Characterization of pulmonary bacterial composition in community-acquired pneumonia depending on sampling methods

Introduction Community-acquired pneumonia (CAP) remains a major global health concern due to its high incidence, morbidity, mortality, and economic impact. Although Streptococcus pneumoniae is the primary bacterial cause, diverse bacterial and viral pathogens contribute to CAP etiology. Conventional microbiological methods provide limited characterization of the pulmonary microbiome, resulting in low diagnostic yield and frequent empirical antibiotic use. This pilot study aimed to characterize the pulmonary microbiome in adult CAP patients using high-throughput sequencing and to determine whether commonly used non-invasive samples accurately represent the lower respiratory tract. Methods Pulmonary samples were obtained using invasive procedures (bronchoalveolar lavage, bronchial aspirate, bronchial brushing, and biopsy) and non-invasive methods (sputum, nasopharyngeal aspirate). Classical microbiological testing identified pathogens in 24.1% of cases. DNA extracted was subjected to 16S rRNA gene amplification and sequencing, and microbial composition and diversity were compared across sampling types. Results Microbiome profiles differed significantly by sampling methodology. Non-invasive samples displayed high variability and limited correspondence with lower airway microbiota. In contrast, invasive samples demonstrated greater consistency and appeared more representative of the lower respiratory tract. The genus Ralstonia was significantly more abundant in invasive samples ( p < 0.05), warranting further investigation of its potential biological and clinical relevance. Conclusions 16S rRNA gene sequencing provides comprehensive detection of pulmonary microorganisms and offers added value in the microbiological diagnosis of CAP. The findings highlight the importance of using invasive specimens for accurate microbiome characterization and for detecting bacterial taxa not identified by routine microbiological methods.