Abstract:Objective To investigate the factors affecting lymph node metastasis in elderly patients with early gastric cancer (EGC), and construct a risk prediction model based on the nomogram. Methods Sixty-two elderly patients diagnosed with EGC in Cangzhou Central Hospital from January 2019 to December 2023 with complete clinicopathological data and follow-up data were selected as research objects, and lymph node metastasis was observed. According to the pathological results,the patients were divided into lymph node metastasis group (LNM group) and non-lymph node metastasis group (NLNM group). Logistic regression analysis was used to screen the independent risk factors affecting lymph node metastasis in elderly EGC patients, and the nomogram model was established and verified with R software. The receiver operating characteristic (ROC) curve was used to evaluate the differentiation of the model, the clinical decision curve and clinical impact curve were used to evaluate the effectiveness of the model. Results Among 62 elderly EGC patients, 13 (20.97%) had lymph node metastasis. Univariate analysis showed significant differences in vascular invasion, degree of differentiation, depth of infiltration and tumor diameter between LNM group and NLNM group ( P ＜0.05). Multivariate logistic regression analysis showed that vascular invasion ( OR =10.50, 95% CI: 1.56-70.64), undifferentiated type ( OR =5.75, 95% CI : 1.07-30.99), infiltration into submucosai> OR =7.37, 95% CI : 1.29-42.15), and diameter≥2 cm ( OR =7.83, 95% CI : 1.27-48.17) were independent risk factors for lymph node metastasis in elderly EGC patients ( P ＜0.05). Based on the independent risk factors, a nomogram model was constructed, the calibration curve showed that the predicted probability was in good agreement with the actual measured probability ( χ2 =7.527,P =0.376), which had high predictive performance. The area under ROC curve was 0.879 (95% CI : 0.782-0.976). When the threshold probability was between 0.01 and 0.96, the nomogram prediction model had good clinical application value. The clinical impact curve showed that when the threshold was greater than 0.4, the prediction model and the actual occurrence were highly matched. Conclusion Vascular invasion, undifferentiation, infiltration into submucosa and diameter≥2 cm are independent risk factors for lymph node metastasis in elderly EGC patients. The nomogram prediction model can improve the diagnostic efficacy and has high clinical application value.