ABSTRACT: Purpose:KRAS mutations and tumor location have been associated with response to targeted therapy among patients with stage IV colorectal cancer (CRC) in various trials. This study performed the first population-based examination of associations between KRAS mutations, tumor location, and survival, and assessed factors associated with documented KRAS testing. Methods: Patients with stage IV adenocarcinoma of the colon/rectum diagnosed from 2010 to 2013 were extracted from SEER data. Analyses of patient characteristics, KRAS testing, and tumor location were conducted using logistic regression. Cox proportional hazards models assessed relationships between KRAS mutations, tumor location, and risk of all-cause death. Results: Of 22,542 patients, 30% received KRAS testing, and 44% of these had mutations. Those tested tended to be younger, married, and metropolitan area residents, and have private insurance or Medicare. Rates of KRAS testing also varied by registry (range, 20%-46%). Patients with right-sided colon cancer (vs left-sided) tended to be older, female, and black; have mucinous, KRAS-mutant tumors; and have a greater risk of death (hazard ratio [HR], 1.27; 95% CI, 1.22-1.32). KRAS mutations were not associated with greater risk of death in the overall population; however, they were associated with greater risk of death among patients with left-sided colon cancer (HR, 1.19; 95% CI, 1.05-1.33). Conclusions: This large population-based study showed that among patients initially diagnosed with stage IV CRC, right-sided colon cancer was associated with greater risk of death compared with left-sided cancer, and KRAS mutations were only associated with risk of death in left-sided colon cancer. An unexpected finding was that among patients with stage IV disease, right-sided cancer was more commonly seen in black patients versus whites. Future studies should further explore these associations and determine the role of biology versus treatment differences. In addition, use of KRAS testing is increasing, but there is wide geographic variation wherein disparities related to insurance coverage and rurality may warrant further study.
