- Email: [email protected]
GE PETtrace cyclotron as a neutron source for boron neutron capture therapy
ARTICLE IN PRESS
Applied Radiation and Isotopes 61 (2004) 1057–1062
GE PETtrace cyclotron as a neutron source for boron neutron capture therapy A. Boskoa,*, D. Zhilchenkovb, W.D. Reecea a
Department of Nuclear Engineering, Texas A&M University, College Station, USA b NuTech, Inc., Tyler, TX, USA
Abstract This paper discusses the use of a General Electric PETtrace cyclotron as a neutron source for boron neutron capture therapy. In particular, the standard PETtrace 18O target is considered. The resulting dose from the neutrons emitted from the target is evaluated using the Monte Carlo radiation transport code MCNP at different depths in a brain phantom. MCNP-simulated results are presented at 1, 2, 3, 4, 5, 6, 7, and 8 cm depth inside this brain phantom. Results showed that using a PETtrace cyclotron in the current configuration allows treating tumors at a depth of up to 4 cm with reasonable treatment times. Further increase of a beam current should significantly improve the treatment time and allow treating tumors at greater depths. r 2004 Elsevier Ltd. All rights reserved. Keywords: BNCT; Cyclotron; MCNP
1. Introduction Boron neutron capture therapy (BNCT) is under active investigation all over the world. 10B attached to a tumor seeking compound is injected intravenously into the patient and is carried to the tumor site. The boron disintegrates after capturing a neutron from a neutron beam and the high-energy particles emitted by the disintegration kill tumor cells while sparing normal tissue. The history of BNCT starts in 1936 when J. Chadwick of Cambridge University and biophysicist G. L. Locher of the Franklin Institute presented the basic theory of BNCT. In 1951 it was suggested that BNCT could be used for treatment of brain tumors, and in particular, treatment of glioblastoma multiforma (GBM). Since then many laboratories, institutes and clinics conducted research and clinic trials using BNCT in efforts to improve the efficiency. One area of research is directed *Corresponding author. E-mail address: [email protected] (A. Bosko).
toward finding a cheap, reliable, and manageable source of neutrons for BNCT. Historically the best source of neutrons at the energy and flux levels required for BNCT was a nuclear reactor. BNCT research has used reactors at Brookhaven medical research reactor (BMRR), at the Massachusetts Institute of Technology reactor (MITR), at the high flux reactor (HFR) in Institute for Energy, Pettern, Netherlands, and other facilities. The advantage of a reactor is that the neutrons from reactors are relatively cheap, if capital costs are discounted. The disadvantages of reactors are that the capital costs are very high and reactors are too complicated for an ordinary clinic to operate, so these clinics cannot afford to build and maintain a small nuclear reactor to use as a neutron source. Another approach to this problem is to use particle accelerators. These machines can accelerate protons to the required energies for a given reaction and the bombarded target emits neutrons at rates and energy levels appropriate for BNCT. These accelerators are cheaper, easier to operate, and pose less overall risk than
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.05.076
ARTICLE IN PRESS 1058
A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
nuclear reactors. Accelerator neutron sources are well understood and implemented in a number of research facilities (Wang et al., 1989). This paper investigates the possibility to make BNCT available for clinics and research groups that do not have nuclear reactor or an accelerator specially designed for BNCT. There are a large number of medical cyclotrons used for medical radioisotope production in the nuclear pharmacies and hospitals. These cyclotrons are designed mostly for fluorodeoxyglucose (FDG) production. FDG is manufactured in an automated synthesis unit from 18F. This isotope is produced in a cyclotron by proton bombardment of enriched water (18O). In this article the General Electric (GE) PETtrace cyclotron, normally used to produce FDG, is examined to study the possibility of using it for BNCT. The GE PETtrace was chosen for this investigation because this type of cyclotron is popular among nuclear pharmacies and clinics in many countries (about 75 PETtraces have been sold worldwide), in part because it is compact and reliable. This cyclotron produces protons with energies large enough to produce neutrons with appropriate energy and flux for BNCT and it does not require significant changes in design to provide these neutrons. The energy of protons made in this cyclotron is fixed at 16.5 MeV. The current on the target runs up to 60 mA. Targets are accessible, easily changeable, and this cyclotron can use dual targets simultaneously, allowing the flexibility to work on radioisotope production and BNCT with minimal investments of time and work force. There are a number of difficulties in modifying this cyclotron for BNCT and they include: *
*
*
There is no known target for BNCT technique, however, as we show in this paper, conventional targets may be acceptable; positioning of target filler (Fig. 1) will scatter neutrons emitted in the forward direction and it could result in less flux than required by BNCT, so the system configuration might have to be changed; some of these cyclotrons are assembled as selfshielded units (stainless tanks filled with borated water surround the cyclotron), and this configuration makes impossible for them to serve as a neutron source.
As a first step to using the GE PETtrace for BNCT, calculations were conducted to study the possibility of using 18O target as a source of neutrons with further reflection, moderation, absorption and collimation to produce appropriate energy and flux for BNCT. The 18O target was studied because about 90–95% of time these cyclotrons use this target already for FDG production. The efficiency of cyclotron use may be significantly increased if unused neutrons produced
Fig. 1. General Electric PETtrace cyclotron.
Fig. 2. Cyclotron vault with channel for filters and collimators proposed for BNCT (side view).
during FDG production could be utilized for other medical modalities such as BNCT at the same time when radioisotopes are produced. This research explores this approach using Monte Carlo codes with standard neutron data files, and the following assumptions: *
*
*
Standard running conditions (60 mA target current, 16.5 MeV protons, standard 18O target); no target filler system behind the target (it can be moved to the side of the cyclotron); cyclotron is in a vault with no self-shielding (Fig. 2).
2. Monte Carlo simulation The general purpose Monte Carlo code, MCNP (Briesmeister, 2000), was used to simulate the transport of neutrons through the geometry shown in Fig. 3. This geometry includes a moderator assembly surrounded by a thick graphite reflector. The moderator
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062 18
assembly consists of an aluminum container filled with D2O. Another popular type of BNCT moderator consisting of aluminum, fluorine and natural lithium was also considered. The heavy water moderator was chosen because it had lower fast-neutron contribution to the total flux in the phantom. Various thicknesses of moderating material were investigated in this study. Again, the thickness was chosen that provided the lowest fast-neutron contribution to the total flux in comparison to the low energy component and that allowed delivery of the necessary dose in a reasonable time. The optimal inner diameter of the moderator container was 30 cm for a fixed 30-cm length. The container wall set at 1 cm for structural considerations. The moderator assembly was surrounded by a 30-cm thick graphite reflector. Also, a 4-cm thick lead shield was used to reduce the dose from the photons emitted from the target. The typical 18O target consists of
O-enriched water kept under a helium over-pressure in a silver container. The neutron spectrum and angular distribution on these 18O-targets are well established (according to PETtrace-source term evaluations by Lundgren and Ingemannson, 2001). The neutron spectrum from such a target is shown in Fig. 4. This data, along with corresponding neutron yield for different angular directions, was used to define the neutron source for the MCNP calculations. The angular distribution from 0 to 180 with bins covering every 20 was constructed for input. The neutrons emitted from the target are moderated by scattering through the heavy water. The neutrons leave the moderator and may then interact within the brain for BNCT treatment. The phantom brain used to model these interactions consists of ellipsoid with major axis of 13.6, 19.6 and 16.6 cm (Harling et al., 1995) filled with a specified composition to simulate brain tissue (ICRU, 1989). The tally cells were 0.5 cm diameter spheres at depths of 1, 2, 3, 4, 5, 6, 7 and 8 cm inside the phantom brain. The flux tally was used to calculate photon and neutron dose, by applying flux-to-dose conversion factors given in the literature (Briesmeister, 2000). The thermal neutron flux as a function of the in phantom depth is plotted on Fig. 5. Most neutrons have been thermalized at the tally depths. However, in order to know the total dose delivered in the tumor and the healthy tissue, the other components of the total dose were calculated. Fig. 6 shows the energy spectrum of neutrons crossing the tally cell at a depth of 3 cm. There are two principal capture reactions for thermal neutrons in tissue—1H(n,g)2H and 14N(n,p)14C. The first capture reaction releases a 2.22 MeV gamma ray, which could deposit a fraction of its energy before escaping the phantom. In contrast, the nitrogen capture reaction releases 0.626 MeV which is deposited by the proton and
Phantom
Lead
Ø 30 cm
Target D 2O
30 cm
Aluminum container
1059
Graphite
Fig. 3. Geometry defined as input for the MCNP calculation.
1.0E+12
1.0E+11
1.1E+111.0E+11 5.9E+10 2.9E+10 1.4E+10 6.4E+09
1.0E+10
n/s
2.9E+09 1.2E+09 1.0E+09
5.1E+08 2.1E+08 1.1E+08
1.0E+08
1.0E+07 0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
Neutron energy, MeV
Fig. 4. Neutron spectrum from the H218O target.
8-9
9 - 10
> 10
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
1060 1.2E+08
8.0E+07
2
Flux , n- cm /s
1.0E+08
6.0E+07 4.0E+07 2.0E+07 0.0E+00 0
1
2
3
4
5
6
7
8
9
Depth, cm Fig. 5. Thermal neutron flux in the brain.
2
Neutron flux, n-cm /s
1.0E+08
1.0E+07
1.0E+06
1.0E+05 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 1.0E+01 1.0E+02
Neutron energy , MeV Fig. 6. Neutron spectrum for 3 cm depth in the phantom.
recoil carbon nucleus in the immediate vicinity of the capture site. The resulting dose rate from exposure to thermal neutrons was calculated using the following formula: D14N ¼ Fth Fn14N f14N ;
ð1Þ
where D14N is the dose rate due to capture on nitrogen (cGy/s), Fth the neutron fluence (n/cm2 s), Fn14N ¼ 0:785 109 cGy cm2/n is the kerma factor for thermal neutrons, f14N ¼ 0:022 is the weight fraction of nitrogen in brain tissue (ICRU, 1989). Kerma factors for this reaction were taken from Caswell et al. (1980). In the neutron reaction, 10B(n,g)7Li, the charged particles are produced with the total kinetic energy of 2.34 MeV on average and with a range in tissue of about 10 mm, the order of the size of a tumor cell. As a result, the cell containing 10B can be destroyed. A similar formula (2)
was used to obtain the dose rate from the neutron reaction with 10B. D10B ¼ Fth Fn10B m;
ð2Þ
where D10B is the dose rate due to capture on boron (cGy/s), Fth the neutron flux (n/cm2 s), Fn10B ¼ 8:6 1012 cGy cm2/n is the corrected kerma factor for thermal neutrons obtained per 1 ppm concentration of 10 B in tissue (Rogus et al., 1994). The total dose to healthy tissue and tumor cells were calculated using the following formula: DðRBEÞtissue ¼ DðRBEÞfast þ DðRBEÞphoton þ D14N RBE14N þ D10B RBE10B ;
ð3Þ
where DðRBEÞtissue is the effective dose rate to tissue (RBE cGy/min), DðRBEÞfast is the simulated neutron
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062 14
Fast Neutron Dose Photon Dose
12
RBE-Dose,cG y/min .
1061
Nitrogen-14 Dose Boron-10 in tumor Dose (40 ppm)
10
Boron-10 in healthy tissue dose (11 ppm) 8
Healthy tissue Dose Tumor Dose
6 4 2 0 0
1
2
3
5
4
6
7
8
9
Depth, cm Fig. 7. RBE-dose rates at different depths into the brain phantom.
Table 1 Estimated treatment times and the dose delivered to the tumor In-phantom depth (cm) Treatment time (min) Tumor dose (Gy)
1
2
3
4
5
6
7
258 30.7
288 29.3
381 30.7
444 29.8
459 26.4
484 24.4
545 22.1
dose rate (RBE cGy/min), DðRBEÞphoton is the simulated photon dose rate (RBE cGy/min), D14N and D10B are the dose from interactions of thermal neutrons with nitrogen and boron, RBE14N ¼ 3:2; RBE10B ¼ 1:3 (for normal tissue) and RBE10B ¼ 3:8 (for tumor) are the relative biological effectiveness given in the literature (Coderre et al., 1993). 10B concentrations in the normal tissue of 11.4 ppm and within the tumor of 40 ppm were assumed (Coderre et al., 1998). The RBE-dose rates obtained from calculations are plotted in Fig. 7. One of the most important clinical requirements according to the Brookhaven Medical Research Reactor protocol (Bleuel et al., 1998) is that the maximum dose to the normal brain should not exceed 12.5 RBE Gy. Treatment times which would satisfy this condition and the corresponding dose to tumor cells are shown in Table 1.
3. Conclusion This study provides a first look at using the GE PETtrace cyclotron with 18O target for BNCT. The results suggest that it is possible to use this particular cyclotron with this target for tumor treatment. The data presented here suggest that this type of neutron source should provide acceptable doses and treatment times for tumor irradiation at a depth of up to 4 cm inside the
8 1107 25.9
brain phantom. Treatment of the tumor at a greater depth requires significant increase of the treatment time. But, since some modifications are proposed by GE in cyclotron design (increased target current, new ion source development, etc.), further research will be continued in order to improve current results.
References Bleuel, D.L., Donahue, R., Ludewigt, B., Vujic, J., 1998. Designing accelerator-based ephitermal neutron beams for boron neutron capture therapy. Med. Phys. 25, 1725–1734. Briesmeister, J.F., 2000. A general Monte-Carlo n-particle transport code. Los Alamos National Laboratory, Los Alamos, NM. Caswell, R.S., Coyne, J.J., Randolph, M.L., 1980. Kerma factors for neutron energies below 30 meV. Radiat. Res. 83, 217–254. Coderre, J., Chanana, A., Joel, D., Elowitz, E., Micca, P., Nawrocky, M., Chadha, M., Gebbers, J., Shady, M., Peress, N., Slatkin, D., 1998. Biodistribution pf boronophenylanine in patients with glioblastoma multiforme: boron concentration correlates with tumor cellularity. Radiat. Res. 149, 163–170. Coderre, J.A., Makar, M.S., Micca, P.L., Nawrocky, M.M., Liu, H.B., Joel, D.D., Slatkin, D.N., Amols, H.I., 1993. Derivations of relative biological effectiveness for the highlet radiations produced during boron neutron capture
ARTICLE IN PRESS 1062
A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
irradiations of the 9L rat gliosarcoma in vitro and in vivo. Int. J. Radiat. Oncol. Biol. Phys. 27, 1121–1129. Harling, O.K., Roberts, K.A., Moulin, D.J., Rogus, R.D., 1995. Head phantoms for neutron capture therapy. Med. Phys. 22 (5), 579–583. International Commission on Radiation Units and Measurements, 1989. Tissue Substitutes in Radiation Dosimetry and Measurement, ICRU-44.
Rogus, R., Harling, O., Yanch, J., 1994. Mixed field dosimetry of epithermal neutron beams for boron neutron capture therapy at MITR-II research reactor. Med. Phys. 21, 1611–1625. Wang, C.K., Blue, T.E., Gahbauer, R., 1989. A neutronic study of an accelerator based neutron irradiation facility for boron neutron capture therapy. Nucl. Technol. 84–93.
Applied Radiation and Isotopes 61 (2004) 1057–1062
GE PETtrace cyclotron as a neutron source for boron neutron capture therapy A. Boskoa,*, D. Zhilchenkovb, W.D. Reecea a
Department of Nuclear Engineering, Texas A&M University, College Station, USA b NuTech, Inc., Tyler, TX, USA
Abstract This paper discusses the use of a General Electric PETtrace cyclotron as a neutron source for boron neutron capture therapy. In particular, the standard PETtrace 18O target is considered. The resulting dose from the neutrons emitted from the target is evaluated using the Monte Carlo radiation transport code MCNP at different depths in a brain phantom. MCNP-simulated results are presented at 1, 2, 3, 4, 5, 6, 7, and 8 cm depth inside this brain phantom. Results showed that using a PETtrace cyclotron in the current configuration allows treating tumors at a depth of up to 4 cm with reasonable treatment times. Further increase of a beam current should significantly improve the treatment time and allow treating tumors at greater depths. r 2004 Elsevier Ltd. All rights reserved. Keywords: BNCT; Cyclotron; MCNP
1. Introduction Boron neutron capture therapy (BNCT) is under active investigation all over the world. 10B attached to a tumor seeking compound is injected intravenously into the patient and is carried to the tumor site. The boron disintegrates after capturing a neutron from a neutron beam and the high-energy particles emitted by the disintegration kill tumor cells while sparing normal tissue. The history of BNCT starts in 1936 when J. Chadwick of Cambridge University and biophysicist G. L. Locher of the Franklin Institute presented the basic theory of BNCT. In 1951 it was suggested that BNCT could be used for treatment of brain tumors, and in particular, treatment of glioblastoma multiforma (GBM). Since then many laboratories, institutes and clinics conducted research and clinic trials using BNCT in efforts to improve the efficiency. One area of research is directed *Corresponding author. E-mail address: [email protected] (A. Bosko).
toward finding a cheap, reliable, and manageable source of neutrons for BNCT. Historically the best source of neutrons at the energy and flux levels required for BNCT was a nuclear reactor. BNCT research has used reactors at Brookhaven medical research reactor (BMRR), at the Massachusetts Institute of Technology reactor (MITR), at the high flux reactor (HFR) in Institute for Energy, Pettern, Netherlands, and other facilities. The advantage of a reactor is that the neutrons from reactors are relatively cheap, if capital costs are discounted. The disadvantages of reactors are that the capital costs are very high and reactors are too complicated for an ordinary clinic to operate, so these clinics cannot afford to build and maintain a small nuclear reactor to use as a neutron source. Another approach to this problem is to use particle accelerators. These machines can accelerate protons to the required energies for a given reaction and the bombarded target emits neutrons at rates and energy levels appropriate for BNCT. These accelerators are cheaper, easier to operate, and pose less overall risk than
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.05.076
ARTICLE IN PRESS 1058
A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
nuclear reactors. Accelerator neutron sources are well understood and implemented in a number of research facilities (Wang et al., 1989). This paper investigates the possibility to make BNCT available for clinics and research groups that do not have nuclear reactor or an accelerator specially designed for BNCT. There are a large number of medical cyclotrons used for medical radioisotope production in the nuclear pharmacies and hospitals. These cyclotrons are designed mostly for fluorodeoxyglucose (FDG) production. FDG is manufactured in an automated synthesis unit from 18F. This isotope is produced in a cyclotron by proton bombardment of enriched water (18O). In this article the General Electric (GE) PETtrace cyclotron, normally used to produce FDG, is examined to study the possibility of using it for BNCT. The GE PETtrace was chosen for this investigation because this type of cyclotron is popular among nuclear pharmacies and clinics in many countries (about 75 PETtraces have been sold worldwide), in part because it is compact and reliable. This cyclotron produces protons with energies large enough to produce neutrons with appropriate energy and flux for BNCT and it does not require significant changes in design to provide these neutrons. The energy of protons made in this cyclotron is fixed at 16.5 MeV. The current on the target runs up to 60 mA. Targets are accessible, easily changeable, and this cyclotron can use dual targets simultaneously, allowing the flexibility to work on radioisotope production and BNCT with minimal investments of time and work force. There are a number of difficulties in modifying this cyclotron for BNCT and they include: *
*
*
There is no known target for BNCT technique, however, as we show in this paper, conventional targets may be acceptable; positioning of target filler (Fig. 1) will scatter neutrons emitted in the forward direction and it could result in less flux than required by BNCT, so the system configuration might have to be changed; some of these cyclotrons are assembled as selfshielded units (stainless tanks filled with borated water surround the cyclotron), and this configuration makes impossible for them to serve as a neutron source.
As a first step to using the GE PETtrace for BNCT, calculations were conducted to study the possibility of using 18O target as a source of neutrons with further reflection, moderation, absorption and collimation to produce appropriate energy and flux for BNCT. The 18O target was studied because about 90–95% of time these cyclotrons use this target already for FDG production. The efficiency of cyclotron use may be significantly increased if unused neutrons produced
Fig. 1. General Electric PETtrace cyclotron.
Fig. 2. Cyclotron vault with channel for filters and collimators proposed for BNCT (side view).
during FDG production could be utilized for other medical modalities such as BNCT at the same time when radioisotopes are produced. This research explores this approach using Monte Carlo codes with standard neutron data files, and the following assumptions: *
*
*
Standard running conditions (60 mA target current, 16.5 MeV protons, standard 18O target); no target filler system behind the target (it can be moved to the side of the cyclotron); cyclotron is in a vault with no self-shielding (Fig. 2).
2. Monte Carlo simulation The general purpose Monte Carlo code, MCNP (Briesmeister, 2000), was used to simulate the transport of neutrons through the geometry shown in Fig. 3. This geometry includes a moderator assembly surrounded by a thick graphite reflector. The moderator
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062 18
assembly consists of an aluminum container filled with D2O. Another popular type of BNCT moderator consisting of aluminum, fluorine and natural lithium was also considered. The heavy water moderator was chosen because it had lower fast-neutron contribution to the total flux in the phantom. Various thicknesses of moderating material were investigated in this study. Again, the thickness was chosen that provided the lowest fast-neutron contribution to the total flux in comparison to the low energy component and that allowed delivery of the necessary dose in a reasonable time. The optimal inner diameter of the moderator container was 30 cm for a fixed 30-cm length. The container wall set at 1 cm for structural considerations. The moderator assembly was surrounded by a 30-cm thick graphite reflector. Also, a 4-cm thick lead shield was used to reduce the dose from the photons emitted from the target. The typical 18O target consists of
O-enriched water kept under a helium over-pressure in a silver container. The neutron spectrum and angular distribution on these 18O-targets are well established (according to PETtrace-source term evaluations by Lundgren and Ingemannson, 2001). The neutron spectrum from such a target is shown in Fig. 4. This data, along with corresponding neutron yield for different angular directions, was used to define the neutron source for the MCNP calculations. The angular distribution from 0 to 180 with bins covering every 20 was constructed for input. The neutrons emitted from the target are moderated by scattering through the heavy water. The neutrons leave the moderator and may then interact within the brain for BNCT treatment. The phantom brain used to model these interactions consists of ellipsoid with major axis of 13.6, 19.6 and 16.6 cm (Harling et al., 1995) filled with a specified composition to simulate brain tissue (ICRU, 1989). The tally cells were 0.5 cm diameter spheres at depths of 1, 2, 3, 4, 5, 6, 7 and 8 cm inside the phantom brain. The flux tally was used to calculate photon and neutron dose, by applying flux-to-dose conversion factors given in the literature (Briesmeister, 2000). The thermal neutron flux as a function of the in phantom depth is plotted on Fig. 5. Most neutrons have been thermalized at the tally depths. However, in order to know the total dose delivered in the tumor and the healthy tissue, the other components of the total dose were calculated. Fig. 6 shows the energy spectrum of neutrons crossing the tally cell at a depth of 3 cm. There are two principal capture reactions for thermal neutrons in tissue—1H(n,g)2H and 14N(n,p)14C. The first capture reaction releases a 2.22 MeV gamma ray, which could deposit a fraction of its energy before escaping the phantom. In contrast, the nitrogen capture reaction releases 0.626 MeV which is deposited by the proton and
Phantom
Lead
Ø 30 cm
Target D 2O
30 cm
Aluminum container
1059
Graphite
Fig. 3. Geometry defined as input for the MCNP calculation.
1.0E+12
1.0E+11
1.1E+111.0E+11 5.9E+10 2.9E+10 1.4E+10 6.4E+09
1.0E+10
n/s
2.9E+09 1.2E+09 1.0E+09
5.1E+08 2.1E+08 1.1E+08
1.0E+08
1.0E+07 0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
Neutron energy, MeV
Fig. 4. Neutron spectrum from the H218O target.
8-9
9 - 10
> 10
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
1060 1.2E+08
8.0E+07
2
Flux , n- cm /s
1.0E+08
6.0E+07 4.0E+07 2.0E+07 0.0E+00 0
1
2
3
4
5
6
7
8
9
Depth, cm Fig. 5. Thermal neutron flux in the brain.
2
Neutron flux, n-cm /s
1.0E+08
1.0E+07
1.0E+06
1.0E+05 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00 1.0E+01 1.0E+02
Neutron energy , MeV Fig. 6. Neutron spectrum for 3 cm depth in the phantom.
recoil carbon nucleus in the immediate vicinity of the capture site. The resulting dose rate from exposure to thermal neutrons was calculated using the following formula: D14N ¼ Fth Fn14N f14N ;
ð1Þ
where D14N is the dose rate due to capture on nitrogen (cGy/s), Fth the neutron fluence (n/cm2 s), Fn14N ¼ 0:785 109 cGy cm2/n is the kerma factor for thermal neutrons, f14N ¼ 0:022 is the weight fraction of nitrogen in brain tissue (ICRU, 1989). Kerma factors for this reaction were taken from Caswell et al. (1980). In the neutron reaction, 10B(n,g)7Li, the charged particles are produced with the total kinetic energy of 2.34 MeV on average and with a range in tissue of about 10 mm, the order of the size of a tumor cell. As a result, the cell containing 10B can be destroyed. A similar formula (2)
was used to obtain the dose rate from the neutron reaction with 10B. D10B ¼ Fth Fn10B m;
ð2Þ
where D10B is the dose rate due to capture on boron (cGy/s), Fth the neutron flux (n/cm2 s), Fn10B ¼ 8:6 1012 cGy cm2/n is the corrected kerma factor for thermal neutrons obtained per 1 ppm concentration of 10 B in tissue (Rogus et al., 1994). The total dose to healthy tissue and tumor cells were calculated using the following formula: DðRBEÞtissue ¼ DðRBEÞfast þ DðRBEÞphoton þ D14N RBE14N þ D10B RBE10B ;
ð3Þ
where DðRBEÞtissue is the effective dose rate to tissue (RBE cGy/min), DðRBEÞfast is the simulated neutron
ARTICLE IN PRESS A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062 14
Fast Neutron Dose Photon Dose
12
RBE-Dose,cG y/min .
1061
Nitrogen-14 Dose Boron-10 in tumor Dose (40 ppm)
10
Boron-10 in healthy tissue dose (11 ppm) 8
Healthy tissue Dose Tumor Dose
6 4 2 0 0
1
2
3
5
4
6
7
8
9
Depth, cm Fig. 7. RBE-dose rates at different depths into the brain phantom.
Table 1 Estimated treatment times and the dose delivered to the tumor In-phantom depth (cm) Treatment time (min) Tumor dose (Gy)
1
2
3
4
5
6
7
258 30.7
288 29.3
381 30.7
444 29.8
459 26.4
484 24.4
545 22.1
dose rate (RBE cGy/min), DðRBEÞphoton is the simulated photon dose rate (RBE cGy/min), D14N and D10B are the dose from interactions of thermal neutrons with nitrogen and boron, RBE14N ¼ 3:2; RBE10B ¼ 1:3 (for normal tissue) and RBE10B ¼ 3:8 (for tumor) are the relative biological effectiveness given in the literature (Coderre et al., 1993). 10B concentrations in the normal tissue of 11.4 ppm and within the tumor of 40 ppm were assumed (Coderre et al., 1998). The RBE-dose rates obtained from calculations are plotted in Fig. 7. One of the most important clinical requirements according to the Brookhaven Medical Research Reactor protocol (Bleuel et al., 1998) is that the maximum dose to the normal brain should not exceed 12.5 RBE Gy. Treatment times which would satisfy this condition and the corresponding dose to tumor cells are shown in Table 1.
3. Conclusion This study provides a first look at using the GE PETtrace cyclotron with 18O target for BNCT. The results suggest that it is possible to use this particular cyclotron with this target for tumor treatment. The data presented here suggest that this type of neutron source should provide acceptable doses and treatment times for tumor irradiation at a depth of up to 4 cm inside the
8 1107 25.9
brain phantom. Treatment of the tumor at a greater depth requires significant increase of the treatment time. But, since some modifications are proposed by GE in cyclotron design (increased target current, new ion source development, etc.), further research will be continued in order to improve current results.
References Bleuel, D.L., Donahue, R., Ludewigt, B., Vujic, J., 1998. Designing accelerator-based ephitermal neutron beams for boron neutron capture therapy. Med. Phys. 25, 1725–1734. Briesmeister, J.F., 2000. A general Monte-Carlo n-particle transport code. Los Alamos National Laboratory, Los Alamos, NM. Caswell, R.S., Coyne, J.J., Randolph, M.L., 1980. Kerma factors for neutron energies below 30 meV. Radiat. Res. 83, 217–254. Coderre, J., Chanana, A., Joel, D., Elowitz, E., Micca, P., Nawrocky, M., Chadha, M., Gebbers, J., Shady, M., Peress, N., Slatkin, D., 1998. Biodistribution pf boronophenylanine in patients with glioblastoma multiforme: boron concentration correlates with tumor cellularity. Radiat. Res. 149, 163–170. Coderre, J.A., Makar, M.S., Micca, P.L., Nawrocky, M.M., Liu, H.B., Joel, D.D., Slatkin, D.N., Amols, H.I., 1993. Derivations of relative biological effectiveness for the highlet radiations produced during boron neutron capture
ARTICLE IN PRESS 1062
A. Bosko et al. / Applied Radiation and Isotopes 61 (2004) 1057–1062
irradiations of the 9L rat gliosarcoma in vitro and in vivo. Int. J. Radiat. Oncol. Biol. Phys. 27, 1121–1129. Harling, O.K., Roberts, K.A., Moulin, D.J., Rogus, R.D., 1995. Head phantoms for neutron capture therapy. Med. Phys. 22 (5), 579–583. International Commission on Radiation Units and Measurements, 1989. Tissue Substitutes in Radiation Dosimetry and Measurement, ICRU-44.
Rogus, R., Harling, O., Yanch, J., 1994. Mixed field dosimetry of epithermal neutron beams for boron neutron capture therapy at MITR-II research reactor. Med. Phys. 21, 1611–1625. Wang, C.K., Blue, T.E., Gahbauer, R., 1989. A neutronic study of an accelerator based neutron irradiation facility for boron neutron capture therapy. Nucl. Technol. 84–93.