Comparison of antral follicle count and serum anti Mullerian hormone level for determination of gonadotropin dosing in in-vitro fertilization: randomized trial

Objective To compare the proportion of women achieving a desired ovarian response following ovarian stimulation when gonadotropin dosing was determined based on antral follicle count (AFC) vs serum anti-Mullerian hormone (AMH) level, in women undergoing in-vitro fertilization (IVF) using the gonadotropin-releasing hormone (GnRH) antagonist protocol. Methods This was a randomized double-blind trial carried out in a university-affiliated assisted reproduction unit. A total of 200 women undergoing their first IVF cycle using the GnRH-antagonist protocol between April 2016 and February 2018 were randomized to determination of gonadotropin dosing based on either AFC or serum AMH level measured in the pretreatment cycle 1 month before the IVF cycle. Patients underwent IVF as per our center's standard protocol. The proportion of subjects achieving a desired ovarian response, defined as retrieval of six to 14 oocytes, was compared between the two study arms. Subgroup analysis of patients with baseline AFC > 5 and those with baseline AFC <= 5 was performed. Concordance in AFC and AMH categorization between the pretreatment cycle and the ovarian-stimulation cycle was assessed using Cohen's kappa (kappa). Results There was no significant difference in the proportion of patients achieving a desired ovarian response between the AFC (54%) and AMH (49%) groups (P = 0.479). The median number of oocytes retrieved was nine vs seven (P = 0.070), and the median follicular output rate was 0.54 vs 0.55 (P = 0.764) in the AFC and AMH groups, respectively. Similar findings were observed on subgroup analysis of subjects with AFC <= 5 and AFC > 5 at the start of ovarian stimulation (P > 0.05 for all comparisons). There was moderate concordance between AFC and AMH measured in the pretreatment cycle and the stimulation cycle (kappa = 0.478 and 0.587, respectively). Conclusion The proportion of women achieving a desired ovarian response following ovarian stimulation using the GnRH-antagonist protocol is similar when the gonadotropin-dosing algorithm used is based on AFC or serum AMH level. Copyright (c) 2019 ISUOG. Published by John Wiley & Sons Ltd.