In a prospective trial conducted by the Gruppo Onco Urologico Piemontese, newly diagnosed prostate cancer patients with bone metastases were randomized to receive goserelin (3.6 mg subcutaneously every 4 weeks) or goserelin plus mitomycin at 14 mg/m2 i.v. every 6 weeks. Treatment was planned to be continued until progression. The study was interrupted because of inadequate accrual rate when 63 patients had been recruited. A long-term follow-up (median, 47 months), performed to counterbalance the limited number of patients included, revealed no difference in time to progression and overall survival between the study treatments. However, 56.5% of assessable patients allocated to the chemotherapy arm presented a > or =90% reduction of prostate-specific antigen levels compared with 36.3% in the goserelin group, and previously elevated levels normalized in 73.9% versus 45.4%. Non-progressing patients received 5-7 cycles of mitomycin C with acceptable toxicity, but the cytotoxic treatment was interrupted early in all cases within the first year due to cumulative myelotoxicity. In conclusion, the results, although inconclusive, fail to support a clear advantage in terms of cost/benefit of chemotherapy plus hormone therapy over hormone treatment alone in advanced prostate cancer with bone involvement.