Abstract 4496: Genomic profiling of late stage tumors powering precision medicine in Japan

Much evidence has shown that genomic profiling of tumors provides patients with treatment options specific to their tumor's biology in the US. Within the Kyoto University Hospital Cancer Center, we offer patients access to these tests through the OncoPrimeTM for those patients of cancer with unknown primary (CUP), rare tumor types or for which standard treatment options have failed. Our institution sees over 6,000 cancer patients each year with over 1,500 receiving chemotherapy and a progressive climate makes this kind of testing favorable. We aim to demonstrate the efficacy of genomic profiling along with successful execution of international specimen logistics as FFPE samples are shipped to EA Genomics and clinical report is provided within 3 weeks. We officially launched the product publicly in March 2015 and the majority of patients have paid for testing out of pocket in lieu of insurance coverage. During this first year, we saw 30% of patients with CUPs, and 16% of both pancreas and biliary tract with the rest being other rare cancers. Patients going through the testing were generally successful (93.5%) with the specimen failures likely indicative with FFPE artifacts or severe tumor heterogeneity. Actionabilty for our population (84.8%) is consistent with previously published numbers for information provided in the clinical report. Additional information was collected after the clinical reports to determine early indicators of follow up and see that the accepted term ‘actionability’ holds different meaning during the practice of patient care. While we had 17% of patients receive the suggested indication we have yet seen positive response from patients who have returned for follow-up. Selected tissues were tested by alternative profiling vendors, and in one case alternate action with Erlotinib was recommended by one vendor; this action has shown promising response. This suggests that heterogeneous interpretation is still a largely unsolved issue with genomic profiling. This may suggested that interpretation is still a largely unsolved issue with genomic profiling. While almost all other patients tested qualified for clinical trials (84.8%) we noted 3 main barriers to entry: geography, cost of intervention and deterioration of patient condition. We show our results in line with previous studies but strive to increase adoption of results and are compiling information of pan cancer biomarker statuses and treatment success/failures to improve interpretatio