[898] Androgen Receptor in Tumor and Stroma in Conservatively Treated Prostate Cancer

Sharanpal S Jeetle, Zi H Yang, Elzbieta Stankiewicz, Gabrielle Fisher, Colin Cooper, Christopher S Foster, Henrik Moller, Peter Scardino, Victor E Reuter, Jack Cuzick, Daniel Berney. Barts Cancer Institute, London, United Kingdom; Memorial Sloan-Kettering Cancer Center, New York; Royal Marsden Hospital, Surrey, United Kingdom; Liverpool University Hospital, Liverpool, United Kingdom; Kings College London, London, United Kingdom

Background: Androgen receptor (AR) is present in the epithelium and stroma of the prostate. Invasive malignancy may be due to down-regulation of stromal AR and maintenance or up-regulation of epithelial AR. Associations between stromal AR and prognosis have been suggested by others. We wished to determine if these changes were apparent in a cohort of 721 patients with conservatively treated prostate cancer (PC).
Design: 721 patients with clinically localized and conservatively treated prostate cancers, diagnosed on TURP from 1991-96 were investigated. Contemporary Gleason score and serum PSA were available with up to 10 years follow up. AR immunohistochemistry was performed on tissue microarrays of tumor and normal tissue and assessed semi-quantitatively in the stroma and epithelium The end point was death from PC. Univariate and multivariate analysis including Gleason score and PSA was performed by proportional hazard (Cox) regression analysis.
Results: In the cancer cores, expression of epithelial AR was marginally but not quite significantly predictive of worse outcome on univariate analysis (HR=1.72, 95%CI 0.96,3.09, P= 0.07) and it did not correlate with Gleason (χ2=3.629,p=0.163) or PSA(χ2=5.67,p=0.225).On Multivariate analysis with Gleason AR in tumor was also not significant (HR=1.64, 95%CI 0.92,2.95, P=0.07). There was no correlation of stromal AR expression with Gleason or PSA and univariate analysis was not significant. In 'normal' cores, epithelial and stromal AR was not associated with outcome on univariate analysis nor was there correlation with Gleason or PSA.
Conclusions: AR in cancer was only marginally and non-significantly predictive of outcome and we were unable to replicate the results of other studies showing the importance of stromal AR as a prognostic factor. Our study weakly supports the general consensus that AR may be important in the progression of prostate cancer but it is not the only mode of paracrine signalling. We suggest the relationship is complex and there may not be a simple two-way signalling between stromal and epithelial AR. Further large scale studies on translational material are needed to understand the specific role of AR in prostate cancer.
Category: Genitourinary (including renal tumors)

Monday, March 19, 2012 9:30 AM

Poster Session I Stowell-Orbison/Surgical Pathology/Autopsy Awards Poster Session # 162, Monday Morning


Close Window