Androgen receptor expression in prostate cancer lymph node metastases is predictive of outcome after surgery

Citation
Sd. Sweat et al., Androgen receptor expression in prostate cancer lymph node metastases is predictive of outcome after surgery, J UROL, 161(4), 1999, pp. 1233-1237
Citations number
16
Categorie Soggetti
Urology & Nephrology","da verificare
Journal title
JOURNAL OF UROLOGY
ISSN journal
00225347 → ACNP
Volume
161
Issue
4
Year of publication
1999
Pages
1233 - 1237
Database
ISI
SICI code
0022-5347(199904)161:4<1233:AREIPC>2.0.ZU;2-C
Abstract
Purpose: Androgens mediate the growth of prostate cancer cells. The predict ive value of androgen receptor immunostaining in patient outcome is controv ersial. We studied the expression of androgen receptors in a large series o f patients with node positive cancer, and correlated the results with clini cal progression and survival. Materials and Methods: We evaluated 197 patients with a mean age of 65.5 ye ars who had node positive adenocarcinoma, and who underwent bilateral pelvi c lymphadenectomy and/or radical prostatectomy at our clinic between 1987 a nd 1992. Mean followup was 6.3 years. Immunohistochemical studies were perf ormed using an antihuman androgen receptor monoclonal antibody. In each cas e 100 nuclei were counted from 3 separate areas (total 300 nuclei per diagn ostic category) of benign epithelium, cancer and lymph node metastases. Mea n androgen receptor expression was determined from the mean of the individu al cases. The intensity of immunoreactivity was evaluated on a scale of 0-n o staining to 3-strong staining. We assessed the correlation of androgen re ceptor immunoreactivity, deoxyribonucleic acid ploidy, Gleason score and pr eoperative serum prostate specific antigen (PSA) with clinical progression, all cause survival and cancer specific survival using the Cox proportional hazards model. Clinical progression was defined as a positive bone scan. Results: There was heterogeneous staining in the majority of cells in benig n and malignant prostatic epithelium. The mean number of immunoreactive nuc lei was similar in all groups (56, 53 and 56% of benign epithelium, cancer and lymph node metastases, respectively). Pairwise comparisons revealed tha t the only significant difference was between benign epithelium and cancer (p = 0.001) with greater immunoreactivity in benign epithelium. Intensity w as lower in benign epithelium than in cancer and lymph nodes (p < 0.05). An drogen receptor expression in lymph node metastases was associated with all cause and cancer specific survival on univariate analysis (p = 0.03 and 0. 04, respectively). The 7-year cause specific survival was 98, 94 and 86% in patients with 51 to 69, less than 50 and greater than 70% androgen recepto r expression in lymph node metastases, respectively (p < 0.05). The associa tion of androgen receptor expression in lymph node metastases was significa nt on multivariate analysis for cancer specific survival (p = 0.021) but no t all cause survival (p = 0.16) after controlling for Gleason score, deoxyr ibonucleic acid ploidy and preoperative PSA. Androgen receptor immunoreacti vity in lymph nodes was not a significant univariate or multivariate predic tor of clinical progression, while androgen receptor expression in the prim ary cancer was not predictive of clinical progression or survival (p > 0.05 ). Conclusions: Androgen receptor expression was similar in benign epithelium, primary cancer and lymph node metastases with approximately half of the ep ithelial cell nuclei staining. Androgen receptor immunoreactivity in lymph node metastases was predictive of cancer specific but not all cause surviva l in univariate and multivariate models. Gleason score was the strongest pr edictor of all cause survival in this cohort of patients. Our results indic ate that it may be clinically useful to determine lymph node androgen recep tor expression in men with advanced prostate cancer when combined with Glea son score and PSA.