Prostate cancer has the highest estimate of heritability of any cancer, with 58% of variability in prostate cancer incidence attributed to inherited genetic factors. Genome wide association studies have validated 269 single nucleotide polymorphisms that are strongly associated with prostate cancer risk. We found that a multiethnic polygenic risk score (PRS) combining these SNPs demonstrate a 9-fold difference in risk of disease comparing men with high vs. low PRS in a both Black and White men. This proposal aims to translate this prostate cancer PRS into clinical practice by addressing four important questions: 1) Can the PRS be integrated with other tools including MRI and rare genetic variants in DNA damage repair (DDR) pathways as part of an early detection strategy to identify clinically-relevant, potentially lethal prostate cancer? 2) At what point in a man's life should an early detection program begin if he is at increased genetic risk? 3) What is the optimal interval of imaging to detect clinically relevant cancer in men at high genetic risk? This collaborative U01 proposal addresses these issues in three specific aims. Aim 1 - we will prospectively determine the ability of a prostate cancer PRS integrated with MRI to identify higher-grade, potentially lethal prostate cancer. We will recruit 1500 men (600 Black, 900 White) from the MGB Biobank, the Walter Reed Biobank, and the primary care network at Howard University and Brigham & Women's Hospital. All men will be stratified into low, average, and high risk on the basis genotyping. PSA, MRI, and DDR variants will be obtained followed by biopsy for elevated PSA or abnormal MRI. We expect to find the PRS identifies a population at risk for prostate cancer while the DDR variants and MRI identifies a subset with clinically relevant disease. In Aim 2, we will evaluate at what point in a man's life an MRI is clinically useful. Our population will be imaged across 5 year age groups from 40-69 years. In addition, men at the high genetic risk without cancer will undergo serial MRI imaging at the NCI at 2 year intervals. In Aim 3 we will determine if deep learning methods applied to mpMRI and informed by genetic risk can more accurately predict significant cancers. This will be the first in field prospective trial to integrate germline genetics with MRI to identify men at risk of clinically-relevant prostate cancer. The results will have short-term impact by establishing an optimal early detection algorithm and show the utility of incorporating information on the germline into an early detection strategy. It will establish the role of MRI in detecting clinically relevant cancers among those with high genetic risk. The longer-term goal will be to use the knowledge gained to design trials of the at-risk populations with longer follow-up to prove that genetic testing can improve our ability to prevent prostate cancer mortality through targeted screening and prophylaxis. Importantly, men at low risk for clinically significant disease could be spared screening, prophylaxis and treatment. This information can be directly translated into patient populations. An additional strength of this proposal is the inclusion in racially diverse patient populations.

PROJECT NARRATIVE The premise of this proposal is that an optimal early detection strategy to identify clinically relevant prostate cancer will involve a two-tiered algorithm that leverages inherited genetic information to determine who is at risk for prostate cancer followed by MRI imaging to determine which of these high risk patients has clinically relevant disease. We propose to test this premise by implementing a polygenic risk score in men and conducting a prospective trial among 1,500 men with MRI to prove the two-tiered algorithm works.