SUN-485 Challenging Current Guidelines For Glucocorticoid-induced Adrenal Insufficiency: A Retrospective Cohort Study

Abstract Disclosure: R. Mehta: None. K. Lazarus: None. A. Sharma: None. K. Narula: None. Z. Htut: None. S. Choudhury: None. D.A. Papadopoulou: None. T. Tan: None. K. Meeran: None. Background: Prolonged glucocorticoid (GC) therapy is associated with increased morbidity and mortality. Discontinuation is a challenge due to the risk of developing GC-induced adrenal insufficiency (GC-AI). Current guidelines recommend against weaning if a morning cortisol is less than 300 nmol/L despite being on a physiological dose of prednisolone. This threshold may be overly cautious and could limit successful weaning. Methods: We conducted a retrospective study of 65 patients with long-term GC use for inflammatory disease, who were undergoing a prednisolone wean between 2019 and 2024. Clinical outcomes and biochemical data were analysed. Serial short Synacthen tests (SSTs; n=52) in 23 patients after dose reductions at ≤5 mg were used to assess hypothalamic-pituitary-adrenal axis (HPA) recovery. Kruskal-Wallis testing compared the effect of different dose reductions on HPA axis function. Results: From the cohort of 65, 19 patients continue to wean. Of the remaining 46 patients, 37 (80%) successfully discontinued prednisolone. The median duration of prednisolone use at referral was 23 months [IQR 6.5-66.5], and the median dose was 5 mg [3.75-5]. Median time to discontinuation was 8 months [5-15.5]. Amongst those who successfully weaned, 41% (n=15) had a morning cortisol <150 nmol/L, including 5 with undetectable levels. Only 24% (n=9) exceeded 300 nmol/L, even once prednisolone had been fully weaned. In patients with a baseline HbA1c ≥42 mmol/mol, prednisolone discontinuation led to a median HbA1c reduction of 6 mmol/mol (n=6, p=0.031).Reasons for weaning failure (n=9) were: inadequate control of the primary disease (n=5), insufficient HPA axis recovery (n=3), and intolerable GC withdrawal symptoms (n=1). The 10 SSTs from the three patients unable to fully discontinue prednisolone due to inadequate HPA axis recovery all showed the following triad: adrenocorticotropic hormone ≤10.7 ng/L, morning cortisol ≤34 nmol/L, and peak cortisol ≤92 nmol/L; yet all reduced to a maintenance dose of 2 mg (n=2) or 1 mg (n=1). No adrenal crises occurred across the cohort during the weaning process.Serial SSTs demonstrated significant increases in both baseline and peak cortisol following dose reductions of 1 mg, 1.5-2 mg, and >2 mg (all p<0.01). Larger dose reductions were associated with greater increases in both morning (p=0.038) and peak (p=0.011) cortisol. Conclusion: HPA axis function in GC-AI is more dynamic than current guidelines suggest and improves with reducing prednisolone dose, even at physiological doses. Additionally, the risk of adrenal crisis appears negligible with structured weaning and appropriate safety-netting. Our findings demonstrate that strict adherence to current GC-AI guidelines prevent successful weaning, leading to unnecessary GC exposure and their associated adverse effects. Presentation: Sunday, July 13, 2025