Lung Disease Biology Over Labels: Two Airway Endotypes Found Across COPD, Asthma, and More

Chronic airway diseases — asthma, COPD, bronchiectasis, and cystic fibrosis (CF) — are typically managed as distinct diagnoses, yet they share overlapping features including chronic inflammation and impaired mucociliary clearance. Up to 50% of COPD patients exhibit asthma-like features, and eosinophilic endotypes have been found in what is classically considered a neutrophilic disease (bronchiectasis). This study asked whether biology-first stratification could better define patient subtypes than traditional diagnostic labels.

Researchers enrolled 271 participants at Ninewells Hospital, Dundee: 91 COPD, 76 asthma, 54 bronchiectasis, 24 CF, and 26 formerly-smoking controls with normal lung function. Spontaneous sputum was collected during clinical stability. Measurements included 18 cytokines via Meso Scale Discovery assay, neutrophil elastase by immunoassay, IL-8 by ELISA, mucin quantification (MUC5B and MUC5AC) via mass spectrometry, rheological parameters (storage modulus G′, loss modulus G′′, complex modulus G∗, and tangent of loss angle δ), sputum dry weight, DNA content, and microbiome profiling via long-read 16S rRNA sequencing. K-means clustering was applied to inflammatory and sputum property data across multiple imputed datasets.

Nine cytokines, neutrophil elastase, dry weight, mucins, and rheological parameters differed significantly between disease groups and controls. K-means clustering identified two optimal clusters: a neutrophilic endotype characterized by higher neutrophil elastase, IL-8, and G-CSF, and a Th2 endotype marked by elevated IL-4, IL-5, eotaxin, eotaxin-3, TARC, and IFN-γ. The Th2 cluster showed higher G′, G′′, and G∗ (more elastic mucus) and higher MUC5B, while the neutrophilic cluster had higher tan delta, indicating more viscous relative to elastic mucus. Sputum dry weight and DNA were lower in Th2 patients. The neutrophilic cluster also exhibited lower alpha diversity and higher Proteobacteria abundance in the airway microbiome.

Critically, both clusters were present in all four disease groups. Neutrophilic inflammation dominated in CF (87%) and bronchiectasis (78%), but 46% of asthma patients and 42% of COPD patients also fell into the neutrophilic cluster — underscoring that disease labels imperfectly predict underlying biology. The Th2 cluster was most prevalent in asthma but present across all diseases.

These findings support the treatable traits paradigm: assessing patients by inflammatory endotype and mucociliary dysfunction biomarkers rather than diagnosis alone could enable more precise therapy selection and facilitate cross-disease repurposing of targeted agents such as biologics. Caveats include single-center design, single-timepoint sampling, exclusion of patients on CFTR modulators, and the relatively small CF cohort.