Objectives The clinical remission (CR) rate of Megestrol acetate (MA) as fertility-sparing treatment was still not optimal. This study was to access whether MA plus metformin could lead to better CR rate for patients with atypical endometrial hyperplasia (AEH) and well-differentiated endometrial cancer (EC).

Methods This was a randomized, single-center, open-label and controlled trial (July 2013-December 2017). Patients were randomized to receive MA (160mg, orally, daily) or MA (160mg, orally, daily) plus metformin (500mg, orally, three times a day), then underwent hysteroscopy every 3–4 months. The primary efficacy parameter was the CR rates at 16th and 30th weeks of treatment (16w-CR rate and 30w-CR rate); the secondary efficacy parameter was rates of recurrence, pregnancy and live-birth.

Results Totally 150 patients received MA (n=74, 62 AEH and 12 EC) or MA/metformin (n=76, 61 AEH and 15 EC). The 16w-CR rates were 34.3% (23/67) and 20.7% (12/58) in MA/metformin- and MA-treated women (p=0.091). However, among 102 AEH patients, MA/metformin yielded significantly higher CR rate (39.6%, 21/53) than MA alone (20.4%,10/49, p=0.032, OR 0.347, 95%CI 0.132–0.914). Regarding 30w-CR rate, it was slightly higher in MA/metformin group than control (69.2% vs 57.4%, p=0.167). Nevertheless, the mean treatment time in MA/metformin group was 4 weeks shorter (27 vs 31 weeks). Particularly, the mean weight gain by MA/metformin (2.5kg) was twice lower than MA alone (5.0kg, p=0.014). No intra-group difference was found in rates of recurrence, pregnancy or live birth.

Conclusions MA/metformin may lead to a higher CR rate, shorter treatment time and less weight gain compared with MA alone.