Background: The prognostic and predictive value of KRAS mutations in patients with lung cancer is controversial. Biases in disease stage, treatment regimen, small-scale patient studies, and biomarker status have led to inconsistent results in previous reports.
Methods: The KRAS and EGFR genes were examined in 1935 consecutive patients with non-small cell lung cancer. All patients were divided into KRAS mutation (KRAS group), EGFR mutation (EGFR group), and KRAS/EGFR wild type (WT group) groups. Randomly selected cases were paired with patients with the KRAS mutation, the EGFR mutation, and KRAS/EGFR wild type patients according to tumor, node, metastasis stage, time of first visit within 1 year, and pathology. Progression-free survival (PFS) and overall survival were evaluated by Kaplan-Meier and Cox models.
Results: The KRAS mutation rate for lung adenocarcinoma was 5.90 %. The overall survival was 14.47, 20.57, and 42.73 months for the KRAS group, WT group, and EGFR group, respectively (P < 0.001). Multivariate analysis indicated that KRAS mutation status was an independent prognostic factor (hazard ratio 2.69, 95 % confidence interval 1.91-3.80, P < 0.001). No difference was found in PFS and tumor responsiveness between patients with a KRAS mutation and those with wild type KRAS/EGFR for chemotherapy and EGFR tyrosine kinase inhibitors (TKI). PFS did not significantly differ for chemotherapy among the three groups (P = 0.270).
Conclusions: KRAS mutation is a poor prognosis factor, but it is not an independent predictor of response to EGFR-TKI or chemotherapy in patients with lung cancer.