Identifying factors associated with prostate cancer progression and survival in African American men: The RESPOND Cohort

Aim 1. The impact of access to health care on PCa outcomes (disease P/R and survival) and patient-reported outcomes in AA men. We will measure and evaluate the impact of the five components of access to care on PCa outcomes (P/R and survival) and quality of life (QoL). Measures of affordability include out of pocket costs, ability to pay, insurance coverage, and financial burden related to care. Availability measures include specialty of physicians providing care and hospital quality metrics. Accessibility measures include geospatial assessments of location of healthcare services in relation to patients residence, supply and demand, mode of transportation and distance to hospitals while accommodation measures include wait times and access to support services. Finally, acceptability involves patients attitudes toward providers, trust in the medical care system and racial concordance between patient and provider. We will also assess the impact of the COVID-19 pandemic on access to care in this population.

Aim 2. The effect of residence in persistently disinvested neighborhoods and modification by chronic exposure to life-course structural and social stressors on PCa aggressiveness, disease P/R and survival. There is increasing recognition that people living in persistently disinvested neighborhoods, as characterized by trajectories in poverty, structural racism, segregation, social, built and environmental hazards, may face even greater stressors that may negatively impact health. We will assess the impacts of residence in persistently disinvested neighborhoods on PCa aggressiveness, P/R, and survival and the extent to which these associations are modified by chronic exposure to life-course structural and social stressors, as well as stress appraisal and resiliency. To promote data dissemination, especially to the RESPOND communities, we will develop a web-based dashboard for visualizing and mapping the structural and social neighborhood data.

Aim 3. The role of modifiable lifestyle factors, co-morbidities and genetics on PCa P/R and survival in AA men. A number of lifestyle and behavioral risk factors for PCa, which are more common in AA men, have been linked with PCa P/R and mortality in White men. We will examine the independent and combined ( score ) associations of lifestyle and behavioral risk factors (e.g., smoking, physical activity and body mass index) and PCa outcomes in RESPOND. In addition, we will investigate whether co-morbid chronic conditions, common among AA men (e.g., type 2 diabetes, hypertension) impact PCa P/R and survival. We will also assess whether germline factors, including the polygenic risk score (PRS) and rare pathogenic variants (PVs), modify or offset risk of P/R and survival associated with lifestyle factors and common co-morbidities.

Aim 4. The application of deep learning (DL) approaches on histopathologic images for PCa prognostication in AA men. While the Gleason grading system is the best prognostic indicator for PCa progression and survival, it is subjective and not all patients have equal access to specialized pathologists who are highly experienced with grading. We propose to utilize DL algorithms to grade, molecularly phenotype and predict disease P/R and PCa survival in AA men, leveraging 3,400 digitalized whole slide images (WSI) of tumor specimens in RESPOND, as well as >2000 additional WSI from Johns Hopkins cohorts (predominantly WH). We will test existing DL Gleason grading algorithms, as well as train and validate DL algorithms to predict high-risk and molecular and etiologic subtypes of PCa based on tumor genomic alterations and associations with germline PRS and PVs. Through this work, we will develop tools to better estimate an AA patient s risk of P/R that could be applied widely across health care facilities to equalize care and reduce health disparities.

For further information about clinical trials, please contact us at 732-235-7356.