Analysis of prognostic factors in men with metastatic prostate cancer. Uro-Oncology Group of Northern Alberta.

We determined the influence of the extent of disease on bone scan, serum testosterone, patient age, performance status, method of initial diagnosis, Gleason grade, clinical stage at diagnosis, serum acid phosphatase, serum prostate specific antigen (PSA) and primary hormonal treatment on survival. The clinical and hormonal data were obtained when the presence of metastatic disease was established and treatment was to be initiated in 162 men with metastatic prostate cancer. Mean followup was 16 months (range 1 to 105 months). A total of 70 men (43.2%) died of the metastatic disease during the evaluation period. Log rank analysis revealed that only serum testosterone (p = 0.035) and extent of disease on bone scan (p = 0.003) significantly affected over-all survival. A trend (p = 0.068) towards decreased survival was observed with increasing values of PSA. Increasing values of acid phosphatase positively correlated with extent of disease on bone scan but was not a significant independent prognostic factor. Patient age, performance status, clinical stage, method of initial diagnosis, Gleason grade and type of hormonal treatment did not significantly influence survival. Upon using multivariate Cox analysis, only extent of disease on bone scan was significantly correlated with over-all survival (p less than 0.014). PSA may also be influential but longer duration of followup will be necessary. We conclude that extent of disease on bone scan is the most important prognosticator of the analyzed factors and that serum testosterone may be of value.