Airway microbiome dysbiosis in severe pneumonia: metagenomic evidence of pathogen expansion and commensal depletion

BACKGROUND: The pulmonary microbiome is increasingly recognized as a key determinant of pneumonia severity, yet its clinical implications remain incompletely understood. Disruption of microbial ecology, or dysbiosis, may impair host immune responses and exacerbate disease progression. This study aimed to characterize microbiome alterations associated with severe pneumonia and their correlation with host inflammatory and coagulative parameters. METHODS: In this multicenter, prospective observational cohort study conducted across nine hospitals in Shanghai (2021-2025), bronchoalveolar lavage fluid (BALF) samples from 306 patients with clinically diagnosed pulmonary infections were analyzed using metagenomic next-generation sequencing (mNGS). Patients were stratified into severe (n = 196) and non-severe (n = 110) groups using WHO-derived severe pneumonia criteria at the time of bronchoalveolar lavage (BAL). Microbial taxonomic profiles, diversity indices, co-occurrence networks, and correlations with clinical markers were comprehensively assessed using standard bioinformatic and statistical approaches. RESULTS: Severe pneumonia was associated with marked microbial dysbiosis, including reorganization of co-occurrence network topology with centrality shifting away from commensals toward opportunistic taxa in severe disease, characterized by reduced α-diversity, altered β-diversity, and enrichment of opportunistic Gram-negative pathogens including Acinetobacter and Klebsiella. In contrast, commensals such as Rothia and Prevotella were depleted. Co-occurrence network analysis revealed fragmentation of microbial interactions in severe cases, with centrality shifting from commensals to opportunists like Corynebacterium striatum. Shannon diversity negatively correlated with SOFA scores, and specific taxa positively associated with systemic inflammation (CRP, PCT) and coagulation abnormalities. Nearly all samples demonstrated polymicrobial infection, with distinct microbial patterns observed across monomicrobial and polymicrobial subgroups. CONCLUSION: Our multicenter observational analysis suggests that severe pneumonia is associated with marked ecological disruption of the lower-airway microbiome, characterized by commensal loss, opportunist expansion, and fragmented interspecies networks, and with concurrent inflammatory and coagulative abnormalities. These hypothesis-generating findings warrant external validation in independent, multi-region cohorts and longitudinal sampling to test directionality and causality before informing clinical decision-making.