Impact of standard-of-truth method on evaluation of a diagnostic PET radiopharmaceutical: Learnings from the phase 3 SPOTLIGHT study.

Abstract

39 Background: SPOTLIGHT (NCT04186845) evaluated newly approved diagnostic PET radiopharmaceutical 18F-flotufolastat (18F-rhPSMA-7.3) in patients (pts) with recurrent prostate cancer (PCa). In the absence of the gold standard histopathology, it can be challenging to verify positive findings from a diagnostic agent under investigation, particularly in recurrent PCa where it is neither feasible nor ethical to obtain multiple biopsies from the same patient. Consequently, conventional imaging is often used as a standard of truth (SoT) despite potential for it to be less sensitive than the investigational agent. Here, we explore the impact of SoT methods used in SPOTLIGHT on the study endpoints. Methods: The SPOTLIGHT coprimary endpoints were patient-level verified detection rate (VDR) and combined region-level positive predictive value (crPPV). Patient-level PPV was an exploratory endpoint. Pts with recurrent PCa underwent PET 50–70 min after IV administration of 296 MBq 18F-flotufolastat. For SoT assessment, the protocol required either histopathology or confirmatory imaging (≤60 and ≤90 days post-PET, respectively). Three blinded readers evaluated the scans, with majority read representing agreement between ≥2 readers. A separate Truth Panel established SoT for all pts. Post-study analysis revealed that, likely due to the study occurring during COVID, some pts had neither histopathology nor post-scan confirmatory imaging. These pts had only baseline/historic imaging as SoT. The present analysis stratified endpoint data according to the SoT method the panel used to verify PET-positive lesions. Results: Of 366 evaluable pts, 69 had histopathology and 297 had only confirmatory imaging for SoT. Of those with an imaging SoT, 90 (30%) had baseline/historic imaging only. As the Table shows, for each endpoint the highest value was obtained when using histopathology as SoT. VDR was ~4-fold higher with histopathology vs baseline/historic imaging, crPPV was 2-fold higher and PPV 2.5-fold. Conclusions: These data illustrate the substantial impact SoT methods can have on clinical trial endpoints and highlight a limitation of evaluating SPOTLIGHT endpoints in a cohort with various SoT methods (Table, column 1).1 Notably lower endpoint values are achieved when including pts with baseline/historic imaging SoT vs gold standard histopathology. 1. J Urol. 2023; 210:299-311. Clinical trial information: NCT04186845 . [Table: see text]