P-644 The Δ4-Androstenedione baseline serum level reflects the ovarian response to stimulation: set-up of a novel index to predict the outcomes of IVF cycles

Abstract

Abstract Study question Aim of this study was to investigate whether baseline androgen levels had a predictive value on stimulation outcomes in IVF cycles. Summary answer Baseline Δ4-Androstenedione (Δ4-A) and particularly the derived G-index, which also reflects the ovarian reserve capability, may serve as additional predictors of ovarian response to stimulation. What is known already Despite the several clinical, serological and ultrasound markers proposed so far, none have proved exempt from a high rate of false positives and negatives. The identification of reliable predictors of ovarian response continues to pose a challenge for reproductive physicians. Study design, size, duration Prospective study. Multiple linear regression and multivariate logistic regression were used to evaluate the predictive value of markers of response to controlled ovarian stimulation (COS). To compare the specificity of different biomarkers, receiver operating characteristic (ROC) curves were constructed to identify the global accuracy of our parameters of interest. Participants/materials, setting, methods The study included 91 infertile patients aged 30-45 years,awaiting the first IVF cycle at the Reproductive Medicine Unit of the San Paolo University Hospital in Milano.All women underwent the same long-protocol stimulation and the same starting dose of recombinant follicle-stimulating hormone (rFSH) was administered. As stimulation outcomes, the number of follicles recruited, Estradiol and Progesterone levels on the day of trigger, the total dose of gonadotropins administered and the number of oocytes collected were recorded. Main results and the role of chance Pearson’s correlation revealed a statistically significant inverse correlation between oocytes collected and age (r= -0.333,p<0.001) and a positive correlation with AMH (r = 0.360,p<0.001), antral follicle count (AFC) (r = 0.639, p < 0.001) and Androstenedione (Δ4-A) (r = 0.359,p<0.001). No significant correlation was reported with FSH (r=-0.133,p=0.207) and total Testosterone (r = 0.180,p=0.088). The G-Index (=AMH ng/ml*AFC/Δ4-A ng/dl) revealed a significant higher level in COS good responders (p < 0.001) compared to AMH, AFC and Δ4-A alone. Limitations, reasons for caution Although the same stimulation regimen and starting dose of gonadotropin were used, subsequent individualization of treatment certainly influenced the results. This was considered as an insurmountable bias. Furthermore, the conclusions of this study, due to the population size, need to be validated in studies with larger populations. Wider implications of the findings Baseline serum Δ4-A, presumably crucial for ensuring a regular early follicular growth, is a reliable marker of ovarian response to stimulation.Since the ovarian capacity to respond to gonadotropins does not depend exclusively on the presence of follicles,we suggest a new index able to contemplate both ovarian reserve and Δ4-A level. Trial registration number N/A