Ovarian hyperstimulation syndrome (OHSS) is a serious and sometimes a life-threatening complication of controlled ovarian stimulation (COS). It has proved difficult to accurately predict OHSS in any particular in vitro fertilization (IVF) treatment cycle. Reportedly, serum levels of anti-müllerian hormone (AMH) are closely related to the ovarian response or ovarian reserve during IVF cycles. The goal of this study was to determine how reliably the basal serum AMH level predicted OHSS in a prospective cohort of 262 IVF cycles. Participating women followed a long or short gonadotropin-releasing hormone (GnRH) agonist protocol, and the luteal phase was supported by human chorionic gonadotropin (hCG). Only micronized progesterone was given after the serum estradiol level exceeded 3000 pg/mL or more than 15 oocytes were retrieved.
Moderate to severe OHSS was noted in 8% of patients, only one of whom followed the short GnRH agonist protocol. The overall rate of clinical pregnancy was 41%. The basal serum AMH level predicted OHSS more accurately than either age or body mass index. Its sensitivity was 90.5%, and its specificity was 81.3%. Logistic regression analysis identified basal serum AMH (odds ratio, 1.7856, P = 0.0003) and the serum estradiol level on the day of hCG treatment (odds ratio, 1.0005, P = 0.0455) as significant predictors of OHSS. On logistic regression analysis, the only significant factor predicting clinical pregnancy was age (odds ratio, 0.9346, P = 0.049). Basal serum AMH did not effectively predict clinical pregnancy.
Estimating the basal serum AMH reliably predicts OHSS and may be used to ensure that patients with high levels receive a mild, patient-friendly protocol for ovarian stimulation.