PurposeNanoparticles appear as a novel tool to enhance the effectiveness of radiotherapy in cancer treatments. Many parameters influence their efficacy, such as their size, concentration, composition, their cellular localization, as well as the photon source energy. The current Monte Carlo study aims at comparing the dose-enhancement in presence of gadolinium (Gd), either as isolated atoms or atoms clustered in nanoparticles (NPs), by investigating the role played by these physical parameters at the cellular and the nanometer scale. In parallel, in vitro assays were performed in presence of either the gadolinium contrast agent (GdCA) Magnevist® or ultrasmall gadolinium NPs (GdNPs, 3 nm) for comparison with the simulations.MethodsPENELOPE Monte Carlo Code was used for in silico dose calculations. Monochromatic photon beams were used to calculate dose enhancements in different cell compartments and low-energy secondary electron spectra dependence with energy. Particular attention has been placed on the interplay between the X-ray beam energy, the Gd localization and its distance from cellular targets. Clonogenic assays were used to quantify F98 rat glioma cell survival after irradiation in the presence of GdNPs or GdCA, using monochromatic X-rays with energies in the 30 keV–80 keV range from a synchrotron and 1.25 MeV gamma photons from a cobalt-60 source. The simulations that correspond to the experimental conditions were compared with the experimental results.ResultsIn silico, a highly heterogeneous and clustered Gd-atom distribution, a massive production of low energy electrons around GdNPs and an optimal X-ray beam energy, above the Gd K-edge, were key factors found to increase microscopic doses, which could potentially induce cell death. The different Gd localizations studied all resulted in a lower dose enhancement for the nucleus component than for cytoplasm or membrane compartments, with a maximum dose-enhancement factor (DEF) found at 65 keV and 58 keV, respectively. In vitro, radiosensitization was observed with GdNPs incubated 5 h with the cells (2.1 mg Gd/mL) at all energies. Experimental DEFs were found to be greater than computational DEFs but follow a similar trend with irradiation energy. However, an important radiosensitivity was observed experimentally with GdNPs at high energy (1.25 MeV), whereas no effect was expected from modeling. This effect was correlated with GdNPs incubation time. In vitro, GdCA provided no dose enhancement at 1.25 MeV energies, in agreement with computed data.ConclusionsThese results provide a foundation on which to base optimizations of the physical parameters in Gd radiation-enhanced therapy. Strong evidence was provided that GdCA or GdNPs could both be used for radiation dose-enhancement therapy. There in vivo biological distribution, in the tumor volume and at the cellular scale, will be the key factor for providing large dose enhancements and determine their therapeutic efficacy.< Réduire