Abstract Background: Recent advances in our understanding of the mutational landscape of prostate cancer (PCa) have occurred through sequencing of primary prostate tumors. Highly aggressive PCa most often does not come to surgery preventing access to frozen sample for high-throughput studies. The purpose of this study is to determine the spectrum of mutations in a subgroup of highly aggressive PCa and to evaluate for targetable molecular alterations using a novel mutation analysis platform from archival formalin-fixed paraffin embedded (FFPE) samples. Methods: 21 PCa (of which 9 were Gleason 5, 8 metastatic), 18 neuroendocrine PCa (NEPC), and 18 benign FFPE prostate samples from 35 unique cases were evaluated and 77 high-density foci were selected for DNA extraction and sequencing library construction. Hybridization-based capture of 3230 exons and 37 intronic intervals of 189 cancer-related genes was performed, followed by deep, massively-parallel sequencing. Sequence data were analyzed for single nucleotide variants (sub), small insertions and deletions (indels), significant copy number changes, and selected re-arrangements/fusion genes. Results: 5/8 biopsies and 62/65 (overall 93%) of prostate or metastatic foci yielded sufficient DNA (≥50 ng) for analysis from 40μ of FFPE tissue. Average uniquely-mapping sequence coverage was 885×. 133 mutations were identified in 30 genes with 58/67 samples containing at least one mutation. TMPRSS2-ERG fusion was present in 31% of cases. Recurrent high confidence cancer alterations included: gain of MYC (9%), gain or sub/indel of AR (43%), deletion or sub/indel of PTEN (23%), RB1 (20%), BRCA2 (11%), and sub/indel of TP53 (40%), CTNNB1 (11%), PIK3CA (6%), ATM (6%). Several novel mutations were identified, including in BRAF and BRCA2. AR alterations were more common in PCa compared to NEPC, and were present even in clinically localized PCa. There was high concordance between NEPC and PCa foci in mixed tumors, as well as between primary tumor and metastases. Conclusions: This study demonstrates feasibility of in-depth DNA analysis using FFPE tissue, and even biopsy material. The spectrum of driver mutations observed highlights potential value of comprehensive genomic profiling for targeted therapy selection in clinical care. These analyses have laid the foundation for ongoing comparison with indolent PCa, and have identified potential new drug targets (e.g., PARP inhibitors for ATM and BRCA2 mutations), with a goal towards better defining biomarkers that distinguish aggressive and indolent PCa and designing more effective treatment approaches for patients with lethal PCa. Citation Format: Himisha Beltran, Philip J. Stephens, Juan Miguel Mosquera, Maureen T. Cronin, Mark A. Rubin, Roman Yelensky, Garrett Frampton, Kyung Park, Sean Downing, Theresa Y. MacDonald, Doron Lipson, Scott T. Tagawa, David M. Nanus. Massively parallel DNA-sequencing of aggressive prostate cancer reveals disease heterogeneity and identifies targetable mutations [abstract]. In: Proceedings of the AACR Special Conference on Advances in Prostate Cancer Research; 2012 Feb 6-9; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2012;72(4 Suppl):Abstract nr C22.
Read full abstract