Endocrine Abstracts

Context: Late-night salivary cortisol (LNSC) measurement has been promoted as an ideal screening test for the diagnosis of Cushing’s syndrome (CS). However, its performance using commercially available assays has not been widely evaluated and limited data are available on its use in population with chronic medical conditions.

Aim: To compare the diagnostic performance of LNSC (routine use) in patients with CS and in a patients’s group with different medical conditions.

Methods: We studied 281 subjects: 117 normal weight healthy volunteers (HV), 47 patients with active CS, 27 patients with uncontrolled diabetes, 61 obese subjects and 29 adrenal incidentaloma. Subjects provided to bedtime saliva samples collected (2300 h) for cortisol measurements by commercially chemiluminescence immunoassay (CLIA, Access Beckman Coulter) and in subjects with abnormal value in CLIA salivary cortisol was measured also by liquid chromatography mass spectrometry (LCMS–MS).

Results: The LNSC concentrations were significantly higher (P<0.001) in CS (1.61±0.83 μ/dl; range: 0.63–4.33) compared with healthy subjects (0.27±0.4 μ/dl; range: 0.018–0.62). The optimal LNSC cut-off value derived from ROC analysis for the differentiation between patients with and without CS was achieved at the level of 0.58 μ/dl (Cushing vs HV; SE 96.7%, SP 96.9%). However, this cut-off showed a lack of specificity when used in obese subjects (SP), and in diabetic patients (SP). Using LCMS-MS, the diagnostic performance in obese and diabetic subjects was increase.

Conclusions: This study confirms the utility of CSN in CS screening, even using a routinary method as CLIA. However, the study underlines the necessity for every single laboratory to revaluate the cut-off, especially in those conditions that can produce an activation of the hypothalamic–pituitary–adrenal axis.