20-and 20-dihydrocortisone may interfere in LC-MS/MS determination of cortisol in saliva and urine

Background LC-MS/MS methods offer high selectivity in cortisol determinations. However, endogenous steroid metabolites may still interfere and compromise the results, for example in the diagnosis of Cushing's syndrome. Erroneously elevated cortisol may, in particular, be misleading at the low concentrations found in salivary samples obtained at late night and after dexamethasone suppression. Methods Interferences in our LC-MS/MS method used for determination of cortisol in saliva and urine were identified by comparing their retention times and mass spectra with those of pure candidate substances. The chromatographic conditions used in our LC-MS/MS method, including column and mobile phase gradient, were varied in order to separate the target compound from the interferences. Results Two interferences, which were co-eluting or eluting close to cortisol in our original method, were successfully separated from cortisol by adjustment of the chromatographic conditions. These interferences were found in both urine and saliva and were identified as the two endogenous cortisol isomers 20- and 20-dihydrocortisone. The isomers share molecular mass and mass spectrometric fragmentation pattern with cortisol using electrospray ionization in the positive-ion mode. Both give rise to the transitions m/z 363.1>121.1, 363.1>115.1 and 363.1>97.1. In our original LC-MS/MS setup, the 20-dihydrocortisone co-eluted with cortisol in the chromatography step resulting in false high determinations. Conclusions Cortisol determination by LC-MS/MS may suffer from erroneously elevated results unless 20- and 20-dihydrocortisone are chromatographically separated from cortisol.