Germline Determinants of Prostate Cancer Evolution

点击翻译按钮获取中文摘要

Abstract Prostate cancer is the most commonly diagnosed non-skin cancer in men, and the second most common cause of cancer death for men. It is a high-incidence, high-fatality, high-morbidity and high-cost cancer that is growing increasingly frequent in our aging population. Clinically, the majority of prostate cancers present as indolent, and are unlikely to significantly influence a man’s health over his lifetime. Nevertheless, a significant number of prostate tumors are diagnosed while localized but have substantial metastatic potential. Once these colonize distant sites, they become almost uniformly lethal. Current standard of care for clinical risk-stratification involves serum abundance of prostate specific antigen (PSA), a digital rectal exam and biopsy. While these are beneficial, ~30-40% of men are over- or under-treated. Therefore, a central problem in prostate cancer remains understanding this dramatic variability in the aggressiveness of localized prostate cancers. We hypothesize that the clinical and molecular evolution of localized prostate cancers are shaped by germline genomic features. Strong preliminary data support this idea. Prostate cancer is the most heritable solid cancer. It shows strong variability across ancestries, and specific genetic features predict both risk of incidence and disease aggression. We will test our hypothesis with three complementary aims. Aim #1 quantifies germline-somatic interactions using transcriptome-wide association studies (TWAS) and pathway-somatic interaction analysis. Aim #2 then identifies ancestry-based germline-somatic interactions. Finally Aim #3 will assess the generalizability of germline-somatic interactions for clinical risk prediction by developing a targeted DNA- sequencing panel and evaluating biomarker potential in large patient cohorts. To achieve these aims, we will leverage publicly available datasets as well as biobanking repositories at UCLA, linked to both established and novel bioinformatics methods. Together, these aims provide three complementary strategies to quantify how the specific clinical and molecular evolutionary features of localized prostate cancer are influenced by specific germline polymorphisms.