Endocrine Abstracts

The diagnosis of Cushing’s Syndrome (CS) requires demonstration of excess circulating cortisol. Measurement of late night salivary cortisol (LNSF) has been advocated as a simple, non-invasive and reliable outpatient diagnostic tool for patients with suspected CS but the usefulness of its metabolite cortisone (LNSE) remains unclear. LNSE levels are approximately six times higher than LNSF in saliva due to the rapid action of 11β-hydroxysteroid dehydrogenase type 2 (11βHSD), and have been shown to have a linear relationship with serum free cortisol. We hypothesised that LNSE concentrations may therefore correlate with degree of metabolic derangement resulting from hypercortisolemia. In this study, we investigated the sensitivity of LNSF and LNSE (measured using liquid chromatography-mass spectrometry, LCMS) as compared to the traditional urine free cortisol (UFC) and overnight dexamethasone suppression test (ONDST) in patients with confirmed CS. We also studied any association between LNSF or LNSE and metabolic parameters affected by cortisol excess including HbA1C, ALT and blood pressure (BP). Eighteen patients (15 female) with confirmed CS attending Beaumont Hospital were included. Sixteen patients had ACTH-dependent CS. Median age was 34 years (range 7 – 65). Both LNSF and LNSE were strongly correlated with UFC (R 0.53, P=0.03 and R 0.69, P=0.01 respectively). Sensitivity of ONDST was 100% (data for 17 patients). UFC was measured in 17 patients – median number of samples 1.5 (range1–6) with a sensitivity of 92%. One patient with confirmed CS had four measurements for UFC, all of which were negative. Median number of LNSF and LNSE samples measured was 3 (range 1–18). LNSF had a sensitivity of 91%, while LNSE had a sensitivity of 84%. Seventeen of 18 LNSF samples were positive in the patient with four negative UFCs. Serial LNSF and LNSE identified four patients with cyclical Cushing’s Disease. Six patients (33%) had diabetes mellitus, 11 (61%) were hypertensive, and 11 (61%) were obese. On multivariate analysis, LNSF was not significantly associated with HbA1C, ALT or BP, while LNSE was negatively associated with HbA1C (P=0.04) but not with BP or ALT. When the ratio of LNSF to LNSE was studied, there was no significant effect on metabolic parameters. Late night salivary cortisol is more sensitive than salivary cortisone as a diagnostic test for CS and may identify cases of cyclical ACTH secretion. We did not identify an association between LNSF and metabolic parameters influenced by hypercortisolemia. The negative association between LNSE and HbA1C requires further elucidation.