Evaluation of a fully automated extended RAS-BRAF test on prospectively collected plasma samples from patients with metastatic colorectal cancer

Background: Extended analysis of RAS genes in metastatic colorectal cancer (mCRC) is not available in many settings and the need for an easy plasma test is high. Methods: Pretreatment plasma samples from 198 mCRC patients enrolled in the prospective multicenter RASANC study (NCT02502656) were retrospectively assessed for the presence of ctDNA mutations in KRAS, NRAS and BRAF by the fully automated Idylla platform.Proper informed consent and statistical analysis were included. The study is ongoing. We used Idylla platform which allows analysis of 21 KRAS, 18 NRAS and 4 BRAF ctDNA mutations from mCRC using 2 ml of plasma. The tests were performed using prototype versions of the Idylla ctKRAS and ctNRAS‐BRAF‐EGFR S492R Mutation Assays (RUO), which contain all reagents for integrated sample preparation and multiplex real time PCR detection. The tests each require 1 ml of plasma, less than 1 minute handson time, and have a turnaround time of less than 130 minutes.As a comparator test, NGS analysis with a sensitivity of 0.2% was performed on extracted plasma ctDNA according to Pecuchet et al. (2016). Results: The Idylla ctDNA assays identified 84 KRAS mutated (42.4%), 6 NRAS mutated (3%) and 13 BRAFmutated (6.6%) cases. The Idylla assays detected 62 KRAS codon 12 (31%), 16 codon 13 (6.5%), 3 codon 61 (1.5%), and 8 codon 146 (4%), and 1 NRAS codon 13 (0.5%) and 5 (2.5%) codon 61 mutated samples. All BRAF mutations detected were V600E. Overall agreement between Idylla and NGS for KRAS, NRAS, and BRAF was 96%, 100%, and 99.5% respectively. OverallRASconcordance with the NGS reference method was 96.5% with a positive agreement of 98% and a negative agreement of 95% (Table 1). Discordancy analysis by means of ddPCR on plasma and/or tissue will be presented. Conclusions: Plasmabased analysis of 41 extended RAS and BRAF mutations in ctDNA on the Idylla platform enables rapid and reliable genotyping in patients with advanced colorectal cancers.