Voxel-wise preclinical dosimetry of 177Lu-NM600 using the Monte Carlo-based dosimetry platform RAPID

Introduction: Accurate estimation of absorbed doses to normal organs and tumors is important for establishing dose-response models in radiopharmaceutical therapy (RPT). In this regard, RAPID (Radiopharmaceutical Assessment Platform for Internal Dosimetry), a subject-specific Monte Carlo internal dosimetry platform, has been developed and utilized for various purposes in preclinical and clinical practice. In this study, the RAPID platform was expanded to utilize high-resolution (0.08 mm voxels) micro-SPECT/CT images to perform dosimetry in mice with intracranial brain tumors treated with RPT.

Methods: C57BL/6 mouse was orthotopically injected with 20k mouse GL261 glioma cells at 1 μL per minute using a stereotactic apparatus. Verification of the brain tumor growth was monitored using both T1-Gd and T2 MRI imaging after 2 weeks of the injection. A theragnostic agent ¹⁷⁷Lu-NM600 was administered to the corresponding mouse intravenously by the tail vein as 18.6 MBq (503.9 μCi). The micro-SPECT scans were acquired in a MiLabs micro-SPECT/CT scanner at 4 different timepoints as 7, 29, 51, and 76 hours post-injection of the agent. Based on the acquired SPECT and CT data, the Monte Carlo simulation was conducted in RAPID by utilizing the MATLAB and AMIRA software was used to perform image co-registration between SPECT (source) and CT (geometry) images. Monte Carlo simulations were conducted using Geant4 by utilizing the co-registered SPECT and CT data. Both CT and MRI were used for creating ROIs. Both the in vivo biodistribution (%ID/g) and integrated absorbed dose (Gy/MBq) were calculated on the high-resolution ROI voxel geometry. The biodistribution and dosimetry results were compared with ex vivo biodistribution data of NM600 performed right after the last imaging time point and RAPID dosimetry results were compared against S-values based on the MIRD formalism. The simulation results were benchmarked with published specific absorbed fractions (SAFs) and partial volume effects (PVEs) were also considered.

Results: The normal brain uptake was relatively low at all 4 timepoints imaged: 0.65 ± 0.29, 1.62 ± 0.95, 1.16 ± 0.93, and 0.96 ± 0.67 %ID/g, while the uptake of the T1-enhanced brain tumor was markedly higher: 6.84 ± 1.83, 9.79 ± 2.31, 9.72 ± 2.46, and 6.96 ± 1.31 %ID/g, and the uptake of the flank tumor was 4.38 ± 1.16, 4.69 ± 1.21, 4.25 ± 1.06, and 3.60 ± 0.80 %ID/g, respectively. The RAPID integrated mean absorbed dose result of the brain was 0.20 Gy/MBq, T1-enhanced brain tumor was 1.21 Gy/MBq, and the flank tumor was 0.66 Gy/MBq, respectively. The ex vivo biodistribution result of the brain was 0.51 %ID/g, the T1-enhanced brain tumor was 6.67 %ID/g, and the flank tumor was 3.98 %ID/g, respectively. The S-values based on the MIRD formalism, the integrated dose of the brain was 0.18 Gy/MBq, the T1-enhanced brain tumor was 1.34 Gy/MBq, and the flank tumor was 0.71 Gy/MBq, respectively. Between the T1-enhanced brain tumor and the flank tumor, the mean ROI volumes were 18 mm³ and 202 mm³, and the mean ROI masses were 19 mg and 205 mg, respectively. In this respect, the T1-enhanced brain tumor exhibits more uptake than the flank tumor with the ¹⁷⁷Lu-NM600 agent. As a result, the biodistribution and integrated absorbed doses showed good agreement with ex vivo biodistribution and OLINDA spheres, respectively, all exhibiting less than 10% difference.

Conclusions: In this study, the RAPID platform was expanded to calculate high-resolution voxel-wise dosimetry for mice gliomas treated with ¹⁷⁷Lu-NM600. This innovation will facilitate future preclinical studies that utilize micro-SPECT/CT to evaluate novel RPT agents.

• Download figure
• Open in new tab
• Download powerpoint