An evolutionary and genomic approach to understand tumor evolution in hepatocellular carcinoma

Abstract

Hepatocellular carcinoma (HCC) is one of the deadliest cancer types with diverse etiological factors across the world and very limited treatment options. Although large-scale genomic and transcriptomic studies have been conducted in different cohorts, an integrative analysis of HCC genomes and the ethnic comparison across ethnic backgrounds is lacking. In the second chapter, we integrated 1349 HCC genomes from five Asian and/or European cohorts and identified a large number of novel drivers (n=29) (e.g. FRG1). Many of these novel drivers are infrequent tumor suppressor genes (TSGs) and tend to enrich in several important pathways including the chromatin remodeling pathway. Novel drivers including PBRM1 and KMT2D often occur in later stages of tumorigenesis indicating their potential roles in driving tumor progression. In the third chapter, in order to understand ethnic differences in HCC, we conducted a systematic comparison across the Asian and European HCCs using the data from the The Cancer Genome Atlas (TCGA) database. We found higher genomic instability in Asians with a series of molecular events ranging from driver genes to immune profiles segregating distinctively between the two ethnic backgrounds. Most strikingly, multiple Asian enriched genomic alterations in particular chromosome 16 deletion, lead to a clinically aggressive RNA subgroup unique to Asians. By integrating information across multiple layers, we found that integrative survival models predict patient prognosis much better in Asians, suggesting a better chance to conduct a precision medicine program in Asia. Taken together, we performed a comprehensive integrative analysis of HCC genomes and uncovered remarkable ethnic differences between Asians and Europeans. In the fourth chapter of the thesis, leveraging the multi-region data of an Asian cohort, PLANET, variable level of intra-tumor heterogeneity (ITH) was found in both genomes and transcriptomes of HCC. Strikingly, we found multiple RNA subtypes within a single patient in the PLANET cohort which might be one reason for the poor treatment response in HCC. Integrative survival analysis highlighted the important roles of DNA and RNA ITH in predicting progression-free patient survival. Taken together, we have drawn a comprehensive landscape of ITH across a prospective cohort for HCC.