Physica Medica xxx (2015) 1e6
Contents lists available at ScienceDirect
Physica Medica journal homepage: http://www.physicamedica.com
Technical notes
A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223 Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals Ernesto Amato a, *, Salvatore Leotta b, Antonio Italiano c, Sergio Baldari a a b c
Department of Biomedical Sciences and of Morphologic and Functional Imaging, University of Messina, Messina, Italy Department of Physics and Earth Sciences, University of Messina, Messina, Italy Istituto Nazionale di Fisica Nucleare, Gr. Coll. Messina, Sez. Catania, Messina, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 5 February 2015 Received in revised form 20 April 2015 Accepted 25 April 2015 Available online xxx
The small-scale dosimetry of radionuclides in solid-tumours is directly related to the intra-tumoral distribution of the administered radiopharmaceutical, which is affected by its egress from the vasculature and dispersion within the tumour. The aim of the present study was to evaluate the combined dosimetric effects of radiopharmaceutical distribution and range of the emitted radiation in a model of tumour microvasculature. We developed a computational model of solid-tumour microenvironment around a blood capillary vessel, and we simulated the transport of radiation emitted by 223Ra, 111In, 131I and 177Lu using the GEANT4 Monte Carlo. For each nuclide, several models of radiopharmaceutical dispersion throughout the capillary vessel were considered. Radial dose profiles around the capillary vessel, the Initial Radioactivity (IR) necessary to deposit 100 Gy of dose at the edge of the viable tumour-cell region, the Endothelial Cell Mean Dose (ECMD) and the Tumour Edge Mean Dose (TEMD), i.e. the mean dose imparted at the 250-mm layer of tissue, were computed. The results for beta and Auger emitters demonstrate that the photon dose is about three to four orders of magnitude lower than that deposited by electrons. For 223Ra, the beta emissions of its progeny deliver a dose about three orders of magnitude lower than that delivered by the alpha emissions. Such results may help to characterize the dose inhomogeneities in solid tumour therapies with radiopharmaceuticals, taking into account the interplay between drug distribution from vasculature and range of ionizing radiations. © 2015 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Keywords: Small-scale dosimetry Monte Carlo Nuclear medicine Angiogenesis Microvasculature
Introduction In the framework of internal radionuclide therapies, small-scale dosimetry and microdosimetric approaches to the tumour structure are fundamental to foresee the therapeutic effects attainable. The biokinetics of therapeutic radiopharmaceuticals follow different pathways, depending upon the route of administration, the molecular targeting mechanism, and the physio-pathologic
* Corresponding author. University of Messina, Department of Biomedical Sciences and of Morphologic and Functional Imaging, Section of Radiological Sciences, Italy. Tel.: þ39 090 221 2942. E-mail address: [email protected] (E. Amato).
conditions of the patient, affecting the uptake in normal tissues and tumour. An administered radiopharmaceutical is distributed systemically by blood flow and microscopically by diffusion and/or other mechanisms of egress from capillaries, as affected by its molecular weight, charge or polarity (if any), and surface topography and by the distribution and accessibility of its molecular target. Any ensuing radiation damage is thus determined by the microscopic biodistribution of the radiopharmaceutical and by the range and ionization density or linear energy transfer (LET) of its emitted radiations. Since several radionuclides are of common use in nuclear medicine therapies, exploiting beta, Auger or alpha emissions, and a variety of molecules is available as carriers with different targeting pathways, a systematic study of the small-scale
http://dx.doi.org/10.1016/j.ejmp.2015.04.015 1120-1797/© 2015 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
2
E. Amato et al. / Physica Medica xxx (2015) 1e6
dosimetry of therapeutic radiopharmaceuticals in and around the tumour microvasculature is therefore necessary. Several approaches have been proposed in the past years, mainly dealing with alpha-particle therapies. In particular, a microdosimetric model of tumour microvasculature was introduced by Humm concerning radioimmunotherapies with 211At [1]. More recently, Huang et al. [2] presented a Monte Carlo study of the effectiveness of Tumour Anti-Vascular Alpha Therapies (TAVAT), in which the doses delivered to the endothelial cell (EC) elements have been evaluated. A more general approach to the small-scale dosimetry of tumour microenvironment and angiogenesis was developed by Zhu et al. [3], in which a model of capillaries in a pathologic tissue was assumed, and two alpha emitters (211At and 213Bi) and six beta emitters (32P, 33P, 67Cu, 90Y, 131I, 188Re) were investigated. The aim of the current study was to extend this model to other, frequently used therapeutic radionuclides, the alpha emitter 223Ra, the beta-emitters 131I and 177Lu, and the Auger electron-emitter 111 In. Our analysis, relying on a specifically developed GEANT4 [4] Monte Carlo code, allowed us to quantify the contributions to the radiation dose of each emitted radiation around the capillary vessel for several models of radionuclide distribution encountered in different therapies: radio-embolization procedures, anti-angiogenic radiopharmaceuticals, and other models involving extra-vascular distribution.
Materials and methods We developed a Monte Carlo simulation in GEANT4, release 10.0.p02, a simulation toolkit originally developed for high energy physics, and now widely applied also in the field of medical radiation physics [5,6]. In particular, the code developed for this study was derived from others previously validated for microdosimetric evaluations of dose enhancement in external beam radiotherapy [7] and for radionuclide dosimetry purposes [8]. The present simulation was aimed to evaluate the dose distribution around a 1-mm-long capillary embedded in soft tissue. In particular, we simulated a cylindrical vessel having 10-mm inner radius and 10-mm-thick wall, composed by the standard NIST material [4] “G4_SKIN_ICRP” (density 1.09 g cm3), filled by blood represented by the material “G4_BLOOD_ICRP” (density 1.06 g cm3), and surrounded by a set of 23 co-axial cylindrical layers made of “G4_TISSUE_SOFT_ICRP” (density 1.03 g cm3), 10mm-thick each. In view of the maximum distribution range of
Table 1 Nuclear properties of the simulated radionuclides. Only the main emissions are listed; the full decay data are available in Ref. [10]. The column Ee-mono lists the energies of the main monoenergetic (Auger or conversion) electrons. Radionuclide
T1/2
Ea (MeV)
(keV)
Ee-mono (keV)
Eg (keV)
223
Ra Rn Po 211 Pb 211 Bi
11.43 d 3.96 s 1.78E3 s 2.16Eþ3 s 3.63Eþ3 s
e e e 450 176 (0.3%) e 496 181.5
See Ref. [10]
Neglected
211
5.7 6.8 7.4 e 6.5 (99.7%) 7.4 e e
3.4 45.6 6.2 47.6 101.7 2.72 19.3
364
219 215
131
I
0.52 s 2.86Eþ2 s 8.02 d
177
Lu
6.71 d
e
132.9
111
In
2.80 d
e
e
207
Po Tl
113 208 171 245
oxygen and nutrients from a capillary vessel [9], we assumed a viable tumour edge located at 250 mm of radial distance from the capillary axis. In Fig. 1 a schematic view of the simulated regions is shown. We considered four radionuclides among those used in nuclear medicine therapies: the alpha emitter 223Ra together with its daughters, the Auger-gamma emitter 111In, and the two betagamma emitters 131I and 177Lu. The decay modes with respective energies, taken from Ref. [10], are summarized in Table 1. In the framework of the GEANT4 simulation code, we compared the three physics models for electromagnetic interactions, namely, the Standard e.m. physics model, and the ones based on Livermore data and Penelope parameterizations [4]. A range cut of 0.5 mm was adopted for all tracked particles, in order to obtain proper spatial accuracy in the energy deposition, consistent with the sizes of the simulated geometry. We tracked 107 primary events of 111In, 131I and 177Lu decays, while 5$105 223Ra decays were chosen, so that the statistical uncertainties associated with the presented results are below 1%, a value lower than the uncertainties on the parameterizations of the experimental cross-sections employed by GEANT4 in the energy range explored in our study [11,12]. Since, at state of the art, the extra-vascular distribution of each radiopharmaceutical is not known exactly, we assumed several configurations of radionuclide radial distribution across the various layers (blood, EC and tissue layers), as reported in Table 2. These assumptions are based on the different realities that may be encountered in the clinics. For example, configuration A (blood only) can represent a radio-embolization procedure; configuration B (EC only) can represent an anti-angiogenic radiopharmaceutical; configuration C results from a mixed effect of the previous configurations; configurations D-G represent other models of extravascular distributions. Energy deposition per primary event was scored for each layer and represented as a function of the radial distance from the
Table 2 Radial distribution configurations of the radionuclides used.
Figure 1. Geometrical lay-out of the simulation.
Configuration#
Regions
Range (mm)
A B C D E F G
Blood Endothelial cells Blood þ E.C. E.C. þ Tumour E.C. þ Tumour E.C. þ Tumour E.C. þ Tumour
0e10 10e20 0e20 10e50 10e100 10e150 10e200
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
E. Amato et al. / Physica Medica xxx (2015) 1e6
3
Figure 2. Radial dose profiles for sources located in blood (configuration A).
capillary axis. In particular, besides the total energy deposition, the energy deposited by each particle type was separately histogrammed. From these data, doses in Gy/event were deduced. In particular, radial dose profiles around the capillary vessel, the Initial Radioactivity (IR) necessary to deposit 100 Gy of dose at the edge of the viable tumour-cell region located at 250 mm of radial distance, the Endothelial Cell Mean Dose (ECMD) and the Tumour Edge Mean Dose (TEMD), i.e. the mean dose imparted at the 250-mm layer, were computed. Results and discussion Firstly, we compared the results relative to 223Ra or 131I in configuration A (source located in blood only) for the three physics models available in GEANT4. We obtained an agreement better than 1% in the whole volume, except for the end of alpha-particle tracks, in which deviations less than 5% were observed. Consequently, the Standard electromagnetic model was adopted for all simulations reported below. In Fig. 2 the peri-vessel radial dose distributions are reported for each nuclide in configuration A. Concerning 223Ra, the alphaparticle contribution dominates up to about 75 mm over the electron contribution, which in turn arises from the high energy b emissions from 211Pb and 207Tl, the multiple mono-energetic electron emissions from 223Ra and daughters and, finally from all secondary electrons emitted. The sharp drop in alpha dose at 75 mm is due to the modest entity of the range straggling of these particles,
which exhibit a stopping power versus distance plot characterized by a Bragg peak. In particular, the contribution from d-rays originating from alpha-particle tracks are prominent below 50 mm. For 111In, 131I and 177Lu, the photon contribution, mainly due to the interaction of primary X and g photons and, at the second order, due to the secondary bremsstrahlung photons, is three to four orders of magnitude lower than that of electrons. This circumstance justifies the choice to neglect primary photons in the emission spectra of 223Ra and daughters. It should be noticed the high self dose to the blood layer from 111In, due to the low energy Auger and conversion electrons. Figures 3e6 present the radial profiles of total radiation absorbed dose for each radionuclide and for all the peri-vessel radionuclide distributions. For 223Ra, in all configurations the alphaparticle dose exceeds by about three orders of magnitude the electron dose. Figure 3 shows that the alpha-particle range extends to about 70 mm beyond the end of the activity distribution. 131 I and 177Lu, whose dose profiles are presented in Figs. 4 and 5, respectively, exhibit a very similar behaviour, because of the similarity of the beta emission spectra. The quite different photon emission spectra give a negligible contribution at these scales. 111In in Fig. 6 is characterized by the expected high localization of dose in the activity distribution, due to the low energy Auger and conversion electrons. In Table 3 and Fig. 7, the ECMD and TEMD expressed in Gy for a linear specific cumulated activity (LSCA) of 1 MBqs/mm, are reported. We recall that the tumour edge is located at 250-mm of
Figure 3. Radial dose profiles for
223
Ra.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
4
E. Amato et al. / Physica Medica xxx (2015) 1e6
Figure 4. Radial dose profiles for
Figure 5. Radial dose profiles for
radial distance from the vessel axis, while LSCA results from the integration over time of activity (cumulated activity, i.e. total number of disintegrations) per unit length. In Fig. 7 (upper panel) it is apparent that the highest dose per unit cumulated activity is imparted by 223Ra and its daughters, while the two beta emitters 131I and 177Lu give comparable contributions. 111In has the lowest specific dose, except when endothelial cells are sources themselves; in this case, the ECMD increases due to the nearly complete absorption of monoenergetic electrons of 111In in endothelial cell. The lower panel shows the behaviour of TEMD when the activity distribution thickness increases. An increase of TEMD is observed for beta and Auger nuclide emitters, while conversely 223Ra TEMD
131
I.
177
Lu.
in configuration G is markedly higher due to the direct irradiation by primary alpha particles. The initial radioactivity (IR) that is needed to deposit a dose of 100 Gy (reference cytocidal dose) at the tumour edge (250-mm layer), was calculated as:
i h IR MBq$mm1 ¼ 100½Gy
ls1
TEMD Gy$MBq1 $s1 $mm
(1)
In this calculation, there is the implicit assumption of physical decay only of each radionuclide (i.e. no biological clearance). For each nuclide and every activity distribution configuration, IR is
Figure 6. Radial dose profiles for
111
In.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
E. Amato et al. / Physica Medica xxx (2015) 1e6
5
Table 3 Endothelial Cell Mean Dose (ECMD) and Tumour Edge Mean Dose (TEMD) [Gy MBq1 s1 mm]. 223
Configuration
A B C D E F G
131
Ra
177
I
111
Lu
In
ECMD
TEMD
ECMD
TEMD
ECMD
TEMD
ECMD
TEMD
9.16Eþ2 1.04Eþ3 1.01Eþ3 4.17Eþ2 1.18Eþ2 5.17Eþ1 2.90Eþ1
4.46E2 4.47E2 4.47E2 4.50E2 4.63E2 4.63E2 6.67E1
1.24Eþ0 1.51Eþ0 1.44Eþ0 5.65E1 2.41E1 1.43E1 9.70E2
1.76E2 1.76E2 1.76E2 1.79E2 1.87E2 2.03E2 2.34E2
1.53Eþ0 1.93Eþ0 1.83Eþ0 6.95E1 2.87E1 1.60E1 1.04E1
1.27E2 1.27E-2 1.27E2 1.30E2 1.38E2 1.58E2 2.03E2
2.38E1 9.56E1 7.78E1 1.85E1 6.82E2 4.03E2 2.76E2
3.19E3 3.23E3 3.22E3 3.48E3 4.10E3 4.69E3 5.30E3
Figure 7. . Plots of Endothelial Cell Mean Dose (ECMD) and Tumour Edge Mean Dose (TEMD) for all the radioactivity distributions considered.
reported in Table 4. As expected, one sees that the IR increases when going from alpha to beta and Auger emitters. IR for 223Ra in configuration G decreases by an order of magnitude with respect to the other configurations due to the direct irradiation of the tumour edge by alpha particles. Conclusions The development of small-scale models of the capillary vasculature of solid tumors is an important tool for understanding the effect of radionuclide therapies, due to both differences in biodistribution at small scales and range of the ionizing radiations. The Monte Carlo simulation developed in the present study allowed us to quantify dose distributions at the microscopic level around a simple model of tumour capillary vessel for some interesting therapeutic radionuclides. The differences between
Table 4 Initial Radioactivity [MBq mm1] needed to deposit 100 Gy of dose at the Tumour Edge. Configuration
223
A B C D E F G
1.57E3 1.57E3 1.57E3 1.56E3 1.51E3 1.51E3 1.05E4
Ra
131
I
5.69E3 5.69E3 5.69E3 5.60E3 5.36E3 4.93E3 4.27E3
177
Lu
9.50E3 9.50E3 9.49E3 9.27E3 8.73E3 7.62E3 5.95E3
111
In
8.97E2 8.86E2 8.89E2 8.22E2 6.98E2 6.10E2 5.40E2
irradiation properties of the alpha, beta and Auger emissions were pointed out and compared for the radionuclides considered. The results of this study contribute the characterization of dose inhomogeneities in solid tumour therapies with radiopharmaceuticals, taking into account the interplay between drug distribution from vasculature and range of the emitted particles. Future studies will be devoted to introduce the actual biodistribution of radiopharmaceuticals labelled with these nuclides in the model of solid tumour capillary vessels.
References [1] Humm JL. A microdosimetric model of astatine-211 labeled antibodies for radioimmunotherapy. Int J Radiat Oncol Biol Phys 1987;13:1767e73. [2] Huang C-Y, Oborn BM, Guatelli S, Allen BJ. Monte Carlo calculation of the maximum therapeutic gain of tumor antivascular alpha therapy. Med Phys 2012;39(3):1282e8. [3] Zhu X, Palmer MR, Makrigiorgos GM, Kassis AI. Solid-tumor radionuclide therapy dosimetry: new paradigms in view of tumor microenvironment and angiogenesis. Med Phys 2010;37(6):2974e84. [4] Agostinelli S, Allison J, Amako K, Apostolakis J, Araujo H, Arce P, et al. GEANT4 e a simulation toolkit. Nucl Instrum Methods A 2003;506:250e303. [5] Amato E, Italiano A, Baldari S. Monte Carlo study of voxel S factor dependence on tissue density and atomic composition. Nucl Instrum Methods Phys Res A 2013;729:870e6. [6] Amato E, Italiano A. An analytical model for calculating internal dose conversion coefficients (DCCs) for non-human biota. Radiat Environ Biophys 2014;53:455e9. [7] Amato E, Italiano A, Leotta S, Pergolizzi S, Torrisi L. Monte Carlo study of the dose enhancement effect of gold nanoparticles during X-ray therapies and evaluation of the anti-angiogenic effect on tumour capillary vessels. J X-ray Sci Technol 2013;21:237e47. [8] Amato E, Italiano A, Baldari S. Absorbed fractions for alpha particles in ellipsoidal volumes. Phys Med Biol 2013;58:5449e59.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
6
E. Amato et al. / Physica Medica xxx (2015) 1e6
[9] Jain RK. Transport of molecules, particles, and cells in solid-tumors. Annu Rev Biomed Eng 1999;1:241e63. [10] Stabin MG, da Luz LCQP. Decay data for internal and external dose assessment. Health Phys 2002;83:471e5.
[11] Amako K, Guatelli S, Ivanchencko V, Maire M, Mascialino B, Murakami K, et al. Geant4 and its validation. Nucl Phys B Proc Suppl 2003;150:44e9. [12] Carrier JF, Archambault L, Beaulieu L, Roy R. Validation of Geant4, an objectoriented Monte Carlo toolkit, for simulations in medical physics. Med Phys 2004;31:484e92.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
Contents lists available at ScienceDirect
Physica Medica journal homepage: http://www.physicamedica.com
Technical notes
A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223 Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals Ernesto Amato a, *, Salvatore Leotta b, Antonio Italiano c, Sergio Baldari a a b c
Department of Biomedical Sciences and of Morphologic and Functional Imaging, University of Messina, Messina, Italy Department of Physics and Earth Sciences, University of Messina, Messina, Italy Istituto Nazionale di Fisica Nucleare, Gr. Coll. Messina, Sez. Catania, Messina, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 5 February 2015 Received in revised form 20 April 2015 Accepted 25 April 2015 Available online xxx
The small-scale dosimetry of radionuclides in solid-tumours is directly related to the intra-tumoral distribution of the administered radiopharmaceutical, which is affected by its egress from the vasculature and dispersion within the tumour. The aim of the present study was to evaluate the combined dosimetric effects of radiopharmaceutical distribution and range of the emitted radiation in a model of tumour microvasculature. We developed a computational model of solid-tumour microenvironment around a blood capillary vessel, and we simulated the transport of radiation emitted by 223Ra, 111In, 131I and 177Lu using the GEANT4 Monte Carlo. For each nuclide, several models of radiopharmaceutical dispersion throughout the capillary vessel were considered. Radial dose profiles around the capillary vessel, the Initial Radioactivity (IR) necessary to deposit 100 Gy of dose at the edge of the viable tumour-cell region, the Endothelial Cell Mean Dose (ECMD) and the Tumour Edge Mean Dose (TEMD), i.e. the mean dose imparted at the 250-mm layer of tissue, were computed. The results for beta and Auger emitters demonstrate that the photon dose is about three to four orders of magnitude lower than that deposited by electrons. For 223Ra, the beta emissions of its progeny deliver a dose about three orders of magnitude lower than that delivered by the alpha emissions. Such results may help to characterize the dose inhomogeneities in solid tumour therapies with radiopharmaceuticals, taking into account the interplay between drug distribution from vasculature and range of ionizing radiations. © 2015 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Keywords: Small-scale dosimetry Monte Carlo Nuclear medicine Angiogenesis Microvasculature
Introduction In the framework of internal radionuclide therapies, small-scale dosimetry and microdosimetric approaches to the tumour structure are fundamental to foresee the therapeutic effects attainable. The biokinetics of therapeutic radiopharmaceuticals follow different pathways, depending upon the route of administration, the molecular targeting mechanism, and the physio-pathologic
* Corresponding author. University of Messina, Department of Biomedical Sciences and of Morphologic and Functional Imaging, Section of Radiological Sciences, Italy. Tel.: þ39 090 221 2942. E-mail address: [email protected] (E. Amato).
conditions of the patient, affecting the uptake in normal tissues and tumour. An administered radiopharmaceutical is distributed systemically by blood flow and microscopically by diffusion and/or other mechanisms of egress from capillaries, as affected by its molecular weight, charge or polarity (if any), and surface topography and by the distribution and accessibility of its molecular target. Any ensuing radiation damage is thus determined by the microscopic biodistribution of the radiopharmaceutical and by the range and ionization density or linear energy transfer (LET) of its emitted radiations. Since several radionuclides are of common use in nuclear medicine therapies, exploiting beta, Auger or alpha emissions, and a variety of molecules is available as carriers with different targeting pathways, a systematic study of the small-scale
http://dx.doi.org/10.1016/j.ejmp.2015.04.015 1120-1797/© 2015 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
2
E. Amato et al. / Physica Medica xxx (2015) 1e6
dosimetry of therapeutic radiopharmaceuticals in and around the tumour microvasculature is therefore necessary. Several approaches have been proposed in the past years, mainly dealing with alpha-particle therapies. In particular, a microdosimetric model of tumour microvasculature was introduced by Humm concerning radioimmunotherapies with 211At [1]. More recently, Huang et al. [2] presented a Monte Carlo study of the effectiveness of Tumour Anti-Vascular Alpha Therapies (TAVAT), in which the doses delivered to the endothelial cell (EC) elements have been evaluated. A more general approach to the small-scale dosimetry of tumour microenvironment and angiogenesis was developed by Zhu et al. [3], in which a model of capillaries in a pathologic tissue was assumed, and two alpha emitters (211At and 213Bi) and six beta emitters (32P, 33P, 67Cu, 90Y, 131I, 188Re) were investigated. The aim of the current study was to extend this model to other, frequently used therapeutic radionuclides, the alpha emitter 223Ra, the beta-emitters 131I and 177Lu, and the Auger electron-emitter 111 In. Our analysis, relying on a specifically developed GEANT4 [4] Monte Carlo code, allowed us to quantify the contributions to the radiation dose of each emitted radiation around the capillary vessel for several models of radionuclide distribution encountered in different therapies: radio-embolization procedures, anti-angiogenic radiopharmaceuticals, and other models involving extra-vascular distribution.
Materials and methods We developed a Monte Carlo simulation in GEANT4, release 10.0.p02, a simulation toolkit originally developed for high energy physics, and now widely applied also in the field of medical radiation physics [5,6]. In particular, the code developed for this study was derived from others previously validated for microdosimetric evaluations of dose enhancement in external beam radiotherapy [7] and for radionuclide dosimetry purposes [8]. The present simulation was aimed to evaluate the dose distribution around a 1-mm-long capillary embedded in soft tissue. In particular, we simulated a cylindrical vessel having 10-mm inner radius and 10-mm-thick wall, composed by the standard NIST material [4] “G4_SKIN_ICRP” (density 1.09 g cm3), filled by blood represented by the material “G4_BLOOD_ICRP” (density 1.06 g cm3), and surrounded by a set of 23 co-axial cylindrical layers made of “G4_TISSUE_SOFT_ICRP” (density 1.03 g cm3), 10mm-thick each. In view of the maximum distribution range of
Table 1 Nuclear properties of the simulated radionuclides. Only the main emissions are listed; the full decay data are available in Ref. [10]. The column Ee-mono lists the energies of the main monoenergetic (Auger or conversion) electrons. Radionuclide
T1/2
Ea (MeV)
(keV)
Ee-mono (keV)
Eg (keV)
223
Ra Rn Po 211 Pb 211 Bi
11.43 d 3.96 s 1.78E3 s 2.16Eþ3 s 3.63Eþ3 s
e e e 450 176 (0.3%) e 496 181.5
See Ref. [10]
Neglected
211
5.7 6.8 7.4 e 6.5 (99.7%) 7.4 e e
3.4 45.6 6.2 47.6 101.7 2.72 19.3
364
219 215
131
I
0.52 s 2.86Eþ2 s 8.02 d
177
Lu
6.71 d
e
132.9
111
In
2.80 d
e
e
207
Po Tl
113 208 171 245
oxygen and nutrients from a capillary vessel [9], we assumed a viable tumour edge located at 250 mm of radial distance from the capillary axis. In Fig. 1 a schematic view of the simulated regions is shown. We considered four radionuclides among those used in nuclear medicine therapies: the alpha emitter 223Ra together with its daughters, the Auger-gamma emitter 111In, and the two betagamma emitters 131I and 177Lu. The decay modes with respective energies, taken from Ref. [10], are summarized in Table 1. In the framework of the GEANT4 simulation code, we compared the three physics models for electromagnetic interactions, namely, the Standard e.m. physics model, and the ones based on Livermore data and Penelope parameterizations [4]. A range cut of 0.5 mm was adopted for all tracked particles, in order to obtain proper spatial accuracy in the energy deposition, consistent with the sizes of the simulated geometry. We tracked 107 primary events of 111In, 131I and 177Lu decays, while 5$105 223Ra decays were chosen, so that the statistical uncertainties associated with the presented results are below 1%, a value lower than the uncertainties on the parameterizations of the experimental cross-sections employed by GEANT4 in the energy range explored in our study [11,12]. Since, at state of the art, the extra-vascular distribution of each radiopharmaceutical is not known exactly, we assumed several configurations of radionuclide radial distribution across the various layers (blood, EC and tissue layers), as reported in Table 2. These assumptions are based on the different realities that may be encountered in the clinics. For example, configuration A (blood only) can represent a radio-embolization procedure; configuration B (EC only) can represent an anti-angiogenic radiopharmaceutical; configuration C results from a mixed effect of the previous configurations; configurations D-G represent other models of extravascular distributions. Energy deposition per primary event was scored for each layer and represented as a function of the radial distance from the
Table 2 Radial distribution configurations of the radionuclides used.
Figure 1. Geometrical lay-out of the simulation.
Configuration#
Regions
Range (mm)
A B C D E F G
Blood Endothelial cells Blood þ E.C. E.C. þ Tumour E.C. þ Tumour E.C. þ Tumour E.C. þ Tumour
0e10 10e20 0e20 10e50 10e100 10e150 10e200
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
E. Amato et al. / Physica Medica xxx (2015) 1e6
3
Figure 2. Radial dose profiles for sources located in blood (configuration A).
capillary axis. In particular, besides the total energy deposition, the energy deposited by each particle type was separately histogrammed. From these data, doses in Gy/event were deduced. In particular, radial dose profiles around the capillary vessel, the Initial Radioactivity (IR) necessary to deposit 100 Gy of dose at the edge of the viable tumour-cell region located at 250 mm of radial distance, the Endothelial Cell Mean Dose (ECMD) and the Tumour Edge Mean Dose (TEMD), i.e. the mean dose imparted at the 250-mm layer, were computed. Results and discussion Firstly, we compared the results relative to 223Ra or 131I in configuration A (source located in blood only) for the three physics models available in GEANT4. We obtained an agreement better than 1% in the whole volume, except for the end of alpha-particle tracks, in which deviations less than 5% were observed. Consequently, the Standard electromagnetic model was adopted for all simulations reported below. In Fig. 2 the peri-vessel radial dose distributions are reported for each nuclide in configuration A. Concerning 223Ra, the alphaparticle contribution dominates up to about 75 mm over the electron contribution, which in turn arises from the high energy b emissions from 211Pb and 207Tl, the multiple mono-energetic electron emissions from 223Ra and daughters and, finally from all secondary electrons emitted. The sharp drop in alpha dose at 75 mm is due to the modest entity of the range straggling of these particles,
which exhibit a stopping power versus distance plot characterized by a Bragg peak. In particular, the contribution from d-rays originating from alpha-particle tracks are prominent below 50 mm. For 111In, 131I and 177Lu, the photon contribution, mainly due to the interaction of primary X and g photons and, at the second order, due to the secondary bremsstrahlung photons, is three to four orders of magnitude lower than that of electrons. This circumstance justifies the choice to neglect primary photons in the emission spectra of 223Ra and daughters. It should be noticed the high self dose to the blood layer from 111In, due to the low energy Auger and conversion electrons. Figures 3e6 present the radial profiles of total radiation absorbed dose for each radionuclide and for all the peri-vessel radionuclide distributions. For 223Ra, in all configurations the alphaparticle dose exceeds by about three orders of magnitude the electron dose. Figure 3 shows that the alpha-particle range extends to about 70 mm beyond the end of the activity distribution. 131 I and 177Lu, whose dose profiles are presented in Figs. 4 and 5, respectively, exhibit a very similar behaviour, because of the similarity of the beta emission spectra. The quite different photon emission spectra give a negligible contribution at these scales. 111In in Fig. 6 is characterized by the expected high localization of dose in the activity distribution, due to the low energy Auger and conversion electrons. In Table 3 and Fig. 7, the ECMD and TEMD expressed in Gy for a linear specific cumulated activity (LSCA) of 1 MBqs/mm, are reported. We recall that the tumour edge is located at 250-mm of
Figure 3. Radial dose profiles for
223
Ra.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
4
E. Amato et al. / Physica Medica xxx (2015) 1e6
Figure 4. Radial dose profiles for
Figure 5. Radial dose profiles for
radial distance from the vessel axis, while LSCA results from the integration over time of activity (cumulated activity, i.e. total number of disintegrations) per unit length. In Fig. 7 (upper panel) it is apparent that the highest dose per unit cumulated activity is imparted by 223Ra and its daughters, while the two beta emitters 131I and 177Lu give comparable contributions. 111In has the lowest specific dose, except when endothelial cells are sources themselves; in this case, the ECMD increases due to the nearly complete absorption of monoenergetic electrons of 111In in endothelial cell. The lower panel shows the behaviour of TEMD when the activity distribution thickness increases. An increase of TEMD is observed for beta and Auger nuclide emitters, while conversely 223Ra TEMD
131
I.
177
Lu.
in configuration G is markedly higher due to the direct irradiation by primary alpha particles. The initial radioactivity (IR) that is needed to deposit a dose of 100 Gy (reference cytocidal dose) at the tumour edge (250-mm layer), was calculated as:
i h IR MBq$mm1 ¼ 100½Gy
ls1
TEMD Gy$MBq1 $s1 $mm
(1)
In this calculation, there is the implicit assumption of physical decay only of each radionuclide (i.e. no biological clearance). For each nuclide and every activity distribution configuration, IR is
Figure 6. Radial dose profiles for
111
In.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
E. Amato et al. / Physica Medica xxx (2015) 1e6
5
Table 3 Endothelial Cell Mean Dose (ECMD) and Tumour Edge Mean Dose (TEMD) [Gy MBq1 s1 mm]. 223
Configuration
A B C D E F G
131
Ra
177
I
111
Lu
In
ECMD
TEMD
ECMD
TEMD
ECMD
TEMD
ECMD
TEMD
9.16Eþ2 1.04Eþ3 1.01Eþ3 4.17Eþ2 1.18Eþ2 5.17Eþ1 2.90Eþ1
4.46E2 4.47E2 4.47E2 4.50E2 4.63E2 4.63E2 6.67E1
1.24Eþ0 1.51Eþ0 1.44Eþ0 5.65E1 2.41E1 1.43E1 9.70E2
1.76E2 1.76E2 1.76E2 1.79E2 1.87E2 2.03E2 2.34E2
1.53Eþ0 1.93Eþ0 1.83Eþ0 6.95E1 2.87E1 1.60E1 1.04E1
1.27E2 1.27E-2 1.27E2 1.30E2 1.38E2 1.58E2 2.03E2
2.38E1 9.56E1 7.78E1 1.85E1 6.82E2 4.03E2 2.76E2
3.19E3 3.23E3 3.22E3 3.48E3 4.10E3 4.69E3 5.30E3
Figure 7. . Plots of Endothelial Cell Mean Dose (ECMD) and Tumour Edge Mean Dose (TEMD) for all the radioactivity distributions considered.
reported in Table 4. As expected, one sees that the IR increases when going from alpha to beta and Auger emitters. IR for 223Ra in configuration G decreases by an order of magnitude with respect to the other configurations due to the direct irradiation of the tumour edge by alpha particles. Conclusions The development of small-scale models of the capillary vasculature of solid tumors is an important tool for understanding the effect of radionuclide therapies, due to both differences in biodistribution at small scales and range of the ionizing radiations. The Monte Carlo simulation developed in the present study allowed us to quantify dose distributions at the microscopic level around a simple model of tumour capillary vessel for some interesting therapeutic radionuclides. The differences between
Table 4 Initial Radioactivity [MBq mm1] needed to deposit 100 Gy of dose at the Tumour Edge. Configuration
223
A B C D E F G
1.57E3 1.57E3 1.57E3 1.56E3 1.51E3 1.51E3 1.05E4
Ra
131
I
5.69E3 5.69E3 5.69E3 5.60E3 5.36E3 4.93E3 4.27E3
177
Lu
9.50E3 9.50E3 9.49E3 9.27E3 8.73E3 7.62E3 5.95E3
111
In
8.97E2 8.86E2 8.89E2 8.22E2 6.98E2 6.10E2 5.40E2
irradiation properties of the alpha, beta and Auger emissions were pointed out and compared for the radionuclides considered. The results of this study contribute the characterization of dose inhomogeneities in solid tumour therapies with radiopharmaceuticals, taking into account the interplay between drug distribution from vasculature and range of the emitted particles. Future studies will be devoted to introduce the actual biodistribution of radiopharmaceuticals labelled with these nuclides in the model of solid tumour capillary vessels.
References [1] Humm JL. A microdosimetric model of astatine-211 labeled antibodies for radioimmunotherapy. Int J Radiat Oncol Biol Phys 1987;13:1767e73. [2] Huang C-Y, Oborn BM, Guatelli S, Allen BJ. Monte Carlo calculation of the maximum therapeutic gain of tumor antivascular alpha therapy. Med Phys 2012;39(3):1282e8. [3] Zhu X, Palmer MR, Makrigiorgos GM, Kassis AI. Solid-tumor radionuclide therapy dosimetry: new paradigms in view of tumor microenvironment and angiogenesis. Med Phys 2010;37(6):2974e84. [4] Agostinelli S, Allison J, Amako K, Apostolakis J, Araujo H, Arce P, et al. GEANT4 e a simulation toolkit. Nucl Instrum Methods A 2003;506:250e303. [5] Amato E, Italiano A, Baldari S. Monte Carlo study of voxel S factor dependence on tissue density and atomic composition. Nucl Instrum Methods Phys Res A 2013;729:870e6. [6] Amato E, Italiano A. An analytical model for calculating internal dose conversion coefficients (DCCs) for non-human biota. Radiat Environ Biophys 2014;53:455e9. [7] Amato E, Italiano A, Leotta S, Pergolizzi S, Torrisi L. Monte Carlo study of the dose enhancement effect of gold nanoparticles during X-ray therapies and evaluation of the anti-angiogenic effect on tumour capillary vessels. J X-ray Sci Technol 2013;21:237e47. [8] Amato E, Italiano A, Baldari S. Absorbed fractions for alpha particles in ellipsoidal volumes. Phys Med Biol 2013;58:5449e59.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
6
E. Amato et al. / Physica Medica xxx (2015) 1e6
[9] Jain RK. Transport of molecules, particles, and cells in solid-tumors. Annu Rev Biomed Eng 1999;1:241e63. [10] Stabin MG, da Luz LCQP. Decay data for internal and external dose assessment. Health Phys 2002;83:471e5.
[11] Amako K, Guatelli S, Ivanchencko V, Maire M, Mascialino B, Murakami K, et al. Geant4 and its validation. Nucl Phys B Proc Suppl 2003;150:44e9. [12] Carrier JF, Archambault L, Beaulieu L, Roy R. Validation of Geant4, an objectoriented Monte Carlo toolkit, for simulations in medical physics. Med Phys 2004;31:484e92.
Please cite this article in press as: Amato E, et al., A Monte Carlo approach to small-scale dosimetry of solid tumour microvasculature for nuclear medicine therapies with 223Ra-, 131I-, 177Lu- and 111In-labelled radiopharmaceuticals, Physica Medica (2015), http://dx.doi.org/10.1016/ j.ejmp.2015.04.015
![[Nuclear medicine and radiopharmaceuticals].](/assets/img/hold.png)
