Immediate versus deferred treatment for advanced prostatic cancer: initial results of the Medical Research Council Trial. The Medical Research Council Prostate Cancer Working Party Investigators Group.

OBJECTIVE To compare the effect on the course of advanced prostate cancer of hormone treatment commenced on diagnosis with that deferred until clinically significant progression occurs. PATIENTS AND METHODS Nine hundred and thirty-eight patients with locally advanced or asymptomatic metastatic prostate cancer were randomized either to immediate treatment (orchidectomy or luteinizing hormone-releasing hormone analogue) or to the same treatment deferred until an indication occurred. Follow-up and management were otherwise according to the participating clinician's normal practice. Information was collected annually on survival, local and distant progression, and major complications (pathological fracture, spinal cord compression, ureteric obstruction and extra-skeletal metastases). RESULTS Follow-up data were returned on 934 patients; 51 deferred patients died from causes other than prostate cancer before treatment was started (but only five of these presented at age < 70 years) and 29 died from prostate cancer before treatment could be started. Treatment was commenced for local progression almost as frequently as for metastatic disease. Progression from M0 to M1 disease (P < 0.001, two-tailed) and development of metastatic pain occurred more rapidly in deferred patients; 141 deferred patients needed transurethral resection for local progression compared with 65 treated immediately (P < 0.001, two-tailed). Pathological fracture, spinal cord compression, ureteric obstruction and development of extra-skeletal metastases were twice as common in deferred patients. Of the patients who died, 67% did so from prostate cancer; 361 patients died in the deferred arm compared with 328 in the immediate arm (P = 0.02, two-tailed), where 257 and 203 were deaths from prostate cancer, respectively (P = 0.001 two-tailed). This difference was seen largely in MO patients, with 119 and 81 deaths from prostate cancer, respectively (P < 0.001 two-tailed). CONCLUSIONS The results consistently favour immediate treatment, although some of the data, especially on MO patients, are immature. The implications for management of advanced prostate cancer are discussed.