ABSTRACT
Purpose: In general, children have less severe disease with SARSCoV2 infection than adults. However, some children do experience life-threatening sequelae from infection. In order to analyze inflammatory responses to SARS-CoV2 and clinical outcomes, this study evaluates the plasma protein profiles of pediatric COVID19 patients compared to HLH, severe sepsis, Kawasaki disease, febrile viral illness and healthy controls Methods: The Luminex platform measured 150 analytes in 108 patients with pediatric COVID19, 32 of which developed MIS-C, 16 with HLH, 14 with severe sepsis, 25 with Kawasaki disease, 21 with febrile viral illness, and 20 health controls. The results were tested for significant proteins with a p<=0.05 and those that pasted with an FDR cut off of 0.1 and 80% confidence. Semi-supervised learning using protein analyte profiles of inflammatory disease were used to predict COVID19 similarities and clinical outcomes were compared Results: Analyte comparison COVID19 patients revealed increase in CXCL9 in patients with MIS-C. Semi-supervised predictions revealed HLH as the most common predictor and showed that an HLH plasma signature in pediatric patients with COVID19 was associated with higher instances of MIS-C, longer hospital stay, and higher instances of respiratory failure Conclusion(s): This evaluation of plasma proteins demonstrated that pediatric patients with SARS-Cov2 infection with inflammatory plasma profile similar to an HLH signature experienced more severe disease These results reflect complex range of immune responses in children with COVID19, and they support potential for prospective risk stratification using plasma biomarkers.