Prognostic nomogram for gastric lymphoepithelioma-like carcinoma: A retrospective cohort analysis

BACKGROUNDThe prognosis of gastric lymphoepithelioma-like carcinoma (LELGC), a unique type of gastric cancer, varies substantially.The lack of individualized prognostic assessment tools for LELGC restricts the ability to precisely optimize clinical treatment strategies. AIMTo construct a model for predicting the prognosis of patients with LELGC. METHODSIn this retrospective cohort study, the clinicopathological data of 69 patients with LELGC who underwent surgical treatment at the First Affiliated Hospital of Soochow University from January 2013 to January 2025 were analyzed.The patients comprised 50 men and 19 women, aged 34 years to 82 (61.7 ± 10.4) years.Independent prognostic risk factors for overall survival (OS) and recurrence-free survival (RFS) were screened using a Cox regression model, and a nomogram prediction model was constructed.The receiver operating characteristic curve (area under the curve), calibration curve, and decision curve analysis were used to assess the model's discrimination, calibration, and clinical net benefit, respectively.The results were compared with the tumor-nodes-metastasis (TNM) stage system outlined in the eighth edition of the American Joint Committee on Cancer. RESULTSDuring the follow-up period, the median OS and RFS of 69 patients with LELGC were 60 (2-117) months and 57 (2-117) months, respectively.The 1-, 3-, and 5-year OS rates were 87.0%, 73.9%, and 49.3%, respectively.In contrast, the 1-, 3-, and 5-year RFS rates were 85.5%, 63.8%, and 42.0%, respectively.Multivariate analysis showed that TNM stage III-4 / 34 IV, human epidermal growth factor receptor 2 positivity, CD3 negativity, and CD20 negativity were independent risk factors affecting OS of LELGC patients.Vascular invasion, Perineural invasion, human epidermal growth factor receptor 2 positivity, carcinoembryonic antigen > 4.62 ng/mL, and CD3 negativity were independent risk factors affecting RFS of LELGC patients.The area under the curve values were 0.833, 0.921, and 0.956 for predicting 1-, 3-, and 5-year OS, respectively.In contrast, the values were 0.940, 0.945, and 0.866 for predicting 1-, 3-, and 5-year RFS, respectively.The calibration curve showed a high degree of consistency between the predicted and real values.Decision curve analysis confirmed that the model's net benefit significantly outperformed the TNM stage. CONCLUSIONThe internally validated prognostic nomogram model demonstrated preliminary predictive capabilities in predicting OS and RFS of patients with LELGC.