Laparoscopic treatment of complete obliteration of the cul-de-sac associated with endometriosis: long-term follow-up of en bloc resection

Transcription

1 FERTILITY AND STERILITY VOL. 76, NO. 2, AUGUST 2001 Copyright 2001 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Laparoscopic treatment of complete obliteration of the cul-de-sac associated with endometriosis: long-term follow-up of en bloc resection David B. Redwine, M.D., a and Jeremy T. Wright, F.R.C.O.G. b St. Charles Medical Center, Bend, Oregon; and Ashford and St. Peter s NHS Trust Hospital, Surrey, United Kingdom Received October 19, 2000; revised and accepted February 16, Supported in part by the Royal College of Obstetricians and Gynaecologists Bernhard Baron Traveling Fellowship, 1999 (Jeremy T. Wright, F.R.C.O.G.) Reprint requests: David B. Redwine, M.D., 2190 N.E. Professional Court, Bend, Oregon (FAX: ; davidbyron@bendcable.com). a Endometriosis Treatment Program, St. Charles Medical Center. b Department of Endoscopic Surgery, Ashford and St. Peter s NHS Trust Hospital /01/$20.00 PII S (01) Objective: To evaluate symptom relief following a laparoscopic technique designed for treatment of complete obliteration of the cul-de-sac associated with endometriosis, with fertility preserved. Design: Preoperative and postoperative questionnaire study of a cohort of patients with complete obliteration of the cul-de-sac undergoing a standardized laparoscopic surgical treatment. Setting: American tertiary referral center for the surgical treatment of endometriosis. Patients: Eighty-four consecutive patients undergoing laparoscopic treatment of endometriosis with complete cul-de-sac obliteration with 67 replying to a postoperative questionnaire. Interventions: Laparoscopic excision of all endometriosis including treatment of complete obliteration of the cul-de-sac by en bloc resection and bowel resection as needed. Main Outcome Measures: Symptom relief as measured on a 5-point ranked ordinal scale administered before and after surgery, as well as perioperative complications, postoperative fertility, and prognostic value of preoperative findings on pelvic examination. Results: Symptom reduction was obtained for all symptoms related to cul-de-sac disease, particularly for patients with severe or debilitating symptoms preoperatively. There was no significant complication, and the postoperative fertility rate was 43%. Seventy-three percent of patients with obliteration of the cul-de-sac had histologically proved rectal endometriosis. Nodularity and tenderness on examination were predictive of symptom improvement. Conclusions: Aggressive laparoscopic excision of endometriosis carried out in a specialist center offers good symptom relief, especially for those with severe or debilitating symptoms. To ensure complete removal of all disease, intestinal surgery is required in most patients with complete obliteration of the cul-de-sac. (Fertil Steril 2001;76: by American Society for Reproductive Medicine.) Key Words: Laparoscopy, endometriosis, symptom relief, rectovaginal endometriosis, cul-de-sac obliteration, surgical treatment, intestine, rectum, transvaginal Complete obliteration of the posterior pelvic cul-de-sac in women with endometriosis was first described by Cullen in 1914 (1) and is identified when the rectosigmoid colon adheres across the posterior cervix, usually with fusion of the rectal wall to the uterosacral ligaments as well. This presentation of endometriosis is considered by most experts to be the most severe form of the disease and the most difficult to treat surgically because of invasive endometriosis hidden under the obliteration. Indeed, this single manifestation of endometriosis is awarded 40 points by the rafs classification system of endometriosis (2), more than any other single factor. Cystic ovarian endometriosis often accompanies complete obliteration of the cul-de-sac (3). Symptoms include dysmenorrhea, dyspareunia, painful bowel movements, painful flatus, and pain when sitting. The occurrence of fibromuscular metaplasia around glandular elements of endometriosis distinguishes this form of the disease from peritoneal endometriosis (4, 5), and the term adenomyoma was coined to reflect the histopathology of this lesion (1). Hormonal treatment alone has not been reported. Because surgery for this condition can 358

2 be difficult, the gynecological surgical management often consists of hysterectomy and bilateral salpingo-oophorectomy, leaving the fibrotic endometriotic areas in situ (6 13). This can result in persistent symptoms in some patients (14), possibly due to local production of aromatase enzyme that converts adrenal precursors into estrogen (15). Since obliteration of the cul-de-sac is uncommon in a general gynecological practice, observational reports on its treatment are usually included in series along with other patients having less severe involvement (16 20). This article details the surgical technique and symptomatic follow-up of women who underwent laparoscopic treatment of this condition. METHOD All surgery was done for a chief complaint of pelvic pain by one surgeon (D.B.R.), with general surgical assistance for segmental bowel resections. Surgery was performed at St. Charles Medical Center in Bend, Oregon. The surgical goals and activities of the Endometriosis Treatment Program have been reviewed and approved by the Executive Committee of the hospital acting as an Institutional Review Board. The surgical technique used for treatment of complete obliteration of the cul-de-sac and the rationale for its use has been described in detail (21 28). This surgical technique was designed with evidence developed from pelvic mapping of anatomic sites of involvement by endometriosis. In patients without obliteration of the cul-de-sac, such mapping (29) had indicated that the cul-de-sac and uterosacral ligaments were among the most common pelvic sites of involvement by endometriosis, while the lower rectosigmoid adjacent to the cul-de-sac peritoneal reflection was among the most common intestinal areas of involvement. This suggested that any surgery for obliteration of the cul-de-sac must take these sites into account, even though the cul-de-sac, uterosacral ligaments, and bowel wall could not necessarily be seen in their entirety in such patients. Preoperative bowel preparation is given to patients with the following conditions: a previous surgical diagnosis of obliteration of the cul-de-sac or ovarian endometriomas, nodularity on pelvic examination, rectal pain with each defecation, or suspicion of ovarian endometrioma on examination. When vaginal involvement is present due to invasion from a rectal nodule or uterosacral ligament nodule, electrosurgery is used to incise the normal vaginal epithelium around the disease, followed by sharp and blunt dissection with scissors or finger into the normal rectovaginal septal tissue lateral and caudad to the lesion. An intrauterine manipulator is placed, the patient is placed in steep Trendelenburg position, and triple-puncture laparoscopy performed. With the uterus in extreme anteversion, the morphology of the rectum is noted. When the rectal wall is rounded (27) at its point of adherence to the cervix, the rectal wall is involved by invasive endometriosis and surgery on the bowel will eventually be necessary. If the surface of the rectum is noted to be flat at its point of adherence to the posterior cervix, rectal wall involvement by endometriosis is superficial or absent. Using 3 mm monopolar scissors passed down the operating channel of a 10 mm operating laparoscope, incisions are made in normal peritoneum lateral and parallel to the involved uterosacral ligaments, using 90 watts of pure cutting current. The uterosacral ligaments are then bluntly undermined. A transverse incision is created across the posterior cervix above the point of adherence of the bowel using either the cutting current or 50 watts of coagulation current. Intrafascial dissection down the posterior cervix toward the rectovaginal septum allows transection of the uterosacral ligaments. Thus, all nodular fibrotic disease to begin to fall away from the posterior cervix still attached to the bowel wall. If the vaginal wall is involved by endometriosis, the rectovaginal septum is initially approached laparoscopically from the right or left side, thus avoiding the mass present in midline. Once the lateral aspects of the rectovaginal septum have been revealed, the intrafascial dissection can be extended down the posterior cervix until the vaginal vault is entered adjacent to the cervix. In patients without vaginal involvement, the transverse incision down the posterior cervix can be continued in midline until the rectovaginal septum is encountered. The normal rectovaginal septum is developed distally, resulting in normal bowel wall being present on all sides of the nodular mass. The cul-de-sac remains obliterated, but is now resting on the anterior wall of the mobilized rectum, lying in the midst of a fibrotic mass of tissue consisting of fused uterosacral ligaments, cul-de-sac, posterior cervix and bowel wall. Removal of this mass from the bowel frequently involves partial or full thickness resection of the anterior rectal wall, or segmental resection and anastomosis in some patients (30, 31). Although the rectal mucosa is rarely involved by endometriosis (32), submucosal fibrosis may lead the dissection through the mucosa into the bowel lumen. The bowel wall is repaired appropriately with interrupted 3-0 silk sutures for partial thickness resection or a double layer closure for full thickness resections, using 3 0 chromic or 3-0 vicryl for the mucosa and 2-0 silk for the seromuscularis. Other areas of pelvic, ovarian, or intestinal endometriosis are excised. Seventeen patients had endometriomas of the right ovary only, 7 of the left ovary only, and 19 had endometriomas of both ovaries. In all cases the diagnosis of endometriosis was confirmed on histological analysis. Presacral neurectomy was performed in eight patients who responded to the postoperative questionnaire mailing. Sixteen patients with complete cul-de-sac obliteration FERTILITY & STERILITY 359

3 requiring conversion to laparotomy during the period of study were not included in this report. Laparoscopy was abandoned in these patients due to multiple sites of bowel involvement that could be more easily treated by laparotomy, or because of huge rectal nodules invading toward the sacrum. Neither preoperative nor postoperative medical therapy specific for endometriosis was prescribed by the operating surgeon. Preoperative barium studies, magnetic resonance, computerized tomographic or endoscopic evaluation of the bowel were not routinely performed and were usually negative when obtained by previous physicians. Such studies are rarely helpful in diagnosis of intestinal endometriosis (32, 33), except in the most severe cases. Intestinal endometriosis remains largely a surgical diagnosis, and the surgeon will still be obligated to treat intestinal involvement when it is found despite preoperative test results. Most patients undergoing bowel surgery in this series had preoperative bowel preparation. Between March 1987 and March 1999, 1,764 patients had undergone surgical treatment of endometriosis by the senior author. Of these, 1,415 had conservative surgical treatment of endometriosis. A programmable computer database (Fox- Pro for the Macintosh; Microsoft Corporation, Redmond, WA) has been used to record and examine information on patients with endometriosis (34 36). Additional information fields have been added to the database as needed. Beginning in December 1989, patients were asked to complete a simple office-based questionnaire before surgery, which included a 5-point ranked ordinal scale asking them to rate the severity of various symptoms thought to be associated with endometriosis (1 none, 2 mild, 3 moderate, 4 severe, 5 debilitating) (Table 1). This information was tabulated in the database prospectively. Four patients had undergone surgery before that time and did not have preoperative pain scales recorded. Beginning in December 1993, fields recording tenderness and nodularity on examination were added. During preoperative pelvic examination, 3-point scales (1 none, 2 moderate, 3 severe) were used to assess tenderness of the cul-de-sac and each uterosacral ligament separately, and a 3-point pain scale (1 none, 2 some, 3 most) was used to attempt to estimate whether any part of the pelvic examination reproduced at least some of a patient s pain. Nodularity in each of these areas was estimated in centimeters. Twenty-nine patients had undergone surgery before the addition of these fields and did not have tabulation of tenderness and nodularity. In April 1999, this database was queried to identify women who had undergone conservative laparoscopic excision of rectovaginal endometriosis associated with complete obliteration of the posterior pelvic cul-de-sac, with retention of the uterus and at least one tube and ovary. The 84 patients so identified were mailed a second questionnaire in which TABLE 1 Preoperative characteristics of questionnaire responders and nonresponders. Variable Statistic Responders (n 67) Nonresponders (n 17) Age (y) Mean % CI Range Previous pregnancy Number, % 22 (33%) 3 (18%) Previous delivery Number, % 8 (12%) 2 (12%) rafs points Median % CI Range Number of pelvic areas Mean involved 95% CI Range Number of intestinal areas Mean involved 95% CI Range Length of surgery (min) Median % CI Range Type of rectal resection required Segmental #, % 6 (9%) 1 (6%) Full thickness #, % 21 (31%) 5 (29%) Deep partial thickness #, % 8 (12%) 3 (18%) Superficial thickness #, % 15 (22%) 2 (12%) Length of follow-up (y) a Mean SD Range Preoperative symptom Median score (number of valid responses) Nonmenstrual pelvic pain 3 (58) 3.5 (16) Menstrual pain other than 3 (59) 4 (15) cramps b Uterine cramps with menses 4 (63) 4.5 (16) Dyspareunia 2 (35) 3 (6) Painful bowel movements 3 (63) 4 (15) Constipation 2 (62) 2 (16) Diarrhea 1 (62) 1.5 (16) Intestinal cramping 3 (61) 2.5 (14) Pelvic pain with exercise 2 (20) 1 (5) Low-back pain 3 (60) 3 (16) Tenderness on exam 3 (56) 3 (16) Fatigue 3 (62) 3 (16) Note: CI confidence interval; SD standard deviation. a For nonresponders, the length of follow-up was calculated up to the date of March 1, b Pain during menses that does not feel like uterine cramping. Redwine. Complete cul-de-sac obliteration. Fertil Steril they were asked to complete symptom scales that were identical to the scales completed preoperatively. Details were also requested of their postoperative fertility history or any further surgery together with questions about symptoms related to altered bowel, urinary, or sexual function. Data from the returned questionnaires was entered into the FoxPro database and then transferred to STATA, a statistical pack- 360 Redwine and Wright Complete cul-de-sac obliteration Vol. 76, No. 2, August 2001

4 age, for analysis (StataCorp Stata Statistical Software; Release 6.0. College Station, TX: Stata Corporation). If a questionnaire was not received from a patient by 30 days after the initial mailing, an attempt was made to reach the patient through contact people listed in the patient s office chart. Statistical Testing To ensure exact pairing of preoperative and postoperative symptom scales, patients not having completed both pain scales for a symptom were excluded from analysis of that symptom. Additionally, patients who stated they were not sexually active or engaged in exercise were excluded from analysis of dyspareunia or pelvic pain with exercise. Patients who had undergone a hysterectomy during the follow-up period were coded as having postoperative uterine cramps to a level of 5 (debilitating) regardless of the level they had indicated. This was to assure that the response of uterine cramping to conservative excision of endometriosis would not be factitiously improved by eliminating those who lacked a postoperative uterine cramping pain score due to intervening hysterectomy. Since preoperative symptoms were skewed toward the higher range of the 5-point symptom scales and postoperative symptoms were skewed toward the lower range, medians were used to characterize preoperative and postoperative symptom levels. The degree of symptom change following surgery was also studied using the Wilcoxon sign rank test for all questionnaire respondents and for the subgroup of respondents with severe or debilitating preoperative symptoms. Median differences and associated confidence intervals were calculated (37) for the paired preoperative and postoperative pain scales using a program written in FoxPro source code. The rafs point totals and length of surgery were shown by a scatter plot diagram to be skewed toward the lower range, so medians were used to characterize these parameters. Scatter plot diagrams of patient age at surgery, number of pelvic areas involved by endometriosis and number of intestinal areas involved by endometriosis all showed a nonskewed distribution, so means were used to characterize these parameters. Time intervals were converted to years by dividing the number of days between two interval events (i.e., birth date, surgery date, date of last follow-up) by Correlation coefficients were calculated for comparisons of preoperative tenderness and nodularity on pelvic examination with postoperative symptom improvement in order to investigate the value of these findings on examination in predicting postoperative symptom improvement. RESULTS Between March 1, 1987 and March 16, 1999, 84 women with complete obliteration of the cul-de-sac (mean age years; range, ; SD 5.17) underwent laparoscopic excision of endometriosis with retention of the uterus and at least one ovary. Eighty-one percent of these had a history of previous medical or surgical therapy. During the same time interval 1,049 women without complete obliteration of the cul-de-sac presented for conservative laparoscopic excision of endometriosis (mean age years; range ; SD 6.49). Sixty-seven of the 84 patients (80%) completed and returned the postoperative questionnaire. Characteristics of questionnaire respondents vs. nonrespondents are shown in Table 1. Nonrespondents tended to be younger and to have slightly more severe presenting symptoms than respondents. Comparison of pre- and postoperative symptom scales for patients with completed paired responses is shown in Table 2. Among all questionnaire responders, some would have had no or only mild symptoms before surgery, so it is of particular interest to study the symptom relief of women rating a preoperative symptom as severe or debilitating (Table 3). Sixty-one of 84 (73%) patients with complete obliteration of the cul-de-sac required some degree of rectal surgery for removal of all disease (Table 1). Postoperatively 28 patients tried to conceive and 12 were successful, five of these requiring ART. There have been five term pregnancies, four spontaneous abortions, one ectopic gestation and two others were still pregnant at the time they completed the questionnaire. Tenderness on preoperative examination that reproduced any of a patient s pain symptoms was found to predict reduction of symptoms following surgery. The correlation coefficient (r) of such tenderness associated with eventual symptom relief ranged from mild (0.19) to strong (0.60) for the following symptoms, ordinally: uterine cramping with menses, pelvic pain away from the menstrual flow, tenderness on pelvic examination, menstrual pain which was not described as uterine cramping, and dyspareunia. Forty of 84 women (48%) had nodularity found on preoperative examination. Nodules ranged in size from an estimated 2 mm up to 2 cm on the uterosacral ligaments and from 4 mm to 3 cm in the cul-de-sac. Fifteen patients had cul-de-sac nodularity without uterosacral ligament nodularity, while 3 had left uterosacral ligament nodularity only and 6 had right uterosacral ligament nodularity only. Cul-de-sac nodularity detected on preoperative pelvic examination was also predictive of symptom relief following surgery. The correlation coefficient (r) of cul-de-sac nodularity associated with eventual symptom relief ranged from 0.20 to 0.71 for the following symptoms, ordinally: painful bowel movements, fatigue, pelvic pain with exercise, uterine cramps with menses, low back pain, intestinal cramping, and diarrhea. There was no unintended injury to bowel or urinary tract. Neither blood transfusion nor reoperation for a surgical complication was necessary in any patient. Postoperative fevers in two patients responded quickly to antibiotics. Two FERTILITY & STERILITY 361

5 TABLE 2 Symptom response as measured by preoperative and postoperative questionnaires for all responders (n 67). Symptom (number of valid responses a ) Median scores Preop Postop Median of Improved differences b (95% CI) b pho T c number (%) Worse number (%) Unchanged number (%) Nonmenstrual pelvic pain (58) ( ) (78%) 2 (3%) 11 (19%) Menstrual pain other than cramps d (59) ( ) (68%) 12 (20%) 7 (12%) Uterine cramps with menses e (63) ( ) (65%) 11 (17%) 11 (17%) Dyspareunia f (35) ( ) (66%) 2 (6%) 10 (29%) Painful bowel movements (63) ( ) (59%) 6 (10%) 20 (32%) Constipation (62) ( ) (53%) 3 (5%) 26 (42%) Diarrhea (62) ( ) (29%) 9 (15%) 35 (56%) Intestinal cramping (61) ( ) (44%) 7 (11%) 27 (44%) Pelvic pain with exercise g (20) ( ) (45%) 1 (5%) 10 (50%) Low-back pain (60) ( ) (65%) 2 (3%) 19 (32%) Tenderness on exam (56) ( ) (59%) 9 (16%) 14 (25%) Fatigue (62) ( ) (44%) 12 (19%) 23 (37%) Note: CI confidence interval. a Patients having both a preop and a postop score for a symptom. b Calculated from all combinations of differences of valid preop and postop responses (responses [responses 1]/2) from Reference 37. c Probability that sets of preop and postop pain scores are not different for a symptom (Wilcoxon sign-rank test). d Pain during menses that does not feel like uterine cramping. e Nine patients with subsequent hysterectomy were assigned a value of 5 for the postop symptom level. f Twenty patients indicated they were not sexually active and were excluded for this reason. g Thirty-six patients indicated they did not exercise and were excluded for this reason. Redwine. Complete cul-de-sac obliteration. Fertil Steril patients with full thickness bowel resection did not have preoperative bowel prep and had no complication. No patient reported either symptoms or signs of uterovaginal prolapse following surgery. DISCUSSION These data suggest that laparoscopic excision of rectovaginal endometriosis associated with complete obliteration of the cul-de-sac is a safe and effective treatment, particularly for those patients with severe or debilitating symptoms. Following the aggressive surgery described here, patients did not report a sense of postoperative uterine prolapse, nor a medical diagnosis of uterine prolapse. This type of aggressive excisional surgery is indicated in any patient who wishes to retain her fertility and who has pain that has not responded well to previous medical or surgical therapy. Preoperative nodularity and tenderness, which reproduce the patient s pain, both predict a greater chance of symptomatic improvement following surgery. Patients with nodularity on examination had a higher initial symptom level and a greater postoperative improvement in every symptom except fatigue, low backache, pelvic pain with exercise, diarrhea, and uterine cramping with menses (data not shown). Findings on examination may vary with the ovarian hormonal cycle, and examination during menses has been advocated as a means of more accurately detecting signs of endometriosis (38). Patients with higher preoperative symptom scores had a greater improvement of postoperative symptom scores (Table 3), which is to be expected since there is greater range for improvement at the high end of the symptom scale. Several intestinal symptoms that might be considered to comprise irritable bowel syndrome were studied (intestinal cramping, diarrhea, constipation, painful bowel movements). These symptoms also improved postoperatively. Since no patient received further therapy directed toward intestinal symptoms by a gastroenterologist or internist, the improvement in these symptoms seems to be a result of surgery alone. The results of this study don t speak directly to an infertile population. However, the postoperative fertility rate is acceptable, particularly given the rather advanced age of patients undergoing surgery, indicating that such aggressive surgery does not result in sterilization due to postoperative adhesion formation. Seventy-three percent of the patients with obliteration of the cul-de-sac in this study had histologically proved involvement of the rectum, which had required some degree of rectal resection (Table 1), and 82% had involvement of one or more intestinal sites. This incidence of intestinal involvement is virtually identical to an incidence of 78% among patients with rectovaginal endometriosis previously reported (39). The surgical finding of obliteration of the cul-de-sac, therefore, must be taken as evidence of likely rectal involve- 362 Redwine and Wright Complete cul-de-sac obliteration Vol. 76, No. 2, August 2001

6 TABLE 3 Symptom response as measured by preoperative and postoperative questionnaires for patients with preoperative levels of 4or5. Symptom (number of valid responses a ) Median scores Preop Postop Median of differences b (95% CI) b pho T c Improved number (%) Worse number (%) Unchanged number (%) Nonmenstrual pelvic pain (23) ( ) (91%) 0 (0%) 2 (9%) Menstrual pain other than cramps (29) ( ) (93%) 1 (4%) 2 (7%) Uterine cramps with menses d (41) ( ) (73%) 5 (12%) 6 (15%) Dyspareunia e (11) ( ) (82%) 1 (9%) 1 (9%) Painful bowel movements (18) ( ) (78%) 0 (0%) 4 (22%) Constipation (13) ( ) (85%) 2 (15%) 0 (0%) Diarrhea (8) ( ) (100%) 0 (0%) 0 (0%) Intestinal cramping (16) ( ) (69%) 1 (6%) 4 (25%) Pelvic pain with exercise f (4) 5 1 f f f 4 (100%) 0 (0%) 0 (0%) Low-back pain (22) ( ) (82%) 0 (0%) 4 (18%) Tenderness on exam (16) ( ) (81%) 2 (13%) 1 (6%) Fatigue (18) ( ) (67%) 1 (6%) 5 (28%) Note: CI confidence interval. a Patients having both a preop and a postop score for a symptom. b Calculated from all combinations of differences of valid preop and postop responses (responses [responses 1]/2) from Reference 37. c Probability that sets of preop and postop pain scores are not different for a symptom (Wilcoxon sign-rank test). d Six patients with subsequent hysterectomy were assigned a value of 5 for the postop symptom level. e Eight patients indicated they were not sexually active and were excluded for this reason. f Twenty-five patients indicated they did not exercise and were excluded for this reason, leaving too few responses for statistical analysis. Redwine. Complete cul-de-sac obliteration. Fertil Steril ment, and surgeons wishing to deal with this condition must be prepared to perform intestinal surgery. Reports of laparoscopic treatment of invasive endometriosis have appeared in the literature, although none has dealt exclusively with complete obliteration of the cul-de-sac. Laparoscopic laser excision of infiltrating cul-de-sac endometriosis was described in 1988 (16) in five patients, none of whom had obliteration of the cul-de-sac. Pain relief was not discussed. Laparoscopic treatment of 23 cases of complete obliteration of the cul-de-sac was reported to result in pain relief in 89% of patients (17). The surgical technique described was based largely on excision by laser or scissors. Surgery was begun by dissecting the rectum away from the cervix until the normal rectovaginal septum was encountered. Nodular endometriosis of the uterosacral ligaments and rectovaginal septum was then excised. Although excision of endometriosis of the anterior rectal wall was performed, it is unclear how many patients had such disease. The methods of symptom evaluation and patient follow-up were not discussed. A large series (19) of laparoscopically treated rectovaginal endometriosis advocates a technique which avoids consideration of rectal wall endometriosis as a pathologic entity and focuses on carbon dioxide laser excision of invasive disease of the cul-de-sac, uterosacral ligaments and vagina. Thickening of the bowel wall, which likely represents endometriosis, is termed perivisceritis. Despite the possibility of leaving rectal wall disease in place, the results of surgery were good in 242 patients followed for more than 2 years, with success rates of 96.3% and 98.2% in eliminating severe pelvic pain and dyspareunia, respectively. The methods of evaluation of symptoms and patient follow-up in this study were not discussed. Laparoscopic excision of endometriosis in 57 patients was found to produce significant pain relief and improvement of quality of life at 4 months after surgery (20). Prospective pre- and postoperative symptom scales similar to those in the current study were used, as well as other standardized quality of life measurements. The surgical technique was similar to that described here. While 36 of these patients were classified as Stage III or Stage IV (2), it is unclear how many had complete obliteration of the cul-desac. Thirty-six patients (56%) had disease of the anterior bowel wall, which was treated at the time of surgery. Not all patients responded to the postoperative questionnaire. Nonresponders were younger and seemed to have slightly more severe preoperative symptoms than responders (Table 1). Since patients with more severe symptoms were found to have more obvious improvement than patients with a lesser degree of symptoms (Table 2, Table 3) the loss of information from nonresponders may have caused underestimation of the benefits of surgery in relieving symptoms. Since fertility declines with advancing age, the conception rate may be an underestimation since the fertility efforts of FERTILITY & STERILITY 363

7 this younger and possibly more fertile group of nonresponders is not included. Not all respondents answered each preoperative or postoperative question. A structured interview by telephone or in person might be more accurate and result in more complete responses. The symptom scales used are necessarily subjective and open to varying interpretations by respondents. A more detailed numeric analysis of symptoms might be possible, but might be more prone to errors or incompletion. We are unaware of any publication showing an effect of time between surgery to completion of a postoperative questionnaire on the accuracy of the questionnaire response. However, with the passage of time patients with good symptom relief tend to forget how much they previously hurt (40). This would not seem to affect our results. While many types of symptom measuring instruments have been promoted, as the complexity of the instrument increases, the response rate and accuracy suffer (41). The 5-point scale used to measure symptoms in this study was used because of its simplicity. Although eight patients receiving presacral neurectomy had a slightly greater reduction in their level of uterine cramping than patients without presacral neurectomy, the small numbers would not greatly exaggerate the benefit of conservative excision of endometriosis alone on this symptom. Thirteen respondents had subsequent surgery for pain and 3 of these also had hormonal treatment for pain. Ten respondents had only subsequent hormonal treatment for pain. For 10 patients having subsequent hysterectomy, the follow-up questionnaire analysis of uterine cramping with menses was adjusted by upcoding the postoperative level of this symptom to the maximum of five. It is possible that hysterectomy may have also improved some of the other symptoms, although there seems to be no way to verify that possibility. The surgery described here is time consuming and requires considerable training and experience, surpassing that of oncological surgery. These patients should be referred to clinicians with the appropriate training and experience to undertake the procedures safely. If complete laparoscopic excision of all pelvic and rectovaginal endometriosis is impossible, excellent results can be achieved by laparotomy (31). CONCLUSION Complete obliteration of the cul-de-sac is a manifestation of endometriosis that causes significant morbidity. Aggressive excision results in long-term pain relief in most patients. Most patients with obliteration of the cul-de-sac have rectal involvement. If carried out in specialist centers by surgeons with proper training and experience, serious complications are rare. Not all pelvic pain is caused by endometriosis, even in patients with obliteration of the cul-de-sac. For this reason, some patients will require subsequent therapy. References 1. Cullen TS. Adenomyoma of the rectovaginal septum. JAMA 1914;62: The American Fertility Society. Revised American Fertility Society classification of endometriosis. Fertil Steril 1985;43: Redwine DB. Ovarian endometriosis. A marker for more severe pelvic and intestinal disease. Fertil Steril 1999;73: Sampson JA. Inguinal endometriosis. Am J Obstet Gynecol 1925;10: Fallon J, Brosnan JT, Manning JJ, Moran WG, Meyers J, Fletcher ME. Endometriosis: A report of 400 cases. Rhode Island Med J 1950;33: Cullen TS. Adenomyoma of the rectovaginal septum. JAMA 1916;67: Meigs JV. Endometrial hematomas of the ovary. Boston Med J 1922; 187: Keene FE, Kimbrough RA. Endometriosis. A review based on the study of one hundred and eighteen cases. JAMA 1930;95: Cattell RB, Swinton NW. Endometriosis. With particular reference to conservative treatment. NEJM 1936;214: Mayo CW, Miller JM. Endometriosis of the sigmoid, rectosigmoid and rectum. Surg Gynecol Obstet 1940;70: Holmes WR. Endometriosis. Am J Obstet Gynecol 1942;43: Counseller VS. Surgical procedures involved in the treatment of endometriosis. Surg Gynecol Obstet 1949;89: Huffman JW. External endometriosis. Am J Obstet Gynecol 1951;62: Redwine DB. Endometriosis persisting after castration: clinical characteristics and results of surgical management. Obstet Gynecol 1994; 83: Takayama K, Zeitoun K, Gunby RT, Sasano H, Carr BR, Bulun SE. Treatment of severe postmenopausal endometriosis with an aromatase inhibitor. Fertil Steril 1998;69: Martin DC. Laparoscopic and vaginal colpotomy for the excision of infiltrating cul-de-sac endometriosis. J Reprod Med 1988;33: Reich H, McGlynn F, Salvat J. Laparoscopic treatment of cul-de-sac obliteration secondary to retrocervical deep fibrotic endometriosis. J Reprod Med 1991;36: Donnez J, Nisolle M, Casanas-Roux F, Bassil S, Anaf V. Rectovaginal septum endometriosis or adenomyosis: Laparoscopic management in a series of 231 patients. Hum Reprod 1995;10: Donnez J, Nisolle M, Gillerot S, Smets M, Bassil S, Casanas-Roux F. Rectovaginal septum adenomyotic nodules: a series of 500 cases. Br J Obstet Gynecol 1997;104: Garry R, Clayton R, Hawe J. The effect of endometriosis: radical laparoscopic excision on quality of life indicators. Br J Obstet Gynae 2000;107: Redwine DB. Laparoscopic en bloc resection for treatment of the obliterated cul-de-sac in endometriosis. J Reprod Med 1992;37: Redwine DB. Bowel resection related to endometriosis. In: Tulandi T, ed. Atlas of laparoscopic technique for gynecologists, London: WB Saunders, 1994: Redwine DB. Treatment of endometriosis of the cul-de-sac. In: Nezhat CR, Berger GS, Nezhat FR, Buttram VC, Nezhat CH, eds. Modern surgical management of endometriosis, New York: Springer-Verlag, 1995: Redwine DB. Treatment of bowel lesions. In: Adamson GD, Martin DC, eds. Endoscopic management of gynecologic disease, Philadelphia: Lippincott-Raven Press, 1996: Redwine DB. Endometriosis. In: Levine RL, Sanfilippo J, eds. Operative gynecologic endoscopy, 2nd ed., New York: Springer-Verlag, 1996: Redwine DB. Monopolar electroexcision of endometriosis. In: Sutton CA, Diamond MP, eds. Endoscopic surgery for gynaecologists, 2nd ed., London: WB Saunders, 1998: Redwine DB. Bowel resection related to endometriosis. In: Tulandi T, ed. Atlas of laparoscopic technique for gynecologists, London: WB Saunders, 2nd ed., 1998: Redwine DB. Deep endometriosis. In: Venturini PL, Evers JLH, eds. Endometriosis: Basic research and clinical practice. Carnforth: Parthenon, 1997: Redwine DB. The distribution of endometriosis in the pelvis by age groups and fertility. Fertil Steril 1987;47: Redwine DB, Sharpe DR. Laparoscopic segmental resection of the sigmoid colon. J Laparoendosc Surg 1991;1: Redwine DB, Koning M, Sharpe DR. Laparoscopically assisted transvaginal segmental bowel resection for endometriosis. Fertil Steril 1996; 65: Gray LA. Endometriosis of the bowel: role of bowel resection, superficial excision and oophorectomy in treatment. Ann Surg 1973;177: Coronado C, Franklin RR, Lotze EC, Bailey HR, Valdes CT. Surgical 364 Redwine and Wright Complete cul-de-sac obliteration Vol. 76, No. 2, August 2001

8 treatment of symptomatic colorectal endometriosis. Fertil Steril 1990; 53: Redwine DB. Database management for personal computers, I. J Am Assoc Gynecol Laparosc 1996;3: Redwine DB. Database management for personal computers, II. J Am Assoc Gynecol Laparosc 1996;3: Redwine DB. Database management for personal computers, III. J Am Assoc Gynecol Laparosc 1996;4: Gardner MJ, Altman DG. Statistics with confidence. Belfast: The Universities Press, 1990, pp Fallon J, Brosnan JT, Moran WG. Endometriosis. Two hundred cases considered from the viewpoint of the practitioner. N Eng J Med 1946; 235: Henriksen E. Endometriosis. Am J Surg 1955;90: Redwine DB. Remote postoperative recollection of preoperative pain in patients undergoing excision of endometriosis. J Am Assoc Gynecol Laparosc 1994;1: Redwine DB. Treatment of endometriosis-associated pain. In: Olive DL, ed., Endometriosis: Infertility and Reproductive Medicine Clinics of North America. Philadelphia: WB Saunders 1992: FERTILITY & STERILITY 365