[Comparative analysis of prognostic performance among different lymph node staging systems in gastric cancer]

Objective: To establish a lymph node staging system based on the number of metastatic lymph nodes and their anatomical distribution (sN), and to evaluate its prognostic predictive performance in gastric cancer patients undergoing D2 radical resection. Methods: A retrospective analysis was conducted on 200 patients who underwent D2 radical gastrectomy for gastric cancer at the Cancer Hospital, Chinese Academy of Medical Sciences, between January 2014 and December 2019. Patients were staged according to four lymph node staging methods: pathological regional lymph node (pN) staging defined by the 8th edition of the American Joint Committee on Cancer (AJCC) TNM classification, lymph node ratio (LNR), log odds of positive lymph nodes (LODDS), and sN staging system. Survival differences among subgroups were compared using the log-rank test. Independent prognostic factors for survival were identified using the Cox proportional hazard model. The prognostic performance and model fit of each staging system were evaluated using the area under the time-dependent receiver operating characteristic (ROC) curve, concordance index (C-index), and Akaike information criterion (AIC). Results: Among the 200 patients, 44, 32, 55, 33, and 34 patients were classified as pN0, pN1, pN2, pN3a, and pN3b, respectively, with corresponding 5-year survival rates of 95.0%, 90.3%, 85.6%, 56.3%, and 40.7%, respectively. No significant differences in survival were observed between the pN0 and pN1 groups, pN1 and pN2 groups, pN0 and pN2 groups, or pN3a and pN3b groups (all P＞0.05). According to LNR staging, 116, 36, and 46 patients were classified as LNR1, LNR2, and LNR3, with 5-year survival rates of 89.5%, 73.3%, and 40.4%, respectively. According to LODDS staging, 117, 35, and 46 patients were classified as LODDS1, LODDS2, and LODDS3, with corresponding 5-year survival rates of 89.5%, 73.2%, and 40.4%, respectively. According to the sN staging system, 93 patients were classified as sN1, 50 as sN2, and 55 as sN3, with corresponding 5-year survival rates of 94.3%, 76.5%, and 46.3%, respectively. Significant survival differences were observed among the three subgroups of the LNR, LODDS, and sN staging systems (all P＜0.001). Multivariate Cox regression analysis demonstrated that pN, LNR, LODDS, and sN staging systems were independent influencing factors for overall survival (all P＜0.05), although no significant differences were observed in the pN1 and pN2 subgroups of the pN staging system (all P＞0.05). The AUCs for predicting the 5-year survival using the pN, LNR, LODDS, and sN staging systems were 0.751, 0.757, 0.757, and 0.789, respectively; the corresponding C-index values were 0.752, 0.754, 0.754, and 0.763; and the AIC values were 384.544, 381.981, 381.913, and 380.395, respectively. Conclusions: The sN staging system, which integrates the number and anatomical location of metastatic lymph nodes, demonstrates superior performance in long-term survival prediction compared with traditional pN, LNR, and LODDS staging systems for gastric cancer patients undergoing D2 radical gastrectomy. It may serve as a practical tool for postoperative risk stratification in this patient population.