ABSTRACT
BackgroundWhile major progress has been made to establish diagnostic tools for the identification of SARS-CoV-2 infection, determining the severity of COVID-19 remains an unmet medical need. There is a limited availability of hospital resources in this or any pandemic, and appropriately gauging severity would allow for some patients to safely recover in home quarantine, while ensuring that sicker patients get needed care. MethodsWe here developed a blood-based generalizable host-gene-expression-based classifier for the severity of viral infections and validated it in multiple viral infection settings including COVID-19. We used training data (N=705) from 21 retrospective transcriptomic clinical studies of influenza and other viral illnesses looking at a preselected panel of host immune mRNAs. ResultsWe selected 6 host mRNAs and trained a logistic regression classifier with a training cross-validation AUROC of 0.90 for predicting 30-day mortality in viral illnesses. Next, in 1,417 samples across 21 independent retrospective validation cohorts the locked 6-mRNA classifier had an AUROC of 0.91 for discriminating patients with severe vs. non-severe infection. Next, in an independent cohort of prospectively enrolled patients with confirmed COVID-19 (N=97) in Athens, Greece, the 6-mRNA locked classifier had an AUROC of 0.89 for identifying patients with severe respiratory failure or 30-day mortality. Finally, we developed an isothermal qRT-LAMP (loop-mediated isothermal gene expression) assay for the 6-mRNA panel to facilitate implementation as a rapid assay. ConclusionsWith further study, the classifier could assist in the risk assessment of patients with confirmed SARS-CoV-2 infection and COVID-19 to determine severity and level of care, thereby improving patient management and healthcare burden.