Predictive Model For Lymph Node Metastasis In Endometrial Cancer

ObjectiveBy analysing the clinical data of patients with endometrial malignancies initially treated at our hospital,independent influencing factors were screened out.The nomograms constructed by R and SPSS was applied to predict the risk of possible eventual lymph node metastasis,providing a meaningful guidance for current clinical and treatment.MethodClinical data were collected from 202 patients who attended our hospital for surgical treatment of "endometrial malignancy" from January 2018 to August 2022.The clinical data collected were analysed retrospectively and a univariate analysis of variables was performed using SPSS software to screen for variables associated with lymph node metastasis in endometrial cancer among various variables.Subsequently,a multifactorial analysis was performed using SPSS software and based on this,a nomograms for predicting lymph node metastasis was developed using R statistical software.The discrimination of the line plot model was assessed using the Receiver Operating Characteristic Curve(ROC)and the Decision Curve Analysis(DCA),and the calibration of the model was assessed using the calibration curve(calibrate).ResultA total of 202 patients were included,including 35 patients with lymph node metastasis(17.3%).A total of 24 different clinical variables have been collected for each patient with endometrial cancer included in the study,and 7 independent risk factors associated with lymph node metastasis from endometrial cancer have been finally screened.These were: preoperative CA125,preoperative HE4,histological type,FIGO stage,depth of myometrial infiltration,bilateral adnexal involvement,and choroidal infiltration as independent risk factors for lymph node metastasis.The above variables were subjected to logistic binary analysis,and subsequently the nomograms of lymph node metastasis in endometrial cancer was constructed by applying R software based on the results obtained from the multifactorial analysis.The nomograms shows that CA125,histological type,FIGO stage,myofilament infiltration,bilateral adnexal involvement,and vascular infiltration are the risk factors;HE4 is the protective factor.CA125,HE4,myofilament infiltration and vascular infiltration were less deterministic for lymph node metastasis;histological type and FIGO stage were more deterministic for the risk of lymph node metastasis.The model was then validated by ROC,DCA and Calibrate curves,which finally showed that the prediction model had good discriminative power and accuracy.ConclusionThe nomograms of lymph node metastasis in endometrial cancer established by applying preoperative CA125,preoperative HE4,histological type,FIGO stage,depth of myometrial infiltration,bilateral adnexal involvement,and choroidal infiltration as variables can be used to predict the risk of lymph node metastasis in endometrial malignancies.To a certain extent,it can assist in the development of treatment plans in the clinic...