- Email: [email protected]
Dose perturbations by high atomic number materials in intravascular brachytherapy
Cardiovascular Radiation Medicine 1:2 (1999) 144–153
PHYSICS ORIGINAL ARTICLE
DOSE PERTURBATIONS BY HIGH ATOMIC NUMBER MATERIALS IN INTRAVASCULAR BRACHYTHERAPY Ravinder Nath, Ph.D.,a,* Ning Yue, Ph.D.,a and Judah Weinberger, M.D., Ph.D.b b
a Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut, USA Cardiac Catheterization Laboratory, Division of Cardiology, Columbia-Presbyterian Medical Center, New York, New York, USA
Received 29 April 1999; accepted 30 April 1999
Purpose. In intravascular brachytherapy, use of high atomic number materials, such as contrast agents and metallic stents, can introduce significant dose perturbations, especially for low energy photons. The purpose of this study is to investigate dose perturbation at the interfaces of high atomic number materials and tissue. Methods. To investigate this issue, the radial dose functions across the interface between different materials and soft tissue were calculated by using Monte Carlo simulations. Various interfaces, including contrast agent to water, stainless steel to water, and bone (simulating a calcified plaque) to water, were investigated for photon energies between 20 keV and 1 MeV. Results. It was found that the dose to water near the interface is enhanced considerably by photons of energies between 0.020 and 0.200 MeV. For example, the maximum dose enhancement factors for the Hypaque–tissue interface ranged from 2.2 to 18.3 for photons in this energy range. The enhancement factor is almost equal to 1 for photon energy between 0.400 and 1.000 MeV. It appears that the maximum enhancement occurs around 60 keV. For 60-keV photons, the maximum dose enhancement factors are about 18.3, 18.7, 19.1, and 3.1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. The dose enhancement decreases exponentially with distance from the interface. The affected tissue thickness is dependent on the photon energy. As expected, the higher the photon energy is, the larger is the affected tissue thickness. Depending on the type of interface and the energy of photons, the dose enhancement distance (defined as the thickness receiving more than twice the dose without interface) ranges from 1.3 to 72 mm for photons of energy from 0.020 to 0.100 MeV, respectively. Conclusions. The existense of high atomic number materials could introduce significant dose enhancement at the interfaces between these materials and tissue. This dose enhancement can be higher than an order of magnitude for photon energies around 60 keV, and should be considered in evaluation of the efficacy of intravascular brachytherapy. © 1999 Elsevier Science Inc. Keywords: Intravascular brachytherapy; Monte Carlo simulation; Restenosis; Interface dosimetry; Dose perturbation; Dosimetry.
* Correspondence to: Ravinder Nath, Ph.D., Department of Therapeutic Radiology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. 1522-1865/99/$–see front matter. © 1999 Elsevier Science Inc. PII S1522-1865(99)0 0007- 4
Supported in part by a USPHS grant number 1-RO1-HL5802201 awarded by the National Health Lung and Blood Institute.
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
145
Figure 1. The product of dose rate for a source with 1 Bq activity and r2 vs. distance r at different energies for the interfaces between tissue and Omnipaque, Hypaque (A, B), stainless steel (SS), and calcified plaque (C, D), respectively.
Introduction Restenosis after arterial interventions such as balloon angioplasty or stent implantation remain a significant long-term limitation of the procedure [1–3]. In coronary arteries of the patients, atherosclerotic plaque gradually reduces the lumen of the arteries, causing reduced blood flow, and consequent myocardial ischemia and infarction. Various techniques have been developed to restore adequate blood flow. Percutaneous techniques to restore normal lumenal geometry, and thus re-establish adequate antegrade blood flow, include balloon dilatation, stent placement, directional atherectomy, rotational atherectomy, and laser angioplasty. However, these techniques, in general, cause some degree of vessel trauma at or around the treated site; this injury frequently occasions smooth muscle proliferation, neointimal proliferation, ad-
ventitial myofibroblast proliferation, and or vascular remodeling. These processes all contribute to arterial restenosis. Recent clinical trials have showed promising results in using radiation therapy to treat restenosis [4]. Radioactive sources, either photon or electron emitters, are positioned at the lesion site for a time sufficient to deliver a prescribed dose. The radiation inhibits neointimal proliferation. In intravascular brachytherapy, the target tissue is usually very close to the radioactive source and the distance away is usually less than a few millimeters. The nature of the treatment and the consideration of normal tissue protection determine the choice of the photon energy used. The dosimetry of various photon and electron emitters have been studied both experimentally and theoretically [5–14] for intravascular brachytherapy. It was found that the radial dose
146
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 1. Cont’d.
function of photons of energy between 0.020 and 0.400 MeV does not differ significantly from unity up to a distance of 10 mm in soft tissues [5]. Therefore, an energy range of 0.020–0.400 MeV is considered to be adequate for photon emitters. However, several kinds of materials of high atomic number (Z) may be present in the blood vessels undergoing radiation treatment. To visualize coronary vessels on fluoroscopy, cardiologists inject contrasts (e.g., Hypaque or Omnipaque), that contain a high percentage of a high Z element, iodine. In some radiation delivery systems, the contrast agents are interposed between the source and the target tissue in the arterial wall during radiation delivery. For the bulk of patients who currently have been stented, the stent, which is commonly made of stainless steel, is between the vessel wall and the radioactive source. Thus, for brachytherapy of stented patients, a high Z material is in direct contact with the target site. Furthermore, in many cases, calcified plaque will be left at the treatment site. The existence of the high Z materials around or in the treatment targets will perturb
the radiation dose delivered by the radioactive sources, especially in the region of interfaces between the high Z materials and tissues [15]. In this study, we investigated this dose perturbation by the high Z materials for monoenergetic photons using Monte Carlo simulations of radiation transport. Materials and Methods Four types of interfaces were considered: soft tissue to Hypaque, soft tissue to Omnipaque, soft tissue to stainless steel stent, and soft tissue to calcified plaque. Water was used to simulate soft tissue in this study. In the calculation, calcified plaque consisted of 3.1% H, 31.26% C, 0.99% N, 37.57% O, 0.05% Cl, and 27.03% Ca. The density used was 1.84 g/cm3. The composition of Hypaque was 7.88% C, 7.42% H, 22.73% I, 1.67% N, 58.92% O, and 1.38% Na. The density of Hypaque was chosen to be 1.32 g/cm3. Omnipaque was consistent of 14.96% C, 6.90% H, 24.89% I, 2.76% N, and 50.49% O. The density of Omnipaque was 1.406 g/
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 1. Cont’d.
147
148
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
cm3. Monte Carlo simulation was performed by using Integrated TIGER Series (ITS) of coupled electron/photon Monte Carlo code system (Version 2.1). The code was run on a DEC AlphaStation 200/ 66. The operating system is VMS Version 6.2. The cutoff energy was 1 keV. The number of histories was 2,000,000, and it usually took about 5–10 h CPU time per simulation. Geometry of the calculation was designed such that radioactive source was at the center and treated as a point source. Multiple layers of spherical shells that could have different thickness and radii and could be of different materials encapsulated the point source. In the calculation, the thickness of the layers in the interface region was set as 1 mm. Two types of calculation were performed. In the first type, high Z material was placed next to the point source, and beyond the high Z materials was water. The thickness of calcified plaque and stainless steel were 0.1 mm, which approximated the thickness of stent and calcified plaque in blood vessels; the thickness of the contrast agents were 1.1 mm, which approximated the radius of a typical coronary blood vessels. In the second type of calculation, all layers were set to be water while the geometry was exactly the same as in the first type. The calculation was performed for 20, 30, 40, 50, 60, 80, 70, 100, 200, and 400 keV, and 1 MeV monoenergetic photons, respectively. ITS code calculated the energy deposited in each layer by a single photon emitted by a photon source. Because thickness of layers was small, dose– distance function was calculated as the average dose in a layer to the average distance of the layer to the source. Assume the energy deposited in a layer is E, the mass density of the material in the layer is r, the inner and outer radius of the layer are R1 and R2, respectively, then the average dose Dm to the material in the layer is: 3E D m = --------------------------------3 3 4πρ ( R 2 – R 1 )
(1)
and the average distance to the source r is: 3 ( R2 – R1 ) r = ------------------------3 3 4 ( R2 – R1 ) 4
4
(2)
The dose to the material Dm was converted to the dose to water Dw for a photon source by using equation: µ en w D w = D m ----- ρ m where µ en ----- ρ m w
(3)
is the ratio of average mass energy absorption coefficients of water and the material m for photons. The mass energy coefficients were obtained for each type of high Z materials and each energy from the work by Hubbell [15]. From the energy deposited by a single photon, the dose rate was calculated for a source with 1 Bq activity, using equations 1, 2, and 3. To investigate the effect of interface due to the existence of high Z materials, two parameters, dose enhancement factor (DEF) and dose enhancement distance (DED), were derived. DEF was defined as the ratio of the peaked dose at the interface between water and high Z material to the dose at the corresponding location without the interface. The DED was defined as the distance from the interface at which the dose with the existence of high Z materials was at least twice the dose without the high Z materials in present. Both enhancement factor and enhancement distance were calculated as a function of photon energy so that the energy dependence of the perturbation could be investigated.
Results and Discussion Soft tissue to high Z material interface introduced significant near-field dose perturbations. Figures 1A–D illustrate the dose rate in cGy/s from a source with 1 Bq activity as a function of distance to the source for monoenergetic photons of energies from 0.020 to 0.200 MeV. The dose rate was expressed as the product of dose rate and the square of the distance to the source to take out the inverse square factor. The interfaces include soft tissue to Hypaque, Omnipaque, stainless steel, and calcified plaque. As stated earlier, the assumed thickness of Hypaque and Omnipaque was 1.1 mm and the thickness of stainless steel and calcified plaque was 0.1 mm. It is evident that, at all four interfaces analyzed, there was an enhanced dose peak, indicating that the dose to the soft tissue at the interfaces was significantly perturbed. As shown in our previous work [5], the radial dose functions of photons of energy from 0.020 to 0.200 MeV do not change significantly within the depth range of 10 mm. Therefore the peaks were caused by the presence of the interfaces. To quantify the dose perturbation due to the existence of high Z material, the ratio of dose in the presence of high Z material to that without high Z material was calculated. Figures 2A–D show the ratio-distance functions for the four types of interfaces at different energies. It is clear that the existence of the interfaces between high Z material and tissue enhanced the dose to tissue at the interfaces for the photon energies between 0.020 and 0.200
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
149
Figure 2. The ratio of doses with the existence of high Z materials to those in pure water at different energies for the interfaces between tissue and Omnipaque, Hypaque (A, B), stainless steel (SS), and calcified plaque (C, D), respectively.
MeV. The increase of the dose could be as much as almost 20 times the unperturbed dose. The magnitude of perturbation, which can be identified by the height of peak, and the perturbation range, were material dependent and energy of incident photon dependent. Because the compositions of Omnipaque and Hypaque were not significantly different, their dose perturbations were similar. On the other hand, the perturbations of stainless steel and calcified plaque were different from each other and from the contrast agents. It is clear that calcified plaque did not change the dose as much as other three high Z materials. The reason is that the effective Z of calcified plaque was closer to that of water than the effective atomic numbers of contrast agents and stainless steel. It appeared that even though the perturbation was energy dependent, its magnitude did not simply increase with photon energy. The magnitude increased first with energy, then decreased with energy after reaching a highest point (Fig. 3). Further-
more, the magnitude changed rapidly with distance from the interface (Fig. 2). The magnitude dropped exponentially with the distance from the interface (Fig. 2). However, the perturbed depth increased with energy (Fig. 2), whereas the dependence of DED on energy was more complicated (Fig. 4). The dependence of dose perturbation on photon energy can be explained as follows. When the photon energy is low, the photon interaction with medium is dominated by the photoelectric effect. Because the photoelectric effect is proportional to Z3, there will be more interactions between photon and high Z materials than low Z material (water). That means there are more secondary electrons ejected in high Z materials than in water. Therefore, within a distance from the interface, which is equal to the practical range of the electrons ejected by photons in the high Z material, more secondary electrons will deposit energy to a point in water than without the interface. This gives rise to a
150
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 2. Cont’d.
higher dose at that point. In another word, within the practical range from the interface, the dose will be enhanced if attenuation is not considered. More precisely, the perturbation of dose to a point in water at the interface is the combination of the effect (interface effect) mentioned above and attenuation by the high Z material. If the interface effect dominates, the dose at the interface will be higher than without the interface. Obviously the interface effect will be more pronounced closer to the interface because of more secondary electrons contributed from the high Z material side. Because the energy of secondary electrons is low and they lose their energy quickly, the magnitude of perturbation decreases dramatically with distance from the interface. On the other hand, because the practical range of the secondary electrons contributed from high Z material increases with photon energy, the perturbed depth increases with photon energy. As the photon energy increases, photoelectrons are ejected preferentially in the forward direction (the direction of photon travel). Although the photoelectric
interaction is inversely proportional to E3, initially the combined effects actually make the high Z material contribute more secondary electrons as photon energy increases. Thus the magnitude of perturbation increases first with energy. As the energy of photon becomes higher and higher, the photoelectric interaction becomes quantitatively less dominant, and the Compton interaction becomes dominant. Because the Compton interaction is almost independent of atomic number, Z, the magnitude of perturbation becomes less pronounced when the photon energy is increased further. Actually, as indicated in Fig. 1D, perturbation became very small for 0.200 MeV photon at the interfaces of tissue to stainless steel and tissue to calcified plaque. According to our calculation, almost no dose perturbation was observed when the energy of photon is higher than 0.400 MeV. The energy dependence of the DEF was plotted for the four types of materials in Fig. 3. As expected, the enhancement factor increased first with photon energy, then decreased with the energy. At 20 keV, the DEF ranged from 2.5 to 8.5, respectively. The precise
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 2. Cont’d.
151
152
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 3. Dose enhancement factors at different energies for the interfaces between tissue and Omnipaque, Hypaque, stainless steel, and calcified plaque, respectively.
Figure 4. Dose enhancement distances at different energies for the interfaces between tissue and Omnipaque, Hypaque, stainless steel (SS), and calcified plaque, respectively.
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
energy at which the DEF reached a maximum value varied between the materials. However, the energy for all materials was about 60 keV, and the corresponding DEFs were 18.3, 18.7, 19.1, and 3.1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. When photon energy increased to 200 keV, the DEFs dropped to 2.2, 2.5, 1.5, and 1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. In the whole range of photon energies investigated, the maximum DEF for calcified plaque interface was about 4.5. The dose perturbation by calcified plaque interface was much less than that caused by other materials studied. The energy dependence of the DED was plotted for the four types of materials in Fig. 4. The DED, as defined previously, increased almost linearly with photon energy for stainless steel interface. The dependence of DED on photon energy showed more complicated patterns with photon energy for the other three interfaces. The DED first increased with photon energy, then decreased a little bit at about 80 keV then increased again for Omnipaque and Hypaque interfaces. For calcified plaque interface, the DED increased with energy initially then started to decrease at about 70 keV. Generally speaking, the DED ranged from about 1.3 mm at 20 keV to about more than 70 mm at 100 keV. Conclusion In intravascular brachytherapy, the existence of high Z materials, like contrast agents and stainless steel, could introduce significant dose perturbations at the interfaces between the high Z materials and tissue. The dose to tissue at these interfaces could be enhanced considerably, sometimes by more than one order of magnitude. The enhancement is photon energy dependent, increasing first then decreasing with energy. The peak appeared to be around 60 keV. The enhancement distance is also energy dependent. Enhancement distance increases almost linearly with photon energy. However, for photon energy above 200 keV, the magnitude of dose perturbation is negligible. In summary, when a photon source with an energy below 200 keV is used in intravascular brachytherapy, the dose perturbation by high Z materials should be taken into consideration for the evaluation of clinical efficacy of the brachytherapy. For this purpose, it is necessary to determine the magnitude of predicted dose perturbation attributable to lower energy photons in available gamma emitters such as 192 Ir in the presence of a previously implanted stent or other high Z materials. The sensitivity of these predictions to thickness of stainless steel struts or other high Z materials also needs to be investigated.
153
Theoretically, the existence of high Z materials could also introduce significant back-scattering effects. However, because in intravascular brachytherapy, the high Z materials usually exist between radioactive sources and tissue target, the back-scattering effect is not as important as forward-scattering effect studied here. Investigations of back-scattering effect are planned in the near future. References [1] Pock SJ, Henderson RA, Rickards AF, et al. Meta analysis of randomized trials comparing coronary angioplasty with bypass surgery. Lancet 1995;346:1184–1189. [2] Serruys PW, DeJaegere P, Kiemeneij F, et al. A comparison of balloon-expandable-stent implantation with balloon angioplasty in patients with coronary artery disease. N. Engl. J. Med. 1994;331:489–495. [3] Fischman DL, Leon MB, Baim DS, et al. A randomized comparison of coronary-stent placement and balloon angioplasty in the treatment of coronary artery disease. N. Engl. J. Med. 1994;331:496–501. [4] Waksman R. Radiation for prevention of restenosis: where are we? Int. J. Radiat. Oncol. Biol. Phys. 1996;36:959–961. [5] Nath R, Yue N, Liu L. On the depth of penetration of photons and electrons for intravascular brachytherapy. Cardiovasc. Radiat. Med. 1999;1:72–79. [6] Dawson JT. Theoretical considerations regarding low-dose radiation therapy for prevention of restenosis after angioplasty. Tex. Heart Inst. J. 1991;18:4–7. [7] Nath R, Amols H, Jani S, et al. Intravascular brachytherapy physics: report of the AAPM Radiation Therapy Committee Task Group No. 60. Med. Phys. 1999;26:119–152. [8] Soares CG, Halpern DG, Wang C-K. Calibration and characterization of beta-particle sources for intravascular brachytherapy. Med. Phys. 1998;25:339–346. [9] Janicki C, Duggan DM, Coffey CW, Fischell DR, Fischell TA. Radiation dose from a phosphorous-32 impregnated wire mesh vascular stent. Med. Phys. 1997;24:437–445. [10] Meigooni AS, Nath R. A comparison of radial dose functions for 103Pd, 125I, 145Sm, 241Am, 169Yb, 192Ir and 137Cs brachytherapy sources. Int. J. Radiat. Oncol. Biol. Phys. 1992;22:1125–1130. [11] Amols HI, Zaider M, Weinberger J, Ennis R, Schiff PB, Reinstein LE. Dosimetry considerations for catheter-based beta and gamma emitters in the therapy of neointimal hyperplasia in human coronary arteries. Int. J. Radiat. Oncol. Biol. Phys. 1996;36:913–921. [12] Xu Z, Almond PR, Deasy JO. The dose distribution produced by 32P source for endovascular irradiation. Int. J. Radiat. Oncol. Biol. Phys. 1996;36:933–939 [13] Popowski Y, Verin V, Papirov I, et al. Intra-arterial Y-90 brachytherapy: preliminary dosimetric study using a specially modified angioplasty balloon. Int. J. Radiat. Oncol. Biol. Phys. 1995;33:713–717. [14] Duggan D, Coffey C, Levit S. Point dose kernels for pure beta emitting intracoronary brachtherapy stents theoretical models versus experimental methods using radiochromic dosimetry. Int. J. Radiat. Oncol. Biol. Phys. 1998;40:713–720. [15] Das IJ. Forward dose pertubation at high atomic number interfaces in kilovoltage x-ray beams. Med. Phys. 1997; 24: 1781–1787. [16] Hubbell JH. Photon mass attenuation and energy-absorption coefficients from 1 keV to 20 MeV. Int. J. Appl. Radat. Isot. 1982;33:1269–1290.
PHYSICS ORIGINAL ARTICLE
DOSE PERTURBATIONS BY HIGH ATOMIC NUMBER MATERIALS IN INTRAVASCULAR BRACHYTHERAPY Ravinder Nath, Ph.D.,a,* Ning Yue, Ph.D.,a and Judah Weinberger, M.D., Ph.D.b b
a Department of Therapeutic Radiology, Yale University School of Medicine, New Haven, Connecticut, USA Cardiac Catheterization Laboratory, Division of Cardiology, Columbia-Presbyterian Medical Center, New York, New York, USA
Received 29 April 1999; accepted 30 April 1999
Purpose. In intravascular brachytherapy, use of high atomic number materials, such as contrast agents and metallic stents, can introduce significant dose perturbations, especially for low energy photons. The purpose of this study is to investigate dose perturbation at the interfaces of high atomic number materials and tissue. Methods. To investigate this issue, the radial dose functions across the interface between different materials and soft tissue were calculated by using Monte Carlo simulations. Various interfaces, including contrast agent to water, stainless steel to water, and bone (simulating a calcified plaque) to water, were investigated for photon energies between 20 keV and 1 MeV. Results. It was found that the dose to water near the interface is enhanced considerably by photons of energies between 0.020 and 0.200 MeV. For example, the maximum dose enhancement factors for the Hypaque–tissue interface ranged from 2.2 to 18.3 for photons in this energy range. The enhancement factor is almost equal to 1 for photon energy between 0.400 and 1.000 MeV. It appears that the maximum enhancement occurs around 60 keV. For 60-keV photons, the maximum dose enhancement factors are about 18.3, 18.7, 19.1, and 3.1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. The dose enhancement decreases exponentially with distance from the interface. The affected tissue thickness is dependent on the photon energy. As expected, the higher the photon energy is, the larger is the affected tissue thickness. Depending on the type of interface and the energy of photons, the dose enhancement distance (defined as the thickness receiving more than twice the dose without interface) ranges from 1.3 to 72 mm for photons of energy from 0.020 to 0.100 MeV, respectively. Conclusions. The existense of high atomic number materials could introduce significant dose enhancement at the interfaces between these materials and tissue. This dose enhancement can be higher than an order of magnitude for photon energies around 60 keV, and should be considered in evaluation of the efficacy of intravascular brachytherapy. © 1999 Elsevier Science Inc. Keywords: Intravascular brachytherapy; Monte Carlo simulation; Restenosis; Interface dosimetry; Dose perturbation; Dosimetry.
* Correspondence to: Ravinder Nath, Ph.D., Department of Therapeutic Radiology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. 1522-1865/99/$–see front matter. © 1999 Elsevier Science Inc. PII S1522-1865(99)0 0007- 4
Supported in part by a USPHS grant number 1-RO1-HL5802201 awarded by the National Health Lung and Blood Institute.
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
145
Figure 1. The product of dose rate for a source with 1 Bq activity and r2 vs. distance r at different energies for the interfaces between tissue and Omnipaque, Hypaque (A, B), stainless steel (SS), and calcified plaque (C, D), respectively.
Introduction Restenosis after arterial interventions such as balloon angioplasty or stent implantation remain a significant long-term limitation of the procedure [1–3]. In coronary arteries of the patients, atherosclerotic plaque gradually reduces the lumen of the arteries, causing reduced blood flow, and consequent myocardial ischemia and infarction. Various techniques have been developed to restore adequate blood flow. Percutaneous techniques to restore normal lumenal geometry, and thus re-establish adequate antegrade blood flow, include balloon dilatation, stent placement, directional atherectomy, rotational atherectomy, and laser angioplasty. However, these techniques, in general, cause some degree of vessel trauma at or around the treated site; this injury frequently occasions smooth muscle proliferation, neointimal proliferation, ad-
ventitial myofibroblast proliferation, and or vascular remodeling. These processes all contribute to arterial restenosis. Recent clinical trials have showed promising results in using radiation therapy to treat restenosis [4]. Radioactive sources, either photon or electron emitters, are positioned at the lesion site for a time sufficient to deliver a prescribed dose. The radiation inhibits neointimal proliferation. In intravascular brachytherapy, the target tissue is usually very close to the radioactive source and the distance away is usually less than a few millimeters. The nature of the treatment and the consideration of normal tissue protection determine the choice of the photon energy used. The dosimetry of various photon and electron emitters have been studied both experimentally and theoretically [5–14] for intravascular brachytherapy. It was found that the radial dose
146
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 1. Cont’d.
function of photons of energy between 0.020 and 0.400 MeV does not differ significantly from unity up to a distance of 10 mm in soft tissues [5]. Therefore, an energy range of 0.020–0.400 MeV is considered to be adequate for photon emitters. However, several kinds of materials of high atomic number (Z) may be present in the blood vessels undergoing radiation treatment. To visualize coronary vessels on fluoroscopy, cardiologists inject contrasts (e.g., Hypaque or Omnipaque), that contain a high percentage of a high Z element, iodine. In some radiation delivery systems, the contrast agents are interposed between the source and the target tissue in the arterial wall during radiation delivery. For the bulk of patients who currently have been stented, the stent, which is commonly made of stainless steel, is between the vessel wall and the radioactive source. Thus, for brachytherapy of stented patients, a high Z material is in direct contact with the target site. Furthermore, in many cases, calcified plaque will be left at the treatment site. The existence of the high Z materials around or in the treatment targets will perturb
the radiation dose delivered by the radioactive sources, especially in the region of interfaces between the high Z materials and tissues [15]. In this study, we investigated this dose perturbation by the high Z materials for monoenergetic photons using Monte Carlo simulations of radiation transport. Materials and Methods Four types of interfaces were considered: soft tissue to Hypaque, soft tissue to Omnipaque, soft tissue to stainless steel stent, and soft tissue to calcified plaque. Water was used to simulate soft tissue in this study. In the calculation, calcified plaque consisted of 3.1% H, 31.26% C, 0.99% N, 37.57% O, 0.05% Cl, and 27.03% Ca. The density used was 1.84 g/cm3. The composition of Hypaque was 7.88% C, 7.42% H, 22.73% I, 1.67% N, 58.92% O, and 1.38% Na. The density of Hypaque was chosen to be 1.32 g/cm3. Omnipaque was consistent of 14.96% C, 6.90% H, 24.89% I, 2.76% N, and 50.49% O. The density of Omnipaque was 1.406 g/
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 1. Cont’d.
147
148
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
cm3. Monte Carlo simulation was performed by using Integrated TIGER Series (ITS) of coupled electron/photon Monte Carlo code system (Version 2.1). The code was run on a DEC AlphaStation 200/ 66. The operating system is VMS Version 6.2. The cutoff energy was 1 keV. The number of histories was 2,000,000, and it usually took about 5–10 h CPU time per simulation. Geometry of the calculation was designed such that radioactive source was at the center and treated as a point source. Multiple layers of spherical shells that could have different thickness and radii and could be of different materials encapsulated the point source. In the calculation, the thickness of the layers in the interface region was set as 1 mm. Two types of calculation were performed. In the first type, high Z material was placed next to the point source, and beyond the high Z materials was water. The thickness of calcified plaque and stainless steel were 0.1 mm, which approximated the thickness of stent and calcified plaque in blood vessels; the thickness of the contrast agents were 1.1 mm, which approximated the radius of a typical coronary blood vessels. In the second type of calculation, all layers were set to be water while the geometry was exactly the same as in the first type. The calculation was performed for 20, 30, 40, 50, 60, 80, 70, 100, 200, and 400 keV, and 1 MeV monoenergetic photons, respectively. ITS code calculated the energy deposited in each layer by a single photon emitted by a photon source. Because thickness of layers was small, dose– distance function was calculated as the average dose in a layer to the average distance of the layer to the source. Assume the energy deposited in a layer is E, the mass density of the material in the layer is r, the inner and outer radius of the layer are R1 and R2, respectively, then the average dose Dm to the material in the layer is: 3E D m = --------------------------------3 3 4πρ ( R 2 – R 1 )
(1)
and the average distance to the source r is: 3 ( R2 – R1 ) r = ------------------------3 3 4 ( R2 – R1 ) 4
4
(2)
The dose to the material Dm was converted to the dose to water Dw for a photon source by using equation: µ en w D w = D m ----- ρ m where µ en ----- ρ m w
(3)
is the ratio of average mass energy absorption coefficients of water and the material m for photons. The mass energy coefficients were obtained for each type of high Z materials and each energy from the work by Hubbell [15]. From the energy deposited by a single photon, the dose rate was calculated for a source with 1 Bq activity, using equations 1, 2, and 3. To investigate the effect of interface due to the existence of high Z materials, two parameters, dose enhancement factor (DEF) and dose enhancement distance (DED), were derived. DEF was defined as the ratio of the peaked dose at the interface between water and high Z material to the dose at the corresponding location without the interface. The DED was defined as the distance from the interface at which the dose with the existence of high Z materials was at least twice the dose without the high Z materials in present. Both enhancement factor and enhancement distance were calculated as a function of photon energy so that the energy dependence of the perturbation could be investigated.
Results and Discussion Soft tissue to high Z material interface introduced significant near-field dose perturbations. Figures 1A–D illustrate the dose rate in cGy/s from a source with 1 Bq activity as a function of distance to the source for monoenergetic photons of energies from 0.020 to 0.200 MeV. The dose rate was expressed as the product of dose rate and the square of the distance to the source to take out the inverse square factor. The interfaces include soft tissue to Hypaque, Omnipaque, stainless steel, and calcified plaque. As stated earlier, the assumed thickness of Hypaque and Omnipaque was 1.1 mm and the thickness of stainless steel and calcified plaque was 0.1 mm. It is evident that, at all four interfaces analyzed, there was an enhanced dose peak, indicating that the dose to the soft tissue at the interfaces was significantly perturbed. As shown in our previous work [5], the radial dose functions of photons of energy from 0.020 to 0.200 MeV do not change significantly within the depth range of 10 mm. Therefore the peaks were caused by the presence of the interfaces. To quantify the dose perturbation due to the existence of high Z material, the ratio of dose in the presence of high Z material to that without high Z material was calculated. Figures 2A–D show the ratio-distance functions for the four types of interfaces at different energies. It is clear that the existence of the interfaces between high Z material and tissue enhanced the dose to tissue at the interfaces for the photon energies between 0.020 and 0.200
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
149
Figure 2. The ratio of doses with the existence of high Z materials to those in pure water at different energies for the interfaces between tissue and Omnipaque, Hypaque (A, B), stainless steel (SS), and calcified plaque (C, D), respectively.
MeV. The increase of the dose could be as much as almost 20 times the unperturbed dose. The magnitude of perturbation, which can be identified by the height of peak, and the perturbation range, were material dependent and energy of incident photon dependent. Because the compositions of Omnipaque and Hypaque were not significantly different, their dose perturbations were similar. On the other hand, the perturbations of stainless steel and calcified plaque were different from each other and from the contrast agents. It is clear that calcified plaque did not change the dose as much as other three high Z materials. The reason is that the effective Z of calcified plaque was closer to that of water than the effective atomic numbers of contrast agents and stainless steel. It appeared that even though the perturbation was energy dependent, its magnitude did not simply increase with photon energy. The magnitude increased first with energy, then decreased with energy after reaching a highest point (Fig. 3). Further-
more, the magnitude changed rapidly with distance from the interface (Fig. 2). The magnitude dropped exponentially with the distance from the interface (Fig. 2). However, the perturbed depth increased with energy (Fig. 2), whereas the dependence of DED on energy was more complicated (Fig. 4). The dependence of dose perturbation on photon energy can be explained as follows. When the photon energy is low, the photon interaction with medium is dominated by the photoelectric effect. Because the photoelectric effect is proportional to Z3, there will be more interactions between photon and high Z materials than low Z material (water). That means there are more secondary electrons ejected in high Z materials than in water. Therefore, within a distance from the interface, which is equal to the practical range of the electrons ejected by photons in the high Z material, more secondary electrons will deposit energy to a point in water than without the interface. This gives rise to a
150
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 2. Cont’d.
higher dose at that point. In another word, within the practical range from the interface, the dose will be enhanced if attenuation is not considered. More precisely, the perturbation of dose to a point in water at the interface is the combination of the effect (interface effect) mentioned above and attenuation by the high Z material. If the interface effect dominates, the dose at the interface will be higher than without the interface. Obviously the interface effect will be more pronounced closer to the interface because of more secondary electrons contributed from the high Z material side. Because the energy of secondary electrons is low and they lose their energy quickly, the magnitude of perturbation decreases dramatically with distance from the interface. On the other hand, because the practical range of the secondary electrons contributed from high Z material increases with photon energy, the perturbed depth increases with photon energy. As the photon energy increases, photoelectrons are ejected preferentially in the forward direction (the direction of photon travel). Although the photoelectric
interaction is inversely proportional to E3, initially the combined effects actually make the high Z material contribute more secondary electrons as photon energy increases. Thus the magnitude of perturbation increases first with energy. As the energy of photon becomes higher and higher, the photoelectric interaction becomes quantitatively less dominant, and the Compton interaction becomes dominant. Because the Compton interaction is almost independent of atomic number, Z, the magnitude of perturbation becomes less pronounced when the photon energy is increased further. Actually, as indicated in Fig. 1D, perturbation became very small for 0.200 MeV photon at the interfaces of tissue to stainless steel and tissue to calcified plaque. According to our calculation, almost no dose perturbation was observed when the energy of photon is higher than 0.400 MeV. The energy dependence of the DEF was plotted for the four types of materials in Fig. 3. As expected, the enhancement factor increased first with photon energy, then decreased with the energy. At 20 keV, the DEF ranged from 2.5 to 8.5, respectively. The precise
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 2. Cont’d.
151
152
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
Figure 3. Dose enhancement factors at different energies for the interfaces between tissue and Omnipaque, Hypaque, stainless steel, and calcified plaque, respectively.
Figure 4. Dose enhancement distances at different energies for the interfaces between tissue and Omnipaque, Hypaque, stainless steel (SS), and calcified plaque, respectively.
R. Nath et al. / Cardiovascular Radiation Medicine 1:2 (1999) 144–153
energy at which the DEF reached a maximum value varied between the materials. However, the energy for all materials was about 60 keV, and the corresponding DEFs were 18.3, 18.7, 19.1, and 3.1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. When photon energy increased to 200 keV, the DEFs dropped to 2.2, 2.5, 1.5, and 1 for Hypaque, Omnipaque, stainless steel, and calcified plaque, respectively. In the whole range of photon energies investigated, the maximum DEF for calcified plaque interface was about 4.5. The dose perturbation by calcified plaque interface was much less than that caused by other materials studied. The energy dependence of the DED was plotted for the four types of materials in Fig. 4. The DED, as defined previously, increased almost linearly with photon energy for stainless steel interface. The dependence of DED on photon energy showed more complicated patterns with photon energy for the other three interfaces. The DED first increased with photon energy, then decreased a little bit at about 80 keV then increased again for Omnipaque and Hypaque interfaces. For calcified plaque interface, the DED increased with energy initially then started to decrease at about 70 keV. Generally speaking, the DED ranged from about 1.3 mm at 20 keV to about more than 70 mm at 100 keV. Conclusion In intravascular brachytherapy, the existence of high Z materials, like contrast agents and stainless steel, could introduce significant dose perturbations at the interfaces between the high Z materials and tissue. The dose to tissue at these interfaces could be enhanced considerably, sometimes by more than one order of magnitude. The enhancement is photon energy dependent, increasing first then decreasing with energy. The peak appeared to be around 60 keV. The enhancement distance is also energy dependent. Enhancement distance increases almost linearly with photon energy. However, for photon energy above 200 keV, the magnitude of dose perturbation is negligible. In summary, when a photon source with an energy below 200 keV is used in intravascular brachytherapy, the dose perturbation by high Z materials should be taken into consideration for the evaluation of clinical efficacy of the brachytherapy. For this purpose, it is necessary to determine the magnitude of predicted dose perturbation attributable to lower energy photons in available gamma emitters such as 192 Ir in the presence of a previously implanted stent or other high Z materials. The sensitivity of these predictions to thickness of stainless steel struts or other high Z materials also needs to be investigated.
153
Theoretically, the existence of high Z materials could also introduce significant back-scattering effects. However, because in intravascular brachytherapy, the high Z materials usually exist between radioactive sources and tissue target, the back-scattering effect is not as important as forward-scattering effect studied here. Investigations of back-scattering effect are planned in the near future. References [1] Pock SJ, Henderson RA, Rickards AF, et al. Meta analysis of randomized trials comparing coronary angioplasty with bypass surgery. Lancet 1995;346:1184–1189. [2] Serruys PW, DeJaegere P, Kiemeneij F, et al. A comparison of balloon-expandable-stent implantation with balloon angioplasty in patients with coronary artery disease. N. Engl. J. Med. 1994;331:489–495. [3] Fischman DL, Leon MB, Baim DS, et al. A randomized comparison of coronary-stent placement and balloon angioplasty in the treatment of coronary artery disease. N. Engl. J. Med. 1994;331:496–501. [4] Waksman R. Radiation for prevention of restenosis: where are we? Int. J. Radiat. Oncol. Biol. Phys. 1996;36:959–961. [5] Nath R, Yue N, Liu L. On the depth of penetration of photons and electrons for intravascular brachytherapy. Cardiovasc. Radiat. Med. 1999;1:72–79. [6] Dawson JT. Theoretical considerations regarding low-dose radiation therapy for prevention of restenosis after angioplasty. Tex. Heart Inst. J. 1991;18:4–7. [7] Nath R, Amols H, Jani S, et al. Intravascular brachytherapy physics: report of the AAPM Radiation Therapy Committee Task Group No. 60. Med. Phys. 1999;26:119–152. [8] Soares CG, Halpern DG, Wang C-K. Calibration and characterization of beta-particle sources for intravascular brachytherapy. Med. Phys. 1998;25:339–346. [9] Janicki C, Duggan DM, Coffey CW, Fischell DR, Fischell TA. Radiation dose from a phosphorous-32 impregnated wire mesh vascular stent. Med. Phys. 1997;24:437–445. [10] Meigooni AS, Nath R. A comparison of radial dose functions for 103Pd, 125I, 145Sm, 241Am, 169Yb, 192Ir and 137Cs brachytherapy sources. Int. J. Radiat. Oncol. Biol. Phys. 1992;22:1125–1130. [11] Amols HI, Zaider M, Weinberger J, Ennis R, Schiff PB, Reinstein LE. Dosimetry considerations for catheter-based beta and gamma emitters in the therapy of neointimal hyperplasia in human coronary arteries. Int. J. Radiat. Oncol. Biol. Phys. 1996;36:913–921. [12] Xu Z, Almond PR, Deasy JO. The dose distribution produced by 32P source for endovascular irradiation. Int. J. Radiat. Oncol. Biol. Phys. 1996;36:933–939 [13] Popowski Y, Verin V, Papirov I, et al. Intra-arterial Y-90 brachytherapy: preliminary dosimetric study using a specially modified angioplasty balloon. Int. J. Radiat. Oncol. Biol. Phys. 1995;33:713–717. [14] Duggan D, Coffey C, Levit S. Point dose kernels for pure beta emitting intracoronary brachtherapy stents theoretical models versus experimental methods using radiochromic dosimetry. Int. J. Radiat. Oncol. Biol. Phys. 1998;40:713–720. [15] Das IJ. Forward dose pertubation at high atomic number interfaces in kilovoltage x-ray beams. Med. Phys. 1997; 24: 1781–1787. [16] Hubbell JH. Photon mass attenuation and energy-absorption coefficients from 1 keV to 20 MeV. Int. J. Appl. Radat. Isot. 1982;33:1269–1290.