- Email: [email protected]
Ion therapy for uveal melanoma in new human eye phantom based on GEANT4 toolkit
Medical Dosimetry ] (2016) ]]]–]]]
Medical Dosimetry journal homepage: www.meddos.org
Ion therapy for uveal melanoma in new human eye phantom based on GEANT4 toolkit Seyed Ali Mahdipour, M.Sc.,* and Ali Asghar Mowlavi, Ph.D.*† *Physics Department, Hakim Sabzevari University, Sabzevar, Iran; and †ICTP, Associate Federation Scheme, Medical Physics Field, Trieste, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 January 2015 Received in revised form 7 August 2015 Accepted 4 October 2015
Radiotherapy with ion beams like proton and carbon has been used for treatment of eye uveal melanoma for many years. In this research, we have developed a new phantom of human eye for Monte Carlo simulation of tumors treatment to use in GEANT4 toolkit. Total depthdose profiles for the proton, alpha, and carbon incident beams with the same ranges have been calculated in the phantom. Moreover, the deposited energy of the secondary particles for each of the primary beams is calculated. The dose curves are compared for 47.8 MeV proton, 190.1 MeV alpha, and 1060 MeV carbon ions that have the same range in the target region reaching to the center of tumor. The passively scattered spread-out Bragg peak (SOBP) for each incident beam as well as the flux curves of the secondary particles including neutron, gamma, and positron has been calculated and compared for the primary beams. The high sharpness of carbon beam's Bragg peak with low lateral broadening is the benefit of this beam in hadrontherapy but it has disadvantages of dose leakage in the tail after its Bragg peak and high intensity of neutron production. However, proton beam, which has a good conformation with tumor shape owing to the beam broadening caused by scattering, can be a good choice for the large-size tumors. & 2016 American Association of Medical Dosimetrists.
Keywords: Ion beams Eye phantom Dose Primary GEANT4 toolkit
Introduction There is increasing evidence that the Monte Carlobased codes or toolkits are the most powerful tools for nuclear particles transport calculations used in many fields such as medical physics.1 The major goal of radiotherapy is to reach a maximum dose to the tumor and to spare the surrounding normal tissue as much as possible.2 Using hadron radiotherapy is advantageous as it provides a spatial form of ionizing energy deposition that is predominant at the end of the pathway of incident particles, represented by a curve whose maximum amplitude is known as the “Bragg peak.” Despite the absence of any tail in proton dose profile after its Bragg peak, there is a tail for the carbon's dose profile, exposing regions immediately posterior to the peak to doses corresponding to approximately 10% of the peak.3 These days, radiotherapy represents an important advance in the treatment of uveal melanoma, including radiotherapeutic plaques4-6 and external beam irradiations,7-11 which are the most commonly used therapeutic modalities. There are 3 types of
Reprint requests to Ali Asghar Mowlavi, Hakim Sabzevari University, P.O. Box 397, Sabzevar, Iran. E-mail: [email protected] http://dx.doi.org/10.1016/j.meddos.2015.10.005 0958-3947/Copyright Ó 2016 American Association of Medical Dosimetrists
tumors in uveal melanoma called T1, T2, and T3 based on their depth. Proton, alpha, and carbon ion therapy is applied in cases of T2 and T3 lesions (4 to 5 and 6 to 10 mm thickness).12 Many reports mentioned that the most uveal melanomas are large with a mean tumor thickness of 6.5 mm; approximately 60% of the patients had tumors that extended anterior to the equator.10-13 In medical dosimetry, usually Monte Carlo codes or toolkits such as Monte Carlo N-Particle Transport Code (MCNP) and GEANT4 are applied by many researchers.14,15 GEANT4 is a Monte Carlobased toolkit for simulating the passage of particles through matter from the high-energy physics community. It is capable of handling all physics processes, including electromagnetic, hadronic, and nucleusnucleus interactions, which are indispensable to calculate 3dimensional (3D) dose distributions in ion therapy.16 In this work, we have designed and presented a new human eye phantom model for use in the GEANT4 toolkit. We have calculated the dose profiles and deposited energy of the beamlets of proton, alpha, and carbon by Monte Carlo simulation. These beamlets irradiated the new eye phantom with a uveal melanoma. Furthermore, the dose profiles of secondary particles in the phantom were evaluated. During the simulation, multiple scattering, ionization, and elastic and nonelastic interactions have been taken into account through GEANT4 Physics references and the
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
2
Fig. 1. (A) Schematic of the simulated human eye phantom. (B) Schematic of the human eye with deep tumor in the vitreous humor. (Color version of figure is available online.)
physical characteristics such as lateral displacement, direction distributions, and deposited energy have been scored for the primary and secondary particles.
Particle source and physics references The particle source and transport parameters in this simulation process are based on the setup of Christovao et al.18 published in 2011:
Methods and Materials
Eye model
A new model of human eye phantom consisting of vitreous humor, choroid, lens, aqueous humor, cornea, and optic nerve has been designed in this study. The phantom geometry is presented in Fig. 1. Dimensions of the human eye are adapted from optics of the human eye published by Atchison and Smith17 in 2003. The model was constructed using concentric spheres with appropriate offset. The vitreous humor was defined to be the volume posterior to the lens, and the anterior chamber as anterior to the lens. The optic nerve is simulated as a cylinder appropriately offset from the posterior pole of the eye. Tumor is considered as a virtual sphere with 8-mm radius. Schematic shape of the eye phantom has been presented in Fig. 1. Compositions of the organs and mass densities are adapted from GEANT4 material database (National Institute of Standards and Technology (NIST) compounds),14,15 based on the reports of International Commission on Radiation Units (ICRU) and International Commission on Radiological Protection (ICRP). The material of the lens and cornea is taken to be the G4_EYE_LENS_ICRP compound; vitreous humor, choroid, and aqueous humor were approximated by G4_BRAIN_ICRP and optic nerves material is G4_MUSCLE_STRAITED_ICRU. Densities of organs and their compositions as mass percentages of elements are listed in Table 1. Most of the uveal melanomas appear near the choroid and sclera structures of the eye. We have considered an 8-mm eye type T3 tumor, which is a part of vitreous humor volume with a diameter along the x-axis. The location of tumor in the phantom region is from x ¼ 1.53 to 2.33 cm, where the initial point of choroid.
The beam source was located at 7.0 mm to the right (z direction) from the center of the eye lens in the voxel model. The beam distance from the eye phantom is 5.0 cm. Its spatial distribution was described by a Gaussian distribution and radius of beam is 2.5 mm.
To obtain depth-dose profiles in the eye phantom, it was equally sliced along the beam axis and deposited energy of each slice was scored. In the calculations, 2 different mesh grids were used: the first grid had 0.06725-mm thickness in X direction and 10 mm lateral dimensions of the phantom all over the organs, and the second one is a 3D rectangular mesh with 0.06725 0.2 0.2 mm3 voxel size, used to generate 3D depthdeposited energy energy profiles for the primary particles. For the clinical investigation, we have not only formed a passively scattered spread-out Bragg peak (SOBP) for proton, alpha, and carbon beams that cover the tumor volume (8 mm) but also used a modulation wheel (polymethyl methacrylate (PMMA) range shifter) that is frequently used in the hadron treatments of the CATANA facility.19 We have
Table 1 Mass density and elemental composition of the tissues in the eye phantom (compositions are expressed as percentage by weight14,15) Tissue type
Density (g/cm3)
H
C
N
O
Ca
Na
P
S
Cl
K
Lens and cornea Choroid, vitreous, and aqueous Optic nerve Water
1.10 1.03 1.04 1.00
9.92 11.06 10.06 11.19
19.38 12.54 10.78 –
5.32 1.33 2.76 –
65.38 73.78 75.47 88.81
– 0.0009 0.0003 –
– 0.18 0.07 –
– 0.35 0.18 –
– 0.18 0.24 –
– 0.23 0.07 –
– 0.31 0.30 –
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
3
calculated the SOBP for proton, alpha, and carbon ions in the box phantom with 0.16 10 10 mm3 slabs. The physics reference of the simulation is hadronic interaction framework within the GEANT4 toolkit. The hadronic physics list was used to simulate nuclear interactions including elastic and inelastic scattering of hadrons. Also, we have used QGSP_BIC_EMY reference physics list from Hadrontherapy example in GEANT4 toolkit.19
Results and Discussion To validate the simulation results, the percentage depth-dose profile of 62-MeV monoenergy proton beam was calculated in the 4 4 30 cm3 water phantom. The result was compared to the recent simulation report result of Gorjiara et al.20 To obtain the total depthdose profile in the phantom, the phantom was equally sliced along the beam axis, and total dose of each slice with 0.3 mm thickness was scored. Figure 2 shows our Monte Carlo results and the values obtained by Gorjiara et al.20 The delivered dose reaches the Bragg peak after passing through just 29.1 mm of their setup compared with 29.13 mm of this study. So the dose decreases to 80% of the Bragg peak at approximately 29.95 and 29.99 mm depths for the setup of Gorjiara et al.20 and for our calculation, respectively. It can be seen that the difference between the 2 compared results is less than 3% and they are in a good agreement. We have illustrated the deposited energy curves for proton, alpha, and carbon beams per particle in Fig. 3. The results show that alpha and carbon particles, which are heavier than a proton, scatter less and hence provide sharper edge definition for the radiation field. Ionization density varies as the square of the ion charge and heavy ions have higher linear energy transfer than protons. In Fig. 4, pristine Bragg peak positions are shown as a function of proton, alpha, and carbon energy. The data are fitted very well by a second-order polynomial function and the evaluated parameters are listed over the figures. From these curves we determined the energy of proton and alpha with the same range as that of the 1060-MeV carbon beam as 47.8 and 190.1 MeV, respectively. Our simulation was run with 2 106 events for proton, alpha, and carbon beams without modulation. Also, we set the cut value parameter equal to 0.01 mm in the GEANT4 toolkit used in the meshed phantom region. The depth-dose profiles of primary,
Fig. 2. Comparison between our result for total percentage depth-dose profile from 62-MeV proton pencil beam and the result of Gorjiara et al.20 (Color version of figure is available online.)
Fig. 3. The deposited energy along the path of the primary particles for different energies (A) for proton, (B) for alpha, and (C) for carbon beams. (Color version of figure is available online.)
4
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
secondary, and total particles for 190.1 MeV alpha and 1060 MeV carbon beams are shown in Fig. 5. These results can be explained by considering the influence of nonelastic nuclear interactions of the primary particles with the target nuclei. In the case of alpha and carbon ions incident beams, the contribution of the secondary particles to the total dose profile is mainly due to projectile fragments, which are characterized by a larger range as shown in Fig. 5. Figure 6A compares the total depthdose distribution with the same ranges at the tumor center (x ¼ 1.9 cm) for different beams. In the nonnormalized dose curves presented in Fig. 6A, the maximum total deposited doses were 8.17 1010, 4.22 109, and 2.82 108 Gy/particle for proton, alpha, and carbon beams respectively. It is necessary to mention that in Fig. 6B we normalized the Bragg peak of proton and alpha to the carbon's peak. Results of the total absorbed dose and total dose percentage in the different organs of the eye phantom are listed in Table 2. The
Fig. 4. The Bragg peak position within the tumor region as a function of beam energy. (Color version of figure is available online.)
Fig. 5. Total depth-dose and depth-dose profiles of the primary and secondary particles: (A) alpha primary particle with 190.1 MeV and (B) carbon primary particle with 1060 MeV energy. (Color version of figure is available online.)
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
5
obtained data show that alpha particle stores a large percentage of the total dose in the plateau area (the entrance region of the depthdose profile), which includes regions before the tumor. In this case, the lowest contribution among beams is for carbon ions. The highest percentage of dose deposition in the tumor region is 43.97% for carbon beam and the lowest is 38.65% for alpha particles. Therefore, it seems that without considering biological aspects, carbon and proton have better cure effects than does the alpha beam. In Fig. 7, SOBPs for proton, alpha, and carbon ions are compared by using physical optimization. Carbon ions showed higher dose behind the target because of the fragmentation tail as depicted in Fig. 7A. Results plotted in Figs. 6 and 7 and listed in Table 2 lead to the conclusion that the absence of tail in the total depthdose profile and SOBP of proton beam causes the proton to store a lower dose in the organs after the tumor-like choroid and optic nerve. This is a benefit in using proton beams in treatment of small sensitive organs like eye. The flux curves of the secondary particles such as positron, neutron, and photon are shown in Fig. 8. It is clear that the curves depend on the energy of the primary beam and it was found to be larger than others for the carbon ions. The highest contribution of
Fig. 6. Comparison among total depthdose profiles with same ranges for different primary particles in eye phantom. (Color version of figure is available online.)
Table 2 Total dose in Gy/particle and the percentage of the absorbed dose in varius organs of the eye phantom for different primary beams Organ
Dose for 47.8 MeV proton beam
Dose percent (%)
Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
20.91 10E12 23.43 10E11 18.22 10E11 14.79 10E16 16.81 10E16 43.74 10E11
4.78 53.57 41.65 0.000381 0.000384 100
Dose for 190.1 MeV alpha beam Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
89.35 10E12 12.08 10E10 83.1 10E11 10.38 10E12 11.35 10E12 21.5 10E10
4.15 56.20 38.65 0.483 0.527 100
Dose for 1060 MeV carbon beam Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
41.30 10E11 53.53 10E10 47.8 10E10 14.21 10E11 17.44 10E11 10.87 10E9
3.79 49.25 43.97 1.3 1.6 100
Fig. 7. Comparison of the spreading of the Bragg peak for proton, alpha, and carbon incident beams: (A) nonnormalized and (B) normalized to carbon's SOBP. (Color version of figure is available online.)
6
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
Fig. 8. The secondary particles flux in the phantom region for proton, alpha, and carbon primary particles: (A) positron flux and (B) neutron flux. (Color version of figure is available online.)
the neutron flux and the secondary positron production corresponded to the carbon beam, due to the high energy of carbon particles. The low-energy neutrons with the high intensity in the flux curves may have further contribution in penetrating the nucleus and treatment problems. In the photon flux curves produced from primary particles interaction with phantom, some of the peaks correspond to excited nuclei like 40Ca, 16O, 31P, 14N, and others, as shown in Fig. 9. Figure 10 presents 2D deposited energy contours per particle for the primary beams. We can see that with the heavier ion beams, the longitudinal extent of deposited energy at the proton Bragg peak location also increases compared with the 2 other beams. In Fig. 11, the lateral distribution of deposited energy at the Bragg peak location is shown for the 3 different beams. We can find out that alpha and carbon beams are scattered less than the proton beam is. Therefore, the greatest lateral scattering and deposited energy of the beams are for the proton and the lowest are for carbon. It seems that in this case, by increasing the lateral diameter of eye tumors, using a proton beam owing to the further lateral scattering and less production of secondary particles has advantages.
Fig. 9. Gamma flux curves in the phantom region for (A) proton, (B) alpha, and (C) carbon beam. (Color version of figure is available online.)
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
7
Fig. 12. The lateral distribution of deposited energy for the proton beam at various depths. (Color version of figure is available online.)
and as shown, the maximum deposited energy and lateral scattering of incident protons are located at the position of its Bragg peak.
Conclusion
Fig. 10. Monte Carlo calculated contours of deposited energy for primary particles: (A) proton, (B) alpha, and (C) carbon. (Color version of figure is available online.)
The lateral distribution of deposited energy plotted for the proton beam at various depths of the phantom is shown in Fig. 12. It is seen that the lateral broadening or the lateral scattering increases with the penetration depth of proton beam. As expected
We have designed and developed a new human eye phantom with all details for the GEANT4 toolkit. By using the phantom, the dose and deposited energy for the primary and secondary particles in ion therapy for deep uveal melanoma have been calculated. For the clinical investigation, passively scattered SOBPs of incident beams covering the tumor volume have been compared. Also, the secondary particles flux and the lateral distribution of deposited dose for the primary proton, alpha, and carbon beams were obtained and compared. The results of this research show that among proton, alpha, and carbon primary beams with the same range in the phantom region (1.9 cm), the heavy carbon ions have the highest dose percentage in the tumor region without considering the risk of secondary particles and the tail dose leakage. Protons can be a good choice for the large-size tumors, to have a good conformation with tumor shape owing to the beam broadening caused by scattering.
Acknowledgment The authors of this work are indebted to M. Severgnini from the Department of Medical Physics, A.O.U. “Ospedali Riuniti” di Trieste, for valuable remarks. References
Fig. 11. A comparison among lateral distributions of deposited energy by proton, alpha, and carbon ion beams having the same range in the eye phantom. (Color version of figure is available online.)
1. Ilic,́ R.D.; Lalić, D.; Stanković, S.J. Srna - Monte Carlo codes for proton transport simulation in combined and voxelized geometries. Nucl. Technol. Radiat. Prot. 17:27–36; 2002. 2. Scholz, M. Heavy ion tumour therapy. Nucl. Instrum. Methods Phys. Res. B 161:76–82; 2000. 3. Amaldi, U; Kraft, G. Radiotherapy with beams of carbon ions. Rep. Prog. Phys. 68:1861–82; 2005. 4. Zhang, H.; Davidorf, F.; Qi, Y. Comparison of 16 mm OSU-Nag and COMS eye plaques. J. Appl. Clin. Med. Phys. 13(3):166–78; 2012. 5. Shields, J.A.; Augsburger, J.J.; Brady, L.W.; et al. Cobalt plaque therapy of posterior uveal melanomas. Ophthalmology 89:1201–7; 1982. 6. Packer, S.; Rotman, M. Radiotherapy of choroidal melanoma with iodine-125. Ophthalmology. 87:582–90; 1980. 7. Egger, E.; Zografos, L.; Munkel, G.; et al. Results of proton radiotherapy for uveal melanomas. Front. Radiat. Ther. Oncol. 30:111–22; 1997.
8
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
8. Castro, J.R.; Char, D.H.; Petti, P.L.; et al. 15 years' experience with helium ion radiotherapy for uveal melanoma. Int. J. Radiat. Oncol. Biol. Phys. 39:989–96; 1997. 9. Desjardins, L.; Levy, C.; D'Hermies, F.; et al. Initial results of proton therapy in choroidal melanoma at the d'Orsey Center for Proton Therapy: The first 464 cases. Cancer Radiother. 1:222–6; 1997. 10. Fuss, M.; Loredo, L.N.; Blacharski, P.A.; et al. Proton radiation therapy for medium and large choroidal melanoma: Preservation of the eye and its functionality. Int. J. Radiat. Oncol. Biol. Phys. 49:1053–9; 2001. 11. Swakon, J.; Olko, P.; Adamczyk, D.; et al. Facility for proton radiotherapy of eye cancer at IFJ PAN in Krakow. Radiat. Meas. 45(10):1469–71; 2010. 12. Sobin, L.H.; Wittekind, Ch. TNM classification of malignant tumours. 6th ed., New York, NY: Wiley-Liss; 2002. 13. Char, D.H.; Castro, J.R.; Kroll, S.M.; et al. Five-year follow-up of helium ion therapy for uveal melonoma. Arch Ophthalmol. 108(2):209–14; 1990.
14. Agostinelli, S.; Allison, J.; Amako, K.; et al. Geant4—A simulation toolkit. Nucl. Instr. Methods. A506:250; 2003. 15. Allison, J.; Amako, K.; Apostolakis, H.; et al. Geant4 developments and applications. Nucl. Sci. 53:270; 2006. 16. Toshito, T.; Bagulya, A.; Lechner, A.; et al. Validation of New Geant4 Electromagnetic Physics Models for Ion Therapy Applications. Nucl. Sci. Technol. 2:918–22; 2011. 17. Atchison, D; Smith, G. Optics of the human eye. Edinburgh, 2003. 18. Christovao, M.T.; Ribeiro de Campos, T.P.; Trindade, B.M. Simulation and dosimetric analysis of proton and carbon ion therapy in the treatment of uveal melanoma. Radiol. Bras. 44(6):367–73; 2011. 19. Cirrone, G.A.P.; et al. Hadrontherapy: A Geant4-Based Tool for Proton/IonTherapy Studies. Prog. Nucl. Sci. Technol. 2:207–12; 2011. 20. Gorjiara, T.; Kuncic, Z.; Doran, S.; et al. Water and tissue equivalence of a new PRESAGE (R) formulation for 3D proton beam dosimetry: A Monte Carlo study. Med. Phys. 39(11):7071–9; 2012.
Medical Dosimetry journal homepage: www.meddos.org
Ion therapy for uveal melanoma in new human eye phantom based on GEANT4 toolkit Seyed Ali Mahdipour, M.Sc.,* and Ali Asghar Mowlavi, Ph.D.*† *Physics Department, Hakim Sabzevari University, Sabzevar, Iran; and †ICTP, Associate Federation Scheme, Medical Physics Field, Trieste, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 January 2015 Received in revised form 7 August 2015 Accepted 4 October 2015
Radiotherapy with ion beams like proton and carbon has been used for treatment of eye uveal melanoma for many years. In this research, we have developed a new phantom of human eye for Monte Carlo simulation of tumors treatment to use in GEANT4 toolkit. Total depthdose profiles for the proton, alpha, and carbon incident beams with the same ranges have been calculated in the phantom. Moreover, the deposited energy of the secondary particles for each of the primary beams is calculated. The dose curves are compared for 47.8 MeV proton, 190.1 MeV alpha, and 1060 MeV carbon ions that have the same range in the target region reaching to the center of tumor. The passively scattered spread-out Bragg peak (SOBP) for each incident beam as well as the flux curves of the secondary particles including neutron, gamma, and positron has been calculated and compared for the primary beams. The high sharpness of carbon beam's Bragg peak with low lateral broadening is the benefit of this beam in hadrontherapy but it has disadvantages of dose leakage in the tail after its Bragg peak and high intensity of neutron production. However, proton beam, which has a good conformation with tumor shape owing to the beam broadening caused by scattering, can be a good choice for the large-size tumors. & 2016 American Association of Medical Dosimetrists.
Keywords: Ion beams Eye phantom Dose Primary GEANT4 toolkit
Introduction There is increasing evidence that the Monte Carlobased codes or toolkits are the most powerful tools for nuclear particles transport calculations used in many fields such as medical physics.1 The major goal of radiotherapy is to reach a maximum dose to the tumor and to spare the surrounding normal tissue as much as possible.2 Using hadron radiotherapy is advantageous as it provides a spatial form of ionizing energy deposition that is predominant at the end of the pathway of incident particles, represented by a curve whose maximum amplitude is known as the “Bragg peak.” Despite the absence of any tail in proton dose profile after its Bragg peak, there is a tail for the carbon's dose profile, exposing regions immediately posterior to the peak to doses corresponding to approximately 10% of the peak.3 These days, radiotherapy represents an important advance in the treatment of uveal melanoma, including radiotherapeutic plaques4-6 and external beam irradiations,7-11 which are the most commonly used therapeutic modalities. There are 3 types of
Reprint requests to Ali Asghar Mowlavi, Hakim Sabzevari University, P.O. Box 397, Sabzevar, Iran. E-mail: [email protected] http://dx.doi.org/10.1016/j.meddos.2015.10.005 0958-3947/Copyright Ó 2016 American Association of Medical Dosimetrists
tumors in uveal melanoma called T1, T2, and T3 based on their depth. Proton, alpha, and carbon ion therapy is applied in cases of T2 and T3 lesions (4 to 5 and 6 to 10 mm thickness).12 Many reports mentioned that the most uveal melanomas are large with a mean tumor thickness of 6.5 mm; approximately 60% of the patients had tumors that extended anterior to the equator.10-13 In medical dosimetry, usually Monte Carlo codes or toolkits such as Monte Carlo N-Particle Transport Code (MCNP) and GEANT4 are applied by many researchers.14,15 GEANT4 is a Monte Carlobased toolkit for simulating the passage of particles through matter from the high-energy physics community. It is capable of handling all physics processes, including electromagnetic, hadronic, and nucleusnucleus interactions, which are indispensable to calculate 3dimensional (3D) dose distributions in ion therapy.16 In this work, we have designed and presented a new human eye phantom model for use in the GEANT4 toolkit. We have calculated the dose profiles and deposited energy of the beamlets of proton, alpha, and carbon by Monte Carlo simulation. These beamlets irradiated the new eye phantom with a uveal melanoma. Furthermore, the dose profiles of secondary particles in the phantom were evaluated. During the simulation, multiple scattering, ionization, and elastic and nonelastic interactions have been taken into account through GEANT4 Physics references and the
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
2
Fig. 1. (A) Schematic of the simulated human eye phantom. (B) Schematic of the human eye with deep tumor in the vitreous humor. (Color version of figure is available online.)
physical characteristics such as lateral displacement, direction distributions, and deposited energy have been scored for the primary and secondary particles.
Particle source and physics references The particle source and transport parameters in this simulation process are based on the setup of Christovao et al.18 published in 2011:
Methods and Materials
Eye model
A new model of human eye phantom consisting of vitreous humor, choroid, lens, aqueous humor, cornea, and optic nerve has been designed in this study. The phantom geometry is presented in Fig. 1. Dimensions of the human eye are adapted from optics of the human eye published by Atchison and Smith17 in 2003. The model was constructed using concentric spheres with appropriate offset. The vitreous humor was defined to be the volume posterior to the lens, and the anterior chamber as anterior to the lens. The optic nerve is simulated as a cylinder appropriately offset from the posterior pole of the eye. Tumor is considered as a virtual sphere with 8-mm radius. Schematic shape of the eye phantom has been presented in Fig. 1. Compositions of the organs and mass densities are adapted from GEANT4 material database (National Institute of Standards and Technology (NIST) compounds),14,15 based on the reports of International Commission on Radiation Units (ICRU) and International Commission on Radiological Protection (ICRP). The material of the lens and cornea is taken to be the G4_EYE_LENS_ICRP compound; vitreous humor, choroid, and aqueous humor were approximated by G4_BRAIN_ICRP and optic nerves material is G4_MUSCLE_STRAITED_ICRU. Densities of organs and their compositions as mass percentages of elements are listed in Table 1. Most of the uveal melanomas appear near the choroid and sclera structures of the eye. We have considered an 8-mm eye type T3 tumor, which is a part of vitreous humor volume with a diameter along the x-axis. The location of tumor in the phantom region is from x ¼ 1.53 to 2.33 cm, where the initial point of choroid.
The beam source was located at 7.0 mm to the right (z direction) from the center of the eye lens in the voxel model. The beam distance from the eye phantom is 5.0 cm. Its spatial distribution was described by a Gaussian distribution and radius of beam is 2.5 mm.
To obtain depth-dose profiles in the eye phantom, it was equally sliced along the beam axis and deposited energy of each slice was scored. In the calculations, 2 different mesh grids were used: the first grid had 0.06725-mm thickness in X direction and 10 mm lateral dimensions of the phantom all over the organs, and the second one is a 3D rectangular mesh with 0.06725 0.2 0.2 mm3 voxel size, used to generate 3D depthdeposited energy energy profiles for the primary particles. For the clinical investigation, we have not only formed a passively scattered spread-out Bragg peak (SOBP) for proton, alpha, and carbon beams that cover the tumor volume (8 mm) but also used a modulation wheel (polymethyl methacrylate (PMMA) range shifter) that is frequently used in the hadron treatments of the CATANA facility.19 We have
Table 1 Mass density and elemental composition of the tissues in the eye phantom (compositions are expressed as percentage by weight14,15) Tissue type
Density (g/cm3)
H
C
N
O
Ca
Na
P
S
Cl
K
Lens and cornea Choroid, vitreous, and aqueous Optic nerve Water
1.10 1.03 1.04 1.00
9.92 11.06 10.06 11.19
19.38 12.54 10.78 –
5.32 1.33 2.76 –
65.38 73.78 75.47 88.81
– 0.0009 0.0003 –
– 0.18 0.07 –
– 0.35 0.18 –
– 0.18 0.24 –
– 0.23 0.07 –
– 0.31 0.30 –
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
3
calculated the SOBP for proton, alpha, and carbon ions in the box phantom with 0.16 10 10 mm3 slabs. The physics reference of the simulation is hadronic interaction framework within the GEANT4 toolkit. The hadronic physics list was used to simulate nuclear interactions including elastic and inelastic scattering of hadrons. Also, we have used QGSP_BIC_EMY reference physics list from Hadrontherapy example in GEANT4 toolkit.19
Results and Discussion To validate the simulation results, the percentage depth-dose profile of 62-MeV monoenergy proton beam was calculated in the 4 4 30 cm3 water phantom. The result was compared to the recent simulation report result of Gorjiara et al.20 To obtain the total depthdose profile in the phantom, the phantom was equally sliced along the beam axis, and total dose of each slice with 0.3 mm thickness was scored. Figure 2 shows our Monte Carlo results and the values obtained by Gorjiara et al.20 The delivered dose reaches the Bragg peak after passing through just 29.1 mm of their setup compared with 29.13 mm of this study. So the dose decreases to 80% of the Bragg peak at approximately 29.95 and 29.99 mm depths for the setup of Gorjiara et al.20 and for our calculation, respectively. It can be seen that the difference between the 2 compared results is less than 3% and they are in a good agreement. We have illustrated the deposited energy curves for proton, alpha, and carbon beams per particle in Fig. 3. The results show that alpha and carbon particles, which are heavier than a proton, scatter less and hence provide sharper edge definition for the radiation field. Ionization density varies as the square of the ion charge and heavy ions have higher linear energy transfer than protons. In Fig. 4, pristine Bragg peak positions are shown as a function of proton, alpha, and carbon energy. The data are fitted very well by a second-order polynomial function and the evaluated parameters are listed over the figures. From these curves we determined the energy of proton and alpha with the same range as that of the 1060-MeV carbon beam as 47.8 and 190.1 MeV, respectively. Our simulation was run with 2 106 events for proton, alpha, and carbon beams without modulation. Also, we set the cut value parameter equal to 0.01 mm in the GEANT4 toolkit used in the meshed phantom region. The depth-dose profiles of primary,
Fig. 2. Comparison between our result for total percentage depth-dose profile from 62-MeV proton pencil beam and the result of Gorjiara et al.20 (Color version of figure is available online.)
Fig. 3. The deposited energy along the path of the primary particles for different energies (A) for proton, (B) for alpha, and (C) for carbon beams. (Color version of figure is available online.)
4
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
secondary, and total particles for 190.1 MeV alpha and 1060 MeV carbon beams are shown in Fig. 5. These results can be explained by considering the influence of nonelastic nuclear interactions of the primary particles with the target nuclei. In the case of alpha and carbon ions incident beams, the contribution of the secondary particles to the total dose profile is mainly due to projectile fragments, which are characterized by a larger range as shown in Fig. 5. Figure 6A compares the total depthdose distribution with the same ranges at the tumor center (x ¼ 1.9 cm) for different beams. In the nonnormalized dose curves presented in Fig. 6A, the maximum total deposited doses were 8.17 1010, 4.22 109, and 2.82 108 Gy/particle for proton, alpha, and carbon beams respectively. It is necessary to mention that in Fig. 6B we normalized the Bragg peak of proton and alpha to the carbon's peak. Results of the total absorbed dose and total dose percentage in the different organs of the eye phantom are listed in Table 2. The
Fig. 4. The Bragg peak position within the tumor region as a function of beam energy. (Color version of figure is available online.)
Fig. 5. Total depth-dose and depth-dose profiles of the primary and secondary particles: (A) alpha primary particle with 190.1 MeV and (B) carbon primary particle with 1060 MeV energy. (Color version of figure is available online.)
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
5
obtained data show that alpha particle stores a large percentage of the total dose in the plateau area (the entrance region of the depthdose profile), which includes regions before the tumor. In this case, the lowest contribution among beams is for carbon ions. The highest percentage of dose deposition in the tumor region is 43.97% for carbon beam and the lowest is 38.65% for alpha particles. Therefore, it seems that without considering biological aspects, carbon and proton have better cure effects than does the alpha beam. In Fig. 7, SOBPs for proton, alpha, and carbon ions are compared by using physical optimization. Carbon ions showed higher dose behind the target because of the fragmentation tail as depicted in Fig. 7A. Results plotted in Figs. 6 and 7 and listed in Table 2 lead to the conclusion that the absence of tail in the total depthdose profile and SOBP of proton beam causes the proton to store a lower dose in the organs after the tumor-like choroid and optic nerve. This is a benefit in using proton beams in treatment of small sensitive organs like eye. The flux curves of the secondary particles such as positron, neutron, and photon are shown in Fig. 8. It is clear that the curves depend on the energy of the primary beam and it was found to be larger than others for the carbon ions. The highest contribution of
Fig. 6. Comparison among total depthdose profiles with same ranges for different primary particles in eye phantom. (Color version of figure is available online.)
Table 2 Total dose in Gy/particle and the percentage of the absorbed dose in varius organs of the eye phantom for different primary beams Organ
Dose for 47.8 MeV proton beam
Dose percent (%)
Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
20.91 10E12 23.43 10E11 18.22 10E11 14.79 10E16 16.81 10E16 43.74 10E11
4.78 53.57 41.65 0.000381 0.000384 100
Dose for 190.1 MeV alpha beam Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
89.35 10E12 12.08 10E10 83.1 10E11 10.38 10E12 11.35 10E12 21.5 10E10
4.15 56.20 38.65 0.483 0.527 100
Dose for 1060 MeV carbon beam Cornea Aqueous, lens, and vitreous Tumor Choroid Optic nerve Total
41.30 10E11 53.53 10E10 47.8 10E10 14.21 10E11 17.44 10E11 10.87 10E9
3.79 49.25 43.97 1.3 1.6 100
Fig. 7. Comparison of the spreading of the Bragg peak for proton, alpha, and carbon incident beams: (A) nonnormalized and (B) normalized to carbon's SOBP. (Color version of figure is available online.)
6
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
Fig. 8. The secondary particles flux in the phantom region for proton, alpha, and carbon primary particles: (A) positron flux and (B) neutron flux. (Color version of figure is available online.)
the neutron flux and the secondary positron production corresponded to the carbon beam, due to the high energy of carbon particles. The low-energy neutrons with the high intensity in the flux curves may have further contribution in penetrating the nucleus and treatment problems. In the photon flux curves produced from primary particles interaction with phantom, some of the peaks correspond to excited nuclei like 40Ca, 16O, 31P, 14N, and others, as shown in Fig. 9. Figure 10 presents 2D deposited energy contours per particle for the primary beams. We can see that with the heavier ion beams, the longitudinal extent of deposited energy at the proton Bragg peak location also increases compared with the 2 other beams. In Fig. 11, the lateral distribution of deposited energy at the Bragg peak location is shown for the 3 different beams. We can find out that alpha and carbon beams are scattered less than the proton beam is. Therefore, the greatest lateral scattering and deposited energy of the beams are for the proton and the lowest are for carbon. It seems that in this case, by increasing the lateral diameter of eye tumors, using a proton beam owing to the further lateral scattering and less production of secondary particles has advantages.
Fig. 9. Gamma flux curves in the phantom region for (A) proton, (B) alpha, and (C) carbon beam. (Color version of figure is available online.)
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
7
Fig. 12. The lateral distribution of deposited energy for the proton beam at various depths. (Color version of figure is available online.)
and as shown, the maximum deposited energy and lateral scattering of incident protons are located at the position of its Bragg peak.
Conclusion
Fig. 10. Monte Carlo calculated contours of deposited energy for primary particles: (A) proton, (B) alpha, and (C) carbon. (Color version of figure is available online.)
The lateral distribution of deposited energy plotted for the proton beam at various depths of the phantom is shown in Fig. 12. It is seen that the lateral broadening or the lateral scattering increases with the penetration depth of proton beam. As expected
We have designed and developed a new human eye phantom with all details for the GEANT4 toolkit. By using the phantom, the dose and deposited energy for the primary and secondary particles in ion therapy for deep uveal melanoma have been calculated. For the clinical investigation, passively scattered SOBPs of incident beams covering the tumor volume have been compared. Also, the secondary particles flux and the lateral distribution of deposited dose for the primary proton, alpha, and carbon beams were obtained and compared. The results of this research show that among proton, alpha, and carbon primary beams with the same range in the phantom region (1.9 cm), the heavy carbon ions have the highest dose percentage in the tumor region without considering the risk of secondary particles and the tail dose leakage. Protons can be a good choice for the large-size tumors, to have a good conformation with tumor shape owing to the beam broadening caused by scattering.
Acknowledgment The authors of this work are indebted to M. Severgnini from the Department of Medical Physics, A.O.U. “Ospedali Riuniti” di Trieste, for valuable remarks. References
Fig. 11. A comparison among lateral distributions of deposited energy by proton, alpha, and carbon ion beams having the same range in the eye phantom. (Color version of figure is available online.)
1. Ilic,́ R.D.; Lalić, D.; Stanković, S.J. Srna - Monte Carlo codes for proton transport simulation in combined and voxelized geometries. Nucl. Technol. Radiat. Prot. 17:27–36; 2002. 2. Scholz, M. Heavy ion tumour therapy. Nucl. Instrum. Methods Phys. Res. B 161:76–82; 2000. 3. Amaldi, U; Kraft, G. Radiotherapy with beams of carbon ions. Rep. Prog. Phys. 68:1861–82; 2005. 4. Zhang, H.; Davidorf, F.; Qi, Y. Comparison of 16 mm OSU-Nag and COMS eye plaques. J. Appl. Clin. Med. Phys. 13(3):166–78; 2012. 5. Shields, J.A.; Augsburger, J.J.; Brady, L.W.; et al. Cobalt plaque therapy of posterior uveal melanomas. Ophthalmology 89:1201–7; 1982. 6. Packer, S.; Rotman, M. Radiotherapy of choroidal melanoma with iodine-125. Ophthalmology. 87:582–90; 1980. 7. Egger, E.; Zografos, L.; Munkel, G.; et al. Results of proton radiotherapy for uveal melanomas. Front. Radiat. Ther. Oncol. 30:111–22; 1997.
8
S.A. Mahdipour, A.A. Mowlavi / Medical Dosimetry ] (2016) ]]]–]]]
8. Castro, J.R.; Char, D.H.; Petti, P.L.; et al. 15 years' experience with helium ion radiotherapy for uveal melanoma. Int. J. Radiat. Oncol. Biol. Phys. 39:989–96; 1997. 9. Desjardins, L.; Levy, C.; D'Hermies, F.; et al. Initial results of proton therapy in choroidal melanoma at the d'Orsey Center for Proton Therapy: The first 464 cases. Cancer Radiother. 1:222–6; 1997. 10. Fuss, M.; Loredo, L.N.; Blacharski, P.A.; et al. Proton radiation therapy for medium and large choroidal melanoma: Preservation of the eye and its functionality. Int. J. Radiat. Oncol. Biol. Phys. 49:1053–9; 2001. 11. Swakon, J.; Olko, P.; Adamczyk, D.; et al. Facility for proton radiotherapy of eye cancer at IFJ PAN in Krakow. Radiat. Meas. 45(10):1469–71; 2010. 12. Sobin, L.H.; Wittekind, Ch. TNM classification of malignant tumours. 6th ed., New York, NY: Wiley-Liss; 2002. 13. Char, D.H.; Castro, J.R.; Kroll, S.M.; et al. Five-year follow-up of helium ion therapy for uveal melonoma. Arch Ophthalmol. 108(2):209–14; 1990.
14. Agostinelli, S.; Allison, J.; Amako, K.; et al. Geant4—A simulation toolkit. Nucl. Instr. Methods. A506:250; 2003. 15. Allison, J.; Amako, K.; Apostolakis, H.; et al. Geant4 developments and applications. Nucl. Sci. 53:270; 2006. 16. Toshito, T.; Bagulya, A.; Lechner, A.; et al. Validation of New Geant4 Electromagnetic Physics Models for Ion Therapy Applications. Nucl. Sci. Technol. 2:918–22; 2011. 17. Atchison, D; Smith, G. Optics of the human eye. Edinburgh, 2003. 18. Christovao, M.T.; Ribeiro de Campos, T.P.; Trindade, B.M. Simulation and dosimetric analysis of proton and carbon ion therapy in the treatment of uveal melanoma. Radiol. Bras. 44(6):367–73; 2011. 19. Cirrone, G.A.P.; et al. Hadrontherapy: A Geant4-Based Tool for Proton/IonTherapy Studies. Prog. Nucl. Sci. Technol. 2:207–12; 2011. 20. Gorjiara, T.; Kuncic, Z.; Doran, S.; et al. Water and tissue equivalence of a new PRESAGE (R) formulation for 3D proton beam dosimetry: A Monte Carlo study. Med. Phys. 39(11):7071–9; 2012.