Hazard rates for progression after radical prostatectomy for clinically localized prostate cancer

Objectives. We calculated the annual hazard rate (HR) for prostate cancer recurrence after radical prostatectomy (RP) to elucidate the pattern of treatment failure over time and to assess the efficacy of definitive therapy. Methods. We calculated the progression-free probabilities (PFP) and HRs after RP for a cohort of 611 consecutive men with clinically localized (cT1-2, NX, MO) prostate cancer and no other treatment before documented progression. Results. PFP for the entire study population was 78% at 5 and 76% at 10 years. The highest HR (0.09) was observed in the year immediately after surgery and dropped to 0 by year 7 (no patient recurred after year 6). Average annual HRs calculated for 3-year intervals resulted in steadily declining HRs over time for the entire study population and for all subsets, except those with a cancer pathologically confined to the prostate. Overall, the more ominous the prognostic factor, the higher the initial HR. For poorly differentiated cancers (biopsy Gleason sum 8 to 10), the HR was high in years 1 and 2 and dropped rapidly to 0 thereafter. Conclusions. Prostate-specific antigen (PSA) progression after RP usually occurred early (77% within the first 2 years) and was largely due to understaging. Late recurrences were rare in patients who were regularly evaluated with PSA. However, because the confidence intervals in our study were broad, larger patient populations with longer follow-up are needed for a definitive establishment of the time, course, and pattern of recurrence after surgery. (C) 1997, Elsevier Science Inc. All rights reserved.