SPECT dosimetry workflow for radionuclide therapy

Clinical radionuclide-therapy dosimetry follows a standard sequence: sequential quantitative SPECT/CT (or hybrid planar-plus-SPECT/CT) imaging at multiple time points after administration, segmentation of source organs and tumours, fitting of time-activity curves to obtain the time-integrated activity in each region, and conversion to absorbed dose using either organ-level S-values (MIRD/OLINDA formalism) or voxel-level convolution/Monte Carlo methods. The EANM Dosimetry Committee's standard operational procedures, most recently the 2022 guideline for 177Lu-labelled somatostatin-receptor- and PSMA-targeting ligands, recommend using the same SPECT/CT system throughout a patient's dosimetry study, medium-energy collimators for 177Lu, iterative reconstruction with CT-based attenuation, scatter and (where available) collimator-response correction, and 60-120 projection angles at roughly 30-40 seconds per projection.

For kidney dosimetry, the guideline recommends three SPECT/CT acquisitions between 24 hours and 7 days after the first therapy cycle (with at least two days between the final two time points), reducing to a single acquisition at a matched time point for subsequent cycles, with the region of interest drawn to include the full renal cortex and medulla but exclude the pelvis. Tumour dosimetry follows a similar multi-time-point protocol, placing particular weight on a late imaging time point because tumour retention is typically slower than that of normal organs. A minimum of three time points is generally advised to characterise each exponential phase of the time-activity curve; two or three points at around 2, 4 and 24 hours are often sufficient for many source organs, but bone marrow and whole-body kinetics may need additional early sampling. Related standard operational procedures exist for 131I-mIBG neuroendocrine tumour therapy and for 90Y radioembolisation (99mTc-MAA pre-therapy and 90Y peri-therapy dosimetry).