An NCI-sponsored, phase II trial of N-(4-hydroxyphenyl)- retinamide (4-HPR) in patients with organ-confined prostate cancer in the period prior to radical prostatectomy was carried out. Thirty-seven men with the histologic diagnosis of prostate cancer planning to have radical prostatectomy entered the study after informed consent and were given 4-HPR (or matching placebo) as a single daily dose (two 100-mg capsules of 4-HPR or two capsules of placebo daily) for 3 weeks prior to surgery. Four men dropped out for unrelated reasons. Thirty-three men completed the study. At the time of surgery, repeat biopsies of the prostate were performed to study the effects of the drug on potential surrogate endpoint biomarkers (SEBs) of malignancy within the tissue. The panel of potential SEBs of malignancy include p53, cytomorphometric indices, ploidy, PNCA, erbB-2, erbB-3, EGF receptor, TGF-alpha tumor-associated glycoprotein-72, fatty acid synthetase and Lewis Y antigen. Twenty-three patients had matching pre- and posttherapy lesions and were considered informative. Results from the patients indicate significant differential expression of biomarkers in pretreatment specimens of uninvolved prostatic tissue (normal-appearing epithelia) prostatic intraepithelial neoplasia (PIN) and prostate cancer. The mean erbB-2 expression was 0.58 in uninvolved vs. 1.04 in PIN (p = 0.002); while the mean erbB-2 expression was 1.35 in prostate cancer (p = 0.0007, uninvolved vs. prostate cancer). A similar pattern of increased biomarker expression between uninvolved and PIN or prostate cancer tissues can be observed for EGF receptor (mean = 1.21, 1.87 and 1.76 for uninvolved, PIN and prostate cancer, respectively) and erbB-3 (mean = 0.81, 1.59 and 1.30 for uninvolved, PIN and prostate cancer, respectively). There were no statistically significant differences in biomarkers observed in the 4-HPR-treated patients when compared with placebo-treated control patients. There was a posttreatment up-regulation of biomarkers observed in both groups of patients. This observation is most likely explained by an effect due to the diagnostic sextant biopsy equally affecting both groups of patients. Results from this study do not demonstrate a chemoprevention effect of 4-HPR on tissue-based SEBs at the dose given.