Metastatic lymph node ratio versus number of metastatic lymph nodes as a prognostic factor in gastric cancer

Objective: Knowledge of prognostic factors in gastric cancer is essential to decide on single patient management. We aim to establish the value of lymph node ratio compared to lymph node involvement in the prediction of gastric cancer survival and treatment approach. Methods: Charts of ninety-six consecutive patients undergoing gastrectomy for resectable gastric cancer were reviewed between January 1996 and December 2005. Receiver operating characteristic (ROC) curves were plotted to verify the accuracy of metastatic lymph node ratio (MLNR) and number of metastatic lymph node (NMLN) cut-off values for survival prediction. Patients were divided into two groups according to ROC curve cut-offs and accuracy in prognosis was reviewed. Results: ROC curves showed that 5 metastatic nodes and a node ratio value of 20% had the best survival prognostic correlation. The median survival of patients with MLNR and NMLN were similar according to cut-off determinations (≤5/>5 metastatic nodes and ≤20/>20% lymph node ratio). Five-year survival rates were 70.9% vs 17.1% and 72.4% vs 15.6%, respectively (p < 0.001). Positive correlation coefficient was found between the number of excised nodes and the number of metastatic nodes. Conclusion: Number of metastatic lymph nodes showed greater accuracy than lymph node ratio for survival prediction in gastric cancer.