A real-world clustering analysis reveals heterogeneous response patterns to biologic therapy in severe asthma

Aim: Despite the revolutionary impact of biologics (Bx) on severe asthma management, predicting individual treatment responses remains challenging. We aimed to characterize the heterogeneous nature of clinical status and disease activity in patients with severe asthma after biologic therapies through a comprehensive evaluation of real-world clinical outcomes. Methods: In this retrospective, multicenter study of 53 patients with severe asthma who received biologic therapies, hierarchical clustering analysis was performed based on three key parameters during treatment: exacerbation, maintenance oral corticosteroid (mOCS) dose, and lung function. Canonical correlation analysis and multinomial logistic regression were used to identify predictors of response patterns. Results: Clustering analysis revealed three distinct control groups: well-controlled (n = 23), moderately controlled (n = 22), and poorly controlled (n = 8). Well-controlled patients exhibited minimal exacerbations, no oral corticosteroid (OCS) use, and optimal or stabilized lung function. Moderately controlled patients showed minimal exacerbations and no mOCS use but variable lung function improvements. Poorly controlled patients exhibited persistent exacerbations, mOCS dependence, or both with limited lung function improvement. Baseline forced expiratory volume in 1 second (FEV1) %predicted (percent predicted FEV1) values and blood eosinophil counts independently differentiated well-controlled from moderately controlled patients, whereas baseline mOCS use distinguished moderately controlled from poorly controlled patients. Conclusions: Our findings reveal distinct patterns of disease control following biologic therapy in severe asthma, with baseline lung function, eosinophilic inflammation, and OCS use as key predictive factors. These results support the need for personalized treatment approaches in severe asthma management.