Sensitive RIA measures testosterone concentrations in prepubertal and pubertal children comparable to tandem mass spectrometry

Background. Immunoassays have been criticized for poor accuracy at low testosterone concentrations. Mass spectrometry (MS) has been proposed as the only reliable method for testosterone determination. The aim of this study was to compare a sensitive testosterone radioimmunoassay (RIA) with results from different MS. Methods. We compared testosterone concentrations determined by a sensitive testosterone RIA, lower limit of detection 0.03 nmol/L and limit of quantitation 0.1 nmol/L, with four tandem MS that were included in an international external quality assessment program for laboratory medicine. We also compared the morning concentrations of testosterone in girls and boys at different pubertal stages, using results from the RIA, with reported values determined by LC-MS/MS, developed for androgen determination in children. Results. The mean (SD), concentrations were similar between RIA and MS: 1.5 (0.3) and 1.4 (0.4) in the child/women range (0.8-2.6 nmol/L) and 16.0 (3.7) and 17.8 (4.5) nmol/L for the adult male range (10.1-30.0 nmol/L), respectively. The ratio between RIA and MS versus results from mean values of the four MS methods was 1.0 (0.18); 1.1 (0.18) for child/women concentrations and 0.9 (0.13) for male testosterone concentrations. Furthermore, compared to the pediatric reference values determined by LC-MS/MS, the sensitive testosterone RIA delivered similar testosterone values across the different pubertal stages. Conclusions. The comparison between different tandem MS methods and a sensitive testosterone RIA illustrates that there are immunoassays that deliver clinically useful information in prepubertal and pubertal children.