The Risk Factors Analysis And Risk Prediction Model Establishment Of Lymph Node Metastasis In Early Gastric Cancer

Research purposeBy analyzing the related risk factors of lymph node metastasis in early gastric cancer,a risk prediction model for lymph node metastasis in early gastric cancer is established to provide a basis for the treatment decision of patients with early gastric cancer.Research methodsPostoperative pathological reports of patients with early gastric cancer who underwent surgical treatment in gastrointestinal surgery department of the first affiliated hospital of fujian medical university from January 2013 to July 2017 are collected.The correlation between the clinicopathological factors and lymph node metastasis is analyzed,including gender,age,tumor size,degree of invasion,degree of differentiation,and vascular tumor-associated thrombosis,so as to obtain the risk factors of lymph node metastasis and establish a risk prediction model to guide clinical diagnosis and treatment.After the establishment of the model,cases of early gastric cancer from August 2017 to now are collected to verify the accuracy and reliability of the model externally.Research results1.In the early stage of this study,a total of 232 patients are included for model construction,including 42 patients with lymph node invasion,190 patients without lymph node invasion,and the positive rate of lymph node is 18.1%.2.Univariate analysis showed that >2cm tumor,infiltration into the submucosa,poor differentiation,invasion into the vessels and accompanied by ulcers is significantly correlated with the invasion of lymph nodes.Multivariate analysis showed that >2cm tumor,infiltration into the submucosa,poor differentiation,and invasion into the vessels are independent risk factors for lymph node metastasis.3.Early gastric cancer lymph node metastasis prediction model is established,P = e ^ x/(1 + e ^ x),x =-5.804 + 1.069×tumor size + 2.939×degree of invasion + 2.114×degree of differentiation + 1.386×vascular tumor emboli.To calculate the probability of lymph node metastasis.The ROC curve was used to verify the diagnostic value of the risk prediction model.The area under the ROC curve was 0.892,and the asymptotic 95% confidence interval was 0.848-0.936.4.The optimal threshold of the predicted value of the model is 0.164.The four-grid table is constructed by predicting the model results and the postoperative pathological examination gold standard.The sensitivity and specificity of the model are calculated to be 80.95% and 81.85%,respectively.5.The accuracy and reliability of the model are verified externally,the area under the ROC curve is 0.889,and the asymptotic 95% confidence interval is 0.810-0.969.The sensitivity and specificity are calculated to be 100% and 73.08%,respectively.Research conclusion1.The risk of lymph node metastasis is higher in patients with early gastric cancer with > 2cm tumor,infiltrating into submucosa,undifferentiated type,and accompanied by vascular carcinoma thrombus.For those who meet the above conditions,lymph node metastasis should be highly suspected,with bias in the treatment decision.2.The lymph node metastasis prediction model established in this paper can be used to predict the risk of lymph node metastasis in patients with early gastric cancer.External verification shows that it has good accuracy and credibility.3.Preoperatively,the model can be used to calculate the probability of lymph node metastasis in early gastric cancer and assess the risk of lymph node metastasis,and urgical treatment is recommended for high risk patients.If the endoscopic treatment is selected,the risk of lymph node metastasis can be assessed after treatment,and the patients with high risk may be considered for additional comprehensive treatment based on surgical treatment.