Basal serum level of Δ4-androstenedione reflects the ovaries' ability to respond to stimulation in IVF cycles: setting up a new reliable index of both ovarian reserve and response.

Purpose: Adequate androgen levels are necessary for regular follicular growth, progression beyond the pre-antral stage, and prevention of follicular atresia. The main purpose of this study was to investigate whether baseline androgen levels had a predictive value on stimulation outcomes in IVF cycles. The secondary purpose was to compare the possible predictive value of androgens with that of already known markers.

Methods: The study included 91 infertile patients aged 30-45 years awaiting the first IVF cycle. All women underwent the same stimulation protocol and the same starting dose of recombinant FSH. 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. Multiple linear regression and multivariate logistic regression were used to evaluate the significant predictive value of the variables for response to controlled ovarian stimulation (COS). By studying the reliability of different markers, an attempt was made to develop a single index with the highest predictive value.

Results: 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 (anti-müllerian hormone) (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). In COS good responders, the G-index (= AMH ng/mL*AFC/Δ4-A ng/dL) revealed a significantly higher level (p < 0.001) than AMH, AFC, and Δ4-A alone.

Conclusion: Baseline serum Δ4-A, presumably crucial for ensuring a regular 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, the G-index, able to contemplate both the ovarian reserve and the Δ4-A level.