Late night salivary cortisol measurements have been increasingly used as an initial diagnostic test for the evaluation of patients with a clinical suspicion of Cushing syndrome (CS). Published studies include varying numbers of cases and controls and importantly various assay methods (usually immunoassays), as well as various methods of generating normative values.
Materials and methods: We examined the diagnostic utility of salivary cortisol measurement in 258 patients evaluated for the possibility of CS because of various clinical conditions using liquid chromatography/tandem mass spectrometry method (LCMS/MS). CS was confirmed in 56 patients (21.7%) and excluded in 202 (78.3%) patients at the time of analysis. The patient population with CS included 39 (69.6%) with pituitary CS, 8 (14.9%) with ectopic CS, 9 (16.1%) with adrenal CS. Salivary specimen were collected at 1100 h using Sarstedt Salivette and LCMS/MS was used for analysis. Normal values <100 ng/dl between 1100 h and midnight were previously established using two standard deviations from the mean of a healthy population.
Results: Late night salivary cortisol was elevated (>100 ng/dl) in 39/56 patients providing a sensitivity of 69.9% for the diagnosis of CS and elevated in 20/202 patients who were found not to have CS, specificity of 90.1%. When only patients with pituitary or ectopic CS were considered the sensitivity was 76.6% with specificity 90%. Using ROC to calculate the cut-off providing optimal sensitivity and specificity a late night salivary cortisol of >61 ng/dl provided a sensitivity of 82.1% and a specificity of 78.2%.
Conclusion: Analysis of data at one referral institution showed the somewhat limited sensitivity of late night salivary collection of cortisol. Therefore in patients with a high suspicion of CS and negative tests results other complementary tests for diagnosis of CS should be used and periodically repeating of testing should be implemented.
Prague, Czech Republic
24 - 28 Apr 2010
European Society of Endocrinology