Discovery and systematic assessment of early biomarkers that predict progression to severe COVID-19 disease

Background: COVID-19 has a wide clinical spectrum from asymptomatic to severe and fatal disease. Early biomarkers that predict progression to severe disease could guide care and enable early intervention. Methods: Plasma protein biomarkers were measured using antibody microarrays (up to 998 proteins) in independent cohorts of SARS-CoV-2–infected individuals early after infection to identify markers predicting severe/critical versus mild/moderate disease. Results: Eleven protein biomarkers consistently differentiated severity groups across two cohorts. Machine learning selected multi-marker panels with strong prognostic performance, including a four-protein panel and two three-protein panels. Selected biomarkers were validated on orthogonal platforms (clinical CRP and ELISA for S100A8/A9). Conclusions: Early biomarker panels can stratify patients at high risk for severe COVID-19 to prioritize early therapeutic interventions (e.g., antivirals, neutralizing antibodies), potentially improving outcomes and reducing hospital burden in future pandemics.

Katrin Hufnagel, Anahita Fathi, Nadine Stroh, Marco Klein, Florian Skwirtblies, Ramy Girgis, Christine Dahlke, Jörg D. Hoheisel, Camille Lowy, Ronny Schmidt, Anne Griesbeck, Uta Merle, Marylyn M. Addo, Christoph Schröder