Association of ARV7 expression with molecular and clinical characteristics in prostate cancer.

Abstract

109 Background: Prostate cancer (PC) is the most common cancer in men over the age of 60 and the second leading cause of cancer mortality in the United States. Clinicopathological parameters such as Gleason score, tumor volume, surgical margins, prostate-specific antigen (PSA), Ki-67 index and clinical stage are used as prognostic markers for clinical outcomes. Identification of novel molecular markers could improve our understanding of the clinical behavior of this disease. Androgen receptor isoforms, in particular variant 7 (ARV7 or AR3) have been recently studied for elucidating their potential role in PC progression, associated with epithelial-mesenchymal transition (EMT), disease aggressiveness, increased proliferation and therapeutic resistance. Our study is analyzing the association of ARV7 mRNA expression to clinical characteristics and is analyzing the genomic data to identify differentially altered genes by ARV expression status, summarized as a potential functional network. Methods: We obtained the TCGA public dataset of prostate adenocarcinoma tumors (N=499) that included the clinical data, gene and isoform expression and mutation data. Cases were categorized into ARV7 over-expressing (ARV+) and normal or low expression (ARV –/N) by using a cut-off of upper 25th percentile of the background genomic expression. Analysis was performed in R and Perl by using custom-made scripts. Differentially altered genes and pathways were identified and were summarized as potential functional networks. Results: We categorized 30 out of the 499 tumors as ARV+. ARV7 over-expression was found to be significantly associated with older age at diagnosis (>70), advanced clinical stage, nodal involvement, high Gleason score and a poor therapeutic response. We also observed a trend towards shorter disease-free survival among ARV+ tumors. In addition, ARV+ tumors showed significantly higher number of mutations in 20 key regulatory pathways including Jak-STAT signaling, homologous recombination, ErbB and Wnt signaling pathways. Conclusions: ARV7 overexpression is associated with genomic alterations in key regulatory pathways and poorer clinical outcome in PC patients.