Rectal bacteriome and virome signatures and clinical outcomes in community-acquired pneumonia: An exploratory study

Background Bacterial intestinal communities interact with the immune system and may contribute to protection against community-acquired pneumonia (CAP). Intestinal viruses are closely integrated with these bacterial communities, yet the composition and clinical significance of these communities in CAP patients are unknown. The aims of this exploratory study were to characterise the composition of the rectal bacteriome and virome at hospital admission for CAP, and to determine if microbiota signatures correlate with clinical outcomes. Methods We performed a prospective observational cohort study in CAP patients, admitted to a university or community hospital in the Netherlands between October 2016 and July 2018, and controls. Rectal bacteriome and virome composition were characterised using 16S ribosomal RNA gene sequencing and virus discovery next-generation sequencing, respectively. Unsupervised multi-omics factor analysis was used to assess the co-variation of bacterial and viral communities, which served as primary predictor. The clinical outcomes of interest were the time to clinical stability and the length of hospital stay. Findings 64 patients and 38 controls were analysed. Rectal bacterial alpha (p = 0•0015) and beta diversity (r(2)=0•023, p = 0•004) of CAP patients differed from controls. Bacterial and viral microbiota signatures correlated with the time to clinical stability (hazard ratio 0•43, 95% confidence interval 0•20–0•93, p = 0•032) and the length of hospital stay (hazard ratio 0•37, 95% confidence interval 0•17–0•81, p = 0•012), although only the latter remained significant following p-value adjustment for examining multiple candidate cut-points (p = 0•12 and p = 0•046, respectively). Interpretation This exploratory study provides preliminary evidence that intestinal bacteriome and virome signatures could be linked with clinical outcomes in CAP. Such exploratory data, when validated in independent cohorts, could inform the development of a microbiota-based diagnostic panel used to predict clinical outcomes in CAP. Funding Netherlands Organization for Scientific Research and Netherlands Organization for Health Research and Development.