Reference ranges of late-night salivary cortisol and cortisone measured by LC–MS/MS and accuracy for the diagnosis of Cushing’s syndrome
F. Ponzetto,F. Settanni,M. Parasiliti-Caprino,F. Rumbolo,A. Nonnato,M. Ricciardo,E. Amante,G. Priolo,S. Vitali,L. Anfossi,E. Arvat,E. Ghigo,R. Giordano,G. Mengozzi
DOI: https://doi.org/10.1007/s40618-020-01388-1
2020-08-09
Journal of Endocrinological Investigation
Abstract:PurposeInternational guidelines recommend salivary cortisol for the diagnosis of Cushing’s syndrome. Despite mass spectrometry-based assays are considered the analytical gold-standard, there is still the need to define reference intervals and diagnostic accuracy of such methodology.Methods100 healthy volunteers and 50 consecutive patients were enrolled to compare LC–MS/MS and electrochemiluminescence assay for the determination of late-night salivary cortisol and cortisone. Moreover, we aimed to determine reference intervals of salivary steroids in a population of healthy individuals and diagnostic accuracy in patients with suspected hypercortisolism and in a population including also healthy individuals.ResultsMethod comparison highlighted a positive bias (51.8%) of immunoassay over LC–MS/MS. Reference intervals of salivary cortisol (0.17–0.97 µg/L), cortisone (0.84–4.85 µg/L) and ratio (0.08–0.30) were obtained. The most accurate thresholds of salivary cortisol for the diagnosis of hypercortisolism were 1.15 µg/L in the population with suspected hypercortisolism (AUC 1) and 1.30 µg/L in the population including also healthy individuals (AUC 1). Cut-off values of salivary cortisone (7.23 µg/L; Se 92.9%, Sp 97.2%, AUC 0.960 and Se 92.9%, Sp 99.1%, AUC 0.985 in suspected hypercortisolism and in overall population, respectively) and cortisol-to-cortisone ratio (0.20; Se 85.7%, Sp 80.6%, AUC 0.820 and Se 85.7%, Sp 85.5%, AUC 0.855 in suspected hypercortisolism and in overall population, respectively) were accurate and similar in both populations.ConclusionLC–MS/MS is the most accurate analytical platform for measuring salivary steroids. Obtained reference intervals are coherent with previously published data and diagnostic accuracy for diagnosis of overt hypercortisolism proved highly satisfactory.
endocrinology & metabolism