The use of epidermal growth factor receptor inhibitors to treat metastatic colorectal cancer (mCRC) patients requires prior confirmation of tumour wild type (WT) RAS mutation status (exons 2/3/4 for KRAS or NRAS). This retrospective pooled analysis aims to robustly estimate RAS mutation prevalence and individual variation patterns in mCRC patients.

Individual patient data from five randomised, controlled panitumumab studies (three phase III, one phase II and one phase Ib/II) were pooled for this analysis. The phase III studies included mCRC patients independent of RAS mutation status; the phase II and Ib/II studies included mCRC patients with confirmed WT KRAS exon 2 status. Four studies conducted RAS testing using Sanger sequencing; one study used a combination of next-generation sequencing and Sanger sequencing. In order to assign overall RAS status, the mutation status of all exons 2/3/4 KRAS or NRAS was required to be known.

Data from 3196 mCRC patients from 36 countries were included in the analysis. The overall unadjusted RAS mutation prevalence in mCRC patients was 55.9% (95% confidence interval (CI): [53.9–57.9%]), with the following distribution observed: KRAS exon 2 (prevalence 42.6% [40.7–44.5%]); KRAS exon 3 (3.8% [2.9–4.9%]); KRAS exon 4 (6.2% [5.0–7.6%]); NRAS exon 2 (2.9% [2.1–3.9%]); NRAS exon 3 (4.2% [3.2–5.4%]); NRAS exon 4 (0.3% [0.1–0.7%]). Differences in RAS mutation prevalence estimates were observed by study (p = 0.001), gender (p = 0.030), and by country (p = 0.028).

This analysis provides robust estimates of overall RAS mutation prevalence and individual variation patterns in mCRC patients.