Introduction: The short Synacthen test (SST) is commonly used to assess the hypothalamic-pituitary-adrenal (HPA) axis. Given variations in corticosteroid binding globulin (CBG) concentration and binding affinity, measuring the total serum cortisol response may misclassify some patients. Salivary cortisol correlates well with serum free cortisol but is easier to measure and widely available in commercial laboratories. The aim of this study was to characterise serum and salivary cortisol response to the SST and compare their respective diagnostic accuracies in patients with suspected cortisol deficiency.
Methods: Synacthen 250µg iv was administered to 114 participants, comprising of patients suspected to have cortisol deficiency (n=83), healthy volunteers (n=21) and healthy women on the oral contraceptive (OCP, n=10). Serum and salivary cortisol were measured at 0, 30 and 60 minutes.
Results: There was a highly significant difference in serum cortisol between the healthy volunteers and the women on the OCP (P<0.001) but no difference in salivary cortisol (P=0.39). The lower limit of normal for salivary cortisol at 60 min (2.5% confidence interval) was 26 nmol/L. 28/83 (34%) patients with suspected HPA axis disorder failed the laboratory serum cortisol cut-off at 60 min of 500 nmol/L. Of these, 3/28 (11%) would have passed the salivary cortisol of ≥26 nmol/L. In contrast, 11/18 patients where the 60 min serum cortisol was 500-599 nmol/L and 4/37 with 60 min cortisol ≥600 nmol/L failed the salivary cortisol cut off of ≥26 nmol/L. Including normal participants, 94% whose 60 min cortisol was ≥600 nmol/L had a 60 min salivary cortisol of ≥26 nmol/L.
Conclusion: There is additional diagnostic value in measuring salivary cortisol during the 250µg SST, particularly in patients considered to have a borderline pass with respect to the 60 min serum cortisol between 500-599 nmol/L and in those with proven or suspected alterations in CBG.