Résumé
Abstract only 10505 Background: We developed a specific method for circulating cell-free DNA (ctDNA) enabling the detection of point mutations. CtDNA exists at high level in patients with different types of cancer, and presents great potential in regards, in particular, to its low invasiveness, rapid data turnaround and cost effectiveness. We described here the first blinded prospective study on the comparison of KRAS and BRAF mutational status data obtained from the analysis of ctDNA and tumor section. Methods: Intplex is a refined Q-PCR-based method specifically designed to analyze ctDNA. IntPlex allows not only the detection of point mutations, but also the determination of tumor-ctDNA concentration and fragmentation index simultaneously. Intplex sensitivity is 0.01% (mutant to WT ratio) and is unprecedented among Q-PCR-based methods. The study was conducted from a multicenter cohort of 79 metastatic CRC patients not having received chemo- or radio-therapy within the month prior to blood sample collection. Results: Mutational status could not be determined in 9 samples with one of the two methods. CtDNA analysis showed 100% specificity and 87% sensitivity for KRAS detection (only three samples are misclassified) as compared to tumor section analysis with 23/70 (33%) positive samples, the concordance value being 96%. For BRAF mutation, the method exhibited a specificity and a sensibility of 100% with 5/70 (7%) positive samples (concordance of 100%). When combining KRAS and BRAF mutations, 100% specificity, 89% sensibility and 98% concordance were determined. The three discordant samples may be explained by the malignant genotype of primary tumor versus metastasis, or by the sensitivity of our detection method. The mutation load (median, 12.4%) expressed as the proportion of mutant allele in ctDNA is highly variable (0.037% to 69%) among mutated samples showing a very high inter-individual heterogeneity. Conclusions: Our blinded prospective multicenter study clearly showed for the first time that tumor section analysis might be advantageously replaced by ctDNA analysis, enlarging personalized medicine power for cancer patients.