INTRODUCTION AND OBJECTIVE:
African American men (AAM) with prostate cancer (PCa) often harbor more aggressive disease compared to European American men (EAM). However, the underlying cause for disparities in PCa outcomes is controversial. We performed gene expression analysis to characterize differences in tumor biology in AAM and EAM with PCa and assess the impact on outcomes.
Microarray analyses using 1,507 probes were performed to assess the expression of 517 genes in PCa radical prostatectomy (RP) specimens from 270 AAM and 369 EAM from 1991-1996. The primary endpoint was biochemical recurrence (BCR) and secondary endpoint was PCa-specific mortality (PCSM). Cox proportional hazard regression analyses were performed adjusting for age, pre-treatment PSA, Gleason grade group (GG), and T-stage. Differential Gene Expression (DEG) analysis was performed using the limma R package, with correction for multiple comparisons using the Benjamini-Hochberg procedure (false-discovery rate (FDR) of <0.05 considered significant). Gene Set Enrichment Analysis (GSEA) was performed to identify enriched hallmark pathways using R package fgsea.
The median age of AAM and EAM was 64 (IQR 56-72) and 62 (IQR 54-70) years, respectively. A total of 66 (24%) AAM and 67 (18%) EAM developed BCR (p=0.002; median follow up 10.5 and 13.8 years, respectively). Sixteen men experienced PCSM in each group (5.9% and 4.3% in AAM and EAM, respectively; p=0.080). Interestingly, among those with GG1-2 disease, AAM were more likely to develop BCR than EAM (HR 2.3, 95% CI 1.3–4, p=0.004). However, the BCR risk was not significantly different between AAM and EAM with GG3-5 disease (p=0.370). DEG and GSEA revealed upregulation of inflammation, TNF-alpha signaling, apoptosis, androgen response, and epithelial to mesenchymal transition, among other pathways, in AAM compared to EAM (Figure 1; FDR <0.05).
AAM with GG1-2 PCa had a higher incidence of BCR compared with EAM following RP, suggesting a more aggressive biologic behavior in AAM in early disease. The molecular profile of PCa in AAM is different from that in EAM, with inflammation playing a prominent role. Prognostic gene signatures accounting for these biological differences are needed to facilitate better disease management.
Source of Funding: