Comparison of PSA, FDG-PET, and Prostate Cancer Working Group 2 (PCWG2) criteria to interpret apparent progression on first post-treatment bone scan.

Abstract

20 Background: The risk of prematurely discontinuing effective therapy in patients with metastatic castrate resistant prostate cancer (mCRPC) because of apparent initial progression on bone scan has repercussions for drug development and clinical practice. We evaluated three methods to distinguish disease progression (POD) from non-progressive bone disease or flare response. Methods: The dataset was comprised of men with mCRPC enrolled in contemporary clinical trials using AR-directed or targeted therapy. To be included, a worsened bone scan (increased size, intensity, or number of lesions) at the time of initial follow-up at 8-12 wks (FU1), a concurrently performed FDG-PET and PSA, and a second follow-up bone scan (FU2) > 6 wks after FU1 were required. Pts were evaluated by three methods: 1) PSA-guided approach: POD was defined as a PSA increase >25% from baseline at FU1; for non-progressors, a special category was created for flare response defined as a PSA decline >50% and a FU2 bone scan documenting stability/improvement; 2) FDG-PET approach: POD was defined as a new bone lesion or increase in SUV >10% at FU1; 3) Bone scan only approach: POD was defined by PCWG2 requiring 2 new bone scan lesions at FU1 and 2 additional lesions at FU2. Results: 66 pts registered to trials conducted between 2007-2011 were examined; 38/66 (57.6%) pts had a worsened bone scan at FU1, 23 of whom had a FU2 bone scan and were considered evaluable for this analysis. Conclusions: Over half of mCRPC pts have a worsening bone scan during the first three months of therapy. PCWG2 controls for flare without use of PSA, confirms radiographic POD, and maintains pts on study longer than use of early post-treatment PSA changes. FDG-PET at first assessment appears to identify an identical percentage of progressors as PCWG2 did at the second assessment. PCWG2 is undergoing prospective validation in phase III trials; the use of FDG PET as an early response biomarker is currently under investigation. [Table: see text]