Multivariate Analysis Of Prognosis In Patients With Upper Urinary Tract Urothelial Carcinoma And Construction And Validation Of Survival Prediction Model

Objective: To explore the prognostic factors of primary upper urinary tract urothelial carcinoma(UTUC)based on the SEER database,and establish a Nomogram survival prediction model based on independent predictors,so that clinicians can better evaluate the prognosis of UTUC patients,and provide reference for individualized clinical treatment.Methods: clinical data of patients diagnosed with UTUC and undergoing surgical treatment between 2004 and 2015 were downloaded from SEER databases,and the final number of patients included was determined by inclusion and exclusion criteria.All included variables(such as age,sex,tumor size,tumor stage,tumor grade,etc.)were coded and descriptive statistical analysis was performed.Death from any cause was the end point of the study.Kaplan-Meier method was further used to draw survival curves,and Log-Rank method was used to test and compare the difference in survival rate between groups.The test level a=0.05 was set to screen out variables affecting the prognosis of upper urinary tract uroepithelial carcinoma,and significant variables were included in Cox multivariate analysis to determine independent risk factors affecting the prognosis.Based on these independent risk factors,the survival nomogram model of UTUC was drawn.The nomogram was then validated using data from Jiangxi Provincial People’s Hospital(patients diagnosed with upper urinary tract urothelial carcinoma between January 2011 and December 2022).Consistency index and area under ROC curve were calculated to evaluate the model differentiation,and the accuracy of the model was tested by internal verification.Results: after screening conditions based on the SEER database,clinical data of8636 patients with upper urinary tract urothelial carcinoma were finally included,including 5127 male patients(59.3%)and 3918 patients(45.4%)older than 75 years.Survival curves were plotted by K-M survival analysis method,and differences between groups were compared by Log-rank test.The results showed that age,tumor size,tumor T stage,tumor N stage,tumor M stage,pathological grade and lymph node dissection were related to the prognosis of upper urinary tract urothelial carcinoma patients.Further Cox multivariate regression analysis showed that the above 7 factors were independent factors influencing the prognosis of UTUC.Based on these factors,a line graph model was constructed,and the consistency index after internal correction was 0.724,the areas under the ROC curve of 3-year and 5-year survival rates were0.741 and 0.739,respectively;the consistency index of external verification set was0.93,and the areas under the ROC curve of 3-year and 5-year survival rates were 0.871 and 0.96,respectively,indicating good model differentiation.The calibration curve is close to the standard curve,and the model conforms well.Conclusions: age,tumor size,tumor T stage,tumor N stage,tumor M stage,pathological grade,and lymph node dissection were used to independently predict the prognosis of upper urinary tract urothelial carcinoma patients.The prognostic survival column map model constructed according to these 7 factors has good differentiation and conformity,which can provide certain reference value for clinicians to assess the survival prognosis of tumor patients.