Endometriosis is a relatively common problem in women of reproductive age. Since it cannot be diagnosed accurately without seeing inside the abdomen it is well nigh impossible to discover the true prevalence of endometriosis but estimates suggest that it affects at least one per cent, but less than five per cent, of women of reproductive age. In selected populations which are more likely to have diagnostic laparoscopy, such as those with chronic pelvic pain or infertility, the prevalence of endometriosis is reported to be much higher.
The management of endometriosis is normally within the jurisdiction of the gynaecologist and it is sensible to involve a gynaecologist in the case at an early stage. In gynaecological practice today endometriosis should rarely have to involve laparotomy since the surgical management of most endometriosis problems is by laparoscopic surgery. If this approach is taken then the surgeon is only likely to find him/herself confronted by significant endometriosis if it is encountered unexpectedly during surgery for an acute abdomen or as a coincidental finding during elective surgery for another problem.
There are undoubtedly pitfalls for the surgeon which could lead to inappropriate and possibly excessive surgery. This may be disastrous (i) for the woman's reproductive potential and may be carried out in a situation where an adequate reproductive history (past experience and future intentions) has not been taken; or (ii) for the woman's future oestrogen production if the ovaries are removed. In gynaecological practice this latter action now demands specific informed consent and discussion if bilateral oophorectomy is a possibility. In general the gynaecological management of endometriosis is surgically conservative.
Radical surgery involving hysterectomy and possibly bilateral oophorectomy is generally regarded as a late stage procedure for chronic intractable pelvic pain. This article will not attempt a comprehensive account of the gynaecological management of endometriosis, but will provide a limited perspective on that management and mainly provide information relevant to a surgeon who might need to be able to suspect endometriosis in advance of surgery or unexpectedly encounters it at laparotomy where it presents the surgeon with a dilemma since open surgery is not usually the treatment of choice.
The nature of endometriosis
Endometriosis is the presence, in sites outside the uterus, of glands and stroma which histologically resemble endometrium. The origin of this tissue remains uncertain but the two most likely explanations are either (i) metaplastic transformation or (ii) seeding and implantation of menstrual endometrium which has reached the peritoneal cavity by retrograde menstruation, which is a common event. The consequences of endometriosis depend on the site: the most common sites are the ovaries, the pelvic peritoneal surfaces, and the posterior aspect of the uterus. Other sites such as bladder and bowel are occasionally affected and some distant sites are rarely involved but often quoted because the cyclical bleeding at sites such as the umbilicus or lung are noteworthy.
An important unresolved current controversy surrounds the question of whether minimal and mild endometriosis usually reflects normal female physiology or is an early stage of disease usually leading on to the more significant forms of advanced disease. These more advanced forms are (i) ovarian endometriotic 'cysts', (ii) extensive peritoneal involvement with adhesion formation, and (iii) deep uterovaginal septum disease. It is argued that localized implants on peritoneal surfaces may be transient and sporadic physiological events, with the retrograde endometrial tissue surviving for a time after retrograde menstruation but if left alone it would eventually clear. In this hypothesis the advanced forms are considered to be true pathological processes but probably distinct from minimal endometriosis.
Endometriotic implants are now recognized to have many appearances in addition to the traditional teaching which represented endometriosis as either 'chocolate cysts' or bluish-black peritoneal nodules or 'powder burn' lesions which may be associated with a local inflammatory reaction, fibrosis, and adhesions. These are now known to represent only part of the spectrum of peritoneal endometriosis. Biopsy studies indicate that much peritoneal endometriosis is not bluish-black in appearance: small haemorrhagic and non-haemorrhagic lesions represent what may be stages in the progression of the disease. These less classical appearances have been correlated with the rate of biopsy confirmation of the diagnosis. There is a close correlation for white opacified peritoneum (81 per cent), red flame-like lesions (81 per cent), and glandular lesions resembling endometrium (67 per cent), and reasonable correlations for subovarian adhesions (50 per cent), yellow-brown patches (47 per cent), and circular peritoneal defects (45 per cent).
The issue of diagnosis has been further compounded by the evidence that biopsy samples from visually normal peritoneum often contain microscopically detectable endometrial tissue. This observation has not made an impact on clinical practice despite having been made several years ago and is of little relevance to the situations which might confront a surgeon.
The inflammation associated with endometriotic implants is probably due to the production of prostanoids and angiogenic cytokines. The peritoneal reaction can be severe, causing the formation of extensive adhesions, particularly after surgery. In such women a second laparotomy may reveal densely adherent organs, matted loops of bowel, and pelvic organs making it difficult to distinguish individual structures).
The forms of advanced disease
Extensive peritoneal involvement with adhesion formation
Some women have extensive involvement of the peritoneal surfaces with an excessive peritoneal reaction. This may take the form of clear blebbing and formation of large peritoneal vesicles which may become detached from the underlying structures. Whole surfaces may be affected and biopsy reveals the pathology to be endometriosis even though there may be no red, blue, or black lesions to be seen. Alternatively the extensive peritoneal involvement may be in the form of widespread reddish implants. The peritoneal involvement may be associated with a tendency to adhesion formation which can lead to obliteration of the pouch of Douglas and in some cases can be so extensive that the pelvic organs can be unrecognizable.
This form of endometriosis can be confused with haemorrhagic luteal cysts which also produce ovarian cysts that contain inspissated blood which has a 'chocolate' appearance. These luteal cysts can be several centimetres in diameter but would not involve adhesion to adjacent structures. Endometriotic cysts frequently involve adhesion to the underlying ovarian fossa and it is thought that many result from an initial adhesion of the ovary to a peritoneal lesion in the ovarian fossa which then bleeds into the adherent portion of ovary. This displaces the ovary so that an apparent 'cyst' pushes into the ovary but the cyst contents are actually outside the ovarian cortex. In such cases any surgical attempt to mobilize the ovary will inevitably rupture the 'cyst'. Some other endometriotic cysts are thought to arise by invagination of the ovarian surface at the site of endometriotic implants. Such 'cysts' would again be extraovarian, being outside the ovarian cortex.
Deep uterovaginal septum disease
This is also referred to as deep invasive endometriosis. This form of endometriosis has attracted much attention in recent years having been relatively ignored previously. It is the form of disease which provides the greatest surgical challenge and carries the highest risk of significant complications such as bowel or ureteric damage. These lesions are deposits of endometriotic tissue deep in the uterovaginal septum which may be invisible at laparoscopy so that a routine diagnostic laparoscopy might be reported as showing no endometriosis. Sadly these lesions can be particularly painful and tender, especially at menstruation. The origin of these deep lesions is another controversial issue. Suggestions include direct invasion from the peritoneal cavity, invagination of implants on the peritoneum in the lowest aspect of the pouch of Douglas, or that they may arise directly in the septum. It is noteworthy that these deep lesions often involve not only glands and stroma but also muscle cells. For this reason it is suggested that this is actually a form of adenomyosis.
Endometriosis detected at a general surgical laparotomy tends to be one of the more severe forms of pelvic disease or intestinal endometriosis, since the more subtle appearances are usually restricted to the pelvis and seen only on close inspection with the laparoscope.
Symptoms of endometriosis
The characteristic symptom of endometriosis is menstrual or immediately premenstrual pain. This pain may reflect stretching of tissues by a local 'menstruation', a local effect of prostaglandins, or might relate to tissue damage and adhesion formation. There is wide variation in the extent of distress, and this correlates poorly with the extent of observed disease. Many women with endometriosis are asymptomatic: some of these have severe disease discovered either after the incidental detection of an ovarian cyst or during investigation of infertility.
Deep dyspareunia can occur, probably resulting from pressure on affected organs such as a fixed retroverted uterus, affected ovaries, uterosacral ligaments, and the rectovaginal septum, but these areas can be affected without such symptoms. In one series of more than 700 women only one in four of those who were sexually active and had disease affecting the pouch of Douglas (cul-de-sac) reported dyspareunia.
Endometriosis is not a common cause of an acute abdomen but if it occurs it could be through bowel involvement causing obstruction or by rupture of an ovarian 'chocolate cyst'. This is an uncommon complication but is more likely than torsion of an endometriotic cyst, since such cysts tend to be locally adherent.
Clinical signs of endometriosis
Clinical examination may increase suspicion of the presence of pelvic endometriosis. An ovarian mass or masses, thickening or tenderness in the uterosacral or pouch of Douglas areas, and fixed uterine retroversion are all features which should warn the surgeon of an increased risk of endometriosis, particularly if accompanied by characteristic symptoms. Non-invasive investigative methods such as serum markers, ultrasonic imaging, or CT scanning have not proved sufficiently specific to have found a place in routine diagnosis. The most promising non-invasive investigation is magnetic resonance imaging. This is now recognized to identify lesions of 1 cm or larger, but lesions of less than 1 cm must still be regarded as equivocal using MRI. On T2-weighted images the endometriotic implants give a bright signal. If there is an ovarian cyst, this gives a bright signal on T2 weighting if the contents of the cyst are fatty (dermoid cyst) or contain haemosiderin (endometriotic cyst). If fat suppression is then applied the fat is no longer bright but the endometrioma remains bright, thus facilitating the characterization of lesions not possible by other non-invasive means. Despite MRI providing the most useful diagnostic information, its widespread use in decision making is limited by cost.
Suspecting pelvic endometriosis: findings at laparotomy
The different findings at laparotomy present the surgeon with different considerations. Factors to be weighed in decision making include whether the woman intends pregnancy in the future, whether she is experiencing subfertility, whether she is symptomatic, and what therapy would have been chosen for that set of circumstances if they had been detected in advance by laparoscopy.
The likely findings at laparotomy can be classified into broad categories as ovarian cysts, adhesive disease, intestinal lesions, and disease limited to the pelvis. Localized deep rectovaginal septum disease is not likely to be a prominent problem at laparotomy although it may be encountered at colorectal surgery.
A laparotomy may be performed following rupture of an ovarian cyst. Endometriotic 'chocolate cysts' can be several centimetres in diameter, and their rupture releases viscous inspissated blood into the peritoneal cavity. Resection of such cysts is appropriate since medical treatment tends to have little impact. The gynaecological management of such cysts would preferably be via laparoscopic surgery. An unruptured endometriotic cyst would be drained and the lining visualized and any endometriotic foci ablated by laser or diathermy. If the cyst is large then some authorities favour drainage by laparoscopy followed by hormonal suppressive therapy (gonadotrophin-releasing hormone (GnRH) agonist) and ablation of endometriotic foci as an interval laparoscopic procedure. Generally attempts to remove the lining prove fruitless since there is no distinct cyst wall, the cyst lining being the invaginated ovarian cortex. Sometimes the cyst lining can be stripped and this may be more likely if the 'chocolate cyst' is actually a haemorrhagic corpus luteum cyst and not endometriosis. If there is already a ruptured cyst and it is discovered by the surgeon at laparotomy, it is still preferable that the management is approximated to the scheme described although if the abdomen is open then even a large cyst can be dealt with by ablating endometriotic foci without the need for a two stage process. The key principle is that the approach should be as conservative as possible.
If there is certainty that the woman does not intend to have children in the future, it can be appropriate to remove the whole ovary, having ensured that the other ovary is healthy. If bilateral cysts are found it is important to preserve some ovarian tissue. The woman's specific consent to the possibility of oophorectomy should have been obtained before the operation. Until relatively recently it was considered inappropriate to remove both ovaries yet leave the uterus intact since the retained uterus would complicate the necessary subsequent hormone replacement therapy. This argument still applies, with consent, if there is no wish for a future pregnancy. If pregnancy may be desired, the retention of the uterus without the ovaries can be considered if the woman is prepared to undergo a pregnancy by ovum donation, where she is not the genetic mother.
Pelvic adhesions are common in association with endometriosis, and may cause extensive matting of tissue including pelvic organs and/or bowel. These adhesions should be separated, taking care to minimize tissue handling, drying of tissues, and haemorrhage. The extent to which adhesions will reform is uncertain, and there is interest in adhesion barrier products which are under evaluation. In current gynaecological practice adhesolysis is electively performed via the laparoscope when possible since laparoscopic techniques are thought to cause fewer subsequent adhesions and the lens of the scope takes the eye of the surgeon close to the operation site with substantial magnification facilitating fine dissection. The carbon dioxide laser is particularly useful for adhesolysis because the energy generated by the laser beam is dissipated within 0.1 mm; thus with experience it can be used to separate bowel from other structures.
Surgeons are occasionally confronted at laparotomy by endometriosis affecting either the superficial serosa or the muscular layer of the bowel. Most series reporting this have been small. One series of 127 cases, which excluded superficial lesions, reported sigmoid and rectal involvement in 52 per cent, and appendix, ileal, and caecal disease in 22, 17, and 5 per cent, respectively. One-quarter of patients with disease of the sigmoid and rectum had rectovaginal septum involvement, and ileal disease always involved the last 50 cm of the ileum. Other sites are rarely affected.
Surgeons are more likely to resect bowel and to perform castration but are less consistent in checking the pelvic organs. They are also less likely to perform selective excision of genital lesions and to prescribe hormonal therapy.
A perimenstrual barium enema with double contrast can be valuable if there is clinical suspicion of intestinal endometriosis. This will reveal features such as filling defects, stricture, and external compression. Sigmoidoscopy can also be useful, revealing the endometriotic lesions and permitting biopsy. Again, MRI may provide valuable information.
At the time of laparotomy, excision of bowel lesions thought to be endometriotic is justified. There is a fairly high recurrence rate for bowel symptoms after medical treatment, as well as a risk of subsequent obstruction or malignant transformation of intestinal endometriosis. If this is thought to be an outside possibility, appropriate consent should have been obtained and bowel preparation given.
Disease limited to the pelvis
The management of endometriosis limited to the reproductive organs and pelvic peritoneum, and without significant ovarian cysts, should take the form of assessment of the extent of disease and action should be guided by a knowledge of the woman's symptoms and fertility requirements.
The first step should be to score the endometriosis as indicated by the revised American Fertility Society classification. This scoring system grades the severity of disease and facilitates comparability of assessment between patients. If the surgeon does not have the scoring system available then it is valuable to record the extent of endometriosis in detail in the operation note. The nature of the pelvic organs render them particularly suitable for a freehand drawing which would include the uterus, tubes, and ovaries. Localized implants can be represented schematically as dots and extensive peritoneal involvement can be represented by shading. Adhesion between structures is often represented as some form of hatching. The important point is that, since endometriosis tends to be a chronic relapsing problem, an accurate record of this type will be valuable to subsequent clinicians who may have to manage the case. Once the assessment has been made, treatment options can be considered.
If the woman is thought to have endometriosis-related pain, the lesions should be ablated using laser, endocoagulation, or electrodiathermy. Any surgeon attempting such ablation must be aware that some peritoneal lesions overlie sensitive structures such as the ureter. The most likely sites for ablation will be the surface of the uterus, the ovaries, and the pelvic peritoneum.
If the woman is thought not to have pain relating to endometriosis then, if the endometriosis is moderate or severe, it is still appropriate to ablate lesions. When the disease is minimal or mild the case for intervention is less clear, particularly in asymptomatic women who are being investigated for infertility. There is no good evidence that medical treatment improves on pregnancy rates achieved in untreated women with minimal or mild disease. There is now evidence from a large, randomized, controlled trial to suggest that ablation of implants by laser or diathermy at diagnostic laparoscopy can improve the subsequent pregnancy rate over the next 9 months in women with minimal or mild endometriosis, but there are conflicting results from other studies. At laparotomy, if lesions are seen, this evidence would appear to support the ablation of implants where it is judged that it can be done without damaging structures such as ureter or bowel.
The extent of surgical intervention must be influenced by desired fertility. Ovaries affected only by surface lesions should not be removed. Lesions can be ablated or treated medically, but unless significant ovarian cysts are present the ovary should not be damaged further since it is of major importance for continued ovarian activity and future pregnancy, whether achieved naturally or by assisted conception.
Postoperative hormone therapy
Since it is difficult to ablate all lesions surgically there is a logical case for a period of hormone treatment of perhaps 3 to 6 months, but there is very limited evidence at present on the combined application of surgical and medical management. Medical therapy should be omitted if there is a desire to conceive quickly.
If medical therapy is required, the best evidence of efficacy is for the use of danazol, gestrinone, or the GnRH-agonist drugs (buserelin, nafarelin, goserelin, or leupolide). These alternatives are equally effective in the treatment of endometriosis but have different side-effects and in general women report a greater acceptance of the hypoestrogenic side-effects of GnRH agonists than of the relatively androgenic side-effects of danazol or gestrinone. There is a growing body of evidence on the use of 'addback' therapy to remove the hypoestrogenic consequences of GnRH-agonist therapy without losing the efficacy of the agonist. Progestogens, such as medroxyprogesterone acetate or norethisterone, are thought to be effective, but there is a less substantial body of efficacy data; however, some women will find the progestogenic side-effects more acceptable. Even less efficacy information is available for the option of prescribing a combined oral contraceptive pill, preferably continuously since this will generally abolish mentrual events and provide relief from dysmenorrhoea. Since the medical management issues can be complex, it is preferable to involve a gynaecologist in the management.
Women who undergo bilateral oophorectomy should receive oestrogen replacement therapy to counter their, otherwise, increased risk of postmenopausal osteoporosis, myocardial infarction, and symptoms of the menopause.
Anxiety that oestrogen replacement might stimulate recurrence of endometriosis has some foundation, but the few published studies suggest that the risk of recurrence is low. The consequences of an untreated premature menopause are sufficiently severe for the risk of oestrogen replacement to be worth taking. The decision may be more difficult in patients with intestinal endometriosis since the one follow-up study in which all patients had bowel endometriosis reported an 18 per cent recurrence rate. It should be remembered, however, that endometriosis may recur, whatever therapy is used. Following open conservative surgery the recurrence rate is reported at 3, 6, and 9 years as 10.5, 26.6, and 47.2 per cent, respectively.
For many women the management of their endometriosis, if it is severe, can become a test of endurance. They suffer chronic pain and dyspareunia and can have recurrent episodes of disease which may eventually so exhaust medical and surgical options (including pelvic clearance) that there is little to offer except the services of a pain clinic.
Bellina JH, Hemmings R, Voros IJ, Ross LF. Carbon dioxide laser and electrosurgical wound study with an animal model. A comparison of tissue damage and healing patterns in peritoneal tissue. American Journal of Obstetrics and Gynecology 1984; 148: 327–34. [Description of the effects of different laparoscopic treatment modalities.]
Brough RJ, O'Flynn K. Recurrent pelvic endometriosis and bilateral ureteric obstruction associated with hormone replacement therapy. British Medical Journal 1996; 312: 1221–2. [Description of a serious complication of endometriosis.]
Delicata RJ, Clarke GW, Roy MK, Shaw RW, Carey PD. Presentation of endometriosis to general surgeons: a 10 year experience. British Journal of Surgery 1996; 83: 711.
Donnez J, Nisolle M, Gillet N, Smets M, Bassil S, Casanas Roux F. Large ovarian endometriomas. Human Reproduction 1996; 11: 641–6. [Discussion of two-stage management of ovarian endometriosis.]
Donnez J, Nisolle M, Gillerot S, Smets M, Bassil S, Casanas Roux F. Rectovaginal septum adenomyotic nodules: a series of 500 cases. British Journal of Obstetrics and Gynaecology 1997; 104: 1014–18. [Large series of rectovaginal disease, discusses morphological nature of this form of endometriosis.]
Evers JL. Endometriosis does not exist; all women have endometriosis. Human Reproduction 1994; 9: 2206–9. [Discussion of possibility that mild endometriosis may be physiological and not pathology.]
Feichtinger W, Kemeter P. Pregnancy after total ovariectomy achieved by ovum donation. Lancet 1985; ii: 722–3. [Justification for conservation of uterus at bilateral oophorectomy.[ Gordon R, Evers K, Kressel, Laufer I, Herlinger H. Double contrast enema in pelvic endometriosis. American Journal of Roentgenology 1982; 138: 549–52.
Goulbourne JA, MacLeod DA. Endometriosis presenting as an acute abdomen. British Journal of Clinical Practice 1986; 40: 124–6.
Gray LA. Endometriosis of the bowel: role of bowel resection, superficial excision, and oophorectomy in treatment. Annals of Surgery 1973; 177: 580–7.
Grodstein F et al. Postmenopausal estrogen and progestin use and the risk of cardiovascular disease. New England Journal of Medicine 1996; 335: 453–61. [Review of the relevance of hormone replacement therapy to cardiovascular disease risk.]
Groff T. Classifications. In: Schenken RS, ed., Endometriosis: contemporary concepts in clinical management, pp. 145–67. Lippincott, Philadelphia, 1989. [Discussions of the issues involved in the classification system of endometriosis.]
Hornstein MD, Hemmings R, Yuzpe AA, Heinrichs WL. Use of nafarelin versus placebo after reductive laparoscopic surgery for endometriosis. Fertility and Sterility 1997; 68: 860–4. [Evidence on the use of postoperative medical therapy.]
Jansen RPS, Russell P. Nonpigmented endometriosis: clinical, laparoscopic, and pathological definition. American Journal of Obstetrics and Gynecology 1986; 155: 1154–9. [Description of the variety of non-pigmented lesions in endometriosis and histological correlations.]
Karthaus M, Prahst A, Geissler RG, Hertenstein B, Degenhardt F, Ganser A. Acute abdomen due to endometriosis as a diagnostic and therapeutic challenge in the treatment of acute myelocytic leukemia. Annals of Hematology 1997; 74: 29–31.
Kennedy SH, Williams IA, Brodribb J, Barlow DH, Shaw RS. A comparison of nafarelin acetate and danazol in the treatment of endometriosis. Fertility and Sterility 1990; 53:924–8. [Description of efficacy and side-effects of medical treatments.]
Koninckx PR, Martin DC. Deep endometriosis: a consequence of infiltration or retraction or possibly adenomyosis externa. Fertility and Sterility 1992; 58: 924–8. [Discussion of the morphology of deep infiltrative endometriosis.]
Koninckx PR, Muyldermans M, Moerman P, Meuleman C, Deprest J, Cornillie F. CA 125 concentrations in ovarian 'chocolate' cyst fluid can differentiate an endometriotic cyst from a cystic corpus luteum. Human Reproduction 1992; 7: 1314–17.
Lansac J, Pierre F, Letessier E. Digestive endometriosis: reults of a multicentre investigation. Contributions to Gynecology and Obstetrics 1987; 16: 192–204.
Macafee C, Greer H. Intestinal endometriosis. A report of 29 cases and a survey of the literature. Journal of Obstetrics and Gynaecology of the British Empire 1960; 67: 549–55.
Murphy AA, Green WR, Bobbie D, dela Cruz, ZC, Rock JA. Unsuspected endometriosis documented by scanning electron microscopy in visually normal peritoneum. Fertility and Sterility 1986; 46: 522–4. [Evidence for the presence of microscopic endometriosis in normal-looking peritoneum.]
Nisolle N, Donnez J. Peritoneal endometriosis, ovarian endometriosis, and adenomyotic nodules of the rectovaginal septum are three different entities. Fertility and Sterility 1997; 68: 585–96. [Discussion of the nature of different forms of endometriosis.]
O'Connor DT. Endometriosis, p. 71. Churchill Livingstone, Edinburgh, 1987. [Monograph on endometriosis which details a large Australian case series.]
Redwine DB. Age-related evolution in color appearance of endometriosis. Fertility and Sterility 1987; 48: 1062–3. [Correlation of different peritoneal lesions with age suggesting a possible pattern of progression appearances of lesions.]
Redwine DB. Conservative laparoscopic excision of endometriosis by sharp dissection: life table analysis of reoperation and persistent or recurrent disease. Fertility and Sterility 1991; 56: 628–34. [A description of disease recurrence risks.]
Revised American Fertility Society classification of endometriosis. Fertility and Sterility 1985; 43: 351–2.
Wheeler JM, Malinak LR. Recurrent endometriosis. Contributions to Gynecology and Obstetrics 1987; 16: 13–21. [A description of disease recurrence.]