| Objective: To explore the relevant factors affecting the prognosis of patients with glioma,and to construct nomogram models of Progression-free Survival(PFS)and Overall Survival(OS)of patients with glioma based on these factors,so as to provide reference for clinical more accurate prediction of survival probability of glioma patients.Methods: To collect clinical data(gender,age,maximum tumor diameter,tumor location,ECOG score,tumor resection degree,WHO)of glioma patients treated in neurosurgery at the Affiliated Tumor Hospital of Guangxi Medical University from January 2013 to December 2019 Grading,Ki67 expression,IDH1 mutation and radiotherapy and chemotherapy)and follow-up data.Use univariate and multivariate COX regression analysis to screen out independent risk factors that affect prognosis,construct postoperative PFS and OS nomogram prediction models,and use the area under the receiver operating characteristic curve(ROC)(AUC)and Hosmer-Lemeshow(H-L)Test to evaluate the prediction effect of the model.Results: The median survival time of PFS after operation was 33 months in142 patients with glioma.The median survival time of OS is 45 months.Univariate and multivariate analysis showed that age(HR=1.030,95%CI:1.011-1.050),tumor resection degree(HR=0.295,95%CI:0.163-0.531)and IDH1 mutation(HR=0.296,95%CI:0.173-0.505)were independent prognostic factors for postoperative PFS of glioma patients(P<0.05).Age(HR=1.026,95%CI:1.002-1.050),tumor resection degree(HR=0.506,95%CI:0.275-0.931),WHO classification(HR=2.939,95%CI:1.301-6.643),IDH1 mutation(HR=0.244,95%CI:0.129-0.464)are independent prognostic factors of OS in glioma patients after surgery(p < 0.05).Based on this,a prediction model for postoperative PFS and OS nomogram of glioma patients is constructed.Through verification,AUC values of PFS models at 1 year,2 year and 3 year after operation are 0.860(95% CI: 0.787-0.933),0.810(95% CI: 0.721-0.900),0.817(95% CI: 0.721-0.914),respectively.AUC values of OS models were 0.804(95% CI: 0.694-0.915),0.904(95% CI: 0.843-0.966)and 0.859(95% CI:0.765-0.952)at 1,2 and 3 years after operation,respectively,with good discrimination.The H-L test of the model showed that the consistency index(C-index)in PFS nomogram model was 0.698(95% CI: 0.622-0.754),0.723(95% CI: 0.675-0.798)and 0.774(95% CI: 0.713-0.835)respectively one year,two years and three years after operation.The C-index in OS nomogram model at 1,2 and 3 years after operation is 0.602(95% CI: 0.560-0.688),0.702(95%CI: 0.622-0.782)and 0.796(95% CI: 0.732-0.860),respectively.The calibration curve in calibration graph is in good agreement with the predicted calibration curve,and the model has good calibration degree.Patients were divided into high risk group,medium risk group and low risk group by nomogram.postoperative PFS and OS in high risk group were lower than those in medium risk group,while postoperative PFS and OS in medium risk group were lower than those in low risk group(P<0.001).Conclusion: Age,tumor resection degree,WHO grade and IDH1 mutation are the prognostic factors of glioma patients.The nomogram model constructed by age,tumor resection degree,WHO classification and IDH1 mutation can effectively predict the survival probability of glioma patients and has good clinical application value. |
PDF Full Text Request