Genomic characterization of primary and metastatic prostate cancer (PC) using a targeted next-generation sequencing assay.

Abstract

254 Background: Genomic alterations in PC have been described across the disease continuum, creating opportunities for selective clinical trial enrollment of patients (pts) with high-risk or metastatic disease based on their tumor profile. MSK-IMPACT is an exon capture-based sequencing assay performed in a CLIA-certified laboratory that targets 410 cancer-associated genes, many of which are potential drug targets. We assessed mutations and copy number alterations (CNAs) in primary and metastatic samples from untreated, hormone-treated and castration resistant pts. Methods: PC pts were enrolled on an IRB-approved protocol for tumor genomic profiling. Fixed tumor and matched germline samples were subjected to DNA sequencing analysis using MSK-IMPACT for the identification of somatic mutations and CNAs. Results: 315 samples from 271 pts were successfully sequenced (Table). Overall success rate was 80% (67% for bone). 14 tumors were pathologically classified as neuroendocrine or had neuroendocrine features. 23% of patients had tumor somatic alterations in DNA repair genes (BRCA2, BRCA1, ATM, FANCA and CDK12). Additional alterations were observed in PI3K, MAPK and Wnt-βCatenin pathway genes. Common alterations in the metastatic tumors include TP53 deletion/mutation (42%), AR amplification/mutation (38%), PTEN deletion/mutation (27%), RB1 deletion/mutation (20%), BRCA2 deletion/mutation (11%), FOXA1 mutation (11%) and SPOP mutation (4%). Among metastatic samples, tumors from pts with castration-resistant disease had higher CNA burden when compared to tumors from non-castration-resistant pts (0.31 vs. 0.19 fraction genome altered, p < 0.05, unpaired t-test), as well as higher rates of AR amplification/mutation (49% vs. 3%, p < 0.01, Fisher’s exact test). Conclusions: Genomic profiling of primary and metastatic prostate tumors is feasible with the clinical MSK-IMPACT sequencing assay and has identified actionable alterations in > 40% of patients. This is allowing for selective trial enrollment and further investigation of PC genomics. [Table: see text]