Aimee L. McNamara

A phenomenological relative biological effectiveness (RBE) model for proton therapy based on all published in vitro cell survival data.

Proton therapy treatments are currently planned and delivered using the assumption that the proton relative biological effectiveness (RBE) relative to photons is 1.1. This assumption ignores strong experimental evidence that suggests the RBE varies along the treatment field, i.e. with linear energy transfer (LET) and with tissue type. A recent review study collected over 70 experimental reports on proton RBE, providing a comprehensive dataset for predicting RBE for cell survival. Using this dataset we developed a model to predict proton RBE based on dose, dose average LET (LETd) and the ratio of the linear-quadratic model parameters for the reference radiation (α/β)x, as the tissue specific parameter. The proposed RBE model is based on the linear quadratic model and was derived from a nonlinear regression fit to 287 experimental data points. The proposed model predicts that the RBE increases with increasing LETd and decreases with increasing (α/β)x. This agrees with previous theoretical predictions on the relationship between RBE, LETd and (α/β)x. The model additionally predicts a decrease in RBE with increasing dose and shows a relationship between both α and β with LETd. Our proposed phenomenological RBE model is derived using the most comprehensive collection of proton RBE experimental data to date. Previously published phenomenological models, based on a limited data set, may have to be revised.

X-ray polarization in relativistic jets

We investigate the polarization properties of Comptonized X-rays from relativistic jets in active galactic nuclei (AGN) using Monte Carlo simulations. We consider three scenarios commonly proposed for the observed X-ray emission in AGN: Compton scattering of blackbody photons emitted from an accretion disc; scattering of cosmic microwave background (CMB) photons and self-Comptonization of intrinsically polarized synchrotron photons emitted by jet electrons. Our simulations show that for Comptonization of disc and CMB photons, the degree of polarization of the scattered photons increases with the viewing inclination angle with respect to the jet axis. In both cases, the maximum linear polarization is ≈20 per cent. In the case of synchrotron self-Comptonization (SSC), we find that the resulting X-ray polarization depends strongly on the seed synchrotron photon injection site, with typical fractional polarizations P ≈ 10–20 per cent when synchrotron emission is localized near the jet base, while P ≈ 20–70 per cent for the case of uniform emission throughout the jet. These results indicate that X-ray polarimetry may be capable of providing unique clues to identify the location of particle acceleration sites in relativistic jets. In particular, if synchrotron photons are emitted quasiuniformly throughout a jet, then the observed degree of X-ray polarization may be sufficiently different for each of the competing X-ray emission mechanisms (synchrotron, SSC or external Comptonization) to determine which is the dominant process. However, X-ray polarimetry alone is unlikely to be able to distinguish between disc and CMB Comptonization.

Extension of TOPAS for the simulation of proton radiation effects considering molecular and cellular endpoints

The aim of this work is to extend a widely used proton Monte Carlo tool, TOPAS, towards the modeling of relative biological effect (RBE) distributions in experimental arrangements as well as patients. TOPAS provides a software core which users configure by writing parameter files to, for instance, define application specific geometries and scoring conditions. Expert users may further extend TOPAS scoring capabilities by plugging in their own additional C++ code. This structure was utilized for the implementation of eight biophysical models suited to calculate proton RBE. As far as physics parameters are concerned, four of these models are based on the proton linear energy transfer, while the others are based on DNA double strand break induction and the frequency-mean specific energy, lineal energy, or delta electron generated track structure. The biological input parameters for all models are typically inferred from fits of the models to radiobiological experiments. The model structures have been implemented in a coherent way within the TOPAS architecture. Their performance was validated against measured experimental data on proton RBE in a spread-out Bragg peak using V79 Chinese Hamster cells. This work is an important step in bringing biologically optimized treatment planning for proton therapy closer to the clinical practice as it will allow researchers to refine and compare pre-defined as well as user-defined models.

Radiation damage on sub-cellular scales: beyond DNA.

This study investigates a model cell as a target for low-dose radiation using Monte Carlo simulations. Mono-energetic electrons and photons are used with initial energies between 10 and 50 keV, relevant to out-of-field radiotherapy scenarios where modern treatment modalities expose relatively large amounts of healthy tissue to low-dose radiation, and also to microbeam cell irradiation studies which show the importance of the cytoplasm as a radiation target. The relative proportions of number of ionizations and total energy deposit in the nucleus and cytoplasm are calculated. We show that for a macroscopic dose of no more than 1 Gy only a few hundred ionizations occur in the nucleus volume whereas the number of ionizations in the cytoplasm is over a magnitude larger. We find that the cell geometry can have an appreciable effect on the energy deposit in the cell and can cause a nonlinear increase in energy deposit with cytoplasm density. We also show that changing the nucleus volume has negligible effect on the total energy deposit but alters the relative proportion deposited in the nucleus and cytoplasm; the nucleus volume must increase to approximately the same volume as the cytoplasm before the energy deposit in the nucleus matches that in the cytoplasm. Additionally we find that energy deposited by electrons is generally insensitive to spatial variations in chemical composition, which can be attributed to negligible differences in electron stopping power for cytoplasm and nucleus materials. On the other hand, we find that chemical composition can affect energy deposited by photons due to non-negligible differences in attenuation coefficients. These results are of relevance in considering radiation effects in healthy cells, which tend to have smaller nuclei. Our results further show that the cytoplasm and organelles residing therein can be important targets for low-dose radiation damage in healthy cells and warrant investigation as much as the conventional focus of a high-dose radiation DNA target in tumour cells.

Characterization of a novel EPID designed for simultaneous imaging and dose verification in radiotherapy

PURPOSE Standard amorphous silicon electronic portal imaging devices (a-Si EPIDs) are x-ray imagers used frequently in radiotherapy that indirectly detect incident x-rays using a metal plate and phosphor screen. These detectors may also be used as two-dimensional dosimeters; however, they have a well-characterized nonwater-equivalent dosimetric response. Plastic scintillating (PS) fibers, on the other hand, have been shown to respond in a water-equivalent manner to x-rays in the energy range typically encountered during radiotherapy. In this study, the authors report on the first experimental measurements taken with a novel prototype PS a-Si EPID developed for the purpose of performing simultaneous imaging and dosimetry in radiotherapy. This prototype employs an array of PS fibers in place of the standard metal plate and phosphor screen. The imaging performance and dosimetric response of the prototype EPID were evaluated experimentally and compared to that of the standard EPID. METHODS Clinical 6 MV photon beams were used to first measure the detector sensitivity, linearity of dose response, and pixel noise characteristics of the prototype and standard EPIDs. Second, the dosimetric response of each EPID was evaluated relative to a reference water-equivalent dosimeter by measuring the off-axis and field size response in a nontransit configuration, along with the off-axis, field size, and transmission response in a transit configuration using solid water blocks. Finally, the imaging performance of the prototype and standard EPIDs was evaluated quantitatively by using an image quality phantom to measure the contrast to noise ratio (CNR) and spatial resolution of images acquired with each detector, and qualitatively by using an anthropomorphic phantom to acquire images representative of human anatomy. RESULTS The prototype EPIDs sensitivity was 0.37 times that of the standard EPID. Both EPIDs exhibited responses that were linear with delivered dose over a range of 1-100 monitor units. Over this range, the prototype and standard EPID central axis responses agreed to within 1.6%. Images taken with the prototype EPID were noisier than those taken with the standard EPID, with fractional uncertainties of 0.2% and 0.05% within the central 1 cm(2), respectively. For all dosimetry measurements, the prototype EPID exhibited a near water-equivalent response whereas the standard EPID did not. The CNR and spatial resolution of images taken with the standard EPID were greater than those taken with the prototype EPID. CONCLUSIONS A prototype EPID employing an array of PS fibers has been developed and the first experimental measurements are reported. The prototype EPID demonstrated a much morewater-equivalent dose response than the standard EPID. While the imaging performance of the standard EPID was superior to that of the prototype, the prototype EPID has many design characteristics that may be optimized to improve imaging performance. This investigation demonstrates the feasibility of a new detector design for simultaneous imaging and dosimetry treatment verification in radiotherapy.

Validation of the radiobiology toolkit TOPAS-nBio in simple DNA geometries

Computational simulations offer a powerful tool for quantitatively investigating radiation interactions with biological tissue and can help bridge the gap between physics, chemistry and biology. The TOPAS collaboration is tackling this challenge by extending the current Monte Carlo tool to allow for sub-cellular in silico simulations in a new extension, TOPAS-nBio. TOPAS wraps and extends the Geant4 Monte Carlo simulation toolkit and the new extension allows the modeling of particles down to vibrational energies (∼2eV) within realistic biological geometries. Here we present a validation of biological geometries available in TOPAS-nBio, by comparing our results to two previously published studies. We compare the prediction of strand breaks in a simple linear DNA strand from TOPAS-nBio to a published Monte Carlo track structure simulation study. While TOPAS-nBio confirms the trend in strand break generation, it predicts a higher frequency of events below an energy of 17.5eV compared to the alternative Monte Carlo track structure study. This is due to differences in the physics models used by each code. We also compare the experimental measurement of strand breaks from incident protons in DNA plasmids to TOPAS-nBio simulations. Our results show good agreement of single and double strand breaks predicting a similar increase in the strand break yield with increasing LET.

Dose enhancement effects to the nucleus and mitochondria from gold nanoparticles in the cytosol

Gold nanoparticles (GNPs) have shown potential as dose enhancers for radiation therapy. Since damage to the genome affects the viability of a cell, it is generally assumed that GNPs have to localise within the cell nucleus. In practice, however, GNPs tend to localise in the cytoplasm yet still appear to have a dose enhancing effect on the cell. Whether this effect can be attributed to stress-induced biological mechanisms or to physical damage to extra-nuclear cellular targets is still unclear. There is however growing evidence to suggest that the cellular response to radiation can also be influenced by indirect processes induced when the nucleus is not directly targeted by radiation. The mitochondrion in particular may be an effective extra-nuclear radiation target given its many important functional roles in the cell. To more accurately predict the physical effect of radiation within different cell organelles, we measured the full chemical composition of a whole human lymphocytic JURKAT cell as well as two separate organelles; the cell nucleus and the mitochondrion. The experimental measurements found that all three biological materials had similar ionisation energies  ∼70 eV, substantially lower than that of liquid water  ∼78 eV. Monte Carlo simulations for 10-50 keV incident photons showed higher energy deposition and ionisation numbers in the cell and organelle materials compared to liquid water. Adding a 1% mass fraction of gold to each material increased the energy deposition by a factor of  ∼1.8 when averaged over all incident photon energies. Simulations of a realistic compartmentalised cell show that the presence of gold in the cytosol increases the energy deposition in the mitochondrial volume more than within the nuclear volume. We find this is due to sub-micron delocalisation of energy by photoelectrons, making the mitochondria a potentially viable indirect radiation target for GNPs that localise to the cytosol.

Predicted ionisation in mitochondria and observed acute changes in the mitochondrial transcriptome after gamma irradiation: a Monte Carlo simulation and quantitative PCR study.

It is a widely accepted that the cell nucleus is the primary site of radiation damage while extra-nuclear radiation effects are not yet systematically included into models of radiation damage. We performed Monte Carlo simulations assuming a spherical cell (diameter 11.5 μm) modelled after JURKAT cells with the inclusion of realistic elemental composition data based on published literature. The cell model consists of cytoplasm (density 1g/cm(3)), nucleus (diameter 8.5 μm; 40% of cell volume) as well as cylindrical mitochondria (diameter 1 μm; volume 0.5 μm(3)) of three different densities (1, 2 and 10 g/cm(3)) and total mitochondrial volume relative to the cell volume (10, 20, 30%). Our simulation predicts that if mitochondria take up more than 20% of a cells volume, ionisation events will be the preferentially located in mitochondria rather than in the cell nucleus. Using quantitative polymerase chain reaction, we substantiate in JURKAT cells that human mitochondria respond to gamma radiation with early (within 30 min) differential changes in the expression levels of 18 mitochondrially encoded genes, whereby the number of regulated genes varies in a dose-dependent but non-linear pattern (10 Gy: 1 gene; 50 Gy: 5 genes; 100 Gy: 12 genes). The simulation data as well as the experimental observations suggest that current models of acute radiation effects, which largely focus on nuclear effects, might benefit from more systematic considerations of the early mitochondrial responses and how these may subsequently determine cell response to ionising radiation.

In Silico Nanodosimetry: New Insights into Nontargeted Biological Responses to Radiation

The long-held view that radiation-induced biological damage must be initiated in the cell nucleus, either on or near DNA itself, is being confronted by mounting evidence to suggest otherwise. While the efficacy of cell death may be determined by radiation damage to nuclear DNA, a plethora of less deterministic biological responses has been observed when DNA is not targeted. These so-called nontargeted responses cannot be understood in the framework of DNA-centric radiobiological models; what is needed are new physically motivated models that address the damage-sensing signalling pathways triggered by the production of reactive free radicals. To this end, we have conducted a series of in silico experiments aimed at elucidating the underlying physical processes responsible for nontargeted biological responses to radiation. Our simulation studies implement new results on very low-energy electromagnetic interactions in liquid water (applicable down to nanoscales) and we also consider a realistic simulation of extranuclear microbeam irradiation of a cell. Our results support the idea that organelles with important functional roles, such as mitochondria and lysosomes, as well as membranes, are viable targets for ionizations and excitations, and their chemical composition and density are critical to determining the free radical yield and ensuing biological responses.

A comparison of X-ray and proton beam low energy secondary electron track structures using the low energy models of Geant4.

Abstract Purpose: Lethal cell damage by ionising radiation is generally initiated by the formation of complex strand breaks, resulting from ionisation clusters in the DNA molecule. A better understanding of the effect of the distribution of ionisation clusters within the cell and particularly in regard to DNA segments could be beneficial to radiation therapy treatment planning. Low energy X-rays generate an abundance of low energy electrons similar to that associated with MeV protons. The study and comparison of the track structure of photon and proton beams could permit the substitution of photon microbeams for single cell ion irradiations at proton facilities used to predict the relative biological effectiveness (RBE) of charged particle fields. Materials and methods: The track structure of X-ray photons is compared with proton pencil beams in voxels of approximate DNA strand size (2 × 2 × 5 nm). The Very Low Energy extension models of the Monte Carlo simulation toolkit GEometry ANd Tracking 4 (Geant4) is used. Simulations were performed in a water phantom for an X-ray and proton beam of energies 100 keV and 20 MeV, respectively. Results: The track structure of the photon and proton beams are evaluated using the ionisation cluster size distribution as well as the radial dose deposition of the beam. Conclusions: A comparative analysis of the ionisation cluster distribution and radial dose deposition obtained is presented, which suggest that low energy X-rays could produce similar ionisation cluster distributions to MeV protons on the DNA scale of size at depths greater than ∼10 μm and at distances greater than ∼1 μm from the beam centre. Here the ionisation cluster size for each beam is less than ∼100. The radial dose deposition is also approximately equal at large depths and at distances greater than 10 μm from the beam centre.