Prognostic Factors Of Primary Stage ? Breast Cancer And The Establishment Of Prognosis Model

Objective:Retrospective analysis of the clinicopathological characteristics and treatment related information of initial stage ? breast cancer patients,screening out the independent prognostic factors that affect the survival of the patients,establishing an intuitive,simple and convenient nomogram prognosis model,to provide individual treatment strategies for the initial stage ? breast cancer patientsMethod:From January 1,2003 to January 1,2019,216 patients with stage ? breast cancer were initially diagnosed in the oncology department of the First Affiliated Hospital of China Medical University.According to the entry criteria,204 patients with initial stage ? breast cancer were screened out.The clinical and pathological data of 204 patients were collected retrospectively,including age at diagnosis,menstrual state,body mass index(BMI),ECOG physical condition score,pathological type,ER,PR,HER2,molecular typing,distant metastasis and treatmentKaplan Meier method was used to determine the relationship between clinicopathological parameters and overall survival(OS).Based on the results of single factor and multi factor analysis of Cox risk proportional regression model,a nomogram is constructed and verified internally,and the accuracy and reliability of the prediction model are measured by the consistency index(c-index).Through the calibration model of 1-year and 3-year and 5-year nomogram,the predicted survival probability value and the actual survival probability value are compared and analyzed,so as to evaluate the prediction accuracy of the nomogram model.According to P<0.05,it is a significant prognostic factor.The time-dependent receiver operating characteristic(ROC)curve was used to evaluate the effectiveness of the prognosis model.The decision curve analysis(DCA)calculates the net benefit of the prognosis of the initial stage ? breast cancer patients under the influence of different clinical histopathological characteristics under Different Pt values,so as to display the clinical application value of the prognosis model of the nomogram more intuitively and accurately.According to the total score shown in the nomogram,take the median as the dividing point,divide the high-risk group and the low-risk group,and use Kaplan Meier method to analyze the OS survival of the two groupsResults:As of August 31,2019,the median follow-up time was 34 months(range of follow-up time was 4-171 months).147 cases(72.1%)of 204 patients had OS events,81.4%,51.5%and 41.2%of them had OS events in one year,three years and five years,respectively.First of all,Kaplan Meier method was used to analyze the single factor influencing the prognosis of patients.The results showed that ECOG score(P 0.002),ER status(P=0.023),molecular typing(P=0.0217),bone and/or soft tissue metastasis(P=0.013),visceral metastasis(P=0.00065),the number of visceral metastasis(P=0.0021)and whether the operation was performed(P=0.0002)were related to OS of patients with initial stage ? breast cancer(P<0.05).The results showed whether the patients received surgery(P=0.001),molecular typing(P 0.015),whether there was visceral metastasis(P=0.001),the number of visceral metastasis(P=0.002),visceral metastasis(P=0.024),ECOG score(P=0.038),histological grade(P=0.01),ER status(P=0.002)and OS was related to the initial stage ? breast cancer patients(P<0.05),which was the prognostic factor affecting the survival of patients.The results of multivariate analysis showed that whether local breast surgery(P=0.001),molecular typing(P=0.009)and visceral metastasis(P 0.019)were independent prognostic factors for OS.The consistency index of the internal validation of the nomogram is 0.671(95%CI 0.647-0.695).The calibration curve shows that the study in this group has good consistency in one year,three years and five years.The AUC value of the time-dependent ROC curve is[0.7,0.8],indicating that the prediction model has high reliability.In the decision curve analysis(DCA),we can see the OS of the initial stage ? breast cancer patients.The net benefit of using this nomogram to make decision is higher,which shows that the model has better clinical application value.According to the total score of the nomogram,set the total score>92.5 as the high-risk group,and set the total score<92.5 as the low-risk group.Analyze the OS survival of the two groups,P=0.0065,which proves that the model has high accuracy and applicationConclusion:The results of univariate analysis in this single center retrospective study showed that:ECOG score,ER status,molecular typing,bone and/or soft tissue metastasis,visceral metastasis,number of visceral metastasis and operation were factors affecting the prognosis of patients with initial stage ? breast cancer According to the results of multivariate analysis,the nomogram model for the prognosis of patients with initial stage ? breast cancer,which is based on the factors such as whether the patients receive breast local focus surgery,molecular typing and visceral metastasis,has better accuracy and practicability,and will be helpful for the selection of individual treatment strategies of patients.