A prospective assessment of the predictive value of basal antral follicles in in vitro fertilization cycles

Objective: To determine the predictive value and define threshold values for basal antral follicle count in patients undergoing IVF. Design: Prospective cohort analysis. Setting: Tertiary care center. Patient(s): Two hundred eighty-nine patients. Intervention(s): Transvaginal ultrasonography before starting gonadotropin administration. Main Outcome Measure(s): Number of oocytes retrieved, basal hormone levels, and cycle outcomes. Result(s): Pregnant patients had significantly more antral follicles (13.8 ± 7.5 vs. 12.4 ± 10.0). Patients in whom cycles were canceled had significantly fewer antral follicles (7.6 ± 4.8 vs. 13.7 ± 8.8). Antral follicle count significantly correlated with most prestimulation and poststimulation IVF variables. Threshold analysis demonstrated a lower pregnancy rate (23.5% vs. 57.6%) and a higher cancellation rate (41% vs. 6.4%) associated with having four or fewer antral follicles. Conclusion(s): The basal antral follicle count identified patients who responded poorly to IVF stimulation. Having four or fewer antral follicles was associated with a high cancellation rate (41%) and, in patients without a cancelled cycle, a low pregnancy rate (23%). However, no antral follicle count absolutely predicted pregnancy or cycle cancellation.