Dysregulated immune networks define distinct respiratory phenotypes of post-acute sequelae of COVID-19

Abstract Long-haul respiratory symptoms are common following severe COVID-19 illness; however, the immune mechanisms contributing to post-acute sequelae of COVID-19 (PASC) remain enigmatic. A comprehensive 2-year longitudinal immune assessment was performed in a cohort of PASC patients receiving pulmonary follow-up care (n=110) by analyzing circulating immune cells using spectral flow cytometry, as well as assessment of virus-specific T cell frequencies, serum antibodies, and plasma mediators. Novel unsupervised machine-learning workflows were used to analyze and integrate multiple clinical and immune measures. Our approach identified discrete respiratory phenotypes according to lung physiology measures that defined the degree of lung fibrosis and its severity. Notably, severe respiratory phenotypes were characterized by immune dysregulation of both innate and adaptive cellular components, including lower ILC-3-like cells and higher activated CD4+ and CD8+ T cells in the blood. On the other hand, all respiratory phenotypes displayed a marked loss of naïve CD8+ T cells and unswitched memory B cells compared to healthy individuals. Spike- and nucleoprotein-specific CD4+ and CD8+ T cells were more abundant in severe respiratory phenotypes. Overall, T cell-derived inflammatory cytokines, along with chemokine receptors and their relationships to cognate ligands, pointed to sustained lung-homing of pathogenic effector cells in fibrotic phenotypes. Integration of high-dimensional data defined discrete cellular and molecular immune networks that provide unconventional and novel mechanistic insight into pulmonary complications post-COVID-19. UO1AI125056 R21 AI160334