Objective To examine the usefulness of a test for prostate specific antigen (PSA) to predict survival in hormonally treated patients with metastatic prostate cancer. Patients and methods The study comprised 49 patients (mean age 72 years, sd 6) who underwent orchidectomy for metastatic prostate cancer. PSA was measured before orchidectomy and after 6 months, and the absolute, differential and proportional decreases calculated. A Cox proportional hazards regression model, which controlled for patient age, tumour (Gleason) grade and the number of skeletal metastases, was then used to evaluate these estimates of PSA as predictors of survival. Results The 6‐month proportional decrease in PSA from the pre‐operative level was the most accurate predictor of patient survival (P=0.006) after hormonal therapy for metastatic prostate cancer. Conclusion This information may help to direct appropriate patients to new and experimental therapies for metastatic disease.

No abstract