OR23-01 Intrapatient ACTH Variability in Cushing’s Disease: Prognostic Significance

Abstract

Introduction: In patients with Cushing’s Disease (CD), intrapatient variability of hormone measurements creates significant clinical challenges, therefore multiple measurements are recommended.1 Urinary and salivary cortisol variation has been well described. However, intrapatient variation of adrenocorticotropic hormone (ACTH) in CD remains unknown. In CD patients, ACTH levels are inherently elevated from baseline but the coefficient of diurnal variation is reduced.2Additionally, at each diurnal time point, these exists a significant variation around the mean for the ACTH levels. In this study, we first analyzed the intrapatient variablility of ACTH at each diurnal timepoint in patients with CD. CD is primarily a disorder of ACTH excess, and treatment directed at pituitary adenomas would presumably perturb ACTH levels prior to affecting serum or urine cortisol. We hypothesized that the coefficient of variation at each diurnal time-point can help predict remission from CD following surgery. Methods: We conducted a retrospective review of patients (n = 645) who had histopathologically confirmed diagnosis of CD from 2005-2019 (NCT NCT00060541). We selected patients that had ≥ 3 plasma ACTH values over a 7 day span prior to surgical or medical intervention. We grouped the ACTH measurements into morning (AM) and midnight (PM) values to account for diurnal variation in ACTH secretion. We then analyzed post-operative hormone measurements performed every 6 hours prior to administration of replacement corticosteroids. Remission was assigned to patients with nadir serum cortisol level ≤5 mcg/dL within ten days post-operatively3,4. Results: We found 54 patients with multiple PM (n = 27) and AM (n = 41) ACTH measurements within a 7 day span. We found that the median coefficient of variation (CV) of intra-patient variability was 19.7% (N=41) (95% CI:12.5-27.5) for the AM and was 24% (N=27) (95% CI: 9.6-31.8) for the PM. Age, the number of tests, or the length of test period were not correlated with CV or absolute levels of ACTH. The intraclass correlation coefficient (ICC) of the AM data set was 0.59 and the PM data set was 0.79 which demonstrates a good and excellent reliability respectively. We found that that, in general, 30-60% decrease from pre-operative ACTH levels predicted remission from CD. ACTH decrease >50% on POD2 and 3 had 100% specificity and sensitivity in predicting remission. The decrease in ACTH preceded cortisol nadir in 3/10 patients by 24 hours. Conclusion: We found significant intra-patient variability in plasma levels of ACTH at individual diurnal timepoints in CD patients. We also found that the change in ACTH >50% on POD2 or 3 is an excellent predictor of remission from CD.