Bronchial microbiome in severe asthma stratified by blood eosinophil count: an exploratory study of diversity and compositional signals

Abstract Asthma is a heterogeneous disease whose pathophysiology underlies distinct inflammatory phenotypes, which in turn influence its clinical course and response to treatment. Neutrophilic inflammation is associated with acute infections and alterations in the microbiome, while data in severe asthma stratified by eosinophilic inflammation remain limited and heterogeneous. To compare clinical-functional characteristics and bronchial microbiome by inflammatory phenotype in patients with severe asthma. This prospective and observational study included 45 patients, 24 non-eosinophilic and 21 eosinophilic (defined as blood eosinophils > 300 cells/µL) subjects. Exacerbations and clinical-functional variables were recorded over 12 months. Bronchial microbiome was analyzed using 16 S rRNA gene amplicon sequencing (MiSeq Illumina), with taxonomic assignment via specific databases, and compositional differences were evaluated using Linear Discriminant Analysis Effect Size (LEfSe). Patients with eosinophilic phenotype had more exacerbations ( p = 0.004) and a non-significant trend toward increased bronchial hypersecretion and bronchopathy signs on chest CT. A significantly higher phylogenetic alpha diversity was observed in eosinophilic patients (Faith-PD, p = 0.006), together with differences in community membership based on unweighted UniFrac distances ( p = 0.005), whereas no differences were observed in abundance-weighted community structure. LEfSe and MaAsLin2 analyses identified Mycoplasma and Escherichia / Shigella as more abundant in eosinophilic patients, using an exploratory false discovery rate threshold (FDR ≤ 0.25). No significant differences in microbial diversity or composition were linked to exacerbation frequency. Blood eosinophil–high asthma was associated with higher phylogenetic diversity and differences in airway microbial community membership, driven mainly by low-abundance taxa rather than shifts in dominant genera. These findings suggest that eosinophilic inflammation may be linked to a more complex bronchial microbiome structure, potentially serving as a starting point for generating new hypotheses in the understanding and management of the disease.