Objectives To evaluate disease outcome using serum prostate-specific antigen (PSA) as an outcome measure in patients with T1 or T2 prostate cancer treated with radiation therapy in the PSA era. Methods. We reviewed the outcome for 461 patients with T1 (n = 205) or T2 (n = 256) prostate cancer followed for a median of 31 months after radiation therapy as the sole initial treatment. Univariate and multivariate analyses were used to delineate significant prognostic factors. Results. The freedom from relapse or rising PSA rate was 70% at 6 years and the survival rate was 83%. Although T stage, Gleason grade, serum prostatic acid phosphatase level, and serum PSA level were each significant determinants of outcome in univariate analysis, pretreatment PSA level was the only clearly independent covariate (P <0.0001) in multivariate analysis. The 5-year actuarial freedom from relapse or from rising PSA levels is shown according to the pretreatment PSA level: 4 ng/mL or less (117 patients), 91%; more than 4 but 10 ng/mL or less (169 patients), 69%; more than 10 but 20 ng/mL or less (118 patients), 62%; and more than 20 ng/ml (57 patients), 38%, PSA doubling times in 75 patients with rising post-treatment profiles ranged from 1.3 to 78.2 months (mean, 14.4; median 11.3). Faster doubling times correlated significantly with adverse pretreatment prognostic factors (high-grade, high pretreatment PSA, and aneuploidy). To date, the survival rate of patients with rising PSA profiles was not depressed below the expected. Conclusions. Radiation therapy is an acceptable modality for treating T1 or T2 disease and produces results comparable to those following radical prostatectomy when patients are stratified according to their pretreatment PSA value. The rapid PSA doubling times observed in patients with relapsing disease are more consistent with a ''selective'' rather than an ''aggravation'' mechanism.
