Serum anti-müllerian hormone (AMH) measurement: inter-assay agreement and temperature stability

Abstract

Two different immunoassay systems are currently in use to measure serum AMH, one produced by Diagnostic Systems Laboratories (DSL) and one produced by Immunotech (both Beckman Coulter, Inc companies). Previously, the Immunotech system was more widely used in publications and is currently in use outside the United States. However, the DSL system is currently the only commercially available system in the United States. Limited published data is available which compares the two systems. There were two objectives to this study: 1) better characterize the relationship between the Immunotech and DSL systems; 2) test the stability of serum specimens over 48 hours at different temperatures. ELISA analysis of serum samples. Objective 1) AMH measurement was performed on samples from serum from 38 different donors using both the Immunotech and DSL AMH assays. Objective 2) Multiple AMH measurements were conducted using the DSL assay and serum from seven different donors. Specimens were separated into different aliquots, the first immediately frozen at −20°C and the subsequent samples frozen after 4, 8, 24, and 48 hours at 4°C and room temperature (22°C). Data from the 38 serum samples comparing the two AMH assays revealed a relationship defined by the line Immunotech = (1.5) DSL + 0.7 (ng/ml) with good agreement (R2 = 0.98). The mean AMH value ratio of samples incubated for 48 hours at 4°C to samples immediately frozen, was 1.01 with a 95% CI of (0.94–1.08). After 48 hours at 22°C, the mean AMH value ratio to immediately frozen samples was 1.09 with a 95% CI of (0.98–1.20). Values obtained using the DSL AMH immunoassay can be linearly converted to Immunotech AMH values with some degree of variation. DSL AMH measurement is stable for at least 48 hours at 4°C, with values likely to differ by less than eight percent of values from immediately frozen samples. With samples at room temperature for 48 hours, AMH values were likely to differ by less than twenty percent of values from immediately frozen samples.