Evaluation of serum and salivary cortisol responses to Synacthen stimulation

Several studies have demonstrated the assay-dependence of cortisol measurement and the potential impact this could have on the interpretation of the Synacthen test. Prompted by clinical concern that hypoadrenalism was being over-diagnosed, this work set out to determine assay-specific serum cortisol and salivary cortisol responses to Synacthen stimulation in healthy volunteers. It begins with an overview of cortisol assays in serum and saliva from their inception 50 years ago to the present day; highlighting the limitations of serum total cortisol as an analyte and immunoassays for its measurement and exploring salivary free cortisol, calculated free cortisol and mass spectrometry as alternatives. The study itself recruited a total of 206 healthy volunteers and patients to undergo serum and salivary Synacthen tests. Serum cortisol was measured by gas chromatography-mass spectrometry (GC-MS) and by 5 contemporary immunoassays, salivary cortisol was measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS) and serum free cortisol was calculated using the Coolens’ equation and the free cortisol index. The results in serum confirmed the assay-dependence of immunoassay cortisol measurement, established valid reference ranges for each of the assays studied, proposed diagnostic cut-off concentrations and evaluated their performance in patients with suspected hypoadrenalism. The salivary results were used to establish adult reference ranges specific to the in-house LC-MS/MS assay used and subsequently demonstrated comparable performance in patients with suspected hypoadrenalism to the serum Synacthen test. Calculated serum free cortisol performed poorly compared to serum total and salivary cortisol and was subject to error because of difficulty measuring cortisol binding globulin (CBG). Comparisons between the three measures in patients with altered serum protein concentrations showed that, of the three, only salivary cortisol had the potential to accurately diagnose hypoadrenalism using a single cut-off; although confirmation of this finding is needed in a larger patient group