Abstract
Metabolic reprogramming is a hallmark of tumors, including hepatocellular carcinoma (HCC). We used data from The Cancer Genome Atlas and the International Cancer Genome Consortium to assess the alterations in glycolytic and cholesterogenic genes in HCC and to determine their association with clinical features in HCC patients. Based on the gene expression profiles from these databases, we established four subtypes of HCC: cholesterogenic, glycolytic, mixed, and quiescent. The prognosis of the cholesterogenic subgroup was poorer than that of the glycolytic group. Tumors in the glycolytic group were more sensitive to chemotherapy. We also explored the relationships between these metabolic subtypes and previously established HCC subgroups. Glycolytic gene expression correlated strongly with poorer prognostic gene expression in the Hoshida classification of HCC. Whole-genome analyses indicated that aberrant amplification of TP53 and MYC in HCC were associated with abnormal anabolic cholesterol metabolism. The mRNA levels of mitochondrial pyruvate carriers 1 and 2 differed among the HCC metabolic subtypes. In a bioinformatics analysis we identified genomic characteristics of tumor metabolism that varied among different cancer types. These findings demonstrate that metabolic subtypes may be valuable prognostic indicators in HCC patients.