A simple liquid chromatography-tandem mass spectrometry method for urinary free cortisol analysis: suitable for routine purpose

Background: The best index of adrenal dysfunction is urinary free cortisol (UFC) measurements performed using a 24-h urine collection. This measurement is also useful in the investigation of Cushing's syndrome. In this paper, we report a simple and selective method for the analysis of UFC by liquid chromatography-tandem mass spectrometry (LC-MS/MS) suitable for use in a high-volume clinical laboratory routine. The results were compared to those obtained using a commercial immunoassay method used in our laboratory. Methods: Urine samples containing 50 ng of internal standard (Cortisol-9,11,12,12-d(4)) were deproteinized using centrifugal filters with a molecular weight 10,000 Da cut-off and injected on a reversed phase column. Cortisol was analyzed in highly selective reaction monitoring in positive atmospheric pressure chemical ionization mode, at a resolution of 0.4 amu full width half maximum, and following the transitions related to the precursor 363.2 for cortisol and 367.2 for deuterated cortisol. The method validation included analysis of precision, linearity, sensitivity, recovery and interference from structurally similar steroids. UFC from 230 subjects was measured using LC-MS/MS and electrochemiluminescence immunoassay (ECLIA) methods. Results: The calibration curves exhibited linearity and reproducibility in the range 7-10,000 nmol/L. Total imprecision was lower than 10\%. The limit of detection and limit of quantification were 2 and 7 nmol/L, respectively. Mean recovery was higher than 90\%. Structurally similar steroids do not interfere in the proposed method, but cause a significant change in the ECLIA results. Cortisol values obtained using the ECLIA method were always higher than those obtained using the LC-MS/MS method, with the bias directly proportional to cortisol concentrations. The reference values calculated using 180 normal subjects were 11-70 mu g/day. Conclusions: The proposed method is sensitive, simple, free from interferences and reliable for routine use. Clin Chem Lab Med 2010;48:1433-7.