TABLE 3.
Metabolomics based potential prognostic markers for COVID-19.
| Specimen | No. of patients | Metabolomics approach (N) | Patient groups | Severity marker/criteria | Metabolic pathway/s affected | Potential metabolomic markers | Association | Strength/weakness | References |
| Serum | 65 | GC-MS (46) | Mild vs. severe COVID-19 patients | Respiratory failure, respiratory rate > 30 bpm, O2 saturation < 92%, PaO2/FiO2 < 300 mmHg52. | Valine and threonine catabolism | α-hydroxyl acids | Correlation with O2 saturation/lung damage FC: 1.8–2.3; adjusted p < 0.05 | Limited data | Paez-Franco et al., 2021 |
| Plasma | 104 | GC-MS UHPLC/MS (77) | COVID-19 PCR positive vs. negative patients with flu-like symptoms | Mild – symptoms, no CT scan or hospitalization Moderate – dyspnea, pneumonia by CT scan, hospitalization, O2 supplementation Critical – ICU admission, respiratory distress, intubation and mechanical ventilation | Kynurenine pathway | Anthranilic acid | Correlation with poor prognosis and high IL-10/18 (R = 0.55/0.46; p = 0.0092/0.037) | Correlation with known immunosuppressive role Small sample size | Danlos et al., 2021 |
| Plasma | 30 | LC-MS/MS, NMR (162) | COVID-19 positive and negative ICU patients and healthy controls | ICU admission, mortality | – | Kynurenine Creatinine/arginine ratio | Prediction of COVID-19 associated death (accuracy = 100%) | KP pathway involvement confirmed in multiple studies Small sample size | Fraser et al., 2020 |
| Serum | 49 | UHPLC-MS | COVID-19 positive and negative patients | Severity inferred from IL-6 levels, CRP and BUN | Tryptophan metabolism/kynurenine pathway | Acylcarnitines, kynurenine, and methionine sulfoxide | Positive correlation with IL-6 [–log(p) > 2] | KP pathway involvement confirmed in multiple studies Small sample size | Thomas et al., 2020 |
| Free fatty acid and tryptophan levels | Negative correlation with CRP [–log(p) > 2] | ||||||||
| Acylcarnitines | Positive correlation with BUN [–log(p) > 2] | ||||||||
| Serum | 187 | GC-MS (75) | Mild vs. severe disease | Dyspnea, respiratory rate ≥ 30/min; O2 saturation ≤ 93%, PaO2/FiO2 < 300 mmHg52, lung infiltrates > 50% | – | 2-hydroxy-3- methylbutyric acid, 3-hydroxybutyric acid, cholesterol, succinic acid, L-ornithine, oleic acid and palmitelaidic acid | Progression from mild to severe COVID-19 (AUC 0.969) | Free fatty acid changes observed in multiple studies Small sample size | Shi et al., 2021 |
| Plasma | 49 | LC-MS/MS (221) | Moderate, severe and critical | O2 saturation, analytical parameters and radiological findings | Ceramides, tryptophan metabolism, and NAD-consuming reactions | Kynurenine/tryptophan, 3-hydroxykynurenine/kynurenine, and 3-hydroxykynurenine/tryptophan | Increase with COVID-19 severity (p ≤ 0.001) | KP pathway involvement confirmed in multiple studies Small sample size | Marin-Corral et al., 2021 |
| Serum | 61 | Targeted: UHPLC-MS/MS (258) Untargeted: UHPLC quadruple TOF high-resolution MS/MS system (155) | Mild, severe COVID-19 patients and healthy control | Respiratory distress, respiratory rate ≥ 30/min; O2 saturation ≤ 93%, PaO2/FiO2 < 300 mmHg5 | Nicotinate and nicotinamide metabolism, tryptophan metabolism, and citrate cycle | – | Correlation with IL-6, IP-10, and M-CSF | KP pathway involvement confirmed in multiple studies Supporting data from longitudinal study Small sample size | Xiao et al., 2021 |
| Plasma | 341 (700 samples – longitudinal over 3 months) | Untargeted: LC-MS/machine learning (235 polar metabolite and 472 lipid metabolite) | COVID-19 positive patients with different levels of severity and COVID-19 negative patients | Symptoms, hospital and ICU admission, mechanical ventilation, death | – | 25 predictor metabolites | Predict disease severity | Results confirmed by animal testing Longitudinal study Predictor metabolites remained uncharacterized | Sindelar et al., 2021 |
| Plasma | 161 | Untargeted: UPLC-MS/MS GCxGC-MS (2075 lipids and 500 small molecules) | Critical and non-critical COVID-19 patients and healthy controls | Mild to severe – O2 supplementation, no mechanical or non-invasive ventilation Critical – respiratory failure, ICU admission, mechanical ventilation | Gluconeogenesis and the metabolism of porphyrins | Arachidonic acid and oleic acid | Correlated to severity of the disease (AUC > 0.98) | Free fatty acid changes observed in multiple studies Small sample size | Barberis et al., 2020 |
| Plasma | 85 | NMR, LC-MS/MS; Multi-omics (348) | Mild to severe vs. critical | Mild – clinical signs of pneumonia but without O2 support Severe – with O2 support using non-invasive ventilation, tracheal tube, tracheotomy assist ventilation, or ECMO | Lipoprotein metabolism | HDL1, HDL4, LDL1, LDL4, LDL5, VLDL5, ApoA1, triglycerides, cholesterol | Disease severity | Preliminary data on potential biomarkers | Chen et al., 2020 |
| Blood, urine | 30 | LC-MS Multi-omics: proteome, amino acids and lipidome (1254 proteins 664 lipids) | Severe vs. non-severe | Common – symptoms, pneumonia Severe – respiratory distress, respiratory rate > 30 bpm, O2 saturation < 92% at rest, PaO2/FiO2 < 300 mmHg52. Critical – respiratory failure, mechanical ventilation, shock, ICU admission | – | 21 lipids and 4 proteins | AUC: 0.993 to classify severe patients | Limited data | Li et al., 2021 |
| – | 96 | Bioinformatics analysis of published metabolomics data of COVID-19 | Non-severe and severe COVID-19 patients, healthy controls and non-COVID-disease controls | Mild – symptoms without pneumonia Typical – fever or respiratory tract symptoms with pneumonia Severe – respiratory distress, respiratory rate > 30 bpm, O2 saturation < 92% at rest, PaO2/FiO2 < 300 mmHg52 Critical – respiratory failure, mechanical ventilation, shock, ICU admission, other organ failure | Nucleic acid and amino acid metabolism | Taurochenodeoxycholic acid 3-sulfate, glucuronate and N,N,N-trimethyl-alanylproline betaine TMAP | Top classifier of severe disease (ROC 0.805) | In silico data, requires experimental and clinical validation | Chen et al., 2021 |
N, number of metabolites assessed; PaO2/FiO2, partial pressure of arterial oxygen to fraction of inspired oxygen; GC-MS, gas chromatography mass spectrometry; LC-MS, liquid chromatography mass spectrometry; LC-MS/MS, liquid chromatography-tandem mass spectrometry; NMR, nuclear magnetic resonance; UHPLC-MS, ultra-high performance liquid chromatography mass spectrometry; UHPLC-MS/MS, ultra-high performance liquid chromatography tandem mass spectrometry; UPLC-MS/MS, ultra-performance liquid chromatography tandem mass spectrometer; GCxGC-MS, comprehensive two-dimensional gas chromatography mass spectrometry; ECMO, extracorporeal membrane oxygenation; FC, fold change; AUC, area under curve; ROC, receiver operating characteristic.