3256 - Re-Defining Risk Stratification in High-Risk Prostate Cancer: Integrating Biopsy Tumor Volume for Improved Prognostic Accuracy Beyond NCCN Criteria

12:45pm - 02:00pm PT

Hall F

Screen: 15

POSTER

Presenter(s)

Abed Rahman Kawakibi, MD - Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL

A. R. R. Kawakibi¹, A. Shimomura², M. Arshad¹, A. A. Solanki³, and S. Liauw¹; ¹Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL, ²Cedars-Sinai Department of Radiation Oncology, Los Angeles, CA, ³Department of Radiation Oncology, Stritch School of Medicine, Cardinal Bernardin Cancer Center, Loyola University Chicago, Maywood, IL

Purpose/Objective(s): Accurate risk stratification is essential to optimize treatment of high-risk prostate cancer treated with radiation therapy (RT). We explored the potential utility of including biopsy tumor volume categorized by Gleason grade group (GG), in addition to the traditionally accepted NCCN risk classification criteria.

Materials/Methods: 298 men with NCCN high-risk (HR) or very-high risk (vHR) prostate cancer treated with RT between 2005 and 2021 were retrospectively reviewed. 232 (78%) patients had pathology reports with detailed biopsy data, including a maximum GG of 1 (3%), 2 (11%), 3 (17%), 4 (35%), or 5 (34%). 93% of men received hormonal therapy for a median length of 20 months and 7% additionally received abiraterone. Median RT dose was 79.2 Gy/1.8 Gy and 17 (7%) patients received a brachytherapy boost. Biopsy data were systematically coded, capturing total millimeters (mm) of each GG category. The final analysis excluded MRI-directed biopsy core data to normalize the assessment of tumor burden. Univariate (UVA) and multivariate analyses (MVA) were performed to evaluate the effects of NCCN risk factors and biopsy core data on freedom from biochemical failure (FFBF) and distant metastasis (FFDM). Median follow-up was 52.1 months.

Results: The median biopsy total tumor volume was 33.5 mm (IQR, 15-64). 5-y FFBF and FFDM were 78%, and 86%, respectively. On UVA, biopsy tumor volume =15 mm was significantly associated with inferior FFBF (5-y 75% vs. 92%, p=0.01) and FFDM (5-y 83% vs. 96%, p<0.01). Associations of biopsy tumor volume with outcomes were weaker when analyzing by individual GG (including total mm of GG4 or GG5). Additionally, there was no minimum amount of GG4, GG5, or GG4-5 disease by which a more favorable risk category could be identified. T3 stage and GG5 disease were strongly associated with FFBF and FFDM (all p<0.01) whereas PSA (>20 and >40), as defined by NCCN for HR and vHR patients, was not associated with outcomes (all p>0.3). On MVA, biopsy tumor volume =15 mm was independently associated with FFDM (RR 6.1, p=0.02) when controlling for NCCN very high-risk disease (RR 3.7, p=0.02), GG5 (RR 3.6, p<0.01), and T3 stage (RR 7.7, p<0.01). Updating the NCCN vHR definition (at least 2 of 3: PSA>40, T3-4, or GG4/5) with 1) biopsy total tumor volume =15 mm, 2) T3, or 3) GG5 resulted in improved risk stratification; this alternate definition yielded a 5-y FFDM of 96% vs. 77% (p<0.01) for HR and vHR patients, respectively, compared to the NCCN definition of 89% vs 81% (p=0.06).

Conclusion: Total tumor volume on biopsy cores may provide additional prognostic value compared to the NCCN v.1.2025 stratification for HR and vHR prostate cancer patients. Our findings suggest more limited utility of PSA thresholds (>20 and >40) and support the study of tumor volume on biopsy in future risk assessment models.