| Objective:By analyzing the high-risk factors of brain metastasis in the patients with HER-2（human epidermal growth factor receptor 2）positive breast cancer and establishing a risk model,we hope to provide active early monitoring and more effective intervention measures for such patients,to provide the best treatment model for the patients and prolong their survival.Methods:Through retrospective analysis of the patients with HER-2 positive breast cancer patients treated in the Department of Radiation Oncology of First Hospital,Jilin University from 2017 to 2020,we compared the correlation between the clinical and pathological information of the patients with brain metastasis,such as age,menopause,mode of operation,stage,vascular invasion,nerve infiltration,hormone receptor status,Ki-67 and regular use of anti-HER-2 therapy.Optimal scaling（CATREG）regression analysis was used to identify the risk factors of brain metastasis in the patients with HER-2 positive breast cancer and a risk prediction model was established according to the risk factors.Results:In this study,a total of 1327 patients who received treatment in the Department of Radiation Oncology of First Hospital,Jilin University from April 2017 to June 2020were selected according to the inclusion and exclusion criteria,and finally included 317patients with HER-2 positive breast cancer.The earliest time for diagnosis was November 2006 and the latest was September 2019.The deadline for follow-up was July 1,2021,with a median follow-up time of 41 months.A total of 48（15.14%）patients developed disease progression（regional lymph node recurrence or distant metastasis）,of which 30 developed brain metastases,and the central nervous system（CNS）was the first site of relapse in 9 of the 30 patients with brain metastases.Through the survival analysis of the patients with and without brain metastasis,it was found that the median overall survival（m OS）of the patients with brain metastasis was significantly shorter than the m OS of patients without that（67 months vs 130 months,P=0.034）.The optimal scaling regression analysis was used to analyze the factors affecting brain metastasis.It was found that the mode of operation,pathological type,lymph node staging,ER,Ki-67,and regular anti-HER-2 therapy were the independent risk factors of brain metastasis.Further analysis revealed that patients who underwent total mastectomy had a high probability of brain metastases,and patients with the pathological type of invasive lobular carcinoma were most likely to develop brain metastases.However,high N stage,high Ki-67 expression,low ER expression,and irregular receipt of anti-HER-2 therapy are closely related to brain metastasis development.Regression models were developed with the above variables associated with the occurrence of brain metastases.R Square（R~2）=0.165,Adjusted R~2=0.127.The ROC curve was drawn by Med Calc software.According to the Youden Index J,the best Cut-off value was 0.5343,the sensitivity was 66.67%,the specificity was 86.76%,the area under the curve（AUC）was 0.820（95%CI:0.773-0.860,P<0.001）,and the prediction ability of the model was good.Conclusion:Patients with HER-2 positive breast cancer had significantly shorter survival after developing brain metastases.The results of the multivariate analysis suggested that surgical approach,pathological type,N stage,ER,Ki-67,and whether they received regular anti-HER-2 therapy were independent risk factors for brain metastases in patients.Based on this,a risk prediction model for brain metastases was established,with a sensitivity of 66.67%,a specificity of 86.76%,and an area under the curve of0.820（P<0.001）.The assessment of this model has good predictive ability and important clinical significance,but it still needs to be supported by larger multicenter clinical data for clinical application. |
PDF Full Text Request