CfDNA for accurate determination of RAS and BRAF mutations using OncoBEAM liquid biopsy in metastatic colorectal cancer patients: Results of the real-world multicentric ColoBEAM study.

Abstract

3542 Background: Determination of KRAS, NRAS ( RAS) and BRAF mutations is a standard of care for the management of patients with metastatic colorectal cancer (mCRC). RAS mutations are well characterized resistance biomarkers to anti-EGFR antibodies and BRAF V600 mutations indicate poor prognosis. Tissue biopsy has traditionally been used to determine RAS and BRAF status, but liquid biopsy analysis of circulating tumor DNA (ctDNA) has demonstrated utility as a less invasive tool to expedite molecular testing results to the clinic. The ColoBEAM study reports the performance of plasma mutation testing in a real-life prospective series of 278 patients across 8 centers. Methods: Plasma derived ctDNA was prepared from 20mL blood samples prospectively collected from mCRC patients who had not received chemotherapy in the prior 15 days. ctDNA was centrally assessed using OncoBEAM and results compared to those obtained by routine analysis of tissue. Both tissue and blood samples with discrepant RAS results were blindly reassessed with OncoBEAM. Results: Of 278 patients enrolled, 202 blood samples were available for OncoBEAM testing. RAS and BRAF V600E mutations were detected in tissue in 132/202 (65.4%) and 4/198 (2.0%) patients, respectively. Analysis of the first ctDNA sample as compared to tissue DNA resulted in a kappa coefficient (κ) of 0.52 [0.41 – 0.63] and accuracy of 75.2% (65.1% sensitivity; 94.3% specificity). OncoBEAM testing of a second sample resulted (κ) of 0.66 [0.56 - 0.76] and accuracy of 83.2% (77.3% sensitivity; 94.3% specificity). Of the 4 samples with a BRAF V600E mutation in tumor tissue 2 were detected in blood. In the subgroup of patients with liver metastasis (n=136), accuracy was 88.2% (87.4% sensitivity; 90.2% specificity) for RAS and BRAF status with (κ) of 0.73 [0.61 – 0.86]. In a subgroup of chemotherapy naïve patients with liver metastasis (n=49), accuracy was 91.8% (93.3% sensitivity; 89.5% specificity) for RAS and BRAF status with (κ) of 0.83 [0.67 – 0.99]. Conclusions: The results of the ColoBEAM study confirm plasma ctDNA as a credible surrogate marker to tissue DNA for RAS and BRAF status assessment and may be incorporated as a first-line theragnostic assessment. New testing on a second sample for wild-type status demonstrated 91.8% concordance between blood and tissue. Clinical trial information: NCT02751177.