Endometriosis-associated pain scores and biomarkers in users of the etonogestrel-releasing subdermal implant or the 52-mg levonorgestrel-releasing intrauterine system for up to 24 months.

The aims of the study were to correlate endometriosis-associated pain, evaluated by visual analogue scale (VAS) scores, with serum levels of etonogestrel (ENG), levonorgestrel (LNG), CA-125 and soluble CD23 in users of the ENG implant or the 52-mg LNG-releasing intrauterine system (52 mg LNG-IUS) for up to 2 years after device placement. A randomised trial was conducted at the University of Campinas Medical School, Brazil. All participants ( = 103) had had endometriosis-associated chronic pelvic pain or dysmenorrhoea, or both, for more than 6 months. Participants were randomly assigned to use an ENG implant (experimental treatment) or a 52-mg LNG-IUS (active comparator). Follow-up was conducted 6 monthly for up to 24 months after device placement. Dysmenorrhoea and chronic pelvic pain were evaluated using a VAS and the scores were correlated with serum levels of ENG, LNG, CA-125 and soluble CD23. Both progestin-only contraceptives significantly reduced VAS scores for dysmenorrhoea and chronic pelvic pain and reduced serum levels of soluble CD23 ( < 0.001). Serum levels of CA-125 decreased only in the ENG implant group after 24 months' use of the device ( < 0.001). No correlation was found between pain scores and ENG or LNG serum levels over time ( > 0.005). Both contraceptives improved dysmenorrhoea and chronic pelvic pain scores in women with endometriosis-associated pain and they reduced serum levels of soluble CD23; however, serum levels of CA-125 were reduced only in ENG implant users over the 24-month study period.