Referral for diagnosis

• Establishing a diagnosis of endometriosis on the basis of symptoms alone is difficult because symptoms are variable and overlap with other conditions (for example pelvic inflammatory disease and irritable bowel syndrome).
• There are no investigations in primary care that are helpful in making a diagnosis of endometriosis. However, CKS expert reviewers commented that:
• A normal ultrasound scan of the pelvis does not exclude endometriosis but can be reassuring, particularly as it can exclude ovarian endometrioma with good certainty.
• A successful trial of hormonal suppression therapy (for example COC or oral progestogens), in the presence of normal pelvic ultrasound, is suggestive that at least mild endometriosis may be present.
• Laparoscopy is the gold standard diagnostic test.
• The Royal College of Obstetrician and Gynaecologists [RCOG, 2006] and the European Society of Human Reproduction and Embryology Special Interest group for Endometriosis and Endometrium guideline development group [Kennedy et al, 2005] recommend visual inspection of the pelvis by laparoscopy as the gold standard diagnostic test [RCOG, 2006] except, rarely, when the disease is visible (such as in the posterior vaginal fornix).
• Other tests are less useful:
• Transvaginal ultrasound has no value in diagnosing peritoneal deposits but is useful to confirm or exclude an ovarian endometrioma [RCOG, 2006].
• There is insufficient evidence to recommend magnetic resonance imaging as a useful diagnostic test for endometriosis [Kennedy et al, 2005; RCOG, 2006].
• Although serum CA-125 may be elevated in endometriosis, it has no value as a diagnostic tool [Kennedy et al, 2005].

Nonsteroidal anti-inflammatory drugs

• Although there is good evidence that NSAIDs are effective in the treatment of primary dysmenorrhoea, there is less evidence regarding their use in women with secondary dysmenorrhoea or confirmed endometriosis. A Cochrane systematic review (that included only one small randomized controlled trial) showed:
• Naproxen may be no more effective than placebo in the management of endometriosis-associated pain.
• Insufficient evidence to indicate whether one NSAID is more effective than any other for the treatment of endometriosis.
• Despite the lack of evidence to support their use, NSAIDs are widely prescribed for the treatment of endometriosis-associated pain and seem to be effective for many women.
• Ibuprofen, diclofenac, naproxen, and mefenamic acid are recommended in the treatment of endometriosis-associated pain because:
• Ibuprofen is considered to have a lower risk of gastrointestinal adverse effects than other NSAIDs [García Rodríguez and Hernández-Díaz, 2001; BNF 57, 2009].
• Naproxen and diclofenac are associated with an intermediate risk of gastrointestinal adverse effects; mefenamic acid, having a short half-life, is likely to be associated with a low-to-intermediate risk [García Rodríguez and Hernández-Díaz, 2001; BNF 57, 2009].

Paracetamol

• CKS found no trials on the use of paracetamol to treat endometriosis-associated pain.
• There is only very limited evidence from one small (n = 35) randomized controlled trial for its use in the relief of primary dysmenorrhoea. This trial showed paracetamol (at a sub-therapeutic dose) to be no better than placebo.

Codeine

• CKS found no direct evidence on the use of codeine in the treatment of endometriosis-associated pain. However, the addition of codeine seems reasonable in women who are unable to take an NSAID or who do not get adequate pain relief from paracetamol plus an NSAID.

Hormonal treatments

• Hormone treatments aim to induce atrophy of ectopic endometrium, either by altering the effect of oestrogen on endometriotic tissue or by reducing circulating oestrogen levels. They reduce endometriosis-associated pain by reducing menstrual blood flow or by inducing amenorrhoea. CKS expert reviewers agreed that although this may alter the laparoscopic appearance of the disease, in practice this was not a concern.

The combined oral contraceptive pill

• The limited evidence available suggests that there is no difference between the effectiveness of the low-dose COC taken cyclically and goserelin in relieving endometriosis-associated pain.
• There is a lack of evidence to guide which COC or treatment regimen to use [Kennedy et al, 2005; RCOG, 2006] therefore, CKS recommends the preparations that are usually used first-line for contraception. See the CKS topic on Contraception for more information. Different treatment regimens may be considered:
• Conventional use of COCs. There is evidence from a small randomized trial supporting the effectiveness of low-dose oral contraception, used cyclically.
• Tricycling COCs (using COCs continuously for 3 months followed by 1 week without pills) has not been studied in women with endometriosis, but as it reduces the frequency of menstrual bleeding it may improve quality of life [Nasir and Bope, 2004]. This practice is an off-label use. The regimen is regarded as having a good safety profile, and is well tolerated and acceptable to women (although this does not specifically relate to women with endometriosis) [FFPRHC, 2005].
• Continuous use (also off-label) could be considered if tricycling does not provide sufficient symptom relief.
• Evidence for continuous use of COCs comes from a prospective, non-randomized, self-controlled trial of women (n = 50) who had experienced recurrence of dysmenorrhoea within a year of having surgery for endometriosis despite taking the COC cyclically [Vercellini et al, 2003]. Women were offered ethinyl estradiol 0.02 mg and desogestrel 0.15 mg daily for 2 years but could choose at each 6 monthly review whether they wanted to continue daily treatment, revert to the standard 21-day cycle, or discontinue treatment. Forty one women completed the study: 80% were either satisfied (54%) or very satisfied (27%) with the treatment. The following adverse effects were reported: amenorrhoea (38%), spotting (36%), and breakthrough bleeding (13%).
• The Faculty of Sexual and Reproductive Healthcare (FSRH), formerly the Faculty of Family Planning and Reproductive Healthcare (FFPRHC), does not report on the safety of this regimen, but other authors [Miller and Hughes, 2003] have commented that it is unlikely that the loss of the pill-free week will greatly alter the safety profile.

Oral progestogens that are not contraceptives

• Progestogens induce endometrial atrophy and reduce oestrogen levels by inhibiting ovulation.
• CKS expert reviewers commented that treatment with oral progestogens (medroxyprogesterone or norethisterone) could be started by primary healthcare professionals provided that the symptoms were suggestive of endometriosis. Some expert reviewers suggested an ultrasound scan, to rule out pathology such as an endometrioma, before prescribing.
• Evidence is limited due to lack of data, but continuous high-dose progestogen (medroxyprogesterone acetate) appears to be effective for the treatment of endometriosis-associated pain.
• Luteal phase dydrogesterone does not appear to be effective and is no longer available.
• There is limited evidence for the use of norethisterone.

Progestogen-only contraceptives

• CKS expert reviewers commented that not all women achieve amenorrhoea with these methods.
• Some expert reviewers stressed that a pelvic examination should be normal before prescribing progestogen-only contraception. They also suggested an ultrasound scan to rule out pathology such as an endometrioma.
• Most CKS expert reviewers recommended that the levonorgestrel-releasing intrauterine system (LNG-IUS) could be initiated by a GP, particularly if the woman needed contraception or had experienced good symptom relief from this method in the past.
• There is limited evidence from two small randomized controlled trials and three small prospective observational studies that the LNG-IUS reduces endometriosis-associated pain, with symptom control maintained over 3 years.
• There is limited evidence from a Cochrane systematic review that found one small open-label study for its use following surgery for endometriosis.
• Most CKS expert reviewers recommended that depot medroxyprogesterone (Depot-Provera^®) could also be initiated by a GP, particularly if the woman needed contraception or had experienced good symptom relief from this method in the past, but that not all women would achieve amenorrhoea.
• There is limited evidence from one small randomized trial that depot medroxyprogesterone is effective in the treatment of pelvic pain associated with endometriosis.
• Some CKS expert reviewers also recommended that Cerazette^® (a progestogen-only pill) or the etonogestrel subdermal implant (Implanon^®) were alternatives if the woman needed contraception or had experienced good symptom relief from this method in the past.
• Note that from mid-October 2010, Nexplanon^® will replace Impanon^®. Nexplanon^® is bioequivalent to Implanon^®. The main differences are that Nexplanon is radio-opaque, and the insertion technique is different [FSRH, 2010].
• CKS found no evidence assessing the effectiveness of Cerazette^® for endometriosis.
• Very limited evidence from a small randomized controlled trial, an open-label study, and five case studies suggests that the etonogestrel subdermal implant might be useful in relieving endometriosis-associated pelvic pain.

Referral if no response to treatment

• Referral if symptoms do not resolve after 3 months' treatment with oral progestogens is based on expert opinion.