The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system

Nuclear Medicine and Biology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Nuclear Medicine and Biology journal homepage: www.elsevier.com/locate/nucmedbio

The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system William D. Webster a,⁎, Geoffrey T. Parks a, Dmitry Titov b, Paul Beasley c a b c

Department of Engineering, University of Cambridge, Cambridge, CB2 1PZ, United Kingdom Siemens Ltd, Corporate Technology, Moscow Siemens CT TIP Technology and Concepts, Oxford

a r t i c l e

i n f o

Article history: Received 4 October 2013 Accepted 21 November 2013 Available online xxxx Keywords: Accelerator production Low-energy 89 Zr 64 Cu 103 Pd

a b s t r a c t Introduction: The field of nuclear medicine is reliant on radionuclides for medical imaging procedures and radioimmunotherapy (RIT). The recent shut-downs of key radionuclide producers have highlighted the fragility of the current radionuclide supply network, however. To ensure that nuclear medicine can continue to grow, adding new diagnostic and therapy options to healthcare, novel and reliable production methods are required. Siemens are developing a low-energy, high-current – up to 10 MeV and 1 mA respectively – accelerator. The capability of this low-cost, compact system for radionuclide production, for use in nuclear medicine procedures, has been considered. Methodology: The production of three medically important radionuclides – 89Zr, 64Cu, and 103Pd – has been considered, via the 89Y(p,n), 64Ni(p,n) and 103Rh(p,n) reactions, respectively. Theoretical cross-sections were generated using TALYS and compared to experimental data available from EXFOR. Stopping power values generated by SRIM have been used, with the TALYS-generated excitation functions, to calculate potential yields and isotopic purity in different irradiation regimes. Results: The TALYS excitation functions were found to have a good agreement with the experimental data available from the EXFOR database. It was found that both 89Zr and 64Cu could be produced with high isotopic purity (over 99%), with activity yields suitable for medical diagnostics and therapy, at a proton energy of 10 MeV. At 10 MeV, the irradiation of 103Rh produced appreciable quantities of 102Pd, reducing the isotopic purity. A reduction in beam energy to 9.5 MeV increased the radioisotopic purity to 99% with only a small reduction in activity yield. Conclusion: This work demonstrates that the low-energy, compact accelerator system under development by Siemens would be capable of providing sufficient quantities of 89Zr, 64Cu, and 103Pd for use in medical diagnostics and therapy. It is suggested that the system could be used to produce many other isotopes currently useful to nuclear medicine. © 2013 Elsevier Inc. All rights reserved.

1. Introduction Radionuclides – such as 99mTc, 18F and 131I – are a key part of the field of nuclear medicine [1]. Since the development of the radiotracer principle and its first application in the early 1920s, to the development of the SPECT and PET imaging modalities in the 1970s and the progression of radiotherapy for cancer treatment, radioactive isotopes have made an increasingly important contribution to medicine [2]. They have advanced our understanding of disease pathways and progression, allowing the development of more advanced methods for their diagnosis. The recent entry to the market of Xofigo [3], a compound containing the alpha-emitter 223Ra targeted

⁎ Corresponding author. Tel.: +44 1223748569. E-mail address: [email protected] (W.D. Webster).

at bone cancer, has opened up a new spectrum of possibilities for targeted radioimmunotherapy (RIT). Theranostics, which involves labelling the same compound with either an imaging or a therapy radionuclide, will allow for better targeting and dose profiling of radiotherapy agents, ultimately improving end treatment for the patient [4]. The range of medically useful radionuclides, and their applications, is continually expanding and the demand for key isotopes such as 99mTc and 18F is rising [5]. Consistency and security in their supply is paramount to the future development and expansion of nuclear medicine. In recent years, shortages of 99mTc, a key isotope in a series of nuclear medicine routines and the dominant isotope in the industry as a whole, led to approximately 30 million patients worldwide having their treatments either delayed or cancelled. Initiated by the shut-down of two major isotope producing reactors, the National Research Universal reactor at Chalk River, Canada, and the High Flux Reactor at Petten, the Netherlands, the shortage highlighted the fragility of this key area of the nuclear medicine

0969-8051/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

2

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

supply chain [6]. With the capabilities and interests of nuclear medicine constantly growing and diversifying, there is the urgent need for the radionuclide production methods to grow and diversify in-kind. Siemens, in association with engineers at the Rutherford Appleton Laboratory in Oxford, are developing a compact DC electrostatic accelerator for the production of medical radionuclides [7]. Intended to deliver high currents of protons, up to 10 MeV in energy, the system aims to be a low-cost, compact alternative to radionuclide production. Unlike reactor facilities or larger accelerators with yields high enough to supply a large area, this accelerator system aims to deliver isotopes to a local area of a few nuclear medicine facilities. A localised production method will significantly reduce the impact of unexpected facility shut-downs and allow nuclear medicine centres to produce only the quantity of radionuclides they need, reducing unnecessary waste from decay. It could also increase access to new radionuclides for fundamental research and clinical trials. Irradiating at energies of 10 MeV and below additionally has the potential to produce very pure radionuclides, with a minimum amount of isotopic impurities and fewer elemental impurities. The range of isotopes that can be produced at these energies is extensive but it is for the moment uncertain how many current and developing medical radionuclides could be produced in sufficient quantities from such an accelerator system. This work will consider current, accelerator based production methods for several medically important radionuclides and the potential of the Siemens accelerator system for their production. Theoretical calculations will be used to estimate the activities of radionuclides that could be produced, and their isotopic and elemental purity. These will be compared to available experimental data with a view to quantifying the potential for their lowenergy production.

2. A low-energy accelerator for medical radionuclides Siemens are currently developing a low-energy DC electrostatic accelerator at the Rutherford Appleton Laboratory in Oxford, UK. Based on the design by Cockroft and Walton, using the Greinacher cascade rectifier, this modern interpretation aims to overcome some of the original design’s practical limits, which prevented it from achieving high output currents [7]. The accelerator uses a novel system of concentric shells around a central high voltage electrode, placed in a high vacuum. Through this design, the DC voltage generator is integrated with the insulator and accelerator structure, significantly reducing the size of the accelerator system while increasing the currents and voltages it can achieve. The accelerator is intended to have a maximum energy of 10 MeV, a spatial footprint of less than 2 m 2 and achieve currents of up to 1 mA. A simple ‘proofof-principle’ system was designed using four air-insulated shells [8]. Beam testing was then carried out using a 7-shell demonstrator in a full-size vacuum vessel that accelerated protons at a low voltage and a reasonably high current. Commissioning of the ion source has also been successful, achieving stable currents of over 300 μA [9]. Other key components such as the power-supplies, transformers and control system have also been tested at full power. Development work now focuses on the shells and insulator design with the aim of achieving much higher voltages, up to the intended 10 MeV. A compact, low-energy, high-current accelerator has many potential industrial applications [7]. Its low energy-consumption and intended low cost would make this accelerator ideal for medical radionuclide production on a smaller, more localised scale. A wide range of radionuclides could be delivered ‘on-demand’ to nearby nuclear medicine facilities, in large enough quantities for medical diagnostics, therapy or fundamental research. To test this hypothesis, information on the production requirements of several

radionuclides – with current or potential application in nuclear medicine – has been collated. 2.1.

89

Zr

89

Zr has a half-life of 78.41 hours and decays by electron capture (76.6%) and β + (22.3%). It is being developed as a new PET isotope, primarily for immunoPET and the imaging of cancerous tumours [10]. It is also ideal for the labelling of monoclonal antibodies (mAbs) [11]. Despite 89Zr’s many promising applications, it has had a slow uptake within nuclear medicine. This has been partly blamed on inefficient methods for separating the produced zirconium from its favoured target material, yttrium [12]. There are four potential routes for 89Zr production: 89Y(p,n), 89 Y(d,2n), natSr(α,X) and 90Zr(n,2n). The yttrium-based irradiation routes are favoured as yttrium is mono-isotopic; it has only one stable isotope, 89Y, so the production of other zirconium isotopes is limited. No expensive target enrichment is required and there is the potential to produce no carrier added, high specific activity 89Zr [10]. In contrast, the alpha irradiation of natural strontium, which has four stable isotopes – 84Sr, 86Sr, 87Sr and 88Sr – leads to the production of more contaminant zirconium isotopes. 2.2.

64

Cu

Cu has a half-life of 12.7 hours and is both a β − (38.5%) and β + emitter (17.6%). This produces 64Zn and 64Ni respectively. The remainder of the decay is electron capture, which also produces 64 Ni. As a result of this, 64Cu has been identified as a bi-functional radionuclide and has applications in both PET imaging and RIT [13]. Copper itself is the third most abundant trace metal in the human body and has many roles in human biochemistry and metabolism [14]. 64Cu is ideal for examining those roles, via PET, that are too slow for analysis by shorter lived isotopes [15]. It also has roles in imaging of peptides and antibodies, exploring cardiovascular disease, inflammation and cancer [16]. The most common mechanism for its production is the 64Ni(p,n) reaction [13,15,16]. 64

2.3.

103

Pd

103 Pd has a half-life of 16.991 days and decays primarily by electron capture to 103mRh, which subsequently decays through internal transition. The combination of Auger-electrons and X-rays emitted by these decay processes has made 103Pd a favoured isotope for brachytherapy [17]. The radionuclide is prepared into small seeds which are then implanted into sites of rapid cancer growth and proliferation. This is known as interstitial implantation. 125I was the dominant isotope in this field of medicine but 103Pd was found to have more suitable properties for treating rapidly growing tumours [18]. It was first proposed for interstitial implants in 1958 but it wasn’t until 1987 that it became commercially available [19]. Since then, 103Pd has been used successfully to treat a wide range of cancers including eye, brain, neck, uterus and colon [20]. It is predominantly produced via the 103Rh(p,n) reaction [17,21]. It can also be produced via neutron capture on 102Pd in nuclear reactors. This was the preferred production mechanism in its early years but the cost of enriching a 102 Pd target and the poor yields of the reaction proved prohibitive and cyclotrons became the dominant source [22].

2.4. Production requirements There are strict regulations on the use of radionuclides in radiopharmaceuticals and nuclear medicine. Radionuclides are required to have a very high standard of radionuclidic and radiochemical purity [23]. This can be achieved through a combination of using enriched target materials (in this work it has been assumed that all

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

target materials are 100% enriched), waiting a suitable length of time after irradiation for short-lived impurities to decay and restricting the irradiation energy to directly limit the production of impurities. To be viable as a production method, the Siemens accelerator would need to be able to satisfy all of these requirements. Some of the requirements are, to an extent, accelerator independent – elemental purity can be achieved through rigorous radiochemical separation procedures. This process can be assisted, again, by careful consideration of the irradiation energy range. There are then minimum requirements on the quantity of radionuclide, in terms of activity, needed for a particular medical procedure (for example, a PET scan or a brachytherapy treatment) to be performed successfully. For assessing the Siemens accelerator’s capabilities to produce the required quantity of radionuclide, data on example applications has been collated. Porrazzo et al. suggested the use of between 24 and 74 mCi (888 and 2738 MBq) with a median of 46 mCi per dose of 103Pd for controlling the growth of rapidly proliferating tumours [19]. In their work considering RIT, Y. Guo et al. suggested that 62 GBq (1680 mCi) of 64Cu would be sufficient for the treatment of colorectal cancer in humans [24]. Börjesson at al. used 75 MBq (2.03 mCi) of 89Zr per patient for the imaging of lymph-node metastases, achieving favourable results in comparison to CT/MRI [25]. 3. Methodology 3.1. Cross-sections and activity calculations The nuclear reaction code TALYS (version 1.4) [26] was used to calculate the excitation functions for each of the irradiations to be considered between 0 and 20 MeV. The irradiation of 89Y, 64Ni, and 103 Rh by protons was considered for producing 89Zr, 64Cu, and 103Pd respectively. Cross-sectional data from the online database EXFOR [27] was retrieved where available and compared to the generated cross-sections. The yield of radionuclides from each reaction were calculated using the following equation [10], R¼

E 2  dE  N AH  −λt I 1−e σ ðE ÞdE ∫ M d ðρx Þ E

ð1Þ

1

where R is the rate of radionuclide production, NA is Avogadro’s number, H is the isotopic abundance of the target material, M is the atomic mass of the target material, I is the current of the irradiating beam, λ is the dE ¼ S ðE Þ half-life of the produced isotope, t is the irradiation time, d ðρx Þ the stopping power and σ(E) is the reaction cross-section. The values for S(E), the stopping power of the target material, were obtained using the code SRIM (Stopping Range in Matter) [28].

Fig. 1. Theoretical cross-sections from TALYS for the

89

Y(p,X) reactions.

3

Fig. 2. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 89Y(p,n) reaction.

4. Results Excitation functions calculated by TALYS were compared to experimental data from the EXFOR database. Theoretical threshold energies were obtained using QCALC [29]. Using values of stopping power obtained from SRIM, the activities of produced radionuclides were calculated. It was assumed that all target materials were 100% isotopically enriched, in a pure metal form. The target was also assumed to be of a suitable thickness, calculated using stopping power data from SRIM, to degrade the proton beam down to the required energy. Radioisotopic purity was calculated as a percentage of the desired radionuclide to the total quantity of radionuclides produced in the energy range. 4.1.

89

Zr production via

89

Y(p,n)

4.1.1. Reaction cross-sections The cross-sections for all of the proton-induced reactions on 89Y up to 20 MeV are shown in Fig. 1. Of the three isotopes of zirconium predicted, only two are produced below 10 MeV; 89Zr, the desired isotope, and the stable 90Zr, produced by the 89Y(p,n) and 89Y(p,γ) reactions respectively. The formation of 89Zr dominates in this energy region, with the cross-section for the stable 90Zr a factor of 10 2–10 3 lower. The 89Y(p,2n) reaction producing the long-lived 88Zr has a threshold energy of 12.6 MeV and its rate of formation remains lower than 89Zr until 16 MeV. The presence of the 89Y(p,γ) cross-section throughout the energy range makes the formation of 90Zr difficult to avoid. 88Zr, which decays by β + emission and has a 83.4 day half-life, can be avoided by irradiating under its threshold energy of 12.6 MeV. Of the three elemental impurities predicted that are likely to be

Fig. 3. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 89Y(p,γ) reaction.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

4

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

Fig. 4. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 89Y(p,2n) reaction.

89

produced in reasonable quantities, only the Y(p,α) reaction forming stable 86Sr occurs below 10 MeV. The stable 88Sr and the long-lived 88 Y, formed from the 89Y(p,2p) and 89Y(p,d) reactions respectively, have theoretical threshold energies between 11 and 13 MeV and their cross-sections remain low below 16 MeV. Fig. 2 displays the TALYS results for the 89Y(p,n) reaction and compares them to experimental data retrieved from EXFOR [30–37]. 89 Zr is produced in both its ground and metastable states from this reaction. The metastable state, 89mZr, has a half-life of 4.161 minutes [38] and two modes of decay: IT (93.77%) forming the 89Zr ground state and β+ emission (6.23%) to 89Y. In this work, the total cross-section for the production of 89Zr has been approximated to the cross-section for 89Y(p,n)89gZr plus 93.77% of the cross-section for the89Y(p,n)89mZr reaction as after one hour following irradiation the majority of the 89mZr will have decayed without significant decay of the 89gZr [39]. EXFOR lists two cross-section data sets for the production of 89mZr: those of Gritsyna et al. and Blaser et al. [37]. These data sets show a very poor agreement with the TALYS results. However, the remaining data sets, which are for the total production of 89Zr, show a reasonable agreement with the TALYS results. The observed threshold for the reaction agrees with the theoretical threshold energy of 3.66 MeV, with the cross-section peaking at approximately 850 mb between 13 and 14 MeV. Fig. 3 compares the TALYS calculated cross-section and experimental cross-sections obtained from EXFOR results for the 89Y(p,γ) reaction. There is only one data set available from EXFOR, Tsagari et al. [40], and it only gives data up to 5 MeV. TALYS agrees well with results up to this point, showing the reaction cross-section peaking

Fig. 5. Activity yield and isotopic purity of proton beam.

Fig. 6. Theoretical cross-sections from TALYS for

64

Ni(p,X) reactions.

around 3 mb at 4 MeV. The 89Y(p,n) reaction quickly dominates it by over a factor of one hundred past this energy. The low values of the cross-section suggests that the quantity of 90Zr produced throughout the energy range under investigation will be negligible in comparison to other isotopes. Fig. 4 compares the TALYS and EXFOR results for the 88Y(p,2n) reaction [30,35,32–34]. TALYS overestimates the cross-section for this reaction by 0.05–0.1 b. This is within the error range of some data sets, however. The reaction begins around the theoretical threshold energy of 12.6 MeV and by 16 MeV the cross-section exceeds that of the 89 Y(p,n). The cross-sectional data from TALYS suggests that production below 10 MeV will result in only three isotopes being produced, with the production of 89Zr dominating significantly. 4.1.2. Calculated yields The reaction yield and isotopic purity of 89Zr was considered between Ep = 10 MeV and 4 MeV and the results are displayed in Fig. 5. With the entering proton energy set at 10 MeV, the exit proton energy was varied from 9.5 MeV to 4 MeV in 0.5 MeV increments. The activity yield increased rapidly for the first few MeV before reaching a maximum at 4 MeV. The 89Y(p,n) cross-section decreases rapidly with decreasing proton energy below 8 MeV. The isotopic purity decreased nearly linearly across the whole energy range – the steady production of the stable 90Zr is responsible for this. The activity yield at an exit energy of 5.5 MeV is 25.20 MBq/μAhr, while at 4 MeV it is 25.89 MBq/μAhr, only an additional 3%. The isotopic purity drops from 99.64% to 99.57% across this energy range, however. An exit energy of 5.5 MeV, or possibly 6 MeV for a slight loss in activity but a gain in

89

Zr vs. proton exit energy for a 10 MeV

Fig. 7. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 64Ni(p,γ) reaction.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

Fig. 8. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 64Ni(p,n) reaction.

purity, would be a reasonable compromise between the two factors, though a more thorough analysis is required. Börjesson et al. [25] used 75 MBq of 89Zr per patient for a clinical comparison of CT/MRI and immuno-PET. An hour irradiation at 100 μA would produce 2.5 GBq of activity, at a proton exit energy of 5.5 MeV, which would be sufficient for over 30 doses at this activity. 89Zr’s longer half-life allows for longer irradiation times before reaching saturation-over 20 GBq could be produced for a ten hour irradiation at 100 μA. 4.2.

64

Cu production via

64

Ni(p,n)

4.2.1. Reaction cross-sections Fig. 6 shows the cross-sections for all the reactions induced by protons on 64Ni between 0 and 20 MeV. TALYS predicts the production of three isotopes of copper: the stable 65Cu, the desired isotope 64Cu and the stable 63Cu via the 64Ni(p,γ), 64Ni(p,n) and 64 Ni(p,2n) reactions, respectively. Below 10 MeV only the 64Ni(p,γ) and 64Ni(p,n) reactions are present, with the cross-section for 64Cu production dominating by a factor of over a hundred through most of the energy range. The 64Ni(p,2n) reaction begins around Ep = 10.5 MeV, matching the theoretical threshold energy of 10.536 MeV, and rapidly dominates the 64Ni(p,n) cross-section, which peaks around 10 MeV and then declines. The presence of stable isotopic impurities can reduce the specific activity. Below 10 MeV, only one elemental impurity is predicted: 61Co from the 64Ni(p,α) reaction. Though the theoretical threshold energy for the reaction is 0 keV, TALYS predicts the cross-section to be 10 −7 mb below 4.3 MeV. The cross-section remains a factor of a hundred lower than that for 64Ni(p,

Fig. 9. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the 64Ni(p,α) reaction.

Fig. 10. Activity yield and isotopic purity of proton beam.

5

64

Cu vs. proton exit energy for a 10 MeV

n) in this lower energy region. Above 10 MeV, TALYS suggests that three other elemental impurities will exist: 63Ni, 62Ni and 60Co via the 64 Ni(p,d), 64Ni(p,t) and 64Ni(p,α+ n) reactions respectively. Fig. 7 compares the TALYS cross-section for the 64Ni(p,γ) reaction to experimental data [41]. TALYS agrees well with the experimental data up to 4 MeV, beyond which no experimental data is listed. Fig. 8 compares the TALYS cross-section for the 64Ni(p,n) reaction to experimental data [34,37,41–47]. TALYS generally shows good agreement with the results though noticeably overestimates the cross-section between 3 and 10 MeV. Experimental data is more limited above 12 MeV, though it suggests that TALYS then underestimates the cross-section in this region. Fig. 9 compares the TALYS cross-section for the 64Ni(p,α) reaction to experimental data [34,48]. TALYS agrees well with the experimental data up to 13 MeV where it appears to underestimate the crosssection. The peak in the reaction cross-section is not clearly defined in the experimental data that is available. The cross-sections suggest that irradiation below 10 MeV has the advantage of avoiding entirely the production of 63Cu, which at higher energies could be a significant impurity, reducing the specific activity of 64Cu achievable. It would also avoid the production of many of the elemental impurities, though 61Co would still need to be removed. 4.2.2. Calculated yields The yield and isotopic purity of 64Cu produced between the energies of 10 and 2 MeV was considered. Fig. 10 shows the activity yield in MBq/μAhr and isotopic purity – the percentage of 64Cu produced to all copper isotopes – against the exit energy of an initially 10 MeV proton beam. The achievable activity increases

Fig. 11. Theoretical cross-sections from TALYS for the

103

Rh(p,X) reactions.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

6

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

Fig. 12. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the

rapidly with decreasing exit energy and peaks at 2 MeV, beyond which there is no further production of 64Cu. The percentage purity of 64Cu increases down to 4 MeV, at which point the production of 65 Cu dominates and the purity decreases. At its maximum, the percentage isotopic purity is 99.92%, with a yield of 340.59 MBq/ μAhr. A current of 1 mA would be able to provide 9205 mCi of 64Cu in one hour of irradiation. The quantity of 64Cu required for a single brachytherapy treatment was estimated to be 1680 mCi by Guo et al. [24]. This suggests that the accelerator would certainly be capable of delivering the quantities of 64Cu required for this brachytherapy treatment. 4.3.

103

Pd production via

103

Rh(p,n)

4.3.1. Reaction cross-sections Fig. 11 displays the TALYS generated cross-sections for the proton irradiation of 103Rh between Ep = 0 and 20 MeV. TALYS predicts the formation of four isotopes of palladium in this energy range: the stable 104Pd, the desired isotope 103Pd, and the stable 102Pd via the 103 Rh(p,γ), 103Rh(p,n) and 103Rh(p,2n) reactions respectively. The

103

Rh(p,n) reaction.

103

Rh(p,n) reaction has a very low threshold of 1.34 MeV leading the production of 103Pd to rapidly dominate over 104Pd. The 103Rh(p,2n) reaction also has a low threshold of 9.04 MeV; at 14 MeV the reaction cross-section dominates the 103Rh(p,n) reaction by a factor of ten. To produce high specific activity 103Pd, irradiation below 10 MeV is preferred, with irradiation below 9 MeV removing the 102Pd impurity entirely, though at the expense of yield. The main elemental impurities suggested by the TALYS crosssections are the stable 100Ru, the long-lived 102Rh and the stable 99Ru via the 103Rh(p,α), 103Rh(p,d) and 103Rh(p,α + n) reactions. While the theoretical threshold energies for the latter two reactions are 7.16 MeV and 3.65 MeV respectively, the excitation functions in TALYS do not begin until over 10 MeV, suggesting that irradiating below 10 MeV will only see the production of 100Ru, with 103Pd production still significantly dominant. No experimental data in a suitable energy range exists to verify the TALYS results for these three reactions, however. Fig. 12 compares the TALYS cross-section for the 103Rh(p,n)103Pd reaction to experimental data from EXFOR [37,49–55]. The TALYS excitation function agrees well with experimental data up to 8–9 MeV

Fig. 13. Comparison between theoretical cross-sections from TALYS and compiled experimental cross-section data from the EXFOR database for the

103

Rh(p,γ) reaction.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

Fig. 14. Activity yield of

103

Pd vs. proton exit energy for 10, 9.5 and 9 MeV proton beams.

but then appears to underestimate the cross-section. The erratic nature of the data makes it difficult to judge how accurate the TALYS excitation function is at energies above 8 MeV. Fig. 13 compares the cross-section results from TALYS and EXFOR for the 103Rh(p,γ) 104Pd reaction [56]. No experimental data exists in EXFOR above 5 MeV for this reaction but the TALYS results agree well with what is available. The cross-sections suggest that, at an irradiation energy of 10 MeV, there could be an appreciable production of 102Pd. This can be avoided by irradiating at a slightly lower energy. The only other isotopic impurity is 104Pd which is likely to be produced only in small quantities. The main elemental impurity would be 100Ru, which could be produced in slightly larger quantities than 104Pd. 4.3.2. Calculated yields The activity yields and isotopic purity of 103Pd produced by the irradiation of 103Rh were calculated. As anticipated, irradiation starting at 10 MeV saw the production of 102Pd, which significantly decreased when the entrance energy was dropped to 9.5 MeV and was eliminated entirely at 9 MeV. Fig. 14 shows the yield of 103Pd in

Fig. 15. Isotopic purity of

7

103

MBq/μAhr for entrance energies of 10 MeV, 9.5 MeV and 9 MeV respectively. The maximum activity achieved was at a proton exit energy of 2 MeV for each irradiation regime: 3.73 MBq/μAhr, 3.01 MBq/μAhr and 2.30 MBq/μAhr at a 10 MeV, 9.5 MeV and 9 MeV entrance energy respectively. Fig. 15 shows the percentage isotopic purity as a function of the exit energy for each of the starting energies. The percentage purity increases from 96% at a 10 MeV entrance energy to 99.1% for a 9.5 MeV entrance energy. The 9 MeV entrance energy offers a 99.8% percentage purity though a 9.5 MeV entrance energy offers a better activity yield while maintaining a high purity. The use of up to 74 mCi of 103Pd was determined by Porrazzo et al. as being suitable for treating rapidly proliferating tumours [19]. A ten hour irradiation at a current of 1 mA would be capable of producing approximately 800 mCi of 103Pd, which would be sufficient for multiple treatments. 5. Conclusions Nuclear medicine has growing and broadening applications throughout medicine in understanding disease mechanisms,

Pd vs. proton exit energy for 10, 9.5 and 9 MeV proton beams.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

8

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx

diagnosis and therapy. Increasing access to currently used radionuclides, as well as making new radionuclides under consideration available for medical trials, is important to ensuring this field can continue to expand. The shortages of 99mTc in 2008 to 2010 highlighted the fragility of the current supply network and the need to expand our production capabilities. Moving away from centralised production systems – such as a single reactor producing 30–40% of supply – to one that is more localised where a single facility only serves a small region of hospitals, will reduce the impact of facility shut-downs and also potentially the cost and legislation involved in radionuclide delivery. A compact DC electrostatic accelerator currently under development by Siemens in collaboration with engineers at the Rutherford Appleton Laboratory has been suggested as an alternative, localised source for radionuclides production. The accelerator is capable of delivering 10 MeV protons at currents of up to 1 mA. The production of three medically important radionuclides from a low energy accelerator system of this design has been considered. Yield calculations performed using data from TALYS and SRIM suggest that all three of the isotopes considered ( 89Zr, 64Cu, and 103Pd) can be produced in sufficient quantities to be medically useful. The isotopic purity predicted for each of these isotopes is over 99%, allowing for very high specific activities of each radionuclide, which will improve image quality and reduce the physical quantity required of each isotope. These isotopes represent just a few of those that are of medical interest, however, and further studies should be conducted to consider more radionuclides. References [1] IAEA. Cyclotron produced radionuclides: principles and practice. Technical Report 465, 2008. [2] Zimmer L, Luxen A. PET radiotracers for molecular imaging in the brain: past, present and future. Neuroimage 2012;61:363–70. [3] Galluzzi L. New immunotherapeutic paradigms for castration-resistant prostate cancer. OncoImmunology 2013;2:e26084. [4] Juweid ME, Mottaghy FM. Current and future aspects of nuclear molecular therapies: a model of theranostics. Methods 2011;55:193–5. [5] EC. Preliminary report on supply of radioisotopes for medical use and current developments in nuclear medicine. Technical report, European Commission, October 2009. [6] Nunan T. A review of the supply of molybdenum-99 the impact of recent shortages and the implication for nuclear medicine service in the UK. Technical report, Administration of Radioactive Substances Advisory Committee, November; 2010. [7] Beasley P, Heid O, Hughes T. A new life for high voltage electrostatic accelerators. Proceedings of IPAC’10, Kyoto, Japan; 2010. [8] Beasley P, Heid O. Progress towards a novel compact high voltage electrostatic accelerator. Proceedings of 2011 Particle Accelerator Conference, New York, NY, USA. New York: NY, USA; 2011. [9] von Jagwitz-Biegnitz H, Faircloth D, Beasley P, Heid O. Commissioning of the ion source for Siemens novel electrostatic accelerator. Proceedings of IPAC2013, Shanghai, China; 2013. [10] Sadeghi M, Enferadi M, Bakhtiari M. Accelerator production of the positron emitter zirconium-89. Ann Nucl Energy 2012;41:97–103. [11] Vugts DJ, van Dongen GAMS. 89Zr-labeled compounds for PET imaging guided personalized therapy. Drug Discov Today Technol 2011;8:e53–61. [12] Holland JP, Sheh Y, Lewis JS. Standardized methods for the production of high specific-activity zirconium-89. Nucl Med Biol 2009;36:729–39. [13] So Le V, Howse J, Zaw M, Pellegrini P, Katsifis A, Greguric I, et al. Alternative method for 64Cu radioisotope production. Appl Radiat Isot 2009;67:1324–31. [14] Sun X, Anderson CJ. Production and applications of copper-64 radiopharmaceuticals. Methods Enzymol 2004;386:237–61. [15] McCarthy DW, Shefer RE, Klinkowstein RE, Bass LA, Margeneau WH, Cutler CS, et al. Efficient production of high specific activity 64Cu using a biomedical cyclotron. Nucl Med Biol 1997;24:35–43. [16] Kume M, Carey PC, Gaehle G, Madrid E, Voller T, Margenau W, et al. A semi-automated system for the routine production of copper-64. Appl Radiat Isot 2012;70:1803–6. [17] Tárkányi F, Hermanne A, Király B, Takály S, Ditrói F, Csikai J, et al. New crosssections for the production of 103Pd; a review of charged particles production routes. Appl Radiat Isot 2009;67:1574–81. [18] Chunfu Z, Yongxian W, Yongping Z, Xiuli Z. Cyclotron production of no-carrieradded palladium-103 by bombardment of rhodium-103 target. Appl Radiat Isot 2001;55:441–5. [19] Porrazzo MS, Hilaris BS, Moorthy CR, Tchelebi AE, Mastoras CA, Shih LL, et al. Permanent interstitial implantation using palladium-103: the New York Medical College preliminary experience. Int J Radiat Oncol Biol Phys 1992;23:1033–6.

[20] IAEA. Cyclotron produced radionuclides: physical characteristics and production methods. Tech Rep 2008;468. [21] Saidi P, Sadeghi M, Enferadi M, Aslani G. Investigation of palladium-103 production and IR07-103Pd brachytherapy seed preparation. Ann Nucl Energy 2011;38:2168–73. [22] Silverman I, Lavie E, Arenshtam A, Kijel D, Vaknin D, Veinguer M, et al. Production of palladium-103 (103Pd) from a thin rhodium foil target - improved cooling concept. Nucl Instrum Methods B 2007;261:747–50. [23] Shivarudrappa V, Vimalnath KV. High purity materials as targets for radioisotope production: needs and challenges. Bull Mater Sci 2005;28:325–30. [24] Guo Y, Parry JJ, Laforest R, Rogers BE, Anderson CJ. The role of p53 in combination radioimmunotherapy with the 64Cu-DOTA-Cetuximab and cisplatin in a mouse model of colorectal cancer. J Nucl Med 2013;54:1621–9. [25] Börjesson PK, Jauw YW, Boellaard R, de Bree R, Comans EF, Roos JC, et al. Performance of immuno-positron emission tomography with zirconium-89labeled chimeric monocolonal antibody u36 in the detection of lymph node metastases in head and neck cancer patients. Clin Cancer Res 2006;12: 2133–40. [26] Koning AJ, Hilaire S, Duijvestijn MC. TALYS-1.0. Proceedings of the International Conference on Nuclear Data for Science and Technology, Nice, France; 2007. [27] Otuka N, Dunaeva S, Dupont E, Schwerer O, Blokhin A. The role of the nuclear reaction data centres in experimental nuclear data knowledge sharing. Proceedings of the International Conference on Nuclear Data for Science and Technology (ND2010), Jeju, Korea; 2010. [28] Ziegler JF, Ziegler MD, Biersack JP. SRIM - the stopping and range of ions in matter. Nucl Instrum Methods B 2010;268:818–1823. [29] T.W. Burrows. National Nuclear Data Centre, Brookhaven National Laboratory, http://www.nndc.bnl.gov/qcalc/. [30] Khandaker MU, Kim K, Lee MW, Kim KS, Kim G, Otuka N. Investigation of 89Y(p, x)86,88,89gZr, 86m+g,87g,87m,88gY, 85gSr, and 84gRb nuclear processes up to 42 MeV. Nucl Inst Methods Phys Res B, 271:72–81, 2012. EXFOR D0675002 and D0675003. [31] Johnson CH, Kernell RL, Ramavataram S. The 89Y(p,n)89Zr cross section near the first two analogue resonances. Nucl Phys A 1968;107:21–34 EXFOR C0774002. [32] Mustafa MG, West HI, O’Brien H, Lanier RG, Benhamou M, Tamura T. Measurements and a direct-reaction-plus-Hauser-Feshbach analysis of 89Y(p, n)89Zr, 89Y(p,2n)88Zr, and 89Y(p,pn)88Y reactions up to 40 MeV. Phys Rev C 1988;38: 1624–37 EXFOR A0403002 and A0403003. [33] Omara HM, Hassan KF, Kandil SA, Hegazy FE, Saleh ZA. Proton induced reactions on 89Y with particular reference to the production of the medically interesting radionuclide 89Zr. Radiochem Acta 2009;97:467–71 EXFOR D0584002 and D0584003. [34] Levkovskij VN. Activation cross sections by protons and alphas. Moscow; 1991 . EXFOR A0510075, A0510076, A0510185 and A0510186. [35] Zhao W, Shen Q, Lu H, Yu W. Investigation of Y-89(p, n)Zr-89, Y-89(p,2n)Zr-88 and Y-89(p, pn)Y-88 reactions up to 22 MeV. Chin J Nucl Phys 1992;14:7 EXFOR S0040002 and S004003. [36] Gritsyna VT, Klyucharev AP, Remaev VV, Reshetova LN. Ratio of the cross sections for the production of the isomer and ground states of nuclei in the (p,n) reaction at the energies from the threshold to 20 MeV. Zhurnal Eksperimental‘noi i Teoret Fiziki 1963;44:1770 EXFOR P0012003. [37] Blaser JP, Boehm F, Marmier P, Scherrer P. Anregungsfunktionen und Wirkungsquerschnitte der (p, n)-Reaktion (ii). Helvetica Phys Acta 1951;24:441 EXFOR P0033004, P0033012 and P0033019. [38] Singh B. Nuclear data sheets for A = 89. Nucl Data Sheets 2013;114:1–208. [39] Infantino A, Cicoria G, Pancaldi D, Ciarmatori A, Boschi S, Fanti S, et al. Prediction of 89Zr production using the Monte Carlo code FLUKA. Appl Radiat Isot 2011;69: 1134–7. [40] Tsagari P, Kokkoris M, Skreti E, Karydas AG, Harissopulos S, Paradellis T, Demetriou P. Cross section measurements of the 89Y(p,γ)90Zr reaction at energies relevant to p-process nucleosynthesis. Phys Rev C, 70:015802–1–015802–10, 2004. EXFOR O1182002. [41] Sevior ME, Mitchell LW, Anderson MR, Tingwell CIW, Sargood DG. Absolute cross sections of proton induced reactions on 65Cu, 64Ni, 63Cu. Aust J Phys 1983;36:463 EXFOR A0198004 and A0198005. [42] Tanaka S, Furukawa M. Excitation functions for (p, n) reactions with titanium, vandium, chromium, iron and nickel up to Ep =14 MeV. J Phys Soc Japan 1959;14: 1269–75 EXFOR B0043009. [43] Guzhovskii BY, Borkin IM, Zvenigorodskii AG, Rudnev VS, Solodovnikov AP, Trusillo SV. Isospin mixing of isobar analog resonances observed for the 59,61,63,65 C u nuclei. Izvestiya Rossiiskoi Akademii Nauk 1969;33:129 EXFOR F0704002. [44] Tanaka S, Furukawa M, Chiba M. Nuclear reactions of nickel with protons up to 56 MeV. J Inorg Nucl Chem 1972;34:2419–26 EXFOR B0020020. [45] Szelecsényi F, Blessing G, Qaim SM. Excitation functions of proton induced nuclear reactions on enriched 61Ni and 64Ni: Possibility of production of no-carrier-added 61 64 Cu and Cu at a small cyclotron. Appl Radiat Isot 1993;44:575–80 EXFOR D4020004. [46] Avila-Rodriguez MA, Nye JA, Nickles RJ. Simultaneous production of high specific activity 64Cu and 61Co with 11.4 MeV protons on enriched 64Ni nuclei. Appl Radiat Isot 2007;65:1115–20 EXFOR C1600002. [47] Adam Rebeles R, Van den Winkel P, Hermanne A, Tárkányi F. New measurement and evaluation of the excitation function of 64Ni(p, n) reaction for the production of 64Cu. Nuclear Instruments and Methods in Physics Research Section B 2009;267:457–61 EXFOR D4207002. [48] Qaim SM, Uhl M, Rösch F, Szelecsényi F. Excitation functions of (p, α) reactions on 64Ni, 78Kr, and 86Sr. Phys Rev Part C 1995;52:733–9 EXFOR D4015002.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007

W.D. Webster et al. / Nuclear Medicine and Biology xxx (2014) xxx–xxx [49] Qaim SM, Uhl M, Rösch F, Albert RD. (p,n) cross section and proton optical-model parameters in the 4- to 5.5-MeV energy region. Phys Rev 1959;115:925–7 XFOR T0130016. [50] Johnson CH, Galonsky A, Inskeep CN. Cross sections for (p, n) reactions in intermediate-weight nuclei. Oak Ridge National Lab Rep 1960;2910:25 EXFOR T0135006. [51] Harper PV, Lathrop K, Need JL, The thick target yield and excitation function for the reaction 103Rh(p, n)103Pd. Nucl Sci Abstr 1961;15:2780 EXFOR D0456002. [52] Harper PV, Lathrop K, Need JL. The thick target yield and excitation function for the reaction 103Rh(p, n)103Pd. Oak Ridge National Lab Reports 1961;61:67 EXFOR C1596002. [53] Hansen LF, Albert RD, Statistical theory predictions for 5- to 11-MeV (p, n) and (p, p’) nuclear reactions in V51, Co59, Cu63, Cu65, and Rh103. Phys Rev 1962;128: 291–9 EXFOR B0066006.

9

[54] Hermanne A, Sonck M, Fenyvesi A, Daraban L, Study on production of 103Pd and characterisation of possible contaminants in the proton irradiation of 103Rh up to 28 MeV. Nuclear Instruments and Methods in Physics Research Section B 2000;170:281–92 EXFOR D4108002. [55] Sudár S, Cserpák F, Qaim SM. Measurements and nuclear model calculations on proton-induced reactions on 103Rh up to 40 MeV: evaluation of the excitation function of the 103Rh(p, n)103Pd reaction relevant to the production of the therapeutic radionuclide 103Pd. Appl Radiat Isot 2002;56:821–31 EXFOR O1010002 and O1010005. [56] Spyrou A, Becker H-W, Lagoyannis A, Harissopulos S, Rolfs C. Cross-section measurements of capture reactions relevant to the p process using a 4π γsumming method. Phys Rev C 2007;76:p015802 EXFOR O1534003.

Please cite this article as: Webster W.D., et al, The production of radionuclides for nuclear medicine from a compact, low-energy accelerator system, Nucl Med Biol (2014), http://dx.doi.org/10.1016/j.nucmedbio.2013.11.007