Click here to


Are you sure ?

Yes, do it No, cancel

Impact of Kernel Truncation On 177Lu-DOTATATE and 131I-MIBG Voxelwise Dosimetry

S Graves*, A Tiwari , D Hyer , R Flynn , J Buatti , J Sunderland , University of Iowa, Iowa City, IA


(Tuesday, 7/16/2019) 9:30 AM - 10:00 AM

Room: Exhibit Hall | Forum 8

Purpose: Patient-specific radiopharmaceutical dosimetry, including voxelwise approaches, has the potential to increase therapeutic benefit. In response to this, several voxelwise dosimetry software packages have become commercially available. These packages tend to convolve an energy deposition kernel (¹��Lu, ¹³¹I, ��Y, etc.) that is dimensionally larger than the beta particle range but insufficient to account for photon emissions involved in these radioactive decays. The purpose of this work was to assess the dosimetric impact of utilizing energy deposition kernels that are sized such that the majority of photon energy deposition is accounted for in the dosimetry calculation.

Methods: ��Ga-DOTATOC PET scans from three patients with advanced neuroendocrine cancer were anonymized and used as representative radiopharmaceutical activity distributions. Activity maps were scaled to a normalized cumulative activity map using a population-based pharmacokinetic estimate, and dose was calculated in two ways: (A) using a truncated kernel (~2.4cm) and (B) using a large kernel (~70cm) that is sufficient to encompass the majority of photon energy deposition. All kernels were derived from in-house Monte Carlo simulations. The resulting differences in mean organ dose (kidneys, liver, lumbar spine and pelvis) were analyzed. Dose to the spine/pelvis was intended to be a surrogate for red marrow dosimetry, which is the most common dose-limiting organ for ¹��Lu-DOTATATE and ¹³¹I-MIBG.

Results: Significant increases in mean organ dose were observed when comparing the full-sized kernel calculation against the truncated kernel calculation. Dosimetric ratios (Dfull/Dtrunc) of 1.08±0.04, 1.09±0.02, and 1.25±0.06 (mean±SD, n=3) were obtained for the kidneys, liver, and spine/pelvis for ¹��Lu. Dosimetric ratios of 1.57±0.22, 1.68±0.06, and 2.90±0.27 were obtained for the kidneys, liver, and spine for ¹³¹I.

Conclusion: Significant dosimetric inaccuracies result from convolution-based voxelwise dosimetry with kernels that do not fully encompass photon energy deposition. Caution should be used when employing these tools clinically.


Radiation Dosimetry, Radial Dose Function, Nuclear Medicine


IM/TH- Radiopharmaceutical therapy: Non I-131 MIRD/analytical dose estimation

Contact Email