Abstract PO-146: Copy number landscape of primary African-American and European-American prostate tumors

Abstract

Abstract African-American (AA) men are more likely to be diagnosed with and to die of prostate cancer than their European-American counterparts. Much of this disparity is due to access to care and social determinants of health; however, there is emerging evidence of molecular differences in prostate tumors arising in patients of different genetic ancestry. We investigated the copy number landscape of 290 grade and race-matched primary prostate tumors through interrogation of Infinium MethylationEPIC Array (Illumina) data derived from primary prostate tumors at radical prostatectomy. Our cohort was composed of 145 self-identified AA patients and 145 self-identified EA surgically-treated patients with accompanying genetic ancestry estimation via SNP array (GSAv3). EPIC array-derived copy number data estimated through the Conumee package on R was significantly correlated with genomic losses and gains identified by exome sequencing of a panel with 100 cancer driver genes. We also observed a significant association between genomic losses of the PTEN gene and PTEN protein loss by immunohistochemistry (P<0.001). Similar observations were made for p53 protein nuclear accumulation detected by immunohistochemistry and TP53 gene copy loss (P<0.001). Next, we derived percent genome altered (PGA) from EPIC array data for our 290 patients. PGA from EPIC array was significantly correlated with PGA obtained from targeted sequencing of a subset of AA tumors (n=119, P<0.0001, R²=0.50). No difference in PGA was found between self-identified AA and EA patients (P<0.05), though PGA levels were significantly associated with Gleason Grade groups for both AA (P<0.0001) and EA (P=0.003). We then used the genetic ancestry estimation derived from SNP arrays to identify regions of genomic loss and gain differing by percent African ancestry. We used a comprehensive genome annotation with >80k coding and non-coding regions from FANTOM CAT. Adjusted LIMMA models with age, Gleason Grade Group, and pre-operative PSA levels showed significant copy number differences by percent African ancestry in chromosomes 6p, 10q, 11p, 12p, and 17p. PTEN losses (as expected) and WT1 gains were more frequent in patients with lower percent African ancestry. Gene expression data from TCGA data showed that WT1 expression was significantly increased for self-identified EA patients. Multivariable Cox regression models adjusted for age, Gleason Grade Group, and pre-operative PSA levels revealed that chromosome 8q gains (including MYC, GRHL2, and FZD6) were significantly associated with biochemical recurrence and metastasis. However, when we stratified the models by self-identified race, only AA tumors showed significant associations with 8q gains and poor outcome. Further in silico and mechanistic validation will be conducted to confirm whether chromosome 8 gains may be a potential biomarker for prostate cancer outcome in AA men. Citation Format: Thiago Vidotto, Eddie Imada, Farzana Faizal, Siqun Zheng, Jianfeng Xu, Karen S. Sfanos, Luigi Marchionni, Tamara L. Lotan. Copy number landscape of primary African-American and European-American prostate tumors [abstract]. In: Proceedings of the AACR Virtual Conference: 14th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2021 Oct 6-8. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2022;31(1 Suppl):Abstract nr PO-146.