Pathological Nipple Discharge In Patients With Breast Duct Tumor Diagnostic Nomogram Prediction Model Building And Validation

Objective:To construct and validate a nomogram for predicting tumor and cancer in breast ducts of nipple discharge.Methods:The clinical data of 3293 patients with papillary discharge as the first symptom and receiving endoscopy from the Fourth Affiliated Hospital of China Medical University from October 2007 to October 2022 were retrospectively analyzed.Statistical analysis was performed on the age of onset,amount of fluid discharge(small amount and large amount),nature of fluid discharge(bloody and non-bloody),volume of fluid discharge pore(single and porous),side of fluid discharge(unilateral and bilateral),microscopic manifestations of milk duct wall(smooth and uneven),and intraductal tumor(regular and irregular shape).Logistic regression was used to analyze the correlation between clinical features of patients and intraductal breast tumors,and the prediction model of line graph was constructed and verified.Receiver operating characteristic curve(ROC)analysis was used to evaluate the sensitivity and specificity of the model’s predictive power.Calibration curves were used to evaluate the consistency of the Logistic regression model and decision curve analysis(DCA)was used to evaluate the diagnostic accuracy.Result:1.To construct a nomogram for predicting intraductal tumors in patients with pathological nipple discharge:From October 2007 to October 2022,an analysis of 3293 cases undergoing mammoscopy suggested that patients aged ≥ 44 years(mean age),unilateral,single-pore,multivolume,and bloody patients were more likely to find intraductal tumors through mammoscopy.A model was constructed based on the above five variables and randomized in a ratio of 7:3.The model was subsequently validated in989 patients and complete data for variables in the validation set was obtained.The area under ROC curve(AUC)of the training set was 0.733,and the area under ROC curve(AUC)of the validation set was 0.702.The analysis results proved that this model had a good ability to predict the occurrence of breast ductal tumors in patients with pathological nipple discharge.2.To construct a nomogram for predicting intraductal tumor carcinogenesis in nipple discharge:A total of 596 patients who underwent surgery from October 2017 to October 2022 were selected for analysis.The results indicated that patients aged ≥ 46 years(mean age),with single orifice,multiple dose,blood,ultrasound and molybdenum target(BI-RADS ≥4b)were more likely to be diagnosed as malignant by surgical pathology.A model was established based on the above six variables and randomized at a ratio of 7:3.The model was subsequently validated in 178 patients with complete data for the variables in the validation set.The area under ROC curve(AUC)of the training set was 0.805,and the area under ROC curve(AUC)of the validation set was 0.766.The analysis results proved that the model had a good ability to predict the carcinosis of breast ductal tumors in patients with pathological nipple discharge.Conclusion: In this study,two kinds of line graph models were constructed,which can effectively predict the occurrence and canceration of breast ductal tumors in patients with pathological nipple discharge,which has good clinical application value and is worth popularizing.