Ovulation suppression for endometriosis

Although the etiology of endometriosis is unknown, several theories exist, the most popular of which is retrograde menstruation. As endometriosis can only be diagnosed by laparoscopy, neither the incidence (annual occurrence) nor the prevalence (proportion of the population affected) of endometriosis is known. The association between endometriosis and infertility isn't clear in Stage I (minimal) and Stage II (mild) endometriosis. Endometriosis appears to be an estrogen dependent condition. At the time of menopause, most endometriosis becomes quiescent. This hormonal dependency prompted researchers to seek agents which would suppress ovarian activity.

To determine effectiveness of a) ovulation suppression with danazol, medroxy progesterone acetate, gestrinone, combined oral contraceptive pills and GnRH analogues versus placebo or no treatment and b) any of the above agents versus danazol, for the treatment of endometriosis explained infertility in terms of clinical pregnancy rate.

The Cochrane Subfertility Review Group specialised register of controlled trials was searched.

Four RCTs with five treatment arms compared an ovulation suppression agent with placebo or no treatment. Eight trials were identified comparing a suppressive agent with danazol.

A diverse search strategy was employed, including hand-search of 43 core journals from 1966 to the present, bibliographies of relevant trials, MEDLINE database, abstracts from North American and European meetings and contact with authors of relevant papers. Relevant data were extracted independently by two reviewers using the standardised data extraction sheet. Validity was assessed in terms of method of randomisation, completeness of follow-up, presence or absence of crossover and co-intervention.

2x2 tables were generated for all relevant outcomes. Odds ratios were generated using the Peto modified Mantel-Haenszel technique. Statistical heterogeneity was assessed using x2.

The common odds ratio for pregnancy following ovulation suppression versus placebo or no treatment was 0.83 (95% CI 0.5-1.39). These data were statistically homogeneous although clinical heterogeneity was present. Their consistency in showing no treatment benefit suggests that this group of interventions is ineffective. Common odds ratio for pregnancy following all agents versus danazol was 1.20 (95% CI 0. 85-1.68). Again these data were homogeneous and suggest no significant treatment benefit in terms of pregnancy rate.

Given the significant period of amenorrhea associated with ovulation suppression, the lack of treatment benefit demonstrated and the adverse effects commonly associated with these treatments, ovulation suppression cannot be recommended as a standard therapy for endometriosis-associated infertility.