MONTE CARLO SIMULATIONS IN RADIO-LIGAND THERAPY WITH LU-177 FOR NEUROENDOCRINE TUMOURS (NETS)

Background: This work aims to apply Monte Carlo (MC) simulations, running in GATE (a GEANT4-based MC simulator), to study dose rate distribution at the voxel level by comparing dose rate curves obtained with GATE with those obtained using the convolution approach. We followed this approach for an ensemble of 7 NET patients treated with Lu-177-DOTATOC. Material and Methods: S-values to be used in the convolution method were calculated by simulating Lu-177 decay in a water box. Then, results have been compared with other works in literature. Then, we calculated the dose rate curves after the convolution of the dose voxel kernel (DVK) with patients’ SPECT images within some regions of interest (ROIs). These results were compared with dose rate curves obtained by performing the full MC simulation of Lu-177 decay in the actual patient’s human body. Preliminary Results: The voxel dose rates compare quite well with MC dose rates. In all cases, we noticed the curves are a linear combination of two decreasing exponentials. Moreover, the MC dose rates obtained 1-hour after radiopharmaceutical injection are lower than those obtained with the convolution approach. This fact leads to an underestimation of the total absorbed doses in ROIs. Both the voxel dosimetry approach and Monte Carlo simulations return comparable results. Of course, in the clinical daily routine, the voxel dosimetry approach is more feasible than MC approach because the last one is highly time-consuming.