KAHIBARO
36.5. Radionuclide Therapy
Table of Contents
I-131
Iodine 131 is one of the most widely used therapeutic radionuclides in nuclear medicine and a central example for internal dosimetry. It emits both beta particles, responsible for most of the local dose, and relatively high energy gamma rays, which contribute to dose at a distance and enable imaging.
From a simulation point of view, I 131 is characterized by its half life, radiation emissions, and typical clinical biodistribution. The physical half life is about 8 days, so activity changes noticeably over typical treatment times. In GATE based internal dosimetry, this time dependence is represented either by explicitly simulating several time points with different activities or by using time dependent source definitions that follow the decay law.
The beta spectrum of I 131 is continuous and extends up to a maximum energy of roughly 600 keV. These beta particles deposit most of their energy within a few millimeters in soft tissue, which means that absorbed dose is relatively localized around the source voxels, but not entirely confined to them. In contrast, the accompanying gamma emissions, especially the 364 keV line, have much longer range and can contribute to cross organ dose and to imaging.
When you define an I 131 source for therapy simulations, you typically start from an activity map that represents the distribution of the tracer in the body. For thyroid therapy, this may be concentrated in the thyroid gland, while for radioimmunotherapy or metastatic disease it may be distributed in multiple organs and lesions. In GATE, such maps are commonly implemented as voxelized sources where each voxel carries an activity proportional to a clinical image, for example a SPECT activity distribution.
Internal dosimetry with I 131 requires careful treatment of energy deposition and transport physics. The beta particles must be transported accurately at low energies to obtain realistic dose kernels in tissue. At the same time, gamma transport must be modeled to capture dose contributions in other organs that do not themselves take up I 131 but are irradiated by photons. This dual nature is important when you interpret organ dose and when you compare with imaging based dose estimates.
Because of I 131 gamma emissions, a single simulation can often serve two purposes. The same underlying Monte Carlo run can produce 3D dose distributions using dose actors and also phase space or detector data that approximate therapeutic imaging. This link between imaging and dosimetry is one of the reasons why I 131 is frequently used in validation studies for internal dosimetry tools.
In clinical practice, I 131 therapy sometimes uses simple administered activity schemes, but patient specific dosimetry is increasingly recommended. In simulation based dosimetry, you typically normalize the voxelized source to a measured time activity curve, then compute absorbed dose per unit administered activity. The Monte Carlo results are then compared with analytical absorbed fraction models or with standardized S values to check consistency.
For I 131 dosimetry, it is essential to model both beta and gamma emissions. Neglecting gamma transport will underestimate cross organ dose, while ignoring beta transport details will distort local dose gradients around source regions.
Lu-177
Lutetium 177 is a beta and gamma emitter that has become a standard radionuclide for targeted radionuclide therapy, especially for treatments using peptides and antibodies. Compared with I 131, Lu 177 has a lower beta energy and a somewhat shorter range in tissue, which makes it well suited for controlled dose deposition at the scale of small tumors and organs.
The physical half life of Lu 177 is about 6.7 days. As for I 131, GATE based simulations usually do not simulate the full decay over days in real time, but instead represent the time integrated effect by choosing an appropriate number of decays or by combining several simulations corresponding to different time points of the time activity curve. The time dependence is then handled analytically outside the Monte Carlo, or through time dependent source definitions.
The beta spectrum of Lu 177 has mean energies around a few hundred keV and a maximum below 500 keV. The resulting penetration in soft tissue is typically in the millimeter range. This makes Lu 177 particularly important for voxel level dosimetry, where you are interested in how dose spills from one voxel to neighboring voxels and how small lesions are irradiated by their own uptake and by surrounding activity.
Lu 177 also emits photons, most notably around 113 keV and 208 keV. These are of sufficient energy to be used for SPECT imaging and also contribute to absorbed dose outside the source region. In internal dosimetry simulations, it is common to focus primarily on the beta component for local dose within organs or tumors, while still including gamma transport if you are interested in whole body dose or dose to organs far from the source.
A typical GATE simulation of Lu 177 therapy begins from patient specific activity distributions obtained from SPECT or hybrid SPECT CT imaging. These images are discretized into voxels, and each voxel is assigned a radionuclide concentration and a corresponding set of materials from CT based material mapping. The Lu 177 emission spectrum is then used to generate particles in each voxel according to its activity, and dose actors score the energy deposition per voxel.
Because Lu 177 is often used in small structures, numerical details become more important. Voxel sizes are often chosen relatively small so that the dose gradient across small tumors or kidney cortex regions is resolved. Range cuts and production thresholds in the physics configuration must be adapted so that secondary electrons are produced and transported at realistic spatial scales. If cuts are too large, energy deposition will be artificially localized and neighboring voxels will receive incorrect dose.
In research and development, Lu 177 is also used to benchmark dose calculation methods and to compare full Monte Carlo with faster convolution or kernel based approaches. Monte Carlo simulations with GATE can produce detailed point dose kernels for Lu 177 in water or tissue like materials. These kernels are then used to build analytical or semi analytical dosimetry tools that approximate Monte Carlo results for larger patient datasets.
For Lu 177 therapy simulations, small voxel sizes and appropriate production cuts are crucial. If voxel dimensions are comparable to or larger than the Lu 177 beta range, coarse discretization and too aggressive cuts can lead to significant errors in estimated dose gradients in and around small lesions.
Y-90
Yttrium 90 is a pure beta emitter used widely in radionuclide therapy, especially in selective internal radiation therapy for liver tumors using microspheres. Unlike I 131 and Lu 177, Y 90 does not produce significant gamma emissions suitable for imaging. Its high beta energy and long range in tissue make it a distinct case in internal dosimetry.
The physical half life of Y 90 is about 64 hours, or approximately 2.7 days. The beta spectrum is much harder than for Lu 177 or I 131, with a maximum energy of about 2.3 MeV and a mean energy around 0.9 MeV. These high energy electrons have ranges of several millimeters in soft tissue and can extend up to about 1 centimeter. As a result, dose deposition is considerably more spread out around the source regions than for lower energy radionuclides.
In GATE simulations, Y 90 therapies are typically modeled with voxelized sources that represent the distribution of microspheres or Y 90 labeled agents in the liver or other organs. Because the beta range is long, the spatial resolution of the dose is governed more by the beta range than by the voxel size once voxels are sufficiently fine. Nonetheless, voxels must be chosen small enough to represent heterogeneous tumor and normal tissue regions that can be only a few millimeters in size.
Since Y 90 has negligible gamma yield, imaging for treatment planning usually relies on surrogate tracers or bremsstrahlung imaging. Monte Carlo simulations can be used not only for dose but also to model bremsstrahlung photons generated by Y 90 beta particles in tissue or in surrounding materials. This can be relevant if you want to understand or optimize imaging protocols or account for bremsstrahlung dose to tissues outside the main target volume.
For pure dosimetry, however, you often focus on electron transport. The high beta energy means that you must use physics models that are accurate in the MeV range and pay attention to step size and multiple scattering parameters. Inadequate transport parameters can lead to incorrect lateral spread of electrons and biased depth dose profiles. In a Y 90 liver therapy simulation, for example, these details influence how much dose spills from lesions into normal liver tissue and into surrounding organs.
In selective internal radiation therapy, Y 90 microspheres are deposited non uniformly in the vascular structure of the liver and tumors. Internal dosimetry with GATE can use detailed micro distributions if available, or can approximate them using coarser distributions derived from imaging. For microdosimetric studies, you may construct small scale geometries that resolve individual vessels or clusters of microspheres and score dose at sub millimeter or even cellular scales.
In clinical scale simulations, Y 90 is often used as a key example for the concept of cross organ dose and non local energy deposition. Because electrons travel several millimeters, a substantial fraction of the absorbed dose in normal liver can originate from activity in adjacent tumor voxels, and vice versa. This contrasts with very short range emitters, for which dose is more tightly confined to the source voxels.
Y 90 has a high energy beta spectrum and negligible gamma emission. For accurate dosimetry you must use appropriate electron transport physics and transport parameters, and you must not assume that dose is confined to source voxels. Ignoring the extended beta range will underestimate dose to tissues surrounding the activity distribution.
Views: 12