The role of prostate specific antigen in the baseline assessment of patients undergoing hormone therapy for advanced prostate cancer.
One hundred and thirty-nine patients with advanced prostate cancer were entered into a randomised trial to test the efficacy and tolerance of goserelin 3.6 mg depot (Zoladex) versus stilboestrol 3 mg/day. As well as the usual clinical and radiological assessments of extent of disease, we used an immunoradiometric assay of prostate specific antigen (PSA) (Hybritech Europe) and normal laboratory enzymatic assays of acid phosphatase (AP) and alkaline phosphatase (ALKP) for biochemical assessment. The upper limit of normal for PSA was taken as 10 micrograms/l. The range of PSA was wide and differed significantly from that of AP and to a lesser extent ALKP in metastatic cases. PSA outperformed both AP and ALKP in both the local and advanced groups in terms of sensitivity. There was no correlation, however, between histological grade and level of PSA, AP or ALKP (the latter in cases with bone disease). In patients with metastatic disease diagnosed by bone scan, nine patients had one abnormal site/one "hot spot", and all of these had a PSA greater than twice the normal upper limit. Early death due to prostate cancer was noted in four patients with levels of PSA greater than 2500 micrograms/l. PSA is more sensitive than either enzymatic AP or ALKP in both locally advanced and metastatic prostate cancer and is useful in identifying those advanced cases who have single lesions on bone scan. In this series PSA gave an overall sensitivity of 89%, compared with 63% for AP and 64% for ALKP in patients with metastatic disease.