Severe COVID-19 is characterised by inflammation and immature myeloid cells early in disease progression

Summary SARS-CoV-2 infection causes a wide spectrum of disease severity. Immune changes associated with severe disease include pro-inflammatory cytokine production and expansion of immature myeloid populations. The relative importance of the immunological changes in driving progression to severe disease remain poorly understood. We aimed to identify and rank clinical and immunological features associated with progression to severe COVID-19. We sought to use tests available in an on-site diagnostic hospital laboratory to identify an immunological signature for severe disease development which could be detected prior to peak severity thereby allowing initiation of therapeutic interventions. We used univariate and multivariate analysis, including unbiased machine learning, to investigate the relationships between clinical and demographic characteristics, inflammatory markers, and leukocyte immunophenotypes with progression to severe disease in 108 patients and to rank these in importance. A combination of four features (elevated levels of interleukin-6 and C-reactive protein, coupled with reduced monocyte HLA-DR expression and reduced neutrophil CD10 expression), were strongly predictive of severe disease with an average prediction score of 0.925. Highlights Severe COVID-19 can be predicted by a combination of emergency myelopoiesis (CD10-neutrophils and HLA DR-monocytes) and inflammation (raised IL-6 and CRP) These changes can be identified from tests carried out prior to peak illness severity in a diagnostic laboratory This predictive model was derived from a cohort of patients with a wide range of ages, frailty and COVID-19 severity Graphical Abstract