Determination of urinary free cortisol by liquid chromatography-tandem mass spectrometry.

Measurement of urinary free cortisol is clinically important in the diagnosis of Cushing's syndrome. While liquid chromatography (LC) with UV detection provides much better specificity than immunologic methods, certain drugs cause interference. Detection by mass spectrometry (MS) is a potentially superior method. Our analysis utilizes 1 mL urine spiked with 6-alpha-methylprednisolone as internal standard. The samples were extracted with dichlormethane and the extract was washed, evaporated to dryness and analyzed by LC-MS/MS operating in the negative mode after separation on a reversed-phase C18 column. The calibration curves for analysis of urinary cortisol exhibited consistent linearity and reproducibility in the range of 10-400 nmol/L. Inter-assay CVs were 4.0-7.6%, at mean concentrations of 21-153 nmol/L. The detection limit was 1 nmol/L (signal-to-noise ratio=3). The mean recovery of cortisol added to urine ranged from 67% to 87% and that of the internal standard from 71% to 76%. The regression equation for the LC-MS/MS (x) and HPLC (y) methods was: y=1.095x+8.0 (r=0.996; n=111). Drugs known to interfere with UV detection did not cause problems here. The sensitivity and specificity of the MS/MS method for urinary free cortisol offer advantages over HPLC with UV detection by eliminating drug interference. The higher equipment costs in comparison with HPLC methods using UV detection are balanced by higher throughput, thanks to shorter chromatographic run times.