High FSH dosing is associated with reduced live birth rate in fresh but not subsequent frozen embryo transfers

Study question Do live birth rates (LBRs) differ between fresh embryo transfer (fresh ET) cycles and their subsequent paired frozen ET (FET) cycles, when comparing cycles based on the total FSH dose used during the fresh cycle? Summary answer When compared to the paired frozen embryo transfer cycles, the LBR in the fresh cycle of the highest total FSH dose group (>2500 IU) was reduced by 38%. What is known already There may be a negative association with high gonadotropin doses and LBR after fresh ET. It is unknown whether a similar effect is seen in FET cycles, which are done with increasing frequency. Study design, size, duration In this retrospective observational paired study, we studied IVF cycles between 10 January 2005 and 19 September 2015, for all patients who underwent a fresh, autologous IVF cycle that resulted in at least one fresh ET and at least one FET. Participants/materials, setting, methods The study included 862 women, treated in our academic medical centre, who underwent 935 fresh ET and 1274 FET cycles. Cycles were allocated into three groups based on the total gonadotropin dose they received during their fresh IVF cycle: Group 1 (≤1800 IU FSH), Group 2 (1801-2500 IU), Group 3 (>2500 IU). The primary outcome was LBR after fresh ET and its subsequent paired FET(s), as well as LBR among fresh ETs and FETs as independent samples, based on the total FSH dose used. Implantation rates obtained from fresh and FET cycles were also compared. Main results and the role of chance The unadjusted fresh LBR was similar between Groups 1 and 2 (46.0% [95% CI: 40.4-51.6] versus 43.8% [38.3-49.4], respectively) but significantly lower in Group 3 (34.4% [29.5-39.8]). The unadjusted frozen transfer LBR was similar among all groups (51.4% [46.7-56.1] versus 46.3% [41.3-51.4] versus 47.5% [42.5-52.4], respectively). When logistic regression analysis with generalized estimating equations was used to control for confounders, the adjusted LBR was found to be similar between the groups both for fresh (odds ratio [OR] = 0.97 [95% CI: 0.61-1.56] Group 2 versus Group 1, OR = 0.69 [0.39-1.21] Group 3 versus Group 1) and FET cycles (OR = 0.87 [0.58-1.31] Group 2 versus Group 1, OR = 0.95 [0.58-1.55] Group 3 versus Group 1). However, for Group 3, the adjusted fresh LBR was 38% lower than its subsequent frozen transfer LBR (OR = 0.62 [0.41-0.93]); this was a statistically significant difference, which was not observed in Group 1 (OR = 0.85 [0.56-1.27]) or Group 2 (OR = 0.95 [0.64-1.41]). Limitations, reasons for caution This study is a retrospective cohort, with all of the associated inherent biases. Wider implications of the findings Fresh LBR is negatively impacted by a high dose of total FSH used, as compared to the LBR in subsequent paired FET cycles. Frozen transfer LBR seems unaffected by the total FSH dose used in the fresh cycle, suggesting that the endometrium may be adversely affected, probably indirectly, by high dose gonadotropin use in the fresh IVF cycle only. Study funding/competing interest(s) No funding source was used for the completion of this project. There are no conflicts of interest.