Abstract
Objective:
Using 153Sm-EDTMP therapy for bone metastases from nasopharyngeal cancer (NPC), we attempted a real-time three-dimensional (3D) dose calculation based on the S-value for voxels. The 3D radionuclide uptake data on SPECT images were factored into the precalculations by the Monte Carlo (MC) method.
Methods:
For the nuclide 153Sm, the S-value for voxels of size 3.45 × 3.45 × 3.45 mm3 in a soft tissue phantom was precalculated on a self-developed program built from the MC program EGS4. Based on the SPECT/CT image of the patient, the 3D dose rate distributions were calculated with the S-value method and compared with the direct MC calculation results. The total volume of the whole calculation region and the bone were chosen as the regions of interest (ROIs); additionally, the dose rate volume histograms (DVHs) for the ROIs were also calculated.
Results:
The iso-dose was administered based on the scan images obtained at 6 hours following injection with the nuclide. In the calculation region, the maximum dose rate was 5.98 × 10−5 and 6.26 × 10−5 mGy/(MBq S) for S-value and direct MC calculations, respectively. Once the dose rate was normalized to the maximum dose rate point, the iso-dose curves for both methods of calculation were similar in most regions and reasonably matched the functional image as well. The DVHs of the ROIs indicated that the dose rate distributions were nonuniform, that is, ∼30% of the bone-ROI volume received 10% of the maximum dose rate; however, only 3.8% volume received 50% of the maximum dose rate.
Conclusions:
Using 153Sm-EDTMP therapy for bone metastases from NPC, we attempted to supplement the current dosimetry work at the image-level by a pragmatic and real-time dosimetry calculation based on S-value and functional imaging. More accurate dose calculations for patients undergoing radionuclide therapy will depend on the development of higher image resolution in nuclear medicine and warrant further studies to optimize the pharmacokinetics model.
Get full access to this article
View all access options for this article.
