Abstract A039: Joint roles of proteomics and neighborhood-level social determinants of metastatic prostate cancer

Abstract Introduction: Striking racial disparities in prostate cancer (PCa) outcomes among American Black men warrant significant efforts to identify early predictors of lethal PCa, to avoid over- treatment of clinically favorable disease with minimal risk of progression. Building on prior work, this study examined independent and joint roles of proteomic markers and social determinants of health (SDOHs) in predicting prostate disease aggressiveness in a racially diverse cohort of men undergoing prostate biopsy. Methodology: A retrospective cohort of 300 men biopsied for prostate cancer was assembled at University Hospitals (UH) Seidman Cancer Center in Cleveland, Ohio between January 1, 2005-May 2022. An extensive medical chart review was performed to identify Black and White men undergoing transrectal ultrasound- guided biopsy that resulted in: (i) a negative biopsy in patients with a history of 1+ prior negative biopsy and no PCa history (“NEG”); (ii) a biopsy-detected PCa of Gleason sum 6 (“GL6”); and (iii) biopsy-detected PCa with nodal and/or distant metastases at initial cancer detection (“METS”). Prostate biopsy tissues were obtained on each patient, balancing 100 men per biopsy group, and balanced on race (1:1) within each biopsy group. Informed by findings from an untargeted analysis of the entire proteome of 60 patients (“Discovery cohort”), we conducted a targeted proteomics analysis of 16 protein markers generated from prostate biopsy tissue, using mass spectrometry, on n=240 patients (“Verification cohort”). Chi-square testing and Analysis of Variance (ANOVA) were used to assess associations between protein expression, neighborhood-level SDOHs, and biopsy group. Race-stratified analyses were performed to examine statistical interactions between study associations. Principal components (PC) analysis was used to identify protein markers that predicted the greatest variance in biopsy group. Multivariable logistic regression (MLR) was used to predict METS (vs. NEG and GL6, combined), as a function of PC-selected protein markers and census block-group level area deprivation index (ADI). Bonferroni correction was used to establish the threshold for the decision rule of statistical significance (alpha error 0.05/16 markers: p 0.0031). Results: Median patient age was 68 years. Protein expression distributions were transformed into quintiles due to non-normality. Of 16 protein markers, 14 were significantly correlated with biopsy group, including PARP 1 and TGFB1, for which METS patients had the highest expression. ADI was strongly correlated with race but not protein expression. ADI was significantly higher in METS patients, followed by GL6 and lowest in NEG group; however no associations were seen for ADI with protein expression. Eight proteins were found to predict METS in PC and MLR analysis. Conclusions: These data support use of proteomics markers to predict a metastatic versus non-metastatic biopsy result. Next steps include deep exploration of proteome findings generated in the Discovery cohort, such as biological pathways. Citation Format: Jennifer Cullen, Tao Liu, Holly Hartman, Fangzhou Liu, Anood Alfahmy, Rini Ghosh, Ayesha Shafi, Julia Payne, Randy Vince, Lee Ponsky, Gregory MacLennan. Joint roles of proteomics and neighborhood-level social determinants of metastatic prostate cancer abstract. In: Proceedings of the 17th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2024 Sep 21-24; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2024;33(9 Suppl):Abstract nr A039.