Abstract
OBJECTIVE: Systemic corticosteroids for 5-7 days are standard care for asthma exacerbations, but the optimal duration and potential for precision prescribing remain unclear. Biomarker-guided approaches could reduce corticosteroid exposure without compromising outcomes. We aimed to evaluate whether the blood eosinophil count can be used to safely reduce systemic corticosteroid exposure in hospitalised asthma exacerbations. METHODS: In this open-label, two-centre randomised trial, adults hospitalised for asthma exacerbation were assigned 1:1 to usual care (5 days prednisolone) or eosinophil-guided care using blood eosinophil counts obtained prior to corticosteroid administration (3 days if eosinophils <300 cells/µL; 5 days if ≥300 cells/µL). The primary outcome was non-inferiority of treatment failure rates (composite of extension of steroid duration, mechanical ventilation or death), with a prespecified 20% non-inferiority margin. RESULTS: Among 110 randomised patients (55 per group), 60% were eosinophilic, and 40% non-eosinophilic. Treatment failure occurred in 6/55 (10.9%) of eosinophil-guided versus 4/55 (7.3%) of usual care patients, with a 3.6% absolute difference (95% CI -8.9% to 16.2%), meeting non-inferiority. Cumulative corticosteroid dose per patient was similar between groups but significantly lower for non-eosinophilic than eosinophilic exacerbations in the eosinophil-guided group (136 vs 214 mg; p=0.0004), a difference not observed in usual care (186 vs 211 mg; p=0.18). Length of stay, Asthma Control Questionnaire-5 change, additional steroid bursts at 14 days or time to next exacerbation up to 1 year showed no significant differences. CONCLUSION: Eosinophil-guided therapy safely reduced systemic corticosteroid exposure in non-eosinophilic exacerbations while maintaining non-inferior outcomes in this proof-of-concept trial. TRIAL REGISTRATION NUMBER: NCT05417906.