Retrospective cohort study: AMH is the best ovarian reserve markers in predicting ovarian response but has unfavorable value in predicting clinical pregnancy in GnRH antagonist protocol

Various ovarian reserve markers have been used to predict ovarian response and pregnancy. However, concerning Chinese population, fewer trials have been performed using the combined ovarian reserve markers to predict ovarian response and pregnancy in GnRH antagonist protocols. Data from a total of 373 patients' in vitro fertilization cycles using GnRH antagonist protocol was retrospectively included. According to our center's daily practice, circulating follicle-stimulating hormone, luteinizing hormone, and estradiol (E2) were tested on menstrual cycle day 2-4 or hCG trigger day, and the concentration of AMH was determined despite of menstrual cycle. The antral follicle count (AFC) was assessed by transvaginal ultrasound on day 2-4 of menstrual cycle. Different ovarian response was defined as 0-4 and 5-15 and > 15 oocyte retrieved for low and normal and high ovarian response, respectively. Gestational sac with fetal heartbeat detected by ultrasound was considered as clinical pregnancy. Serum AMH levels was the most accurate marker in predicting ovarian response [area under the receiver operating characteristic (ROC) curve = 0.767]. Significant difference was found in age between non-clinical pregnancy and clinical pregnancy groups (p < 0.001). Our data demonstrated that the circulating AMH despite of menstrual cycle was preferable in prediction of oocyte retrieved outcome during GnRH antagonist protocol than age, AFC and the other currently used hormone markers. Furthermore, age is the only marker in predicting clinical pregnancy.