Untargeted metabolomics and lipidomics in COVID-19 patient plasma reveals disease severity biomarkers.

Abstract

Introduction: Coronavirus disease 2019 (COVID-19) has widely varying clinical severity. Currently, no single marker or panel of markers is considered standard of care for prediction of COVID-19 disease progression. The goal of this study is to gain mechanistic insights at the molecular level and to discover predictive biomarkers of severity of infection and outcomes among COVID-19 patients.

Method: This cohort study (n = 76) included participants aged 16-78 years who tested positive for SARS-CoV-2 and enrolled in Memphis, TN between August 2020 to July 2022. Clinical outcomes were classified as Non-severe (n = 39) or Severe (n = 37). LC/HRMS-based untargeted metabolomics/lipidomics was conducted to examine the difference in plasma metabolome and lipidome between the two groups.

Results: Metabolomics data indicated that the kynurenine pathway was activated in Severe participants. Significant increases in short chain acylcarnitines, and short and medium chain acylcarnitines containing OH-FA chain in Severe vs. Non-severe group, which indicates that (1) the energy pathway switched to FA β-oxidation to maintain the host energy homeostasis and to provide energy for virus proliferation; (2) ROS status was aggravated in Severe vs. Non-severe group. Based on PLS-DA and correlation analysis to severity score, IL-6, and creatine, a biomarker panel containing glucose (pro-inflammation), ceramide and S1P (inflammation related), 4-hydroxybutyric acid (oxidative stress related), testosterone sulfate (immune related), and creatine (kidney function), was discovered. This novel biomarker panel plus IL-6 with an AUC of 0.945 provides a better indication of COVID-19 clinical outcomes than that of IL-6 alone or the three clinical biomarker panel (IL-6, glucose and creatine) with AUCs of 0.875 or 0.892.