Discordance of KRAS Mutational Status between Primary Tumors and Liver Metastases in Colorectal Cancer: Impact on Long-Term Survival Following Radical Resection

Simple Summary KRAS mutational heterogeneity between primary colorectal cancer and liver metastases may present a challenge in assessing prognostic information prior to the multimodal treatment. Aim of our study is to assess the incidence of KRAS discordance in a single-center series by comparing primary colorectal tumor specimens with the corresponding liver metastasis. Mutation analyses in all patients were performed at the same institution. Impact of discordance on overall survival and risk factors associated with discordance were analyzed. Our study showed that KRAS discordance was observed in 15.9% of patients. In patients with wild-type colorectal primary tumors, subsequent acquisition of mutation in the corresponding liver metastasis was associated with worse overall survival and was significantly more frequently found in patients with multiple liver metastases. These results suggested that, in the era of precision medicine, the possibility of KRAS discordance should be taken into account within the multidisciplinary management of patients with metastatic colorectal cancer. If KRAS mutation status of primary colorectal tumor is representative of corresponding colorectal liver metastases (CRLM) mutational pattern, is controversial. Several studies have reported different rates of KRAS discordance, ranging from 4 to 32%. Aim of this study is to assess the incidence of discordance and its impact on overall survival (OS) in a homogenous group of patients. KRAS mutation status was evaluated in 107 patients resected for both primary colorectal tumor and corresponding CRLM at the same institution, between 2007 and 2018. Discordance rate was 15.9%. Its incidence varied according to the time interval between the two mutation analyses (p = 0.025; Pearson correlation = 0.2) and it was significantly higher during the first 6 months from the time of primary tumor evaluation. On multivariable analysis, type of discordance (wild-type in primary tumor, mutation in CRLM) was the strongest predictor of poor OS (p < 0.001). At multivariable logistic regression analysis, the number of CRLM >3 was an independent risk factor for the risk of KRAS discordance associated with the worst prognosis (OR = 4.600; p = 0.047). Results of our study suggested that, in the era of precision medicine, possibility of KRAS discordance should be taken into account within multidisciplinary management of patients with metastatic colorectal cancer.