Author: Mercy Iyabode Akerele, Yuni K. Dewaraja, Richard Laforest, Justin K. Mikell, Wendy Siman, Allan Thomas, Daniel Thorek 👨🔬
Affiliation: University of Michigan, Washington University School of Medicine in St. Louis, Washington University School of Medicine, University of Colorado Anschutz Medical Campus, School of Medicine, Washington University in St. Louis, Mallinckrodt Institute of Radiology, Washington University School of Medicine 🌍
Purpose: Post-therapy quantitative 90Y SPECT and PET measurements show ~25% lower activity than vendor-reported values for glass microspheres. We hypothesize this discrepancy is due to bremsstrahlung production differences between aggregated microspheres in vendor vials and 90YCl3 solution used for calibration. This study uses simulations and experiments to assess how 90Y microsphere distribution affects bremsstrahlung production and dose calibrator measurements.
Methods: We used a 5-GBq 90Y Therasphere vial and acrylic pig to evaluate two geometries: 1) the original vial with aggregated microspheres and an added gel layer and 2) a configuration with microspheres suspended in gel, achieved by mixing the original vial. X-ray imaging confirmed microsphere distribution in each geometry. Activities were measured using a Capintec CRC-25W dose calibrator, and planar images were acquired with a Siemens SymbiaT gamma camera using MELP and an open energy window. Measured activities were compared to Monte Carlo simulations and published biases. Simulations were performed using EGSnrc/egs++, incorporating internal bremsstrahlung and assuming a sphere packing ratio of 0.64. Cross sections were generated for packed and dispersed 90Y glass microsphere geometries for comparison.
Results: Initial x-rays of settled microspheres showed packing at the vial's bottom, while imaging in-suspension revealed dispersion with no visible aggregated microspheres.
Measured 90Y activity in the packed geometry exceeded the dispersed geometry by 16%. Similarly, planar image counts were 26% higher in the packed setup. Simulated 90Y activity showed a 34% increase in the packed versus dispersed geometry. Minimal discrepancies (≤1.5%) were observed between simulated dispersion in water and gel.
Conclusion: Dispersing 90Y microspheres reduced dose calibrator and gamma camera measurements, consistent with the hypothesis that dispersion decreases bremsstrahlung yield by lowering the effective atomic number of the surrounding matter. Future efforts will aim to reconcile vendor-reported and actual 90Y activity through improved activity standards, calibration corrections, and sterile, non-invasive assay methods.