Abstract 4949: 18q copy number variation in primary prostate tumors as a predictor of long-term outcome

Abstract

Abstract The goal of this study is to use primary prostate tumors with long-term outcome to identify regions of copy number variation (CNV) on 18q along with other known loci to determine which regions correlate with poor prognosis. Most prostate cancers (PCa) grow slowly and remain indolent, yet some become aggressive and metastasize. Clinical decision-making requires prognostic markers that can be utilized at the time of diagnosis to determine which tumors will become aggressive. Previous studies have shown a correlation between alterations on the long arm of chromosome 18 (18q) and metastatic PCa. Prostatectomy specimens were collected and banked in the University of Texas Health Science Center San Antonio Genitourinary tissue biorepository. Subjects were categorized as No Evidence of Disease recurrence (NED) or those who developed Metastases (MET) with a minimum of five years follow-up. Tumor foci were marked after H&E staining. Using Agilent's SureDesign software, a custom comparative genomic hybridization (CGH) array was created containing high-density tiling of 18q, as well as 15 genes located elsewhere in the genome that have been implicated in PCa prognosis. Tumor DNA was isolated and analyzed for CNV using standard CGH methodology and the Agilent Cytogenomics software. Chromosome-wide (18q) and genome-wide statistical significance of the relative CNV anomaly prevalence (NED vs. MET) was determined with permutation testing. Thirty-five primary prostate tumors were analyzed; 17 were classified as NED and 18 as MET. Two significant regions of copy number gains were found on 18q. One gain at 18q21.31 was found only in MET samples while another gain at 18q11.2 was found in NED samples. Additionally, confirmatory gains and losses in the regions containing 15 PCa related genes were found, only three of which were statistically different between the two sets. Our arrayCGH results implicate DNA gain of copy at 18q21.31 as being predictive of poor outcome. MicroRNA-122, which lies within this region, has been shown to have increased expression in prostate tumors with higher Gleason score and has been implicated in breast cancer metastasis. Our data suggests it may be an important prognostic marker for aggressive PCa. The 18q11.2 region found to be amplified in NED samples warrants further study to identify genes in this region that may predict good outcome. Of the 15 gene regions included, MYC and Cyclin D1 regions both exhibit gain of copy number in primary tumors with poor outcome, in line with other reports that these gains that these are early events in PCa progression. Citation Format: Keith A. Ashcraft, Teresa L. Johnson-Pais, Jonathan A.L Gelfond, Javier Hernandez, Ian M. Thompson, Robin J. Leach. 18q copy number variation in primary prostate tumors as a predictor of long-term outcome. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4949.