| Objective:Clinical Data of Early Stage Gastric Carcinoma who underwent radical resection of gastric cancer were analyzed retrospectively,the risk factors affecting lymph node metastasis in patients with early gastric cancer were studied and summarized,and the prediction model of lymph node metastasis in early gastric cancer was established.in order to provide patients with accurate treatment.Method: The clinical data of 126 patients with early gastric cancer in The Second Affiliated Hospital of Kunming Medical University from January 2013 to December 2022 were retrospectively analyzed.All patients received standard radical gastrectomy(D2)and recovered well.First of all,the relationship between LNM and sex,age,drinking history,family history and body mass index(Body Mass Index)of patients with early gastric cancer.BMI),smoking history,Helicobacter pylori infection history,focus location,depth of invasion detected,number of lymph nodes,nerve invasion,tumor size,degree of differentiation,lymphatic and vascular invasion,prealbumin content,hemoglobin content,NRS2002 score,albumin content,neutrophil-lymphocyte ratio(Neutrophil To Lymphocyte Ratio,NLR),platelet-lymphocyte ratio(Platelet-Lymphocyte Ratio).Univariate analysis was used to analyze the relationship between PLR)and other factors.Then the indicators with significant correlation are analyzed by binary Logistics multi-factor regression analysis,and the prediction model is established according to the results of multi-factor analysis.The risk factor score in the model is determined by the weight of β coefficient obtained by Logistic multi-factor risk regression.The 126 patients collected were scored according to the predictive model to get the total score,and then the risk of lymph node metastasis was predicted according to the total score,taking the lymph node metastasis in the postoperative pathological report as the gold standard.The accuracy of the prediction model was tested by the area under the characteristic curve.Finally,the external verification of the model was carried out,and the model was used to evaluate 94 patients with early gastric cancer.According to the clinical data of the patients,the ROC curve was drawn,and the postoperative pathological examination was taken as the gold standard to verify the correctness and credibility of the model.Through this model,patients with early gastric cancer were divided into high-risk group and low-risk group,and the sensitivity and specificity of each group were calculated according to the pathological results as the gold standard.Using bilateral statistical test,P < 0.05 was considered to be statistically significant.Results: We analyzed the clinicopathological data of 126 patients with early gastric cancer,including 24 patients with lymph node metastasis and 112 patients without lymph node metastasis.the overall rate of lymph node metastasis was 19.05%.Univariate analysis showed that tumor differentiation,nerve infiltration,tumor size,depth of invasion,lymphatic and vascular infiltration were significantly correlated with lymph node metastasis(P < 0.002,< 0.001,0.011,0.001,< 0.001).Compared with the control group,there was no significant difference in onset time,sex,family history,smoking,BMI,HP infection history,metabolic disease history,focus location,ALB content,PA content,RBP content,HP content,NLR,PLR and NRS2002 score between the patients with lymph node metastasis and the control group.Binary Logistic multivariate regression analysis showed that tumor diameter ≥ 2cm,poor differentiation,tumor invasion to submucosa,tumor with lymphatic and vascular infiltration were independent risk factors for LNM in EGC(P = 0.028,0.006,0.001,<0.001).According to the four independent risk factors,a predictive model with a full score of 100 was established,and each patient with early gastric cancer was scored according to the model.The diagnostic value of the model is verified by the ROC curve,and the area under the curve(Areaunderthecurve,AUC)of ROC is 0.914,and the asymptotic 95% confidence interval(Confidence Interval,CI)is 0.855-0.973.The optimal threshold of the model is 29.50.Patients with early cancer below this threshold are divided into low risk group of lymph node metastasis,and those higher than this threshold are divided into high risk group of lymph node metastasis.A cross table is established with the results of the prediction model and the gold standard of postoperative pathological examination.the sensitivity of the model is 75.00% and the specificity is 93.87%.The accuracy and credibility of the model are verified externally.the lower product of the ROC curve is 0.851,the asymptotic 95%confidence interval is 0.737-0.964,the sensitivity is 90%,and the specificity is75.34%.Conclusion: Through this study,we found that tumor invasion to submucosa,tumor diameter ≥ 2cm,poor differentiation,lymphatic and vascular invasion were independent risk factors for LNM in EGC.In this paper,a model for predicting lymph node metastasis of early gastric cancer was established based on the above four independent risk factors,and the LNM status of EGC was predicted by LNM high-risk group and LNM low-risk group.After external verification,this method has high accuracy and credibility,and the model can be used to evaluate the risk of lymph node metastasis before operation,and standard radical gastrectomy plus regional lymph node dissection(D2)is recommended for LNM high-risk group.In the low-risk group of LNM,endoscopic resection can be performed first and fully communicated with the surgery,and the intraoperative condition and postoperative pathological specimens can be evaluated again to evaluate the risk of LNM.If the re-evaluation shows that the LNM high-risk group,the standard radical gastrectomy(D2)should be added. |
PDF Full Text Request