Risk Factor Analysis For Lymph Node Metastasis In Early Gastric Cancer Patients With Mucosal Invasion

ObjectiveLymph node status is an important factor affecting the treatment of early gastric cancer patients with mucosal invasion.This study aims to analyze the risk factors related to lymph node metastasis of early gastric cancer patients with mucosal invasion,and build a risk prediction model for lymph node metastasis,so as to provide a reference for the selection of further treatment for early gastric cancer patients with mucosal invasion after endoscopic treatment.MethodsRetrospectively analyze the clinicopathological data in early gastric cancer patients with mucosal invasion treated by general surgery in the First Affiliated Hospital of Nanchang University from January 2008 to September 2019,including gender,age,tumor size,tumor type,vascular infiltration,presence or absence of ulcer formation,histological grade and depth of tumor invasion.Univariate logistic regression analysis was used to preliminarily analyze the correlation between the above clinicopathological parameters and the risk of lymph node metastasis in patients with intramucosal early gastric cancer,and then multivariate logistic regression analysis by using the R programming language were performed to construct the prediction model of lymph node metastasis in early gastric cancer patients with mucosal invasion.The prediction accuracy of the model was calculated and a nomogram was constructed to visualize the risk of lymph node metastasis.ResultAmong the 882 patients of early gastric cancer with mucosal invasion,62 patients had lymph node metastasis.Univariate logistic regression analysis showed that gender(p=0.002),tumor size(p=0.000),histological grade(p=0.000),and vascular invasion(p=0.000)were related to lymph node metastasis of early gastric cancer patients with musosal invasion.Multivariate logistic regression analysis were used to construct the prediction model as follows: logit(P)=-5.886092 + 0.030732 ×tumorsize + 3.025813×vascular invasion + 0.982832×histological grade.the model distinguish degree analysis showed that calculate the area under the ROC curve(AUC)is 0.738.The calibration plot demonstrated an almost perfect agreement between the predicted probability and the observed outcome fitted to the ideal line(mean absolute error is 0.009),which showed that the prediction model constructed in this study predicting the risk ability of lymph node metastasis has a good consistency and it has clinical practical significance.The predictive nomogram model constructed with R language showed that if the total score of early gastric cancer patients with mucosal invasion reached 205 or higher,the probability of lymph node metastasis was ≥80%.Conclusion1、gender,tumor size,histological grade,and vascular invasion were correlated with lymph node metastasis of early gastric cancer patients with musosal invasion.2、This study trained a risk prediction model for the prediction of lymph node metastasis of early gastric cancer patients with musosal invasion,logit(P)=-5.886092 + 0.030732 × tumorsize + 3.025813 × vascular invasion + 0.982832 ×histological grade,which has a good predictive and calibration capabilities and provided a reference for the selection of further treatment for early gastric cancer patients with mucosal invasion after endoscopic treatment.