Objective: Liquid chromatography-tandem mass spectrometry (LC-MS/MS) eliminates cross-reactivity, which is a major limitation of immunoassays used for the measurement of salivary cortisol (SalC). We aimed to evaluate the potential of SalC measured by LC-MS/MS in patients undergoing assessment of the HPA axis.
Design and patients: Cross-sectional study of 78 patients admitted for routine testing in a specialized endocrine unit.
Measurements: Matched serum and saliva samples were collected from 68 patients who had a short synacthen test (SST, 250 mcg im) and 10 patients who had an insulin tolerance test (ITT, insulin 0.15 U/kg iv). Serum cortisol (SerC) was measured with an automated immunoassay and SalC with LC-MS/MS. Adequate SerC responses were >500 nmol/l.
Results: In all patients with adequate responses, the relative increase in SalC was significantly higher than that in SerC [6.4(0.3-26.1) vs. 1.0(0.3-4.9), P < 0.0001)]. The SerC-SalC relationship was better explained by an exponential rather than a linear model (R(2)=0.83 vs. R(2)=0.65, both P < 0.0001). Based on 59 patients with adequate SerC responses to an SST, an adequate SalC response was defined as 8.3 nmol/l. Seven patients following an SST and three patients following an ITT showed inadequate responses in both SerC and SalC, but two patients with CBG deficiency showed a low SerC with normal SalC.
Conclusions: We have shown an excellent diagnostic sensitivity and specificity of LC-MS/MS SalC in the assessment of the HPA axis and superiority over SerC when CBG levels are altered. The exponential relationship between SerC and SalC supports the concept of CBG binding capacity saturation.