Protein expressions of multi-gene for predicting the prognosis of early breast cancer with COX model analysis / 中国癌症杂志

ABSTRACT

Background and purpose:The prognosis in patients with early breast cancer(EBC) was poorly defined by clinical and histopathological features.There were more than 200 genes that might be correlated with development,progression,recurrence and prognosis of breast cancer in the published literature.This study evaluated the prognosis of patients with EBC at molecule level according to the protein expressions of 16 selecting genes(HER2,ER,PR,BCL2,Ki-67,BMYB,Cyclin B1,STK15,MMP11,BAG1,NM23,PTEN,P53,P27,VEGF,PCN A).Methods:The immunohistochemical method was used to detect the protein expressions of these genes in 76 patients with EBC and the statistical analysis was performed by COX proportional hazards model.Results:COX proportional hazard model revealed that BCL2,VEGF and STK15 were independent prognostic factors for patients with EBC.A formula of PI was set up according the three genes.The predictive outcomes with PI were compared with the actual follow-up outcomes in 48 patients with EBC.As a result,the predictive accuracy of good and bad prognosis was 86.67% and 91.67%,respectively.Conclusion:A formula of PI derived from protein expressions of the 16 genes correlated with breast cancer was ultimately set up and might be used to predict the outcomes of the patients with EBC.