Purpose: The purpose of this study is to investigate whether individual response of anti-Mullerian hormone (AMH) to gonadotropin-releasing hormone (GnRH) treatment is associated with difference in ovarian stimulation outcomes.
Methods: The retrospective study included 1058 non-polycystic ovary syndrome (PCOS) women undergoing long agonist protocol in a single in vitro fertilization unit from January 1, 2016, through December 31, 2016. Patients were grouped according to AMH changes from day 3 to the day of stimulation (group 1, change < 1 ng/ml, n = 714; group 2, decrease ≥ 1 ng/ml, n = 143; group 3, increase ≥ 1 ng/ml, n = 201). A generalized linear model including Poisson distribution and log link function was used to evaluate the association between AMH response and the number of oocytes retrieved.
Results: Group 2 was characterized by higher basal AMH level and increased AMH to AFC ratio in comparison with two other groups. However, the number of oocytes and ovarian sensitivity index in group 2 was significantly lower than group 3. Adjusted for age, BMI, ovarian reserve markers, and stimulation parameters, the population marginal means (95% confidence interval) of oocyte number in groups 1 through 3 were 9.51 (9.17, 9.86), 8.04 (7.54, 8.58), and 10.65 (10.15, 11.18), respectively. For patients from group 2 and group 3, basal AMH is no longer significantly associated with oocyte yield.
Conclusions: AMH change in response to GnRH agonist varies among individuals; for those undergoing significant changes in AMH following GnRH agonist treatment, basal AMH may not be a reliable marker for ovarian response in long agonist protocol.
Keywords: Anti-Mullerian hormone; Controlled ovarian hyperstimulation; GnRH agonist; Long agonist protocol; Pituitary downregulation.