Integrated Genomic Studies to Discern Tumor Heterogeneity and Prognostic Subgroups in Hepatocellular Carcinoma

Abstract

Globally, hepatocellular carcinoma (HCC) accounts for 70-85% of primary liver cancers and is the second leading cause of male cancer death. Among patients there is widespread tumor heterogeneity, yet only a single FDA-approved therapeutic, sorafenib, for standard of care in advanced HCC. Identifying homogeneous subgroups of patients will allow development of effective treatments targeted to specific pathways that drive poor survival. In this study, I took both a specific and global approach to characterize HCC tumors with poor prognosis. In the specific approach, I focused on serum alpha-fetoprotein (AFP), a diagnostic biomarker for HCC. AFP, normally only expressed in the liver during fetal development, is reactivated in 60% of HCC tumors and associated with poor outcome. Therefore, I hypothesized that AFP+ and AFP- tumors differ biologically. Using microRNA and mRNA microarray-profiling in predominantly Hepatitis B+ HCC cases from Zhongshan Hospital, I found that miR-29 family members were the most significantly down-regulated miRNAs in AFP+ HCC. Moreover, global DNA methylation profiling revealed that increased methylation was associated with AFP+ HCC. Experimentally, I demonstrated that AFP inhibited transcription of the miR-29a/b-1 locus and this effect is mediated through the c-MYC protein. These findings suggest that AFP is a functional antagonist of miR-29, which may contribute to epigenetic alterations and poor prognosis in AFP+ HCC. In the more global approach to clarify tumor heterogeneity, I aimed to integrate data from multiple platforms to define prognostic tumor subgroups. First, we identified the tumor-specific genes in two complementary platforms: Illumina 27k methylation and Affymetrix gene expression. An epigenetic-driven signature was derived from the tumor-specific genes that inversely correlated with methylation. Employing the epigenetic-driven gene signature along with a somatic copy number alteration (SCNA)-driven signature developed by the lab, we found that epigenetic changes drive poor outcome in one subset of patients while SCNA-related changes drive poor outcome in a different group. In summary, I found that HCC tumors could be separated into prognostic subgroups based on AFP level, as well as epigenetic-driven or SCNA-driven genetic changes. Future work will elucidate key signaling pathways activated in each subgroup to better characterize the differences in HCC tumors.