Inflammatory and Mucociliary Dysfunction-Based Endotypes Across the Spectrum of Chronic Airway Diseases.

Abstract

There is substantial overlap between features of COPD, asthma, bronchiectasis, and cystic fibrosis (CF). Each is characterized by inflammation and mucociliary dysfunction. Is there a relationship between inflammation and mucociliary clearance in chronic respiratory conditions, and can biology, rather than disease labels, stratify patients into therapeutically relevant subtypes? Patients were categorized according to primary disease and clinical characteristics recorded. Spontaneous sputum was collected, and inflammatory markers (neutrophil elastase and 18 cytokines), sputum properties (DNA content, mucins, rheology, dry weight), and microbiome (long-read 16S sequencing) were measured. K-means clustering was performed and parameters compared between and within disease groups. Control participants were individuals who had formerly smoked but were without respiratory disease. The study included patients with asthma (n = 76), COPD (n = 91), bronchiectasis (n = 54), CF (n = 24), and control participants (n = 26). Nine cytokines (interferon-γ, IL-4, IL-5, eotaxin, eotaxin-3, thymus and activation regulated chemokine, granulocyte colony-stimulating factor, fractalkine, IL-22), neutrophil elastase, dry weight, mucins, and sputum rheology parameters were significantly different between disease groups and control participants (P < .05). K-means clustering identified 2 clusters defined by neutrophilic or T helper 2 (Th2) inflammation. The Th2 cluster was associated with lower sputum dry weight and DNA content and higher mucin-5B. Rheological parameters G', G'', and G∗ were significantly higher in the Th2 group, whereas the tangent of the loss angle δ was higher in the neutrophilic group, indicating a higher viscous to elastic ratio (P < .05 all comparisons). The neutrophilic cluster was associated with decreased alpha diversity (P = .04) and increased presence of Proteobacteria in their sputum microbiome compared with the Th2 cluster (P = .01). More neutrophilic inflammation was present in CF and bronchiectasis (42% of patients with COPD and 46% of patients with asthma were neutrophilic vs 78% of bronchiectasis and 87% of CF (P < .0001). Both clusters were present in all disease groups. Our results indicate that airway diseases have heterogeneous mucus properties. Patients were shown to cluster according to inflammatory endotype rather than disease label. Assessment based on disease labels may be aided by endotyping using inflammatory and mucociliary clearance biomarkers.