MP53-03: Effect of Targeted Cores on Risk Stratification in the Era of Image-guided Prostate Biopsy

Samuel Washington*, Janet Cowan, San Francisco, CA, Peter Lonergan, Dublin, Ireland, Alexander Bell, Robert Parker, Katsuto Shinohara, Hao Nguyen, Peter Carroll, Matthew Cooperberg, San Francisco, CA

Introduction: Targeted biopsy on transrectal ultrasound (TRUS) or magnetic resonance imaging (MRI) increases detection of high-risk disease, but can oversample indolent lesions. We utilize various weighting strategies to assess the impact of targeted cores on clinical risk stratification for men undergoing image-guided prostate biopsy.

Methods: Men with clinically localized (cT1-2/No/xMo/x) prostate cancer from 2005-2020, who underwent systematic biopsy with additional TRUS and/or MRI guided targeting, and subsequently treated with radical prostatectomy at our institution were included. We estimated tumor volume on biopsy as percentage of positive cores using three weighting strategies for targeted biopsy cores by location: standard percentage of all cores positive vs. limit to maximum 1 core per location vs. limit to 2 cores per location. We then computed the Cancer of the Prostate Risk Assessment (CAPRA) score using each of these approaches. We compared concordance of risk scores and risk groupings (low, intermediate, high) using Bowker’s test of symmetry. Cohen’s kappa coefficients were used to measure inter-rater agreement between systematic cores alone vs. systematic+targeted cores.

Results: At diagnosis 534 men averaged 62.6 years (SD 6.9) with median PSA of 6.4 ng/ml (IQR 4.7-9.1). Most had cT1 (35%) or cT2 (59%) disease, with Gleason Grade Group (GG) 1 (21%) or GG2 (39%) and 20 cores (IQR 18-22) obtained per biopsy session, including 4 targeted cores (IQR 2-6). Median percentage of cores positive was 41% overall (IQR 30-52) and 100% for targeted lesions (IQR 57-100).
Using the standard computation of percentage of positive cores, 18% of CAPRA scores computed from systematic+targeted cores were higher than CAPRA based on systematic sampling only (symmetry p<0.01, weighted K=0.70 (0.65, 0.75)). After limiting the number of positive cores by location, 15% (maximum 1 core) and 20% (maximum 2 cores) of CAPRA scores using systematic+targeted cores were higher than systematic-only CAPRA scores (symmetry p<0.01 for both, K=0.74 (0.69, 0.78) and K=0.79 (0.74, 0.83) respectively.)

Conclusions: Inclusion of targeted cores in clinical risk score results in higher risk estimation, with greatest agreement between systematic and systemic+targeted when weighting to 1 core per targeted lesion, with the lowest percentage of risk estimate elevation. Standardization of incorporating data from targeted biopsy cores is necessary to appropriately estimate clinical risk for men with undergoing an image-guided prostate biopsy.

Source of Funding: UCSF Goldberg-Benioff Program in Translational Cancer Biology