Abstract
Objectives: We aimed to evaluate the effectiveness of the discrimination ability of newly proposed nodal classifications based on the location or number of metastatic lymph nodes in non-small-cell lung cancer (NSCLC). Methods: A retrospective analysis was conducted on 1487 patients who underwent complete resection with systematic lymph node dissection for NSCLC between 2000 and 2010. Five nodal classifications were analysed based on the following categories: current nodal classification (N1, N2), zone classification (single-zone N1, multiple-zone N1 or single-zone N2, and multiple-zone N2), number classification (1–2, 3–6, and more than 7), ratio classification (15%, 15%–40%, and more than 40%), and location with number classification (N1-number from 1 to 3, N1-number more than 4, N2-number from 1 to 3, and N2-more than 4). C-index was used as a means of assessing discrimination by the models. Results: In multivariate analysis, all of the newly proposed classifications were independent predictors (P < 0.001) for overall survival after adjustment by significant variables (age, cell type, pathologic T stage). The C-index of the classifications based on current nodal system, zone, number, ratio, and location with number was 0.541, 0.562, 0.575, 0.555 and 0.577, respectively, although the differences of C-indices were statistically insignificant. Conclusion: Although discrimination ability is not significantly different between the newly proposed classifications, discrimination ability tends to increase in models considering the number of metastatic lymph nodes. These results may justify the inclusion of number of metastatic lymph nodes in new nodal stage classification in the future.