Elevated basal progesterone levels are associated with increased preovulatory progesterone rise but not with higher pregnancy rates in ICSI cycles with GnRH antagonists

Abstract

ObjectiveTo ascertain the association between basal progesterone (P) levels and the occurrence of preovulatory progesterone rise (PPR) and clinical pregnancy rates (CPRs) in ICSI cycles with GnRH antagonists. Study designSerum P levels of 464 patients were measured on day 2 and day of hCG of cycles. Cycles with basal P levels>1.6ng/mL were cancelled. All embryos were cryopreserved in cycles with P levels≥2ng/mL on the day of hCG. The primary outcome measures were the incidence of PPR (P>1.5ng/mL) and CPR with regard to basal P. ResultsBasal P levels were significantly higher in cycles with PPR than in those without PPR (0.63±0.31 vs. 0.48±0.28ng/mL). Area under the curve for basal P according to ROC analysis to discriminate between elevated and normal P levels on the day of hCG was 0.65 (0.58–0.71 95% CI, p<0.01). The cut-off value for basal P levels that best discriminates between cycles with and without PPR was 0.65ng/mL. Cycles with basal P levels above 0.65ng/mL had a significantly higher incidence of PPR (30.9% vs. 13.5%) but similar clinical and cumulative pregnancy rates (38.8% vs. 31.1% and 41.7% vs. 32.6%, respectively) in comparison to cycles with basal P levels below 0.65ng/mL. In multivariate regression analysis, basal P levels, LH level on the first day of antagonist administration, and estradiol levels on the day of hCG trigger were the variables that predicted PPR. ConclusionBasal P levels were associated with increased incidence of PPR but not with CPR.