Multicenter Evaluation of a New, Fully Automated Androstenedione Electrochemiluminescence Immunoassay: Precision Analysis, Method Comparison, and Determination of Reference Ranges.

Abstract

Androstenedione (ASD) levels can aid diagnosis of hyperandrogenism together with other clinical/laboratory findings. We evaluated performance of the new, automated Elecsys® ASD assay vs an ASD isotope dilution-liquid chromatography-tandem mass spectrometry (ID-LC-MS/MS) reference measurement procedure and determined reference ranges. Repeatability/intermediate precision were assessed using 3 control levels and 5 human serum pools (n = 75 each; Clinical and Laboratory Standards Institute EP05-A3). Method comparisons vs commercially available immunoassays [IMMULITE ASD (Siemens) and LIAISON ASD (DiaSorin)] and an ID-LC-MS/MS measurement procedure method were conducted using 421 serum samples; Passing-Bablok regression and Pearson's correlation coefficients were calculated. Reference ranges and distribution of values associated with polycystic ovary syndrome (PCOS) were determined in five clinical cohorts using samples from several sites/vendors. Repeatability/intermediate precision coefficients of variation across all sites were 2.01% to 3.91% and 2.43% to 4.30%, respectively (mean ASD: 7.80-34.7 nmol/L). The Elecsys ASD assay showed poor agreement with IMMULITE ASD (slope = 0.459; r = 0.856; n = 320), fair agreement with LIAISON ASD (slope = 0.625; r = 0.984; n = 327), and very good agreement with ID-LC-MS/MS (slope = 1.040; r = 0.996; n = 332). Reference ranges (2.5th-97.5th percentiles) were: children (≤8 years; n = 140), <0.525 to 1.81 nmol/L; males (≥18 years; n = 138), 0.979 to 5.32 nmol/L; and postmenopausal females (n = 140), 0.654 to 3.74 nmol/L. Reference range (5th-95th percentiles) for females with fertile cycle (≥18 years; n = 84) was 1.71 to 4.58 nmol/L. The distribution of values (2.5th-97.5th percentiles) in females with PCOS (n = 125) was 2.26 to 12.1 nmol/L. Elecsys ASD assay demonstrated excellent precision and very good agreement with ID-LC-MS/MS. Reference ranges were established to support results interpretation in routine practice.