| Objective:Estrogen receptor-positive(ER-)progesterone receptor(PR+)positive breast cancer is an uncommon and controversial subtype.The aim of this study was to identify prognostic factors affecting non-metastatic estrogen receptor-positive(ER-)progesterone receptor(PR+)positive breast cancer and to develop and validate a clinical prediction model to predict 3-and 5-year overall survival(OS)and(BCSS)for non-metastatic ER-/ PR+ breast cancer to help guide the treatment of non-metas tatic ER-/ PR+ breast cancer.Methods:This study screened 1374 eligible patients with early ER-/ PR+ breast cancer from the Surveillance,Epidemiology,and End Results(SEER)database through strict inclusion and exclusion criteria.The study was internally validated,and the included patients were divided into a training cohort population(N=964)and a validation cohort population(N=410)using a random resampling method.The prognostic impact of multiple clinicopathological variables on survival in the training cohort was assessed by applying the Kaplan-Meier method and constructing a Cox proportional risk model.Significant prognostic factors were combined to construct column line plots,and the predictive power of the column line plots was examined by internal validation using the concordance index(C-index),calibration plots,and area under the ROC curve.Results:1.72.9% of the total population of patients with non-metastatic ER-/ PR+ breast cancer were Caucasian,84.2% of patients with histological grade III/IV,66.6% of N0 patients,95.1% of patients who underwent surgery,and 74.0% of patients who received chemotherapy.The mean age of the total patient population was 54.97 years(20-99 years)and the median survival months was 38 months(0-83 months).The median survival time was 40 months(0-83 months)in the modeling group and 35months(0-83 months)in the validation group.2.In Cox proportional risk model multifactorial showed histological grade,T stage,N stage,chemotherapy,radiotherapy and surgery as prognostic influencing factors for OS in ER-/ PR+ breast cancer.Histologic grade,clinicopathologic stage,T stage,N stage,chemotherapy and surgery were prognostic influencing factors for BCSS in ER-/ PR+ breast cancer.3.the C indices of the column line plots created in the training cohort for predicting 3-year and 5-year OS and BCSS were 0.771 and 0.7718,respectively,and in the validation cohort for 3-year and 5-year OS and BCSS were 0.7804 and 0.7877,respectively.the area under the ROC curve for 3-year and 5-year OS in the training cohort in the ER-/ PR+ breast cancer 0S prediction model AUC values were 0.79 and0.783 for the training cohort and 0.814 and 0.757 for the validation cohort for 3-year and 5-year OS,respectively;in the BCSS prediction model the area under the ROC curve AUC values were 0.803 and 0.789 for the training cohort and 0.789 for the validation cohort for 3-year and 5-year BCSS,respectively,and the area under the ROC curve AUC values were 0.814 and 0.757 for the validation cohort for 3-year and5-year BCSS,respectively.The calibration curves for both the OS and BCSS column line plots crossed the 45° dashed line and fit well.The column line plots in this study showed good accuracy in predicting OS and BCSS at 3 and 5 years,which is a good tool for predicting individual prognosis.4.In subgroup analysis it was shown that undergoing surgery,chemotherapy had better OS and BCSS for both ER-/ PR+ breast cancer patients.based on the predicted total score of column plots divided into low risk and high risk groups,Kaplan-Meier survival curves were plotted and differences in OS and BCSS were tested by Log-rank test.The results showed a worse prognosis in the high-risk group.Conclusion:Histologic grade,T-stage,N-stage,chemotherapy,and surgery are prognostic influences on OS and BCSS in patients with ER-/ PR+ breast cancer.Columnar plots predicting 3-and 5-year OS and BCSS in non-metastatic ER-/ PR+ breast cancer have good predictive power and can assist clinicians in making treatment decisions for their patients. |
PDF Full Text Request