Sigmoid endometriosis and ovarian stimulation: Case reports

Vincent Anaf1,4, Issam El Nakadi2, Philippe Simon1, Yvon Englert1, Marie-Odile Peny3, Isabelle Fayt3 and Jean-Christophe Noel3

1 Departments of Gynaecology, 2 Digestive Surgery and 3 Pathology, Hospital Erasme, Universite Libre de Bruxelles (ULB), Brussels, Belgium


    Abstract
 Top
 Abstract
 Introduction
 Case reports
 Discussion
 References
 
In-vitro fertilization (IVF) and ovarian stimulation are frequently performed in patients with endometriosis. Although endometriosis is a hormone-dependent disease, the rate of IVF complications related to endometriosis is low. We report four cases of severe digestive complications due to the rapid growth of sigmoid endometriosis under ovarian stimulation. In three patients, sigmoid endometriosis was diagnosed at laparoscopy for sterility. Because of the absence of digestive symptoms or repercussion on the bowel, no bowel resection was performed before ovarian stimulation. All patients experienced severe digestive symptoms during ovarian stimulation, and a segmental sigmoid resection had to be performed. Analysis of endoscopic and radiological data demonstrated that bowel lesions of small size may rapidly enlarge and become highly symptomatic under ovarian stimulation. At immunohistochemistry, these infiltrating lesions displayed high populations of steroid receptors and a high proliferative index (Ki-67 activity), suggesting a strong dependence on circulating ovarian hormones and a potential for rapid growth under supraphysiological oestrogen concentrations. Clinicians should be aware of this rare but severe digestive complication of ovarian stimulation. The early diagnosis of such lesions may help the patients to avoid months of morbidity falsely attributed to ovarian stimulation side effects. Further experience is necessary to determine the optimal attitude when diagnosing a small and asymptomatic endometriotic bowel lesion before ovarian stimulation.

Key words: deep infiltrating endometriosis/IVF complication/ovarian stimulation/sigmoid endometriosis


    Introduction
 Top
 Abstract
 Introduction
 Case reports
 Discussion
 References
 
The bowel represents the third extrauterine location of endometriosis after the ovaries and the peritoneum. In a series of 3037 laparotomies for endometriosis (Weed and Ray, 1987Go), bowel endometriosis was found to be present in 5.4% of cases. The real prevalence of bowel endometriosis is probably underestimated because this condition may remain undiagnosed for long periods. In infertile patients with endometriosis who are not immediate candidates for in-vitro fertilization (IVF), the therapeutic recommendations are to destroy as much diseased tissue as possible, without producing postoperative adhesions where none was previously present. However, endometriosis outside of the reproductive tract plays little (if any) role in fertility and should therefore be conservatively treated according to location, depth, and the patient's symptoms (Diamond and Osteen, 1997Go). Although the effects of endometriosis and therapy for endometriosis on the success of assisted reproductive treatment have been well studied, few data exist on the consequences of ovarian stimulation on endometriosis. Moreover, while complications related to pregnancies after IVF have been widely published, there is scant information concerning the medicosurgical complications of ovarian stimulation in IVF. Because endometriosis is a hormone-dependent disease, one might expect endometriosis to increase under the influence of supraphysiological circulating concentrations of oestrogens. In a retrospective study on IVF complications, bowel complications appeared to be a rare condition, occurring in 0.13% of oocyte retrieval procedures for IVF and embryo transfer (Govaerts et al., 1998Go). We report four cases of patients who severely worsened during ovarian stimulation and presented severe digestive complications resulting from the rapid growth of an asymptomatic small sigmoid endometriotic lesion.


    Case reports
 Top
 Abstract
 Introduction
 Case reports
 Discussion
 References
 
Patients
Four patients (mean age 32 years; range 30–34 years) who were either consulting in our infertility department (n = 3) or referred (n = 1) complained of severe increasing abdominal pain, cramps, increasing constipation and rectorrhagia after one to seven cycles (mean 3.5 cycles) of ovarian stimulation for IVF (Table IGo). IVF indications were endometriosis stage 4 in all patients; this was associated with male infertility in two cases. All patients had severe cyclical dysmenorrhoea during their menses and dyspareunia for more than 2 years. Patients were previously prescribed eithre gonadotrophin releasing hormone (GnRH) agonists for 6 months (n = 2) or continuous progestagens (n = 2). Both these medications were effective in correcting dysmenorrhoea and dyspareunia, but the symptoms recurred when the treatment was interrupted or discontinued.


View this table:
[in this window]
[in a new window]
 
Table I. Pre- and post-procedural findings
 
Before ovarian stimulation, all four patients underwent a laparoscopy where endometriosis stage 4 was found and treated (Table IGo). No cul-de-sac obliteration was observed at laparoscopy, but a high sigmoid lesion was found in three patients. The lesion appeared as a solid dark-brown nodule of 2 cm diameter (measured with an atraumatic graduated probe) located at the mesenteric side of the sigmoid colon. Biopsies showed the presence of endometrial glands and stroma. The lesions were located at the level of the iliac vessels. Three patients underwent a barium enema and a colonoscopy before starting the stimulation protocol. At a retrospective consultation, none of these three patients presented any specific symptoms of bowel endometriosis. Moreover, these patients had neither significant bowel lumen calibre reduction at barium enema and colonoscopy, nor suspect submucosal lesions at colonoscopy. Therefore it was decided not to perform sigmoid resection before ovarian stimulation.

The ovarian stimulation protocol included GnRH analogues (buserelin acetate, Suprefact spray; Hoechst Inc., Frankfurt-Main, Germany) and human menopausal gonadotrophins (HMG) (Humegon; Organon, Oss, Netherlands). Ovulation induction was performed with 10 000 IU of human chorionic gonadotrophin (HCG) (Pregnyl; Organon, or Profasi; Serono, Aubonne, Switzerland).

In the referred patient (no. 4; Table IGo), the diagnosis of sigmoid endometriosis was not made at laparoscopy. At retrospective consultation, this patient presented symptoms of chronic mid-left quadrant pain and long-term alternation of constipation and diarrhoea, but these symptoms were most likely under-rated because of the predominance of pelvic endometriosis symptoms. When subocclusion and rectorrhagia occurred during ovarian stimulation, gonadotrophin injections were immediately stopped in all four patients. A barium enema and a colonoscopy were then performed (Figure 1Go). An anorectal examination was also performed in order to rule out other causes of rectorrhagia. One patient became pregnant during the cycle where rectorrhagia occurred (patient 1; Table 1Go). Her symptoms progressively regressed during the pregnancy, but symptoms of subocclusion and rectorrhagia recurred in the second cycle after breastfeeding was interrupted. No patient required any temporary faecal diversion for total bowel occlusion or peritonitis due to bowel perforation. All patients underwent a segmental sigmoid resection by laparotomy (n = 1) or by laparoscopy (n = 3) with end-to-end anastomosis (n = 2) or lateroterminal anastomosis (n = 2) without diverting colostomy.



View larger version (109K):
[in this window]
[in a new window]
 
Figure 1. Barium enema performed after ovarian stimulation, showing almost complete sigmoid occlusion and significant circular stricture.

 
Immunohistochemistry
All patients were operated on in the mid-secretory phase and underwent segmental sigmoid resection after resolution of their bowel symptoms. After sigmoid resection the bowel was rinsed and fixed in 4% neutral buffered formalin and the lesions embedded in paraffin. Immunohistochemistry with monoclonal antibody against oestrogen receptors (ER) and progesterone receptors (PR) (ER ID5, PR ID5; working dilution 1/50; Immunotech, Marseilles, France) were used to detect the presence of ER and PR. Immunohistochemistry with monoclonal antibody against the Ki-67 antigen (MIB 1; working dilution 1/100; Immunotech) was used to detect the presence of the Ki-67 antigen and subsequently to measure the labelling index, a good marker of cell proliferation. To control for non-specific binding of the primary antibody, non-immune mouse serum at the same concentration as the primary monoclonal antibody was substituted as the first layer for staining of the serial sections. Positive controls for ER and PR were the normal breast, and for Ki-67 the crypts of the colonic glands within the same section. Immunohistochemical ER, PR and Ki-67 results were expressed as the mean percentage of cells exhibiting definite nuclear immunoreactivity over at least 2000 glandular and stromal cells in more than 10 non-overlapping randomly selected high-power fields (x400). All sections were scored by two observers (V.A. and J.-C.N.) and the discrepancies discussed and resolved.


    Discussion
 Top
 Abstract
 Introduction
 Case reports
 Discussion
 References
 
Postoperative infections, intraperitoneal bleedings, adnexal torsions and ovarian hyperstimulation syndrome are the most common medicosurgical complications of IVF. However, few authors have described endometriosis-related IVF complications. In 1993, two patients were reported among 3656 oocyte retrievals with endometriosis (0.06%) who became severely symptomatic at 6 and 8 h after the procedure (Dicker et al., 1993Go); laparoscopy revealed ruptured endometriotic cystic masses. A case of ureterohydronephrosis due to the growth of (peri)ureteral endometriosis has been described (Renier et al., 1995Go) during ovarian stimulation. In the available literature, only two cases of IVF complications due to bowel endometriosis have been reported (Govaerts et al., 1998Go). One patient complained of severe dysmenorrhoea, and the other required intestinal resection because of the occurrence of rectorrhagia. In unstimulated patients, a rapid return of gastrointestinal symptoms has been reported when residual bowel endometriosis was stimulated by ovarian, exogenous or even peripherally converted endogenous oestrogens (Clayton et al., 1999Go). Cases have also been reported of recurrent bowel endometriosis due to exogenous administration of oestrogens in women previously treated by bilateral oophorectomy (Redwine, 1991Go). In our series, one patient became pregnant and experienced a rapid return of the digestive symptoms after interruption of breastfeeding.

The most frequent symptoms suggestive of bowel involvement are cyclical diarrhoea (26%) followed by cyclical rectorrhagia (16%), dyschezia (15%) and cyclical constipation (11.6%) (Weed and Ray, 1987Go). A distinction must be drawn, however, between the `endometriotic disease of the bowel'—which is generally highly symptomatic—and the incidental finding of endometriotic implants on the bowel during laparotomy or laparoscopy. Of course, not all endometriotic lesions of the bowel require surgery. The available literature on bowel endometriosis suggests that the indications for surgery are the treatment of symptomatic lesions and the exclusion of malignancy (Meyers et al., 1979Go; Amano and Yamada, 1981Go; Graham and Mazier, 1988Go; Prystowsky et al., 1988Go). In case of suspicion of neoplasm, perioperative frozen sections must be prepared. In the absence of obstructive symptoms, treatment can be initiated with progestagens, danazol or oral contraceptives. GnRH agonists have been used with success in some cases of deep endometriosis (Markham et al., 1991Go).

In this series, the sigmoid lesions had been diagnosed at laparoscopy in three cases. Preliminary results on the role of rectal endoscopic ultrasonography have shown that this simple and non-invasive technique provides reliable information as to the presence of deep bowel infiltration in patients with retroperitoneal endometriotic lesions (Chapron et al., 1998Go). However, mid and high sigmoid endometriotic lesions, as reported here, are located higher than rectovaginal or rectosigmoid lesions and are inaccessible to rectal endoscopic ultrasonography. Because these three patients were asymptomatic and their lesions did not have repercussions on the bowel lumen calibre and were not visible at colonoscopy, we decided not to perform segmental sigmoid resection before ovarian stimulation.

Many consider the hard fibrotic lesions associated with this type of endometriosis to be `old' or `burnt out'. In fact, the immunohistochemical study strongly suggested that these lesions were biologically very active and could invade the bowel muscularis (Figure 2Go), the submucosa and the mucosa. The populations of ER and PR were very high in both the glands and the stroma, suggesting that such lesions are highly responsive to exogenous oestrogens (Figure 2Go).



View larger version (145K):
[in this window]
[in a new window]
 
Figure 2. Sigmoid wall section (oestrogen receptor, ER) immunohistochemistry and Mayer's haematoxylin counterstaining). High ER positivity in glandular and stromal tissue of a lesion located in the bowel muscularis. Scale bar = 280 µm.

 
Deep infiltrating endometriosis is defined as the presence of endometrial glands and stroma more than 5 mm under the peritoneum (Cornillie et al., 1990Go). Sigmoid lesions such as those resected represented deep infiltrating nodular lesions. Such lesions have escaped from the predominant influence of the peritoneal fluid and are probably mainly influenced by circulating steroids. In another report (Clayton et al., 1999Go), five cases were reported of patients who presented with increasing pelvic pain, deep dyspareunia and rectal pain due to the presence of deep infiltrating endometriosis that was left behind after hysterectomy and bilateral salpingo-oophorectomy for endometriosis. In that series, symptoms progressively occurred under unopposed oestrogen replacement therapy, and the immunohistochemical study also showed the presence of positive immunoreactivity for ER and PR (Clayton et al., 1999Go).

Ki-67 is an antigen that corresponds to a nuclear non-histone protein expressed by cells in the proliferative phases G1, G2, M and S. The Ki-67 labelling index is therefore a good marker of cell proliferation (Rosai, 1996Go), and was found to be elevated in all four patients in both stromal and glandular epithelial cells, with more than 20% of the cells being identified as `cycling' (Table IIGo). According to the immunohistochemical results and the `deep infiltrating nature' of the lesions, retrospectively it is not surprising that these patients—who had no digestive symptoms before ovarian stimulation—severely worsened under supraphysiological concentrations of oestrogens. Indeed, in all four patients, abdominal pain with cramps, distension, tenderness, constipation and rectorrhagia occurred when the oestradiol concentrations were between 2230 and 2635 pg/ml (Table IGo). The symptoms partially regressed with the progressive return to physiological concentrations of oestradiol, but the lesions remained enlarged because some of the lesion components were irreversible. These lesions are composed not only of endometrial glands and stroma but also of a smooth muscle hyperplasia and an intramural fibrosis that are irreversible (Figure 3Go).


View this table:
[in this window]
[in a new window]
 
Table II. Immunohistochemical resultsa using monoclonal antibodies to detect oestrogen receptors (ER), progesterone receptors (PR) and Ki-67 antigen in lesions from four patients after sigmoid resection of the bowel
 


View larger version (116K):
[in this window]
[in a new window]
 
Figure 3. Macroscopic section through the lesion at the mesenteric side. From left to right: the mesosigmoid fat delimited by the normal bowel wall (1), the haemorrhagic nodular lesion (2) and the thickened bowel wall containing haemosiderin clusters, the bowel lumen on the right (3).

 
This report poses the question of which strategy to adopt when an asymptomatic endometriotic bowel lesion is diagnosed in a patient who will undergo ovarian stimulation. Unfortunately, the rarity of this situation precludes the possibility of guidelines based upon trials or retrospective comparison of therapy options: `evidence-based medicine' is not a real option. It has been shown that full-thickness resection of the colon for the treatment of deep bowel endometriosis is a safe procedure with low morbidity, good postoperative relief of symptoms and favourable pregnancy rates (Coronado et al., 1990Go). In the context of infertility, resection of deep endometriotic lesions infiltrating the uterosacral ligaments has been shown to be safely associated with classic treatments of endometriosis (adhesiolysis, intraperitoneal cystectomies, bipolar coagulation of peritoneal implants) (Chapron et al., 1999Go), and enables most of the treated patients to become pregnant. We recently described a safe and feasible, laparoscopically assisted sigmoid resection technique for sigmoid endometriosis where other endometriotic locations can be simultaneously treated (Anaf et al., 2000Go). However, it must be emphasized that such a procedure prolongs the operating time and might lead to an increase in morbidity of the laparoscopic procedure because of the potential complications of bowel resection.

This report shows that patients, even with small asymptomatic sigmoid endometriotic lesions, can develop severe digestive complications under ovarian stimulation. More cases and experience are necessary to determine which is the optimal strategy to adopt in order to avoid this rare but severe complication. An effort should be made to achieve early diagnosis because in this way potential severe complications may be avoided, as well as months of morbidity falsely attributed to ovarian stimulation side effects, or recurrence of pelvic endometriosis under high concentrations of circulating steroids. Indeed, in one patient (Table IGo, patient no. 4) where the diagnosis was missed at laparoscopy, it took seven cycles with increasing abdominal pain before the diagnosis was made.


    Notes
 
4 To whom correspondence should be addressed at: Department of Gynaecology, Hospital Erasme, Free University of Brussels, 808, Route de Lennik, 1070 Brussels, Belgium Back


    References
 Top
 Abstract
 Introduction
 Case reports
 Discussion
 References
 
Amano, S. and Yamada, N. (1981) Endometrioid carcinoma arising from endometriosis of the sigmoid colon. A case report. Hum. Pathol., 12, 845–848.[ISI][Medline]

Anaf, V., Sperduto, N., Simon, Ph. et al. (2000) Laparoscopically assisted segmental sigmoid resection (LASSR) for sigmoid endometriosis. Gynaecol. Endosc. (in press).

Chapron, C., Dumontier, I., Dousset, B. et al. (1998) Results and role of rectal endoscopic ultrasonography for patients with deep pelvic endometriosis. Hum. Reprod., 13, 2266–2270.[Abstract]

Chapron, C., Fritel, X. and Dubuisson, J.B. (1999) Fertility after laparoscopic management of deep endometriosis infiltrating the uterosacral ligaments. Hum. Reprod., 14, 329–332.[Abstract/Free Full Text]

Clayton, R.D., Hawe, J.A., Lowe, J.C. et al. (1999) Recurrent pain after hysterectomy and bilateral salpingo-oophorectomy for endometriosis: evaluation of laparoscopic excision of residual excision. Br. J. Obstet. Gynaecol., 106, 740–744.[ISI][Medline]

Cornillie, F.J., Oosterlynck, D., Lauweryns, J.M. et al. (1990) Deeply infiltrating endometriosis: histology and clinical significance. Fertil. Steril., 53, 978–983.[ISI][Medline]

Coronado, C., Franklin, R.R., Lotze, E.C. et al. (1990) Surgical treatment of symptomatic colorectal endometriosis. Fertil. Steril., 53, 411–416.[ISI][Medline]

Diamond, M.P. and Osteen, K.G. (eds) (1997) Endometrium and Endometriosis. Blackwell Science, London, pp. 202–211.

Dicker, D., Ashkenasi, J., Feldberg, D. et al. (1993) Severe abdominal complications after transvaginal ultrasonographically guided retrieval of oocytes for in vitro fertilization and embryo transfer. Fertil. Steril., 53, 1313–1315.

Govaerts, I., Devreker, F., Delbaere, A. et al. (1998) Short-term medical complications of one thousand five hundred oocyte retrievals for in vitro fertilization and embryo transfer. Eur. J. Obstet. Gynecol. Reprod. Biol., 77, 239–243.[ISI][Medline]

Graham, B. and Mazier, W. (1988) Diagnosis and management of endometriosis of the colon and rectum. Dis. Colon Rectum, 31, 952–956.[ISI][Medline]

Markham, S.M., Welling, D.R., Larsen, K.S. et al. (1991) Endometriosis of the rectum treated with long term Gn-Rh agonist and surgery. N.Y. State J. Med., 91, 69–71.

Meyers, W.C., Kelvin, F.M. and Jones, R.S. (1979) Diagnosis and surgical treatment of colonic endometriosis. Arch. Surg., 114, 169–175.[Abstract]

Prystowsky, J.B., Stryker, S.J., Ujiki, G.T. et al. (1988) Gastrointestinal endometriosis: incidence and indications for resections. Arch. Surg., 123, 855–858.[Abstract]

Redwine, D.B. (1991) Conservative laparoscopic excision of endometriosis by sharp dissection: life table analysis of reoperation and persistent or recurrent disease. Fertil. Steril., 56, 628–634.[ISI][Medline]

Renier, M., Verheyden, B. and Termote, L. (1995) An unusual coincidence of endometriosis and ovarian stimulation. Eur. J. Obstet. Gynecol. Reprod. Biol., 63, 187–189.[ISI][Medline]

Rosai, J. (ed.) (1996) Ackerman's Surgical Pathology, Vol. 1, 8th edn. Mosby, St Louis.

Weed, J.C. and Ray, J.E. (1987) Endometriosis of the bowel. Obstet. Gynecol., 123, 855–858.

Submitted on September 17, 1999; accepted on January 5, 2000.