Discriminative Ability of Decipher to Predict Biopsy Upgrade on Active Surveillance
Benjamin Press, MD, Ghazal Khajir, MD, Michael Leapman, MD, Preston Sprenkle, MD.
Yale University School of Medicine, New Haven, CT, USA.
BACKGROUND: Although initially validated as prognostic tools in the setting of definitive prostate cancer (PCa) treatment, tissue-based gene expression tests have become increasingly utilized in active surveillance (AS) of favorable-risk PCa. We aimed to assess the prognostic significance of the Decipher assay, a 22-gene signature associated with PCa outcome, on biopsy grade progression during AS.
METHODS: Between July 2016 and November 2020, 133 men on AS underwent 146 separate MRI - ultrasound fusion targeted biopsies and systematic biopsies with their prostate tissue sent for Decipher testing. Demographic information, PIRADS, and Decipher scores were prospectively recorded. Decipher risk categories (DR) were labeled by commercial risk categories as low (<0.45), intermediate (0.45-0.60), or high (>0.60) score. Pathologic upgrade was defined as an increase in Gleason Grade Group (GG) on subsequent biopsy. We assessed the association between Decipher score and upgrade using univariate statistics and logistic regression adjusted for clinical and pathologic factors.
RESULTS: Median age was 67.7 years (IQR = 62.4 - 71.4). Median PSA was 5.6 ng/mL (IQR = 4.3 - 7.1). In this cohort, 75.9% and 24.1% of men had GG1 and GG2 PCa, respectively. Median time between biopsy was 13.6 months (IQR = 11.9 - 16.9). Median Decipher score was 0.39 (IQR 0.25 - 0.48) and upgrade occurred in 32.3%. Decipher scores were similar among patients who did and did not experience upgrade (0.36 vs 0.40, p = 0.27). Upgrade rates were also similar using DR (31.9% [low] vs 32.4% [intermediate] vs 46.7% [high], p = 0.455). Multivariable logistic regression revealed increasing Decipher score was associated with greater odds of upgrade (OR 1.31 per 0.10 unit increase, p = 0.033) (Table 1). When stratifying by GG, Decipher score was associated with upgrade among patients with diagnostic GG1, (OR 1.43 per 0.10, p = 0.023), but not GG2 disease. Time to biopsy was not independently associated with upgrade. Decipher score remained predictive of upgrade with time to biopsy incorporated in our model (OR 1.36 per 0.10, p = 0.021).
CONCLUSIONS: Increasing Decipher scores was associated with greater odds of upgrade among men AS for PCa. The lower spectrum of Decipher scores among men on AS may suggest that distinct categories may be appropriate for risk grouping in this population. 
Back to 2021 Abstracts