Abstract Massive tumor sequencing allows us to progress beyond cataloging driver mutations, to discovery of the intricate patterns of genomic alterations that are required to enable a successful tumor. We used patterns of mutation co-occurrence to map activation of the RAS pathway in over 16,000 lung adenocarcinomas. This revealed a network of 40 highly mutually exclusive mutations in 19 genes, 18 of which are known RAS pathway genes (the RTKs EGFR, ERBB2, MET, ALK, RET, ROS1, KIT, FGFR2; NRAS and KRAS, BRAF and CRAF, MEK1/2, CBL, NF1, SOS1 and RIT1), indicating that each mutation is a strong pathway driver, and is redundant with other pathway mutations. Together, these mutations cover 71% of all lung adenocarcinomas, and include several previously uncharacterized mutations. Analysis of the remaining tumors uncovers additional patterns of RAS pathway activation. Some have additional mutations in RAS pathway genes that have the opposite pattern—they are co-occurent with each other, either in an additive manner where several mutations are enriched with each other, such as rare RAF and RAS family mutations, or synergistic, where a particular pair of mutations is highly co-occurrent, often in the same gene. These pairs are exclusive of other pathway mutations, suggesting that they are also strong pathway activators. We also see partially exclusive mutations, including amplifications of several genes and loss of function RAF family mutations that are likely to be weak activators of the RAS pathway in additional tumors. Finally, this approach also identifies recurrent mutations that are independent of other pathway mutations, giving clinical evidence that their function in human tumors is distinct from the common function of other RAS pathway mutations. We validated these patterns of exclusivity, additivity, and synergy in 80,000 genomic profiles from 400 other tumor types, and identified additional exclusive RAS pathway alterations in several other tissues. Across all cancers, 30% of all tumors had exclusive mutations indicating that they were RAS pathway driven. The individual mutations varied greatly between tissues, suggesting that there exist extensive tissue-specific enhancers and inhibitors of each oncogenic mutation, which may be of great relevance to drug and biomarker development. This initial foray into the future of massive tumor genomic datasets shows that we can transform genomic data from individual mutations into patterns of pathway activation, and distinguish pathway-driving mutations from other hotspots, providing a better understanding of the oncogenic strategies used by each human tumor. The degree of mutational exclusivity seen can predict the degree of oncogene addiction, and patterns of co-mutation can suggest strategies for combination therapies, as well as allow reverse translation to enable better matching of preclinical and basic research models to oncogenic strategies that are most relevant to human tumors. This abstract is also being presented as Poster B37. Citation Format: Gerard Manning, Nicholas Lounsbury, Ryan Hartmaier, Sally Trabucco, Ethan Sokol, Jacob Rinaldi, Jinfeng Liu, Florian Gnad. Dissecting the playbook of cancer: Genomic analysis of 100,000 human tumors reveals elaborate patterns of activation of the RTK-RAS-MAPK pathway [abstract]. In: Proceedings of the Fifth AACR-IASLC International Joint Conference: Lung Cancer Translational Science from the Bench to the Clinic; Jan 8-11, 2018; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(17_Suppl):Abstract nr PR03.
Read full abstract