Background: Emerging data suggest that somatic KRAS mutation in advanced colorectal cancer is a strong predictor of non-response to anti-epidermal growth factor receptor antibody (anti-EGFR) therapy.
Patients and methods: A comprehensive search through March 2010 identified randomized controlled trials in metastatic colorectal cancer that evaluated chemotherapy regimens or best supportive care, with and without anti-EGFR therapy. Outcomes included progression-free survival (PFS), median overall survival (OS), and predictive test performance.
Results: In pooled data from 8 trials with 5325 patients, the addition of anti-EGFR to standard chemotherapy resulted in improved PFS (HR 0.66 [95% CI, 0.53-0.82]) in patients with wild-type KRAS in the tumor tissue, but not in patients with KRAS mutation (HR 1.07 [95% CI, 0.91-1.27]). Anti-EGFR treatment in the wild-type group did not significantly improve median OS. As a predictive biomarker, KRAS mutation had a positive likelihood ratio of 2.0 (95% CI, 1.45-2.76) in predicting nonresponse to anti-EGFR treatment.
Conclusion: In patients with advanced colorectal cancer, the addition of anti-EGFR treatment to standard chemotherapy improves PFS for those with wild-type, but not mutant KRAS status. KRAS gene mutation testing provides a fair biomarker in predicting non-response to anti-EGFR treatment.