PO-320 Gene panel mutation screening for a better molecular stratification of colorectal cancer patients

Abstract

Introduction Colorectal carcinoma (CRC) is one of the most commonly diagnosed cancers worldwide. The metastatic disease contributes to the high mortality rate reported for such tumours. Significant benefit on overall survival was brought about the introduction of monoclonal antibodies anti-EGFR and anti-VEGF used in combination with chemotherapy in metastatic CRC (mCRC). While anti-VEGF treatment does not require biomarker-based selection criteria, the potential efficacy of anti-EGFR antibodies is neglected to patients with activating mutations in KRAS and NRAS (RAS) genes, whose molecular analysis became a clinical routine. The advent of Next Generation Sequencing (NGS) instruments, able to reach quick testing of multiple clinically-relevant hotspots, yet maintaining precision and low costs, allows the simultaneous determination of the mutation status of an expanding number of genes. Despite only few of these molecular biomarkers have gained clinical utility in the routine oncological practice, the acquisition of more complex cancer mutational patterns may provide more efficient tumour characterisation for prognostic and predictive purposes and highlight actionable targets. Material and methods We sequenced 639 mCRC samples by IT-PGM platform using a panel of hotspots and targeted regions of 22 genes (including RAS) commonly involved in CRCs. MSI analyses on 89 patients have been performed with a single fluorescent system comprising BAT25 and BAT26 mononucleotide repeats. Results and discussions We identified recurrent mutations (≥1%) in 12/22 genes, being KRAS, TP53 and PIK3CA the most frequently mutated ones. Statistical analysis, indicated that the mutation associations follow a non-random distribution. Categorization of the cases on the base of KRAS and p53 mutation status led us the definition of 8 Mutation Association Patterns (MAPs) characterised by specific mutation associations. Analysis of the clinicopathological data available for 89 out of 639 cases indicates interesting trends for the associations of MAPs with specific parameters, some of which reached statistical significance. Conclusion Application of NGS gene panel as a routine for the characterisation of RAS/BRAF status required for predictive purposes in CRC patients, may provide additional prognostic/predictive information, with no significant extra-costs.