Introduction
Over the last few decades, radiotherapy has become increasingly conformal through improvements in treatment planning and the introduction of new delivery techniques. This accurate dose delivery has led to a growth in research into how to further protect healthy tissues, including the use of hypofractionated treatments [
1,
2] and high dose rate FLASH radiotherapy treatments [
3‐
5]. However, research is also discovering ways to enable greater tumour radiosensitisation through drug radiotherapy combination approaches [
6], including the use of metal-based nanoparticles [
7].
Nanoparticles (NPs), particularly high Z number metal elements, have been extensively studied due to their high absorption of X-ray photons leading to large increases in secondary electrons and subsequent increase levels of DNA damage by physically enhancing the dose [
8‐
10]. However, high Z metal NPs have also been shown to act through distinct biological mechanisms including higher levels of oxidative stress, increasing levels of reactive oxygen species (ROS) and inducing highly reactive hydroxyl radicals which go on to cause further DNA damage [
10‐
17].
High Z number metal NPs have significant potential in radiotherapy as they are preferentially taken up by the tumour, ensuring that there is no increase in DNA damage to healthy surrounding tissue. This is due to the enhanced permeability and retention (EPR) effect, by which the leaky vasculature of the tumour takes up the NPs, where the size of the particles would otherwise be too large for healthy cells with normal vasculature [
18‐
20]. As NPs are normally larger than the renal excretion threshold, they are also likely to stay in the tumour for a longer period, potentially causing greater levels of radiosensitisation [
21].
The focus of much of the research into nanoparticle radiosensitisation has involved gold nanoparticles. They have proved to be effective in a range of sizes, with various coatings and chelating agents [
22‐
34]. However, more recently, gadolinium nanoparticles have been studied [
35‐
45], with AGuIX (a 5 nm gadolinium nanoparticle) currently being tested in the NanoRAD2 clinical trial [
46]. The optimal element for NP radiosensitisers is under debate, with McMahon et al
. demonstrating both the micro- and macroscopic factors for a range of elements to determine the optimal nanoparticle element[
47].
Superparamagnetic Iron Oxide Nanoparticles (SPIONs) have been studied for a range of different treatments, from drug delivery to antibody-binding [
48], however, they are primarily investigated for their ability to provide magnetic hyperthermia treatments [
49‐
52]. The superparamagnetic properties of SPIONs mean that, when placed in an alternating magnetic field, they vibrate and heat up, but once removed from the magnetic field, this magnetism is lost [
50]. When placed inside tissues, it is thought that this hyperthermia effect induces cell death [
48].
The rationale for selecting SPIONs in this study is due to its applicability to the ongoing development of MR-Linacs. Iron oxide contrast agents have been used primarily in T2-weighted MRI scans [
48,
53], and so, with the combination of MRI and radiotherapy now becoming clinically available, it is possible that these MRI contrast agents could be used both diagnostically and therapeutically, both for contrast enhancement but also for increasing the radiosensitivity of a tumour for more effective radiotherapy treatments. A study by Hu et al. demonstrated this capability with the gadolinium-based nanoparticle, AGuIX [
54].
Some previous studies have evaluated the radiosensitising properties of SPIONs, including Klein et al. indicating radiosensitisation with MCF7 breast cells through cytotoxicity studies and production of reactive oxygen species [
55]. Kirakli et al. also investigated the radiosensitisation of citrate coated SPIONs using MV radiation on 3 cancer cell lines using clonogenic assays, finding the greatest radiosensitisation at a dose of 2 Gy [
56]. Rashid et al. investigated radiosensitisation of colon cancer cells from SPIONs in combination with a 150 MeV proton beam, investigating the reactive oxygen species generation [
13]. The use of iron oxide nanoparticles in combination with proton beams have also been investigated in vivo by Seo et al. [
57]. Choi et al. found 13 nm iron oxide nanoparticles to significantly reduce the viability of colon tumour cells in mice compared with control cells when treated with 7.1 keV monochromatic x-rays [
58].
There is a clinical need to determine the efficacy of SPIONs for the development of MR-Linacs with potential capabilities of utilising them both diagnostically and therapeutically, which should be determined using further studies. The aim of this study is to investigate the impact of SPIONs on the radiosensitisation in vitro when combined with 225 kVp X-rays through clonogenic and DNA damage immunofluorescence assays. The SPIONs’ hydrodynamic radius was characterised using Dynamic Light Scattering (DLS). Inductively-Coupled Plasma Mass-Spectrometry (ICP-MS) has also been used to quantify the iron oxide taken into cells, before correlating with the levels of DNA damage. This radiosensitisation was then investigated in vivo for H460 subcutaneous tumours.
It is also important to note the difference between the kVp source used for this investigation and the MV energies which are conventionally used for radiotherapy. kVp energies are dominated by the photoelectric effect which is proportional to Z
3 compared with the Compton effect for MV energies which is proportional to Z [
59], suggesting that radiosensitisation from the nanoparticles would be greater for kVp energies. However, the largest differences in the ratios of absorption coefficients between soft tissue and nanoparticles occur in the region below 100 keV [
60] which are strongly attenuated by the 2 mm Cu filter used for the kVp beams (Additional file
2). In particular, the k-edge of iron oxide is < 10 keV, further reducing the effect [
60]. Whilst it is important to acknowledge the differences in the physical dose absorption caused by the differences in energy, comparison can be made regarding the extent of the biological effects.
Materials and methods
Cell culture and nanoparticles
6 human cancer cell lines were selected to be investigated in vitro; H460 (large cell lung cancer), MiaPaCa2 (pancreatic carcinoma), DU145 (prostate carcinoma), MCF7 (breast adenocarcinoma), U87 (brain glioblastoma) and HEPG2 (liver carcinoma). All cell lines were acquired from the American Type Culture Collection (ATCC), and were routinely tested for mycoplasma. All cell lines have also been authenticated by ATCC using Short Tandem Repeat (STR) Screening. Cells were incubated at 37 °C in 5% CO2, and all tissue culture was performed in a Class II laminar flow cabinet (Thermo, US). All cell lines, except for MiaPaCa2, were cultured in Dulbecco’s Modified Eagle Media (DMEM) (Sigma, US), supplemented with 10% Foetal Bovine Serum (FBS) (Sigma, US) and 1% Penicillin Streptomycin (Pen-Strep) (Sigma, US). MiaPaCa2 cells were incubated in High Glucose DMEM media, supplemented with 10% FBS, 1% Pen-Strep, and 0.2% Sodium Pyruvate. Cells were passaged every 2–3 days to maintain exponential grown.
NPs used for this study were SPIONs with a 5 nm diameter, acquired from Sigma, US (product number 725331), suspended in H
2O at a concentration of 5 mg/ml, with the molecular weight of Fe
3O
4 being 231.53 g/mol. For the purpose of this study, all assays were performed at a SPION concentration of 23.5 μg/ml, which was determined as being within the range investigated by Kirakli et al. [
56].
Dynamic light scattering
The SPIONs used for this study are commercially available and quoted as being 5 nm in diameter, determined by Transmission Electron Microscopy (TEM) [
61]. However, the size of the molecules when dispersed in an aqueous solution is unclear. DLS was used to quantify the diameter of hydrodynamic radius and provide detail about the nature of the dispersion.
The experiment was carried out at the National Physical Laboratory, and followed the methods described by Minelli et al. [
62]. This is a validated method for determining nanoparticle size. A Zetasizer (Zetasizer Nano ZS 3600, Malvern Panalytical Ltd, Malvern, UK) was used which performs measurements using a 633 nm Helium-Neon laser, measuring the light at a scattered angle of 173°. The particles were diluted to a range of concentrations; stock dilution, 1:1, 1:4, 1:6 and 1:20, with measurements carried out at temperature of 25°C and 3 repeats for each sample. The average diameter for each sample was noted along with peaks in the measurements to determine any variation in size of agglomeration of the particles.
ICP-MS uptake measurements
ICP-MS was used to determine the quantity of the SPIONs that were entering the cells. For preparation of ICP-MS samples, 1 × 106 cells were plated into 30 mm petri-dishes, before adding SPIONs at a concentration of 23.5 μg/ml in 500 μl media, 24 h later. After a further 24 h, in order to adhere to the incubation times used for in vitro experiments, the cells were washed with phosphate buffered solutions (PBS) (Sigma, US), suspended using 2 ml trypsin (Sigma, US), and the number of cells counted using a Coulter Counter (Beckman Coulter Life Sciences, US). Cells were covered in 500 μl deionised water (ddH2O) and left to dry overnight. Cells were then exposed to 500 μl of Aqua Regia (3: 1 hydrochloric acid: nitric acid) and left in a fume hood overnight to evaporate, to dissolve the cells so that only the NPs remained. The remaining SPIONs were resuspended into 30 ml of ddH2O prior to running the samples.
The ICP-MS (Agilent 8800, Agilent Technologies, UK) ran the samples consecutively using an autosampler, with a regular sample blank to test for any iron contamination. Also, one SPION sample was sampled repeatedly as an internal standard over the course of the run to test for evidence of SPIONs falling out of solution. Any decrease in detected signal found was then corrected. The ICP-MS machine is equipped with two quadrupole mass filters and a collision-reaction cell, but in this study was operated with only one mass filter (termed single quad mode). Helium was introduced into the cell at a flow rate of 4 ml/min as a collision gas to remove
40Ar
16O formed in the plasma, which would otherwise overlap with the
56Fe signal. The instrument was fitted with a quartz double-pass spray chamber as well as a MicroMist nebuliser (Glass Expansion, Australia), with nickel sample and skimmer cones (Crawford Scientific, UK). This setup was in accordance to Russell et al. [
63]. Between each of the samples the probe was rinsed twice, firstly with 0.3 M HNO
3 to wash Fe from the previous sample and prevent any cross contamination between samples, and then with ddH
2O to match the matrix of the next sample to be measured.
Using a calibration curve made from a range of samples with known concentrations of SPIONs, the counts per second detected by the ICP-MS were converted to a corresponding concentration to determine the mass of iron oxide found on average within each cell.
Immunofluorescence DNA damage assay
DNA double strand breaks (DSBs) were detected using a 53BP1 immunofluorescence assay. Cells were seeded onto sterilised cover slips in 6-well plates (Starstedt, Germany), at a cell density of 1 × 105 cells per well, covered with 2 ml media. 24 h later, SPIONs were added at a concentration of 23.5 μg/ml in 500 μl. After a further 24 h, cells were irradiated with doses of 0–2 Gy, and immediately afterwards the nanoparticles were replaced with fresh media. Only doses up to 2 Gy were used because higher doses would result in a high number of overlapping foci leading to inaccurate determinations of the level of DNA damage within the cell. All irradiations were carried out using an X-RAD225 X-ray cabinet source (Precision X-Ray, US). Irradiations were performed at an SSD of 50 cm, with a dose rate of 0.589 Gy/min and an energy of 225 kVp filtered with a copper filter of 2 mm Cu that gives a half value layer (HVL) of 2.3 mm Cu. The mean energy of the x-ray spectrum is 113 kV, with a peak energy occurring around 100 kV.
Cells were fixed at time points of 1 h and 24 h post-irradiation, adding 2 ml of 50: 50 methanol: acetone to each well for an incubation time of 20 min at 4 °C before cells were washed 3 times for 5 min in PBS. Cells were permeabilised using 1 ml methanol per well for 10 min before being washed again 3 times with PBS. Blocking was performed using 2 ml blocking buffer (5% FBS, 0.5% Triton X-100 in PBS), incubated for 1 h at 4 °C. An anti-53BP1 antibody derived from rabbits was used as the primary antibody, at a concentration of 1:3000 suspended in blocking buffer, with 500 μl added per well. After an incubation time of 1 h at room temperature, the primary antibody was washed off 3 times in washing buffer (0.1% Triton X-100 in PBS) for 5 min. The secondary antibody, goat-anti-rabbit, was used (GAR 488) at a concentration of 1:2000 suspended in blocking buffer, at room temperature for 1 h, in darkness to not fade the fluorescence. Samples were washed in washing buffer 3 times and mounted onto glass slides using 20 μl ProLong Gold Antifade Mountant with DAPI (Thermo, US).
To quantify levels of DSBs, the number of foci was counted for 50 randomly selected cells for each slide and averaged over 3 repeats. Statistics were carried out as a Student’s t test for any significant increase in DNA damage. When counting foci, cells were discounted if they were enlarged, had very large numbers of foci, or appeared to be undergoing mitosis or apoptosis.
DNA damage was further analysed by comparing with the uptake data using ICP-MS. A correlation was determined between the numbers of foci and the mass of iron oxide taken up using the Spearman’s Rank correlation coefficient, used for non-parametric data, with a coefficient between + / − 1 with 1 being an entirely positive correlation and − 1 being entirely negative. A correction of greater than + / − 0.7 is considered to be a strong correlation [
64].
Clonogenic assay
Cell survival was quantified using the clonogenic assay for a range of doses from 0 to 8 Gy for the combination of radiation combined with SPIONs, using the method described by Puck and Marcus [
65]. Firstly, cells were seeded into 6 well plates with cell densities depending on the dose of radiation received; 0–2 Gy: 500 cells, 4 Gy: 1000 cells, and 8 Gy: 2000 cells. Cells were incubated for 24 h to adhere to the wells, before SPIONs were added both 1 h and 24 h prior to radiation at a concentration of 23.5 μg/ml in 500 μl media. After irradiation with 225 kVp X-rays, the SPIONs were replacing with fresh media. They were left for an incubation time ranging from 6 to 13 days, depending on the cell line. Cells were stained using Crystal Violet (0.4% in 95% Ethanol), for 20 min before the excess was rinsed off.
The number of colonies was then counted for each well, with a colony determined to be a group of > 50 cells. The plating efficiency (PE) was calculated to be the percentage of the cells plated that went on to develop colonies. This was used to determine the surviving fraction (SF) for each dose, as the proportion of colonies counted compared with controls. Graphs of radiation dose against surviving fraction were then fitted with the Linear Quadratic Model, given by Eq.
1, where D is the radiation dose (Gy) and
\(\alpha\) and
\(\beta\) are the linear and quadradic components of cell killing.
$$SF = {\text{exp}}\left( { - \alpha D - \beta D^{2} } \right)$$
(1)
Statistical significance in the clonogenic assays was carried out using the Extra Sum-of-Squares F-Test using GraphPad™ Prism, by which it is determined whether one curve of best fit could accurately fit the other data sets.
Predicted dose enhancement factor (DEF) and sensitisation enhancement ratio (SER)
The SER was calculated using the clonogenic assay and is a ratio of the dose of radiation in combination with SPIONs that gives the same level of cell survival for a specific radiation dose only [
66]. This was carried out both for the 8 Gy dose and for a 10% surviving fraction for controls, in which the
\(\alpha\) and
\(\beta\) values for the control and the treated linear quadratic curves were used to determine the dose that would give the same surviving fraction for treated cells. An SER of 1 would represent the same biological effect as the control samples.
To predict the DEF, prior to performing the clonogenic assay, it was approximated as the ratio of mass energy absorption coefficients between the media alone and the media plus iron oxide, using the method described by McMahon et al. for macroscopic dose enhancement calculations [
47]. Firstly, the attenuation coefficients were determined for a range of energies from 0 to 225 keV, and then a scaling factor accounted for the fluence of the X-Ray spectrum at each energy for both water and iron oxide. The density of the SPIONs in the media was then used to predict the DEF.
In vivo model
Severe Combined Immunodeficient (SCID) mice at 8–10 weeks old were obtained from Charles River Laboratories (Oxford, UK), 1 × 106 H460 cells were subcutaneously injected into the dorsum of each mouse. Animals were randomly assigned to the following experimental groups; control (n = 5), SPIONs only (n = 2), Radiation only (n = 4), and combined radiation and SPIONs (n = 5). 23.5 μg/ml SPIONs in 50 μl sterile ddH2O were intra-tumourally injected 15 min prior to irradiations of 12 Gy in a single fraction with X-ray energy 220 kVp using a Small Animal Radiation Research Platform (SAARP) (Xstrahl, UK), with Cone Beam CT (CBCT) image-guided and parallel opposed beam geometry. Following treatment, tumour volume was measured every 2–3 days using callipers until experimental endpoints were reached which were defined as a tumour geometric mean diameter (GMD) of 12 mm3 or a 10% loss in body weight, measured by taking the diameter of the tumour in 3 planes. All procedures were carried out in accordance with United Kingdom Department of Health approval for in vivo experimentations.
Discussion
This study aimed to assess the radiosensitising properties of SPIONs as potential theranostic agents. An inherent increased DNA damage and decreased survival was observed with 23.5 μg/ml SPIONs in all cell lines excluding U87 cells, for a 24 h incubation time. The most significant effects were seen in MCF7 and HEPG2 cells (
P < 0.0001). These data are in contrast to a previous study by Khoshgard et al. which reported no cytotoxicity for SPIONs following 24 h exposures in MCF7 and HeLa cells, using an MTT assay [
67]. Also, the inherent cytotoxicity observed in our study did not correlate DNA damage, suggesting that it is possible that in cell lines other than U87 and HEPG2, any decrease in cell survival post-irradiation was due to radiosensitisation, along with other possible causes of DNA damage.
The impact of combinations of SPIONs with radiation on DNA damage yields is presented in Fig.
4. All cell lines saw a significant increase in damage at 1 Gy, and all cell lines except U87 cells at 2 Gy, 1 h post-irradiation. It suggests an increase in DNA damage driven by SPIONs alone on the cells, as well as a radiosensitising effect. At the 24 h timepoint, once again, all cell lines except for U87 cells see a significant increase in DNA damage at 1 Gy. However, significant increases were only seen for H460, MiaPaCa2, MCF7 and HEPG2 cells at 2 Gy. DNA damage decreased at 2 Gy for DU145 cells. While still significant, HEPG2 cells saw higher levels of DNA damage at 1 Gy than 2 Gy, however the foci numbers at the 24 h timepoint are small, and so the differences in the residual levels of DNA damage are small, suggesting repair is effective.
No strong correlation was found between ICP-MS uptake and DNA damage at the both the 1 h and 24 h timepoint, which contrasts previous work with a gadolinium contrast agent (Dotarem) in which there was a positive correlation [
68]. This difference may be due to the higher concentrations of Dotarem used in comparison to the SPIONs, and also that, for iron and irradiation with 100 keV beams, photoelectrons can have energy of around 90 keV, providing a range in water of > 20 microns, which is greater than the dimensions of a cell. Cell lines had taken up varying quantities of SPIONs (measured by ICP-MS), but the cell-specific sensitivity of the cell to the SPIONs seems not be solely due to the concentration within the cell, as demonstrated by U87 cells. Ahmad et al. appears to disagree with these results, finding a positive correlation between levels of DNA damage and uptake with ICP-MS, for Iron Oxide nanoparticles with a mean diameter of 140 ± 4 nm at a concentration of 0.5 mg/ml, which is much higher than the concentration used for this study [
69]. However, Ahmad et al. finds a complex relationship between uptake and radiosensitivity, as there is a negative relationship between dose enhancement and uptake, and this was carried out for 6 MV, not the 225 kVp used in this study. Increasing the SPION concentration may impact the relationship between DNA damage through both physical and biological dose enhancement, therefore further work is necessary to determine this relationship.
Exposure times for the DNA damage and clonogenic assay were 24 h exposure (with the clonogenic assay also having a 1 h exposure time). All cell lines except for MCF7 and U87 cells showed some level dose enhancement by the SER, correlating directly for U87 with the DNA damage results for the 1 h timepoint, yet appears to disagree with the DNA damage seen for MCF7 cells. However, a significant change in the survival curve for 24 h exposure was seen only for H460 and DU145 cells, suggesting higher levels of DNA damage may be repairable in certain cell lines.
The predicted DEF was found only to be 1.0001, compared with SERs found to vary up to 1.28 for HEPG2 cells for 1 h exposure time to SPIONs. This emphasises an important point, that the radiosensitisation of nanoparticles have 2 factors; the physical dose enhancement due to an increase in secondary electrons, and the biological radiosensitisation. For SPIONs, the biological effects are much more prominent, and has implications on the biological functions, likely increasing oxidative stress by increasing levels of ROS and hydroxyl radicals [
12,
15,
16], depending on variables such as the concentration and the energy of the radiation.
Kirakli et al. investigated the impact of SPIONs on radiosensitisation in vitro, through both clonogenic assays and metabolic activity assays for 3 cell lines; MCF7, MDA-MB-231 and MDAH-2774 [
56]. The SER, which they termed nanoparticle-mediated enhancement ratio (NER) was greatest at 2 Gy, and the radiosensitising effects had diminished by 8 Gy [
56]. However, Kirakli et al. performed the study using 6 MV X-rays and there may be cell-type dependent effects. Overall, there is a paucity of data in regard to the radiosensitisation of SPIONs in-vitro, in particular for the kV energy range and specific cell types.
For the in vivo experimentation, all subgroups showed a significant decrease in the rate of tumour growth (
P < 0.05), measured as the time for the tumour volume to reach 300 mm
3. This shows that SPIONs alone impaired the ability for the tumour cells to grow and divide (Fig.
9), however this disagrees with Fig.
8, in which no decrease in clonogenic viability was found. In combination with radiation, SPIONs have a greater impact on tumour growth (Fig.
10), however, it must be noted that the small sample size can lead to uncertain statistics, and should therefore be used only as a guide. SPIONs may be leading to the production of higher levels of ROS and greater DNA damage, as suggested by Dayem et al. [
11] Iron oxide has the potential to cause an increase in ROS production through the Haber–Weiss reaction, in which iron ions generate hydroxyl free radicals [
55,
70]. The results can be compared with a similar in vivo experiment with AGuIX, a gadolinium sub 5 nm nanoparticle currently in a Phase 1 clinical trial [
46]. Dufort et al. performed the in vivo experiment on rats bearing gliosarcoma tumours, demonstrating an increase in the survival time of the rats with microbeam radiation, up to 6.4 keV [
71]. The median survival time was found to increase by a factor of 3.10 and 4.75 for AGuIX injections 1 h and 24 h prior to irradiation respectively, compared with controls [
71], compared with 5.71 for SPIONs plus radiation in this study. The in vivo results validate the in vitro methods, in which H460 cells consistently indicated toxicity as well as a decrease in cell survival and an increase in DNA damage.
One limitation of comparing the in vitro and in vivo experimentation of this study is in the difference between concentrations of SPIONs used and the different incubation times. Due to the nature of the in vivo experiment being a pilot study, it was unknown for how long the SPIONs would remain inside the tumour, therefore a shorter incubation time was chosen. Also, the number of cells in a solid tumour is much greater than that used for the in vitro experiments, hence the much higher concentration of SPIONs used. The fact that the SPIONs were injected intratumourally means also that the EPR effect was not tested in this experiment. These differences should be considered when directly comparing these results.
It is also important to note the difference between the kVp source used for this investigation and the MV energies which are conventionally used for radiotherapy. kVp energies are dominated by the photoelectric effect which is proportional to Z
3 compared with the Compton effect for MV energies which is proportional to Z [
59], suggesting that radiosensitisation from the nanoparticles would be greater for kVp energies. However, the largest differences in the ratios of absorption coefficients occur in the region below 100 keV [
60] which are strongly attenuated by the 2 mm Cu filter used. In particular, the k-edge of iron oxide is < 10 keV, further reducing the effect [
60]. Whilst it is important to acknowledge the differences in the physical dose enhancement caused by the differences in energy, some comparison can be made due to the biological effects being of greater impact.
DLS measurements indicated a hydrodynamic radius of 124 ± 6 nm for the SPIONs, when averaged across a range of concentrations. This is much greater than the 5 nm nominal diameter reported by the manufacturer and may influence the mechanism by which the SPIONs are taken up by cells, although the exact mechanism of uptake is not known. This result is surprising, although comparable with previously published work where Lanier et al. reported differences of the same order of magnitude for hydrodynamic radius’ compared with physical nanoparticle size as quoted by a manufacturer [
72]. This large hydrodynamic radius may suggest the possibility of aggregation of the nanoparticles, however, the measurements also indicated that the SPIONs are predominantly mono-disperse and do not aggregate. Measurements for polydispersity (see Additional file
1) which indicate the distribution of nanoparticle sizes, show a perfect sigmoidal curve starting at ~ 0.9 with a steep decay suggesting no signs of aggregation. Aggregation has been discussed in relation to magnetic nanoparticles by Gutiérrez et al., with evidence that aggregation impacts the interaction of the nanoparticles with cells [
73]. What is not clear from this study, however, is whether aggregation would occur when the SPIONs are exposed to biological fluids, possibly resulting in protein coronas surrounding the nanoparticles (
74).
From this initial study, the next step is to investigate radiosensitising effects with the superparamagnetic properties of SPIONs by combining radiation exposures with an alternating magnetic field to mimic scenarios relevant to MR-Linac exposures, which are expanding in use clinically. Also, further studies using non-cancer cell lines and MV energies should be carried out, and an investigation of the impact of varying SPION concentration on radiosensitisation. To determine in greater detail how radiosensitisation occurs, the mechanism of uptake should also be investigated.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.