Purpose: Genetic alterations that are closely associated with patient prognosis can be landmarks of definitive therapeutic targets as well as useful biomarkers in human cancer clinics.
Methods: Three hundred seventy-eight colorectal cancer (CRC) patients were examined for K-ras mutations by single-strand conformation polymorphism (SSCP), with a subsequent 144 young colon cancer (YCC) patients added to validate its prognostic significance.
Results: K-ras mutations were identified in 161 (43%) of the 378 CRC patients and were significantly associated with tumor location (colon vs. rectum; 80/218 = 37% vs. 81/160 = 51%; P = 0.0068) and age (≥60 vs. <60; 103/220 = 47% vs. 58/158 = 37%; P = 0.049). The incidence of K-ras mutations was 30% in YCC patients as compared to 55% in elderly rectal cancer patients (P = 0.0004). K-ras mutations significantly correlated with a worse prognosis (P = 0.0014) only in 73 curatively resected YCC with stages I-III, but not in other CRCs, which was further validated in the independent set of the corresponding 144 YCC patients (P = 0.024). Both univariate and multivariate analyses identified K-ras mutations as an independent prognostic factor (HR = 5.5, P = 0.029; HR = 3.6, P = 0.011) in both learning and validation sets of the curatively resected YCC with stages I-III, respectively, and the prognostic relevance was marked in stage III YCC patients (P = 0.002), but not in stages I, II, and IV.
Conclusion: In curative YCC, K-ras mutations could have excellent prognostic value. Hence, the K-ras mutation status could be a good indicator to predict the clinical outcome in curatively resected stage III YCC patients, and K-ras pathway inhibition may be a relevant therapeutic target in CRC, excluding YCC patients with no K-ras mutation.