Next generation sequencing of primary prostate cancer tumors to reveal potentially actionable opportunities for clinical management: The UNC experience

305 Background: The Strata trial (NCT03061305) is a multi-institutional precision oncology collaboration structured as an observational protocol that aims to match patients to genomically-guided therapies. Methods: Selected University of North Carolina (UNC) metastatic prostate cancer (mPC) patients were enrolled on this IRB-approved study. Formalin fixed paraffin-embedded primary tumor specimens, without matched germline controls, were sent for targeted next generation sequencing (NGS) to detect actionable variants, including: mutations in 87 genes, copy number variations in 31 genes, and gene fusions in 46 gene drivers. mPC-related genes of particular interest included: AR, ATM, BRCA1/2, ERG, MSH2, MSH6, PTEN, RB1, and TP53. Results: Of the 92 cases sequenced, 5 [5%] failed testing. Of the 87 mPC patients (median age 69 years [47-86]) enrolled: 53 [61%] were white, 28 [32%] were black, 1 [1%] was Asian, and 5 [6%] declined to be identified. NGS data revealed 106 variants in 27 genes: 62 patients (71%) had at least one variant, 21 (24%) had 2 variants, 7 (8%) had 3 variants, and 4 (3%) had 4 variants. Among the 62 patients with at least 1 identified variant, TMPRSS2-ERG fusion occurred most frequently (50%), followed by TP53 (40%), and PTEN (16%). 6% of all sequenced patients had variants in DNA damage repair genes including ATM (3%), BRCA2 (2%) and MSH2 (1%). One patient had a SLC45A3-ERG fusion combined with PTEN deep deletion, which has been associated with a more aggressive phenotype. One patient with a microsatellite-instability high tumor was treated with pembrolizumab. Conclusions: The UNC experience shows that a high proportion of primary prostate cancer tumors from mPC patients have genomic variants, and one patient was treated based on these data. Limited actionability may reflect the landscape of currently FDA approved mPC treatments, and available clinical trials. It may also be due to a short follow-up, and these data could inform treatment planning upon progression.