DEvelopment of cancer predictive models informed by geneTic and EnviRonMental exposures and their interactions to Improve caNcer screening and prEvention strategies (DETERMINE)

Given the uniqueness of the VA population, any serious attempt to produce precision screening or treatment for cancer risk must account for and understand the effects of military and lifetime environmental exposures. However, uncertainty exists around the specifics of genetic susceptibility to exposures. Furthermore, patterns of mutations in cancer left by military exposures have not been characterized. The overarching goal of this proposal is to improve clinical screening and treatment of Veterans by accounting for both exposures and genetics. Cancers chosen for analysis in this proposal are prostate, lung, and bladder cancer. Exposures chosen for analysis in this proposal are exposure to Agent Orange, contaminated groundwater at Camp Lejeune, cigarette smoke, ambient air pollution, and lifetime radon. We propose to identify genetic loci which more strongly predispose individuals to the carcinogenic effects of these exposures, and to build precision risk prediction models capable of more accurately predicting cancer risk in a VA population. We will additionally characterize mutational changes in the cancer genome that are attributable to specific prior exposures and use the patterns of these mutational changes—and their relationship to germline genetics—to further identify key genetic loci for exposure susceptibility and to improve risk prediction. This proposal is unprecedent in the scale at which it combines exposure, germline, and somatic genotyping data. We additionally propose entirely novel approach to identify mutational signatures that represent prior exposures. This proposal will improve Veteran health care through 1) creation of personalized risk scores for targeted cancer screening 2) unlocking the ability to identify subtypes and signatures of cancer specific to exposures 3) constructing an exposure derived signature map in which clinicians can search for driver mutations in individual patients to drive their individual therapeutic strategies.