Construction Of Prediction Model For Postoperative Survival Rate Of Uterus Serous Endometrial Carcinoma Based On SEER Database

Objective:This study analyzed the relevant clinical data of patients with serous endometrial carcinoma(USC)in the"surveillance,epidemiology,and end results"(SEER)database of the National Cancer Institute of the United States,established a prediction model for postoperative survival rate,and evaluated its predictive efficacy.Methods:In this study,765 patients with serous endometrial carcinoma from 2006 to 2015 were selected by inclusion and exclusion criteria according to the original data of SEER database,and the data of included patients were randomly divided into a training set(535cases)and a test set(230 cases)according to a ratio of 7:3,the differences in the inclusion indicators of the two sets was compared.Based on the training set,prognostic factors of serous endometrial carcinoma were selected,and univariate Cox regression was used to select prognostic factors of serous endometrial carcinoma patients.Subsequently,indicators with statistically significant differences in univariate Cox regression(P<0.05)were included in the multivariate Cox regression model.Finally,the factors influencing the independent prediction of prognosis were obtained,and the specific weights of each index were determined to build a column graph,and a prediction model for the overall survival of patients with serous endometrial carcinoma was established,and the model was evaluated.Results:In the 2338 USC patients diagnosed pathologically in the SEER database,765 patients were selected according to the nano-pathology criteria,including535 in the training set and 230 in the test set.In training concentration,age(χ~2=5.165,P=0.023),diagnosis year(χ~2=9.861,P=0.007).Transfer(χ~2=170.004,P<0.001),AJCC stage(χ~2=181.993,P<0.005),radiotherapy method(χ~2=42.516,P<0.001),tumor size(χ~2=20.262,P<0.001),lymph node resection area(χ~2=66.009,P<0.001)is a risk factor affecting the prognosis of patients with USC.The stepwise regression method found that age,diagnosis year,metastasis,AJCC stage,radiotherapy method and lymph node resection area were independent risk factors for the prognosis of USC patients.The independent risk factors were used to construct the patient survival nomograph model.The C index of the training set was 0.758,the area under the 3-and 5-year survival ROC curve was 0.812 and 0.794,respectively.The C index of the test set was 0.763,and the area under the 3-and 5-year survival ROC curve was 0.814 and 0.798,respectively.The results showed that the nomograph model had good consistency in predicting the postoperative survival of patients in the training set and the test set.Conclusions:This study showed that age,diagnosis year,AJCC stage,metastasis,radiotherapy method and lymph node resection area were independent risk factors for the prognosis of patients with USC.This nomogram model has certain reference value for predicting the survival rate of patients after surgery,and needs to be verified in actual clinical practice.