Multivariate Risk Analysis of RAS, BRAF and EGFR Mutations Allelic Frequency and Coexistence as Colorectal Cancer Predictive Biomarkers.

Abstract

Simple SummaryThe colorectal cancer (CRC) stage and evolution should be described by biomarker profiles. In 60 CRC cases, KRAS, NRAS, BRAF, and EGFR mutations were analyzed by droplet digital PCR (ddPCR). KRAS G12/G13 mutation was present in all patients with variable allelic frequencies. KRAS Q61 mutation was correlated with tumor invasion beyond the subserosa and poor differentiation, both associated with worst prognosis. Tumors with NRAS and BRAF mutations were prevalently localized on the right segment colon. The discovery of the KRAS Q61 role in tumor phenotypes provides the foundation for new therapeutic strategies and perspectives on molecular subtypes classification of CRC.Background: Biomarker profiles should represent a coherent description of the colorectal cancer (CRC) stage and its predicted evolution. Methods: Using droplet digital PCR, we detected the allelic frequencies (AF) of KRAS, NRAS, BRAF, and EGFR mutations from 60 tumors. We employed a pair-wise association approach to estimate the risk involving AF mutations as outcome variables for clinical data and as predicting variables for tumor-staging. We evaluated correlations between mutations of AFs and also between the mutations and histopathology features (tumor staging, inflammation, differentiation, and invasiveness). Results: KRAS G12/G13 mutations were present in all patients. KRAS Q61 was significantly associated with poor differentiation, high desmoplastic reaction, invasiveness (ypT4), and metastasis (ypM1). NRAS and BRAF were associated with the right-side localization of tumors. Diabetic patients had a higher risk to exhibit NRAS G12/G13 mutations. BRAF and NRAS G12/G13 mutations co-existed in tumors with invasiveness limited to the submucosa. Conclusions: The associations we found and the mutational AF we reported may help to understand disease processes and may be considered as potential CCR biomarker candidates. In addition, we propose representative mutation panels associated with specific clinical and histopathological features of CRC, as a unique opportunity to refine the degree of personalization of CRC treatment.