Hyperinflammatory ARDS is characterized by interferon-stimulated gene expression, T-cell activation, and an altered metatranscriptome in tracheal aspirates

BackgroundTwo molecular phenotypes of the acute respiratory distress syndrome (ARDS) with divergent clinical trajectories and responses to therapy have been identified. Classification as "hyperinflammatory" or "hypoinflammatory" depends on plasma biomarker profiling. Differences in pulmonary biology underlying these phenotypes are unknown. MethodsWe analyzed tracheal aspirate (TA) RNA sequencing (RNASeq) data from 41 ARDS patients and 5 mechanically ventilated controls to assess differences in lung inflammation and repair between ARDS phenotypes. In a subset of subjects, we also analyzed plasma proteomic data. We performed single-cell RNA sequencing (scRNASeq) on TA samples from 9 ARDS patients. We conducted differential gene expression and gene set enrichment analyses, in silico prediction of pharmacologic treatments, and compared results to experimental models of acute lung injury. FindingsIn bulk RNASeq data, 1334 genes were differentially expressed between ARDS phenotypes (false detection rate < 0.1). Hyperinflammatory ARDS was characterized by an exaggerated innate immune response, increased activation of the integrated stress response, interferon signaling, apoptosis, and T-cell activation. Gene sets from experimental models of lipopolysaccharide lung injury overlapped more strongly with hyperinflammatory than hypoinflammatory ARDS, though overlap in gene expression between experimental and clinical samples was variable. ScRNASeq demonstrated a central role for T-cells in the hyperinflammatory phenotype. Plasma proteomics confirmed a role for innate immune activation, interferon signaling, and T-cell activation in the hyperinflammatory phenotype. Predicted candidate therapeutics for the hyperinflammatory phenotype included imatinib and dexamethasone. InterpretationHyperinflammatory and hypoinflammatory ARDS phenotypes have distinct respiratory tract biology, which could inform targeted therapeutic development. FundingNational Institutes of Health; University of California San Francisco ImmunoX CoLabs; Chan Zuckerberg Foundation; Genentech