Use of metformin before and during assisted reproductive technology in non-obese young infertile women with polycystic ovary syndrome: a prospective, randomized, double-blind, multi-centre study.

BACKGROUND To study the effect of metformin before and during assisted reproductive technology (ART) on the clinical pregnancy rate (CPR) in non-obese women with polycystic ovary syndrome (PCOS). METHODS A multi-centre, prospective, randomized, double-blind study was conducted in eight IVF clinics in four Nordic countries. We enrolled 150 PCOS women with a body mass index <28 kg/m(2), and treated them with 2000 mg/day metformin or identical placebo tablets for ≥ 12 weeks prior to and during long protocol IVF or ICSI and until the day of pregnancy testing. The primary outcome measure was CPR. Secondary outcome measures included spontaneous pregnancy rates during the pretreatment period, and the live birth rate (LBR). RESULTS Among IVF treated women (n = 112), biochemical pregnancy rates were identical in both groups (42.9%), and there were no significant differences in the metformin versus the placebo group in CPR [39.3 versus 30.4%; 95% confidence interval (CI): -8.6 to 26.5]. The LBR was 37.5 versus 28.6% (95% CI: -8.4 to 26.3). However, prior to IVF there were 15 (20.3%) spontaneous pregnancies in the metformin group and eight (10.7%) in the placebo group (95% CI: -1.9 to 21.1; P = 0.1047). According to intention to treat analyses (n = 149); significantly higher overall CPR were observed in the metformin versus placebo group (50.0 versus 33.3%; 95% CI: -1.1 to 32.3; P = 0.0391). LBR was also significantly higher with use of metformin versus placebo (48.6 versus 32.0; 95% CI: 1.1 to 32.2; P = 0.0383). No major unexpected safety issues or multiple births were reported. More gastrointestinal side effects occurred in the metformin group (41 versus 12%; 95% CI: 0.15 to 0.42; P < 0.001). CONCLUSIONS Metformin treatment for 12 weeks before and during IVF or ICSI in non-obese women with PCOS significantly increases pregnancy and LBRs compared with placebo. However, there was no effect on the outcome of ART per se. TRIAL REGISTRATION ClinicalTrials.gov Identifier: NCT00159575.