3286 - Clinical vs. MRI Risk Signals In Intermediate-Risk Prostate Cancer Treated with Radiotherapy: Associations with PSA Density and Radiographic ECE/SVI
Presenter(s)
J. Janopaul-Naylor1, S. Lobaugh2, S. Pandya1, D. J. Gorovets1, M. A. Kollmeier3, S. M. McBride1, M. B. Bernstein1, J. Haseltine1, T. P. Kole1, V. S. Brennan1, Z. R. Moore4, B. R. Mychalczak1, B. A. Mueller1, D. M. Guttmann4, A. R. Barsky1, Z. S. Zumsteg5, D. E. Spratt6, Z. Zhang2, S. G. Zhao7, and H. Nagar1; 1Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, 2Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY, 3DepMemorial Sloan Kettering Cancer Center, New York, NY, 4Memorial Sloan Kettering Cancer Center, New York, NY, 5Cedars-Sinai Medical Center, Los Angeles, CA, 6University Hospitals Seidman Cancer Center, Case Western Reserve University, Cleveland, OH, 7University of Wisconsin, Madison, WI
Purpose/Objective(s): Intermediate-risk prostate cancer (IR PCa) is a biologically diverse group ranging from indolent to aggressive disease. Risk stratification for IR PCa historically has relied on PSA, Gleason Score (GS), and DRE. We examined the addition of MR-derived features to established IR PCa clinical risk stratification.
Materials/Methods: A multi-site, single institution retrospective study was conducted including 1,280 men with IR PCa (GS 6-7, PSA <20) who received dose-escalated prostate RT between 1998-2020 with or without concurrent androgen deprivation therapy (ADT) and had a pre-treatment diagnostic MRI. Outcomes included overall survival (OS), distant metastases (DM), and biochemical recurrence (BCR) evaluated using Kaplan-Meier analysis. Univariate (UVA) and multivariable (MVA) cox-proportional hazard models used to evaluate MR derived covariates including: extracapsular extension (ECE) & seminal vesicle invasion (SVI) categorized as no, intermediate (possible/probable) or high risk (definite/gross); composite radiographic risk (CRR) based on highest ECE or SVI grade; and PSA density grouping defined as Low Density (<0.15), Low PSA-High Density (PSA <10 and density >=0.15), and High PSA-High Density (PSA >=10 and density >=0.15).
Results: The median PSA was 6.9 ng/mL (IQR 5.1-10.2), PSA density was 0.20 ng/mL/cc (IQR 0.13-0.30), and follow-up was 7 years (IQR 3-9). Most patients had clinical T1c (61%), <50% cores involved (63%), no ECE (54%), and no SVI (92%) while 49% had GS 4+3, 46% had GS 3+4, and 15% and 32% had CRR high or intermediate, respectively. RT included conventionally fractionated RT (62%), SBRT (33%), and moderately hypofractionated (5%). A brachytherapy boost was used in 22% and 45% of patients received ADT. On UVA, ECE, SVI, CRR, PSA density group, GS, percent positive cores were all significantly associated with DM but not OS. MVA showed GS 4+3 and High PSA-High Density features were independently associated with increased risk of DM. MVA and 10-year BCR, DM, and OS in Table.
Conclusion: PSA density—an NCCN feature in low risk PCa for surveillance—adds prognostic value for men with IR PCa treated with RT. However, there is no strong correlation between radiographic ECE or SVI and DM or BCR. Future studies incorporating dominant intraprostatic lesion (DIL) features and volume Gleason pattern 4 could help refine IR PCa prognostication.
| MVA for DM HR (95% CI) | p-value | 10-yr BCR | 10-yr DM | 10-yr OS | |
| GS 6 | Ref | - | 11% | 1.7% | 71% |
| GS 3+4 | 2.70 (0.66-11.1) | 0.17 | 20% | 5.6% | 78% |
| GS 4+3 | 5.26 (1.28-21.6) | 0.021 | 23% | 14% | 71% |
| <50% Positive Cores | Ref | - | 17% | 6.1% | 75% |
| >=50% Positive Cores | 1.26 (0.80-1.98) | 0.32 | 26% | 13% | 73% |
| CRR Low | Ref | - | 17% | 6.6% | 77% |
| CRR Int | 1.04 (0.61-1.77) | 0.88 | 21% | 8.9% | 72% |
| CRR High | 1.48 (0.87-2.54) | 0.15 | 26% | 12% | 72% |
| Low PSA Density | 0.60 (0.30-1.18) | 0.14 | 13% | 5.1% | 73% |
| Low PSA-High Density | Ref | - | 20% | 7.8% | 75% |
| High PSA-High Density | 2.51 (1.56-4.04) | <0.001 | 29% | 14% | 75% |