Dynamic Patterns of serum metabolites in fulminant hepatic failure pigs

Fulminant hepatic failure (FHF) is still an intractable disease associated with serious metabolic disorder. Investigating the dynamic changes of serum metabolites during the development of FHF would facilitate revealing the pathogenesis and also promote its treatment. Therefore, this study characterized the dynamic metabonome of serum from FHF Pigs using ultra performance liquid chromatography-mass spectrometry. Based on multiple statistical analysis of the resulting dataset, three types of up-regulated and one type of down-regulated patterns were delineated. Each pattern demonstrated distinct trends at different stages during the whole process of FHF, implying the differential clinical significance of them. Specifically, aromatic amino acids (Pattern 1) and lysophosphatidylcholines (LPCs) (Pattern 4) might be good markers for evaluating the severity of FHF, while some conjugated bile acids, long chain acylcarnitines (Pattern 2) and Glycocholic acid (Pattern 3) could indicate liver injury in the early stage. Inspired from the PCA plot that the pathogenetic condition of FHF aggravated with sampling time, a linear discriminant analysis (LDA) model based on phenylalanine and LPC 18:1 were further constructed for evaluating the severity of FHF. The leave-one-out cross-validation accuracy of 91.67% for the training set and the prediction accuracy of 92.31% for the external validation set confirmed its excellent performance. In conclusion, findings obtained from the present study, including four types of Dynamic Patterns of serum metabolites during FHF development and an LDA model for evaluating the severity of FHF, will be of great help to the research and management of FHF in the future.