Identification of Potential Prognosticators for Sepsis through Expression Analysis of Transcriptomic Data from Sepsis Survivors and Nonsurvivors.

BACKGROUND

Infection can be severely complicated by a dysregulated, whole-body inflammatory response known as sepsis. While previous research showed that genetic predisposition is linked to outcome differences, current patient characterization fails to determine which septic patients have greater tendencies to develop into severe sepsis or go into septic shock. As such, the identification of prognostic biomarkers may assist in identifying these high-risk patients and help improve the clinical management of the disease.

OBJECTIVE

In this study, we aimed to identify molecular patterns involved in sepsis. We also aimed to identify essential genes associated with the disease's survival which could serve as potential prognosticators for the disease.

METHODS

We used weighted gene co-expression analysis (WGCNA) to analyze GSE63042, an RNA expression dataset from 129 patients with systemic inflammatory response syndrome or sepsis, including 78 sepsis survivors and 28 sepsis nonsurvivors. This analysis included identifying gene modules that differentiate sepsis survivors from nonsurvivors and qualitatively assessing differentially expressed genes. We then used STRING's protein-protein interaction and gene ontology analysis to determine the functional and pathway relationships of the genes in the top modules. Lastly, we assessed the prognosticator abilities of the hub genes using ROC analysis.

RESULTS

We found four diverse co-expression gene modules significantly associated with sepsis survival. Our differential gene expression analysis, combined with protein-protein interaction and gene ontology analysis, revealed that the hub genes of these modules - , and may serve as candidate markers for sepsis prognosis. These markers were significantly downregulated in sepsis nonsurvivors compared with sepsis survivors.

CONCLUSION

Weighted gene co-expression analysis, gene ontology enrichment analysis, and proteinprotein network interaction analysis of transcriptomic data from sepsis survivors and nonsurvivors revealed , and as potential biomarkers for sepsis prognosis. These genes are associated with functions related to proper immune response, and their downregulation in sepsis nonsurvivors suggests eventual immune exhaustion in late sepsis. Further analyses, however, are necessary to validate their roles in sepsis progression and patient survival.