A novel clinical prognostic score incorporating the number of resected lymph-nodes to predict recurrence and survival in non-small-cell lung cancer

Background: The number of resected lymph-nodes (#RNs) has proven prognostic in breast and colorectal cancer. Here we evaluated its prognostic impact in a series of resected NSCLC patients. Methods: A panel of established prognostic factors plus (1) #RNs or (2) the ratio between the number of metastatic nodes and #RNs (NR) were correlated to overall- (OS), cancer-specific- (CSS), and disease-free-survival (DFS), using the Cox-model. Risk-classes according to hazard ratios (HR) were generated. Internal and external validation was accomplished. Results: A dataset of 415 resected NSCLC patients was retrieved. At multivariate analysis, #RNs and NR were independent factor for longer OS, CSS and DFS (p<0.0001). Patients with a #RNs> 10 (identified optimal cut-off) had a statistically significant CS (p = 0.02) and DFS (p = 0.0005) benefit. In node-positive patients, a NR < 9% significantly correlated with better outcome. Stratification into High-, Medium-, and Low-Risk classes, based on High- (HRFs: stage, N-status, age, #RNs) and Intermediate-Risk Factors (IRFs: sex, grading, histology), efficiently predicted outcomes (p < 0.0001). The risk class model performance was externally validated in and independent dataset of 297 patients. Conclusions: These results contribute to complete the panel of prognostic factors for resected NSCLC. A prospective larger validation and comparison with molecular prognostic tools is warranted. (C) 2009 Elsevier Ireland Ltd. All rights reserved.