Identification of clinically relevant subgroups of COPD based on airway and circulating autoantibody profiles.

Autoimmunity may serve a role in the pathological features of a subgroup of patients with chronic obstructive pulmonary disease (COPD); however, in immunological subgroups of COPD patients, the interrelationships between airway and circulating autoantibody responses, and clinical parameters, remain unclear. The present study was undertaken to evaluate these interrelationships in various immunological subgroups of COPD patients. Sputum supernatant and serum obtained from 102 patients with stable COPD were assayed for the presence of immunoglobulin G antibodies against ten autoantigens via Luminex multiplex technology. Hierarchical clustering based on principal components was performed on autoantibody profiles to classify patients into clusters. Network‑based and module analyses were conducted to explore interrelationships among autoantibodies and clinical variables in each cluster. Topological characteristics were compared between clusters. Unsupervised clustering identified four clusters: No significant differences in the majority of clinical characteristics were observed among clusters. In cluster 1, retrospective exacerbation was only positively associated with COPD assessment test score. Lung functions (predicted % of forced expiratory volume in 1 sec and maximal mid‑expiratory flow) were negatively associated with exacerbation risk only in cluster 2. Sputum autoantibodies (against U1 small nuclear ribonucleoprotein, proteinase‑3 and Ro/Sjögren syndrome type A antigen) were negatively associated with exacerbation risks in cluster 2, but positively associated in cluster 3. The four networks also exhibited distinct topological properties. In COPD, autoantibody responses were heterogeneous and differentially associated with exacerbation risk in certain subgroups; their dual character should be considered in future research.