# Characteristics of pathogenic microorganisms in COPD-related infections: prognostic correlations and implications

## TL;DR

This study identifies distinct microbial patterns in COPD patients with respiratory infections, linking specific pathogens to prognosis and suggesting tailored clinical approaches.

## Contribution

The study reveals unique microbial profiles in COPD-related infections and their prognostic implications using next-generation sequencing and machine learning.

## Key findings

- COPD patients showed higher prevalence of gram-negative bacteria and fungi like Pseudomonas aeruginosa and Haemophilus influenzae.
- Non-COPD patients had more atypical pathogens such as Mycoplasma pneumoniae.
- Machine learning identified SARS-CoV-2, Veillonella parvula, and Achromobacter xylosoxidans as strong predictors of adverse outcomes.

## Abstract

Chronic obstructive pulmonary disease (COPD) significantly impacts global health, primarily due to frequent acute exacerbations caused by respiratory infections. Precise microbial characterization may inform prognostic insights and optimize clinical management.

We conducted a prospective observational study from December 2023 to February 2025 involving 1146 patients (259 COPD; 887 non-COPD) with suspected respiratory infections. Bronchoalveolar lavage fluid samples underwent next-generation sequencing (NGS) and conventional microbiological testing. Multivariate logistic regression identified COPD predictors, and machine learning modeled prognostic outcomes based on microbial profiles.

Distinct pathogen distributions emerged between COPD and non-COPD groups, with COPD patients exhibiting higher prevalence of gram-negative bacteria, particularly Pseudomonas aeruginosa and Haemophilus influenzae, and fungal pathogens. Non-COPD patients demonstrated increased occurrence of atypical pathogens, notably Mycoplasma pneumoniae. COPD patients also presented higher loads of traditionally commensal microorganisms, such as Veillonella parvula and Schaalia odontolytica. Age, dyspnea, smoking duration, elevated leukocyte and neutrophil counts, and decreased lymphocyte levels were significantly associated with COPD presence. Machine learning identified specific microorganisms as strong predictors of adverse outcomes, such as SARS-CoV-2, Veillonella parvula, and Achromobacter xylosoxidans.

Comprehensive microbial profiling using NGS effectively distinguishes pathogen differences between COPD and non-COPD patients, revealing key associations with clinical prognosis. These insights can inform tailored clinical interventions aimed at mitigating COPD exacerbations and improving patient outcomes.