Endoscopic versus laparotomy management of endometriomas*

FERTILITY AND STERILITY Copyright e 1994 The American Fertility Society Printed on acid-free paper in U. S. A. Endoscopic versus laparotomy management of endometriomas* Bruce G. Bateman, M.D.t:j: Lisa A. Kolp, M.D. t Stacey Mills, M.D. University of Virginia Health Sciences Center, Charlottesville, Virginia Objective: To compare the surgical management and follow-up of patients with endometriomas managed by endoscopic surgery versus laparotomy using a retrospective case control format. Design: Endoscopic oophorocystectomies were performed on 36 patients. Chart review of laparotomy oophorocystectomies from 21 patients was conducted. Six-week and 12-month follow-up for evaluation of symptoms, evidence of recurrence, and fertility was available on all subjects. Results: In the endoscopy group, 39 patients had screening laparoscopy for possible endoscopic surgery. Three of this group required laparotomy and 36 patients underwent endoscopic surgery. Chart review identified 21 patients who had undergone primary laparotomy for endometriomas. Patient groups were matched for age, severity of disease, presence of other infertility factors, and absence of perioperative medical suppression. Outcome parameters for each group were: operating time-endoscopy 2.8 hours (±1.2), laparotomy 3.1 hours (±1.8); estimated blood loss-endoscopy 40 cc (±45); laparotomy, 240 cc (±107); recovery time-endoscopy, 6.2 days (±2.5), laparotomy 30 days (±6.8); endometrioma recurrence rate-endoscopy 11.1 %, laparotomy 19%; and pregnancy rate-endoscopy 42.8%, laparotomy 46.6%. Conclusion: A high percentage of patients with endometriomas associated with advanced endometriosis can be managed effectively by endoscopic surgery. Fertil Steril1994;62:690-5 Key Words: Endometriosis, surgical treatment, laparoscopic treatment Endoscopic surgery is rapidly replacing laparotomy for many gynecologic procedures. The advantages of endoscopy are clear in terms of recovery time and postoperative discomfort. There is a growing concern about the rush to accept this technique in the absence of documented risks and comparative data related to laparotomy. Some degree of reporting bias in success versus failure data seems probable. Endoscopic ovarian surgery has been a focus of dispute related to adequacy of surgery and spill of cystic malignant lesions. Although endoscopic management of stage I and II endometriosis is widely accepted, endoscopic treatment of endometriomas is controversial. This report compares the surgical management and follow-up of patients with endometriomas associated with stage III or IV (revised American Fertility Society [AFS] (1) endometriosis managed by endoscopic surgery versus laparotomy. Received August 30, 1993; revised and accepted May 5, 1994. * Presented at the 49th American Fertility Society Conjoint Meeting with the Canadian Fertility and Andrology Society, Montreal, Canada, October 9 to 14, 1993. t Department of Obstetrics and Gynecology. :j: Reprint requests: Bruce G. Bateman, M.D., Department of Obstetrics and Gynecology, Box 387, University of Virginia Health Sciences Center, Charlottesville, Virginia 22908 (FAX: 804-982-1840) 0 Department of Pathology. MATERIALS AND METHODS The surgical procedures in this report were performed by the authors at the University of Virginia Health Sciences Center. Endoscopic surgeries were performed between 1989 and 1992. Thirty-nine patients underwent screening laparoscopy for possible endoscopic management of endometriomas. 690 Bateman et al. Endoscopy for endometriomas Fertility and Sterility

Table 1 Characteristics of Patient Groups Age* Revised AFS stage and score Stage III:j: Stage IV:j: Score* Other infertility factors present(%) Perioperative medical suppression Laparoscopy 33 ± 4 20 (56) 16 (44) 39 ± 29 43 None *Values are means± SEM. t p > 0.05. :j: Values in parentheses are percentages. Laparotomy 34 ± 4t 11 (52)t 10 (48lt 44 ± 32t 47t None Thirty-six had definitive endoscopic procedures and three required immediate laparotomy. Retrospective chart review of 31 cases identified 21 patients who had undergone laparotomy (before 1989) for conservative surgical management of endometriomas and satisfied matching criteria, which included age, severity of disease, presence of other infertility factors, and absence of perioperative medical suppression. In the laparoscopy group, 15 ( 42%) had pain as the primary surgical indication and 21 (58%) had infertility (10 of 21 [48%] also had pain). In the laparotomy group, 6 (29%) had pain as the primary surgical indication and 15 (71%) had infertility (7 of 15 [ 4 7%] also had pain). General characteristics of the study groups are shown in Table 1. The laparoscopy versus laparotomy groups were matched for age (P > 0.05, t-test), severity of disease (P > 0.05, x 2 ), presence of other infertility factors (P > 0.05, x 2 ), and absence of perioperative ovarian suppression. All patients undergoing elective surgery in both groups had preoperative pelvic sonography. Sonographic criterion for endoscopic surgery was absence of papillary excrescences or septations on ultrasound. All procedures were performed under general anesthesia. Elective procedures (n = 31) were performed during the follicular phase of the menstrual cycle. Five cases were seen initially as intraoperative consultations. Operative laparoscopy was accomplished by video control via subumibilical incision, a lower abdominal midline incision, and two lateral lower abdominal incisions. Instrumentation included 5-mm scissors, graspers, needle holders, and C0 2 laser with wave guide. Irrigation was performed with lactated Ringer's solution containing 2,000 U Heparin/1,000 cc. Endoscopic technique included thorough inspec- tion of the pelvic organs, pelvic peritoneum, diaphragm, and liver. Endometriomas were decompressed by incision and suction drainage. Lysis of adhesions then was performed, using sharp dissection, to fully mobilize the ovaries. Ovarian incisions were made in the thinnest area of the cyst wallovarian cortex. Cyst walls were stripped from normal ovarian tissue and/or incised by sharp dissection and removed at the conclusion of the procedure. Hemostasis was maintained with endocoagulation and suture ligatures of 4-0 polydioxanone with intracorporal knot tying. Areas of superficial active endometriosis involving the other ovary or pelvic peritoneum were vaporized by C0 2 laser or cauterized by endotherm coagulation. Ovaries with endometriomas >3 em in diameter were reconstructed with 4-0 polydioxanone on a 2-cm taper needle using intracorporal knot tying. Interrupted sutures were placed just under the edge of the cortex to invert the suture line. When adhesions to the pelvic sidewall were present, Interceed (Johnson and Johnson Co., Somerville, NJ) was placed at the conclusion of the procedure. Laparotomies were performed by a Pfannensteil (81%) or midline (19%) incision. Microsurgical technique was used in all procedures. Intact removal of endometriomas was attempted. Lysis of ovarian adhesions was performed by microtip electrosurgery. Endometriomas were excised by sharp dissection and microtip electrosurgery. Ovaries were reconstructed with a layered closure of 3-0 and 5-0 polyglactin using inverting technique. Superficial implants of endometriosis elsewhere in the pelvis were fulgurated by electrosurgery. Postoperative parenteral dexamethasone suppression was used for adhesion prophylaxis in all cases. No patient in either the laparotomy or laparoscopy group had perioperative ovarian suppression. Follow-up was accomplished at 6 weeks and 12 months in both groups. Follow-up included pelvic examination and review of history with regard to pain resolution. Patients with pain (dysmenorrhea, dyspareunia, general pain) as a presenting symptom were asked to compare the pain level before surgery with the level at 12-month follow-up. Pain was described as worse, unchanged, improved, or resolved. In patients attempting to conceive, pregnancies were documented by presence of an intrauterine gestational sac on ultrasound (US). Second-look,laparoscopy was performed for recurrent or persistent ovarian enlargement, continued infertility, or recurrent pain. Patients who underwent second Bateman et al. Endoscopy for endometriomas 691

surgeries during the follow-up period had recurrent endometriomas documented histologically. Forty-three percent of patients in the laparoscopy group and 4 7% of patients in the laparotomy group had other infertility factors, including anovulation, male factor, and cervical factor. These factors were treated and corrected in all patients. RESULTS Thirty-nine patients underwent screening laparoscopy for possible endoscopic surgery. Three required laparotomy because of severe adhesions or bowel involvement. Thirty-six patients underwent endoscopic surgery with removal of 49 endometriomas. Endometrioma cyst walls were stripped easily from normal ovarian tissue in 51%, additional sharp dissection was required in 45%, and sharp dissection was required throughout removal in 4%. In 4% of endometriomas, cyst wall dissection was judged incomplete due to lack of a clear tissue plane and the remaining lesions were endocoagulated. Seventy-one percent of the ovaries were affected by adhesions. Mean blood loss, operating room and recovery time are displayed in Table 2. One of 36 patients was hospitalized overnight for control of pain and vomiting. The remainder of the procedures were performed on an outpatient basis. Of the 36 patients undergoing endoscopic surgery, one required a laparotomy 3 months after the original procedure. That patient underwent oophorectomy and segmental resection of the sigmoid colon for a recurrent endometrioma and colonic en- Table 2 Characteristics of Patient Groups Primary indication Infertility Pain Estimated blood loss (cc)t Endometrioma diameter 2to3cm 4 to 6 em 7to9cm Operating time (h) Recovery time (d) Endometrioma recurrence (%) Pregnancy rate(%) Laparoscopy 21 (58)* 15 (42) 40 ± 45:j: 12 (24) 29 (58) 9 (18) 2.8 ± 1.2 6.2 ± 2.5 11.1 42.8 * Values in parentheses are percentages. Laparotomy 15 (71) 6 (29) 240 ± 107 7 (23)11 14 (46)11 9 (30) II 3.1 ± 1.811 30 ± 6.8 19.011 46.611 t Blood loss estimated by irrigation versus suction volumes and sponge counts. :j: Values are means± SEM. p < 0.05. II P > o.o5. dometriosis. By one year follow-up visit, of the remaining 35 patients, 18 (72%) patients presenting with pain were pain free, 5 (20%) had decreased pain, and 2 (8%) had no improvement. Nine (43%) of 21 patients attempting pregnancy conceived (7 of 15 [ 4 7%] patients with stage III disease, 2 of 6 (33%) patients with stage IV disease). Of the remaining 27 patients, including the 15 not attempting conception and the 12 who failed to conceive, 3 developed ovarian enlargement in the previously operated ovary. These three patients underwent a second laparoscopic oophorocystectomy with removal of three histologically confirmed endometriomas. Of the remaining 24 patients, 13 had repeat laparoscopy for pain or infertility and, of that group, 6 had recurrent or persistent superficial disease. There were no recurrent endometriomas in this group. Eleven (69%) patients who underwent second laparoscopy had recurrent ovarian adhesions. The mean adhesion score (revised AFS) was reduced from 10.8 ± 2.2 to 1.7 ± 0.6 (P = 0.01) (1). In the laparotomy group, 30 endometriomas were removed from 21 patients. Intact endometrioma removal was accomplished in two (7%) cases. Cyst wall removal was complete in 28 (93%) cases. In the remaining two (7%), removal was incomplete and the remaining tissue was fulgurated by electrosurgery. Thirty-one (79%) ovaries were affected by adhesions. Mean blood loss, operating room time, and recovery time are displayed in Table 2. By the 1- year follow-up visit, 10 of 13 (77%) patients presenting with pain were pain free, 1 (8%) had decreased pain, and 2 (15%) had no improvement. Seven of the 15 (46%) patients attempting pregnancy conceived ( 5 [50%] of 10 patients with stage III disease; 2 [40%] of 5 patients with stage IV) (1). Of theremaining 14 patients, including the 6 not attempting to conceive and the 8 who failed to conceive, four developed ovarian enlargement in the previously operated ovary. All four underwent repeat laparotomy and oophorectomy for recurrent endometriomas. Of the remaining 10 patients, 5 had laparoscopy for pain or infertility during the follow-up period. Four of five had recurrent superficial endometriosis and one had a 2-cm endometrioma in the contralateral ovary. Six (67%) of nine patients who underwent repeat laparotomy or follow-up laparoscopy had recurrent ovarian adhesions. The mean adhesion score (revised AFS) was reduced from 9.2 (± 4.3) to 2.6 (± 1.9) (P = 0.013) (1). There were no major intraoperative complications in either group. In the endoscopy group, one 2-cm straight needle was sheered off in the anterior 692 Bateman et al. Endoscopy for endometriomas Fertility and Sterility

abdominal wall during removal because of improper technique and two patients undergoing prolonged procedures developed hypothermia. In the laparotomy group there was one postoperative wound infection. Endometrial glands and stroma were present in 86% of the endometriomas in the study (85% endoscopy versus 88% laparotomy). The surgical and follow-up parameters for laparoscopy versus laparotomy groups showed significant differences in blood loss, (P = 0.022, t-test) and recovery time (P = 0.001, t-test). Recovery time to full activity was 6.3 (± 2.5) days for the endoscopy group versus 30.1 (± 6.8) for laparotomy patients (P = 0.001, Wilcoxon Rank Sum Test). Thirty patients from our practice who had undergone diagnostic laparoscopy (no positive findings) were interviewed by telephone and reported a return to full and normal activity with pain resolution by 2 to 5 days (2.9 ± 0.9) compared with the endoscopic surgery group with return to full activity at 3 to 10 days (6.3 ± 2.5) (P < 0.05, Wilcoxon Rank Sum Test). There were no differences in endometrioma size (P > 0.05, t-test), endometrioma recurrence rate (P > 0.05, Fishers Exact Test), presence of glands and stroma on histology (P > 0.05, Fishers Exact Test), or pregnancy rate (P > 0.05, Fishers Exact Test). DISCUSSION Mettler and Semm (2) pioneered endoscopic management of endometriomas. Several techniques have been described, including excision of the cyst wall with and without suture repair of the ovary (2); incision and drainage without removal of the cyst wall (3); laser or cautery ablation of the cyst wall (2, 4). There is little basis for comparison of results in most endoscopic studies because of different grading systems, follow-up, and expression of data. In general, results of most laparoscopic series are similar to laparotomy (5). Adamson et al. (6), using a prospective, controlled study design, compared laparoscopic to laparotomy management of endometriomas and found similar pregnancy rates. Prospective randomized comparisons have not been performed. Our data suggest that most endometriomas can be managed by laparoscopy with results comparable to laparotomy. Four of our 39 (10%) patients required a laparotomy at the time of initial endoscopic evaluation or within 3 months of originallaparoscopic surgery. Severity of adhesions was the principle determinate of approach. The laparoscopic view of some areas deep in the pelvis is clearly better than laparotomy and facilitates an endoscopic approach. On the other hand, a loss of the sense of palpation is a shortfall of laparoscopy. Bowel involvement in our patient requiring interval bowel surgery was not appreciated at the original procedure. Our endoscopic technique did not emphasize the use of lasers or disposable instruments. Once the proper plane between normal ovarian tissue and the endometrioma cyst wall was identified, gentle traction and sharp dissection provided clear separation of the cyst in most cases. Lasers were not used in this aspect of the procedure. The only disposable instruments in routine use were 5-mm accessory trocars, which we found convenient because of their small size and built-in grips. Laparoscopic suture repair of the ovary recently has been questioned (7). We attempted to mimic laparotomy technique with carefully placed inverting sutures. There were very few patients with absence of adhesions at initial surgery who had a second-look laparoscopy. This prevented a meaningful analysis of de novo adhesion formation. Overall, our study design did not address the question of ovarian suture repair. In our experience, after removal of large endometriomas, the remaining ovarian tissue does not necessarily fall together in a normal configuration. Adhesion prophylaxis differed between the study groups. Interceed was used in the laparoscopy group between the ovary and pelvic sidewall. Systemic glucocorticoid suppression was used in the laparotomy group. Adhesion reformation was similar in frequency and severity. The severity of adhesions at second look laparoscopy generally was less than the initial procedure (P 0.03). In most cases, recurrent adhesions were minimal, although in a few cases they were more severe. Detailed analysis of these data is not useful because of selection bias. Patients returning for second-look laparoscopy had continued infertility or recurrent pain. With the advent of endoscopic surgical treatment for ovarian masses, concern about rupture or delayed recognition of ovarian cancer has developed. In this regard, careful attention was focused on US findings in the endoscopy group. Transvaginal US was highly accurate in predicting endometrioma presence and size. The internal echo pattern of most endometriomas was homogeneous and medium to low level (ground glass), although two were impossible to differentiate from solid lesions-that Bateman et al. Endoscopy for endometriomas 693

CA 125 ulml 200 180 160 140 120 100 80 60 40 20 PREOPERATIVELY &WEEKS POSTOPERATIVELY Figure 1 Preoperative and 6 weeks postoperative CA 125levels (laparoscopy group). distinction was made intraoperatively. Two patients (not included in the 39) were seen for possible endoscopic surgery and disqualified because of papillary excrescences or septations. Both underwent laparotomies. One had a benign lesion and the other had a serous papillary carcinoma, stage 1C. Rupture and spill of cystic malignant lesions and delay of definitive diagnosis are obvious concerns with regard to ovarian malignancy. Intuitively, spill of the fluid from cystic malignancies is undesirable, however, there is controversy regarding the effect of spill on prognosis ( 8, 9). As a matter of technique, concern with intentional rupture and drainage of endometriomas by laparoscopy seems academic considering the 93% rupture rate during attempted intact removal at laparotomy. Preoperative CA-125 was performed on patients undergoing elective endoscopic surgery (normal range < 35 U/mL). Fifty-two percent of the group had elevations ranging from 39 to 220, as shown in Figure 1. Seventy-five percent returned to normal by the time of the 6-week follow-up visit and 25% remained elevated. Persistent elevation was not linked to persistence of symptoms or evidence of persistent disease in patients who underwent a second laparoscopy. The value of preoperative CA -125 in premenopausal women is limited because of its nonspecific nature (10). We did not attempt toestablish a breakpoint above which surgery would not be performed. The patient in the endoscopy group who required a laparotomy during the follow-up period exhibited persistently high levels of CA -125, which ultimately returned to normal after resection of colonic endometriosis (Fig. 2). Fayez and Vogel (3) have reported endoscopic excision of endometriomas from 50 patients, none of which showed glands and stroma on histopathology. Incision and drainage was suggested as adequate surgical therapy. Glands and stroma were identified in 86% of our cases. This would seem to support excision or destruction of cyst walls. Hasson (11) reported recurrence of eight of nine endometriomas managed by drainage alone. Vercellini et al. (12) reported recurrent endometriomas in 100% of patients managed by aspiration followed by 4 months of GnRH analog therapy. Aspiration or incision and drainage also fails to establish a tissue diagnosis. As treatment modalities are evolving, the controversy of whether endometrial glands and stroma are required to make the diagnosis of endometriosis is rising again. Pathologists have long theorized that pressure from the expanding endometrioma mass may obscure recognition of glands and stroma within the cyst wall. Current opinion varies widely on this point. In our series, lesions that lacked glands and stroma were "chocolate" cysts lined by hemosiderin-laden macrophages, persistent through two menstrual cycles, and associated with biopsy-proven endometriosis elsewhere in the pelvis. Postoperative recovery time is clearly decreased in endoscopic surgery compared with laparotomy. On the other hand, endoscopic surgeries for advanced endometriosis have significantly longer recovery times than diagnostic laparoscopies. An awareness of this difference should aid clinicians in advising patients what to expect during their recovery. In retrospective studies of this type there may be a bias toward more extensive disease in the laparotomy group. The authors acknowledge this situation and it may relate in part to our conclusion that all cases are not appropriate for endoscopic surgery. 800 500 CA 125 ulml 400 700 * LAPAROSCOPY 600 300 200 100 0 Feb. April June Aug. MONTHS Figure 2 Serial CA 125levels in a patient requiring segmental colon resection. 694 Bateman et al. Endoscopy for endometriomas Fertility and Sterility

However, our data support the findings of a recent larger study (6) that many endometriomas can be treated safely and appropriately at laparoscopy. REFERENCES 1. The American Fertility Society. Revised American Fertility Society classification for endometriosis. Fertil Steril 1985;43:351-2. 2. Mettler L, Semm K. Three step medical and surgical treatment of endometriosis. Ir J Med Sci 1983;152:2-4. 3. Fayez J, Vogel MF. Comparison of different treatment methods of endometriomas by laparoscopy. Obstet Gynecol 1991;78:660-5. 4. Daniell JF, Kurtz BR, Gurley LD. Laser laparoscopic management of large endometriomas. Fertil Steril 1991;55: 692-5. 5. Rock JA, Guzick DS, Sengos C, Schweditsch M, Sapp KC, Jones HW Jr. The conservative surgical treatment of endometriosis: evaluation of pregnancy success with respect to the extent of disease as categorized using contemporary classification systems. Fertil Steril1981;35:131-7. 6. Adamson GD, Subak LL, Pasta DJ, Hurd SJ, vonfraque 0, Rodriguez BD. Comparison of C0 2 laser laparoscopy with laparotomy for treatment of endometriomata. Fertil Steril 1992;57:965-73. 7. Brumsted JR, Deaton J, Lavigne E, Riddick DH. Postoperative adhesion formation after ovarian wedge resection with and without ovarian reconstruction in the rabbit. Fertil Steril 1990;53:723-6. 8. Webb MJ, Decker DG, Mussey E, Williams TJ. Factors influencing survival in stage I ovarian cancer. Am J Obstet Gynecol 1973;116:222-8. 9. Dembo AJ, Davy M, Stenwig AE, Berle EJ, Bush RS, Kjorstad K. Prognostic factors in patients with stage I epithelial ovarian cancer. Obstet Gynecol1990;75:263-72. 10. Vasilev SA, Schlaerth JB, Campeau J, Morrow CP. Serum Ca 125 levels in preoperative evaluation of pelvic masses. Obstet Gynecol1988;71:751-6. 11. Hasson HM. Laparoscopic management of ovarian cysts. J Reprod Med 1990;35:863-7. 12. Vercellini P, Vendola N, Boccio lone L, Colombo A, Rognoni MT, Bolis G. Laparoscopic aspiration of ovarian endometriomas. Effect with postoperative gonadotropin releasing hormone agonist treatment. J Reprod Med 1992;37:577-82. Bateman et al. Endoscopy for endometriomas 695