Encrypted login | home

Program Information

Validation of a Patient-Specific Monte Carlo Targeted Radionuclide Therapy Dosimetry Platform

A Besemer

A Besemer* and B Bednarz , University of Wisconsin, Madison, WI


SU-E-T-345 Sunday 3:00PM - 6:00PM Room: Exhibit Hall

Purpose: There is a compelling need for personalized dosimetry in targeted radionuclide therapy given that conventional dose calculation methods fail to accurately predict dose response relationships. To address this need, we have developed a Geant4-based Monte Carlo patient-specific 3D dosimetry platform for TRT. This platform calculates patient-specific dose distributions based on serial CT/PET or CT/SPECT images acquired after injection of the TRT agent. In this work, S-values and specific absorbed fractions (SAFs) were calculated using this platform and benchmarked against reference values.
Methods: S-values for 1, 10, 100, and 1000g spherical tumors with uniform activity distributions of I-124, I-125, I-131, F-18, and Ra-223 were calculated and compared to OLINDA/EXM reference values. SAFs for monoenergetic photons of 0.01, 0.1, and 1 MeV and S factors for monoenergetic electrons of 0.935 MeV were calculated for the liver, kidneys, lungs, pancreas, spleen, and adrenals in the Zubal Phantom and compared with previously published values. Sufficient particles were simulated to keep the voxel statistical uncertainty below 5%.
Results: The calculated spherical S-values agreed within a few percent of reference data from OLINDA/EXM for each radionuclide and sphere size. The comparison of photon SAFs and electron S-values with previously published values showed good agreement with the previously published values. The S-values and SAFs of the source organs agreed within 1%.
Conclusion: Our platform has been benchmarked against reference values for a variety of radionuclides and over a wide range of energies and tumor sizes. Therefore, this platform could be used to provide accurate patient-specific dosimetry for use in radiopharmaceutical clinical trials.

Contact Email: