Integrated multiomics analysis of hepatoblastoma unravels its heterogeneity and provides novel druggable targets

Abstract

Although hepatoblastoma is the most common pediatric liver cancer, its genetic heterogeneity and therapeutic targets are not well elucidated. Therefore, we conducted a multiomics analysis, including mutatome, DNA methylome, and transcriptome analyses, of 59 hepatoblastoma samples. Based on DNA methylation patterns, hepatoblastoma was classified into three clusters exhibiting remarkable correlation with clinical, histological, and genetic features. Cluster F was largely composed of cases with fetal histology and good outcomes, whereas clusters E1 and E2 corresponded primarily to embryonal/combined histology and poor outcomes. E1 and E2, albeit distinguishable by different patient age distributions, were genetically characterized by hypermethylation of the HNF4A/CEBPA-binding regions, fetal liver-like expression patterns, upregulation of the cell cycle pathway, and overexpression of NQO1 and ODC1. Inhibition of NQO1 and ODC1 in hepatoblastoma cells induced chemosensitization and growth suppression, respectively. Our results provide a comprehensive description of the molecular basis of hepatoblastoma and rational therapeutic strategies for high-risk cases.