Systematic review of the effects of aromatase inhibitors on pain associated with endometriosis

Sir, We read with interest the systematic review of the effects of aromatase inhibitors on pain associated with endometriosis by Patwardhan et al.1 It seems that in relation to pelvic pain, only a single case study supports the use of aromatase inhibitors alone as fully effective in alleviating endometriosisassociated pelvic pain.2 A single randomised control trial (RCT) has demonstrated an improved pain score in the combined arm, but did not assess other outcome measures such as quality of life.3 Other studies have compared the use of combined therapy (hormone/gonadotrophin-releasing hormone [GnRH] analogue + aromatase inhibitor) with the use of a hormone/GnRH analogue alone. However, no studies are available comparing the efficacy of a hormone or GnRH analogue/aromatase inhibitor combination with that of a hormone or GnRH analogue/nonsteroidal antiinflammatory drug (NSAID) combination. Another aspect worth considering is the efficacy of aromatase inhibitors alone in comparison with NSAIDs or placebo for that matter, in treating endometriosis-associated pain. A Cochrane systematic review in 2005 stated that the evidence regarding the use of nonsteroidal anti-inflammatory drugs for endometriosis-associated pain was inconclusive because of insufficient studies.4 In the absence of proper RCTs, is there enough evidence to establish the principle that aromatase inhibitors can be effective in alleviating pelvic pain associated with endometriosis, especially with the obvious difference in cost-effectiveness when compared with NSAIDs and possible detrimental effect on bone mineral density? Despite the aforementioned limitations of this review, and the questionable quality of the evidence, we were somewhat surprised to see that in the review of medical management of endometriosis by Panay5 in the same issue, aromatase inhibitors were deemed to have a sufficient evidence base for him to recommend their usage. One further issue relates to the advisability of symptom relief without disease suppression. The use of nonhormonal therapy presumably risks achieving the relief of symptoms but without suppression of the disease process. While we realise that the review of aromatase inhibitors does not address this issue more specific evidence advice on this aspect would have been welcome, particularly on the relative merits of GnRHa, aromatase inhibitors and nonsteroidal therapy. j