Abstract
We simulate the production of high-LET particles from different boron nuclear reactions in proton therapy. This study is motivated by neutron capture-enhanced particle therapy (NCEPT) (Safavi-Naeini et al. in Sci Rep 8(1):1–13, 2018) and the recently proposed proton boron capture therapy (PBCT) (Cirrone et al. in Sci Rep 8(1):1–15, 2018), where both therapy types suggest to utilize the production of high-LET particle for enhancing the RBE of the treatment. From Monte Carlo simulations performed in TOPAS (Perl et al. in Med Phys 39(11):6818–6837, 2012), we find that NCEPT reaction is the more prominent reaction of the two with a factor of \(4000 \pm 700\) more particles produced. The reason for this is the relative high fluence of thermal and epithermal neutrons (neutrons \(< 1 \hbox { keV}\)) present in clinical relevant proton fields. The creation of these neutrons is shown to be highly influenced by the target phantom and proton field size. By varying the field and phantom size, it was found that the fluence of low-energy neutrons can increase/decrease two orders of magnitude.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
In 2018, the first experimental evidence of proton boron capture therapy was published [2]. Here, a biological effect was observed when DU-145 prostate cancer cells had been exposed to sodium borocaptate \({\hbox {Na}_{2}\hbox {B}_{12}\hbox {H}_{11}\hbox {SH}}\) (BSH) during irradiation. The BSH used in this experiment was consisted of natural boron, i.e., roughly 19 % \({{}^{11}\hbox {B}}\) and 81 % \({{}^{10}\hbox {B}}\). The observed effect was a decrease in the cell survival fraction of the BSH-exposed DU-145 cells. The effect was confined to the spread out Bragg peak and not observed in the entry region of the proton beam. It was originally attributed to the production of additional high-LET alpha particles from the \({\hbox {p} + {}^{11}\hbox {B} \longrightarrow {}^{4}\hbox {He} + {}^{4}\hbox {He} + {}^{4}\hbox {He}}\) (boron proton capture, BPC) reaction. As has been mentioned in the original publication as well as others (e.g., [4]), the yield from this reaction in a proton beam is likely very small, since the maximum total cross section is \(\approx 1\) barn at \(600 \hbox { keV}\) [5]. The question of how many \({{}^{4}\hbox {He}}\)-particles are needed to see a biological effect is still unanswered. Independently, another boron reaction has been proposed as a source of high-LET particles in proton therapy, namely, the \({\hbox {n}_{\textrm{th}} + {}^{10}\hbox {B} \longrightarrow {}^{7}\hbox {Li} + {}^{4}\hbox {He}}\) (Boron neutron capture, BNC) reaction, driven by the thermal secondary neutrons, ubiquitous in targets exposed to clinical proton beams [1]. The previous work comparing the BPC and BNC reaction mechanisms in boron-enriched proton therapy is found in ([6, 7]). They both conclude the dose contribution from the BPC reaction to be higher than the BNC reaction. These studies, however, only investigated small pristine proton beams on small phantoms or partially incomplete physics packages. In this study, we will discuss the use of physics packages and also quantify the neutron fluence dependency on field size, energy, and phantom size. When simulated with clinically more realistic parameters, we find a factor of \(4000 \pm 700\) high-LET particle production from NCEPT than PBCT.
2 Methodology
We used the Monte Carlo particle transport code TOPAS 3.7 [3] for our simulations, which were additionally replicated in SHIELD-HIT12A [8] for cross-checking. The field shape used in these simulations is box-shaped fields. The physics modules used in the TOPAS simulations are as follows:
-
g4em-extra
-
g4em-standardopt4
-
g4h-elasticHP
-
g4h-QGSPBICAllHP
-
g4decay
-
g4stopping
We here highlight the importance of g4h-elasticHP. The thermalization of neutrons occurs through elastic scattering; therefore, if this is not included, only high-energy neutrons from the g4h-QGSPBICAllHP module are generated. By scoring the reactions directly in TOPAS, we are able to examine how the g4h-QGSPBICAllHP package handles the BPC and BNC reactions. This is done by scoring the projectile particle, target nucleus, and produced secondaries. We simulate a thin target of pure \({{}^{10}\hbox {B}}/{{}^{11}\hbox {B}}\) to acquire the cross section directly from TOPAS. The acquired cross sections are shown in Figs. 1 and 2 together with experimental cross section from the EXFOR database. It should be noted that Fig. 2 includes two decay channels to the finale triple \(\alpha\) state, which can be seen as the two groupings of experimental data. Namely, \({\hbox {p} + {}^{11}\hbox {B} \longrightarrow 3\,\, {}^{4}\hbox {He}}\) and \({\hbox {p} + {}^{11}\hbox {B} \longrightarrow {}^{12}\hbox {C} \longrightarrow {}^{8}\hbox {Be} + {}^{4}\hbox {He} \longrightarrow 3\,\,{}^{4}\hbox {He}}\), where the prior is the more dominant. The geometry of our setup consists of a 20 \(\times\) 20 \(\times\) 20 \({\hbox {cm}^{3}}\) phantom, with a cubic-shaped field of 3 \(\times\) 3 \({\hbox {cm}^{2}}\) wide with the SOBP positioned at a depth of 7.7–13.3 cm in the beam direction. Our scoring volume has a width of 1 cm \(\times\) 1 cm in the lateral directions, i.e., lateral equilibrium is retained. Scoring along depth is in 300 steps along the entire phantom (20 cm).
After determining the relevant physics modules need for the simulations, we made a parameter study where we varied the phantom and field size while recording the fluence of low-energy protons and epithermal neutrons. This was done by scoring the averaged fluence inside the target volume covered by the SOBP, that is, to say the average fluence between 7.7 and 13.3 cm in Fig. 5. We did simulations with increasing phantom size of 1 cm in each lateral direction pr step. A similar setup can be made with increasing field sizes. Here, the phantom size is fixed at 100 cm \(\times\) 100 cm \(\times\) 100 cm, and the field is increased in the lateral directions 1 cm for each step. For increasing field/phantom size, we normalize the average dose in the SOBP to 1 Gy. We then investigate the fluence of relevant energies of proton and neutrons needed to drive the BPC and BNC reactions.
Next, we investigate the produced secondaries. We note that other work tend to limit the scope to \(\alpha\) particle production but we will here also include \({}^{7}\text {Li}\) from NBC, since its mean free path range is also on the order of a cell diameter (2–6 \({\upmu \hbox {m}}\)). Including a fragmentation model for the PBC reaction in TOPAS is not trivial. Experimentally obtained cross sections disagree with a factor of two in the relevant energy range \(< 2 \hbox { MeV}\) [5, 9, 10], and a correct quantum mechanical implementation of the BPC reaction with break-up amplitudes described in [11] and [12] is out of scope for this work. Instead, we implement a custom scorer in TOPAS for scoring the yield of a reaction provided a given cross-section table. Despite the experimental cross-section disparity of the BPC reaction, the energy range where the maximum cross section is located is consistent. From the EXFOR database, the largest cross section is below \(1 \hbox { MeV}\), peaking at \(\approx 600 \hbox { keV}\) [5]. For energies larger than \(1 \hbox { MeV}\), the cross section decreases with \(\approx 90\%\). We, therefore, chose to score the fluence of protons \(< 2 \hbox { MeV}\) to quantify the protons available for the PBC reaction. For neutrons, the BNC reaction still has a cross section > 10 barns at \(1 \hbox { keV}\) [13], so scoring fluence of neutrons \(< 1 \hbox { keV}\) is representative of the BPC reaction yield. The maximum cross section for BCN is well established, found at \(3839 \pm 9\) barns for \(0.253 \hbox { eV}\) [14]. Additionally, the yield of the two reactions was estimated. The proton and epithermal fluence were also obtained independently using SHIELD-HIT12A [8].
3 Results
We here present the findings of work described in the methods section. The results will be divided into three sections.
3.1 Geant4 physics package verification
When investigating the Geant4 physics package g4h-QGSPBICAllHP, we start by plotting the cross section for the BNC and BPC reactions. These are shown in Figs. 1 and 2. Figure 1 shows good agreement with the experimental results of the BNC reaction for protons \(< 1 \hbox { MeV}\). Above \(1 \hbox { MeV}\) was not simulated. For the BPC reaction, there is no cross section for protons below \(1 \hbox { MeV}\) available in the g4h-QGSPBICAllHP package (since it is based on TENDL version 1.3.2). Furthermore, the reaction products simulated with this package are inconsistent: From a given nuclear interaction between protons and \({}^{11}\text {B}\), particles different from alphas are frequently generated, violating baryon number conservation. This observation has only been seen in the g4h-QGSPBICAllHP physics package for the BPC reaction. Other reactions from the same and different physics packages have not shown this issue. This makes the g4h-QGSPBICAllHP physics package unfit for simulating BPC reactions (for TOPAS 3.7), since both the reaction yield and dose contribution are likely to be incorrect. We, therefore, provide a custom extension for TOPAS to calculate the yield of a given reaction provided a cross section as an input file. This extension does not generate the reaction products, and therefore, the particle dose contribution from this reaction is not established in this work. The yield can, however, provide a relative comparison between the BPC and BNC reactions.
3.2 Geometrical parameter study of low-energy neutron and proton fluence
We now turn to the geometrical effects of neutron thermalization in our setup. We plot the average fluence of low-energy neutrons and protons in the SOBP inside the proton beam vs lateral phantom size, which is shown in Fig. 3.
From Fig. 3, it is observed that increasing the phantom size in the lateral directions can increase the average fluence of low-energy neutrons in the SOBP with up to an order of magnitude. This can explain why articles like [6] do not find the BNC reaction to produce more high-LET particles than the BPC reaction, since there simulated phantom size was 4 \(\times\) 4 \(\times\) \(10 {\hbox { cm}^{3}}\). Which from figure 3 means that they are producing close to an order of magnitude less low-energy neutrons, than in a clinically more realistic phantom size. For this specific beam profile, it is shown that you need at least a 10-cm wide and tall phantom to reach lateral equilibrium for neutron thermalization. This effect should be understood in terms of the thermal neutron “soup” as described in [1]. Thermal neutrons are generated from scattering of higher ( \(> 10 \hbox { MeV}\)) energy neutrons. The neutrons are generated mainly from different inelastic scattering on oxygen directly or decay of \({{}^{5}\hbox {He}}\) secondaries (based on simulations). The thermal neutrons move in the phantom without a common direction, thereby filling the phantom as a kind of “soup”. Equilibrium is then reached when there is as many neutron moving out of the SOBP area as moving in. The equilibrium point is specific for each beam size and energy. Meaning you could also have the phantom size constant and increase the field size and have a similar result. This is shown in Fig. 4. In [7], they used a field diameter of 2 mm, which compared to Fig. 4 is not representative of the potential neutron fluence obtained with a more clinically relevant field size. This could explain why they find the BNC reaction to have a lower impact than the BPC reaction.
We conclude that increasing the field size can gain another order of magnitude increase in neutron fluence in the SOBP, for larger phantoms. The field and phantom size chosen for experiments/simulations can thereby change the produced low-energy neutrons with up to two orders of magnitude.
3.3 \({}^{11}\)B and \({}^{10}\)B reaction yield comparison
We investigate the low-energy neutron fluence inside the proton field only, where the BPC reaction competes with the BNC reaction. A convenient way to visualize this is to plot the fluence of the particles needed for the two reactions, that is, neutrons \(< 1 \hbox { keV}\) and protons \(< 2 \hbox { MeV}\). This can be plotted along the beam axis within the proton field, as shown in Fig. 5.
Here we have a close to two orders of magnitude difference between neutrons \(< 1 \hbox { keV}\) and protons \(< 2 \hbox { MeV}\). The yield for the two reactions can be calculated as follows:
Where \(Y_x\) is the yield for reaction x, \(\Phi\) is the fluence of the participating projectile, \(\sigma _x\) is the microscopic cross section for reaction x, and \(\rho\) is the density of atoms in the target. We can calculate the ratio between the two reaction yields as follows:
We estimate the expected R value based on Fig. 5 and determine each fraction in (2) separately. \(\Phi _n/\Phi _p = 10^{2}\) on average in the SOBP. The \(\sigma _{\textrm{BNC}}/\sigma _{\textrm{BPC}}\) = 10, from choosing the lowest possible \(\sigma _{\textrm{BNC}}=10\) barn and the highest possible \(\sigma _{\textrm{BPC}}=1\) barn. \(\rho _{\textrm{B10}}/\rho _{\textrm{B11}} = 1/4\) since we work with natural boron (\(80\% {{}^{11}\hbox {B}}\) and \(20\% {{}^{10}\hbox {B}}\)). Since BPC produces three high-LET particles and BNC produces two, we multiply R with 2/3 to get the ratio of produced particles and not just ratio of number of reaction. Here, it should be noted that we ignore the increased range of the reactions products from BPC reaction compared to BNC. The result is \(R =17.5\), which means that we produce a factor of 17.5 more high-LET particles from the BNC reaction than from the BPC reaction. If we chose the highest \(\sigma _{\textrm{BNC}}\) and lowest \(\sigma _{\textrm{BNC}}\), we get \(R = 46.67 \cdot 10^{3}\). From the preceding discussion, using a 2-mm thick beam as in [7], the neutron and proton fluences are approximately \(14 \times 10^{4}/\text {cm}^{2}\) and \(6 \times 10^{4}/\text {cm}^{2}\), respectively (derived by fitting a second-order polynomial to Fig. 4). This signifies that compared to an 8-cm field, we see from Eq. 2 that the neutron contribution is underestimated by nearly two orders of magnitude.
This suggests that the main reason for the observed enhanced RBE in [2, 15] arrives from the BNC reaction. Note also that the thermal neutron fluence is reduced by one order of magnitude in 1–2 cm from the phantom surface, such that the dominating BNC reaction is suppressed near phantom surfaces (and therefore in particular small phantoms). While the BPC reaction undoubtedly will add to the high-LET particle yield, it is unlikely to be the main contributor by at least an order of magnitude.
To obtain a more precise estimation of the R value, we developed a custom code for calculating the yield of a specified reaction using TOPAS. For the PBC reaction, we utilized the cross-section data from Fig. 2 from the EXFOR database, while for the BNC reaction, we relied on the g4h-QGSPBICAllHP physics package that was verified in Fig. 2. Our boron target was a cell flask placed in the SOBP (at 9-cm depth, see Fig. 5) with dimensions of 92 \(\times\) 51 \(\times\) 29 mm, which corresponds to the size of an actual culture flask used at the Danish Centre of Particle Therapy. We selected a boron medium concentration of 17 mg/mL, which is 100 times higher than the concentration used in the original experiments [2, 15]. This concentration, irradiated with a 10 cm \(\times\) 10 cm field (SOBP positioned at a depth of 7.7–13.3 cm as shown in Fig. 5), produced approximately 1 PBC reaction for every 10.000.000 simulated protons. We simulated 200.000.000 protons, resulting in an R value of \(4000 \pm 700\), with the uncertainty calculated from counting statistics. On average, we generate 4.65 alpha particles per 10.000.000 protons for the PBC reaction, compared to 12,440.6 alphas and \({}^{7}\text {Li}\) per 10,000,000 protons for the BNC reaction. Using the highest possible experimental cross sections from Fig. 2 (approximately 1 barn) from 0.14 to 45 MeV yields an R value of \(22.2 \pm 1\).
4 Conclusion
We show that the fluence of low-energy neutrons (\(< 1 \hbox { keV}\)) strongly depends on the lateral phantom size as well as the field size, which was not taken into account in the previous studies of PBCT. Changing either may affect the low-energy neutron fluence with up to two orders of magnitude. The BNC reaction produced \(4000 \pm 700\) times more high-LET particles than the BPC reaction in the SOBP, when considering clinical dimensions for phantom and field size. Whether the BNC (and BPC) are biologically relevant is out of scope for this work, but remain to be investigated.
Data availability
Any data and/or code that were not included in the manuscript can be made available upon reasonable request to the corresponding author.
Availability of data and materials
Yes.
Code availability
Yes.
References
M. Safavi-Naeini, A. Chacon, S. Guatelli, D.R. Franklin, K. Bambery, M.-C. Gregoire, A. Rosenfeld, Opportunistic dose amplification for proton and carbon ion therapy via capture of internally generated thermal neutrons. Sci. Rep. 8(1), 1–13 (2018)
G. Cirrone, L. Manti, D. Margarone, G. Petringa, L. Giuffrida, A. Minopoli, A. Picciotto, G. Russo, F. Cammarata, P. Pisciotta et al., First experimental proof of proton boron capture therapy (pbct) to enhance protontherapy effectiveness. Sci. Rep. 8(1), 1–15 (2018)
J. Perl, J. Shin, J. Schümann, B. Faddegon, H. Paganetti, Topas: an innovative proton Monte Carlo platform for research and clinical applications. Med. Phys. 39(11), 6818–6837 (2012)
A. Mazzone, P. Finocchiaro, S.L. Meo, N. Colonna, On the (un) effectiveness of proton boron capture in proton therapy. Eur. Phys. J. Plus 134(7), 361 (2019)
H. Becker, C. Rolfs, H. Trautvetter, Low-energy cross sections for11b (p, 3\(\alpha\)). Z. für Phys. A At. Nuclei 327(3), 341–355 (1987)
M.-S. Kim, M. Wai-Ming Law, S.-K. Djeng, H.-B. Shin, M.-G. Choi, Y.-J. Kim, B.-Y. Choe, T.S. Suh, D.-K. Yoon, Synergy effect of alpha particles by using natural boron in proton therapy: computational verification. AIP Adv. 9(11), 115017 (2019)
F. Tabbakh, N.S. Hosmane, Enhancement of radiation effectiveness in proton therapy: comparison between fusion and fission methods and further approaches. Sci. Rep. 10(1), 1–12 (2020)
N. Bassler, D.C. Hansen, A. Lühr, B. Thomsen, J. Petersen, N. Sobolevsky, Shield-hit12a-a Monte Carlo particle transport program for ion therapy research. J. Phys. Conf. Ser. 489, 012004 (2014)
R. Segel, S. Hanna, R. Allas, States in c 12 between 16.4 and 19.6 mev. Phys. Rev. 139(4B), 818 (1965)
O. Beckman, T. Huus, Č Zupančič, Excitation curves for \(\alpha\) particles from b 11 bombarded with protons. Phys. Rev. 91(3), 606 (1953)
D.P. Balamuth, R.W. Zurmühle, S.L. Tabor, Isospin-forbidden alpha decay of the 15.11-mev state in \({}^{12}{{\rm C}}\). Phys. Rev. C 10, 975–986 (1974). https://doi.org/10.1103/PhysRevC.10.975
H.O.U. Fynbo, Y. Prezado, U.C. Bergmann, M.J.G. Borge, P. Dendooven, W.X. Huang, J. Huikari, H. Jeppesen, P. Jones, B. Jonson, M. Meister, G. Nyman, K. Riisager, O. Tengblad, I.S. Vogelius, Y. Wang, L. Weissman, K.W. Rolander, J. Äystö, Clarification of the three-body decay of \({}^{12}{{\rm C}}\) (12.71 mev). Phys. Rev. Lett. 91, 082502 (2003). https://doi.org/10.1103/PhysRevLett.91.082502
E. Bavarnegin, Y. Kasesaz, F. Wagner, Neutron beams implemented at nuclear research reactors for bnct. J. Instrum. 12(05), 05005 (2017)
A. Prosdocimi, A. Deruytter, A precise determination of the thermal neutron absorption cross section of b10 and natural boron by time of flight. J. Nucl. Energy Parts A/B React. Sci. Technol. 17(2), 83–87 (1963)
P. Bláha, C. Feoli, S. Agosteo, M. Calvaruso, F.P. Cammarata, R. Catalano, M. Ciocca, G.A.P. Cirrone, V. Conte, G. Cuttone, et al., The proton-boron reaction increases the radiobiological effectiveness of clinical low-and high-energy proton beams: Novel experimental evidence and perspectives. Front. Oncol. 11, 682647 (2021)
S. Agostinelli, J. Allison, K.A. Amako, J. Apostolakis, H. Araujo, P. Arce, M. Asai, D. Axen, S. Banerjee, G. Barrand et al., Geant4-a simulation toolkit. Nucl. Instrum. Methods Phys. Res. Sect. A Accel. Spectrom. Detect. Assoc. Equip. 506(3), 250–303 (2003)
Acknowledgements
This research was supported in part by PL-Grid Infrastructure (Grant plgccbmc11).
Funding
Open access funding provided by Aarhus University Hospital. Novo Nordisk Fonden: NNF195A0059372, Danish Comprehensive Cancer Center.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
No conlflict to declare.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Jacobsen, V.L., Johansen, J.G., Fynbo, H.O.U. et al. A Monte Carlo study of high-LET particle production in proton boron therapy. Eur. Phys. J. Plus 138, 625 (2023). https://doi.org/10.1140/epjp/s13360-023-04235-3
Received:
Accepted:
Published:
DOI: https://doi.org/10.1140/epjp/s13360-023-04235-3