Objectives:Ovarian endometrioid cancer(OEC)is rare in clinical practice.Most patients are diagnosed in the early stage and the prognosis is good.At present,the prognosis of OEC patients is mainly determined according to the International Federation of Gynecology and Obstetrics(FIGO)stage.However,existing studies have shown that age,residual tumor size after surgery,peritoneal dissemination,related gene mutations,surgery,and chemotherapy also significantly affect its prognosis.The aim of this study is to explore the prognostic factors of cancer-specific survival(CSS)in OEC patients,establish and validate a nomogram prediction model for OEC patients,and establish a risk stratification system to distinguish low-,intermediate-,and high-risk groups.Methods:We retrieved data of ovarian cancer patients from The Surveillance,Epidemiology and End Results(SEER)database between 2000 and 2018.The incidence,proportion and CSS survival curve of different histological subtypes of ovarian cancer were analyzed.The OEC patients were randomly divided into the model group and the validation group at a ratio of 7:3.Univariate and multivariate Cox regression analyses were used to determine the independent prognostic factors of OEC in the model group,which were used to create a nomogram prediction model of OEC patients for 3-,5-and 10-year CSS.Then the Cindex,and area under ROC curve(AUC),net reclassification index(NRI),and integrated discrimination improvement(IDI)were used to evaluate the predictive ability of the nomogram prediction model.Calibration curves were used to evaluate the accuracy of the nomogram prediction model.Decision curve analysis(DCA)was used to compare the clinical utility of the nomogram prediction model and the FIGO stage model.Based on the nomogram prediction model,a webpage prediction applet was created for clinical use.Finally,a risk stratification system was established to distinguish low-,intermediate-and high-risk groups.Results:1.The age-adjusted incidence of both ovarian cancer and OEC decreased between2000 and 2018.OEC accounted for 12.22% of ovarian cancer.The prognosis of different histological subtypes of ovarian cancer was significantly different(P < 0.001),and the prognosis of OEC was better,with a 5-year CSS rate of 81.3%.2.Multivariate analysis showed that the independent predictors of CSS in OEC patients included age,marital status,tumor laterality,grade,FIGO stage,tumor size,initial CA125 and surgical procedure(P < 0.05).3.The C-index of the nomogram prediction model was 0.819 in the model group and0.812 in the validation group.The AUC values of the 3-and 5-year prediction models(model group: 0.840,0.841 and 0.833;validation group: 0.803,0.804 and 0.795)were both greater than 0.80,and the NRI(model group: 0.267,0.287 and 0.313;validation group:0.250,0.314 and 0.355)and IDI(model group: 0.052,0.063 and 0.075;validation group:0.061,0.081 and 0.087)were both greater than 0,which indicated that the nomogram prediction model has better predictive ability compared with the classical FIGO stage system.The calibration curve showed that the predicted values were highly consistent with the actual observed values,indicating that the nomogram prediction model had good accuracy.The DCA curve also showed that the nomogram prediction model had better clinical utility than the FIGO stage system.4.The risk stratification system based on the nomogram prediction model could effectively distinguish patients in the low-,intermediate-and high-risk groups(P<0.001).Conclusions:1.Independent prognostic factors for CSS in OEC patients included age,marital status,tumor laterality,grade,FIGO stage,tumor size,initial CA125,and surgical procedure.2.The nomogram prediction model can accurately predict the survival and prognosis of OEC patients individually.Compared with the FIGO staging system,the nomogram prediction model has better predictive ability and clinical practicability.3.The risk stratification system can effectively distinguish low-,intermediate-and high-risk groups,and provide reference for clinical treatment. |