Exploration of the optimal lymph node count retrieved for patients with node-negative colonic mucinous cancer

This study aims to explore the effect of lymph node (LN) count retrieved on the survival of patients with node-negative colonic mucinous adenocarcinoma (MAC). All registered patients with colon cancer from the Surveillance, Epidemiology, and End Results Program (SEER) database were collected for analysis in this study. X-tile plots were used to find the optimal cutoff values of LN counts in MAC patients. Based on these numbers, patients were divided into high, middle and low risk subsets in terms of colon cause specific survival (CCSS). Chi-square (chi(2)) test was applied to analyze the effect of LN count on CCSS. Chi-square (chi(2)) test was applied to analyze the clinicopathologic parameters in different LN subgroups. Univariate survival analysis and multivariate Cox proportional hazards model were used to assess the risk factors affecting survival. A total of 17,280 patients with node-negative colonic MAC between 1988 and 2013 were included. Based on the LN count, all patients were stratified into high (1-7), middle (8-17), and low (> 18) risk subsets, and their 5-year CCSS rates were 80.8%, 85.0% and 89.5%, respectively. We found that the LN counts of each subset was an independently prognostic factor using multivariate Cox analysis (HR=0.765, 95% CI=0.702-0.833; HR=0.518, 95% CI=0.465-0.577. P<0.001). Our study demonstrated that the LN count was an independent prognostic factor for MAC. The higher the LN counts retrieved, the longer the survival would be. The longest CCSS was observed in patients with more than 17 LNs.