Introduction:

Gleason grade group (GGG) is a primary driver of the decision to select initial active surveillance versus treatment for prostate cancer. The role of genomic classifiers (GC) to sub-stratify risks of grade classification among patients with biopsy GG1 and GG2 disease selecting active surveillance has not been well defined. This study aims to evaluate how Decipher GC testing impacts the risk of progression to treatment over a 5-year period in active surveillance patients.

Methods:

We conducted a retrospective study to identify patients with biopsy GG 1 or 2 who underwent Decipher GC testing, and had a minimum of 2 biopsies. Patients were stratified into risk groups based on their GG and GC categories (0-0.45 [low] and >0.45 [int/high]), with four major risk groups of GG1 low GC, GG1 intermediate GC, GG1 high GC, and GG2 low GC. Risk groups of GG2 intermediate and higher were analyzed, but not the focus of this study given natural progression to treatment. Progression to treatment criteria was defined as development of GG2 intermediate / high GC or GG3+ pathology upon subsequent surveillance biopsy. Descriptive statistics, Fine & Gray competing risks regression, and cumulative incidence functions were used to evaluate progression risk across risk groups, adjusted for PSA density (PSAD) and PIRADS in binary fashion (4-5 vs. ≤ 3).

Results:

We identified 224 eligible patients including 19 (8.5%) GG1/GC low, 54 (24.1%) GG1/GC intermediate, 51 (22.8%) GG1/GC high, and 100 (44.6%) GG2/GC low. Progression to treatment criteria differed by GG and GC categories (Figure 1). Progression to treatment occurred in 114 (50.9%) patients, with 10.5%, 13%, 29.4%, and 90% progression rates across the risk groups, respectively. Fine & Gray regression showed GG1 intermediate GC was not significantly different from GG1 low GC (SHR 0.84, p = 0.83). Notably, GG1 high GC had significantly greater progression risk (SHR 5.12, p = 0.036), as well as GG2 low GC with the highest risk (SHR 12.38, p = 0.001) to treatment criteria relative to GG1/GC when adjusted for PIRADS 4-5 (SHR 1.36, p = 0.112) and PSAD (SHR 13.0, p = 0.009). Elevated baseline PSAD was independently associated with higher progression risk, emphasizing PSAD use in AS risk assessment as a potential biomarker.

Conclusion:

Our findings suggest that in AS patients, GG1 intermediate GC behaves similarly to GG1 low GC, as expected in active surveillance. However, GG1 high GC demonstrated significantly higher progression risk, as well as GG2 low GC in terms of progression to treatment criteria. These results suggest higher GC category with GG1 as well as GG2 low GC may warrant earlier definitive treatment or cautious surveillance. The incorporation of Decipher GC with Gleason grade group serves to improve risk stratification of the AS prostate cancer cohort and provides promise to guide personalized management for the AS cohort.  

Funding: N/A