| Objective: To study the pathological characteristics of patients with de novo stage Ⅳ breast cancer as a specific group of breast cancer and explore the risk factors affecting the prognosis of patients,and to establish a practical,convenient and intuitive line chart prediction model.The model can be used to quickly calculate the survival probability of these patients in clinical work,so as to evaluate individual conditions and provide patients with accurate and individualized treatment.Methods: Patients’ information and clinicopathological features were extracted from the SEER database in the United States.7433 female patients with stage Ⅳ breast cancer from 2010 to 2015 were randomly divided into training group(n=5,205)and validation group(n=2,228).The training group established the model,and the factors related to the prognosis were obtained by the first round of screening with LASSO regression,and then the survival prediction model was established by combining the factors obtained by Cox univariate analysis and multivariate analysis with the results of LASSO regression.Then the validation group was used to verify the training group: the consistency between the model-specific survival rate and the actual specific survival rate was measured by the calibration curve reaction,the reliability of the model was measured by C-index,and the sensitivity and specificity of the model were reflected by ROC curve and AUC size.The risk score of each patients with novo stage Ⅳ breast cancer was calculated by the corresponding score of each factor in the nomogram,and the population was divided into low and high risk groups by using X-tile to find the best Cutoff value.Kaplan-Meier method was used to analyze the breast cancer-specific survival of the two groups and the effective of surgery or radiation therapy in the patients with de novo stage Ⅳ breast cancer.Results: Fourteen clinical features of patients with de novo stage Ⅳ breast cancer were included in the nomogram.The slope of the calibration map was close to 45 °,the prognostic C-index of the model was 0.719(95% CI:0.711-0.727),and AUC of the 1-year,3-year and 5-year were 0.785,0.765 and 0.765.Then,we divided into low-risk groups and high-risk groups by the cutoff nomogram score 155.Radiotherapy can bring survival benefits in the total population HR=1.094(1.028-1.168,p=0.006).And in the low-risk group,patients received radiotherapy had long-term prognosis(HR=1.102,p=0.026).While radiotherapy in high-risk groups can not improve the prognosis of patients(p=0.31).Patients can benefit from surgery,and the median survival time of patients with low risk can be extended by 22.8 months.The median survival time of operation group and nonoperation group in high risk group was 18.9 months and 13.3 months respectively(p <0.001).In the subgroup analysis,the patients who received surgery and radiotherapy had the best prognosis,while those who received radiotherapy only had the worst prognosis.Conclusion: Age,marital status,histopathological grade,histopathological type,breast subtype,chemotherapy,tumor staging and distant metastasis are independent risk factors for survival in patients with de novo stage Ⅳ breast cancer.The prognosis of the elderly,triple negative breast cancer,high histological grade and brain metastasis have poor survival prognosis.Patients with de novo stage Ⅳ breast cancer from Low-risk group should do surgery and radiotherapy.Surgical treatment of the primary site and radiotherapy can bring a good prognosis.The nomogram of de novo stage Ⅳ breast cancer established in this study under large samples and multiple factors has a certain accuracy,and can objectively evaluate the survival and prognosis of patients with de novo stage Ⅳ breast cancer,which provides operable and more convenient individual treatment information... |
PDF Full Text Request