A randomized trial comparing fluorocholine-PET/CT with conventional imaging in prostate cancer.

• DOI: 10.1200/jco.2019.37.7_suppl.2
• Publication Date: 2019-03-07T21:20:31Z
• Authors: Scott Williams; Jean-Mathieu Beau...; Peter Roselt; Kate Moody; Richard Fisher; Elizabeth Drummond; Rodney J Hicks
• Keywords: 03 medical and health sciences; 0302 clinical medicine; 3. Good health

2 Background: We conducted a randomised trial comparing 18Flourocholine-PET/CT (FCH) to Computed Tomography (abdomen and pelvis) plus 99mTc-Whole Body Bone Scan (Conventional Imaging [CIm]) to determine imaging performance in prostate cancer (PC). Methods: This prospective two-arm 1:1 randomised trial enrolled men with newly diagnosed or biochemically recurrent PC to first-line imaging (FLI) with either CIm or FCH. Participants without evidence of metastases proceeded to second-line imaging (SLI) using the alternative imaging strategy. The primary aim was to determine whether FCH was more effective as a FLI approach in changing management. Secondary endpoints included incremental utility of SLI and negative predictive value (NPV) based on progression-free survival (PFS). Australian New Zealand Clinical Trials Registry ACTRN12608000641392. Results: 108 men were enrolled; 44% were for staging of newly-diagnosed PC and median follow-up 43 months. Imaging impacted clinical management in 32.4% of men (95% CI=23.7-42.1%), mostly with FLI (n=30). High-impact management changes occurred in 27.8% (95% CI=16.5-41.6%) of FCH cases compared with 11.1% (95% CI=4.2-22.6%) in the CIm arm (p=0.032). The final management plan was derived using FCH in 98.1% (95% CI = 90.1-100%) of cases and 92.6% (95%CI = 82.1-97.9%) of CIm cases (p=0.242). FLI with FCH showed unequivocally N1 or M1 disease in 22.2% (95% CI = 12-35.6%), and 16.7% (95% CI = 7.9-29.3%; p= 0.531) of CIm cases. The overall NPV for stage TxN0M0 (from all imaging) was 26.3% (95% CI: 13.9 - 41.2%), with no significant difference between arms (p=0.9). For N1M0 cases, the NPV was 14.3% (95% CI: 7.1 - 35.7%). The identification of N1M0 by FCH resulted in a longer time to identification of progressive disease, with a median PFS of 32 months (95% CI=2-68months) compared with 3 months (95% CI=1-16 months) in the CIm N1M0 cohort (p=0.05). Conclusions: FCH-PET/CT identifies more high-clinical-impact lesions than CIm as first-line imaging. All imaging modalities were poor at predicting subsequent progressive disease. Isolated node-positive disease seen with FCH is associated with a longer time to - but similarly high rates of - recurrence, suggesting a lead-time bias. Clinical trial information: ACTRN12608000641392.