Ultra-low dose Na¹⁸F using a digital PET/CT in a preclinical phase I study - How low can we go?

Objectives: Next-generation digital photon counting PET (dPET) enables Na¹⁸F ultra-low dose while still advancing image quality, and/or reducing acquisition time compared to current standard-of-care PET. We have previously studied the use of a standard dose (3 mCi Na¹⁸F) in this preclinical model. We customized acquisitions by optimizing high definition reconstructions. We have then pushed these optimizations to adjust a Na¹⁸F ultra-low dose protocol in our canine model. The use of PET/CT in animals nowadays is a growing image modality in both the veterinary community (animal patients) and the research community (animal subjects). Therefore, it is also necessary to define and reduce PET tracer doses to reach as low as reasonably achievable (ALARA) levels. The aim of the study was to assess the feasibility of ultra-low Na¹⁸F in a translational preclinical model, and to demonstrate that Na¹⁸F ultra-low dose could provide diagnostic acceptable image quality and reasonable quantification. Methods: Eight adult male beagles (weight (kg) mean ± SD; 14± 1.54) were scanned. The dogs were administered 3 different Na¹⁸F doses: 1 mCi (standard dose/SD), 0.1 mCi (low dose/LD), and 0.05 mCi (ultra- low dose/ULD). Imaging started ≃ 45 min post injection for 33 min total acquisition time. Covering the whole body, 11 bed positions, acquiring 180 seconds per bed position. All imaging was performed on a digital photon counting system (Philips Vereos). PET list mode data were reconstructed using Time-of-Flight with 1mm³ (ultra-high definition/UHD) voxel volumes. Three iterations and 3 subsets. Point spread function and Gaussian filtering were applied. Two experienced blinded readers evaluated overall quality, tissue characterization, and quality of background in the whole body skeleton. Outcome parameters included SUV_max from three-dimensional (3D) regions of interest (ROI) traced over the distal femur, first lumbar vertebra, and a portion of the liver. Results: All the scans and reconstructions were successfully completed in all subjects. Reducing Na¹⁸F dose from the standard dose (1 mCi) performed in previous studies, to the ultra-low dose/ULD (0.05 mCi), demonstrated acceptable image quality and quantitative accuracy. When comparing the impact of reconstruction, HD and UHD continued to demonstrate improved image quality even on the LD and ULD dose level scans. This was also found for the qualitative overall and quantitative ROI assessment of the healthy canine skeletons. Conclusion: Ultra-low dose Na¹⁸F at a level of 50 microCi for a 14 kg canine appears to be diagnostically feasible and a robust option to reduce radiotracer doses in a translational animal model using a dPET system. It allows us to move preclinical nuclear medicine imaging forward with substantial reduced exposure levels while preserving image quality. This also reduced radiation exposure to staff and accompanying persons in accordance with ALARA.