Tumor mutational burden modulates the prognostic effect of RAS mutations in metastatic colon cancer: mechanistic insights and genotype-phenotype correlations.

BACKGROUND

RAS mutations, present in 40-50% of metastatic colorectal cancer (mCRC) cases, drive oncogenic signaling and confer resistance to anti-EGFR therapies. Tumor mutational burden (TMB), a marker of genomic instability, has recently emerged as a predictive biomarker of response to immunotherapy. However, the prognostic interaction between RAS status and TMB in mCRC remains poorly defined.

PATIENTS AND

METHODS

We analyzed 108 patients with microsatellite-stable metastatic colon cancer (mCC). Tumor samples were profiled using the TruSight Oncology® platform. Eligible patients had an ECOG Performance Status < 2, a cachexia risk score < 1, and no peritoneal carcinomatosis. TMB and RAS mutation status were assessed, and the prognostic significance of the different RAS/TMB combinations was evaluated for overall survival (OS) using Kaplan-Meier and Cox proportional hazards models. Biological differences across selected subgroups were explored using Gene Ontology (GO) enrichment and Phenolyzer network analyses.

RESULTS

RAS mutations were associated with reduced OS (46.4 vs. 67.9 months for mutant vs. wild-type; HR 1.76; P = 0.0495). Stratified analysis showed that the adverse effect of RAS mutations was restricted to patients with low TMB (< 10 mutations/Mb). The subgroup with both RAS mutations and low TMB had the poorest OS (28.0 months; HR 2.34; P = 0.0058), whereas patients with either RAS wild-type or high TMB showed comparable survival. GO analysis revealed enrichment of receptor-mediated signaling pathways in RAS-mutant/TMB-low tumors. Phenolyzer highlighted distinct molecular networks, with APC, TP53, and ERBB2 as central hubs in RAS-mutant/TMB-low tumors, and APC, TP53, and BRCA1 in RAS-wild-type/TMB-high tumors.

CONCLUSIONS

This study demonstrates a prognostic interaction between RAS mutations and TMB in mCC, identifying the RAS-mutant/TMB-low subgroup as having the poorest outcomes. Integrative bioinformatic analyses suggest distinct biological mechanisms underlying these differences. These findings support the development of tailored therapeutic and monitoring strategies for specific molecular subgroups.