Progestagens and anti‐progestagens for pain associated with endometriosis

Julie Brown; Sari Kives; Muhammad Akhtar

doi:10.1002/14651858.CD002122.pub2

. 2012 Mar 14;2012(3):CD002122. doi: 10.1002/14651858.CD002122.pub2

Progestagens and anti‐progestagens for pain associated with endometriosis

Julie Brown ^1,^✉, Sari Kives ², Muhammad Akhtar ³

Editor: Cochrane Gynaecology and Fertility Group

PMCID: PMC6885053 PMID: 22419284

Abstract

Background

Endometriosis is a chronic inflammatory condition defined by the presence of glands and stroma outside the uterine cavity. It occurs in 7% to 10% of all women of reproductive age and may present as pain or infertility. The pelvic pain may be in the form of dysmenorrhoea, dyspareunia or pelvic pain. Initially a combination of estrogens and progestagens was used to create a pseudopregnancy and alleviate the symptoms associated with endometriosis. Progestagens alone or anti‐progestagens have been considered as alternatives because they are inexpensive and may have a better side effect profile than other choices.

Objectives

To determine the effectiveness of both the progestagens and anti‐progestagens in the treatment of painful symptoms ascribed to the diagnosis of endometriosis.

Search methods

We used the search strategy of the Menstrual Disorders and Subfertility Group to identify all publications which described or might have described randomised controlled trials (RCTs) of any progestagen or any anti‐progestagen in the treatment of symptomatic endometriosis. We updated the review in 2011.

Selection criteria

We considered only RCTs which compared the use of progestagens and anti‐progestagens with other interventions, placebo or no treatment for the alleviation of symptomatic endometriosis.

Data collection and analysis

We have added six new studies, bringing the total of included studies to 13 in the update of this review. The six newly included studies evaluated progestagens (comparisons with placebo, danazol, oral or subdermal contraceptive, oral contraceptive pill and danazol, gonadotrophin‐releasing hormone (GnRH) analogue and other drugs). The remaining studies compared the anti‐progestagen gestrinone with danazol, GnRH analogues or itself.

Main results

The progestagen medroxyprogesterone acetate (100 mg daily) appeared to be more effective at reducing all symptoms up to 12 months of follow‐up (MD ‐0.70, 95% CI ‐8.61 to ‐5.39; P < 0.00001) compared with placebo. There was evidence of significantly more cases of acne (six versus one) and oedema (11 versus one) in the medroxyprogesterone acetate group compared with placebo. There was no evidence of a difference in objective efficacy between dydrogesterone and placebo.

There was no evidence of a benefit with depot administration of progestagens versus other treatments (low dose oral contraceptive or leuprolide acetate) for reduced symptoms. The depot progestagen group experienced significantly more adverse effects.

There was no overall evidence of a benefit of oral progestagens over other medical treatment at six months of follow‐up for self‐reported efficacy. Amenorrhoea and bleeding were more frequently reported in the progestagen group compared with other treatment groups.

There was no evidence of a benefit of anti‐progestagens (gestrinone) compared with danazol. GnRH analogue (leuprorelin) was found to significantly improve dysmenorrhoea compared with gestrinone (MD 0.82, 95% CI 0.15 to 1.49; P = 0.02) although it was also associated with increased hot flushes (OR 0.20, 95% CI 0.06 to ‐0.63; P = 0.006).

Authors' conclusions

There is only limited evidence to support the use of progestagens and anti‐progestagens for pain associated with endometriosis.

Plain language summary

Progestagens and anti‐progestagens for pain associated with endometriosis

Endometriosis is a painful condition where tissue from the lining of the womb (uterus) is found outside the uterus as well. It can cause pain in the abdomen, generally and during periods (menstruation) or sex. Endometriosis can also lead to infertility. Treatments include surgery or drugs to try and shrink the tissue. Progestagens and anti‐progestagens are some of the hormonal drugs used for treatment. This systematic review of trials found limited evidence for the effectiveness of these drugs in the reduction of pain from endometriosis. This was due to the limited number of randomised controlled trials comparing each drug.

Summary of findings

Background

Description of the condition

Endometriosis presents either with the problem of infertility (Haney 1993; Prentice 1996) or with painful symptoms (Barlow 1993). The painful symptoms may take the form of dysmenorrhoea (painful periods), dyspareunia (pain during or after sexual intercourse) or pelvic or lower abdominal pain. Typically the pain precedes the onset of menses and lasts for the duration of the cycle. Less commonly patients also present with cyclical pain at other sites, relating to endometriosis at extra‐pelvic sites (Augoulea 2008; Lancaster 1995). Although the exact incidence of endometriosis is unknown, endometriosis is a significant problem for the affected individual and the cost of the disease is high both in human and economic terms (Mathias 1996).

Description of the intervention

The clinical observation of an apparent resolution of symptoms during pregnancy gave rise to the concept of treating patients with a pseudopregnancy regime (Kistner 1959). Initially combinations of high dose estrogens and progestagens were used, but this was subsequently replaced by progestagens alone (Kistner 1958). More recently anti‐progestagens have been developed and they have also been employed in the treatment of endometriosis (Thomas 1987a). The main side effects of progestagens include irregular menstrual cycles or cessation of menstruation, weight gain and breast tenderness. Cytoproterone acetate is associated with liver toxicity. The main side effects associated with anti‐progesterones include breakthrough bleeding, acne, fluid retention, weight gain and other androgenic symptoms.

How the intervention might work

The precise pathogenesis (mode of development) of endometriosis remains unclear, but it is evident that endometriosis arises by the dissemination of endometrium to ectopic sites (sites other than its normal location within the uterus) and the subsequent establishment of deposits of ectopic endometrium (Kruitwagen 1993; McLaren 1996). The assumption is made that these deposits of ectopic endometrium are responsible for the symptoms of endometriosis. Conventional treatments, therefore, are directed at the removal of all ectopic tissue. Surgical treatments achieve this by destroying or removing the implant, whilst medical therapies induce atrophy within the hormonally dependent ectopic endometrium so that they shrink in size and number.

Medical treatments theoretically have the ability to treat those implants not visible to the naked eye. Traditionally the oral contraceptive pill has been first line treatment for patients with presumed endometriosis (Davis 2007). Progestagens alone, however, can induce decidualisation (an adaption of the uterus to enable implantation of the embryo), atrophy of implants and resolution of symptoms. The progestagens result in the creation of a pseudopregnancy. The clinical observation of apparent resolution of symptoms of endometriosis during pregnancy gave rise to treatment with a medication containing a progestagen (Moghissi 1990). Gonadotrophin‐releasing hormone analogues and danazol are also used for the treatment of endometriosis but have a less favourable profile in terms of safety, tolerability and cost (Rodgers 2008). Anti‐progestagens are a substance that prevents cells from making or using progesterone. They may also be beneficial in treating endometriosis as they display anti‐proliferative effects in the endometrium but serum estradiol levels remain in the early to mid‐follicular phase range. For this reason they avoid the bone loss and hypoestrogenism associated with progestagens alone (Spitz 2003).

Why it is important to do this review

Progestagens are readily available, inexpensive and may have a better side effect profile than other choices (such as danazol), and antiprogestagens may have even fewer side effects. This review evaluates the role of both progestagens and anti‐progestagens in the treatment of symptomatic endometriosis.

Objectives

To determine the effectiveness and adverse effects of both progestagens and anti‐progestagens in the treatment of painful symptoms associated with endometriosis.

Methods

Criteria for considering studies for this review

Types of studies

We have included only randomised controlled trials (RCTs) which compared the use of progestagens and anti‐progestagens in the treatment of symptomatic endometriosis. We considered trials with placebo arms, no treatment arms and comparison to other medical therapies or surgical therapies, but have analysed these separately. We have not included quasi‐RCTs.

Types of participants

This review considered studies that included women of reproductive years with painful symptoms and a laparoscopic diagnosis of endometriosis.

We considered painful symptoms associated with endometriosis as follows: cyclical pain associated with menstruation (dysmenorrhoea), or not associated with menstruation; deep dyspareunia (pain during or following sexual intercourse); lower abdominal or pelvic pain of a non‐cyclical nature; pain on defecation; and any other painful symptom ascribed to endometriosis that was studied in any trial. We included all studies, whether the duration of symptoms was specified (three or six months) or not.

We excluded trials where participants had asymptomatic disease or infertility alone.

Types of interventions

We considered only those treatments where the aim was to achieve symptom control through disease resolution either medically or postoperatively regardless of dose, route of administration or duration of treatment.

Depomedroxyprogesterone acetate, cytoproterone acetate, medroxyprogesterone acetate, gestagen and dienogest were all evaluated in the literature as different progestagens for the treatment of endometriosis. Gestrinone was the only anti‐progestagen identified that has been evaluated for the treatment of endometriosis.

The comparisons considered were head‐to‐head drug comparisons, conservative surgery, non‐steroidal anti‐inflammatories, placebo or no treatment, oral contraceptive or danazol or GNRHa.

We did not consider any trial where the symptom relief was not documented, either through an objective or subjective measure, or if the surgical procedure was not conservative. We also excluded alternative or complementary therapies.

We have not included any trial where the progesterone intrauterine system was used as a treatment for endometriosis as a separate Cochrane Review addresses this question. Similarly we have not included any trial where danazol was used as a treatment as a separate Cochane Review addresses this.

Types of outcome measures

Primary outcomes

We considered relief or reduction of symptoms of endometriosis, measured either subjectively or objectively, for each pain symptom when possible. We considered outcomes measures at the end of the treatment and, when possible, at three, six, nine, 12 and 18 months following completion of treatment.

Subjective outcome: relief of any or all symptoms of endometriosis using quantitative measures such as visual analogue scales or qualitative measures such as cured, better, same, or worse.
Objective outcome: resolution of endometriotic implants assessed by either the revised American Fertility Society (AFS) score or implant score. Although this is neither a direct or indirect measure of pain, it is an independent assessment of disease resolution.

Secondary outcomes

We considered the occurrence of any adverse effects either during treatment or following treatment.

Search methods for identification of studies

Electronic searches

We utilised the search strategy of the Menstrual Disorders and Subfertility Group to identify all publications which described or might have described RCTs of any progestagen or anti‐progestagen in the treatment of symptomatic endometriosis (refer to Appendix 1). For a full outline of the Review Group search strategy see Review Group details.

In addition, we conducted electronic searches in the following electronic databases:

1) CENTRAL (Appendix 2) (to 23 August 2011);

2) MEDLINE (Appendix 3) (1950 to 23 August 2011);

3) EMBASE (Appendix 4) (1980 to 23 August 2011);

4) PsycINFO (Appendix 5) (1806 to 23 August 2011).

CINAHL was not searched in the 2011 update.

Searching other resources

We searched conference proceedings and reference lists of retrieved articles, and also contacted authors for additional information and data.

Data collection and analysis

Selection of studies

Two review authors (SK, JB) independently selected studies. Where uncertainty existed regarding suitability for inclusion, or discrepancy existed between the initial two authors, a third author made a further assessment. If required, we sought additional information from the principal or corresponding investigator of the trial.

Data extraction and management

The same two assessors extracted data; at least one of the assessors was an expert in the content matter. For data extraction, we used forms developed according to Cochrane guidelines. In some papers data were presented in graphical form. Where this was the case, we have approached the authors for clarification and, if necessary, extracted the data from the graphs.

Assessment of risk of bias in included studies

We assessed the quality of trials for inclusion using a standard risk of bias checklist (refer to Figure 1 and Figure 2). We collected the following information: the method of randomisation, allocation concealment, blinding, completeness of data and selective reporting. Both JB and SK extracted this information independently and resolved disagreements through consensus.

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Measures of treatment effect

We performed statistical analyses according to the statistical guidelines for review authors in the Menstrual Disorders and Subfertility Review Group. We used relative risk as the measure of effect for dichotomous data. For continuous data, we used weighted differences whenever outcomes were measured in a standard way across studies. However, as many different methods exist for assessing pain, we used standardised mean differences when comparing multiple methods. Although different methods give different absolute values, they are conceptually measuring the same parameter. We considered the different methods of measuring pain together, not subjected to separate subgroup analyses. Where there were sufficient data, we calculated a summary statistic for each outcome using a fixed‐effect model.

Unit of analysis issues

We presented data as per woman randomised and there were no anticipated concerns over unit of analysis issues between studies.

Dealing with missing data

We requested from the original authors any data that could not be analysed because they were in graph form or were missing. We planned a sensitivity analysis if significant data were missing.

Assessment of heterogeneity

We noted heterogeneity in the data and cautiously explored it using the previously identified characteristics of the studies, particularly assessments of quality. We undertook sensitivity analyses to examine the viability of the results in relation to a number of factors including study quality and the source of the data (published or unpublished). See the Review Group module details for more information. We determined statistical heterogeneity using the I² statistic.

Assessment of reporting biases

There were insufficient studies to determine the existence of publication bias via a funnel plot. However, the review authors have attempted to obtain data from unpublished as well as published sources.

Data synthesis

We carried out meta‐analysis using a fixed‐effect model.

Subgroup analysis and investigation of heterogeneity

Data for pain associated with endometriosis are often presented as an overall score and then subgrouped according to pelvic pain, dysmenorrhoea and dyspareunia. Data were also subgrouped according to whether pain was objectively or subjectively determined.

Sensitivity analysis

Where heterogeneity was more than 50%, we considered sensitivity analysis based on the quality of the individual trials to attempt to explain it.

Results

Description of studies

Results of the search

We identified 26 studies as possibilities for inclusion in the systematic review. We identified nine additional studies in this 2011 Cochrane Review update.

Included studies

We have included a total of 13 studies in this 2011 Cochrane Review update (Bergvist 2001; Bromham 1995; Fedele 1989; GISG 1996; Harada 2009; Hornstein 1990; Overton 1994; Razzi 2007; Schlaff 2006; Strowitzki 2010; Telimaa 1987b; Vercellini 1996; Vercellini 2002). There were seven studies in the last published version from 2000.

Progestagens

We identified eight RCTs that considered the role of progestagens alone in the treatment of endometriosis (Bergvist 2001; Harada 2009; Overton 1994; Razzi 2007; Schlaff 2006; Strowitzki 2010; Vercellini 1996; Vercellini 2002). We identified other studies but excluded them because many participants had received operative treatment at the time of study entry, the drug formulation was unknown, or the patients studied were too specific (see Characteristics of excluded studies). We included one study although a small percentage of patients had received operative treatment at the time of diagnostic laparoscopy (Telimaa 1987b).

Overton 1994 considered three groups of women with endometriosis who wished to achieve pregnancy but also complained of pain. Patients were randomised to two doses of dydrogesterone (40 mg or 60 mg) once daily or placebo. The endpoints of this study that were relevant to this review were the reduction in pain scores (derived from diary cards) and the reduction in the AFS score at second look laparoscopy (performed within three months of completing treatment). Only 39 out of 62 women completed the study and underwent a second look laparoscopy.

Vercellini 1996 compared 150 mg of depot medroxyprogesterone every three months with a 20 µg oral contraceptive pill (OCP) with 50 mg danazol. Both the pill and danazol were taken for three weeks out of four. The primary endpoint was the degree of satisfaction at the end of therapy. A change in severity of symptoms was also measured using a 10 cm visual analogue score and a 0 to 3 point verbal rating scale.

Vercellini 2002 similarly compared 12.5 mg cytoproterone acetate once daily versus a continuous monophasic OCP once daily (0.02 µg ethinyl estradiol and 0.15 mg desogestrel). The primary endpoint, as in their previous study, was the degree of satisfaction at the end of therapy. A change in severity of symptoms was also measured by a 100 mm visual analogue score and a 0 to 3 point verbal rating scale.

Bergvist 2001 compared 15 mg medroxyprogesterone twice daily versus 200 µg nafarelin intranasally twice daily. Each group also received a placebo nasal spray or placebo tablets. In this way each group took the same number of tablets daily as the those in the active medroxyprogesterone group and the same number of nasal sprays as those in the active nafarelin group. The endpoint relevant to this review was the endometriosis severity score. Of the 48 who participated only 30 competed the study.

Schlaff 2006 compared 104 mg subcutaneous depot medroxyprogesterone every three months versus 11.25 mg leuprolide intramuscularly (IM) every three months. The primary endpoint was the reduction in five endometriosis symptoms or signs. Of the 274 participants only 190 completed the six months of active treatment.

Razzi 2007 compared desogestrel 75 μg daily with ethinylestridiol plus desogestrel daily for six months in 40 women with Stage I to III endometriosis. The primary endpoint was self‐reported pain using a visual analogue scale.

Strowitzki 2010 compared 2 mg of dienogest daily with leuprolide acetate 3.75 mg depot (IM four weekly) for six months in 252 women with Stage I to IV endometriosis. The primary endpoint was self‐reported pain using a visual analogue scale.

Harada 2009 compared 2 mg of dienogest daily with 300 μg of buserelin acetate (intranasally) daily in 271 women with confirmed endometriosis. The primary endpoint was self‐reported pain.

Telimaa 1987b compared three groups of participants with mild to moderate endometriosis. They were randomised to either 100 mg medroxprogesterone once daily, 200 mg danazol three times daily or placebo for six months. Participants received identical packets of tablets so that each group took the same number of tablets daily as the active medroxyprogesterone or active danazol group. Twenty‐seven per cent of participants did receive a surgical co‐intervention at the study entry point but as they were evenly distributed in all three groups they were still included in the review. Change in the American Fertility Score and four‐point verbal pain scores at the end of treatment were the relevant endpoint.

No other studies comparing progestagens with surgical therapy were identified.

Anti‐progestagens

We identified no placebo controlled trial or no therapy trials comparing the anti‐progestagen gestrinone. In addition, we identified no studies comparing gestrinone to any progestagen.

We identified two studies that compared gestrinone with danazol (Bromham 1995; Fedele 1989). Fedele 1989 reported on 39 infertile women with laparoscopically confirmed endometriosis. Patients received either 2.5 mg gestrinone twice weekly or 600 gm danazol per day. If amenorrhoea was not achieved, danazol was increased to 800 mg per day and gestrinone was increased to three times per week. The prevalence of pain symptoms as well as the change in the American Fertility Score at laparoscopy following treatment were considered the relevant endpoints. Bromham 1995 was a larger study, comparing 269 women who received either 2.5 mg gestrinone twice weekly or 200 mg danazol twice daily. American Fertility Scores at the laparoscopy following treatment and pain scores during treatment were similar endpoints. In this study 69 women withdrew during treatment.

We identified one multi‐centre study (GISG 1996) comparing 2.5 mg gestrinone twice weekly with 3.75 mg leuprolin depot intramuscular monthly. Both groups also received a placebo pill or injection depending on their allocation. A change in severity of symptoms was measured by a 100 mm visual analogue score and a 0 to 3 point verbal rating scale.

One small study (Hornstein 1990) compared two doses of gestrinone, 1.25 mg versus 2.5 mg twice weekly. A total of six participants in each arm were assessed for a change in the Revised American Fertility Society Score of endometriosis as well as symptom scores.

Excluded studies

We excluded 13 studies from this Cochrane Review update (Cosson 2002; Dawood 1997; Harrison 2000; Mettler 1987; Nieto 1996; Noble 1980; Regidor 2001; Telimaa 1987a; Thomas 1987a; Vercellini 2005; Walch 2009; Worthington 1993; Yang 2006).

Risk of bias in included studies

Refer to the 'Risk of bias' tables and Figure 1 and Figure 2.

Allocation

Eight studies (Bergvist 2001; Bromham 1995; GISG 1996; Harada 2009; Hornstein 1990; Overton 1994; Schlaff 2006; Strowitzki 2010) included allocation concealment in their study design. In four studies (Razzi 2007; Telimaa 1987b; Vercellini 1996; Vercellini 2002) it was unclear whether allocation concealment was performed. The remaining trial failed to demonstrate allocation concealment in the study design (Fedele 1989).

Blinding

Six studies (Bergvist 2001; Bromham 1995; GISG 1996; Harada 2009; Schlaff 2006) included blinding in their study design. In four studies (Razzi 2007; Telimaa 1987b; Vercellini 1996; Vercellini 2002) it was unclear if blinding occurred. The remaining trials did not use blinding in their study design (Fedele 1989; Strowitzki 2010).

Incomplete outcome data

In three trials (Bergvist 2001; Schlaff 2006; Vercellini 2002) it was not possible to analyse the outcome data as they were in graphic or tabular form only. In the prior Cochrane review, the authors had successfully reported (with the exception of Vercellini 1996) all outcome data by contacting the appropriate authors.

Three studies reported large losses to follow‐up. In Bromham 1995, 124 out of 265 did not complete the trial: five conceived before treatment, 69 withdrew during treatment and 50 were lost during the 12 months of follow‐up. Similarly in Overton 1994, five patients were excluded post‐randomisation (four conceived) and 23 were lost to follow‐up out of a total of 62 patients. Finally, in Schlaff 2006 84 out of 247 did not complete the trial. Losses were equally distributed between groups in all three studies. All patients appear to have been followed up in the trial conducted by Razzi 2007 and the remaining trials provided numbers and reasons for losses (Harada 2009; Strowitzki 2010).

Selective reporting

All of the studies reported on a priori outcomes which had been stated in the methods section of the studies with the exception of Strowitzki 2010 who did not report on individual symptoms for the Biberglu and Behrman scores, which had been reported as an outcome in the trial methodology. The original protocols for each study were not accessed.

Other potential sources of bias

The authors are not aware of any other sources of bias.

Effects of interventions

See: Table 1; Table 2; Table 3; Table 4

Summary of findings for the main comparison. Progestagen compared to placebo for pain associated with endometriosis.

Progestagen compared to placebo for pain associated with endometriosis
Patient or population: patients with pain associated with endometriosis   Settings: gynaecology clinics   Intervention: progestagen   Comparison: placebo
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Assumed risk	Corresponding risk
	Placebo	Progestagen
AFS score	The mean AFS score in the control groups was   1.76	The mean AFS score in the intervention groups was   0.58 lower   (1.41 lower to 0.25 higher)		33   (1 study)	⊕⊕⊝⊝   low^1,2
Patient assessed efficacy, sum of all symptoms   Follow‐up: mean 6 months	The mean patient assessed efficacy, sum of all symptoms in the control groups was   ‐5.20	The mean patient assessed efficacy, sum of all symptoms in the intervention groups was   5.2 lower   (6.8 to 3.6 lower)		33   (1 study)	⊕⊕⊝⊝   low^1,2
Patient assessed efficacy, sum of all symptoms   Follow‐up: mean 12 months	The mean patient assessed efficacy, sum of all symptoms in the control groups was   ‐7.0	The mean patient assessed efficacy, sum of all symptoms in the intervention groups was   7 lower   (8.61 to 5.39 lower)		29   (1 study)	⊕⊕⊝⊝   low^1,2
Side effects ‐ acne	59 per 1000	375 per 1000   (59 to 852)	OR 9.6   (1 to 91.96)	33   (1 study)	⊕⊝⊝⊝   very low^1,2,3
Side effects ‐ oedema	59 per 1000	688 per 1000   (184 to 956)	OR 35.2   (3.6 to 344.19)	33   (1 study)	⊕⊝⊝⊝   very low^1,2,4
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).   CI: Confidence interval; OR: Odds ratio;
GRADE Working Group grades of evidence   High quality: further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: we are very uncertain about the estimate.

Depot progestagen compared to other treatment for pain associated with endometriosis
Patient or population: patients with pain associated with endometriosis   Settings: gynaecology clinics   Intervention: depot progestagen   Comparison: other treatment
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Assumed risk	Corresponding risk
	Other treatment	Depot progestagen
Improvement in dysmenorrhoea   Follow‐up: mean 6 months	978 per 1000	895 per 1000   (692 to 969)	OR 0.19   (0.05 to 0.69)	274   (1 study)	⊕⊕⊝⊝   low^1,2
Improvement in dysmenorrhoea   Follow‐up: mean 12 months	768 per 1000	676 per 1000   (551 to 782)	OR 0.63   (0.37 to 1.08)	274   (1 study)	⊕⊕⊝⊝   low^2,3
Side effects ‐ hot flushes	90 per 1000	29 per 1000   (11 to 76)	OR 0.3   (0.11 to 0.83)	354   (2 studies)	⊕⊕⊝⊝   low^4,5
Side effects ‐ amenorrhoea	0 per 1000	0 per 1000   (0 to 0)	OR 21.18   (1.18 to 380.9)	80   (1 study)	⊕⊝⊝⊝   very low^1,2,6
Side effects ‐ breakthrough bleeding/spotting	28 per 1000	373 per 1000   (157 to 655)	OR 20.56   (6.44 to 65.56)	354   (2 studies)	⊕⊕⊝⊝   low^1,4
Side effects ‐ bloating	275 per 1000	625 per 1000   (393 to 811)	OR 4.39   (1.71 to 11.3)	80   (1 study)	⊕⊝⊝⊝   very low^1,2,6
Side effects ‐ injection site reaction	0 per 1000	0 per 1000   (0 to 0)	OR 20.64   (1.19 to 358.23)	274   (1 study)	⊕⊕⊝⊝   low^1,2
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).   CI: Confidence interval; OR: Odds ratio;
GRADE Working Group grades of evidence   High quality: further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: we are very uncertain about the estimate.

Oral progestagens versus other treatment for pain associated with endometriosis
Patient or population: patients with pain associated with endometriosis   Settings:   Intervention: oral progestagens versus other treatment
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of Participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Assumed risk	Corresponding risk
	Control	Oral progestagens versus other treatment
Patient assessed efficacy ‐ pain   Follow‐up: mean 6 months	The mean patient assessed efficacy ‐ pain in the control groups was   21.1	The mean patient assessed efficacy ‐ pain in the intervention groups was   0.1 higher   (0.26 lower to 0.46 higher)		286   (2 studies)	⊕⊕⊕⊝   moderate¹
Objective efficacy at end of follow up (12 months) ‐ AFS score	The mean objective efficacy at end of follow up (12 months) ‐ AFS score in the control groups was   1.31	The mean objective efficacy at end of follow up (12 months) ‐ AFS score in the intervention groups was   0.34 higher   (0.01 lower to 0.7 higher)		302   (2 studies)	⊕⊕⊕⊝   moderate²
Side effects ‐ headache	244 per 1000	158 per 1000   (109 to 220)	OR 0.58   (0.38 to 0.87)	613   (3 studies)	⊕⊕⊕⊝   moderate¹
Side effects ‐ hot flushes	306 per 1000	178 per 1000   (120 to 251)	OR 0.49   (0.31 to 0.76)	613   (3 studies)	⊕⊕⊝⊝   low^1,3
Side effects ‐ genital bleeding	634 per 1000	891 per 1000   (811 to 939)	OR 4.69   (2.47 to 8.9)	271   (1 study)	⊕⊕⊝⊝   low^4,5
Side effects ‐ amenorrhoea	387 per 1000	758 per 1000   (645 to 843)	OR 4.95   (2.88 to 8.52)	252   (1 study)	⊕⊝⊝⊝   very low^4,5,6
Sleep disorder	78 per 1000	16 per 1000   (3 to 71)	OR 0.19   (0.04 to 0.90)	252   (1 study)	⊕⊝⊝⊝   very low^4,5,6
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).   CI: Confidence interval; OR: Odds ratio;
GRADE Working Group grades of evidence   High quality: further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: we are very uncertain about the estimate.

Anti‐progestagen compared to other treatment for pain associated with endometriosis
Patient or population: patients with pain associated with endometriosis   Settings: gynaecology clinics   Intervention: anti‐progestagen   Comparison: other treatment
Outcomes	*Illustrative comparative risks (95% CI)**		Relative effect   (95% CI)	No of participants   (studies)	Quality of the evidence   (GRADE)	Comments
	Assumed risk	Corresponding risk
	other treatment	Anti‐progestagen
Patient assessed efficacy none or mild painful periods (dysmenorrhoea)   Follow‐up: mean 6 months	667 per 1000	673 per 1000   (524 to 794)	OR 1.03   (0.55 to 1.93)	176   (2 studies)	⊕⊕⊝⊝   low¹
Objective assessment of efficacy at end of treatment (6 months) ‐ rAFS scores	The mean objective assessment of efficacy at end of treatment (6 months) ‐ rAFS scores in the control groups was   11.8	The mean objective assessment of efficacy at end of treatment (6 months) ‐ rAFS scores in the intervention groups was   1.4 higher   (6.76 lower to 9.56 higher)		16   (1 study)	⊕⊝⊝⊝   very low^2,3,4
Patient assessed efficacy painful periods   visual analogue scale   Follow‐up: mean 12 months	The mean patient assessed efficacy painful periods in the control groups was   4.76	The mean patient assessed efficacy painful periods in the intervention groups was   3 lower   (4.79 to 1.21 lower)		55   (1 study)	⊕⊕⊕⊝   moderate⁴
Side effects ‐ seborrhoea	204 per 1000	413 per 1000   (303 to 534)	OR 2.74   (1.69 to 4.46)	357   (3 studies)	⊕⊕⊝⊝   low¹
Side effects ‐ hirsutism	248 per 1000	465 per 1000   (346 to 588)	OR 2.63   (1.6 to 4.32)	302   (2 studies)	⊕⊝⊝⊝   very low^1,5
Side effects ‐ hot flushes	464 per 1000	360 per 1000   (267 to 462)	OR 0.65   (0.42 to 0.99)	357   (3 studies)	⊕⊝⊝⊝   very low^1,6
Side effects ‐ amenorrhoea	962 per 1000	500 per 1000   (200 to 905)	OR 0.04   (0.01 to 0.38)	49   (1 study)	⊕⊕⊝⊝   low^3,4
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).   CI: Confidence interval; OR: Odds ratio;
GRADE Working Group grades of evidence   High quality: further research is very unlikely to change our confidence in the estimate of effect.   Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.   Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.   Very low quality: we are very uncertain about the estimate.

Patient assessed efficacy during and at end of treatment (6 and 12 months)
Study	Heading 1	Heading 2	Heading 3	Heading 4	Heading 5
painful periods, visual analogue scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 7 (5‐10) and 6.5 (5.1‐8.2) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐3) and 2 (0.5‐3.3) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 0.5 (0‐1.5) respectively
painful periods, verbal rating scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 2 (1‐3) and 2 (1‐3) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 1 (0‐1) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 0 (0‐0) respectively
pain on intercourse, visual analogue scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 4 (0‐8) and 3.5 (0‐8.1) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐2.7) and 0 (0‐3.2) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 0 (0‐0.5) respectively
pain on intercourse, verbal rating scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 1 (0‐2) and 1 (0‐2) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐1) and 0 (0‐1) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 0 (0‐0) respectively
non‐menstrual pain, visual analogue scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 4 (0‐7.5) and 4.1 (1‐7.3) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0.2 (0‐3) and 0 (0‐2) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐1) and 0 (0‐0.5) respectively
non‐menstrual pain, verbal rating scale
Vercellini 1996	Baseline values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 1 (0‐2) and 1 (0‐2) respectively	Month six values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐1) and 0 (0‐0.1) respectively	Month twelve values (range) for depot medroxyprogesterone acetate and oral contraceptive plus danazol were 0 (0‐0) and 0 (0‐0) respectively

Pain symptom scores
Study	6 months	Cyproterone acetate Visual analogue scale	Cyproterone acetate Verbal rating scale	Oral contraceptiveVisual analogue scale	Oral contraceptiveVerbal scale
Vercellini 2002	Dysmenorrhoea Median (IQR)	0 (0 ‐ 0) n=39	2 (1 ‐ 2) n=39	74 (59 ‐ 83) n=36
Vercellini 2002	Deep dyspareunia Median (IQR)	13 (10 ‐ 30) n= 23	0 (0 ‐ 1) n=23	15 (0 ‐ 20) n = 25	0 (0 ‐ 1 ) n = 25
Vercellini 2002	Non Menstrual pain Median (IQR)	14 (0 ‐ 40) n = 22	0 (0 ‐ 1) n = 22	20 (0 ‐ 30) n = 20	0 (0 ‐ 1) n = 20

Self reported pain
Study
Razzi 2007	Both desogestrel and the oral contraceptive showed significant decreases in self reported pain compared to baseline P <0.001. After 6 months the mean VAS score for desogestrel alone was 2.5 and for the oral contraceptive was 2.3. There was no statistical comparison between groups calculated.

Side effects
Study
Razzi 2007	The authors report on breakthrough bleeding in 4/20 patients randomised to desogestrel and increased body weight in 3/20 randomised to oral contraceptive. no other details provided

Date	Event	Description
16 February 2012	New citation required but conclusions have not changed	Included studies have not led to change in conclusions
29 August 2011	New search has been performed	We have included six new studies in this update (Bergvist 2001; Harada 2009; Razzi 2007; Schlaff 2006; Strowitzki 2010; Vercellini 2002)

Date	Event	Description
7 November 2008	Amended	Converted to new review format.
17 January 2000	New citation required and conclusions have changed	Substantive amendment

Outcome or subgroup title	No. of studies	Statistical method	Effect size
1 AFS score	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2 AFS score (improved or remission)	1	Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
3 Patient assessed efficacy, 4 point verbal rating scale at end of treatment (6 months)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 pelvic pain	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 sum of all symptoms	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Patient assessed efficacy, 4 point verbal rating scale at end of follow‐up (12 months)	1	Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 pelvic pain	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 sum of all symptoms	1	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Change in pain score at 12 months follow‐up (Improvement)	1	Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6 Side effects	1	Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
6.1 acne	1	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.2 oedema	1	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.3 muscle cramps	1	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
6.4 spotting	1	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]

Methods	Randomised double blind, multi‐centre trial
Participants	Japan (24 centres) N = 271 Inclusion: 20 years or older, regular menstrual cycles, endometriosis diagnosed by laparotomy, laparoscopy or imaging analysis, the presence of subjective symptoms during menstruation, the presence of subjective symptoms during non‐menstruation, presence of objective findings Exclusion: undiagnosed genital bleeding, class 3 or more on Pap test within 3 months before enrolment, use of GnRH agonists, testosterone derivatives, hormonal therapy with progesterone and/or oestrogen, oestrogen antagonists, or aromatase inhibitors within 16 weeks before enrolment. Pregnant or nursing, history of severe adverse reaction or hypersensitivity to steroid hormone or GnRH agonists, past use of GnRH agonists with low BMD, having undergone surgery therapy or surgical examination for endometriosis within a menstrual cycle before the start of medication, use of drugs that could be expected to affect the release of sex hormones, a history or complication of thrombosis/embolism or depression, malignant tumour complication or findings suggestive of malignancy, complication of serious heart, liver, kidney, blood or endocrine disease, participating in another clinical trial in previous 4 months, deemed to be unsuitable
Interventions	Treated for 24 weeks with 2 mg dienogest daily PO (n = 137) versus 300 µg buserelin acetate IN TDS (n = 134)
Outcomes	Self‐reported pain, QoL, BMD, adverse events
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation by permuted block
Allocation concealment (selection bias)	Low risk	' allocation sequence...was kept centrally...'
Blinding (performance bias and detection bias)   All outcomes	Low risk	Double blind, patients were blinded using a double dummy
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Numbers and reasons for withdrawals given in paper
Selective reporting (reporting bias)	Low risk	A priori outcomes reported as per methods section. Protocol not accessed

Methods	Double blind randomised multi‐centre study   Randomisation method stated   Pharmaceutical company stated
Participants	62 British women aged 21‐42 years   Inclusion criteria: minimal ‐ mild endometriosis (AFS classification score 1‐15, stage 1 or 2) diagnosed at laparoscopy within preceding 3 months, women with azoospermic partners who had had more than 12 cycles of unsuccessful donor insemination, women taking clomiphene citrate or cyclofenil for ovulation induction also included   Exclusion criteria: women taking corticosteroids, hormones, danazol, or GnRH agonists in month before admission to the study
Interventions	1. 40 mg dydrogesterone for 12 days starting 2 days after LH surge   2. 60 mg dydrogesterone given as above   3. Placebo given as above.   Duration of treatment: 6 months
Outcomes	Conception rates   Change in AFS scores at laparoscopy following treatment   Pain scores   Bleeding
Notes	Follow‐up: 12 months   Second laparoscopy within 3 months of completing treatment   Exclusions post randomisation: 5 never treated, 1 refused, 4 conceived   Losses to follow‐up: 23
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	'women were allocated randomly' 'using computer generated randomization lists'
Allocation concealment (selection bias)	Unclear risk	No details in paper
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	Double blind, no details in paper
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Exclusions: 5 never treated, 1 refused, 13 conceived, 5 had unwanted side effects, 5 withdrew for miscellaneous/social reasons
Selective reporting (reporting bias)	High risk	No details of live birth

Methods	Multi‐centre, open label, randomised trial
Participants	Germany, Poland, Portugal, Spain and Austria (17 centres) n = 252 Inclusion: women aged 18‐45 years, experiencing pain with histologically confirmed endometriosis stage I‐IV. Laparoscopic diagnosis Exclusion: pregnancy or breast feeding, amenorrhoea within 3 months of screening, a primary need for surgical treatment, previous use of hormonal agents (GnRH agonists ≤, progestins/danazol ≤ 3 months or oral contraceptives ≤ 1 month), abnormal gynaecological examination or smear test result or risk factors for decreased bone mineral density
Interventions	Dienogest 2 mg daily PO (n = 124) versus leuprolide acetate 3.75 mg depot IM every 4 weeks (n = 128) Treatment for 24 weeks
Outcomes	Absolute change in pelvic pain using VAS (0‐100); improvement in pain (VAS); responder rates, Biberoglu & Behrman (B&B) scores; QoL; adverse effects, BMD
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	'randomization blocks'
Allocation concealment (selection bias)	Low risk	Randomisation done centrally
Blinding (performance bias and detection bias)   All outcomes	High risk	'open label'
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Drops outs recorded and reasons given in text
Selective reporting (reporting bias)	High risk	Did not show data on individual symptoms for B&B scores

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Patient assessed efficacy during and at end of treatment (6 and 12 months)			Other data	No numeric data
1.1 painful periods, visual analogue scale			Other data	No numeric data
1.2 painful periods, verbal rating scale			Other data	No numeric data
1.3 pain on intercourse, visual analogue scale			Other data	No numeric data
1.4 pain on intercourse, verbal rating scale			Other data	No numeric data
1.5 non‐menstrual pain, visual analogue scale			Other data	No numeric data
1.6 non‐menstrual pain, verbal rating scale			Other data	No numeric data
2 Improvement in symptoms	1		Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
2.1 dysmenorrhoea 6 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 dysmenorrhoea 12 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.3 dyspareunia 6 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.4 dyspareunia 12 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.5 pelvic pain 6 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.6 pelvic pain 12 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.7 pelvic tenderness 6 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.8 pelvic tenderness 12 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.9 induration 6 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.10 induration 12 months	1		Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Side effects	2		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
3.1 acne/greasy skin (seborrhoea)	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	4.75 [0.94, 23.98]
3.2 hot flushes	2	354	Odds Ratio (M‐H, Fixed, 95% CI)	0.30 [0.11, 0.83]
3.3 breast pain/tension	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	1.24 [0.34, 4.43]
3.4 headaches	2	354	Odds Ratio (M‐H, Fixed, 95% CI)	0.91 [0.48, 1.73]
3.5 dizziness	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	3.08 [0.12, 77.80]
3.6 nausea	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	3.86 [1.12, 13.26]
3.7 weight gain	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	2.58 [1.03, 6.46]
3.8 amenorrhoea	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	21.18 [1.18, 380.90]
3.9 breakthrough bleeding/spotting	2	354	Odds Ratio (M‐H, Fixed, 95% CI)	20.56 [6.44, 65.56]
3.10 bloating	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	4.39 [1.71, 11.30]
3.11 depression	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	1.18 [0.38, 3.63]
3.12 asthenia	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	1.0 [0.06, 16.56]
3.13 peripheral oedema	1	80	Odds Ratio (M‐H, Fixed, 95% CI)	2.05 [0.18, 23.59]
3.14 injection site reaction	1	274	Odds Ratio (M‐H, Fixed, 95% CI)	20.64 [1.19, 358.23]
3.15 insomnia	1	274	Odds Ratio (M‐H, Fixed, 95% CI)	0.42 [0.11, 1.67]
3.16 decreased libido	1	274	Odds Ratio (M‐H, Fixed, 95% CI)	0.42 [0.11, 1.67]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Patient assessed efficacy (6 months)	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
1.1 pain	2	286	Mean Difference (IV, Fixed, 95% CI)	0.10 [‐0.26, 0.46]
1.2 sum of all symptoms	1	34	Mean Difference (IV, Fixed, 95% CI)	0.50 [‐1.10, 2.10]
2 Patient assessed efficacy, 4 point verbal rating scale at end of follow‐up (12 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
2.1 pelvic pain	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
2.2 sum of all symptoms	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Objective efficacy at end of follow‐up (12 months)	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
3.1 AFS score	2	302	Mean Difference (IV, Fixed, 95% CI)	0.34 [‐0.01, 0.70]
4 Improved VAS score	1		Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
5 Quality of life	2		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 physical health summary scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 mental health summary scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.3 bodily pain	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Severe/very severe signs and symptoms (24 weeks)	1		Odds Ratio (M‐H, Fixed, 95% CI)	Totals not selected
7 Change in pain from baseline to 24 weeks	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
7.1 abdominal pain	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
7.2 lumbago	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
8 Pain symptom scores			Other data	No numeric data
9 Self reported pain			Other data	No numeric data
10 Side effects			Other data	No numeric data
11 Side effects	4		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
11.1 acne	2	286	Odds Ratio (M‐H, Fixed, 95% CI)	0.60 [0.24, 1.49]
11.2 oedema	1	34	Odds Ratio (M‐H, Fixed, 95% CI)	2.75 [0.67, 11.24]
11.3 muscle cramps	1	34	Odds Ratio (M‐H, Fixed, 95% CI)	0.46 [0.09, 2.27]
11.4 spotting	2	124	Odds Ratio (M‐H, Fixed, 95% CI)	0.73 [0.35, 1.54]
11.5 headache	3	613	Odds Ratio (M‐H, Fixed, 95% CI)	0.58 [0.38, 0.87]
11.6 weight gain	2	342	Odds Ratio (M‐H, Fixed, 95% CI)	1.09 [0.51, 2.33]
11.7 depression	2	342	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.37, 1.97]
11.8 decreased libido	2	342	Odds Ratio (M‐H, Fixed, 95% CI)	1.24 [0.52, 2.94]
11.9 hair loss	1	252	Odds Ratio (M‐H, Fixed, 95% CI)	0.58 [0.16, 2.02]
11.10 migraine	1	252	Odds Ratio (M‐H, Fixed, 95% CI)	0.50 [0.12, 2.06]
11.11 sleep disorder	1	252	Odds Ratio (M‐H, Fixed, 95% CI)	0.19 [0.04, 0.90]
11.12 vaginal dryness	2	342	Odds Ratio (M‐H, Fixed, 95% CI)	0.47 [0.15, 1.48]
11.13 hot flushes	3	613	Odds Ratio (M‐H, Fixed, 95% CI)	0.49 [0.31, 0.76]
11.14 study withdrawal due to side effects	1	252	Odds Ratio (M‐H, Fixed, 95% CI)	1.25 [0.37, 4.21]
11.15 genital bleeding	1	271	Odds Ratio (M‐H, Fixed, 95% CI)	4.69 [2.47, 8.90]
11.16 amenorrhoea	1	252	Odds Ratio (M‐H, Fixed, 95% CI)	4.95 [2.88, 8.52]
11.17 bloating or swelling	1	90	Odds Ratio (M‐H, Fixed, 95% CI)	0.90 [0.37, 2.19]
11.18 irritability	1	90	Odds Ratio (M‐H, Fixed, 95% CI)	3.14 [0.31, 31.42]
11.19 nausea	1	90	Odds Ratio (M‐H, Fixed, 95% CI)	0.10 [0.01, 1.94]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Patient assessed efficacy at end of treatment (6 months)	2		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
1.1 none or mild pelvic pain	2	230	Odds Ratio (M‐H, Fixed, 95% CI)	0.71 [0.33, 1.56]
1.2 none or mild painful periods (dysmenorrhoea)	2	214	Odds Ratio (M‐H, Fixed, 95% CI)	0.72 [0.39, 1.33]
1.3 none or mild pain on intercourse (dyspareunia)	2	222	Odds Ratio (M‐H, Fixed, 95% CI)	0.83 [0.37, 1.86]
2 Patient assessed efficacy 6 months after the end of treatment	2		Peto Odds Ratio (Peto, Fixed, 95% CI)	Subtotals only
2.1 none or mild pelvic pain	2	202	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.20 [0.58, 2.48]
2.2 none or mild painful periods (dysmenorrhoea)	2	176	Peto Odds Ratio (Peto, Fixed, 95% CI)	1.03 [0.55, 1.93]
2.3 none or mild pain on intercourse (dyspareunia)	2	192	Peto Odds Ratio (Peto, Fixed, 95% CI)	0.94 [0.42, 2.09]
3 Objective assessment of efficacy at end of treatment (6 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
3.1 rAFS scores	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
3.2 implant score	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4 Patient assessed efficacy at end of treatment (6 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
4.1 painful periods, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.2 painful periods, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.3 pain on intercourse, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.4 pain on intercourse, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.5 non‐menstrual pain, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
4.6 non‐menstrual pain, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Patient assessed efficacy at end of follow‐up (12 months)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
5.1 painful periods, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.2 painful periods, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.3 pain on intercourse, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.4 pain on intercourse, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.5 non‐menstrual pain, visual analogue scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
5.6 non‐menstrual pain, verbal rating scale	1		Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
6 Side effects	3		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
6.1 acne	2	302	Odds Ratio (M‐H, Fixed, 95% CI)	1.45 [0.90, 2.33]
6.2 seborrhoea	3	357	Odds Ratio (M‐H, Fixed, 95% CI)	2.74 [1.69, 4.46]
6.3 hirsutism	2	302	Odds Ratio (M‐H, Fixed, 95% CI)	2.63 [1.60, 4.32]
6.4 voice problems	2	302	Odds Ratio (M‐H, Fixed, 95% CI)	0.70 [0.34, 1.43]
6.5 swelling hands/feet	2	319	Odds Ratio (M‐H, Fixed, 95% CI)	1.48 [0.88, 2.48]
6.6 hot flushes	3	357	Odds Ratio (M‐H, Fixed, 95% CI)	0.65 [0.42, 0.99]
6.7 sweating problems	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	1.44 [0.88, 2.35]
6.8 loss of libido	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	1.32 [0.80, 2.19]
6.9 decreased breast size	2	302	Odds Ratio (M‐H, Fixed, 95% CI)	0.62 [0.38, 0.98]
6.10 leg or muscle cramps	3	357	Odds Ratio (M‐H, Fixed, 95% CI)	0.48 [0.30, 0.77]
6.11 headaches	2	319	Odds Ratio (M‐H, Fixed, 95% CI)	0.99 [0.64, 1.53]
6.12 nausea	3	357	Odds Ratio (M‐H, Fixed, 95% CI)	1.35 [0.84, 2.16]
6.13 vomiting	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	0.67 [0.32, 1.43]
6.14 loss of appetite	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	1.31 [0.72, 2.37]
6.15 hunger	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	0.59 [0.36, 0.97]
6.16 dizziness	2	319	Odds Ratio (M‐H, Fixed, 95% CI)	1.27 [0.77, 2.08]
6.17 tiredness	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	1.44 [0.84, 2.45]
6.18 faintness	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	1.23 [0.54, 2.76]
6.19 skin rash	2	319	Odds Ratio (M‐H, Fixed, 95% CI)	1.76 [0.95, 3.24]
6.20 weight gain	1	38	Odds Ratio (M‐H, Fixed, 95% CI)	0.34 [0.09, 1.27]
6.21 vaginal dryness	2	93	Odds Ratio (M‐H, Fixed, 95% CI)	0.19 [0.02, 1.66]
6.22 raised liver transaminases	1	38	Odds Ratio (M‐H, Fixed, 95% CI)	0.18 [0.01, 4.00]
6.23 stopped treatment because of side effects	1	264	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.47, 1.57]
6.24 asthenia	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.8 [0.19, 3.36]
6.25 mood changes	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.67 [0.10, 4.34]
6.26 dermatitis	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	8.14 [0.40, 165.53]
6.27 joint pain	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	2.16 [0.18, 25.32]
6.28 drowsiness	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	2.16 [0.18, 25.32]
6.29 tachycardia	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	1.04 [0.06, 17.49]
6.30 insomnia	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	3.23 [0.13, 82.71]
6.31 hypertrichosis	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	3.23 [0.13, 82.71]
6.32 constipation	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	3.23 [0.13, 82.71]
6.33 itching	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.33 [0.01, 8.55]
6.34 vaginal discharge	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.33 [0.01, 8.55]
6.35 parasthesia	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.33 [0.01, 8.55]
6.36 suffered any side effect	1	55	Odds Ratio (M‐H, Fixed, 95% CI)	0.59 [0.20, 1.77]
6.37 amenorrhoea	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	0.04 [0.01, 0.38]
6.38 spotting or bleeding	1	49	Odds Ratio (M‐H, Fixed, 95% CI)	22.92 [2.64, 198.66]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Subjective improvement in pain	1		Peto Odds Ratio (Peto, Fixed, 95% CI)	Subtotals only
2 Objective efficacy ‐ rAFS scores at 6 months	1		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
3 Side effects	1		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
3.1 noted any side effect	1	12	Odds Ratio (M‐H, Fixed, 95% CI)	23.40 [0.89, 612.98]
3.2 discontinued treatment because of headaches	1	12	Odds Ratio (M‐H, Fixed, 95% CI)	3.55 [0.12, 105.82]
3.3 discontinued treatment because of continuing pain	1	12	Odds Ratio (M‐H, Fixed, 95% CI)	0.28 [0.01, 8.42]
3.4 suffered from irregular bleeding	1	12	Odds Ratio (M‐H, Fixed, 95% CI)	0.4 [0.03, 6.18]

Methods	Randomised single centre Double dummy parallel study Method of randomisation not described
Participants	48 Swedish women 18‐46 years Inclusion criteria: diagnosis of endometriosis by laparoscopy or laparotomy within 3 months regular menstruating and complaining of dysmenorrhoea, dyspareunia and/or pelvic pain Exclusion criteria: extensive adhesions, pelvic pain for other reasons, no surgery within the last 12 months with the exception of removal of an endometrioma, no use of laser or diathermy, steroid medication within 3 months or 1 month of diagnostic laparoscopy, previous use of any GnRH agonists, pregnant, breastfeeding or hysterectomy within 6 months prior to inclusion, use of concomitant contraceptive steroids, androgenic hormones, estrogens, progestagens, danazol,GnRh analogs, anxiolytics, cortizone and hypnotics,women with other concurrent disease either oncologic or psychiatric
Interventions	1. Nafarelin 200 µg intranasally (IN) BID and 'dummy' medroxyprogesterone tablets (23 women) 2. Medroxyprogesterone 15 mg PO BID and 'dummy' nafarelin nasal spray (25 women) Duration of treatment: 6 months
Outcomes	Pain scores using Biberoglu and Behrman scoring at 3, 6 and 12 months
Notes	18 withdrew from study Follow up: 6 months Unable to calculate means given data in current form
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of randomisation not described
Allocation concealment (selection bias)	Unclear risk	Not detailed in paper
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	Double dummy, no details and no details of blinding
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Detail 18 women who withdrew, six from nafarelin group and 12 from MPA group ‐ reasons not stated in paper
Selective reporting (reporting bias)	High risk	Main outcomes described, no details of side effects

Methods	Randomised double blind multi‐centre study   Method of randomisation not described   Pharmaceutical company stated
Participants	269 British women aged 18‐45   Inclusion criteria: endometriosis confirmed by laparoscopy or laparotomy.   Exclusion criteria: those requiring surgical excision, serious systemic disease, those requiring long‐term treatment, previous failure of danazol treatment, other hormonal treatment within 2 months, unwillingness to use mechanical contraception
Interventions	1. Gestrinone 2.5 mg twice weekly plus 'dummy' danazol for 6 months (132 women)   2. Danazol 200 mg bd plus 'dummy' gestrinone for 6 months (137 women)   Duration of treatment: 6 months
Outcomes	AFS scores at laparoscopy following 6 months treatment   Pain scores during treatment and 1 year follow‐up   Side effects   Fertility
Notes	Repeat laparoscopy 23 days (median) after end of treatment   Follow up: 12 months   5 women became pregnant before commencing treatment   69 withdrew during treatment   50 withdrew from follow‐up phase
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	'Allocated at random'; no other details
Allocation concealment (selection bias)	Unclear risk	Unclear, no details in paper
Blinding (performance bias and detection bias)   All outcomes	Low risk	'Double blind' 'Double dummy'. Patients received two identical tablets. Authors state that patients were blinded but do not reveal who else was also blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Details provided of those women not included in the analysis and at what time point
Selective reporting (reporting bias)	Low risk	Include main outcomes and side effects

Methods	Open randomised trial   No source of funding stated
Participants	39 Italian women aged 23‐35   Inclusion criteria: infertility, laparoscopic diagnosis of endometriosis in preceding 3 months   Exclusion criteria: bilateral tubal occlusion, severe dyspermia in partner, use of danazol or other sex steroids in preceding 6 months, severe systemic or endocrine disease
Interventions	1. Gestrinone 2.5 mg twice weekly (20 women) increasing to 3 times a week if no amenorrhoea by 1 month (7 of the 20)   2. Danazol 600 mg per day (19 women) increasing to 800 mg per day if no amenorrhoea by 1 month (2 of the 19)   Duration of treatment: 6 months
Outcomes	rAFS scores at laparoscopy 1 month after end of treatment   Pain scores during treatment and 18 month follow‐up   Plasma hormone levels before and during treatment   Pregnancy rates post treatment   Side effects
Notes	Only 7 gestrinone and 9 danazol patients had repeat laparoscopy   Follow up: 12 months   Losses to follow‐up: 1
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	'patients were randomly assigned'
Allocation concealment (selection bias)	High risk	No details provided
Blinding (performance bias and detection bias)   All outcomes	High risk	No blinding
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All women accounted for
Selective reporting (reporting bias)	High risk	All important outcomes reported with the exception of live birth

Methods	Randomised double blind double dummy multi‐centre trial   Method of randomisation described   Pharmaceutical company stated
Participants	55 Italian women aged 18‐40   Inclusion criteria: chronic pelvic pain, laparoscopic diagnosis of endometriosis with no attempts at endometriosis reduction other than biopsy up to 3 months before study entry, no medical or surgical treatment for endometriosis between laparoscopy and study entry, not wanting pregnancies in the immediate future   Exclusion criteria: treatment for endometriosis other than non steroidal anti inflammatory drugs in the previous 6 months, concomitant pelvic pain causing disorders, contraindications to the use of gestrinone or GnRH analogues, abnormal baseline bone density values, unwillingness to use barrier contraception
Interventions	1. Gestrinone 2.5 mg twice weekly plus placebo injections (27 women)   2. Intramuscular (IM) leuprolide acetate 3.75mg once a month plus placebo tablets (28 women)   Duration of treatment: 6 months
Outcomes	Pain symptoms   Bone mineral density   Lipid profile
Notes	Follow up: 6 months   6 withdrawals during treatment period   7 lost to follow‐up   8 pregnancies
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	'randomized' 'allocating consecutively numbered anonymous packages'
Allocation concealment (selection bias)	Low risk	Sealed envelopes containing randomization codes'
Blinding (performance bias and detection bias)   All outcomes	Low risk	'Double blind, double dummy'. Each patient received an active drug and a dummy placebo. Patients and clinicians were blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Losses to follow up detailed, 6 withdrawals during treatment period   7 lost to follow‐up
Selective reporting (reporting bias)	High risk	Did not include live births

Methods	Randomised double blind trial   Pharmaceutical company stated
Participants	12 American women   Inclusion criteria: endometriosis (stage 2‐3 disease according to rAFS classification) diagnosed on videotaped laparoscopy within previous 6 weeks   Exclusion criteria: none specified
Interventions	1. Gestrinone 1.25 mg twice weekly (6 women)   2. Gestrinone 2.5 mg twice weekly (6 women)   Duration of treatment: 6 months
Outcomes	rAFS scores of endometriosis at laparoscopy following treatment   Symptom scores during treatment and follow‐up   Side effects   Bone densitometry   Hormonal, lipoprotein, haematological and biochemical measurements
Notes	Second laparoscopy within 4 weeks of completing treatment   Follow‐up: 6 months   Losses to follow‐up: 2
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomised trial
Allocation concealment (selection bias)	Low risk	A ‐ Adequate
Blinding (performance bias and detection bias)   All outcomes	Low risk	Double blind
Incomplete outcome data (attrition bias)   All outcomes	Low risk	Losses to follow‐up: 2
Selective reporting (reporting bias)	High risk	Not addressed live births

Methods	RCT
Participants	Italy n = 40 women with mild endometriosis (stage I‐II) Age range 23 to 35 years Diagnosed by laparoscopy and clinical symptomology
Interventions	Desogestrel 75 μg per day (n = 20) versus ethinylestrdiol plus desogestrel (EE 20 μg + desogestrel 150 μg per day) Follow‐up: 6 months
Outcomes	Pain score (VAS 0‐10), serum glucose, cholesterol and triglycerides, side effects
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	'randomized' no other details
Allocation concealment (selection bias)	Unclear risk	Not stated
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	Not stated
Incomplete outcome data (attrition bias)   All outcomes	Low risk	All patients appear to have been followed up
Selective reporting (reporting bias)	Low risk	All a priori outcomes were reported on

PERMALINK

Progestagens and anti‐progestagens for pain associated with endometriosis

Julie Brown

Sari Kives

Muhammad Akhtar

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

1.

2.

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Results

Description of studies

Results of the search

Included studies

Progestagens

Anti‐progestagens

Excluded studies

Risk of bias in included studies

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Effects of interventions

Summary of findings for the main comparison. Progestagen compared to placebo for pain associated with endometriosis.

Summary of findings 2. Depot progestagen compared to other treatment for pain associated with endometriosis.

Summary of findings 3. Oral progestagens versus other treatment for pain associated with endometriosis.

Summary of findings 4. Anti‐progestagen compared to other treatment for pain associated with endometriosis.

1. Progestagens versus no treatment or placebo

Efficacy

3.

4.

Adverse effects

5.

6.

7.

2. Depot progestagens versus other treatment

Efficacy

2.1. Analysis.

8.

Adverse effects

9.

3. Oral progestagens versus other treatment

Efficacy

Methods	Randomised, evaluator blinded, multi‐centre (7 sites) trial
Participants	274 Candian and North American pre‐menopausal women aged 18‐49. Mean age DMPA 29.2±6.3, Leuprolide 32.1±6.6 (P < 0.001) Inclusion criteria: endometriosis surgically diagnosed within 42 months and pain within 30 days of diagnostic laparoscopy or after 3 months following laparoscopy or laparotomy; Biberoglu & Behrman score ≥ 6 including at least 2 in symptoms of dysmenorrhoea, dyspareunia and pelvic pain; pain must persist more than 3 months Exclusion criteria: BMD at lumbar spine or hip < v1.0SD below mean for peak adult bone mass
Interventions	Depomedroxyprogesterone acetate 104 mg SC every 3 months (n = 136) versus leuprolide 11.25 mg IM every 3 months (n = 138) Treatment duration ‐ 6 months
Outcomes	Pain scores during treatment at 12 months post treatment, BMD, adverse events, hyperoestrogenic symptoms, bleeding, and quality of life
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Authors state 'randomised', no other details
Allocation concealment (selection bias)	Low risk	Centrally randomised by an independent investigator
Blinding (performance bias and detection bias)   All outcomes	Low risk	Principle investigator and sub investigators were blinded
Incomplete outcome data (attrition bias)   All outcomes	Low risk	DMPA drop out was 48/136 and leuprolide was 36/138, non‐specific reasons given
Selective reporting (reporting bias)	Low risk	A priori outcomes presented as per methods section of paper. Protocol not accessed

Methods	Double blind double dummy single centre study   Randomisation method not clear
Participants	59 participants aged 26‐38 with mild to moderate endometriosis   No previous medical or surgical treatment   No exclusion criteria specified   9 participants lost to follow‐up
Interventions	Danazol 200 mg PO TDS   Medroxyprogesterone acetate 100 mg PO daily   Placebo   All medications taken for 180 days
Outcomes	Change in AFS scores   Patient reported pain symptoms   Side effects
Notes	27% of patients had electro‐coagulation of implants at initial diagnostic laparoscopy   2nd look laparoscopy was performed 6 months after completion of treatment
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Authors state 'randomised' but no other details
Allocation concealment (selection bias)	Unclear risk	No details in paper
Blinding (performance bias and detection bias)   All outcomes	Unclear risk	State 'double blind' but no other details
Incomplete outcome data (attrition bias)   All outcomes	High risk	Numbers of patients not completing study in placebo group does not add up correctly
Selective reporting (reporting bias)	Unclear risk	A priori outcomes reported but original protocol not sighted

Methods	RCT
Participants	90 women with recurrent moderate or severe pelvic pain after conservative surgery for symptomatic endometriosis Inclusion: 18‐40 years, not desiring pregnancy, who had undergone conservative surgery at laparoscopy or laparotomy for stage I‐IV symptomatic disease in the previous 12 months. Only included women with confirmed surgical eradication and who had recurrent pelvic pain for more than 6 months Exclusion: therapies other than non‐steroidal anti‐inflammatories
Interventions	6 months treatment Oral cyproterone acetate 12.5mg/d versus oral contraceptive ‐ ethinyl estradiol 0.02 mg and desogestrel 0.15 mg
Outcomes	Biberoglu and Behrman scores and VAS for pain
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Unclear
Allocation concealment (selection bias)	Unclear risk	Unclear
Blinding (performance bias and detection bias)   All outcomes	High risk	No blinding ‐ open label study
Incomplete outcome data (attrition bias)   All outcomes	Low risk	6 in the cyproterone acetate group and 9 in the oral contraceptive group withdrew due to side effects (n = 9), treatment inefficacy (n = 4) or loss to follow‐up (n = 2)
Selective reporting (reporting bias)	Unclear risk	Unclear

Study	Reason for exclusion
Cosson 2002	All patients received surgery immediately prior to medical therapy
Dawood 1997	Pain data not reported separately for the two groups. Relief of pain was not a primary endpoint
Harrison 2000	Relief of pain was not an outcome in this study
Mettler 1987	The "three step" therapy discussed in this study is a mixture of surgical and medical therapy
Nieto 1996	23/25 patients on gestrinone and 18/18 patients on danazol had surgery prior to medical treatment
Noble 1980	Comparison of danazol with oral contraceptive pill
Regidor 2001	All patients had received surgery immediately prior to medical therapy
Strowitzki 2009	This is a conference abstract that has been superseded by a full text paper which has been included in the review
Telimaa 1987a	Patients were recruited to the study following surgical treatment
Thomas 1987a	This study does not have relief of pain as an outcome measure; it concentrates on effects on fertility
Vercellini 2005	Patients had rectovaginal endometriosis only
Walch 2009	Comparison was between 2 progestagens
Worthington 1993	Relief of pain is not an outcome considered in this study
Yang 2006	The comparison group received a complementary therapy intervention