Long-term hazard of progression after radical prostatectomy for clinicallylocalized prostate cancer: Continued rise of biochemical failure after 5 years

Purpose: Cure from malignancy is commonly defined as a disease-free state l asting 5 years after treatment. We analyzed clinical and biochemical progre ssion rates after radical prostatectomy for men with clinically localized p rostate cancer with particular attention to recurrence beyond 5 years. Annu al hazard rates of progression were calculated to determine the probability of recurrence at specific intervals following surgery. Materials and Methods: The records of 2,782 men with clinically localized p rostate cancer (cT1-T2) undergoing radical prostatectomy between 1987 and 1 993 were reviewed. All patients were treated in the prostate specific antig en (PSA) era so that serial followup PSA values were available from the tim e of surgery. Analysis was limited to patients who did not receive adjuvant treatment within 90 days of radical prostatectomy. Disease progression was defined as documented local recurrence, systemic progression and/or PSA 0. 4 ng./ml. or greater. Lymph node positive cases were eliminated from analys is since almost all received adjuvant hormonal therapy. Annual hazard rates for progression were calculated using the formula: [No. events No. patient s at risk] x 100. Progression-free survival probabilities were determined u sing the Kaplan-Meier method. Results: Pathological stage was pT2a-b, N0 (68%), pTSa, N0 (21%) and pT3b, N0 (11%). Biochemical progression-free survival at 5 and 10 years was 76% a nd 59%, respectively, for the entire study population while those with path ologically organ confined (pT2, N0) cancers had progression-free survival r ates of 82% and 68% at 5 and 10 years, respectively. A total of 819 patient s (29%) eventually had disease progression, including 160 (6%) with progres sion after 5 years. Annual hazard rates were highest during the first 2 yea rs after radical prostatectomy for the entire population. Patients with adv erse prognostic features (pT3b, PSA 10 ng./ml. or greater, Gleason score 8- 10 and nondiploid cancers) had high initial hazard rates that decreased wit h time to lower levels. Those with pathologically organ confined cancer had low but constant hazard rates throughout followup. Conclusions: Although progression after radical prostatectomy usually occur s early, reflecting the impact of clinical under staging, a significant num ber of men, including those with organ confined cancers, will continue to h ave disease progression after 5 years. Patients undergoing radical prostate ctomy should be subjected to long-term followup to allow the option of earl y intervention should progression occur.