Activity of docetaxel with or without estramustine phosphate versus mitoxantrone in androgen dependent and independent human prostate cancer xenografts.

PURPOSE: Prostate cancer is known to be refractory to chemotherapy with only mitoxantrone showing some benefit. Recent clinical data indicate the antitumoral efficacy of taxanes alone or combined with estramustine phosphate. We compared the response to these treatments of hormone dependent and independent human prostate cancer xenografts. MATERIALS AND METHODS: PAC120, an androgen dependent human prostate cancer xenograft, and several HIDs, which are androgen independent variants, were established in nude mice. Human gene expression was determined by semiquantitative reverse transcriptase-polymerase chain reaction. Androgen deprivation was achieved by surgical castration. Tumor bearing mice received 20 mg./kg. docetaxel on day 2, 1 mg./kg. hydrocortisone on days 1 to 3, 4 mg./kg. estramustine phosphate on days 1 to 4 or 1 mg./kg. mitoxantrone on day 1 by the intraperitoneal route for 3-week cycles. Relative tumor volume and growth delay were evaluated. Histological examination of tumors was done before and after treatment. RESULTS: Mitoxantrone transiently decreased the growth rate of HID xenografts but did not affect that of PAC120. Estramustine phosphate alone inhibited the growth of PAC120 but not that of HID variants. Docetaxel inhibited the growth of all prostate cancer xenografts (PAC120 and HIDs) and increased survival. PAC120 showed distinct response patterns during prolonged treatment. Efficacy was significantly decreased by splitting docetaxel into 2 doses given at a 7-day interval (p = 0.01). The docetaxel effect was potentiated by estramustine phosphate in 1 of the 2 HID variants tested. In castrated mice docetaxel induced a greater growth delay than in intact male mice (p = 0.01). High Her2/neu and beta2-tubulin transcripts were detected in all samples. Prostate specific antigen, androgen receptor and multidrug related protein-1 mRNA did not correlate with the drug response, while CYP3A4 mRNA inversely correlated with the response. Docetaxel treated tumors had an increased number of mitotic cells with centrosome alterations and multinuclei, an increased number of Ki67 labeled cells and a strong decrease in beta-tubulin without evidence of apoptosis. CONCLUSIONS: Docetaxel showed a significant antitumoral effect on hormone dependent and tumors, which was largely superior to that of mitoxantrone. Estramustine phosphate alone had a modest effect. The drug response was associated with high Her2/neu expression, low CYP3A4 expression and the induction of numerous mitotic abnormalities.