Particle therapy and nanomedicine: state of art and research perspectives

Conventional radiotherapy is applied in 50% of cancer treatments. Based on the properties of high-energy photons to traverse the entire body, this non-invasive method is used to treat deeply seated tumours. However, as the interaction of photons is not tissue specific, severe side effects or even secondary cancers may be induced when healthy tissues are damaged. It is thus a major challenge to develop new strategies and improve the tumour selectivity of radiation effects.

The enrichment of tumours with high-Z compounds has been proposed as a new strategy to improve the effects of radiation as due to the amplification of primary (electronic) processes. To avoid confusion with radiosensitizing drugs, those compounds that make cells more sensitive to radiation, such as DNA repair inhibitors, oxygen transporters [see for instance (Lawrence et al. 2003)], in this review, we use the term “nano-radio-enhancers” (NRE) to distinguish these compounds.

The principle of radio-enhancement was first demonstrated using metallic complexes to increase the effects of high-energy photons [see (Kobayashi et al. 2010) for a review]. The clinical use of these compounds is, however, limited by the lack of tumour selectivity. Hence, nanoparticles (NPs) have been proposed as a more efficient means to improve the concentration of active products in the tumour and, as a consequence, to improve the tumour targeting of radiation effects. The selective delivery of NPs is due to the enhanced permeability and retention effect (EPR) when the systems are small enough (diameter <200 nm) to permeate through the tumour blood vessel walls (Jäger et al. 2013). Tumour targeting may also be achieved when nanoparticles are functionalized with tumour specific agents such as antibodies or other peptides [see (Friedman et al. 2013) for review]. Thus, the combination of radiation therapies with nanomedicine opens a new range of treatments (Kong et al. 2008). Hainfeld et al. (2008) were the first to show that 1.9 nm gold core NPs prolong the life of mice treated with 160 kV X-rays. Gold NPs are presently the most well studied agents [see (Her et al. 2017) and (Haume et al. 2016) for review]. Other sophisticated NPs, composed of other heavy elements such as hafnium (Maggiorella et al. 2012) and gadolinium (Sancey et al. 2014) developed by Nanobiotix (Paris, France) and NH TherAguix (Villeurbanne, France) respectively, are already being transferred to the clinic.

Although conventional radiotherapy has been tremendously improved (e.g., with the IMRT technique), the use of highly penetrating photons remains critical for the treatment of tumours located in close vicinity of sensitive organs (i.e. eyes, brain, neck) and the treatment of paediatric cases, where damage of surrounding tissues can have severe consequences. The latter are mainly related to the geometry of the irradiation (e.g. in a typical craniospinal irradiation for a medulloblastoma, the dose to the spine is extremely dangerous) and to the young age of the patients, which emphasizes later risk effects (Armstrong et al. 2010). Moreover, conventional radiotherapy is not able to eradicate rare but highly aggressive radioresistant cancers such as glioblastoma and chordoma, for which the treatment outcomes remain poor. For these cases, treatment by high-energy ions such as protons (proton therapy) and carbon ions (carbon therapy) is being proposed as an alternative (Durante et al. 2017). The main advantage of ion beams (70–400 MeV/amu) stems from their property to penetrate tissues over several centimeters and deposit the maximum energy at the end of their track, where the ionization cross section of the medium is extremely large and at a depth dependent from their initial energy, forming the so called Bragg peak in a depth dose profile (Schardt et al. 2010). Thus, the beam may be tuned by modulating its energy to target the tumour without damaging the tissues located at a deeper position [see Fig. 1)]. Moreover, thanks to a larger relative biological effectiveness (RBE) associated to ion beam radiation as compared to X-rays due to its more densely ionizing feature providing greater cell killing for the same amount of delivered dose (Scifoni 2015), particle therapy is also the most efficient method to treat radioresistant tumours (Ares et al. 2009; Schlaff et al. 2014; Kamada et al. 2015; Durante et al. 2017). Carbon ions in particular can, in some cases, be four times more efficient than X-rays (Loeffler and Durante 2013; Kamada et al. 2015). Particle therapy is thus considered, at least for a number of indications, superior to conventional radiotherapy (Baumann et al. 2016) and, in spite of the high costs, new centres of proton therapy and carbon therapy are developing worldwide. In fact, beyond the 74 centres already in operation as of April 2017, 83 new centres have already started construction [e.g. in Dallas (USA) and Lanzhou (China)] and at least another 40 (e.g. in Australia, India, Denmark and Netherlands) are in the planning stages [see (Jermann 2015; Zietman 2016) for recent printed reviews and the PTCOG dedicated website for most updated data: https://www.ptcog.ch/index.php/facilities-in-operation].

Particle therapy is delivered with two different modalities. One is the passively modulated broad beam modality, which consists of a beam shaped to the target with a spread out Bragg peak (SOBP). The second is the recent pencil beam active scanning mode, where a beamlet of a few mm is scanned, spot by spot, on the tumour, modulating the energy for each depth slice (Schardt et al. 2010). Because of its larger degradation of the beam through the beamline materials, the broad beam modality usually provides a larger entrance channel dose, as compared to the pencil beam (Shiomi et al. 2016).

Hence, because of the physical profile of the beam, a low but significant dose deposited by the ions in the tissues located before reaching the tumour [see Fig. 1b] is unavoidable. Moreover, damage to surrounding tissues may be caused by motion and a range of other uncertainties.

To overcome these limitations, the addition of NREs to the tumour is proposed as a challenging strategy to amplify the effect of ion radiation locally and thus reduce the total dose to the patient. The use of contrast agents, in particular, offers the possibility to follow the biodistribution of the agent as well as to image the tumour just prior to or during the treatment. While nanomedicine is now approaching a clinical stage in conventional radiotherapy, only few studies have been dedicated to the combination of high-Z NREs with ion beam modalities.

This review summarizes the first experimental and modelling studies that display and tentatively describe the effects of different radio-enhancers, including metallic complexes and NPs, used to improve the performance of particle beam treatments, e.g. protons, helium and carbon ion radiation. The first section exposes the major results reported on the effects of (i) platinum complexes activated by different ion radiations (helium, carbon, iron), (ii) gold NPs combined with proton radiation and (iii) platinum NPs and gadolinium-based nanoagents (AGuiX) combined with carbon radiation. In the second section, the recent modelling and simulation studies dedicated to radio-enhancement induced by ion radiation are collected together with a summary of the known results and the remaining open questions to be faced.

Modelling of the nanoscopic mechanisms involved in nanoparticle induced radio-enhancement was first undertaken in the case of photon irradiation. The amplification of radiation effects in this case was explained in terms of a nanoscale enhancement of the local dose in close vicinity to the NPs. This was demonstrated by McMahon et al. (2011) and recently refined by Brown and Currell (2017), thus explaining the results of several experiments through adapting the Local Effect Model (LEM) (Scholz and Kraft 1996) initially developed for ion beams. This model, in its simpler formulation (LEM I), predicts a higher cell killing for higher densely ionizing (LET) radiation, correlating a higher spatial concentration of ionizations on a biological target, and then the induction of more severe damage to a higher probability to induce a lethal effect and than cell death. It was seen that simply including the high local enhancement of the dose due to Auger electrons can lead to a significant effect on the radial dose, which then induces an increase in cell killing quantified by a Sensitizing Enhancement Ratio (SER), i.e. a ratio of doses giving the same biological effect with and without sensitizer, in a way similar to an RBE (McMahon et al. 2011).

In the case of ion beam irradiation, an enhancement of radiation effects was observed in the presence of nanoparticles either at the molecular (DNA damage), in vitro (cell killing) and in vivo (mouse tumour regression) levels, as discussed in the previous section and listed in Table 1. However, the mechanistic explanation of local dose enhancement provided for photons is not the same as for ions. In the studies with photons, it was shown that a large increase in the radial dose profile was induced in the presence of NP as compared to photon irradiation in water, enough to justify the sizeable difference in the yield of severe damage. However, in the case of ions, the dose is already highly localized along the tracks, and an extremely high local dose would be required to induce an additional impact on the damage concentration, without even accounting for over-kill effects. In this case, the enhancement of radiation effects is not, as yet, fully understood. The first study approaching this problem (Wälzlein et al. 2014) was conducted using the particle track structure code TRAX (Krämer and Kraft 1994) to analyse, at a nanoscale level, a possible dose enhancement in high-Z nanoparticles (Au, Pt, Ag, Fe and Gd) traversed by proton beam (see Fig. 4). It was found that a relevant increase in local dose around the nanoparticle could be computed, but the relative enhancement was much smaller than that observed in photon irradiation. Moreover, the simulation was performed in the condition of ion traversing across the nanoparticle, which with typical fluences adopted in proton therapy (10⁶ to 10⁹ cm⁻²) is very rare. Thus, the dose enhancement effect occurring in the case of an ion traversal should be weighted by this very low probability to occur (≈10⁻³ to 10⁻⁴). In total, this would lead to a noticeably reduced overall dose enhancement effect. This study has shown a larger effect of gold and platinum, as compared to other high-Z materials, in acting as dose enhancers. More importantly, it demonstrated that, for proton radiation, a significant dose enhancement effect can be observed, mostly due to Auger electrons and consecutive cascades. However, this process is not sufficient to justify any overall macroscopic effect such as those observed in several experiments.

The amplification effect of ion radiation by high-Z NPs may be explained by other mechanisms, such as modification of the radiation chemistry pathways and enhancement of radical mediated component of radiation damage, as suggested with X-rays (Sicard-Roselli et al. 2014).

Gao and Zheng (2014) explored different proton energies and found that a larger number of electrons escape the nanoparticles for lower primary ion energy. These electrons have lower energies and shorter ranges compared to those induced by more energetic protons (Gao and Zheng 2014). Lin et al. (2014) attempted to establish comparative figures of merit between protons and different types of photon radiation (Lin et al. 2014) and proposed a model for biological effect calculation (Lin et al. 2015) based on the Local Effect Model. The result pointed out the need of a much higher nanoparticle uptake in the case of protons as compared to photons, in order to observe a similar enhancement effect. This concentration should be even higher for protons of lower energies for the emitted electrons of lower range to reach and affect sensitive cell components.

Verkhovtsev et al. (2015a, b) proposed the idea of a new  channel through surface plasmon excitation, which was shown to strongly link to a large production of secondary electrons, thus arguing a new pathway for dose enhancement [Verkhovtsev et al. (2015a, b]. The authors showed, for 1 MeV protons, an increase of an order of magnitude in the emitted electron spectra, as compared to direct ionization.

Other studies, using Monte Carlo calculations, have been performed focusing on macroscopic dose enhancement due to the absorbed physical dose only (Ahmad et al. 2016; Cho et al. 2016). The effect was found to be very small for realistic values of NP concentrations.

A recent study (Martínez-Rovira and Prezado 2015) confirmed that a nanoscale dose enhancement, based on physical boost of electron production alone, cannot explain the amplification effect observed in experiments and that radiation chemistry or biological pathways should also be taken into account (Wälzlein et al. 2014). A critical summary of Monte Carlo studies on proton interaction with NP has been collected in Verkhovtsev et al. (2017).

A recent study attempted to include the physico-chemical and chemical stage in this process for protons of 2 to 170 MeV traversing a gold NP, using a combination of GEANT4 and GEANT4-DNA (Tran et al. 2016). Despite the underestimation of secondary electrons production at low energy inherent to the model, this study emphasized an interesting “radiolysis enhancement factor “, i.e., an increased radical production due the presence of the gold NP, which increases with the energy of the incident particle.

In Fig. 5, we show a scheme that summarizes all the mechanisms proposed in these studies.

Thus, despite the fact that several questions have been answered, modelling of the enhancement of ion beam effects with NPs is just at its initial stage. There is a large need for further studies. In particular, before entering the radiobiological effects, the first parameters to be verified are the cross sections of the pure physical processes, which are needed in the simulation codes. While many studies are focused on detecting a biological effect, the physics itself has still to be fully elucidated. For example, both elastic and inelastic cross sections in high-Z materials like gold have still not been characterized in detail, and relevant differences appear, e.g. when using the standard Livermore library (Wälzlein et al. 2014). Studies in this direction are now ongoing, providing, for the moment, a partial confirmation of the validity of the cross section sets used in TRAX (Hespeels et al. 2017).

As for the search of the ideal conditions of radio-enhancement, only effects of incident protons have been simulated, and there is no indication of a possible trend of the track structure effect, thus emphasizing an ion type dependence (beyond pure LET), as has been demonstrated for the RBE (Friedrich et al. 2013). As for the pure energy (or LET) dependence, despite some indications, there is still not a complete explanation of the enhancement effect. In particular, from experiments, this dependence appears counter-intuitive, pointing to a larger effect for higher LET, while one should expect a larger enhancement for a more “photon-like” radiation type. The challenges arising from these studies will probably stimulate research not only to shed light on the specific mechanism, but also on reconsidering the general paradigm of radiation bio-damage (Scifoni 2015).

In addition, the role of oxygenation of the medium (quantified by the Oxygen Enhancement Ratio—OER) may be significant. The OER with ion beams shows a strong peculiarity, decreasing with high LET (Furusawa et al. 2000). Thus far, the OER effect associated with the presence of nanoparticles has not yet been considered, aside from a study with photons where anoxic cells appeared to be not sensitized by NPs (Jain et al. 2014). However, this effect could be different with ion beams, and the potential to additionally sensitize hypoxic cells with NPs is very attractive. Last, but not least, it will be necessary to explicitly study the case of radio-enhancement mediated by NPs in the cytoplasm. In fact, as discussed above, it is now almost established, from most of the prior studies, that the enhancement of cell killing is induced by nanosensitisers located in the cytoplasm (Usami et al. 2008b; Porcel et al. 2010; Stefančíková et al. 2014), despite the fact that, as mentioned in the previous section, a few studies have also found NPs in the nucleus (Li et al. 2016). This type of study was initiated for photons, pointing to mitochondria as possible sensitive targets (McMahon et al. 2017). In the case of ions, these targets will have a completely different and probably more complex scenarios.

The development of nanoagents to improve the performance of particle therapy is only at its beginning. Several studies already demonstrated the feasibility of this strategy, but the efficacy of nanoparticles must be further optimized to be of clinical interest for radio-oncologists.

The results obtained with several nanoparticles are already promising but greater efforts are needed to improve active tumour targeting, renal clearance, and detection of the agents by medical imaging (CT or MRI). The nanoagents of the future will have various designs (i.e. nanoparticles, nanocages, nanocarriers [see for instance (Horcajada et al. 2010; Yu et al. 2012; Kunz-Schughart et al. 2017)] and will offer unique perspectives to combine different modalities using the same compound. For instance, NPs able to act on the immune system, such as those proposed for some cancer treatments (Dimitriou et al. 2017; Ebner et al. 2017), will be of particular interest for particle therapy.

In parallel, the mechanistic sequences involved in the enhancement of ion radiation effect, which is needed for predictive assessments, are not yet fully revealed, but a number of clear pictures are emerging. However, in order to appropriately simulate the enhancement effect and introduce the concept in treatment planning, the explicit description of the radiation chemistry, initiated after the physical step, will be required.

The association of particle therapy and nanomedicine is a new era. Its evolution depends on the capacity of the different communities to share their expertise in developing competitive nanoagents and predictive models. In this context, a collaborative European research programme entitled Marie Curie ITN “ARGENT” (http://itn-argent.eu) has been initiated (Bolsa Ferruz et al. 2017).