SAT-741 Salivary Cortisol and Cortisone Measurement Provide a Novel and Non-Invasive Method of Monitoring Medical Therapy in Cushing’s Syndrome

Introduction: Salivary glucocorticoids (cortisol [SalF], cortisone [SalE]) are collected non-invasively, unaffected by variation in cortisol binding globulin (CBG) and an established tool in the investigation of Cushing’s syndrome (CS). We have previously shown a strong correlation between salivary glucocorticoids and circulating free serum cortisol, better than that with total serum cortisol. Measurement by liquid chromatography with tandem mass spectrometry (LC-MS/MS) permits lower limits of quantification, eliminates cross-reactivity both between cortisol and cortisone, and by metyrapone-induced elevations in 11-deoxycortisol. Previous studies in CS have demonstrated that a mean (based on 5 samples during a single day) serum cortisol (serumF) of 150-300 nmol/l equates to a normal isotopically calculated cortisol production rate. We report the use of salivary glucocorticoid measurement in metyrapone-treated CS patients undergoing dose titration to achieve a mean serumF of 150-300 nmol/L. Methods: Seventeen (11 females; age-range 24-74 years) patients with CS undergoing dose titration with metyrapone were studied on 44 occasions: 15 ACTH-dependent (5 ectopic) and 2 adrenal. 24 healthy male volunteers (HV) were also studied. Both cohorts had paired serumF and SalE and SalF samples collected at 5 time-points (10:00; 11:30; 13:00; 14:30; 16:00). Serum and salivary glucocorticoids were measured using LC-MS/MS. The TR for salivary glucocorticoids was determined from the mean SalE and SalF in HV whose mean serumF (n=20) was in the desired range (150-300 nmol/L). In CS patients the metyrapone dose had been titrated to achieve a mean serumF of 150-300 nmol/L on 14 (out of 44) occasions. Results: Mean SalE (r=0.70; p<0.001) and mean SalF (r=0.51; p<0.001) showed a significant correlation with serumF. The TR from HV for mean SalE and mean SalF were 5.8–24.7 nmol/L and 0.6-5.4 nmol/L respectively. In CS patients, SalE had greater sensitivity (79% vs. 75%) and specificity (80% vs. 57%) in predicting a mean serumF in the TR than SalF. Conclusion: We have demonstrated a close correlation between mean SalE and mean serumF in metyrapone treated CS patients. Salivary glucocorticoids within the derived TR were highly predictive for a target serumF. It is unsurprising, due to the effect of CBG on serumF measurements, that the sensitivity and specificity of SalE and SalF are not greater than reported. Consequently, SalE and SalF, as surrogates for free serum F, have the potential to be superior than serumF when assessing adequacy of medical therapy in CS and may permit out-of-hospital monitoring. Further work is required to validate these findings.