Correlation analysis of androgen receptor expression with breast cancer prognosis
Yu Yan,Maimaiti Yusufu·,Changwen Wang,Jie Tan,Qiuyang Zhao,Tao Huang
DOI: https://doi.org/10.3877/cma.j.issn.1674-0807.2016.03.005
2016-01-01
Abstract:Objective To investigate the relationship of androgen receptor ( AR ) expression with clinicopathological characteristics and the prognosis of breast cancer .Methods Using tissue microarray ( purchased from Shanghai Xinchao Biotechnological Co ., Ltd., No.K14-051 ) , we collected the data of 310 patients with breast invasive ductal carcinoma who underwent modified mastectomy from January 2001 to July 2008 .The expression of AR was detected by immunohistochemistry .The relationship of AR expression with patients’age, lymphatic metastasis, ER, P53, and Ki67 expression was analyzed using χ2 test.The relationship of AR expression with molecular subtype , tumor size and histological grade was analyzed by nonparametric rank sum test .The Kaplan-Meier survival curve was used to investigate the impact of AR on breast cancer prognosis , and Cox regression model was used to explore independent factors in the prognosis of breast cancer.Results In all 310 patients, there were 155 patients with AR positive, 100 AR weakly positive, 55 AR negative, accounting for 50%, 32% and 18% respectively.AR expression was correlated with patients’age, molecular subtype, histological grade, ER, and P53 expression (χ2=7.128, P=0.028;Z=56.099 , P<0.001;Z=10.144 , P=0.006;χ2=65.577 , P<0.001;χ2=16.127 , P<0.001 ) .There was a significant difference in overall survival among three groups with different AR expression (χ2=11.965 , P=0.003 ) .The overall survival in AR positive group ( 81.3%) or AR weakly positive group ( 87.0%) was significantly higher than that in AR negative group ( 67.3%) (χ2 =5.991 , P=0.014;χ2 =10.335 , P=0.001 ) , while the overall survival showed no statistical difference between AR positive group and AR weakly positive group (χ2=2.386, P=0.122).Cox regression showed that AR , tumor size, lymph node status and molecular subtypes were the independent predictive factors in the prognosis of breast cancer ( OR=0.559 , 95%CI:0.318-0.981 , P=0.043;OR=1.505 , 95%CI:1.047-2.161 , P=0.027;OR=1.616, 95%CI:1.046-2.497, P=0.030; OR=1.457 , 95%CI: 1.023-2.076 , P=0.037 ) .Conclusion AR expression is associated with the prognosis of breast cancer , so it can be used as an independent predictive factor in the prognosis of breast cancer .