Oxidative stress markers and prediction of severity by a machine learning approach in hospitalized patients with COVID-19 and severe lung disease: a feasibility study (Preprint)

Abstract

BACKGROUND Oxidative Stress (OS) is an imbalance between the production of free radicals and the body's ability to neutralize them, leading to damages of cells, proteins and deoxyribonucleic. OBJECTIVE To identify the relevant biomarkers of OS which could be associated to severity of hospitalized patients and to identify a possible correlation between OS biomarkers and clinical status of hospitalized COVID-19 patients with severe lung disease at hospital admission. METHODS All adult patients hospitalized with COVID-19 at the Infirmerie Protestante (Lyon, France) from 9th February 2022 to 18th May 2022 were included, regardless of the care service. The final sample consisted in 28 patients. Ten biomarkers were collected per patient (Zinc (Zn), Copper (Cu), Cu/Zn, Selenium, Uric acid, CRplus, Oxidized LDL, Glutathione peroxidase, Glutathione reductase and Thiols), as well as demographic variables and comorbidities. A support vector machine (SVM) model was used to predict the severity grade per patient, based on the collected data as training set. RESULTS : Three biomarkers of OS were associated with severity; Zn, Cu/Zn and Thiols especially for grade 0 (asymptomatic) and grade 1 (mild to moderate severity). The SVM model predicted the level of severity from the biological analysis of the OS biomarkers with only 7.14% of discrepancy in the training dataset. CONCLUSIONS In case of COVID-19 infection, moderate to severe symptomatic patients are associated with a lowered zinc level, a lowered plasma thiol level, an increased CRPus and an increased Cu/Zn ratio among a panel of ten biomarkers of OS.