3365 - Prognostic Impact of Grade Group Discordance between MRI-Targeted and Systematic Biopsy Cores in Localized Prostate Cancer
Presenter(s)
H. J. Ryu1,2, M. P. Dykstra1,2, D. Elliot2, R. T. Dess1, W. C. Jackson1, K. D. Stensland3,4, and A. K. Bryant1,2; 1Department of Radiation Oncology, University of Michigan, Ann Arbor, MI, 2Department of Radiation Oncology, Veterans Affairs Ann Arbor Healthcare System, Ann Arbor, MI, 3Department of Urology, University of Michigan, Ann Arbor, MI, 4Department of Urology, Veterans Affairs Ann Arbor Healthcare System, Ann Arbor, MI
Purpose/Objective(s): MRI-guided targeted biopsy (TBx) combined with systematic biopsy (SBx) increases detection of clinically significant prostate cancer (Grade Group [GG] =2) and frequently results in grade upgrading. Current NCCN risk stratification—and thus treatment selection—is based on the single highest GG across all cores. However, it remains unclear whether patients with discordant grading between TBx and SBx cores have similar oncologic risk as those with concordant maximum grade. We evaluated the prognostic significance of TBx-SBx GG discordance on near-term oncologic outcomes in a large national VA cohort.
Materials/Methods: We identified 3,564 patients with localized (cN0M0) prostate cancer diagnosed by combined SBx and TBx within the VHA (2013–2025) who were treated with radical prostatectomy (RP) or definitive radiotherapy (RT). Patients with maximum GG1 across all cores were excluded as discordance is impossible for these patients. Discordance was defined as TBx GG > SBx GG or vice-versa. Random-effects Cox regression evaluated associations between discordance and biochemical recurrence (BCR); random-effects logistic regression assessed adverse pathology (pT3a, pT3b, or pT4) in surgical patients. Models adjusted for NCCN risk group (assigned based on maximum GG across all TBx and SBx cores), age, race/ethnicity, calendar year, and hospital-level clustering. Sensitivity analyses adjusted for individual NCCN components (baseline PSA, percent positive cores, clinical tumor stage, and maximum GG) instead of risk group.
Results: Among 2,194 RT and 1,370 RP patients (median age 69; 40% non-Hispanic Black; median PSA 7.2 ng/mL), 55% had GG discordance (22% TBx > SBx; 33% SBx > TBx). Median follow-up for BCR was 31 months (RT) and 35 months (RP). Discordance was more common for patients with cT1c tumors (compared to high clinical T stages) and larger prostate volumes. After adjustment, discordance was independently associated with lower odds of adverse pathology at RP (aOR 0.60, 95% CI 0.47–0.76, p<0.001) and lower risk of BCR following RP (aHR 0.81, 95% CI 0.66–0.99, p=0.037). In the RT cohort, discordance was similarly associated with reduced BCR risk (aHR 0.58, 95% CI 0.38–0.90, p=0.016). Findings were consistent when adjusting for individual NCCN components instead of NCCN risk group. Each pattern of discordance (TBx > SBx vs. SBx > TBx) was associated with a similar prognostic effect.
Conclusion: In this large national cohort, TBx-SBx biopsy grade discordance was independently associated with more favorable short-term oncologic outcomes after surgery or radiation. These findings suggest that reliance on the single highest GG across all cores may overestimate oncologic risk in discordant cases and, if confirmed with longer follow-up, warrant reconsideration of risk stratification and treatment intensity in this population.