Higher live birth rate with stimulated rather than artificial cycle for frozen-thawed embryo transfer

Objective: To study which endometrial preparation allows a better ongoing pregnancy rates (OPR) and live birth rate (LBR) after frozen-thawed embryo transfer (FET) between mild gonadotropin ovarian stimulation (OS) and artificial cycles (AC). Study design: Retrospective follow-up study including all FET performed in one fertility center from 2013 to 2016. In the OS group, gonadotropins were followed by r-hCG triggering. Vaginal micronized progesterone (200 mg/day) was given systematically. In the AC group, estradiol (E2) was started on Day 1. Vaginal micronized progesterone (600 mg/d) was added to E2 for 12 weeks. Data were analyzed using a multiple regression model. Results: Among 1021 FETs, 35% underwent OS preparation, 65% had an AC. As expected, patients in the AC group suffered more from endometriosis (18.5% vs. 12.9%; p = .021) and polycystic ovarian syndrome (21.7% vs. 10.9%; p < .0001) than patients in the OS group. There was no difference between groups with respect to endometrial thickness, number of embryos transferred, development stage at FET, cryopreservation technique. Despite a similar clinical pregnancy rate (CPR) (24.4% vs. 20.8%; p = .189), the OPR was significantly higher in the OS than in the AC group (17.9% vs. 11%; p = .002), leading to an increased LBR (17.1% vs. 9.8%; p <.001). After adjusting for parameters usually linked to early pregnancy losses or potential bias (patient age at freezing, smoking status, PCOS, endometriosis, rank of transfer and previous miscarriages), the results remained significant. Conclusion: Despite a similar CPR, LBR was significantly higher with mild OS than with the AC preparation, even after adjusting for potential confounders. In light of these results, the first-line endometrial preparation could be OS instead of an AC. In an AC, a potential defect of the luteal phase may exist, treatment could be optimized to avoid pregnancy losses. A randomized controlled trial should be undertaken to assess the role of OS and ACs in FET. (C) 2019 Elsevier B.V. All rights reserved.