SAT-382 Glucocorticoid Tapering in Monthly 1-mg Decrements Does Not Result in Clinically Manifest Adrenal Insufficiency in Patients with Rheumatoid Arthritis: Learnings from the Phase 3/4 SEMIRA Study

Introduction: Systemic glucocorticoids (GCs) are used to treat serious inflammatory diseases but are associated with adverse events. Guidelines recommend tapering GCs to the lowest possible dose and discontinuing as soon as possible. Physicians have concerns that reductions, even from low doses, may increase disease symptoms or cause adrenal insufficiency (AI), especially in patients (pts) receiving long-term GCs. The expectation of AI risk may be inflated by false positives in cortisol testing of pts without relevant symptoms and by a lack of robustly designed GC taper trials. The international, multicenter SEMIRA trial evaluated a taper scheme in rheumatoid arthritis (RA) pts receiving tocilizumab (TCZ) ± conventional synthetic disease-modifying antirheumatic drugs (csDMARDs).(1)Methods: Eligible pts had low disease activity or remission for ≥4 weeks (wks) and were receiving a stable prednisone regimen (5 mg/day) + TCZ ± csDMARDs for ≥4 wks. Pts had to have received ≥6 months’ total TCZ + GCs (prednisone equivalent 5-15 mg/day). Pts were randomized (1:1) to double-blind continued prednisone 5 mg/day (n=128) or prednisone taper (n=131). TCZ ± csDMARDs remained stable during the 24-wk study. GC tapering consisted of 1-mg decrements from randomization and every 4 wks thereafter until discontinuation. RA flares were treated with prednisone 5 mg/day for 2 wks. The primary assessment was maintenance of disease control with GC discontinuation. The protocol provided evidence-based guidance for diagnosis and management of AI. Confirmatory testing was recommended for pts with suspected AI, but routine precautionary ACTH stimulation testing was not mandated. Results: In the taper arm, 65% of pts achieved “treatment success,” meeting all key secondary endpoint components (maintained low disease activity, experienced no flares, and had no confirmed AI requiring replacement therapy) at wk 24 versus 77% of continued pts (risk ratio for treatment success 0.833 [95% CI: 0.714, 0.972] p=0.021). None of the taper pts required ACTH stimulation testing, and no cases of AI were reported. Conclusions: SEMIRA demonstrated the usefulness of a new standardized GC dose taper scheme. Two-thirds of pts receiving TCZ underwent successful tapering and could stop GCs entirely, which is higher than a spontaneous 35% discontinuation rate observed in real-world RA pts(2) and underscores the potential to further reduce steroid burden. Clinical AI was not observed; thus, routine laboratory testing may be unnecessary in real-world applications of this taper scheme. References: 1. Burmester GR et al. Arthritis Rheumatol. 2018;70(suppl 10):L18. 2. Yagız B et al. Arthritis Rheumatol. 2018;70(suppl 10):627.