Predicting long-term survival, and the need for hormonal therapy: A meta-analysis of RTOG prostate cancer trials

Purpose: To assess the impact of short-term and long-term androgen suppression on the disease-specific and overall survival of 2200 men treated with radiotherapy on one of 5 prospective randomized trials when stratified by prognostic risk groups. Methods and Materials: Between 1975 and 1992, 2742 men were treated for clinically localized prostate cancer on one of 5 consecutive prospective Phase III randomized trials. Patients were selected for this analysis if they were deemed evaluable and eligible for the trial, and if follow-up information was available. For this analysis patients were stratified into four previously described prognostic risk groups: Group 1 patients had a Gleason score (GS) = 2-6, and T1-2Nx; Group 2: GS = 2-6, T3Nx; or GS = 2-6, Nf, or GS = 7, T1-2Nx; Group 3: T3Nx, GS = 7; or Nf, GS = 7, or T1-2Nx, GS = 8-10; and Group 4 patients were T3Nx, GS = 8-10, or N+, GS = 8-10. The median pretreatment prostate-specific antigen (PSA) was 25 ng/ml for the 434 evaluable patients for whom this information was available. The median follow-up times for patients treated on early studies exceeded 11 years, and for more recent studies 6 years. Results: Risk group 2 patients with "bulky" or T3 disease appeared to have a disease-specific survival benefit at 8 years with the addition of 4 months of goserelin and flutamide. Group 3 and 4 patients were noted to have an approximately 20% higher survival at 8 years with the addition of long-term hormonal therapy (p less than or equal to 0.0004). Conclusions: Based on this meta-analysis of RTOG trials, subsets of patients can be identified who either do not appear to benefit from the use of hormonal therapy, benefit from short-term hormonal therapy, or who benefit only from long-term hormonal therapy. These observations should be confirmed by prospective randomized trials before they can be considered conclusive. In the meantime, however, these observations provide rational guidelines for deciding who should receive hormonal therapy and for how long. (C) 2000 Elsevier Science Inc.