Abstract
OBJECTIVE: Autonomous cortisol secretion (ACS) is a condition with ACTH-independent cortisol overproduction from adrenal incidentalomas (AI) or adrenal hyperplasia. The hypercortisolism is often mild, and most patients lack typical clinical features of overt Cushing's syndrome (CS). ACS is not well defined and diagnostic tests lack validation. METHODS: Retrospective study of 165 patients with AI evaluated clinically and by assay of morning plasma ACTH, late-night saliva cortisol, serum DHEA sulphate (DHEAS), 24-h urine-free cortisol, and cortisol after dexamethasone suppression. RESULTS: Patients with AI (n = 165) were diagnosed as non-functioning incidentalomas (NFI) (n = 82) or ACS (n = 83) according to current European guidelines. Late-night saliva cortisol discriminated poorly between NFI and ACS, showing a high rate of false-positive (23/63) and false-negative (38/69) results. The conventional low-dose dexamethasone suppression test (LDDST) did not improve the diagnostic specificity, compared with the 1 mg overnight DST. Receiver operating characteristic curve analysis of DHEAS in the two cohorts demonstrated an area under the curve of 0.76 (P < 0.01) with a sensitivity for ACS of 58% and a specificity of 80% using the recommended cutoff at 1.04 µmol/L (40 µg/dL). CONCLUSION: We here demonstrate in a large retrospective cohort of incidentaloma patients, that neither DHEAS, late-night saliva cortisol nor 24-h urine free cortisol are useful to discriminate between non-functioning adrenal incidentalomas and ACS. The conventional LDDST do not add further information compared with the 1 mg overnight DST. Alternative biomarkers are needed to improve the diagnostic workup of ACS.