Laparoscopic myomectomy for symptomatic uterine myomas

Transcription

1 MODERN TRENDS Edward E. Wallach, M.D. Associate Editor Laparoscopic myomectomy for symptomatic uterine myomas Bradley S. Hurst, M.D., Michelle L. Matthews, M.D., and Paul B. Marshburn, M.D. Division of Reproductive Endocrinology, Department of Obstetrics and Gynecology, Carolinas Medical Center, Charlotte, North Carolina Objective: To evaluate the safety, efficacy, and techniques of laparoscopic myomectomy as treatment for symptomatic uterine myomas. Design: Medline literature review and cross-reference of published data. Results: Results from randomized trials and clinical series have shown that laparoscopic myomectomy provides the advantages of shorter hospitalization, faster recovery, fewer adhesions, and less blood loss than abdominal myomectomy when performed by skilled surgeons. Improvements in surgical instruments and techniques allows for safe removal and multilayer myometrial repair of multiple large intramural myomas. Randomized trials support the use of absorbable adhesion barriers to reduce adhesions, but there is no apparent benefit of presurgical use of GnRH agonists. Pregnancy outcomes have been good, and the risk of uterine rupture is very low when the myometrium is repaired appropriately. Conclusion(s): Advances in surgical instruments and techniques are expanding the role of laparoscopic myomectomy in well-selected individuals. Meticulous repair of the myometrium is essential for women considering pregnancy after laparoscopic myomectomy to minimize the risk of uterine rupture. Laparoscopic myomectomy is an appropriate alternative to abdominal myomectomy, hysterectomy, and uterine artery embolization for some women. (Fertil Steril 2005;83: by American Society for Reproductive Medicine.) Key Words: Laparoscopy, endoscopy, uterine myomas, leiomyoma, hysterectomy, myomectomy, adhesions, menometorrhagia, clinical studies, infertility, pregnancy It is well recognized that uterine leiomyomas may cause menorrhagia, dysmenorrhea, and pelvic pressure. The impact of uterine leiomyomas on reproduction is more controversial and the benefit of myomectomy for these patients is not clear. Leiomyomas may impair fertility through several mechanisms including distortion of the uterine cavity, obstruction of the tubal ostia, and alterations in the endometrium affecting embryo implantation and growth. However, because the incidence of uterine leiomyomas increases with age, fertility declines with age, and many women with fibroids conceive spontaneously, it is difficult to assess the direct impact of leiomyomas on fertility. Therefore, myomectomy is indicated only after a complete evaluation of other potential factors for infertility (1). A recent comprehensive review of the literature (23 studies) on leiomyomas and reproduction reported an overall conception rate of 57% after myomectomy among prospective studies (2). Among women with otherwise unexplained infertility, the conception rate was 61% after myomectomy. Received August 25, 2004; revised and accepted September 3, Reprints requests: Bradley S. Hurst, M.D., Department of Obstetrics and Gynecology, P.O. Box 32861, Charlotte, NC (FAX: ; bhurst@carolinas.org). Fibroids that distort the endometrial cavity may impair fertility by several mechanisms including the creation of an abnormal site for placental implantation and growth resulting in infertility, an increased risk of spontaneous abortions, preterm labor and delivery (3). The conception rate is approximately 53% 70% after myomectomy for submucus myomas, and 58% 65% after myomectomy with intramural or subserosal leiomyomas (2). Unfortunately, information on duration of infertility, surgical technique, number and size of leiomyomas, or the increase in pregnancy rate after myomectomy is limited. MYOMAS AND ASSISTED REPRODUCTION Information from studies in IVF cycles controlling for other fertility factors is helpful in evaluating the direct impact of leiomyomas on fertility. Submucosal leiomyomas significantly decrease IVF pregnancy rates (4, 5). Hysteroscopic resection of submucus leiomyomas resulted in a significantly higher pregnancy rate compared to controls with a normal uterine cavity (48% vs. 26%) (6). There appears to be a significant impact on fertility when fibroids distort the uterine cavity, and surgery may be warranted. Fortunately, in many cases, surgery may be performed by hysteroscopy, /05/$30.00 Fertility and Sterility Vol. 83, No. 1, January 2005 doi: /j.fertnstert Copyright 2005 American Society for Reproductive Medicine, Published by Elsevier Inc. 1

2 which provides a less invasive method for leiomyoma treatment. The effect of medium and large intramural myomas on fertility is less clear. A retrospective study showed that IVF live birth rates were not improved by myomectomy, as IVF ongoing pregnancy rates were 16.9% after myomectomy (n 47), 20.8% with fibroids diagnosed but not removed (n 11), and 19% in nonfibroid controls (5). However, there was a 50% spontaneous abortion rate with fibroids compared to 34% after myomectomy in this study, possibly suggesting compromised pregnancy outcome in the presence of fibroids. The impact on fertility of fibroids that do not deform the uterine cavity is also uncertain. The presence of subserosal or intramural leiomyomas without distortion of the endometrial cavity did not impact live birth rates in one study of 46 patients undergoing IVF when compared to controls; however, the impact of leiomyoma size was not evaluated (7). Ramzy et al. (8) also found that myomas, 73% of which were subserosal, had no effect on conception in 39 women. Another study found a significant decrease in IVF live birth rates in women less than 40 years old with intramural leiomyomas (49% vs. 58%) (9). However, the investigators did not support routine myomectomy for these patients. A regression analysis was performed to assess the impact of leiomyoma size on implantation, including patients with multiple myomas but no correlation was found. However, this study included primarily patients with small leiomyomas (mean diameter, 2.0 cm) (9). In contrast to these reports, some studies have shown that intramural myomas reduce IVF success. Stovall et al. (10) evaluated 91 matched IVF cycles in women with subserosal or intramural leiomyomas and found a decreased pregnancy rate of 37% compared to controls 53%. Ninety-five percent of the fibroids in this study were intramural, and size ranged from 8 to 54 mm, with mean diameter 29 mm. Implantation rates were 13.8% with fibroids, significantly lower (P.05) than the 19.7% implantation rate without fibroids. In another study, women with intramural myomas had significantly lower pregnancy rates with intramural myomas compared to women without fibroids (16% vs. 34%, P.05) (4). Implantation rates were 6.4% with intramural fibroids, compared to 15.8% in the control group (P.005), although the total fibroid diameter was only 24 mm. Finally, a recent study by Oliveira et al. (11) evaluated the contribution of intramural leiomyoma size on pregnancy rates and found that the pregnancy rate was significantly decreased only when intramural leiomyomas were 4 cm or larger. Intramural leiomyomas larger than 7 cm were not evaluated routinely as this group recommends removal of larger leiomyomas before IVF. Marchionni et al. (12) retrospectively evaluated the reproductive performance in 72 women before and after abdominal myomectomy of intramural or subserosal myomas. Seventy-five percent achieved a live birth after surgery, significantly improved compared to the preoperative rate of 31%. They correctly pointed out the limitations of this analysis in a retrospective study. Women aged 30 years or younger had the highest pregnancy rates, as did those with only one myoma resected. The size and location of myomas did not affect the prognosis in this study. MYOMAS AND OBSTETRICAL OUTCOMES Obstetrical outcomes are compromised by uterine fibroids, according to a population-based retrospective study by Sheiner et al. (13). Compared to controls, women with uterine myomas during pregnancy had a 3.5-fold increase in the incidence of intrauterine growth restriction (6.8% vs. 1.9%), a 4-fold increase in placental abruption (2.8% vs. 0.7%), a 5-fold increase in the incidence of transverse lie or breech presentation (16.9% vs. 2.4%), a 5-fold increase in the cesarean section rate (57.7% vs. 10.8%), a 70% increase in premature rupture of membranes (9.6% vs. 5.5%), and were three times more likely to receive transfusion (4.2% vs. 1.4%). All of these outcomes were significant (P.001). Adjusting for maternal age, parity, gestational age, and malpresentation, pregnancies in the myoma group still had a 6.7-fold risk of cesarean section (95% confidence interval [CI] , P.01), 2.5-fold increase of placental abruption (95% CI , P.001), and a 40% increase in preterm deliveries compared to pregnancies without uterine fibroids (95% CI , P.009). The size and location of the myomas were not described in this study, but other studies have shown that fibroids adjacent to the placenta increase the risk of bleeding and premature rupture of membranes (14). It would appear that myomectomy could be justified in some circumstances to reduce the risk of adverse pregnancy outcomes in some patients. Further studies stratifying patients by leiomyoma number, size, and location will clarify the contribution of leiomyoma size on fertility. The majority of the evidence appears to support an impairment of fertility with intramural leiomyomas, particularly in larger leiomyomas. It is important to clarify this issue, as myomectomy for intramural fibroids has a risk of morbidity and adhesion formation, and surgery should not be considered unless the benefits outweigh the risks. In spite of the limited data, myoma size, location, and number are key factors when discussing the outcomes after myomectomy, and these are considered separately in this discussion. Unfortunately, these factors are not separable for an individual patient, and the surgeon must weigh the cumulative impact of all three factors when deciding how and when to perform a myomectomy. LAPAROSCOPIC MYOMECTOMY Although laparoscopic myomectomy has been performed since Semm and colleagues described the procedure in late 1970s (15 18), the role of laparoscopic myomectomy as a treatment option for symptomatic uterine fibroids has been questioned. Many fibroids that can be easily removed lapa- 2 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

3 TABLE 1 Laparoscopic myomectomy surgical outcomes. Study type Number L/S (abd) Advantages laparoscopy Disadvantages laparoscopy Notes RCT Mais, 1996 (22) 20 (20) Less pain, shorter hospitalization and recovery Seracchioli, 2000 (23) Rossetti, 2001 (24) 41 (40) 84 nonrandomized L/S 66 (65) Less fever, less blood loss, shorter hospitalization None (no difference in recurrence) None Up to 4 myomas, largest 3 6 cm None None Myoma 5 cm, up to 3 total; recurrence not different Increased myoma recurrence with GnRH-a RCT-GnRHa Zullo, 1998 (36) 67 1 OR time, 2 blood loss with GnRH-a Campo, 1999 (37) 60 1 OR time with GnRH-a RCT-Tibolone Palomba, 2002 (38) 66 2Blood loss with GnRH-a, no benefit with tibolone RCT-Bupiv. epi. Zullo, 2004 (99) 56 2 EBL, 2postoperative pain, 2 surgical difficulty, 2 surgical time with bupivicaine epinephrine Case control Stringer, 1997 (26) 49 (49) Shorter hospitalization, fewer complications Longer surgical time, higher blood loss No difference in cost Silva, 2000 (25) 25 (51) Shorter Longer surgical 20/25 L/S assisted hospitalization time Case series Landi, 2001 (27) 368 Complication rate 3.3%: 12 fever, 10 transfusions, 2 2nd OR; full recovery by 10 days Malzoni, 2003 (31) 144 Complication rate 2% including 1 transfusion, 1 conversion to laparotomy Ou, 2002 (30) hyst., 1 conversion to laparotomy Adamian, 1996 (29) 176 Myoma 7 16 cm Doridot, 2001 (28) % recurrence rate by ultrasound, 4% reoperation, recurrence 1 yr, Seinera, 1999 (32) 30 Uterine scar 272% by day 30 Ostrzenski, 1997 (34) 32 No complications, postoperative hospitalization 7.5 hours Seinera, 1997 (33) 89 1 conversion to laparotomy Sinha, 2003 (35) 51 Myomectomy for myomas 9 cm, largest 21 cm Note: EBL estimated blood loss; GnRH-a GnRH agonist. Hurst. Laparoscopic myomectomy. Fertil Steril Fertility and Sterility 3

4 roscopically may not require surgical intervention (19). Laparoscopic myomectomy was given lukewarm support in the May 2000 American College of Obstetricians and Gynecologists (ACOG) guidelines (20): The two major concerns with laparoscopic myomectomy versus hysterectomy are the removal of large myomas through small abdominal incisions and the repair of the uterus. The introduction of more efficient morcellators has made the removal easier, although skilled operative technique is necessary because injury to other organs is possible. Although there are multiple techniques available for laparoscopic suturing, there is controversy as to whether the closure techniques available are equal to those achieved at laparotomy. This is most relevant to women contemplating a future pregnancy. In spite of these reservations, recently published studies indicate that laparoscopic myomectomy may be an appropriate alternative to abdominal myomectomy in wellselected patients. Many reproductive surgeons have the prerequisite skills to perform laparoscopic myomectomy, and advances in instruments and techniques have made this approach more accessible to physicians and patients. However, before embracing laparoscopic myomectomy, this approach must be proven to be safe, effective, and in some ways superior to abdominal myomectomy and other treatment options. Indications, preoperative evaluation, and exclusion criteria must be established for the procedure. Pain, cost, surgical risks, recovery time, and obstetrical outcomes are important considerations. The major studies and clinical series for laparoscopic myomectomy are summarized in Table 1. Because laparoscopic myomectomy can be a technically challenging procedure, techniques to simplify the surgery are provided in this review. Indications for Myomectomy (American College of Obstetricians and Gynecologists [ACOG] and American Society for Reproductive Medicine [ASRM]) Laparoscopic myomectomy is only appropriate when indications for surgery have been met. Pelvic pain, pressure, and abnormal uterine bleeding are the most common symptoms that lead women to seek surgery for fibroids (20). Fibroids may also compress adjacent structures, and cause urinary frequency or urgency, constipation, or dyspareunia, or poor reproductive outcomes. Indications for surgical management of uterine myomas include the following: [1] abnormal uterine bleeding not responsive to conservative treatments, [2] high level of suspicion of malignancy, [3] growth after menopause, [4] infertility with distortion of the endometrial cavity or tubal occlusion, [5] pain or pressure that interferes with quality of life, and [6] urinary tract frequency or obstruction, or iron deficiency anemia related to abnormal uterine bleeding (21). Randomized Controlled Trials: Laparoscopic vs. Abdominal Myomectomy There have been three prospective randomized studies comparing abdominal and laparoscopic myomectomy (22 24). From these studies, laparoscopic myomectomy is clearly associated with shorter hospitalization, faster recovery, less expense, less pain, less blood loss, less fever, and fewer surgical complications compared to abdominal myomectomy. Pregnancy rates and recurrence rates appear to be comparable between laparoscopic and abdominal myomectomy. The first randomized controlled trial published in 1996 by Mais et al. (22) compared open (n 20) and laparoscopic (n 20) myomectomy and found less pain, shorter hospitalization, and shorter recovery with laparoscopic surgery. Subjects were aged years and had one to four subserous or intramural myomas, with the largest measuring 3 6 cm. Submucus myomas were excluded by hysteroscopy. No patient received preoperative ovarian suppression. Surgery was performed by placing a 10-mm umbilical laparoscopic port, a 10- to 12-mm port to the right of the umbilicus, and a 5-mm left port. The myoma was secured with a myoma drill, and bipolar cautery and scissors were used to dissect the fibroid from the pseudocapsule. No vasoconstrictive agents were used. The myometrial defect was closed with interrupted 1-0 Vicryl (Ethicon, Somerville, NJ) with extracorporeal knots. The serosa was not closed in the laparoscopic group, but was closed in the open myomectomy group. Myomas were removed laparoscopically by morcellation. In the Mais et al. study (22), the number of myomas ( vs ) and mean diameter of the largest myoma ( vs cm) were similar between the laparoscopic and open technique, respectively. Women undergoing laparoscopic myomectomy had comparable blood loss ( vs ml) and surgical time ( vs minutes) compared to open myomectomy. However, postoperative pain was significantly less with laparoscopic myomectomy using a visual analogue scale testing and narcotic requirements (P.05). Ninety percent of patients were discharged by day 3 in the laparoscopy group, compared to 10% in the open group (P.05), and 90% reported complete recovery in the laparoscopic group compared to only 5% in the open group (P.05). There were no major complications in either group. In 2000, Seracchioli et al. (23) reported a randomized controlled trial of 65 abdominal and 66 laparoscopic myomectomies in infertile women with at least one fibroid of 5 cm or larger and found less febrile morbidity and anemia postoperatively with laparoscopy. Hysteroscopy was performed when preoperative ultrasonography showed endometrial distortion. Cases with more than three myomas 5 cm or with a uterine size above the umbilicus were excluded. Laparoscopic myomectomy was performed through an umbilical incision, two 5-mm suprapubic trocars lateral to the deep inferior epigastric arteries were placed slightly higher than usual, a 10-mm trocar was inserted in the midline, and a uterine cannula was always used. The serosal incision was made with a monopolar pointed knife and the myoma was dissected with claw forceps and scissors while traction was 4 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

5 kept on the myoma and countertraction on the uterus. Vessels were coagulated with bipolar cautery. The uterus was sutured in one or two layers, according to the depth of the incision, with interrupted or continuous sutures of 1-0 or 2-0 absorbable suture using intracorporeal knots. The myomas were removed with a manual morcellator. In this study, febrile morbidity was significantly higher with abdominal myomectomy compared to laparoscopy (26% vs. 12%, P.05), and abdominal myomectomy caused a larger drop in hemoglobin level (2.2 g/dl vs. 1.2 g/dl, P.001) (23). Three women required transfusion after abdominal myomectomy, but none were transfused after laparoscopy. Hospitalization was longer with abdominal myomectomy (143 vs. 76 hours, P.001). There was no difference between the two groups concerning patient age, size, number, or location of the largest myoma. The surgical time was slightly lower with abdominal myomectomy (89 27 vs minutes), but the difference was not significant. Three laparoscopic cases were converted to abdominal procedures due to difficulties of hemostasis or difficulties in suturing. The recurrence rates were similar. Rossetti et al. (24) reported long-term follow-up of 81 patients randomized to abdominal or laparoscopic myomectomy plus 84 nonrandomized patients and found similar recurrence rates, 23% and 27%, respectively, between the laparoscopic and abdominal myomectomy. In the prospective study, 81 women aged 42 years with at least one but no more than seven symptomatic myoma 3 cm were randomized to abdominal (n 40) or laparoscopic (n 41) myomectomy after a submucus myoma was excluded by hysteroscopy. Techniques for laparoscopic myomectomy included placement of three suprapubic ports and a manipulator probe placed into the uterus. A dilute solution of vasopressin was used to reduce blood loss. Pedunculated myomas were secured with pretied suture, coagulated, and the base cut with bipolar forceps and scissors. Subserous and intramural myomas were excised with a monopolar hook and mechanical dissection. Grasping forceps were used to apply traction to the myoma, and bipolar cautery was used to obtain hemostasis. The myometrium was closed with interrupted simple or cross-stitches of 1-0 or 0-polyglactin sutures. In the study by Rossetti et al. (24) the open and laparoscopy groups were similar regarding age, preoperative symptoms, myoma size, and number. The recurrence rates were comparable, 27% after laparoscopy and 23% in the open group, with most recurrences seen within 24 months of surgery. There were no major complications. Two laparoscopies were converted to open procedures: one due to anesthesia problems and one due to the size and number of myomas. Surgical Outcomes of Laparoscopic Myomectomy: Case- Control and Case Series Comparative studies and case-series have supported the advantages of laparoscopic myomectomy over abdominal myomectomy (Table 1). Two case-control studies have evaluated surgical outcomes comparing laparoscopic and abdominal myomectomy. Silva and colleagues (25) performed a case-control study of 25 laparoscopic myomectomy and 51 abdominal myomectomy procedures. Of the 25 in the laparoscopy group, 20 were laparoscopy assisted. Hospital stay was significantly shorter with laparoscopy compared to laparotomy (30.5 vs hours, P.001, as was duration of postoperative narcotics use (14.8 vs. 24 hours, P.001). However, surgical time was longer in the laparoscopy group (222.5 vs. 180 minutes, P.001) compared to the open group. There was no difference in blood loss. Stringer et al. (26) performed a retrospective case-control study comparing 49 laparoscopic myomectomies performed between 1993 and 1995 and 49 consecutive abdominal myomectomies performed from 1983 to Surgical time for laparoscopic myomectomy was almost double compared to open procedures (264 vs. 133 minutes, P.001). Mean blood loss was also higher for laparoscopic myomectomies (340 vs. 110 ml, P.001), but three patients were transfused in the open group, compared to no transfusions with laparoscopy. Benefits included a considerably shorter hospitalization with laparoscopic myomectomy (0.6 vs. 5.6 days, P.001). There were 17 complications in the abdominal myomectomy group compared to 5 in the laparoscopic group, a significant increase (P.007). Several case series have described laparoscopic myomectomy. Landi et al. (27) studied laparoscopic myomectomy in a series of 386 women. Mean operating time was 101 minutes, the decrease in postoperative hemoglobin was 1.4 g/100 ml, and the hospital stay was 2.9 days. Short-term complications occurred in 3.3% and included bleeding requiring autologous transfusion in 2.7% and fever in 3.3%. Three women required transfusions, and a second laparoscopy was performed in two cases. The average time to full recovery was 10.5 days. Postoperative sonography showed an echogenic area with ill-defined margins at the myometrial surgical site. By 30 days postoperative, 3% of women had an anechoic area attributed to a hematoma. Doridot et al. (28) assessed the recurrence of leiomyomata in 192 women after laparoscopic myomectomy. In a mean follow-up of 47 months, the cumulative recurrence risk was approximately 13% at 2 years, and 17% at 5 years. They concluded that the recurrence rate may be higher after laparoscopic myomectomy, but considered the low morbidity to be a major advantage of the laparoscopic approach. Adamian et al. (29) described laparoscopic myomectomy in 178 cases. Surgical time ranged from 70 to 200 minutes (mean, 105 minutes), with prolonged cases attributed to removal of myomas. The mean blood loss was 56 ml (range, ml); no vasoconstrictors were used. The investigators concluded that laparoscopic myomectomy was effective for even large myomas. Fertility and Sterility 5

6 Ou and colleagues (30) retrospectively evaluated 168 women treated for uterine fibroids, including 165 women who were successfully treated laparoscopically. They found that blood loss was less when myomas were resected by the Harmonic Scalpel (Ethicon Endo-Surgery, Inc., Cincinnati, OH), and greater when resected by cautery, 243 ml and 378 ml, respectively. Women whose myomas were removed by colpotomy had significantly shorter operating time than those whose fibroids were removed by morcellation, 144 minutes compared to 168 minutes, respectively, although a mean of 7 myomas were removed by colpotomy compared to 4 by morcellation. Malzoni et al. (31) performed laparoscopic myomectomy in 144 women with myomas 5 cm or larger. In this series, 108 myomas ranging in size from 5 to 18 cm were intramural or submucosal, and the remaining myomas were subserosal, intraligamentous, or pedunculated. The surgical time ranged from 58 to 180 minutes, with a mean of 95 minutes, and two women required conversion to laparotomy. The researchers concluded that large myomas could be resected laparoscopically. Seinera and colleagues (32) evaluated myomectomy scar healing by prospectively evaluating 30 women after laparoscopic myomectomy by Doppler ultrasound. The myomas were 3 4 cm in 9 of the 30 cases, 5 6 cm in 18, and 7 cm in 3 patients. Eighty percent of the myomas were intramural, and the remainder was submucosal. Scar diameter determined by ultrasound was 37% smaller on postoperative day 1 compared to the preoperative myoma diameter, and was 72% smaller than the original myoma by postoperative day 30. Postoperative bleeding was minimal; all patients had 20 ml of blood in the cul-de-sac, as estimated by ultrasound. In an earlier study of 89 laparoscopic myomectomies by the same group (33), 1 4 myomas were resected per patient and most of these were intramural. The mean diameter of the largest myoma was 4 cm. The defect was closed in one or two layers, depending on the depth of the myoma. The mean blood loss was 84 ml. Fewer than 2% of the procedures required conversion to laparotomy, and the overall complication rate was 1.8%. These investgators concluded that laparoscopic myomectomy was associated with a low incidence of complications and appeared to be a feasible treatment. Ostrzenski (34) evaluated laparoscopic myomectomy in 32 women with intramural myomas that caused endometrial distortion. The myometrium was repaired in three layers, and no case was converted to laparotomy. The mean operative time was 163 minutes, and ranged from 127 to 246 minutes. The mean drop of hemoglobin was 0.74 g/dl. No patient required transfusion, and no complications were observed. The average stay was 7.5 hours. The researcher concluded that laparoscopic myomectomy was a safe alternative to laparotomy. Sinha et al. (35) described a case series of 51 laparoscopic myomectomies for large myomas 9 21 cm. The mean myoma weight was 700 g, and mean surgical time was 2 hours 17 minutes. The blood loss was 322 ml, but the range was 100 2,000 ml. Although they concluded that laparoscopy is a safe alternative to laparotomy for very large myomas, complications included one broad ligament hematoma and one postoperative hysterectomy for dilutional coagulopathy. Overall, these studies consistently show an overall low complication rate with laparoscopic myomectomy. Surgical times may be longer than with open procedures, but the recovery time is shorter. In these studies, the conversion rate to laparotomy was typically less than 2% of cases, even when large deep myomas were resected. Finally, several studies demonstrated the feasibility of a multilayered closure during laparoscopic myomectomy. GnRH Agonist Pretreatment Three prospective randomized studies have evaluated the effect of preoperative treatment with GnRH agonists (GnRH-a) for laparoscopic myomectomy (36 38). In theory, pretreatment could shrink the myomas and simplify myomectomy. However, GnRH-a therapy might, in some circumstances, increase the difficulty of the procedure and increase operating times. In addition, one study found that GnRH-a therapy significantly increased the risk of recurrence after laparoscopic myomectomy (24). The primary advantages of GnRH-a therapy appear to be correction of anemia before surgery, and a slight reduction in blood loss during laparoscopic myomectomy (39). The GnRH-a therapy may also reduce adhesions (40 43). However, whether the expense and side effects of GnRH-a pretreatment without anemia are justified based on a slight clinical benefit is debatable (44). Zullo and co-investigators (36) randomized 67 women who were undergoing laparoscopic myomectomy for symptomatic fibroids to leuprolide acetate (LA) or direct surgery and found marginal benefits for GnRH-a. The GnRH-a group received two intramuscular doses of depot leuprolide (3.75 mg) given 4 weeks apart, and laparoscopic myomectomy was performed 2 5 weeks after the second injection. The control group directly underwent surgery. Preoperative uterine volume was significantly reduced with GnRH-a therapy. The mean operative time was 99 minutes with GnRH-a and 113 minutes in controls. The mean drop of hemoglobin was 0.9 g/dl in the agonist group and 1.3 g/dl in the controls. Both of these measurements were statistically significant, but the clinical difference was marginal. Large fibroids with volume 60 cm 3 or multiple ( 3) myomas had a significantly shorter operative time with GnRH-a therapy, but LA increased the surgical time for hypoechogenic myomas compared to controls. Two cases were converted to laparotomy, and one patient required transfusion. During a minimum 6-month follow-up, 43% of infertile patients conceived and 16% of those with abnormal uterine bleeding experienced 6 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

7 recurrent symptoms. The investigators concluded that GnRH-a pretreatment might be beneficial, except for those patients with markedly hypoechogenic ultrasound patterns. Palomba and investigators (38) randomized 66 women with symptomatic myomas to three groups: GnRH-a plus tibolone and iron, GnRH-a plus placebo, and iron-only controls. Subjects were treated for 2 months before surgery. Operative time and blood loss were lower in both GnRH-a groups. Tibolone provided no additional benefit for surgical outcomes. Campo and Garcea (37) randomized 60 women with symptomatic myomas to three cycles of preoperative GnRH-a (decapeptyl 3.75 mg) or controls. Patients were excluded if myoma diameter was 10 cm or if there were more than 8 myomas. Operative time was more than 40 minutes longer in the GnRH-a group, 158 compared to 112 minutes, a difference that was statistically significant. Blood loss tended to be lower with GnRH-a pretreatment, but this difference was not significant. The number of myomas (mean, 2.9) removed was similar in each group, as was the mean diameter of the largest myoma (4.8 cm). There were no transfusions in either group, and no serious complications. Postoperatively, the pregnancy rate in infertile patients was 54%, and there was no difference in pregnancy rates between GnRH-a therapy and controls. Campo and Garcea suggested that GnRH-a therapy increased the difficulty of myoma enucleation, possibly by altering the myoma myometrium interface. Because GnRH-a therapy may be associated with higher recurrence and may increase the difficulty of fibroid enucleation, and provides only slight benefit of reducing blood loss during surgery, preoperative GnRH-a should not be routinely used except in cases of preoperative anemia (36 38). ADHESIONS Laparoscopic vs. Abdominal Myomectomy There are no prospective randomized controlled trials comparing adhesion formation after laparoscopic and abdominal myomectomy. However, prospective studies and surgical case series seem to point to a benefit with laparoscopy (Table 2). Adhesions form in more than 90% of abdominal myomectomies (45). The incidence is highest (94%) with posterior incisions, and lower (56%) with fundal or anterior uterine incisions (46, 47). One case-control series compared 16 abdominal myomectomies with laparoscopic myomectomy and found fewer adhesions and significantly reduced adhesion scores with laparoscopic myomectomy (48). In the study by Stringer et al. (26) women undergoing subsequent surgery had fewer adhesions with laparoscopic myomectomy and higher adhesion rates with abdominal myomectomy. Several case series have evaluated adhesion formation after laparoscopic myomectomy. Although these studies are limited by the absence of a control group and lack of systematic performance of second-look laparoscopy, the incidence of adhesions in all of these studies is well below the expected occurrence seen in abdominal myomectomy cases (45). Dubuisson et al. (49) performed second-look laparoscopy in 45 women after laparoscopic myomectomy and assessed 72 myomectomy sites. Adhesions were found in 36% of patients and at 17% of each myomectomy site. The adhesion rate was highest with posterior incisions, but the rate of adhesion formation was only 33% at this site. The sigmoid colon was the leading site for uterine adhesions, followed by the bladder, adnexa, small bowel, and peritoneum. Of those who had adnexal adhesions, most had preexisting adnexal adhesions, another surgical procedure carried out at the same time as myomectomy, or a posterior myomectomy site. The investigators concluded that the rate of adhesions after laparoscopic myomectomy was low and rarely involved the adnexa. Takeuchi and Kinoshita (50) performed second-look laparoscopy in 51 women after laparoscopic myomectomy. The mean size of the largest myoma removed was more than 6 cm, and a mean of three myomas were removed per patient. Fibrin glue spray was applied to prevent postoperative adhesions. At second-look, adhesions were found in 29% of patients, and at 11% per myomectomy site. The incidence of adhesions was greatest at posterior myomectomy sites, and with resection of intramural myomas. The sigmoid colon was the organ most commonly adhered to the uterus. Adnexal adhesions were identified in 18% of patients. They concluded that the rate of adhesion formation after laparoscopic myomectomy was low. Keckstein and co-investigators (51) performed secondlook laparoscopy in 22 cases after laparoscopic myomectomy. Myomectomy sites were closed with two or three layers. Overall, the rate of adhesion formation was 28% per patient, with adhesion formation greatest with posterior myomectomies. Malzoni et al. (31) described second-look laparoscopy in 18 women after laparoscopic myomectomy and found adhesions in 33% of patients. The mean myoma diameter in this study was 7.8 cm, and most of the myomas resected were intramural. No adnexal adhesions were found at the time of second-look laparoscopy. Other studies have demonstrated consistent results regarding adhesions after laparoscopic myomectomy. DiGregorio and co-investigators (52) reported second-look laparoscopy in 121 women after laparoscopic myomectomy, and found adhesions in only two cases. Hasson and colleagues (53) described second-look laparoscopy in 24 women and found adhesions in 66%. The incidence of adhesions was greatest in women with adhesions noted at the time of initial surgery. Of those with no adhesions at the time of myomectomy, adhesions were found in 54%. Nezhat et al. (54) found that suturing the incision site increased the incidence of adhesions. Fertility and Sterility 7

8 TABLE 2 Adhesions after laparoscopic myomectomy. Study type (author) Number Adhesion rate Comments RCT 25 Interceed 40% Significant 2 with Interceed Mais, 1995 (55) 25 controls 88% Pellicano, 2003 (57) 18 hyaluronic acid 28% Significant 2 with hyaluronic acid 18 controls 78% Assaf, 1999 (98) 21 vasopressin Significant 1 adhesions with vasopressin 17 no injection Case control 16 l scope Significant 2 with l scope Bulletti, 1996 (48) 16 abdominal Case series 45 36% Adhesions 17% per site Dubuisson, 1998 (49) Takeuchi, 2002 (50) 51 29% Adhesions 11% per site with fibrin glue Keckstein, 1994 (51) 22 28% Significant 1 adhesions with posterior incisions Malzoni, 2003 (31) 18 33% No adnexal adhesions DiGregorio, 2002 (52) 121 2% Interceed always used Hasson, 1992 (53) 24 66% 54% de novo adhesion Note: l scope laparoscope. Hurst. Laparoscopic myomectomy. Fertil Steril Adhesion Barriers Two prospective, randomized controlled studies have evaluated the efficacy of adhesion barriers during laparoscopic myomectomy, and both studies found intervention to be beneficial. Interestingly, when systematic second-look laparoscopy is used in these prospective studies, the overall incidence of adhesions seems to be higher than in the retrospective case series. The explanation for this discrepancy may lie in the tendency of case series to be self-selected, with mainly the best outcomes submitted or accepted for publication. The true incidence of adhesions after laparoscopic myomectomy is likely to be more reliable in these small but informative randomized studies. Mais and colleagues (55) evaluated the effect of oxidized regenerated cellulose, Interceed (Gynecare, Somerville, NJ), on adhesion formation after laparoscopic myomectomy in a prospective randomized study of 50 women. The mean size of the largest myoma was 4.4 cm in each group. A mean of 2.0 myomas were resected in the control group, and 2.2 resected in the Interceed group. The two groups were similar for the size, number, and location of myomas resected. Interceed was placed to cover all incisions and suture material with a 1-cm margin, and the barrier then moistened with saline. The time required for placement of the barrier was 5 minutes in all cases. Mais et al. did not state if there was any evidence of bleeding, or if the Interceed membrane turned brown during placement. During the second-look laparoscopy, 60% of the Interceed group was free of adhesions, compared to only 12% adhesion-free in the control group, and the difference was significant. Adhesions tended to be less dense, and filmier with Interceed. Adnexal adhesions were found in 40% with Interceed, and 64% in the control group. The investigators concluded that Interceed significantly reduced, but did not prevent, adhesions after laparoscopic myomectomy. They speculated that incomplete hemostasis might explain, in part, the impairment of the effectiveness of Interceed in preventing adhesions in some patients, as had been found in other studies (56). Pellicano et al. (57) recently showed that hyaluronic acid gel reduced adhesions after laparoscopic myomectomy in a prospective randomized study of 36 infertile women with symptomatic uterine fibroids. The size, number, and location of the fibroids were similar between the treatment and control groups. Vasopressin was injected before myomectomy, and the incisions were closed with either figure-of-eight sutures or subserosal sutures of 2-0 polyglactin. At the conclusion of the myomectomy and uterine closure, treated patients were administered autocrosslinked hyaluronic acid gel (5 ml) to the surface of the uterus. During second-look, 72% of patients were adhesion-free with hyaluronic acid gel treatment, compared to an adhesion-free rate of only 22% in the control group, a significant difference (57). Adhesions were increased with figure-of-eight closures, and occurred in 44% of those treated with hyaluronic acid, and 89% of the controls. This adhesion rate was significantly higher than with subserosal closure; adhesions formed in 11% with hy- 8 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

9 TABLE 3 Pregnancy outcomes after laparoscopic myomectomy (patients attempting pregnancy). Study Number pregnant Pregnancy rate SAB Live birth rate C/S rate Uterine rupture RCTs 30 L/S 54% 20% 77% 65% 0 Seracchioli, 2000 (23) 33 abd 56% 12% 88% 78% 0 Case control 44 L/S 42% 7% 93% 0 Bulletti, 1999 (64) 12 No Tx 11% 45% 55% 0 27 Unexpl 25% 7% 93% 0 Case series Ribeiro, 1999 (68) 18 64% 12% 78% 57% 0 Landi, 2003 (69) 72 17% 79% 46% 0 Campo, 1999 (37) 13 54% 15% 85% 45% 0 Malzoni, 2003 (31) 21 55% 15% 81% 57% 0 Seracchioli, 2003 (61) 9 39% 22% 78% 0 DiGregorio, 2002 (52) 65 44% 11% 86% 92% 0 Dubuisson, 2000 (76, 77) % 31% 69% 42% 1 surgical site Seinera, 1997, 2000 (33, 63) 64 12% 86% 80% 0 Stringer, 1997, 2001 (26, 60) 7 28% 72% 57% 0 Rossetti, 2001 (73) 21 66% 22% 78% 71% 0 Dessolle, 2001 (74) 44 41% 14% 82% 32% 0 Darai, 1997 (70) 17 39% 23% 58% 33% 0 Nezhat, 1999 (71) 42 20% 75% 78% 0 Dubuisson, 1996 (62) 7 33% 0% 100% 57% 0 Miller, 1996 (72) 30 75% 13% 87% 0 Campo, 2003 (65) 22 61% 14% 86% 40% 0 Total L/S (n) Hurst. Laparoscopic myomectomy. Fertil Steril aluronic acid, and 67% in the control group with this technique. The location of myomas did not appear to affect the rate of adhesion formation in this study. In a randomized controlled trial, Imai et al. (43) found fewer adhesions with GnRH-a pretreatment compared to no-treatment after laparoscopic and abdominal myomectomies. The utility of GnRH-a in reducing adhesions has also been shown in other clinical and animal studies (40 43). Mettler et al. recently assessed adhesion prevention with SprayGel (Confluent Surgical, Waltham, MA), a synthetic absorbable adhesion barrier, in a randomized controlled trial in both abdominal and laparoscopic myomectomies (58, 59). Lower adhesion scores were found with this gel, and there were no adverse events found with its application. This product is currently under investigation in the United States. Overall, the results of all of these studies indicate that the adhesion rate with laparoscopic myomectomy is consistently lower than the rate expected with abdominal myomectomy, although the adhesion rates with laparoscopy appear to approach the expected rate with abdominal myomectomy in prospective randomized studies with no intervention. Data are limited, but oxygenized regenerated cellulose and autocrosslinked hyaluronic acid gel both appear to reduce adhesions, and GnRH-a therapy may also be effective for this purpose. It remains to be proven if any of these interventions enhance pregnancy rates or reduce clinical symptoms of pelvic adhesive disease. PREGNANCY AFTER LAPAROSCOPIC MYOMECTOMY Overall pregnancy rates and spontaneous abortion rates with laparoscopic myomectomy are comparable to abdominal myomectomy (Table 3). This appears to be true for patients who require laparoscopic repair of the endometrial cavity (60, 61), have large myomas (62), and for those who require IVF (63). One randomized controlled trial and three casecontrol studies are described. The only prospective randomized study to evaluate pregnancy rates after laparoscopic and abdominal myomectomy was published in 2000 by Seracchioli et al. (23), and there was no difference in pregnancy rates or outcomes. The study included 56 women who attempted to conceive after laparoscopic myomectomy, and 59 after abdominal myomectomy. All women were instructed to delay pregnancy for 6 months postoperatively. The pregnancy rate was 54% with laparoscopic myomectomy, not different compared to 56% with abdominal myomectomy. The cumulative pregnancy Fertility and Sterility 9

10 rates by Kaplan-Meier analysis was approximately 26% by 12 months in both groups, approximately 45% by 2 years, and approximately 50% by 3 years. The abortion rate was 20% in the laparoscopic myomectomy group and 12% with open myomectomy, but this difference was not significant. There was no difference in the incidence of preterm deliveries. The cesarean section rate was 78% in the abdominal myomectomy group, and 65% with laparoscopic myomectomy, corresponding to a vaginal delivery rate of 22% and 35%, respectively. Three case control studies have evaluated pregnancy outcomes after laparoscopic myomectomy. Bulletti et al. (64) retrospectively compared pregnancy outcomes in three groups of infertile women: [1] women with uterine fibroids and no surgery, [2] laparoscopic myomectomy, and [3] unexplained infertility. Delivery rates were significantly higher with laparoscopic myomectomy (42%) compared to untreated women with fibroids (11%) and women with unexplained infertility (25%). The spontaneous abortion rate was higher in the untreated myoma group compared to either the laparoscopic myomectomy group or those with unexplained infertility, although the difference did not meet statistical significance. Stringer and colleagues retrospective case-control study compared 49 women with abdominal myomectomy and 49 women with laparoscopic myomectomy (26). Seven pregnancies were reported in the laparoscopic group, including three who had delivered by the time of publication. All three were delivered by elective cesarean section at term, and no evidence of uterine dehiscence was found. Campo and co-investigators (65) retrospectively studied 41 women with infertility attributed to uterine fibroids, including 22 treated with laparoscopic myomectomy, and 19 abdominal myomectomy patients. Postoperative outcomes were similar in both groups. Overall the pregnancy rate was 61%, and included a live birth rate of 86% and a spontaneous abortion rate of 14%. Patients who conceived were younger, had larger myomas resected, and the myomas tended to be intramural. There were no cases of uterine rupture. Goldberg et al. (66) evaluated published series and found more pregnancy complications in 53 pregnancies after uterine artery embolization compared to 139 pregnancies after laparoscopic myomectomy. Pregnancies after uterine artery embolization had significantly higher preterm delivery (odds ratio 6.2, 95% CI ) and malpresentation rates (odds ratio 4.3, 95% CI ). Postpartum hemorrhage and spontaneous abortion rates tended to be higher after embolization, but the sample size was too small to be conclusive. Many case series have described pregnancy rates and pregnancy outcomes after laparoscopic myomectomy, and the overall pregnancy rates and pregnancy outcomes compare favorably to outcomes expected with abdominal myomectomy (2, 67). Ribeiro et al. (68) retrospectively studied pregnancy outcomes in 28 infertile women who had at least one uterine myoma of 5 cm diameter resected by laparoscopy. The postoperative pregnancy rate was 65%. Seventy-eight percent of these delivered viable term infants, 43% delivered vaginally and 57% delivered by cesarean section. Twentytwo percent of pregnancies ended in spontaneous abortion. There were no complications related to myomectomy in any of these pregnancies. Landi et al. (69) evaluated pregnancy outcomes in 359 women after laparoscopic myomectomy for various indications. Seventy-two women conceived after surgery. Of these, 17% ended in first trimester spontaneous abortion, and there was one case each of ectopic pregnancy, molar pregnancy, and elective termination. The remaining pregnancies were delivered at term, 54% delivered vaginally and 46% delivered by cesarean section. There were no cases of uterine rupture or dehiscence. Campo and Garcea (37) described a prospective randomized study of 60 women undergoing laparoscopic myomectomy to determine the efficacy of preoperative GnRH-a therapy. Of the 24 infertile patients, 13 spontaneous pregnancies occurred, a pregnancy rate of 54%. The live birth rate was 85%, and 55% of these pregnancies were delivered vaginally and 45% by cesarean section. The spontaneous abortion rate was 15%. There were no cases of uterine rupture. Malzoni and co-investigators (31) evaluated the reproductive outcome of 38 infertile women with large leiomyomata undergoing laparoscopic myomectomy. The pregnancy rate at 6 months after surgery was 34%, and was 55% at 12 months. Forty-three percent of women delivered vaginally, and 57% delivered by cesarean section. The miscarriage rate was 15%, and there was one ectopic pregnancy. There were no cases of uterine rupture. Seracchioli and colleagues (61) studied pregnancy outcomes in 34 women with fibroids penetrating the uterine cavity treated with laparoscopic myomectomy. Of the 23 women who attempted pregnancy, 39% conceived within 1 year, and 78% of these delivered at term without complications. There were no cases of uterine rupture. DiGregorio et al. (52) described laparoscopic myomectomy in a series of 148 infertile women with one or more myomata 3 cm or larger. Forty-four percent of these women conceived, including 11 who conceived with in vitro fertilization. The delivery rate was 86%. No cases of uterine rupture or dehiscence was observed. Seinera and colleagues (63) reported 65 pregnancies in 54 patients after laparoscopic myomectomy. Of these, 32% were conceived with IVF. Miscarriage occurred in 12%, and there was one interstitial pregnancy. Two women delivered before 37 weeks. Cesarean section was performed in 80% of cases. There were no cases of uterine rupture or dehiscence. 10 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

11 Darai et al. (70) retrospectively described outcomes in 102 women undergoing laparoscopic myomectomy, 44 of whom wished to conceive. Of the 44, 16 conceived spontaneously and 1 by IVF, for an overall pregnancy rate of 39%. The delivery rate was 58%. There were eight vaginal deliveries, three cesarean sections, four miscarriages, two abortions, one ectopic pregnancy, and one therapeutic abortion. There were no cases of uterine rupture or dehiscence. Nezhat and colleagues (71) retrospectively reported pregnancy outcomes after laparoscopic myomectomy in 115 women, including 34 with infertility. Forty pregnancies in 31 women were described. Twenty-nine delivered at term, including 6 vaginal deliveries, 21 cesarean sections at term, and 2 with unrecorded methods of delivery. There was 1 cesarean delivery at 26 weeks, 8 first trimester losses, 1 ectopic, 1 elective termination, and 3 ongoing pregnancies. There were no cases of uterine rupture. Miller and co-authors (72) followed 40 infertile women after laparoscopic myomectomy. The pregnancy rate was 75% by 1 year, and 87% of women delivered at term. There were no cases of uterine rupture. Rossetti and colleagues (73) evaluated 29 women desiring pregnancy after laparoscopic myomectomy for symptomatic myomas. The pregnancy rate was 66%, and two women conceived twice during the follow-up period. Of those with no other associated infertility factors, 7/10 became pregnant, and 9/10 of those who had a myoma encroaching the cavity or rapid myoma growth conceived. Nine women had a cesarean section, four delivered vaginally, one had placental failure at 28 weeks, and four miscarried. No uterine ruptures were identified. Dessolle et al. (74) retrospectively evaluated 103 infertile women with 2 or more years of infertility who underwent laparoscopic myomectomy. The overall pregnancy rate was 41%, and 80% of these conceived naturally. Age had a significant effect on prognosis. The pregnancy rate for women under age 35 years was 28/34 (82%), and was 0/22 in women aged 40 years. The pregnancy rate was high (73%) in women with otherwise unexplained infertility, compared to 23% in those with multifactorial infertility. Duration of infertility was likewise important, as the pregnancy rate was 75% with less than 3 years of infertility, but only 15% with 3 or more years of infertility. Size and location of myomas did not affect pregnancy rates in this study, although there was a trend toward higher pregnancy rates in women who had preoperative distortion of the uterine cavity. Seventy-seven percent delivered vaginally in this series, and there were no cases of uterine rupture. A follow-up report from this group found no difference in pregnancy rates with 88 laparoscopic myomectomies compared to 18 laparoscopic conversions to abdominal myomectomy (75). Dubuisson et al. published several reports describing pregnancy outcomes after laparoscopic myomectomy (62, 76 78). The first, published in 1996, retrospectively reviewed 21 infertile patients who underwent laparoscopic myomectomy for a myoma measuring 5 cm or greater (62). Seven (33%) women conceived, four spontaneously and three with IVF. Four were delivered by cesarean section, and there were no cases of uterine rupture in this series. Two reports in 2000 by this group surveyed 91 women who had previously undergone laparoscopic myomectomy (76, 78). Of the 81 responders, 53% conceived. The 2-year cumulative conception rate was 51%, including 7% who conceived by assisted reproduction techniques (ART). Of those with no other associated infertility factors, the 2-year pregnancy rate was 70%, and this number fell to 32% if other confounding infertility factors were present. These combined studies indicate that laparoscopic myomectomy is a feasible option for infertile women. The best prognosis for future fertility is found in young women with otherwise unexplained infertility when a myoma distorts the endometrial cavity. The data are not sufficient to determine whether routine vaginal delivery should be attempted, or whether cesarean section should be advised. Uterine Rupture After Laparoscopic Myomectomy One of the major concerns about laparoscopic myomectomy in a woman of reproductive age is the risk of uterine rupture during pregnancy or labor due to insufficient closure or healing of a laparoscopic myomectomy incision. There are several case reports of uterine rupture after laparoscopic myomectomy (79 86). The incidence is difficult to determine, as most of these reports do not describe the incidence per number of procedures performed, with the exception of the studies by Dubuisson (62) and Mecke (79) and their colleagues. In addition, the method of closure is not cited in all reports. However, when laparoscopic myomectomy is performed by experienced surgeons, uterine rupture or dehiscence is a very infrequent complication after laparoscopic myomectomy (Table 3). Few clinical series have found uterine rupture during pregnancy. A long-term survey by Dubuisson et al. (77) found 145 pregnancies in 98 women resulting in 101 deliveries. Seventy-two women were allowed a trial of labor, and 81% of these women delivered vaginally. There were no cases of uterine rupture during labor. However, three cases of spontaneous uterine rupture were identified, but only one occurred at a laparoscopic myomectomy site. This rupture at the myomectomy site occurred at 34 weeks in a 3-cm posterior myomectomy closed initially with one layer of four stitches of 3-0 Vicryl, and later found to be a fistula site that was repaired with a figure-of-eight suture. The second rupture occurred at 25 weeks in an abdominal myomectomy site in a woman who had undergone laparoscopic and abdominal myomectomies in two separate procedures (77). The third rupture occurred at 34 weeks in a tubocornual reanastomosis site in a women who had separate open tubal reanastomosis Fertility and Sterility 11

12 and laparoscopic myomectomy. In all cases of uterine rupture, the mother and child reported a favorable outcome. Mecke et al. (79) performed 117 laparoscopic myomectomies, and converted an additional 14 cases from laparoscopy to laparotomy. The mean diameter of the myomas was 5 cm, and an average of two myomas were removed per patient. Complications included repeat laparoscopy in one patient due to secondary hemorrhage, and one patient required hysterectomy on postoperative day 3 due to ileus caused by intestinal adhesions at a uterine wound dehiscence site. One uterine rupture at 28 weeks was also reported. Uterine rupture appears to be a rare occurrence in large clinical series. Based on the clinical trials and case series, it would appear that the risk of uterine rupture during pregnancy is no higher than 1% when the myometrial incision is appropriately repaired (62). The prognosis for pregnancy in women who undergo laparoscopic myomectomy for infertility appears to be comparable to published reports of abdominal myomectomy. The indications for laparoscopic myomectomy for infertility are difficult to define based on the current literature, although most studies indicate that the most likely women to benefit from laparoscopic myomectomy are those with large (4 cm or larger) intramural myomas, especially those that distort the endometrial cavity. There is little evidence to support laparoscopic myomectomy of subserosal myomas to enhance fertility. Finally, although the risk of uterine dehiscence during pregnancy is low with proper myometrial closure, data are insufficient to determine whether cesarean section should be routinely advised for delivery after resecting large, deep, or multiple leiomyomata. ADENOMYOSIS The unexpected finding of adenomyosis presents a particular challenge during laparoscopic myomectomy due to the lack of a clearly defined dissection plane. Nezhat and colleagues (87) showed that pregnancy is possible after laparoscopic resection of adenomyosis in a cytoreduction procedure followed by myometrial repair, with a total of nine pregnancies, five term deliveries, one delivery at 36 weeks, one ongoing pregnancy, and two miscarriages. Laparoscopic resection of myometrial adenomyosis has also been shown to reduce pain, menorrhagia, and dysmenorrhea in 15 women (88). Encouraging results were also found in three women who underwent laparoscopic excision of myometrial adenomyosis with relief of dysmenorrhea and menorrhagia (89). Other surgeons describe a more conservative approach, with laparoscopy-directed biopsy, but not resection (90, 91). If resection of myometrial adenomyosis is attempted, the surgeon must be careful to resect only as much myometrium as can be repaired, being mindful that residual myometrial/ adenomyosis tissue may not pull together as well as healthy myometrium. Incomplete or inappropriate closure could increase the possibility of fistula formation, or possibly increase the risk of causing extensive adenomyosis (92). ENTRY INTO THE CAVITY Entry into the endometrial cavity presents another technical challenge during laparoscopic myomectomy. Small series show that the cavity can be adequately repaired and the prognosis for future pregnancy is good in these circumstances. Delivery by cesarean section is generally recommended in these cases. Ostrzenski (34) reported myomectomies in 32 women with fibroids penetrating into the uterine cavity. The uterus was closed in three layers. There were no complications. The researcher concluded that the operation was safe, provided good outcomes, and allowed the advantages of minimal access surgery. Stringer et al. (60) reported seven women who required endometrial repair after laparoscopic myomectomy. The uterus was repaired in three layers by an Endo Stitch device (United States Surgical, Norwalk, CT). All of these patients conceived. Of the completed pregnancies, four were delivered by cesarean section at term, one delivered vaginally at 28 weeks, and two had elective first trimester terminations. There were no cases of uterine rupture. Seracchioli et al. (61) reported on laparoscopic myomectomy in 34 women with uterine fibroids penetrating the uterine cavity. Of the 23 women who attempted pregnancy, 9 conceived within 1 year, and 7 delivered at term without complications. They concluded that laparoscopic myomectomy for fibroids penetrating into the uterine cavity was safe and provided the advantages of minimal access surgery. SURGICAL TECHNIQUES Preoperative Evaluation Women with leiomyomata present with a history of pain, pressure, bowel or bladder symptoms, and may have infertility or a history of pregnancy wastage (1). Couples with infertility should undergo a full evaluation to identify confounding factors such as male factor infertility or tubal occlusion. For example, hysterectomy may be considered for symptomatic fibroids if the male is azoospermic and donor insemination is objectionable to the couple. The pelvic examination may confirm uterine fibroids in some cases, but ultrasound must always be performed to evaluate the size, number, and location of fibroids when laparoscopic myomectomy is considered. Saline infusion sonohysterography is indicated when the relationship between the myomas and the endometrium is not clearly defined by transvaginal ultrasound evaluation. Hysteroscopy is appropriate if a submucus myoma or polyp is identified by the hysterosalpingogram or saline infusion sonohysterography. When the ultrasound findings or clinical presentation is atypical, magnetic resonance imaging (MRI) or computerized tomography (CT) scanning may provide enhanced diagnostic information and confirm the diagnosis of myomas or adenomyosis. 12 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

13 The decision to perform laparoscopic myomectomy must be made based on the size, number, and location of the myomas and the experience of the surgeon. In our practice, we usually limit laparoscopic myomectomy to women with four or fewer myomas. Furthermore, we determined that laparoscopic myomectomy for infertility would be most appropriate for women when an intramural myoma of 4 cm or greater is identified, or if an intramural myoma compromises the uterine cavity, regardless of size. Submucus myomas are treated by hysteroscopic resection. Surgery is not performed for subserosal myomas unless indicated by other symptoms including pain, pressure, bladder or bowel symptoms. Laparoscopic Myomectomy Techniques The goal of laparoscopic myomectomy should be a repair that is comparable, or superior, to closure of abdominal myomectomy. The deep intramural myomectomy incision is closed in three layers deep myometrium, superficial myometrium, and serosa (93). If comparable results are to be obtained with laparoscopic myomectomy, a three-layer closure (or four-layer, if the endometrium is entered) should be performed. Only a reproductive surgeon who has the technical experience to repair a myomectomy incision with multilayer uterine closure and extensive experience with advanced laparoscopic surgery should attempt laparoscopic myomectomy. In addition, a surgical assistant with advanced laparoscopic skills is essential. Although many techniques of laparoscopic myomectomy have been described in the literature, recent advances in instrumentation and techniques make this complex procedure more feasible for surgeons with advanced laparoscopic skills. In this section, we describe techniques and instruments that have simplified laparoscopic myomectomy, generally following techniques described by Koh and Janik (94). Trocar placement is critical to allow myoma dissection and countertraction during enucleation, facile needle passage, pick-up, and assistance during suturing, the most difficult aspect of laparoscopic myomectomy. The laparoscope is placed through an umbilical incision, although placement in the left upper quadrant at Palmer s point, a midclavicular position under the lower left rib, may be used to evaluate extremely large myomas or when pelvic adhesions are expected (95, 96). Two 5-mm operating trocars are placed on the primary surgeon s side of the table, with one suprapubic trocar placed lateral to the insertion of the round ligament and the other trocar placed lateral to the umbilical trocar, parallel to the suprapubic trocar and lateral to the inferior epigastric vessels (Fig. 1). A 10- to 11-mm trocar is placed on the contralateral side, approximately midway between the pubic symphysis and the umbilicus, and lateral to the inferior epigastric vessels. Initially, a 5-mm reducer is placed over the large 10- to 11-mm port. Eventually, the large port will be replaced by the morcellator trocar. However, the morcellator port valve is easily damaged and leaks with extensive FIGURE 1 Placement of trocars for laparoscopic myomectomy. The laparoscope is placed through an umbilical incision (D), although the left upper quadrant may be used as an alternate site (95, 96 Golan 2003; Jansen 2004). Two 5-mm trocars are placed on the primary surgeon s side of the table: a suprapubic trocar (B) lateral to the insertion of the round ligament and the other trocar (C) placed lateral to the umbilical trocar, parallel to the suprapubic trocar and lateral to the inferior epigastric vessels. A 10- to 11-mm trocar (A) is placed on the assistant s side midway between the pubic symphysis and the umbilicus and lateral to the inferior epigastric vessels (Inf. epigastric a., v.). These trocar sites (A, B, C, D) are used in subsequent figures. Artwork by Nancy Marshburn. Hurst. Laparoscopic myomectomy. Fertil Steril use; therefore we prefer to place a standard 10- to 11-mm operative trocar initially. A manipulator probe is placed into the cervix, and this manipulator can be a tremendous benefit to move the uterus into optimal position during enucleation and suturing. Once the instruments are placed, a thorough assessment of the pelvis is performed, and the feasibility of performing laparoscopic myomectomy is reassessed. Conversion to laparotomy in difficult cases should be considered a sign of wisdom, not evidence of defeat. The preoperative consent should reflect the possibility that abdominal myomectomy may be needed if the procedure cannot be completed safely by laparoscopy. Fertility and Sterility 13

14 Occasionally, a deep intramural myoma cannot be localized by laparoscopy. Sometimes a probe can be brushed over the uterus to provide some tactile information about the location of a myoma, but this technique is less reliable than the direct manual palpation during abdominal myomectomy. When needed, intraoperative transvaginal ultrasound can help direct the surgeon to identify the location of the myoma, and identify the optimal site for the uterine incision (97). Ideally, a transvaginal ultrasound is immediately available in the operating suite, and the surgical assistant can break scrub to assist with the ultrasound. Ultrasound visualization may be enhanced by filling the pelvis with saline. Once the myoma is identified by ultrasound, the primary surgeon passes a probe over the surface of the myoma to confirm the size and location of the myoma, and to determine the incision site. The myomas are injected with a dilute solution of vasopressin (1 U/mL). Injection with a vasopressor helps to reduce blood loss, decreases surgical time, and increases the ease of resection. However, it may increase the incidence of adhesions and could potentially be a factor in delayed bleeding (98). We inject 5 10 ml into the myoma interface at one or more sites with an 18-gauge spinal needle placed directly through the abdominal wall and wait until blanching occurs. Alternately, 0.25% bupivicaine with epinephrine can be injected. A recent prospective randomized study showed that this regimen reduces blood loss, operative time, and postoperative pain compared to injection of saline for laparoscopic myomectomy (99). Although many instruments have been used to facilitate myoma resection, the Harmonic Scalpel provides a very clean dissection and appears to be associated with less blood loss compared to unipolar cautery (30). When placed in the myometrial/myoma interface, the Harmonic Scalpel passes through the plane of dissection smoothly and easily. We prefer a hook, and use the hook to make the initial incision and the smooth side to dissect the myoma from the myometrium. FIGURE 2 Myoma enucleation. A transverse elliptical incision exposes the myoma after injection with vasopressin. The Harmonic Scalpel is placed through port B. An aspirator/irrigator placed through the ipsilateral port C for countertraction against the myometrium, to identify tissue planes, and rinse the surgical site. The assistant places a tenaculum or a myoma screw through the 10- to 11-mm port A to provide tension on the myoma (M). Gentle pressure is placed on the edge of the Harmonic Scalpel blade to facilitate cutting and coagulation along the pseudocapsule. If needed, the Harmonic Scalpel is inserted through ports A or C to facilitate dissection. Hurst. Laparoscopic myomectomy. Fertil Steril The direction of the incision is extremely important (Fig. 2). A transverse incision can be sutured easily, whereas suturing a vertical incision can be a frustrating challenge. The surgeon places the Harmonic Scalpel hook through either of the parallel ports, and places an aspirator/irrigator probe through the ipsilateral sheath to provide countertraction against the myometrium, help identify tissue planes, and rinse blood from the surgical site. The aspirator/irrigator can bluntly dissect the myoma from the myometrium by pushing the myometrium directly away from the myoma. The assistant places a large self-locking tenaculum or a myoma screw through the contralateral 10- to 11-mm port to provide tension. An initial transverse elliptical incision is preferable to a simple linear incision, as it allows resection of some of the myometrium covering the myoma and facilitates later myometrial closure by reducing the amount of redundant tissue. Once the myoma is exposed, countertraction is accomplished with the tenaculum grasper; the primary surgeon s irrigator probe pushes away the myometrium with an aspiration/irrigator. The Harmonic Scalpel blade is placed with the flat surface on the myoma and gentle pressure is placed on the edge of the Harmonic Scalpel blade to facilitate cutting and coagulation along the pseudocapsule. If needed, the Harmonic Scalpel can be eventually placed through the surgical assistant s port, and the myoma retracted from the primary surgeon s side, to facilitate dissection on the assistant s side. Extreme tension on the myoma during dissection is avoided to prevent the risk of endometrial avulsion. Keeping the scalpel blade in direct contact with the myoma reduces the chance of entering the endometrial cavity. The myomectomy dissection by laparoscopy is usually remarkably clean and relatively bloodless (Fig. 3). Once removed, the myoma is placed in the anterior or posterior cul-de-sac. Closing the incision is technically the most challenging aspect of laparoscopic myomectomy, but advances in surgical instruments allow for excellent approximation of all layers. In the past, simply loading a curved needle properly into a laparoscopic needle driver could be frustrating, as laparoscopy provides a two-dimensional image. As a result, 14 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

15 FIGURE 3 Uterine defect after myoma enucleation. The combination of vasopressin and the Harmonic Scalpel provides good hemostasis. The myoma M has been placed in the anterior cul de sac. FIGURE 4 Technique of laparoscopic suturing. The deep myometrial layer is closed with a continuous running suture. A Lapra-Ty clip has been placed at the beginning of the suture line. A self-righting needle driver is placed through port B and held in a right-handed surgeon s right hand. A second needle driver is inserted through port C and held in the left hand. The assistant uses a third-needle driver inserted through port A to provide traction on the suture. Sutures are placed anterior to posterior, from the distal end of the incision to the proximal end. The needle is placed into the superior margin, then is pulled out by needle driver C and immediately reloaded into driver A. Artwork by Nancy Marshburn. Hurst. Laparoscopic myomectomy. Fertil Steril the needle may be placed at an improper angle with a standard laparoscopic needle driver causing the needle to move during suturing and making precise suturing difficult or impossible. This difficulty is overcome with a self-righting needle driver, which has an open basket for the lower jaw that snaps a curved needle securely into place with the correct right angle. The self-righting needle driver is held in the primary surgeon s right hand, and a second grasper is held in the primary surgeon s left hand. A right-handed surgeon sews from right to left, in other words, passing the needle into and out of the tissue layers with the self-righting needle driver in the right hand, then grasps the needle with the left hand needle driver. The assistant uses a third suturegrasping instrument (such as a needle driver) or irrigator, as needed, to expose the tissues. The surgeon holds both needle drivers, and the assistant holds the camera and grasper/ irrigator. If the endometrium is entered during the myomectomy, it is closed with interrupted sutures of 4-0 Vicryl. The lateral sites are sutured and tied first, and the sutures held to allow slight tension to be placed at the corners to provide optimal exposure of the middle of the endometrium. For the endometrial closure, the interrupted sutures are tied either intracorporally or extracorporally. We prefer intracorporeal ties, as tension on the knot and visualization is good with intracorporeal tying. The principles for intracorporeal knot tying are the same as those used during microsurgery, and most surgeons with microsurgical experience can become proficient at laparoscopic intracorporeal knot tying with practice. Placement of the two operative ports on the surgeon s dominant side facilitates this process. A key point: a self-righting Hurst. Laparoscopic myomectomy. Fertil Steril needle driver should be replaced by a standard laparoscopic needle driver any time the suture must be held, as the self-righting driver will fray the suture. The endometrium closes without tension; therefore knot slippage usually is not a problem. After tying all the knots to approximate the endometrium, the surgeon holds the sutures as the assistant cuts them close to the knot. The deep myometrial layer is closed with a continuous running layer of 2-0 Vicryl or 2-0 Monocryl on a cutting needle (Fig. 4). We find Lapra-ty clips (Ethicon Endo-Surgery, Inc.) a tremendous help at this stage; these clips provide good tension and comparable strength to three tied knots (100). A Lapra-ty clip, made of resorbable polyglycolic acid, is placed on the free end of the suture, then the assistant introduces the needle and suture into the abdomen through the 10- to 11-mm port. The surgeon grasps the suture near the needle with the left needle driver and rotates the needle into approximately the desired position. Next, the needle is grasped with the self-righting needle driver in the right hand. Alternately, the needle can be held by the left Fertility and Sterility 15

16 driver and stabilized while grasping the needle with the right driver. A curved needle allows the surgeon to take a deep bite into the upper then the lower margins of the myometrium. Sewing proceeds from an anterior to posterior direction, from the distal end of the incision to the proximal end. Initially, the needle is placed into the superior margin, then is pulled out by the left grasper, and immediately reloaded into the right self-righting driver to suture the inferior margin of the deep myometrium. The left needle driver pulls the suture through, and the Lapra-ty clip secures the suture into the myometrium, without requiring a knot. Alternately, the first suture can be tied, but as this bite may be under tension, the knot may slip and result in an air knot. The clip holds the distal end of the suture securely. The assistant holds the suture firmly at a 90-degree angle with a needle driver to prevent slippage of the suture line. The needle is reloaded into the right self-righting needle driver, and the second bite is made, top to bottom, using the principles described for the first knot, in a continuous running layer. The assistant follows by keeping constant tension on the suture line to avoid slippage. At the conclusion of the deep running layer, another Lapra-ty clip is placed at the end of the running suture, or the suture may be tied, according to the surgeon s preference. To use the clip, the primary surgeon firmly holds the suture line with tension, and the assistant inserts the Lapra-ty clip through the 10- to 11-mm port. The clip is placed around the suture, then pushed down slightly against the uterus to ensure that proper tension is maintained, then clipped onto the suture. The free ends are cut. If the suture is tied, a high level of technical skill is needed to avoid slippage of the suture line, and the selfrighting driver must be replaced with a standard laparoscopic needle driver (or an extracorporeal knot tied). The superficial myometrial layer is closed with a continuous running layer of 2-0 Vicryl or 2-0 Monocryl on a cutting needle, using the same principles as for the deep myometrium. Lapra-ty clips may be placed on the proximal and distal ends of the running suture, or the suture may be tied, according to the surgeon s preference. The serosa is closed with a continuous running layer of 4-0 Vicryl or 4-0 Monocryl on a cutting or taper needle (Fig. 5). If possible, the Lapra-ty clips are buried under the serosal layer. We prefer a continuous running layer, rather than a locking layer, as a continuous closure, in theory, provides hemostasis without compromising tissue perfusion. Lapra-ty clips may be placed on the proximal and distal ends of the running suture, or the suture may be tied, according to the surgeon s preference. At this step, we prefer to tie the suture, if possible, as a tied knot exposes less foreign material on the surface than the clip. The fibroids are removed with a mechanical morcellator. The 10- to 11-mm trocar sheath is removed from the assistant s side, and replaced with the morcellator trocar. The assistant grasps the myomas with the morcellator tenaculum and pulls the myomas into the activated morcellator. The FIGURE 5 Uterus after three-layer closure. The serosa is closed with a continuous running layer. If possible, the Lapra-ty clips are buried under the serosal layer. A Lapra-ty clip (arrow) has been placed at the terminal end of the running suture. Hurst. Laparoscopic myomectomy. Fertil Steril primary surgeon uses graspers to feed the myomas into the morcellator, ideally with peripheral myoma resection in a fashion similar to peeling an apple. Myoma strips and fragments are removed through the morcellator channel. Extreme care must be taken to avoid injury to abdominal organs during morcellation, and the cutting blade is retracted except when morcellating the myomas. Alternately, the myomas may be removed by alternate methods including a colpotomy incision in the posterior cul-de-sac or by minilaparotomy. After removal of myomas, the uterus is irrigated and hemostasis confirmed. The myomectomy incisions are covered by a single sheet of Interceed to reduce adhesions after hemostasis has been achieved (55). The sheet is cut, if needed, then rolled and held with graspers. The assistant places the sheet onto the anterior uterus, and then the primary surgeon unrolls the sheet. After it has been unrolled, the sheet is moistened with saline (Fig. 6). Use of an adhesion barrier is most important for posterior and fundal myomectomy sites, as the risk of adhesions are greatest at these locations (45). The 10- to 11-mm and morcellator trocars increase the risk of herniation into the internal fascial sheath, and this layer should be closed separately to avoid herniation. We use a suture passer needle, and place the needle through one side of the internal fascia under laparoscopic guidance. The primary surgeon then grasps and holds the free intraabdominal suture with a needle driver. The needle passer is removed from the abdomen, and then placed through the other edge of the fascia. The primary surgeon places the free suture into the needle passer, and the assistant removes the passer with 16 Hurst et al. Laparoscopic myomectomy Vol. 83, No. 1, January 2005

17 FIGURE 6 Placement of Interceed (arrow) over the ectomy incision to reduce adhesions. Hurst. Laparoscopic myomectomy. Fertil Steril the free end. A simple suture knot is tied, or a figure-of-eight suture can be placed if there is still a fascial defect. The skin and all other incisions are closed per normal routine. Postoperative recovery after laparoscopic myomectomy is longer than for diagnostic laparoscopy, but most women may return to full activities within 1 2 weeks of surgery. Active ambulation is advised. Regular diet is initiated as tolerated, usually after the postoperative nausea from anesthesia is resolved. ALTERNATIVES TO MYOMECTOMY Although hysterectomy and myomectomy has been the primary treatment for the management of symptomatic fibroids, new treatments provide alternatives to surgical extirpation. Eventually, current surgical strategies including laparoscopic myomectomy may be replaced by approaches available now or in the near future. Promising treatments in some circumstances may include minimally invasive surgical approaches such as hysteroscopic myomectomy and myolysis, ultrasound and other imaging-directed means of myoma ablation including uterine artery embolization, and pharmaceutical agents. Hysteroscopic Myomectomy Symptomatic submucosal fibroids are effectively managed by transcervical resection or ablation. Hysteroscopic myomectomy affords safe, efficient, minimally invasive, outpatient treatment. However, clinical outcomes are reported primarily in case series. A summary of 11 retrospective case series indicated that hysteroscopic myomectomy resulted in satisfactory resolution of abnormal uterine bleeding in 80% of women (101). Repeat surgery was required in 16% at 4 years after the initial hysteroscopic resection (102). Coexistent intramural fibroids and myoma size greater than 4 cm were associated with more treatment failures after hysteroscopic myomectomy. However, pretreatment with GnRH-a in women with myomas more than 3 cm or myomas with a significant intramural component reduced rates of abnormal uterine bleeding after hysteroscopic myomectomy (103). Submucosal fibroids with deep intramural extension can be safely resected even when the distance from the outer myoma margin to the uterine serosa is 5 mm or more (104). Intraoperative ultrasound guidance is a useful aid for fibroid localization and directing resection (105). Adding endometrial ablation or resection after hysteroscopic myomectomy improves the long-term outcome symptomatic relief from menorrhagia if future fertility is not desired (2). The effect on fertility after hysteroscopic myomectomy is unclear (106). Assessing the impact of hysteroscopic myomectomy on subsequent fertility is complicated by difficulty in controlling for multiple infertility factors, the number of submucus myomas resected, and the presence of coincident intramural fibroids (107). In one review, pregnancy occurred in 59% of infertile women with submucosal myomas after hysteroscopic myomectomy (108). Endometrial Ablation Most endometrial ablation techniques, including the thermal balloon, bipolar mesh, and cryogenic devices are approved by the U.S. Food and Drug Administration (FDA) for women with intramural fibroids less than 2 cm, but not for the treatment of submucosal myomas. One study of hysteroscopic endometrial ablation had a failure rate of 40% when myomas were present, compared with a failure rate of 5% in women with a normal uterus (109). In a 5-year follow-up of 255 patients from a randomized control trial comparing rollerball endometrial ablation with thermal balloon ablation, one-third of 42 hysterectomies were for myoma-related symptoms (110). The hydrothermal ablation (HTA) system circulates free-flowing 90 C saline under hysteroscopic view, and has FDA approval for use with women having submucosal myomas up to 4 cm. The HTA system was used for endometrial ablation in 22 patients with submucosal fibroids. A 91% patient satisfaction rate was achieved at months follow-up with a 9% failure rate requiring additional surgery (111). Equipment modification to prevent activation of the ablation energy source unless optimum safety parameters are achieved has reduced complications from use of the endometrial ablation devices. Myolysis Electrical, thermal, and ultrasound energy sources have been used to coagulate and devascularize symptomatic myomas. Current methods of myolysis have achieved success in relieving symptoms relating to myoma volume, but its safety for women desiring pregnancy is unknown. Fertility and Sterility 17