Article written by Jeremy Wright, M.D., edited by David Redwine, M.D.
Photos courtesy of David Redwine, M.D. and Thomas Lyons, M.D.

Rectovaginal endometriosis is perhaps the greatest gynecological surgical challenge, exceeding the complexity of gynecologic oncological surgery. If such disease represented cancer, it would be deemed inoperable, whereas endometriosis is benign and locally invasive and the only effective treatment is excision. The gold standard effective treatment is surgical removal of the disease from the body. Many have a long history of sub-fertility [1]. For the majority of women of childbearing age surgery can be accomplished with preservation of the uterus and adnexae. Pelvic endometriosis affects 5-7% of menstruating women [1]. The incidence of bowel involvement however varies greatly because of methods of case selection, with a reported incidence of between 3% and 34%.

Rectovaginal endometriosis is frequently associated with endometriosis elsewhere in the pelvis, particularly ovarian endometriomas and ovarian fossae endometriosis, which will require concurrent removal [2, 3, 4, 5]. This chapter explores the presentation, clinical findings, pathology and treatment of this difficult manifestation of endometriotic disease.

Presentation

Women with rectovaginal endometriosis frequently present with severe backache, lower abdominal pain, dyschezia and constipation [6], all of which can be aggravated before or during menses. They may also suffer with rectal pain with flatus, bowel movements and even sitting down. The frequent use of laxatives can give rise to diarrhoea, and some patients complain of diarrhoea during menstruation. Symptomatology is usually associated with severe dyspareunia and heavy and painful periods, during which time some patients may also report a slight temperature elevation. A detailed history will often find that the symptoms date from the patient's menarche and have been ignored or treated as primary dysmenorrhoea, often with oral contraceptives. One helpful distinguishing point is that with involvement of the rectum by endometriosis, patients frequently have pain with every bowel movement during the month, whereas patients with cul-de-sac or uterosacral ligament disease without rectal disease may complain of painful bowel movements only with menses.

When taking a history from these patients, it is useful to ask them to score their pain using a ranked ordinal scale or a visual analogue scale, or other quality of life measures. Preoperative symptom levels may easily be compared to post-operative levels to obtain an objective measure of symptom relief. Reports using such scales are rarely used in the literature to measure response of symptoms to treatment of rectovaginal endometriosis [1, 7].

Systemic upset is rare and general examination is usually normal apart from mild lower abdominal tenderness and tenderness over the sigmoid colon. Rectovaginal endometriosis is often associated with considerable psychological morbidity, a high dependence on family members, difficulties with relationships and under-achievement associated with a poor sickness record in employment. Women with endometriosis are usually reluctant to undergo pelvic examination as they know this will be painful. However, examination of the pelvis is essential in aiding diagnosis.

Imaging techniques such as ultrasound, computed tomography (CT) and magnetic resonance imaging (MRI) may produce confirmatory evidence but are of themselves not diagnostic and interpretation of the findings can be difficult [8, 9, 10]. A normal scan result does not relieve the patient's pain nor eliminate the eventual need for surgical diagnosis and treatment. A positive scan result will usually show only ovarian cysts when they are present but will often fail to diagnose invasive disease of the pelvic floor and bowel wall. Thus, the scan may bias the surgeon to look for and treat only ovarian cysts and leave possibly highly symptomatic disease behind.

Pelvic examination

Speculum examination is sometimes useful and should concentrate on the area of the posterior vaginal fornix by tipping a bivalve speculum posterior to the cervix. The speculum is opened so as to expose the epithelium of the posterior fornix. In this area, rectovaginal disease may be seen invading the vaginal epithelium from an underlying nodule of the uterosacral ligament or rectal nodule. Vaginal endometriosis can be identified by disruption of the normal rugae in the overlying vagina and by epithelial piling, distortion, possibly small bluish cysts and occasionally larger polypoid reddish lesions (Fig. 1). It is always retained between the confines of the uterosacral ligaments and the space between them. This type of lesion is frequently missed on speculum examination in the office because only the cervix is visualized.

A thorough but gentle pelvic examination is necessary. Bimanual examination should assess size, position and mobility of the uterus, followed by an assessment of adnexal tenderness and the presence of any ovarian enlargement. The remainder of the examination should only be performed by monomanual vaginal digitation and careful observation of the facial expression and body language of the patient as she reacts to palpation of the cul-de-sac and uterosacral ligaments. Women with this condition will usually at least grimace, if not cry out and push up on the table away from the examining finger. The uterosacral ligaments should be gently palpated for any nodularity, which may be exquisitely tender. Palpation of the pouch of Douglas will often reveal a tender fixed nodule of endometriosis which may seem to involve the rectovaginal septum [11]. This can frequently be confirmed by rectal examination or by concomitant rectal and vaginal palpation, although this is uncomfortable and may cause the patient some distress. Not all nodularity which might be noted later at surgery will necessarily be found on pelvic examination [2, 12].

Indications for osmotic and mechanical bowel prep include symptoms or pelvic examination suggesting colon involvement by endometriosis, known bowel involvement or obliteration of the cul-de-sac or ovarian endometrioma cysts. Bowel preparation is recommended but can sometimes be problematic since the colon is frequently filled with watery fluid which can contaminate the pelvis if full thickness bowel resection is performed.

Etiology and pathogenesis

The etiology and pathogenesis of endometriosis remains controversial, but there is general agreement between the major theorists that rectovaginal endometriosis is different in morphology and histology from peritoneal disease, and most likely to arise from tracts of embryologically patterned rests of Müllerian tissue within the uterosacral ligaments, rectovaginal septum and anterior bowel wall [6, 13, 14]. There is a variation in oestrogen receptor and progesterone receptor content in this tissue compared to eutopic endometrium, suggesting different regulatory mechanisms for rectovaginal endometriosis as well as a different origin compared to peritoneal disease. The concept of embryologically patterned metaplasia is crucial to the surgical management of the condition since if all disease and all tracts which may form disease are removed, the disease can be cured surgically. If the disease were caused by retrograde menstruation, one would expect rapid regrowth following surgical excision [15, 16]. As it is, a protracted clinical response can be expected to excision of endometriotic tissue when undertaken in specialist centres. The disease process seems to begin, often in a field effect, in parenchymal structures, such as the uterosacral ligaments, posterior cervix and anterior bowel wall. While many cases may invade the posterior vaginal fornix, the fatty and areolar tissue of the recto- vaginal septum itself is not a site of origin, nor a site of frequent involvement. Thus, the term 'endometriosis of the rectovaginal septum' may be slightly in error.

Surgical diagnosis by laparoscopy

The majority of patients with the symptoms of endometriosis will be advised to undergo a diagnostic laparoscopy which needs to be adequate and thorough. The following points may help to identify what is required:

• The patient must be appropriately placed with her buttocks well over the edge of the operating table so that vaginal, uterine or rectal manipulation is not impeded.
• The patient must be put in sufficiently steep Trendelenburg to allow proper inspection of the pouch of Douglas. This should routinely include assessment of rectal tethering or mobility, which can be achieved by placing a ring forceps within the rectum and observing its mobility.
• Laparoscopic inspection should include the whole of the pelvic peritoneum including both ovarian fossae and the anterior and posterior cul-de-sacs. This requires manipulation of these structures and moving the bowel out of the posterior cul-de-sac. This cannot be carried out using a suprapubically placed Verres needle.
• Ideally, a laterally placed 5mm port should be employed to allow access for a subsequent irrigation probe. Laterally placed ports, however, run the risk of epigastric injury. As these vessels are very constant in position they can usually be avoided. The inferior epigastric vessels are always situated in the triangle between the obliterated umbilical vessels and the insertion of the round ligament and can usually be identified visually by examining them laparoscopically through the anterior parietal peritoneum.
• The laterally placed ports should be inserted when the abdomen is fully inflated and the point directed medially. This can either be under direct vision or by supporting the abdominal wall from the peritoneal side by holding the laparoscope just medial to the point of insertion of the 5mm trocar and directing the trocar so it passes directly under the laparoscope into the pneumoperitoneum. Control of the tension on the abdominal wall ensures that there is no rapid or explosive insertion of the trocar leading to possible damage to underlying viscera or great vessels.
• Following inspection of the pelvic peritoneum, a suction irrigation probe or laparoscopic forceps should be used to palpate the pouch of Douglas and uterosacral ligaments for areas of nodularity [17], while the uterus is upwardly displaced by a uterine manipulator. Nodular endometriosis is classically hard so the probe will appear to 'click' over it. Ideally, there should be concomitant excision biopsy of all these areas of endometriosis, but there are few centres where the equipment and expertise exist to undertake this procedure, and in these circumstances, a biopsy of at least one of the endometriotic lesions should be taken to confirm histologically the diagnosis, so that the patient can be advised to seek the appropriate specialist help.
• The cecum, appendix and terminal ileum should all be examined since if obliteration of the cul-de-sac is present, there is a possibility these other bowel areas could be involved.

Rectovaginal disease may not be associated with florid flame haemorrhages or active vascular change (Fig. 2). Occasionally, an apparently small vascular lesion of the uterosacral ligament may be the tip of a large rectovaginal nodule. The associated fibromuscular metaplasia and fibrosis give a yellowish or whitish appearance, occasionally with overlying haemorrhagic discolorations. Frank rectovaginal endometriosis will be diagnosed by the presence of dense adhesive fibrotic disease with obliteration of the pelvic cul-de-sac. Involvement of the rectum can be identified visually as the 'rounded rectum' indicating a round bulge in the rectal wall at the point of attachment to the posterior cervix (Fig. 3) as the uterus is held in extreme anteversion. Rectal involvement is absent or superficial if the rectum is flat at its point of adherence to the posterior cervix [18].

Sometimes, the presence of bilateral ovarian cysts meeting in the middle of the pelvis and dense peritubal adhesions with tethering of the sigmoid colon to the fundus of the uterus and the left pelvic side wall will make a complete initial visual survey impossible.

Treatment

The pathology is that of dense fibrosis or fibromuscular metaplasia with relatively small areas of endometriosis (Fig. 4), which are poorly responsive to hormones. In bowel lesions, there may be striking hypertrophy of smooth muscle reminiscent of the changes seen in adenomyosis of the uterus [19, 20]. Hormone manipulation does little to suppress the disease, despite the presence of oestrogen and progesterone receptors [21, 22]. Even if there is some suppression during hormone therapy, symptoms resume once hormone suppression is stopped. Many clinicians view obliteration of the cul-de-sac as an adhesive process only: the colon is stuck to the back of the cervix and it must be unstuck. The important consideration for surgery is that the adhesive process is the result of inflammation from underlying invasive endometriosis. This invasive endometriosis involves the uterosacral ligaments, posterior cervix, cul-de-sac, and usually the anterior wall of the rectum. Accordingly, surgeons who treat only the adhesions will fail in any attempt at treating the disease. Some surgeons may observe obliteration of the cul-de-sac, describe 'dense adhesions' and attempt to dissect the rectum from the cervix to 'restore normal anatomy'. Such treatment completely misinterprets the pathology and leaves 100% of invasive disease behind as well as a broad, raw area extending from the posterior cervix and down across the cul-de-sac and on to the anterior rectal wall. The depth of invasion associated with obliteration of the cul-de-sac can extend several centimetres beneath the visible surface, and just because this is not visible to the surgeon's eye does not make it acceptable for the surgeon not to think about it. This depth of invasion makes thermal ablation techniques, such as laser vaporization or electrocoagulation, inappropriate choices for surgical treatment.

The most efficient technique for treatment of obliteration of the cul-de-sac is en bloc resection [23] using unipolar electrodiathermy delivered by 3mm hook scissors which are utilized via an operating laparoscope. High power settings (90watts cut, 50watts coag) allow a high current density beneath the tip of the active electrode with rapid cutting of tissue by vaporization and little lateral thermal spread.

With involvement of the vaginal mucosa it is helpful to first delineate this (Fig. 5), using a vaginal approach prior to laparoscopy. Using a pencil electrode and coagulation current, the nodule is outlined by incision of the vaginal mucosa. The incision is then extended into the softer tissue of the rectovaginal septum followed by blunt finger dissection laterally and inferiorly.

The laparoscopic procedure should begin with the excision of other areas of endometriosis, such as endometriomas or other pelvic sidewall disease. This begins to isolate the obliterated cul-de-sac in the bottom of the pelvis. Following any ovarian surgery, the ovaries can be suspended from the round ligaments using an absorbable suture, such as 3-0 Vicryl, as this will both improve access to the posterior cul-de-sac and prevent the ovary becoming adherent to the denuded pelvis postoperatively. The stitch will dissolve when the pelvic peritoneum has regrown, thus preventing dense adherence of the ovary into the cul-de-sac.

Laparoscopic dissection of the cul-de-sac begins with incisions lateral and parallel to the uterosacral ligaments in relatively normal non-fibrotic peritoneum (Figs 6, 7), followed by blunt undermining of the uterosacral ligaments. A transverse incision (Fig. 8) is created across the cervix or uterine body above the line of adherence of the bowel. Intrafascial dissection with electrosurgery is then carried down the posterior cervix towards the rectovaginal septum, with the dissection proceeding more deeply if hidden pockets of chocolate-coloured fluid are found (Fig. 9), indicating more deeply invasive endometriosis. The uterosacral ligaments are transected at the cervical insertions in this process and the dissection is continued caudally down the rectovaginal septum until normal areolar tissue of the rectovaginal septum is clearly identified. No attempt has been made to dissect in the plane between the rectum and cervix. Instead, the cul-de-sac remains obliterated but comes to lie eventually on the anterior wall of the rectum as one mass containing the utero- sacral ligaments, the posterior cervix and the obliterated cul-de-sac (Figs 10, 11). The lateral fatty attachments of the rectum to the pelvic sidewall are severed (Fig. 12) with the result that all the diseased area is now isolated centrally onto the rectum. The fatty tissue overlying the bowel wall adjacent to the nodule is shaved away further isolating the nodule and exposing normal bowel wall for later suturing (Fig. 13).

If the endometriotic disease involves the vaginal mucosa, the dissection is slightly altered. Since the central portion of the vagina is involved by the nodule which also involves the posterior cervix, the intrafascial dissection down the posterior cervix is postponed, because the plane of dissection will be more fibrotic and obscured due to the vaginal lesion. Instead, after transecting one or both uterosacral ligaments, the vagina is entered on the right or left side of the apex, alongside the vaginal nodule (Fig. 14). The vaginal epithelial incisions that were previously created help guide the vaginal entry. Once the vagina has been entered from the side, it is easier to ensure that the rectovaginal nodule can be removed completely under direct vision, working immediately adjacent to the posterior lip of the cervix at first, then following the epithelial incision line with the dissection to ensure that the vaginal lesion has been incorporated on to the mass which remains attached to the bowel wall. The normal rectal wall distal to the nodule is exposed by dissection of its areolar tissue. It is important that all fatty tissue be cleaned from the wall of the bowel in areas where sutures will eventually be placed.

If the vagina is not opened, dissection is continued laparoscopically. The colonic wall has four layers: the serosa, outer longitudinal muscularis, inner circular muscularis, and mucosa. The serosa is absent distal to the peritoneal cul-de-sac reflection. These layers can be used to the surgeon's advantage. The hypertrophied muscular layers of the bowel frequently allow the surgeon to lightly peel the affected layers off the underlying mucosa by a technique called 'mucosal skinning' (Fig. 15). In many cases rectal mucosa is not penetrated during resection, in which case the seromuscular layer is closed with interrupted 2-0 silk. If dense submucosal fibrosis causes the mucosa to be penetrated during dissection, the mass of diseased tissue can be removed transanally with a ring forceps and a rectal repair undertaken; otherwise, the diseased tissue can be removed after slight enlargement of the 10 mm umbilical port. The defect to be repaired can be suspended between sutures which exit the lower incisions alongside the trocar sheaths (Fig. 16). The bowel is closed in two layers, the first being a running layer of 3-0 Vicryl beginning at one lateral angle and working toward the other (Fig. 17). Each end of the mucosal suture line is buried beneath the bowel wall surface using a purse-string suture of 2-0 silk (Figs 18, 19). The seromuscular layer is closed from each angle to midline with interrupted 2-0 silk (Fig. 20). The repair can be checked for leaks by underwater transanal air pressure examination. If surgery has been lengthy and copious irrigation used, a drain can be pulled through one of the 5 mm ports and placed into the cul-de-sac (Fig. 21) for post-operative drainage. This approach, associated with meticulous attention to haemostasis and peritoneal lavage, is associated with a very low morbidity.

Occasionally a patient with rectosigmoid endometriosis may have a lesion so large that segmental bowel resection is necessary. The steps of the procedure are identical to the point of isolating the nodule to the anterior wall of the mobilized bowel. At this point, coagulation current is used to isolate the involved segment from its mesentery by coagulating and severing the mesentery immediately adjacent to the bowel wall (Fig. 22). The vessels of the mesentery are smallest and easy to coagulate as they enter the bowel wall, although occasionally a larger vessel will be encountered in the centre of the posterior bowel wall. These larger branches of the inferior mesenteric artery can still be controlled with a longer application of 50watts of coagulation current. Since the segment which is being isolated will be removed, there is no concern over thermal damage to the bowel wall, although it is possible to avoid any thermal damage if the tissue is quickly coagulated before being bluntly stripped from the bowel wall. Once the mesentery has been completely separated from the segment to be removed, the normal distal bowel wall is stripped of its enveloping fatty tissue so clean muscularis is present 360 degrees around the bowel (Figs 23, 24). A linear endoscopic stapler is used to staple across this area of normal bowel wall just distal to the mass (Fig. 25). The umbilical incision is enlarged slightly and the proximal segment of rectosigmoid colon with the mass at its stapled end is delivered on to the abdomen (Fig. 26). The bowel wall is transected proximal to the mass, then the anvil of a circular endoscopic stapler is placed into the normal bowel lumen and secured with a purse-string suture, then returned to the abdomen (Fig. 27). The mating spike of the stapler is now forced through the stapled rectal stump (Fig. 28), and the anvil fitted on to it using a laparoscopic grasper (Fig. 29). The stapler is then closed (Fig. 30), and fired and the anastomosis is complete. The tissue within the stapler should have two complete rings of tissue to ensure proper bowel repair. The integrity of the bowel can be checked with underwater air pressure examination.

Another technique which can sometimes be used with low lying rectal nodules is a transvaginal approach, particularly if hysterectomy is undertaken as well. This approach can be used for anterior disc resection or repair, or even segmental bowel resections in some cases [24, 25].

Morbidity from bowel leakage is low when a deliberate decision is taken to perform full-thickness bowel resection for endometriosis. This is because the hole which is created is obvious and must be repaired. Where there is an attempt to dissect in the adhesive plane binding the rectum to the posterior cervix in an effort to avoid bowel entry, the likelihood of occult bowel injury is increased. Management of complications is discussed in the article, Intestinal endometriosis: symptoms, diagnosis, and treatment.

Excision of rectovaginal disease with these techniques is associated with good symptom relief, low morbidity and a short one or two day hospital stay [1, 7].

Other forms of treatment for rectovaginal endometriosis have been described. A large series has been published by Nisolle and Donnez et al in Belgium [26]. Their surgical technique is based on the belief that endometriotic involvement of the rectum does not occur, despite decades of published evidence to the contrary. Plaque disease lying on the anterior wall of the rectum identified by double contrast barium enema is diagnosed as 'peri-visceritis'. Since no biopsy is typically taken from the bowel wall, this diagnosis is a presumption which many experts believe is incorrect. Using a carbon dioxide (CO2) laser as a dissecting tool and staying within the boundaries of the utero-sacral ligaments, they dissect the rectum free from the posterior cervix. Once the rectum is freed, the vaginal adenomyoma is removed transvaginally and the defect sutured. Any residual endometriosis in the uterosacral ligaments is then ablated using the CO2 laser to vaporize the tissue. No attempt is made to remove the plaque from the anterior bowel wall. Good symptomatic relief is claimed for this method, although objective measurement of symptom response was not mentioned. Leaving intestinal disease behind may be evidence that symptoms arise more commonly from the uterosacral ligaments and rectovaginal adenomyoma than from the rectal wall itself. The bowel lumen was entered on three occasions in their series with no significant morbidity resulting, since entry was identified and suture repaired.

Summary

There are few gynaecologists who have had sufficient training in complex laparoscopic techniques to feel confident to undertake rectal repair. There is an equally small number of coloproctologists who understand the pathology of endometriotic disease involving the bowel, particularly the disease of the seromuscular layer which rarely involves the mucosa and which can frequently be treated by limited disc resection. Women with a diagnosis of endometriosis of the bowel or rectovaginal septum, or those with a diagnosis of obliteration of the cul-de-sac, should be referred to specialist centres where this surgery can be carried out most successfully. Specialist surgical treatment centres for endometriosis should allow open access to surgeons wishing to learn these techniques, and where possible, have adequate supervised operative training, as well as training in the correct assessment of patients with this disease. Record-keeping in a standardized form allows valuable information on symptoms and signs to be collected and coded so that comparisons of treatment efficacy can be made.

References

1. Mathias SD, Kuppermann M, Liberman RF, Lipschutz RC, Steege JF. Chronic pelvic pain: prevalence, health-related quality of life, and economic correlates. Obstet Gynecol 1996; 87(3): 321-7.
2. Redwine DB, Wright JT. Laparoscopic treatment of complete obliteration of the cul-de-sac associated with endometriosis: long-term follow-up of en bloc resection. Fertil Steril 2001; 76(2):358-65.
3. Hemmings R, Bissonnette F, Bouzayen R. Results of laparoscopic treatments of ovarian endometriomas: laparoscopic ovarian fenestration and coagulation. Fertil Steril 1998; 70(3):527-9.
4. Jones KD, Sutton CJ. Laparoscopic management of ovarian endometriomas: a critical review of current practice. Curr Opin Obstet Gynecol 2000; 12(4):309-15.
5. Wright JT. The diagnosis and management of infiltrating nodular recto-vaginal endometriosis. Curr Opin Obstet Gynecol 2000; 12(4):283-7.
6. Garry R, Clayton R, Hawe J. The effect of endometriosis and its radical laparoscopic excision on quality of life indicators. BJOG 2000; 107(1):44-54.
7. Boog G, Penot P, Momber A. Ultrasound as a diagnostic aid in endometriosis. Contrib Gynecol Obstet 1987;16:119-24.
8. Dumontier I, Roseau G, Vincent B, Chapron C, Dousset B, Chaussade S et al. [Comparison of endoscopic ultrasound and magnetic resonance imaging in severe pelvic endometriosis]. Gastroenterol Clin Biol 2000; 24(12):1197-204.
9. Roseau G, Dumontier I, Palazzo L, Chapron C, Dousset B, Chaussade S et al. Rectosigmoid endometriosis: endoscopic ultrasound features and clinical implications. Endoscopy 2000;32(7): 525-30.
10. Momoeda M. [Recto-vaginal endometriosis]. Nippon Rinsho 2001;59(Suppl 1):192-5.
11. Stevens TG. Adenomyoma of the vaginal wall. Proc Roy Soc Med 1910;3:57-8.
12. Chapron C, Dubuisson J-B, Pansini V, Vieira M, Fauconnier A, Barakat H, Dousset B. Routine clinical examination is not sufficient for diagnosing and locating deeply infiltrating endometriosis. J Am Assoc Gynecol Laparosc 2002;9: 115-9.
13. Redwine DB. Mulleriosis: the single best fit model of origin of endometriosis. J Reprod Med 1988;33:915-20.
14. Halme J, Hammond MG, Hulka JF, Raj SG, Talbert LM. Retrograde menstruation in healthy women and in patients with endometriosis. Obstet Gynecol 1984;64(2):151-4.
15. Koninckx PR, Barlow D, Kennedy S. Implantation versus infiltration: the Sampson versus the endometriotic disease theory. Gynecol Obstet Invest 1999;47 (Suppl 1):3-9 (Review, 87 refs).
16. Brosens I, Puttemans P, Deprest J. Appearances of endometriosis. Baillière's Clin Obstet Gynaecol 1993;7(4):741-57.
17. Martin DC, Hubert GD, Vander ZR, el Zeky FA. Laparoscopic appearances of peritoneal endometriosis. Fertil Steril 1989;51(1):63-7.
18. Redwine DB. Bowel resection related to endometriosis. In: Tulandi T (ed) Atlas of laparoscopic technique for gynecologists (2nd edn). London: WB Saunders, 1998:85-92.
19. Brosens I, Puttemans P, Deprest J. Appearances of endometriosis. Baillière's Clin Obstet Gynaecol 1993;7(4):741-57.
20. Bardos A. [Pathogenesis of endometriosis and notes on its histological diagnosis]. [Slovak]. Cesk Gynekol 1973;38(1):25-7.
21. Howell RJ, Dowsett M, Edmonds DK. Oestrogen and progesterone receptors in endometriosis: heterogeneity of different sites. Hum Reprod 1994;9(9):1752-8.
22. Jones RK, Bulmer JN, Searle RF. Immunohistochemical characterization of proliferation, oestrogen receptor and progesterone receptor expression in endometriosis: comparison of eutopic and ectopic endometrium with normal cycling endometrium. Hum Reprod 1995;10(12): 3272-9.
23. Redwine DB. Laparoscopic en bloc resection for treatment of the obliterated cul de sac in endometriosis. J Reprod Med 1992;37:695-8.
24. Redwine DB, Koning M, Sharpe DR. Laparoscopically assisted transvaginal segmental bowel resection for endometriosis. Fertil Steril 1996;65: 193-7.
25. Varol N, Maher P, Woods R. Laparoscopic management of intestinal endometriosis. J Am Assoc Gynecol Laparosc 2000;7:405-9.
26. 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(3): 630-5.