Early studies demonstrated that gold nanoparticles are effective radiosensitizers in mice. But it is still unclear whether gold is the optimum material, as only a few other elements have been examined and no systematic comparisons have been performed. Now, Stephen McMahon from Queen's University, Belfast, and colleagues have used Monte Carlo modelling to evaluate the radiosensitization effects of nanoparticles made from a wide range of elements. They found that the dependence on Z did not follow the expected simple steady increase, suggesting that gold may not provide the best biological impact (Nanoscale 8 581).

"When interest in nanoparticle radiosensitizers was picking up about a decade ago, the focus was primarily on increased radiation absorption," McMahon told medicalphysicsweb. "For that, gold was an obvious choice, as it was a very heavy but biocompatible element, and quickly had experimental evidence supporting it. However, absorption only tells one small part of the story."

In radiotherapy, increased dose deposition does not necessarily correlate with an increase in biological effects. The recent observation that extremely high doses are deposited in the immediate vicinity of a nanoparticle prompted the researchers to examine the differences between macro- and nano-scale dose enhancement.

A question of scale

McMahon and colleagues first calculated the macroscopic dose enhancement following X-ray irradiation at imaging (kilovoltage) and therapeutic (megavoltage) energies, for elements from silicon (Z = 14) to mercury (Z = 80). At kilovoltage energies, they observed a general trend for higher-Z atoms to exhibit greater absorption. At megavoltage energies, dose enhancement was far less dependent on material, as energy deposition is dominated by Compton scatter, which is largely independent of Z.

Next, the researchers examined nano-scale dose enhancements by modelling radiation–nanoparticle interactions and the resulting radial dose distributions. For kilovoltage X-rays, energies were tailored to each material, set to 20 keV above the material's K-edge (ranging from 22 to 102 keV). For therapeutic X-rays, they modelled clinically-relevant 6 MV linac exposures.

Plotting the nanoscale radial energy and dose distributions for an average radiation–nanoparticle interaction revealed considerable variability among the different elements for kilovoltage irradiations. On a macroscopic scale, energy deposition at low and high kilovoltage energies is dominated by photoelectrons and fluorescence photons, respectively. However, for dose deposited within 1 μm of the nanoparticle, Auger electrons are the primary source of energy deposition. The observed fluctuations are attributed to variations in secondary Auger electron spectra between elements.

"While the total amount of energy deposited goes up steadily with increasing Z, the dense ionization clusters that are believed to cause higher levels of biological damage are driven by the Auger electron yield," McMahon explained.

For a clinical megavoltage source, total energy deposits in the vicinity of the nanoparticle are dominated by electron scattering, which has little dependence on Z with almost no Auger electrons generated. This was reflected in the simulated dose and energy distributions, which varied only slightly with material for megavoltage irradiations.

Biological effects

To evaluate the potential biological impact of the nanoscale dose depositions, the researchers used an approach based on the local effect model. Examining the effective dose deposited by a single ionizing event revealed a complex material dependency for kilovoltage exposures, with two distinct local maxima centred around Z = 34 and Z = 68. For megavoltage exposures, little change was seen over the entire range of atomic number.

Finally, they calculated a relative biological effectiveness (RBE) for each element, defined as a reference X-ray dose (2 Gy) divided by the dose that yields equal survival in the presence of nanoparticles. Here again, significant variation was seen at kilovoltage energies and little variation for megavoltage exposures.

As megavoltage photons don't benefit from the dense ionization clusters or higher absorption rates seen for kilovoltage exposures, this translates into a very small impact on overall sensitization compared with kilovoltage exposures. This result is somewhat surprising, McMahon notes, as experiments have demonstrated significant sensitization when combining gold nanoparticles with megavoltage exposures.

"This tells us that our models are incomplete," he said. "There's growing evidence that what's missing is knowledge of the biochemistry of these particles – while gold nanoparticles were initially thought to be chemically inert, research has suggested that they cause a variety of biological stresses, which can lead to sensitization through other pathways. These may prove to be the major effect for high-energy exposures for many particle types."

The researchers hope to validate their models by looking in more detail at the sensitizing properties of particles made from different materials, or even combinations of materials. "This may open up a lot of interesting opportunities for the design of nanoparticles with different properties," said McMahon. "We also plan to further explore the biochemical impact of these particles in more detail to see if we can develop a more holistic understanding that combines both physical and biological processes, which may provide much better predictions of the effects that we might see clinically from these particles."

Related stories

• Nanoparticles enhance AMD radiotherapy
• Nanoparticles target tumour vasculature
• Inhaled nanoparticles target lung cancer
• Nanoparticles boost proton dose