Comparative study of midnight salivary cortisol measurements: Impact on clinical interpretation

Despite the other first-line tests for investigating hypercortisolism, saliva analysis is rather easy to collect and it does not require hospitalization, which represents an important economic issue and better comfort for the patient. Indeed, the patient can easily collect the saliva sample from home by using a specific device, assuring the good conditions of the specimen until arrival to the laboratory. Also, in restricted and crucial categories (renal failure patients and pregnant women), the midnight salivary cortisol proved to be more robust and reliable analysis. In this work, we compared the immuno-analysis methods commercialized for the salivary cortisol measurements, the Elecsys Cortisol II assay on COBAS 8000 (ROCHE Diagnostics) with a LC-MS/MS method previously developed in our laboratory for urinary free cortisol measurements. Analytical performance (imprecision, linearity and limit of detection) was evaluated. Particularly the sensitivity, the effect of the matrix and the null interferences with other steroids compared to immuno-analysis were taken into account. Consequently, a comparison of approximately 60 saliva samples from patients with suspicious of Cushing syndrome was performed on the two systems. Interestingly, the comparison allowed to show that the LC-MS/MS method presented similar results for the studied range of concentrations with lower limit of detection (< 1.5 nmol/L) and no particular problems of saliva matrix effects. Results from external quality controls showed a lower signal of our LC-MS/MS system compared to the pairwise, probably due to the extraction method of the different LC-MS/MS systems.