Prevalence of RAS mutations among patients with metastatic colorectal cancer by country and region in randomized clinical trials: A pooled analysis.

Abstract

520 Background: Global mCRC treatment guidelines for EGFR inhibitors require prior confirmation of wild-type RAS status (KRAS exons 2, 3, 4 and NRAS 2, 3, 4). The aim of this study was to estimate the prevalence of RAS mutations among mCRC patients by country, demographic characteristics, and clinical risk factors. A secondary aim was to estimate BRAF and KRASexon 2 mutation prevalence. Methods: Data from 5 published Amgen-sponsored randomized clinical trials (RCTs) were merged in a retrospective pooled analysis. There were 3 phase III, 1 phase II, and 1 phase Ib/II studies. For 4 out of 5 RCTs, RAStesting was conducted in a U.S. laboratory (Transgenomic Inc.) using Sanger sequencing on DNA extracted from tumor samples. For the remaining trial, a combination of next-generation sequencing and Sanger sequencing was used. Results: A total of 3,196 patients from 36 countries were included. The overall unadjusted prevalence of RAS mutations among mCRC subjects was 55.9% (95% CI, 53.9%, 57.9%); KRAS exon 2 mutation prevalence was 42.6% (40.7%, 44.5%). The prevalence by exon is given in the table below. BRAFmutation prevalence was 8.1% (6.7%, 9.6%). There were no statistically significant differences in RAS mutation prevalence by gender, age, or clinical factors such as performance status, tumour site, biopsy origin, or metastasis characteristics. Statistically significant differences in RAS mutation prevalence estimates were observed by country and by region with rates for Central West Europe being significantly lower than Eastern Europe (49.4% [44.4%, 54.3%] and 61.5%; [55.3%; 67.4%] respectively). Statistically significant RASmutation prevalence differences were observed between studies which could be due to varying patient characteristics. Conclusions: By merging data from RCTs, the analysis provides robust estimates of RASmutation prevalence. Studies using observational data are needed to confirm these findings. [Table: see text]