Sputum microbiome is associated with 1-year mortality after chronic obstructive pulmonary disease hospitalizations

Abstract

Rationale: Lung dysbiosis promotes airway inflammation and decreased local immunity, potentially playing a role in the pathogenesis of acute exacerbations of chronic obstructive pulmonary disease (AECOPD). Objectives: We sought to determine the relationship between sputum microbiome at the time of AECOPD hospitalization and 1-year mortality in a COPD cohort. Methods: We used sputum samples from 102 patients hospitalized because of AECOPD. All subjects were followed for 1 year after discharge. The microbiome profile was assessed through sequencing of 16S rRNA gene. Microbiome analyses were performed according to 1-year mortality status. To investigate the effect of a-diversity measures and taxon features on time to death, we applied Cox proportional hazards regression models and obtained hazard ratios (HRs) associated with these variables. Measurements and Main Results: We observed significantly lower values of a-diversity (richness, Shannon index, evenness, and Faith’s Phylogenetic Diversity) among nonsurvivors (n = 19, 18.6%) than survivors (n = 83, 81.4%). b-Diversity analysis also demonstrated significant differences between both groups (adjusted permutational multivariate ANOVA, P = 0.010). The survivors had a higher relative abundance of Veillonella; in contrast, nonsurvivors had a higher abundance of Staphylococcus. The adjusted HRs for 1-year mortality increased significantly with decreasing a-diversity. We also observed lower survival among patients in whom sputum samples were negative for Veillonella (HR, 13.5; 95% confidence interval, 4.2–43.9; P, 0.001) or positive for Staphylococcus (HR, 7.3; 95% confidence interval, 1.6–33.2; P = 0.01). Conclusions: The microbiome profile of sputum in AECOPD is associated with 1-year mortality and may be used to identify subjects with a poor prognosis at the time of hospitalization.