Molecular genomic landscapes of hepatobiliary cancer

Abstract

Hepatocellular carcinoma (HCC) and biliary tract cancer are more frequent in East Asia including Japan than in Europe or North America. A compilation of 1340 multi-ethnic HCC genomes, the largest cohort ever reported, identified a comprehensive landscape of HCC driver genes, comprised of three core drivers (TP53, TERT, and WNT signaling) and combinations of infrequent alterations in various cancer pathways. In contrast, five core driver genes (TP53, ARID1A, KRAS, SMAD4, and BAP1) with characteristic molecular alterations including fusion transcripts involving fibroblast growth factor receptor 2 and the protein kinase A pathway, and IDH1/2 mutation constituted the biliary tract cancer genomes. Consistent with their heterogeneous epidemiological backgrounds, mutational signatures and combinations of non-core driver genes within these cancer genomes were found to be complex. Integrative analyses of multi-omics data identified molecular classifications of these tumors that are associated with clinical outcome and enrichments of potential therapeutic targets, including immune checkpoint molecules. Translating comprehensive molecular-genomic analysis together with further basic research and international collaborations are highly anticipated for developing precise and better treatments, diagnosis, and prevention of these tumor types.

Keywords: biliary tract cancer; driver gene; hepatocellular carcinoma; molecular classification; mutational signature.

Figure 1

Frequencies of significantly mutated driver genes in 1340 hepatocellular carcinoma genomes. Genes were sorted by q‐value. Somatic mutations in the coding exons of 1340 hepatocellular carcinoma genomes were accumulated and the statistical significance of mutation frequencies in driver genes were estimated. Briefly, the P‐values from each individual significance test were calculated as the geometric mean, as previously reported.5 Multiple testing adjustments (q‐value) were carried out as described.45 Note that the frequency of TERT promoter mutations was calculated by combining data from four studies5, 7, 8, 9

Figure 2

Landscapes of major driver genes in hepatocellular carcinoma (HCC) and biliary tract cancer (BTC). A, Combination of three core driver genes (TP53,TERT, and WNT signal) and other driver pathways in HCC. B, Combination of five core driver genes (TP53,ARID1A,KRAS,SMAD4, and BAP1) and other driver pathways in BTC. TGFβ, transforming growth factor‐β

Figure 3

Mutational signatures in the hepatocellular carcinoma (HCC) and biliary tract cancer (BTC) genomes. Mutational signatures extracted from previous sequencing analyses of HCC and BTC genomes are classified by their mechanisms. Direct causes of these signatures, if identified, are indicated. Representative mutational spectra with sequence context (from the Catalogue of Somatic Mutations in Cancer database) are shown

Figure 4

Specific driver genes across anatomical subtypes of biliary tract cancer (BTC), and comparison with core driver genes in hepatocellular carcinoma (HCC) and pancreatic cancer. ECC, extrahepatic cholangiocarcinoma; GB, gallbladder cancer; ICC, intrahepatic cholangiocarcinoma

Figure 5

Diverse fusion partners with FGFR2 reported in biliary tract cancer. Nineteen fusion partner genes with FGFR2 identified so far are shown with structural rearrangements. References are listed in Table S3

Figure 6

Molecular classifications of hepatocellular carcinoma (HCC) and biliary tract cancer (BTC) by integrative molecular analyses. (a) Clinical and molecular features in subtypes of HCC reported in three studies8, 42, 43 are summarized. TPM,TERT promoter mutation. (b) Clinical and molecular features in subtypes of BTC or cholangiocarcinoma (CCA) reported in four studies24, 27, 30, 43 are summarized. Subtypes colored green showed better prognosis; those colored orange showed poorer prognosis. ECC, extrahepatic cholangiocarcinoma; ICC, intrahepatic cholangiocarcinoma; JP‐UM, Japanese‐specific ICC subtype; TCGA, The Cancer Genome Atlas