|Year : 2017 | Volume
| Issue : 6 | Page : 974-980
Evaluation of the effect of soft tissue composition on the characteristics of spread-out Bragg peak in proton therapy
Mahdi Ghorbani1, Sayyed Bijan Jia2, Mohsen Khosroabadi3, Hamid Reza Sadoughi3, Courtney Knaup4
1 Medical Physics Research Center, Mashhad University of Medical Sciences, Mashhad, Iran
2 Department of Physics, University of Bojnord, Bojnord, Iran
3 Faculty of Medicine, North Khorasan University of Medical Sciences, Bojnurd, Iran
4 Comprehensive Cancer Centers of Nevada, Las Vegas, Nevada, USA
|Date of Web Publication||13-Dec-2017|
Dr. Sayyed Bijan Jia
Department of Physics, University of Bojnord, Bojnord
Source of Support: None, Conflict of Interest: None
Aim: The aim of this study is to evaluate the effect of soft tissue composition on dose distribution and spread-out Bragg peak (SOBP) characteristics in proton therapy.
Subjects and Methods: Proton beams with nominal energies of 70, 120 and 210 MeV were considered. The soft tissues and tissue equivalent materials implemented in this study are: 9-component soft tissue, 4-component soft tissue, adipose tissue, muscle (skeletal), lung tissue, breast tissue, A-150 tissue equivalent plastic, perspex and water. Each material was separately defined inside a 20 cm × 20 cm × 40 cm phantom. A multilayer phantom was evaluated as well. The effect of tissue composition on the relative dose in SOBP region (relative to the dose in SOBP region in water), range of SOBP, length of SOBP, and uniformity index of SOBP was evaluated.
Results: Various soft tissues and tissue equivalent materials have shown different dose level in SOBPs, ranges of SOBPs, lengths of SOBPs and uniformity indices.
Conclusions: Based on the obtained results, various soft tissues and tissue equivalent materials have quite different SOBP characteristics. Since in clinical practice with proton therapy, only the range of SOBP is corrected for various tissues, omission of the above effects may result in major discrepancies in proton beam radiotherapy. To improve treatment accuracy, it is necessary to introduce such effects in treatment planning in proton therapy.
Keywords: Dose distribution, proton therapy, soft tissue, spread-out Bragg peak, tissue-equivalent material
|How to cite this article:|
Ghorbani M, Jia SB, Khosroabadi M, Sadoughi HR, Knaup C. Evaluation of the effect of soft tissue composition on the characteristics of spread-out Bragg peak in proton therapy. J Can Res Ther 2017;13:974-80
|How to cite this URL:|
Ghorbani M, Jia SB, Khosroabadi M, Sadoughi HR, Knaup C. Evaluation of the effect of soft tissue composition on the characteristics of spread-out Bragg peak in proton therapy. J Can Res Ther [serial online] 2017 [cited 2020 Sep 19];13:974-80. Available from: http://www.cancerjournal.net/text.asp?2017/13/6/974/220420
| > Introduction|| |
X-rays have been used for the treatment of cancer since 1895. Linear accelerators have been developed as a source of high energy photons for deeper tumors. Other advances in radiotherapy with X-rays are development of blocking techniques to conform the beam to the shape of the tumor, application of multiple beams and angles to adapt the dose distribution to the tumor and also to spare healthy organs, advances in imaging and the improved tumor contouring, image-guided radiotherapy, adaptive radiotherapy techniques, more accurate dose calculations with faster and more powerful computers, etc.
In 1946, Robert Wilson suggested that a proton beam may be suitable for cancer radiotherapy due to its advantageous dose distribution. When a proton beam interacts with matter, it deposits energy with a relatively low dose in the shallow regions in the beam's path. Near the end of the proton range, the dose has a sharp peak and then rapid fall-off to zero. Therefore, a high radiation dose is delivered to a deep tumor, while sparing the shallow normal tissues, and no dose to distal normal tissues. With the use of range modulation wheels or by changing the beam energies extracted from accelerators in a proper manner, it has been possible to spread out Bragg peaks (SOBPs). With SOBPs, it is feasible to cover larger targets that extend in depth.
Proton therapy offers clinical advantages compared to conventional radiotherapy with X-rays or electrons for many cancer patients. These advantages are mainly due to a more favorable dose distribution with proton beams. The risk of normal tissue damage is decreased. Therefore, in this way, it is possible to scale the dose and to increase the probability for cure. Considering the treatment effect in cancer patients, in some cases, proton therapy is superior to conventional radiotherapy, resulting from its higher curative dose with the same side effects or the same curative dose with lower side effects, in some patients compared to conventional radiotherapy.
In particle therapy, total mass stopping power ((S/ρ)tot) is the sum of mass collision stopping power ((S/ρ)col) and mass radiation stopping power ((S/ρ)rad). Mass collision stopping power is the result of interactions of the particle beam with orbital electrons. Dose in a medium is the multiplication of fluence and mass collision stopping power. According to the report No. 37 of International Commission on Radiation Units and Measurements (ICRU), mass collision stopping power is calculated from the following formula:
It is evident from the above equation that mass collision stopping power (therefore absorbed dose) depends on various parameters of the particle and medium. In other words, the atomic number and atomic weight of the material effect the dose distribution in particle beam therapy. Various soft tissues have different mass densities, compositions, effective atomic numbers, and effective atomic weights. This will result in differences in dose distributions inside different tissues. In addition, in the commissioning step of radiotherapy, dosimetry is normally performed in a water phantom. Then, this dosimetry data are introduced into the treatment planning system to calculate dose distributions inside the body. Quantification of the effect of tissue composition on dose distribution will be of interest in radiotherapy, since it estimates the discrepancies in dose delivery due to taking the same composition for all soft tissues and tissue-equivalent materials in dosimetry and treatment planning calculations.
In a Monte Carlo study by White et al., the dosimetric effect of trace elements in normal and cancerous soft tissues was evaluated for low-energy photons emitting from brachytherapy sources. Trace elements (with atomic numbers ranging from 11 to 30) were studied in soft tissue. Prostate, adipose, and mammary gland tissues were also simulated to quantify the contribution of each trace element to dose distribution. The results indicated that the presence of trace elements in a soft tissue causes differences in the dose distribution, which depends on the atomic number and the concentration of the trace element. There was also a significant difference in dose distribution between cancerous and healthy prostate tissues. Therefore, trace elements have a non-negligible effect on the tissue dose in brachytherapy with low-energy photon sources. It was concluded that trace elements introduce uncertainties in dose calculations. Ghorbani et al. compared the dose in various soft tissues with 103 Pd,125 I,169 Yb, and 192 Ir brachytherapy sources. Based on Monte Carlo simulations of a spherical phantom with 50 cm radius, they calculated absolute dose and relative dose with respect to the dose in nine-component soft tissue for various materials, sources, and radial distances. Some of the studied materials showed major differences with 103 Pd,125 I, and 169 Yb sources. Furthermore, greater distances from the source had higher dose differences. It was concluded that the ignorance of the density and composition of various soft tissues by treatment planning systems introduces significant discrepancy in brachytherapy with photon-emitting sources. The discrepancy depends on various parameters including type of soft tissue, source, and the distance from the source.
In a study by Khosroabadi et al., the effect of the compositions of various soft tissues and tissue-equivalent materials on dose distribution in neutron brachytherapy/neutron capture therapy was assessed. A low dose rate 252 Cf source was used as the neutron source, put inside a spherical phantom. Total and neutron dose rates were calculated for the materials. There were observed differences in dose rate with respect to dose rate in nine-component soft tissue, which varied with compositions of the materials and the radial distance from the source. It was concluded that taking the same composition for various media can lead to discrepancy in treatment planning in neutron brachytherapy/neutron capture therapy, and these discrepancies should be considered in dose calculation and treatment planning in neutron brachytherapy/neutron capture therapy. In a study by Ghorbani et al., the same effect was evaluated in electron beam radiotherapy. In a cylindrical phantom, various types of soft tissues and tissue-equivalent materials were investigated separately. Relative electron dose to the dose in nine-component soft tissue was determined for 8, 12, and 14 MeV electron beams. Negligible differences were observed between dose distributions in various soft tissues and tissue-equivalent materials, which were related to the uncertainties in Monte Carlo calculations.
Commercial treatment planning systems use approximations in their dose calculations. These approximations are applied to simplify and to speed up their calculations. The discrepancies due to these simplifications compared to a reference method for dose calculation have been the subject of a number of studies in proton therapy.,, These discrepancies are more significant in the cases of inhomogeneities inside the body or phantom. In a study by Ciangaru et al., a proton analytical computational algorithm was introduced. The algorithm was based on the superposition of narrow proton beamlets. The accuracy of the computational algorithm was verified through comparison with experimental data and Monte Carlo simulation. For this purpose, parallel wide proton beams were delivered in water phantoms in the presence of air and bone materials as inhomogeneities. The results have shown that the presented algorithm was useful, reasonably fast, and accurate for quality assurance in treatment planning in the presence of homogeneous materials or containing laterally extended inhomogeneities positioned in depths far from the Bragg peak.
In a study by Moskvin et al., a simple semi-empirical model was developed to estimate the range shift in the presence of high atomic number inhomogeneity in the pathway of proton beam. This model was based on the logarithmic function of stopping power versus atomic number of the inhomogeneity. The range shift due to metallic plates with various thicknesses was measured with a parallel plate ionization chamber and was calculated with FLUKA Monte Carlo code. The results were compared with those calculated by the semi-empirical model. The results have indicated that the proposed model requires the knowledge of the effective atomic number and the thickness of the inhomogeneity. The model can be applied in clinical proton therapy for estimation of proton range shift and for quality assurance in the treatment planning. In a study by Szymanowski and Oelfke, a detailed derivation was developed which is based on an additional scaling of the lateral proton fluence. The new two-dimensional scaling method was evaluated for different phantom materials including bone, muscle, fat, and air. A detailed comparison was performed between the calculations by the new pencil beam scaling method, standard pencil beam approach, and Monte Carlo simulations by GEANT code. It was shown that the new method was capable of calculation of dose with accuracy almost equal to that by Monte Carlo method, while it required increased calculation time compared to the standard pencil beam method.
As it was reviewed herein, there are various studies presenting the effect of soft tissue composition on dose distribution in brachytherapy with photon and neutron sources, neutron capture therapy, and electron beam radiotherapy. In addition, there are also studies on the effect of air and bone inhomogeneities in proton therapy. However, to the best of our knowledge, there is no study on the effect of tissue composition on dose distribution in proton therapy. The aim of the present study is to evaluate the effect of soft tissue and tissue-equivalent materials on dose distribution and characteristics of SOBP in proton therapy in various proton beam energies.
| > Subjects and Methods|| |
Constitution of spread-out Bragg peaks
MCNPX (version 2.6.0) was used for the simulations which is a modern general- purpose Monte Carlo code developed at Los Alamos National Laboratory in USA. A 20 cm × 20 cm × 40 cm water phantom was simulated including a tumor with 2 cm and 2 cm width and length in the directions perpendicular to the beam's direction, respectively. The proton beam was a 2 cm × 2 cm planar proton beam at 5 cm above the phantom surface [Figure 1]. Designed tumor depths ranged from 3.5 to 4.5 cm, 9.5 to 11.5 cm, and 25 to 29 cm for the 70 MeV, 120 MeV, and 210 MeV nominal proton energies, respectively. For each nominal proton energy to constitute the corresponding SOBP, a number of monoenergetic proton beams were modeled and their dose results in water phantom were summed up with according weighting factors in a manner to obtain the SOBPs conform with the tumor dimension. The weight of each pristine peak was evaluated from a mathematical algorithm that solves a series of coupled linear equations. These beams are considered to have the lateral dimension such as they conform to the target volume as it is in an active or in a passive scattering delivery system with completely different method. The former uses scanning method to move pencil beams across the target volume and the latter uses some physical components to spread the beam laterally. For nominal 70 MeV beam energy, the number of monoenergetic beams was 17 beams with energies ranging from 63.356 to 73.2472 MeV; for 120 MeV beam energy, 14 beams were required with energies ranging from 112.7602 to 125.7607 MeV; and for 210 MeV beam energy, 12 beams with energies ranging from 197.1357 to 215.6029 MeV were required. In these simulations, protons, neutrons, photons and electrons were transported. The energy cutoff for these particles was 10 keV, except for neutrons for which this value was set to zero. The type 3 mesh tally was used to score the proton dose in the phantom. The dimension of the mesh was 20 cm × 20 cm × 40 cm. This mesh included 20 cm × 20 cm × 0.1 mm voxels. The input Monte Carlo programs were run for 108 particles, and the average type A uncertainty in the Monte Carlo statistical calculations was 5.14%.
|Figure 1: A schematic geometry illustrating the proton beam, phantom, and tumor depth|
Click here to view
Evaluation the effect of soft tissues
So far, we have planned some uniform dose distributions with the proper modulation lengths in water. Now, we would like to investigate the use of these beam properties that were planned for water in some tissues with density near water. For evaluation of soft tissue and tissue-equivalent material, a 20 cm × 20 cm × 40 cm phantom was simulated using MCNPX Monte Carlo code. The soft tissue and tissue-equivalent materials that were evaluated are adipose tissue, four-component soft tissue, breast tissue, muscle (skeletal), lung tissue, nine-component soft tissue, water, A-150 tissue-equivalent plastic, and perspex. The mass densities and chemical compositions of these materials that are presented in [Table 1] were extracted from ICRU report number 44. The phantom included separately, each of the above materials.
|Table 1: Mass densities and compositions of the soft tissues and tissue-equivalent materials|
Click here to view
The other details of the simulation were the same as described above on the constitution of the SOBPs, with the difference that the energy spectra obtained for constitution of the SOBPs in nominal energies of 70 MeV, 120 MeV, and 210 MeV were used in this step. The beam was a 2 cm × 2 cm planar proton beam at 5 cm above the phantom surface [Figure 1]. The effect of soft tissue and tissue-equivalent material on dose distribution and characteristics of SOBP was evaluated. The characteristics include the effect on the dose in the SOBP for tissue relative to the dose in the SOBP for water, range of SOBP, the length of SOBP, and uniformity index of SOBP. The input Monte Carlo programs were run for 108 particles, and for all the nominal energies and materials, the maximum of type A uncertainty in the Monte Carlo statistical calculations was 16.38%. This uncertainty is the maximum value and the uncertainty in the Monte Carlo calculation in the SOBP region was considerably less than this level.
SOBP for a multilayer phantom was also obtained through Monte Carlo simulations and the results were compared with SOBP in a water phantom. The multilayer phantom was 20 cm × 20 cm × 40 cm, while in the beam's direction, it was composed of 2 mm thick skin, 4 mm thick adipose tissue, 3.4 mm thick muscle (skeletal), and nine-component soft tissue (the remaining thickness). The characteristics for the multilayer phantom were evaluated for 70 MeV, 120 MeV, and 210 MeV nominal proton energies. The other details of the simulations in this case were the same as described above for soft tissue and tissue-equivalent materials. The Monte Carlo input programs were run for 108 particles and the maximum type A uncertainty was 20.40%. It should be noticed that the uncertainty in the SOBP region was considerably less than this level.
| > Results|| |
As mentioned in the materials and methods section, we obtained the beam properties in a way to have the flat dose regions with the desired modulation ranges and penetration depths in the water phantom. Then, these beam properties were used to construct flat dose regions in the tissues with density near water. The SOBPs in various soft tissues and tissue-equivalent materials had their middle correspond to about 70, 120, and 210 MeV proton beam energies and are illustrated in [Figure 2]. The dose value in each case was normalized to the average dose in the SOBP of water with the same incident beam property.
|Figure 2: Spread-out Bragg peak in various soft tissues and tissue-equivalent materials for 70, 120, and 210 proton beams. The dose value in each energy level was normalized to the average dose in water on spread-out Bragg peak in that energy level|
Click here to view
Dose values in SOBPs relative to water, range of SOBPs, length of SOBPs, and uniformity index of SOBPs in various soft tissues and tissue-equivalent materials for the three mentioned cases are shown in [Table 2].
|Table 2: Relative dose in spread-out Bragg peak relative to water, range of spread-out Bragg peak, length of spread-out Bragg peak, and uniformity index of spread-out Bragg peak in various soft tissues and tissue-equivalent materials in 70, 120, and 210 proton beams|
Click here to view
Range of SOBP divided by mass density for various soft tissues and tissue-equivalent materials is also listed in [Table 3]. The depth dose profiles in the multilayer phantom for the three different modulation ranges and penetration depths are plotted in [Figure 3].
|Table 3: Range of spread-out Bragg peak/ρ for various soft tissues and tissue-equivalent materials|
Click here to view
|Figure 3: Spread-out Bragg peak in multilayer phantom for 70, 120, and 210 proton beams. The dose value in each energy level was normalized to the average dose in water on spread-out Bragg peak in that energy level|
Click here to view
| > Discussion|| |
In the present study, the effect of composition of soft tissue on SOBP characteristics of various proton beam energies was evaluated. The data presented in [Table 2] indicate that while the SOBPs were designed primarily based on a water phantom, the ranges of SOBPs for various soft tissues and tissue-equivalent materials are different. As another effect, the doses of these materials relative to the dose in water phantom indicate that there are also differences in doses in different materials. The amount of differences is as high as 5% for the adipose tissue. The differences are related to the mass densities and compositions of the materials. Ignoring these differences can result in dose calculation discrepancies. In some proton therapy centers that treat tissue with density near water such as ocular melanoma, a flat SOBP in water is used. In such a passive delivery beam line, the range modulators for creating SOBPs are designed using experimental or simulated depth dose profiles in a water phantom. In the cases of presence of a tissue with composition different than water (for example eye), the range of SOBP is corrected by taking account of the density of that tissue. However, the existence of tissues may affect the flatness of the SOBP. It should be noted that since there is a limitation in the uncertainties in radiotherapy, these discrepancies can affect the outcomes of the proton therapy. The results in the multilayer phantom also approve such points. It should be noticed that in a real situation in the human body, the geometry is different from that used herein in the case of a homogeneous soft tissue phantom and the multilayer phantom has a closer geometry to the real situation.
The lengths of SOBPs of various soft tissues and tissue-equivalent materials are different [Table 2]. The differences in the lengths are minor in the 70 MeV and 120 MeV energies, but it amounts up to about 6 mm for the 210 MeV proton beam. Since in proton beam therapy the SOBP is designed based on the target volume in a water phantom, ignoring such differences results in dose delivery discrepancies to the target volume in proton therapy.
It can be understood from [Table 2] that the uniformity index depends on the composition of the soft tissue or tissue-equivalent material. The uniformity index shows the amount of absorbed dose uniformity inside the tumor. Therefore, taking the uniformity index measured in water phantom for all the other materials results to have a nonuniformity of dose distribution inside the tumor. The level of this nonuniformity depends on the soft tissue.
A comparison of the data presented in [Table 2] implies that there are similarities and differences in SOBP characteristics in various energies. As an example, while the relative dose values are the same, there are differences in the ranges of SOBPs, lengths of SOBPs, and minor differences in the uniformity indexes in various energies. In addition, the differences in the latter vary with energy for different materials. This implies that to correct these effects, the energy of the proton beam or in the other words, the depth of tumor should be considered.
In the clinical practice of proton therapy, a simple correction which is applied is to account for the difference of the composition of an organ relative to water and to correct for the density of that organ. By comparing the ratio of the SOBP ranges to mass density in [Table 3], one can understand that this ratio is not equal to the range of SOBP in water in a specific energy. This indicates that a simple correction based on the mass density of a soft tissue does not have enough accuracy in treatment planning calculation for proton therapy. In general, accounting for the effect of soft tissue composition on dosimetric parameters and characteristics of SOBP is not simple. Current treatment planning in proton therapy cannot fully consider the influence of tissue composition to dose distribution. While a methodology for corrections of discrepancies mentioned above are not presented herein, the results of the present study show the existence of such basic uncertainties in proton therapy dosimetric calculations. With Monte Carlo-based treatment planning systems, it would be possible to introduce the mass densities and compositions of various soft tissues into the treatment planning system and to calculate the dose distribution more accurately than the conventional algorithms. However, Monte Carlo treatment planning systems require high processing speed computers, parallel processing capability, and the related high costs.
As it was reviewed in the introduction section, there are various studies on the effect of soft tissue on dose distribution in different radiotherapy modalities including brachytherapy with photon sources, neutron brachytherapy, neutron capture therapy, and electron beam radiotherapy.,,, Although the level of dose differences in various soft tissues in these studies and the current studies are not the same in these modalities, they indicate that there is a basic unsolved issue in radiotherapy that the soft tissue composition effects the dose distribution and this should be taken into account with further improvement in the technology of treatment planning systems and computer processing power.
| > Conclusions|| |
Based on the obtained results, various soft tissues and tissue-equivalent materials have quite different SOBP characteristics. These differences depend on the characteristics of proton beams and tissue compositions. It is seen that designing beam characteristics in a way as to have a perfect SOBP region in water can cause inaccuracy in dose calculation even for tissues with mass density similar to water. Since in clinical practice with proton therapy, only the range of SOBP is corrected for various tissues by considering the mass density of the tissue, and also not considering the exact tissue composition in the treatment planning systems, these introduce major discrepancies in proton beam radiotherapy. To enhance the treatment accuracy, it is recommended that the analytical algorithms that are implemented in commercial treatment planning systems should be modified and upgraded to include more parameters except electron density and the aforementioned effects have to be introduced to have more accurate dose calculation in proton therapy.
Financial support and sponsorship
Mashhad University of Medical Sciences, Mashhad, Razavi Khorasan, Iran.
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Foote RL, Stafford SL, Petersen IA, Pulido JS, Clarke MJ, Schild SE, et al
. The clinical case for proton beam therapy. Radiat Oncol 2012;7:174.
Smith AR. Proton therapy. Phys Med Biol 2006;51:R491-504.
Lundkvist J, Ekman M, Ericsson SR, Jönsson B, Glimelius B. Proton therapy of cancer: Potential clinical advantages and cost-effectiveness. Acta Oncol 2005;44:850-61.
ICRU Report No. 37. Stopping powers for electrons and positrons. Bethesda, Maryland, USA: International Commission on Radiation Units and Measurements (ICRU); 1984.
White SA, Landry G, van Gils F, Verhaegen F, Reniers B. Influence of trace elements in human tissue in low-energy photon brachytherapy dosimetry. Phys Med Biol 2012;57:3585-96.
Ghorbani M, Salahshour F, Haghparast A, Moghaddas TA, Knaup C. Effect of tissue composition on dose distribution in brachytherapy with various photon emitting sources. J Contemp Brachytherapy 2014;6:54-67.
Khosroabadi M, Farhood B, Ghorbani M, Hamzian N, Moghaddam HR, Davenport D. Tissue composition effect on dose distribution in neutron brachytherapy/neutron capture therapy. Rep Pract Oncol Radiother 2016;21:8-16.
Ghorbani M, Tabatabaei ZS, Vejdani Noghreiyan A, Vosoughi H, Knaup C. Effect of tissue composition on dose distribution in electron beam radiotherapy. J Biomed Phys Eng 2015;5:15-24.
Ciangaru G, Polf JC, Bues M, Smith AR. Benchmarking analytical calculations of proton doses in heterogeneous matter. Med Phys 2005;32:3511-23.
Moskvin V, Cheng CW, Fanelli L, Zhao L, Das IJ. A semi-empirical model for the therapeutic range shift estimation caused by inhomogeneities in proton beam therapy. J Appl Clin Med Phys 2012;13:3631.
Szymanowski H, Oelfke U. Two-dimensional pencil beam scaling: An improved proton dose algorithm for heterogeneous media. Phys Med Biol 2002;47:3313-30.
Pelowitz D. MCNPX users manual, LA-CP-07-1473. Ver. 2.6.0. USA: Los Alamos National Laboratory; 2008.
ICRU. ICRU Report No. 44, Tissue Substitutes in Radiation Dosimetry and Measurement. Bethesda, MD: ICRU; 1989.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3]