Biological modeling of gadolinium-based nanoparticles radio-enhancement for kilovoltage photons: a Monte Carlo study

Abstract

Background: Gadolinium-based nanoparticles (GdNPs) are clinically used agents to increase the radiosensitivity of tumor cells. However, studies on the mechanisms and biological modeling of GdNP radio-enhancement are still preliminary. This study aims to investigate the mechanism of radio-enhancement of GdNPs for kilovoltage photons using Monte Carlo (MC) simulations, and to establish local effect model (LEM)-based biological model of GdNP radiosensitization. Methods: The spectrum and yield of secondary electrons and dose enhancement around a single GdNP and clustered GdNPs were calculated in a water cube phantom by MC track-structure simulations using TOPAS code. We constructed a partial shell-like cell geometry model of pancreatic cancer cell based on transmission electron microscope (TEM) observations. LEM-based biological modeling of GdNP radiosensitization was established based on the MC-calculated nano-scale dose distributions in the cell model to predict the cell surviving fractions after irradiation. Results: The yield of secondary electrons for GdNP was 0.16% of the yield for gold nanoparticle (GNP), whereas the average electron energy was 12% higher. The majority of the dose enhancement came from the contribution of Auger electrons. GdNP clusters had a larger range and extent of dose enhancement than single GdNPs, although GdNP clustering reduced radial dose per interacting photon significantly. For the dose range between 0 and 8 Gy, the surviving fraction predicted using LEM-based biological model laid within one standard deviation of the published experimental results, and the deviations between them were all within 25%. Conclusions: The mechanism of radio-enhancement of GdNPs for kilovoltage photons was investigated using MC simulations. The prediction results of the established LEM-based biological model for GdNP radiosensitization showed good agreement with published experimental results, although the deviation of simulation parameters can lead to large disparity in the results. To our knowledge, this was the first LEM-based biological model for GdNP radiosensitization.