Targeted Metabolomics Analysis of Bile Acids in Patients with Idiosyncratic Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is rare but clinically important due to a high rate of mortality. However, specific biomarkers for diagnosing and predicting the severity and prognosis of DILI are lacking. Here, we used targeted metabolomics to identify and quantify specific types of bile acids that can predict the severity of DILI. A total of 161 DILI patients were enrolled in this prospective cohort study, as well as 31 health controls. A targeted metabolomics method was used to identify 24 types of bile acids. DILI patients were divided into mild, moderate, and severe groups according to disease severity. A multivariate analysis was performed to identify characteristic bile acids. Then the patients were divided into severe and non-severe groups, and logistic regression was used to identify bile acids that could predict DILI severity. Among the enrolled DILI patients, 32 were in the mild group, 90 were in the moderate group, and 39 were in the severe group. Orthogonal partial least squares-discriminant analysis (OPLS-DA) modeling clearly discriminated among the different groups. Among the four groups, glycochenodeoxycholate (GCDCA), taurochenodeoxycholate (TCDCA), deoxycholic acid (DCA), Nor Cholic acid (NorCA), glycocholic acid (GCA), and taurocholic acid (TCA) showed significant differences in concentration between at least two groups. NorCA, GCDCA, and TCDCA were all independent risk factors that differentiated severe DILI patients from the other groups. The area under the receiver operating characteristic curve (AUROC) of GCDCA, TCDCA, and NorCA was 0.856, 0.792, and 0.753, respectively. Together, these three bile acids had an AUROC of 0.895 for predicting severe DILI patients. DILI patients with different disease severities have specific bile acid metabolomics. NorCA, GCDTA, and TCDCA were independent risk factors for differentiating severe DILI patients from less-severe patients and have the potential to predict DILI severity.