Author: Spencer Behr, Joseph Grudzinski, Youngho Seo, Jaehoon Shin, Yiran Wang, Frederick J. Wilson 👨🔬
Affiliation: University of California San Francisco, University of California, San Francisco, Voximetry, Inc., Voximetry, Inc 🌍
Purpose: Voxel-level imaging-based dosimetry enables a significant improvement in accurate and personalized treatment planning in radiopharmaceutical therapy, potentially leading to optimized disease management. The imaging data used for dosimetry is critical as it directly affects the dose evaluation and depends on the methods applied during image reconstruction. This study investigated the impact of time-of-flight (TOF) and motion correction (MC) on PET imaging-based dosimetry and their effects on dose quantification in 90Y radioembolization for hepatocellular carcinoma (HCC).
Methods: Six subjects with HCC underwent 90Y radioembolization of 2.8±1.1 GBq and had PET/CT imaging approximately 24 hours post-90Y administration. PET images were reconstructed using three approaches: (1) without TOF, (2) with TOF, and (3) with TOF combined with data-driven respiratory gating for MC (Fig. 1). Imaging-based dosimetry was performed using the Monte Carlo-based software, Torch® (Voximetry, Madison, WI). Voxelized dose distributions generated from different PET reconstructions were compared. A consistent region of interest (ROI) was placed for each subject to assess the mean dose in the overall perfused liver area.
Results: The mean dose of the overall liver perfused area was estimated to range 200-500 Gy (Fig. 2). We observed a 7.7%±2.0% dose difference (range: 5.2%-10.1%) with and without TOF, and a 5.2%±2.2% difference (range: 1.9%-8.3%) with and without MC (Fig. 3). These discrepancies were further validated by examining the dose maps generated from different PET images (Fig. 4), highlighting variations in both high- and low-dose regions.
Conclusion: Preliminary results reveal that PET reconstruction methods, including time-of-flight (TOF) and motion correction (MC), can affect dosimetry in 90Y radioembolization both at regional and voxel levels. These findings highlight the need for standardized protocols to ensure reliable dose quantification. Future investigations aiming to optimize image reconstruction parameters for accuracy will be pivotal for modernizing dosimetry and personalized treatment using 90Y radioembolization.