Case Reports Retroperitoneal Hemorrhage After a Vaginal Mesh Prolapse Procedure performed with a pre-cut nonabsorbable polypropylene (Prolene, Ethicon Endo-Surgery, Inc., Cincinnati, OH) soft mesh. CASE Nirupa Gangam, HBSc, MD, and Allan Kanee, MD, FRCSC BACKGROUND: The present study describes a complication of a transvaginal pelvic floor repair system for the treatment of anterior vaginal prolapse. CASE: A postmenopausal woman with anterior vaginal wall prolapse was treated with a transvaginal nonabsorbable polypropylene mesh system. The procedure was complicated by a large perioperative retroperitoneal hematoma. CONCLUSION: The insertion of synthetic meshes in gynecologic surgery is gaining popularity, but all pelvic surgeons should be aware of the potential complications associated with these new techniques. (Obstet Gynecol 2007;110:463–4) T he Gynecare Prolift Pelvic Floor Repair System (Gynecare Prolift Pelvic Floor Repair System, Gynecare Worldwide, Ethicon Inc., Johnson & Johnson Company, Somerville, NJ) is a surgical technique that has been refined over a 5-year period. It is used for tissue reinforcement and long-lasting stabilization of fascial structures of the pelvic floor in vaginal wall prolapse where surgical treatment is indicated, either as mechanical support or bridging material for the fascial defect. Complications associated with the use of synthetic mesh include mesh erosion, infection, fistulae and dyspareunia.1 Our case describes the management of anterior vaginal wall prolapse repair See related article on page 455. From McMaster University Hospital, Hamilton, and Credit Valley Hospital, Mississauga, Ontario, Canada. Corresponding author: Nirupa Gangam, HBSc, MD, McMaster University Health Centre, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada; e-mail: nirupagangam@hotmail.com. Financial Disclosure Dr. Kanee is a preceptor for Johnson & Johnson, Somerville, NJ. He is financially compensated to provide training to other physicians on the Prolift System. The other author has no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 This was a 59-year-old white woman (gravida 4, para 3) who underwent a vaginal hysterectomy and anterior vaginal repair in at a different hospital. Nine months later she developed a grade III rectocele and enterocele, which was repaired at our center without complication. Approximately 2 1/2 years later, she returned with recurrent anterior vaginal wall prolapse. She had no demonstrable stress urinary incontinence or other vaginal compartment prolapse. Her past obstetric history included two spontaneous vaginal deliveries and a cesarean delivery. Prolift vaginal mesh was used to repair the anterior vaginal wall prolapse. A fully trained, experienced urogynecologist performed the procedure. During the procedure, the four polypropylene mesh straps were passed through the obturator membrane using the Prolift guide and cannula. Meticulous adherence to protocol1 was followed, with an estimated blood loss of 200 mL. While in the recovery room, the patient began reporting lower abdominal discomfort and became hypotensive and pale; however, she did not demonstrate a responsive tachycardia. Her hemoglobin dropped from 140 g/dL preoperatively, to 68 g/dL in the recovery room. The decision was made to take her back to the operating room for examination under anesthesia and possible mesh revision. She received two units of packed red blood bells and vigorous fluid resuscitation. In the operating room, the patient was found to have a tense hematoma on the anterior wall of the vagina beneath the suture line. The suture was removed, 300 mL of clot was expressed, multiples areas on the vaginal mucosa were cauterized, and the vaginal mucosa was subsequently closed. She again became hypotensive before her transfer back to the recovery room. A decision was made to proceed to diagnostic laparoscopy. A total of 200 mL of blood was present in the cul-de-sac; however, the origin of the bleeding could not be localized. Large bilateral hematomas were noted emanating from the pararectal spaces extending to the level of the infundibulopelvic ligament, and 400 mL of clot was present in the space of Retzius. Vascular surgery was consulted intraoperatively and felt that a laparotomy was not indicated, because no ongoing bleeding was seen, and 4 hours had passed since the original surgery. She was then taken to the angiography suite to rule out and potentially embolize a vascular injury. The angiogram was negative. The hematomas described above were then assumed to be caused by a self-limiting injury to a low pressure venous plexus. She was observed in OBSTETRICS & GYNECOLOGY 463 the intensive care unit overnight, where she remained stable. She was transfused a total of three units of packed red blood cells and was discharged home on postoperative day 3 in stable condition. She was assessed weekly in an outpatient setting until her 6-week postoperative visit. She encountered no further difficulties, and her anterior vaginal wall remains supported at the time of this report. CONCLUSION Surgical approaches to anterior vaginal wall defect are notorious for high recurrence rates.2 However, more recent approaches for anterior vaginal wall repair anchor an allograft, xenograft, or polypropylene mesh without tension.3 These techniques await large efficacy studies, but are increasing in use. The potential advantage of this approach is that all defects (central, lateral, proximal, and distal) can be treated in a time-efficient and minimally-invasive manner. The Anterior Prolift procedure involves implantation of a large sheet of high-porosity monofilament polypropylene “tension-free” mesh featuring anterior intervesicovaginal prosthesis. The anterior prosthesis is retained by two nonsecured bilateral transobturator arms anteriorly at a point 1 to 2 cm from the proximal arcus tendineus fasciae pelvis and posteriorly at a point 1 to 2 cm from the arcus tendineus fasciae pelvis. As the guides are passed through the obturator membrane in the lateral to medial direction, care must be taken to avoid injury to the anterior and posterior branches of the obturator artery and vein, because they are both located on the lateral surface of the obturator foramen. Cure rates for the transobturator mesh range from 75–100%,3 but comparative trials with more traditional repairs in the literature are limited. Weber et al4 found a success rate of 45%, when comparing the efficacy of three different techniques, but further trials are necessary to determine long-term efficacy4 Complications with vaginal mesh include high erosion rates, with a range of 2.3–13%,2 accidental cystotomy, urinary tract infections, voiding difficulties, and dyspareunia, but there have been few docu- 464 Gangam and Kanee mented cases of large retroperitoneal hemorrhages. LaSala et al5 recently published two case reports describing postoperative hematomas using a similar mesh augmentation system. Altman et al6 reviewed 106 anterior vaginal mesh procedures over a 6-month period and noted complications, including bladder perforation, urethral perforation, and four cases of hemorrhage, one greater than 1,000 mL. It is likely that the hemorrhage in our patient was related to a ruptured venous sinus, and not arterial injury. However, recent literature suggests that surgeons must be aware of the risk of a large perioperative hemorrhage as a complication of this “blind” procedure. Because the insertion of synthetic meshes in gynecologic surgery is gaining popularity, all surgeons should be aware of the potential complications that are associated with these new techniques. Until further data on the safety and efficacy of these new surgical techniques is uncovered, the procedure should be used cautiously by pelvic surgeons.7 REFERENCES 1. Gynecare Worldwide. Gynecare Prolift Manual. Somerville (NJ): Ethicon Inc.; 2004. 2. Debodinance P, Berrocal J, Clave H, Cosson M, Garbin O, Jacquetin B, et al. Changing attitudes on the surgical treatment of urogenital prolapse: birth of the tension-free vaginal mesh [in French]. J Gynecol Obstet Biol Reprod (Paris) 2004;33: 577–88. 3. Walters MD, Paraiso MF. Anterior vaginal wall prolapse: innovative surgical approaches. Cleve Clin J Med 2005;72: S20–7. 4. Weber AM, Walters MD, Piedmonte MR, Ballard LA. Anterior colporrhaphy: a randomized trial of three surgical techniques. Am J Obstet Gynecol 2001;185:1299–304. 5. LaSala CA, Schimpf MO. Occurrence of postoperative hematomas after prolapse repair using a mesh augmentation system. Obstet Gynecol 2007;109:569–72. 6. Altman D, Falconer C. Perioperative morbidity using transvaginal mesh in pelvic organ prolapse repair. Obstet Gynecol 2007;109:303–8. 7. Baessler K, Hewson AD, Tunn R, Schuessler B, Maher CF. Severe mesh complications following intravaginal slingplasty.Obstet Gynecol 2005;106:713–6. Retroperitoneal Vaginal Mesh Hemorrhage OBSTETRICS & GYNECOLOGY Electroconvulsive Therapy in Pregnancy Michael G. Pinette, MD, Camille Santarpio, DO, Joseph R. Wax, MD, and Jacquelyn Blackstone, DO BACKGROUND: Electroconvulsive therapy for the treatment of depressive and bipolar disorders has been advocated as being safe and effective in pregnancy. CASE: A primigravida underwent multiple electroconvulsive treatments during pregnancy for the diagnosis of major depression. The infant was subsequently born with multiple deep interhemispheric infarcts. CONCLUSION: Despite reassuring statements regarding the safety of electroconvulsive therapy, this case report and a review of the literature suggests that electroconvulsive therapy during pregnancy should be performed with caution. (Obstet Gynecol 2007;110:465–6) T he American Psychiatric Association (APA) published text, The Practice of Electroconvulsive Therapy, Recommendations for Treatment, Training, and Privileging,1 states that electroconvulsive therapy (ECT) is a “treatment with low risk and high efficacy in the management of specific disorders in all three trimesters of pregnancy.” We present a case of a primigravida who received ECT every 2 weeks throughout her pregnancy at an outside hospital for major depression and bipolar disorder and subsequently delivered an infant with neurologic abnormalities associated with several brain infarcts. CASE A 22-year-old primigravida with a long history of bipolar disorder and major depression received electroconvulsive bifrontal electroconvulsive therapy (ECT) every 2 weeks during her entire pregnancy, with no complications reported. The patient had been maintained before pregnancy on intermittent ECT with good response. Therapy was See related editorial on page 451. From the Division of Maternal–Fetal Medicine, Department of Obstetrics and Gynecology, Maine Medical Center, Portland, Maine. Corresponding author: Michael G. Pinette, MD, MMC Ob/Gyn Associates, 887 Congress Street, Suite 200, Portland, ME 04102; e-mail: cartia@mmc.org. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 continued during the pregnancy for ongoing severe depression, because the patient was only partially responsive to conventional medical therapy. Between 20 and 34 weeks of gestational age, seven documented ECT treatments occurred. Fetal heart rates were recorded after each treatment with no reported abnormalities. It was unclear how soon after the treatments the testing was performed or at what gestational age this testing began. The hospital records contain no evidence of a formal nonstress test being performed. Details of the maternal ECT were not recorded. At 36.1 weeks of gestation she was noted to have an elevated blood pressure of 162/100 mm Hg before her scheduled ECT. She also had mildly elevated protein on a 24-hour urine analysis, with all other laboratory values being normal. She underwent induction of labor for preeclampsia and vaginally delivered a male infant weighing 2,550 g with 1- and 5-minute Apgar scores of 4 and 7, without complications. The fetal heart tracing throughout the labor and delivery was unremarkable. Anatomic survey at 20 weeks was normal. The infant was stimulated with no response, and his upper extremities were tonic with extension posturing. An umbilical cord blood gas analysis showed a pH of 7.24, PCO2 of 42, HCO3 of 18, PO2 of 27. Results of the infant’s infectious disease workup, toxicity screens, and blood work were negative. A head computed tomography and magnetic resonance image performed on the second and third day of life showed a small left cerebellar, bihemispheric deep white matter, and cortical infarcts. The infant was discharged home at 12 days of life. A neurology consult stated that the prognosis was guarded with expected long-term motor control issues. Long-term follow-up was not available. DISCUSSION Although a cause-and-effect relationship cannot be established in this case, the multiple ECT treatments this patient received in pregnancy and the final neonatal outcome are temporally related. The potential for ECT to evoke uterine contractions and effect fetal heart rates has been previously suggested and could potentially lead to fetal compromise. Although infectious disease and labor-induced hypoxia seem unlikely, other causes of brain injury predating labor are possible. Whether electroconvulsive therapy in this case led to tetanic uterine contractions causing a fetal bradycardia or possibly induced a fetal arrhythmia, resulting in fetal brain injury, is purely speculative. The data in the literature to support the recommendations of the APA include numerous case reports and reviews of case reports. We performed a search of MEDLINE for articles written in English between January 1966 and January 2007, using the keywords “depression,” “pregnancy,” and “electro- Pinette et al Electroconvulsive Therapy in Pregnancy 465 convulsive therapy.” Our review showed no prospective and or controlled studies of ECT in pregnancy. There are also only three case reports reporting no adverse long-term developmental defects in children exposed to ECT during pregnancy. These reports are limited by small numbers of patients, variable follow-up periods, and publication before 1964.2 The largest literature review of ECT in pregnancy reviewed 300 case reports from 1942 to 1991.3 Twentyeight cases (9.3%) were noted to have complications, including fetal cardiac arrhythmias (5 cases), vaginal bleeding (5 cases), uterine contractions (2 cases), abdominal pain (3 cases), premature labor (4 cases), miscarriage (5 cases), stillbirth and neonatal death (3 cases), neonatal respiratory distress (1 case), and teratogenicity (5 cases). There have been an additional six cases in the literature with obstetric outcomes reported.2– 6 In addition, there was a case series of 27 patients undergoing ECT as a primary treatment, four of whom were pregnant. No details were given in regard to pregnancy outcomes.7 Included in the six cases are reports of fetal heart rate decelerations in three of six cases, uterine contractions in three of six cases, one requiring tocolytic medication, and one case of first trimester bleeding with subsequent miscarriage. There is broad heterogeneity in regard to the details of diagnosis, treatments, pregnancy trimester, other contributing factors and outcomes reported between the case reports. In addition, 25 of the 28 cases that reported complications were more than 30 years old at the time of publication of the review. The data do not provide adequate evidence for definitive conclusions to be made regarding ECT in pregnancy. The APA has published guidelines for management of a pregnant patient undergoing ECT.1 Numerous review articles have recommended similar guidelines.2,3,8 These guidelines are aimed at minimizing potential complications including aspiration and altered uteroplacental blood flow. Specific recommendations include 1) Consultation with an obstetrician before initiation of treatment. 2) Treatments performed in a facility with immediate access to obstetric 466 Pinette et al Electroconvulsive Therapy in Pregnancy care for emergencies. 3) Monitoring of fetal heart rate before and after treatments. Increase in monitoring at viability to include a nonstress test with tocometry after treatments. Perform a Level 2 ultrasonography between 18 weeks and 22 weeks gestational age. 4) Routine anesthetic measures (leftward tilt of trunk, adequate oxygenation, hydration, and muscle relaxation, nonparticulate antacid, consider intubation in the third trimester). Despite statements by the APA regarding the safety of ECT in pregnancy, the data to support this conclusion is limited. Some case reports suggest possible occasional potentially serious complications. The rate of these complications is clearly unknown; however, available data suggests a number as high as 9%. In the absence of prospective outcome studies, ECT in pregnancy should be performed specifically according to APA guidelines, and only after medical therapy has failed in cases of severe debilitating depression. REFERENCES 1. American Psychiatric Association. Committee on Electroconvulsive Therapy. The practice of electroconvulsive therapy, recommendations for treatment, training, and privileging: a task force report of the American Psychiatric Association. 2nd ed. Washington (DC): American Psychiatric Association; 2001. 2. Bhatia SC, Baldwin SA, Bhatia SK. Electroconvulsive therapy during the third trimester of pregnancy. J ECT 1999;15:270–4. 3. DeBattista C, Cochran M, Barry JJ, Brock-Utne JG. Fetal heart rate decelerations during ECT-induced seizures: is it important? Acta Anaesthesiol Scand 2003;47:101–3. 4. Polster DS, Wisner KL. ECT-induced premature labor: a case report. J Clin Psychiatry 1999;60:53–4. 5. Echevarria Moreno M, Martin Munoz J, Sanchez Valderrabanos J, Vazquez Gutierrez T. Electroconvulsive therapy in the first trimester of pregnancy. J ECT 1998;14:251–4. 6. Livingston JC, Johnstone WM Jr, Hadi HA. Electroconvulsive therapy in a twin pregnancy: a case report. Am J Perinatol 1994;11:116–118. 7. Maletzky BM. The first-line use of electroconvulsive therapy in major affective disorders. J ECT 2004;20:112–7. 8. Walker R, Swartz CM. Electroconvulsive therapy during highrisk pregnancy. Gen Hosp Psychiatry 1994;16:348–53. OBSTETRICS & GYNECOLOGY Cervical Adenosquamous Carcinoma mous carcinoma diagnosed at delivery with subsequent implantation in the episiotomy scar. Tumor Implantation in an Episiotomy Scar CASE Gudrun Neumann, MD, Kjeld L. Rasmussen, MD,† and Lone Kjeld Petersen, DMSc BACKGROUND: We report a rare case of a cervical adenosquamous carcinoma, initially diagnosed during delivery, with subsequent implantation in the episiotomy scar 5 weeks postpartum. CASE: A 35-year-old woman with cervical adenosquamous carcinoma diagnosed during delivery was treated with radical abdominal hysterectomy with bilateral pelvic lymphadenectomy. Five weeks later the metastatic tumor at the episiotomy site was excised, and the patient received adjuvant chemotherapy and radiation therapy. Relapse occurred rapidly, and surgical exenteration was initiated but abandoned intraoperatively due to the presence of intra-abdominal carcinomatosis. The patient was declared terminal 6 months postpartum and died 2 months later. CONCLUSION: This case illustrates the importance of inspection of the perineal area during delivery in patients diagnosed with cervical cancer. (Obstet Gynecol 2007;110:467–9) E pisiotomy scar implantation from a primary cervical cancer is a rare condition. The incidence of cervical cancer during pregnancy is low (approximately 3%).1–3 There are very few cases in the literature regarding the subject.1–9 All but three cases report a previous negative Pap test and were diagnosed either during pregnancy or at delivery. We report a case of rapid progression of a cervical adenosquaSee related editorial on page 453. From the Department of Gynecology & Obstetrics, Herning Hospital, Herning, Denmark; and Department of Gynecology & Obstetrics, Aarhus University Hospital, Skejby, Aarhus N, Denmark. † Deceased. Corresponding author: Gudrun Neumann, Lavendelvej 5, 7400 Herning, Denmark; e-mail: gudrunneumann@dadlnet.dk. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 A 35-year-old woman (gravida 2, para 1) at 39 weeks of gestation was admitted to the hospital while in labor. The patient was healthy and her pregnancy was uncomplicated. Specifically, no symptoms of vaginal discharge or vaginal bleeding were reported. During pregnancy she was followed by her general practitioner and the county midwife (according to the standard Danish practice for uncomplicated pregnancies), and underwent an ultrasonographic examination early in pregnancy. During delivery, the midwife felt a tumor on the anterior aspect of the cervix. A piece of tumor was delivered together with the infant The delivery was otherwise uncomplicated and a healthy boy was born after a mediolateral episiotomy was performed. The episiotomy was repaired with a continuous resorbable suture under local anesthesia. Histopathologic examination of the cervical tumor revealed a low differentiated adenosquamous carcinoma. The placenta and umbilical cord were histologically normal. After the diagnosis of cervical cancer, the patient was immediately referred to an oncologist for further treatment. Previous Pap test results were obtained and all were normal. A reevaluation by a senior pathologist concluded the same. The last Pap test taken 10 months before delivery consisted of autolyzed cells with atypical epithelia without sign of malignant cells. During the 9-year period before delivery, no Pap tests were taken. At clinical examination during preoperative work-up an exophytically growing tumor measuring 5 cm on a small base located on the anterior part of the cervix was removed. Cystoscopy and rectovaginal exploration were without suspicion. Computed tomography (CT) scan of the abdominal cavity and pelvis was normal. Preoperatively, the tumor staging was thus determined as International Federation of Gynecology and Obstetrics (FIGO) stage IB2. Twenty-one days after delivery, the patient underwent an uncomplicated radical abdominal hysterectomy with bilateral pelvic lymphadenectomy. Histology showed free resection margins but there was tumor cell embolism in the paracervical vessels. The patient subsequently received adjuvant chemotherapy and radiation therapy. Eighteen days after radical abdominal hysterectomy the patient complained about perineal pain. A tumor measuring 5⫻6 cm located in the episiotomy scar and a small metastasis located near the orificium externum urethrae were seen at clinical examination (Fig. 1). Two days later a local surgical excision was done. Pathologic examination confirmed the diagnosis of a low differentiated adenosquamous carcinoma. During the next 5 weeks the patient received radiotherapy with 50/45 Gy on 25 doses and concomitant chemotherapy with cisplatin 50 mg/kg in 5 series followed by brachytherapy. The response of the radiation therapy was effectual. Two months later the patient was readmitted with vaginal discharge and pain. Clinical examination revealed recurrence Neumann et al Episiotomy Scar Tumor Implantation 467 Fig. 1. Tumor implantation (arrow) in the episiotomy scar. Neumann. Episiotomy Scar Tumor Implantation. Obstet Gynecol. 2007 of metastases in the perineum, located at the episiotomy scar and in the area between the orificium externum urethrae and the clitoris. Positron emission tomography and CT scans showed bilateral metastases in the lymph nodes in the groin. An exenteration operation was planned but explorative laparotomy revealed widespread intra-abdominal metastases and further surgical treatment was abandoned. The disease progressed rapidly and only palliative chemotherapy was an option. Within a week the patient had clinical signs of bowel obstruction, confirmed by CT scan. Stenosis was evident distally in the sigmoid colon, due to a tumor mass measuring 5⫻5 cm squeezing the bowel against the os sacrum. An endoscopic attempt to place a stent was unsuccessful and a sigmoidostomy was needed. The patient’s condition worsened rapidly and palliative analgesic treatment was initiated soon thereafter. The patient died after a brief stay in a hospice for palliative care. COMMENT The occurrence of cervical cancer in pregnancy is a rare condition.1–3 Even rarer is tumor implantation in the episiotomy scar.1–9 There are 13 such cases reported to date.1–9 The patients’ ages were between 21 years and 33 years. Of these, only three had a positive prenatal Pap test. Squamous cell carcinoma was diagnosed in eight patients and five patients had an adenocarcinoma. Only one author reported a villoglandular carcinoma located in an episiotomy scar.1 Nine patients were diagnosed with FIGO stage IB 468 Neumann et al Episiotomy Scar Tumor Implantation disease and one patient had FIGO stage IIIB disease. The rest had FIGO stage IA-IIA disease. Survival time after recurrence was 3– 60 months.1,4 –5 When a Schauta-Amreich radical vaginal hysterectomy is chosen, a Schuchardt incision is often performed to facilitate access to the parametrium. A recently published case reported a recurrence of adenocarcinoma in a Schuchardt incision. The patient died of the disease 51 months after the initial operation.10 We report the case of rapidly progressing cervical cancer detected during delivery. Previous Pap tests were normal and the patient had routine gynecologic examinations, but had no Pap tests during the 9 years before delivery. There was no complaint of vaginal discharge or bleeding. At delivery of her second child, the patient was diagnosed with an invasive adenosquamous cervical carcinoma and tumor embolism in the parametria, which may indicate that the tumor was aggressive. Pathologic examination confirmed this assumption. Immune suppression during pregnancy may facilitate the rapid progression of cervical cancers, but evidence is difficult to come by.3 Vaginal discharge and spotting is common during pregnancy and therefore the condition is sometimes underdiagnosed. Furthermore, performing a colposcopy is difficult during pregnancy. Wound metastasizing has been previously described in breast cancer and gastrointestinal cancers.2 Nonbiologic exfoliate tumor implantation is rare, and the tumor is probably implanted during delivery.1,3,6,8 It is assumed that the mechanism of tumor implants is related to mechanical spread rather than lymphatic or hematogenous spread. Tumor emboli are small and break off as tumors, which penetrate vessels. They do not usually cause immediate problems the way other emboli do, but it is a major route of dissemination of malignancies. However, the cause of recurrences in surgical incisions seems to be through tumor cell implantations rather than metastasis through vascular or lymphatic channels.10 When one examines the available literature, there are no randomized controlled trials showing the benefits of a cesarean compared with vaginal delivery regarding the occurrence of hemorrhage, dystocia, or tumor spread.9 It is unknown whether delivery by cesarean carries the same possible risk for tumor implantation in the abdominal scar or in the lymph nodes in the groin.2,6 A matched case– control study published in 2000 concluded that women diagnosed postpartum had worse survival than those diagnosed during pregnancy and were at significant risk of OBSTETRICS & GYNECOLOGY recurrent disease, with poor outcome, particularly after vaginal delivery.9 In conclusion, the development of tumor implantation arising from a primary cervical cancer in an episiotomy scar is a rare but possible event. It is important to inspect the perineal area of a patient diagnosed with cervical cancer during pregnancy, especially after a normal vaginal delivery with an episiotomy. An early biopsy is recommended. Moreover, it is important to avoid these incisions. In case of known cervical cancer, reexamination of a necessary episiotomy scar should be considered. Prenatal cytologic screening should be considered and a biopsy should be taken on wide indications. REFERENCES 1. Van den Broek NR, Lopes AD, Ansink A, Monaghan JM. “Microinvasive” adenocarcinoma of the cervix implanting in an episiotomy scar. Gynecol Oncol 1995;59:297–9. 2. Gordon AN, Jensen R, Jones HW 3rd. Squamous carcinoma of the cervix complicating pregnancy: recurrence in episiotomy after vaginal delivery. Obstet Gynecol 1989;73:850–2. Isolated Clear Cell Adenocarcinoma in Scar Endometriosis Mimicking an Incisional Hernia Vanessa N. Harry, MBBS, MRCOG, Smruta Shanbhag, MBBS, MRCOG, Matthew Lyall, MBchB, Gordon V. Narayansingh, FRCOG, and David E. Parkin, MD, FRCOG BACKGROUND: The development of a mass in association with a previous surgical scar can pose a diagnostic dilemma due to similarities in appearance to hernias, ab- 3. Goldman NA, Goldberg GL. Late recurrence of squamous cell cervical cancer in an episiotomy site after vaginal delivery. Obstet Gynecol 2003;101:1127–9. 4. Heron DE, Axtel A, Gerszten K, Amortegui A, Kelley J, Comerci J, et al. Villoglandular adenocarcinoma of the cervix recurrent in an episiotomy scar: a case report in a 32-year-old female. Int J Gynecol Cancer 2005;15:366–71. 5. Khalil AM, Khatib R, Mufarrij AA, Tawil AN, Issa PY. Squamous cell carcinoma of the cervix implanting in the episiotomy site. Gynecol Oncol 1993;51:408–10. 6. Cliby WA, Dodson MK, Podratz KC. Cervical cancer complicated by pregnancy: episiotomy site recurrences following vaginal delivery. Obstet Gynecol 1994;84:179–82. 7. Burgess SP, Waymont B. Implantation of a cervical carcinoma in an episiotomy site. Case report. Br J Obstet Gynaecol 1987;94:598–9. 8. Copeland LJ, Saul PB, Sneige N. Cervical adenocarcinoma: tumor implantation in the episiotomy sites of two patients. Gynecol Oncol 1987;28:230–5. 9. Sood AK, Sorosky JI, Mayr N, Anderson B, Buller RE, Niebyl J. Cervical cancer diagnosed shortly after pregnancy: prognostic variables and delivery routes. Obstet Gynecol 2000;95: 832–8. 10. Bader AA, Bjelic-Radisic V, Tamussino KF, Pristauz G, Winter R. Recurrence in a Schuchardt incision after Schauta-Amreich radical vaginal hysterectomy for cervical cancer. Int J Gynecol Cancer 2006;16:1479–81. scesses, hematomas, or desmoid tumors. Scar endometriosis is an uncommon cause of such a lump, but malignant change within this ectopic tissue is exceptionally rare. CASE: We present a case of a 55-year-old woman who was found to have an isolated clear cell adenocarcinoma in an area of scar endometriosis more than 30 years after an open tubal sterilization. This mass was initially thought to be an incisional hernia, but was later diagnosed intraoperatively by frozen section and then incompletely excised, highlighting the difficulties in preoperative diagnosis as well as surgical treatment. CONCLUSION: Malignant change within scar endometriosis is rare, but increased awareness of this phenomenon is required. Vigilance is paramount and a mass located in or close to a surgical scar should be treated with suspicion. (Obstet Gynecol 2007;110:469–71) See related editorial on 453. From the Departments of Gynaecological Oncology and Pathology, Aberdeen Royal Infirmary, Aberdeen, Scotland, United Kingdom. Corresponding author: Vanessa N. Harry, Ward 43, Department of Gynaecological Oncology, Aberdeen Royal Infirmary, Foresterhill, Aberdeen, Scotland, United Kingdom, AB25 2ZN; e-mail: v.harry@abdn.ac.uk. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 I t is well known that endometriosis can occur both in extra- as well as intra-abdominal sites. However, its occurrence within a surgical scar is uncommon and is usually the result of a previous caesarean delivery or abdominal hysterectomy.1 Malignancy arising in scar endometriosis is exceptionally rare, and has been found almost exclusively in caesarean delivery scars.2– 6 We present an unusual case of clear cell adenocarcinoma arising in endometriosis in an open tubal sterilization scar. Harry et al Clear Cell Cancer in Scar Endometriosis 469 CASE A 55-year-old woman was referred to the general surgery outpatient department at Aberdeen Royal Infirmary with a history of a small subumbilical abdominal lump. This had been present for about 15 years, but had only recently become larger and intermittently painful. Her medical history included hypothyroidism and acute poliomyelitis as a child. She never took hormone replacement therapy and was a heavy smoker for the past 40 years. She had three children, all by vaginal deliveries, and in 1973 had an open sterilization performed through a subumbilical midline incision. This patient gave no prior history of endometriosis and documentation of the sterilization procedure showed no evidence of endometriosis or other pelvic abnormality at that time. Physical examination revealed a 4⫻4 cm mass just lateral to the sterilization scar, which was immobile, and tender. This was thought to be an incisional hernia, and the patient was later admitted to have a repair operation. During this procedure however, the lump was found to be not a hernia, but a firm, fixed mass arising from the rectus sheath. Frozen section revealed an adenocarcinoma, and the mass was subsequently excised with the remaining defect closed by using a polypropylene (Prolene, Ethicon Endo-Surgery, Inc., Cincinnati, OH) mesh. Final histopathology showed an incompletely excised clear cell adenocarcinoma, with adjacent small foci of endometriosis, as demonstrated in Figure 1. A computed tomography scan of the abdomen, pelvis, and chest was done, and a diagnostic laparoscopy with the taking of peritoneal washings was performed, all of which showed no evidence of malignancy or endometriosis. Serum CA 125 testing proved to be normal at a level of 9. The patient then received 10 fractions of radiotherapy to the wound site. She had no detectable recurrence at ongoing follow-up 18 months later. COMMENT In 1925, Sampson7 was the first to document the ability of ectopic endometrial tissue to undergo malignant change and suggested criteria to define this. Mostoufizadeh and Scully8 later proposed that the simple coexistence of a tumor and endometriotic tissue was sufficient to demonstrate the endometriotic origin of the tissue. The ovary is by far the commonest site for this phenomenon, with malignant change occurring in 0.7% of cases of ovarian endometriosis,9 and the histologic type is predominantly clear cell. This is even more uncommon at extraovarian sites, where the histologic differentiation is invariably endometrioid, with clear cell carcinoma representing only 4.5%.10 On review of the literature, there are very few documented cases of clear cell carcinoma arising in extraovarian sites of endometriosis such as episiotomy scars and sigmoid colon.6 Similarly, there are few reported cases of clear cell carcinoma arising in scar endometriosis, and most have followed caesarean deliveries.2– 6 Our report of this occurrence arising within a sterilization scar is highly unusual. The cause of the development of carcinoma is unclear, and whether the presence of endometriosis is causal or casual is controversial. The use of exogenous estrogen has been reported as a possible risk factor and in a study by Modesitt et al11 62% of the 21 patients with extraovarian cancers arising in endometriosis were taking hormone replacement therapy (mostly unopposed estrogen), compared with 10% of the cases of ovarian cancer associated with endometriosis. Presenting symptoms and signs can be variable, but usually include abdominal pain accompanied by an abdominal or pelvic mass. The uncommon nature of this occurrence poses a diagnostic dilemma because an abdominal wall swelling can masquerade not only as hernia, but also an abscess, hematoma, desmoid tumor, lymphoma, or sarcoma. Suspiciously, our patient reported a recent increase in size of the abdominal lump. Preoperative diagnosis is difficult, and routine imaging may not be helpful in detecting malignant Fig. 1. A. Clear cell adenocarcinoma (thin arrow), represented as glandular structures lined by hobnail cells, emanating from a focus of endometriosis (thick arrow) set in fibroconnective tissue (hematoxylin-eosin; ⫻40, original magnification). B. This image demonstrates a high-power view of the tumor elsewhere, displaying typical clear cell morphology (hematoxylin-eosin; x400, original magnification). Harry. Clear Cell Cancer in Scar Endometriosis. Obstet Gynecol 2007. 470 Harry et al Clear Cell Cancer in Scar Endometriosis OBSTETRICS & GYNECOLOGY change. Although 2-[18F] Fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) is known to be useful in the preoperative characterization of various masses, one report has suggested that the inflammation associated with endometriosis may cause a positive uptake of FDG on PET scans, similar to that of malignancy, giving a falsely positive result.12 Ischida et al3 proposed that fine-needle aspiration cytology of a mass arising in a surgical scar could be a useful tool in diagnosing a malignancy. Increased awareness of the possibility of endometriosis-associated malignant change is required, and a mass located in or close to a surgical scar should be treated with suspicion, especially with a history of recent change in size or increased tenderness. A comprehensive evaluation of the patient is essential to exclude other sites of disease, and should include investigations such as CA 125 and computed tomography scans of abdomen and pelvis. Treatment of isolated malignancy in scar endometriosis is surgical, but as demonstrated in our case, may not be entirely straightforward, due to incomplete excision. Hull et al13 recently reported on the use of a hook-wire under ultrasound guidance to facilitate the complete excision of scar endometriosis, but whether this may be useful for improving the completeness of excision of malignant lesions is still to be investigated. We believe that despite the uncommon occurrence of scar endometriosis, this diagnosis should be entertained in all cases where abdominal lumps have developed in association with surgical scars. This is compounded by our revelation of malignant change and adds to the ongoing enigma surrounding endometriosis and challenges in its management. Hemolytic Transfusion Reaction After Preoperative Prophylactic Blood Transfusion for Sickle Cell Disease in Pregnancy Christine L. Proudfit, MD, Emad Atta, MD, and Nora M. Doyle, MD, MPH BACKGROUND: Preoperative transfusions are frequently given to prevent morbidity in nonpregnant pa- From the Department of Gynecology and Obstetrics, and Division of Maternal Fetal Medicine, Emory University School of Medicine, Atlanta, Georgia. VOL. 110, NO. 2, PART 2, AUGUST 2007 REFERENCES 1. Gunes M, Kayikcioglu F, Ozturkoglu E, Haberal A. Incisional endometriosis after cesarean section, episiotomy and other gynecologic procedures. J Obstet Gynaecol Res 2005;31: 471–5. 2. Alberto VO, Lynch M, Labbei FN, Jeffers M. Primary abdominal wall clear cell carcinoma arising in a Caesarean section scar endometriosis. Ir J Med Sci 2006;175:69–71. 3. Ishida GM, Motoyama T, Watanabe T, Emura I. Clear cell carcinoma arising in a cesarean section scar. Report of a case with fine needle aspiration cytology. Acta Cytol 2003;47: 1095–8. 4. Park SW, Hong SM, Wu HG, Ha SW. Clear cell carcinoma arising in a Cesarean section scar endometriosis: a case report. J Korean Med Sci 1999;14:217–9. 5. Miller DM, Schouls JJ, Ehlen TG. Clear cell carcinoma arising in extragonadal endometriosis in a caesarean section scar during pregnancy. Gynecol Oncol 1998;70:127–30. 6. Hitti IF, Glasberg SS, Lubicz S. Clear cell carcinoma arising in extraovarian endometriosis: report of three cases and review of the literature. Gynecol Oncol 1990;39:314–20. 7. Sampson JA. Endometrial carcinoma of the ovary arising in endometrial tissue in that organ. Arch Surg 1925;10:1–72. 8. Mostoufizadeh M, Scully RE. Malignant tumors arising in endometriosis. Clin Obstet Gynecol 1980;23:951–63. 9. Nishida M, Watanabe K, Sato N, Ichikawa Y. Malignant transformation of ovarian endometriosis. Gynecol Obstet Invest 2000;50 suppl:18–25. 10. Heaps JM, Nieberg RK, Berek JS. Malignant neoplasms arising in endometriosis. Obstet Gynecol 1990;75:1023–8. 11. Modesitt SC, Tortolero-Luna G, Robinson JB, Gershenson DM, Wolf JK. Ovarian and extraovarian endometriosis-associated cancer. Obstet Gynecol 2002;100:788–95. 12. Jeffry L, Kerrou K, Camatte S, Metzger U, Lelievre L, Talbot JN, et al. Endometriosis with FDG uptake on PET. Eur J Obstet Gynecol Reprod Biol 2004;117:236–9. 13. Hull ML, Gun MT, Ritossa M. Hook-wire insertion facilitates the excision of scar endometriosis. BJOG 2006;113:744–6. tients with sickle cell disease. We describe a case of a life-threatening delayed hemolytic transfusion reaction with hyperhemolysis syndrome in pregnancy. CASE: A multigravida with sickle cell disease underwent prophylactic blood transfusion before repeat cesarean delivery. Her immediate postpartum course was uneventful, but on postoperative day number 6 she presented in grave condition with what was thought initially to be an infection versus crisis. Delayed hemolytic transCorresponding author: Nora M. Doyle, MD, MPH, MS, 69 Jesse Hill Drive NE, Atlanta, GA 30303; e-mail: ndoyle@emory.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 Proudfit et al Delayed Transfusion Reaction 471 fusion reaction with hyperhemolysis was ultimately diagnosed. CONCLUSION: In the gravida with sickle cell disease and known multiple red cell antibodies, blood transfusion may incur a higher risk for delayed transfusion reaction, hyperhemolysis syndrome, and possible death. Blood transfusion should be used cautiously in these patients. (Obstet Gynecol 2007;110:471–4) S ickle cell anemia, an autosomal recessive disease, is one of the structural hemoglobinopathies caused by inheriting two copies of hemoglobin (Hb)S which results in the production of HbSS (␣2/␤S2). Sickle cell disease affects 1 in 600 African Americans and people of Mediterranean or Southeast Asian origin.1 Patients suffer from lifelong complications, in part due to the markedly shortened life span of their red blood cells (RBCs). Virtually all signs and symptoms are secondary to hemolysis, vasoocclusive disease, or an increased susceptibility to infection. Repeated blood transfusions are frequently the norm for these patients, with the untoward effect of multiple red cell antibody formation.2 Preoperative transfusions are commonly given to prevent perioperative morbidity in patients with sickle cell anemia.3 Risks of blood transfusion include transmission of viral or bacterial infections, hemolytic reactions, transfusion-related acute lung injury and both acute and delayed transfusion reactions. Rarely, hyperhemolysis syndrome can follow a delayed transfusion reaction. Hyperhemolysis syndrome is a separate clinical entity that can only follow a delayed hemolytic transfusion reaction and is characterized by the destruction of both donor and recipient RBCs.4 Continuation of blood transfusion may be lethal because this exacerbates the hemolysis. In part because of these risks, there is no consensus on the necessity nor the best regimen of transfusions for this purpose, particularly in the pregnant patient undergoing cesarean delivery. A MEDLINE search was performed covering the period from 1966 to August 2006 and using a combination of the keywords “sickle cell disease,” “pregnancy,” “cesarean,” “blood transfusion,” “red cell I antigen,” “delayed transfusion reaction,” and “hyperhemolysis syndrome.” No cases of delayed hemolytic transfusion reaction after preoperative blood transfusion involving I antigen in pregnant sickle cell patients were found. We report a pregnant sickle cell patient with known multiple red cell antibodies, given a prophylactic pre– cesarean delivery crossmatched blood transfusion, who experienced a delayed hemo- 472 Proudfit et al Delayed Transfusion Reaction lytic transfusion reaction with hyperhemolysis syndrome from a rare I antibody. CASE A young African-American gravida known sickle cell patient transferred to the care of our High Risk Obstetrical Clinic from another state. Her sickle cell disease was complicated by multiple episodes of acute chest syndrome, bilateral hip replacement surgery secondary to avascular necrosis of the femoral heads with bilateral revisions, osteomyelitis, numerous alloantibodies from repeated blood transfusions, and delayed transfusion reactions, all documented from outside records. Her most recent transfusion was 5 years ago. Her current pregnancy was uneventful, apart from one crisis 8 days before delivery. Elective repeat cesarean delivery was performed. The indications were previous cesarean delivery as well as limited range of motion in both hips because of her bilateral hip replacement surgery. Preoperatively, hemoglobin and hematocrit were 8.2g/dL and 24%, respectively, and after consultation with the anesthesiologist, the patient was transfused with 1 unit of apparently crossmatched packed RBCs. An uncomplicated cesarean delivery with bilateral tubal ligation was performed resulting in a healthy 2,920-g male infant with Apgars of 8 and 9. Estimated blood loss was 500 mL. Postoperative hemoglobin remained stable at 8.5 g/dL. Initial postpartum course was unremarkable, and she was discharged home in good condition on postoperative day 3. On postoperative day 6, the patient presented to Obstetrics Triage with a chief complaint of 10/10 upper and lower extremity crisis-like pain for 1 day, not relieved by acetaminophen with hydrocodone. She reported a 38.3ºC temperature at home on the evening before her admission. Review of systems was otherwise negative. Initial assessment revealed temperature of 39.4°C, heart rate 128, and blood pressure 143/70, respiratory rate of 22 with 95% oxygen saturation on 8 L of oxygen. Physical examination revealed the patient to be in visible distress with a 3/6 systolic ejection murmur at the left upper sternal border and diffuse abdominal pain. Initial laboratory studies demonstrated a white blood cell count of 32⫻103/␮L (normal range 3.6 –11.1), hemoglobin 7.1 g/dL (normal range 11.4 –14.4), hematocrit 21% (normal range 33.3– 41.4), and a reticulocyte count of 14.8% ( normal less than 2%). She had an increased total bilirubin of 2.4 mg/dL (normal range 0.3–1.2) with indirect component, and aspartate transaminase of 90 units/L (normal range 15– 41). Chest X-ray and computed tomography scans were normal, with no evidence of pulmonary embolus, lung consolidation, pelvic abscess, or osteomyelitis of the femoral heads. Transthoracic echocardiogram showed mild tricuspid regurgitation and enlargement of both atria but was otherwise normal, with no valvular vegetations. Blood cultures, urine culture, and throat culture were obtained with subsequent negative results. She was admitted to our Obstetrical Intensive Care OBSTETRICS & GYNECOLOGY Unit in guarded condition with what was thought initially to be an infection versus crisis, and antibiotic therapy with ampicillin, gentamicin, and clindamycin was initiated. Supportive care, intravenous fluids, oxygen support, iron, folate, and narcotic analgesia were given. She remained febrile with tachycardia to the 150s until hospital day 5. Blood pressure and urine output were stable throughout. Serial complete blood counts revealed decreasing hemoglobin and hematocrit to a nadir of 4.9 g/dL and 14.6% on hospital day 4, postoperative day 14. Hemolysis was present on peripheral blood smear. Hemoglobin electrophoresis revealed S 89%, F 4%, A 0%, A2 4.4%. Her direct antiglobulin test was positive, and red cell serology identified anti-I antibodies that had not been detected preoperatively. Further advice was sought from the sickle cell team, blood bank, and American Red Cross. A blood type and crossmatch found no compatible blood available in the southeastern United States for the patient at this time secondary to her multiple auto- and alloantibodies, including anti-E, anti-Fy(A), anti-C, anti-c, anti-Jk(A), anti-Jk(B), and anti-s, and the new finding of high levels of anti-I antibodies. She was begun on erythropoietin injections. Supportive care was weaned as tolerated. On hospital day 10, postoperative day number 16, she was discharged home in stable condition with hemoglobin of 5.0 with plans to continue erythropoietin as an outpatient. COMMENT Sickle cell disease is the most common inherited blood disorder in the United States, affecting over 70,000 individuals.1 Growing numbers of women with sickle cell disease are choosing to have children, which presents the obstetrician with many unique issues. The often profound anemia characteristic of this disease raises the question of blood transfusion versus conservative management for optimal maternal outcomes, particularly in the operative setting. The American College of Obstetricians and Gynecologists notes that there is no consensus regarding the hematocrit level below which pregnant sickle cell patients should be transfused.5 Moreover, the blood loss experienced during normal delivery may worsen anemia and trigger crisis. Each red cell membrane contains millions of antigens. Of these, I is a rare antigen with a reported distribution of 500,000 per RBC. Approximately one in 260,000 patients will have a delayed transfusion reaction due to antibodies to minor red cell antigens6 that were not detected by the routine pretransfusion antibody assay. This is particularly true in sickle cell patients who often have a history of multiple blood transfusions and an increased incidence of alloimmunization. The rate of alloimmunization in sickle cell VOL. 110, NO. 2, PART 2, AUGUST 2007 patients is estimated to be between 18% and 36%.2 It is estimated that six to eight deaths per year in the United States occur as a result of delayed transfusion reactions.6 Differential diagnosis of pain and fever in post– cesarean delivery sickle cell patients includes crisis, infection, and postoperative complications including pulmonary embolism. Delayed transfusion reaction should be included in this differential diagnosis. Although rare, it can also be quite serious. It is the result of stimulation of an alloantibody by a foreign antigen on transfused erythrocytes. Such alloantibodies may fall below serologically detectable levels, and thus, they are not recognized when blood is being crossmatched. Levels rise when the appropriate antigen is again introduced to the patient’s immune system. A delayed transfusion reaction is classically characterized clinically by a triad of fever, hyperbilirubinemia, and anemia occurring 3–10 days after transfusion, as was demonstrated by our patient. Laboratory tests that aid diagnosis also include increased reticulocyte count, free hemoglobin in the urine, fragmented RBCs on peripheral smear, with the most convincing evidence being discovery of previously undetected alloantibodies. Although our patient’s crisis was initially thought to be an infection versus crisis, delayed hemolytic transfusion reaction with hyperhemolysis was ultimately diagnosed. Delayed transfusion reaction may occur with or without hyperhemolysis syndrome. The addition of hyperhemolysis syndrome poses a more serious threat to the patient because fatal outcomes have been reported. The exact mechanism of hemolytic transfusion reaction is not well understood, and the pathophysiology of hemolytic transfusion reaction seems to be far more complex because it involves the destruction of both patient and donor RBCs. Hemolysis on peripheral smear and increased reticulocyte count, as demonstrated by our patient, help to confirm the diagnosis. Hemolytic transfusion reaction is characterized by severe anemia after transfusion—a drop in posttransfusion hemoglobin below the pretransfusion levels, suggesting that there is destruction of both transfused and autologous cells. The lack of any A hemoglobin on hemoglobin electrophoresis despite transfusion 12 days before confirmed the diagnosis of a delayed transfusion reaction with hyperhemolysis syndrome because one would expect hemoglobin A in a recently transfused patient unless hemolysis has resulted in its extinction. Succumbing to the temptation to further transfuse these patients may have Proudfit et al Delayed Transfusion Reaction 473 lethal consequences because the reaction is worsened by the addition of further blood. Perhaps the most humbling lesson to be learned from our patient is the importance of listening to the patient herself. The patient repeatedly stated throughout her antepartum, intrapartum, and postpartum course that she was extremely resistant to blood transfusion, given her history of multiple transfusion reactions in the past. Because of this case, we have, in consultation with our anesthesiology department, revised our guidelines for blood transfusion in patients with sickle cell disease. Blood transfusion is at times unavoidable in sickle cell patients, and any transfused erythrocytes should be matched for minor antigens. As this cautionary tale demonstrates, a patient’s alloimmunization status and previous transfusion history must be reviewed before transfusion in weighing risks and benefits of prophylactic blood transfusions for the gravid sickle cell patient. REFERENCES Hereditary Hemorrhagic Telangiectasia and Pregnancy tesis was performed at the 32nd week for hydramnios. Because of a nonreassuring fetal heart rate pattern, she was delivered by cesarean of a female infant. At the beginning of the postpartum period, her pulmonary function spontaneously improved, and mother and female infant were discharged in good condition. Christof Worda, MD, Irene Lang, MD, Peter Husslein, MD, and Meinhard Kneussl, MD BACKGROUND: Hereditary hemorrhagic telangiectasia is a rare but life-threatening disease characterized by telangiectasias and arteriovenous malformations in different organs. When it is associated with pregnancy, this disease causes significant morbidity and mortality. CASE: We describe the pregnancy course of a primigravida with severe hereditary hemorrhagic telangiectasia. Her antepartum course was complicated by reduced pulmonary function, mild tricuspid regurgitation, and mild pulmonary hypertension. An amniocen- From the Department of Obstetrics and Gynecology; Department of Internal Medicine II, Division of Cardiology; and Department of Internal Medicine IV, Pulmonary Division, Vienna General Hospital, Medical University of Vienna, Vienna, Austria. Corresponding author: Christof Worda, MD, Department of Obstetrics and Gynecology, Medical University of Vienna, Waehringer Gurtel 18-20, A-1090 Vienna, Austria; e-mail: christof.worda@meduniwien.ac.at. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 474 Worda et al Hereditary Hemorrhagic Telangiectasia 1. National Institutes of Health, National Heart, Lung, and Blood Institute. Facts about sickle cell anemia (NIH publication 96-4057). Washington, DC: National Institutes of Health; 1996. 2. Aygun B, Padmanabhan S, Paley C, Chandrasekaran V. Clinical significance of RBC alloantibodies and autoantibodies in sickle cell patients who received transfusions. Transfusion 2002;42:37–43. 3. Vichinsky EP, Haberkem CM, Neumayr L, Earles AN, Black D, Koshy M, et al. A comparison of conservative and aggressive transfusion regimes in the perioperative management of sickle cell disease. N Engl J Med 1995;333:206–13. 4. Win N, Doughty H, Telfer P, Wild BJ, Pearson TC. Hyperhemolytic transfusion reaction in sickle cell disease. Transfusion 2001;41:323–8. 5. Hemoglobinopathies in pregnancy. ACOG Practice Bulletin No. 64. American College of Obstetricians and Gynecologists. Obstet Gynecol 2005;106:203–11. 6. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. First of two parts— blood transfusion N Engl J Med 1999;340:438–47. 7. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion medicine. Second of two parts— blood conservation. N Engl J Med 1999;340:525–33. CONCLUSION: Pregnancy is complicated in women with severe hereditary hemorrhagic telangiectasia, but can be managed successfully. (Obstet Gynecol 2007;110:474–7) H ereditary hemorrhagic telangiectasia (RenduOsler-Weber syndrome) is a rare disease with a prevalence of one in 10,000. When it is associated with pulmonary and brain lesions it carries a high morbidity and mortality. The clinical manifestations, such as nasal and gastrointestinal bleeding, cutaneous lesions, dyspnea, cyanosis, headache, brain abscesses, and transient ischemic stroke, are caused by arteriovenous malformations in the affected organs. Telangiectasias are dilated and convoluted venules with many layers of smooth muscle cells lacking elastic fibers. Arteriovenous malformations are seen in certain forms of this disease and represent direct shunts between arteries and veins with larger diameters than telangiectasias.1 Hereditary hemorrhagic telangiectasia is inherited in an autosomal dominant pattern, with an age-related penetrance. Mutations in endoglin and activin receptor- like kinase 1 genes, encoding transforming OBSTETRICS & GYNECOLOGY growth factor ␤ receptors, have been found associated with the disease,2 being particularly associated with pulmonary hypertension as a complication of hereditary hemorrhagic telangiectasia. Hereditary hemorrhagic telangiectasia may become life-threatening for mother and infant during pregnancy and delivery because of the cardiovascular consequences of extensive arteriovenous connections. These can lead to the development of significant arteriovenous shunts. We present the successful obstetric management of a mother with hereditary hemorrhagic telangiectasia and describe in detail the evolution of pulmonary function parameters at different stages of pregnancy. CASE A 28-year-old primigravida in her 12th week of pregnancy with a family history of hereditary hemorrhagic telangiectasia was referred to our hospital. Her medical history revealed a miscarriage in the eighth gestational week and an upper left lobectomy. On admission the patient was cyanotic and dyspneic at rest (functional class: New York Heart Association III to IV). Pulmonary function tests showed a severely reduced total lung capacity and vital capacity, with a moderately increased residual volume. A respiratory, noncompensated alkalosis with reduced arterial oxygen partial pressure at rest and decreased arterial carbon dioxide partial pressure was observed (Table 1). Echocardiographic examination showed mild tricuspid regurgitation (Vmax 2.9 m/s) with increased systolic pulmonary artery pressure of 40 mm Hg at rest, indicating mild pulmonary arterial hypertension. Fetal measurements were appropriate for gestational age. Blood test results were normal, and arterial blood pressure was within the normal range (120/80 mm Hg). A multislice computed axial tomographic scan demonstrated enlarged pulmonary arteries with signs of pulmonary hyperperfusion and numerous small arteriovenous malformations. In addition, multiple telangiectases of the skin were present. After discussing the potential complications of hereditary hemorrhagic telangiectasia in pregnancy, the patient decided to continue with her pregnancy. When 100% oxygen over a tight facial mask was given, incomplete saturation was observed, and a 40% shunt was calculated (Table 1). She was hospitalized for 2 weeks, and nasal oxygen was given at 3 L/min. The patient was examined every other week from then on (Table 1), and fetal biometry was performed at two weekly intervals. In the 28th week of gestation, hydramnios was diagnosed. In the 32nd week of gestation, the patient underwent amniocentesis because of the increasing hydramnios (maximal vertical pocket greater than 13 cm), and steroids were given to mature the fetal lungs. Amniocentesis yielded 1,250 mL of amniotic fluid with normal insulin, VOL. 110, NO. 2, PART 2, AUGUST 2007 glucose, leukocytes, and interleukin 8 concentrations. During the same week, a pathological cardiotocography, with repeated late decelerations, was the indication for delivery by cesarean. A 1,767-g female infant was delivered with Apgar scores of 9, 10, and 10 at 1, 5, and 10 minutes, respectively, and was taken to the neonatology department. Three days later the infant was transferred to the postnatal ward. There were no complications in the postpartum period although, at the beginning, the mother had worsening of pulmonary function parameters. This improved spontaneously. Echocardiography of the mother showed an enlarged left and right atrium, with normal hemodynamic parameters and an estimated systolic pulmonary artery pressure of 28 mm Hg at rest. A second multislice computed axial tomographic scan was performed 3 weeks after the cesarean delivery. It demonstrated no progression of the arteriovenous malformations. Mother and baby were discharged after 4 weeks in good condition. COMMENT Approximately 5–15% of patients with hereditary hemorrhagic telangiectasia have pulmonary arteriovenous malformations. These are associated with high morbidity and mortality.3,4 During the pregnancy circulating blood volume increases by more than 50%, resulting in an additional burden for the mother’s cardiovascular system. In combination with a decreasing functional lung capacity and the growing fetus, this results in an increased mismatch of oxygen requirements and delivery. In the postpartum period, additional fluid is shifted to the maternal circulation, causing increasing shunt volumes with worsening hypoxia. We have reported pulmonary function parameters at different stages of pregnancy and postpartum period complicated by hereditary hemorrhagic telangiectasia. Despite a 50% increase in the shunt volume during the course of the pregnancy and the postpartum period, our patient was successfully treated with oxygen. The fetus did not show any signs of hypoxemia. Furthermore, our case demonstrates that the early postpartum period is characterized by worsening of the pulmonary function parameters. Signs of volume overload of the right ventricle rapidly improved after delivery. On the whole, the outcomes of mother and infant were excellent. REFERENCES 1. Guttmacher AE, Marchuk DA, White RI Jr. Hereditary hemorrhagic telangiectasia. N Engl J Med 1995;333:918–24. Worda et al Hereditary Hemorrhagic Telangiectasia 475 476 Worda et al Hereditary Hemorrhagic Telangiectasia OBSTETRICS & GYNECOLOGY 1.87 (61) 1.68 (54) 1.69 (55) 1.79 (58) 1.76 (48) 2.02 (55) FEV1 (L) 2.02 (55) 1.97 (54) VC (L) 89 (100) 96 (108) 93 (104) 85 (96) FEV1% (F) VC (%) 2.37 (54) 2.8 (64) 2.87 (65) 2.51 (57) MEF 50% FVC (L/s) 3.4 (70) 3.48 (71) 3.52 (72) 3.16 (65) TLC (L) 63 70 60 61 41 50 43 38 FRC% RV% TLC (%) TLC (%) 3.17 2.84 3.35 3.32 DLCO (Units) 64 65 63 56 76 72 47 50 54 36 26 29 30 32 32 29 32 19 44 60 53 56 36 39 68 62 62 195 63 323 32 40 23 Shunt PaO2 PaO2 PaCO2 AaDO2 (100% O2) Volume (%) (mm Hg) (mm Hg) (mm Hg) (mm Hg) VC, vital capacity; L, liter; FEV1, forced expiratory volume in 1 second; FEV1%(F)VC, forced expiratory volume in 1 second as a percentage of (forced) vital capacity; MEF 50% FVC, maximal expiratory flow 50% of forced vital capacity; L/s liters per second; TLC, total lung capacity; FRC, functional residual capacity; FRC%TLC, functional residual capacity as a percentage of total lung capacity; RV%TLC, residual volume as a percentage of total lung capacity; DLCO, pulmonary diffusing capacity for carbon monoxide; PaO2, arterial oxygen partial pressure; PaCO2, arterial carbon dioxide partial pressure; AaDO2, alveolar-arterial oxygen difference; PaO2 (100% O2), arterial oxygen partial pressure with 100% O2 insufflations; 3L/O2, 3 liters per minute oxygen supplied. Percentages in parentheses are percentages of the normal values. 11⫹0 wk (3L/O2) 13⫹1 wk 20⫹2 wk 27⫹0 wk 29⫹1 wk (3L/O2) 31⫹1 wk (3L/O2) 12 d postpartum 39 dpostpartum 72 dpostpartum Time of Examination Table 1. Pulmonary Function Parameters 2. Shovlin CL, Letarte M. Hereditary haemorrhagic telangiectasia and pulmonary arteriovenous malformations: issues in clinical management and review of pathogenic mechanisms. Thorax 1999;54:714–29. 3. Swinburne AJ, Fedullo AJ, Gangemi R, Mijangos JA. Hereditary Pregnancy After Tumor Debulking and Intraperitoneal Cisplatin for Appendiceal Carcinoid Tumor Gina M. Mantia Smaldone, MD, Scott D. Richard, MD, Thomas C. Krivak, MD, Joseph L. Kelley III, MD, and Robert P. Edwards, MD BACKGROUND: Pregnancy after cytoreductive surgery and intraperitoneal chemotherapy is rare. CASE: We present the case of a 25-year-old woman with appendiceal carcinoid tumor treated with intraperitoneal cisplatin for peritoneal recurrence after a fertility-sparing cytoreductive procedure. Five years after her procedure, she conceived with the help of assisted reproductive technologies and delivered a viable term fetus via cesarean delivery. She subsequently had a successful second pregnancy and is currently alive without evidence of her disease. CONCLUSION: Conception is possible after tumor debulking and intraperitoneal chemotherapy. In unique clinical situations, fertility-sparing surgery may be considered in young patients who desire future pregnancy. (Obstet Gynecol 2007;110:477–9) I ntraperitoneal (IP) chemotherapy is often an adjunct to therapy for patients diagnosed with peritoneal malignancies. The high local drug concentration in IP therapy produces greater therapeutic ratios than those obtained with intravenous treatment. Intraperi- From the Division of Gynecologic Oncology, Department of Obstetrics, Gynecology and Reproductive Sciences, Magee-Womens Hospital, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania. Corresponding author: Gina M. Mantia Smaldone, MD, Division of Gynecologic Oncology, Department of Obstetrics, Gynecology and Reproductive Sciences, Magee-Womens Hospital, University of Pittsburgh Medical Center, 300 Halket Street, Pittsburgh, PA 15213; e-mail: mantiagm@upmc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 telangiectasia and multiple pulmonary arteriovenous fistulas. Clinical deterioration during pregnancy. Chest 1986;89:459–60. 4. Shovlin CL, Winstock AR, Peters AM, Jackson JE, Hughes JM. Medical complications of pregnancy in hereditary haemorrhagic telangiectasia. QJM 1995;88:879–87. toneal chemotherapy along with surgical cytoreduction may offer the best chance for survival in those individuals with ovarian cancer as well as with other peritoneal cancers. With improved survival stemming from advances in cancer therapy, there has been an emphasis on improving the quality of life for cancer survivors. Assisted reproductive technologies (ART) have been investigated as means for cancer patients to preserve reproductive capabilities. We report a case of successful pregnancy via in vitro fertilization (IVF) after tumor debulking and IP cisplatin therapy for a recurrent, metastatic appendiceal carcinoid tumor. CASE A 25-year-old nulliparous woman initially presented to an outside facility for infertility evaluation. Her laparoscopic evaluation revealed a right hemorrhagic ovarian cyst and small nodules on both uterosacral ligaments. Uterosacral biopsies revealed an epithelial tumor with glandular neuroendocrine differentiation with carcinoid features; this diagnosis was verified by pathologists at Massachusetts General Hospital as metastatic carcinoid tumor. The patient was then referred to our care for definitive management. She underwent an exploratory laparotomy and fertility-sparing tumor debulking. Gross disease was noted in the appendix and its stump, along with a nodule in the posterior cul-de-sac. The abdomen was carefully explored, and all palpable disease was completely removed. Final pathology revealed a 1.7-cm appendiceal carcinoid tumor with metastasis in the cul-de-sac nodule. This patient refused any adjuvant treatment. Unfortunately, 11 months later she represented with increasing right lower quadrant pain. A second-look laparoscopy revealed nodules on both uterosacral ligaments, with pathology consistent with recurrent disease. After consultation with medical oncology, IP chemotherapy was recommended and accepted by the patient. A Bard peritoneal port was placed, and the patient underwent four courses of monthly IP cisplatin (75 mg/m2). Follow-up computed tomography (CT) scan at the completion of her treatment revealed no evidence of disease. One year after IP chemotherapy, the patient reported regular menses. Given her history of primary infertility and IP chemotherapy, the patient explored ART options to achieve conception. She underwent a clomiphene citrate challenge test, suggesting appropriate ovarian reserve, and ovulation was indicated by endometrial biopsy during the VOL. 110, NO. 2, PART 2, AUGUST 2007 Mantia Smaldone et al Pregnancy After Intraperitoneal Chemotherapy 477 478 Mantia Smaldone et al Pregnancy After Intraperitoneal Chemotherapy OBSTETRICS & GYNECOLOGY Appendectomy, resection of uterosacral nodule Uterosacral ligaments (bilateral) Omentum, uterine deposits Malignant ascites Malignant ascites Recurrence None None Cyclophosphamide Doxorubicin vinblastine Methotrexate actinomycin D Cyclophosphamide cisplatin Intravenous Chemotherapy Hyperthermic cisplatin (428 mg),and postoperative paclitaxel, 5FU Cisplatin (75 mg/m ) Cisplatin (50 mg/m2) Cyclophosphamide (100 mg bid ⫻ 10 days) Intraperitoneal Chemotherapy 60 ART Clomid/IUI; IVF Embryo cryopreservation NA 26 NA NA 60 Time to Pregnancy (mo) ART, assisted reproductive technology; LSO, left salpingo-oophorectomy; NA, not applicable; RSO, right salpingo-oophorectomy; 5FU, 5-fluorouracil; IUI, intrauterine insemination; IVF, in vitro fertilization. Mantia Appendiceal Smaldone carcinoid et al Initial Resection Omentectomy (partial); argon ablation of implants RSO Borderline ovarian papillary serous cystadenocarcinoma Peritoneal epithelioid mesothelioma Shaves et al5 Kyser et al6 LSO Endodermal sinus tumor Ward et al4 Case Primary Malignancy Table 1. Pregnancy After Intraperitoneal Chemotherapy luteal phase. She had a normal postcoital examination. The patient was able to conceive on three occasions with help of clomiphene citrate and intrauterine insemination but spontaneously aborted all three pregnancies. Five years after the completion of her IP chemotherapeutic regimen, the patient achieved successful conception via IVF. Because of her repetitive pregnancy losses, a thrombophilia evaluation was performed, revealing a MTHFR gene variant. She received empiric treatment with aspirin and folic acid and weekly progesterone intramuscular injections. After an uncomplicated pregnancy, she delivered a vigorous male infant by a primary low transverse cesarean secondary to arrest of active phase. The patient continues to be followed by serial tumor markers and CT scans on an annual basis. The patient has now been disease free for 9 years. Coincidentally, she subsequently had a successful second pregnancy achieved by natural conception. COMMENT Appendiceal neoplasms are rare, accounting for 0.4 – 1.0% of all gastrointestinal tract malignancies, with the most common being carcinoid tumors.1 Sandor and Modlin2 reported an incidence of regional metastases ranging from 3.8% to 26.8%. Appendectomy is often curative, especially for lesions smaller than 2 cm. Currently, there is no clear consensus on the use of postoperative chemotherapy. The effects of intravenous systemic chemotherapy on reproductive potential have been investigated; the effects of IP chemotherapy, however, have not been specifically reported. Chemotherapeutic agents, in particular alkylating agents, have a direct cytotoxic effect on the ovaries, which may result in premature ovarian failure and, thus, infertility. A model for comparison is the use of high-dose alkylating chemotherapy during the preoperative preparation for cancer patients receiving bone marrow transplantation.3 Tauchmanová and colleagues3 published a retrospective analysis of 40 patients, including 21 women, with hematologic diseases who were treated with highdose alkylating therapy before allografic bone marrow transplantation. Ninety-five percent of all women studied had permanent ovarian insufficiency. To date, only three cases exist in the literature that document pregnancy after IP chemotherapy (Table 1).4 – 6 The first two cases document successful pregnancies in two individuals treated with a combination of intravenous and IP chemotherapy.4,5 It is unclear from these case reports whether reproductive technologies were used to facilitate conception. How- ever, Kyser and colleagues6 reported the use of embryo cryopreservation and the clomiphene citrate challenge test in a young woman with peritoneal epitheliod mesothelioma. Ginsburg and colleagues7 performed a retrospective review to compare outcomes of those patients using ART before chemotherapy with those using ART after chemotherapy. Even though patients who received chemotherapy before IVF were often younger, they experienced poorer responses to ovulation induction and subsequent lower pregnancy rates. Similarly, Dolmans and colleagues8 reported a statistically higher number of cryopreserved embryos for individuals receiving IVF before chemotherapy. Currently, there are no reported studies specifically examining the use of ART in cancer patients of reproductive age previously treated with IP chemotherapy. With improving cancer survival rates in women of reproductive age, the importance of fertility conservation is increasing. Assisted reproductive technology options to preserve reproductive capabilities should be discussed with patients who have new cancer diagnoses during their pretreatment evaluation. REFERENCES 1. Connor SJ, Hanna GB, Frizelle FA. Appendiceal tumors: retrospective clinicopathologic analysis of appendiceal tumors from 7,970 appendectomies. Dis Colon Rectum 1998;41: 75–80. 2. Sandor A, Modlin IM. A retrospective analysis of 1570 appendiceal carcinoids. Am J Gastroenterol 1998;93:422–8. 3. Tauchmanova L, Selleri C, Rosa GD, Pagano L, Orio F, Lombardi G, et al. High prevalence of endocrine dysfunction in long-term survivors after allogeneic bone marrow transplantation for hematologic diseases. Cancer 2002;95:1076–84. 4. Ward BG, Harvey VJ, Shepherd JH. Pregnancy after treatment of endodermal sinus tumour. Case report with five-year survival. Br J Obstet Gynaecol 1982;89:769–70. 5. Shaves M, Kamps RR, Laufman LR, Runowicz CD. A successful term pregnancy following the systemic and intraperitoneal administration of cisplatin chemotherapy. Gynecol Oncol 1990;39:378–80. 6. Kyser K, Bidus MA, Rodriguez M, Rose GS, Elkas JC. Spontaneous pregnancy following cytoreduction with peritonectomy and hyperthermic intraperitoneal chemotherapy. Gynecol Oncol 2006;100:198–200. 7. Ginsburg ES, Yanushpolsky EH, Jackson KV. In vitro fertilization for cancer patients and survivors. Fertil Steril 2001;75: 705–10. 8. Dolmans MM, Demylle D, Martinez-Madrid B, Donnez J. Efficacy of in vitro fertilization after chemotherapy. Fertil Steril 2005;83:897–901. VOL. 110, NO. 2, PART 2, AUGUST 2007 Mantia Smaldone et al Pregnancy After Intraperitoneal Chemotherapy 479 Carnitine Deficiency in Pregnancy Christopher T. Donnelly, MD, Afshan B. Hameed, MD, Jose E. Abdenur, MD, and Deborah A. Wing, MD BACKGROUND: Carnitine deficiency is a potential cause of metabolic crisis during periods of high energy demand or stress. Affected individuals have very low carnitine levels in blood, decreased carnitine transport in fibroblasts, and commonly have mutations in the OCTN2 gene. CASE: We report management through pregnancy and delivery of a patient with carnitine deficiency who had reduced carnitine transport in fibroblasts, but no mutations in the OCTN2 gene. CONCLUSION: Carnitine deficiency can be treated with exogenous carnitine in select patients during pregnancy. This is especially helpful, because carnitine levels decrease during pregnancy in normal individuals, and neonates are dependent on exogenous carnitine. (Obstet Gynecol 2007;110:480–2) organic cation and carnitine transporter with high specificity for carnitine, OCTN3, has been cloned in mice.2 Carnitine deficiency can be secondary to several causes, including metabolic diseases (ie, organic acidemias or other fatty acid oxidation defects), renal tubular defects, vegetarian diets, severe intestinal malabsorption, hemodialysis, or valproate treatment. A carnitine transporter defect due to mutations in the OCTN2 gene is the most common cause for inherited carnitine deficiency. In most patients with carnitine transporter defect, exogenous carnitine supplementation yields an excellent prognosis.3 Carnitine deficiency can be suspected to be due to defects in other transporters in individuals who have no mutation in the OCTN2 gene, yet have low plasma carnitine and reduced carnitine transport in fibroblasts. Pregnancy poses an additional risk to patients with carnitine deficiency, because plasma carnitine levels normally decline during pregnancy.4 We describe a pregnant patient with carnitine deficiency of unknown cause who was managed with carnitine supplementation through pregnancy and delivery. CASE C arnitine is an amino acid derivative. About 75% of carnitine sources are dietary (red meat and dairy), and the remainder is endogenously synthesized. Carnitine has several important intracellular functions, including the transport of long chain fatty acids across the mitochondrial membranes, which allows the ␤-oxidation process in various tissues, including skeletal and cardiac muscle. Transport of carnitine into the cells is an active process carried by high- and low-affinity human carnitine transporters (OCTN2 and OCTN1, respectively).1 They belong to a family of organic cation and carnitine transporters that function in different tissues, including heart, muscle, kidney, liver, intestine, and placenta. A third From the Department of Obstetrics and Gynecology, University of California, Irvine, California and Division of Metabolic Disorders, Children’s Hospital of Orange County, Orange, California. The authors thank Nicola Longo, MD, PhD, University of Utah for sequencing of the OCTN2 gene and Kelly King Covault, RN, Children’s Hospital of Orange County. Corresponding author: Deborah Wing, MD, University of California, Irvine, Department of Obstetrics and Gynecology, 101 the City Drive, Building 22A, Third Floor, Route 81, Orange, CA 92868; e-mail: dwing@uci.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 480 Donnelly et al Carnitine Deficiency in Pregnancy Our patient was a 28-year-old primigravida who presented in the first trimester. She was diagnosed with carnitine deficiency by muscle biopsy at age 12, with symptoms of muscle weakness and chronic fatigue. Oral carnitine treatment led to prompt resolution of her symptoms. Review of patient records disclosed four carnitine levels from ages 12–20 years, while on carnitine treatment, with results ranging 16.0 to 25.0 (normal 30 –73) ␮M for total carnitine and 12.3–20.8 (normal 19 – 60) ␮M for free carnitine. Over time, the patient decided to discontinue treatment. Later, she became pregnant and was referred to the Metabolic Clinic at Children’s Hospital of Orange County. Her family history is significant for a younger brother diagnosed with carnitine deficiency by muscle biopsy, who received carnitine treatment (oral and intravenous) for several years. The patient’s mother was diagnosed with cardiomyopathy and was also treated with carnitine therapy. Our patient’s physical examination results were normal. Laboratory tests, including renal function (blood urea nitrogen, creatinine, urine analysis, electrolytes), creatine kinase, chemistry panel, urine organic acids, and acyl carnitine profile were all normal, except for a borderline elevation of alanine transaminase of 63 (normal 1–55) international units/L, which subsequently normalized. The patient’s EKG and echocardiogram remained within normal limits. Free plasma carnitine at 7 weeks gestation was 14 ␮M (normal mean for pregnant women 25 ␮M),5 whereas urine carnitine levels were elevated at 39.3 (normal 22.6) and 24.0 (normal less than 10.3) mmol/L creatinine for total and free carnitine respectively. The patient’s carnitine OBSTETRICS & GYNECOLOGY transport assay in fibroblasts was reduced (22%), but no mutation in the OCTN2 gene could be identified (N. Longo, University of Utah, personal communication). She was initiated on carnitine at 3 g/d, and titrated up to 4.5 g/d at 15 weeks, then to 8 g/d from 19 weeks of gestation until delivery. Her free carnitine levels were initially low (range 11–16 ␮mol/L) but normalized to 18 ␮mol/L (normal for the gestational age) with the increased dose of carnitine to 8 g/day (Table 1). The pregnancy was uncomplicated until 39 weeks, at which time the patient presented with elevated blood pressures, proteinuria, and a mild headache unresponsive to conventional measures. The patient was diagnosed with preeclampsia and the decision was made to effect delivery. Upon admission, she was given intravenous 5% dextrose lactated Ringer’s solution and intravenous carnitine supplementation at a dose of 4 g over 24 hours. Because of the potential for energy exhaustion during a long labor induction, the patient was offered and consented for primary cesarean delivery. As such, the patient was never in labor before her cesarean. She delivered a viable male infant weighing 2,830 g with Apgar scores of 8 at 1 minute and 9 at 5 minutes. Her postoperative course was unremarkable. She resumed oral carnitine at 4 g/d on the first postoperative day. She was discharged to home on her third postoperative day with the newborn. A free carnitine level in the newborn at time of discharge was 13 (normal 15–55) ␮mol/L. The newborn’s total and esterified carnitine were normal at 31 (normal 21– 83) and 18 (normal 4 –29) ␮mol/L respectively. COMMENT Carnitine deficiency can be secondary to strict vegetarian diets, malabsorption, renal tubular defects or metabolic diseases (ie, organic acidemias). Those conditions were excluded in our patient by appropriate laboratory tests. Her clinical presentation, family history, and the reported results of her muscle biopsy are consistent with carnitine deficiency. The patient’s reduced carnitine transport (22%) is in the range for individuals with a heterozygous carnitine transporter defect gene mutation. However, a complete bidirectional sequence of the OCTN2 gene failed to detect any mutations. It is known that more than one transporter is involved in the intracellular carnitine transport,2 and therefore it is possible that a gene different from the OCTN2 may be responsible for the carnitine deficiency in our patient. The incidence of carnitine transporter defect is unknown, but observation studies in a Japanese population suggest a carrier frequency of about 1%.6 The availability of newborn screening with tandem mass spectrometry allows the detection of presymptomatic patients, and preliminary data suggests that the disease is more frequent than previously recognized. As with all disorders of fatty acid metabolism, clinical symptoms of carnitine transporter defect become evident during fasting or stress conditions. Cardiomyopathy, fasting hypoketotic hypoglycemia, hyperammonemia, Reye’s-like syndrome or skeletal muscle weakness may occur. Approximately 50% of patients present later in childhood with cardiomyopathy and rapidly progressive heart failure as the first manifestations of the disease.2,6,7 It is also common to find a family history of unexplained or sudden cardiac deaths.5–7 Individuals with a heterozygous carnitine transporter defect gene mutation have low plasma carnitine and may be at risk for cardiomyopathy in adult life, particularly if additional risk factors such as hypertension are present.2 In classical carnitine transporter defect, exogenous carnitine supplementation allows for increased diffusion of carnitine into the tissues (possibly through low-affinity carnitine transporters) and compensates for the renal wastage that is characteristic for the disorder.6 Plasma carnitine levels in our patient were low, even when adjusted for normal pregnant women. The dose of 8 g/d required to achieve a normal plasma value is in the range reported for the treatment of carnitine transporter defect. Absorption of oral carnitine is approximately 30%, therefore the intravenous administration of 4 g/d given during delivery represents indeed a higher dose than 8 g/d given orally and was given due to anticipated increased metabolic demand during this period. Our patient was given a solution with 5% dextrose during labor and delivery. In cases of prolonged stress or fasting it would be prudent to provide 10% dextrose solutions to provide more calories and decrease lipolysis. Carnitine is also known to be transported across Table 1. Carnitine Levels During Pregnancy Weeks of Pregnancy Normal values of free plasma carnitine (␮mol/L) 4 VOL. 110, NO. 2, PART 2, AUGUST 2007 7 11 15 19 24 32 25 25 22.5 20 16.5 16 Donnelly et al Carnitine Deficiency in Pregnancy 481 the placenta.8 Studies of cord blood of pregnancies that delivered at different gestational ages indicate that both free and esterified carnitine levels decrease during pregnancy.4 The neonate is especially sensitive to carnitine shortage given limited carnitine reserve after birth. Neonates of unaffected mothers obtain their carnitine supply from the mother’s breast milk. In a carnitine-deficient mother, decreased carnitine uptake by the neonate could result in failure to thrive and hypotonia and occasionally has been linked to sudden infant death syndrome.4 In summary, we report successful treatment with carnitine supplementation during pregnancy and the increased metabolic demands of delivery in a patient with carnitine deficiency. Studies in this patient also raise the possibility of carnitine deficiency due to genetic defects other than the common OCTN2 gene mutation. REFERENCES 1. Longo N, Amat di San Filippo C, Pasquali M. Disorders of carnitine transport and the carnitine cycle. Am J Med Genet C Semin Med Genet 2006;142:77–85. Stillbirth Due to Placental Hypoperfusion After SalpingoOophorectomy for an Ovarian Cyst Robert W. Bendon, MD, and Divya B. Cantor, MD BACKGROUND: Gravid oophorectomy past mid-pregnancy may be necessary, but the alterations of blood flow to supply the placenta may present risks to the mother and fetus. CASE: A salpingo-oophorectomy for a mucinous cystadenoma resulted in a postoperative hemorrhage of 2 L and fetal death. The placenta demonstrated a unique lesion that was consistent with global hypoperfusion of the placenta. From Kosair Children’s Hospital and Norton Suburban Hospital, Louisville, Kentucky. Corresponding author: Robert Bendon, MD, Department of Pathology, Kosair Children’s Hospital, 231 E. Chestnut Street, Louisville, KY 40202; e-mail: robert.bendon@nortonhealthcare.org. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 482 Bendon and Cantor Stillbirth Postoophorectomy 2. Tein I. Carnitine transport: pathophysiology and metabolism of known molecular defects. J Inherit Metab Dis 2003;26: 147–69. 3. Cederbaum SD, Koo-McCoy S, Tein I, Hsu BY, Ganguly A, Vilain E, et al. Carnitine membrane transporter deficiency: a long-term follow up and OCTN2 mutation in the first documented case of primary carnitine deficiency [published erratum appears in Mol Genet Metab 2003;78:82]. Mol Genet Metabol 2002;77:195–201. 4. Winter SC, Linn LS, Helton E. Plasma carnitine concentrations in pregnancy, cord blood, and neonates and children. Clin Chim Acta 1995;243:87–93. 5. Rinaldo P, Stanley CA, Hsu BY, Sanchez LA, Stern HJ. Sudden neonatal death in carnitine transporter deficiency. J Pediatr 1997;131:304–5. 6. Roe CR, Ding J. Mitochondrial fatty acid oxidation disorders. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular basis of inherited disease. New York (NY): McGraw-Hill; 2001. p. 2297–326. 7. Brivet M, Boutron A, Slama A, Costa C, Thuillier L, Demaugre F, et al. Defects in activation and transport of fatty acids. J Inherit Metab Dis 1999;22:428–41. 8. Lahjouji K, Elimrani I, Lafond J, Leduc L, Qureshi IA, Mitchell GA. L-carnitine transport in human placental brushborder membranes is mediated by the sodium-dependent organic cation transporter OCTN2. Am J Physiol Cell Physiol 2004;287:C263–9. CONCLUSION: The postoperative hemorrhage occurred despite good immediate operative hemostasis. Blood flow was shunted from the uteroplacental circulation due to the large utero-ovarian collateral circulation. (Obstet Gynecol 2007;110:482–4) D espite risks of preterm labor and other complications, mid-pregnancy abdominal surgery may be necessary. Salpingo-oophorectomy is complicated by the collateral circulation of the uterine and ovarian arteries and the massive increase in uterine blood flow to the placenta. CASE A 23-year-old white woman (gravida 3, para 2) at 27 weeks of gestation (based on a 19-week ultrasound examination) presented with a 3-day history of constant right lower quadrant pain. The patient denied fever, emesis, vaginal bleeding, or leaking of fluid. She experienced occasional nausea and received pain relief with narcotics. The patient had good fetal movement upon admission. Her prenatal history included an ultrasound examination at 19 weeks of gestation that demonstrated a 5⫻7 cm cystic mass in the right ovary. The patient failed to have a recommended follow-up ultrasonography until her arrival at the hospital at 27 weeks. Her past pregnancies were delivered by cesarean at term and at 34 weeks. She smoked one pack of cigarettes per day and OBSTETRICS & GYNECOLOGY did not use drugs or alcohol. She had no other significant medical history. She was admitted to the hospital to exclude appendicitis. A pelvic ultrasound examination confirmed the multiloculated cystic right adnexal mass, now measuring 6.0⫻9.2⫻10.0 cm. There were arterial and venous circulation of the mass, no malignant features, and no maternal ascites. No other fetal abnormalities were detected during this scan, and the amniotic fluid was within normal limits. Doppler velocimetry was not performed. The placenta was fundal. Because of the patient’s pain and increased size of the mass, she was taken to the operating room for excision after appropriate consents were obtained. A right salpingooophorectomy was performed under general anesthesia through a midline abdominal incision. The suture used for the salpingo-oophorectomy was O polyglactin (Vicryl, Ethicon, Summerville, NJ) free ties placed twice. Minimal blood loss was encountered, and hemostasis was confirmed before closure. Fetal heart tones were documented in the recovery room. The patient had significant coughing and writhing while in the recovery room due to general discomfort despite appropriate comfort measures and stable vital signs documented. After arriving on the antepartum floor, fetal heart tones could not be heard, and an ultrasound examination confirmed fetal demise. During this time, the patient had decreased urine output with other vital signs stable and a decrease in hemoglobin from her preoperative level (11.2 g/dL to 6.7 g/dL). She was taken to the operating room to explore for suspected intraabdominal bleeding. Intraoperatively, the serosa was seen to be denuded around the area of the excised adnexa, and therefore, this was the suspected site of bleeding even though no active bleeding was encountered. The ties were still visualized but no longer providing hemostasis. There was approximately 2 L of blood in the abdominal cavity. The adnexa were oversewn for hemostasis. Approximately 24 hours later, the patient vaginally delivered a stillborn female infant after a spontaneous onset of labor. The patient’s postoperative course was uncomplicated, and she recovered well. The cystic ovarian mass was a benign mucinous cystadenoma measuring 10.3⫻7.2⫻4.5 cm. The placenta had a centrally inserted, 22 cm long, 1.4 cm diameter, three-vessel umbilical cord. The membranes had ruptured apically and had no lesions. The placental disc measured 15⫻14⫻1.5 cm and weighed 180 g. The fetal and maternal surfaces of the placenta were unremarkable. The placenta was sliced at 1-cm intervals. The sections demonstrated a pattern of three to six areas of pallor in each slice with a pyramidal base similar to an infarction. Each of these discrete areas was surrounded by a halo of deeper red tissue (Fig. 1). On microscopic examination, the deeper red areas were due to vasodilation of acutely infarcted villi in the periphery of each placentome. The central areas were normal immature villi. In total, an estimated 40% of the placenta was acutely infarcted. The autopsy demonstrated a 27-week gestation, 950 g, appropriate-weight female infant without abnormality. Based on histological criteria, the death had occurred approximately 24 – 48 hours before delivery.1 There was no anatomic evidence of acute asphyxia (intrathoracic petechiae), consistent with asphyxia occurring over an hour from loss of placental respiration.2 Fig. 1. A slice of the gross placenta (A) and two hematoxylin and eosin–stained micrographs from the corresponding dark (B) and light (C) areas (⫻ 20, original magnification). The dark peripheral areas marked with a dark arrow on the gross correspond to the vasodilated, necrotic villi on the micrograph. B. The pale basal areas, marked with a white arrow on the gross image, correspond to micrograph C, which demonstrates normal villi for 27 weeks of gestation. Bendon. Stillbirth Postoopherectomy. Obstet Gynecol 2007. VOL. 110, NO. 2, PART 2, AUGUST 2007 Bendon and Cantor Stillbirth Postoophorectomy 483 Fig. 2. A slice of gross placenta showing the dark areas which are acute infarctions from spiral artery thrombi surrounded by pale normal tissue. This figure is for comparison with the reported case. Bendon. Stillbirth Postoopherectomy. Obstet Gynecol 2007. COMMENT The abdominal hemorrhage after obtaining surgical hemostasis was unexpected. Removal of ovarian tumors in the second trimester or at cesarean delivery has not been associated with lethal complications or hemorrhage.3 Our surgery was performed at a later gestation, 27 weeks, than most reported cases. The uterine artery blood flow in pregnancy increases from 40 mL/min to 500 mL/min. This flow is achieved by lowering resistance and hence pressure in the uterine arteries. The anastomoses of uterine and ovarian arteries are well known. After occlusion of ovarian arteries and after some initial arteriolar vasospasm, retrograde uterine arterial blood could begin to flow through small ovarian vessels that were not bleeding at the end of the procedure. We hypothesize that the increased uterine blood flow later in gestation increases the risk for postoperative hemorrhage after ovarian surgery in the gravid patient. 484 Bendon and Cantor Stillbirth Postoophorectomy This placenta demonstrated a pattern of infarction that inversely mirrors the usual pattern (Fig. 2). The lesion is explained by comparing it with the anatomy of normal intervillous perfusion.4 The intervillous space is perfused by numerous spiral arteries, each acting as an end artery to an overlying volume of villi. The spiral artery flow resembles fountains squirting blood into the intervillous space in overlapping patterns. In the typical placental infarction, the infarction is pyramidal, with the base on the maternal floor. Each infarction is the result of occlusion of one spiral artery. The surviving periphery of this infarction toward the fetal surface corresponds to overlapping peripheral blood flow from adjacent spiral arteries. In this placenta, the centers of spiral artery blood flow are viable, but the overlapping peripheries are all infarcted. This pattern can be explained if all spiral artery flow was at low pressure and was able to perfuse only a small area of villi near its inflow. Villi in the periphery of the flow did not receive adequate blood flow to survive. The pattern of infarction in the placenta explains the fetal death before the mother was in severe shock. The pelvic hemorrhage became a sink for uterine flow diverting it from the placenta, which we hypothesize resulted in infarction of 40% of the placenta. REFERENCES 1. Genest DR. Estimating the time of death in stillborn fetuses: II. Histologic evaluation of the placenta; a study of 71 stillborns. Obstet Gynecol 1992;80:585–92. 2. Bendon RW. Review of some causes of stillbirth. Pediatr Dev Pathol 2001;4:517–31. 3. Sherard GB, Hodson CA, Williams HJ, Semer DA, Hadi HA, Tait DL. Adnexal masses and pregnancy: a 12-year experience. Am J Obstet Gynecol 2003 Aug;189:358–62. 4. Ramsey EM, Donner MW. Placental vasculature and circulation: anatomy, physiology, radiology, clinical aspects. Philadelphia (PA): W. B. Saunders; 1980. OBSTETRICS & GYNECOLOGY Management of Cold Agglutinin– Immune Hemolytic Anemia in Pregnancy Simi Dhingra, J. J. Wiener, and Helen Jackson BACKGROUND: Cold hemagglutinin disease is an acquired autoimmune hemolytic anemia caused by an immunoglobulin M autoantibody directed against the polysaccharide antigens on the red blood cell surface. This case presents the challenges surrounding the management of cold hemagglutinin disease in pregnancy. CASE: A pregnant woman in her thirties with type-2 diabetes, reporting shortness of breath and productive cough, was found to have anemia, reticulocytosis, bilirubinemia, positive direct Coombs test result, positive cold agglutinin antibody, and raised lactate dehydrogenase levels. As the infection screen and autoimmune serology results were negative, she was diagnosed as having idiopathic cold hemagglutinin disease. The management included keeping the patient warm and hydrated and treating the anemia with warm packed red blood cell transfusion. CONCLUSION: Cold hemagglutinin disease is a rare condition. Investigations to rule out infections help determine the diagnosis of cold hemagglutinin disease of unknown origin. (Obstet Gynecol 2007;110:485–6) C old hemagglutinin disease is an acquired autoimmune hemolytic anemia caused by an immunoglobulin M autoantibody directed against the polysaccharide antigens on the red blood cell surface. It is commonly associated with infections, lymphoproliferative diseases, malignancies, and immunodeficiency syndromes.1 The associated hemolysis occurs at low temperature and can be difficult to treat. Patients with infections who have cold hemagglutinin disease often have a short clinical course, whereas those with lymphoma require therapy. Therapeutic maneuvers successful in patients with warm agglutinin hemolytic anemia such as corticosteroids, intravenous immunoglobulin G, and splenectomy, are usu- From the Department of Obstetrics and Gynecology and Hematology, Royal Gwent Hospital, Newport, United Kingdom. Corresponding author: Simi Dhingra, 15, Heol Mynydd Bychan, Cardiff, South Glamorgan, CF14 4NL; e-mail: sdhingra@doctors.org.uk. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 ally ineffective in cold hemagglutinin disease. We report here a rare case of primary cold hemagglutinin disease and its management in a pregnant woman. CASE A primigravida in her early thirties with diet-controlled type-2 diabetes was referred to the physicians at 22 weeks of gestation with a 1-week history of shortness of breath, productive cough, and tiredness. She was not taking any medications and denied any abdominal pain or vaginal bleeding. At a routine antenatal visit, 8 days before this presentation, her hemoglobin was 11.8 g/dL. The patient was pale, her chest was clear on auscultation, and her oxygen saturation was 100% on air. The uterus corresponded to 22–24 weeks of gestation, and the fetal heart sounds were heard. There was no spleenomegaly or lymphadenopathy. Hematological investigations showed severe anemia (hemoglobin 6g/dL), marked reticulocytosis (7.3%), bilirubinemia (41 mg/dL), a strongly positive direct Coombs test result, a positive cold agglutinin antibody, and raised lactate dehydrogenase levels (453 mg/dL). Examination of the peripheral smear showed red cell agglutination and polychromasia. The provisional diagnosis was cold agglutinin hemolytic anemia secondary to Mycoplasma pneumoniae, because of her respiratory symptoms. Infection screen and autoimmune serology results were reported negative, so the diagnosis was changed to idiopathic cold hemagglutinin disease. The management plan was conservative and included keeping the patient warm and hydrated and to treat the anemia with packed red blood cell transfusion using blood warmers. The upper respiratory chest infection was treated with oral erythromycin. She was also started on folic acid 5 mg daily. Because of an increased risk of thrombosis due to hemolytic anemia2 and systemic risk factors (pregnancy, obesity, and immobility) prophylactic subcutaneous heparin (40 mg daily) was commenced. She was monitored clinically and hematologically by daily full blood count, reticulocyte count, serum bilirubin, and serum lactic dehydrogenase. After transfusion of 4 units of blood, the clinical symptoms improved and the hemoglobin stabilized between 8.9 and 10.8 g/dL. On discharge, the patient was advised to keep herself warm and avoid exposure to cold. She was rehospitalized 3 days later because of a significant drop in hemoglobin to 6.7 g/dL, elevated reticulocyte percentage (7.9%), elevated lactate dehydrogenase level (456 units/L), and elevated bilirubin level (32 mg/dL). Additional prewarmed packed red blood cell transfusion (3 units) stabilized the hemoglobin between 8.6 and 8.8g/dL. Ten days later, the hemoglobin dropped to 6.9 g/dL and 3 more units of warm blood were transfused. Altogether, 10 units of packed red blood cells were transfused over a period of 18 days (Table 1). After this, the hemoglobin stabilized between 9 and 10.4 g/dlL and the patient was followed up in the antenatal clinic fortnightly. Only one more transfusion (2 units) was needed at 30 weeks of gestation and serial Dhingra et al Cold Agglutinin–Immune Hemolytic Anemia 485 Table 1. Gestational Age, Blood Levels, and Transfusion Gestational Age (wk) Hemoglobin (g/dL) Reticulocytosis (%) Lactate Dehydrogenase (units/L) Bilrubin (␮mol/L) Packed Cell Volume Transfusion (units) 6 6.7 6.9 8.4 7.3 7.9 6.4 6.2 453 456 350 322 41 32 24 24 4 3 3 2 22 23 24⫹4 30 growth scans were within normal limits. A planned induction was carried out at 39 weeks of gestation because of diabetes. She delivered vaginally a healthy baby weighing 2.9 kg. COMMENT Autoimmune hemolytic anemia is an uncommon disorder, and the incidence is reported to be 1 in 100,000.3 Warm auto antibodies are responsible for 48 –70% of autoimmune hemolytic anemia and cold agglutinins account for 16 –32%.4 Primary cold agglutinin anemia generally affects older adults, and it has a slight female predilection. This case presents a challenging problem of cold hemagglutinin disease in a pregnant woman who needed 12 units of warmed blood to treat the anemia. The management was mainly conservative and expectant. The initial diagnosis was thought to be cold hemagglutinin disease secondary to Mycoplasma pneumoniae because the respiratory symptoms preceded the anemia. Secondary cold agglutinins regularly occur in the course of Mycoplasma pneumoniae and infectious mononucleosis5 and less commonly with other viral infections like cytomegalovirus and varicella. In this case, further investigations to rule out infections (Paul Bunnell test; Mycoplasma serology; toxoplasmosis, rubella, cytomegalovirus, herpes; cytomegalovirus, parvovirus; Epstein Barr virus; hepatitis screen; serology for influenza; and adeno virus) along with autoimmune serology was negative. The diagnosis was therefore changed to cold hemagglutinin disease of unknown origin (idiopathic or primary). Hematology consultation was obtained at an early stage as sudden and life-threatening anemia is known to occur, requiring urgent coordination among clinicians, clinical pathologists, and blood bank personnel for appropriate management. Hemolysis may also be the first sign of an underlying systemic disorder, such as thrombotic thrombocytopenic purpura, lupus erythematosus, or chronic lymphocytic leukemia and may require urgent intervention to prevent death or disease-related complications. Once this patient’s anemia was corrected with 486 Dhingra et al packed red blood cell transfusion, she was further instructed to keep her warm as avoidance of cold is the most useful single therapy in cold agglutinin disease.6 Cytotoxic agents, such as cyclophosphamide and chlorambucil, in combination with corticosteroids have been used to reduce the production of antibody, especially when treating patients with cold hemagglutinin disease secondary to lymphoma.7 A number of reports have also shown the usefulness of rituximab, a monoclonal anti-CD20 antibody, in the treatment of patients with cold agglutinin disease and severe hemolysis, not responding to treatment with conventional therapy.8 However, its safety in pregnancy is not known, and it has been employed in only three cases. Two of these women had lymphoma and one had warm autoimmune hemolytic anemia.9 So far, the use of rituximab in pregnant women with cold hemagglutinin disease has not been reported. REFERENCES 1. Sokol RJ, Hewitt S. Autoimmune hemolysis: a critical review. Crit Rev Oncol Hematol 1985;4:125–54. 2. Hendrick AM. Auto-immune haemolytic anaemia: a high-risk disorder for thromboembolism? Hematology 2003;8:53–6. 3. Gehrs BC, Friedberg RC. Autoimmune hemolytic anemia. Am J Hematol 2002;69:258–71. 4. Sokol RJ, Hewitt S, Stamps BK. Autoimmune hemolysis: an 18-year study of 865 cases referred to a regional transfusion centre. Br Med J (Clin Res Ed) 1981;282:2023–7. 5. Horwitz CA, Moulds J, Henle W, Henle G, Polesky H, Balfour HH Jr, et al. Cold agglutinins in infectious mononucleosis and heterophil-antibody-negative mononucleosis-like syndromes. Blood 1977;50:195–202. 6. Petz LD. Treatment of autoimmune hemolytic anemias. Curr Opin Hematol 2001;8:411–6. 7. Azuma E, Nishihara H, Hanada M, Nagai M, Hiratake S, Komada Y, et al. Recurrent cold hemagglutinin disease following allogeneic bone marrow transplantation successfully treated with plasmapheresis, corticosteroid and cyclophosphamide. Bone Marrow Transplant 1996;18:243–6. 8. Engelhardt M, Jakob A, Ruter B, Trepel M, Hirsch F, Lubbert M. Severe cold hemagglutinin disease (CHD) successfully treated with rituximab. Blood 2002;100:1922–3. 9. Ojeda-Uribe M, Gilliot C, Jung G, Drenou B, Brunot A. Administration of rituximab during the first trimester of pregnancy without consequences for the newborn. J Perinatol 2006;26:252–5. Cold Agglutinin–Immune Hemolytic Anemia OBSTETRICS & GYNECOLOGY Alopecia Associated With Birth Injury Efstathios G. Lykoudis, MD, PhD, Georgia-Alexandra Ch. Spyropoulou, MD, Lazaros G. Lavasidis, MD, Minas E. Paschopoulos, MD, PhD, and Evangelos A. Paraskevaidis, MD, PhD BACKGROUND: Alopecia after birth-related caput succedaneum is an extremely rare complication. CASE: The case of a child with permanent alopecia due to birth-related caput succedaneum is presented. After delivery with vacuum extraction, caput succedaneum at the left occipitoparietal region of the neonate’s head was noted, which subsided within a week, leaving a circular necrotic crust and finally a circular bald area. At age 4, the child was referred at a tertiary center for the management of alopecia. Treatment initially consisted of the expansion of the hairbearing skin adjacent to the bald area, which was excised at a second stage and covered with the expanded skin. A pleasing esthetic result was achieved. CONCLUSION: Neonatal alopecia is a rare birth-associated complication. Premature rupture of the membranes, prolonged second stage of the labor, and prolonged vacuum extraction time may be important features in the pathogenesis of this complication. In case of permanent alopecia, excellent esthetic results can be achieved with the use of reconstructive plastic surgery techniques. (Obstet Gynecol 2007;110:487–90) B irth-associated fetal head injuries are classified, on the basis of their anatomic site, as extracranial (scalp), cranial, and intracranial. Scalp injuries are mainly related to pressure exerted by the cervix or instruments used during difficult vaginal deliveries. They range from superficial abrasions to life-threatening hemorrhages.1 The cosmetic effects of extracranial injuries are usually transient and have neither clinical significance nor long-term sequelae to the infant. Alope- From the Ioannina University School of Medicine, Departments of Plastic Surgery and Burns, and Obstetrics and Gynecology, Ioannina, Greece. Corresponding author: Efstathios G. Lykoudis, Assistant Professor of Plastic Surgery, Ioannina University School of Medicine, University Campus, 45110 Ioannina, Greece; e-mail: elykoudi@cc.uoi.gr. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 cia represents an extremely rare, but very distressing, complication of birth trauma. It is almost exclusively associated with caput succedaneum, although cephalhematoma and infection may also be implicated. 2–15 Caput succedaneum is a serosanguineous, subcutaneous, extraperiosteal fluid collection, associated with head molding, and characteristically extending across the midline and over suture lines, as opposed to cephalhematoma. Usually, it does not cause any complications, needs no management except for observation, and resolves within a few days. The pathophysiologic mechanisms related to the formation of caput succedaneum include the mechanical pressure on the fetal head exerted by the application of the vacuum extractor and the pressure exerted on the presenting head in utero, especially during a prolonged second stage of labor3 or after premature rupture of membranes.4 The occurrence of caput succedaneum associated with vaginal delivery may be underdiagnosed or underreported. In a retrospective study of 4,589 instrumental vaginal deliveries, 14 cases of caput succedaneum were reported (incidence 0.3%).16 The incidence of caput succedaneum related alopecia has not been reported thus far. We present a case of alopecia associated with caput succedaneum after vacuum extraction, and we review the existing literature, focusing on the delivery technique and the avoidance of this complication. We also present the etiology, pathophysiology, and the reconstructive management of this rare condition. CASE At 40 weeks of gestation, a 28-year-old, gravida 1 para 0, woman delivered vaginally a male neonate weighing 3,800 g. The patient had a history of ruptured amniotic membranes for 48 hours. At birth, a vacuum extractor (metallic ventouse, size 5) was used because of vertex presentation and fetal distress. The mother reported at least two cup detachments during the vacuum extraction process. According to the mother’s medical record, the second stage of labor lasted for 85 minutes. The Apgar scores were 5 and 7 at 1 and 5 minutes, respectively. The neonate was admitted to the neonate intensive care unit for 48 hours. Pertinent physical findings of the neonate included mild respiratory distress and a circular swollen area at the left occipitoparietal region of the scalp, with surrounding ecchymosis and excessive bruising, which were diagnosed as caput succedaneum. The caput succedaneum subsided within a week, leaving a circular necrotic crust. After the apoptosis of the crust, a circular bald area remained on the left occipitoparietal region of the scalp. No microbial contamination was diagnosed. Within the first year of life, Lykoudis et al Alopecia Induced by Birth Injury 487 hair regrowth was noticed centrally, as a tuft, with no further improvement of the alopecia. At age 4 years, the child was referred to the plastic surgery department of Ioannina University Hospital because of the persistent alopecia. On physical examination, the alopecia was located at the left occipitoparietal area and measured 10 cm in diameter. The skin was thin, scarred, and easily wrinkling, with a single tuft of hair located centrally (Fig. 1). The skin biopsy revealed excessive fibrosis. Reconstruction was achieved in two stages. Initially, tissue expansion of normal hair bearing scalp was performed. Two “sausage” type tissue expanders, 300 cm3 of volume each, were placed beneath the normal hair-bearing scalp adjacent to the area of alopecia. The expanders were then inflated on a weekly basis, until they reached the aforementioned volume within 3 months (Fig. 2). On completion of the expansion procedure, the bald area was excised and covered with the adjacent expanded hair-bearing scalp (Fig. 3). The postoperative course was uneventful, and a very pleasing esthetic result was achieved (Fig. 4). Fig. 2. Appearance of the patient during inflation period of the expanders. Lykoudis. Alopecia Induced by Birth Injury. Obstet Gynecol 2007. COMMENT Caput succedaneum, defined as edema in and under the fetal scalp, often occurs at birth, especially in vaginal delivery. Usually, it is a condition with good prognosis, with spontaneous regression within a few days. Risk factors for the development of caput succedaneum include nulliparity, prolonged second stage of labor, premature rupture of membranes, and vaginal delivery by vacuum extractor (notably duration of vacuum application longer than 10 minutes,17 cup misplacement or improper placement,18 –20 and repetitive cup detachments from the fetal head20,21). Here we report a rare case of nontransient neonatal alopecia, associated with mechanical extraction from the birth canal. Of note, in the present report, Fig. 1. Initial appearance of the patient. Lykoudis. Alopecia Induced by Birth Injury. Obstet Gynecol 2007. 488 Lykoudis et al Alopecia Induced by Birth Injury three predisposing factors were present: premature rupture of membranes, prolonged second stage of the labor, and vaginal delivery by vacuum extractor. Birth injury is any trauma to the infant resulting from mechanical forces during labor. Since the process of birth is a blend of compression, contractions, torques, and traction, these forces may inflict trauma. In addition, the use of obstetric instrumentation may amplify the effects of the aforementioned forces or may even induce birth trauma.4 Birth-related permanent alopecia is a rare obstetric complication. Vilanova5 reported the first case of alopecia associated with birth injury in 1961. The alopecia was attributed to bacterial contamination of caput succedaneum, coexisting with cephalhematoma, both induced by vacuum extraction. The alopecia was permanent, circular in pattern, with a tuft of hair centrally. Subsequently, five similar cases have been reported. In these cases, common findings were the use of vacuum extractor and the presence of a long extraction time.2,6,7,14,15 In 1984, Neal et al4 reported the first two cases of alopecia associated with caput succedaneum due to cervical pressure to the vertex during difficult, longlasting labors. It characteristically had a “halo ring” pattern because the pressure exerted by the cervix at the outer rim of the caput induced scalp ischemia, leading to hair loss. As a sequence, the term “halo scalp ring” alopecia was introduced. In both cases, the alopecia was temporary and normal hair regrowth was noticed within the first year of life. In addition, five more articles have been published, referring to seven cases of “halo scalp ring” alopecia.8 –11,13 It is OBSTETRICS & GYNECOLOGY Fig. 4. Final result. Lykoudis. Alopecia Induced by Birth Injury. Obstet Gynecol 2007. Fig. 3. Second stage operation. Excision of the bald area (A), followed by removal of the expanders and coverage with hair-bearing skin (B). Lykoudis. Alopecia Induced by Birth Injury. Obstet Gynecol 2007. worth mentioning that, in three of these cases, the alopecia was permanent. Therefore, it appears that two main pathophysiologic mechanisms are implicated in the formation of caput succedaneum. The first mechanism includes the pressure exerted on the presenting part of the fetal cranium against the dilating cervix after premature rupture of the membranes. Indeed, the formation of caput succedaneum is very frequently caused by the obstruction of the fetal descend during the second stage of labor.4 If there is rupture of the membranes in a full-term pregnancy, then labor is usually about to commence within the following 48 hours. During this interval, the fetal head is no longer protected by the amniotic fluid; when labor begins, the cervix exerts pressure upon the head, thus causing edema and microhemorrhages. The caput can occur more frequently in prolonged and unsupervised labors. VOL. 110, NO. 2, PART 2, AUGUST 2007 The second mechanism is attributed to the use of vacuum extraction during vaginal delivery (complications may include direct trauma, pressure ischemia to the underlying scalp, and venous congestion leading to interstitial fluid and microhemorrhages).3,12 The vacuum extractor (ventouse) has been designed to assist delivery by the application of traction to a suction cup, properly attached to the fetal scalp. If not properly used, the ventouse may cause the appearance of caput succedaneum. Indeed, vacuum extraction vaginal delivery may carry substantial risks. Therefore, only obstetricians who are adequately trained or are under supervision should undertake vacuum extraction delivery. In general, if the fetal head is not delivered after five attempts using the vacuum extractor, within a period of 15–20 minutes,20 the process should be considered a failure. Then, cesarean delivery should be performed, unless the failure occurs at the introitus, where forceps may be used. Interestingly, it has been shown that the risk of caput succedaneum formation is greater when the extraction procedure is greater than 10 minutes compared with when the procedure lasts for 10 minutes or less.17 It appears that two pathophysiologic mechanisms are implicated in the development of caput succedaneum-related alopecia. First, the pressure exerted on the presenting part of the fetal cranium against the dilating cervix or the cup of the vacuum extractor inflicts direct trauma and pressure ischemia to the underlying scalp, in a halo ring pattern. The severity of this lesion ranges from temporary atrophy of hair follicles to skin necrosis. Second, within the compres- Lykoudis et al Alopecia Induced by Birth Injury 489 sion ring, interstitial fluid and microhemorrhages accumulate and form the caput succedaneum, due to venous congestion. In case of excessive venous congestion, the severity of tissue damage can range from temporary atrophy or permanent scarring of hair follicles to ischemic full thickness skin necrosis (necrotic form).4,12 Treatment modalities include the initial care of caput succedaneum and the late reconstruction of any residual alopecia. The initial care depends on the severity of the trauma inflicted by the caput succedaneum. If the lesion of the skin is superficial, no more than local cleansing is needed. In case of deeper, but not full-thickness, skin lesions with microbial contamination, local application and systemic use of antibiotics is added. Finally, in case of full-thickness skin lesions, surgical debridement is needed followed by coverage with either split-thickness skin grafts or cultured autologous keratinocytes.3,12 Upon completion of acute phase treatment, the child should be examined on a regular basis for any hair regrowth, because the natural course of the alopecia depends on the depth of the lesion. Hair regrowth can be noticed in superficial skin lesions. On the contrary, deep lesions destroy hair follicles and result in permanent scarring alopecia. Also, in case of full-thickness skin necrosis treated with skin graft or keratinocytes, no hair regrowth is anticipated. Definite treatment of any permanent alopecia should be undertaken at school age to prevent psychological trauma to the child. Reconstruction can be achieved either with excision of the area of alopecia and coverage with previously expanded adjacent normal hair-bearing skin, or with hair transplantation.3,11,12 In conclusion, the contribution of the obstetrician is crucial in the prevention of complications related with caput succedaneum formation, including alopecia. Awareness of the pathophysiologic mechanisms inducing those complications dictates minimization of the duration of the second stage of labor and proper use of the vacuum extractor. The latter includes proper placement of the cup on the fetal head and minimization of repetitive cup detachments. If the formation of a complicated form of caput succedaneum (eg, contaminated, necrotic, etc) is not avoided, the role of the neonatologist and the plastic surgeon is of utmost importance, because early recognition and treatment of the lesion minimize permanent sequelae. In case of development of permanent alopecia, excel- 490 Lykoudis et al Alopecia Induced by Birth Injury lent esthetic results can be achieved with the use of reconstructive plastic surgery techniques. REFERENCES 1. Matthews M. Prenatal pressure necrosis of the scalp. Ann Plast Surg 1999;43:74–6. 2. Aron R, Gordon W. Cicatricial alopecia: a complication of vacuum extraction. S Afr Med J 1972;45:671. 3. Kulshrestha S, Kulshrestha M, Sarkar B, Chandra M, Singh P, Agarwal N, et al. Ischaemic gangrene of the scalp in the neonate: a complication of obstructed labour. BJOG 2005;112: 1334–5. 4. Neal P, Merk P, Norins A. Halo scalp ring: a form of localized scalp injury associated with caput succedaneum. Pediatr Dermatol 1984;2:52–4. 5. Vilanova X. A new type of cicatricial alopecia: alopecia caused by obstetrical vacuum extractor [in French]. Bull Soc Fr Dermatol Syphiligr 1961;68:202–3. 6. Broese I. Cicatricial alopecia and reduced pigmentation following vacuum extraction [in German]. Dermatol Monatsschr 1976;162:254–7. 7. Hall-Smith P, Foulkes JF. Traumatic cicatricial alopecia in an infant girl. Result of the use of a vacuum extractor (ventouse). Arch Dermatol 1964;89:473–4. 8. Das S. Permanent baldness following caput succedaneum. J R Coll Gen Pract 1980;30:428–9. 9. Beutner KR. Halo ring scalp associated with caput succedaneum. Pediatr Dermatol 1985;3:83. 10. Prendiville JS, Esterly NB. Halo scalp ring: a cause of scarring alopecia. Arch Dermatol 1987;123:992–3. 11. Tanzi E, Hornung RL, Silverberg NB. Halo scalp ring: a case series and review of the literature. Arch Pediatr Adolesc Med 2002;156:188–90. 12. Morykwas MJ, Beason ES, Argenta LC. Scalp necrosis in a neonate treated with cultured autologous keratinocytes. Plast Reconstr Surg 1991;87:549–52. 13. Patrizi A, Savoia F, Neri I, Bardazzi F. An incomplete circle of alopecia: a new case of halo scalp ring? Acta Derm Venereol 2006;86:65–6. 14. Byanov B. Cicatricial alopecia following vacuum extraction of the fetus. Derm Vener (Sofia) 1970;9:134. 15. Glavanakova V, Vulov V. Traumatic alopecia following vacuum extraction. Derm Vener (Sofia) 1965;4:262. 16. Dupuis O, Silveira R, Redarce T, Dittmar A, Rodrigoz RC. Instrumental extraction in 2002 in the “AURORE” hospital network: incidence and serious neonatal complications [in French]. Gynecol Obstet Fertil 2003;31:920–6. 17. Teng FY, Sayre JW. Vacuum extraction: does duration predict scalp injury? Obstet Gynecol 1997;89:281–5. 18. Bird GC. The importance of flexion in vacuum extraction delivery. Br J Obstet Gynaecol 1976;83:194–200. 19. Johanson RB, Rice C, Doyle M, Arthur J, Anyanwu L, Ibrahim J, et al. A randomised prospective study comparing the new vacuum extractor policy with forceps delivery. Br J Obstet Gynaecol 1993;100:524–30. 20. Bird GC. The use of vacuum extractor. Clin Obstet Gynaecol 1982;9:641–61. 21. Plauche WC. Fetal cranial injuries related to delivery with the Malmstrom vacuum extractor. Obstet Gynecol 1979;53:750–7. OBSTETRICS & GYNECOLOGY Nonpuerperal Uterine Inversion Associated With an Immature Teratoma of the Uterus in an Adolescent Veronica Gomez-Lobo, MD, Will Burch, MD, and Paritosh C. Khanna, MD BACKGROUND: We report a case of nonpuerperal uterine inversion associated with an immature teratoma of the uterus. CASE: An adolescent nullipara with prolonged vaginal bleeding, severe abdominal pain, symptomatic anemia, and a presumed diagnosis of retained products of conception was found to have a large mass in the vagina. Uterine inversion was diagnosed and corrected using the Haultain procedure. The inversion catalyst was found to be an immature teratoma of the uterus. admission demonstrated a 13.0⫻6.9⫻7.6 cm–sized uterus with heterogeneous material visualized within a thickened endometrial cavity measuring 2.9 cm (Fig. 1A and 1B). Based on the patient’s history and ultrasound findings, a diagnosis of retained products of conception was entertained, and the patient was brought to the operating room for an examination under anesthesia and dilation and curettage. Examination under anesthesia revealed a mass resembling myometrium and adherent necrotic tissue with small fragments of cartilage filling the vagina. The mass was noted to extend beyond a palpable dilated cervix with no fundus present and nonpuerperal uterine inversion was diagnosed. Curettage was completed and frozen section revealed myometrium with products of conception. Nitroglycerine was administered to attempt conservative repositioning of the uterus with upward pressure, using the finger CONCLUSION: Reproductive age women with the rare finding of nonpuerperal uterine inversion are likely to have a malignancy. However, uterine-sparing surgery to correct the inversion should be attempted in young women until final pathology is known. (Obstet Gynecol 2007;110:491–3) N onpuerperal uterine inversion is rare, difficult to diagnose preoperatively, and likely associated with malignancy in young women. We report a case of nonpuerperal uterine inversion in a adolescent female associated with an immature teratoma of the uterus. CASE A 15-year-old presented with abnormal uterine bleeding and pain. Her hemoglobin and hematocrit were 5.5 mg/dL and 18.8% and urine pregnancy test result was negative. The patient’s history included a spontaneous abortion 6 months before her current hospitalization, when passage of tissue with cartilage was interpreted as products of conception, although the pregnancy test was negative. No procedures where undertaken at that time, but birth control pills as well as the contraceptive patch had been prescribed to treat persistent vaginal bleeding. The ultrasonogram at From the Washington Hospital Center; and Children’s National Medical Center, Washington, DC. Corresponding author: Will Burch, MD, Washington Hospital Center, 110 Irving Street NW, Washington, DC 20010; e-mail: willburch@aol.com. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 Fig. 1. A. Longitudinal ultrasound image of the uterus demonstrates thick, heterogeneous, mixed-echogenic endometrium (black arrow) that is more prominent in the lower uterine segment and cervix. Visualized myometrium appears normal (white arrow). The upper vagina, normally seen as parallel echogenic curvilinear echoes is not seen. UB denotes urinary bladder. B. The appearance of the uterus (white arrow) and thick, heterogeneous, mixedechogenic endometrium (black arrow) on transverse images can be likened to the “target” or “doughnut” sign of intussusception. This illustrates the telescoping of the uterine fundus through its cavity with the teratoma as the lead-point, much like a similar process in bowel. Burch. Nonpuerperal Uterine Inversion. Obstet Gynecol 2007. Gomez-Lobo et al Nonpuerperal Uterine Inversion 491 tips and hand. Repositioning was unsuccessful and the case was converted to an exploratory laparotomy which revealed a viable inverted uterus, tubes, and ovaries abnormally positioned within a crater created by the inverted uterine fundus. The Huntington procedure, eversion of the uterus by placing traction bilaterally on the round ligaments was attempted and failed.1 Finally, the Haultain procedure was successfully carried out by incising through the full thickness of the posterior uterine wall at the site of intussusception then grasping the invaginated portion of the uterus with forceps and applying combined traction on the uterine wall from above and operator pressure from below. The resulting uterine defect was closed in layers using absorbable suture. The postoperative course was unremarkable, and final pathology revealed an immature uterine teratoma, high grade. The patient was referred to gynecology– oncology service for follow-up care. At the time of this report, follow-up care included chemotherapy, and she was continuing to retain her pelvic organs. COMMENT A review of OVID from 1966 to June 2006 and PubMed 1966 to June 2006 using the term “uterine inversion” revealed an article published in Gynecologic Oncology in June 2005 by Lupovitch et al2 reporting 139 cases of nonpuerperal uterine inversion from 1887 to October 2004; in addition, 10 subsequent cases where noted.3– 8 In general, nonpuerperal uterine inversion presents after age 45 years and in relation to uterine fibroids in approximately 85% of cases.2 Other causes of nonpuerperal uterine inversion include sarcomas, endometrial cancer, and polyps.2– 8 Patients tend to present with excessive vaginal bleeding, which may lead to anemia and the need for transfusion, as well as pelvic pain, pressure, and abdominal discomfort. A proposed mechanism of action is that the tumor distends the uterine cavity, thus irritating the uterine wall which in turn contracts to expel the tumor, creating an intussusception.2 Although this condition is difficult to diagnose preoperatively, a pelvic ultrasonogram may identify the condition. Unfortunately, because of the rare nature of the disorder, uterine inversion frequently goes undetected until surgery unless a high index of suspicion is maintained. In our patient, on retrospective review of the images, there was thought to be a bulge in the cervix and possibly the upper vagina on the longitudinal scan (Fig. 1A) that is consistent with an inverted uterus. The transverse image (Fig. 1B) is consistent with the “target” or “doughnut” appearance of intussuscepted tubular organs, such as bowel, wherein the outer normal myometrium can be likened to an “intussuscipiens” and the inverted segment 492 Gomez-Lobo et al Nonpuerperal Uterine Inversion of uterine fundus to an “intussusceptum.” In the uterus, however, this appearance can be simulated by a hyperplastic endometrium or endometrial carcinoma. Magnetic resonance imaging may have a role in elucidating a diagnosis of uterine inversion, although this would need to be studied further. Of the total 149 cases of nonpuerperal uterine inversion reported, only three have been reported in women aged younger than 45 years.3– 8 Lupovitch et al2 reported a case of a 26 year old with an endometrial stromal sarcoma diagnosed before surgery, found to have a nonpuerperal uterine inversion at the time of radical hysterectomy. Case et al9 reported uterine inversion in a 21-year-old diagnosed with a rhabdomyosarcoma who underwent the Haultain procedure, followed shortly thereafter by a hysterectomy secondary to rapidly progressing disease.2,8 Finally, there was a case reported in 1924 of nonpuerperal uterine inversion in a 15-year-old girl with a müllerian cyst that was treated with eversion; however, due to the age of the case, the specific pathology could not be confirmed.2 Our case is the only one in which the inversion was associated with an immature teratoma of the uterus. These cases support the notion that nonpuerperal uterine inversion in women aged younger than 45 years is associated with malignancy. In summary, uterine inversion in women aged younger than 45 years is very rare and has a significant association with malignancy. A high index of suspicion is necessary for timely diagnosis, and several fertility-sparing procedures exist. In light of the fact that many urogenital cancers in young women are amenable to treatment with chemotherapy, fertility-sparing surgery should be attempted until the final pathology is known in these patients. REFERENCES 1. Hankins GD, Clark SL, Cunningham FG, Gilstrap LC, editors. Operative obstetrics. Norwalk (CT): Appleton & Lange; 1995. p. 1238–41. 2. Lupovitch A, England ER, Chen R. Non-puerperal uterine inversion in association with uterine sarcoma: case report in a 26-yearold and review of the literature. Gynecol Oncol 2005;97:938–41 3. Oguri H, Maeda N, Yamamoto Y, Wakatsuki A, Fukaya T. Non-puerperal uterine inversion associated with endometrial carcinoma—a case report. Gynecol Oncol 2005;97:973–5. 4. Kumar G, Reynolds K. Uterine inversion associated with endometrial carcinoma: a case report. J Obstet Gynaecol 2005;25:91–2. 5. Rosales Aujang E, Gonzalez Romo R. Non-puerperal uterine inversion. Report of a case [in Spanish]. Ginecol Obstet Mex 2005;73:328–31. 6. Shivkumar PV, Barick RD, Shambharkar C. Fundal leiomyoma presenting as acute on chronic uterine inversion. J Obstet Gynaecol 2005;25:832–3. OBSTETRICS & GYNECOLOGY 7. Eftekhar Z, Rahimi-Moghaddam P, Izadi-Mood N, Yarandi F. Non-puerperal uterine inversion caused by uterine sarcoma. Aust N Z J Obstet Gynaecol 2005;45:82–3. 8. Hanprasertpong J, Wootipoom V, Hanprasertpong T. Nonpuerperal uterine inversion and uterine sarcoma (malignant mixed mullerian tumor): report of an unusual case. J Obstet Gynaecol Res 2004;30:105–8. 9. Case AS, Kirby TO, Conner MG, Huh WK. A case report of rhabdomyosarcoma of the uterus associated with uterine inversion. Gynecol Oncol 2005;96:850 –3. Discordant Middle Cerebral Artery Peak Systolic Velocity Doppler Studies in a Fetus With RhD Alloimmunization F Eliza M. F. Berkley, MD, Valerie J. Rappaport, MD, and Timothy J. Hurley, MD BACKGROUND: Doppler measurement of the fetal middle cerebral artery peak systolic velocity is a valuable tool in detecting the presence of fetal anemia in Rh-sensitized pregnancies. We present a case in which discordant left and right middle cerebral artery Dopplers complicated clinical management. CASE: An RhD-alloimmunized patient had middle cerebral artery Dopplers at 30 weeks of gestation, which showed an elevated peak systolic velocity in the left middle cerebral artery, greater than 1.55 multiples of the mean, but the right middle cerebral artery was within the normal range. The amniotic fluid change in optical density at a wavelength of 450 nm was consistent with the right middle cerebral artery Doppler. When both Dopplers were greater than or equal to 1.5 multiples of the mean, fetal blood sampling revealed a hematocrit of 28%. Postnatal cranial ultrasound examination showed normal architecture, but there was persistent discordant Dopplers in the left versus the right middle cerebral artery. CONCLUSION: Measurement of both left and right middle cerebral artery peak systolic velocities may identify patients with intrinsic variations in cranial blood vessels resulting in abnormal Doppler flows. (Obstet Gynecol 2007;110:493–5) From the Division of Maternal–Fetal Medicine, Department of Obstetrics and Gynecology, University of New Mexico School of Medicine, Albuquerque, Mexico. Corresponding author: Eliza M. F. Berkley, MD, Department of Obstetrics and Gynecology, Division of Maternal–Fetal Medicine, MSC 10 5580, 1 University of New Mexico, Albuquerque, NM 87131-5286; e-mail: eberkley@salud.unm.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 etal anemia leads to decreased blood viscosity, increased cardiac output, and subsequent increased blood flow velocity. With proper technique, color flow Doppler measurement of the middle cerebral artery peak systolic velocity has been demonstrated to correlate with the degree of fetal anemia.1 Because the middle cerebral artery peak systolic velocity accurately estimates the fetal hematocrit and it is noninvasive, it is generally accepted as the method of choice to follow and manage pregnancies complicated by red cell alloimmunization and other causes of fetal anemia.2 Peak systolic velocity of the middle cerebral artery is performed in an axial section with the sample volume being acquired most optimally 2 mm after its origin from the internal carotid artery with the angle of Theta being as close to 0° as possible. Good interobserver and intraobserver reproducibility of the middle cerebral artery peak systolic velocity measurement has been demonstrated.3 As such, waveforms of the peak systolic flow velocities in the middle cerebral artery should be similar with the highest peak velocity measured and compared with established normograms.4 Measurements of the fetal middle cerebral artery peak systolic velocity are generally taken from the artery which is nearest to the ultrasound transducer. This can be either the left or right middle cerebral artery depending on fetal position. However, measurements of the near or far artery flow rates are felt to be similar and interchangeable. Discordance between the near and far middle cerebral artery peak systolic velocities was not felt to significantly influence the middle cerebral artery peak systolic velocity values. Therefore, if the artery flow closest to the transducer cannot be obtained for technical reasons, measurement of the contralateral artery has been recommended as an alternative.5 We present a case of fetal anemia with persistent discordance of the right and left middle cerebral artery peak systolic velocity irrespective of near or far location. CASE A gravida 3 para 1011 was referred to our institution at 22 weeks of gestation for management of RhD alloimmunization. The first pregnancy was an uncomplicated term delivery. The patient’s second pregnancy resulted in a first-trimester miscarriage after which the patient did not receive Rhesus-immune globulin. In Berkley et al Discordant Middle Cerebral Artery Dopplers 493 her third pregnancy, her initial antibody screen was positive for anti-D with a titer of 1:128. The patient lived remote from a tertiary center, and Doppler evaluation was not available in her hometown. Because travel to a tertiary center for Doppler follow-up would be burdensome, her referring provider confirmed fetal D antigen status by amniocentesis before referral. The prenatal course was otherwise uncomplicated. Ultrasound examination revealed no evidence of fetal hydrops:ascites, pleural or pericardial effusions, or skin edema. The initial left and right middle cerebral artery peak systolic velocities were concordant at 36 –37 cm/s, 1.29 multiples of the mean (MOM). Per our protocol, a cutoff value of 1.5 MOM is used, above which fetal blood sampling with potential transfusion is performed. The patient was followed weekly for ultrasound assessments. In addition she underwent twice weekly antenatal testing per our protocol. Weekly ultrasound examinations revealed no evidence of fetal hydrops and middle cerebral artery peak systolic velocity Doppler values continued to measure at 1.3 MOM. These readings were taken from the artery closest to the transducer, which by chance, was primarily the left middle cerebral artery. A Doppler discrepancy was first noted at 30 weeks of gestation when the near (right) middle cerebral artery peak systolic velocity measured 47 cm/s, 1.1 MOM but the far (left) middle cerebral artery peak systolic velocity measured 65 cm/s, well above 1.55 MOM (Fig. 1XIII). The far Doppler was measured because initial positioning of the fetal head prevented optimal assessment of the near middle cerebral artery. The ultrasound findings still revealed no evidence of fetal decompensation or hydrops. Because of this discordance, the patient underwent an amniocentesis to obtain a delta optical density measurement at a wavelength of 450 nm to assess the bilirubin level in the amniotic fluid. The findings revealed a normal amniotic fluid change in optical density at 450 nm of 0.05, Liley Curve Zone I. Discordant middle cerebral artery peak systolic velocity Doppler measurements were noted throughout the remainder of pregnancy with the right being significantly less than the left, irrespective of near or far placement. Only at 33 weeks of estimated gestational age, when both Doppler indices were 1.5 or greater MOM was umbilical vein blood sampling performed. At this time, the fetus was found to have a hematocrit of 28% and an intravascular intrauterine transfusion was carried out. At 34 weeks of gestation, secondary to premature rupture of membranes, labor was induced and a newborn boy was delivered and transferred to the neonatal intensive care unit. He had a hematocrit of 30%, which required phototherapy and one exchange transfusion for an elevated bilirubin level. Postnatal cranial ultrasonography with Doppler flow studies showed a difference between the right and left middle cerebral artery resistive indices, 0.64 compared with 0.80, but otherwise normal anatomy. After phototherapy and one neonatal transfusion, the newborn was doing well and discharged home. COMMENT Use of Doppler studies to obtain middle cerebral artery peak systolic velocities in the assessment of fetuses with suspected anemia has essentially replaced the conventional management of amniocentesis for optical density at 450 nm values. Middle cerebral artery peak systolic velocity Doppler measurement can accurately predict fetal anemia and reduce the number of invasive procedures such as amniocentesis and cordocentesis.6 Correct sampling of the middle cerebral artery peak systolic velocity is essential for the proper diagnosis and management of fetal anemia. Doppler examination should be performed on the middle cerebral artery in an axial section of the brain, which includes the thalamus and the cavum septi pellucidi, during a period of fetal rest, with color Doppler imaging of the circle of Willis. Studies comparing peak systolic velocities at different locations along the course of the middle cerebral artery, as well as the peak systolic velocity of the contralateral middle cerebral artery have shown that the distance from the circle of Willis is critical to the reproducibility of the measurements, with 2 mm from the origin of the middle cerebral artery being optimal. In contrast, no Fig. 1. Trends in the left and right middle cerebral artery peak systolic velocities (MCA PSV) by gestational age. MoM, multiples of the mean. Berkley. Discordant Middle Cerebral Artery Dopplers. Obstet Gynecol 2007. 494 Berkley et al Discordant Middle Cerebral Artery Dopplers OBSTETRICS & GYNECOLOGY difference was demonstrated related to sampling the near versus the far artery.3,5 We believe that the current case illustrates a situation in which minor anatomic variation in the left and right middle cerebral artery Doppler readings resulted in discordancy of the near and far velocity measurements. Our case would suggest that the assumption regarding the equivalence of the left and right middle cerebral artery peak systolic velocities may not be true in all cases. In patients undergoing serial middle cerebral artery Doppler studies, evaluating the equivalence of the left and right middle cerebral artery may avoid seemingly erratic peak systolic velocity trend results as well as avoid unnecessary invasive testing. In the current case, by awaiting deterioration of the peak systolic velocity, in the “normal” middle cerebral artery, we were able to delay definitive fetal blood sampling until 33 weeks of gestation, without the development of fetal hydrops or life-threatening fetal anemia. REFERENCES Pelvic Varicosities and Inferior Vena Cava potential thrombotic and hemorrhagic complications, a spontaneous vaginal delivery was achieved. Tiki Bakhshi, MD, Angela M. Glaser, RDMS, and Joan M. Mastrobattista, MD BACKGROUND: Large pelvic varicosities and an absence of the inferior vena cava below the renal veins were identified in pregnancy. CASE: A young woman with a history of hemitruncus repair in infancy was noted to have large pelvic varicosities on a transvaginal ultrasonogram in early pregnancy. Magnetic resonance imaging confirmed these findings. Her protein S activity was mildly depressed, and an methylene tetrahydrofolate reductase mutation was present with normal fasting homocysteine levels. Despite concerns regarding the presence of these varicosities and From the Department of Obstetrics, Gynecology, and Reproductive Sciences, University of Texas Health Science Center at Houston, Houston, Texas. Presented as a poster at the American College of Obstetricians and Gynecologists Armed Forces District Meeting, Sonthofen, Germany, October 28 –November 1, 2006. Corresponding author: Tiki Bakhshi, MD, University of Texas Health Science Center at Houston Medical School, 6431 Fannin Street, Suite 3.604, Houston, TX 77030; e-mail: Tiki.Bakhshi@uth.tmc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 1. Mari G, Detti L, Oz U, Zimmerman R, Duerig P, Stefos T. Accurate prediction of fetal hemoglobin by Doppler ultrasonography. Obstet Gynecol 2002;99:589–93. 2. Pereira L, Jenkins TM, Berghella V. Conventional management of maternal red cell alloimmuization compared with management by Doppler assessment of middle cerebral artery peak systolic velocity. Am J Obstet Gynecol 2003;189:1002–6. 3. Mari G, Abuhamad A, Cosmi E, Segata M, Altaye M, Akiyama M. Middle cerebral artery peak systolic velocity: technique and variability. J Ultrasound Med 2005;24:425–30. 4. Mari G, Deter RL, Carpenter RL, Rahman F, Zimmerman R, Moise KJ, et al. Noninvasive diagnosis by Doppler Ultrasonography of fetal anemia due to maternal red-cell alloimmunization. N Engl J Med 2000;342:9–14. 5. Abel DE, Grambow SC, Brancazio LR, Hertzberg BR. Ultrasound assessment of the fetal middle cerebral artery peak systolic velocity: a comparison of the near-field versus far-field vessel. Am J Obstet Gynecol 2003;189:986–9. 6. Oepkes D, Seaward PG, Vanderbussche FPHA, Windrim R, Kingdom J, Beyene J, et al. Doppler Ultrasonography versus amniocentesis to predict fetal anemia. N Engl J Med 2006;355:156–64. CONCLUSION: Large pelvic varicosities may be present in women with congenital heart disease. Although these women are at risk for complications, vaginal delivery may be safely achieved. (Obstet Gynecol 2007;110:495–7) P arametrial phlebectasia is a pathologic entity in which veins become tortuous, dilated, and even thrombosed. Richet1 first described pelvic varicosities in 1857 in a report of a tuboovarian varicocele. Large studies have been reported in the literature, but most involved gynecologic patients with findings noted at the time of hysterectomy.2 Microscopic changes in vessel walls include dilatation with medial muscular hypertrophy and occasional thrombosis.2 The normal female pelvis has a vast network of venous structures that act together to provide drainage of blood into the internal iliac veins. When varicosities or thrombosis occur, venous flow becomes altered and collateral flow and alternate routes of drainage may develop. As compared with earlier authors2 who made the diagnosis at the time of surgery, Frede3 diagnosed pelvic varicosities using ultrasonography. Ultrasonography is an excellent, noninvasive tool in diagnosing pelvic varicosities. We describe a case where large pelvic varicosities were identified during a first trimester ultrasonogram in a woman with congenital heart disease. Inherent risks of thrombosis, emboli, and the Bakhshi et al Pelvic Varicosities and Inferior Vena Cava 495 potential for hemorrhage impart challenging management options. CASE A young woman with a history of hemitruncus repair in infancy was referred for an ultrasonogram to assess viability. She had experienced two prior first trimester pregnancy losses. Transvaginal ultrasonography revealed a single, viable intrauterine pregnancy at 5 5/7 weeks gestation. No uterine abnormalities were noted. A large left-sided pelvic floor varix was identified measuring 5.7⫻3.9⫻3.8 cm (Fig. 1). Using both gray-scale and power Doppler, prominent, slow swirling venous blood flow was visualized within the varicosity. Also, several large collateral vessels were identified on both sides of the uterus (Fig. 2). To determine the causes and significance of these findings, a complete history was elicited. The patient was born with Fig. 2. Two-dimensional ultrasound image of the uterus, with prominent collateral vessels (arrow). Bakhshi. Pelvic Varicosities and Inferior Vena Cava. Obstet Gynecol 2007. Fig. 1. A. Two-dimensional ultrasound image of large pelvic varicosity (arrow). B. Two-dimensional ultrasound image of large pelvic varicosity (arrow). Bakhshi. Pelvic Varicosities and Inferior Vena Cava. Obstet Gynecol 2007. 496 Bakhshi et al aortic origin of the right pulmonary artery or hemitruncus and underwent cardiac catheterization on day of life five. Subsequently, cardiac surgery was performed, and a 5.0-mm graft was inserted between the main pulmonary artery and the right pulmonary artery. A follow-up cardiac catheterization at age 1 year revealed normal pressures in the right ventricle, main pulmonary artery, and distal right pulmonary artery. However, she was found to have bilateral femoral vein occlusion and large, tortuous, anastomotic collateral venous vessels present in both groins. A third catheterization performed at age 5 years showed an absence of the inferior vena cava, with large tortuous pelvic vessels draining into the azygous system. At age 13 years, she underwent graft revision, and a 12.0-mm graft replaced the initial one. Clinically, the patient has done well since that time, denies symptoms of chest pain or shortness of breath, and leads an active life. Following initial ultrasound evaluation, magnetic resonance imaging (MRI) of the pelvis was performed to further evaluate the maternal vasculature (Fig. 3). An MRI venogram using noncontrast time-of-flight technique was performed at 12 weeks of gestation and showed a large pelvic phlebectasia measuring 5.2⫻5.0⫻3.5 cm arising from the left paravaginal venous plexus at the left posterior uterus extending to the left posterior lateral aspect of the gestational sac, demonstrating slow flow (corresponding to the large varix identified by ultrasonography), absence of the infrarenal inferior vena cava with bilateral prominent collateralization of the ovarian veins, and the paravertebral venous plexus In addition to our patient’s history of hemitruncus and prior miscarriages, her mother had also experienced recurrent miscarriages and one stillbirth. Therefore, the patient’s karyotype with fluorescent in situ hybridization of the 22nd chromosome was obtained. The karyotype was normal and fluorescent in situ hybridization analysis was negative for DiGeorge syn- Pelvic Varicosities and Inferior Vena Cava OBSTETRICS & GYNECOLOGY Fig. 3. Magnetic resonance image of large pelvic varicosity (arrow). Bakhshi. Pelvic Varicosities and Inferior Vena Cava. Obstet Gynecol 2007. drome. A fetal echocardiogram performed at 22 weeks was normal. A thrombophilia workup was sent due to concerns for venous stasis with the large pregnant uterus compressing the venous system and possible genetic predisposition. A mildly depressed protein S activity (51%) and a methylene tetrahydrofolate reductase mutation (1 copy C677T; 1 copy A1298C) were identified. The fasting homocysteine level was normal. We were concerned for the potential of both thrombosis and spontaneous rupture of the varicosities, and the patient was begun empirically on 81 mg of aspirin and additional folic acid, as well as daily prenatal vitamins. The patient underwent frequent office visits and serial ultrasound exams to assess the size of the varicosities and to follow fetal growth. The vascular surgery team evaluated her in case of catastrophic rupture during delivery and for postpartum management suggestions. The patient observed modified activities, and her antepartum course was otherwise uneventful. She presented in active labor at 39 weeks of gestation and delivered a 3,600-g neonate vaginally without complication. The postpartum course was without incident. COMMENT Large pelvic varicosities are rare in pregnancy. However, with the advances in cardiac surgery over the past 30 years, a generation of women of child bearing age having had extensive cardiac surgery is now upon us. We present a rare case of a woman with large pelvic varicosities identified in the first trimester of pregnancy. Possible causes of these varicosities stem from the patient’s genetic predisposition to thrombosis and history of cardiac catheterization before hemitruncus repair. It is conceivable that during her neonatal catheterization, both iliac veins and distal inferior vena cava sustained VOL. 110, NO. 2, PART 2, AUGUST 2007 vessel injury, and thrombus formation occurred after the catheterization and surgery. This scenario could have led to the development of large pelvic varicosities, which drain by way of the inferior and posterior cardinal system with azygous drainage to the superior vena cava. The patient’s varicosities were first identified by ultrasonography. Both ultrasonography and MRI were valuable tools to demonstrate the location of these vessels and their proximity to the uterus and surrounding structures. This case exemplifies the usefulness of using multiple imaging modalities to make a diagnosis during pregnancy. A multidisciplinary team in a large tertiary-care center managed the pregnancy. There were many theoretical concerns about possible complications of pregnancy, such as the risk of thromboemboli due to venous stasis, risk for spontaneous rupture of the varicosities with profuse hemorrhage, and the appropriate route of delivery. A careful evaluation, in consultation with a pediatric cardiologist, was done to review the patient’s medical history, operative reports, and catheterizations as a child. Vascular surgery was available for any evidence of decompensation or hemorrhage. The appropriate route of delivery was also of concern, and the risk of spontaneous vessel rupture during the second stage of labor was considered. However, cesarean delivery is certainly associated with an increased risk of thromboemboli and hemorrhage if vessels are disrupted. The lack of a series of similar patients in the literature left us to rely on expert opinion on which to base our decisions. A controlled vaginal delivery seemed the most prudent option. Although some might advocate the use of forceps to shorten the second stage and to minimize increased intra-abdominal pressure while pushing, concerns about possible vaginal lacerations with extension deterred their use. Although identification of large pelvic varicosities is possible using ultrasonography, other imaging modalities such as MRI are helpful in confirming the diagnosis. A multidisciplinary team approach and delivery in a tertiary care center are suggested in managing cases like ours. Although women with large pelvic varicosities are at risk for complications, vaginal delivery may be safely achieved. REFERENCES 1. Richet, MA. Traite’practique d’anat.medico-chir. (France): F. Chamerot Libraire Editeur; 1857. Paris 2. Hanzlik H, Leissring JC. Parametrial phlebectasia: a clinicalpathologic entity. Am J Obstet Gynecol 1976;125:431–4. 3. Frede TE. Ultrasonic visualization of varicosities in the female genital tract. J Ultrasound Med 1984;3:365–9. Bakhshi et al Pelvic Varicosities and Inferior Vena Cava 497 Presence of Talc in Pelvic Lymph Nodes of a Woman With Ovarian Cancer and Long-Term Genital Exposure to Cosmetic Talc Daniel W. Cramer, MD, ScD, William R. Welch, MD, Ross S. Berkowitz, MD, and John J. Godleski, MD BACKGROUND: Although epidemiologic studies suggest talc use may increase ovarian cancer risk, there is no proof that talc used externally reaches the pelvis. CASE: A 68-year-old woman with stage III ovarian papillary serous carcinoma revealed she had used talc daily for 30 years to powder her genital area. Examination of her pelvic lymph nodes under polarized light microscopy showed diffuse areas of birefringence compatible with talc, confirmed by scanning electron microscopy and X-ray spectroscopy. CONCLUSION: This description of talc in pelvic lymph nodes of a woman with ovarian cancer and decades of exposure to talc may prompt new studies and offer new insights into the biologic basis for the consistent, but debated, association between talc use and ovarian cancer. (Obstet Gynecol 2007;110:498–501) A n epidemiologic association between the use of cosmetic talc in genital hygiene and ovarian cancer was first described in 1982, and many subsequent studies found talc use to increase risk for ovarian cancer.1 However, the causality of the relationship has been challenged for several reasons.2 From the Obstetrics and Gynecology Epidemiology Center, Women’s and Perinatal Division, Department of Pathology, and Division of Gynecology Oncology, Department of Obstetrics, Gynecology, and Reproductive Biology, Brigham and Women’s Hospital, Harvard Medical School; and Department of Environmental Health, Harvard School of Public Health, Boston, Massachusetts. Supported by R01CA054419, Genes, Hormones & Environment in an Ovarian Cancer Model from the National Cancer Institute and 1P50CA105009, Ovarian SPORE, from the National Cancer Institute. The authors thank Ms. Rebecca Stearns for the scanning electron microscopy and energy dispersive X-ray spectroscopy studies. Corresponding author: Daniel W. Cramer, MD, ScD, Obstetrics, Gynecology and Reproductive Biology, Brigham and Women’s Hospital, 221 Longwood Ave, Boston MA 02115; e-mail: dcramer@partners.org. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 498 Cramer et al Talc in Pelvic Lymph Nodes First, the association is a relatively weak one (ie, summary relative risk of approximately 1.3). Second, no clear increase in risk with duration of use has been found in most studies. Third, the ability of talc used in the genital area to enter the pelvic cavity has not been conclusively proven. At the time of pelvic surgery for ovarian cancer, pelvic lymph nodes are commonly sampled for staging purposes, but pathologic examination of the nodes is focused on the presence or absence of metastatic disease. More careful examination of pelvic lymph nodes from women with ovarian cancer may contribute to new perspectives in the debate regarding the role of talc in the causation of ovarian cancer, as illustrated by the following case. CASE A 68-year-old, married woman presented with abdominal swelling. A computed tomographic scan revealed a 13-cm pelvic mass, and her serum CA 125 level was more than 1,000. She was referred to the Gynecologic Oncology Service at the Brigham and Women’s Hospital, where cytoreductive surgery was performed, including total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and pelvic lymph node sampling. A stage III papillary serous carcinoma with a minor clear cell component was found. Metastatic serous carcinoma was described in two of six right external iliac and obturator nodes. Postoperatively, the patient was referred for chemotherapy. She also consented to our interview about risk factors for ovarian cancer. This study is approved by the Dana Farber–Harvard Cancer Center Institutional Review Board and permits administration of general and dietary questionnaires, blood donation, and investigation of surgical specimen(s) after written informed consent. The patient’s past history included three term deliveries followed by a tubal ligation. She had not smoked, used oral contraceptives, or used postmenopausal hormone therapy other than 6 months of progesterone therapy to regulate periods around the time of menopause, which occurred at age 50. There is a family history of colon cancer in a sister and maternal grandmother. At our interview, the patient stated she had used talc daily for 30 years as a body powder on the perineum and also applied it to underwear and sanitary napkins. In searching for ideas to help clarify the association between talc use and ovarian cancer, we consulted with an expert on mesothelioma (J.G.), who pointed out that asbestos and other particulate material commonly migrates to lymph nodes.3,4 We decided that a more systematic examination of pelvic lymph nodes from ovarian cancer cases might be in order, beginning with this case. In examining the patient’s pelvic lymph nodes, no distinct particulates were seen under regular light microscopy, although a diffuse histocytic reaction was noted, even in a node without metastases (Fig. 1A). Under polarized light, diffuse OBSTETRICS & GYNECOLOGY Fig. 1. Hematoxylin and eosin–stained section of a lymph node from the case showing morphologic findings with no polarization of the microscope light and with combinations of polarized and incident light at several different levels. A. Nodal morphology is illustrated and reveals no particulates seen without polarized light, but clusters of histiocytes are more prominent than usual. B. This panel shows the same field with polarized light plus additional light to view tissue context; birefringence is noted especially in areas of histiocyte clusters. Arrows are used to call attention to the birefringent particles. C. This shows the same field without added light, revealing the particulate nature of the birefringent material. Arrows highlight the particulate. Magnification bar shows 100 ␮m and applies to all three panels. Cramer. Talc in Pelvic Lymph Nodes. Obstet Gynecol 2007. birefringence was seen corresponding to the areas of histiocyte infiltration (Fig. 1B). Figure 1C shows the same field under polarization with no added light, revealing the particulate nature of the material, compatible with talc. Three of this patient’s four nodes (not containing metastases) displayed polarizing material. Using methods described by Shelburne et al,5 we next examined lymph nodes from this patient by combined scanning electron microscopy and energy dispersive X-ray spectroscopy. Scanning electron microscopy revealed plate-like particulates in the 5–10 ␮m range within the lymph node, in which energy dispersive X-ray spectroscopy showed a magnesium and silicate signature— compatible with talc (Fig. 2A,B). Dystrophic calcium deposits were also found within her nodes, probably a consequence of nodal aging. Of nodes from the next 12 patients examined, this case was strongest for birefringence; but these nodes have not yet been subjected to scanning electron microscopy or energy dispersive X-ray spectroscopy. Figure 3 illustrates a node negative for polarization (or histiocyte reaction) from a patient with ovarian cancer who had not used talc. COMMENT Talc is a hydrous magnesium silicate chemically similar to asbestos but structurally quite different. Asbestos has a fiber-like structure and talc a plate-like one. Because of this difference, it has been argued that the relationship between asbestos and mesothelioma should not be invoked to explain how talc might cause ovarian cancer. However, one feature of expo- Fig. 2. Analytical microscopy. A. Scanning electron microscopy of a histologic section of the lymph node from the case shows a large collection of plate-like particulates in the 5–10 ␮m range (arrows) as well as scattered individual particulates. Magnification bar shows 100 ␮m. B. X-ray spectrum taken from the central bright area with particles reveals a Magnesium (Mg), Silicon (Si), and Oxygen (O) signature compatible with talc. A Carbon (C) signal is coming from the tissue or the underlying Carbon plancette or both. Cramer. Talc in Pelvic Lymph Nodes. Obstet Gynecol 2007. VOL. 110, NO. 2, PART 2, AUGUST 2007 Cramer et al Talc in Pelvic Lymph Nodes 499 Fig. 3. Comparative node section illustrated from a woman reporting no talc use. A. Hematoxylin and eosin stained section showing fewer macrophages than seen in the node in Figure 1A. Magnification bar shows 100 ␮m. B. Polarized light examination of the same area of the node showing only some birefringence in the node capsule (arrows) compatible with collagen. Cramer. Talc in Pelvic Lymph Nodes. Obstet Gynecol 2007. sure that the minerals do have in common is nodal dissemination. Migration and entrapment in lymph nodes is observed in human asbestos exposure and correlates with the asbestos burden.3 Talc has also been described in pulmonary lymph nodes of talc miners.4 However, a MEDLINE search of (all language) publications between January 1950 and February 2007 using the search terms, “talc,” “birefringence,” “histiocytosis,” “lymph nodes,” and “ovarian neoplasms,” revealed no reports of talc in lymph nodes of ovarian cancer patients. In one of the few studies in women to evaluate the potential for talc to migrate into the pelvis, Heller et al studied normal ovaries from women having oophorectomy for benign disease.6 The protocol involved a multistep process of tissue rehydration, blotting, drying, digestion, rehydration, centrifugation, and multiple washes. After this process, polarizing bodies were found in all ovarian specimens examined by light microscopy. By electron microscopy, tissues from 5 of 12 women who regularly used talc and 6 of 12 who had not were found to have particles consistent with talc. The investigators concluded that talc can be found in ovaries but that this does not correlate with genital talc use. Contamination that might have been introduced during extensive processing is a potential weakness of this study. In this case report, we describe examination of pelvic lymph nodes from a woman with ovarian cancer who had been a long-term talc user. Particles compatible with talc were clearly visible under polar- 500 Cramer et al Talc in Pelvic Lymph Nodes ized light in regular hematoxylin and eosin–stained sections from her pelvic nodes, which were then shown by scanning electron microscopy and energy dispersive X-ray spectroscopy to be talc. Thus, as opposed to the aforementioned study, we focused on pelvic lymph nodes rather than ovaries; and talc was shown to be present in macrophages within the actual tissue, ruling out contamination during processing. In reporting this case, we are not proposing that pelvic lymph nodes from women with ovarian cancer must now be subjected to electron microscopy. However, pathologists may wish to examine pelvic lymph nodes with evidence of histiocytic infiltrates by polarized light microscopy. Clear evidence of polarization may be reported so that clinicians can obtain information about potential talc exposure, if this information has not already been collected. Also we are not claiming that a causal relationship between ovarian cancer and talc use is proven for this case or in general. Because case reports cannot establish causality, we have begun a more extensive study of nodes with two purposes. First it is necessary to establish in a quantitative manner the likelihood of finding talc in lymph nodes of women with ovarian cancer and correlate this by whether they did or did not use talc. Second, studies of immune markers in nodes may help make the case for a causal connection. What we do hope this case report accomplishes is to infuse a fresh perspective on the talc and ovarian cancer association. Previous biologic arguments linking talc and ovarian cancer have been based upon: similarities between talc and asbestos, the ability of talc to reach the ovaries through the open female tract, and induction of a mesothelioma-like cancer from the ovarian epithelium. Our new perspective would not depend upon structural similarities between talc and asbestos. The adverse effects of talc may relate to its ability to induce an inflammatory reaction, a well-established property of talc, independent of any similarity to asbestos.7 Also, we don’t believe that talc needs to reach the ovaries to affect ovarian cancer risk; rather, the harmful effects of talc may involve inflammatory reactions in the lower genital tract, including the upper vagina, cervix, and endometrium. These tissues express the surface glycoprotein human mucin 1, MUC1, whose function is to protect cells from environmental stressors. It is likely that chronic talc exposure is one factor that upregulates MUC1 expression. Human mucin 1 is related to CA 125 (MUC16), and like CA 125 is overexpressed in ovarian cancer. It is known that women with ovarian cancer who have anti-MUC1 antibodies survive longer, leading us to propose that OBSTETRICS & GYNECOLOGY many risk factors for ovarian cancer may be explained by their ability to raise or lower MUC1 immunity.8 Looking at predictors of anti-MUC1 antibodies, talc use was a factor that lowered anti-MUC1 antibodies. Thus, rather than a direct carcinogenic effect on ovarian epithelium, immune dysregulation involving MUC1 may be induced by chronic talc use that may lower protective immunity. Furthermore, sequestration of talc in nodes may affect antigen processing and be another important element in the postulated immune dysregulation. In conclusion, this description of talc in pelvic lymph nodes of a long-term talc user with ovarian cancer may begin to reshape understanding about the relationship between talc and ovarian cancer and shed new light on whether talc used externally in the genital area is capable of migrating into the pelvis. REFERENCES 1. Cramer DW, Liberman RF, Titus-Ernstoff L, Welch WR, Greenberg ER, Baron JA, et al. Genital talc exposure and risk of ovarian cancer. Int J Cancer 1999;81:351–6. Postpartum Sudden Death From Pulmonary Hypertension in the Setting of Portal Hypertension 3. Friedrichs KH. Electron microscopic analyses of dust from the lungs and the lymph nodes of talc-mine employees. Am Ind Hyg Assoc J 1987;48:626–33. 4. Roggli VL, Benning TL. Asbestos bodies in pulmonary hilar lymph nodes. Mod Pathol 1990;3:513–7. 5. Shelburne JD, Estrada H, Hale M, Ingram P, Tucker JA. Correlative microscopy and microprobe analysis in pathology. In: Bailey GW, editor. Proceedings of the 47th annual meeting of the Microscopy Society of America. Vol 900. San Francisco (CA): San Francisco Press; 1989. 6. Heller DS, Westhoff C, Gordon RE, Katz N. The relationship between perineal cosmetic talc usage and ovarian talc particle burden. Am J Obstet Gynecol 1996;174:1507–10. 7. van den Heuvel MM, Smit HJ, Barbierato SB, Havenith CE, Beelen RH, Postmus PE. Talc-induced inflammation in the pleural cavity. Eur Respir J 1998;12:1419–23. 8. Cramer DW, Titus-Ernstoff L, McKolanis JR, Welch WR, Vitonis AF, Berkowitz RS, et al. Conditions associated with antibodies against the tumor-associated antigen MUC1 and their relationship to risk for ovarian cancer. Cancer Epidemiol Biomarkers Prev 2005;14:1125–31. initially responsive, within minutes she was unresponsive and resuscitation was unsuccessful. Postmortem examination showed cirrhosis and plexogenic pulmonary arteriopathy. Carlie S. Sigel, MD, Teresa C. Harper, MD, and Leigh B. Thorne, MD BACKGROUND: Pulmonary arterial hypertension carries a high maternal mortality rate in the peripartum period. Pulmonary hypertension may arise as a complication of portal hypertension with poor patient survival. CASE: A young primigravida with chronic autoimmune hepatitis and portal hypertension presented at 26 4/7 weeks of gestation with contractions and bleeding. Within 48 hours, an 892-g female fetus was delivered vaginally without complications. On postpartum day 2, the mother was found on the floor by her bed. Although From the Department of Pathology, University of North Carolina-Chapel Hill, Chapel Hill, North Carolina; and Perinatal Associates of New Mexico, Albuquerque, New Mexico. Corresponding author: Leigh B. Thorne, MD, Department of Pathology, University of North Carolina, 101 Manning Drive, CB#7525, Chapel Hill, NC 27599-7525; e-mail: lthorne@unch.unc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 2. Wehner AP. Cosmetic talc should not be listed as a carcinogen: comments on NTP’s deliberations to list talc as a carcinogen. Regul Toxicol Pharmacol 2002;36:40–50. CONCLUSION: Increased awareness of pulmonary hypertension as a complication of portal hypertension and a high index of clinical suspicion are necessary to diagnose pregnant women with this condition and provide appropriate prenatal counseling and peripartum intervention. (Obstet Gynecol 2007;110:501–3) P ulmonary hypertension is an under-recognized complication of portal hypertension. We present an individual with known autoimmune hepatitis with cirrhosis and portal hypertension where underlying pulmonary hypertension was identified after her postpartum sudden death. Pulmonary hypertension may present in a subtle manner, but is important to appreciate in this high-risk obstetric patient population. CASE A young primigravida with a 10-year history of autoimmune hepatitis with chronic thrombocytopenia presented to the hospital at 26 4/7 weeks of gestation with contractions and bleeding. Before her pregnancy, she was a noncompliant transplantation candidate not using birth control. Prenatal care had been initiated at 6 weeks of Sigel et al Portopulmonary Hypertension and Pregnancy 501 gestation in a maternal–fetal medicine specialty clinic, and her transplant team had been consulted. Upon her presentation, the prenatal chart was obtained. Her clinical diagnoses included autoimmune hepatitis with end-stage cirrhosis, history of cholangitis, recurrent pancreatitis, portal hypertension, and thrombocytopenia. She was afebrile but tachycardic (pulse 118 beats per minute [bpm]) with a blood pressure of 121/90 mm Hg. She was thrombocytopenic (14⫻109/L platelets) with elevated liver enzymes (total bilirubin 4.1 mg/dL [0.0 –1.2], aspartate aminotransferase 61 units/L [14 –38]), and alkaline phosphatase 129 units/L [38 –126]). Physical examination revealed a gravid abdomen, serosanguinous vaginal discharge, and intact prolapsing membranes. Fetal ultrasound showed a viable fetus in vertex presentation. She was treated with magnesium sulfate, betamethasone, penicillin, and azithromyocin. Forty-eight hours later, an 892-g female infant was delivered vaginally without complications. The next day, the patient showed persistent vaginal bleeding and passed some retained products of conception. Early on postpartum day two, after a call for maternal assistance, the patient was found on the floor with a small forehead laceration. She was responsive and tachycardic (110 –120 bpm), and a fingerstick glucose and basic chemistries were within normal range. Hemoglobin, hematocrit, and platelets were 12.4 g/dL (12.0 –16.0), 37.5% (36.0 – 46.0), and 15⫻109/L (150 – 440), respectively. Within 10 minutes she postured with right-sided eye deviation and became unresponsive. Autopsy findings included a 1.2-cm laceration above the left eyebrow. Reflection of the scalp showed no underlying skull fracture. There was mild-to-moderate right ventricular dilatation and a mild pericardial effusion (100 mL). The lungs had no pulmonary embolus, hemorrhage, or consolidation. Abdominal examination revealed a cirrhotic liver, splenomegaly, and an intact biliary stent placed 5 years prior for recurrent cholecystitis in the setting of primary sclerosing cholangitis. Neuropathologic examination showed no evidence of an intracranial bleed, tumor, or herniation. Microscopic examination of the lungs showed a plexogenic pulmonary arteriopathy characterized by proliferative arterial lesions throughout all lobes, variably present small thrombi within the lesions, and rare foci of acute vasculitis indicating pulmonary hypertension.1 The liver showed cirrhosis with minimal interface hepatitis. The kidneys showed patent glomerular capillary loops. Evaluation of the systemic microvasculature, specifically the renal and superficial maternal spiral arteries, showed no histologic evidence of preeclampsia, underlying connective tissue disorder, or thrombotic microangiopathic changes. At delivery, the placenta had not been retained for evaluation. Changes suggesting early ischemic neuronal injury were seen in the brain. The postmortem review of the patient’s medical history including records obtained from the outside hospital described her 6-year history of portal hypertension secondary to end-stage liver disease which was due to autoimmune hepatitis and primary sclerosing cholangitis overlap syn- 502 Sigel et al drome. Her medication compliance was poor. Six months before pregnancy, she was treated for pancreatitis. At that time, an echocardiogram showed an overall ejection fraction of 55%, cardiac chambers within normal limits, trace mitral regurgitation, mild tricuspid regurgitation, and moderate pulmonary hypertension (estimated 55 mm Hg). During the 6 months preceding pregnancy, she was evaluated at the same regional hospital on five occasions for complaints of chest pain, shortness of breath, and intermittent confusion. Two ventilation-perfusion scans were negative for thromboembolic pulmonary disease. A pulmonologist diagnosed her with early asthma. Her pregnancy was found after she presented to our emergency department with altered mental status and syncope. She was counseled regarding the risks of complications in pregnancy in patients with chronic liver disease but decided to continue the pregnancy. Throughout gestation, her only complication was chronic thrombocytopenia. COMMENT Pregnant women with pulmonary hypertension, regardless of etiology, have a high risk of maternal mortality (30 –56%), particularly in the postpartum period.2 Pulmonary hypertension is an uncommon complication of portal hypertension occurring in 2–12.5% of patients undergoing liver transplantation.4 – 6 On average, it is seen 4 and 7 years after the diagnosis of portal hypertension.4 Identifying portopulmonary hypertension (pulmonary hypertension associated with portal hypertension) in gravid individuals with cirrhosis is difficult because the symptoms may be nonspecific and attributed to either the pregnancy or the liver disease. Dyspnea on exertion is the most common presenting symptom, but others include fatigue, syncope, angina, peripheral edema, and abdominal distension.6 Physical examination findings may include an accentuated pulmonary component of the second heart sound or a systolic murmur of tricuspid regurgitation.6 In severe disease, there may be signs of right heart failure on chest X-ray, including increased main pulmonary artery size or right atrial and ventricular dilatation. The most common electrocardiographic abnormalities are right atrial enlargement, right ventricular hypertrophy, or right axis deviation; however, electrocardiography is not universally considered an adequate screening tool.7 Formerly classified as a form of secondary pulmonary hypertension, portopulmonary hypertension has been reclassified as pulmonary arterial hypertension in association with hepatic disease or portal hypertension.4 Portopulmonary hypertension is defined as an elevated pulmonary arterial pressure with Portopulmonary Hypertension and Pregnancy OBSTETRICS & GYNECOLOGY increased vascular resistance and a normal pulmonary capillary wedge pressure and portal hypertension.8 Other causes of pulmonary hypertension such as congenital heart disease, human immunodeficiency virus (HIV), collagen vascular disease, or drugs such as anorectic agents should be excluded. A recent European Respiratory Journal Task Force recommended that patients with portal hypertension who report dyspnea at rest or during exercise should be assessed for pulmonary hypertension by Doppler echocardiography.8 Routine screening is not recommended.8 Echocardiography is used to diagnose pulmonary hypertension by measuring the velocity of the tricuspid regurgitant jet, estimating the pulmonary artery systolic pressure. A pulmonary arterial systolic pressure higher than 25 mm Hg at rest or higher than 30 mm Hg during exercise indicates pulmonary hypertension.7 Transthoracic echocardiography correlates strongly with right-heart catheterization findings,7 but right-heart catheterization remains the gold standard for confirming the diagnosis and assessing severity.4,8 In pregnancy, echocardiography almost always overestimates pulmonary arterial systolic pressures.2 Ideally, preconceptional evaluation would be recommended for initial diagnosis. Recently, maternal mortality due to pulmonary hypertension was reexamined to assess the efficacy of a multidisciplinary approach to treatment.3 An overall 36% death rate was found, concordant with rates reported previously.3 Pulmonary hypertension creates a state of fixed pulmonary vascular resistance which poorly accommodates the major hemodynamic shifts occurring postpartum.2 Consequently, close attention to patients during the 72 hours after delivery is crucial as these individuals are at risk for refractory right heart failure and sudden death.2 A multidisciplinary team approach to care, early diagnosis and hospital admission, close monitoring at delivery with electrocardiography, pulse oximetry, and invasive arterial blood pressure monitoring are recommended.2,3 A MEDLINE search in any language from January 2001 to December 2006 using the search terms “pregnancy,” “pulmonary hypertension,” and “guidelines” produced no results for guidelines regarding effective pharmaceutical management, mode of delivery, or anesthesia for pregnant patients with pulmonary hypertension. The strategy for caring for these individuals is guided by the VOL. 110, NO. 2, PART 2, AUGUST 2007 successful outcomes others have shared. Epoprostenol, a synthetic prostacyclin, is a potent vasodilator that has shown improvement in both pulmonary hypertension and portopulmonary hypertension and, thus, has been introduced with some success in pregnancy management.2– 4 Inhaled nitrous oxide, calcium channel blockers, and oral anticoagulants are also used.2,3 Portopulmonary hypertension is an uncommon, but important, entity with a subtle presentation and high patient mortality. Without therapy, mean and median survivals have been shown to be 15 and 6 months, respectively.5 Despite up-to-date approaches to management, pregnancy raises mortality substantially. In general, pregnancy is not advised and termination should be offered especially when the disease appears to be progressive.3 Counseling reproductiveaged women with pulmonary hypertension on the risks of pregnancy with emphasis on consistent contraception is of paramount importance for preventing death. A multidisciplinary approach with close collaboration between transplant specialists and high-risk obstetricians is necessary to minimize fetal and maternal morbidity and mortality. REFERENCES 1. Pietra GG, Capron F, Stewart S, Leone O, Humbert M, Robbins IM, et al. Pathologic assessment of vasculopathies in pulmonary hypertension. J Am Coll Cardiol 2004;43:25S–32S. 2. Budev MM, Arroliga AC, Emery S. Exacerbation of underlying pulmonary disease in pregnancy. Crit Care Med 2005;33: S313–8. 3. Bonnin M, Mercier FJ, Sitbon O, Roger-Christoph S, Jais X, Humbert M, et al. Severe pulmonary hypertension during pregnancy: mode of delivery and anesthetic management of 15 consecutive cases. Anesthesiology 2005;102:1133–7. 4. Halank M, Ewert R, Seyfarth H-J, Hoeffken G. Portopulmonary hypertension. J Gastroenterol 2006;41:837–47. 5. Robalino BD, Moodie DS. Association between primary pulmonary hypertension and portal hypertension: analysis of its pathophysiology and clinical, laboratory and hemodynamic manifestations. J Am Coll Cardiol 1991;17:492–8. 6. Ramsay MA, Simpson BR, Nguyen AT, Ramsay KJ, East C, Klintmalm GB. Severe pulmonary hypertension in liver transplant candidates. Liver Transpl Surg 1997;3:494–500. 7. Barst RJ, McGoon M, Torbicki A, Sitbon O, Krowka MJ, Olschewski H, et al. Diagnosis and differential assessment of pulmonary arterial hypertension. J Am Coll Cardiol 2004;43: 40S–7S. 8. Rodriguez-Roisin R, Krowka MJ, Herve P, Fallon MB. Pulmonary-hepatic vascular disorders (PHD). Eur Respir J 2004;24: 861–80. Sigel et al Portopulmonary Hypertension and Pregnancy 503 Pregnancy Outcome After Transcervical Hysteroscopic Sterilization Marie Hastings-Tolsma, PhD, CNM, Priscilla Nodine, CNM, PhDc, Stephanie B. Teal, MD, MPH, and Julia Embry, MD BACKGROUND: Hysteroscopic, transcervical sterilization involves placement of microinserts in tubal ostia. As with any contraceptive method, pregnancy can occur. This case reports the outcome when pregnancy occurred after microinsert placement. CASE: A multiparous woman presented at 16 weeks of gestation. Hysteroscopic sterilization was performed 2 years earlier, although a postprocedure hysterosalpingogram was not done to verify tubal occlusion. The patient had a normal-term pregnancy. Postpregnancy hysterosalpingogram revealed both microinserts were embedded in the uterine fundus and myometrium. CONCLUSION: This case demonstrates how pregnancy can occur after hysteroscopic microinsert placement and details how it might be avoided. (Obstet Gynecol 2007;110:504–6) E ssure hysteroscopic, transcervical sterilization (Conceptus, Mountain View, CA) was approved by the Food and Drug Administration in 2002. An incisionless, hormone-free method, Essure placement allows quick patient recovery with few, if any, side effects.1 Because of the recent introduction of the method and a very low failure rate, little is known about the effect of the microinserts on gestation. As with any contraceptive method, pregnancy can From the University of Colorado at Denver & Health Sciences Center, Denver, Colorado. The authors thank Brian Yuhnke Jr, Instructional Technology and Web Media Production Coordinator at the University of Colorado at Denver & Health Sciences Center for his assistance with graphics. Corresponding author: Marie Hastings-Tolsma, PhD, CNM, Associate Professor, Nurse Midwifery, University of Colorado at Denver & Health Sciences Center, 4200 East Ninth Avenue #C-288, Denver, CO 80262; e-mail: marie.hastingstolsma@uchsc.edu. Financial Disclosure The manufacturer of Essure (Conceptus, Inc.) paid for the patient to have a hysterosalpingogram. The University of Colorado Hospital was paid for the procedure costs, but the researchers were not paid. The authors have no financial affiliation with Conceptus, Inc. Conceptus, Inc., provided financial support for Drs. Hastings-Tolsma and Teal to be presenters at the 52nd Annual Meeting of the American College of Nurse Midwives, May 29, 2007. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 504 Hastings-Tolsma et al occur. Between 1997 and 2005, there were 64 pregnancies of an estimated 50,000 placements that were reported to the manufacturer (M. Childers, personal communication, February 7, 2007). Of these known pregnancies, approximately 33% planned or reported termination; over one half (53%) were either undecided, planned to go to term, or plans were unknown. Five pregnancies (8%) resulted in a live, uncomplicated birth Physicians have few data upon which to base counseling recommendations for women who become pregnant with the devices in place. A PubMed review of the literature from 1995 to 2006 using search terms Essure microinsert, pregnancy, birth control, genitourinary tract imaging, and hysterosalpingogram, revealed no reports of pregnancy outcomes after the Essure procedure. This case report presents how pregnancy can occur after device placement, evidence of normal pregnancy outcome, and strategies to avoid it. CASE A 31-year-old multiparous woman presented to the obstetric clinic at 16 weeks of gestation. She had Essure sterilization performed 2 years earlier and 8 weeks after a spontaneous birth of twins. The devices were placed after intravenous sedation. The operative report indicated “a thickened endometrial lining with possible retained placental products and difficulty locating the tubal ostia.” The patient stated she had been told that the procedure took longer than usual because her “tubes were higher up than normal.” The physician who inserted the devices noted that 3 coils extended into the uterine cavity from the left ostium; 8 coils were noted from the right ostium. These coil extensions are within the appropriate range for successful placement.2 The patient denied problems after placement of the devices. Not realizing that it was “that big of a deal”, the patient did not schedule the requisite hysterosalpingogram at 3 months after the procedure to verify tubal occlusion. However, at about that same time, the physician did perform office ultrasonography, and believed the devices were in place bilaterally. Two years later, ultrasonography was again done after her new obstetric examination at 16 weeks gestation, and the devices appeared to be in place bilaterally (Fig. 1). The patient was counseled regarding the lack of data regarding the effect, if any, of the devices on the pregnancy. She opted to continue the pregnancy and was scheduled for routine care. An additional ultrasonogram at 20 weeks gestation showed normal fetal growth and development; Essure placement was not evaluated. Results of prenatal laboratory tests were normal. Except for obesity, past medical and surgical history was negative. The obstetric history revealed two prior term uncomplicated vaginal births. The patient did not breastfeed. The gynecologic history was unremarkable, with no infertility issues. She had no history of abnormal Pap tests, and other than treatment for trichomonas vaginalis in early Pregnancy Hysteroscopic Sterilization OBSTETRICS & GYNECOLOGY Fig. 1. Ultrasonogram of right microinsert at approximately 20 weeks of gestation. The microinsert appears as a white linear echogenic line (arrow) that corresponds to the outer coil of the device. Fig. 2. Location of Essure microinsert devices on hysterosalpingogram. Smaller arrows demonstrate incorrect placement in the fundal myometrium. The longer arrow shows free spillage of the contrast agent from the right fimbria. Hastings-Tolsma. Pregnancy Hysteroscopic Sterilization. Obstet Gynecol 2007. Hastings-Tolsma. Pregnancy Hysteroscopic Sterilization. Obstet Gynecol 2007. pregnancy, there was no other sexually transmitted infection history. Menses were monthly with normal duration and flow. There was a history of moderate to severe dysmenorrhea which improved after Essure placement. After placement of the devices, the patient reported regular coitus with only occasional use of condoms. With the current pregnancy, there was spontaneous onset of labor at 39 5/7 weeks of gestation. After an 8-hour first stage labor, the patient had an uncomplicated spontaneous vaginal birth of a viable infant with Apgar scores of 8 and 9. Estimated blood loss was 100 mL. The patient requested postpartum tubal ligation rather than repeat Essure sterilization or another temporary contraceptive method. On postpartum day 1, a plain film abdominal X-ray was obtained. The devices appeared to be correctly placed, but tubal occlusion could not be verified. Modified Pomeroy tubal ligation was then performed with ligature well distal to the presumed device locations. Histology confirmed complete cross-section of both fallopian tubes. A hysterosalpingogram at 5 months postpartum to evaluate tubal occlusion from the Essure microinserts found both devices visible on the radiographic films (Fig. 2). However, neither of the coils were in the ostia or tubal lumina; rather, they were observed to be superior to the ostia, apparently embedded in the myometrium. In addition, free spillage of fluoroscopic dye was seen from the distal portion of the right tube beyond the point of tubal ligation. This latter finding made sterility questionable and the patient was advised to use a temporary method until a repeat right tubal or Essure device placement could be effected. She requested hormonal contraception, refusing repeat sterilization. VOL. 110, NO. 2, PART 2, AUGUST 2007 COMMENT The manufacturer identifies pregnancy as a known risk, and ectopic pregnancy as a theoretical risk.2 One reference to a pregnancy during clinical trials was found,1 but no data could be found regarding outcome. Although the cumulative 5-year failure rate for Essure is low (2.6 of 1,000 procedures), risk could be further reduced by following placement protocols and recommended postprocedure follow-up (Levey B, Levie MD, Childers MA. A summary of reported pregnancies following hysteroscopic sterilization. J Minim Invasive Gynecol. In press 2007). Levey, Levie, and Childers reviewed the causes for reported pregnancies after Essure placement and found patient or physician noncompliance as the cause nearly one half of the time (46%).Other reasons included misreading the X-ray or hysterosalpingogram (25%), pregnant at the time of placement (12%), and other (15%), such as patient failure to use alternate contraceptive after placement. Where noncompliance was the primary factor, patient failure to return for follow-up and incorrect physician advice (eg, instructed patient there was no need to return for follow-up) were the most common reasons. Although it is possible for the coil devices to migrate from the proximal tubes, it is likely that for this patient, the devices were never correctly placed Hastings-Tolsma et al Pregnancy Hysteroscopic Sterilization 505 in the ostia. Careful training and adherence to protocols in how to place the microinserts is crucial, and increased experience likely improves placement rates.3 It is also important to place the devices when there is good visibility within the endometrial cavity; that is, in the early proliferative phase of the menstrual cycle or when the patient is using hormonal contraception to effect a thin endometrium. This patient was 8 weeks postpartum and without hormonal contraception before placement; this likely made visualization of ostia difficult. It is important to stress with patients the need for careful timing of the procedure when hormonal contraception has not been used. Similarly, reliable contraception until hysterosalpingogram follow-up is crucial. Compliance with hysterosalpingogram follow-up may be improved through telephone or mail reminders, as well as providing printed materials. Success rates for device placement on first attempt are reported between 85 and 92%.3,4 Placement difficulty due to tubal stenosis or obstruction or difficulty accessing the proximal tubal lumen or visualizing the ostia should prompt consideration of procedure abandonment. Where insertion is difficult, a high index of suspicion for placement failure should exist. Care should be taken to confirm device location and tubal occlusion. For this patient, imaging studies were done to determine correct device location and verification of ostia occlusion. Office ultrasonography at 3 months after placement and ultrasonography and plain abdominal X-ray during and immediately after pregnancy suggested that the devices were in place. Contrary to literature reports,5–7 ultrasonography and X-ray were not adequate measures of appropriate device placement and the extent of tubal occlusion in this patient. Adequate imaging requires views from at least two angles to confirm the device position, typically anterior-posterior and lateral views, as well as visualization of an echogenic structure spanning the myometrial layer of the uterotubal junction.5 Imaging to confirm appropriate placement likely requires additional training, as well as a willingness to obtain further imagining where uncertainty exists. Pregnancy occurred over two years after Essure placement. Aside from imaging studies, no additional interventions were initiated. Although an uncomplicated pregnancy outcome was achieved, it is unknown whether there are risks for adverse outcome, and insufficient data exist for counseling patients. There have been reports, albeit few, of failed tubal occlusion after correct placement of the Essure devices.8 It is possible that a woman with a history of 506 Hastings-Tolsma et al failed hysteroscopic sterilization will desire sterilization by tubal ligation. Per manufacturer information,2 there is risk of electrical current near the microinserts during tubal surgery. A major consideration for this patient was how the postpartum tubal ligation procedure would be carried out. The microinserts are made of titanium, stainless steel, and nickel, containing Dacron (E.I. du Pont de Nemours and Company, Wilmington, DE) fibers.2 The microinserts will conduct energy, and patient injury could occur during tubal ligation. Electrosurgery should be avoided in procedures conducted on the uterine cornua and proximal fallopian tubes with Essure coils present. Hysteroscopic sterilization offers an advantage because it may be performed in the outpatient setting with local anesthesia.9 With easy, convenient access to the devices, strategies to ensure correct placement and tubal occlusion are imperative. Should pregnancy occur, increased vigilance may not be necessary, although careful tracking of perinatal outcomes are needed to determine risk. REFERENCES 1. Duffy S, Marsh F, Rogerson L, Hudson H, Cooper K, Jack S, et al. Female sterilisation: a cohort controlled comparative study of ESSURE versus laparoscopic sterilisation. BJOG 2005;112:1522–8. 2. Conceptus Inc. Essure Permanent Birth Control. Prescribing information. Available at: http://www.essuremd.com/Portals/0/ Skins/Conceptus_Skin/PDFs/CC-0366-prescribinginfo.pdf#search⫽%22essure%20prescribing%20information%22. Retrieved October 31, 2006. 3. Kerin JF, Carignan CS, Cher D. The safety and effectiveness of a new hysteroscopic method for permanent birth control: results of the first Essure pbc clinical study. Aust N Z J Obstet Gynaecol 2001;41:364–70. 4. Cooper JM, Carignan CS, Cher D, Kerin JF. Selective Tubal Occlusion Procedure 2000 Investigators Group. Microinsert nonincisional hysteroscopic sterilization. Obstet Gynecol 2003;102:59–67. 5. Wittmer MH, Brown DL, Hartman RP, Famuyide AO, Kawashima A, King BF. Sonography, CT, and MRI appearance of the Essure microinsert permanent birth control device. AJR Am J Roentgenol 2006;187:959–64. 6. Thiel JA, Suchet IB, Lortie K. Confirmation of Essure microinsert tubal coil placement with conventional and volumecontrast imaging three-dimensional ultrasound. Fertil Steril 2005;84:504–8. 7. Veersema S, Vleugels MP, Timmermans A, Brolmann HA. Follow-up of successful bilateral placement of Essure microinserts with ultrasound. Fertil Steril 2005;84:1733–6. 8. Karthigasu KA, Garry R, Hart R. Case report of failed tubal occlusion using Essure pbc (permanent birth control) hysteroscopic sterilisation procedure. Aust N Z J Obstet Gynaecol 2006;46:365–7. 9. Levie MD, Chudnoff SG. Prospective analysis of office-based hysteroscopic sterilization. J Minim Invasive Gynecol 2006;13:98–101. Pregnancy Hysteroscopic Sterilization OBSTETRICS & GYNECOLOGY Trastuzumab Use for Metastatic Breast Cancer in Pregnancy Renuka Sekar, FRANZCOG, and Peter R. Stone, FRANZCOG, CMFM BACKGROUND: Trastuzumab is approved for first-line treatment for breast cancer in combination with docetaxel for stage 2 tumors positive for human epidermal growth factor receptor 2. The effects of trastuzumab on the fetus are mostly unknown. CASE: Our case report focuses on a woman who was treated for invasive ductal carcinoma 1 year before pregnancy. She presented at 20 weeks of gestation with metastases and was treated with docetaxel and trastuzumab. She underwent two cycles of chemotherapy, and an ultrasound scan at 30 weeks showed anhydramnios. There was no history of ruptured membranes. Reappearance of amniotic fluid was noted at 33 weeks of gestation, 7 weeks after cessation of treatment. CONCLUSION: Treatment with trastuzumab during midgestation may be associated with anhydramnios. (Obstet Gynecol 2007;110:507–10) B reast cancer is becoming more common in women who desire pregnancy, partly due to the delay in childbearing. Social, personal, educational, and lifestyle factors have contributed to this, resulting in the postponement of first pregnancies to the third or even fourth decade. Seven percent to 14% of breast cancers are diagnosed in women less than 40 years of age.1 Gestational breast cancer includes 7–14% of breast cancers in this age group. The Western Australian population-based study showed gestational breast cancer to account for 6.25% of breast cancers in Western Australian women less than 45 years of age and complicated 23.6 per 100,000 pregnancies.2 Breast cancer tumors in women under the age of 40 tend to be large, with a median tumor size of 2 cm in young women compared with 1.5 cm in older women. They are likely to be both lymph node– positive and aggressive, while also being less differentiated than in older women.3 This results in younger women having both a higher mortality and a shorter disease-free survival after treatment. Trastuzumab (Herceptin, F. Hoffmann–La Roche, Basel, Switzerland) is approved for first-line treatment of breast cancer in combination with docetaxel for stage 2 tumors that are positive for human epidermal growth factor receptor 2. The effects of trastuzumab on the fetus are mostly unknown. Trastuzumab is a recombinant humanized monoclonal antibody (immunoglobulin G [IgG]) that blocks the human epidermal growth factor receptor 2 protein.4 Human epidermal growth factor receptor 2 is a member of a family of four transmembrane receptor tyrosine kinases that regulate cell growth, survival, and differentiation via multiple signal transduction pathways. The mean half-life for a weekly maintenance dose is 5.8 days (range 1–32 days). This may define a timeline for recovery of anhydramnios after trastuzumab treatment. There appear to have been two case reports of trastuzumab use in pregnancies of less than 20 weeks of gestation5 and one where the drug was given at 27 weeks.6 In one of the early reports, anhydramnios was described, but the exact gestation and the time to recovery of amniotic fluid was not detailed.7 CASE © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 A 28-year-old primigravid was referred to National Women’s Hospital (NWH), Auckland, New Zealand, at 30 weeks of gestation with anhydramnios, fetal growth at the fifth percentile, and a normal umbilical artery Doppler resistance index (0.58). History and examination did not confirm ruptured membranes. The patient’s history was complicated by a left breast infiltrative ductal carcinoma, grade 2, and stage T2N2M0 in December 2004. She underwent radical mastectomy of the left breast and lymphadenectomy. This was followed by chemotherapy and radiotherapy. She had 18 of 18 nodes that were positive. The cells were estrogen receptor– and progesterone receptor–negative and human epidermal growth factor receptor 2–positive. She had no other significant medical history or family history. She had smoked about 5–10 cigarettes per day, having stopped since becoming pregnant. Brachial plexus and pulmonary metastases were diagnosed at 20 weeks of gestation by magnetic resonance imaging. The patient was treated with adjuvant therapy docetaxel and trastuzumab. Her first cycle was commenced VOL. 110, NO. 2, PART 2, AUGUST 2007 Sekar and Stone From the Department of Maternal and Fetal Medicine, National Women’s Health Auckland Hospital, Auckland, New Zealand; and Department Maternal and Fetal Medicine, Faculty of Medical and Health Science, University of Auckland, New Zealand. The authors thank the obstetricians at the Auckland Obstetric Centre, Auckland, New Zealand, for the clinical care provided by them. Corresponding author: Dr. Renuka Sekar, Department of Obstetrics and Gynecology, Faculty of Medical and Health Science, University of Auckland, P.O. Box 92019, Auckland, New Zealand; e-mail: renukas@adhb.govt.nz. Financial Disclosure The authors have no potential conflicts of interest to disclose. Trastuzumab Use in Pregnancy 507 at 23 weeks and consisted of docetaxel 190 mg and trastuzumab 695 mg (8 mg/m2) as a single parenteral dose. The next cycle was 21 days later at 26 weeks of gestation and consisted of docetaxel 190 mg and trastuzumab 520 mg, followed by a final dose of 170 mg trastuzumab a week later. An ultrasound examination was performed at 30 weeks of gestation after clinical suspicion of fetal growth restriction showed growth measurements at the fifth percentile, with anhydramnios (Figs. 1 and 2). Trastuzumab was omitted because of the presence of anhydramnios in the fetus during the patient’s third cycle of chemotherapy, which was administered at 30 weeks of gestation. The fourth cycle of chemotherapy was deferred until after delivery because of concerns about fetal welfare. Two doses of betamethasone, 11.4 mg intramuscularly, 24 hours apart, were administered because of the concern that preterm delivery might be required. Growth of the fetus remained at the fifth percentile. At 33 weeks of gestation, one pocket of amniotic fluid of 2.2 cm was noted (Fig 3). Because of the advanced nature of the patient’s tumor and the fact that therapy had been curtailed, elective cesarean delivery for breech presentation occurred at 36⫹2 weeks of gestation so that that treatment could be resumed. At cesarean delivery, there was a small amount of clear amniotic fluid. A male infant weighing 2,230 g was delivered, with Apgar scores of 7 at 1 minute and 9 at 5 minutes. The infant showed no signs of prolonged oligohydramnios, such as positional deformities or respiratory abnormalities. Neonatal urine output was normal, as has been subsequent development. COMMENT Management of gestational breast cancer is best done with a multidisciplinary approach. The preferred surgical treatment of breast cancer in pregnant women is Fig. 1. Ultrasound scan at 30 weeks of gestation showing absence of amniotic fluid. Arrow indicates the bladder. Sekar. Trastuzumab Use in Pregnancy. Obstet Gynecol 2007. 508 Sekar and Stone Trastuzumab Use in Pregnancy Fig. 2. Ultrasound scan at 30 weeks of gestation showing absence of amniotic fluid. Arrow indicates the stomach. Sekar. Trastuzumab Use in Pregnancy. Obstet Gynecol 2007. Fig. 3. Ultrasound scan at 33 weeks of gestation showing reappearance of amniotic fluid. Arrow indicates amniotic fluid. The outline in the figure is a Doppler box showing the absence of umbilical cord in the pocket of amniotic fluid. Sekar. Trastuzumab Use in Pregnancy. Obstet Gynecol 2007. mastectomy because radiotherapy usually cannot be safely performed. Chemotherapy is generally considered to be safe in the second and third trimester but may be associated with low birth weight. To date, long-term follow-up has demonstrated normal growth and development of children exposed to antineoplastic agents in utero.8 As yet, no clear evidence has supported the concern of delayed malignancies or infertility in the offspring, but other effects such as those described in this case, although transient, may nevertheless have important implications for fetal welfare. The new agent Herceptin (trastuzumab) blocks the human epidermal growth factor receptor 2 protein. Human epidermal growth factor 2 is known to have an important role in embryonic development. Studies involving embryonic, fetal, and adult OBSTETRICS & GYNECOLOGY rat kidneys found that heparin-bound epidermal growth factor was expressed strongly in the embryonic and neonatal rat kidney, compared with the adult kidney. Heparin-bound epidermal growth factor may play a role in the development of renal epithelial cells in rat kidneys. Blocking the human epidermal growth factor receptors may have an effect on human renal epithelial cell function. Blocking human epidermal growth factor receptor 2 proteins in the renal epithelial cell membranes may produce anhydramnios by effects on aquaporins or by an as yet undiscovered mechanism. Water channels, known as aquaporins, are a family of intrinsic membrane proteins that span the phospholipid bilayer of cell membranes and serve as selective pores for the transfer of water. Ten mammalian aquaporins have been cloned, four highly expressed (aquaporin 1, 2, 3, 4) in the kidney. The cell membrane water channel protein aquaporins may be important in regulating the intramembranous pathway of amniotic fluid (AF) resorption. A recent study provides evidence of aquaporin-3 expression in human fetal membranes. Aquaporin-3 is permeable to water, urea, and glycerol, and modulation of its expression in fetal membranes may contribute to AF homeostasis.9 Aquaporin-9 mRNA is also expressed in the human chorion and amnion and may play an important role in amniotic fluid volume regulation.10 The U.S. Food and Drug Administration approved Herceptin in 1998 for the treatment of advanced breast cancer. It is a category B drug in pregnancy (ie, animal reproduction studies have failed to demonstrate a risk to the fetus, and there are no adequate and well-controlled studies in pregnant women or animal studies that have shown an adverse effect, but adequate and well-controlled studies in pregnant women have failed to demonstrate a risk to the fetus in any trimester).11 Preclinical models demonstrated that this antibody has significant antitumor activity as a single agent and has synergy with certain chemotherapeutic drugs. Phase 2 and 3 clinical trials performed in women with metastatic breast cancers that overexpress human epidermal growth factor receptor 2 have shown that trastuzumab improves survival when used as first-line therapy in combination with chemotherapy.12,13 Trastuzumab was generally well tolerated in all the trials, but reversible cardiotoxicity was an unexpected adverse effect (particularly combined with anthracyclines). Studies in cynomolgus monkeys showed no harm VOL. 110, NO. 2, PART 2, AUGUST 2007 to the fetus. Some studies did reveal placental transfer of trastuzumab in the monkeys. There are three previous case reports of Herceptin use in pregnancy,5–7 of which there is only one associated with decreased amniotic fluid volume when used at less than 30 weeks of pregnancy.7 Only one report considered the mechanisms by which this effect may occur. That report did not comment on the recovery time of the amniotic fluid volume. We have shown, by close follow-up that it may take up to 7 weeks for a normal (that is more than 2 cm in depth) volume of amniotic fluid to reappear after the cessation of the trastuzumab therapy. Trastuzumab use in pregnancy may be associated with a reduction in amniotic fluid volume and fetal growth. Serial monitoring of these pregnancies is recommended. It is hypothesized that trastuzumab may affect aquaporin channels that line the basement membrane of the renal tubules, leading to decreased fetal urinary production and oligohydramnios. The transplacental transfer and mechanism of action in humans warrants further study. REFERENCES 1. Nugent P O’Connell TX. Breast cancer and pregnancy. Arch Surg 1985;120:1221–4. 2. Ives A, Saunders C, Semmens J. The Western Australian gestational breast cancer project: a population-based study of incidence, management and outcomes. Breast 2005;14: 276–82. 3. Foxcroft LM, Evans EB, Porter AJ. The diagnosis of breast cancer in women younger than 40. Breast 2004;13:297–306. 4. Tokunaga E, Oki E, Nishida K, Koga T, Egashira A, Morita M, et al. Trastuzumab and breast cancer: developments and current status. Int J Clin Oncol 2006;11:199–208. 5. Waterston AM, Graham J. Effect of adjuvant trastuzumab in pregnancy. J Clin Oncol 2006;24:321–2. 6. Fanale MA, Uyei AR, Theriault RL, Adam K, Thompson RA. Treatment of metastatic breast cancer with trastuzumab and vinorelbine during pregnancy. Clin Breast Cancer 2005;6: 354–6. 7. Watson WJ. Herceptin (trastuzumab) therapy during pregnancy; association with reversible anhydramnios. Obstet Gynecol 2005;105:642–3. 8. Aviles A, Niz J. Long-term follow-up of children born to mothers with acute leukaemia during pregnancy. Med Pediatr Oncol 1988;16:3–6. 9. Wang S, Amidi F, Beall M, Gui L, Ross MG. Aquaporin 3 expression in human fetal membranes and its up-regulation by cyclic adenosine monophosphate in amnion epithelial cell culture. J Soc Gynecol Investig 2006;13:181–5. 10. Wang S, Chen J, Beall M, Zhou W, Ross MG. Expression of aquaporin 9 in human chorioamniotic membranes and placenta. Am J Obstet Gynecol 2004;191:2160–7. 11. Meadows M. Pregnancy and the drug dilemma. U.S. Food and Drug Administration. Available at: http://www.fda.gov/fdac/ features/2001/301_preg.html. Retrieved April 20, 2007. Sekar and Stone Trastuzumab Use in Pregnancy 509 12. Tokunaga E, Oki E, Nishida K, Koga T, Egashira A, Morita M, et al. Trastuzumab and breast cancer developments and current status. Int J Clin Oncol 2006;11:199 –208. 13. Yeon CH, Pegram MD. Anti-erb B-2 antibody trastuzumab in the treatment of HER-2-amplified breast cancer. Invest New Drugs 2006;23:391– 409. Nitrofurantoin Neuropathy ciated with intermittent, then continuous, use of nitrofurantoin for recurrent UTIs. Medical history was significant for occasional uncomplicated UTIs and mild lumbar disc disease, treated conservatively after her second delivery. Her pertinent history is described here. The patient’s third pregnancy was complicated by eight culture-positive UTIs, mainly Escherichia coli and Enterococcus. Most were treated with nitrofurantoin and she was placed on continuous suppression during the third trimester. During the latter half of the pregnancy, she developed bilateral thigh pain and left-sided chest pain. The chest pain did not respond to antacids. The thigh pain was diagnosed as meralgia paresthetica by the neurosurgeon she had seen previously for her disc disease. Over the next 16 months, while breastfeeding, the patient contracted nine UTIs, mostly treated with nitrofurantoin. Her thigh and chest pain worsened. A gastroenterologist performed a modified barium swallow and endoscopy, with no gastrointestinal cause found for her chest pain. Her urologist performed a complete urologic evaluation with no anatomic cause found for the recurrent UTIs. She conceived 16 months after the third delivery and experienced only three UTIs during this pregnancy, which were treated with penicillin and fosfomycin tromethamine. She had no recurrence of thigh or chest pain during this pregnancy. During 2 years of breastfeeding, the patient had 14 culture-positive UTIs, most treated with nitrofurantoin. Two years postpartum, she developed numbness of the left vulva and vagina, sexual dysfunction, and severe thigh pain. She had urinary frequency in the absence of infection. She consulted her neurosurgeon, who found focal areas of hypesthesia in the lower extremities. Strength and reflexes were normal. Evaluation included normal magnetic resonance imaging scans of the brain and spinal cord. Simultaneously, she developed tingling and pain in the fingers and toes and muscle spasms and paresthesias over the entire body. A consulting neurologist rendered the diagnosis of a postviral syndrome and prescribed gabapentin for pain. Despite this treatment, the patient’s symptoms continued to worsen, and her family physician suspected early multiple sclerosis (MS). She was referred to a neurologist with expertise in this disorder, who adopted a “wait and see” approach. Over the next 18 months, he followed her expectantly with serial magnetic resonance imaging scans, which remained normal. Her neuropathic pain, paresthesias, and numbness of the vulva and vagina intensified, and the urinary frequency worsened to the point where she was voiding over 20 times daily. She frequently used nitrofurantoin during this time for presumed UTIs, although cultures were often negative. Her urologist suspected detrusor insta- A Forgotten Adverse Effect Leslie D. Kammire, MD, and Peter D. Donofrio, MD BACKGROUND: Nitrofurantoin is a widely prescribed antibiotic used to treat uncomplicated lower urinary tract infections. Unknown in the obstetric and gynecologic literature is the complication of peripheral neuropathy as an adverse effect. CASE: We describe a patient who developed a severe sensory neuropathy after taking nitrofurantoin intermittently and then continuously over a 7-year period. She recovered almost completely after its discontinuation. CONCLUSION: Peripheral neuropathy is a rare and potentially reversible adverse effect, unreported in the obstetric and gynecologic literature, and commonly unrecognized by physicians who prescribe it. (Obstet Gynecol 2007;110:510–2) N itrofurantoin is a synthetic antibacterial agent that has been used for more than 50 years to treat uncomplicated urinary tract infections (UTIs). Advantages of its use include low cost, safety during pregnancy and breastfeeding, and few drug interactions.1 Neurologic adverse effects of nitrofurantoin are rare but include neuropathy, dizziness, vertigo, diplopia, cerebellar dysfunction, and benign intracranial hypertension.2 We submit this case report to alert obstetricians and gynecologists to the often unrecognized adverse effect of neuropathy. CASE A 38-year-old multiparous woman developed a variety of sensory neuropathic symptoms over a 7-year period asso- From the Departments of Obstetrics and Gynecology and Neurology, Wake Forest University School of Medicine, Winston-Salem, North Carolina. Corresponding author: Leslie D. Kammire, MD, Department of Obstetrics and Gynecology, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157-1066; e-mail: lkammire@wfubmc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 510 Kammire and Donofrio Nitrofurantoin Neuropathy OBSTETRICS & GYNECOLOGY bility despite a normal urodynamic evaluation and started her on oxybutynin chloride and daily nitrofurantoin for UTI prophylaxis, which she took for the next year. During this year of nitrofurantoin prophylaxis, the patient experienced worsening of all her neurologic symptoms. She developed a tremor, myoclonus, anosmia, and weakness of the hands. Her neuropathic pain became more constant and spread to the upper extremities and torso. Gabapentin only helped partially despite high doses. She experienced progressive numbness and loss of proprioception in her feet and hands, causing her to frequently stumble and drop things. Her chest pain recurred, leading to further gastrointestinal workup of esophageal manometry and 24hour pH monitoring, both normal. At the end of this year, the nitrofurantoin prophylaxis was stopped because she had not had any further UTIs. Because of her worsening neurologic symptoms, she was treated empirically for possible MS with two courses of high-dose prednisone without success. She was referred to another neurologist, who documented on examination decreased sensation to pinprick, vibration, temperature, and proprioception distally in both upper and lower extremities. The electromyogram showed features consistent with chronic denervation in an intrinsic foot muscle, suggestive of a small fiber neuropathy. The patient acknowledged that she had taken nitrofurantoin intermittently for years, and more recently on a continual basis. She was diagnosed with a sensory polyneuropathy secondary to prolonged nitrofurantoin use and was advised to permanently discontinue the drug. The patient has made almost a complete neurologic recovery 21 months after stopping nitrofurantoin. She has mild neuropathic pain, numbness, and decreased temperature sensation distally in the fingers and toes but has not required symptomatic treatment. Muscle cramps, chest and thigh pain, tremors, and myoclonus have abated. She still has slight numbness of the left vulva but has recovered normal sexual function. She voids normally. She uses methenamine mandelate-sodium biphosphate (a urinary acidifier), cranberry extract, and intravaginal estrogen cream for UTI prophylaxis and is doing well, with only three culture-positive UTIs during this time Nitrofurantoin is a synthetic antibiotic introduced in 1953 for treatment and suppression of lower urinary tract infections. It is effective against Escherichia coli, Enterococcus, and Staphylococcus aureus, as well as some strains of Klebsiella and Enterobacter. Because of multiple sites of action, resistance has not been a problem, despite over 50 years of use. Nitrofurantoin is not associated with impaired fertility or teratogenicity and is considered safe in pregnancy and breastfeeding.1 It has few drug interactions, making it an attractive choice for treating elderly patients on multiple med- ications. Primary adverse effects are nausea, emesis, and anorexia.1 Despite its overall safety, rare but serious adverse effects are reported. The most widely known is pulmonary toxicity.1 Reports also exist of toxic hepatitis and blood dyscrasias.1 Neurotoxicity from nitrofurantoin is less recognized and is estimated to occur in 0.0007% of courses of therapy.1 Case reports of drug-induced peripheral neuropathy appeared in the general medicine and neurologic literature in the late 1950s.3–5 Toole and Parrish6 reviewed the medical literature in 1973 and found 137 cases from 58 manuscripts. Reports were published through the mid 1980s, mostly from England, Australia, New Zealand, and Scandinavia. A PubMed search from January 1992 to March 2007, using the search terms “nitrofurantoin” and “neuropathy” and limited to the English language, revealed only one case report and letter to the editor (Spring PJ, Sharpe DM, Hayes MW. Nitrofurantoin and peripheral neuropathy: a forgotten problem [letter]? Med J Aust 2001;174:153– 4).7 Although nitrofurantoin is frequently prescribed for women, particularly during pregnancy, no published articles on nitrofurantoin neuropathy can be found in the obstetrics and gynecology literature. The neuropathy from nitrofurantoin exposure is typically sensory and motor in distribution and usually begins distally in the extremities. Patients complain of numbness, tingling, and pain in the toes and fingers, and over time this moves proximally. The pain often manifests as sharp, jabbing pain. In the severest form, there is motor involvement causing distal weakness and atrophy. Because the genitourinary system receives generous sensory innervation, patients with nitrofurantoin neuropathy may have urinary retention, frequency, nocturia, and sexual dysfunction. In mild presentations of nitrofurantoin neuropathy, nerve conduction studies can be normal, and the only abnormality will be features of distal denervation on the needle examination. The pathogenesis of nitrofurantoin neuropathy is primarily axon loss, but the exact mechanism of disease is not well understood.2 Nitrofurantoin neuropathy affects women more than men, especially the elderly. Its evolution is independent of dose and treatment duration.2,6 It is more likely in patients with impaired renal function, but it also occurs in patients with normal renal function. Neuropathy most commonly presents in the first 3 months of continuous therapy but can occur at any time. It may develop after intermittent treatment or even after nitrofurantoin has been stopped. Some patients develop a severe motor and sensory neurop- VOL. 110, NO. 2, PART 2, AUGUST 2007 Kammire and Donofrio COMMENT Nitrofurantoin Neuropathy 511 athy after short-term use, whereas others have a mild neuropathy after years of drug use. The prognosis for recovery appears to relate to the severity of the neuropathy when diagnosed, and some patients do not recover fully.6 In summary, nitrofurantoin is a widely prescribed drug, particularly in women, and is generally regarded as safe. Neuropathy is a rare form of toxicity, unrecognized by many physicians who prescribe it. The patient we present saw eight physicians over a 3-year period, and none recognized the toxicity to nitrofurantoin, despite the patient’s disclosure of her chronic use of nitrofurantoin. An informal poll of obstetricians, gynecologists, and urologists at our medical center did not reveal any individual who had knowledge of this neurotoxicity. Clinicians in obstet- rics and gynecology should be aware of this rare but serious complication of a widely used medication. Suggested Approach for Management of Placenta Percreta Invading the Urinary Bladder CONCLUSION: We suggest an alternative approach for managing placenta percreta invading the bladder, which may reduce blood loss and preserve an intact bladder. Rita Faranesh, MD, Romano Shabtai, MD, Shalev Eliezer, MD, and Salim Raed, MD P BACKGROUND: Management of placenta percreta invading the urinary bladder usually requires radical surgery, which may include partial or total resection of the bladder. CASE: A multigravida presented with placenta previa percreta invading the urinary bladder. Preoperatively, prophylactic occlusive balloon catheters were placed in the internal iliac arteries and were inflated after cesarean delivery of a healthy newborn. Subtotal hysterectomy with removal of a large part of the placental volume was then performed. Part of the placenta that was adherent to the bladder was left in situ. This way we were able to avoid severe blood loss and preserve the bladder intact. The woman was discharged healthy, on day 9 postoperatively. From the Department of Obstetrics and Gynecology, Ha’Emek Medical Center, Afula, and Rappaport Faculty of Medicine, Technion, Haifa, Israel. Corresponding author: Eliezer Shalev, MD, Department of Obstetrics and Gynecology, Ha’Emek Medical Center, Afula, Israel 18101; e-mail: shaleve@tx.technion.ac.il. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 512 Faranesh et al Placenta Percreta Invading the Bladder REFERENCES 1. D’Arcy PF. Nitrofurantoin. Drug Intell Clin Pharm 1985;19: 540–7. 2. Yiannikas C, Pollard JD, McLeod JG. Nitrofurantoin neuropathy. Aust N Z J Med 1981;11:400–5. 3. De Olivarius BF. Polyneuropathy due to nitrofurantoin therapy [in Danish]. Ugeskr Laeger 1956;118:753–5. 4. Willett RW. Peripheral neuropathy due to nitrofurantoin. Neurology 1963;13:344–5. 5. Vickers FN. Peripheral neuropathy due to nitrofurantoin. J Ky Med Assoc 1965;63:38–40. 6. Toole JF, Parrish ML. Nitrofurantoin polyneuropathy. Neurology 1973;23:554–9. 7. Electrolyte disorders and sodium phosphates. Prescrire Int 2002;11:49. (Obstet Gynecol 2007;110:512–5) lacenta percreta, which invades through the myometrium to the serosa and occasionally to adjacent organs, is a rare and complicated form of pathologic placentation. Due to the rising rate of cesarean deliveries, more cases of abnormal placentation are being reported. Wu et al1 reported that the incidence of placenta accreta was 1:533 for the period 1982–2002, much greater than previous reports ranging from 1:4,027 in the 1970s to 1:2,510 in the 1980s. The most important risk factors are previous cesarean delivery, placenta previa, and advanced maternal age.1 Besides increasing the rate of maternal and fetal death, placenta accreta that invades the urinary bladder may cause urinary fistula, ureteral transection, and bladder laceration requiring partial or total cystectomy.2 Management options for placenta percreta consist of surgical removal of the uterus and involved tissues or, alternatively, conservative treatment with the placenta left in situ after delivery. O’Brien et al3 reported, according to a questionnaire issued to members of the Society of Perinatal Obstetricians, that 93% of cases of placenta percreta were managed with total hysterectomy and removal of involved tissues. We suggest a new approach for management of placenta percreta invading the bladder. This includes prophylactic occlusive balloon catheters combined with subtotal hysterectomy, leaving that portion of the placenta that is adherent to the bladder in situ to OBSTETRICS & GYNECOLOGY reabsorb spontaneously. This may avoid severe blood loss and preserve an intact bladder. CASE A 38-year-old woman (gravida 5 para 4) with four previous cesarean deliveries was admitted at 27 weeks of gestation with gross hematuria. An anterior central placenta previa was diagnosed by ultrasound examination 3 weeks before admission. On admission, the woman was medically stable and had no vaginal bleeding. External fetal heart rate monitoring demonstrated a reactive tracing. Besides complete placenta previa, ultrasonography revealed hypoechogenic masses resembling blood clots in the bladder. There was a protrusion of the placenta into the bladder, with no clear myometrial tissue separating the placenta and the urinary bladder wall. Color flow ultrasonogram revealed placental lacunae and large blood vessels located at the level of the serosa-bladder interface. Given her history of multiple cesarean deliveries and these clinical and sonographic findings, the diagnosis of placenta percreta invading the bladder was most probable. The hematuria subsided spontaneously by the next day and did not recur again for the rest of her pregnancy. Clotting tests were normal. The hemoglobin level was stable. Accordingly, she was managed expectantly. The woman received one course of corticosteroids to promote fetal lung maturation. Urine culture was sterile. After two weeks of continuous hospitalization, she was discharged to home. At 33 weeks of gestation, the woman developed recurrent episodes of preterm uterine contractions. We decided to schedule an elective delivery at 34 weeks of gestation and not wait to term to avoid a situation where an emergency delivery would be required due to the onset of labor or hemorrhage. Amniocentesis was offered for documenting fetal lung maturity, but the woman refused the procedure. Because the patient did not desire future fertility, preparations were made for a cesarean hysterectomy and preoperative uterine artery catheterization. Informed consent, including a discussion of risks, benefits, and alternatives of the planned procedure, were discussed with the woman. We took the precaution of having a skilled urologist present during the procedure in case a bladder resection and repair might be required. The blood bank team was also prepared in case multiple blood components would be required. The morning of the operation, double lumen catheters were introduced through both the right and left femoral artery under fluoroscopic guidance by an experienced interventional radiologist. Each catheter was successfully placed into the anterior division of the internal iliac artery. Each catheter had an inflatable balloon localized at the tip of the catheter. The other lumen was reserved for the injection of embolizing material. In the operating room, a cystoscopy was performed by a skilled urologist under general anesthesia with placement of ureteric stents that might aid in identification of the ureters and prevent ureteric injury. Hyperemia and congestion of blood vessels were noted on the posterior wall of the bladder. A vertical midline skin incision was made through a previous VOL. 110, NO. 2, PART 2, AUGUST 2007 incision scar, extending cephalad up to the level of the umbilicus. When the peritoneal cavity was entered, we encountered mild adhesions of the omentum to the anterior uterine serosa, and these were lysed without difficulty. The placenta projected through the uterine wall and occupied the lower half of the anterior uterus. The upper margin of the placenta was visible. The lower part of the placenta was attached to the posterior wall of the bladder. The uterus was incised horizontally, above the level of the placenta, to avoid spontaneous expanding of the incision toward the placenta. A female infant weighing 1,812 g was delivered. The Apgar score was 7 at 5 minutes. The umbilical artery pH was 7.39. The occlusive balloons were inflated immediately, and a marked diminution of bleeding from the incised arteries was noticed. A running suture of Polysorb 1 (Tyco, Tokyo, Japan) was rapidly placed around the cut edge of the incision for rapid temporary hemostasis, leaving the placenta in situ. Subsequently we performed a subtotal hysterectomy, leaving the anterior lower uterine segment and the cervix with the attached cotyledon that was adherent to the posterior bladder wall in situ (Fig. 1). Besides aggravating blood loss, it was apparent that any attempt to remove this portion of the placenta might cause a bladder injury and the need for partial or total cystectomy. The incision was then closed in two layers using Vicryl 1 (Ethicon, Somerville, NJ). At this point we noted no active bleeding intraabdominally or through the vagina. Postoperatively, the woman was transferred to the angiography suite. Uterine artery embolization was performed on both sides by way of the catheters that had previously been placed, using absorbable gelatin sponge (Gelfoam, Pharmacia & Upjohn, Kalamazoo, MI), and the balloons were deflated. The woman received a total of 4 units of red blood cells. Her hemoglobin concentration stabilized at a level of 8.5 g/dL. The ␤-hCG level was 303 International Units/L, 2 days postoperatively. No complications related to the catheterization were observed. The diagnosis of abnormal placentation was confirmed histologically. On day 9 postoperatively, the patient was discharged home in good condition. She was seen after 1 week and again every 3 weeks for follow-up and for serial ␤-hCG levels, hemoglobin concentration, and ultrasound evaluation. The ␤-hCG levels dropped spontaneously to zero 3 weeks postdelivery, and her hemoglobin concentration remained stable. She reported intermittently passing tissue and clots vaginally during the first few weeks postpartum. The remaining cotyledon shrank spontaneously from 5.5⫻3.6 cm, measured sonographically before discharge, to 2.5⫻2.7 cm 3 months later. No other complications were reported as of the writing of this paper 4 months postoperatively. COMMENT Placenta percreta that invades the myometrium and serosa to the urinary bladder is a rare condition that may cause catastrophic obstetric and urologic complications.2 The largest meta-analysis performed in- Faranesh et al Placenta Percreta Invading the Bladder 513 Fig. 1. Diagram illustrating the uterus after delivery of the fetus. Illustration: Julian Maack. Faranesh. Placenta Percreta Invading the Bladder. Obstet Gynecol 2007. cludes 54 cases, of which 44% of the women were reported to have had partial or total cystectomy. Three maternal deaths were reported among the 54 cases.2 The management of this rare situation is complicated and still a challenge. A collaborative team approach to management is imperative, because it is necessary to discuss all the various options with the patient and the appropriate consultants. We report a case of placenta previa percreta invading the urinary bladder in a woman who did not desire future fertility. We used two techniques to preserve the bladder intact. The first was to reduce the blood perfusion to the uterus by occluding the anterior division of the internal iliac arteries. The inflation of the balloons localized in the internal iliac arteries immediately and rapidly decreased bleeding in the surgical field. Besides decreasing blood flow to the uterus and reducing hemorrhage, this action allowed for better exposure for the surgical staff. The second technique was performing a subtotal hysterectomy, removing most of the placental volume 514 Faranesh et al Placenta Percreta Invading the Bladder and leaving in situ a portion of the placenta that was attached to the posterior wall of the bladder. We believe that by not attempting to remove this portion of the placenta, we prevented bladder injury and the need for a partial or total cystectomy. The use of methotrexate for conservative treatment of placenta percreta has been suggested by some4 but argued against by others.5 In our case, the initial ␤-hCG level was low, and because the use of the drug is still disputed, we chose not to use it postoperatively. The ␤-hCG levels fell spontaneously, and the placental tissue shrank. Several reports have been published recently regarding the use of prophylactic occlusive balloon catheters and embolization for the treatment of severe obstetric hemorrhage.4,5 Ojala et al6 reported the use of this technique in seven patients with abnormal placentation diagnosed prenatally. There were no complications associated with the procedure when performed in a prophylactic setting, and good outcomes were reported. Several complications were associated with OBSTETRICS & GYNECOLOGY emergency embolization though, such as vaginal necrosis, paresthesia of the leg, and thrombosis of the popliteal artery. However, conservative management of cases of placenta percreta invading the bladder, including leaving both the uterus and placenta in situ, use of methotrexate, and uterine artery embolization, has not always been so successful. Severe hemorrhage and emergency hysterectomy have occasionally resulted from such conservatism.7,8 On the other hand, the surgical approach, ie, total hysterectomy, has been reported as the treatment of choice in cases of abnormal placentation. However, in cases where the placenta invades the urinary bladder, it compounds the problems associated with the surgical approach because hysterectomy alone cannot effectively remove the placenta in its entirety. Hemostasis is difficult to achieve. Bladder resection makes the operation technically more complex, exposing the bladder and the ureters to the risk of injury. It seems to us that applying prophylactic occlusive balloon catheters in cases of placenta percreta invading the urinary bladder combined with subtotal hysterectomy, leaving that portion of the placenta that is adherent to the bladder in situ, is a good strategy that may reduce blood loss and preserve an intact bladder. However, the patient should be informed of the reported complications of the procedure and the limited experience in achieving this goal before asking for her consent. Profound HypothyroidismInduced Acute Menorrhagia Resulting in Life-Threatening Anemia CASE: We present the case of a 31-year-old patient with acute menorrhagia causing life-threatening anemia that resulted from profound hypothyroidism. Despite timely institution of thyroid replacement, the patient required emergent embolization of a uterine arteriovenous malformation after dilatation and curettage failed to control her bleeding. She was stabilized and discharged to home on the sixth hospital day. Three years later, she successfully conceived and delivered a healthy infant. Vasiliki A. Moragianni, MD, and Stephen G. Somkuti, MD, PhD BACKGROUND: Menorrhagia is a very common gynecologic entity with a broad differential diagnosis which includes both hypothyroidism and hyperthyroidism. From the Department of Obstetrics and Gynecology, Abington Memorial Hospital, Abington, Pennsylvania. Corresponding author: Stephen G. Somkuti, MD, PhD, Division of Reproductive Medicine, Department of Obstetrics and Gynecology, Abington Memorial Hospital, 1200 Old York Road, Abington, PA 19001; e-mail: ssomkuti@abington-repromed.com. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 REFERENCES 1. Wu S, Kocherginsky M, Hibbard JU. Abnormal placentation: twenty-year analysis. Am J Obstet Gynecol 2005;192:1458–61. 2. Washecka R, Behling A. Urologic complications of placenta percreta invading the urinary bladder: a case report and review of the literature. Hawaii Med J 2002;61:66–9. 3. O⬙Brien JM, Barton J, Donaldson E. The management of placenta percreta: conservative and operative strategies. Am J Obstet Gynecol 1996;175:1632–8. 4. Clement D, Kayem G, Cabrol D. Conservative treatment of placenta percreta: a safe alternative. Eur J Obstet Gynecol Reprod Biol 2004;114:108–9. 5. Weeks S, Stroud T, Sandhu J, Mauro M, Jaques P. Temporary balloon occlusion of the internal iliac arteries for control of hemorrhage during cesarean hysterectomy in a patient with placenta previa and placenta increta. J Vasc Interv Radiol 2000;11:622–4. 6. Ojala K, Perala J, Kariniemi J, Ranta P, Raudaskoski T, Tekay A. Arterial embolization and prophylactic catheterization for the treatment for severe obstetric hemorrhage. Acta Obtet Gynecol Scand 2005;84:1075–80. 7. Dinkel HP, During P, Schnatterbeck P, Triller J. Percutaneous treatment of placenta percreta using coil embolization. J Endovasc Ther 2003;10:158–62. 8. Butt K, Gagnon A, Delisle MF. Failure of methotrexate and internal iliac balloon catheterization to manage placenta percreta. Obstet Gynecol 2002;99:981–2. CONCLUSION: Our case demonstrates the importance of thyroid evaluation in the patient who presents with menorrhagia. Timely management with medical treatment as well as conservative surgical or radiological interventions can facilitate resolution of symptoms and preserve the patient’s fertility potential. (Obstet Gynecol 2007;110:515–7) M enorrhagia, as defined by blood loss of 80 mL or more per cycle, affects 9% to 14% of otherwise healthy women and has multiple causes (Table 1).1 For the reproductive-age patient who presents with menorrhagia, initial laboratory evaluation should include at a minimum a complete blood count and pregnancy test. Additionally, consideration should be given to checking Moragianni and Somkuti Hypothyroidism and Life-Threatening Anemia 515 thyroid and liver function tests, coagulation profile, serum prolactin and follicle-stimulating hormone levels.2 Among the endocrine causes, thyroid disease merits special mention. It is the second most common endocrine disease in women of reproductive age and has long been linked to menstrual disturbances. Hypothyroidism alone afflicts up to 3% of women by age 60.3 We report a case of life-threatening anemia as a result of severe hypothyroidism in a patient presenting with menorrhagia. CASE A 31-year-old nulligravida presented to our hospital with heavy vaginal bleeding and feeling faint. Her past medical and surgical history was unremarkable, with the exception of presenting with similar complaints at another institution 9 years prior. At that time the patient was admitted to the intensive care unit with anemia (hemoglobin on presentation was 7.1 g/dL) secondary to menorrhagia, and highoutput cardiac failure. She was stabilized and discharged home on hospital day 7 after multiple blood transfusions and a failed angiographic attempt to identify and embolize a presumed uterine arteriovenous malformation (AVM). The patient was subsequently lost to follow-up. On initial examination at our hospital, the patient’s blood pressure was 66/36 mm Hg, pulse 92 beats per minute, and respirations 22 per minute. Physical findings included generalized pallor, marked periorbital edema, alopecia, macroglossia, edematous extremities, and delayed deep tendon reflexes, all suggesting the possibility of a thyroid disorder. Pelvic examination revealed active bleeding from the cervical os. Her hemoglobin on admission was 3.9 g/dL, hematocrit 11.1%, white blood cell count 4,800/mm3, and platelets 149,000/mm3. The prothrombin time (PT) was 10.7 seconds and activated partial thromboplastin time (aPTT) 21 seconds. Serum hCG was negative. In addition to fluid resuscitation, the patient emergently received 2 units of packed red blood cells (RBCs) and was started on intravenous (IV) conjugated estrogens (25 mg every 4 hours). On hospital day 2, the patient continued to pass large clots and feel extremely tired. The hemoglobin was 6.6 g/dL, hematocrit 18.9%, and the patient was transfused an additional 2 units of packed RBCs. Thyroid stimulating hormone (TSH) was elevated at 881.8 micro-international units/mL, with a total triiodothyronine of 2 ng/mL and a total thyroxine of less than 0.5 mcg/mL. The patient was started on IV levothyroxine (100 mcg every 24 hours) and IV dexamethasone (2 mg once) followed by IV hydrocortisone (60 mg every 8 hours) for myxedema and presumed adrenal insufficiency. On hospital day 3, she continued to saturate multiple sanitary pads, and the hemoglobin dropped to 6.2 from 8.4 g/dL after the fourth unit of blood. Two additional units of packed RBCs were transfused at that time. Pelvic ultrasonography was also obtained, revealing a normal-sized 516 Moragianni and Somkuti uterus and right ovary, a cystic left ovary measuring 4.2⫻8.8⫻5.9 cm, and a 6-cm vaginal blood clot. After the placement of a right femoral line for additional venous access, oozing was noted from all IV access lines. Platelets dropped to 55,000/mm3, PT was elevated to 13.5 seconds and aPTT to 22 seconds. The presumptive diagnosis of disseminated intravascular coagulation was thus made and the patient was transfused another 2 units of packed RBCs, 1 unit of platelets, and 4 units of cryoprecipitate. That afternoon, the patient underwent emergent dilation and curettage under local anesthesia. Pathologic findings were consistent with benign proliferative endometrium and anovulatory changes. Despite a vigorous curettage the patient continued to bleed. Alternative conservative interventions were then considered as she desired to maintain her reproductive potential. Interventional radiology performed a bilateral uterine artery embolization using Gelfoam (Pharmacia & Upjohn, Bridgewater, NJ) and hemostasis was achieved. A uterine arteriovenous malformation was identified next to the uterus that was noted to have increased myometrial vascularity. The patient also received an additional 2 units of fresh frozen plasma and 4 units of cryoprecipitate. She started on 1 mg of norethindrone and 0.035 mg of ethinyl estradiol (E2)– containing oral contraceptive tablets four times daily. On hospital day 4, the patient appeared to be improving clinically and reported no vaginal bleeding. The hemoglobin was 8.5 g/dL, hematocrit 24.4%, platelets 113,000/mm3, TSH 359 micro-international units/mL, free thyroxine 0.31 mcg/ mL, PT 11 seconds, and aPTT 11.2 seconds. Prolactin was 195 ng/mL, fibrin split products 9 mcg/mL, and fibrinogen 215 mg/dL. Brain magnetic resonance imaging revealed mild pituitary hypertrophy without an identifiable microadenoma and mild cerebral atrophy. Later that day hemoglobin was 10.1 g/dL and hematocrit 28.4%. The patient was tolerating a regular diet and denied any vaginal bleeding. Table 1. Causes of Menorrhagia Category Specific Cause Reproductive tract disorders Complications of pregnancy Infection Malignancy Benign lesions Anovulation Arteriovenous malformations Systemic diseases Coagulopathy Hepatorenal disease Endocrinologic disturbances Iatrogenic causes Medications Intrauterine devices Miscellaneous Hypothyroidism and Life-Threatening Anemia Smoking Stress Excessive exercise Trauma OBSTETRICS & GYNECOLOGY On hospital day 5, she continued to improve and was transferred out of the intensive care unit. Antithyroid antibody was negative and hemoglobin was stable at 10.3 g/dL. The patient remained stable and was discharged home on hospital day 6. Her medications on discharge included oral levothyroxine (100 mcg daily) and iron (325 mg thrice daily). She was also to decrease the oral contraceptive tablets to thrice daily for 3 days followed by twice daily for 3 days and once daily thereafter. The patient returned pregnant in March of 1998, after having continued appropriate thyroid replacement. Her hemoglobin at that time was 12.5 g/dL and TSH was 3.13 micro-international units/mL. She underwent uncomplicated vaginal delivery of a viable, healthy, full-term male infant. COMMENT This case demonstrates the highest TSH value we have ever encountered as well as one of very rare reports of successful full-term vaginal delivery after embolization of uterine AVM. Fifteen percent of women will be diagnosed with thyroid dysfunction at some point of their life, with menorrhagia affecting 32% to 80% of them. Primary hypothyroidism is typically characterized by normal gonadotropins but elevated TSH levels. Thyroidstimulating hormone has both follicle-stimulating hormone and luteinizing hormone–like effects through their shared ␣-subunit. As a result, negative feedback is down-regulated, leading to a decreased secretion of luteinizing hormone and an ensuing decrease in progesterone.3 At the same time, decreased sex hormone binding globulin results in increased serum estriol and unbound E2, which in turn leads to elevated levels of circulating free estrogens. This prolonged, unopposed effect on the endometrium results in failure of ovulation and corpus luteum formation, ultimately leading to menometrorrhagia.4,5 Additional considerations in the evaluation of these patients should also include ectopic pituitary adenomas, acquired von Willebrand’s disease, and uterine AVM. To date, there have been less than 50 reported cases in the literature of ectopic TSH-secreting pituitary tumors.6 Several investigators7 have described patients with newonset bleeding diatheses (commonly menorrhagia) who fulfill criteria for the diagnosis of hypothyroidism; yet VOL. 110, NO. 2, PART 2, AUGUST 2007 upon further investigation they have been found also to exhibit laboratory findings consistent with acquired von Willebrand’s disease. Notably, both symptoms and laboratory aberrations tend to normalize after treatment with thyroxine. Only case reports of uterine AVMs have been described in the literature; their incidence is thus unknown. They consist of aberrant fistulous arteriovenous communications that result in very large, tortuous vessels. The most common presenting symptom is menometrorrhagia and uncontrolled bleeding after uterine instrumentation. Cause can be congenital or acquired, most likely secondary to trauma, surgery, neoplasm, or infection. Ultrasonography is the diagnostic modality of choice, and the traditional treatment of uterine AVM has been hysterectomy. However, in the fertility-desiring patient, embolization is a tried alternative, with only a few reported cases of uncomplicated ensuing pregnancies.8 In conclusion, a prompt diagnosis of the cause of menorrhagia is of paramount importance for timely, effective therapy and preservation of reproductive function. REFERENCES 1. Van Eijkeren MA, Christiaens GC, Sixma JJ, Haspels AA. Menorrhagia: a review. Obstet Gynecol Surv 1989;44:421–9. 2. Long CA. Evaluation of patients with abnormal uterine bleeding. Am J Obstet Gynecol. 1996;175:784–6. 3. Larsen PR, Kronenberg HM, Melmed S, Polonsky KS. Williams textbook of endocrinology. 10th ed. Philadelphia (PA): Saunders; 2003. 4. Wilansky DL, Greisman B. Early hypothyroidism in patients with menorrhagia. Am J Obstet Gynecol 1989;160:673–7. 5. Bohnet HG, Fiedler K, Leidenberger FA. Subclinical hypothyroidism and infertility. Lancet 1981;2:1278. 6. Pasquini E, Faustini-Fustini M, Sciarretta V, Saggese D, Roncaroli F, Serra D, et al. Ectopic TSH-secreting pituitary adenoma of the vomerosphenoidal junction. Eur J Endocrinol 2003;148:253–7. 7. Michiels JJ, Schroyens W, Berneman Z, van der Planken M. Acquired von Willebrand syndrome type 1 in hypothyroidism: reversal after treatment with thyroxine. Clin Appl Thromb Hemost 2001;7:113–5. 8. Hoffman MK, Meilstrup JW, Shackelford DP, Kaminski PF. Arteriovenous malformations of the uterus: an uncommon cause of vaginal bleeding. Obstet Gynecol Surv 1997;52:736–40. Moragianni and Somkuti Hypothyroidism and Life-Threatening Anemia 517 Metastatic Uterine Leiomyosarcoma Regression Using an Aromatase Inhibitor Mary Pat Hardman, MD, Juan J. Roman, MD, Alexander F. Burnett, MD, and Alessandro D. Santin, MD, BACKGROUND: Metastatic and disseminated uterine leiomyosarcoma has limited chemotherapeutic options. CASE: A 45-year-old nulligravida was originally diagnosed with a primary uterine leiomyosarcoma and treated with a total abdominal hysterectomy, followed by six courses of adjuvant platinum and doxorubicin chemotherapy. Three years later she developed multiple metastatic tumor nodules in the lungs. Because of the estrogen receptor positivity of the leiomyosarcoma by immunohistochemistry, she was treated with 1 mg of anastrozole daily. Progressive tumor regression was demonstrated by serial imaging in all metastatic lung tumor deposits, with an objective response lasting at least 12 months. CONCLUSION: Uterine leiomyosarcoma metastatic to the lungs regressed with the use of the anastrozole aromatase inhibitor. (Obstet Gynecol 2007;110:518–20) U terine leiomyosarcomas are rare smooth-muscle tumors accounting for less than 2–3% of all uterine malignancies.1 Although approximately half of women diagnosed with uterine leiomyosarcoma will present with disease confined to the uterus, the majority of these patients will experience recurrence of disease, for which limited chemotherapeutic options are available.2 Aromatase inhibitors have been widely adopted as a safe and effective treatment for metastatic breast cancer. However, to the best of our knowledge, there are no data in the literature about the use of aromatase inhibitors in metastatic leiomyosarcoma. From the Department of Obstetrics and Gynecology, Division of Gynecologic Oncology, University of Arkansas for Medical Sciences, Little Rock, Arkansas. Corresponding author: Alessandro D. Santin, MD, UAMS Medical Center, Division of Gynecologic Oncology, University of Arkansas, Slot 518, 4301 W. Markham, Little Rock, AR 72205-7199; e-mail: SantinAlessandroD@ uams.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 518 Hardman et al CASE A 45-year-old nulligravida (body mass index 42 kg/m2) underwent a total abdominal hysterectomy and bilateral salpingo-oophorectomy for pelvic pain and menstrual irregularity associated with an increasing enlargement of the uterus. On clinical examination her abdomen was swollen with a uterus of about 18 weeks in size. Routine preoperative chest X-ray revealed normal lungs while a computed tomography (CT) scan of the abdomen and pelvis confirmed the presence of a markedly enlarged uterus measuring 14 cm in cephalocaudad extent, 15 cm in anteroposterior, and 21 cm in transverse dimension, with multiple masses showing cystic and necrotic degeneration. Pathologic examination revealed a massively enlarged and grossly distorted uterus, with the uncut specimen measuring 25⫻15⫻13 cm. Microscopic examination revealed a leiomyosarcoma (Fig. 1A) characterized by spindled neoplastic cells arranged into well-formed intersecting fascicles, diffuse smooth muscle actin and desmin immunoreactivity, and diffuse permeation of the myometrium with extension into the parametrial soft tissues, cervical stroma, and uterine serosa. Pelvic washings disclosed no tumor. After surgery, the patient was treated with adjuvant chemotherapy based on three courses of cisplatin and doxorubicin, followed by three courses of carboplatin and doxorubicin (because of a major allergic reaction to cisplatin). She then received regular follow-up with pelvic examination and Pap smears every 3 months and CT scans and chest X-rays every 6 –12 months. The patient remained disease free for almost 3 years when, at the time of a routine CT scan of the chest, multiple metastatic tumor nodules, ranging from 2–3 mm to 1 cm in size, were identified in both lungs (Fig. 2A). No sign of recurrence was evident in the abdomen and pelvis. Salvage chemotherapy was recommended at that time, but was declined by the patient. Immunohistochemical stains for estrogen receptor (ER), progesterone receptor, and c-kit ligand (CD117) showed strong nuclear immunoreactivity for both estrogen and progesterone receptors in more than 90% of tumor cells of the primary uterine leiomyosarcoma (Fig. 1B and not shown, respectively) and no reactivity for CD117 (not shown). Because of these findings, the patient was offered treatment with the aromatase inhibitor anastrozole (Arimidex, AstraZeneca, Wilmington, DE), 1 mg once daily. As shown in Figure 2B, a CT scan of the chest performed 4 months from the beginning of therapy documented a generalized decrease in the size of the metastatic tumor deposits in the lungs, with disappearance of the smaller tumor nodules. A chest X-ray performed after 8 months from the beginning of aromatase inhibitor treatment showed further reduction of the disease in the chest (not shown). Finally, a CT scan of the chest, abdomen, and pelvis performed 1 year after the beginning of salvage therapy with Arimidex showed no sign of recurrence in the pelvis (not shown) and a durable response to the aromatase inhibitor in the chest (Fig. 2C). The patient’s performance status remained 100% for the Leiomyosarcoma Regression With Anastrozole OBSTETRICS & GYNECOLOGY Fig. 1. Hematoxylin-eosin stain of uterine leiomyosarcoma showing malignant smooth muscle spindles and increased mitosis (arrow) (A, original magnification, ⫻400) with positive estrogen receptor immunostains (B, original magnification, ⫻200). Hardman. Leiomyosarcoma Regression With Anastrozole. Obstet Gynecol 2007. duration of the therapy, with no clinical or adverse effect reported to date. COMMENT Advanced and recurrent or metastatic uterine leiomyosarcomas are poor prognosis tumors with limited treatment options. Because of their propensity for early hematogenous spread, uterine leiomyosarcoma may metastasize to distant areas years after therapeutic hysterectomy. Adjuvant radiation therapy has not shown any survival advantage, and the use of singleagent and combination chemotherapy in the systemic treatment of these cancers has not resulted in clinically meaningful responses.2 The Gynecologic Oncology Group (GOG) has investigated multiple single- VOL. 110, NO. 2, PART 2, AUGUST 2007 Fig. 2. Representative chest computed tomography scan before the beginning of anastrozole therapy (A) and after 4 (B) and 12 (C) months. Arrows indicate a representative metastatic nodule showing regression with anastrozole therapy (eg, A, 1 cm; B, 0.6 cm; C, 0.6 cm). Hardman. Leiomyosarcoma Regression With Anastrozole. Obstet Gynecol 2007. agent drugs, as well as combinations, with mixed results. Doxorubicin still remains the most highly Hardman et al Leiomyosarcoma Regression With Anastrozole 519 active single agent in these cancers, with a response rate of approximately 25%.3 Memorial Sloan-Kettering Cancer Center’s phase II trial of gemcitabine and docetaxel in leiomyosarcoma patients has recently published a response rate of 53%.4 Unfortunately, the median duration of response is less than 6 months. Thus, despite these reports, no demonstration of a survival advantage has been shown, and new targeted therapies are under investigation. Anastrozole (Arimidex) is a third-generation nonsteroidal aromatase inhibitor endowed with marked therapeutic efficacy in metastatic breast cancer.5 Based on its efficacy and favorable adverse-effect profile, anastrozole is approved for first-line treatment of patients with ER-positive metastatic breast cancer. Similarly to the other third-generation nonsteroidal aromatase inhibitor, letrozole (reversible, type II), and to the steroidal aromatase inhibitor, exemestane (irreversible, type I), in use today, anastrozole acts by blocking the aromatase-P450 enzyme in the final step of estrogen synthesis, thus lowering circulating estrogen levels and depriving the ER of its substrate. Anastrozole, unlike tamoxifen (a selective estrogen receptor modulator), lacks partial agonist activity of the ER and induces a profound suppression in circulating estrogen of approximately 95–98%.5 There have been few active agents identified in the treatment of leiomyosarcomas that have resulted in durable responses. However, several facts support the potential hormonal dependency of a large subset of these tumors. Consistent with this view, estrogen receptor expression has been previously reported by several investigators in uterine leiomyosarcomas.6,7 Indeed, Bodner et al6 and Rao et al7 reported 57% and 71% ER positivity, respectively, in leiomyosarcoma series tested by immunohistochemistry. Furthermore, previous reports have shown that aromatase-P450 overexpression in uterine leiomyomas may be responsible for their growth advantage over the surrounding myometrium.8 To the best of our knowledge, however, there are no data in the literature about the use of aromatase inhibitors in leiomyosarcomas. Thus, our report is the first to describe a role for anastrozole in the treatment of metastatic leiomyosarcomas. Given the limited chemotherapy options available and because of their high safety profile, salvage therapy for patients with advanced or metastatic 520 Hardman et al ER-positive leiomyosarcoma with aromatase inhibitor may represent a novel, potentially effective, therapeutic approach. Consistent with this view, the case presented here demonstrates a dramatic and durable response of a metastatic uterine leiomyosarcoma to Arimidex, with no progression of disease documented to date. Importantly in this patient, adjuvant chemotherapy, based on the administration of two of the most effective cytotoxic drugs available against leiomyosarcoma (ie, platinum and doxorubicin), failed to sterilize microscopic tumor deposits and avoid metastatic tumor recurrence in the lungs, once again documenting the generalized resistance of leiomyosarcoma to adjuvant chemotherapy. In conclusion, we report a durable response to anastrozole in an ER-positive metastatic leiomyosarcoma to the lungs. The result of the current study demonstrates the potential responsiveness of metastatic leiomyosarcomas to aromatase inhibitor therapy. Randomized, controlled trials in ER-positive advanced or recurrent and metastatic leiomyosarcoma are warranted. REFERENCES 1. DiSaia PJ, Creasman WT. Cervical cancer. In: DiSaia PJ, Creasman WT, editors. Clinical gynecologic oncology. 5th ed. St. Louis (MO): Mosby; 1997. p. 1–106. 2. Look KY, Sandler A, Blessing JA, Lucci JA, Rose PG. Phase II trial of gemcitabine as second-line chemotherapy of uterine leiomyosarcoma: a Gynecologic Oncology Group (GOG) study, Gynecol Oncol 2004;92:644–7. 3. Omura GA, Major FJ, Blessing JA, Sedlacek TV, Thigpen JT, Creasman WT, et al. A randomized study of Adriamycin with and without dimethyl triazenoimidazole carboxamide in advanced uterine sarcomas. Cancer 1983;52:626–32. 4. Hensley ML, Maki R, Venkatraman E, Geller G, Lovegren M, Aghajanian C, et al. Gemcitabine and docetaxel in patients with unresectable leiomyosarcoma: results of a phase II trial. J Clin Oncol 2002;20:2824–31. 5. Ryan PD, Goss PE. Adjuvant hormonal therapy in peri- and postmenopausal breast cancer. Oncologist 2006;11:718–31. 6. Bodner K, Bodner-Adler B, Kimberger O, Czerwenka K, Leodolter S, Mayerhofer K. Estrogen and progesterone receptor expression in patients with uterine leiomyosarcoma and correlation with different clinicopathological parameters. Anticancer Res 2003;23:729–32. 7. Rao UN, Finkelstein SD, Jones MW. Comparative immunohistochemical and molecular analysis of uterine and extrauterine leiomyosarcomas. Mod Pathol 1999;12:1001–9. 8. Shozu M, Murakami K, Segawa T, Kasai T, Inoue M. Successful treatment of a symptomatic uterine leiomyoma in a perimenopausal woman with a nonsteroidal aromatase inhibitor. Fertil Steril 2003;79:628–31. Leiomyosarcoma Regression With Anastrozole OBSTETRICS & GYNECOLOGY Recurrent Hemorrhagic Ascites A Rare Presentation of Endometriosis Meena Palayekar, MD, Joseph Jenci, MD, PhD, and John A. Carlson Jr, MD BACKGROUND: Endometriosis is rarely a cause of recurrent hemorrhagic ascites. This report draws attention to this uncommon condition, which could present a diagnostic dilemma. CASE: A young African-American woman who had experienced recurrent hemorrhagic ascites for more than 2 years underwent laparotomy and was found to have extensive pelvic endometriosis. After a hysterectomy and bilateral salpingo-oophorectomy, her ascites resolved. CONCLUSION: Endometriosis should be considered in the differential diagnosis of recurrent hemorrhagic ascites in premenopausal women. The diagnosis always requires operative assessment and histologic confirmation. (Obstet Gynecol 2007;110:521–2) O vert clinical ascites is rarely associated with endometriosis. When it does occur it mimics ovarian cancer and often presents a diagnostic dilemma. CASE A young, primiparous African-American woman was referred for a gynecologic consultation 2 years after she initially presented with abdominal distension secondary to hemorrhagic ascites. Before this, she had been seen by specialists in gastroenterology, nephrology, cardiology, and infectious diseases, and none of the ensuing evaluations had been able to detect a cause for her recurrent ascites. Paracenteses, yielding 4 – 6 L of hemorrhagic fluid, had been performed for palliation on several occasions. The ascitic fluid cytology was negative for malignancy, and cultures were negative for bacterial and fungal infections. The fluid contained protein and red cells and was an exudate. Other than an iron deficiency anemia and mild hypoalbuminemia, all her laboratory tests were normal. Her CA 125 was 33.6 units/mL. Multiple imaging studies From the Department of Obstetrics and Gynecology, Saint Peter’s University Hospital, New Brunswick, New Jersey. Corresponding author: Meena Palayekar, MD, Department of Obstetrics and Gynecology, Saint Peter’s University Hospital, MOB 4th floor, 254 Easton Avenue, New Brunswick, NJ 08901; e-mail: mpalayekar@hotmail.com. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 revealed only ascites without any organ abnormalities, including uterus and ovaries. The patient had one prior uncomplicated pregnancy many years before. She had had an unrevealing diagnostic laparoscopy for dysmenorrhea many years before the onset of ascites. At the initial presentation of ascites, she had a second diagnostic laparoscopy during which a left ovarian cyst was identified and cauterized without biopsy. The surgeon did not document endometriosis. There was no family history of gynecologic malignancy, but both her mother and sister had endometriosis and uterine leiomyomata. On examination, the patient was a young AfricanAmerican woman of average build with moderate abdominal distension. Her vital signs were normal. There was no peripheral adenopathy. There was no pleural effusion. The abdominal examination was significant only for distension and ascites, without organomegaly or signs of peritonitis or bowel obstruction. On pelvic examination, the external genitalia, vagina, and cervix were visibly and palpably normal. On rectovaginal and bimanual examination, the uterus and adnexa were difficult to palpate because of the ascites, but did not seem abnormal. Due to failure to establish a diagnosis from multiple prior procedures, including laparoscopy, and her worsening clinical condition, an exploratory laparotomy was performed, at which time advanced pelvic endometriosis was identified, involving both tubes and ovaries and the sigmoid colon. Multiple biopsies were taken from the peritoneal surfaces. Frozen section confirmed the clinical diagnosis of endometriosis (Fig. 1). Because the patient had requested a definitive surgical treatment of her condition and was not desirous of future child bearing, a hysterectomy with bilateral salpingo-oophorectomy was performed. At 1 year after the surgery, the patient had not had recurrence of the ascites. Hormonal therapy was suggested, but declined. COMMENT Endometriosis is defined as the presence of endometrial tissue outside the uterine cavity.1 The usual symptoms of endometriosis include dysmenorrhea, pelvic pain, dyspareunia, and infertility. However, multiple reports have documented a wide variety of clinical symptoms, including bowel obstruction or hematochezia, ureteral strictures and hematuria, and hemorrhagic pleural effusion.2 Endometriosis presenting as recurrent massive hemorrhagic ascites has been rarely reported. When present, most patients are African American, nulliparous, aged 22–39 years, and have a history of infertility.3 Many cases clinically resemble ovarian cancer, presenting with unilateral or bilateral ovarian masses, ascites, an elevated CA 125 value, and weight loss. The diagnosis of endometriosis can be established only Palayekar et al Endometriosis With Hemorrhagic Ascites 521 resolution of ascites after six months of gonadotropin releasing hormone (GnRH) agonist therapy.6 However, because there is a risk of recurrence of ascites after cessation of GnRH agonist therapy, bilateral oophorectomy with or without hysterectomy may be an optimal management for patients not desiring preservation of fertility. The pathogenesis of ascites associated with endometriosis is uncertain. Bernstein et al7 suggested that the ascites may be a reaction to the blood and endometrial cells shed into the peritoneal cavity. This theory is supported by the fact that the ascitic fluid is an exudate. Diagnosis is made by histology showing benign endometrial glands and stroma in patients where other causes of ascites, such as heart failure, cirrhosis, renal insufficiency, tuberculosis, and hepatic or ovarian cancer, have been excluded. Fig. 1. Photomicrograph showing endometrial glands with stroma (arrows) in biopsy of tissue. Hematoxylin-eosin stain, ⫻40, original magnification. Palayekar. Endometriosis With Hemorrhagic Ascites. Obstet Gynecol 2007. after surgery and histologic confirmation.4 Although our patient did not have enlarged ovaries, the ovaries were found to be infiltrated with endometriosis. Our patient elected definitive surgical treatment with a total abdominal hysterectomy with bilateral salpingo-oophorectomy. However, many patients are nulliparous and may desire preservation of fertility. Therefore, once malignancy has been excluded, endocrine therapy may be considered even for patients with apparently advanced disease.5,6 Some authors have reported a complete 522 Palayekar et al REFERENCES 1. Brosens IA. Endometriosis. Current issues in diagnosis and medical management. J Reprod Med 1998;43 suppl:281–6. 2. Ekoukou D, Guilherme R, Desligneres S, Rotten D. Endometriosis with massive hemorrhagic ascites: a case report and review of literature [in French]. J Gynecol Obstet Biol Reprod (Paris) 2005;34:351–9. 3. Spitzer M Benjamin F. Ascites due to endometriosis. Obstet Gynecol Surv 1995;50:628–31. 4. Goumenou A, Matalliotakis I, Mahutte N, Koumantakis E. Endometriosis mimicking advanced ovarian cancer. Fertil Steril 2006; 86:219. 5. Fortier D, Dedecker F, Gabriele M, Graesslin O, Barau G. Endometriosis with ascites and pleural effusion: a case report [in French]. Gynecol Obstet Fertil 2005;33:508–10. 6. Dias C, Andrade JM, Ferriani RA, Villanova MG, Meirelles RS. Hemorrhagic ascites associated with endometriosis. A case report. J Reprod Med 2000;45:688–90. 7. Bernstein JS, Perlow V, Brenner JJ. Massive ascites due to endometriosis. Am J Dig Dis 1961;6:1. Endometriosis With Hemorrhagic Ascites OBSTETRICS & GYNECOLOGY Clitoroplasty Robyn A. Sayer, MD, Aaron Deutsch, MD, and Mitchel S. Hoffman, MD BACKGROUND: Clitoroplasty, especially in an adult, is a rare procedure. The goals of clitoroplasty are to achieve a normal genital appearance and to preserve sensation with a satisfactory sexual response. CASES: We present two cases of acquired clitoromegaly. Case 1 is a 46-year-old woman with clitoromegaly caused by an androgen-producing ovarian tumor. The second case is a 49-year-old woman with clitoromegaly resulting from a prolonged history of self-injected anabolic steroid use. Both women underwent a clitoral reduction procedure with preservation of the neurovascular supply to the glans clitoris. CONCLUSION: Clitoroplasty is an uncommon procedure that is useful in correcting acquired clitoromegaly. The results are cosmetically acceptable, and sexual function can be preserved. (Obstet Gynecol 2007;110:523–5) A cquired clitoromegaly is a relatively rare condition, and data in the literature are sparse.1 The etiologies of acquired clitoromegaly can be categorized as either hormonal or nonhormonal. In the hormonal causes, an androgen excess is the main contributing factor of the clitoral enlargement. Three groups should be distinguished within this group: endocrinopathies, androgen-producing tumors, or self-injection of long-acting synthetic androgens. The most important endocrinopathies are nonpolycystic ovarian hypertestosteronism and polycystic ovarian syndrome (PCOS). In PCOS, patients generally have lower serum androgen levels, with most testosterone levels less than 150 ng/dL, and dehydroepiandrosterone sulfate values are usually either normal or slightly elevated; therefore, clitoral enlargement is usually not very apparent.2 The more pronounced clitoral enlargement is observed in hormonally active, masculinizing tumors, such as Leydig cell tumors, From the Department of Gynecologic Oncology, University of South Florida, Tampa, Florida. Corresponding author: Robyn A. Sayer, MD, Harbourside Medical Tower, Department of Obstetrics and Gynecology, 4 Columbia Drive, Suite 500, Tampa, FL 33606; e-mail: rsayer@hsc.usf.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 lipid-cell tumors, or arrhenoblastomas. Virilizing cortical carcinoma or adenoma of the adrenal gland can also lead to clitoromegaly. Nonhormonal causes of acquired clitoral enlargement are even rarer, with the only reported cause being neurofibromatosis.3 Evaluation of acquired clitoral enlargement must take into consideration all of the above-mentioned etiological factors. Clitoroplasty, especially in an adult, is a rare procedure. The goals of clitoroplasty are to achieve a normal genital appearance and to preserve sensation with a satisfactory sexual response. We present two cases of adult women who underwent a clitoral reduction procedure. CASES Case 1 is a 46-year-old African-American woman who was referred for a 3-year history of increasing hair growth, deepening voice, and an enlarging clitoris causing significant depression and jeopardizing her marriage. Medical history was notable for obesity, diet-controlled diabetes, hypertension, and polycythemia. Total serum testosterone was 697. An MRI of the abdomen revealed normal adrenal glands. An ultrasound examination of the pelvis showed a normal uterus and a right ovary with ill-defined nodules. The patient was taken to surgery and underwent a laparoscopic bilateral salpingo-oophorectomy and a reduction clitoroplasty. Pathology examination revealed a 4-cm steroid-producing tumor confined to the right ovary. The postoperative recovery was without complication. The patient has been happy with the cosmetic outcome and sexual function following the procedure. Clitoral orgasm has been achieved without pain. Case 2 is a 49-year-old white woman who was referred for an enlarged clitoris. The patient is a former body-builder with a long history of synthetic anabolic steroid use, which she reportedly discontinued 2 years prior. Medical history is notable for human immunodeficiency virus (HIV) with an undetectable viral load on medications. A computed tomography scan of the abdomen and pelvis revealed normal adrenal glands and ovaries. A preoperative serum testosterone level was 14. The patient underwent a reduction clitoroplasty. Postoperative recovery was uneventful. The patient has been happy with the cosmetic result and has yet to resume sexual relations primarily due to a recent divorce and her HIV status. The procedure is performed with the patient under general anesthesia in the high dorsal lithotomy position. A Foley catheter is placed into the bladder before the procedure. The excess skin of the clitoral hood is marked for an incision which starts in the skin angle at the base of the clitoris and extends 1 cm posterior down either side of the hood. A second incision line is drawn starting 0.5 cm proximal to the glans on the clitoral hood and extends to intercept the first line at the base of the clitoris Sayer et al Clitoroplasty 523 Fig. 1. First and second incision line. The second incision line extends to intercept the first line at the base of the clitoris. Sayer. Clitoroplasty. Obstet Gynecol 2007. (Fig. 1). Incisions are made following the lines to remove the excess skin and expose the suspensory ligament of the clitoris and the corpora cavernosa (Fig. 2). The suspensory ligament is isolated, transected at its distal end, tagged, and retracted upwards for later reattachment. The corpora cavernosa are then carefully dissected from the ventral tissue which contains the neurovascular supply of the clitoris. During this dissection it is important to stay close to the corpora cavernosa to avoid the neurovascular tissue. The corpora are clamped next to Fig. 3. The corpora clamped next to the pubic bone and distally 5 cm from the prepuce of the clitoris. Sayer. Clitoroplasty. Obstet Gynecol 2007. the pubic bone and at a point distal, approximately 0.5 cm, from the prepuce of the clitoris (Fig. 3). The segment of corpora between the clamps is excised, and the pedicles are suture ligated. The distal end of the corpora is reattached to the periosteum of the pubis with 0 synthetic absorbable suture. The distal end of the suspensory ligament is trimmed and reattached to the proximal suspensory ligament pedicle using the same suture. The subcutaneous tissue along the clitoral hood is reapproxi- Fig. 2. The suspensory ligament of the clitoris (blue tag) and the corpora cavernosa (yellow tag) are exposed. The neurovascular tissue of the clitoris is indicated with an arrow. Fig. 4. Skin edges of the clitoral hood are reapproximated with absorbable sutures. Sayer. Clitoroplasty. Obstet Gynecol 2007. Sayer. Clitoroplasty. Obstet Gynecol 2007. 524 Sayer et al Clitoroplasty OBSTETRICS & GYNECOLOGY mated with interrupted 2-0 synthetic absorbable suture. The skin edges of the clitoral hood are reapproximated with a series of interrupted 3-0 synthetic absorbable sutures (Fig. 4). The Foley catheter remains in place until the following morning. The patient is discharged from the hospital on postoperative day one. She is seen in the office 2 weeks and 6 weeks postoperatively. Normal sexual activity may resume 6 weeks postoperatively. tion-preserving procedure described above was originally described by Graves in 1982, with a series of eight successful procedures performed over the course of 9 years.6 By using this technique of reduction clitoroplasty, acquired clitoromegaly in adult women can be corrected with excellent cosmetic results and preservation of sexual function. COMMENT REFERENCES A surgical method for the correction of clitoromegaly was first described in 1934 by Young,4 who performed a clitorectomy on a child with congenital adrenal hyperplasia. Until the 1960s, clitorectomy or clitoral amputation was widely accepted as standard treatment.5 Studies in the late 1960s demonstrated the importance of the clitoris in sexual function, leading to modifications in the procedure to preserve sexual function. These goals are accomplished by shortening the corpora cavernosa while preserving the neurovascular supply to the glans clitoris. This modern func- 1. Horejsi J. Acquired clitoral enlargement: diagnosis and treatment. Ann N Y Acad Sci 1997;816:369–72. HELLP Syndrome diagnosis of severe HELLP syndrome was supported by the findings of hemolysis, elevated liver enzymes, and severe thrombocytopenia, all of which resolved after termination of the pregnancy. A Rare, Early Presentation at 17 Weeks of Gestation Eran Bornstein, MD, Yoni Barnhard, MD, Russell Atkin, MD, and Michael Y. Divon, MD 2. Sheehan MT. Polycystic ovarian syndrome: diagnosis and management. Clin Med Res 2004;2:13–27. 3. Sutphen R, Galan-Gomez E, Kousseff BG. Clitoromegaly in neurofibromatosis. Am J Med Genet 1995;55:325–30. 4. Young HH. Genital abnormalities, hermaphroditism and related adrenal disease. Baltimore (MD): Williams and Wilkins; 1937. p. 103–5. 5. Allen LE, Hardy BE, Churchill BM. The surgical management of the enlarged clitoris. J Urol 1982;128:351–4. 6. Graves KL, Wilson EA, Greene JW. Surgical technique for clitoral reduction. Obstet Gynecol 1982;59:758–61. CONCLUSION: Clinicians should consider the diagnosis of HELLP syndrome in women presenting with clinical manifestations or laboratory abnormalities consistent with this diagnosis, even before 20 weeks of gestation. (Obstet Gynecol 2007;110:525–7) BACKGROUND: The occurrence of hemolysis, elevated liver enzymes, low platelets (HELLP) syndrome before 20 weeks of gestation is extremely rare. This condition has been reported in only few cases, and always in conjunction with other comorbidities such as fetal triploidy and the antiphospholipid syndrome. CASE: A 24-year-old woman was admitted at 15 weeks and 3 days of gestation with uncontrollable hypertension and proteinuria. An extensive workup did not reveal any underlying medical or obstetric conditions. Within 11 days of admission her condition deteriorated and the From the Department of Obstetrics and Gynecology, Lenox-Hill Hospital, New York, New York. Corresponding author: Eran Bornstein, MD, Department of Obstetrics and Gynecology, Lenox-Hill Hospital, 100 E. 77th Street, New York, NY 10021; e-mail: eranbor@yahoo.com. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 T he syndrome of hemolysis, elevated liver enzymes, and a low platelet count (HELLP syndrome) is a known severe complication of preeclampsia-eclampsia. This syndrome is associated with an increased risk of both maternal and fetal adverse outcomes, and its management is often challenging. Most often, HELLP syndrome occurs in the third trimester of pregnancy and, less frequently, in the late second trimester or in the postpartum period. A few case reports describe the rare diagnosis of this syndrome before 20 weeks of gestation. However, in all previous reports, the pregnancy was complicated by severe underlying medical or obstetric conditions such as fetal triploidy or the antiphospholipid syndrome.1– 4 We report an unusual case of preeclampsia that presented at 15 weeks and 3 days of gestation with uncontrollable hypertension and proteinuria that progressed to severe HELLP syndrome within 11 Bornstein et al Early Presentation of HELLP Syndrome 525 days in a patient with no known underlying medical or obstetric complications. CASE A 24-year-old African-American woman (gravida4 para0030) was admitted to our antenatal unit secondary to extremely elevated blood pressure (BP) at 15 weeks and 3 days of gestation, which was detected during her second prenatal visit. The pregnancy was dated by her last menstrual period, which was consistent with firsttrimester sonographic measurements. Her baseline blood pressure at 11 weeks of gestation was 120/80 mm Hg. Upon admission, the patient’s systolic and diastolic BPs ranged from 160 to 200 mm Hg and from 91 to 114 mm Hg, respectively. She reported a severe headache. She denied any nausea, vomiting, right upper quadrant or epigastric pain, visual disturbances or scotoma, chest pain, or palpitations. The patient denied any history of hypertension before this event or any other medical conditions associated with hypertension. She reported a history of three elective, early first-trimester terminations in which no abnormal blood pressures were documented. The present pregnancy was conceived with a new partner. Her family history was significant for chronic hypertension affecting both her parents and her 36-year-old sister. The patient denied ever using antihypertensive medications as well as smoking or substance abuse. The physical examination on admission revealed an afebrile, hypertensive, morbidly obese (body mass index 38.9) gravida. Her examination was otherwise normal, with no evidence of peripheral edema or any neurological deficits and normal deep tendon reflexes. The initial laboratory evaluation revealed hemoglobin 11.4 g/dL, platelet count 271,000, creatinine level 0.6 mg/dL, lactate dehydrogenase (LDH) 594 Units/L, uric acid 4.2 mg/dL, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) of 19 Units/L and 31 Units/L, respectively, and a urinalysis with ⫹2 proteinuria. The patient was admitted for evaluation and control of hypertension. Over the ensuing 11 days diligent management by a multidisciplinary team (including maternal–fetal medicine, nephrology, and intensive care) using multidrug regimens (including labetalol, hydralazine, nifedipine, and methyldopa) failed to achieve adequate BP control. Her BP remained extremely labile, ranging from 140 to 210 mm Hg systolic and 90 to140 mm Hg diastolic. A thorough investigation included a 24-hour urine collection, serum catecholamine levels, urine metanephrines, plasma renin aldosterone activity, thyroid profile, lipid profile, a connective tissue disease workup, and a complete thrombophilia workup, as well as assessment for renal artery stenosis. All of the above tests were normal except for mild proteinuria (446 mg of protein/24 hour). At 16 4/7 weeks of gestation a gradual decrease in the platelet count, to a level of 128,000, was observed. This decrease progressed abruptly to severe thrombocytope- 526 Bornstein et al Early Presentation of HELLP Syndrome nia (platelet count of 21,000) by 17 weeks of gestation. At that time the patients was afebrile, her BP was still extremely labile, significant pedal edema was noticed, she became oliguric, although she did not develop renal failure, and no neurological deficits were detected. A peripheral blood smear demonstrated schistocytes (a classic finding of microangiopathic hemolysis), and the metabolic panel detected several abnormal values (LDH of 945 Units/L, elevated liver enzymes, with ALT of 99 Units/L and AST of 58 Units/L, respectively, uric acid of 7.0 mg/dL, and creatinine level of 0.9 mg/dL). A random urinalysis sample detected ⫹4 proteinuria. Although an extensive differential diagnosis was considered, including thrombotic thrombocytopenic purpura, hemolytic uremic syndrome, acute exacerbation of systemic lupus erythematosus, and paroxysmal nocturnal hemoglobinuria, the diagnosis of HELLP syndrome was made, and the patient was counseled appropriately. After discussion with her family, the patient elected to terminate the pregnancy. Seizure prophylaxis with magnesium sulfate was administered, as well as dexamethasone 10 mg intramuscularly for management of the severe thrombocytopenia. A successful dilatation and evacuation was performed after maternal stabilization, which necessitated transfusion of platelets. The patient improved rapidly, with normalization of her platelet count on postoperative day 4 (to 166,000). Additional work-up revealed mild bilateral hypertensive retinopathy, and an echocardiogram demonstrated left ventricular hypertrophy, both supporting the postpartum diagnosis of chronic hypertension with end-organ damage complicated by superimposed preeclampsia with HELLP syndrome. Her blood pressure was finally controlled with methyldopa, enalapril, and metoprolol, and she was discharged home on postoperative day 6 with a BP of 160/86 mm Hg. The patient was counseled regarding a markedly increased risk of preeclampsia/HELLP syndrome in subsequent pregnancies.6 Additionally, she was informed of an increased risk for other adverse pregnancy outcomes (such as preterm delivery, fetal growth restriction, abruptio placenta, and fetal death) in subsequent pregnancies. Ongoing follow-up with a nephrologist excluded known secondary causes of hypertension. Genetic studies of the fetus revealed a normal male fetal karyotype (46XY). COMMENT Hemolysis, elevated liver enzymes, low platelets syndrome was named by Dr. Luis Weinstein, who in 1982 described a case series in which patients with severe preeclampsia-eclampsia also demonstrated thrombocytopenia, abnormal peripheral blood smear, and abnormal hepatic function test results.5 This constellation of abnormal laboratory findings usually presents after 20 weeks of gestation and is often accompanied by the findings of new onset hypertension and proteinuria in previously normo- OBSTETRICS & GYNECOLOGY tensive and nonproteinuric women. The syndrome is associated with life threatening maternal and fetal complications, and therefore, management and optimization of maternal and fetal well-being is challenging. Maternal mortality is about 1%, and the most common severe maternal morbidities are disseminated intravascular coagulopathy, placental abruption, pulmonary edema, acute renal failure, and liver hemorrhage.6 Only about 15% of cases complicated by HELLP syndrome occur between 24 and 28 weeks of gestation, and its presence before 24 weeks of gestation is extremely rare. Hemolysis, elevated liver enzymes, low platelets syndrome before 20 weeks of gestation has been reported in a few cases and always in conjunction with other comorbidities such as fetal triploidy and the antiphospholipid syndrome.1– 4 A PubMed review of the English literature from January 1982 through March 2007 for the terms “preeclampsia,” “HELLP syndrome,” “prematurity,” “second/ first trimester,” and “preterm” did not reveal any cases of “pure” HELLP syndrome before 20 weeks of gestation. Our patient presented at 15 weeks of gestation with findings consistent with mild preeclampsia, which progressed to severe HELLP syndrome by 17 weeks of gestation. Meticulous investigation ruled out other comorbidities associated with early-onset HELLP syndrome or preeclampsia.1– 4,7 In fact, the only known risk factors our patient had were obesity, ethnicity, family history of hypertension, and undiagnosed primary chronic hypertension. It is likely that the undiagnosed chronic hypertension was an essential factor predisposing her to such an early manifestation of HELLP syndrome. Several obstetric and medical conditions may present with similar clinical manifestations and laboratory results as HELLP syndrome creating a confusing and difficult diagnostic challenge. As described by Sibai8 recently, these “imitators” include acute fatty liver of pregnancy, thrombotic thrombocytopenic purpura, hemolytic uremic syndrome, acute exacerbation of systemic lupus ery- VOL. 110, NO. 2, PART 2, AUGUST 2007 thematosus, and paroxysmal nocturnal hemoglobinuria. Although these conditions share a similar spectrum of signs, symptoms, and laboratory results with HELLP syndrome, each has some distinguishing clinical or laboratory findings. An interdisciplinary approach may be helpful in obtaining the correct diagnosis and therefore optimizing the outcome in such complex cases. Early presentation, such as in our case, may result in severe maternal morbidity and mortality especially if HELLP syndrome is not suspected. Awareness, intensive maternal monitoring, and early intervention may be life saving. Therefore, our experience should guide obstetricians in considering the diagnosis of HELLP syndrome in women presenting with clinical manifestations or laboratory abnormalities consistent with this condition even before 20 weeks of gestation. REFERENCES 1. Sherer DM, Dalloul M, Stimphil R, Hellmann M, KhouryCollado F, Osho J, et al. Acute onset of severe hemolysis, elevated liver enzymes, and low platelet count syndrome in a patient with a partial hydatidiform mole at 17 weeks gestation. Am J Perinatol 2006;23:163–6. 2. Stefos T, Plachouras N, Mari G, Cosmi E, Lolis D. A case of partial mole and atypical type 1 triploidy associated with severe HELLP syndrome at 18 weeks’ gestation. Ultrasound Obstet Gynecol 2002;20:403–4. 3. Haram K, Trovik J, Sandset PM, Hordnes K. Severe syndrome of hemolysis, elevated liver enzymes and low platelets (HELLP) in the 18th week of pregnancy associated with the antiphospholipid-antibody syndrome. Acta Obstet Gynecol Scand 2003;82:679–80. 4. McMahon LP, Smith J. The HELLP syndrome at 16 weeks gestation: possible association with the antiphospholipid syndrome. Aust N Z J Obstet Gynecol 1997;37:313–4. 5. Weinstein L. Syndrome of hemolysis, elevated liver enzymes, and low platelet count: a severe consequence of hypertension in pregnancy. Am J Obstet Gynecol 1982;142:159–67. 6. Sibai BM. Diagnosis, controversies, and management of the syndrome of hemolysis, elevated liver enzymes, and low platelet count. Obstet Gynecol 2004;103:981–91. 7. Alfadda A, Tamilia M. Preeclampsia-like syndrome that is associated with severe hypothyroidism in a 20-week pregnant woman. Am J Obstet Gynecol 2004;191:1723–4. 8. Sibai BM. Imitators of severe preeclampsia. Obstet Gynecol 2007;109:956–66. Bornstein et al Early Presentation of HELLP Syndrome 527 Secondary Amenorrhea Resulting From Traumatic Separation of the Cervix From the Uterine Corpus Joshua Kesterson, MD, Jennifer Dietrich, MD, Marvin Yussman, MD, and S. Paige Hertweck, MD BACKGROUND: Amenorrhea resulting from crushing trauma of the pelvis is exceptionally rare. The purpose of this case report is to describe the diagnosis of and successful surgical correction of traumatic separation of the cervix from the uterine corpus. CASE: A nulligravida presented with a complaint of secondary amenorrhea after a motor vehicle accident in which she sustained a crush-type injury to the pelvis. Ultrasonography and laparotomy revealed a complete separation of the uterine corpus from the cervix. The uterine corpus was approximated to the cervix with circumferentially placed sutures to establish a patent outflow tract from the endometrial cavity to the cervical canal. CONCLUSION: This case demonstrates the successful surgical correction of secondary amenorrhea resulting from traumatic separation of the uterine corpus from the cervix. Normal menstruation resumed 6 weeks after surgery. (Obstet Gynecol 2007;110:528–30) E xcluding pregnancy, secondary amenorrhea is most commonly due to a hormonal abnormality. In their series of 262 patients with adult onset amenorrhea, Reindollar et al1 reported the most frequent etiologies as being hypothalamic suppression, hyperprolactinemia, and ovarian failure, respectively. Amenorrhea resulting from trauma to the pelvis occurs rarely, with the available literature limited to a few case reports.2,3 In their study of traumatic reproductive tract injuries in women, Fallat et al4 found that a majority of noncoital injuries resulted from vehicular trauma and that late sequelae were infrequent. This case report stresses the importance of a detailed history in diagnosing an anatomic cause of secondary amenorrhea. It also describes the surgical correction of traumatic avulsion of the uterine From the Department of Obstetrics, Gynecology, and Women’s Health, University of Louisville, Louisville, Kentucky. Corresponding author: Joshua Kesterson, MD, Department of Gynecologic Oncology, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263; e-mail: joshuapkesterson@yahoo.com. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 528 Kesterson et al Amenorrhea After Trauma corpus from the cervix with restoration of the genital outflow tract. CASE A 15-year-old nulligravida presented to the gynecology clinic with a history of secondary amenorrhea for the past 2 years. She gave a history of menarche at 11 years of age with subsequent monthly menses. Although now amenorrheic, the patient continued to experience cyclical abdominal pain and molimina. On examination, she had normal secondary sexual characteristics with Tanner 5 breasts and Tanner 5 external genitalia. The vagina and vulva were normal. The cervix was palpable and the uterus noted to be small and deviated slightly to the left of midline. A detailed history revealed that the patient had been involved in a motor vehicle accident 2 years previously that resulted in a crush injury to the pelvis. There were no intraabdominal abnormalities noted at time of exploratory laparotomy. She was diagnosed with both a posterior and anterior pelvic ring fracture, for which she underwent open reduction with internal fixation of her posterior iliac fracture and external fixation of the pelvis. The patient made a full recovery. However, she failed to resume menses. Serum levels of human chorionic gonadotropin (hCG), thyroid-stimulating hormone (TSH), luteinizing hormone (LH), follicle-stimulating hormone (FSH), and prolactin were within normal limits. A magnetic resonance imaging (MRI) scan of the head and pelvis was ordered at another clinic before the patient’s consultation in our gynecology clinic. The MRI of the brain did not show evidence of a suprasellar lesion or disruption of the pituitary stalk. An attempt at hysterosalpingography was unsuccessful, secondary to the inability to cannulate the cervical canal. The MRI of the pelvis failed to show a communication between the uterus and cervix. A pelvic ultrasound examination showed the corpus to be avulsed from the cervix, with the lower uterine segment 1.7 cm from the internal cervical os (Fig. 1). At laparotomy, the cervix was noted to be completely separated from the uterine corpus. The right broad ligament was lacerated and the right uterine artery obliterated. The intact bladder had been reflected inferiorly. Ovaries and fallopian tubes were normal. There was no grossly visible endometriosis. The repair was initiated by making a vertical incision through the anterior uterine corpus, exposing the endometrial cavity. The scarred lower segment of the uterine corpus and the scarred superior segment of the separated cervix were excised to provide fresh surfaces for anastomosis. A uterine sound was passed vaginally through the cervix into the uterine cavity. Once this communication was established, a latex T-tube was positioned (Fig. 2), extending from the fundal portion of the endometrial canal through the cervix and into the vaginal vault. The cervix was then approximated to the uterus by placing stay sutures that fixed the cervix to the fundus at the 3:00 and 9:00 positions. This was followed by two layers of circumferentially placed sutures in the manner of a Tompkins metroplasty. The remaining vertical defect in the uterus was closed in two layers with interrupted stitches. The OBSTETRICS & GYNECOLOGY COMMENT Fig. 1. Pelvic ultrasonography showing the uterine corpus separate from the cervix. A, bladder; B, uterus; C, endometrial canal; D, cervical canal; arrows, space between inferior limit of endometrial canal and superior aspect of cervical canal. Kesterson. Amenorrhea After Trauma. Obstet Gynecol 2007. Fig. 2. Photograph showing the T-tube with its superior end within the fundal portion of endometrial cavity, exiting the lower segment of the uterine corpus, and entering the endometrial canal. A, uterus; B, T-tube. Kesterson. Amenorrhea After Trauma. Obstet Gynecol 2007. serosa was then approximated with a running suture. A good anatomical result was achieved. The patient did well postoperatively. She was discharged home on an estradiol transdermal patch and doxycycline. Two weeks later, at time of diagnostic laparoscopy, the fallopian tubes were noted to be lying directly on the hysterotomy repair site, attached by filmy adhesions. Blunt adhesiolysis was performed, restoring the fallopian tubes to their normal anatomic position. The surgical sites appeared to be healing well, and there were otherwise no abnormalities visualized within the pelvis or abdomen. Normal menstruation occurred approximately 6 weeks after surgical repair. After two normal cycles, the T-tube was removed, and hysteroscopy was performed. A normal endometrial cavity was observed. The patient continues to have normal monthly menstruation. VOL. 110, NO. 2, PART 2, AUGUST 2007 The bony architecture of the pelvis normally provides a protective basket for the female reproductive organs, making injury to these organs rare. In their 5-year review of 220 exploratory laparotomies performed for blunt abdominal trauma, Stone et al5 reported only 15 gynecologic injuries. A majority of these involved ovarian cysts, and there was a single patient who sustained uterine and vaginal lacerations. Niemi and Norton6 reported their 10-year experience involving 114 female patients with pelvic fractures in which only four patients sustained vaginal lacerations. Several mechanisms can account for compromise of the reproductive organs with a pelvic fracture, including a shearing of the lateral attachments of the uterus and cervix as well as a break in the pelvic ring resulting in a crush or guillotine-type action.7 Considering this patient’s history and intraoperative findings of uterine avulsion and a lacerated right broad ligament, the most likely mechanism of action of her injury was a compression fracture of the pelvic ring resulting in a vertical shearing force sufficient to cause her documented injuries. There was no evidence of hematometra or endometriosis at the time of operation. This is consistent with the findings reported by Toaff and Ballas.8 They described 31 cases of amenorrhea resulting from traumatic damage to, and subsequent occlusion of, the cervico-isthmic area. There were no cases of hematometra if the occlusion was limited to the level of the uterocervical outflow tract. They proposed that this type of lesion causes the endometrium to be refractory to hormonal stimuli, possibly due to a visceral reflex which may originate in the isthmic portion of the uterus. Our ability to counsel this patient regarding her future reproductive options has been limited secondary to a paucity of similar cases. Of note, in the two documented cases involving women who experienced traumatic injuries and repairs similar to those of our patient, both subsequently carried pregnancies to term. One patient became pregnant several years after her original reconstruction and underwent an elective cesarean delivery at term.3 She subsequently had two more successful pregnancies that also ended by cesarean delivery. The other patient became pregnant approximately 14 years after the surgical restoration of her uterine outflow tract, with a single cycle of in vitro fertilization performed because of tubal factor infertility.2 She underwent an elective cesarean delivery at term. Our patient has been counseled regarding the importance of carefully planning any desired pregnancy in the future and the need for cesarean delivery Kesterson et al Amenorrhea After Trauma 529 3. Murphy D, Bonnar J. An unusual case of traumatic primary amenorrhea. Br J Obstet Gynecol 1993;100:784–5. 4. Fallat M, Weaver J, Hertweck S, Miller F. Late follow-up and functional outcome after traumatic reproductive tract injuries in women. Am Surg 1998;64:858–61. 5. Stone N, Ances I, Brotman S. Gynecologic injury in the nongravid female during blunt abdominal trauma. J Trauma 1984;24:626–7. 6. Niemi T, Norton L. Vaginal injuries in patients with pelvic fractures. J Trauma 1985;25:547–51. 7. Smith R. Avulsion of the nongravid uterus due to pelvic fracture. South Med J 1989;82:70–3. 8. Toaff R, Ballas S. Traumatic hypomenorrhea-amenorrhea (Asherman’s syndrome). Fertil Steril 1978;30:379–87. before the onset of labor. Because of her uterine scar she is at increased risk for uterine rupture, placenta previa, and placenta accreta. This patient presents an unusual case of secondary amenorrhea resulting from crush trauma to the pelvis REFERENCES 1. Reindollar R, Novak M, Tho S, McDonough P. Adult-onset amenorrhea: a study of 262 patients. Am J Obstet Gynecol 1986;155:531–43. 2. Donner G, Pel M, Lammes F. Primary amenorrhea caused by crushing trauma of the pelvis. Am J Obstet Gynecol 2000;183: 500–1. Radical Excision of Inguinal Endometriosis Luigi Fedele, MD, Stefano Bianchi, MD, Giada Frontino, MD, Giovanni Zanconato, MD, and Tommaso Rubino, MD BACKGROUND: The diagnosis of inguinal endometriosis can be complex, and patients are often first operated by a general surgeon for a hernia. We present five cases of inguinal endometriosis in which primary surgery resulted in recurrence and a second correct procedure resulted in a cure. CASES: Five patients with inguinal endometriosis, operated between 1996 and 2002, were seen for the recurrence of symptoms. All underwent excision of the extraperitoneal portion of the round ligament and pelvic exploration. In all cases, both pelvic and round ligament endometriosic lesions were confirmed at histology. No complications or recurrence of inguinal endometriosis occurred. CONCLUSION: The complete excision of inguinal endometriosis must also include the extraperitoneal portion of the round ligament. (Obstet Gynecol 2007;110:530–3) From the Department of Obstetrics, Gynecology and Neonatology, Fondazione “Policlinico-Mangiagalli-Regina Elena,” University of Milan, Milan, Italy; and Department of Obstetrics and Gynecology, University of Verona, Ospedale Borgoroma, Verona, Italy. The authors thank Nevio Ciappina, MD, for data gathering and photographic contributions. Corresponding author: Luigi Fedele, MD, Clinica Ostetrica e Ginecologica II, Università di Milano, Istituto Luigi Mangiagalli, Via della Commenda 12, 20122 Milano, Italy; e-mail: luigi.fedele@unimi.it. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 530 Fedele et al I nguinal endometriosis is an extremely rare condition1–5 that is characterized by the presence of endometrial stroma and glands in the extraperitoneal portion of the round ligament and in the surrounding connective and lymphatic tissues. The diagnosis of this variant of extragenital endometriosis is seldom straightforward and the patient is frequently brought by the general surgeon to the operating table with a diagnosis of incarcerated hernia. Catamenial pain is usually the first appearing pathognomic symptom. Magnetic resonance imaging (MRI) helps in the differential diagnosis of a tender inguinal swelling. Although a medical treatment can be considered in premenopause, the treatment of choice in younger patients should be that of excisional surgery to avoid symptom and lesion relapse.3,4,6 Our study presents five cases of inguinal endometriosis in which a first inadequate treatment resulted in a recurrence of the symptoms and lesions, whereas a correct excisional procedure was followed by a complete and long-standing resolution of symptoms. CASES The patients assessed in the present study are all the women (n⫽5) with inguinal endometriosis who have been operated by the first author (L.F.) between 1996 and 2002. These patients were referred for the recurrence of symptoms and physical findings associated with inguinal endometriosis to the Centre for Endometriosis of the University of Verona, Italy, between 1996 and 2000 and to the Centre for Endometriosis of the University of Milan, Italy, between 2000 and 2002. This study was approved by the Consiglio di Istituto at the University of Milan, Italy. The patients were assessed preoperatively with a transabdominal and transvaginal ultrasound examination and MRI. All subjects underwent a repeat surgery in the inguinal region and a laparoscopic or laparotomic exploration of the pelvis. Patient 2 also underwent a laparotomic bilateral adnexectomy and hysterectomy. At a Radical Surgery for Inguinal Endometriosis OBSTETRICS & GYNECOLOGY variable distance from the first-line inguinal surgery, the patients (Table 1) had a relapse of a painful mass in the inguinal region, with worsening of pain symptoms and catamenial onset of skin reddening and edema. Four of the five patients had been first operated by a general surgeon for a suspected incarcerated inguinal hernia (n⫽3) or for primary or secondary lymphoadenopathy (n⫽1). Only one patient with a preoperative diagnosis of inguinal endometriosis had undergone a first surgical procedure by the gynecologist. Excision of the subcutaneous tissue had been apparently radical in the former four cases, and neither opening of the inguinal canal nor identification of the round ligament was performed. Patient 5 had undergone removal of the subcutaneous lesions as well as of the superficial and deep inguinal lymph nodes. The surgical approach we used consists of the following steps: an oblique cutaneous incision is performed along the line running from the pubic tubercle to the anterior superior iliac spine, at the superior margin of the inguinal ligament for a brief portion that is proportional to the size of the lesion. The subcutaneous tissue is dissected under a visual and palpatory guide, until the endometriosic lesions are identified and isolated. These generally have a fibrous consistency with hemorrhagic extravasations and are sometimes bluish in appearance. The lesions are removed, and the aponeurosis of the external oblique muscle and the anterior rectus fascia are identified, along with the external inguinal ring and the terminal portion of the round ligament which inserts on the pubic tubercle. The round ligament is cut at this level, and the pubic periosteum is carefully cleaned of the endometriosic tissue (Fig. 1). The inguinal canal is opened and the extraperitoneal portion of the round ligament is removed by applying traction to it and then cutting it at the level of the internal inguinal ring. The proximal stump is fixed to the fibrous structure of the inguinal canal. The superficial and deep inguinal lymph nodes are then inspected by palpation and removed if suspected to be positive. A small net of nonresorbable material is placed in the inguinal canal, after which the abdominal wall is closed. All patients had pelvic endometriosis that was treated endoscopically. Follow-up lasted from a minimum of 5 to a maximum of 12 years and consisted of semestral clinical examinations. Two patients took a monophasic oral contraceptive for 2 years, while another patient had a term pregnancy 2 years after surgery. Patient number 2 has initiated transdermal hormonal replacement therapy at age 52. Endometriosis of the round ligament was present in all cases and histologic examination demonstrated smoothmuscle fibers infiltrated by endometrial glandular structures and/or microcystic formations covered by simple columnar epithelium and surrounded by stroma containing hemosiderin. None of the patients had intra- or postoperative complications. There was no clinical evidence of recur- VOL. 110, NO. 2, PART 2, AUGUST 2007 rence of the inguinal endometriosic lesions in all five patients during the follow-up period. COMMENT Our observational study demonstrates that, to obtain a radical excision of inguinal endometriosis, the surgical procedure must always include the removal of the extraperitoneal portion of the round ligament. Such procedure must involve an adequate resection to avoid symptom recurrence and persistence of lesions, which if left in place may also undergo oncogenic transformation.7 Although our case series is not a large one due to the rarity of these lesions, it is characterized by a long follow-up, a homogeneous second-line surgical treatment, and a systematic exploration of the abdominal cavity. The thorough pelvic exploration was in all cases important to staging the presence of pelvic endometriosis. Pelvic endometriosis was indeed associated with almost all cases of inguinal endometriosis found in the literature in which surgical exploration of the abdomen was performed. Only Quagliarello et al,2 reporting a case of isolated endometriosis in an inguinal hernia, and Moore et al,8 showing that one case out of five had inguinal endometriosis, did not have any evidence of concurrent pelvic localizations. Pelvic endometriosis was also demonstrated in all of six cases of inguinal endometriosis in our study of 1991.6 Such nearly constant coexistence of pelvic endometriosis in all cases described also constitutes the basis of the pathogenetic theory that is currently most recognized for the pathogenesis of inguinal endometriosis. The endometrial cells that are dispersed in the peritoneal fluid in the internal inguinal ring might cross the peritoneal mesothelium into the lymphatic and venous vessels of the round ligament. In the literature the histologic examination showed the presence of endometriosis on the extraperitoneal portion of the round ligament in all cases in which this was examined.2– 6,8 This finding, along with the recognition of the mechanisms of implantation, have brought us to believe that the extraperitoneal excision of the entire portion of round ligament, from the internal inguinal ring to the terminal insertion at the pubic tubercle, is an essential element of the surgical procedure to avoid recurrence. A strong prevalence of lesions on the right side was shown, both in our case series and in those published in the literature. This feature seems to be due to the clockwise circulation of the peritoneal fluid and to the presence of the sigmoid colon that shields the left inguinal ring. A correct preoperative diagnosis Fedele et al Radical Surgery for Inguinal Endometriosis 531 532 Fedele et al Radical Surgery for Inguinal Endometriosis OBSTETRICS & GYNECOLOGY LPS: normal pelvic findings 12 Excision of the subcutaneous lesion (twice) Right inguinal multinodular mass 28 35 Right Nulligravid Catamenial right inguinal pain; pain at flexion of right thigh Patient 4 LPT: pelvic endometriosis LPS: pelvic endometriosis LPS: pelvic endometriosis 4th stage* 3rd stage* 4th stage* 11 5 5 Excision of the subcutaneous lesion Simple excision of the Simple excision of the subcutaneous lesion subcutaneous lesion 28 Right Nulligravid Catamenial inguinal pain Relapse of mass in the inguinal region 36 43 Left Nulligravid Inguinal pain Patient 3 Small tender and deep inguinal nodule 4 40 Right G4, 2-0-1-2 Catamenial inguinal pain with paresthesia Non-tender inguinal mass 8 Patient 2 Radical pelvic superficial and deep lymphadenectomy LPS: endometriosis 2nd stage* 10 Nodule of the inguinal scar 14 34 Right Nulligravid Catamenial inguinal pain Patient 5 LPS, laparoscopy; LPT, laparotomy. * American Society for Reproductive Medicine: Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil Steril 1997;67:817–21. Follow-up (y) Intraabdominal findings Interval between first and second surgery (mo) Previous surgery Physical findings Age at second surgery (y) Side of inguinal endometriosis Gravidity and parity Presenting complaints Patient 1 Table 1. Clinical and Surgical Characteristics of Five Patients With Inguinal Endometriosis metriosis requires an increased awareness by gynecologists and especially by general surgeons, to whom patients may frequently present for a primary evaluation. Finally, to be truly radical, the surgical excision has to include the entire extraperitoneal portion of the round ligament. REFERENCES 1. Sataloff DM, LaVorgna KA, McFarland MM. Extrapelvic endometriosis presenting as a hernia: clinical reports and review of the literature. Surgery 1989;105:109–12. 2. Quagliarello J, Coppa G, Bigelow B. Isolated endometriosis in an inguinal hernia. Am J Obstet Gynecol 1985;152:688–9. 3. Kapan M, Kapan S, Durgun AV, Goksoy E. Inguinal endometriosis. Arch Gynecol Obstet 2005;271:76–8. Fig. 1. Extraperitoneal excision of inguinal endometriosis. The pubic periosteum is cleaned of the endometriosic tissue, and the round ligament is excised along with the endometriosic tissue. Arrow shows the inguinal ligament. Fedele. Radical Surgery for Inguinal Endometriosis. Obstet Gynecol 2007. 4. Licheri S, Pisano G, Erdas E, Ledda S, Casu B, Cherchi MV, et al. Endometriosis of the round ligament: description of a clinical case and review of the literature. Hernia 2005;9:294–7. 5. Terada S, Miyata Y, Nakazawa H, Higashimori T, Arai T, Kikuchi Y, et al. Immunohistochemical analysis of an ectopic endometriosis in the uterine round ligament. Diagn Pathol 2006;1:27–31. can be found in the literature only in a few cases.1,4 The preoperative suspicion is most often incarcerated hernia, lymphadenitis, or hydrocele of the inguinal canal, while the diagnosis is usually made only after the histopathologic result. Our experience suggests two considerations. First, an appropriate and timely diagnosis of inguinal endo- 6. Candiani GB, Vercellini P, Fedele L, Vendola N, Carinelli S, Scaglione V. Inguinal endometriosis: pathogenetic and clinical implications. Obstet Gynecol 1991;78:191–4. Aortic Puncture With a Laparoscopic Fascial Closure Device Veress needle or the trocar. We report an aortic puncture from the use of a laparoscopic fascial closure device. Gene Lee, MD, Alexander Nguyen, MD, Seth Kivnick, MD, and John R. Marshall, MD BACKGROUND: Vascular injury during laparoscopic surgery typically occurs with the blind insertion of either the From the Departments of Obstetrics and Gynecology and General Surgery, Harbor–University of California Los Angeles, Torrance, California; and Department of Obstetrics and Gynecology, Kaiser-Permanente West Los Angeles, Los Angeles, California. Corresponding author: Gene Lee, MD, Harbor Mail Box #3, Harbor-UCLA Medical Center, 1000 W. Carson Street, Torrance, CA 90509-2910; e-mail: glee@obgyn.humc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 7. Milam MR, Atkinson JB, Currie JL. Adenosarcoma arising in inguinal endometriosis. Obstet Gynecol 2006;108:753–5. 8. Moore JB, Binstock MA, Growdon WA. The clinical implications of retroperitoneal endometriosis. Am J Obstet Gynecol 1988;158:1291–8. CASE: After a laparoscopic-assisted vaginal hysterectomy and during closure of the 10-mm umbilical port with a fascial closure device, attempted passage of the needle encountered unusual resistance in the abdominal wall. Application of increased force resulted in uncontrolled entry of the needle into the abdominal cavity and a 1-mm puncture of the aorta. After emergency laparotomy the puncture was successfully repaired by a vascular surgeon. CONCLUSION: A previously unreported complication of this laparoscopic fascial closure device is aortic injury. Unusual resistance to passage of the needle should engender extra caution. The use of assistive closure devices should be reserved for patients with difficult anatomy. (Obstet Gynecol 2007;110:533–5) A ortic injury is rare in laparoscopic surgery, where the combined incidence of major vascular injury is 0.03– 0.07%.1 The Manufacturer and User Facility Lee et al Aortic Puncture With a Laparoscopic Device 533 Device Experience (MAUDE) database includes no reports of complications related to the device. A MEDLINE search in English from 1966 to March 2007 using the terms “Carter-Thomason” or “aortic puncture and laparoscopy” also reported no complications of this type. Vascular injury during laparoscopic surgery typically occurs with the blind insertion of either the Veress needle or the trocar. We report an aortic puncture occurring as a complication of using the Carter-Thomason (Inlet Medical Inc, Trumball, CT) laparoscopic fascial closure device (Fig. 1). hematoma was seen to expand, an aortic injury was suspected. A vascular surgeon was called, and the abdomen opened through a midline incision. The hematoma was easily visualized; bleeding was controlled by direct pressure. The vascular surgeon identified a 1-mm anterior aortic wall puncture, which was repaired with a 5-0 Prolene “U” stitch; blood loss was minimal. The postoperative course was uneventful. The patient’s postoperative hemoglobin concentration nadir was 7.8 (preoperative 9.8) and remained stable. She was discharged on postoperative day 3. CASE Previous reports describe puncture of the aorta, inferior vena cava, and common iliac arteries and veins, all during blind insertion of the Veress needle or trocar.1,2 In all cases, immediate recognition and repair was necessary to avoid morbidity and mortality. Because the incidence of bowel herniation through 10- to 12-mm trocar wounds is 0.5–3%,3 fascia of all 10- to 12-mm trocar wounds needs to be closed.3 Bowel herniations through 5-mm trocar wounds are rare and appear to occur after lengthy procedures.4 The location of the port does not influence herniation. Several instruments specially designed to facilitate fascial closure of port sites have been described.5,6 All provide direct visualization and permit full thickness closure of the trocar wound. They are especially useful in obese patients where fascial closure can be particularly difficult. The Carter-Thomason device is widely used. In 2004, it received the Innovative Product of the Year Award from the Society of Laparoendoscopic Surgeons. A randomized, but very small, study of 32 patients compared eight different techniques and reported that closure using the Carter-Thomason device was faster and did not appear to have any difference in complications.6 According to Carter, there were no reports of failure to close the fascia in over 4,000 closures occurring over 4 years.7 The Carter-Thomason device facilitates fascial closure after laparoscopy. However, the exposed needle tip creates the potential for intraabdominal injury. Unusual resistance was noted during use of the Carter-Thomason device in this case, and the associated force translated into an uncontrolled entry of the needle tip into the abdominal cavity with subsequent aortic puncture. Our experience suggests three lessons. First, if the patient is thin and the fascia can be closed easily under direct visualization without use of any device, Our patient was a 41-year-old woman (gravida 4, para 4) with persistent cervical dysplasia. She had no other significant medical or surgical history. Because of the dysplasia, she underwent a laparoscopic-assisted vaginal hysterectomy. The uterus was 6 cm in length and was associated with minimal descensus. Before the closure of the anterior abdominal wall, there were no complications. The estimated blood loss was 300 mL. Closure of the 10-mm umbilical port site was started with the Carter-Thomason device. Closure was monitored through a 5-mm scope placed in one of the lateral ports. After insertion of the device, more than usual force was required to pass the suture through the abdominal wall fascia. With the second pass of the needle through the fascia, the needle passed uncontrollably into the peritoneal cavity. A small retroperitoneal hematoma located about 7 cm superior to the sacral promontory was immediately recognized through the 5-mm laparoscope. When the Fig. 1. Carter-Thomason laparoscopic fascial closure device. While carrying suture, the tip (arrow) passes through the full thickness of the wound on one side of the wound. The user then withdraws, inserts the tip through the opposite side of the wound, and grasps the suture. When the user withdraws the suture, the port trocar wound is closed with a tie. Lee. Aortic Puncture With a Laparoscopic Device. Obstet Gynecol 2007. 534 Lee et al Aortic Puncture With a Laparoscopic Device COMMENT OBSTETRICS & GYNECOLOGY closure without use of a device is probably preferable. Second, complications can occur with the CarterThomason device. And third, significant resistance while passing the needle should alert the operator to the possibility of an uncontrolled entrance of the needle into the abdominal cavity with possible resultant injury to vessel or viscus. REFERENCES 1. Guloglu R, Dilege S, Aksoy M, Alimoglu O, Yavuz N, Mihmanli M, et al. Major retroperitoneal vascular injuries during laparoscopic cholecystectomy and appendectomy. J Laparoendosc Adv Surg Tech 2004;14:73–6. 2. Montero M, Tellado MG, Rios J, Mendez R, Somoza I, Pais E, et al. Aortic injury during diagnostic pediatric laparoscopy. Surg Endosc 2001;15:519. Postpartum Iliopsoas Pyomyositis Due to Community-Acquired Methicillin-Resistant Staphylococcus aureus Karen M. Sokolov, MD, Eva Kreye, MD, Loren G. Miller, MD, Chester Choi, MD, and Angela W. Tang, MD BACKGROUND: Community-acquired, methicillin-resistant Staphylococcus aureus (MRSA) infections are on the rise among patients without risk factors for resistant microorganisms. A new, serious community-acquired MRSA manifestation, postpartum iliopsoas pyomyositis is described. CASE: A 24-year-old Hispanic female presented with back pain 9 days after a normal vaginal delivery. Magnetic resonance imaging showed extensive ill-defined edema of the left iliopsoas. Blood cultures yielded communityacquired MRSA. The patient received intravenous vancomycin for 6 days, followed by intravenous, then oral, trimethoprim-sulfamethoxazole. She was discharged on day 8 and made a full recovery. From the Department of Radiation Oncology, University of California Los Angeles; Department of Medical Education, Saint Mary Medical Center, Long Beach, California; and Division of Infectious Diseases, Harbor-UCLA Medical Center, Torrance, California. Corresponding author: Angela W. Tang, MD, Department of Medical Education, St. Mary Medical Center, 1050 Linden Avenue, Long Beach, CA 90813. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 3. Kadar N, Reich H, Liu CY, Manko GF, Gimpelson R. Incisional hernias after major laparoscopic gynecological procedures. Am J Obstet Gynecol 1993;168:1493–5. 4. Reardon PR, Preciado A, Scarborough T, Matthews B, Marti JL. Hernia at 5-mm laparoscopic port site presenting as early postoperative small bowel obstruction. J Laparoendosc Adv Surg Tech A 1999;9:523–5. 5. Carter JE. A new technique of fascial closure for laparoscopic incisions. J Laparoendosc Surg 1994;4:143–8. 6. Elashry OM, Nakada SY, Wolf JS Jr, Figenshau RS, McDougall EM, Clayman RV. Comparative clinical study of portclosure techniques following laparoscopic surgery. J Am Coll Surg 1996;183:335–44. 7. Carter JE. Editorial comment on Carter-Thomason subcutaneous tissue closure device for laparoscopic procedure. Minim Invasive Ther Allied Technol 1996;5:495–7. CONCLUSION: Iliopsoas pyomyositis is a new manifestation of community-acquired MRSA in the obstetric population that may masquerade as benign musculoskeletal back pain. Obstetricians must be alert to the range of presentations of this emerging pathogen. (Obstet Gynecol 2007;110:535–8) M ethicillin-resistant Staphylococcus aureus (MRSA) has long caused infections in patients with risk factors such as hospitalization, surgery, residence in chronic care facilities, and injection drug use. More recently, community-dwelling patients lacking traditional risk factors have acquired MRSA infections.1,2 Many of these infections affect young, healthy people and cause significant morbidity. Community-acquired MRSA is a pathogen that differs from hospital-acquired MRSA in genotype, phenotype, and treatment options. This strain can be distinguished from hospital acquired MRSA by molecular typing methods. Risk factors for communityacquired MRSA include close living conditions, sharing personal items, young age, and association with health care workers, although many patients have no risk factors. Unlike traditional healthcare-associated MRSA, these community-acquired MRSA strains are susceptible to numerous antibiotics including clindamycin, trimethoprim-sulfamethoxazole, and fluoroquinolones. Thus, these infections are amenable to treatment if suspected and properly identified. Recently, community-acquired MRSA has emerged in the healthy obstetric population.3 In the obstetric population, the vast majority (96%) of these infections involve the skin and superficial soft tissue such as cellulitis, abscesses, and mastitis.4 As this organism becomes more prevalent, new life-threatening presenta- Sokolov et al Postpartum Iliopsoas Pyomyositis 535 tions are being described, such as necrotizing fasciitis, Waterhouse-Friderichsen syndrome, and purpura fulminans. We present a new, difficult-to-diagnose, deepseated manifestation of community-acquired MRSA in the postpartum setting to raise awareness of this emerging pathogen among obstetricians. CASE A 24-year-old previously healthy, G3P3 Hispanic female presented 9 days after normal spontaneous vaginal delivery with 2 days of progressively worsening left back pain. Her pain radiated to her left posterior thigh, limiting her ability to ambulate or to sit up from a supine position. She was unable to lift up her legs or flex her hips. The patient denied focal neurologic deficits and tingling or numbness in her legs. She reported subjective fevers and chills for the last two days. The patient’s pregnancy and delivery had been uneventful. She was afebrile throughout her peripartum period, without signs of infection, and had tested negative for human immunodeficiency virus (HIV). She also tested negative for Group B Streptococcus before delivery. She had an epidural catheter inserted for analgesia during delivery. There was no history of sprain or other injury to her back. The patient appeared well developed and nourished. Her temperature was 38.9°C, heart rate 130 beats per minute, and blood pressure 121/75 mm Hg. Her neck was supple without evidence of meningismus. The patient did not have any skin lesions. Her breasts showed no sign of engorgement or infection. There was no swelling or erythema at the catheter insertion site. The patient’s examination was limited by extreme pain on any movement of her hips, which were without deformity, swelling, redness, or warmth. She was able to move normally at the right hip. There was no sensory deficit and reflexes were normal throughout. Her rectal examination showed normal sphincter tone. Routine blood tests showed leukocyte count of 14.1⫻103/␮L, neutrophils 84%, lymphocytes 9%, monocytes 4%, eosinophils 1%. Urinalysis was negative. A CT of her thoracic and lumbar spine was negative for abscess or a protruding intervertebral disc. The patient was started on empiric antibiotic therapy with intravenous ceftriaxone. On hospital day 2, the patient showed no improvement in her symptoms. She remained febrile at 39.8°C. Her white blood cell count had increased to 17.8⫻103/␮L and peaked at 19.1⫻103/␮L on hospital day 3. Vancomycin and metronidazole were added. Pelvic ultrasound revealed an enlarged uterus consistent with recent pregnancy. An MRI of the left hip showed extensive ill-defined soft tissue swelling and edema of the left iliopsoas extending to the piriformis muscle (Fig. 1). On hospital day 3, two of two sets of blood cultures identified MRSA sensitive to clindamycin, trimethoprimsulfamethoxazole, and vancomycin and resistant to cefazo- 536 Sokolov et al Postpartum Iliopsoas Pyomyositis Fig 1. T1 fat suppressed after contrast magnetic resonance imaging showing ill-defined soft tissue swelling and edema of the left iliopsoas muscle (A), extending to the medial aspect of the pelvis lateral to the uterus (B). L, left. Sokolov. Postpartum Iliopsoas Pyomyositis. Obstet Gynecol 2007. lin, erythromycin, and oxacillin. This sensitivity profile was consistent with that of community-acquired MRSA. Metronidazole and ceftriaxone were discontinued on hospital day 4. By day 7, her clinical picture had improved dramatically. She was able to tolerate passive flexion of her left hip to 30 degrees and right hip to 45 degrees. Her white blood cell count decreased to 11.8, and vancomycin was changed to oral trimethoprim-sulfamethoxazole alone. The patient was discharged on hospital day 8 on oral antibiotic trimethoprim-sulfamethoxazole to complete a 15-day course of antibiotic therapy. On discharge she was able to sit up in a wheelchair and since then made a complete recovery. COMMENT This case illustrates a rare but important postpartum complication, iliopsoas pyomyositis. In addition, it highlights community-acquired MRSA, an emerging pathogen in the obstetric population. To our knowledge, this is the first published case of postpartum iliopsoas pyomyositis due to community-acquired MRSA. A PubMed review of the literature with no OBSTETRICS & GYNECOLOGY language restrictions from January 1966 to March 2007, using search terms “postpartum,” “community acquired methicillin resistant Staphylococcus aureus,” “obstetric,” “iliopsoas,” “psoas,” and “pyomyositis,” revealed no other cases of iliopsoas pyomyositis in the postpartum population related to this emerging pathogen. Community-acquired MRSA can be distinguished from hospital acquired MRSA by molecular typing methods. Four staphylococcal cassette chromosomes (SCC) have been identified for Staphylococcus aureus. Staphylococcal cassette chromosomes is a mobile genetic element that carries the gene (mecA), resulting in methicillin resistance in staphylococci. The type IV staphylococcal cassette chromosomes mec is unique to community-acquired MRSA.5 As a rapidly emerging problem in the obstetric and general population, community-acquired MRSA presents predominantly with skin and soft tissue infections. Eight postpartum women with skin and soft tissue infections due to communityacquired MRSA have been described, including a postoperative wound infection, a cellulitis, a pustulosis, and four cases of mastitis.3 Similarly, a retrospective review of 57 pregnant patients diagnosed with community-acquired MRSA found that it most commonly manifests as skin or soft tissue infections, particularly cellulitis and skin abscesses.4 This pathogen was diagnosed in all trimesters as well as the postpartum period. This infection occurred more often in Hispanic patients and patients with comorbidities of HIV infection. Although it is most often found in the extremities and buttocks, nearly one fourth of these patients had mastitis. There are no previous reports of iliopsoas pyomyositis as a manifestation of community-acquired MRSA in the obstetric setting. As a deep-seated infection, pyomyositis is harder to diagnose and results in more morbidity than previously described, superficial community-acquired MRSA infections. It is important for obstetricians to be aware of the more aggressive manifestations of community-acquired MRSA, such as pyomyositis. Pyomyositis is a rare bacterial infection of skeletal muscle more commonly seen in the tropics but increasingly recognized in more temperate climates, especially in HIV-positive patients.1 In more than 90% of tropical pyomyositis and 65–70% of cases in temperate regions, Staphylococcus aureus is the responsible pathogen.1 The main theory for the development of pyomyositis proposes that muscle injury during a period of bacteremia leads to muscle infection.6 In this VOL. 110, NO. 2, PART 2, AUGUST 2007 patient, the proximity of the uterus to the iliopsoas suggests another route of transmission for community-acquired MRSA may be direct extension. Animal models have shown that pyomyositis, in the setting of staphylococcal bacteremia, does not develop without preceding muscle injury.7 In this patient, muscle injury may have resulted from strain during delivery. The presentation of pyomyositis varies depending on the affected muscles. Psoas infections may cause fever, flank or abdominal pain, and difficulty walking. Because the psoas muscle is innervated by L2– 4, pain may radiate anteriorly to the hip and thigh. Nonspecific symptoms of malaise, nausea, and weight loss may be present. On examination, the patient will be most comfortable in the supine position, with the knee moderately flexed and the hip somewhat externally rotated. Stretching or contraction of the psoas muscle leads to pain. Laboratory testing is helpful but not specific. Leukocytosis, elevated erythrocyte sedimentation rate, and elevated blood urea nitrogen are expected. Blood cultures may reveal the most common causative organisms, Staphylococcus aureus or enteric bacteria. The best diagnostic imaging tool for psoas pyomyositis is CT scan, which detects 80 –100% of cases. Ultrasound detects 60%, and MRI is a more cumbersome test that is not more sensitive than CT.8 In our case, the initial CT focused on the thoracic and lumbar spine, missing the abnormality lying just outside this area. The treatment of pyomyositis consists of intravenous antibiotics alone if the infection is identified in its early stages and an abscess has not yet developed. Antibiotic treatment should be based on cultures obtained from blood or from the abscess directly. If MRSA is clinically suspected, inpatients can be treated empirically with intravenous vancomycin. If community-acquired MRSA is identified, antibiotic treatment may be changed to trimethoprim-sulfamethoxazole or clindamycin,5 depending on antimicrobial susceptibility testing. It is not clear at this time which is the preferred agent. Other antibiotics that may be effective but for which there are limited data in breastfeeding and pregnant women include linezolid and daptomycin. Although tetracycline may be active against some strains, prolonged use is generally avoided in breastfeeding and pregnant patients.5 If there is a significant abscess or the patient is not improving on antibiotics alone, percutaneous or open drainage may be necessary. This case of iliopsoas pyomyositis illustrates a rare cause of the common problem back and leg pain. Sokolov et al Postpartum Iliopsoas Pyomyositis 537 It also reflects a new manifestation of an emerging pathogen, community-acquired MRSA, in the obstetric population. Iliopsoas or piriformis pyomyositis should be considered in the differential diagnosis of back and lower extremity pain. Furthermore, the rise of community-acquired MRSA in the obstetric population necessitates that obstetricians consider this organism even in patients without well-defined risk factors for MRSA. With the proper level of suspicion, obstetricians can identify and treat community-acquired MRSA infections promptly, preventing serious morbidity. REFERENCES 1. Maguire GP, Arthur AD, Boustead PJ, Dwyer B, Currie BJ. Emerging epidemic of community-acquired methicillin-resistant Staphylococcus aureus infection in the Northern Territory. Med J Aust 1996;164:721–3. Pyometra After Thermal Endometrial Ablation Matthew Schlumbrecht, MD, Sunil Balgobin, MD, and Larry Word, MD BACKGROUND: Pyometra is a rare but serious complication after thermal endometrial ablation. CASE: A 48-year-old woman with type 2 diabetes and a prosthetic mitral valve underwent a thermal endometrial ablation for abnormal uterine bleeding. She subsequently developed pyometra that induced sepsis and cervical necrosis necessitating hysterectomy. CONCLUSION: The development of a pyometra after endometrial ablation is likely multifactorial. (Obstet Gynecol 2007;110:538–40) E ndometrial ablation is a widely used procedure to treat menstrual disorders in women who have completed child bearing. Since its introduction in the early 1980s, it has become a more popular means of From the Department of Obstetrics and Gynecology, University of TexasSouthwestern Medical Center, Dallas, Texas. Corresponding author: Matthew Schlumbrecht, MD, 5323 Harry Hines Boulevard, Dallas, TX 75204; e-mail: mpschl@parknet.pmh.org. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 538 Schlumbrecht et al Pyometra After Thermal Ablation 2. Hussain FM, Boyle-Vavra S, Bethel CD, Daum RS. Current trends in community-acquired methicillin-resistant Staphylococcus aureus at a tertiary care pediatric facility. Pediatr Infect Dis J 2000;19:1163–6. 3. Saiman L, O’Keefe M, Graham PL 3rd, Wu F, Said-Salim B, Kreiswirth B, et al. Hospital transmission of communityacquired methicillin-resistant Staphylococcus aureus among postpartum women. Clin Infect Dis 2003;37:1313–9. 4. Laibl V, Sheffield J, Roberts S, McIntire D, Trevino S, Wendel GD Jr. Clinical presentation of community-acquired methicillin-resistant Staphylococcus aureus in pregnancy. Obstet Gynecol 2005;106:461–5. 5. Deresinski S. Methicillin-resistant Staphylococcus aureus: an evolutionary, epidemiologic, and therapeutic odyssey. Clin Infect Dis 2005;40:562–73. 6. Kinahan AM, Douglas MJ. Piriformis pyomyositis mimicking epidural abscess in a parturient. Can J Anaesth 1995;42:240–5. 7. Miyaki H. Beitraege zur Kenntnis der sog Myositis Infectiosa. Mitteilungen aus den Grenzgebieten der Medizin und Chirurgie 1904;13:155–98. 8. Taiwo B. Psoas abscess: a primer for the internist. South Med J 2001;94:2–5. surgical management due to its decreased costs, shortened hospital stay, and fewer associated major complications compared with hysterectomy. We present here a rare but severe complication associated with the use of thermal endometrial ablation in a patient with multiple medical comorbidities. CASE A 48-year-old female with no significant gynecologic history presented initially to our gynecology clinic with symptomatic anemia and heavy uterine bleeding. Her medical history was significant for controlled type 2 diabetes and a mechanical mitral valve that required oral anticoagulant therapy. She was admitted to the hospital with an initial hematocrit of 29.0% and an international normalized ratio of 2.0. She continued to bleed despite transfusion, and after anticoagulation reversal, she underwent a dilation and sharp curettage followed by thermal endometrial ablation without hysteroscopy under conscious sedation. The Thermachoice Uterine Balloon Therapy System (Gynecare, Division of Ethicon, Somerville, NJ) was used for ablation, and treatment lasted for 8 minutes per instrument instructions. Appropriate endocarditis antibiotic prophylaxis was administered before the procedure and 6 hours after it. The patient’s bleeding stopped, and oral anticoagulation was restarted with a concurrent low-molecular-weight heparin bridge. She was discharged home on postoperative day 4 without antibiotic therapy. On postoperative day 9, the patient was seen in clinic with a benign examination, although some pieces of endometrium were noted to be coming from her cervical os. On postoperative day 20, our patient presented to the emergency room with complaints of high spiking fevers, lower abdominal pain, and foul-smelling vaginal discharge. OBSTETRICS & GYNECOLOGY On examination, the cervix appeared irregular and friable, with extrusion of purulent material. Bimanual examination revealed cervical motion tenderness, and transvaginal ultrasonography identified a 6-cm abscess involving the lower uterine segment and endocervix. The large size of the abscess and the high risk for bacterial seeding of her prosthetic heart valve prompted a decision to proceed urgently with a total abdominal hysterectomy with bilateral salpingo-oophorectomy. Vancomycin, gentamicin, and clindamycin were administered intravenously. Additionally, 10 units of fresh frozen plasma were given over 16 hours to correct coagulation function to a safe level for surgery. Upon exploration in the operating room, full-thickness necrosis of the lower uterine segment and cervix were noted, with significant inflammation and edema involving the bladder. The procedure was complicated by an incidental cystotomy at the bladder’s dome, which was repaired. The patient was taken to the surgical intensive care unit postoperatively. Final pathologic evaluation of the specimen revealed acute and chronic cervicitis, and cultures from the abscess grew Escherichia coli and Enterococcus. During the first postoperative week our patient developed acute renal failure, with a maximal creatinine level of 1.9 mg/dL. She was febrile, with temperatures spiking to more than 39.0°C, and was continued on antibiotics. On postoperative day 5, she developed atrial fibrillation with rapid ventricular response, which was spontaneously reverted to normal sinus rhythm with an intravenous amiodarone drip. On the morning of postoperative day 7, abdominal examination revealed a large amount of serous fluid and complete disruption of the wound with evisceration. The patient was emergently taken to the operating room, and the wound was closed in an en-bloc fashion with nylon retention sutures. The subcutaneous tissues were left open. During the next 10 days, the patient defervesced, and she was extubated without difficulty. On postoperative day 18, she was transferred to the routine care surgical ward. The retention sutures were cut, and a wound vacuum assisted closure device was placed on the abdominal wound. During her third postoperative week, our patient complained of urine leakage from her vagina. Speculum examination revealed a 2-cm defect at the superior vaginal cuff and a vesicovaginal fistula connecting the cuff to the bladder dome. A Foley catheter was left in place to aid with healing and patient comfort. The remainder of the patient’s hospitalization was unremarkable, and involved wound care and physical and occupational therapy. Her wound was closed with the wound vacuum assisted closure, and the patient was discharged on postoperative day 53 with plans to surgically repair her vesicovaginal fistula at a later date. VOL. 110, NO. 2, PART 2, AUGUST 2007 COMMENT The mechanism of pyometra formation in our patient remains unclear in both its cause and delayed presentation. A pyometra results from a combination of infection and obstruction or as a result of secondary infection of a hematometra. The latter mechanism seems more likely in our patient, given the onset of bleeding after postoperative day nine with passage of clots, followed in a few days by foul odor and purulent egress. We treated our patient with thermal ablation for acute bleeding. Thermal balloon endometrial ablation is a well-established treatment for chronic bleeding problems in low-risk patients, but its role for acute bleeding and in patients with serious medical comorbidities is less clear. There is one case report that documents the use of thermal ablation in an acute setting with no complications.1 However, residual functioning or even regenerating endometrium is known to occur after ablation,2 and in the setting of acute bleeding there is potential for the balloon to coagulate blood in the cavity and fail to ablate the endometrium to the desired depth, leaving residual functioning endometrium. Despite having her anticoagulant state reversed at the time of operation, our patient continued to bleed. This mechanism may also explain the reported higher failure rate of balloon ablation in patients on anticoagulation therapy.3 At the time of the ablation, we noted the cervix to be dilated to approximately 2 cm, with extrusion of blood clots. The mechanical structure of the cervix in addition to the cervical mucus serves to protect the endocervix and uterine cavity from infection under normal circumstances. The loss of these barriers may have predisposed our patient to subsequent infection by facilitating the ascent of vaginal flora. Additionally, on postoperative day 9 our patient was seen in the clinic and noted to have strips of endometrium in the os and scattered throughout the vagina. During the healing process, endometrial casts and debris appear to slough, and this material at 1-week office examination has been shown histologically to contain abundant bacteria.4 One of these fragments may be responsible for occlusion of the os, in addition to providing a high bacterial inoculum to infect a collection of blood present. This obstruction of the cervical os after postoperative day 9, in addition to the bleeding at her therapeutic international normalized ratio, likely caused a hematometra that evolved into a pyometra. The authors of a recent case report recommended prophylactic and postablation antibiotics as a result of a serious infective complication.5 We decided not to give prophylactic antibiotics specifically for the abla- Schlumbrecht et al Pyometra After Thermal Ablation 539 tion. However, the patient received vancomycin and gentamicin for endocarditis prophylaxis both preprocedure and 6 hours postprocedure. Our initial thoughts focused on the possible selection of predominantly resistant or anaerobic organisms, but the heavy growth of E coli and Enterococcus in the abscess that were sensitive to the antibiotics given argued against this idea. A course of therapeutic antibiotics may not have prevented this complication, but may have aided in reducing the bacterial inoculum or limiting our complication to a hematometra only. Furthermore, therapeutic antibiotics may have postponed the pyometra or selected for more ominous or resistant organisms. Pyometra is a serious complication that can result from endometrial ablation, and its cause is likely multifactorial. REFERENCES Mirror Syndrome resolved the fetal hydrops and maternal mirror syndrome. A Novel Approach to Therapy With Fetal Peritoneal–Amniotic Shunt CONCLUSION: In utero treatment of hydropic fetus can result in the cure of maternal mirror syndrome. Jeffrey C. Livingston, MD, Khurram M. Malik, MD, Timothy M. Crombleholme, MD, Foong-Yen Lim, MD, and Baha M. Sibai, MD M BACKGROUND: Mirror syndrome is a rare entity characterized by maternal disease mimicking fetal hydrops. In mirror syndrome, there is maternal hypertension, edema, and often proteinuria in association with fetal hydrops. The causal link between mirror syndrome and hydrops fetalis remains elusive. CASE: This is a case report of a pregnant woman who developed mirror syndrome associated with fetal hydrops. A fetal pelvic mass resulted in bladder outlet obstruction, subsequent bladder rupture, and massive urinary ascites. The resultant massive ascites caused thoracic and cardiac compression and subsequent hydrops fetalis. Placement of a peritoneal–amniotic shunt From the Department of Obstetrics and Gynecology, Maternal Fetal Medicine Division, University of Cincinnati, College of Medicine, and Fetal Care Center of Cincinnati, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio. Corresponding author: Jeffrey C. Livingston, MD, Department of Obstetrics and Gynecology, University of Cincinnati, 231 Albert Sabin Way, Cincinnati, OH 45267-0526; e-mail: livingjc@ucmail.uc.edu. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 540 Livingston et al Mirror Syndrome 1. Nichols CM, Gill EJ. Thermal balloon endometrial ablation for the management of acute uterine hemorrhage. Obstet Gynecol 2002;100:1092–4. 2. McCausland AM, McCausland VM. Frequency of symptomatic cornual hematometra and postablation tubal sterilization syndrome after total rollerball endometrial ablation: a 10-year follow-up. Am J Obstet Gynecol 2002;186:1274–80. 3. Aletebi FA, Vilos GA, Eskandar MA. Thermal balloon endometrial ablation to treat menorrhagia in high-risk surgical candidates. J Am Assoc Gynecol Laparosc 1999;6:435–9. 4. Vilos GA, Vilos EC, Pendley L. Endometrial ablation with a thermal balloon for the treatment of menorrhagia. J Am Assoc Gynecol Laparosc 1996;3:383–7. 5. Roth TM, Rivlin ME. Tuboovarian abscess: a postoperative complication of endometrial ablation. Obstet Gynecol 2004; 104:1198–9. (Obstet Gynecol 2007;110:540–3) irror syndrome, also called “Ballantyne’s syndrome” or “triple edema syndrome,” was first described in 1892 by John William Ballantyne as a triad of fetal hydrops, generalized maternal edema, and placentomegaly. The mother “mirrors” the edema of the fetus and the placenta.1 It can occur at any time during gestation and often results in significant perinatal morbidity and mortality. The incidence and mechanism of this syndrome is unknown. Vidaeff et al2 recently published a literature review of 20 cases of mirror syndrome since 1956. The mechanism of hydrops in our case is due to decreased venous return to the right side of heart, chest compression from ascites, and elevated diaphragm. The clinical manifestations of mirror syndrome are variable, but patients often present with peripheral edema, rapid weight gain, progressive shortness of breath, and hypertension, a clinical presentation and clinical course that is similar to preeclampsia. However, in contrast to preeclampsia, a dilutional anemia is common. In preeclampsia, there is usually hemoconcentration. Hydramnios and elevated maternal plasma uric acid concentration are common in both conditions. Similar to severe preeclampsia, the traditional definitive treatment of mirror syndrome is delivery regardless of gestational age, which often results in OBSTETRICS & GYNECOLOGY neonatal loss or long-term burdens from prematurity. Treating the underlying cause of fetal condition has been reported to result in maternal resolution in conjunction with remission of fetal hydrops.3 CASE A 26-year-old healthy gravida 5, para 4, at 29 2/7 weeks of gestation, was referred to the Fetal Care Center of Cincinnati for the management of fetal pelvic mass thought to be sacrococcygeal teratoma. During her evaluation, she was found to have bilateral swelling of lower extremities, shortness of breath, chest pain, palpitations, and blurred vision. On examination, she had new onset hypertension (blood pressure was 150/105 mm Hg), as well as peripheral and pulmonary edema. Laboratory findings were significant for dilutional anemia (hemoglobin 8.5 g/dL, hematocrit 25.3%) and hyperuricemia (plasma uric acid increased overnight from 5.9mg/dL to 7.9mg/dL). A 12-hour urine collection contained 381 mg of protein. The differential diagnosis includes preeclampsia or mirror syndrome. Ultrasonography and ultrafast fetal magnetic rosonance imaging revealed the presence of a 71⫻81⫻50-mm heterogeneous mass protruding in the sacrococcygeal area, fetal abdominal urinary ascites, skin edema, placentomegaly, oligohydramnios, hydroureter, hydronephrosis, and evidence of bladder rupture. Electrocardiogram of the mother showed sinus rhythm, and chest X-rays showed mild cardiomegaly. Maternal echocardiogram and cardiac enzymes were normal (Fig. 1). At 29 5/7 weeks of gestation, hospital day 3, fetal paracentesis was performed to evaluate renal functions, which demonstrated normal urinary electrolytes. All preeclampsia blood tests remained normal. On hospital day 6, patient developed hypoxia from pulmonary edema. She was treated with intravenous furosemide and fluid restriction. On hospital day 7, the patient had worsening dyspnea and hypoxia. Her oxygen saturation was 79%. Computed tomography pulmonary angiogram was consistent with the clinical pulmonary edema. Doppler studies of lower extremities were negative. Because hydrops fetalis is uniformly lethal at a gestational age of less than 30 weeks, she opted for fetal surgery in an attempt to treat maternal and fetal hydrops. A fetal peritoneal–amniotic shunt (Cook Medical Inc., Bloomington, IN) was placed under direct ultrasonic guidance, in an effort to correct decreased venous blood return to the fetal heart and to reduce intrathoracic pressure. For several days, the patient experienced dyspnea, orthopnea, tachypnea, tachycardia, and frequent runs of asymptomatic premature ventricular contractions. Supportive care and intravenous diuretics were given to treat recurrent episodes of pulmonary edema. Fetal ascites was successfully shunted and the fetal thorax expanded. Fetal hydrops improved, as did placentomegaly. By the third postshunt day, she experienced a 15.9-kg weight loss since admission. Her shortness of breath and edema ultimately resolved, and fetal ultrasonography showed normalization of amniotic fluid (Fig. 2). At 31 weeks of gestation, hospital day 12, the patient was discharged. On her repeat ultrasonography at 33 weeks of gestation, fetal hydrops was completely resolved. The patient continued to be asymptomatic. At 34 4/7 weeks of gestation, she underwent fifth cesarean and delivered a female neonate weighing 3,231 g. Apgar scores were 6 and 8 at 1 and 5 minutes. Maternal postpartum course was unremarkable. The neonate underwent a successful resection of large pelvic mass. The histopathology reports exhibited endodermal sinus tumor of the sacrococcygeal region. (Fig. 3) Fig. 1. Large pelvic mass and urinary ascites at 29 2/7 weeks of gestation. Arrows indicate thick-walled bladder. Fig. 2. Arrows indicate urinary ascites at 29 weeks of gestation. Livingston. Mirror Syndrome. Obstet Gynecol 2007. Livingston. Mirror Syndrome. Obstet Gynecol 2007. VOL. 110, NO. 2, PART 2, AUGUST 2007 Livingston et al Mirror Syndrome 541 Fig. 3. Arrow shows no evidence of fetal ascites at 33 5/7 weeks of gestation; ST indicates fetal stomach. Livingston. Mirror Syndrome. Obstet Gynecol 2007. COMMENT This case represents one of the few successful attempts at in utero fetal therapy for mirror syndrome. Midgely et al3 reported another case of mirror syndrome in which there was complete resolution of maternal and fetal signs and symptoms after successful treatment of fetal tachyarrhythmia. Fetal hydrops is a known complication of vascular tumors due to arteriovenous shunts within the tumor.4 In the context of this case report, we are confident that our patient had mirror syndrome. One of the evidences of mirror syndrome includes maternal volume expansion with subsequent anemia, a most important pathologic feature in contrast to preeclampsia. A case reported by Carbillon et al,5 in which the hematocrit on the presentation was low (26%) and blood volume increased with low plasma protein and no significant proteinuria, was similar to this case. The authors considered the presence of hemodilution and maternal edema to be important distinguishing cgaracteristics between the mirror syndrome and preeclampsia. Previous case reports and small studies confirm anemia as a hallmark sign of the mirror syndrome.2,5 Van Selm et al6 described three pregnancies with severe immunologic fetal–placental hydrops, resulting in fetal death. Mothers experienced severe hydrops syndrome and had anemia, low hematocrit, and elevated plasma uric acid levels. Van Selm suggested the low hematocrit as an important pathophysiologic factor in mirror syndrome. In this case, mild proteinuria and normal renal functions also shift the diagnosis from preeclampsia to mirror syndrome, whereas in preeclampsia decreased renal blood flow leads to decreased clearance of metabolic wastes and elevated levels of creatinine. The pathogenesis of mirror syndrome is unclear. 542 Livingston et al Mirror Syndrome Espinoza et al7 recently suggested the high plasma concentrations of soluble vascular endothelial growth factor receptor-1 (sVEGFR-1), an antiangiogenic factor, is implicated in the pathophysiology of mirror syndrome. Hypoxia of the villous trophoblast in cases of villous edema leads to increased production and release of sVEGFR-1 and other antiangiogenic factors into the maternal circulation. Excessive concentrations of these products may be responsible of maternal edema in mirror syndrome. There are reported cases of mirror syndrome and fetal germ cell tumors. In most of the reported cases of the mirror syndrome, perinatal outcome is poor.8 The other published case of germ cell tumors and mirror syndrome demonstrated postdelivery surgical intervention at an early gestational age. The neonate died of pulmonary immaturity, and maternal illness resolved in next 2 days.8 Other treatable causes of mirror syndrome such as fetal tachyarrhythmia, sacrococcygeal teratoma, and viral infections have shown resolution after successful resolution of the fetal condition.2,3 In our case, reversal of hydrops fetalis and subsequent resolution of maternal disease resulted in a favorable perinatal outcome. The fetus benefited by having reversal of hydrops and a 3-week increase in gestation age period during which time fetal hydronephrosis and pleural and intra-abdominal effusions disappeared. Significant maternal peripheral and pulmonary edema also resolved. This case is also important because it demonstrates an innovative way of complete reversal of maternal and fetal morbidities. Placement of a peritoneal–amniotic shunt treated not only the urinary ascites, but also the maternal symptoms. The key element in this syndrome is to recognize and identify a treatable cause, because timely intervention can lead to the correction of the maternal syndrome and improve perinatal outcomes. Future work is needed to better determine candidates for fetal therapeutics when maternal mirror syndrome is diagnosed. REFERENCES 1. Goeden AM, Worthington D. Spontaneous resolution of mirror syndrome: Obstet Gynecol 2005;106:1183–6. 2. Vidaeff AC, Pschirrer ER, Mastrobattista JM, Gilstrap LC 3rd, Ramin SM. Mirror syndrome: a case report. J Reprod Med 2002;47:770–4. 3. Midgley DY, Harding K. The Mirror syndrome. Eur J Obstet Gynecol Reprod Biol 2000;88:201–2. 4. Inoue M, Kubota A, Hasegawa T, Hata S, Takahashi E, Kawahara H, et al. Antenatal diagnosis of sacrococcygeal teratoma with hydrops fetalis; a case report. Eur J Pediatr Surg 1994;4:125–7. OBSTETRICS & GYNECOLOGY 5. Carbillon L, Oury JF, Guerin JM, Azancot A, Blot P. Clinical biological features of Ballantyne syndrome and the role of placental hydrops. Obstet Gynecol Surv 1997;52: 310–4. 6. Van Selm M, Kanhai HH, Gravenhorst JB. Maternal hydrops syndrome: a review. Obstet Gynecol Surv 1991;46: 785–8. 7. Espinoza J, Romero R, Nien JK, Kusanovic JP, Richani K, Gomez R, et al A role of antiangiogenic factor sVEGFR-1 in the “mirror syndrome” (Ballantyne’s syndrome). J Matern Fetal Neonatal Med 2006;19:607–13. Uterine Carcinosarcoma Associated With Hereditary Nonpolyposis Colorectal Cancer H Stacey A. South, MD, Mollie Hutton, MS, Carolyn Farrell, MS, CNP, Paulette Mhawech-Fauceglia, MD, and Kerry J. Rodabaugh, MD BACKGROUND: Hereditary nonpolyposis colorectal cancer (HNPCC) was originally described as a genetic disorder predominantly involving colorectal cancer. Numerous neoplasms are known to be associated with this condition. Sarcomas have also been reported within families with HNPCC. The challenge is determining if these cancers are sporadic or hereditary. CASE: We report on a 46-year-old woman with uterine carcinosarcoma and a family history suspicious for HNPCC. Genetic testing identified a germline MLH1 mutation. Immunohistochemistry testing of the carcinosarcoma revealed loss of MLH1 expression with preservation of MSH2 expression. CONCLUSION: The loss of MLH1 protein expression suggests the germline mutation contributed to the development of the carcinosarcoma. Hereditary nonpolyposis colorectal cancer should be included in the differential diagnosis of persons with uterine carcinosarcoma when noted within a family history suspicious for HNPCC. (Obstet Gynecol 2007;110:543–5) From the Division of Gynecologic Oncology, Department of Surgical Oncology, Clinical Genetics Service, and Department of Pathology, Roswell Park Cancer Institute, Buffalo, New York. Corresponding author: Kerry J. Rodabaugh, MD, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263; e-mail: kerry.rodabaugh@ roswellpark.org. Financial Disclosure The authors have no potential conflicts of interest to disclose. © 2007 by The American College of Obstetricians and Gynecologists. Published by Lippincott Williams & Wilkins. ISSN: 0029-7844/07 VOL. 110, NO. 2, PART 2, AUGUST 2007 8. Flake AW. Fetal sacrococcygeal teratoma. Semin Pediatr Surg 1993;2:113–20. ereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal-dominant genetic disorder characterized by early onset colorectal carcinoma and is caused by germline mutations in DNA mismatch repair genes, most commonly MLH1 and MSH2, with smaller contributions from MSH6 and PMS2.1,2 It is also associated with extracolonic malignancies, including tumors of the endometrium, stomach, ovary, hepatobiliary tract, and urinary tract. The Amsterdam criteria for the clinical diagnosis of HNPCC were expanded to include cancers of the endometrium, small bowel, ureter, and renal pelvis as HNPCC-associated tumors.2 Sarcomas have also been reported within families with HNPCC. The challenge is determining if these cancers are sporadic or hereditary. Sarcomas have rarely been described in HNPCC and are considered a coincidental finding. We report a case of carcinosarcoma of the uterus within an HNPCC family harboring an MLH1 mutation. CASE A 46-year-old nulligravida presented with an enlarged right inguinal lymph node, which upon biopsy revealed carcinosarcoma. She underwent total abdominal hysterectomy, bilateral salpingo-oophorectomy, and bilateral pelvic and para-aortic lymphadenectomy, confirming stage IV carcinosarcoma of the uterus. She received pelvic and groin radiation, followed by six cycles of cisplatin, ifosfamide, and gemcitabine and has remained disease free for more than 2 years. A comprehensive family history (Fig. 1) revealed her father was diagnosed with colorectal cancer at age 43, and her sister with ovarian cancer at age 44. This information, coupled with our patient’s diagnosis, does not meet modified Amsterdam criteria for the clinical diagnosis of HNPCC.2 The criteria require three relatives to have an HNPCC-associated cancer, with two generations affected and at least one diagnosed before age 50. Her extended family history, however, does meet the criteria as a paternal great aunt and two of her children were diagnosed with colorectal cancer, one before age 50. Genetic testing of the HNPCC-associated genes, MLH1 and MSH2, was pursued through our patient’s father, as he was the closest affected relative with an HNPCC-associated cancer. This was performed using a commercially available South et al Uterine Carcinosarcoma in HNPCC 543 Fig. 1. Comprehensive family pedigree. The proband is denoted by the arrow. The age of death, age at diagnosis, and type of cancer or medical condition for specific individuals are listed when known. Individuals diagnosed with an Hereditary nonpolyposis colorectal cancer–related cancer are shown by solid symbols, while other individuals are depicted by open symbols. The asterisk identifies which family members had genetic testing performed. South. Uterine Carcinosarcoma in HNPCC. Obstet Gynecol 2007. test from Myriad Genetic Laboratories, Inc (Salt Lake City, UT). A mutation in the MLH1 gene, designated E632D (1896G⬎C), was identified. E632D is the result of a nucleotide substitution at the last base pair of exon 16, a location usually highly conserved and necessary for proper mRNA splicing. A similar substitution at this location was identified by the testing laboratory in multiple affected individuals of two HNPCC kindreds and was shown to cause skipping of exon 16 (personal communication with Myriad Genetics Laboratory, Inc, October 5, 2005). E632D is predicted to have a similar effect on mRNA splicing and thus suspected to be deleterious and clinically significant, likely responsible for HNPCC. Site-specific testing for this mutation was performed on our patient and her sister, both confirmed to carry the familial MLH1 mutation found in their father. To determine the MLH1 mutation’s association with the uterine carcinosarcoma, immunohistochemistry testing was performed on the uterine tumor. Formalin-fixed tissue was incubated with monoclonal antibodies, MLH1 (1:30, Zymed, CA), and MSH2 (1:200, Zymed), for 60 minutes at room temperature. A subsequent reaction was performed with biotin-free HRP Enzyme-labeled polymer of the Envision plus detection system (Dakocytomation, CA). Diaminobenzidine complex was used as chromogen. Positive controls were normal colon mucosa. Neg- 544 South et al Uterine Carcinosarcoma in HNPCC ative controls used mouse serum instead of primary antibody. The tumor showed characteristic nuclear staining for MSH2 expression (Fig. 2A); however, no MLH1 expression was noted (Fig. 2B). The functional loss of MLH1 protein product suggests the germline MLH1 mutation contributed to the development of the uterine carcinosarcoma. COMMENT Most cancers associated with HNPCC are adenocarcinomas. A few case reports have described sarcomas in HNPCC families. Hirata et al3 described a patient with rectal adenocarcinoma who developed liposarcoma of the thigh. Genetic testing identified a novel MSH2 mutation. Both the rectal tumor and liposarcoma lacked MSH2 expression by immunohistochemistry, while two sporadic liposarcomas showed normal MSH2 expression. The authors concluded that, since the sporadic tumors retained MSH2 expression, the loss of MSH2 expression in the case suggested the germline MSH2 mutation played a role in the development of the liposarcoma. Similarly Sijmons et al4 documented a case of malignant fibrous histiocytoma in a known OBSTETRICS & GYNECOLOGY radic malignant fibrous histiocytoma tumors retained expression. These reports indicate that mutations in mismatch repair genes may not play a major role in sporadic liposarcoma or malignant fibrous histiocytoma, but they do seem to contribute to the development of these tumors in patients with HNPCC. In conclusion, we have reported a patient with uterine carcinosarcoma and a familial germline MLH1 gene mutation resulting in loss of MLH1 protein expression. The literature reports similar findings in other sarcomas associated with mismatch repair gene mutations, while sporadic tumors retain protein expression.3,4 The absent MLH1 expression in our patient’s tumor, in combination with her MLH1 germline mutation, supports that her carcinosarcoma was related to her diagnosis of HNPCC. Thus, carcinosarcoma should be considered and evaluated as an HNPCC-associated tumor when noted in a family clinically suspicious for HNPCC. Identification of individuals at risk for HNPCC will provide access to the proven benefits of screening programs and prophylactic surgery.5–7 Current data do not support alternate therapies for HNPCC-associated cancers. REFERENCES 1. Peltomaki P, Vasen H. Mutations associated with HNPCC predisposition: update of ICG-HNPCC/INSiGHT mutation database. Dis Markers 2004;20:269–76. 2. Vasen HF. Clinical diagnosis and management of hereditary colorectal cancer syndromes. J Clin Oncol 2000;18: 81S–92S. 3. Hirata K, Kanemitsu S, Nakayama Y, Nagata N, Itoh H, Ohnishi H, et al. A novel germline mutation of MSH2 in a hereditary nonpolyposis colorectal cancer patient with liposarcoma. Am J Gastroenterol 2006;101:193–6. 4. Sijmons R, Hofstra R, Hollema H, Mensink R, van der Hout A, Hoekstra H, et al. Inclusion of malignant fibrous histiocytoma in the tumour spectrum associated with hereditary nonpolyposis colorectal cancer. Genes Chromosomes Cancer 2000;29:353–5. Fig. 2. MSH2 and MLH1 immunohistochemistry staining of our patient’s uterine carcinosarcoma. A. Positive MSH2 nuclear staining is demonstrated by the brown staining designated by the arrow. B. No MLH1 protein expression is noted by the lack of brown cellular staining (arrow). Original magnification, 200⫻. 5. Jarvinen HJ, Aarnio M, Mustonen H, Aktan-Collan K, Aaltonen LA, Peltomaki P, et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000;118: 829–34. South. Uterine Carcinosarcoma in HNPCC. Obstet Gynecol 2007. 6. Lindor NM, Petersen GM, Hadley DW, Kinney AY, Miesfeldt S, Lu KH, et al. Recommendations for the care of individuals with an inherited predisposition to Lynch syndrome: a systematic review. JAMA 2006;296:1507–17. MSH2 mutation carrier. The tumor lacked MSH2 expression by immunohistochemistry, but five spo- 7. Schmeler KM, Lynch HT, Chen LM, Munsell MF, Soliman PT, Clark MB, et al. Prophylactic surgery to reduce the risk of gynecologic cancers in the Lynch syndrome. N Engl J Med 2006;354:261–9. VOL. 110, NO. 2, PART 2, AUGUST 2007 South et al Uterine Carcinosarcoma in HNPCC 545