ROUGH DRAFT authorea.com/15943

Abstract

# Introduction

In recent years there has been considerable interest on the use of gold nanoparticles (GNPs) as novel agents for cancer therapy, with many studies in diverse fields including preclinical and clinical trials (Zhang 2008, Rahman 2009, Cao-Milán 2014). In particular, the use of GNPs showed promising results for cancerous cells radiosensitization at both kilo- and mega-voltage energies (Hainfeld 2004, Kong 2008, Rahman 2009, Jain 2011).

As the concentration of nanoparticles increases the radiosensitization increases (Rahman 2009, Ranjbar 2010, Brun 2009), both due to the higher number of photoelectric interactions, and consequently higher dose deposition, and to the additional oxidative stress induced by the presence of nanoparticles (Pan 2009, Kang 2010). Experimental studies have revealed that GNPs radiosensitization is also highly sensitive to photon source energy (Chithrani 2010), cancer cell type (Patra 2007, Jain 2011), nanoparticle size (Butterworth 2012) and localization relative to cellular DNA (Brun 2009).

The premise of GNPs radiosensitization relies on the higher photoelectric absorption cross-section of gold relative to tissues. The high radiosensitization induced at kilovoltage energies by GNPs is well documented by both in vitro and in vivo studies (mice with implanted tumors) (Hainfeld 2004, Kong 2008, Rahman 2009, Jain 2012). However, the radiosensitization observed at kilovoltage energies by increased photoabsorption cannot help predict the effects occurring at clinically relevant megavoltage energies, where Compton interactions are dominant and photon energy absorption weakly depends on the atomic number. Ionization rates at megavoltage energies seem to be extremely low, meaning that, for doses typically used in radiotherapy, a fair amount of nanoparticles present in the system (i.e., more than 99% (McMahon 2011)) does not contribute to the dose deposition processes. Despite this minimal dose increase, it is was observed that GNPs could lead to significant levels of radiosensitization when irradiated with MV photons (Jain 2011).

Even for kV photons, the observed dose enhancement factor (DEF) cannot be completely justified only by the higher attenuation cross-section of gold relative to tissues (see for example (Rahman 2009)), and other mechanisms in addition to the strong photoelectric absorption have to be considered, such as different distributions of energy deposition at the nanometric scale, compared to those in tissues without GNPs (McMahon 2011, Butterworth 2012), and/or additional oxidative stress induced by the nanoparticles presence (Pan 2009, Kang 2010).

Additional studies are essential in order to understand the underlying processes of radiosensitization so as to use GNPs as a clinical therapeutic tool. In particular, the full mechanism under radiosensitization in radiotherapy still needs to be investigated in a patient-like geometry. Of note, the efficacy of a treatment depends on the nanoparticles concentration and spatial distribution in the tumor cells, as well as on the incident beam energy and delivered dose.

When metallic nanoparticles are localized in the tumor a greater fraction of the incident photon energy can be imparted to it without escalating the damage to the surrounding healthy tissues. Since GNPs are easily synthesized and can be designed to interact with various biomolecules, improved diagnosis and treatments efficacy can be obtained by using labeled nanoparticles that target specific cell receptors. That is, it is expected that nanoparticles bound to targeting agents should accumulate in higher concentrations in the tumor than in other organs, therefore amplifying the dose release in the cancerous cells while keeping constrained the cellular damage in the surrounding healthy tissues. So far, GNPs uptake has been found to be highest when bound to sugars (Kong 2008) and peptides (Kim 2011). It was also observed that, even in the absence of any surface modification, nanoparticles are able to passively accumulate in cancerous cells due to the enhanced permeability and retention effect (EPR) (Maeda 2000) of the abnormal tumor microvasculature (Dudley 2011). The combined effects of the intrinsic passive targeting (EPR) with the actual functionalization of the particle surface providing active targeting highly improves GNPs concentration inside the tumor volume.

In this paper, a radiobiological model is introduced to investigate the mechanisms underlying radiosensitization. The model, benchmarked with in vitro data taken from the literature, was included in the simulations of breast cancer IMRT treatments, with both 6 and 15 MV photons. In these simulations different uptake scenarios were also modeled to quantify the expected effectiveness of radiotherapy treatments with GNPs targeted dose enhancement.

# Materials and Methods

## GNPs dose profile simulations

Knowledge of range and type of secondary particles such as Auger electrons, photoelectrons or characteristic X-rays, and their variation with the primary photon energy is fundamental to develop a model delineating GNPs-photons interactions. Monte Carlo simulations were carried out using Geant4 simulation toolkit in which a 2 nm GNP placed at the center of a $$10^3$$ $$\mu\text{m}^3$$ cube of water was uniformly irradiated with different photon energies ranging from 50 keV to 15 MeV. Auger electrons, photoelectric emission, and interaction of secondaries in nearby atoms were simulated to evaluate the ionization density following irradiation. Energetic cutoff of 10 eV were used for the production of secondary particles.

The radial distribution of energy deposited from secondary electrons emitted from the nanoparticle was assumed symmetrical (McMahon 2011). The phase-space of secondary electrons was sampled on the GNP surface and the following energy deposits in the water volume surrounding the nanoparticle were histogrammed radially, putting the GNP at the origin, with a bin width of 1 nm up to a distance of 5000 nm. The actual shape of the radial dose profile to be found in realistic treatments was obtained through a weighted superposition of monoenergetic profiles, using specific photon spectra. For this work, spectra resulting from 160 kVp, 6 and 15 MV irradiation in water were analyzed to evaluate GNPs ionization rate (figure \ref{radial_dose}). The obtained dose distribution profiles were used as input to a radial function which was exploited to calculate the number of lethal events upon a determinate distance from the nanoparticles center.