Endocrine Abstracts

Salivary cortisol (SalC) potentially has advantages over serum cortisol (SerC) as it is easier to collect and possibly reflects unbound (free) cortisol. SalC levels are approximately 10 times lower than SerC and require high measurement precision best addressed by liquid chromatography–tandem mass spectrometry (LC-MS/MS), which in contrast to previously used radioimmunoassays eliminates cross-reactivity by other steroids. We describe the use of LC-MS/MS to evaluate the potential of SalC in a series of patients undergoing routine assessment of the HPA axis by short synacthen test (250 mcg, sampling 0, 30 and 60 min, normal SerC response >500 nmol/l).

Forty-two patients (26 female, mean age 42, 4 on oestrogens/SERMs, 1 CBG deficiency) had simultaneous salivary and blood samples collected. SerC was measured with an automated immunoassay (Siemens Centaur, CV 5.5, 4, 4.1% at 260, 675 and 1055 nmol/l, respectively) and SalC with LC-MS/MS (lower limit of quantitation 2nmol/l, CV 9.4, 4.4, 5.5% at 4.6, 28 and 56 nmol/l).

Median (range) SerC (nmol/l) was 344.5 (30–658), 601 (367–950) and 725 (115–1089) and SalC (nmol/l) was 3.91 (2–95), 23.93 (5.66–345) and 39.8 (7.4–66.8) at baseline, 30 and 60 min, respectively. The relative increase in SalC was significantly higher than in SerC (6.73 (1.95–23.3) vs 1.14 (0.27–2.83), P<0.0001). SerC and SalC were significantly correlated (r=0.87, P<0.0001) and showed an exponential relationship. Thirty-five patients had normal serum responses and a cut-off level of 10 nmol/l was established for salivary cortisol. One patient on triamcinolone had both SerC and SalC below the cut-offs (467, 7.4 nmol/l, respectively), and a patient with CBG deficiency showed a subnormal SerC (115 nmol/l) but normal SalC (41.5 nmol/l).

These data illustrate the correlation of SalC measured by LC-MS/MS with SerC. The exponential rise in SalC supports the concept of CBG binding capacity saturation. SalC may have advantages over SerC when CBG levels are altered.