50-Gene Risk Profiles in Peripheral Blood Predict COVID-19 Severity

B. M. Juan-Guardela,J. Sun,T. Zhang,B. Xu,Y. Huang,S. Ma,P. L. Molyneaux,T. M. Maher,I Noth,G. C. Michaud,J. D. Herazo-Maya
DOI: https://doi.org/10.1164/ajrccm-conference.2021.203.1_meetingabstracts.a4180
IF: 24.7
2021-01-01
American Journal of Respiratory and Critical Care Medicine
Abstract:Rationale: In this study we aimed to determine whether a gene signature predictive of idiopathic pulmonary fibrosis mortality could be repurposed to predict COVID-19 severity. Methods: We analyzed expression level of 50 genes of the 52-gene signature in publically available datasets from COVID-19-Discovery (N=11), COVID-19-Validation (N=100), IPF-University of Chicago (N=45), and IPF-Imperial College London (N=55) cohorts. A COVID-19-Single cell cohort (N=7 subjects, N=145 single cells) was used to identify the cellular sources of the signature. The Scoring Algorithm of Molecular Subphenotypes (SAMS) was used to calculate Up and Down scores. Results: Up and Down Scores derived from the COVID-19-Discovery cohort were used to identify 50-gene risk profiles in the COVID-19-Validation cohort. A High-risk profile significantly predicted ICU admission (AUC: 0.78, 95%CI:0.68-0.85,P<0.0001) and mechanical ventilation (0.77, 95%CI:0.675-0.84,P<0.0001). High-risk subjects were significantly older, had higher Apache II, SOFA, Charlson Disease Severity Index, CRP, and d-dimer levels. They also had shorter ventilator-free and hospital-free days. Same COVID-19, Up and Down score cutoffs were predictive of transplant-free survival and mortality in IPF-University of Chicago (HR:4.16, 95%CI 1.3-13.28,P=0.0012) and IPF-Imperial College London (HR:4.0,95%CI 1.61-9.92,P=0.001) cohort. 50-gene expression analysis in single cells demonstrated higher proportion of CD14+ and CD16+ monocytes, red blood cells, neutrophils, and dendritic cells in high-risk subjects. Higher proportions of CD4 and CD8 T cells, natural killer, B cells, and immunoglobulin-producing plasmablasts were seen in low-risk subjects. Conclusion: 50-gene risk profiles predict ICU admission and need for mechanical ventilation in COVID-19. Single-cell analysis revealed potentially related immune responses associated with COVID-19 and IPF severity.
What problem does this paper attempt to address?