungs and fluid shifting during the surgery.
In conclusion, large ovarian mucinous cystadenomas
can result in increased intraabdominal pressure and sub-
sequent abdominal compartment syndrome. In turn,
abdominal compartment syndrome can present as a true
renal and cardiopulmonary emergency. Immediate and
careful surgical evacuation is required to decompress the
abdomen, restore the anatomy, and allow for the return
of normal physiological function.
REFERENCES
1. Ivatury RR, Diebel L, Porter JM, Simon RJ. Intra-abdomi-
nal hypertension and the abdominal compartment syn-
drome. Surg Clin N Am 1997;77:783– 800.
Treatment of Endometrial
Stromal Sarcoma With a
Gonadotropin-Releasing
Hormone Analogue
Cathy Burke, MRCOG, and
Kevin Hickey, MRCOG, FRCS(Edin)
Department of Gynaecology, Regional General Hospital, Limerick City, Ireland
BACKGROUND: Endometrial stromal sarcoma can present
management difficulties due to its lack of response to con-
ventional chemotherapy and radiotherapy. Various hor-
monal therapies have been shown to reduce tumor volume
in both primary and recurrent disease.
CASE: A woman who underwent myomectomy was discov-
ered to have a low-grade endometrial stromal sarcoma.
Treatment with the gonadotropin-releasing hormone
(GnRH) analogue triptorelin before surgery had produced
reduction in uterine size. The woman developed tumor
recurrence six months after definitive surgical treatment.
The tumor enlarged rapidly during a 2-month period, with
development of a right-sided hydronephrosis. Repeat ad-
ministration of triptorelin was accompanied by resolution
of the hydronephrosis and reduction in tumor volume.
Biopsy results confirmed recurrent low-grade endometrial
stromal sarcoma with moderate estrogen and progesterone
receptor positivity.
Address reprint requests to: Dr. Cathy Burke, National Maternity
Hospital, Holles Street, Dublin 2, Ireland; e-mail: catheeburke200@
yahoo.co.uk.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1182 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
2. Celoria G, Steingrub J, Dawson JA, Teres D. Oliguria from
high intra-abdominal pressure secondary to ovarian mass.
Crit Care Med 1987;15:78 –9.
3. von Gruenigen VE, Coleman RL, King MR, Miller DS.
Abdominal compartment syndrome in gynecologic surgery.
Obstet Gynecol 1999;94:830 –2.
4. Cil T, Tummon IS, House AA, Taylor B, Hooker G,
Franklin J, et al. A tale of two syndromes: ovarian hyper-
stimulation and abdominal compartment. Hum Reprod
2000;15:1058 – 60.
5. Morris JA Jr, Eddy VA, Blinman TA, Rutherford EJ,
Sharp KW. The staged celiotomy for trauma: issues in
unpacking and reconstruction. Ann Surg 1993;217:576–86.
6. Shelly MP, Robinson AA, Hesford JW, Park GR. Haemo-
dynamic effects following surgical release of increased intra-
abdominal pressure. Br J Anaesth 1987;59:800 –5.
Received January 5, 2004. Received in revised form February 27, 2004.
Accepted March 3, 2004.
CONCLUSION: Control of progression of a recurrent
endometrial stromal sarcoma was achieved with the GnRH
analogue triptorelin. This is the first report in the English-
language literature during a 30-year period of single-agent
GnRH analogue being an effective treatment intervention in
this context. (Obstet Gynecol 2004;104:1182–4. © 2004 by
The American College of Obstetricians and Gynecologists.)
Uterine sarcoma is a rare mesodermal tumor, account-
ing for approximately 3% of uterine cancers. Endome-
trial stromal sarcoma constitutes 15–25% of this group.
The diagnosis may be made after uterine curettage or
unexpectedly after hysterectomy. Primary treatment is
by total abdominal hysterectomy and bilateral salpingo-
oophorectomy.
Low-grade endometrial stromal sarcoma is character-
istically unpredictable in its behavior, often recurring or
metastasizing in the pelvis or beyond. Prolonged remission
or even cure is possible, however, after surgical resection of
metastases. Whole-abdomen radiotherapy has been re-
ported to decrease locoregional relapse rate,1 although it
fails to increase overall survival. Chemotherapy agents
such as doxorubicin have been used to treat uterine sar-
coma without improving either progression-free or overall
survival.2 The use of hormonal agents such as medroxy-
progesterone acetate, megestrol acetate, letrozole, and leu-
prolide acetate have been associated with a reduction in
tumor volume in both primary and recurrent disease.
CASE
A 29-year-old woman with menorrhagia and uterine
fibroids underwent elective myomectomy after 10
months of treatment with the gonadotropin-releasing
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000133533.05148.aa

Fig. 1. Tumor recurrence (arrow) seen posterior to the
bladder before commencement of gonadotropin-releasing
hormone analogue treatment.
Burke. GnRH Analogue/Stromal Sarcoma. Obstet Gynecol 2004.
hormone (GnRH) analogue triptorelin, which had
caused reduction in uterine size. At the time of surgery,
large intramural fibroids replaced most of the uterine
wall and extended to the mucosal surface. Subsequent
histology revealed a low-grade endometrial stromal sar-
coma. The patient underwent definitive surgery of total
abdominal hysterectomy, unilateral salpingo-oophorec-
tomy for coincident endometriosis and pelvic lymphad-
enectomy. No residual sarcoma was found in the uterine
specimen and all resected pelvic nodes were free of
tumor.
Six months later a 3 cm vaginal vault mass was found
on clinical examination and confirmed by computed
tomography (CT) scan. There was also dilatation of the
right mid ureter to 8 mm diameter. Within 2 months the
Fig. 2. Eight months after commencement of analogue
treatment, there is significant reduction in the size of tumor
recurrence.
Burke. GnRH Analogue/Stromal Sarcoma. Obstet Gynecol 2004.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
vaginal mass had increased in size to 6 cm, with accom-
panying right-sided hydronephrosis and hydroureter on
repeat CT imaging (Fig. 1). A cystoscopy and right
ureteric stenting was peformed and a vaginal vault nee-
dle biopsy taken. Histology of the biopsy specimen re-
vealed fibrous tissue only. The ureteric stent was passed
spontaneously through the urethra within 3 days and
was not replaced. Instead it was decided to recommence
treatment with the GnRH analogue triptorelin, given the
fact that uterine size had decreased during its initial use
before myomectomy.
A repeat CT scan performed 2 months after com-
mencing treatment showed no further increase in tumor
size. A vaginal vault wedge biopsy confirmed recurrent
low-grade endometrial stromal sarcoma. Moderate es-
trogen and progesterone receptor positivity was estab-
lished. By 8 months after commencement of the GnRH
analogue, the right hydroureter and hydronephrosis had
greatly resolved and tumor size had decreased (Fig. 2).
After discussion it was decided to proceed once more
to surgical management. Twenty-three months after her
initial surgery and after 12 months of treatment with trip-
torelin, the patient underwent resection of the vaginal vault
recurrence, distal right ureter, and posterior bladder wall,
with ureteric reimplantation and left salpingo-oophorec-
tomy. Twenty-one months elapsed after this surgery with
no evidence of tumor recurrence on follow-up.
DISCUSSION
Low-grade endometrial stromal sarcoma is typically a
slow-growing tumor for which chemotherapy and radio-
therapy seem to offer no survival benefit. Various med-
ical therapies working on the estrogen and progesterone
receptor have also been used in the management of this
cancer. These include the gonadotrophin-releasing hor-
mone analogue leuprolide acetate,3 medroxyprogester-
one acetate4 and the aromatase inhibitor letrozole.5
Scribner et al6 demonstrated a significant reduction in
tumor size of an initially inoperable low-grade endome-
trial stromal sarcoma with the use of a combination of
leuprolide acetate and megestrol acetate. To what extent
each of these compounds were individually responsible
for tumor shrinkage is unknown. Mesia and Demopou-
los3 retrospectively noted decrease in uterine size in a
woman who underwent a vaginal hysterectomy for uter-
ine fibroids after 2 once-monthly injections of leuprolide
acetate. Histologic examination in this case confirmed
leiomyomata and a coexistant low-grade endometrial
stromal sarcoma. Gonadotrophin–releasing analogue
treatment has successfully reduced uterine size in other
tumors such as leiomyosarcoma, where treatment with
Burke and Hickey GnRH Analogue/Stromal Sarcoma 1183
leuprolide acetate was instigated for presumed benign
uterine fibroids.7
A systematic literature search was performed of En-
glish language publications during the past 30 years
(1973–2003) using the MEDLINE search engine and
entering the terms “endometrial stromal sarcoma” and
“treatment.” This is the first report in the English language
literature of a woman being effectively treated with single-
agent GnRH analogue for recurrent endometrial stromal
sarcoma. Treatment effectiveness in this case was evi-
denced by arrest of tumor growth followed by reduction in
tumor size and resolution of an accompanying hydrone-
phrosis and hydroureter. We postulate that downregula-
tion of gonadotrophin-releasing hormone receptors at pitu-
itary level and the subsequent hypoestrogenic state
achieved by the administration of triptorelin to a woman
with one functioning ovary caused the observed reduction
in size of the recurrent tumor.
REFERENCES
1. Echt G, Jepson J, Steel J, Langholz B, Luxton G, Hernandez
W. Treatment of uterine sarcomas. Cancer 1990;66:35–9.
Hemolytic Uremic Syndrome
Presenting After Treatment of
Endodermal Sinus Tumor
Dele A. Ogunleye, MD, Sharyn N. Lewin, MD,
David G. Mutch, MD, Helen Liapis, MD, and
Thomas J. Herzog, MD
Department of Obstetrics and Gynecology, Good Samaritan Hospital, Cincinnati,
Ohio; Departments of Gynecologic Oncology and Pathology, Washington University
School of Medicine, St. Louis, Missouri; and Division of Gynecologic Oncology,
Columbia University, New York, New York
BACKGROUND: Hemolytic uremic syndrome is a rare multi-
system disorder that is caused by infections, preeclampsia,
autoimmune disorders, or oral contraceptive agents, and
rarely in association with different cancers and chemother-
apeutic agents.
CASE: A 34-year-old woman who presented for evaluation
of a pelvic mass received a diagnosis of International Fed-
eration of Gynecology and Obstetrics (FIGO) stage 1c
endodermal sinus tumor at laparotomy. Three months
after receiving 3 courses of bleomycin, etoposide, and cis-
platinum, she presented with renal failure, thrombocyto-
penia, anemia, and severe hypertension. Cancer-associated
Address reprint requests to: Thomas J. Herzog, MD, Division of
Gynecologic Oncology, Columbia University, 161 Ft. Washington
Avenue, New York, NY 10032; e-mail: th2135@columbia.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1184 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
2. Omura GA, Blessing JA, Major F, Lifshitz S, Ehrlich CE,
Mangan C, et al. A randomized clinical trial of adjuvant
adriamycin in uterine sarcomas: a Gynecologic Oncology
Group Study. J Clin Oncol 1985;3:1240 –5.
3. Mesia AF, Demopoulos RI. Effects of leuprolide acetate on
low-grade endometrial stromal sarcoma. Am J Obstet
Gynecol 2000;182:1140 –1.
4. Rand RJ, Lowe JW. Low-grade endometrial stromal sarcoma
treated with a progestogen. Br J Hosp Med 1990;43:154–6.
5. Maluf FC, Sabbatini P, Schwartz L, Xia J, Aghajanian C.
Endometrial stromal sarcoma:objective response to letro-
zole. Gynecol Oncol 2001;82:384 – 8.
6. Scribner DR, Walker JL. Low-grade endometrial stromal
sarcoma preoperative treatment with Depo-Lupron and
Megace. Gynecol Oncol 1998;71:458 – 60.
7. Meyer WR, Mayer AR, Diamond MP, Carcangiu ML,
Schwartz PE, DeCherney AH. Unsuspected leiomyosar-
coma: treatment with a gonadotrophin-releasing hormone
analogue. Obstet Gynecol 1990;75:529 –32.
Received November 18, 2003. Received in revised form January 26,
2004. Accepted February 81, 2004.
hemolytic uremic syndrome was diagnosed, and the pa-
tient was treated with plasmaphoresis, blood transfusion,
and hemodialysis.
CONCLUSION: Cancer-associated hemolytic uremic syn-
drome has a high mortality rate; thus, prompt diagnosis is
critical to survival. (Obstet Gynecol 2004;104:1184 –7.
© 2004 by The American College of Obstetricians and
Gynecologists.)
Complications that could result from chemotherapy in-
clude myelosuppression, which might present as anemia
and thrombocytopenia, and nephrotoxicity. However, a
rare, but rapidly fatal, complication of chemotherapy is
hemolytic uremic syndrome. Both these entities usually
present in the same fashion. We performed a MEDLINE
database search from 1966 to 2003, using the key words
“hemolytic uremic syndrome” and “chemotherapy,” in
English language medical journals and could find only
one other case report linked to use of bleomycin, etopo-
side, and cisplatinum, and none linked to chemotherapy
for treating endodermal sinus tumors of the ovary. The
difficulty regarding diagnosis and treatment of chemo-
therapy-associated hemolytic uremic syndrome is illus-
trated in the following case.
CASE
A 34-year-old woman presented to Washington Univer-
sity Medical Center for evaluation of a pelvic mass. She
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000142694.74553.4b
was seen by her gynecologist 4 weeks earlier with com-
plaints of pressure on urination, cramping, and occa-
sional diarrhea and bloating. Her medical history was
unremarkable except for a prior cesarean delivery.
Clinical examination revealed a pelvic mass that was
later confirmed on computed tomography scan to be a
10.8 cm ! 8.5 cm ! 7.9 cm mass containing multiple
solid and cystic components compressing the bladder.
Serum CA 125 was 119 U/mL, and alpha-fetoprotein
was 12,600 ng/mL. Staging laparotomy was performed.
Upon pathologic examination, the left ovary was re-
placed by a 435-g endodermal sinus tumor. Although
Schiller-Duval bodies were not identified, the tumor
showed the typical reticular histologic pattern. Abundant
hyaline globules corresponding to production of alpha- Fig. 1. Kidney biopsy showing concentric proliferation of
fetoprotein were present. An International Federation of smooth muscle cells in the wall (onion skinning). Adjacent
Gynecology and Obstetrics (FIGO) stage 1c was as- glomerulus contains fibrin thrombi in the capillary lumen
signed on the basis of involvement of the ovarian surface (arrow). Hematoxylin-eosin, ! 100, original magnification.
Ogunleye. Hemolytic Uremic Syndrome. Obstet Gynecol 2004.
by the tumor. Four weeks after surgery, the patient
began chemotherapy with a plan to receive bleomycin,
etoposide, and cisplatinum for 5 days every 3 weeks for Despite transfusion and aggressive diuresis, the pa-
a total of 3 courses. Subsequent follow-up visits showed tient’s respiratory status improved only marginally, and
no clinical evidence of recurrence. During chemother- thus she was intubated emergently and transferred to the
apy, her alpha-fetoprotein levels fell rapidly, and by 2 intensive care unit. The nephrology team was consulted
months after chemotherapy, it was 4.5 ng/mL. and, because of the constellation of signs and symptoms
Three months after chemotherapy, the patient was of renal failure, thrombocytopenia, anemia, and severe
seen at an outside facility for severe persistent nausea, hypertension, hemolytic uremic syndrome was sus-
vomiting, and profound weakness. Initial evaluation pected. A renal biopsy was performed and showed arte-
showed her vital signs to be stable. During transfer to our riolar changes ranging from onion skinning with com-
facility, however, she became acutely short of breath, her plete occlusion of arterial lumina to subendothelial
oxygen saturation dropped to approximately 65% on 6 edema and dilated glomerular capillary lumens with
liters of oxygen, blood pressure (BP) was 210/110 mm microthrombi (Fig. 1 Arrow). Electron microscopy dem-
onstrated subendothelial edema, focal fibrin deposits and
Hg, and pulse rate was 140 –150 beats per minute (bpm).
glomerular capillary wall thickening. This finding is
On admission, physical examination revealed an ic-
typical of thrombotic microangiopathy and is commonly
teric, thin, but well-nourished female. She was tachy-
seen in hemolytic uremic syndrome.
pneic, tachycardic, and in severe respiratory distress.
The patient was placed on nitroglycerin by continu-
Her vital signs revealed a pulse of 140 –150 bpm, tem-
ous intravenous drip and was subsequently stabilized on
perature of 36.3°C, BP 210/120 mm Hg, and a respira-
metoprolol and hydralazine. She underwent 3 courses of
tory rate of 40 cycles/min. Oxygen saturation by pulse
plasmapheresis and required dialysis to correct her rap-
oximetry was 47% on 10 L/min of O2. On auscultation,
idly worsening electrolyte status. She was extubated 2
her lungs had coarse crackles throughout lung fields, and days afterward and was observed for an additional 7
her heart sounds were S1, S2, with a grade 3/6 systolic days, during which time her liver enzymes returned to
ejection murmur. Abdomen was soft, nontender, with normal. She was discharged home with a slightly ele-
decreased bowel sounds. vated creatinine level of 1.8 mg/dL.
Significant laboratory findings were hemoglobin of 6.4
g/dL and hematocrit of 17.9%, platelets 88,000/#L, se-
rum alanine transaminase 1,943 IU, serum aspartate COMMENT
transaminase 4,478 IU, and creatinine 5.2 mg/dL. Re- Hemolytic uremic syndrome is a rare disorder that
sults of arterial blood gas analysis were as follows: pH may be primary or secondary to infectious diseases (eg,
–7.29, carbon dioxide partial pressure (PCO2) –27 mm Escherichia coli, Shigella sp), preeclampsia, autoimmune
hg, oxygen partial pressure (PO2) –54 mm Hg on 10 disorders, or use of contraceptive or cytotoxic agents.
L/min. Chest X-ray showed diffuse pulmonary edema. Diagnosis is made by demonstrating evidence of mi-

VOL. 104, NO. 5, PART 2, NOVEMBER 2004 Ogunleye et al Hemolytic Uremic Syndrome 1185
croangiopathic hemolytic anemia, thrombocytopenia,
and renal failure. In addition to this triad, other charac-
teristic signs are unexplained severe hypertension and
neurologic manifestations. The most common cytotoxic
agent implicated in the development of cancer-associated
hemolytic uremic syndrome is mitomycin C. In a review
of patients with this syndrome by Lesesne et al,1 84 of 85
patients had been treated with mitomycin C, and this
association was dose-dependent. This study also noted
that 88% of these patients had an adenocarcinoma,
mainly from nongynecological sites, and 35% of these
patients were in remission at the time of diagnosis. The
average elapsed time for the development of cancer-
associated hemolytic uremic syndrome from last mito-
mycin C dose was 3 months, and a 52% mortality was
observed from cancer-associated hemolytic uremic syn-
drome–related renal failure. One patient in Lesesne et
al’s1 study was treated with bleomycin and cisplatinum.
There have not been any case reports so far with
cancer-associated hemolytic uremic syndrome compli-
cating cisplatinum therapy without bleomycin, and there
are only 2 reports with bleomycin as a single agent.2 This
suggests that the role of these 2 drugs in the development
of hemolytic uremic syndrome might be synergistic.
Cisplatinum and bleomycin are both known to cause
vascular injury, which has led to the hypothesis that the
mechanism of cancer-associated hemolytic uremic syn-
drome with these 2 chemotherapeutic agents is endothe-
lial injury and vasospasm.3 Injured endothelial cells re-
lease large amounts of von Willebrand factor multimers,
resulting in focal platelet aggregation and thrombotic
microangiopathy that characterizes hemolytic uremic
syndrome pathology and promotes glomerular capillary
lumen occlusion and ischemia.4 An alternative theory of
pathogenesis involves the accumulation of soluble im-
mune complexes formed in the presence of saturating
amounts of tumor-associated antigens that bind to com-
plement receptors, causing platelet aggregation around
renal microvasculature and formation of occluding
thrombi.5
Mortality from cancer-associated hemolytic uremic
syndrome approaches 75%.6 The delay in making the
diagnosis may be a contributing factor. The difficulty in
making the diagnosis of this syndrome is highlighted by
this case in which there was a profound overlap in the
signs and symptoms of cancer-associated hemolytic ure-
mic syndrome and those of chemotoxicity-induced my-
elosuppression. In this case, the distinguishing signs of
cancer-associated hemolytic uremic syndrome were se-
vere hypertension and hypoxemia, secondary to noncar-
diogenic pulmonary edema. Our patient presented 3
months after receiving bleomycin-etoposide-cisplatin
chemotherapy with her alpha-fetoprotein levels normal-
1186 Ogunleye et al Hemolytic Uremic Syndrome
izing from 12,600 ng/mL to 4.5 ng/mL. Delayed diagno-
sis is common, with 76% of cancer-associated hemolytic
uremic syndrome patients presenting within a range of 1
day to 7 months following chemotherapy.7
The cornerstone of treatment is prompt diagnosis,
followed by aggressive supportive care and plasma-
pheresis or dialysis to manage fluid and electrolyte im-
balances. Transfusion in these severely anemic patients
should be carefully considered because this can often
exacerbate their symptoms.2,6 The most common thera-
pies have been based on inhibition of platelet function
with salicylates, suppression of immune function with
corticosteroids, cytotoxic agents, and plasmapheresis, or
symptom treatment (dialysis, antihypertensives). Unfor-
tunately, these treatments, either individually or com-
bined, often do not achieve more than transient clinical
improvement.6 The method of therapy that has been the
most encouraging is staphylococcal protein A immuno-
adsorption. Protein A is a staphylococcal cell-wall pro-
tein that binds to the Fc fragment of immunoglobulin G
(IgG) 1, 2, and 4. It preferentially binds to IgG immune
complexes. A multicenter study by Snyder et al8 estab-
lished that extracorporeal protein A immunoadsorption
was a safe and effective way of treating cancer-associated
hemolytic uremic syndrome. It was found to be most
effective in patients who developed this syndrome in
remission, with a 1-year survival of 74%, compared with
27% in patients with stable or progressing tumors and
22% in historical controls.
In summary, cancer-associated hemolytic uremic syn-
drome is a rare but real syndrome associated with bleo-
mycin, cisplatin, and other chemotherapies. It has a high
mortality rate, and early diagnosis is critical to improved
survival. Cancer-associated hemolytic uremic syndrome
must be included in the differential diagnosis of patients
with anemia, thrombocytopenia, and acute renal failure.
REFERENCES
1. Lesesne JB, Rothschild N, Erickson B, Korec S, Sisk R,
Keller J, et al. Cancer-associated hemolytic uremic syn-
drome: analysis of 85 cases from a national registry. J Cin
Oncol 1989;7:781–9.
2. Torra R, Poch E, Torras A, Bombi JA, Revert L. Pulmo-
nary hemorrhage as a clinical manifestation of hemolytic-
uremic syndrome associated with mitomycin C therapy.
Chemotherapy 1993;39:453– 6.
3. Doll DC, List AF, Greco FA, Hainsworth JD, Hande KR,
Johnson DH. Acute vascular ischemic events after cisplatin
based combination therapy for germ-cell tumors of the
testis. Ann Interm Med 1986;105:48 –51.
4. Licciardello JT, Moake JL, Rudy CK, Karp DD, Hong WK.
OBSTETRICS & GYNECOLOGY
 Elevated plasma von Willebrand factor levels and arterial
occlusive complications associated with cisplatin-based che-
motherapy. Oncology 1985;42:296 –300.
5. Korec S, Schein PS, Smith FP, Neefe JR, Woolley PV,
Goldberg RM, et al. Treatment of cancer-associated hemo-
lytic uremic syndrome with staphylococcal protein A immu-
noperfusion. J Clin Oncol 1986;4:210 –5.
6. Gardner G, Mesler D, Gitelman HJ. Hemolytic uremic
syndrome following cisplastin, bleomycin, and vincristin
chemotherapy: a case report and review of literature. Ren
Fail 1989;11:133–7.
Protrusio Acetabuli Presenting
as a Complex Pelvic Mass After
Total Hip Arthroplasty
Charles A. Leath III, MD,
Mack N. Barnes III, MD, and
Warner K. Huh, MD
Department of Obstetrics and Gynecology, Division of Gynecologic Oncology,
University of Alabama at Birmingham, Birmingham, Alabama
BACKGROUND: The differential diagnosis in an elderly
woman with a complex pelvic mass includes benign and
malignant diseases.
CASE: A woman in her ninth decade was discovered to have
both a breast mass and a pelvic mass at examination. After
diagnostic breast biopsy results that confirmed invasive
breast carcinoma, pelvic examination and ultrasonogra-
phy were performed. The ultrasonogram demonstrated a
large complex pelvic mass. The patient underwent an ex-
ploratory laparotomy at the time of her mastectomy. The
pelvic mass was a protrusio acetabuli resulting from a prior
right total hip arthroplasty.
CONCLUSION: A complex pelvic mass secondary to protru-
sio acetabuli is a rare clinical finding. (Obstet Gynecol
2004;104:1187–9. © 2004 by The American College of
Obstetricians and Gynecologists.)
Pelvic masses are often encountered in the geriatric
population. Characteristics of the mass, such as size,
unilateral versus bilateral masses, presence or absence of
ascites, Doppler features, and presence of solid compo-
nents, combined with family history, personal history,
and the presence or absence of symptoms related to the
mass, play an important role in the decision-making
process on the need for surgical exploration. In patients
diagnosed with breast cancer, a pelvic mass may repre-
Address reprint requests to: Charles A. Leath III, MD, Division of
Gynecologic Oncology, 619 19th Street South, OHB 538, Birming-
ham, AL 35249-7333; e-mail: Trey_Leath@yahoo.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
7. Canpolat C, Pearson P, Jaffe N. Cispaltin-associated hemo-
lytic uremic syndrome. Cancer 1994;74:3059 – 62.
8. Snyder HW Jr, Mittelman A, Oral A, Messerschmidt GL,
Henry DH, Korec S, et al. Treatment of cancer chemother-
apy-associated thrombotic thrombocytopenic purpura/he-
molytic uremic syndrome by protein A immunoadsorption
of plasma. Cancer 1993;71:1882–92.
Received November 17, 2003. Received in revised form January 7,
2004. Accepted February 18, 2004.
sent, among others, a colonic process such as diverticular
disease, unrelated uterine pathology, a benign ovarian
process, a primary ovarian malignancy, or metastatic
disease from the breast.
CASE
An ambulatory, healthy, postmenopausal female in her
ninth decade presented to her primary health care pro-
vider reporting a 16-month history of worsening urinary
frequency and urge incontinence symptoms. Initially,
her care had been managed by a urologist, who insti-
tuted nightly intermittent self-catheterization and 4 mg of
daily tolterodine therapy. These measures were success-
ful for approximately 14 months; however, her symp-
toms began to worsen over the previous 2 months, and
she sought additional medical care.
An evaluation by her primary health care provider
noted both a breast mass and a pelvic mass. Diagnostic
mammography was ordered, the results of which dem-
onstrated a mass in the patient’s right breast. Biopsy
results of her right breast mass confirmed invasive breast
carcinoma, and she was scheduled for breast surgery at
our institution. Additionally, the patient was referred to
the Continence Center at our institution for evaluation
of her worsening incontinence.
A review of the patient’s medical history revealed
well-controlled hypertension on &-blocker therapy, sco-
liosis, cataract surgery, total right hip arthroplasty, and a
history of breast cancer in her daughter. The patient
denied a personal or family history of other malignan-
cies, including all gynecologic cancers. During her incon-
tinence evaluation 1 day before breast surgery, a pelvic
examination by a urogynecologist noted a deviated cer-
vix and a right-sided pelvic mass, with bladder impinge-
ment and extension both into the posterior cul-de-sac
and to the umbilicus. Pelvic ultrasonography confirmed
a right-sided complex pelvic mass measuring 12.4 cm !
11.2 cm ! 11.0 cm that appeared to arise from the
adnexa, with minimal free peritoneal fluid present (Fig.
1). The differential diagnosis included a synchronous
0029-7844/04/$30.00 1187
doi:10.1097/01.AOG.0000121825.49900.2d
Fig. 1. Transvaginal ultrasound image of complex pelvic
mass appearing to arise from the right adnexa.
Leath. Protrusio Acetabuli in a Breast Cancer Patient. Obstet Gynecol 2004.
primary ovarian malignancy, benign ovarian process, or
metastatic breast carcinoma. The decision was made to
proceed with exploratory laparotomy with possible gy-
necologic cancer staging immediately after the patients’
mastectomy to limit the patient’s exposure to general
anesthesia.
Preoperative laboratory values were normal, and a
chest X-ray revealed no evidence of pulmonary disease.
After completion of the patient’s breast cancer surgery,
consisting of a right total mastectomy, sentinel lymph
node biopsy, and right axillary node dissection, she
underwent an exploratory laparotomy. Initial evaluation
of the abdomen and pelvis confirmed normal peritoneal
surfaces, without gross evidence of tumor. Additionally,
inspection of the mesentery, large and small intestines,
and appendix confirmed normal anatomy. The uterus
and adnexa were normal in appearance, but were dis-
placed to the left by a retroperitoneal pelvic mass arising
from the right paravesical space, extending to the sacral
promontory. The retroperitoneum was entered, and co-
pious yellowish fluid and foreign-body material were
encountered. A biopsy sample of the wall of the mass
was sent for permanent section.
The mass was copiously irrigated, and all foreign
body material was removed. After irrigation, screws and
orthopedic cement were visible emanating from the right
pelvic sidewall (Fig. 2). The orthopedic surgery service
was consulted and confirmed the diagnosis of protrusio
acetabuli. No additional recommendations were given
other than thorough irrigation of the site, and the lapa-
rotomy was abandoned. The final pathology report on
the right breast specimen revealed an invasive ductal
carcinoma, stage IIB (T2N1M0), and tamoxifen therapy
was instituted. The pelvic mass pathology report was
consistent with fibrinous and necrotic debris with local
1188 Leath et al Protrusio Acetabuli in a Breast Cancer Patient
Fig. 2. Intraoperative photograph of the right paravesical
space with orthopedic screw from total hip arthroplasty
identified.
Leath. Protrusio Acetabuli in a Breast Cancer Patient. Obstet Gynecol 2004.
calcifications. No metastatic disease was identified. A hip
X-ray on postoperative day 1 provided radiographic
confirmation of the diagnosis of protrusio acetabuli. The
patient was discharged on postoperative day 4, able to
bear weight on her lower extremities.
COMMENT
Multiple tumors, including breast carcinoma, may me-
tastasize to the ovary and present as an adnexal mass.1,2
Contemporary management of breast cancer consists of
breast conservation surgery when appropriate, followed
by adjuvant therapy consisting of a combination of che-
motherapy, radiation therapy, and hormonal therapy
with tamoxifen.3,4 More recently, sentinel lymph node
dissection has been shown to be a reasonable alternative
to a complete axillary dissection in appropriate candi-
dates.5 Findings suspicious for metastatic disease should
be evaluated, because they influence both prognosis and
treatment decisions.
Protrusio acetabuli is defined as an inward displace-
ment of the medial wall of the acetabulum with accom-
panying migration of the femoral head and was initially
described in 1824 by Otto,6 a German pathologist. Al-
though protrusio acetabuli may be a primary idiopathic
process, it is most often a secondary process attributed to
various benign and malignant conditions, including infec-
tious, inflammatory, metabolic, traumatic, genetic, and neo-
plastic ones, and including breast carcinoma.7 Postopera-
tive protrusio has been reported to have a prevalence as
high as 15% in a high-risk group of patients with rheuma-
toid arthritis undergoing total hip arthroplasty.8
Although associated with multiple medical conditions,
patients with protrusio acetabuli commonly present with
OBSTETRICS & GYNECOLOGY
activity-related pain in the groin as well as groin stiff-
ness.7 It is exceedingly rare for protrusio acetabuli to
present as a complex pelvic mass. A review of the litera-
ture listed in MEDLINE from 1966 to 2003 using the
search terms “protrusio acetabuli,” “pelvic mass,” and
“cancer” failed to identify any cases of protrusio acetabuli as
a presenting physical finding in the setting of malignancy.
Hattrup et al9 reported on a patient presenting with urinary
symptoms who underwent both abdominal exploration
and arthroplasty revision. Intraoperative findings con-
firmed a communication from the patient’s left hip to the
pelvis, which was present on hip arthrogram. Additionally,
a case of protrusio presenting as a solid adnexal mass in a
patient being given chronic steroid therapy has been report-
ed.10 Laparotomy confirmed a foreign body granuloma-
tous reaction to a polyethylene insert.
Management of protrusio acetabuli is variable and
dependent on several factors, such as diagnosis and
treatment of underlying medical illnesses that may con-
tribute or be the cause of the protrusion, patient age, and
skeletal maturity.7 Generally, the surgical approach used
is based on the maturity of the patient’s skeleton and
may include corrective surgery to close the triradiate
cartilage of the acetabulum, a valgus intertrochanteric
osteotomy procedure, resection arthroplasty and arthro-
desis, or revision arthroplasty.7 Although rare, the pos-
sibility of protrusio acetabuli should be included in the
differential diagnosis in patients with a history of a total
hip arthroplasty presenting with a pelvic mass.
REFERENCES
1. Mazur MT, Hsueh S, Gersell DJ. Metastases to the female
genital tract: analysis of 325 cases. Cancer 1984;53:1978–84.
Delayed Hemorrhage After
Cervical Conization Unmasking
Severe Factor XI Deficiency
Sarah H. Kim, MD, Sindhu K. Srinivas, MD,
Stephen C. Rubin, MD,
Louis J. Freedman, MD, and Heidi J. Gray, MD
Department of Obstetrics and Gynecology, University of Pennsylvania Medical
Center, Philadelphia, Pennsylvania; and Wilkes-Barre, Pennsylvania
BACKGROUND: Factor XI deficiency, a rare bleeding disor-
der found most commonly in patients of Ashkenazi Jewish
Address reprint requests to: Heidi J. Gray, MD, University of Pennsyl-
vania, Department of Obstetrics and Gynecology, 3400 Spruce Street,
Philadelphia, PA, 19104; e-mail: hgray@mail.obgyn.upenn.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
2. Curtin JP, Barakat RR, Hoskins WJ. Ovarian disease in
women with breast cancer. Obstet Gynecol 1994;84:449–52.
3. Fisher B, Anderson S, Bryant J, Margolese RG, Deutsch
M, Fisher, et al. Twenty-year follow-up of a randomized
trial comparing total mastectomy, lumpectomy, and
lumpectomy plus irradiation for the treatment of invasive
breast cancer. N Engl J Med 2002;347:1233– 41.
4. Systemic treatment of early breast cancer by hormonal, cyto-
toxic, or immune therapy: 133 randomised trials involving
31,000 recurrences and 24,000 deaths among 75,000 women.
Early Breast Cancer Trialists’ Collaborative Group. Lancet
1992;339:71–85.
5. Veronesi U, Paganelli G, Galimberti V, Viale G, Zurrida
S, Bedoni M, et al. Sentinel-node biopsy to avoid axillary
dissection in breast cancer with clinically negative lymph-
nodes. Lancet 1997;349:1864 –7.
6. Pomeranz MM. Intrapelvic protrusion of the acetabulum
(Otto pelvis). J Bone Joint Surg Am 1932;14:663– 86.
7. McBride MT, Muldoon MP, Santore RF, Trousdale RT,
Wenger DR. Protrusio acetabuli: diagnosis and treatment.
J Am Acad Orthop Surg 2001;9:79 – 88.
8. Ranawat CS, Dorr LD, Inglis AE. Total hip arthroplasty in
protrusio acetabuli of rheumatoid arthritis. J Bone Joint
Surg Am 1980;62:1059 – 65.
9. Hattrup SJ, Bryan RS, Gaffey TA, Stanhope CR. Pelvic
mass causing vesical compression after total hip arthro-
plasty: case report. Clin Orthop 1988;227:184 –9.
10. Mak KH, Wong TK, Poddar NC. Wear debris from total
hip arthroplasty presenting as an intrapelvic mass. J
Arthroplasty 2001;16:674 – 6.
Received August 1, 2003. Received in revised form September 7, 2003.
Accepted September 30, 2003.
background, may be present in patients with a history of
abnormal bleeding after elective surgery.
CASE: A patient of Ashkenazi Jewish descent presented
12 days after cervical conization for adenocarcinoma in
situ with severe vaginal bleeding requiring multiple
transfusions and uterine artery embolization. After a
thorough workup, a severe factor XI deficiency was
found. The patient ultimately required modified radical
hysterectomy for treatment of early cervical cancer.
With appropriate perioperative management, the pa-
tient underwent abdominal surgery without further
bleeding complications.
CONCLUSION: Factor XI deficiency can present with severe
bleeding episodes after elective surgery. Adequate pre-
operative assessment and perioperative management
are necessary to prevent bleeding complications in these
patients. (Obstet Gynecol 2004;104:1189 –92. © 2004
by The American College of Obstetricians and
Gynecologists.)
0029-7844/04/$30.00 1189
doi:10.1097/01.AOG.0000125882.72624.2c
Deficiency of factor XI is an autosomal recessive inher-
ited coagulopathy that leads to a more variable bleeding
tendency than hemophilia A or B. Although severely
deficient women are likely to bleed after trauma, surgery,
or other hemostatic challenge, there is evidence that
some partially deficient women are also at risk of signif-
icant bleeding. Therefore, careful planning of gyneco-
logic procedures is warranted in these women.
CASE
A 33-year-old multigravida of Ashkenazi Jewish back-
ground presented with excessive vaginal bleeding 12
days after conization of the cervix. She had undergone
an uncomplicated cold-knife cone biopsy for adenocarci-
noma in situ diagnosed by cervical biopsy. Initially, the
patient tolerated the outpatient procedure well, without
any significant blood loss. Twelve days after her proce-
dure, she presented at a hospital with heavy vaginal
bleeding. The patient was hemodynamically stable, and
coagulation studies demonstrated a hemoglobin level of
8.4 mg/dL and normal bleeding time, platelet count, and
prothrombin time (PT), but a prolonged activated partial
thromboplastin time (APTT) of 98 seconds (normal,
22.0 seconds to 33.3 seconds). A cervical stitch and
vaginal packing were attempted, but the patient contin-
ued to bleed heavily and was transferred to our hospital
for further management. On arrival, her hemoglobin
level was 5.4 mg/dL, requiring transfusion with 3 units of
packed red blood cells (PRBCs) while undergoing imme-
diate bilateral uterine artery embolization by interven-
tional radiology. The vaginal bleeding subsided initially
but then continued. Repeat laboratory studies revealed
an appropriate rise in hemoglobin, normal PT, and
persistently prolonged APTT of 50.5 seconds.
Further detailed history was obtained, and a hematol-
ogy consultation was arranged. The patient denied a
family history of bleeding disorders. She had 3 full-term
uncomplicated vaginal deliveries and described normal
periods. However, she did report a history of blood
transfusion after ovarian cyst rupture years ago, for
which she did not know the cause. The hematology
workup included testing for PT, APTT, coagulation
factor levels, von Willebrand’s factor (factor VIII), anti-
cardiolipin antibody, and &-2 glycoprotein assays. The
patient was noted to have an extremely low level of
factor XI at 3% (normal, 60 –140%) and a persistently
elevated APTT of 56.1 seconds. The rest of the labora-
tory results were normal. Therefore, she was diagnosed
with severe factor XI deficiency and received transfusion
of PRBCs and fresh frozen plasma (FFP) to maintain
factor XI levels greater than 25%. The patient ultimately
required 15 units of FFP and 4 units of PRBCs and was
1190 Kim et al Factor XI Deficiency
discharged on hospital day 6 with resolution of vaginal
bleeding.
Review of the cone biopsy pathology results revealed
an early invasive adenocarcinoma, necessitating a hys-
terectomy for curative treatment. The patient was admit-
ted 1 day before surgery. Her factor XI level was noted
to be 2% and APTT was 51.1 seconds. She received a
loading dose of 20 mL/kg FFP (6 U) 2 hours preopera-
tively and was maintained on 6 mL/kg every 12 hours
postoperatively. The goal of the treatment was to keep
factor XI levels above 25%. Factor XI levels and PTT
were followed up every 12 hours. She underwent a
modified radical hysterectomy with pelvic lymph node
sampling without complications. Postoperatively, factor
XI levels were in the range of 28% to 36%. The patient
was discharged on postoperative day 3 and continued
daily treatment with 3 U FFP for the next 7 days as an
outpatient without further bleeding episodes.
COMMENT
As shown in Fig. 1, the coagulation cascade is initiated
when factor VIIa is exposed to tissue factor at a site of
vessel injury. Factors VIII, IX, and XI are required for
production of factor Xa, because factor VIIa/tissue factor
complex undergoes feedback inhibition by tissue factor
pathway inhibitor. The resulting sequence of events
ultimately leads to the production of active enzyme
thrombin, which causes fibrin formation from fibrino-
gen. Thrombin has feedback activation by stimulating
factors V, VII, and XI. Thus, factor XI provides a
back-up coagulation pathway, because the factor VIIa/
tissue factor complex is inactivated by the extrinsic path-
way inhibitor.1
The exposure of blood to negatively charged surfaces,
such as collagen, results in activation of the contact
system of the intrinsic pathway of coagulation. Assembly
of prekallikrein, factor XII, and factor XI on a negatively
charged surface leads to activation of factor XI, thereby
propagating the intrinsic coagulation pathway. There-
fore, in factor XI– deficient patients, the secondary burst
of thrombin is decreased.
Recent studies have also shown that factor XI plays a
role in the down-regulation of fibrinolysis. Patients with
a factor XI deficiency are prone to bleed from tissues
with high local fibrinolytic activity, such as genitourinary
tract, oral cavity, nose, and tonsils. Bleeding occurs from
these tissue sites because factor XI deficiency does not
provide adequate down-regulation of fibrinolysis.2,3
Factor XI deficiency was originally described in 1953
by Rosenthal et al4 after several family members of
Ashkenazi Jewish background presented with excessive
bleeding after tonsillectomy and dental extractions. Fur-
OBSTETRICS & GYNECOLOGY

Fig. 1. Coagulation cascade. Coagulation is initiated at a site
of vessel damage when factor VIIa is exposed to tissue factor.
Factors VIII, IX, and XI are required for the production of
additional factor Xa due to feedback inhibition of the factor
VIIa/tissue factor complex by TFPI. The assembly of coagula-
tion proteins occurs on phospholipid surfaces, such as plate-
lets. Coagulation inhibitors (rectangles): TFPI, tissue factor
pathway inhibitor; APC, activated protein C; PS, protein S;
AT, antithrombin. Coagulation cofactors (ellipses): HMWK,
high molecular weight kininogen. Reprinted from Illustrated
Textbook of Paediatrics, Tom Lissauer and Graham Clayden,
page 308, ©1997, with permission from Elsevier.
Kim. Factor XI Deficiency. Obstet Gynecol 2004.
ther studies have confirmed that factor XI deficiency is
particularly common in the Ashkenazi Jewish popula-
tion, with a gene frequency of approximately 8%. This
disorder is sometimes referred to as hemophilia C and is
distinguished from hemophilia A (factor VIII deficiency)
and B (factor IX deficiency) by its occurrence in both
sexes and the absence of spontaneous bleeding into joints
and muscles.5
The inheritance is autosomal recessive, with severe
deficiency in homozygotes or compound heterozygotes
and partial deficiency in heterozygotes. Homozygotes
usually have factor XI levels less than 4%, whereas
heterozygotes have a wide range of levels between 15%
and 65%.4 However, severity of bleeding tendency is not
well correlated with factor XI levels. Several studies have
reported between 20% and 50% of individuals with only
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
partial deficiency may bleed excessively.1,5 Furthermore,
excessive bleeding may not occur immediately after sur-
gery but may present a week or so thereafter. The
reasons for the unpredictable nature of bleeding with
factor XI deficiency are not well understood. Coinciden-
tal inheritance of von Willebrand’s disease (vWD) has
been reported5; however, there is no increased incidence
in vWD over that expected in the general population.
Some patients also have been described with long bleed-
ing times and platelet defects. Therefore, the assessment
of any patient with factor XI deficiency should include
measurement of factor VIII and von Willebrand factor,
the bleeding time, and platelet function studies.
Prolonged and heavy menstrual bleeding is the most
common symptom reported by women with hemato-
logic deficiencies. In one study, the prevalence of menor-
rhagia in patients with factor XI deficiency was reported
as 59%, compared with 9% in the general population.1
During menstruation, hemostasis in the uterus is the result
of a delicate balance between platelet aggregation, fibrin
formation, vasoconstriction, and tissue regeneration on one
end and prostaglandins-induced platelet inhibition, vasodi-
latation, and fibrinolysis on the other.6 Therapeutic inter-
vention such as dilatation and curettage can worsen bleed-
ing, because curettage disrupts any platelet plug or fibrin
clots that may be in place.1
Individuals with factor XI deficiency are prone to
bleed after surgery or trauma to areas of the body where
the fibrinolytic activity is high, such as the oral cavity,
nasal passages, and urogenital tract. Therefore, frequent
clinical manifestations of factor XI deficiency in gyneco-
logic patients are bleeding after dental extractions and
nasal surgeries such as rhinoplasty and bleeding from the
uterus, either as menorrhagia or as bleeding in relation to
child birth and gynecologic surgery. The endometrium
has high fibrinolytic activity, and any surgical interven-
tions, even if minor, may be associated with risk of
bleeding.2 There are several case reports describing
abnormal postoperative bleeding leading to the diagno-
sis of factor XI deficiency after such diverse proce-
dures as sterotactic breast biopsy and transvaginal
oocyte retrieval.7,8
There is no consensus regarding whether factor XI
levels increase or decrease in pregnancy, but given the
lack of correlation between factor XI levels and bleeding
in partial deficiency, screening for the level in an asymp-
tomatic individual will not predict bleeding more than
the previous medical history alone.1
Fresh frozen plasma is the treatment of choice for
factor XI deficiency. In patients undergoing urologic
procedures, single doses between 10 mL/kg and 20
mL/kg given each day for 14 days has been shown to be
effective in preventing bleeding complications.4,9 The
Kim et al Factor XI Deficiency 1191
half-life of factor XI in plasma is generally considered to
be between 40 hours and 80 hours.4 There is no clear
consensus on the factor XI level required for hemostasis
during surgical procedures, but it has been suggested
that a nadir of 45 IU/dL and 30 IU/dL should be the goal
for major and minor surgeries, respectively.2 Mainte-
nance of factor XI levels at approximately 30% is gener-
ally sufficient, but some patients may need higher levels.4
One study showed the recovery of factor XI was 93%
and the rise in factor XI averaged 37% for each liter of
plasma infused.5 Notably, excessive bleeding in patients
with factor XI levels of approximately 40% has been
reported.4 The disadvantages of FFP are the large vol-
ume required, allergic reactions, and the potential for
transmission of infectious agents. Our patient developed
a mild transfusion reaction to the plasma, necessitating
use of intravenous steroids before transfusion for the
remainder of her treatment.
Factor XI concentrates have been developed and
given to some patients with factor XI deficiency. Some
reports of thrombotic events, such as severe cardiac
complications and pulmonary embolism, using factor XI
concentrates have been published, and this mode of
therapy is reserved for patients with severe deficiency in
factor XI. The dose of factor XI concentrates should not
exceed 30 U/kg and peak factor XI levels no more than
50 U/dL to 70 U/dL. Treatment with these concentrates
should be carried out in collaboration with hemophilia
centers, and it may be prudent to monitor thrombotic
markers.2
Antifibrinolytic drugs are effective and may be an
important adjunct in patients undergoing surgery. Tran-
examic acid is the drug of choice in treatment of menor-
rhagia in factor XI– deficient women. Few isolated cases
of intracranial thrombosis have been reported, but other
studies have shown that the incidence of thrombosis is
not greater than spontaneous thrombosis in women.10
Desmopressin can be used in those patients with factor
XI deficiency with vWD or von Willebrand factor levels
toward the lower end of the normal range. One study
reported a rise in factor XI by 15 U/dL to 20 U/dL,
whereas factor VIIIc and von Willebrand factor levels
1192 Kim et al Factor XI Deficiency
rose higher than expected.10 Desmopressin has major
advantages of safety and ease of use, but further studies
are needed to establish its role in factor XI deficiency.
As shown in the case presented, use of a preoperative
loading dose of FFP (20 mL/kg) followed by postopera-
tive maintenance consisting of daily FFP (6 mL/kg)
transfusions for 1 week to maintain factor XI levels may
be sufficient to prevent major bleeding complications
after major elective abdominal surgery in patients with
severe factor IX deficiency.
REFERENCES
1. Bolton-Maggs PH. Bleeding problems in factor XI defi-
cient women. Haemophilia 1999;5:155–9.
2. Kadir RA, Economides DL, Lee CA. Factor XI deficiency
in women. Am J Hematol 1999;60:48 –54.
3. Bouma BN, Meijers JC. Fibrinolysis and the contact sys-
tem: a role for factor XI in the down-regulation of fibrino-
lysis. Thromb Haemost 1999;82:243–50.
4. Steinberg MH, Saletan S, Funt M, Baker D, Coller BS.
Management of factor XI deficiency in gynecologic and
obstetric patients. Obstet Gynecol 1986;68:130 –3.
5. Bolton-Maggs PH. The management of factor XI defi-
ciency. Haemophilia 1998;4:683– 8.
6. Paper R. Gynecological complications in women with
bleeding disorders. Haemophilia 2000;6(suppl):28 –33.
7. Deutch BM, Schwartz MR, Fodera T, Ray DM. Stereotac-
tic core breast biopsy of a minimal carcinoma complicated
by a large hematoma: a management dilemma. Radiology
1997;202:431–3.
8. Battaglia C, Regnani G, Giulini S, Madgar L, Genazzani
AD, Volpe A. Severe intraabdominal bleeding after trans-
vaginal oocyte retrieval for IVF-ET and coagulation factor
XI deficiency: a case report. J Assist Reprod Genet 2001;
18:178 – 81.
9. Martlew VJ. Peri-operative management of patients with
coagulation disorders. Br J Anaesth 2000;85:446 –55.
10. Bolton-Maggs PH. Factor XI deficiency and its manage-
ment. Haemophilia 2000;6(suppl):100 –9.
Received August 26, 2003. Received in revised form November 14,
2003. Accepted January 5, 2004.
OBSTETRICS & GYNECOLOGY
Interstitial Pregnancy and
Transcervical Curettage
Xinmei Zhang, MD, Xinchang Liu, MD, and
Huaguang Fan, MD
Women’s Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang,
China
BACKGROUND: Interstitial pregnancies too large to be
treated with methotrexate are usually managed surgi-
cally, and that may adversely affect future fertility and
pregnancies. Transcervical curettage under laparoscopic
guidance may be possible in some cases if the pregnancy is
accessible vaginally.
CASE: Three women with interstitial pregnancy were
treated by transcervical suction curettage under laparo-
scopic guidance. In all cases, the procedure was quick,
bleeding was minimal, and there were no complications.
Removal was complete, and the serum "-hCG quickly
became undetectable.
CONCLUSION: Transcervical curettage under laparoscopic
guidance provides an alternative conservative treatment
for interstitial pregnancy. (Obstet Gynecol 2004;104:
1193–5. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
Interstitial pregnancy is uncommon, accounting for
2– 4% of all extrauterine pregnancies, and has a maternal
mortality rate as high as 2% to 2.5%.1 Although metho-
trexate can successfully resolve the pregnancy if it is
detected early enough,2 the usual treatment is cornual
resection carried out by laparotomy or via laparos-
copy.3–5 However, surgical treatment may result in a
blocked cornu or weakened uterine wall with reduced
fertility and a risk of uterine rupture in subsequent
pregnancies.6,7
The availability of high-resolution transvaginal ultra-
sonography and sensitive serum &-hCG measurements
allow earlier detection of interstitial pregnancy. This and
advances in endoscopic technology provide an opportu-
nity for a more conservative approach, such as transcer-
vical removal. The avoidance of the cornual transection
reduces the risk of tubal obstruction and subsequent
infertility while maintaining the integrity of the uterine
wall reduces the risk of rupture in future pregnancies.
Address reprint requests to: Xinmei Zhang, MD, Women’s Hospital,
Zhejiang University School of Medicine, 2 Xue Shi Road, Hangzhou
Zhejiang, P.R. China 310006; e-mail: xinmei6@yahoo.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
We have successfully used transcervical suction curet-
tage of the interstitial pregnancy under laparoscopic
guidance.
CASES
Between December 2000 and June 2002, 18 cases of
interstitial pregnancy presented to the Zhejiang Wom-
en’s Hospital. Three were suitable for treatment by
transcervical suction curettage under laparoscopic guid-
ance. The characteristics of these cases are presented in
Table 1. The institutional review board of Zhejiang
Women’s Hospital approved this study.
All the cases had a gestation between 7 and 10 weeks
and were diagnosed before the pregnancy ruptured or
presented as an acute crisis. The main presenting symp-
tom was vaginal bleeding, and the estimated loss was
between 30 to 40 mL. The diagnosis was made with
observation of a cornual mass with vaginal ultrasonog-
raphy and an elevated hCG level.
Laparoscopy in all the cases confirmed the diagnosis
that the cornua had not ruptured and that the tubes and
ovaries were otherwise normal (Fig. 1A).
Transcervical suction was carried out with a tech-
nique similar to that for intrauterine pregnancy. The
uterus was sounded to reach the cornu and to confirm
that a passage to the pregnancy could be established.
This was guided by laparoscopy to make sure that the
uterine wall was not perforated. A 5-mm suction cannula
was then inserted into the uterus, and maneuvered to
reach the gestational site. A negative pressure of 180–200
mm Hg was then applied to evacuate the products of
conception (Fig. 1B). Finally, the gestational site was ex-
plored by using a 3-mm curette to ensure that no debris
remained. This was carried out carefully by an experienced
operator and under the guidance of the laparoscopy to
ensure that the cornu was not perforated in the process.
In all cases, operative time was short (less than 18
minutes) and bleeding minimal (less than 50 mL). No
patient required a transfusion, and there were no subse-
quent complications. Chorionic villi were identified dur-
ing the curettage and confirmed by pathological exami-
nation. All patients were discharged from the hospital
within 3 days after the operation.
In all cases, serum &-hCG levels returned to undetect-
able levels by the fourth week postoperatively. Transvagi-
nal ultrasonography 6 weeks postoperatively showed the
cornual region to have a normal appearance. Hysterosal-
pingography 3 months postoperatively showed the cornu
concerned to be normal and patent.
Within 11⁄2 years, one of the patients had become
pregnant and delivered vaginally a healthy male infant of
3,300 g.
0029-7844/04/$30.00 1193
doi:10.1097/01.AOG.0000132807.44055.f5
Table 1. Interstitial Pregnancies Treated by Transcervical
Curettage
Case 1 Case 2 Case 3
Age (y) 26 27 30
Gestational age (d) 50 63 68
Serum &-hCG level (mIU/mL) 9,897 8,593 16,820
Diameter (cm) 3.1 3.0 3.6
Fetal cardiac activity Absent Absent Present
Vaginal bleeding (mL) 40 30 40
Blood loss (mL) 50 45 50
Time of procedure (min) 15 12 18

COMMENT
Newly developed technology has allowed early detec-
tion of interstitial pregnancy before its rupture and the
greater use of conservative medical and surgical manage-
ment.
Although cornuostomy (or salpingostomy) by lapa-
roscopy can be successfully carried out,3–5 the integrity
of the cornu is disrupted, and this compromises future
fertility and the safety of future pregnancies.
We report 3 cases of interstitial pregnancies that were
too large to be treated by methotrexate and showed that
they could be successfully managed with suction curet-
tage guided by laparoscopy. The procedure was easy
with little bleeding, and no complications, although care
and skill are needed during the final exploration of the
gestational cavity with a fine curette.
The main advantages of this method of treatment are
the preservation of uterine wall integrity and the main-
taining of cornual patency, avoiding the inevitable com- Fig. 1. A. Laparoscopic finding shows a large bulging
promise to future fertility and pregnancies. Additional ectopic mass in the right cornual region before evacuation
advantages are that the procedure is less invasive, the and an arrow indicates the gestational site. The interstitial
overall costs are less, the procedure takes little time, and pregnancy is easily distinguished from an angular preg-
there are no complications or prolonged hospital stays. nancy by its lateral location to the insertion of the ipsilat-
However, transcervical suction curettage is only ap- eral round ligament. B. A bulging section in the intersitial
propriate under a restricted set of circumstances. The region is largely reduced, and the intersitial region is clearly
patient must be hemodynamically stable. The pregnancy displayed after evacuation. Arrow 1 indicates the previous
needs to be situated in the proximal portion of the gestational site, arrow 2 indicates the ipsilateral fallopian
interstitium, preferably with a dilated proximal tubal tube, and arrow 3 indicates the ipsilateral round ligament.
Zhang. Interstitial Pregnancy and Transcervical Curettage. Obstet Gynecol
ostium, so that it is accessible vaginally. The pregnancy 2004.
must be diagnosed before its rupture.
In addition to making a correct diagnosis, the use of
laparoscopy is important in this procedure. It evaluates ultrasonography may not be able to detect minor perfo-
the size of the pregnancy and whether it is likely to rations during the procedure.
rupture during the procedure. If the uterine wall is
perforated during the procedure, the operator can take REFERENCES
remedial action, and in most occasions the uterine wall 1. Chen CL, Wang PH, Chiu LM, Yang ML, Hung JH.
can be repaired laparoscopically. Successful conservative treatment for advanced interstitial
It is possible to use ultrasonography rather than the pregnancy: a case report. J Reprod Med 2002;7:424 – 6.
laparoscope to guide this procedure. The main advan- 2. Fisch JD, Ortiz BH, Tazuke SI, Chitkara U, Giudice LC.
tage will be its less-invasive nature. However, the diag- Medical management of interstitial ectopic pregnancy: a case
nosis is less certain than direct visual inspection, and the report and literature review. Hum Reprod 1998;13:1981–6.

1194 Zhang et al Interstitial Pregnancy and Transcervical Curettage OBSTETRICS & GYNECOLOGY
3. Matsuzaki S, Fukaya T, Murakami T, Yajima A. Laparo-
scopic cornuostomy for interstitial pregnancy: a case report.
J Reprod Med 1999;44:981–2.
4. Katz DL, Barrett JP, Sanfilippo JS, Badway DM. Combined
hysteroscopy and laparoscopy in the treatment of interstitial
pregnancy. Am J Obstet Gynecol 2003;188:1113– 4.
5. Yoo EH, Chun SH, Kim JI. Endoscopic treatment of inter-
stitial pregnancy. Acta Obstet Gynecol Scand 2003;82:
189 –91.
Heterotopic Pregnancy in a
Natural Conception Cycle
Presenting as Hematometra
Po-Jen Cheng, MD, Ho-Yen Chueh, MD, and
Jian-Tai Qiu, PhD
Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital,
Taipei, Taiwan
BACKGROUND: Heterotopic pregnancy is a rare event in
natural conception cycles. Diagnosis of heterotopic preg-
nancy requires a high index of suspicion. Described herein
is a reported case of a heterotopic pregnancy presenting as
hematometra.
CASE: A young, multiparous woman, with her last men-
strual period 6 weeks before presentation, complained of a
dark reddish-brown vaginal discharge and progressive left
lower-quadrant discomfort during early pregnancy follow-
ing a natural conception cycle. An intrauterine mass was
observed, and subsequently a heterotopic pregnancy com-
plicated by hematometra was diagnosed with the help of
transvaginal ultrasonography and 3-dimensional power
Doppler ultrasound angiography.
CONCLUSION: Women with intrauterine fluid accumula-
tion during pregnancy may be at risk for coexistent ectopic
pregnancy. High resolution transvaginal color Doppler
sonography may be useful to identify a heterotopic preg-
nancy preoperatively. (Obstet Gynecol 2004;104:1195– 8.
© 2004 by The American College of Obstetricians and
Gynecologists.)
Heterotopic pregnancy, a rare event in which the coex-
istent gestations occur at 2 or more implantation sites, is
associated with a significant maternal morbidity and
mortality. The frequency is reported to be from 1 in
30,000 to 1 in 3,889.1 Tal et al2 have proposed that the
incidence of heterotopic pregnancy may rise up to 1 in 100
Address reprint requests to: Po-Jen Cheng, MD, Department of
Obstetrics and Gynecology, Chang Gung Memorial Hospital, 5, Fu-
Shin Street, Kwei-Shan, Tao-Yuan, 333, Taiwan, ROC; e-mail:
pjcheng@cgmh.org.tw.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
6. Broome JD, Vancaillie TG, Torode H. Conservative treat-
ment of interstitial pregnancy. Gynecol Endosc 1999;8:1– 4.
7. Moon HS, Choi YJ, Park YH, Kim SG. New simple endo-
scopic operations for interstitial pregnancies. Am J Obstet
Gynecol 2000;182:114 –21.
Received October 6, 2003. Received in revised form January 1, 2004.
Accepted February 5, 2004.
pregnancies where conception is assisted by ovulation in-
duction and in vitro fertilization. Diagnosis of heterotopic
pregnancy is often delayed and requires a high index of
suspicion. Clinical symptoms and signs, history, physical
examination, and laboratory and ultrasonographic features
are often nonspecific. Thanks to newer imaging modalities,
we are able to expand investigations of early embryonic
vascular features of heterotopic pregnancy implanta-
tion sites.3 After a thorough literature review (MEDLINE;
keywords: “heterotopic,” “pregnancy,” “hematometra”;
English and non-English languages; 1966 to December
2003), we describe the first case report of heterotopic preg-
nancy that was presented as a hematometra following a
natural conception cycle. The diagnosis was confirmed by
transvaginal ultrasonography and 3-dimensional power
Doppler ultrasound angiography.
CASE
A young, multiparous woman visited our gynecology clinic
with a dark reddish-brown vaginal discharge and progres-
sive left lower-quadrant discomfort. The patient reported 6
weeks of amenorrhea. The pregnancy test was positive. She
had no history of pelvic inflammatory disease, abortion,
venereal disease, intrauterine device use, abdominal sur-
gery, or treatment for induction of ovulation.
The patient was normotensive; pelvic examination
showed that she had an enlarged uterus with a closed
cervix and a tender left adnexum. Her blood type was A,
Rh-positive. Her hemoglobin concentration was 8.9
g/dL, leukocyte count was 8.8 ! 109/L, and platelet
count was 178 ! 109/L. Serum &-hCG concentration
was 4,878 IU/dL. Transvaginal ultrasonography dem-
onstrated the presence of an intrauterine pregnancy. The
small gestational sac was surrounded by a homogenous
sonolucent fluid-filled space measuring 5.4 ! 3.9 cm
(Fig. 1). In addition, a small amount of fluid was noted in
the cul-de-sac, and a 1.7 ! 1.7 cm eccentric echogenic
mass was present in the left fallopian tube. Three-dimen-
sional power Doppler ultrasound angiography was used
for diagnosis, using a fully digitized 3D device with a
complete storage of 3D power Doppler data (Kretz 730;
Kretz-Medison, Zipf, Austria and Seoul Korea) and a
0029-7844/04/$30.00 1195
doi:10.1097/01.AOG.0000142698.17433.cd
 Fig. 1. Transvaginal ultrasonography
demonstrating a small gestational
sac (G) and hematic fluid (E) within
the markedly dilated endometrial
cavity of the uterus. Some fluid is
present in the cul-de-sac (F).
Cheng. Heterotopic Pregnancy and Coexistent
Hematometra. Obstet Gynecol 2004.

6.25-MHz transvaginal 3D probe. The result showed 0.408) within the intrauterine sac and the left tubal
prominent blood flow with low impedance and high lesion. This finding was compatible with normal subtro-
diastolic blood flow velocity (resistance index: 0.412, phoblastic flow (Fig. 2).

Fig. 2. Three-dimensional power Doppler image of intrauterine gestation. The upper left image (A) is a sagittal image, the
upper right image (B) is a coronal image, the lower left image (C) is an axial image through the uterine corpus, and the lower
right image (D) is a rendered image of the villous vascular supply.
Cheng. Heterotopic Pregnancy and Coexistent Hematometra. Obstet Gynecol 2004.

1196 Cheng et al Heterotopic Pregnancy and Coexistent Hematometra OBSTETRICS & GYNECOLOGY
 Shortly after hospital admission, the patient com-
plained of a sudden worsening of abdominal pain. Re-
peated transvaginal ultrasonography showed an increase
in the amount of fluid rapidly accumulating within the
uterine cavity and cul-de-sac. Emergency laparoscopy
confirmed that the patient had an extrauterine preg-
nancy in the ampullary part of the left tube and approx-
imately 650 mL of blood in the peritoneal cavity. A
conservative salpingostomy with an evacuation of the
intratubal conceptus was performed, followed by a dila-
tion and curettage to remove gestational tissue and he-
matic fluid in the uterine cavity. The pathology report
confirmed a heterotopic pregnancy. The patient had an
uneventful postoperative course, and the serum &-hCG
concentration regressed to less than 3 IU/dL on day 23
after the surgical procedure. Her menstruation resumed
its normal course on day 43, lasting for 4 days, with
normal menstrual flow.
COMMENT
The preoperative diagnosis of a heterotopic pregnancy is
undoubtedly a major challenge for modern reproductive
medicine. Reece et al4 regard the common presenting
signs and symptoms for heterotopic pregnancy as ab-
dominal pain, adnexal mass, peritoneal irritation, and an
enlarged uterus. These presentations are, however, non-
specific and may be confused with other normal or
abnormal pregnancy manifestations. The ultrasound vi-
sualization of heart activity in both intrauterine and
extrauterine gestations is important for diagnosis, but
rare.5 Moreover, the appearance of the heartbeat may
differ in its time of onset.6 Fa and Gerscovich7 regard
transvaginal ultrasonography as an important aid in the
diagnosis of heterotopic pregnancy, but ultrasono-
graphic identification of ectopic pregnancy is low in its
sensitivity. It is difficult to differentiate an anembryonic
adnexal gestational sac from a hemorrhagic corpus luteal
cyst.
Recent developments in Doppler sonography enable
obstetricians to assess the physiology of feto-maternal
circulation and the noninvasive uteroplacental blood
flow velocity waveform.3 The process of placentation in
early pregnancy is a key period of hemodynamic
changes, mainly characterized by the establishment of
a continuous maternal blood flow in the intervillous
space. Fetal blood flow velocity waveform can be de-
tected as early as 5 weeks of gestation. The cutting edge
of advancement in ultrasound technology enables us to
expand investigations into placental flow during early
gestation.8
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
Three-dimensional power Doppler is able to depict
the integral 3D image of placental vascular network and
provides quantifiable data. Numerical data illustrate vas-
cular signals in the investigated volume of normal and
abnormal pregnancies. In this case, we found the tropho-
blastic blood flow velocity in both the intrauterine gesta-
tions and the heterotopic tubal pregnancy. With the use
of high resolution transvaginal color Doppler sonogra-
phy, an abnormal location of the gestational sac and a
high diastolic vascular flow were diagnosed, which aided
in the early preoperative diagnosis of unruptured heter-
otopic pregnancy.
In our patient, a well-defined sonolucent homogenous
mass surrounding an eccentrically located intrauterine
gestational sac had distended and thinned the outer
myometrium wall. This suggested a coexistent hema-
tometra. A thorough search of the literature using MED-
LINE, with the keywords “heterotopic pregnancy” and
“hematometra,” found no reference. This makes our
case report unique and significant. Accumulation of
hemorrhagic bleeding retrograde from the ectopic preg-
nancy and threatened intrauterine abortion may be the
possible causes of such an idiopathic complication. The
incidence of heterotopic pregnancy has been reported to
increase with assisted reproductive technologies.2 This
case should alert obstetricians that intrauterine fluid
accumulation during pregnancy should prompt an inves-
tigation for a coexistent ectopic pregnancy, with or with-
out assisted reproductive technology.
REFERENCES
1. Habana A, Dokras A, Giraldo JL, Jones EE. Cornual hete-
rotopic pregnancy: contemporary management options.
Am J Obstet Gynecol 2000;182:1264 –70.
2. Tal J, Haddad S, Gordon N, Timor-Tritsch I. Heterotopic
pregnancy after ovulation induction and assisted reproduc-
tive technologies: a literature review from 1971–1993. Fertil
Steril 1996;66:1–12.
3. Kurjak A, Hafner T, Kupešić S, Kostovic L. Three-dimen-
sional power Doppler in study of embryonic vasculogen-
esis. J Perinat Med 2002;30:18 –25.
4. Reece EA, Petrie RH, Sirmans MF, Finster M, Todd WD.
Combined intrauterine and extrauterine gestations: a
review. Am J Obstet Gynecol 1983;146:323–30.
5. van Dam PA, Vanderheiden JS, Uyttenbroeck F. Applica-
tion of ultrasound in the diagnosis of heterotopic preg-
nancy: a review of the literature. J Clin Ultrasound 1988;
16:159 – 65.
6. Hirsch E, Cohen L, Hecht BL. Heterotopic pregnancy with
discordant ultrasonic appearance of fetal cardiac activity.
Obstet Gynecol 1992;79:824 –5.
Cheng et al Heterotopic Pregnancy and Coexistent Hematometra 1197
7. Fa EM, Gerscovich EO. High resolution ultrasound in the
diagnosis of heterotopic pregnancy: combined transabdom-
inal and transvaginal approach. Br J Obstet Gynaecol 1993;
100:871–2.
8. Kurjak A, Kupešić S. Doppler assessment of the intervillous
Tuboovarian Abscess: A
Postoperative Complication of
Endometrial Ablation
Ted M. Roth, MD, and Michel E. Rivlin, MD
Division of Gynecology, Department of Obstetrics and Gynecology, University of
Mississippi Medical Center, Jackson, Mississippi
BACKGROUND: Global endometrial ablation may be associ-
ated with serious complications.
CASE: We present a case of bilateral tuboovarian abscesses
that developed 50 days postoperatively after a thermal
ablation. The patient underwent total abdominal hysterec-
tomy and bilateral salpingo-oophorectomy.
CONCLUSION: Infectious morbidity is known to occur after
thermal endometrial ablation. Further investigation is re-
quired into ways of reducing the risks of endometrial abla-
tion. (Obstet Gynecol 2004;104:1198 –9. © 2004 by The
American College of Obstetricians and Gynecologists.)
Late complications of endometrial resection and abla-
tion include incomplete resection, postresection pain re-
sulting from hematometra and synechiae, uterine rup-
ture during subsequent pregnancy, or postablation tubal
sterilization syndrome.1– 4 A recent review of complica-
tions associated with global endometrial ablation and a
search of the U.S. Food and Drug Administration
MAUDE (Manufacturer and User Facility Device Expe-
rience) database, yielded 85 complications in 62 pa-
tients.5 The majority of these complications arose soon
after the ablation procedure was performed. We present
a case of bilateral tuboovarian abscesses that developed
50 days after a thermal endometrial ablation and discuss
the mechanisms likely to be responsible for them.
CASE
A 46-year-old para 2 with a history of menorrhagia under-
went an uncomplicated thermal endometrial ablation
(ThermaChoice; Gynecare, Menlo Park, CA). Her history
was significant for 2 cesarean deliveries. A preoperative
physical examination and ultrasonogram were unremark-
Address reprint requests to: Ted M. Roth, MD, Department of Ob-
stetrics and Gynecology, UMC, 2500 North State Street, Jackson, MS
39206; e-mail: TROTH@jam.rr.com.
1198 Roth and Rivlin Tuboovarian Abscess
blood flow in normal and abnormal early pregnancy.
Obstet Gynecol 1997;89:252– 6.
Received January 13, 2004. Received in revised form February 23,
2004. Accepted March 3, 2004.
able. The uterus sounded to 8.5 cm, and her endometrial
biopsy was benign. She received no pretreatment with
gonadotropin-releasing hormone (GnRH) agonists.
The patient missed her 4-week return visit and pre-
sented on postoperative day 50 with complaints of ab-
dominal pain and fever for 1 week. On admission, her
physical examination was notable for a temperature of
38.5°C, moderate tenderness in each lower quadrant,
and exquisitely tender bilateral adnexal masses on bi-
manual examination. Ttransvaginal ultrasonography re-
vealed bilateral multicystic adnexal masses: the right was
6.9 ! 5.6 cm, and the left was 6.4 ! 5.3 cm. Notable
laboratory studies included a white blood cell count of 13
mm3 with 40% bands. She was started on intravenous
levofloxacin and metronidazole and failed to defervesce
after 72 hours. The decision was made to proceed with
surgery because of her continued fever and abdominal
pain in the setting of the adnexal masses. At laparotomy,
the patient had bilateral tuboovarian abscesses, extensive
adhesive disease, and an obliterated cul-de-sac. She un-
derwent lysis of adhesions, total abdominal hysterec-
tomy, and bilateral salpingo-oophorectomy. The vaginal
cuff was left open to drain the pelvis. Blood cultures were
negative, and cultures from the abscess grew Escherichia coli
and Bacteroides fragilis. Pathology confirmed the surgical
findings. Her postoperative course was unremarkable.
COMMENT
In their review of the MAUDE database, Gurtcheff and
Sharp5 found that the use of global endometrial ablation
devices was associated with significant complications: 9
of 85 complications reported were thermal injuries to
bowel or other structures, and there were 30 uterine
perforations. Although there were only 6 infectious com-
plications, 1 case in particular was associated with signif-
icant morbidity and mortality.
Our patient had bilateral tuboovarian abscesses 50
days postoperatively after thermal endometrial ablation.
Possible explanations for this occurrence include dissem-
ination of vaginal flora (despite thermal destruction),
tubal inoculation of vaginal organisms and transient
tubal occlusion (secondary to the ablation) leading to a
nidus for abscess formation, or preexisting salpingitis
that was somehow exacerbated by the procedure.
Although bacteremia after hysteroscopic surgery (en-
dometrial laser ablation and transcervical resection of
OBSTETRICS & GYNECOLOGY
the endometrium) has been shown to be significantly
higher without prophylactic antibiotics in one report, the
organisms obtained were of dubious clinical impact and
contamination could not be excluded in the majority of
cases.6 Although prophylactic antibiotics are typically rec-
ommended for hysteroscopic procedures, there is no con-
vincing evidence of their value. Our patient did not receive
antibiotics at the time of her procedure. Gurtcheff and
Sharp5 only recently recommended prophylactic antibiot-
ics at the time of endometrial ablation, based on the infec-
tions and the subsequent morbidity in their series. There
may also be a role for postablation antibiotics, which are
used for metritis prophylaxis when a pregnant uterus is
instrumented for abortion.
In addition to the recommendation by Gurtcheff and
Sharp,5 we feel that selected patients would benefit from
preoperative cervical cultures and a saline preparation of
the vaginal flora, specifically for bacterial vaginosis. Op-
erators should also limit the number of passes of the
ablation device into the cervix/uterus to limit the possi-
bility of ascending spread of vaginal flora. Ensuring a
proper balloon “fit” may also aid in thermal sterilization
of the uterine cavity and prevent infectious complica-
tions. Further study is required to help reduce the risks of
endometrial ablation.
Surgical Management of Vaginal
Vault Prolapse in a Woman With
a Neovagina and Pelvic Kidneys
Tristi W. Muir, MD, and Mark D. Walters, MD
Department of Obstetrics and Gynecology, The Cleveland Clinic Foundation,
Cleveland, Ohio
BACKGROUND: Women with Mayer-Rokitansky-Kuster-
Hauser syndrome have congenital absence of the uterus
and upper two-thirds of the vagina, which is frequently
accompanied by skeletal and renal anomalies. Mechanical
dilation or surgical creation of a vagina allows for function
but does not provide endopelvic fascial support of the
vagina. Vaginal prolapse may occur.
CASE: A 32-year-old woman presented with pelvic kidneys
and a 5-year history of prolapse of her mechanically cre-
ated neovagina. She underwent a sacrospinous ligament
suspension with a cadaveric fascia lata bridge. The apex of
the neovagina was 5 cm above the hymen 30 months
postoperatively.
Address reprint requests to: Tristi W. Muir, MD, Brooke Army
Medical Center, Attention: MCHE-OG, 385 Roger Brooke Drive,
Fort Sam Houston, TX 78234--6200; e-mail: muirtw@aol.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
REFERENCES
1. Brooks PG. Complications of operative hysteroscopy: how
safe is it? Clin Obstet Gynecol 1992;35:256 – 61.
2. Howe RS. Third-trimester uterine rupture following hystero-
scopic uterine perforation. Obstet Gynecol 1993;81(suppl):
827–9.
3. Hill DJ, Mahrer P. Pregnancy following endometrial abla-
tion. Gynaecol Endosc 1992;1:47–9.
4. Townsend DE, McCausland V, McCausland A, Fields G,
Kauffman K. Post-ablation tubal sterilization syndrome.
Obstet Gynecol 1993;82:422– 4.
5. Gurtcheff SE, Sharp HT. Complications associated with
global endometrial ablation: the utility of the MAUDE
database. Obstet Gynecol 2003;102:1278 – 82.
6. Bhattacharya S, Parkin DE, Reid TM, Abramovich
DR, Mollison J, Kitchener HC. A prospective randomised
study of the effects of prophylactic antibiotics on the inci-
dence of bacteraemia following hysteroscopic surgery. Eur J
Obstet Gynecol Reprod Biol 1995;63:37– 40.
Address reprint requests to: Ted M. Roth, MD, Department of
Obstetrics and Gynecology, UMC, 2500 North State Street,
Jackson, MS 39206; e-mail: TROTH@jam.rr.com.
Received January 25, 2004. Received in revised form March 1, 2004.
Accepted March 11, 2004.
CONCLUSION: An allograft colpopexy to the sacrospinous
ligament is an effective method of surgical treatment of
women with a prolapsed shortened vagina and an inacces-
sible presacral space. (Obstet Gynecol 2004;104:
1199 –201. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
Approximately 1 in 4,000 to 5,000 women have the
clinical syndrome of congenital absence of the uterus and
upper two thirds of the vagina, described as Mayer-Roki-
tansky-Kuster-Hauser syndrome. Skeletal and renal anom-
alies are also common in women with this syndrome.
Approximately one-third of women have abnormal kid-
neys.1 Methods of vaginal construction for these women
include mechanical dilation or reconstructive surgical pro-
cedures. The vagina created by dilation is often functional
for intercourse, but is not apically or laterally suspended in
the pelvis by endopelvic fascial attachments. Subsequent
prolapse may develop. We present a case of prolapse of a
neovagina in a woman with Mayer-Rokitansky-Kuster-
Hauser Syndrome. Fused pelvic kidneys complicated the
surgical management of the prolapse.
CASE
A 32-year-old woman presented with a 5-year history of
symptomatic vaginal vault prolapse. She described bulging
0029-7844/04/$30.00 1199
doi:10.1097/01.AOG.0000133534.85084.ea
of the vagina beyond the hymen. She had been diagnosed
with Mayer-Rokitansky-Kuster-Hauser syndrome at the
age of 13 years. She presented with a right pelvic mass and
was found to have vaginal agenesis on examination. Ultra-
sound evaluation revealed pelvic kidneys (the source of the
pelvic mass) and no evidence of a uterus. Chromosomal
analysis revealed a normal 46 XX karyotype. She later
proceeded with nonsurgical creation of a functional vagina
with the Ingram bicycle seat dilator system.2 Ultrasound
evaluation of the renal system identified crossed, fused
ectopic kidneys located in the right side of the pelvis. She
desired surgical management of her prolapse.
On examination in the dorsal lithotomy position, the
total vaginal length was 4 cm with eversion of the vaginal
apex 2 cm beyond the hymen with strain (pelvic organ
prolapse quantification system stage III apical prolapse).3
With the kidneys located in the pelvis, the patient was
not a candidate for suspension of the vagina to the
uterosacral ligaments or sacrum. The vagina was not
long enough to be directly sutured to the sacrospinous
ligament, therefore a fascia lata allograft was constructed
and sutured to the anterior and posterior vaginal walls
and secured to the right sacrospinous ligament.
The perineum was opened with a transverse incision.
Dissection was performed underneath the neovagina
posteriorly to the apex and mobilizing the entire vaginal Fig. 1. Fascia lata colpopexy to the sacrospinous ligament
tube. Allograft fascial strips (2 pieces measuring 10 cm ! in a woman with pelvic kidneys and stage III vaginal vault
3 cm) were secured with permanent suture to 2 cm of the prolapse of a neovagina. Illustration by David Schumick.
Reprinted with the permission of The Cleveland Clinic
proximal anterior vaginal wall and 3 cm of the posterior
Foundation.
vaginal wall. The fascial strips were trimmed to bridge
Muir. Sacrospinous Suspension With Allograft. Obstet Gynecol 2004.
the distance between the apex of the vagina and the
sacrospinous ligament. Three permanent sutures were
placed in the sacrospinous ligament. These were passed ectopic or absent kidneys, abnormalities of the renal
through the fascial graft to provide support without pelvis or ureters, a solitary fused kidney, atrophic kid-
tension (Fig. 1). neys, and malrotation.1 The urinary system of women
The patient did well postoperatively and was dis- with this syndrome should be evaluated at the time of
charged home on her second postoperative day. The diagnosis. Before proceeding with gynecologic surgery,
patient was sexually active before the surgery. She expe- it is essential to know the location of the kidneys and the
rienced dyspareunia in the first few months after the course of the ureters.
procedure and resumed use of a vaginal dilator. The A variety of methods are available to create a neova-
pain had improved by her 6-month postoperative evalu- gina. Although the incidence of prolapse of a neovagina
ation. Thirty months postoperatively, she was satisfied is unknown, these women are at risk for vaginal prolapse
with the results. On pelvic examination in the dorsal because the methods used to create a functional vagina
lithotomy position, the apex was 5 cm above the hymen do not generally address the suspensory or lateral sup-
at rest and 3 cm above the hymen with maximal strain port of the vaginal tube. Suspension of a prolapsed
(pelvic organ prolapse quantification system measure- neovagina has been successfully managed with abdomi-
ments Aa –1, Ba –1, C –3, tvl 5, Ap –2, Bp –2). nal sacral colpopexy.4 – 6 In our patient, crossed fused
ectopic kidneys located in the right pelvis limited the
surgical options for the treatment of her prolapse.
COMMENT The sacrospinous ligament suspension is an operation
Mayer-Rokitansky-Kuster-Hauser syndrome may in- that provides an excellent (more than 90%) long-term
volve vaginal atresia and urinary tract abnormalities. cure rate of apical prolapse.7 Complications associated
Congenital abnormalities of the renal system include with a sacrospinous ligament suspension are uncommon

1200 Muir and Walters Sacrospinous Suspension With Allograft OBSTETRICS & GYNECOLOGY
but include injury to the pudendal and sacral neurovas-
cular structures, recurrent prolapse, and de novo urinary
and fecal incontinence.8 The addition of an allograft
secured to the entire anterior and posterior vaginal walls,
bridged to the sacrospinous ligament allows this liga-
ment to provide support to the entire neovagina.
The allograft suspension of the vaginal vault to the
sacrospinous ligament provides a method of treatment that
does not compromise vaginal length and provides excellent
anatomic results. This procedure should be considered as a
surgical option for women with vaginal prolapse of a short
vagina, especially when the presacral area is not accessible.
REFERENCES
1. Griffin JE, Edwards C, Madden JD, Harrod MJ, Wilson JD.
Congenital absence of the vagina: the Mayer-Rokitansky-
Kuster-Hauser syndrome. Ann Intern Med 1976;85:224–36.
2. Ingram JM. The bicycle seat stool in the treatment of
vaginal agenesis and stenosis: a preliminary report. Am J
Obstet Gynecol 1981;140:867–73.
3. Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey
Milk-Alkali Syndrome in
Pregnancy
Michalis K. Picolos, MD, Charles R. Sims, MD,
Joan M. Mastrobattista, MD,
Mary A. Carroll, MD, and Victor R. Lavis, MD
Department of Internal Medicine, Division of Endocrinology, Diabetes, and
Metabolism, and Department of Obstetrics, Gynecology, and Reproductive Sciences,
The University of Texas Medical School at Houston, Houston, Texas
BACKGROUND: Severe hypercalcemia, a potentially life-
threatening medical emergency, is rare in pregnancy.
CASE: We report a 32-year-old woman presenting early in
the second trimester with severe hypercalcemia (total cal-
cium 22mg/dL), alkalosis, and acute renal insufficiency re-
sulting from excessive ingestion of calcium carbonate–con-
taining antacid for gastroesophageal reflux. The patient was
treated with aggressive hydration and furosemide, and re-
ceived 1 dose of intravenous etidronate, leading to short-term
symptomatic hypocalcemia. To our knowledge, this is the
third reported case of milk-alkali syndrome in pregnancy.
CONCLUSION: Milk-alkali syndrome is an uncommon cause
of hypercalcemia in pregnancy. Intravenous hydration
with saline should be the cornerstone of treatment, reserv-
ing bisphosphonates for selected cases. (Obstet Gynecol
Address reprint requests to: Michalis K. Picolos, MD, The University
of Texas Medical School at Houston, 6431 Fannin Street, MSB 6.100,
Houston, Texas 77030; e-mail: Michalis.K.Picolos@uth.tmc.edu.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
JO, Klarskov P, et al. The standardization of terminology of
female pelvic organ prolapse and pelvic floor dysfunction.
Am J Obstet Gynecol 1996;175:10 –7.
4. Matsui H, Seki K, Sekiya S. Prolapse of the neovagina in
Mayer-Rokitansky-Kuster-Hauser syndrome: a case report.
J Reprod Med 1999;44:548 –50.
5. Peters WA 3rd, Uhlir JK. Prolapse of a neovagina created
by self-dilatation. Obstet Gynecol 1990;76:904 – 6.
6. Schaffer J, Fabricant C, Carr BR. Vaginal vault prolapse
after nonsurgical and surgical treatment of Mullerian agen-
esis. Obstet Gynecol 2002;99:947–9.
7. Paraiso MF, Ballard LA, Walters MD, Lee JC, Mitchinson
AR. Pelvic support defects and visceral and sexual function
in women treated with sacrospinous ligament suspension
and pelvic reconstruction. Am J Obstet Gynecol 1996;175:
1423–30.
8. Lovatsis D, Drutz HP. Safety and efficacy of sacrospinous
vault suspension. Int Urogynecol J Pelvic Floor Dysfunct
2002;13:308 –13.
Received December 19, 2003. Received in revised form February 3,
2004. Accepted February 11, 2004.
2004;104:1201– 4. © 2004 by The American College of
Obstetricians and Gynecologists.)
Milk-alkali syndrome is a cause of hypercalcemia pro-
duced by ingestion of large amounts of calcium and
absorbable alkali. It is very rare in pregnancy. The most
common cause of hypercalcemia in pregnant women is,
by far, primary hyperparathyroidism (109 cases re-
ported from 1930 to 1990).1 Other causes, such as famil-
ial hypocalciuric hypercalcemia, parathyroid hormone
(PTH)-related peptide-secreting mammary gland hyper-
plasia, and malignancies (parathyroid, breast, biliary
tract, renal cell, pancreatic neuroendocrine tumor, and
multiple myeloma), have been described but are also
extremely uncommon. We report a case of milk-alkali
syndrome in the second trimester of pregnancy.
CASE REPORT
A 32-year-old white woman, gravida 2 para 0, presented
at 16 weeks of gestation with a 2-week history of nausea
and occasional vomiting. This progressed to 5 days of
intractable vomiting, weakness, and constipation and 1
day of lethargy and disorientation. Weight loss of 17
pounds was noted in a 2-week period. The patient had a
history of gastroesophageal reflux and had been taking a
proton pump inhibitor for 4 years before her pregnancy.
When pregnancy was confirmed, she discontinued the
proton pump inhibitor and used Tums (GlaxoSmith-
Kline, Pittsburgh, PA) instead (6 –10 tablets daily). Each
0029-7844/04/$30.00 1201
doi:10.1097/01.AOG.0000128109.44291.e2
Table 1. Laboratory Values of our Patient During Hospitalization and Follow-up.
Hospitalization Follow-up
Day 5 1 2 6
Laboratory value (normal range) Admission Day 1 Day 3 Day 5 (pm) Day 7 week weeks weeks
Total calcium (8.5–10.5 mg/dL) 22.0 20.2 10.5 8.6 7.7 8.0 7.4 9.5 9.2
Ionized calcium (4.65–5.20 mg/dL) 12.16 ... ... ... 4.00 4.56 4.10 4.80 ...
Sodium (135–145 mg/dL) 134 135 141 138 137 138 138 140 136
Potassium (3.5–5.1 mg/dL) 2.7 2.3 4.0 2.8 3.1 3.4 3.6 3.8 4.0
Chloride (95–109 mg/dL) 93 95 115 115 110 113 109 106 103
Bicarbonate (24–32 mmol/L) 35 35 22 18 19 20 20 24 21
Urea nitrogen (10–20 mg/dL) 18 17 14 8 5 4 6 9 11
Creatinine (0.5–1.4 mg/dL) 1.4 1.4 1.4 1.2 1.0 0.9 0.9 0.7 0.6
Magnesium (1.8–2.4 mg/dL) 0.8 0.8 1.1 1.3 1.5 2.1 1.1 1.4 1.7
Phosphorus (2.5–4.5 mg/dL) 0.8 1.5 1.1 2.8 2.4 2.8 2.8 3.5 2.6
Albumin (3.5–5.0 g/dL) 3.3 ... ... 2.2 ... 2.3 3.4 3.6 3.7
25-hydroxyvitamin D (20–57 ng/mL) ... 20 ... ... ... 17 ... ... 25
1,25-dihydroxyvitamin D (15–75 pg/mL) ... 8 ... ... ... 32 ... ... 49
Intact PTH (10–65 pg/mL) ... 1 $ 1.0 ... 148 ... ... ... 43
PTH-related peptide ($ 0.5 pmol/L) ... $ 0.3 ... ... ... ... ... ... ...
PTH, parathyroid hormone.

tablet contained 200 mg elemental calcium. Her intake of
milk and other dairy products was not excessive, be-
cause she was consuming less than 1,000 mg of elemental
calcium daily in the diet. Her past medical history was
significant for hypertension of 5 years, controlled with
labetalol during her pregnancy and atenolol with hydro-
chlorothiazide before that. Upon hospital admission,
examination revealed a lethargic woman oriented only
to person. Her weight was 217 pounds, blood pressure
142/95 mm Hg, pulse 93 beats per minute, and respira-
tory rate 20 breaths per minute. The remainder of the
physical examination was significant for dry oral mu-
cosa, flat jugular veins, and epigastric tenderness. Fetal
heart rate was 160 beats per minute. A maternal electro-
cardiogram showed nonspecific ST-T wave changes.
Laboratory results (Table 1) included total serum cal-
cium of 22 mg/dL, ionized calcium of 12.16 mg/dL
(normal, 4.65–5.20), and albumin 3.1 g/dL. Creatinine
was 1.4 mg/dL, urea nitrogen 18 mg/dL, and bicarbonate
35 mmol/L, values that are significantly elevated for
pregnancy. Magnesium was 0.8 mg/dL (normal 1.8 –
3.0), and phosphorus 0.8 mg/dL (normal 2.5– 4.5). Se-
rum amylase, lipase, alkaline phosphatase, liver en-
zymes, and thyroid function results were normal. Urine
pH was 5.0. Serum PTH was 1 pg/mL (normal, 10 – 65),
PTH-related peptide was less than 0.3 pmol/L (normal
less than 0.5), 25-hydroxyvitamin D was 20 ng/mL
(normal, 20 –57), and 1,25-dihydroxyvitamin D was 8
pg/mL (normal, 15–75). The above hormonal levels
were appropriately suppressed for the degree of hyper-
calcemia. Arterial blood gas revealed metabolic alkalosis,
with a pH of 7.49, and compensatory respiratory acido-
sis, with PCO2 of 43 mm Hg. Chest x-ray and mammog-
raphy results were normal.
The patient was aggressively hydrated with intrave-
nous isotonic saline (9 L during the first 24 hours and a
total of 22 L during the first 5 days of her hospital stay).
Clinical indices of volume status (urine output and con-
centration, neck veins and oral mucosa examination,
body weight, and thirst) were followed up to guide
hydration therapy. Because of the degree of severe
symptomatic hypercalcemia, it was thought appropriate
at the time to administer a single dose of 600 mg etidr-
onate intravenously on hospital day 1. After volume
repletion, she received a total of 2 doses of intravenous
furosemide (40 mg each) on day 3 and day 4 to enhance
calciuresis. Magnesium, potassium, and phosphorus
were supplemented. She required 20 grams of intrave-
nous magnesium sulfate to correct hypomagnesemia and
keep the magnesium values in the normal range during
the hospitalization. Additional treatments included cime-
tidine and antiemetics. She had clinically improved by
the second hospital day, with cessation of vomiting, and
was back to baseline status by day 4. The serum calcium
level normalized on day 5, and serum creatinine slowly
decreased to 0.9 mg/dL by day 6. Symptomatic hypocal-
cemia occurred on the evening of day 5, with tingling of
the extremities and a positive Chvostek sign (percussion
of the facial nerve at the top of the cheek in front of the
ear below the zygomatic bone caused contraction of the
facial muscle). Ionized serum calcium was 4.0 mg/dL,
and intact parathyroid hormone was appropriately ele-
vated to 148 pg/mL. The patient received calcium glu-
conate intravenously (180 mg of elemental calcium)
which relieved her symptoms, and she was started on
oral calcium carbonate.
At the time of discharge on day 7, the patient’s ionized
calcium level was borderline low, 4.56 mg/dL, and the

1202 Picolos et al Milk-Alkali Syndrome OBSTETRICS & GYNECOLOGY

remainder of the electrolytes were normal. She was
instructed to take 2 Tums daily (elemental calcium 200
mg/tablet) in addition to a prenatal vitamin. One week
later, the patient was asymptomatic. Ionized calcium was
4.1 mg/dL, and serum magnesium was 1.1 mg/dL. Tums
dosage was increased to 3 tablets daily, and oral magne-
sium was prescribed. Supplemental calcium carbonate
(Tums) was discontinued within the following 2 weeks.
The serum calcium remained within the normal range 6
weeks after hospital discharge.
COMMENT
Milk-alkali syndrome is characterized by the triad of
hypercalcemia, alkalosis, and renal insufficiency associ-
ated with the ingestion of large amounts of calcium and
absorbable alkali. The diagnosis is one of exclusion.2
Once a classic cause of hypercalcemia, due to ingestion
of large amounts of milk and alkali for treatment of
peptic ulcer disease according to the regimen introduced
by Sippy in 1915, the syndrome virtually disappeared by
the early 1980s due to the advent of H2 blockers. With
the use of calcium carbonate for osteoporosis, gastro-
esophageal reflux, peptic ulcer disease, and as a phos-
phate binder in renal disease, the incidence of the syn-
drome might be on the rise.
The initiating event for the development of milk-alkali
syndrome is probably the increased gastrointestinal input
paired with decreased renal excretion of calcium. Patients
with milk-alkali syndrome are volume depleted due to
increased renal tubular clearance of sodium and free water3
caused by the hypercalcemic effects on the renal medulla.
The situation is worsened by vomiting. In turn, volume
depletion causes contraction alkalosis and decreased glo-
merular filtration rate, resulting in reduction of calcium
excretion and worsening of the hypercalcemia.
The increase in intestinal calcium absorption observed
in pregnancy,4 a major maternal adaptation to the fetal
need for calcium, might be an important predisposing
factor to the development of milk-alkali syndrome in the
setting of high calcium intake. This increased calcium
absorption is thought to be mediated by 1,25-dihy-
droxyvitamin D, prolactin, and placental lactogen.4 De-
spite increased intestinal absorption, serum ionized cal-
cium levels are maintained at nonpregnancy values. This
is probably a result of the fetal uptake of calcium and the
increased maternal glomerular filtration rate.4
Prolonged hypercalcemia of maternal origin reported
in patients with primary hyperparathyroidism affects the
fetal circulation and can lead to suppression of fetal
parathyroid function and thus to neonatal hypocalcemia
and tetany.1 It has also been associated with spontaneous
abortion and stillbirth.4 Fetal effects of the short-lived
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
hypercalcemia of milk-alkali syndrome, if identified and
treated promptly, are unknown.
Through MEDLINE search and review of the pro-
posed cases we have identified 2 previously reported
patients with milk-alkali syndrome in pregnancy. In both
cases vomiting was one of the cardinal features. Both
patients were ingesting large amounts of calcium carbon-
ate and milk. The first patient presented in the second
trimester with pancreatitis and total serum calcium of
14.3 mg/dL.5 She was successfully treated with hydra-
tion. At 37 weeks she delivered a stillborn fetus with
short limbs, low-set ears, and no evidence of tissue
calcification at autopsy. If present, tissue calcification
could signify severe or prolonged fetal hypercalcemia.
The chromosomal analysis was normal, and no other
cause for the stillbirth was proposed. The second patient
presented in the third trimester with total serum calcium
of 22.5 mg/dL and was treated with hydration, furo-
semide, and hemodialysis.6 Delivery was uneventful 4
weeks later, with an uncomplicated neonatal course.
Our patient exhibited the classic metabolic changes of
milk-alkali syndrome: hypercalcemia, alkalosis, and acute
renal insufficiency. She was ingesting 1.2 to 2 grams of
elemental calcium in the form of calcium carbonate daily
and had appropriately suppressed PTH, PTH-related pep-
tide, and 1,25-dihydroxyvitamin D levels. She received 1
dose of etidronate (category C for pregnancy if used intra-
venously), which probably contributed to the development
of symptomatic hypocalcemia. Interestingly, hypocalcemia
has been reported not infrequently—even when conven-
tional treatment (hydration and diuretics) is used—in the
recovery phase of milk-alkali syndrome,7 but seems more
prevalent when bisphosphonates are used.8
After discontinuing the source of calcium and alkali,
the mainstay of treatment of hypercalcemia in the setting
of milk-alkali syndrome is aggressive volume repletion
with normal saline. It breaks the vicious circle of hyper-
calcemia, alkalosis, volume depletion, and renal insuffi-
ciency. Approximately 4 – 6 L of saline should be admin-
istered intravenously in the first 6 – 8 hours. The rate of
saline administration can be reduced later to 200 –300
mL/h until normocalcemia is achieved. Volume status,
electrolytes, and urine output should be monitored
closely. Sodium diuresis increases the excretion of potas-
sium and magnesium. Low levels of these electrolytes
can result in cardiac arrhythmias, so supplements should
be administered as needed to maintain normal serum
concentrations. If fluid overload occurs, the rate of the
saline infusion should be reduced, and furosemide
should be administered 20 – 80 mg intravenously on an
as-needed basis. A few doses of furosemide can also be
used after volume repletion to promote calciuresis. Fu-
rosemide is also indicated in patients with known con-
Picolos et al Milk-Alkali Syndrome 1203
gestive heart failure or those at risk for congestive heart
failure resulting from the aggressive volume replace-
ment. Loop diuretics should be used with caution, how-
ever, because they may have a more profound effect on
sodium excretion than on calcium excretion, resulting in
normal physiological renal sodium– conserving mecha-
nisms that can aggravate the hypercalcemia.
Calciuric therapy is not effective in patients with renal
failure from causes other than dehydration, so in these
situations dialysis with a calcium-free or low-calcium
solution can be considered. This is not usually the case in
patients with milk-alkali syndrome, but in cases of severe
hypercalcemia nonresponsive to other measures, it re-
mains as an option.
Calcitonin (category C for pregnancy), a peptide pro-
duced by the neuroendocrine C cells of the thyroid,
inhibits osteoclastic bone resorption and has a moderate
calciuric effect. It has a rapid onset of action, is safe in
dehydrated patients and patients with preexisting renal
failure, and can be used concomitantly with calciuric ther-
apy. Serum calcium concentrations may decline by 2–3
mg/dL within a few hours, and the nadir may be reached
within 24 hours after therapy is initiated. The usefulness of
calcitonin is limited to the first days of therapy due to
tachyphylaxis, but it can play a key role in treating patients
with life-threatening increases in calcium concentration.
Calcitonin 4–8 IU/kg can be used subcutaneously or intra-
venously every 6 –12 hours for 2–3 days. Common ad-
verse effects include nausea, rash, flushing, and malaise. A
skin test should be performed before initiating treatment to
evaluate for hypersensitivity reaction.
The bisphosphonates are a class of drugs that directly
inhibit resorption of bone by osteoclasts. Etidronate is
the least potent of the clinically evaluated bisphospho-
nates and is fairly effective in reducing calcium concen-
trations by more than 1 mg/dL per course of therapy. It
normalizes serum calcium in 40 –92% of patients with
hypercalcemia of malignancy. In patients with moderate
to severe hypercalcemia, 7.5 mg/kg/d can be adminis-
tered intravenously over 2– 4 hours, and the dose is
usually repeated for a minimum of 3 days. Serum cal-
cium concentrations begin to decline within 2 days and
reach a nadir within 7 days after administration of the
first dose. Duration of action is variable and ranges from
4 days to 6 weeks. Because the drug is excreted in the
kidneys it is important to adjust the dose in patients with
renal failure. The administration of bisphosphonates as
well as dialysis can be contemplated in patients with
1204 Picolos et al Milk-Alkali Syndrome
milk-alkali syndrome in situations of severe hypercalce-
mia with no or minimal improvement of calcium levels
after 36 – 48 hour use of conventional treatments. Lower
than usual doses of bisphosphonates can be used to
avoid the risk of hypocalcemia, because these patients do
not have a sustained calcium increasing factor (like pa-
tients with malignancy).
Milk-alkali syndrome is an uncommon cause of severe
hypercalcemia in pregnancy. Hyperemesis and in-
creased intestinal calcium absorption seem to play a
significant role in the development of the syndrome in
the setting of increased calcium intake.
Aggressive hydration with isotonic saline is critical in the
management of milk-alkali syndrome and should be paired
with close observation and serial laboratory evaluation.
Treatment with etidronate may carry a significant risk of
hypocalcemia and should probably be reserved for patients
unresponsive to aggressive hydration, diuretic therapy, and
calcitonin.
REFERENCES
1. Kelly TR. Primary hyperparathyroidism during preg-
nancy. Surgery 1991;110:1028 –33.
2. Orwoll ES. The milk-alkali syndrome: current concepts.
Ann Intern Med 1982;97:242– 8.
3. Fiorino AS. Hypercalcemia and alkalosis due to the milk-
alkali syndrome: a case report and review. Yale J Biol Med
1996;69:517–23.
4. Kovacs CS, Kronenberg HM. Maternal-fetal calcium and
bone metabolism during pregnancy, puerperium, and lacta-
tion. Endocr Rev 1997;18:832–72.
5. Ullian ME, Linas SL. The milk-alkali syndrome in pregnancy:
case report. Miner Electrolyte Metab 1988;14:208–10.
6. Kleinman GE, Rodriquez H, Good MC, Caudle MR.
Hypercalcemic crisis in pregnancy associated with excessive
ingestion of calcium carbonate antacid (milk-alkali syn-
drome): successful treatment with hemodialysis. Obstet
Gynecol 1991;78:496 –9.
7. Beall DP, Scofield RH. Milk-alkali syndrome associated
with calcium carbonate consumption: report of 7 patients
with parathyroid hormone levels and an estimate of preva-
lence among patients hospitalized with hypercalcemia.
Medicine (Baltimore) 1995;74:89 –96.
8. Camidge R, Peaston R. Recommended dose antacids and
severe hypercalcaemia. Br J Clin Pharmacol 2001;52:341–2.
Received December 16, 2003. Received in revised form February 10,
2004. Accepted March 18, 2004.
OBSTETRICS & GYNECOLOGY
Obstetric Management of
Klippel-Trenaunay Syndrome
Andrei Rebarber, MD,
Ashley S. Roman, MD, MPH,
Daniel Roshan, MD, and Francine Blei, MD
Departments of Obstetrics & Gynecology, Pediatrics, and Plastic Surgery, New
York University School of Medicine, New York, New York
BACKGROUND: Klippel-Trenaunay syndrome is a rare con-
genital disease characterized by extensive cutaneous vascu-
lar malformations, venous varicosities, focal abnormalities
of the deep venous system, and underlying soft tissue or
bony hypertrophy. Given the rarity of the disease, there is
little information available to counsel patients with Klippel-
Trenaunay syndrome regarding obstetric outcome.
CASES: We report our experience with 3 patients in whom
Klippel-Trenaunay syndrome complicated 4 pregnancies.
Successful delivery of a healthy infant at or beyond 36 weeks
of gestation was achieved in all pregnancies. One of the 4
pregnancies was complicated by pulmonary embolism.
CONCLUSION: Klippel-Trenaunay syndrome was once
thought to be a contraindication to pregnancy. With care-
ful management, successful pregnancies can be achieved.
(Obstet Gynecol 2004;104:1205– 8. © 2004 by The Amer-
ican College of Obstetricians and Gynecologists.)
Klippel-Trenaunay syndrome is a rare congenital disease
characterized by extensive cutaneous vascular malfor-
mations, venous varicosities, focal abnormalities of the
deep venous system, and underlying soft tissue or bony
hypertrophy. The morbidity of the disease is primarily
related to vascular abnormalities, which can result in
venous insufficiency, thrombophlebitis, cellulitis, limb
disparity, and thromboembolic disease.1 Additionally,
hemorrhage from underlying arteriovenous malforma-
tions in patients with a subtype of Klippel-Trenaunay-
Weber syndrome can result in thrombocytopenia and a
severe consumptive coagulopathy.2 Pregnancy in pa-
tients with Klippel-Trenaunay syndrome is believed to
be associated with exacerbation of these complications
and a high risk of thromboembolism and hemorrhagic
complications. For this reason, pregnancy has been dis-
couraged by some physicians in these patients.
Reprints are not available. Address correspondence to: Andrei Rebar-
ber, MD, Associate Professor and Associate Director of the Division of
Maternal Fetal Medicine, Department of Obstetrics & Gynecology,
New York University School of Medicine, 550 First Avenue, Suite 7N,
New York, NY 10016; e-mail: ar53@nyu.edu.
The authors thank the Klippel Trenaunay Support Group, especially Judy Vessey
and Erin Rosas, who participated in data retrieval and collection.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
Given the rarity of the disease, there is little medical
outcome information available to counsel patients with
Klippel-Trenaunay syndrome regarding obstetric care.
To date, 13 reports in the world literature describing
pregnancy in women with Klippel-Trenaunay syndrome
were identified using a MEDLINE search (keywords:
“Klippel Trenaunay syndrome,” “pregnancy,” “angioos-
teohypertrophy syndrome,” “Klippel Trenaunay Weber
syndrome”). Due to reporting bias, these cases may
represent the more severe outcomes. We report the
successful management of 3 women with Klippel-
Trenaunay syndrome through 4 pregnancies.
CASE 1
A 21-year-old gravida 1 para 0 previously diagnosed
with Klippel-Trenaunay syndrome presented at 11
weeks of gestation reporting worsening shortness of
breath upon exertion. Lower extremity Doppler studies
were obtained but were inconclusive due to multiple
superficial varicosities and a lack of a demonstrable deep
venous system. A ventilation and perfusion scan re-
vealed multiple small pulmonary emboli. Therapeutic
anticoagulation therapy was initiated using unfraction-
ated heparin, and an emergent Greenfield filter was
placed in the inferior vena cava. the patient transferred
care to our institution at this time. A thrombophilia
workup was initiated, with negative results. The patient
remained stable for the remainder of the pregnancy
receiving therapeutic anticoagulation therapy.
At 38 weeks gestation, ultrasound evaluation revealed
suspected intrauterine growth restriction, and the patient
was admitted to labor and delivery for induction of
labor. Under regional anesthesia, she delivered vaginally
a male infant weighing 2,940 g with Apgar scores of 9 at
1 minute and 9 at 5 minutes. The patient’s postpartum
course was complicated by endomyometritis, urinary
tract infection, and bacteremia. She was given ampicillin
with sulbactam for 3 days and was discharged home on
postpartum day 8. She was continued on anticoagulation
for 6 weeks postpartum.
CASE 2
A 32-year-old gravida 5 para 0050 with a known history
of Klippel-Trenaunay syndrome initially presented for
preconception counseling. Her disease was characterized
by right leg and breast cutaneous vascular malforma-
tions and limb hypertrophy, leading to multiple episodes
of thromboembolism. She had previously undergone
right breast reconstruction, toe amputations, right leg
stripping, and a Hickman catheter placement for venous
access. Her 5 prior spontaneous pregnancies resulted in
0029-7844/04/$30.00 1205
doi:10.1097/01.AOG.0000141649.11305.4b
medical terminations after counseling regarding the risks
of thromboembolism. At the end of this visit, it was
recommended that she self-administer low–molecular-
weight heparin anticoagulation therapy throughout the
entire pregnancy after an intrauterine pregnancy was
documented sonographically. Subsequent to this visit,
she conceived spontaneously, and prophylactic anticoag-
ulation was initiated after confirmation of an intrauterine
pregnancy.
Her pregnancy was complicated by progressive, se-
vere lower extremity pain, for which she was prescribed
narcotics for pain control. At 36 weeks of gestation, she
was noted to have massive vulvovaginal varicosities
which obstructed the birth canal. After amniocentesis
confirmed fetal lung maturity, she underwent a primary
cesarean delivery of a live female infant with a weight
appropriate for gestational age. She recovered without
complication and was discharged in stable condition. She
continued taking prophylactic doses of low-molecular-
weight heparin for 6 weeks postpartum.
Several months after delivery, a thrombophilia evalu-
ation was performed and revealed that she was heterozy-
gous for the prothrombin gene mutation. A subsequent
pregnancy in the same patient occurred 2 years later and
was uncomplicated as managed above.
CASE 3
A 29-year-old gravida 0 with Klippel-Trenaunay syn-
drome initially presented for preconception counseling.
She had a past medical history significant for left leg and
breast capillary malformation, left limb hypertrophy
leading to multiple episodes of thromboembolism,
splenic arteriovenous malformations, and pelvic vein
engorgement. She had undergone extensive surgeries,
including a left leg amputation (below the knee), splenec-
tomy, hemorrhoidectomy, pelvic vein embolization,
right ovarian cystectomy for an endometrioma, and
Greenfield filter placement. The filter was further com-
plicated by chronic obstruction, and inferior vena cava
wall stenting was necessary to maintain inferior vena
cava patency.
She conceived spontaneously despite previous failed
infertility cycles. A thrombophilia workup revealed that
the patient was heterozygous for the prothrombin gene
mutation. Throughout gestation, she was maintained on
therapeutic doses of low–molecular-weight heparin with-
out thromboembolic complications. After confirmation
of fetal lung maturity by amniocentesis, she delivered a
live female infant at 36 weeks of gestation by primary
cesarean secondary to large vulvovaginal varicosities.
Her postoperative course was complicated by a superfi-
1206 Rebarber et al Klippel-Trenaunay Syndrome
cial wound separation secondary to a thrombosed capil-
lary hemangioma bordering the right aspect of the verti-
cal skin incision.
COMMENT
Klippel-Trenaunay syndrome, the hereditary triad of
venous varicosities or abnormalities of the deep venous
system, superficial capillary malformations, and under-
lying osseous or soft tissue hypertrophy, presents diffi-
cult obstetric management issues. Sperandeo and col-
leagues3 described a family in which one first cousin had
Klippel-Trenaunay syndrome and the other had Beck-
with-Wiedemann syndrome. The data supported the
notion that insulin-like growth factor–2 overexpression
is involved in the cause of tissue hypertrophy observed
in different overgrowth disorders, including Klippel-
Trenaunay syndrome. Although information regarding
the prognosis and treatment of this disease in pregnancy
is limited, the normal physiologic changes associated
with pregnancy (increased plasma volume, cardiac out-
put, venous pressure, leg edema, and venous stasis) seem
to exacerbate the problems that characterize this syn-
drome, particularly the risks of thromboembolism and
hemorrhage. In addition, these patients are at risk for
bleeding diathesis from pelvic vascular malforma-
tions.4 – 6 Hemorrhage into the rectum, vagina, or vulva
can occur from abnormally dilated veins. Sonography
has been shown in several case reports to be an impor-
tant adjunct in following up these vascular malforma-
tions and developing a delivery plan.7,8 Massive vulvo-
vaginal varicosities can also lead to obstruction of the
introitus, mandating cesarean delivery.2 The placement
of regional anesthesia may be complicated by unrecog-
nized, underlying vascular anomalies and the develop-
ment of spinal hematomas.5,6
Fetal risks in these patients are not well understood.
The mode of inheritance for this disorder is not known,
and it is uncertain whether offspring of patients with
Klippel-Trenaunay syndrome are at increased risk of
inheriting the disorder. Prenatal diagnosis of Klippel-
Trenaunay syndrome using sonography has been de-
scribed in several case reports.9 Limb hemihypertrophy
and multiloculated cystic lesions at the level of the tho-
racic and abdominal wall have been described as mark-
ers for this disease.10 –12 In addition, it has been sug-
gested that fetuses in patients with Klippel-Trenaunay
syndrome may be at increased risk of developing intra-
uterine growth restriction.13
The arteriovenous malformations (AVM) as are seen
in Klippel-Trenaunay-Weber syndrome may further
complicate the pregnancy. These AVMs may involve
the pelvis, vulva, and uterus as well as the central ner-
OBSTETRICS & GYNECOLOGY
vous system and can lead to platelet destruction and
subsequent thrombocytopenia and coagulopathy.14
These malformations are also associated with an in-
creased risk of cerebrovascular accidents.14 Theoreti-
cally, the increase in blood pressure in labor (as is seen in
the second stage with Valsalva maneuver, for example)
may increase the risk of intracranial bleeding in patients
with CNS anomalies. If vascular AVMs are diagnosed
prenatally, the management of second stage of labor
should include operative delivery with avoidance of
pushing. Additionally, medications such as ergot alka-
loids should be avoided, because they may increase
arterial pressure and lead to an increased risk of cerebro-
vascular accidents.
We have successfully managed 3 women through 4
pregnancies. One of these patients was diagnosed with
pulmonary embolism during her pregnancy. Two of
these women were identified to be heterozygous carriers
of the prothrombin gene mutation. These findings indi-
cate that there may be an increased risk of thrombophilic
mutations in the Klippel-Trenaunay syndrome popula-
tion. As a result, we suggest that patients with Klippel-
Trenaunay syndrome be screened for hereditary and
acquired thrombophilic disorders, which include anti-
thrombin III deficiency, protein S deficiency, protein C
deficiency, factor V Leiden mutation, prothrombin gene
mutation (G20210A), plasminogen activator inhibitor
4G/4G homozygosity, the thermolabile variant of meth-
ylenetetrahydrofolate reductase, and antiphospholipid
antibody syndrome. Anticoagulation therapy should be
tailored to the patient’s individual history. Patients with
either an identified thrombophilic disorder or a prior
history of thromboembolic disease in the nonpregnant
state should be prophylactically anticoagulated with un-
fractionated heparin or low–molecular-weight heparin
during the pregnancy and in the postpartum period. If
the methylenetetrahydrofolate reductase variant is diag-
nosed, Vitamin B12, B6, and folate should be adminis-
tered. Patients with pulmonary embolism or deep ve-
nous thrombosis in the index pregnancy should be
prescribed therapeutic anticoagulation with these agents.
Those patients not fulfilling the above criteria may be
considered for daily baby aspirin therapy alone. These
recommendations are made based upon the cases re-
viewed and in the context of the American College of
Obstetricians and Gynecologists practice bulletin on an-
ticoagulation therapy during pregnancy.15
Although pelvic varicosities can complicate cesarean
delivery just as they complicate vaginal delivery, no
difficulty was encountered in the 3 cesareans performed
on patients 2 and 3. For prevention of hemorrhagic
complications, careful preconceptional and antenatal de-
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
lineation of vascular malformations in the pelvis are an
essential step in devising an appropriate delivery plan.
Widespread uterine vascular malformations in pregnant
Klippel-Trenaunay syndrome patients can be diagnosed
with prenatal ultrasound.8 Magnetic resonance imaging
of the maternal spine and brain to identify subclinical
vascular malformations is advised. Furthermore, cesar-
ean delivery should be considered in the event of ob-
structive vulvar vascular malformations. Anesthesia and
vascular team consultation is strongly advised before
delivery to optimize patient care and safety.
At the present time, there is a dearth of information
to describe the natural course of Klippel-Trenaunay
syndrome in pregnancy. Although patients with Klippel-
Trenaunay syndrome are at high risk for thromboem-
bolic events and hemorrhage during pregnancy, success-
ful maternal and fetal outcomes can be achieved, as
demonstrated by our experience with 3 patients.
REFERENCES
1. Gloviczki P, Stanson AW, Stickler GB, Johnson CM,
Toomey BJ, Meland NB, et al. Klippel-Trenaunay syn-
drome: the risks and benefits of vascular interventions.
Surgery 1991;110:469 –79.
2. Neubert AG, Golden MA, Rose NC. Kasabach-Merritt
coagulopathy complicating Klippel-Trenaunay-Weber syn-
drome in pregnancy. Obstet Gynecol 1995;85(5 Pt 2):831–3.
3. Sperandeo MP, Ungaro P, Vernucci M, Pedone PV, Cer-
rato F, Perone L, et al. Relaxation of insulin-like growth factor
2 imprinting and discordant methylation at KvDMR1 in two
first cousins affected by Beckwith-Wiedemann and Klippel-
Trenaunay-Weber syndromes. Am J Hum Genet 2000;66:
841–847.
4. Fishman A, Paldi E. Klippel-Trenaunay syndrome with
complications during pregnancy "in Hebrew#. Harefuah
1989;116(3):147– 8.
5. Dobbs P, Caunt A, Alderson TJ. Epidural analgesia in an
obstetric patient with Klippel-Trenaunay syndrome. Br J
Anaesth 1999;82:144 – 6.
6. Felten ML, Mercier FJ, Bonnet V, Benhamou D. Obstetric
analgesia in patients with Klippel-Trenaunay syndrome
"in French#. Ann Fr Anesth Reanim 2001;20:791– 4.
7. Verheijen RH, vanRijen-de Rooij HJ, van Zundert AA, de
Jong PA. Pregnancy in a patient with the Klippel-Trenau-
nay-Weber syndrome: a case report. Eur J Obstet Gynecol
Reprod Biol 1989;33(1):89 –94.
8. Richards DS, Cruz AC. Sonographic demonstration of
widespread uterine angiomatosis in a pregnant patient
with Klippel-Trenaunay-Weber syndrome. J Ultrasound
Med 1997;16:631–3.
9. Christenson L, Yankowitz J, Robinson R. Prenatal diagno-
sis of Klippel-Trenaunay-Weber syndrome as a cause for
Rebarber et al Klippel-Trenaunay Syndrome 1207
 in utero heart failure and severe postnatal sequelae. Prenat
Diagn 1997;17:1176 – 80.
10. Hatjis CG, Philip AG, Anderson GG, Mann LI. The in-utero
ultrasonographic appearance of Klippel-Trenaunay-Weber
syndrome. Am J Obstet Gynecol 1981;139:972– 4.
11. Warhit JM, Goldman MA, Sachs L, Weiss LM, Pek H.
Klippel-Trenaunay-Weber syndrome: appearance in
utero. J Ultrasound Med 1983;2:515– 8.
12. Lewis BD, Doubilet PM, Heller VL, Bierre A, Bieber FR.
Cutaneous and visceral hemangiomata in the Klippel-
Trenaunay-Weber syndrome: antenatal sonographic
detection. AJR Am J Roentgenol 1986;147:598 – 600.
13. Fait G, Daniel Y, Kupferminc MJ, Gull I, Peyser MR,
Detection of Fetal Lactate With
Two-Dimensional-Localized
Proton Magnetic Resonance
Spectroscopy
Julian N. Robinson, MD,
Jane Cleary-Goldman, MD,
Fernando Arias-Mendoza, MD, PhD,
Jaime Cruz-Lobo, MD, Clare Tempany, MD,
Robert V. Mulkern, PhD,
Bruce B. Feinberg, MD, and
Truman R. Brown, PhD
Department of Obstetrics and Gynecology, Columbia Presbyterian Medical Center,
New York, New York; Department of Radiology, Brigham and Women’s
Hospital, Boston, Massachusetts; and Hatch Research Unit, Department of
Radiology, Columbia Presbyterian Medical Center, New York, New York
BACKGROUND: Antenatal surveillance is inefficient for accu-
rately detecting fetal compromise. A noninvasive tech-
nique for assessing fetal metabolic status would be useful
for clinical management.
CASE: Fetal magnetic resonance spectroscopy was per-
formed at 20 weeks in a pregnancy complicated by severe
intrauterine growth restriction to determine if lactate, a
metabolite associated with fetal hypoxia, was present. Two-
dimensional, single-slice proton magnetic resonance spec-
troscopy was carried out at 1.5 T using a volume-selective,
double-spin echo technique. Lactate was detected in fetal
back muscle. Fetal death occurred the next day.
CONCLUSIONS: Although this initial report is purely exper-
imental, further development of this technique may prove
to be a valuable noninvasive tool in the management of
Reprints are not available. Address correspondence to: Julian N.
Robinson, Department of Obstetrics and Gynecology, Brigham and
Women’s Hospital, 75 Francis Street, Boston, MA 02115; e-mail:
JNRobinson@partners.org.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1208 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
Lessing JB. Klippel-Trenaunay-Weber syndrome associ-
ated with fetal growth restriction. Hum Reprod 1996;
11(11):2544 –5.
14. Pollack RN, Quance DR, Shatz RM. Pregnancy compli-
cated by the Klippel-Trenaunay syndrome. A case report.
J Reprod Med 1995;40(3):240 –2.
15. American College of Obstetricians and Gynecologists.
Thromboembolism in pregnancy. ACOG Practice Bulle-
tin No. 19. Washington, DC: ACOG; 2000.
Received January 4, 2004. Received in revised form March 3, 2004.
Accepted March 18, 2004.
suspected fetal hypoxia. (Obstet Gynecol 2004;104:
1208 –10. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
The management of severe intrauterine growth restric-
tion (IUGR) remote from delivery is one of the most
challenging dilemmas in modern obstetrics. Although
diverse factors may impair fetal well-being, uteroplacen-
tal insufficiency resulting in fetal hypoxia and lactic
acidosis is a recognized etiology of IUGR that can lead to
multiorgan failure, including adverse neurological se-
quelae. The obstetrician balances expectant manage-
ment to avoid the complications of prematurity versus
interruption of the pregnancy before irreversible multi-
organ injury ensues.
Indirect methods, including the nonstress test, the
biophysical profile, and Doppler velocimetry, are cur-
rently used to assess fetal well-being. None of these
examinations discriminate accurately between pregnan-
cies requiring delivery to optimize neonatal well-being
and those that do not. This deficiency in fetal testing may
be due to the inherent limitations of predicting a meta-
bolic state from the physiological parameters assessed
with indirect methods. Fetal serum acid base balance can
be quantified before delivery using percutaneous fetal
umbilical cord blood sampling. Such a strategy is difficult
to extrapolate to routine antenatal testing.
Lactate is a metabolite readily accessible by proton
magnetic resonance spectroscopy and has been detected
in human magnetic resonance spectroscopy studies.1
Reproducibility and reliability have been studied and are
reasonably understood in the adult population.2 Kok et
al3 have performed fetal magnetic resonance spectros-
copy studies, thus establishing its feasibility. We report a
case of antenatal fetal lactate detection in a severely
compromised pregnancy, using magnetic resonance
spectroscopy as a noninvasive test of fetal metabolic
status.
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000142697.09330.77
Fig. 1. Spectra of fetal back muscu-
lature.
Robinson. Detection of Fetal Lactate. Obstet
Gynecol 2004.
CASE
A 35-year-old nulliparous woman transferred to our
care at 19 weeks of gestation. The pregnancy was dated
by a first-trimester ultrasonography. A previous preg-
nancy had resulted in intrauterine death at 20 weeks,
secondary to IUGR. Medical history was significant for
a body mass index (BMI) of 39 kg/m2. Sonographic
assessment revealed a 3-week discrepancy (187 g and
biometry $ 3rd percentile) and reduced amniotic fluid.
Doppler interrogation disclosed reversed diastolic flow
in the umbilical artery waveform, reversed flow in the
ductus venosus, and increased diastolic flow in the mid-
dle cerebral artery. Fetal breathing and movement were
present. The patient declined amniocentesis for fetal
karyotype, and she desired expectant management. Ul-
trasound surveillance at 20 weeks revealed anhydram-
nios, no fetal movement, and no fetal breathing.
The patient agreed to participate in our study, which
was approved by the Columbia University Medical Cen-
ter Institutional Review Board. Written consent was
obtained. The plan was to obtain 2 data sets; one placed
in a coronal plane across the fetal torso and another of
the fetal brain. The magnetic resonance spectroscopy
study was carried out at 1.5 T in a clinical imager (Philips
Intera; Philips Medical Systems, Best, Netherlands). The
patient was positioned in the magnet bore, and the body
coil was used because of the patient’s large BMI. Two-
dimensional, single-slice proton magnetic resonance
spectroscopy was carried out using a volume-selective,
double-spin echo technique, saturating the proton water
signal with an optimized chemical shift–selective imag-
ing. The repetition time was 1,500 ms, and the echo time
was 136 ms. According to these parameters, the signal of
the methyl protons in lactate is shown as an inverted
doublet at 1.3 ppm. Localization was carried out with the
chemical shift imaging using 16 phase encoding steps in
2 dimensions and a slice-selective pulse in the third. The
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
field of view was a square of 240 mm per side, and the
slice was 20 mm in thickness. Each nominal voxel size
was 15 mm ! 15 mm ! 20 mm, for a volume of 4.5 mL
per voxel. Only one data set was acquired because the
patient declined further imaging. The spectral data set
was Fourier transformed in the 2 spatial dimensions and
in time, using a Gaussian filter of 2 Hz in the time
domain. Spectral signals were referenced, assigning the
water signal to 4.7 ppm.
The sum of the 4 spectra are shown in Fig. 1. The
remnant of the proton water signal after saturation is
shown at 4 ppm. No signals are present between 1.7 and
3.6 ppm. A wide, complex signal is shown between 0.6
and 1.7 ppm. An inverted doublet at 1.3 ppm is above a
wide, mostly positive, signal. We believe that the in-
verted signal corresponds to the methyl protons in lac-
tate, whereas the mostly positive wide signal is contam-
inant lipids. To demonstrate that the inverted signal is
lactate, the distance between the 2 components of the
doublet was measured. The value obtained (( 7 Hz) is in
agreement with the J-coupling of the methyl protons of
lactate.4
The following day, the mother sensed fetal death,
which was confirmed by ultrasonography. The patient
underwent an uncomplicated induction. Autopsy was
significant for growth restriction. The placenta was small
for gestational age, the karyotype was normal, and the
thrombophilia and infectious etiology work-ups were
negative.
COMMENT
Magnetic resonance spectroscopy for the examination of
fetal brain metabolism has been established. Brain spec-
tra of term fetuses demonstrated 4 dominating metabo-
lites, which included inositol, choline, creatinine, and
N-acetylaspartate. We have previously detected choline,
Robinson et al Detection of Fetal Lactate 1209
creatinine, and N-acetylaspartate in preterm fetal brains
(unpublished data). Lactate has been detected in fetal
lamb brains made hypoxic under experimental condi-
tions.5 In this report, we describe the in utero detection of
lactate in a small area of the back musculature in a
severely growth restricted fetus.
Following the physiological model of preferential
blood distribution, a sequence cascade can be predicted
in the setting of hypoxia.6 The fetal brain, heart, and
adrenals have preferential blood flow at the expense of
peripheral tissues, which leads to hypoxia and lactic
acidosis. Considering this model, we targeted a large
muscle area, the back, to probe for lactate by proton
magnetic resonance spectroscopy. We avoided the heart
because of its motion. The patient withdrew from the
study before acquiring brain data. Fetal death precluded
us from corroborating our findings with a conventional
assay. Ethical considerations prevented us from using an
invasive technique antenatally to obtain data not contrib-
uting to management.
Limitations of the study include the fact that lactate
was found in only one muscle group. The lack of lactate
detection in other tissues could be due to limitations of
the maternal body habitus (BMI of 39). It is possible that
the fetus was so compromised that the other metabolites
found in normal muscle by magnetic resonance spectros-
copy, such as creatine,7 were not present. It is also
feasible that the placenta eliminates lactate. We previ-
ously failed to detect lactate within the brain of a fetus
with congenital pyruvate dehydrogenase deficiency.8 At
birth, serum lactate levels were normal. Once the neo-
nate began relying on its own excretion system, serum
lactate levels rose rapidly leading to death.8 Another
limitation of this study includes the fact that the mother
did not complete the study. In addition, recruitment was
difficult. Clinical suspicion of asphyxia may indicate a
need for delivery, thus precluding this study.
In this case, the weight of circumferential evidence
that there should be fetal hypoxia and lactic acidosis was
compelling. Significant growth restriction, anhydram-
nios, and abnormal Doppler studies were present. No
fetal movement and no fetal breathing movements were
noted by ultrasound examination.
1210 Robinson et al Detection of Fetal Lactate
A literature search of the National Library of Medi-
cine’s PubMed database from 1953 to March 2004, using
the keywords “lactate” and “magnetic resonance spec-
troscopy” in combination, did not reveal any reports of
the noninvasive detection of fetal lactate by magnetic
resonance spectroscopy in the human fetus. Additional
research and development is required to determine
whether there is potential clinical use for this application.
If successful development is achieved, the clinical bene-
fits may be substantial.
REFERENCES
1. Toft PB, Christiansen P, Pryds O, Lou HC, Henriksen O.
T1, T2, and concentrations of brain metabolites in neonates
and adolescents estimated with H-1 MR spectroscopy. J
Magn Reson Imaging 1994;4:1–5.
2. Marshall I, Wardlaw J, Cannon J, Slattery J, Sellar RJ.
Reproducibility of metabolite peak areas in 1H MRS of
brain. Magn Reson Imaging 1996;14:281–92.
3. Kok RD, van den Berg PP, van den Bergh AJ, Nijland R,
Heerschap A. Maturation of the fetal brain as observed by
1
H MR spectroscopy. Magn Reson Med 2002;48:611– 6.
4. Govindaraju V, Young K, Maudsley AA. Proton NMR
chemical shifts and coupling constants for brain metabolites.
NMR Biomed 2000;13:129 –53.
5. Van Cappellen AM, Heerschap A, Nijhuis JG, Oeseburg B,
Jongsma HW. Hypoxia, the subsequent systemic metabolic
acidosis, and their relationship with cerebral metabolite
concentrations: an in vivo study in fetal lambs with proton
magnetic resonance spectroscopy. Am J Obstet Gynecol
1999;181:1537– 45.
6. Behrman RE, Lees MH, Peterson EN, de Lannoy CW,
Seeds AE. Distribution of the circulation in the abnormal
and asphyxiated fetal primate. Am J Obstet Gynecol 1970;
108:956 – 69.
7. Bottomley PA, Lee YH, Weiss RG. Total creatine in mus-
cle: imaging and quantification with proton MR spectros-
copy. Radiology 1997;204:403–10.
8. Robinson JN, Norwitz ER, Mulkern R, Brown SA, Rybicki
F, Tempany CMC. Prenatal diagnosis of pyruvate dehy-
drogenase deficiency using magnetic resonance imaging.
Prenat Diagn 2001;21:1053– 6.
Received November 21, 2003. Received in revised form March 2, 2004.
Accepted March 18, 2004.
OBSTETRICS & GYNECOLOGY
Repair of Uterine Dehiscence
With Continuation of Pregnancy
Jon S. Matsunaga, MD, Cornelia B. Daly, MD,
Clifford J. Bochner, MD, and
Connie L. Agnew, MD
St. John’s Hospital and Health Center, Santa Monica, California
BACKGROUND: Uterine dehiscence in the past has been
treated with delivery of the pregnancy and repair of the
uterus or cesarean hysterectomy. Uterine repair and con-
tinuation of the pregnancy has not been attempted to our
knowledge.
CASE: A patient with a history of a laparoscopic myomec-
tomy presented at 28 weeks of gestation with a uterine
dehiscence. This was repaired and the pregnancy contin-
ued until fetal lung maturity at 34 weeks.
CONCLUSION: Repair of a uterine dehiscence in a hemody-
namically stable patient and continuation of the pregnancy
should be considered in a very premature pregnancy to
improve neonatal outcome. (Obstet Gynecol 2004;104:
1211–2. © 2004 by The American College of Obstetri-
cians and Gynecologists.)
Uterine dehiscence usually occurs after prior uterine
surgery. This is most commonly seen after a prior cesar-
ean delivery but is being increasingly reported after a
laparoscopic myomectomy.1– 4 Uterine dehiscence is of-
ten a catastrophic event requiring delivery of the preg-
nancy and uterine repair or cesarean hysterectomy. If
the pregnancy is very premature at the time of uterine
dehiscence, there is a high risk of neonatal morbidity and
mortality. Uterine repair with continuation of the preg-
nancy has the potential to lower the risk of neonatal
morbidity and mortality from prematurity. We present a
case of a uterine dehiscence after a previous laparoscopic
myomectomy, which was stabilized and repaired to al-
low continuation of the pregnancy to fetal lung maturity.
CASE
A 40-year-old gravida 2, para 0, presented at 28 weeks of
gestation with abdominal pain. She had a history of a
laparoscopic myomectomy 7 years earlier and was told
at that time to request a primary cesarean delivery if she
became pregnant in the future. An ultrasound examina-
tion was performed, and a large amount of complex free
Address reprint requests to: Jon S. Matsunaga, MD, 2001 Santa
Monica Boulevard, Suite 970W, Santa Monica, CA 90404; e-mail:
FMDRTMDS@aol.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
© 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
fluid was noted. She also had a change in hematocrit
from 32% to 25%. The fetal heart rate was normal.
The patient underwent an exploratory laparotomy
and an approximately 4 ! 1 cm uterine dehiscence was
found on the posterior fundus, with active bleeding from
the superior margin of the dehiscence. The serosa, ex-
cept for the bleeding vessel, and amniotic membranes
were intact. Bleeding abated with pressure after approx-
imately 20 minutes while a neonatal team was assembled
and intraoperative consultation was obtained from the
perinatal staff. It was elected to attempt a repair of
the dehiscence because of the early gestational age and
the patient’s stable condition. The dehiscence was re-
paired with a running vertical imbricating stitch of num-
ber 1 chromic. The stitches were placed approximately 1
cm away from the dehiscence on each side where the
myometrium was thick enough to sustain the stitch. Four
units of packed red blood cells were transfused.
The patient was started on magnesium sulfate preoper-
atively, and this was continued postoperatively. She contin-
ued to have uterine contractions, however, and indometh-
acin was added for additional tocolysis. Her uterine
contractions resolved, and the magnesium sulfate was dis-
continued on postoperative day 2 and the indomethacin
discontinued on postoperative day 3. She was continuously
monitored initially on the labor and delivery unit and then
on the antepartum unit. She remained stable on bedrest,
with good fetal growth by serial ultrasound examinations.
At 34 5/7 weeks of gestation, ultrasonography esti-
mated fetal weight was 2,862 g, and pulmonary lung
maturity was documented by amniotic fluid analysis. A
primary low transverse cesarean delivery was per-
formed, and a 2,451-g female infant was delivered with
Apgar scores of 8 at 1 minute and 9 at 5 minutes. The site
of the prior uterine dehiscence showed a shallow depres-
sion in the uterine wall, but myometrium had reestab-
lished itself over the dehiscence site. This was reinforced
with a running horizontal imbricating layer of number 1
chromic. Postoperatively, the patient and baby did well
and were discharged home on postoperative day 4.
COMMENT
Uterine dehiscence after a prior laparoscopic myomec-
tomy has been reported1– 4 and reviewed recently.5
There appears to be an greater risk of uterine dehiscence
after a laparoscopic myomectomy than after an open
myomectomy. This is probably because the closure of
the uterine wall is often not as meticulous during a
laparoscopic myomectomy because of the greater tech-
nical difficulty of laparoscopic suturing. It has been there-
fore recommended that a multilayer closure be used
whenever a laparoscopic myomectomy is performed.6
0029-7844/04/$30.00 1211
doi:10.1097/01.AOG.0000142696.84491.ae
 This is the first report of the successful repair of a
uterine dehiscence and continuation of the pregnancy
that we are aware of (MEDLINE English language
search from January 1966 to April 2004 with the key-
word “uterine dehiscence”). In a hemodynamically sta-
ble patient, this option should be considered to allow the
pregnancy to continue, thus improving perinatal out-
come in very premature pregnancies. Repair of the preg-
nant uterine wall has been demonstrated to be safe and
successful for fetal surgery.7 The most significant com-
plications are premature labor and premature rupture of
the membranes.7,8 In this case, premature contractions
developed but were successfully managed with magne-
sium sulfate and indomethacin. Premature rupture of the
membranes was not a complication, but this may have
been less of a risk in this case because the amniotic
membranes were not penetrated.
REFERENCES
1. Harris WJ. Uterine dehiscence following laparoscopic myo-
mectomy. Obstet Gynecol 1992;80:545– 6.
Management of Maternal Cor
Triatriatum During Pregnancy
Loı̈c Sentilhes, MD, Eric Verspyck, MD, PhD,
Fabrice Bauer, MD, and
Loı̈c Marpeau, MD, PhD
Departments of Obstetrics and Gynecology and Cardiology, Rouen University
Hospital –Charles Nicolle, Rouen, France
BACKGROUND: Cor triatriatum is a rare congenital cardiac
abnormality, usually diagnosed in childhood. We describe
the first case of atrial fibrillation secondary to maternal
cor triatriatum diagnosed during the first trimester of
pregnancy and its successful management until postpar-
tum (MEDLINE "1966 to 2003# and Embase "1988 to
2003#, using MeSH terms for “cor triatriatum” and
“pregnancy”).
CASE: A 31-year-old gravida 1 complained of progressive
dyspnea on exertion and palpitations, which occurred at
the end of the first trimester of the pregnancy. Atrial fibrilla-
tion was observed on electrocardiogram. A transesophageal
echocardiography examination revealed a cor triatriatum
that was responsible for the arrhythmia. "-adrenergic block-
ing agents and digitalis glycosides were used to control su-
Address reprint requests to: Dr. Loı̈c Sentilhes, Department of Obstet-
rics and Gynecology, Pavillon Mère-Enfant, Rouen University Hospi-
tal –Charles Nicolle, 1, rue de Germont, 76031 Rouen-Cedex, France;
e-mail: loicsentilhes@hotmail.com.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
1212 © 2004 by The American College of Obstetricians and Gynecologists.
Published by Lippincott Williams & Wilkins.
2. Dubuisson JB, Chavet X, Chapron C. Uterine rupture
during pregnancy after laparoscopic myomectomy. Hum
Reprod 1995;10:1475–7.
3. Pelosi MA III, Pelosi MA. Spontaneous uterine rupture at
33 weeks subsequent to previous superficial laparoscopic
myomectomy. Am J Obstet Gynecol 1997;177:1547–9.
4. Friedmann W, Maier RF, Luttkus A, Schafer AP, Duden-
hausen JW. Uterine rupture after laparoscopic myomec-
tomy. Acta Obstet Gynecol Scand 1996;75:683– 4.
5. Miller CE. Myomectomy: comparison of open and laparo-
scopic techniques. Obstet Gynecol Clin North Am 2000;27:
407–20.
6. Tulandi T. Gynecologic operative endoscopy. Obstet
Gynecol Clin North Am 1999;26:135– 48.
7. Johnson MP. Fetal myelomeningocele repair: short-term
clinical outcomes. Am J Obstet Gynecol 2003;189:482–7.
8. Quinn TM. Fetal surgery. Obstet Gynecol Clin North Am
1997;24:143–57.
Received February 3, 2004. Received in revised form February 26,
2004. Accepted March 18, 2004.
praventricular arrhythmia, while low-molecular-weight hep-
arin was administered to prevent thromboembolic events.
Low-molecular-weight heparin was discontinued at 37 weeks
of gestation, and subcutaneous unfractionated heparin was
administered instead. Pregnancy continued a normal course
until full-term vaginal delivery with epidural anesthesia and
close hemodynamic monitoring.
CONCLUSION: A standard treatment for atrial fibrillation
could be effective in preventing maternal hemodynamic
complications secondary to cor triatriatum during preg-
nancy. Moreover, this case illustrates the American consen-
sus in neuraxial anesthesia and anticoagulation, which
supports the opinion that there is a limited risk associated
with the use of epidural and spinal anesthesia in the pres-
ence of subcutaneous heparin treatment. (Obstet Gy-
necol 2004;104:1212–5. © 2004 by The American College
of Obstetricians and Gynecologists.)
Cor triatriatum is a rare congenital cardiac abnormality
that is usually diagnosed in infancy or childhood. This
congenital heart disease can be associated with other
cardiac defects and complicated by supraventricular ar-
rhythmia, exposing patients to heart failure. Physiologi-
cal hemodynamic changes subject pregnant women with
pre-existing cardiac disease to heart failure during preg-
nancy and during the peripartum period.
We report a case of atrial fibrillation secondary to
maternal cor triatriatum, diagnosed during the first tri-
mester of pregnancy, and its management until the post-
partum period.
0029-7844/04/$30.00
doi:10.1097/01.AOG.0000142695.22437.0d

Fig. 1. Transesophageal view. LV, left ventricle; LA, left
atrium; MV, mitral valve; LAA, left atrial appendage.
Sentilhes. Maternal Cor Triatriatum During Pregnancy. Obstet Gynecol 2004.
CASE
A 31-year-old, 46-kg, gravida 1 was referred to our
hospital for cardiac evaluation at 12 weeks of gestation.
The patient had no previous medical history and com-
plained of palpitations, asthenia, and progressive dys-
pnea on exertion, which started at the beginning of
pregnancy. Physical examination revealed a rapid irreg-
ular pulse of 150 beats per minute (bpm), blood pressure
120/70 mm Hg, and respiratory rate 15 per minute. On
cardiac auscultation, a grade 2/4 ejection holosystolic
murmur along the left sternal border was present, with
no changes in position or respiration. There were no
signs of heart failure, ie, no jugular venous distension,
hepatomegaly, or presence of edema. Electrocardiogram
showed atrial fibrillation at 200 bpm. Both transthoracic
echocardiography and transesophageal echocardiogra-
phy demonstrated a cor triatriatum with a clear mem-
brane separating a large atrium (Fig. 1). No cardiac
malformations were found on fetal ultrasonography.
Digitalis glycosides (0.25 mg/d) were administered to
decrease the heart rate, and low-molecular-weight hepa-
rin (5,000 IU twice daily) was given to prevent throm-
boembolic events. A &-blocker (atenolol 50 mg/d) was
added 1 week later because of persistent rapid atrial
fibrillation. Two weeks later, symptoms suddenly disap-
peared, and electrocardiogram displayed a permanent
sinus rhythm. Both pregnancy and fetal biometry pro-
gressed normally until delivery. The use of low-molecu-
lar-weight heparin caused no clinical bleeding episodes,
such as epistaxis. Moreover, any measurable changes
occurred on platelet count, prothrombin time, and acti-
vated partial thromboplastin time. The anti-Xa level was
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
prolonged of 0.3 to 0.6 which demonstrated an anti-
thrombotic effect.
After multidisciplinary assessment, a trial of labor
with epidural analgesia was decided. Low-molecular-
weight heparin was discontinued at 37 weeks of gesta-
tion, and subcutaneous unfractionated heparin (0.6 mL
twice daily) was administered. Five days later, the patient
experienced spontaneous onset of labor. The activated
partial thromboplastin time was 1.7 times the baseline
value. Because of cardiac malformations, antibiotic pro-
phylaxis by ampicillin (2 g, then 1 g every 4 hours) was
given. Epidural analgesia was performed 8 hours after
the last administration of subcutaneous unfractionated
heparin, with a loading dose of 10 #g of sufentanil and a
continuous intrathecal infusion of 5 #g/h of sufentanil.
Hemodynamic measurements after initiation of the infu-
sion showed no significant changes. Five hours later, the
patient gave birth to a girl weighing 2,400 g and in good
condition. Postdelivery was uneventful.
Digitalis glycosides, &-blocker, and low-molecular-
weight heparin were reintroduced the day after delivery
and continued until cardiac surgical repair 6 months later.
COMMENT
Cor triatriatum is a rare congenital cardiac abnormality,
accounting for 0.1% of all congenital cardiac defects and
usually diagnosed in childhood.1 Pulmonary veins and
the mitral valve are separated by a perforated fibromus-
cular membrane dividing the left atrium into 2 cham-
bers.1 The severity of symptoms depends on the size of
the orifice area of the membrane and coexisting congen-
ital defects. If symptoms are present, they relate to de-
generation of aging membrane with sclerosis and calcifi-
cation, the development of mitral regurgitation leading
to left atrial enlargement, and atrial fibrillation.2 Adults
usually present with progressive dyspnea, hemoptysis,
and fast irregular heart beat, confounded with symptoms
of mitral stenosis. However, patients with cor triatriatum
may be completely asymptomatic, and the diagnosis is a
coincidental finding on echocardiography. Transesoph-
ageal echocardiography is preferable to transthoracic
echocardiography because of its ability to differentiate
cor triatriatum from supravalvular mitral ring.1
Pregnancy and the peripartum period involve major
cardiovascular changes. After 5 weeks of gestation, car-
diac output gradually increases and attains its maximum
at the third trimester.3 This increasing intravascular
volume can be responsible for heart failure in women
with pre-existing cardiac disease. Moreover, pregnancy
increases the incidence of arrhythmia in women,
whether or not they are affected by cardiomyopathy.3
Thus, similarly to mitral stenosis, left atrium overload
Sentilhes et al Maternal Cor Triatriatum During Pregnancy 1213
and tachycardia predispose patients to atrial fibrillation
and congestive heart failure in cor triatriatum during
pregnancy and postdelivery.4
During the patient’s pregnancy, to perform medical
and obstetrical management based on the literature, we
carried out a computerized literature search in the gen-
eral bibliographic databases: MEDLINE (1966 to 2003)
and Embase (1988 to 2003), using MeSH terms for “cor
triatriatum” and “pregnancy.” The search was not lim-
ited by language or publication type (full articles or
abstracts). Only 2 cases were found: one observation of
asymptomatic maternal cor triatriatum diagnosed inci-
dentally during pregnancy, in which both the pregnancy
and postpartum period were uneventful and one case of
maternal cor triatriatum diagnosed during the postpar-
tum period because of severe pulmonary edema occur-
ring after cesarean delivery.4,5 We did not find in the
literature any reports of maternal cor triatriatum during
pregnancy, which could have been helpful in the man-
agement of our patient’s pregnancy. During the prepara-
tion of this manuscript, we completed our bibliographic
investigation with a manual search of the citation lists of
many studies of cor triatriatum in adults (all studies are
not cited in this manuscript). Therefore, we conclude
that, to our knowledge, this represents the first reported
case of atrial fibrillation management secondary to ma-
ternal cor triatriatum during pregnancy.
Although no antiarrhythmic drug is completely safe
during pregnancy, most are well tolerated and can be
given with relatively low risk. In the present case report,
we administered all medications after taking into account
all safety considerations and neonatologist assessment.6
During pregnancy, digitalis glycosides and &-blockers
were used to control the heart rate. Anticoagulation by
low-molecular-weight heparin was used to prevent
thromboembolic events. During labor, uterine tender-
ness, anxiety, pain, and dorsal decubitus increase the
heart rate by 50%.3 We chose an epidural anesthesia
with continuous infusion of sufentanil to minimize he-
modynamic changes during labor.7 Therefore, a trial of
labor and vaginal delivery could be attempted success-
fully. There is a well-known risk of epidural hematoma
from insertion of an epidural catheter when the patient’s
coagulation function has been disrupted with heparin.
Nevertheless, the widespread use of subcutaneous hepa-
rin and paucity of complications suggest that there is a
little risk of spinal or epidural hematoma associated with
this therapy. There are 9 published series, totaling over
9,000 patients who have received this therapy without
complications, as well as extensive experience in both
Europe and United States, without a significant fre-
quency of complication.8 There are only 4 case reports of
neuraxial hematomas, 3 epidural and 1 subarachnoid,
1214 Sentilhes et al Maternal Cor Triatriatum During Pregnancy
during neuraxial block with the use of subcutaneous
heparin.8 Therefore, based on the American consensus,
during subcutaneous heparin prophylaxis, there are no
contraindications to the use of neuraxial techniques.8
However, a delay after administration of subcutaneous
heparin and the needle placement may be preferable to
reduce the risk of neuraxial bleeding.
Multidisciplinary assessment by a cardiologist, an ob-
stetrician, and an anesthesiologist was essential to opti-
mize cardiac function during the peripartum period and
to make informed decisions regarding mode of delivery
and anesthetic technique.9 In our patient, the cardiac
output was normal, and there were no other concomitant
congenital heart defects. The medical management was,
therefore, the same as for any case of maternal atrial
fibrillation during pregnancy. For patients whose cardiac
output is compromised as a result of a restricted mem-
brane orifice area, the aim of medical treatment is to
reduce the heart rate and blood volume with complete
rest and &-adrenergic blocking agents. If signs of mater-
nal heart failure occur despite medical management,
cardiac surgical repair could be considered until 32
weeks of gestation, as regards the risks of prematurity,
whereas caesarean delivery may be preferred beyond
this term. During labor, invasive arterial pressure mon-
itoring could be performed to facilitate early recognition
and treatment of possible deleterious changes in arterial
pressure. Epidural anesthesia is preferred to general
anesthesia, whatever the mode of delivery, because it
results in a decrease of arterial pulmonary and left auric-
ular pressures secondary to systemic vasodilatation. Epi-
dural anesthesia also avoids the need for general anes-
thesia with its well-known risk of hypertensive response
to intubation. Occurrence of pulmonary edema after
delivery due to aortocaval decompression is possible,
and cardiac follow-up is mandatory within the first 48
hours when cor triatriatum is present.4
REFERENCES
1. Vuocolo L, Stoddard M, Longaker R. transesophageal two-
dimensional and Doppler echocardiographic diagnosis of
cor triatriatum in the adult. Am Heart J 1992;124:791–3.
2. Beller B, Childers R, Ecjner F, Duchelle R, Ranniger K,
Robinowitz M. Cor triatriatum in the adult: complicated by
mitral insufficiency and aortic dissection. Am J Cardiol
1967;19:749 –54.
3. Elkayam U, Gleicher N. Hemodynamics and cardiac func-
tion during normal pregnancy and the puerperium. In:
Elkayam U, Gleicher N, editors. Cardiac problems in preg-
nancy: diagnosis and management of maternal and fetal dis-
ease. 2nd ed. New York (NY): Alan R. Liss, Inc; 1990. p. 5.
OBSTETRICS & GYNECOLOGY
4. LeClair SJ, Funk KJ, Goff DR. Cor triatriatum presenting as
postcesarean section pulmonary edema. J Cardiothorac
Vasc Anesth 1996;10:638 –9.
5. Sajeev CG, Roy TNS, Krishnan MN, Venugopal K. A case
of cor triatriatum diagnosed during pregnancy. Int J Cardiol
2003;90:127– 8.
6. Joglar JA, Page RL. Treatment of cardiac arrhythmias dur-
ing pregnancy: safety considerations. Drug Saf 1999;20:85–94.
7. Ransom DM, Leicht CH. Continuous spinal analgesia with
sufentanyl for labor and delivery in a parturient with severe
pulmonary stenosis. Anesth Analg 1995;80:418 –21.
VOL. 104, NO. 5, PART 2, NOVEMBER 2004
8. Horlocker TT, Wedel DJ, Benzon H, Brown DL, Kayser
Enneking F, Heit JA, et al. Regional anesthesia in the
anticoagulated patient: defining the risks (the second ASRA
Consensus Conference on Neuraxial Anesthesia and Anti-
coagulation). Reg Anesth Pain Med 2003;28:172–97.
9. Iserin L. Management of pregnancy in women with congen-
ital heart disease. Heart 2001;85:493– 4.
Received January 12, 2004. Received in revised form February 27,
2004. Accepted March 18, 2004.
Sentilhes et al Maternal Cor Triatriatum During Pregnancy 1215