We have recently sequenced 142 exomes from a prospectively accrued cohort of early stage pancreatic ductal adenocarcinoma (PDAC), 99 of which had sufficient tumor cellularity for detailed analysis [1]. From these, we identified 1642 mutated genes, >99% of which were mutated at low frequency (1-5%).
Prioritising driver mutations remains a major challenge in genome sequencing projects. We have used a combination of algorithms, pathway enrichment analyses, and functional genomics to identify driver mutations. The functional genomics strategy involved the integration of mutation data from human tumours, with shRNA functional screens from 102 cancer cell lines, as well as murine in vivo Sleeping Beauty mutagenesis screens. Through these analyses we defined 16 significantly frequently mutated genes, reaffirming known mutations (KRAS, TP53, CDKN2A, SMAD4), and uncover novel mutated genes including additional genes involved in chromatin modification (EPC1 and ARID2), DNA damage repair (ATM) and other mechanisms, all of which have functional support from at least one of the functional screens. Through pathway analyses we identified a robust enrichment of genes traditionally annotated as embryonic regulators of axon guidance, especially ROBO/SLIT signaling. Through the functional genomics strategy, we identified SOX9 as a candidate oncogene in Pancreatic Cancer, which appears to be a driver of early carcinogenesis in our preliminary follow up experiments.
Despite the apparent heterogeneity of PDAC, we have begun to identify low (1-5%) prevalence phenotypes, which may be clinically relevant. We identified a HER2-amplified patient in our cohort, which we have confirmed as having a prevalence of 2.1% in an independent cohort of 469 PDAC tumours. HER2-amplified PDAC patients have markedly atypical patterns of recurrence, the tumor appears to be driven by the HER2 amplicon, and possibly had a clinical response to Tarceva, an EGFR inhibitor.