3265 - PSMA-PET Radiopharmaceutical Image Scoring Method (PRISM): A Practical Tool for Predicting ¹⁷⁷lu-PSMA-617 Therapy Outcomes in Metastatic Castration-Resistant Prostate Cancer (mCRPC)

Purpose/Objective(s): Lutetium-177-PSMA-617 (Lu-PSMA), a radiopharmaceutical therapy, targets prostate specific membrane antigen (PSMA). It has shown improved overall survival (OS) and progression-free survival (PFS) in PSMA-PET-positive mCRPC patients (pts), but treatment (Tt) success is variable. No tool currently predicts Lu-PSMA Tt outcome. We developed a novel PSMA-PET Radiopharmaceutical Image Scoring Method (PRISM), utilizing visual assessment of the baseline pre-therapy PSMA-PET and tested its predictive ability.

Materials/

Methods: We included 47 pts who received =1 cycle of Lu-PSMA between July 2022 and January 2024. All pts had a positive baseline PSMA-PET and standard Lu-PSMA Tt. PRISM scores were calculated as the sum of skeletal (range 0-27) and lymphatic-visceral (0-32) metastatic extent. We categorized PRISM scores <10 as low disease burden (LDB) and scores = 10 as high disease burden (HDB) by using the ROC curve for OS and PFS. Primary endpoints, OS and PFS, were analyzed using the Kaplan-Meier method, and univariate comparisons of survival curves were made using the log-rank test. Secondary endpoints included PSA response (PSA 50 =50% decline in PSA level; PSA 90 =90% decline in PSA level), radiographic response by RECIST 1.1 criteria, and side effects (SEs).

Results: Median age at start of Lu-PSMA was 73 years (50-88), median baseline PSA was 24.44 ng/mL (0.05-5580), and median PRISM score was 12 (2-38). Metastatic sites included bone (93.6%), lymph nodes (68.1%), liver (17%), lung (4.3%), and brain (2.1%). Regarding Lu-PSMA Tt: 49% completed all 6 cycles, 23.4% Tt in progress, 27.7% discontinued early and 12.7% required dose reduction. One-year OS and PFS were 100% and 42.8% for pts with LDB, and 62.3% and 12.9% for HDB, respectively (OS; p=0.05 and PFS; p= 0.03). Secondary endpoint analysis (Table 1) revealed better PSA and radiological responses in pts with PRISM<10. The 3 most common SEs were fatigue (27.7%), nausea (17%), and dry mouth (12.8%).

Conclusion: PRISM is a practical and broadly adaptable tool for dichotomizing disease burden in mCRPC pts. Our findings show that pts with LDB respond better to Lu-PSMA Tt than those with HDB. We intend to validate these findings in larger cohorts and investigate this tool to determine the correlation between the Tt response and the scoring continuum.

Abstract 3265 – Table 1