- Email: [email protected]
Dose effects of guide wires for catheter-based intravascular brachytherapy
Int. J. Radiation Oncology Biol. Phys., Vol. 51, No. 4, pp. 1103–1110, 2001 Copyright © 2001 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/01/$–see front matter
PII S0360-3016(01)01763-1
PHYSICS CONTRIBUTION
DOSE EFFECTS OF GUIDE WIRES FOR CATHETER-BASED INTRAVASCULAR BRACHYTHERAPY X. ALLEN LI, PH.D.,*
AND
ROMPIN SHIH, PH.D.*†
*Department of Radiation Oncology, University of Maryland, Baltimore, MD; †Department of Radiation Oncology, Tri-Service General Hospital & National Defense Medical College, Taipei, Taiwan, R.O.C. Purpose: Guide wires with high torquability and steerability are commonly used to navigate through a tortuous and/or branching arterial tree in a catheter-based intravascular brachytherapy procedure. The dosimetric effects due to the presence of metallic guide wires have not been addressed. This work investigates these dose effects for the three most commonly used  and ␥ sources (90Sr, 32P, and 192Ir). Methods and Materials: The EGS4 Monte Carlo codes were used to calculate the dose distributions for the 90 Sr(NOVOSTE), 32P (Guidant), and 192Ir (BEST Ind.) with and without a guide wire in place. Energy spectra for particles exiting the sources were calculated from the full phase-space data obtained from the Monte Carlo simulations of the source constructions. Guide wires of various thicknesses and compositions were studied. Results: The dose perturbations due to the presence of guide wires were found to be far more significant for the 90 Sr/90Y and 32P beta sources than those for the 192Ir gamma source. Because of the attenuation by the guide wires, a dose reduction of up to 60% behind a guide wire was observed for the beta sources, whereas the dose perturbation was found to be negligible for the ␥ source. For a  source, the dose perturbations depend on the thickness and the material of the guide wire. When the region behind a guide wire is part of an intravascular brachytherapy target, the presence of the guide wire results in a significant underdosing for  sources. The underdosed region can extend a few mm behind the guide wire and up to 1 mm in other directions. Conclusion: Significant dose perturbations by the presence of a metallic guide wire have been found in catheter-based intravascular brachytherapy using  sources. The dose effects should be considered in the dose prescription and/or in analyzing the treatment outcome for  sources. Such precautions are not necessary if using a gamma source. © 2001 Elsevier Science Inc. Intravascular brachytherapy, Dosimetry, Guide wire, EGS4 Monte Carlo.
for each specific patient” (13). However, the uncertainty in the three-dimensional (3-D) dosimetry of IVBT is generally high at the current time. This is because dosimetric measurements are difficult due to the extremely high dose gradient, small region of interest, and complicated source and patient geometry. Also, analytic and empiric dose calculations are inaccurate, because of the perturbations caused by delivery devices, heterogeneities, and scatter and selfabsorption of low-energy secondary radiation (13, 14). It has been reported that the dose distribution is perturbed by the presence of metallic stents (15–19), calcified plaques (15, 19, 20), and radiopaque contrast media (19, 21) used during the course of the treatment. These perturbations are found to be significant for beta sources and can be as high as 50%. During an angioplasty procedure, a steerable guide wire (GW) inserted into guiding catheter is advanced to
INTRODUCTION Intravascular brachytherapy (IVBT) has recently been recognized as a treatment modality for reducing coronary restenosis after angioplasty (1–10). Many clinical trials have already proven the efficacy of IVBT for in-stent restenosis, and several other clinical trials are exploring its efficacy for other lesions (11, 12). At this stage, the optimum dose and treatment target are not well understood. Accurate dosimetry is crucial in analyzing data from these trials to determine the most effective dose window and to improve our understanding of treatment failure. Without such accurate data, the derived dose–response relationship would contain flaws, and key information to improve the treatment would be missed. The recently published AAPM TG-60 protocol for IVBT physics states that “ideally a three-dimensional dose distribution over the entire irradiated volume should be generated
This investigation was supported in part by a grant from the National Institutes of Health (No. R1HL62213). Received Feb 5, 2001, and in revised form Jul 3, 2001. Accepted for publication Jul 12, 2001.
Reprint requests to: X. A. Li, Department of Radiation Oncology, University of Maryland, 22 South Greene St., Baltimore, MD 21201-1595. Tel: (410) 328-7165; Fax: (410) 328-5279; E-mail: [email protected] Acknowledgments—The authors are grateful to Dr. Mohan Suntharalingam for the useful discussion on the topic. 1103
1104
I. J. Radiation Oncology
● Biology ● Physics
pass the target region under fluoroscopic guidance. This guide wire is necessary, because balloon and other intracoronary catheters generally lack the torquability and steerability necessary to navigate through a tortuous and/or branching arterial tree. Even if intracoronary catheters had these properties, interventional procedures often require that catheters be placed across the lesion and withdrawn multiple times, so that a guide wire over which these catheters can be quickly and safely advanced saves time. A wide variety of guide wires is available for various situations and from various manufacturers. The GW lengths range from 175 to 300 cm. Their diameters are specified in inches, typically from 0.009 to 0.018 inches. Thicker wires are required for peripheral vessels. The guide wires are usually metallic (typically stainless steel). During irradiation in a catheter-based IVBT procedure, a guide wire is frequently present inside the target region, which will inevitably perturb the dose distribution. Quantitative analyses of such perturbations are lacking in the current literature. The purpose of this work was to calculate the dosimetric perturbations caused by the presence of guide wires during the procedure of catheter-based IVBT using a Monte Carlo method. The study was performed for the three most commonly used  and ␥ sources, 90Sr, 32P, and 192Ir. The dependence of these dose perturbations on the composition and thickness of the guide wire was studied. The relative dosimetric merits of using  or ␥ sources for IVBT in the presence of a guide wire were discussed. METHODS AND MATERIALS The EGS4 Monte Carlo system (22) with the PRESTA algorithm (23) for electron transport was used in this project. This system has been extensively benchmarked against measurement in a variety of situations (22, 24), including IVBT (25, 26). The EGS4 user codes BEAM (24) and DOSXYZ (27) were employed. The component module SIDETUBE in BEAM code was used to simulate a cylindrical IVBT source. The BEAM code generates phase-space data, which include all the particle information, such as the charge, position, direction, energy, and history tag for each particle. The phase-space data were scored on a plane immediately outside the source capsule. A BEAM utility code, BEAMDP (24), was then used to analyze the phase-space data for beam characterization, which includes the spectra for each type of particle. These phase-space files were also used as the input data for the 3D dose calculation in water, with or without the presence of a guide wire, using the DOSXYZ code. The doses were calculated in 0.2 ⫻ 0.2 ⫻ 0.2 mm3 voxels. A dose perturbation factor (DPF), defined as the ratio of the doses with and without the presence of a guide wire, was introduced to quantify the effects. A dose enhancement is observed when DPF ⬎ 1.0. Guide wires of different composition, thickness, and shape were studied. The central axes of the source and
Volume 51, Number 4, 2001
the guide wire were parallel to one another and were placed 1 mm apart. This is a typical distance, because guide catheters with diameters 3.0 –7.0 Fr (1 Fr ⫽ 0.33 mm) are normally used in an IVBT procedure. BEAM code was used to calculate dose in one dimension (in the radial direction passing through the GW) for guide wires with either cylindrical or square shapes. The results showed that the DPF values at positions of 0.1 mm or farther from the guide wire are almost identical for the two shapes, provided that their cross-sections are equal. Because the DOSXYZ code takes Cartesian geometry as input, it was necessary to replace the cylindrical crosssection of the guide wire with a rectangular cross-section in the following calculations. Guide wires with crosssections of 0.2 ⫻ 0.2 and 0.3 ⫻ 0.3 mm2, which correspond to the cylindrical guide wires with diameters of 0.009 and 0.014 inches, respectively, were considered in this work. The guide wires were assumed to be made of either stainless steel or titanium. Three most commonly used IVBT sources, 90Sr (NOVOSTE Corporation), 32P (Guidant Vascular Intervention), and 192Ir (BEST Ind.) were studied. The compositions and sizes used in the Monte Carlo simulation for these three sources are tabulated in Table 1. The core material for the 90 Sr seed was taken as SiO2 with density of 3.0 g/cm3. The input -particle energy spectra (i.e., those from the bare nuclides) for the 90Sr/90Y source and the 32P source were reported by Cross et al. (28) and by Mourtada et al. (29), respectively. (Note that practical sources of 90Sr always have 90Y in equilibrium.) The ␥-particle energy spectrum for the 192Ir source was from the previous paper of Glasgow and Dillman (30). For the ␥ source, secondary electron transport was used in the Monte Carlo calculation. A sketch showing the source geometry and the location of a guiding wire is presented in Fig. 1. The 90Sr, 32P, or 192Ir sources
Fig. 1. Sketch of calculation geometry in the presence of a guide wire. A source consisting of a core and a capsule is shown.
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1105
Fig. 2. Energy spectra for (a) Novoste 90Sr/90Y beta source, (b) Guidant 32P beta source, and (c) BEST Ind. 192Ir gamma source. The spectra from the corresponding bare nuclides (i.e., the incident spectra for the Monte Carlo calculations) and the average energy for each spectrum are also included.
consist of a core and a capsule as shown in the figure. For more detailed characterizations of these sources, readers are referred to Refs. (25, 26, 29, 31). A set of EGS4 transport parameters, carefully chosen
for the small geometry and the low energies in question, was used in the calculation (e.g., AE ⫽ ECUT ⫽ 0.521 MeV, AP ⫽ PCUT ⫽ 0.01 MeV). The statistical uncertainty in the doses was less than 2.0% for the region less
Table 1. The simulated source configurations (materials and sizes) Core Source type 90
Sr P 192 Ir 32
Capsule
Materials
Outer diameter (mm)
Height (cm)
Materials
Outer diameter (mm)
Thickness (mm)
Height (cm)
Ceramic C2H6O2 70% Pt and 30% Ir
0.56 0.24 0.1
2.88 2.7 0.3
Stainless steel NiTi Stainless steel
0.64 0.46 0.5
0.04 0.11 0.2
2.88 2.7 0.3
1106
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Fig. 3. Dose perturbation factor versus radial distance in the presence of a stainless steel or titanium guide wire with a thickness of 0.009 inches. The data are calculated for 90Sr/90Y and 32P beta sources. The center of the sources and the position of the guide wire are shown.
than 3.0 mm from the source and less than 5% for the region in the range of 3 to 6 mm from the source. This requires up to 1011 incident particle histories to obtain statistical accuracy. RESULTS 90
90
Energy spectra of Sr/ Y, 32P, and 192Ir sources Because the energy spectra of the sources can be affected by the enclosing cores and capsules, it is necessary to use the spectra of the electrons and photons exiting the source capsules for the dose calculations. These spectra were derived from the phase-space data obtained in the plane immediately outside the source capsule by the Monte Carlo simulation. Figure 2 shows the energy spectra for (a) the Novoste 90Sr/90Y beta source, (b) the Guidant 32P beta source, and (c) the BEST 192Ir gamma source. The spectra from the bare nuclides (the incident spectra for the Monte Carlo calculations) and the average energy for each spectrum are also included in the figure. For the beta sources [Fig. 2 (a) and (b)], the electron energy spectra were altered significantly by the source cores and capsules. The fluence
of secondary photons generated by the source cores and capsules was also significant. For the 90Sr/90Y source, the average energy of exit beta particles was approximately 20% lower than that from the bare nuclides, whereas it was lower by 27% for the 32P source. For the 192Ir gamma source [Fig. 2 (c)], the energy spectrum for gamma particles is changed slightly (3% on average energy) by the source materials. The fluence of secondary electrons generated by the source structure was approximately 2 orders of magnitude lower than that of the photons. The average energy for these electrons was 0.115 MeV. Effects of guiding wires In Fig. 3, the DPF values are plotted as a function of the radial distance in the presence of a stainless steel or titanium guide wire having a 0.009⬙ diameter. The data were calculated for the 90Sr/90Y and 32P beta sources. The center of the sources and the position of the guide wire are shown in the figure. It is seen that the dose beyond the guide wire is reduced significantly. For the 32P source with the steel guide wire, this reduction is approximately 50% for a distance of 0.1 mm and 15–20% for distances farther than 1 mm from
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1107
Fig. 4. Dose perturbation factor versus radial distance in the presence of a stainless steel or titanium guide wire with a thickness of 0.014 inches. The data are calculated for 90Sr/90Y and 32P beta sources.
the guide wire. The dose reduction is in the range 10 –30% for the 90Sr/90Y source with the steel guide wire. Compared to steel, the titanium guide wire yields a 5–10% smaller dose reduction for both sources. In the region between the guide wire and the source, the dose near the guide wire is enhanced (DPF ⬎ 1). This enhancement, which can be as high as 10%, is primarily due to the backscatter from the GW. For IVBT, this region is normally inside the guide catheter. Thus, the dose enhancement should not be a concern. The values of DPF calculated for the 90Sr/90Y and 32P beta sources in the presence of a stainless steel or titanium guide wire of 0.014⬙ diameter are plotted in Fig. 4. For the steel guide wire, the dose reduction in the region beyond the guide wire was in the range 20 –55% or 25– 65% for the 90Sr/90Y or the 32P source, respectively. For the titanium guide wire, the dose reduction is 5–10% smaller compared with the steel GW for both beta sources. Because the DPF profiles in Figs. 3 and 4 transverse the central axis of the guide wires, the values can be considered the maxima for the 3D distributions. To see the dose perturbation in other directions, we have plotted in Fig. 5 the relative isodose distributions in two dimen-
sions for the 32P source, with or without (a) a steel guide wire of 0.009⬙, (b) a titanium guide wire of 0.009⬙, (c) a steel guide wire of 0.014⬙, and (d) a titanium guide wire of 0.014⬙. The solid and dashed lines represent isodose lines calculated with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The central axis of the source is located at (0, 0). The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. These isodose levels were arbitrarily normalized for presentation purposes. The position of the guide wire is indicated by the solid square in the figure. Because of the symmetry, only data in half of the plane are shown. Similar isodose distributions calculated for the 90Sr/90Y source are presented in Fig. 6. It is clear from Figs. 5 and 6 that the dose perturbation occurs in the region surrounding the guide wire, with the maximum perturbation occurring along the radial direction across the guide wire (the data shown in Figs. 3 and 4). This region can be extended over 1 mm from the guide wire in the radial direction. The region behind a guide wire, a portion of the IVBT target, is partially shielded from the irradiation. A similar calculation was performed for the 192Ir gamma
1108
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Fig. 5. Relative isodose distributions calculated for a 32P source with or without (a) a steel guide wire of 0.009 inches, (b) a titanium guide wire of 0.009 inches, (c) a steel guide wire of 0.014 inches, and (d) a titanium guide wire of 0.014 inches. The solid and dashed lines represent isodose lines for calculations with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. The central axis of the source is located at (0, 0). The position of the guide wire is indicated by the solid square. Because of symmetry, only data in half the plane are shown.
source. The dose perturbation was found to be statistically negligible for the guide wires studied (data not shown). DISCUSSION AND CONCLUSIONS The spectra for both primary and secondary particles (electrons or photons) exiting source structures were calculated with the aid of the Monte Carlo simulation. For the beta sources, the source cores and capsules affect the electron spectra, generating substantial secondary photons. The spectra of the exiting particles may be used for analytic or other Monte Carlo calculations for these sources without detailed simulations of their source structures. The dose perturbation because of the presence of guide wires was found significantly larger for the 90Sr/90Y and 32P beta sources than those for the 192Ir gamma source. Because of attenuation in the guide wires, a dose reduction behind the guide wire of up to 60% was observed for the beta sources. The dose perturbation was found to be negligible for the ␥ source. When the region behind the guide wire is part of the IVBT target, the presence of the guide wire may result in a significant underdosing for  sources. The underdosed region can extend a few mm behind the guide wire
and up to 1 mm in other directions. The effect may be ignored for the ␥ source. For a given  source, the dose perturbations by a guide wire depend upon the thickness and material of the guide wire. As expected, a thicker or denser guide wire leads to a higher dose perturbation. For a given thickness, the dose reduction calculated for titanium was 5–10% smaller than that for stainless steel. For a given material, the dose reduction for a 0.009⬙ guide wire was 10 –15% lower than that for a 0.014⬙ guide wire. For a given guide wire (thickness and material), a lower-energy source results in a more significant dose effect. The dose reduction for the 32P source was slightly larger than that for the 90Sr/90Y source for a given guide wire. In summary, the significant dose perturbation found for a beta source in the presence of a guide wire should be recognized in intravascular brachytherapy. If possible, the guide wire should be retracted from the treatment region during irradiation. If the guide wire has to be left inside the treatment region, the dose effects revealed in this work may be used in analyzing the treatment outcome and/or should be considered in the dose prescription. Such precautions are not necessary if using a gamma source.
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1109
Fig. 6. Relative isodose distributions calculated for a 90Sr/90Y source with or without (a) a steel guide wire of 0.009 inches, (b) a titanium guide wire of 0.009 inches, (c) a steel guide wire of 0.014 inches, and (d) a titanium guide wire of 0.014 inches. The solid and dashed lines represent isodose lines for calculations with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. The central axis of the source is located at (0, 0). The position of the guide wire is indicated by the solid square.
REFERENCES 1. Califf RM, Vetrovec GW, Kent KM, et al. Restenosis after coronary angioplasty: An overview. J Am Coll Cardiol 1991; 17:2B–13B. 2. Abbas MA, Afshari NA, Stadius ML, et al. Effect of x-ray irradiation on neointimal proliferation following balloon angioplasty. Clin Res 1993;41:79. 3. Bottcher HD, Schopohl B, Liermann D, et al. Endovascular irradiation—A new method to avoid recurrent stenosis after stent implantation in peripheral arteries: Technique and preliminary results. Int J Radiat Oncol Biol Phys 1994;29:183– 186. 4. Fischell TA, Kharma BK, Fishell DR, et al. Low-dose, beta particle emission from “stent” wire results in complete, localized inhibition of smooth muscle cell proliferation. Circulation 1994;90:2956 –2963. 5. Wiederman J, Leavy J, Amols H, et al. Effects of high dose intracoronary irradiation on vasomotor function and smooth muscle histopathology. Am J Physiol 1994;267:H125–H132. 6. Hehrlein C, Gollan C, Donges K, et al. Low-dose radioactive endovascular stent prevents smooth muscle cell proliferation and neointimal hyperplasia in rabbits. Circulation 1995;92: 1570 –1575. 7. Mayberg MR, Luo Z, London S, et al. Radiation inhibition of intimal hyperplasia after arterial injury. Radiat Res 1995;142: 212–220. 8. Popowski Y, Verin V, Papirov I, et al. High dose rate brachytherapy for prevention of restenosis after percutaneous transluminal coronary angioplasty: Preliminary dosimetric tests of
9.
10.
11.
12. 13.
14.
15.
16.
a new source presentation. Int J Radiat Oncol Biol Phys 1995;33:211–215. Condado JA, Waksman R, Gurdiel O, et al. Long-term angiographic and clinical outcome after percutaneous transluminal coronary angioplasty and intracoronary radiation therapy in humans. Circulation 1997;96:727–732. Waksman R, Robinson KA, Crocker IR, et al. Endovascular low-dose irradiation inhibits neointima formation after coronary artery balloon injury in swine: A possible role for radiation therapy in restenosis prevention. Circulation 1995;91: 1533–1539. King SB III, Waksman R, Grocker I, eds. Discoveries in radiation for restenosis. Atlanta: Emory University School of Medicine Press; 1995. Waksman R, Serruys P, eds. Handbook of vascular brachytherapy. 2nd ed. Dunitz; 2000. Nath R, Amols H, Coffey C, et al. Intravascular brachytherapy physics: Report of AAPM Radiation Therapy Committee Task Group No. 60. Med Phys 1999;26:119 –151. Amols HI, Zaider M, Weinberger J, et al. Dosimetric 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. Li XA, Wang R, Yu C, Suntharalingam M. BETA versus GAMMA for intravascular brachytherapy: Dosimetric perspectives in the presence of metallic stents and calcified plaques. Int J Radiat Oncol Biol Phys 2000;46:1043–1049. Ye SJ, Li XA, Zimmer JR, et al. Dosimetric perturbations of
1110
17. 18. 19. 20.
21. 22. 23.
I. J. Radiation Oncology
● Biology ● Physics
linear array of  emitters and metallic stents in intravascular brachytherapy. Med Phys 2000;27:374 –380. Fan P, Chiu-Tsao S-T, Patel NS, et al. Effect of stent and radiation dosimetry in an in-stent restenosis model. Cardiovasc Radiat Med 2000;2:18 –25. Amols HI, Trichter F, Weinberger J. Intracoronary radiation for prevention of restenosis: Dose perturbations caused by stents. Circulation 1998;98:2024 –2029. Nath R, Ning Y, Weinberger J. Dose perturbations by high atomic number materials in intravascular brachytherapy. Cardiovasc Radiat Med 1999;1:144 –153. Rahdert D, Sweet WL, Tio FO, et al. Measurement of density and calcium in human atherosclerotic plaque and implications for arterial brachytherapy. Cardiovasc Radiat Med 1999;1: 358 –367. Li XA. Dosimetric effects of contrast media for catheter-based intravascular brachytherapy. Med Phys: 2001;28:757–763. Nelson WR, Hirayama H, Rogers DWO. The EGS4 code system. SLAC-265 (Stanford Linear Accelerator Center, Stanford, CA), 1985. Bielajew AF, Rogers DWO. PRESTA: The parameter reduced electron-step transport algorithm for electron Monte Carlo transport. Nucl Instru Methods B 1987;18:165–181.
Volume 51, Number 4, 2001
24. Rogers DWO, Faddegon BA, Ding GX, et al. Beam: A Monte Carlo code to simulate radiotherapy treatment units. Med Phys 1995;22:503–524. 25. Soares CG, Halpern DG, Wang CK. Calibration and characterization of beta-particle sources for intravascular brachytherapy. Med Phys 1998;25:339 –346. 26. Wang R, Li XA. Monte Carlo calculation of a Sr/Y-90 and an Ir-192 sources for intravascular brachytherapy. Med Phys 2000;27:2528 –2535. 27. Ma CM, Reckwerdt P, Holmes M, Rogers DWO, Geiser B. DOSXYZ users manual. NRC Report PIRS 509b (1995). 28. Cross WG, Ing H, Freedman N. A short atlas of beta-ray spectra. Phys Med Biol 1983;28:1251–1260. 29. Mourtada FA, Soares CG, Seltzer SM, Lott SH. Dosimetry characterization of 32P catheter-based vascular brachytherapy source wire. Med Phys 2000;27:1770 –1776. 30. Glasgow GP, Dillman LT. Specific ␥-ray constant and exposure rate constant of 192Ir. Med Phys 1979;6:49 –52. 31. Wang R, Li XA. Monte Carlo dose calculations of betaemitting sources for intravascular brachytherapy: A comparison of EGS4, EGSnrc and MCNP. Med Phys 2001;28:134 – 141.
PII S0360-3016(01)01763-1
PHYSICS CONTRIBUTION
DOSE EFFECTS OF GUIDE WIRES FOR CATHETER-BASED INTRAVASCULAR BRACHYTHERAPY X. ALLEN LI, PH.D.,*
AND
ROMPIN SHIH, PH.D.*†
*Department of Radiation Oncology, University of Maryland, Baltimore, MD; †Department of Radiation Oncology, Tri-Service General Hospital & National Defense Medical College, Taipei, Taiwan, R.O.C. Purpose: Guide wires with high torquability and steerability are commonly used to navigate through a tortuous and/or branching arterial tree in a catheter-based intravascular brachytherapy procedure. The dosimetric effects due to the presence of metallic guide wires have not been addressed. This work investigates these dose effects for the three most commonly used  and ␥ sources (90Sr, 32P, and 192Ir). Methods and Materials: The EGS4 Monte Carlo codes were used to calculate the dose distributions for the 90 Sr(NOVOSTE), 32P (Guidant), and 192Ir (BEST Ind.) with and without a guide wire in place. Energy spectra for particles exiting the sources were calculated from the full phase-space data obtained from the Monte Carlo simulations of the source constructions. Guide wires of various thicknesses and compositions were studied. Results: The dose perturbations due to the presence of guide wires were found to be far more significant for the 90 Sr/90Y and 32P beta sources than those for the 192Ir gamma source. Because of the attenuation by the guide wires, a dose reduction of up to 60% behind a guide wire was observed for the beta sources, whereas the dose perturbation was found to be negligible for the ␥ source. For a  source, the dose perturbations depend on the thickness and the material of the guide wire. When the region behind a guide wire is part of an intravascular brachytherapy target, the presence of the guide wire results in a significant underdosing for  sources. The underdosed region can extend a few mm behind the guide wire and up to 1 mm in other directions. Conclusion: Significant dose perturbations by the presence of a metallic guide wire have been found in catheter-based intravascular brachytherapy using  sources. The dose effects should be considered in the dose prescription and/or in analyzing the treatment outcome for  sources. Such precautions are not necessary if using a gamma source. © 2001 Elsevier Science Inc. Intravascular brachytherapy, Dosimetry, Guide wire, EGS4 Monte Carlo.
for each specific patient” (13). However, the uncertainty in the three-dimensional (3-D) dosimetry of IVBT is generally high at the current time. This is because dosimetric measurements are difficult due to the extremely high dose gradient, small region of interest, and complicated source and patient geometry. Also, analytic and empiric dose calculations are inaccurate, because of the perturbations caused by delivery devices, heterogeneities, and scatter and selfabsorption of low-energy secondary radiation (13, 14). It has been reported that the dose distribution is perturbed by the presence of metallic stents (15–19), calcified plaques (15, 19, 20), and radiopaque contrast media (19, 21) used during the course of the treatment. These perturbations are found to be significant for beta sources and can be as high as 50%. During an angioplasty procedure, a steerable guide wire (GW) inserted into guiding catheter is advanced to
INTRODUCTION Intravascular brachytherapy (IVBT) has recently been recognized as a treatment modality for reducing coronary restenosis after angioplasty (1–10). Many clinical trials have already proven the efficacy of IVBT for in-stent restenosis, and several other clinical trials are exploring its efficacy for other lesions (11, 12). At this stage, the optimum dose and treatment target are not well understood. Accurate dosimetry is crucial in analyzing data from these trials to determine the most effective dose window and to improve our understanding of treatment failure. Without such accurate data, the derived dose–response relationship would contain flaws, and key information to improve the treatment would be missed. The recently published AAPM TG-60 protocol for IVBT physics states that “ideally a three-dimensional dose distribution over the entire irradiated volume should be generated
This investigation was supported in part by a grant from the National Institutes of Health (No. R1HL62213). Received Feb 5, 2001, and in revised form Jul 3, 2001. Accepted for publication Jul 12, 2001.
Reprint requests to: X. A. Li, Department of Radiation Oncology, University of Maryland, 22 South Greene St., Baltimore, MD 21201-1595. Tel: (410) 328-7165; Fax: (410) 328-5279; E-mail: [email protected] Acknowledgments—The authors are grateful to Dr. Mohan Suntharalingam for the useful discussion on the topic. 1103
1104
I. J. Radiation Oncology
● Biology ● Physics
pass the target region under fluoroscopic guidance. This guide wire is necessary, because balloon and other intracoronary catheters generally lack the torquability and steerability necessary to navigate through a tortuous and/or branching arterial tree. Even if intracoronary catheters had these properties, interventional procedures often require that catheters be placed across the lesion and withdrawn multiple times, so that a guide wire over which these catheters can be quickly and safely advanced saves time. A wide variety of guide wires is available for various situations and from various manufacturers. The GW lengths range from 175 to 300 cm. Their diameters are specified in inches, typically from 0.009 to 0.018 inches. Thicker wires are required for peripheral vessels. The guide wires are usually metallic (typically stainless steel). During irradiation in a catheter-based IVBT procedure, a guide wire is frequently present inside the target region, which will inevitably perturb the dose distribution. Quantitative analyses of such perturbations are lacking in the current literature. The purpose of this work was to calculate the dosimetric perturbations caused by the presence of guide wires during the procedure of catheter-based IVBT using a Monte Carlo method. The study was performed for the three most commonly used  and ␥ sources, 90Sr, 32P, and 192Ir. The dependence of these dose perturbations on the composition and thickness of the guide wire was studied. The relative dosimetric merits of using  or ␥ sources for IVBT in the presence of a guide wire were discussed. METHODS AND MATERIALS The EGS4 Monte Carlo system (22) with the PRESTA algorithm (23) for electron transport was used in this project. This system has been extensively benchmarked against measurement in a variety of situations (22, 24), including IVBT (25, 26). The EGS4 user codes BEAM (24) and DOSXYZ (27) were employed. The component module SIDETUBE in BEAM code was used to simulate a cylindrical IVBT source. The BEAM code generates phase-space data, which include all the particle information, such as the charge, position, direction, energy, and history tag for each particle. The phase-space data were scored on a plane immediately outside the source capsule. A BEAM utility code, BEAMDP (24), was then used to analyze the phase-space data for beam characterization, which includes the spectra for each type of particle. These phase-space files were also used as the input data for the 3D dose calculation in water, with or without the presence of a guide wire, using the DOSXYZ code. The doses were calculated in 0.2 ⫻ 0.2 ⫻ 0.2 mm3 voxels. A dose perturbation factor (DPF), defined as the ratio of the doses with and without the presence of a guide wire, was introduced to quantify the effects. A dose enhancement is observed when DPF ⬎ 1.0. Guide wires of different composition, thickness, and shape were studied. The central axes of the source and
Volume 51, Number 4, 2001
the guide wire were parallel to one another and were placed 1 mm apart. This is a typical distance, because guide catheters with diameters 3.0 –7.0 Fr (1 Fr ⫽ 0.33 mm) are normally used in an IVBT procedure. BEAM code was used to calculate dose in one dimension (in the radial direction passing through the GW) for guide wires with either cylindrical or square shapes. The results showed that the DPF values at positions of 0.1 mm or farther from the guide wire are almost identical for the two shapes, provided that their cross-sections are equal. Because the DOSXYZ code takes Cartesian geometry as input, it was necessary to replace the cylindrical crosssection of the guide wire with a rectangular cross-section in the following calculations. Guide wires with crosssections of 0.2 ⫻ 0.2 and 0.3 ⫻ 0.3 mm2, which correspond to the cylindrical guide wires with diameters of 0.009 and 0.014 inches, respectively, were considered in this work. The guide wires were assumed to be made of either stainless steel or titanium. Three most commonly used IVBT sources, 90Sr (NOVOSTE Corporation), 32P (Guidant Vascular Intervention), and 192Ir (BEST Ind.) were studied. The compositions and sizes used in the Monte Carlo simulation for these three sources are tabulated in Table 1. The core material for the 90 Sr seed was taken as SiO2 with density of 3.0 g/cm3. The input -particle energy spectra (i.e., those from the bare nuclides) for the 90Sr/90Y source and the 32P source were reported by Cross et al. (28) and by Mourtada et al. (29), respectively. (Note that practical sources of 90Sr always have 90Y in equilibrium.) The ␥-particle energy spectrum for the 192Ir source was from the previous paper of Glasgow and Dillman (30). For the ␥ source, secondary electron transport was used in the Monte Carlo calculation. A sketch showing the source geometry and the location of a guiding wire is presented in Fig. 1. The 90Sr, 32P, or 192Ir sources
Fig. 1. Sketch of calculation geometry in the presence of a guide wire. A source consisting of a core and a capsule is shown.
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1105
Fig. 2. Energy spectra for (a) Novoste 90Sr/90Y beta source, (b) Guidant 32P beta source, and (c) BEST Ind. 192Ir gamma source. The spectra from the corresponding bare nuclides (i.e., the incident spectra for the Monte Carlo calculations) and the average energy for each spectrum are also included.
consist of a core and a capsule as shown in the figure. For more detailed characterizations of these sources, readers are referred to Refs. (25, 26, 29, 31). A set of EGS4 transport parameters, carefully chosen
for the small geometry and the low energies in question, was used in the calculation (e.g., AE ⫽ ECUT ⫽ 0.521 MeV, AP ⫽ PCUT ⫽ 0.01 MeV). The statistical uncertainty in the doses was less than 2.0% for the region less
Table 1. The simulated source configurations (materials and sizes) Core Source type 90
Sr P 192 Ir 32
Capsule
Materials
Outer diameter (mm)
Height (cm)
Materials
Outer diameter (mm)
Thickness (mm)
Height (cm)
Ceramic C2H6O2 70% Pt and 30% Ir
0.56 0.24 0.1
2.88 2.7 0.3
Stainless steel NiTi Stainless steel
0.64 0.46 0.5
0.04 0.11 0.2
2.88 2.7 0.3
1106
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Fig. 3. Dose perturbation factor versus radial distance in the presence of a stainless steel or titanium guide wire with a thickness of 0.009 inches. The data are calculated for 90Sr/90Y and 32P beta sources. The center of the sources and the position of the guide wire are shown.
than 3.0 mm from the source and less than 5% for the region in the range of 3 to 6 mm from the source. This requires up to 1011 incident particle histories to obtain statistical accuracy. RESULTS 90
90
Energy spectra of Sr/ Y, 32P, and 192Ir sources Because the energy spectra of the sources can be affected by the enclosing cores and capsules, it is necessary to use the spectra of the electrons and photons exiting the source capsules for the dose calculations. These spectra were derived from the phase-space data obtained in the plane immediately outside the source capsule by the Monte Carlo simulation. Figure 2 shows the energy spectra for (a) the Novoste 90Sr/90Y beta source, (b) the Guidant 32P beta source, and (c) the BEST 192Ir gamma source. The spectra from the bare nuclides (the incident spectra for the Monte Carlo calculations) and the average energy for each spectrum are also included in the figure. For the beta sources [Fig. 2 (a) and (b)], the electron energy spectra were altered significantly by the source cores and capsules. The fluence
of secondary photons generated by the source cores and capsules was also significant. For the 90Sr/90Y source, the average energy of exit beta particles was approximately 20% lower than that from the bare nuclides, whereas it was lower by 27% for the 32P source. For the 192Ir gamma source [Fig. 2 (c)], the energy spectrum for gamma particles is changed slightly (3% on average energy) by the source materials. The fluence of secondary electrons generated by the source structure was approximately 2 orders of magnitude lower than that of the photons. The average energy for these electrons was 0.115 MeV. Effects of guiding wires In Fig. 3, the DPF values are plotted as a function of the radial distance in the presence of a stainless steel or titanium guide wire having a 0.009⬙ diameter. The data were calculated for the 90Sr/90Y and 32P beta sources. The center of the sources and the position of the guide wire are shown in the figure. It is seen that the dose beyond the guide wire is reduced significantly. For the 32P source with the steel guide wire, this reduction is approximately 50% for a distance of 0.1 mm and 15–20% for distances farther than 1 mm from
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1107
Fig. 4. Dose perturbation factor versus radial distance in the presence of a stainless steel or titanium guide wire with a thickness of 0.014 inches. The data are calculated for 90Sr/90Y and 32P beta sources.
the guide wire. The dose reduction is in the range 10 –30% for the 90Sr/90Y source with the steel guide wire. Compared to steel, the titanium guide wire yields a 5–10% smaller dose reduction for both sources. In the region between the guide wire and the source, the dose near the guide wire is enhanced (DPF ⬎ 1). This enhancement, which can be as high as 10%, is primarily due to the backscatter from the GW. For IVBT, this region is normally inside the guide catheter. Thus, the dose enhancement should not be a concern. The values of DPF calculated for the 90Sr/90Y and 32P beta sources in the presence of a stainless steel or titanium guide wire of 0.014⬙ diameter are plotted in Fig. 4. For the steel guide wire, the dose reduction in the region beyond the guide wire was in the range 20 –55% or 25– 65% for the 90Sr/90Y or the 32P source, respectively. For the titanium guide wire, the dose reduction is 5–10% smaller compared with the steel GW for both beta sources. Because the DPF profiles in Figs. 3 and 4 transverse the central axis of the guide wires, the values can be considered the maxima for the 3D distributions. To see the dose perturbation in other directions, we have plotted in Fig. 5 the relative isodose distributions in two dimen-
sions for the 32P source, with or without (a) a steel guide wire of 0.009⬙, (b) a titanium guide wire of 0.009⬙, (c) a steel guide wire of 0.014⬙, and (d) a titanium guide wire of 0.014⬙. The solid and dashed lines represent isodose lines calculated with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The central axis of the source is located at (0, 0). The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. These isodose levels were arbitrarily normalized for presentation purposes. The position of the guide wire is indicated by the solid square in the figure. Because of the symmetry, only data in half of the plane are shown. Similar isodose distributions calculated for the 90Sr/90Y source are presented in Fig. 6. It is clear from Figs. 5 and 6 that the dose perturbation occurs in the region surrounding the guide wire, with the maximum perturbation occurring along the radial direction across the guide wire (the data shown in Figs. 3 and 4). This region can be extended over 1 mm from the guide wire in the radial direction. The region behind a guide wire, a portion of the IVBT target, is partially shielded from the irradiation. A similar calculation was performed for the 192Ir gamma
1108
I. J. Radiation Oncology
● Biology ● Physics
Volume 51, Number 4, 2001
Fig. 5. Relative isodose distributions calculated for a 32P source with or without (a) a steel guide wire of 0.009 inches, (b) a titanium guide wire of 0.009 inches, (c) a steel guide wire of 0.014 inches, and (d) a titanium guide wire of 0.014 inches. The solid and dashed lines represent isodose lines for calculations with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. The central axis of the source is located at (0, 0). The position of the guide wire is indicated by the solid square. Because of symmetry, only data in half the plane are shown.
source. The dose perturbation was found to be statistically negligible for the guide wires studied (data not shown). DISCUSSION AND CONCLUSIONS The spectra for both primary and secondary particles (electrons or photons) exiting source structures were calculated with the aid of the Monte Carlo simulation. For the beta sources, the source cores and capsules affect the electron spectra, generating substantial secondary photons. The spectra of the exiting particles may be used for analytic or other Monte Carlo calculations for these sources without detailed simulations of their source structures. The dose perturbation because of the presence of guide wires was found significantly larger for the 90Sr/90Y and 32P beta sources than those for the 192Ir gamma source. Because of attenuation in the guide wires, a dose reduction behind the guide wire of up to 60% was observed for the beta sources. The dose perturbation was found to be negligible for the ␥ source. When the region behind the guide wire is part of the IVBT target, the presence of the guide wire may result in a significant underdosing for  sources. The underdosed region can extend a few mm behind the guide wire
and up to 1 mm in other directions. The effect may be ignored for the ␥ source. For a given  source, the dose perturbations by a guide wire depend upon the thickness and material of the guide wire. As expected, a thicker or denser guide wire leads to a higher dose perturbation. For a given thickness, the dose reduction calculated for titanium was 5–10% smaller than that for stainless steel. For a given material, the dose reduction for a 0.009⬙ guide wire was 10 –15% lower than that for a 0.014⬙ guide wire. For a given guide wire (thickness and material), a lower-energy source results in a more significant dose effect. The dose reduction for the 32P source was slightly larger than that for the 90Sr/90Y source for a given guide wire. In summary, the significant dose perturbation found for a beta source in the presence of a guide wire should be recognized in intravascular brachytherapy. If possible, the guide wire should be retracted from the treatment region during irradiation. If the guide wire has to be left inside the treatment region, the dose effects revealed in this work may be used in analyzing the treatment outcome and/or should be considered in the dose prescription. Such precautions are not necessary if using a gamma source.
Effects of guide wires in IVBT
● X. A. LI AND R. SHIH
1109
Fig. 6. Relative isodose distributions calculated for a 90Sr/90Y source with or without (a) a steel guide wire of 0.009 inches, (b) a titanium guide wire of 0.009 inches, (c) a steel guide wire of 0.014 inches, and (d) a titanium guide wire of 0.014 inches. The solid and dashed lines represent isodose lines for calculations with and without a guide wire, respectively. The dimensions on both the abscissa and ordinate are mm. The isodose lines, from the innermost to the outermost, correspond to 90%, 70%, 50%, 40%, 30%, 20%, and 10% levels. The central axis of the source is located at (0, 0). The position of the guide wire is indicated by the solid square.
REFERENCES 1. Califf RM, Vetrovec GW, Kent KM, et al. Restenosis after coronary angioplasty: An overview. J Am Coll Cardiol 1991; 17:2B–13B. 2. Abbas MA, Afshari NA, Stadius ML, et al. Effect of x-ray irradiation on neointimal proliferation following balloon angioplasty. Clin Res 1993;41:79. 3. Bottcher HD, Schopohl B, Liermann D, et al. Endovascular irradiation—A new method to avoid recurrent stenosis after stent implantation in peripheral arteries: Technique and preliminary results. Int J Radiat Oncol Biol Phys 1994;29:183– 186. 4. Fischell TA, Kharma BK, Fishell DR, et al. Low-dose, beta particle emission from “stent” wire results in complete, localized inhibition of smooth muscle cell proliferation. Circulation 1994;90:2956 –2963. 5. Wiederman J, Leavy J, Amols H, et al. Effects of high dose intracoronary irradiation on vasomotor function and smooth muscle histopathology. Am J Physiol 1994;267:H125–H132. 6. Hehrlein C, Gollan C, Donges K, et al. Low-dose radioactive endovascular stent prevents smooth muscle cell proliferation and neointimal hyperplasia in rabbits. Circulation 1995;92: 1570 –1575. 7. Mayberg MR, Luo Z, London S, et al. Radiation inhibition of intimal hyperplasia after arterial injury. Radiat Res 1995;142: 212–220. 8. Popowski Y, Verin V, Papirov I, et al. High dose rate brachytherapy for prevention of restenosis after percutaneous transluminal coronary angioplasty: Preliminary dosimetric tests of
9.
10.
11.
12. 13.
14.
15.
16.
a new source presentation. Int J Radiat Oncol Biol Phys 1995;33:211–215. Condado JA, Waksman R, Gurdiel O, et al. Long-term angiographic and clinical outcome after percutaneous transluminal coronary angioplasty and intracoronary radiation therapy in humans. Circulation 1997;96:727–732. Waksman R, Robinson KA, Crocker IR, et al. Endovascular low-dose irradiation inhibits neointima formation after coronary artery balloon injury in swine: A possible role for radiation therapy in restenosis prevention. Circulation 1995;91: 1533–1539. King SB III, Waksman R, Grocker I, eds. Discoveries in radiation for restenosis. Atlanta: Emory University School of Medicine Press; 1995. Waksman R, Serruys P, eds. Handbook of vascular brachytherapy. 2nd ed. Dunitz; 2000. Nath R, Amols H, Coffey C, et al. Intravascular brachytherapy physics: Report of AAPM Radiation Therapy Committee Task Group No. 60. Med Phys 1999;26:119 –151. Amols HI, Zaider M, Weinberger J, et al. Dosimetric 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. Li XA, Wang R, Yu C, Suntharalingam M. BETA versus GAMMA for intravascular brachytherapy: Dosimetric perspectives in the presence of metallic stents and calcified plaques. Int J Radiat Oncol Biol Phys 2000;46:1043–1049. Ye SJ, Li XA, Zimmer JR, et al. Dosimetric perturbations of
1110
17. 18. 19. 20.
21. 22. 23.
I. J. Radiation Oncology
● Biology ● Physics
linear array of  emitters and metallic stents in intravascular brachytherapy. Med Phys 2000;27:374 –380. Fan P, Chiu-Tsao S-T, Patel NS, et al. Effect of stent and radiation dosimetry in an in-stent restenosis model. Cardiovasc Radiat Med 2000;2:18 –25. Amols HI, Trichter F, Weinberger J. Intracoronary radiation for prevention of restenosis: Dose perturbations caused by stents. Circulation 1998;98:2024 –2029. Nath R, Ning Y, Weinberger J. Dose perturbations by high atomic number materials in intravascular brachytherapy. Cardiovasc Radiat Med 1999;1:144 –153. Rahdert D, Sweet WL, Tio FO, et al. Measurement of density and calcium in human atherosclerotic plaque and implications for arterial brachytherapy. Cardiovasc Radiat Med 1999;1: 358 –367. Li XA. Dosimetric effects of contrast media for catheter-based intravascular brachytherapy. Med Phys: 2001;28:757–763. Nelson WR, Hirayama H, Rogers DWO. The EGS4 code system. SLAC-265 (Stanford Linear Accelerator Center, Stanford, CA), 1985. Bielajew AF, Rogers DWO. PRESTA: The parameter reduced electron-step transport algorithm for electron Monte Carlo transport. Nucl Instru Methods B 1987;18:165–181.
Volume 51, Number 4, 2001
24. Rogers DWO, Faddegon BA, Ding GX, et al. Beam: A Monte Carlo code to simulate radiotherapy treatment units. Med Phys 1995;22:503–524. 25. Soares CG, Halpern DG, Wang CK. Calibration and characterization of beta-particle sources for intravascular brachytherapy. Med Phys 1998;25:339 –346. 26. Wang R, Li XA. Monte Carlo calculation of a Sr/Y-90 and an Ir-192 sources for intravascular brachytherapy. Med Phys 2000;27:2528 –2535. 27. Ma CM, Reckwerdt P, Holmes M, Rogers DWO, Geiser B. DOSXYZ users manual. NRC Report PIRS 509b (1995). 28. Cross WG, Ing H, Freedman N. A short atlas of beta-ray spectra. Phys Med Biol 1983;28:1251–1260. 29. Mourtada FA, Soares CG, Seltzer SM, Lott SH. Dosimetry characterization of 32P catheter-based vascular brachytherapy source wire. Med Phys 2000;27:1770 –1776. 30. Glasgow GP, Dillman LT. Specific ␥-ray constant and exposure rate constant of 192Ir. Med Phys 1979;6:49 –52. 31. Wang R, Li XA. Monte Carlo dose calculations of betaemitting sources for intravascular brachytherapy: A comparison of EGS4, EGSnrc and MCNP. Med Phys 2001;28:134 – 141.