HHelium beam particle therapy facility
Mariusz Sapinski ∗ GSI, Darmstadt, GermanySEEIIST, CERN, Geneva, Switzerland
October 1, 2020
Abstract
Due to its precision and limited side effects, the particle therapy of canceris gaining popularity. The number of patients treated with protons and lightions reached 150,000 worldwide. There are currently more than 80 facilities,which offer this treatment and several dozen of new ones are in construction.Mostly they are cyclotron-based facilities, which provide proton beams, andonly several are synchrotron-based facilities that provide both: proton andcarbon-ion beams. The advantage of carbon ions is their higher efficiency indestroying the cancer cells, more localized shape of the Bragg peak and smallerbeam scattering in the body. Several of the large centers, which can providecarbon, have experimented with other ions including helium. The advantageof helium over protons is significantly larger precision of the treatment due tolower scattering of alpha particles in the body, what is especially importantfor pediatric patients. A facility with therapeutic beams limited to helium,would be smaller and cheaper than carbon-based facility. Here we discuss thisoption. ∗ mail: [email protected] a r X i v : . [ phy s i c s . acc - ph ] S e p Introduction
In comparison to protons, the helium ions provide reduced lateral scattering andenhanced biological damage (relative biological effectiveness, RBE) to tumors. Atthe same time the particle fragmentation effects in distal healthy tissues are muchsmaller than for heavier ions like carbon. Multiple experiments with helium beamshave been carried out in various facilities [1]. The enhancement of RBE with respectto protons is moderate [2], however the precision of the helium beam due to reductionof lateral scattering is significantly improved [3]. This is an important factor forpediatric patients, where tumors are small and sensitive organs are very close tothe tumor. The reduction of the fragmentation tail with respect to heavier ions isillustrated on the left plot of Figure 1. The comparison of lateral scattering, shownon the right plot of the Figure 1, means increased precision of the dose delivery.Figure 1: Comparison of properties of the ions used in cancer radiotherapy. Leftplot shows absolute longitudinal dose profile. Lack of fragmentation tail is visiblefor protons and helium. Right plot shows the width of the lateral dose fall-off dueto multiple scattering. Figures from [4] (left) and [5] (right plot).The energy of the ion beams required by the therapy are determined by their rangein the body. The irradiation procedures require the range in the body to be from3 cm (shallow tumors) to 33 cm (deep-seated tumors) what corresponds to He beam energies between 60 MeV/u and 250 MeV/u. The relation between energy E k and range R (position of the Bragg peak) has approximately a form R = α · E pk ,where α is a constants and p is in the range 1.78-1.82. The range tables for protonsand helium ions can be found in [6].The maximum He beam energy of 250 MeV/u corresponds to magnetic rigidityof 4 .
85 Tm what is only about 70% of 6 .
62 Tm required for carbon ion beams. The70% scaling factor gives an approximate idea how the size and cost of the helium-beam facility compares to carbon-beam facilities.2
Current trends in accelerators for radiotherapy
Presently, the particle radiotherapy facilities are dominated by cyclotrons for protonsand by synchrotrons for heavier ions. The main reason is that proton cyclotrons arevery compact and easy-to-operate machines. For instance the IBA S2C2 supercon-ducting synchrocyclotron has a diameter of about 2.5 meters and weights around50 tons [7]. The synchrotrons and linacs are larger than cyclotrons, with the mostcompact proton synchrotron - new Hitachi design - occupying a square of 5 × . m surface, without the injector [8] (see Figure 2).One of the most bulky element of a radiation therapy facility is shielding, whichprotects the patients, the staff and the public from the radiation produced due tobeam losses. The most troublesome component of this radiation are high energyneutrons, which are difficult to shield because they go through multiple elastic scat-tering events before they are stopped, therefore they can ’leak’ through cracks andopenings in the radiation shield. This radiation is mainly generated in extractionprocess and by Energy Selection Systems (ESS). The compact medical cyclotronscannot use charge exchange extraction, which is very efficient and minimizes beamlosses, because negative ions dissociate in strong magnetic fields. Therefore, the mostcommon is extraction in which orbits of the last turns are excited using resonance,creating separation of the last turn orbit and peeling off the beam by electrostaticdeflector. Synchrotrons also use resonant extraction, but applied to the whole beamin the machine. In both cases significant percentage (10-30%) of the beam is lost.However, the main source of beam losses are degrades of the ESS system, which isused to tune the beam energy. Those degraders also deteriorate the beam quality.The ESS is needed only for cyclotrons, because synchrotrons can vary their extrac-tion energy. Therefore, the synchrotrons usually require thinner shielding, however,due to their larger size, a much larger space must be shielded. Therefore, the protonsynchrotrons still compete on the market with cyclotrons.A relatively new technology on the proton therapy market are the high-frequencylinacs, which offer a fast pulse-to-pulse energy variation with frequency of 200 Hz [9].This is an emerging technology and one of those linacs is currently under construction(AVO-ADAM, UK).A summary of the technologies is presented in Table 1. Note that comparisonTable 1: Comparison of the main features of particle therapy accelerators.cyclotrons synchrotrons linacsbeam energy fixed variable variablebeam structure continuous spills (0.2-10 s) pulses of 10 µ s at 200 Hzsize (p,He,C)[m](diameter/length) 2.5, 5.2, 6.9 5.5, 15, 20 24, 40, 54technology state established established emerging3f sizes is not straightforward; the linacs are much longer than synchrotrons andcyclotrons, but their width is much smaller. In the Table the length of not foldedlinac is given. The final footprint of facilities based on linacs is smaller than normal-conducting synchrotrons.For ions heavier than protons, the advantage of synchrotrons over cyclotronsbecome more emphasized. The larger maximum rigidities of the beams require muchlarger cyclotron magnets, therefore currently there is only one project of carbon-cyclotron therapy machine [10], while the other facilities are based on synchrotrons.A helium linac was proposed in [11] but currently, to author’s knowledge, thereare no projects focused on delivering helium-only therapeutic beams. Because ofmedical advantages of helium over protons and significant reduction of cost andsize of the facility in comparison with carbon ions, a helium-beam cancer treatmentfacility is investigated. All three types of accelerators can be considered, however,because the linac technology is still emerging and the helium cyclotron is a massivedevice, here an example of synchrotron-based facility is drafted. The proposed system is based on classical concept of injector linac and synchrotron,but makes use of exciting recent developments. The crucial development is theinjector, made of 750 MHz RFQ developed at CERN, which is the most compactRFQ developed to date. This RFQ could be used as an injector alone, withoutadditional accelerating structures. This was already suggested in the 1991 [12],however, at that time the RFQ output energies were too low for efficient injectioninto synchrotron. The modern RFQ can provide higher beam energy and the directinjection to the synchrotron, without additional accelerating structures, should bereconsidered. The use of RFQ direct injection alone makes a significant difference inthe complexity and price of an ion therapy facility.
The beam is produced by two Electron-Beam Ion Sources (EBIS). Those sourcesproduce beams with much smaller emittance than the Electron-Cyclotron ResonanceIon Sources (ECRIS) used in ion-therapy facilities. Due to lower pressure EBISsources generate less impurities what particularly important as from the beginningthe accelerators are tuned to charge-to-mass ratio of 1:2, what could lead to manyother ions being accelerated, for instance H +2 , C or O . The main disadvantageof EBIS source is a low beam current. For instance a compact Dresden-EBIS-SCsource can produce only 2 · He ions per pulse. The pulse length can beregulated in a broad range 0 . µ s. The total number of particles, which areneeded in the synchrotron to treat 2 dm tumor, is about 8 · , what corresponds to40 pulses, however the transmission through the injector is only about 50%. Multi-turn injection over such large number of turns is challenging. Nevertheless these4arameters are close to the requested ones and ongoing developments give hope thatEBIS sources can replace ECRIS in medical facilities in the next few years [13]. Twoof such sources should be installed to allow for fast switching in case one of themfails.The typical injector linac is a complex system with RFQ and IH structures andaccounts for about 30% of the cost of the facility. The 750 MHz RFQs featuressmall cross section of the cavity and innovate beam dynamics which allow for avery efficient acceleration [14]. One of these RFQs, designed for protons, has beeninstalled on LIGHT proton therapy linac. The design based on similar assumptionshas been prepared for C ions [15] and the same design could be used for He beam.This design foresees the output energy of 2.5 MeV/u or 5 MeV/u. Especiallyin the latter case, the RFQ could be the only accelerating element of the injectorlinac. The length of the RFQ is 2.7 m for output energy of 2.5 MeV/u or 5.8 m for5 MeV/u. The injection energy to the synchrotron at 5 MeV/u is close to the optimaldefined by efficiency of multi-turn injection. Probably slightly lower injection energyshould be studied. Here the output energy of 4 MeV/u and the length of 4 m areassumed. No additional IH structure is needed and therefore, the cost of the systemshould be reduced by about 50% with respect to contemporary carbon injector linacs.The transfer line between the injector and synchrotron includes a debunchercavity in order to decrease the momentum spread of the beam and to facilitate theRF-capture in the synchrotron.Table 2: Parameters of the injector linacSource Twin-EBISSource current 2 · particles per pulsePulse duration 100 ns − µ s (verify)RFQ frequency 750 MHzRFQ Transmission 50%Overall length about 7 mFinal energy 4 MeV/u Numerous designs of carbon and proton synchrotrons have been developed. Scalingup or down one of these machines should provide a good design for helium machine.The current commercial proton synchrotrons often use four 90 ° main dipoles (forexample Hitachi design, see Figure 2, ProTom splits them into 8 keeping the squareshape) to reduce its footprint and fit into typical rectangular rooms. Very compactdesigns usually relay on edge focusing to reduce number of quadrupoles and suffer5rom lack of dispersion-free sections, but the circumference of the machine is as smallas 18 m [16] or ever 16 m [17].Heidelberg Ion Therapy center (HIT) synchrotron has 6 main dipoles, what is thesmallest number among carbon machines. Circumference of this machine is about65 meters but dipoles are very heavy and there are no have dispersion-free sections.Figure 2: Hitachi proton synchrotron. Figure from [16]The progress in synchrotron design in the last 20 years comes from synchrotronlight sources. Those machines are optimized for minimum beam size and there-fore their lattices provide the smallest β functions. Those lattices are multi-bendachromats. The small beam size is also useful for hadron therapy machine, as thesmall vertical aperture leads to cheaper and lighter dipoles, which are one of the costdrivers of the facility. Therefore, a carbon-therapy machine based on double-bendachromat lattice has been proposed recently [18]. This preliminary design featuresalso a reduced-size of machine (55 m of circumference) and dispersion-free regions,which are ideal locations for RF cavities and injection/extraction devices.In this exercise an unconventional lattice made of double-bend achromatic (DBA)and triple-bend achromatic sections (TBA) is investigated. Five main dipoles isprobably the smallest number of bends which can be reasonably used to construct asynchrotron at the required rigidities.A warm synchrotron can typically operate with bending fields up to 1.5 T, there-fore, for the maximum required rigidity of helium beam B ρ = 4 .
85 Tm the bendingradius of the dipoles is only 3.23 m. The length of these magnets is around 4.1 m.These are large magnets, however they should be compared with 60 ° dipoles installedin HIT synchrotron, with length of 4.6 m and mass of 23.5 tons.The synchrotron was designed using MAD-X [19]. The layout of the machine isshown in Figure 3. It has a form of a pentagon with two sides elongated. Those aredispersion-free regions, each about 5 meters long, which can fit comfortably injection6nd extraction equipment and RF-cavity. The circumference of the whole machineis 50 meters.Figure 3: Left plot shows the layout of helium synchrotron. Right one shows thesynchrotron lattice.Table 3: Parameters of the proposed synchrotron for helium particle therapy beams.Tunes Q H , Q V β H , β V
88, 64 mNatural chromaticity -9.65, -8.15Transition γ . µ sRevolution period at extraction 0 . µ sIn this preliminary design there are 13 quadrupoles in 7 families. Sextupolesmust be installed in dispersive regions to control the chromaticity. The advantage oflow β is seen only in the TBA part of the lattice, while in the dispersion-free region β s reach high values what helps to increase the separation of particles being on theirlast turn before extraction.The RF system consists of 0.5-3.7 MHz Finemet loaded wideband cavity poweredby solid state amplifiers, similar to the systems installed in carbon ion facilities. They7re state of the art cavities for low energy accelerators [20] and allow operation inmulti-harmonics mode.The slow extraction is done by RF-KO method, which does not require beamdebunching and therefore simplifies Multi-Energy Extraction, what is currently con-sidered as one of the most important improvements in operation of carbon ion fa-cilities. The electrostatic septum (ES) is about 60-cm long and is installed in a2-meter drift space between main dipole and quadrupole TBA section. The phaseadvance between ES and magnetic septum is 267 ° . It is strongly dependent on theoptics. In order to fulfill Hard’t condition the additional sextupoles are installed indispersion-free sections.The lattice is preliminary and there is space for improvements. It could profitstronger from weak focusing or use combined function magnets and the dipoles couldbe split to more manageable size. The final design could be 10-15% more compact.The main parameters of the proposed synchrotron are summarized in Table 3. Transfer lines connect the extraction from the synchrotron with the patient. Thebeam can be delivered using a fixed beam line but many medical treatment plansforesee use of a gantry. Proton gantries are common, commercially available prod-ucts. Typical gantry has about 3.5 meters of diameter and weights around 100 tons.Carbon gantries are much larger and currently only two such objects exist: one basedon normal conducting magnets at HIT (Germany) [21] and one based on supercon-ducting magnets at NIRS (Japan) [22]. New projects of superconducting carbongantries are ongoing. Helium gantry could be a scaled down version of one of thesenew gantries.Figure 4 shows the whole facility in a single room configuration with a gantry.
The goal of this article is to build a case for helium ion beam therapy facility. Heliumbeams offer a significant increase in precision of the irradiation process over protons.With the final cost of about two to three times of a synchrotron based proton therapysystems, it is an interesting alternative, which could be commercialized easier thancarbon ion therapy systems. Helium-beam therapy can complement the carbon ther-apy but will not replace it because radioresistant tumors require high RBE providedonly by carbon ions.The particular solution discussed here is a conceptual design of the synchrotronbased on double and triple-bend achromats connected by spacious dispersion-freeregions. The big advantage is an injector made of RFQ only, what greatly simplifiesthe system. Development of such a machine should be accompanied by develop-ment of EBIS ion sources, which produce beams with much smaller emittance thancurrently used ECR sources. 8igure 4: The single-room layout of helium irradiation facility.
References [1] B. Knusl, H. Fuchs, K. Dieckmann, and D. Georg, “Can particle beam ther-apy be improved using helium ions? - a planning study focusing on pediatricpatients,”
Acta oncologica , no. 55, pp. 751–759, 2016.[2] S. Mein, I. Dokic, C. Klein, T. Tessonnier, T. T. Bhlen, G. Magro, J. Bauer,A. Ferrari, K. Parodi, T. Haberer, J. Debus, A. Abdollahi, and A. Mairani,“Biophysical modeling and experimental validation of relative biological effec-tiveness (RBE) for 4 He ion beam therapy,”
Radiation Oncology , vol. 14, no. 14,p. 123, 2019.[3] T. Tessonnier, A. Mairani, S. Brons, T. Haberer, J. Debus, and K. Parodi,“Experimental dosimetric comparison of (1)H, (4)He, (12)C and (16)O scannedion beams,”
Physics in medicine and biology , vol. 62, pp. 3958–3982, 04 2017.[4] F. Tommasino, E. Scifoni, and M. Durante, “New ions for therapy,”
Int J PartTher , vol. 2, no. 3, 2015.[5] J. Boldeman, R. Banati, H. Bttner, D. Cohen, and R. Garrett, “Preliminarydetails of a light ion facility for hadron therapy and research,” pp. ANSTO/E–770, 03 2010.[6] M. J. Berger, M. Inokuti, H. H. Andersen, H. Bichsel, D. Powers, S. . M. Seltzer,D. . Thwaites, and D. E. Watt, “Report 49,”
Journal of the International Com-mission on Radiation Units and Measurements , vol. 25, pp. NP–NP, 04 2016.97] E. Pearson, M. Abs, S. Henrotin, W. Kleeven, J. Van de Walle, P. Verbruggen,and S. Zaremba, “The new IBA superconducting synchrocyclotron (S2C2):From modelling to reality,” in
Proceedings of Cyclotrons2013, Vancouver, BC,Canada , Aug 2013.[8] M. Umezawa, R. Fujimoto, T. Umekawa, Y. Fujii, T. Takayanagi, F. Ebina,T. Aoki, Y. Nagamine, K. Matsuda, K. Hiramoto, T. Matsuura, N. Miyamoto,H. Nihongi, K. Umegaki, and H. Shirato, “Development of the compact protonbeam therapy system dedicated to spot scanning with real-time tumor-trackingtechnology,”
AIP Conference Proceedings , vol. 1525, pp. 360–363, 04 2013.[9] D. Ungaro, A. Degiovanni, and P. Stabile, “LIGHT: A Linear Accelerator forProton Therapy,” p. FRB1IO02. 5 p, 2017.[10] Y. Jongen et al. , “Current Status of the IBA C400 Cyclotron Project for HadronTherapy,”
Conf. Proc. C , vol. 0806233, 2008.[11] S. Benedetti,
High-gradient and high-efficiency linear accelerators for hadrontherapy . PhD thesis, Lausanne, 2018.[12] R. Becker, M. Kleinod, A. Schempp, E. D. Donets, and A. I. Pikin, “AnEBIS/RFQ-injector for a cancer therapy synchrotron,”
Review of Scientific In-struments , no. 63, p. 2812, 1992.[13] A. Shornikov and F. Wenander, “Advanced Electron Beam Ion Sources (EBIS)for 2-nd generation carbon radiotherapy facilities,”
JINST , vol. 11, no. 04,p. T04001, 2016.[14] A. Lombardi et al. , “Beam Dynamics in a High Frequency RFQ,” in
Proc. 6th International Particle Accelerator Conference (IPAC’15), Rich-mond, VA, USA, May 3-8, 2015 , no. 6 in International Particle Accelera-tor Conference, (Geneva, Switzerland), pp. 2408–2412, JACoW, June 2015.https://doi.org/10.18429/JACoW-IPAC2015-WEYB2.[15] V. Bencini, M. Breitenfeldt, J.-B. Lallement, A. Lombardi, H. Pahl, A. Pikin,J. Pitters, and F. Wenander, “High Frequency RFQ Design and LEBT Matchingfor the CERN TwinEBIS Ion Source,” no. CERN-ACC-2018-186, p. TUPO016.4 p, 2018.[16] F. Ebina, M. Uwezawa, H. Nishiuchi, T. Aoki, K. Hiramoto, K. Matsuda,K. Umegaki, and M. Furusaka, “Development of a compact synchrotron for pro-ton beam therapy,”
Electronics and Communications in Japan , vol. 100, no. 8,pp. 34–42, 2017.[17] V. Vostrikov, S. Karnaev, and Y. Pupkov, “Novel Approach to Design of theCompact Proton Synchrotron Magnetic Lattice,” in , p. TUPSA17, 2018.1018] X. Zhang, “Lattice design of a carbon-ion synchrotron based on double-bendachromat lens,” 2020. arXiv: 2007.11787 [physics.acc-ph].[19] “CERN MADX webpage.” http://madx.web.cern.ch/madx/ . Accessed: 2020-08-10.[20] C. Ohmori, M. Kanazawa, K. Noda, M. Kawashima, T. Misu, Y. Mori, A. Sug-iura, A. Takagi, and T. Uesugi, “A multi-harmonic rf system using a ma cavity,”
Nuclear Instruments and Methods in Physics Research Section A: Accelerators,Spectrometers, Detectors and Associated Equipment , vol. 547, pp. 249–258, 082005.[21] M. Galonska, S. Scheloske, R. Cee, K. Hppner, T. Winkelmann, A. Peters, andT. Haberer, “Commissioning of the ion beam gantry at hit,”
IPAC 2011 - 2ndInternational Particle Accelerator Conference , 01 2011.[22] Y. Iwata et al. , “Superconducting Gantry for Carbon-Ion Radiotherapy,” in