Establishing thresholds for cytokine storm and defining their relationship to disease severity in respiratory viral infections (preprint)

Previous studies have identified cytokines associated with respiratory virus infection illness outcome. However, few studies have included comprehensive cytokine panels, longitudinal analyses, and/or simultaneous assessment across the severity spectrum. This, coupled with subjective definitions of cytokine storm syndrome (CSS), have contributed to inconsistent findings of cytokine signatures, particularly with COVID severity. Here, we measured 38 plasma cytokines and compared profiles in healthy, SARS-CoV-2 infected, and multisystem inflammatory syndrome in children (MIS-C) patients (n = 169). Infected patients spanned the severity spectrum and were classified as Asymptomatic, Mild, Moderate or Severe. Our results showed acute cytokine profiles and longitudinal dynamics of IL1Ra, IL10, MIP1b, and IP10 can differentiate COVID severity groups. Only 4% of acutely infected patients exhibited hypercytokinemia. Of these subjects, 3 were Mild, 3 Moderate, and 1 Severe, highlighting the lack of association between CSS and COVID severity. Additionally, we identified IL1Ra and TNFa as potential biomarkers for patients at high risk for long COVID. Lastly, we compare hypercytokinemia profiles across COVID and influenza patients and show distinct elevated cytokine signatures, wherein influenza induces the most elevated cytokine profile. Together, these results identify key analytes that, if obtained at early time points, can predict COVID illness outcome and/or risk of complications, and provide novel insight for improving the conceptual framework of hypercytokinemia, wherein CSS is a subgroup that requires concomitant severe clinical manifestations, and including a list of cytokines that can distinguish between subtypes of hypercytokinemia.

Convergent epitope-specific T cell responses after SARS-CoV-2 infection and vaccination (preprint)

SARS-CoV-2 mRNA vaccines, including Pfizer/Biontech BNT162b2, were shown to be effective for COVID-19 prevention, eliciting both robust antibody responses in naive individuals and boosting pre-existing antibody levels in SARS-CoV-2-recovered individuals. However, the magnitude, repertoire, and phenotype of epitope-specific T cell responses to this vaccine, and the effect of vaccination on pre-existing T cell memory in SARS-CoV-2 convalescent patients, are still poorly understood. Thus, in this study we compared epitope-specific T cells elicited after natural SARS-CoV-2 infection, and vaccination of both naive and recovered individuals. We collected peripheral blood mononuclear cells before and after BNT162b2 vaccination and used pools of 18 DNA-barcoded MHC-class I multimers, combined with scRNAseq and scTCRseq, to characterize T cell responses to several immunodominant epitopes, including a spike-derived epitope cross-reactive to common cold coronaviruses. Comparing responses after infection or vaccination, we found that T cells responding to spike-derived epitopes show similar magnitudes of response, memory phenotypes, TCR repertoire diversity, and {beta}TCR sequence motifs, demonstrating the potency of this vaccination platform. Importantly, in COVID-19-recovered individuals receiving the vaccine, pre-existing spike-specific memory cells showed both clonal expansion and a phenotypic shift towards more differentiated CCR7-CD45RA+ effector cells. In-depth analysis of T cell receptor repertoires demonstrates that both vaccination and infection elicit largely identical repertoires as measured by dominant TCR motifs and receptor breadth, indicating that BNT162b2 vaccination largely recapitulates T cell generation by infection for all critical parameters. Thus, BNT162b2 vaccination elicits potent spike-specific T cell responses in naive individuals and also triggers the recall T cell response in previously infected individuals, further boosting spike-specific responses but altering their differentiation state. Overall, our study demonstrates the potential of mRNA vaccines to induce, maintain, and shape T cell memory through vaccination and revaccination.

Targeted Immunosuppression Distinguishes COVID-19 from Influenza in Moderate and Severe Disease (preprint)

Coronavirus disease 2019 (COVID-19) is characterized by a high incidence of acute respiratory failure. The underlying immunopathology of that failure and how it compares to other causes of severe respiratory distress, such as influenza virus infection, are not fully understood. Here we addressed this by developing a prospective observational cohort of COVID-19 and influenza subjects with varying degrees of disease severity and assessing the quality and magnitude of their immune responses at the cellular and protein level. Additionally, we performed single-cell RNA transcriptional profiling of peripheral blood mononuclear cells from select subjects. The cohort consists of 79 COVID-19 subjects, 26 influenza subjects, and 15 control subjects, including 35 COVID-19 and 7 influenza subjects with acute respiratory failure. While COVID-19 subjects exhibited largely equivalent or greater activated lymphocyte counts compared to influenza subjects, they had fewer monocytes and lower surface HLA-class II expression on monocytes compared to influenza subjects and controls. At least two distinct immune profiles were observed by cytokine levels in severe COVID-19 patients: 3 of 71 patients were characterized by extreme inflammation, with greater than or equal to ~50% of the 35 cytokines measured greater than 2 standard deviations from the mean level of other severe patients (both influenza and COVID-19); the other immune profile, which characterized 68 of 71 subjects, had a mixed inflammatory signature, where 28 of 35 cytokines in COVID-19 patients had lower mean cytokine levels, though not all were statistically significant. Only 2 cytokines were higher in COVID-19 subjects compared to influenza subjects (IL-6 and IL-8). Influenza and COVID-19 patients could be distinguished statistically based on cytokine module expression, particularly after controlling for the significant effects of age on cytokine expression, but again with lower levels of most cytokines in COVID-19 subjects. Further, high circulating levels of IL-1RA and IL-6 were associated with increased odds of intubation in the combined influenza and COVID-19 cohort [OR = 3.93 and 4.30, respectively] as well as among only COVID-19 patients. Single cell transcriptional profiling of COVID-19 and influenza subjects with respiratory failure identified profound suppression in type I and type II interferon signaling in COVID-19 patients across multiple clusters. In contrast, COVID-19 cell clusters were enriched for alterations in metabolic, stress, and apoptotic pathways. These alterations were consistent with an increased glucocorticoid response in COVID-19 patients compared to influenza. When considered across the spectrum of innate and adaptive immune profiles, the immune pathologies underlying severe influenza and COVID-19 are substantially distinct. The majority of COVID-19 patients with acute respiratory failure do not have a cytokine storm phenotype but instead exhibit profound type I and type II IFN immunosuppression when compared to patients with acute influenza. Upregulation of a small number of inflammatory mediators, including IL-6, predicts acute respiratory failure in both COVID-19 and influenza patients.