"A morning cortisol is the most effective clinical predictor of short synacthen test outcome": A tertiary care centre...

Increasing referrals to Endocrinology with nonspecific symptoms of suspected adrenal insufficiency (AI) has increased use of the short-synacthen test (SST). Prevailing resource and safety concerns emphasise importance of patient selection criterion to optimise SST use. This study aimed to (1) document the adverse event profile of the SST (2) identify any pretest predictors of SST outcome.

Retrospective data analysis of all patients referred for SST in Oxford from 2017 to 2021. Pretest clinical variables (age, sex, BMI, blood pressure and electrolytes), symptoms (fatigue, dizziness, weight loss) and pretest morning cortisol were included in the statistical model with the aim of identifying any variables that could predict SST outcome in Group 1 primary AI, Group 2 central AI and Group 3 glucocorticoid induced AI. Symptoms and signs during and post SST were also noted with the aim of describing adverse effects to synacthen across a large cohort.

A total 1480 SSTs (Males:38%, age 52 [39-66] years) were performed: 505 (34.1%) in Group 1, 838 (57%) in Group 2, and 137 (9.3%) in Group 3. Adverse-effects were recorded in 1.8% of tests, including one episode of anaphylaxis. Pretest morning-cortisol was the only predictor for an "SST pass" (whole cohort: B = 0.015, p < 0.001, Group 1: B = 0.018, p < .001; Group 2: B = 0.010, p < 0.012; Group 3: B = 0.018, p = <.001). A threshold of ≥343 nmol/l (receiver-operating characteristic [ROC] area under the curve [AUC] = 0.725, 95% confidence interval [CI] 0.675-0.775, p < 0.001) for the whole cohort, ≥300 nmol/L (ROC AUC = 0.763, 95% CI 0.675 to 0.850, p < 0.001) for Group 1, ≥340 nmol/L (ROC AUC = 0.688, 95% CI 0.615 to 0.761, p < 0.001) for Group2, and ≥376 nmol/L [baseline cortisol] (ROC AUC = 0.783, 95% CI 0.708 to 0.859, p < 0.001) for Group 3, predicted an 'SST pass' with 100% specificity.

Adverse effects to synacthen are rare. Pretest morning cortisol is a reliable predictor for SST outcome and is a helpful tool to rationalise use of the SST. Predictive morning-cortisol thresholds vary according to the aetiology of AI.