Abstract
Steroid analysis in clinical laboratories is dominated by immunoassays (IAs) that have a high sample turnover but are inherently limited in trueness, precision, and sensitivity. Liquid chromatography coupled to mass spectrometry (LC-MS/MS) has proved to be a far more capable tool, delivering better sensitivity, specificity, and the possibility of parallel analysis of multiple steroids and metabolites, providing the endocrinologist with more reliable and comprehensive diagnostic information. An LC-MS/MS assay with gradient elution over less than eight minutes and a one-step sample preparation combining protein precipitation with phospholipid removal of off-line solid-phase extraction was developed and validated. It allowed the quantification of 11-deoxycorticosterone (11-DOC), 11-deoxycortisol (11-DF), 17-OH-progesterone (17P), 21-deoxycortisol (21-DF), androstenedione (ANDRO), aldosterone (ALDO), corticosterone (CC), cortisol (CL), cortisone (CN), dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEAS), dihydrotestosterone (DHT), estradiol (E2), progesterone (PROG), and testosterone (TES) in human serum. Interday imprecision was generally better than 15%, trueness was proven by recovery experiments with ISO 17034-certified reference materials, proficiency testing (UK NEQAS), and measuring serum reference standards. In-house comparison against IVD-CE-certified immunoassays (IA) for 17P, ANDRO, CL, DHEAS, E2, PROG, and TES was conducted by assessing leftover routine patient samples and purpose-built patient serum pools. None of the compared routine IAs were meeting the standards of the LC-MS/MS. Insufficient overall comparability was found for ANDRO and 17P (mean bias > +65%). Accuracy limitations at lower concentrations were present in IAs for PROG, E2, and TES.