Metabolic markers distinguish COVID-19 from other intensive care patients and show potential to stratify for disease risk

2021 
Coronavirus disease 2019 (COVID-19) is a viral infection affecting multiple organ systems of great significance for metabolic processes. Thus. there is increasing interest in metabolic and lipoprotein signatures of the disease and early analyses have demonstrated metabolic pattern typical for atherosclerotic and hepatic damage in COVID-19 patients. However, it remains unclear whether these are specific for COVID-19 or a general marker of critical illness. To answer this question, we have analyzed 276 serum samples from 92 individuals using NMR metabolomics, including longitudinally collected samples from 5 COVID-19 and 11 cardiogenic shock intensive care patients, 18 SARS-CoV-2 antibody-positive individuals, and 58 healthy controls. COVID-19 patients showed a distinct metabolic serum profile, including changes typical for severe dyslipidemia and a deeply altered metabolic status compared to healthy controls. Specifically, VLDL parameters, IDL particles, large-sized LDL particles, and the ApoB100/ApoA1 ratio were significantly increased, whereas HDL fractions were decreased. Moreover, a similarly perturbed profile was apparent, even when compared to other ICU patients suffering from cardiogenic shock, highlighting the impact of COVID-19 especially on lipid metabolism and energy status. COVID-19 patients were separated with an AUROC of 1.0 when compared to both healthy controls and cardiogenic shock patients. Anti-SARS-CoV-2 antibody-positive individuals without acute COVID-19 did not show a significantly perturbed metabolic profile compared to age- and sex-matched healthy controls, but SARS-CoV-2 antibody-titers correlated significantly with metabolic parameters, including levels of glycine, ApoA2, and small-sized LDL and HDL subfractions. Our data suggest that NMR metabolic profiles are suitable for COVID-19 patient stratification and post-treatment monitoring.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    38
    References
    1
    Citations
    NaN
    KQI
    []