Host–Microbe Multiomic Profiling Predicts Mortality in Sepsis

RATIONALE: Sepsis is a leading cause of mortality and involves a dysregulated host response to infection. The host and microbe have historically been considered independently in studies of sepsis, limiting our understanding of key relationships driving mortality. OBJECTIVES: We sought to identify host and microbial factors associated with sepsis mortality and build prognostic classifiers. METHODS: We studied 321 critically ill adults and adjudicated sepsis status. From whole blood collected within 24 hours of admission, we performed transcriptional profiling, and from plasma, we performed proteomic and metagenomic analyses. We evaluated associations between in-hospital mortality and gene expression, protein levels, and microbial metagenomic data and built support vector machine-based prognostic classifiers. MEASUREMENTS AND MAIN RESULTS: In patients with sepsis, mortality was associated with increased expression of genes related to neutrophil degranulation, lower expression of genes related to T-cell signaling, and higher IL-8 levels. Mortality was also associated with greater microbial mass and greater bacterial relative dominance. Similar findings were observed in a broader group that also included patients with culture-negative sepsis or indeterminate sepsis status. An integrated host-microbe metagenomic classifier predicted sepsis mortality with an area under the curve (AUC) of 0.79, and a host transcriptomic classifier performed comparably, with an AUC of 0.75. Both performed better (P < 0.05 by paired DeLong tests) that the Acute Physiology, Age, Chronic Health Evaluation III score (AUC of 0.69). CONCLUSIONS: Taken together, our findings provide a conceptual advance in the understanding of host and microbial factors associated with mortality in critical illness and demonstrate a new approach to mortality prediction in sepsis.