The Diagnostic Utility of Host RNA Biosignatures in Adult Patients With Sepsis: A Systematic Review and Meta-Analysis.

Abstract

Objectives: Sepsis is a life-threatening medical emergency, with a profound healthcare burden globally. Its pathophysiology is complex, heterogeneous and temporally dynamic, making diagnosis challenging. Medical management is predicated on early diagnosis and timely intervention. Transcriptomics is one of the novel "-omics" technologies being evaluated for recognition of sepsis. Our objective was to evaluate the performance of host gene expression biosignatures for the diagnosis of all-cause sepsis in adults.

Data sources: PubMed/Ovid Medline, Ovid Embase, and Cochrane databases from inception to June 2023.

Study selection: We included studies evaluating the performance of host gene expression biosignatures in adults who were diagnosed with sepsis using existing clinical definitions. Controls where applicable were patients without clinical sepsis.

Data extraction: Data including population demographics, sample size, study design, tissue specimen, type of transcriptome, health status of comparator group, and performance of transcriptomic biomarkers were independently extracted by at least two reviewers.

Data synthesis: Meta-analysis to describe the performance of host gene expression biosignatures for the diagnosis of sepsis in adult patients was performed using the random-effects model. Risk of bias was assessed according to the Quality Assessment of Diagnostic Accuracy Studies-2 tool. A total of 117 studies (n = 17,469), comprising 132 separate patient datasets, were included in our final analysis. Performance of transcriptomics for the diagnosis of sepsis against pooled controls showed area under the receiver operating characteristic curve (AUC, 0.86; 95% CI, 0.84-0.88). Studies using healthy controls showed AUC 0.87 (95% CI, 0.84-0.89), while studies using controls with systemic inflammatory response syndrome (SIRS) had AUC 0.84 (95% CI, 0.78-0.90). Transcripts with excellent discrimination against SIRS controls include UrSepsisModel, a 210 differentially expressed genes biosignature, microRNA-143, and Septicyte laboratory.

Conclusions: Transcriptomics is a promising approach for the accurate diagnosis of sepsis in adults and demonstrates good discriminatory ability against both healthy and SIRS control subjects.