Does prostate biopsy Gleason score accurately express the biologic features of prostate cancer?

Abstract

Purpose To correlate Gleason grading in prostate biopsies with the final Gleason score in radical prostatectomy specimens, and to investigate predictors for concordance and preoperative undergrading. Materials and methods The charts of 303 patients who underwent radical retropubic prostatectomy between 1992 and 2002 were retrospectively reviewed. Prostate biopsy and surgical specimen Gleason scores and correlative clinical data were recorded, and a multivariate analysis model was applied. Results Data were available in 293 cases (97%). The preoperative biopsy predicted the prostatectomy Gleason score accurately in 51% and undergraded them in 41% of the patients. Accuracy rates were significantly higher for Gleason scores 7–10 compared to low Gleason scores (2–4), concordance 90% and 6%, respectively ( P < 0.01). Moreover, accuracy rates were higher in patients with prostate-specific antigen (PSA) higher than 10 ng/ml (85% vs. 40%; P < 0.01) and prostate glands smaller than 55 g (68% vs. 38%; P < 0.01). In 233 patients, the biopsy Gleason score did not include 4 or 5 components. Upgrading to 4 or 5 in 1 of the components was noted in 32 patients (14%). Multivariate analysis revealed that upgrading is associated with preoperative serum PSA (odds ratio 1.05; P < 0.05) and the percentage of positive cores in the biopsy (odds ratio 1.47; P < 0.001). Conclusions Biopsy Gleason scores of 2–4, low PSA, and a low percentage of positive cores in the biopsy can predict the biopsy driven biologically significant undergrading of 1 of the components of the Gleason score that may adversely affect therapeutic decisions.