Outcome after radical prostatectomy with a pretreatment prostate biopsy Gleason score of >/=8.

Objective: To determine the outcome and predictors of recurrence in patients with a pretreatment prostate biopsy Gleason score (GS) of >/= 8 and treated with radical prostatectomy (RP).
Patients and Methods: We retrospectively reviewed 1048 consecutive patients who underwent RP by one surgeon (M.S.S.); patients who had a pretreatment biopsy GS of >/= 8 were identified. Information was recorded on patient age, initial prostate specific antigen (PSA) level, clinical stage, biopsy GS, pathology GS, extraprostatic extension (EPE), tumour volume, surgical margin status, seminal vesicle invasion (SVI), and lymph node involvement. The results were assessed statistically using the Kaplan-Meier method, univariate log-rank tests and multivariate analysis using Cox's proportional hazards regression.
Results: In all, 123 patients met the initial selection criteria; 44 were excluded from further analyses (five salvage RP, 23 < 1 year follow-up and 16 adjuvant treatment). Thus 79 patients were included in the uni- and multivariate analyses; 25 (31%) patients had a GS of </= 7 in the RP specimen and 54 (69%) remained at GS >/= 8. The mean follow-up was 55 months, the age of the patients 63 years and the mean (sd) initial PSA level 13 (12) ng/mL. The overall biochemical failure rate was 38% (41% if the final GS was >/= 8 and 32% if it was </= 7). For those with a GS of >/= 8 in the RP specimen, 20% (11/54) were organ-confined; two patients (2.5%) in this group developed local recurrence. If the final GS was </= 7, 52% (13/25) were organ-confined. In the univariate analysis, significant risk factors for recurrence were PSA >/= 20 ng/mL, EPE, SVI, a positive surgical margin and tumour volume. Cox's proportional regression indicated that a PSA of >/= 20 ng/mL (hazard ratio 7.9, 95% confidence interval 2.6-24.2, P < 0.001), the presence of EPE (4.2, 1.6-10.9, P = 0.004) and a positive surgical margin (3.8, 1.5-9.7, P = 0.005) were significant independent predictors in a multivariate analysis.
Conclusion: RP is a reasonable treatment option for patients with a prostate biopsy GS of >/=8 and clinical stage T1-2. These patients have a high chance of remaining disease-free if their PSA level is </= 20 ng/mL. Patients with a pretreatment biopsy GS of >/= 8 should be counselled about the potential differences between the biopsy and the RP specimen GS.