A head-to-head comparison between two commercial software packages for hybrid dosimetry after peptide receptor radionuclide therapy

Abstract

Abstract Background Dosimetry after peptide receptor radionuclide therapy (PRRT) is increasing; however, comparing or pooling of dosimetric results can be challenging since different approaches are used. The aim of this study was to perform a head-to-head comparison of post-PRRT curve fitting and dosimetry obtained from two commercial software Hybrid Viewer Dosimetry and PLANET Dose. Methods Post-therapy imaging included planar scintigraphy at 0.5, 4, 24 and 72 h post-injection of [177Lu]Lu-DOTA-TATE for kinetics and SPECT/CT at 24 h for quantification. On planar imaging, 2 cm regions-of-interest were positioned within the inferior pole of the kidneys and kidney cortex was segmented on low-dose CT. On both planar and SPECT/CT, 2 cm spheres were positioned in the proximal humerus (red marrow equivalent) and in the region with the highest uptake in tumour lesions. TACs were estimated with mono- and bi-exponential fits in both software systems, after which tissue absorbed (kidney, red marrow, tumour) and biological effective doses (kidney) were calculated. Agreement-ICC, Spearman correlation and Bland-Altman plots were used to compare results. Results Mono-exponential fits showed the most comparable correlation between the measured and fitted data between both software. The ICC between absorbed dose outcomes was > 0.7 in tumour lesions and kidneys, but negative for the red marrow. Spearman correlation was > 0.9 for mono-exponential fits in kidneys and tumour lesions, and −0.7 in red marrow. Bi-exponential fits resulted in lower correlations and agreement values. Concordance between both software packages concerning the number of PRRT cycles with 7.4 GBq was observed based on a biological effective dose limit of 27 Gy to the kidneys. Conclusion [177Lu]Lu-DOTA-TATE dosimetry results of two software packages were comparable in the same dataset, despite the limited number of imaging time-points. However, these results should be verified in a larger cohort before pooling of clinical data, as the obtained results will depend on acquisition protocol, timing and lesions definition.