Leveraging Diversity in Cancer Epidemiology Cohorts and Novel Methods to Improve Polygenic Risk Scores

点击翻译按钮获取中文摘要

Abstract There are stark differences in the burden of certain cancers across racial/ethnic populations. For example, in comparison to individuals of European ancestry, African American men have a ~67% higher incidence rate of prostate cancer and Asian/Pacific Islander men and women have a 70% and 95% higher incidence rate of liver cancer, respectively. These disparities in the burden of cancer across racial/ethnic groups have been attributed to an interplay of genetic, environmental, and social factors. Despite such disparities, a majority of genetic research has focused on individuals of European ancestry. While genome-wide association studies (GWAS) have successfully identified >1000 risk loci for cancer, they have focused primarily on individuals of European ancestry. The inadequate representation of diverse racial/ethnic populations limits the translational potential of GWAS findings to the world's populations. Applying PRS developed in European ancestry individuals to other populations may result in biased risk prediction, and further exacerbate health disparities due to inaccurate assessment of individuals at high risk of disease. Here, we propose to address the drastic need for appropriate PRS construction and evaluation across multiple race/ethnic groups by applying new PRS approaches to the following six large-scale, longstanding cohorts: the Multiethnic Cohort (MEC); the Kaiser Resource for Genetic Epidemiology Research on Aging (GERA) cohort; the Women's Health Initiative (WHI); the Harvard Nurses Health Studies (NHS); the Harvard Health Professionals Follow-Up Study (HPFS); and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO). Together, these cohorts include over 300,000 individuals (100,000 non-Europeans) and 91,000 incident cancer cases (24,000 non-Europeans). The individuals in these cohorts are from five racial/ethnic groups: African Americans, Latinos, Japanese, Native Populations, and European ancestry. While focusing on cancer outcomes, we will utilize these unique and extensive resources to develop methods to construct and evaluate PRS, and importantly for translation, estimate absolute and excess relative risk of cancer jointly for PRS and established risk factors in multiethnic populations. To facilitate access to developed pipelines and data resources, we will follow F.A.I.R. analytic principles while participating with the Coordinating Center and other study sites. Ultimately, constructing and evaluating risk models in non-European ancestry populations is essential to broaden the impact of genomic medicine on human health.