Robust classification of bacterial and viral infections via integrated host gene expression diagnostics

Abstract

Improved diagnostics for acute infections could decrease morbidity and mortality by increasing early antibiotics for patients with bacterial infections and reducing unnecessary antibiotics for patients without bacterial infections. Several groups have used gene expression microarrays to build classifiers for acute infections, but these have been hampered by the size of the gene sets, use of overfit models, or lack of independent validation. We used multicohort analysis to derive a set of seven genes for robust discrimination of bacterial and viral infections, which we then validated in 30 independent cohorts. We next used our previously published 11-gene Sepsis MetaScore together with the new bacterial/viral classifier to build an integrated antibiotics decision model. In a pooled analysis of 1057 samples from 20 cohorts (excluding infants), the integrated antibiotics decision model had a sensitivity and specificity for bacterial infections of 94.0 and 59.8%, respectively (negative likelihood ratio, 0.10). Prospective clinical validation will be needed before these findings are implemented for patient care.

Shared and Distinct Aspects of the Sepsis Transcriptomic Response to Fecal Peritonitis and Pneumonia

“…Additionally, the extent to which the source of infection contributes to heterogeneity in the transcriptomic response is unclear, although there is evidence that gene expression signatures can distinguish between gram-positive, gramnegative, and viral etiologies, and these signatures may be useful in the diagnosis of CAP (6)(7)(8)(9)(10)(11)(12)(13). In fecal peritonitis (FP), sepsis is triggered by a polymicrobial infection within the peritoneal cavity, complicated by the release of damage-associated molecular patterns (DAMPs) and the effects of anesthesia (14).…”

“…These differences seem to be driven predominantly by viral respiratory infection within the CAP cohort; however, given current difficulties in pathogen detection, the biological and clinical interpretation of such transcriptomic differences remains challenging. Whereas some previous studies have supported a common transcriptional septic response independent of pathogen (33), others have reported that expression signatures can discriminate between infecting organisms (6)(7)(8)13), although these findings remain controversial (9,10). For patients admitted to the ICU with suspected CAP, a 78-transcript signature has been reported to differentiate cases of CAP from non-CAP, with the FAIM3: PLAC8 gene expression ratio proposed as a diagnostic biomarker (11).…”

Empowering Multi-Cohort Gene Expression Analysis to Increase Reproducibility

“…Many of these results also been further validated in experimental settings. 8,11,16 These results have further demonstrated the ability of our framework to create "Big Data" by combining multiple smaller studies that are collectively representative of the real word patient population heterogeneity.…”

“…We have repeatedly demonstrated the utility of our framework for identifying novel diagnostic and prognostic biomarkers, drug targets, and repurposing FDA-approved drugs in diverse diseases, including organ transplantation, cancer, infection, and neurodegenerative diseases. [8][9][10][11][12][13][14][15][16] In each of these analyses, we analyzed more than a thousand human samples from more than 10 independent cohorts to generate and validate data-driven hypotheses. Many of these results also been further validated in experimental settings.…”

Multi-Omics Research Trends in Sepsis: A Bibliometric, Comparative Analysis Between the United States, the European Union 28 Member States, and China

“…This is especially relevant for complex diseases and syndromes, such as sepsis, where the disruption of the pathophysiological homeostasis is often caused by numerous, intertwined, subcellular biological events. Integration of data generated by various ''-omics'' technologies sheds light on such disorders within the different ''-omics'' regimes, allowing for precision diagnosis and targeted treatment (Langley and Wong, 2017;Sweeney et al, 2016). In this sense, integration of research efforts (either in terms of ''-omics'' data generated by single ''-omics'' technologies or in terms of multi-omics approaches) has a significant impact on the daily clinical practice; thus, the degree of integration may be considered a potential qualitative indicator of scientific performance on multiple levels, ranging from the researchers and the scientific journals to the research institutes and the (supra)national settings.…”