Association between PET-based TheraP eligibility and 177Lu-PSMA-617 (LuPSMA) outcomes in VISION-eligible patients with metastatic castration-resistant prostate cancer (mCRPC).

Abstract

41 Background: The optimal threshold based on pre-treatment PET for patient selection for LuPSMA is yet to be defined. TheraP trial applied a high threshold for PSMA-expression on PSMA PET and utilized FDG PET to exclude FDG-avid/non-PSMA-avid disease while the VISION trial used only PSMA PET with a lower threshold of PSMA-expression. The stricter criteria in TheraP trial doubled the screening failure rate compared to VISION trial, 28% vs. 12.6%. In this study among VISION-eligible patients, we aimed to compare the outcome of those who were TheraP-Eligible (TheraP-E) to TheraP-Ineligible (TheraP-I). Further, the prognostic value of FDG-PET phenotyping is investigated. Methods: This retrospective cohort study evaluated consecutive VISION-eligible mCRPC patients who received at least one cycle of LuPSMA and had both PSMA- and FDG-PET within one month of each other between June 2022 and January 2023. Patients were blindly classified as TheraP-E vs. TheraP-I. We compared whole-body semiquantitative parameters including total tumor volume (TTV), SUVmean and SUVmax and outcomes including PSA decline of ≥50% relative to baseline (PSA50), PSA-progression-free survival (PSA-PFS), and overall survival (OS) between the groups. Separately, the patients were dichotomized to having dominant FDG-avid disease (uptake > liver in most sites of disease) vs. non-FDG-avid disease and OS were compared by Cox-regression. Results: 46 patients (median age: 72, ECOG 0-2: 89%, visceral disease: 28%) were included with a median follow-up of 10 months (IQR: 8-14). 14/46 (30%) patients assessed as TheraP-I and had lower PSMA-PET SUVmean and SUVmax compared to TheraP-E patients (5.6 vs. 8.8 and 23 vs. 53, p < 0.001 for both) while other PET parameters including PSMA TTV and FDG parameters (SUVmax, SUVmean and TTV) were not statistically different. TheraP-I patients had lower PSA50 response (21% vs. 56%, p = 0.029) but PSA-PFS and OS were not significantly different compared to TheraP-E patients (Table). 37/46 (80%) patients assessed as FDG-avid had an increased risk of death compared to non-FDG-avid counterparts (Table). Conclusions: In a VISION-eligible population receiving LuPSMA, TheraP imaging-based ineligibility was associated with worse PSA response but with no significant difference in PSA-PFS or OS. Clinical investigation of the more permissive VISION inclusion criteria and the importance of FDG-avid disease using a larger sample size is warranted. [Table: see text]