Cluster Analysis of Inflammatory Biomarker Expression in the International Severe Asthma Registry

Background: Allergy, eosinophilic inflammation, and epithelial dysregulation are implicated in severe asthma pathogenesis. Objective: We characterized biomarker expression in adults with severe asthma. Methods: Within the International Severe Asthma Registry (ISAR), we analyzed data from 10 countries in North America, Europe, and Asia, with prespecified thresholds for biomarker positivity (serum IgE ≥ 75 kU/L, blood eosinophils ≥ 300 cells/μL, and FeNO ≥ 25 ppb), and with hierarchical cluster analysis using biomarkers as continuous variables. Results: Of 1,175 patients; 64% were female, age (mean ± SD) 53 ± 15 years, body mass index (BMI) 30 ± 8, postbronchodilator forced expiratory volume in 1 second (FEV1) predicted 72% ± 20%. By prespecified thresholds, 59% were IgE positive, 57% eosinophil positive, and 58% FeNO positive. There was substantial inflammatory biomarker overlap; 59% were positive for either 2 or 3 biomarkers. Five distinct clusters were identified: cluster 1 (61%, low-to-medium biomarkers) comprised highly symptomatic, older females with elevated BMI and frequent exacerbations; cluster 2 (18%, elevated eosinophils and FeNO) older females with lower BMI and frequent exacerbations; cluster 3 (14%, extremely high FeNO) older, highly symptomatic, lower BMI, and preserved lung function; cluster 4 (6%, extremely high IgE) younger, long duration of asthma, elevated BMI, and poor lung function; cluster 5 (1.2%, extremely high eosinophils) younger males with low BMI, poor lung function, and high burden of sinonasal disease and polyposis. Conclusions: There is significant overlap of biomarker positivity in severe asthma. Distinct clusters according to biomarker expression exhibit unique clinical characteristics, suggesting the occurrence of discrete patterns of underlying inflammatory pathway activation and providing pathogenic insights relevant to the era of monoclonal biologics. © 2021 The Authors