Hepatol Res. 2020 Nov 23. doi: 10.1111/hepr.13594. Online ahead of print.
AIM: The aim of this study was to use a metabonomics approach to identify potential biomarkers of exhaled breath condensate (EBC) for predicting the prognosis of acute-on-chronic liver failure (ACLF).
METHODS: Using liquid chromatography mass spectrometry, EBC metabolites of ACLF patients surviving without liver transplantation (n=57) and those with worse outcomes (n=45), and controls (n=16) were profiled from a specialized liver disease center in Beijing. The metabolites were used to identify candidate biomarkers, and the predicted performance of potential biomarkers was tested.
RESULTS: 41 metabolites, involving glycerophospholipid metabolism, sphingolipid metabolism, arachidonic acid metabolism and amino acid metabolism, as candidate biomarkers for discriminating the different outcomes of ACLF were selected. A prognostic model was constructed by a panel of four metabolites including phosphatidylinositol [20:4(5Z,8Z,11Z,14Z)/13:0], phosphatidyl ethanolamine (12:0/22:0), l-metanephrine and ethylbenzene, which could predict the worse prognosis in ACLF patients with sensitivity (84.4%) and specificity (89.5%) (area under the receiver operating characteristic curve [AUC] = 0.859, 95% confidence interval [CI] = 0.787-0.931). Compared with MELD score (AUC = 0.639, 95% CI = 0.526-0.753) and MELD-Na score (AUC = 0.692, 95% CI = 0.582-0.803), EBC-associated metabolite signature model could better predict worse outcomes in patients with ACLF (p<0.05). Using the MELD-Na score and EBC metabolite signatures, a decision tree model was built for predicting the prognosis of ACLF identified on logistic regression analyses (AUC = 0.906, 95% CI = 0.846-0.965).
CONCLUSION: EBC metabolic signatures show promise as potential biomarkers for predicting worse prognosis of ACLF.