Oncology - Prostate
Introduction & Objective :
Gleason Score (GS) assignment from prostate biopsy has a significant rate of discordance with that assigned from post-radical prostatectomy (RP) pathology. The Prolaris Score (PS), derived from a genomic Cell Cycle Progression analysis and Prostate Cancer Risk Assessment (CAPRA) score, has been validated in numerous settings as an independent predictor of cancer-related death and biochemical recurrence. The score is used to predict whether an individual’s cancer is more aggressive, less aggressive, or consistent with others in his AUA risk group. Thus far, the ability of the PS to predict a change in GS between biopsy and RP has not been evaluated.
Methods : We evaluated 66 men with prostate cancer who underwent treatment with RP at one tertiary care center between 2015 and 2017. Patients were stratified by AUA risk score with representation in low (LR)-, intermediate (IR)-, and high-risk (HR) groups. They were subsequently sub-grouped by change in GS as upgraded, downgraded, or no change. A PS on a continuum of 1 (less aggressive) to 5 (more aggressive) was obtained for each patient. The mean PS for each subgroup was then calculated to assess for correlation between Prolaris-predicted risk and grade-change.
Results : The analyzed cohort included 9 HR, 45 IR, and 12 LR patients with biopsy GS of 3+3 (15.4%), 3+4 (33.0%), 4+3 (15.0%), 4+4 (9.2%), and 4+5 (1.5%). 86% of patients had pre-biopsy PSA levels of ≤10.0. PS distribution included 6.2% with scores of 1-2.4, 53.8% 2.5-3.5, and 40% 3.6-5. Overall, upgraded (n=18), no change (n=34), and downgraded (n=12) groups had average PS (with standard deviations) of 3.48 ± 0.76, 3.34 ± 0.58, and 3.84 ± 0.71, respectively. A one-way analysis of variance (ANOVA) was calculated on patients' Prolaris Scores, stratified by grade change. The analysis was not significant, F(2, 63) = 2.46, p = 0.094.
When analyzed as a whole, and when stratified by AUA risk, there was no correlation between average PS and grade change from biopsy to post-RP pathology. While the average PS for the “no change” group was within the 2.5-3.5 range, there was high degree of variance within the group. Furthermore, the ANOVA revealed that the average PS was not significantly different than those of the upgraded or downgraded groups. Our data suggest that the Prolaris Score may not be used to reliably predict a change in biopsy GS. Both its independence from GS and its validated efficacy in predicting mortality and biochemical recurrence make the Prolaris Score a valuable tool to be used in risk stratification and shared decision-making.