Quantitative Gene Expression Analysis of Androgen Receptor in Benign and Neoplastic Prostate Epithelial Cells May Predict PSA Recurrence

Inger Rosner, Lakshmi Ravindranath, Bungo Furusato, Yongmei Chen, Jennifer Cullen, Isabell Sesterhenn, David Mcleod, Shiv Srivastava, Gyorgy Petrovics

Walter Reed Army Medical Center Washington, DC and the Center for Prostate Disease Research Rockville, MD

Abstract

Introduction: Alterations of androgen receptor (AR) functions due to overexpression, amplification, or mutation have been described in a significant subset of advanced prostate cancer (CaP). Quantitative determination of AR expression levels in the prostate tumor cells may have potential to predict aggressive clinical behavior in newly diagnosed patients. In this study laser capture microdissected (LCM) benign and neoplastic epithelial cells from prostatectomy specimens were analyzed for correlation of AR expression with disease progression.

Methods: 105 hormone naïve patients post radical prostatectomy were studied. Benign and neoplastic prostate epithelial cells were collected with LCM from frozen tissue slides. Expression of AR and GAPDH genes were measured by duplex quantitative real-time RT-PCR in 210 specimens. The expression of AR, normalized to GAPDH expression in the same specimens, was compared in tumor and benign epithelial cells, and a correlation with clinico-pathological features was evaluated.

Results: Paired t-test was used to compare AR expression between tumor and benign prostate cells of each patient. Overall a 50% lower AR expression was detected in tumor tissue than in benign tissue (p=0.0037). However, patients with the highest AR expression in their tumor cells, compared to matched benign cells, were more likely to have biochemical disease recurrence (p=0.0183). Stepwise logistic regression analysis with age, race, PSA at diagnosis, pathologic stage, Gleason sum, follow up time and AR expression suggested that increased tumor versus benign AR expression ratio is an independent factor predicting PSA recurrence (p=0.0381). With each 2-fold increase in AR expression ratio the odds of PSA recurrence increased by 36.4%.

Conclusions: AR expression, as determined in LCM-derived malignant and benign prostate cells, was lower overall in tumor cells, however patients with increased expression of AR in their tumor versus benign cells have an increased risk of PSA recurrence. Therefore, quantitative determination of AR gene expression levels in prostate epithelial cells may be useful for predicting PSA recurrence. This study supports the accumulating data suggesting that elevated AR functions may contribute to CaP progression.

Print Window | Close Window