Genetic Predictors of Prostate Cancer Survival

Prostate cancer (PC) has a very heterogenous clinical course. Many men have an indolent PC that can be safely watched for years without progression. Alternatively, other men have an aggressive form of PC and can progress very rapidly. Understanding PC risk stratification and separating indolent from aggressive disease is a

crucial unmet clinical need. Increasing data suggest that genetic single nucleotide polymorphisms (SNPs) may aid in this effort. Indeed, this is the fundamental basis of the parent R01 (Klein PI; Freedland site PI). In line with that overarching view, this supplement supports Dr. Asilonu's development to focus on two areas

of risk stratification: 1) Using genetic SNP data; 2) using epidemiological data (obesity and diabetes status). In Aim 1, we aim to build upon our genetic analyses proposed in the parent grant. Specifically, we will use Mendelian Randomization (MR) to determine causal relationships between type 2 diabetes and obesity with

prostate cancer survival. Building upon recent work classifying type 2 diabetes into different subtypes, we will ask use genetic predictors of these diabetes subtypes, as well as predictors of obesity, as instrumental variables in the Mendelian Randomization analysis. By comparing the effect size of SNPs on these variabes with the

effect size of these same SNPs on prostate cancer survival (derived from data from the parent grant), we will determine the extent to which these subtypes of diabetes and obesity directly lead to worse prostate cancer survival. In Aim 2, we will build upon the novel finding by Dr. Freedland and his team that diabetes and obesity appear

to act synergistically to create a more aggressive PC. This was not seen for overall PC risk, but specifically for aggressive PC. This synergistic interaction was noted for both diagnosis of high-grade PC and PC mortality after surgery for early-stage disease. Based upon these findings, we hypothesize that diabetes and obesity will

interact to synergistically increase the risk of aggressive PC (high-grade PC and PC mortality) but not low-grade PC. We further hypothesize these associations will be independent of screening patterns and access to care suggesting a biological basis. We will test this hypothesis using nationwide data from the Veterans Affairs (VA)

Health System. We received a separate grant to create a nationwide database to study the potential link between obesity and race in predicting aggressive PC. The vast majority of work for this other grant is now complete. This creates a great opportunity for Dr. Asilonu to build upon his current statistical knowledge and complete the

complicated analyses proposed in this supplement. From the research proposed, Dr. Asilonu will develop new skills in advanced statistical analyses, genetic analyses, working with nationwide data, and manuscript preparation. We also built-in didactic work to help him prepare grants. At the end of the nearly 3-year supplement, not only will we have answered key questions in the

field, but he will be ready to take the next step in his career – applying for a K-award or R-series grant.