Prediction of Different Ovarian Responses Using Anti-Mullerian Hormone Following a Long Agonist Treatment Protocol for Ivf Publisher Pubmed



Heidar Z¹ ; Bakhtiyari M^{2, 3} ; Mirzamoradi M¹ ; Zadehmodarres S¹ ; Sarfjoo FS⁴ ; Mansournia MA³ Show Affiliations
Source: Journal of Endocrinological Investigation Published:2015

Abstract

Objective: The purpose of this study was to predict the poor and excessive ovarian response using anti-Mullerian hormone (AMH) levels following a long agonist protocol in IVF candidates. Research design and methods: Through a prospective cohort study, the type of relationship and appropriate scale for AMH were determined using the fractional polynomial regression. To determine the effect of AMH on the outcomes of ovarian stimulation and different ovarian responses, the multi-nominal and negative binomial regression models were fitted using backward stepwise method. The ovarian response of study subject who entered a standard long-term treatment cycle with GnRH agonist was evaluated using prediction model, separately and in combined models with (ROC) curves. Results: The use of standard long-term treatments with GnRH agonist led to positive pregnancy test results in 30 % of treated patients. With each unit increase in the log of AMH, the odds ratio of having poor response compared to normal response decreases by 64 % (OR 0.36, 95 % CI 0.19-0.68). Also the results of negative binomial regression model indicated that for one unit increase in the log of AMH blood levels, the odds of releasing an oocyte increased 24 % (OR 1.24, 95 % CI 1.14-1.35). The optimal cut-off points of AMH for predicting excessive and poor ovarian responses were 3.4 and 1.2 ng/ml, respectively, with area under curves of 0.69 (0.60-0.77) and 0.76 (0.66-0.86), respectively. Conclusion: By considering the age of the patient undergoing infertility treatment as a variable affecting ovulation, use of AMH levels showed to be a good test to discriminate between different ovarian responses. © 2015 Italian Society of Endocrinology (SIE).