P-459 Impact of a drop in serum estradiol levels after gonadotropin-releasing hormone (GnRH) antagonist on clinical outcomes of ovarian stimulation cycles in assisted reproductive technology (ART)

Abstract

Abstract Study question Does a drop in serum estradiol level after the administration of a GnRH antagonist affect outcomes in ovarian stimulation cycles in ART? Summary answer A drop in serum estradiol after administration of a GnRH antagonist is related to changes in luteinizing hormone (LH) and to inferior clinical outcomes. What is known already In recent years, the GnRH antagonist protocol has become widely used to prevent a premature LH surge during ovarian stimulation in ART, with many advantages such as immediate pituitary suppression and a lower risk of ovarian hyperstimulation. However, one phenomenon intriguing professionals monitoring these cycles and reviewing blood analyses, is the decrease in serum estradiol that occasionally occurs after the administration of the antagonist. With the natural menstrual cycle in mind, this could be an undesirable effect possibly affecting clinical outcomes. Study design, size, duration This was a non-interventional, retrospective, observational cohort study. We analyzed 1678 ovarian stimulation cycles in ART in 1143 patients. Patients were treated between January 1st 2015 and November 1st 2022. Patients were 20 to 45 years old, were stimulated with human menopausal gonadotropin (hMG) or recombinant follicle stimulating hormone (follitropin α, β or δ, corifollitropin α and follitropin α + lutropin α) and a GnRH antagonist was used for pituitary suppression in a flexible protocol. Participants/materials, setting, methods Data were extracted from medical records at a tertiary infertility clinic. The difference between serum estradiol level before and after the first antagonist administration was calculated. When this value was negative, the cycle was marked as having an estradiol drop. Influence of treatment and cycle characteristics on estradiol drop, as well as influence of the drop on clinical outcomes were analyzed using generalized linear mixed effects models. Main results and the role of chance The presence of a serum estradiol drop could be calculated in 1552 cycles, of which 153 cycles (9.9%) showed a drop. There was no significant difference in the odds on cycle cancellation between cycles with and without drop (p = 0.87). Type of gonadotrophin used was significantly associated with the presence of an estradiol drop (p < 0.0001), with hMG cycles showing the lowest amount of drops (3.0%). More estradiol drops were observed with the start of antagonist administration on day six of stimulation (10.3% of cycles), compared to starting later than day six (9.1%)(p = 0.001). Serum LH at the start of the cycle was significantly higher in cycles with estradiol drop (mean±SD=4.10±4.13IU/L) than in cycles without (3.88±2.74IU/L) (p = 0.01). Decrease in serum LH levels after antagonist administration was significantly larger in cycles with estradiol drop (-5.30±7.58IU/L versus -1.66±3.13IU/L,p<0.0001). Serum LH on the day of ovulation triggering was also lower in estradiol drop cycles (1.29±1.59IU/L versus 2.24±8.13IU/L,p=0.0007). Estradiol drop was not significantly associated with the number of oocytes retrieved, 2PN fertilized, utilization rate or embryo transfer (p-values>0.05). Ongoing pregnancy and live birth were also significantly lower in cycles with an estradiol drop (13.9%) than in cycles without (25%, p = 0.01). When correcting for age, p-values remained unchanged. Limitations, reasons for caution A limitation of the retrospective study design is that stimulation protocols and any adjustments were chosen by the treating physician according to the patient’s characteristics. The authors did adjust for different cycles per patient by using generalized mixed models. Absolute numbers of estradiol drop are low, limiting optimal multivariable analysis. Wider implications of the findings Because serum estradiol drop after GnRH antagonist administration is related to changes in LH, prospective research could focus on potential optimization of the LH serum levels which could potentially ameliorate clinical outcome. Trial registration number not applicable