Addressing racial disparities in colorectal cancer risk prediction

PROJECT SUMMARY With Black men and women experiencing a disproportionately high burden of colorectal cancer (CRC), there is a critical need for new improved risk prediction tools to identify those at highest risk to undergo screening at earlier ages or more frequently as needed. Black adults have the highest incidence and mortality rates of any

racial and ethnic group in the US for reasons only partly understood and their risk factors have not been well- studied to date. Further, Black Americans experience persistently high rates of early-onset CRC diagnosed in younger adults aged less than 50-years. Early-onset CRC is an emerging public health concern due to rising

incidence rates, leading to a recently reduced routine screening age of 45-years in the US. Established risk prediction tools such as the NCI's Colorectal Cancer Risk Assessment Tool were developed and validated on white populations and have been shown to perform poorly in Black women and have not yet been tested in

Black men. In this grant our overarching goal is to develop new personalized CRC risk tools with improved accuracy that are uniquely tailored for Black Americans, and are built and validated on the data of Black men and women. Our new web-based risk models will be developed using the rich longitudinal data of large

prospective established US cohorts of Black adults in the Southern Community Cohort Study and Multiethnic Cohort Study with external validation in three further cohorts of Black adults. Our three aims are: (a) to develop and externally validate new risk prediction models for colorectal cancer in Black men and women based upon

easily obtainable lifestyle, demographic, social and screening factors; (b) to build dynamic risk models for CRC for Black adults using modern statistical and machine-learning methods to obtain optimal prediction performance, and (c) to determine if accounting for the site of the colorectal cancer tumor improves the

individualized prediction accuracy of our new risk models in Black adults. Due to the high CRC incidence and death rates in Black adults, our new risk prediction models will have significant and far-reaching implications for colorectal cancer screening and mortality in the Black community. Our new clinical risk tools will provide

accurate personalized risk assessments of CRC which can be used by Black adults and their physicians to identify those at high risk and to target optimal, timely and effective CRC screening and lifestyle prevention strategies, leading to earlier detection of CRC, removal of precancerous polyps during colonoscopy and

reduced mortality. Our proposed work closely aligns with current key initiatives of NCI to support research on improving health equity in cancer.