管腔型乳腺癌伴肺转移女性患者预后列线图模型的建立与验证OA

Objective Based on the surveillance,epidemiology,and end results(SEER)database,this study aimed to con-struct a nomogram model to evaluate the prognosis of patients with luminal breast cancer with lung metastasis and to provide a basis for individualized treatment.Methods In this study,2 607 patients with ductal breast cancer with lung metastases were screened from the SEER database and randomly assigned to the training group(n=1 824)and validation group(n=783).Variables with significant effects on patients'overall survival were screened through the use of univariate and multivariate COX risk-proportional modeling.Based on these key variables,nomogram models were constructed to predict patient survival at 1,3,and 5 years.To ensure the accu-racy and reliability of the nomogram,the efficacy of the models was evaluated by subject work characteristics(ROC)curves,calibra-tion plots,and clinical decision curves.Results Eleven variables,including age,chemotherapy,marital status,bone metastasis,liver metastasis,brain metastasis,tumor grade,race,surgery,estrogen receptor,and progesterone receptor,were identified as inde-pendent risk factors for the prognosis of luminal breast cancer with lung metastasis by multifactorial COX regression analysis.In the training group,the AUC values of the area under the ROC curve at 1,3,and 5 years were 0.744,0.740,and 0.711,respectively,while in the validation group,the corresponding AUC values were 0.735,0.703,and 0.718,respectively.The calibration plots further validated that the agreement between the overall survival predicted by the nomogram and the actual observation was good,and the deci-sion curve results indicated that the nomogram model has a high net benefit.Conclusion Based on the variables that have a significant effect on patients,a nomogram for predicting 1,3,and 5 years'survival rates of patients with ductal breast cancer with lung metastases has been successfully constructed,and it has been verified that its predictive performance is good,accurate,and reliable.It provides an auxiliary decision-making tool for clinical prediction of prognosis and optimizes patient management and treatment.