SFEBES2022 Poster Presentations Adrenal and Cardiovascular (66 abstracts)
1Division of Endocrinology, Digestion & Metabolism, Imperial College London, London, United Kingdom; 2Department of Endocrinology & Metabolic Medicine, Imperial College Healthcare NHS Trust, London, United Kingdom
Background: Adrenal Insufficiency (AI) presents a diagnostic and clinical challenge. While short Synacthen Tests (SSTs) are most commonly utilised to diagnose AI, the value of a baseline cortisol is being explored. We aimed to review indications for performing an SST and to determine the baseline cortisol that predicted a preserved cortisol response to Synacthen.
Methods: Eight hundred and sixty SSTs performed in 621 individuals at a tertiary endocrinology unit were retrospectively reviewed (April 2017-December 2021, Abbott Architect i-2000 immunoassay analyser). Tests started at 09:00-10:00. ROC curve analysis was used to determine the 0-minute cortisol threshold that predicted a subsequent adequate response to synacthen. The thresholds that have previously been determined by our Trust were used to define a preserved cortisol response: 30-minute value 370 nmol/l and 60-minute value 420 nmol/l.
Results: Mean (SD) age of the cohort was 53 (±16.0) years and mean (SD) body mass index measured 27.6 (±6.57) kg/m2. Of the 860 SSTs, 30.2% (n=260) were conducted to investigate for primary, 51.4% (n=442) for secondary and 16.3% (n=141) for tertiary AI. Table 1 illustrates the 0-minute cortisol values that predicted a preserved adrenal axis according to type of AI investigated for.
0 min Cortisol Compared with | N | Basal(0 min) Cortisol Measurement (nmol/l) at Sensitivity=0.974 | Specificity at Sensitivity=0.974 |
Primary 30 min | 251 | 147 | 0.607 |
Primary 60 min | 258 | 144 | 0.552 |
Secondary 30 min | 422 | 135 | 0.556 |
Secondary 60 min | 425 | 130 | 0.353 |
Tertiary 30 min | 135 | 94 | 0.377 |
Discussion and Conclusions: The type of adrenal insufficiency being investigated is associated with variations in baseline cortisol. In determining thresholds, the balance of missing a potentially serious condition against the risks of over-exposure to steroids needs to be considered. In this cohort, 0-minute cortisol has a limited ability to accurately predict a preserved cortisol response due to the low associated specificity.