In a recent intergroup study under the auspices of the National Cancer Institute, 603 eligible patients with newly diagnosed disseminated adenocarcinoma of the prostate were prospectively randomized in a double-blinded clinical trial to receive either a gonadotropin-releasing hormone analogue (leuprolide) and a nonsteroidal antiandrogen (flutamide) or leuprolide and placebo. Of the 603 eligible patients, 300 were in the leuprolide and placebo arm and 303 were in the leuprolide and flutamide arm. At the time of disease progression, the code was broken: Those patients in the placebo arm were given the opportunity to receive flutamide, and the patients in the flutamide arm were treated at their physician's discretion. There was no survival time distribution difference, based on survival measured from the progression data, between the patients who were received flutamide after progression and those who were treated at their physician's discretion after progression. Furthermore, the addition of flutamide to leuprolide at the time of disease progression resulted in a survival-time distribution that is similar to other treatments of hormone-refractory prostate cancer.