Abstract 3393: Genomic analysis of multi-site fresh prostate samples

Abstract

Abstract Prostate cancer is the most common cancer in males in the UK with >40,000 cases diagnosed every year and >10,000 deaths. Recent multi-platform genomic studies have revealed a very complex picture of the disease, with no clear subcategories linking to the histopathological markers currently used in clinical practice. The multifocal and heterogeneous nature of the disease suggests that single biopsy sites may be missing valuable subclones which contribute to the etiology of the disease. To address this, we obtained fresh core biopsies from multiple sites (4-8) in the prostate from eight high-risk patients undergoing prostatectomies at the Christie NHS Foundation Trust. The tissues were immediately processed by the pathology department and each core divided in two, one for genomic analysis and one for parallel disease modeling (patient derived xenografts (PDXs), patient-derived cell lines and organoids). The cores were cryo-sectioned and H&E analysis performed at the top, middle and bottom of each core. These were reviewed by a pathologist and tissue was micro dissected prior to simultaneous DNA and RNA extraction of normal and tumor tissue. Blood for germline DNA and plasma for circulating free DNA (cfDNA) was also obtained. Our study comprises eight patients with two or more tumor sites (2-4), some bifocal, which were analyzed using whole exome sequencing (WES), copy number aberration (CNA) profiling, transcriptomic analysis and methylation profiling. The most frequent aberration identified was loss of 8p (NKX3-1) in 6/8 patients (12/24 tumor cores). Loss of 13q (RB1) was observed in 5/8 patients (8/24 tumor cores), but never in all cores from a single patient, suggesting a late event. Loss of 5q (CHD1) was identified in 4/8 patients (11/24 tumor cores). Loss of 6q (MAP3K7), 10q (PTEN), amplification of 8q (MYC) and TMPRSS2-ERG fusion were identified in 3/8 patients. Some copy number events were only observed in either one patient or one tumor core demonstrating extensive inter and intra-patient heterogeneity. SPOP was the only gene affected by recurrent mutations across patients, but different amino acids were affected within and between patients. Mutations in PTEN, TP53, APC, BRAF and ERCC3, were also identified, among many others. The most significantly overexpressed gene was ERG, seen in patients with the TMPRSS2-ERG fusion. Changes in gene expression differed between cores from the same patients, reflecting the heterogeneity at the DNA level. These data will be analyzed in conjunction with the results from disease modeling to investigate the functional impact of these changes and cfDNA analysis is underway to understand which tumor clones are entering the bloodstream. The multifocal and genomically heterogeneous nature of prostate cancer is highlighted by this data and is likely to impact on precision medicine approaches for this disease. We will correlate the molecular profiling of our patient tumors with their clinical data in order to identify targets for further validation. Citation Format: Marina A. Parry, Shambhavi Srivastava, Alessio Cannistraci, Hui Sun Leong, Syed Ali, Jenny Antonello, Vijay Ramani, Maurice Lau, Jonathan Shanks, Daisuke Nonaka, Pedro Oliveira, Noel Clarke, Crispin Miller, Ged Brady, Nathalie Dhomen, Esther Baena, Richard Marais. Genomic analysis of multi-site fresh prostate samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3393. doi:10.1158/1538-7445.AM2017-3393