RESUMO
<p><b>OBJECTIVE</b>To explore the association between rs659366 polymorphisms and the outcomes of patients after surgery for colorectal cancer.</p><p><b>METHODS</b>The study was conducted among a cohort of 501 patients with primary colorectal cancer who had surgery in Sichuan Cancer Hospital during March 2010 and July 2013. The outcomes of the patients were followed up. The polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) method was applied to detect rs659366 genotypes. The log-rank test was performed to analyze the effects of clinical features on patients' outcomes. The correlation between rs659366 polymorphisms and the outcomes of patients was analyzed using the Cox proportional hazard model.</p><p><b>RESULTS</b>In this study, the median of follow-up time was 44.23(0.13-78.53)months, and 101 out of 501 (20.2%) patients failed to follow-up. The log-rank test showed the tumor site, TNM stage, vascular invasion, perineural invasion and the preoperative carcino-embryonic antigen(CEA) level were significantly associated with the outcome of colorectal cancer (<0.05 or <0.01). The overall survival rate of patients with AA, GA and GG genotypes were 62.7%, 69.9% and 75.5%, respectively. Multivariate analysis according to Cox proportional hazard model taking the GG genotype as the reference indicated that the AA genotype increased risks for survival of patients (=1.823); under the dominant genetic model taking GG genotype as reference, GA+AA genotypes increased risks for the poorer outcomes of patients (=1.498); the addictive genetic model showed that allele A increased the hazard for the poorer outcomes (=1.787).</p><p><b>CONCLUSIONS</b>The rs659366 polymorphisms are significantly associated with the outcome of patients with colorectal cancer.</p>