Prostate cancers with distinct transcriptional programs in Black and White men

Abstract

Abstract Background Black men are at a higher risk of prostate cancer (PC) diagnosis and present with more high-grade PC than White men in an equal access setting. This study aimed to identify differential transcriptional regulation between Black and White men with PC. Methods We performed microarray of radical prostatectomy tissue blocks from 305 Black and 238 White men treated at the Durham Veterans Affairs Medical Center. Differential expression, gene set enrichment analysis, master regulator analysis, and network modeling were conducted to compare gene expression by race. Findings were validated using external datasets that are available in the Gene Expression Omnibus (GEO) database. The first was a multi-institutional cohort of 1152 prostate cancer patients (596 Black, 556 White) with microarray data (GEO ID: GSE169038). The second was an Emory cohort of 106 patients (22 Black, 48 White, 36 men of unknown race) with RNA-seq data (GEO ID: GSE54460). Additionally, we analyzed androgen receptor (AR) chromatin binding profiles using paired AR ChIP-Seq datasets from Black and White men (GEO IDs: GSE18440 and GSE18441). Results We identified 871 differentially expressed genes between Black and White men. White men had higher activity of MYC-related pathways, while Black men showed increased activity of inflammation, steroid hormone responses, and cancer progression-related pathways. We further identified the top 10 transcription factors (TFs) in Black patients, which formed a transcriptional regulatory network centered on the AR. The activities of this network and the pathways were significantly different in Black vs. White men across multiple cohorts and PC molecular subtypes. Conclusions These findings suggest PC in Black and White men have distinct tumor transcriptional profiles. Furthermore, a highly interactive TF network centered on AR drives differential gene expression in Black men. Additional study is needed to understand the degree to which these differences in transcriptional regulatory elements contribute to PC health disparities.