Work

RECIP 1.0 more predictive of overall survival than PSMA PET progression criteria in biochemically recurrent prostate cancer

Year: 2025

Type: article

Abstract: PURPOSE: Up to 40% of prostate cancer (PCa) patients experience biochemical recurrence (BCR) after primary treatment, but reliable tools for risk stratification in this setting are limited. This study compared Response Evaluation Criteria in PSMA PET/CT (RECIP 1.0) and PSMA PET Progression (PPP) criteria in predicting overall survival (OS) and prostate-specific antigen progression-free survival (PSA-PFS). As a secondary aim, PROMISE-based nomograms were assessed as tools for OS prediction. METHODS: Ga]Ga-PSMA-11 PET/CT was analysed. Disease progression was defined by RECIP 1.0 and three PPP variants (PPP-Volume, PPP-Mean SUV, PPP-Max SUV). Associations with OS and PSA-PFS were tested using Kaplan-Meier and Cox regression models. PROMISE-based visual and quantitative nomograms were applied to follow-up scans to predict OS. Subgroup analyses by treatment type were performed for RECIP and PPP criteria. RESULTS: RECIP-defined progressive disease (RECIP-PD) was most strongly associated with OS, identifying patients at highest risk of death (median OS 53.2 months; HR 4.62; 95% CI: 2.41-8.87; p < 0.001). PPP-based criteria were also prognostic for OS, and were the only framework associated with PSA-PFS. PROMISE-based nomograms were predictive of OS, with performance comparable to PPP criteria. By treatment type, RECIP-PD performed best in androgen deprivation therapy, PPP-Max SUV in radiotherapy, and PPP-Volume in combined ADT + radiotherapy. SUV-based PPP criteria were most predictive of PSA-PFS in combined ADT + radiotherapy. CONCLUSION: In BCR PCa, RECIP 1.0 best predicts OS, PPP criteria better predict PSA progression, and PROMISE nomograms provide a single-timepoint approach to OS risk stratification. TRIAL REGISTRATION NUMBER: ACTRN ACTRN12615000608561. Registered 11 June 2015. Retrospectively registered.

Source: European Journal of Nuclear Medicine and Molecular Imaging

DOI: 10.1007/s00259-025-07592-6