Evaluating tertiary adrenal insufficiency in rheumatology patients on long‐term systemic glucocorticoid treatment

Abstract Objective Patients with rheumatic diseases are often treated with prolonged, high‐dose systemic glucocorticoids which can cause hypothalamic‐pituitary‐adrenal (HPA) axis suppression and development of tertiary adrenal insufficiency. Adrenal insufficiency carries the risk of serious, potentially life‐threatening adrenal crisis. Our study evaluated the prevalence, characteristics and recovery of patients with underlying rheumatology conditions who had received prolonged glucocorticoid treatment. Design and patients Retrospective, cross‐sectional study. We evaluated 238 patients seen in outpatient rheumatology clinic, managed in accordance with current nationally and internationally accepted clinical guidelines. Measurements Data collected included patient demographics, historical steroid data, 09.00 h cortisol/short synacthen test (SST) results and follow‐up data on those with repeat assessments. Results Overall, 65% of our cohort had a 09.00 h cortisol <350 nmol/L. On SST, 43% of patients demonstrated evidence of possible tertiary adrenal insufficiency. Prednisolone equivalent dose at time of SST was significantly higher in the group who failed SST vs. those who passed; mean of 5.57 mg vs. 3.59mg ( p = .005). 09.00 h cortisol result correlated with 30‐min cortisol on SST ( R 2 = .20, p = .002). 0‐min cortisol on SST correlated more strongly with 30‐min cortisol than 09.00 h cortisol ( R 2 = .59, p ‐value < .001). Patients with 0‐min cortisol >350 nmol/L, all passed their SST. Patients who remained on prednisolone were more likely to recover (71%) vs. those switched to hydrocortisone (27%), P = .02. Peak steroid dose was predictive of recovery; significantly lower in those who recovered (mean of 22.3 mg vs. 33.8 mg, P = .03). Steroid duration was not found to be a predictor of recovery [recovery (64.2 months) vs. non‐recovery (55.6 months), P = .58]. There was no correlation found to outcome on SST with age, sex, peak steroid dose, steroid duration, underlying rheumatological condition, additional exogenous steroid use or serum sodium. Conclusions Forty three percent of our patients demonstrated sub‐optimal adrenal function on SST. Steroid dose at the time of SST was the only significant predictive risk factor for tertiary adrenal insufficiency. 09.00 h cortisol demonstrated good correlation with outcome on SST and could represent a valid screening test to reduce need for SST if 09.00 h >350 nmol/L. Further prospective data are required to further characterize risk factors, predictive features of recovery and establish optimal management strategy of steroids (prednisolone vs hydrocortisone) to encourage adrenal recovery.