Plasma ctDNA RAS mutation analysis for the diagnosis and treatment monitoring of metastatic colorectal cancer patients

Abstract

Background: RAS assessment is mandatory for therapy decision in metastatic colorectal cancer (mCRC) patients. This determination is based on tumor tissue, however, genotyping of circulating tumor (ct)DNA offers clear advantages as a minimally invasive method that represents tumor heterogeneity. Our study aims to evaluate the use of ctDNA as an alternative for determining baseline RAS status and subsequent monitoring of RAS mutations during therapy as a component of routine clinical practice.

Patients and methods: RAS mutational status in plasma was evaluated in mCRC patients by OncoBEAM™ RAS CRC assay. Concordance of results in plasma and tissue was retrospectively evaluated. RAS mutations were also prospectively monitored in longitudinal plasma samples from selected patients.

Results: Analysis of RAS in tissue and plasma samples from 115 mCRC patients showed a 93% overall agreement. Plasma/tissue RAS discrepancies were mainly explained by spatial and temporal tumor heterogeneity. Analysis of clinico-pathological features showed that the site of metastasis (i.e. peritoneal, lung), the histology of the tumor (i.e. mucinous) and administration of treatment previous to blood collection negatively impacted the detection of RAS in ctDNA. In patients with baseline mutant RAS tumors treated with chemotherapy/antiangiogenic, longitudinal analysis of RAS ctDNA mirrored response to treatment, being an early predictor of response. In patients RAS wt, longitudinal monitoring of RAS ctDNA revealed that OncoBEAM was useful to detect emergence of RAS mutations during anti-EGFR treatment.

Conclusion: The high overall agreement in RAS mutational assessment between plasma and tissue supports blood-based testing with OncoBEAM™ as a viable alternative for genotyping RAS of mCRC patients in routine clinical practice. Our study describes practical clinico-pathological specifications to optimize RAS ctDNA determination. Moreover, OncoBEAM™ is useful to monitor RAS in patients undergoing systemic therapy to detect resistance and evaluate the efficacy of particular treatments.

Keywords: RAS mutations; colorectal cancer; ctDNA; heterogeneity; liquid biopsy; tumor dynamics.

© The Author 2017. Published by Oxford University Press on behalf of the European Society for Medical Oncology. All rights reserved. For permissions, please email: journals.permissions@oup.com.

Figures

Figure 1.

Study flowchart. Number of patients included in each of the analysis endpoints and reasons for exclusion are depicted.

Figure 2.

Comparison of RAS mutations detected in tissue versus plasma and analysis of discrepancies. Overall concordance analysis between RAS mutations detected in tumor by SoC and BEAMing plasma. Positive agreement (patients RAS mutated in plasma and tissue analysis) and negative agreement (patients wild-type in tissue and plasma). Clinico-pathological and treatment characteristics from eight discordant cases.

Figure 3.

Correlation of circulating RAS mutations and clinico-pathological characteristics. (A) Differences in RAS ctDNA mutant allele fraction (MAF) in patients with concordant RAS plasma and tissue determination compared with discordant cases (RAS wild-type in plasma and RAS mutated in tissue or RAS mutated in plasma and RAS wild-type in tissue). (B) Differences in RAS ctDNA MAF according to patients with liver metastasis versus patients without liver involvement. (C) Differences in RAS ctDNA MAF according to patients without peritoneum involvement, patients with only peritoneum metastasis and patients with peritoneum plus another metastatic site. (D) Differences in RAS ctDNA MAF according to treatment naïve patients versus patients with previous systemic treatment received within a month prior ctDNA blood extraction.

Figure 4

Longitudinal analysis of plasma RAS ctDNA to evaluate response to treatment with chemotherapy ± antiangiogenic. (A) RAS ctDNA dynamics in nine patients with RAS mutated tumors treated with chemotherapy ± antiangiogenic that initially respond to treatment. Frequency of circulating RAS mutant alleles at baseline, at time of first CT-scan to evaluate treatment response and at disease progression. Decline and increase in circulating RAS MAF correlate with response and progression to treatment, respectively. (B) RAS ctDNA dynamics in 5 patients with RAS mutated tumors that progressed at first CTscan at 8–12 weeks from beginning of treatment. (C) Patient diagnosed with stage IV rectal cancer with liver metastasis. An NRAS codon 61 mutation was detected in tissue and plasma. After 4 weeks of treatment with FOLFOX + bevacizumab, plasma MAF dramatically decreased correlating with a stable disease observed in the CT-scan at week 12. The patient underwent surgery of the primary tumor and liver metastasis, and plasma RAS became undetectable. Eight months later, the patient relapsed and RAS MAF increased accordingly. Three months after initiating second line treatment with FOLFIRI + aflibercept, the patient achieved a stable disease by CT scan, and no plasma ctDNA RAS mutations were detected. (D) Monitoring ctDNA KRAS codon 146 mutation during treatment with FOLFOX-Bevacizumab in patient #3, diagnosed with a stage IV colon cancer with lung and liver metastasis. The colonoscopic biopsy analysis was RAS wt but plasma ctDNA showed a KRAS codon 146 mutation. Following removal of the primary tumor, re-analysis of RAS in the surgical sample confirmed the plasma result. The patient received FOLFOX + bevacizumab with an early decrease in RAS ctDNA that became undetectable at 12 weeks, alongside at the first CT scan. Treatment was discontinued and a subsequently increase in KRAS codon 146 MAF was observed, which then rapidly decreased when the chemotherapy was reintroduced. Gray area indicates tumor load. Blue line indicates changes in ctDNA KRAS146 frequency.