A spatial dispersion index of mean organ S-values for preclinical dosimetry

Objectives Preclinical experiments have become a key point in the development of new radiopharmaceuticals for use in targeted radionuclide therapy. Several organ S-values based on the MIRD scheme are available for various radionuclides and 3D voxel-based rodent models. However murine organs are of the same order of magnitude as β-particle range, and mean organ S-values may not always be relevant. This study aimed at providing reference rodent S-values associated with a spatial dispersion index for radionuclides of interest in preclinical imaging and therapy.

Methods A previously proposed mouse model (Moby Software; advanced organ definition; 200-μm³ spatial-sampling) was used to calculate S-values, at voxel and organ levels, by Monte Carlo simulation with GATE (v6.2). A spatial dispersion index was calculated for each source/target organ S-value by an iterative approach, consisting in the simulation of an increasing number of primary particles (typically from 10⁶ to 10¹¹) until each associated standard deviation reached a steady-state value. The spatial dispersion was also assessed by dose volume histograms.Radionuclides of various electron/photon energy spectra were considered. Specific corrections were implemented in the code and validated to improve voxel energy scoring.

Results S-value and spatial-dispersion-index tables were generated for all considered radionuclides and source/target organs. High spatial dispersion indexes (>300%) were observed for total body and diffused organs while lower but significant (30-200%) indexes were obtained for mid source/target distances. Indexes of 5% to 30% were obtained for self S-values (e.g. ^99mTc, ¹³¹I, ¹⁸F, ⁹⁰Y). Differences for given source/target organs and various radionuclides were also observed.

Conclusions This study showed that heterogeneous absorbed dose distribution in mice at the sub-organ level may prevent the use of mean organ S-values. The proposed method and generated S-value tables may be used to perform more accurate and reliable preclinical dosimetry, provided that the spatial dispersion index is first considered.