Association of Metastatic Lymph Node Count and Primary Tumor Size with Disease-Free Survival in Stage II–III Gastric Cancer: A Retrospective Cohort Study

Abstract

Background and Objectives: The relative prognostic information provided by different nodal metrics and the independent role of primary tumor size remain unclear in gastric cancer. This study evaluated the association of metastatic lymph node count and primary tumor size with disease-free survival (DFS) in stage II–III gastric cancer (GC) treated with a surgery-first approach. Materials and Methods: This retrospective single-center study included 128 patients who underwent curative-intent gastrectomy between 2003 and 2019 without neoadjuvant treatment. A multivariable Cox model with clinically prespecified covariates was fitted. Metastatic lymph node count, pathological N (pN) category, lymph node ratio (LNR), and log odds of positive lymph nodes (LODDS) were each added separately to the same base clinical model and compared using the Akaike information criterion (AIC). Results: During a median follow-up of 114.3 months, 86 DFS events occurred. Each additional metastatic lymph node was independently associated with a higher hazard of a DFS event (HR, 1.049; 95% CI, 1.022–1.076; p < 0.001), and this association persisted across four sensitivity analyses. Primary tumor size was not independently associated with DFS. All four nodal metrics improved model fit, with LNR yielding the lowest AIC. Conclusions: Metastatic lymph node count was independently and robustly associated with DFS, whereas primary tumor size was not. Although LNR provided the best relative model fit, these within-cohort comparisons do not establish clinical superiority, and external validation is required.