Abstract A018: Geographic disparities in aggressive prostate cancer varies by region-specific environmental and sociodemographic characteristics: A statewide analysis

Abstract Background: Although the overall 5-year survival rate for prostate cancer (PCa) is relatively high, there are substantial differences in survival by stage at diagnosis. In contrast to men diagnosed with localized or regional stage PCa, both of which have a 5-year survival rate of nearly 100%, only 31% of patients survive five years after being diagnosed with distant or metastatic disease. Given PCa remains the second leading cause of cancer death among U.S. men, targeted PCa screening interventions centered on early detection for men at risk for more aggressive PCa subtypes appear warranted. Under a Precision Public Health approach, the application of geospatial analysis to cancer surveillance data can help to identify and “narrow down” geographic areas where rates of aggressive PCa are higher than expected to target for intervention. Thus, the primary objective of this study was to conduct a geospatial scan to identify clusters of aggressive PCa in a State-wide analysis. Given well-known racial disparities and associations between socioeconomic status and PCa stage at diagnosis and survival, the secondary objective was to evaluate geographic variation, or changes in aggressive PCa clusters, by area-level sociodemographics and environmental exposures. Methods: The study population included 82,580 patients diagnosed with incident PCa from the Pennsylvania (PA) State Registry from 2005-2015. Patient addresses were geocoded. Patient clinical data from the registry (age, race/ethnicity, insurance status) were linked via geocodes to area-level sociodemographic measures generated from the US Census Bureau at the census tract level, and to Environmental Quality Indices (Air, water, soil) from the Environmental Protection Agency (EPA) at the county level. Aggressive PCa was evaluated using 3 separate definitions based on TNM staging and/or Gleason grade. A Bayesian geoadditive modeling approach identified census tract clusters with higher than expected rates of aggressive PCa, adjusted for individual-level factors (age, race, diagnosis year) (Baseline model). The baseline model was then compared to models that included sociodemographic and/or environmental measures by evaluating changes in model fit (DIC), cluster size, and relative risk (RR). Results: We identified 3 main clusters in PA where rates of aggressive PCa were higher than expected near Philadelphia, Altoona, and Pittsburgh. Clusters of aggressive PCa in Philadelphia were eliminated and could be fully explained after adjustment for sociodemographic characteristics (insurance, poverty, etc), but not environmental measures. In contract, clusters of aggressive PCa in Pittsburgh were only reduced after adjustment for environmental variables, particularly water quality. Conclusion: PCa screening interventions could be prioritized in high risk cluster areas. Results are novel in that they preliminarily suggest that the type of intervention and focus of future etiologic studies in aggressive PCa might vary by geographic location, even within a single State. Citation Format: Daniel Wiese, Kevin A. Henry, Carolyn Fang, Adam Reese, Mary Daly, Camille Ragin, Shannon M. Lynch. Geographic disparities in aggressive prostate cancer varies by region-specific environmental and sociodemographic characteristics: A statewide analysis [abstract]. In: Proceedings of the 15th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2022 Sep 16-19; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr A018.