Modeling gold nanoparticle radiosensitization using a clustering algorithm to quantitate DNA double‐strand breaks with mixed‐physics Monte Carlo simulation

Abstract

Purpose: The radiosensitization properties of gold nanoparticles (GNPs) are investigated using a simple Geant4 cell model considering a realistic cell geometry and a clustering algorithm to characterize the number of DNA double-strand breaks (DSBs). Materials and methods: A mixed-physics approach is taken for accurate modeling of low-energy photon interactions in the different regions of the model using Geant4-DNA physics within the cell, and Livermore physics within gold. Density-based spatial clustering of applications with noise (DBSCAN), a clustering algorithm, is used to directly quantitate DNA DSBs after irradiation. The simulation was run using different sizes of GNPs, different distances of GNPs from the cell nucleus, and several combinations of these two conditions. Results: Four types of radiation were simulated in the work: 80-keV monoenergetic photons, 100-keV monoenergetic photons, a 250-kVp photon spectrum, and a 6-MV flattening filter free (FFF) photon spectrum. A variable enhancement in DSB yield, nucleus dose, and cell dose was observed when there are GNPs in the cell cytoplasm, and increases with larger GNPs and proximity to the nucleus. The distance of the GNPs from the nucleus has a large impact on the DSB yield and nucleus dose, but little to no effect on the cell dose. The cell dose enhancement factor of 80 keV photons varies from 1.037-1.125 at 0.2 µm for 30-100 nm GNPs to 1.040-1.127 at 4 lm. The DSB enhancement factor varies from 1.050 to 1.174 at 0.2 µm to a marginal effect of <1.01 at 4 lm. For 100 keV, the dose enhancement factor is from 1.142-1.470 at 0.2 µm to 1.106-1.371 at 4 lm. The DSB enhancement factor varies from 1.249-1.813 at 0.2 µm to almost no effect at 4 lm. For 250 kVp, the dose enhancement factor is from 1.117-1.393 at 0.2 lm to 1.110-1.342 at 4 lm. The DSB enhancement factor varies from 1.183-1.600 at 0.2 lm to a marginal effect of~1.03 at 4 lm. A 6-MV FFF shows a dose enhancement factor of 1.061-1.103 at 0.2 lm and 1.053-1.107 at 4 lm. The DSB yield varies from 1.070-1.143 at 0.2 lm to a marginal effect at 4 lm. Conclusion: The stark difference in behavior for DSB yield when compared to cell dose highlights the importance of evaluating more complex radiobiological quantities rather than dose alone when evaluating the radiosensitization properties from metallic nanomaterials. The nucleus dose showed similar characteristics to the DSB yield demonstrating the ability of the method to predict DNA damage and its relationship with nuclear dose. The proposed method provides a way to explore the radiobiological mechanisms of radiation-induced DNA damages, and it aids to evaluate the physical radiosensitization properties of GNP-aided radiotherapy, which can be easily combined with radiochemical DSB quantitation in order to better understand the intricate DNA damage induction mechanisms that are involved in GNP-aided radiotherapy.

A Computational Method to Estimate the Effect of Gold Nanoparticles on X-Ray Induced Dose Enhancement and Double-Strand Break Yields

“…Compared with the study by Cai et al [10] whose radius of cell nucleus is 6.3 μm, the dose enhancement using our approach (with the same radius) is ~1.4 and is in good agreement with Cai's calculated values 1.3 (Table 1 in Cai's study) for the concentration of 7 mg/ml in cells irradiated with 100 keV X-rays. For DSB induction, the RBE of cells mixed with GNP irradiated by 80 keV X-rays is reported to be in the range of 1.05-1.17 [40] while our result is 1.1 (Table 2).…”

“…The penmain program in PENELOPE code (2011 version) [29] was used to calculate the deposited amount of photon energies (10,15,40,68,69,82, and 250 keV and 2 MeV) in cells. The photon energies used in this study are all mono-energetic.…”

Ionizing Radiation and Complex DNA Damage: Quantifying the Radiobiological Damage Using Monte Carlo Simulations

“…The first method is to estimate potential DSBs by superimposing DNA geometry to the radiation track structure [42,44,[96][97][98]. The other method is to use clustering algorithms based on probabilistic models to estimate DSB yield [99][100][101]. The first method is more direct way to estimate DNA damage accurately, but it is also a time-consuming procedure.…”

IDDRRA: A novel platform, based on Geant4‐DNA to quantify DNA damage by ionizing radiation

“…Several incorporate chemistry models [6][7][8]10,[12][13][14]16,17,19 in contrast to others that can only simulate the physical interactions. 5,22,23 Some of the developed tools are easily accessible, while others require special licenses, some of which as purchasable options. Furthermore, quite a few of those packages require advanced computational user skills to install and run.…”