SAT-166 Advantage and Trustworthy of Cortisol and Dexamethasone Evaluation in Different Biological Matrices in Patients with Adrenal Masses

Biochemical function of adrenal masses is currently based on 1mg post-overnight dexamethasone suppression test (pDST). Several approaches are recently developed, in order to reduce false positive/negative samples, only in retrospective series. They are based on the correlation of some different parameters, i.e. late-night salivary cortisol (LNSC) vs serum and salivary cortisol pDST; LNSC vs serum and salivary cortisol and serum dexamethasone pDST; LNSC and cortisone vs serum cortisol and salivary cortisol and cortisone pDST. Although these findings offer a better diagnostic performance, several conditions are still disappointed. No information is traceable about the harvest time of diurnal salivary and serum samples and no study include neither the levels of salivary nor urinary dexamethasone pDST. Aim of our study is to combine all these strategies in order to avoid the underestimated biases and obtain more precise information about the true “cortisol condition” of the patients. To reach this purpose we assess both cortisol and dexamethasone concentrations in several samples: saliva at 11PM before the drug administration, diurnal saliva and serum at 8AM and also the urine collection from 11PM to 8AM. Analytes levels are measured using a validated liquid chromatography-tandem mass spectrometry method. In this study we included 20 subjects without morphological adrenal alteration (MRI assessment), dyslipidemia, hypertension and impaired glucose tolerance (healthy controls) and 20 patients with adrenal incidentaloma showing different cortisol levels ranging from normal to ACTH-independent hypercortisolism. In both series, LNSC were similar to salivary cortisol pDST, even if they were greater in the patients with adrenal incidentalomas and subclinical cortisol secretion. Serum dexamethasone levels were in reference ranges, while salivary and urinary dexamethasone found in these matrices require additional sample numbers in order to establish appropriate cut-offs. Our preliminary results suggest that the combination of these findings could represent an improvement to assess the individual cortisol status.