Testosterone Measured with an Automatic Immunoassay Compares Reasonably Well to Results Obtained by LC-MS/MS.

Background: Previous studies have reported problems measuring testosterone with immunological assays. Here we explore an automatic second generation immunoassay compared to a LC-MS/MS method.
Methods: We collected blood samples from 76 women and measured testosterone, progesterone, gender hormonebinding globulin (SHBG), and albumin employing Cobas e601/c501. Testosterone, androstenedione (andro), dehydroepiandrosterone sulphate (DHEAS), and 17-hydroxyprogesterone (17-OHP) concentrations were measured employing LC-MS/MS. We evaluated the difference between testosterone measured by the two methods and examined the potential interference from the selected steroids and bindings proteins.
Results: Testosterone concentrations measured by the two methods yielded: Cobas e601 = 1.240 x (LC-MS/MS) - 0.197, r = 0.84, for testosterone concentrations between 0.22 - 4.9 nmol/L. A positive correlation was observed for the difference between results obtained by the two methods and the sample concentration of DHEAS and andro: Diff (Cobas e601 - LC-MS/MS) = 0.116 x DHEAS - 0.396, r = 0.84 and Diff (Cobas e601 - LC-MS/MS) = 0.08 andro - 0.380, r = 0.58. No statistically significant interference was observed for progesterone, 17-OHP, SHBG, and albumin.
Conclusions: We report significant differences between testosterone measurements employing an automatic second generation immunoassay and LC-MS/MS. The difference can be correlated with the measured concentrations of DHEAS and andro, and its magnitude is judged to be of limited clinical relevance. Thus we judge that the automatic second generation immunoassay can be used for routine measurement of testosterone.