Complete androgen ablation suppresses prostate stem cell antigen (PSCA) mRNA expression in human prostate carcinoma

Prostate stem cell antigen (PSCA) is a recently identified glycosylphosphatidylinositol (GPI)-anchored cell surface protein belonging to the Thy-1/Ly-6 family of cell surface antigens. Prior data in prostate cancers indicated that PSCA is directly regulated by androgens and PSCA expression increases with high-tumor grade, advanced stage, extracapsular invasion, and androgen-independent progression. The effect of complete androgen ablation (CAA) on tumor PSCA mRNA expression has not been elucidated. The purpose of the present study was to investigate the variations in the expression levels of PSCA mRNA before and after CAA, and further evaluate the clinically prognostic value of PSCA in human prostate carcinoma.

PSCA in situ hybridization (ISH) was performed on the cancerous pretreatment biopsy or transurethral resection of prostate (TURP) tissue of 42 men with primarily organ-confined prostate cancer before CAA, and on their tumor tissue from radical retropubic prostatectomy after CAA with bicalutamide and goserelin acetate for 3 months prior to undergoing radical prostatectomy. Tumor cytoplasmic staining of PSCA mRNA was evaluated by two independent pathologists and the differences of PSCA mRNA expression levels between the samples before and after CAA were analyzed using the Student's t-test. Thirty-six to forty months follow-up studies after radical retropubic prostatectomy were performed and aimed at assessing the correlation of PSCA mRNA expression level with local recurrences or metastases from the cancer.

The percent of cells positive for PSCA mRNA by ISH labeling declined from 67.3% (0-89%)+/-9.4% before CAA to 33.8% (0-92%)+/-7.7% after CAA (P<0.001). Before CAA, 40 of 42 cases (95.2%) were positive for PSCA mRNA labeling, however, after CAA the percentage of positive reactivity of PSCA mRNA was decreased to 27 of 40 cases (67.5%), in which none was found with local recurrences or distant metastases after radical prostatectomy on follow-up. This decline in PSCA mRNA labeling was dependent on the original tumor grade with Gleason score of or=8: 73.4%+/-13.8% (P<0.05, respectively). The rest 13 cases had the increased percentage of cells positive for PSCA mRNA after CAA, in which 3 cases were found with local recurrences and 4 cases with distant metastases from tumor on follow-up.

Our data demonstrate that CAA for prostate cancer can suppress PSCA mRNA expression with a tumor grade dependence and the increased expression of PSCA mRNA after CAA may be a clinically adverse predictor for tumor recurrences or distant metastases.