Redefining the prognostic significance of RAS and BRAF V600E mutations on disease free survival in patients with colorectal cancer in the era of ct-DNA defined minimal residual disease: Results from the MD Anderson INTERCEPT Program.

Abstract

196 Background: In patients (pts) with resected colorectal cancer (CRC), circulating tumor DNA (ctDNA) is an emerging tool for the detection of minimal residual disease (MRD) following the completion of curative intent therapy. The presence of MRD is highly prognostic, with MRD+ pts having worse disease-free survival (DFS). Prior to the advent of ctDNA, the modest negative prognostic implications of BRAF and RAS mutations (mt) on DFS were established in some studies. However, it is unclear if these previously noted prognostic effects of RAS/BRAF mts will be maintained with the information provided with ctDNA testing for MRD. Methods: As part of the INTERCEPT program, pts with Stage II-IV CRC treated with curative intent at MD Anderson Cancer Center were evaluated with a tissue-informed ctDNA assay (Signatera). This ctDNA testing was performed along with routine surveillance visits (post-operatively and Q3 months). A cohort of patients from the INTERCEPT program also had correlative analysis performed including whole exome sequencing called by GATK-Mutect2 to assess for BRAFV600E and RAS mt. Results: Within the INTERCEPT program, 265 pts had whole exome sequencing and were included for this analysis, with RAS mutations in 44% (n=116) and BRAF V600E in 3.8% (n=10). After a median follow up of 16-m, the median was DFS 7.93-m (95% CI 5.52 – 10.33-m) in ctDNA+ vs not reached in ctDNA-ve. There was no significant difference between ctDNA+ rates in RAS mt vs RAS wt pts (χ2 p=0.41); 217 pts were ctDNA-ve with a DFS HR=1.81 for RAS mut vs wt (95% CI 0.57-5.73, p=0.30). The 12-m DFS in this cohort was 95.2% in RAS mt vs 97.9% in RAS wt (OR 1.25, 95% CI 0.4-4.4, p=NS). 48 pts were ctDNA+, with no difference in DFS by RAS mt status with HR=1.42 (95% CI 0.76 – 2.64, p = NS), which was unchanged after excluding those with radiographically evident disease at or shortly after the finding of ctDNA+. In a multivariate model, after controlling for ctDNA status, there was no prognostic impact of RAS on DFS in the larger population with HR 1.51 (95% CI 0.88 – 2.60; p = NS). While the number of pts with BRAF V600E mt was low, there was no evidence of a prognostic effect after controlling for ctDNA status. Conclusions: In CRC pts treated with curative intent, positive ctDNA at any point post-operatively or during surveillance is highly predictive of disease recurrence. In these pts, RAS mt status was not associated with DFS. Moreover, RAS mt pts are not more likely to be ctDNA+ve. Thus, our data suggest that the presence of ctDNA defined MRD has greater prognostic value than the presence or absence of RAS or BRAF V600E mt.