Practical application of whole genome and transcriptome tumour analysis to guide chemotherapy decision-making for patients with advanced cancers.

e22020 Background: We propose that applying personal genomic information prospectively, in a clinically realistic timeframe can aid chemotherapy decision-making and result in more effective cancer treatment. We are investigating this approach in a variety of cancers to examine timeliness, deliverability, and rate of actionable targets identified. Methods: Eligible subjects with incurable cancer and limited chemo options have a tumour biopsy and “normal” blood taken for analysis. Archival specimens are concurrently analyzed to look for changes with time and treatment. Samples are subject to both an Ampliseq amplicon panel and in-depth whole genome DNA and RNA sequencing (WGS). Bioinformatics approaches identity genes with somatic and copy number variations, and expression changes. Variants are integrated into a pathway analysis to identify tumour specific processes that may drive the tumour, these are then matched to drug databases, with manual literature reviews, to indentify drugs that may be useful or even contra-indicated. Results: Between July 2012 -Jan 2013, 9 subjects (of 30 planned) are enrolled: 2 cases each of: colorectal and breast and 1 each of: squamous skin, squamous ethmoid sinus, nasopharyngeal, lung, and CLL-peripheral mantle cell cancer. 5 have completed analyses. Cancer panel results correlated well with WGS; although the panel is more rapid, it provides less comprehensive information and has not been as informative for identifying candidate druggable drivers. Extensive pathway mapping uncovered potential drug targets in each case that would not have necessarily been considered without this analyses. To date, 4 subjects have started chemo based on the analyses and 1 patient has had his diagnosis radically changed. There are significant genomic differences between archival and fresh tumour samples. Conclusions: This approach is feasible and yields actionable targets that can inform real-time chemotherapy decision-making. Archival samples do not appear to adequately represent post-treatment cancers. The impact of WGS vs. panel sequencing will require more subjects but it appears a panel may be insufficient for detailed treatment guidance.