Elevated follicular cortisone level is a negative predictor of clinical pregnancy in women undergoing fresh embryo transfer

BACKGROUND: Although numerous studies have investigated the potential correlation between follicular fluid (FF) steroid concentrations and in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) outcomes, few have accounted for the effect of controlled ovarian hyperstimulation regimes on FF steroid concentrations. OBJECTIVE: To comprehensively compare follicular steroid concentrations between women stimulated with gonadotropin-releasing hormone agonist (GnRHa) and antagonist (GnRHant) protocols and to explore the associations between FF steroid concentrations and IVF/ICSI outcomes. METHODS: A total of 295 infertile women undergoing IVF/ICSI from January 2018 to May 2020 were enrolled. Eighty-four and 211 women received GnRHa and GnRHant protocols, respectively. Seventeen steroids in FF were quantified by liquid chromatography tandem mass spectrometry (LC–MS/MS), and the correlation of follicular steroids with clinical pregnancy was explored. RESULTS: Follicular steroid concentrations were similar between the GnRHa and GnRHant groups. Follicular cortisone levels were adversely associated with clinical pregnancy in fresh embryo transfers. Receiver operating characteristic (ROC) analysis revealed an area under the ROC curve (AUC) of 0.639 (95% confidence interval = 0.527–0.751, p = 0.025) for predicting non-pregnancy, with an optimal cutoff value of 15.81 ng/mL (sensitivity = 33.3%, specificity = 94.1%). Women with FF cortisone concentrations ≥15.81 ng/mL were fifty times less likely to achieve clinical pregnancy in fresh embryo transfers than those with FF cortisone levels below this threshold (adjusted OR = 0.019, 95% confidence interval = 0.002–0.207, p = 0.001) after adjusting for age, body mass index, baseline serum progesterone levels, serum levels of luteinizing hormone, estradiol and progesterone on human chorionic gonadotropin day, ovarian stimulation protocols, and the number of transferred embryos. CONCLUSIONS: There was no significant difference in intrafollicular steroid levels between GnRHa and GnRHant protocols, and intrafollicular cortisone level ≥15.81 ng/mL was found to be a strong negative predictor of clinical pregnancy in fresh embryo transfers with high specificity.