Randomized, multicenter, phase II trial of two multicomponent regimens in androgen-independent prostate cancer

Purpose: Several multicomponent regimens have been reported to be useful in advanced androgen-independent prostate cancer. We used a randomized phase II design to evaluate and compare two such regimens. Patients were accrued primarily in the community setting. Patients and Methods: Patients with progressive, androgen-independent prostate cancer were randomly assigned to one of two treatments: either ketoconazole/doxorubicin alternating with vinblastine/estramustine (KA/VE) or paclitaxel, estramustine, and oral etoposide (TEE). Patients were prospectively stratified on the basis of disease volume. The primary end points were response and overall survival time. Results: A total of 75 patients were registered; 71 are included in the analysis. By the criterion of an 80% prostate-specific antigen reduction maintained for at least 8 weeks, 11 (30%) of 37 patients in the TEE arm responded, whereas 11 (32%) of 34 assigned to KA/VE responded. Median survival was 16.9 months (95% confidence interval [CI], 10.5 to 21.2 months) in the TEE arm and 23.4 months (95% CI, 12.9 to 30.6 months) for patients treated with KA/VE. Many patients (24%) failed to complete at least 6 weeks of therapy, including five (8%) treatment-related early deaths. Conclusion: Each of these regimens produced clinically significant responses, and the observed median survival (18.9 months for all 71 patients) compares favorably with previously published results, especially in the community setting. Nonetheless, it is apparent that these first-generation regimens must be applied judiciously, and thus we view efforts at better patient selection and the development of more tolerable therapies as higher priorities than carrying either of these regimens to phase III evaluation in the cooperative group setting.