Hair Glucocorticoids as a Biomarker for Endogenous Cushing's Syndrome: Validation in Two Independent Cohorts

Abstract

Background/Aims: The current diagnostic workup of Cushing's syndrome (CS) requires various tests which only capture short-term cortisol exposure, whereas patients with endogenous CS generally have elevated cortisol levels over longer periods of time. Scalp hair assessment has emerged as a convenient test in capturing glucocorticoid concentrations over long periods of time. The aim of this multicenter, multinational, prospective, case-control study was to evaluate the diagnostic efficacy of scalp hair glucocorticoids in screening of endogenous CS. Methods: We assessed the diagnostic performances of hair cortisol (HairF), hair cortisone (HairE), and the sum of both (sumHairF+E), as measured by a state-of-the-art LC-MS/MS technique, in untreated patients with confirmed endogenous CS (n = 89) as well as in community controls (n = 295) from the population-based Lifelines cohort study. Results: Both glucocorticoids were significantly elevated in CS patients when compared to controls. A high diagnostic efficacy was found for HairF (area under the curve 0.87 [95% CI: 0.83-0.92]), HairE (0.93 [0.89- 0.96]), and sumHairF+E (0.92 [0.88-0.96]) (all p < 0.001). The participants were accurately classified at the optimal cutoff threshold in 86% of the cases (81% sensitivity, 88% specificity, and 94% negative predictive value [NPV]) by HairF, in 90% of the cases (87% sensitivity, 90% specificity, and 96% NPV) by HairE, and in 87% of the cases (86% sensitivity, 88% specificity, and 95% NPV) by the sumHairF+E. HairE was shown to be the most accurate in differentiating CS patients.