Prediction of ovarian response using the automated Elecsys anti-Müllerian hormone assay in gonadotrophin-releasing hormone antagonist cycles.

RESEARCH QUESTION: What is the capability of serum anti-Müllerian hormone (AMH) measured using the automated Elecsys® AMH immunoassay to (Roche Diagnostics International Ltd) determine ovarian response after fertility treatment? DESIGN: Single-centre, retrospective, observational, cohort study including women undergoing ovarian stimulation. Serum AMH concentrations were determined using the Elecsys AMH immunoassay based on one blood sample drawn 6 months or less before treatment. Stimulation was conducted in accordance with a gonadotrophin-releasing hormone (GnRH) antagonist protocol. Patients were divided into four ovarian response categories based on their oocyte yield: low (0-3), suboptimal (4-9), optimal (10-15) and high (>15). Areas under the curve were calculated for each ovarian response group. RESULTS: Overall, 1248 patients were enrolled. The AMH concentration had a strong positive correlation with oocyte yield (Spearman's rho = 0.74, P < 0.001). Areas under the curve (95% CI) for AMH predicting ovarian response were 0.85 (0.83 to 0.88) for low and 0.89 (0.87 to 0.91) for high response. Optimal serum AMH cut-offs for predicting a low and high response using the Elecsys AMH immunoassay were 6.4 pmol/l (0.89 ng/ml) and 14.2 pmol/l (1.99 ng/ml), respectively. Multivariable regression analysis showed that 47% (R2 = 0.470) of variation in ovarian response could be attributed to AMH alone, increasing to 50.9% (R2 = 0.509) with the addition of age, body weight, and total dose of gonadotrophin. CONCLUSION: Ovarian response and oocyte yield after stimulation in a GnRH antagonist cycle can be predicted with high accuracy using a single determination of serum AMH before ovarian stimulation.