Project 1

Project Summary Understanding the spectrum of germline mutations underlying hereditary cancers has clear clinical implications, and while mutations in at least four dozen genes have been associated with heritable cancers, the prevalence, characteristics, and clinical implications of most of these genes have not been well described in African

Americans (AAs). Due to the lack of racial diversity in genetic databases, we and others have shown that AAs have been shown to harbor a disproportionate number of variants of uncertain significance (VUS) compared to whites. A VUS result has the potential to not only increase psychological burden but also exclude African

Americans from gene-specific cancer treatments and prevention interventions, further exacerbating existing racial disparities in cancer incidence and survival. VUS reclassification is a time- and labor-intensive process. The international efforts for VUS reclassification include aggregation of epidemiological and family data, as well

as evaluation of functional data applying molecular biology techniques to in vitro and in vivo systems. Unfortunately, the variants most likely to be examined are those seen most frequently in populations of European ancestry and there is limited throughput for evaluating the thousands of VUS that have been identified to date.

The overall goal of the study is to develop and implement an integrative approach for re-classifying VUS and identifying novel pathogenic variants through targeted germline sequencing, familial segregation analyses, and somatic mutation profiling of tumors. The specific aims of the project are to (1) quantify the prevalence of

pathogenic variants and VUS in AAs with breast, prostate, colorectal, endometrial, and other young onset cancers in metropolitan Detroit, (2) characterize segregation patterns of VUS in these genes among family members and evaluate in silico and RNA-based predictions of pathogenicity of these VUS, and (3) develop a

tumor expression- and somatic mutation-based classifier to predict breast and colorectal tumors with PVs versus non-PVs in a DNA repair gene (homologous recombination or mismatch repair) to be used for VUS reclassification. This approach represents a powerful tool for characterizing the clinical implications of VUS in

cancer susceptibility genes using population science approaches. We will leverage data and biospecimens from the Detroit Research on Cancer Survivors (Detroit ROCS) cohort, a population-based study of AAs with cancer within the Metropolitan Detroit Cancer Surveillance System (MDCSS) registry catchment area. This study is

impactful as it will as it will result in a robust approach for identifying clinically actionable variants among African American families affected with cancer.