Weighted Gleason scores do not outperform standard clinical Gleason scores

Predicting patient survival rates following radical prostatectomy remains an area of clinical interest. We compared the ability of standard clinical Gleason scores and alternative 'weighted' Gleason scores to predict pathology, margin status and recurrence in prostate cancer.

Patients who underwent robotic radical prostatectomy performed by a single surgeon between Jan 2007 - Dec 2008 were included. Tumor at the inked margin in pathologic samples was considered a positive margin. Recurrence was defined as PSA ≥ 0.2 or the institution of salvage therapy. Standard pathologic Gleason scores were recorded. The proportion of tumor in each core was used to calculate 'weighted' and 'rounded weighted ' Gleason scores. The ability of each Gleason score to predict pathology, margin status and recurrence were statistically compared.

Of 433 cases, 281 with uniform Gleason 6 cores were excluded. One hundred and fifty-two cases had Gleason scores ≥ 7, of which complete data were unavailable for three patients. In the final cohort of 149 cases, 72 (48.3%) patients had uniformly scored biopsies, while 77 (51.7%) had biopsies with non-uniform Gleason scores. The positive margin rate and recurrence free rates were 30.2% and 77.2%, respectively. Analyses of the entire patient cohort, and patients with non-uniform cores, found no significant difference between the predictive capacities of each scoring system. The alternative algorithms were not shown to be better predictors of pathologic Gleason score, margin status or recurrence.

Using the highest standard Gleason score of all cores to define a preoperative Gleason score remains an appropriate clinical practice.