# Radiological Phenotypes of Bronchiectasis Based on Airway Generation

## TL;DR

This study identifies two bronchiectasis subtypes based on airway location, showing differences in disease severity, inflammation, and microbes, which could help tailor treatments.

## Contribution

The paper introduces a novel classification of bronchiectasis based on airway generation, linking it to distinct clinical and biological features.

## Key findings

- The DA group had more severe lung function decline and structural damage compared to the PIA group.
- PIA was associated with allergic bronchopulmonary aspergillosis and eosinophilic inflammation.
- Metabolomic analysis found higher 4-hydroxynonenal levels in the DA group, linked to more frequent exacerbations.

## Abstract

Background: High-resolution computed tomography reveals a marked radiological heterogeneity in bronchiectasis; however, the clinical characteristics have not been clearly elucidated. Method: We conducted a prospective observational cohort of 334 bronchiectasis patients at Wuhan Union Hospital. Patients were classified into distal airway (DA) and proximal–intermediate airway (PIA) phenotypes and followed every six months for exacerbations. Clinical, inflammatory, microbial, and metabolic features were compared between groups. Results: Among 334 patients, 206 were classified as DA and 128 as PIA. Most allergic bronchopulmonary aspergillosis cases belonged to the PIA group (p < 0.001). The DA group showed a lower FEV1%pred (p = 0.010) and Bhalla scores (p < 0.001), higher BSI (p = 0.003) and FACED scores (p < 0.001), more frequent exacerbations (p = 0.002), and a greater prevalence of Pseudomonas aeruginosa (PA) colonization (p < 0.001). Radiologically, the DA group exhibited more extensive structural lung damage (all p < 0.05). Inflammatory profiling showed higher neutrophil counts (p = 0.047) and elevated CRP levels (p = 0.006) in DA, whereas the PIA group was characterized by eosinophilic inflammation (p = 0.026); no significant differences were observed in inflammatory cytokine levels. Microbial interaction network analysis revealed distinct ecological structures between phenotypes. The PIA group showed strong negative correlations with Streptococcus, Rothia, and other commensal taxa, whereas the DA group exhibited no significant associations between Pseudomonas aeruginosa and other species. Furthermore, metabolomic analyses revealed elevated 4-hydroxynonenal levels in the DA group, which also experienced a higher rate of acute exacerbations during follow-up (p = 0.003). Conclusions: Distinct radiological phenotypes based on airway generation in bronchiectasis are associated with different clinical severity, inflammatory profiles, and microbiome features which enable personalized bronchiectasis management.