Carcinoembryonic antigen trajectory predicts pathological complete response in advanced gastric cancer after neoadjuvant chemotherapy

Aims This study aims to develop a simple, clinically applicable classification system to predict pCR based on carcinoembryonic antigen (CEA) trajectory during NAC. Methods This study included 366 AGC patients who received NAC followed by radical gastrectomy. CEA levels were measured before, during, and after NAC, with changes classified into three trajectory types: Type I (>=80% decline), Type II (>=40% but <80% decline), and Type III (<40% decline or increase). We analyzed associations between these CEA trajectories, pCR, lymph node remission, and survival. Results pCR was achieved in 10.4% (38/366) of patients. pCR rates were significantly higher in Type I (41%) and Type II (15.8%) trajectories compared to Type III (6.7%). Lymph node remission also correlated with CEA trajectories, with Type I having the highest proportion of ypN0 (79.2%). Multivariate analysis identified CEA trajectory subtypes and tumor differentiation as independent predictors of pCR. This classification system proved robust across subgroups. Although no significant differences in overall survival were observed between subtypes, higher initial CEA levels were associated with worse survival. Conclusion The trajectory of CEA change during NAC is a promising predictor of pCR in AGC. This simple and accessible classification system may facilitate personalized surgical strategies for patients with AGC.