Gadolinium neutron capture brachytherapy (GdNCB), a new treatment method forintravascular brachytherapyShirin A. Enger, Arash Rezaei, Per Munck af Rosenschöld, and Hans Lundqvist

Citation: Medical Physics 33, 46 (2006); doi: 10.1118/1.2146050 View online: http://dx.doi.org/10.1118/1.2146050 View Table of Contents: http://scitation.aip.org/content/aapm/journal/medphys/33/1?ver=pdfcov Published by the American Association of Physicists in Medicine Articles you may be interested in Limitations of the TG-43 formalism for skin high-dose-rate brachytherapy dose calculations Med. Phys. 41, 021703 (2014); 10.1118/1.4860175 Monte Carlo parametric study of stent impact on dose for catheter-based intravascular brachytherapy with 90 Sr /90 Y Med. Phys. 31, 1964 (2004); 10.1118/1.1753431 Beta dose calculation in human arteries for various brachytherapy seed types Med. Phys. 30, 2561 (2003); 10.1118/1.1595871 Dose distributions in bifurcated coronary vessels treated with catheter-based photon and beta emitters inintravascular brachytherapy Med. Phys. 30, 1628 (2003); 10.1118/1.1582813 GAF film dosimetry of a tandem positioned -emitting intravascular brachytherapy source train Med. Phys. 30, 1004 (2003); 10.1118/1.1573206

Gadolinium neutron capture brachytherapy „GdNCB…, a new treatmentmethod for intravascular brachytherapy

Shirin A. Engera�

Division of Biomedical Radiation Sciences, Department of Oncology, Radiology and Clinical Immunology,Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden

Arash RezaeiDivision of Medical Radiation Physics, Department of Oncology-Pathology, Karolinska Institutet,SE-171 76 Stockholm, Sweden and Studsvik Medical AB, SE-612 82 Nyköping, Sweden

Per Munck af RosenschöldDepartment of Radiation Physics, Lund University Hospital, SE-22185 Lund, Sweden

Hans LundqvistDivision of Biomedical Radiation Sciences, Department of Oncology, Radiology and Clinical Immunology,Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden

�Received 27 June 2005; revised 4 November 2005; accepted for publication 7 November 2005;published 19 December 2005�

Restenosis is a major problem after balloon angioplasty and stent implantation. The aim of thisstudy is to introduce gadolinium neutron capture brachytherapy �GdNCB� as a suitable modality fortreatment of stenosis. The utility of GdNCB in intravascular brachytherapy �IVBT� of stent stenosisis investigated by using the GEANT4 and MCNP4B Monte Carlo radiation transport codes. To studycapture rate, Kerma, absorbed dose and absorbed dose rate around a Gd-containing stent activatedwith neutrons, a 30 mm long, 5 mm diameter gadolinium foil is chosen. The input data is a neutronspectrum used for clinical neutron capture therapy in Studsvik, Sweden. Thermal neutron capture ingadolinium yields a spectrum of high-energy gamma photons, which due to the build-up effectgives an almost flat dose delivery pattern to the first 4 mm around the stent. The absorbed dose rateis 1.33 Gy/min, 0.25 mm from the stent surface while the dose to normal tissue is in order of0.22 Gy/min, i.e., a factor of 6 lower. To spare normal tissue further fractionation of the dose isalso possible. The capture rate is relatively high at both ends of the foil. The dose distribution fromgamma and charge particle radiation at the edges and inside the stent contributes to a nonuniformdose distribution. This will lead to higher doses to the surrounding tissue and may prevent stentedge and in-stent restenosis. The position of the stent can be verified and corrected by the treatmentplan prior to activation. Activation of the stent by an external neutron field can be performed daysafter catherization when the target cells start to proliferate and can be expected to be more radiationsensitive. Another advantage of the nonradioactive gadolinium stent is the possibility to avoidradiation hazard to personnel. © 2006 American Association of Physicists in Medicine.�DOI: 10.1118/1.2146050�

Key words: gadolinium neutron capture brachytherapy, radioactive stent, intravascularbrachytherapy, restenosis

I. INTRODUCTION

Four years after the discovery of neutrons in 1932, G.L.Locher introduced the concept of Neutron Capture Therapy�NCT�.1 The physical principle of NCT is simple andelegant. It is a two component or binary system, based onthe nuclear reaction that occurs when a stable isotope is ir-radiated with thermal neutrons to yield highly energetic par-ticles.

Several nuclides have properties that make them candi-dates for NCT. Due to its relatively high cross section forthermal neutrons �3800 barn� and short range of the reactionproducts, 10B is the most studied nuclide in NCT.2 The ele-ment gadolinium is another candidate.3–7 The thermal neu-tron cross section of natural gadolinium �49 000 barns� is

157

dominated by Gd �254 000 barns�. Thermal neutron cap-

46 Med. Phys. 33 „1…, January 2006 0094-2405/2006/33

ture in 157Gd causes an exited state of 158Gd which decays tothe ground state through internal conversion �IC� by emis-sion of thousands cascade gamma rays.

Restenosis is a major complication after percutaneoustransluminal coronary angioplasty �PTCA�. The lesioncaused by PTCA triggers a complex molecular process thatstimulates a proliferative response �intimal hyperplasia� andremodeling of the vessel wall. This may lead to narrowing ofthe vessel area and restenosis within months afterangioplasty.8

Immediately after PTCA a metal stent may be placed in-side the lesion to hold the vessel open and act as a supportfor the vessel wall. Stents were believed to overcome rest-enosis problem and have been widely used. The implant ofstents do not decrease but, in fact, increase the proliferative

component of restenosis, which in a substantial fraction of

46„1…/46/6/$23.00 © 2006 Am. Assoc. Phys. Med.

47 Enger et al.: Gadolinium neutron capture brachytherapy 47

patients may lead to a so-called in stent restenosis �ISR�. Toprevent this undesired effect several methods have been sug-gested and tried, among them vascular brachytherapy.Catheter-based gamma and beta irradiation and beta emittingradioactive stents have been suggested as adjunct treatmentto reduce the incidence of ISR.9–13

In conventional brachytherapy, the target is 1–10 cm insize and the dosimetry of applied radiation sources is reason-ably well established. In intravascular brachytherapy�IVBT�, the target of interest is within the thin arterial walland at a distance less than a few millimeters from the source.At these distances, due to large dose gradients and problemswith centering of applied radiation sources, the dosimetry isdifficult. Complicated target geometry with heterogeneitiesdue to calcified plaque and vessel curvatures also adds to theproblems.14,15

In this paper the use of gadolinium neutron capturebrachytherapy �GdNCB� in IVBT for treatment of stenosis issuggested. GdNCB has several advantages that may over-come some of the described problems. Other advantages ofthe concept are that the stenting procedure is made with anonradioactive Gd-containing stent, which eliminates radia-tion hazard to personnel. The position of the stent can beverified and corrected post implantation by the treatmentplan. Activation of the stent can be performed by an externalneutron field a couple of days after catherization when thetarget cells start to proliferate and can be expected to bemore radiation sensitive. Fractionation of the total dose isalso possible which will spare healthy tissue. The radiationmay be repeated at any time when clinically needed. Theseadvantages may compensate for a somewhat larger irradiatedvolume due to the neutron field used to activate the gado-linium in the stent and limited access to epithermal tele-therapy neutron sources. All these potential clinical advan-tages have stimulated a study of Gd-containing stentsactivated with neutrons. The aim of the present study is toinvestigate if GdNCB is a suitable treatment modality forIVBT.

FIG. 1. Schematic picture of the beam filter and beam collimators at Studs-vik viewed from above. Our simulations include the last part of the beamline, i.e., a collimator delimiting a 14�10 cm2 field and an eight mm thick6Li filter �Region III�. A 15�20�20 cm PMMA phantom is placed in front

of the beam port.

Medical Physics, Vol. 33, No. 1, January 2006

II. MATERIALS AND METHODS

Simulated neutron and gamma spectra from an epithermalneutron beam built at the R2-0 reactor in Studsvik, Sweden,are used as input data in the Monte Carlo calculations. Thereactor geometry is divided into three regions.16 The usedspectra are taken from the end of region II, which means thatour simulations include the last part of the beam line, i.e., acollimator delimiting a 14�10 cm2 field and an eight mmthick 6Li filter as it is illustrated in Fig. 1. A 15�20�20 cm poly�methylmethacrylate� �PMMA� phantom isplaced in front of the beam port. The beam direction is alongthe x axis.

The diameter of the vessel lumen is 3–6 mm in the coro-nary arteries and up to 10 mm in peripheral arteries. In vas-cular brachytherapy the dose target is considered to be within1 mm from the vessel wall. However, calcifications andother inhomogeneities have a large influence on the dose-distribution. The dose may be delivered along the full lengthof the stent or it may be concentrated to the ends of the stentwhere the restenosis effect will clog the vessel. To studygadolinium containing stents the geometries described inTable I are simulated.

The Kerma from the prompt gamma produced by the neu-tron capture reactions in Gd is modeled. Low-energy elec-trons are not considered due to following reason. Besidegamma photons, the neutron capture in 157Gd produces con-version electrons of varying energies as well as Auger elec-trons and characteristic x rays.17 The range of electrons withenergy �70 keV is low and the abundance of conversionelectrons with energy �400 keV is only a few pro mille andthey will probably not contribute to the stent surface dose.The medium energy conversion electrons �abundance �6%�

TABLE I. Gadolinium-tubes used in the calculations.

Length�cm�

Diameter�mm�

Thickness��m�

Depth in the phantom�cm�

3 3 25 33 5 25 33 5 25 53 5 25 10

FIG. 2. The relative number of neutron captures in a cylindrical naturalGd-foil �diameter 5 mm, length 30 mm� with varying thickness. The Gd-tube is placed in a phantom at 3 cm depth and irradiated with an epithermal

field simulated with GEANT4.

48 Enger et al.: Gadolinium neutron capture brachytherapy 48

have a varying range depending of the stent material but willbe less than 40 �m. This implies that the electron contribu-tion to the GDNCB is negligible compare to the high-energygamma spectrum, especially in calcified regions.

Thermal neutrons will also cause nuclear reactions in nor-mal tissue mainly through 14N�n , p�14C and 1H�n ,��2H. Thecross sections of these reactions are small compared to thatof Gd, but the number of H and N atoms are considerablyhigher, and their contribution has to be considered in theevaluation. Gamma contamination of the neutron field due tothe production of the neutrons also contributes to the patientdose. The Kerma contribution from the beam port gamma,from the epithermal neutron beam and capture reactions gen-erated in the phantom are taken into account. The Kermadistributions are calculated in 100 slices, each one with a1 mm thickness in the beam direction �x axis�, 1 mm perpen-dicular to the beam direction �y axis� and 1 cm in the z axis.The Kerma is normalized to 1 for the highest value of thetotal Kerma.

The dose distribution for the average energy from gado-linium neutron absorption reaction �1.8 MeV� is compared to

FIG. 3. Relative capture rate in a 25 �m thick, 3 cm long Gd-foil with5 mm diameter. The depth 25 �m is the outer surface of the Gd-foil whiledepth 0 is the inner surface of the foil, i.e., the depth is reversed. Thecalculation is performed with GEANT4.

FIG. 4. Relative capture rate as a function of reverse foil depth for a 25 �m

thick, 3 cm long and 5 mm in diameter Gd-foil simulated with GEANT4.

Medical Physics, Vol. 33, No. 1, January 2006

dose distributions from two different mono-energetic �28 and375 keV� gamma radiations to show the difference betweenlow, intermediate, and high gamma energies. These energiesare chosen to represent commonly used radio nuclides forIVBT �125I and 192Ir�. The gamma radiations are sampleduniformly from the same geometry as the Gd-foil.

The Monte Carlo code MCNP18 is one of the most accurate

and commonly used codes in low energy neutron transportcalculations and is well investigated. All Kerma calculationsare performed with MCNP4B in this paper. The object-oriented Monte Carlo code GEANT4

19 has a flexible structureand openness that makes it preferable to work with. GEANT4

5.0 patch1 is used for all other calculations.

III. RESULTS

The simulations in Figs. 2–6 are performed with GEANT4

with a statistical uncertainty less than 5%. Most of the simu-lations in this study are performed for a 25 �m thick Gd-foil.For comparison a few simulations are carried out for theGd-foil thickness of 5 �m, which is a more probable thick-ness in an eventual application.

The relative number of thermal neutron captures for acylindrical Gd-foil �diameter 5 mm, length 3 cm� as a func-tion of thickness is given in Fig. 2. The number of captures

FIG. 5. Capture rate in a 5 �m thick, 3 cm long Gd-foil with 5 mm diameterirradiated with epithermal neutrons simulated with GEANT4. Also in thissimulation the depth is reversed.

FIG. 6. Relative radial dose distributions from high, intermediate, and low-gamma energy, normalized to 1 at 0.5 mm from the vessel wall simulated

with GEANT4.

49 Enger et al.: Gadolinium neutron capture brachytherapy 49

reaches a plateau already after 10 �m and a foil of 5 �mthickness yields about 80% of the total number of captures.

A �3D�-distribution of the neutron capture rate inside thewall of a 25 �m thick, 3 cm long Gd-tube with 5 mm diam-eter is given in Fig. 3. Normalization is made to the highestcapture value, which is at the outer surface of the Gd-tube.Capture rate is also relatively higher at both ends of the tube,where thermal neutrons had a high possibility to activateboth the outer and the inner Gd-tube surface.

The capture rate profile in a cross section of the sameGd-foil is presented in Fig. 4. From a high value on the outersurface there is a rapid decrease in capture rate. Already5 �m deep into the foil, i.e., at 20 �m reverse depth thecapture rate is as low as 30% compared to that of the outersurface, and reaches a minimum of 10% before it increasessomewhat to 20% in the inner surface.

A simulation of the capture rate in a 5 �m thick Gd-foilformed as a cylinder is showed in Fig. 5. The capture rate ishere more evenly distributed in the foil and is as high as 60%of the inner surface. This indicated that the epithermal neu-trons are thermalized when passing through the Gd-tube butalso in the water inside the cylinder.

The dose distribution for the average energy from gado-linium neutron absorption reaction �1.8 MeV� is compared todose distributions from two different mono-energetic �28 and375 keV� gamma radiations in Fig. 6. For the low and me-

FIG. 7. Relative Kerma distributions from different contributions for �a� a 3a 3 cm long Gd-tube with 5 mm diameter placed at 5 cm depth in the phanta 3 cm long Gd-tube with 3 mm diameter placed at 3 cm depth in the phan

dium energies the highest dose values are on the stent sur-

Medical Physics, Vol. 33, No. 1, January 2006

face, whereas for the gadolinium gamma spectra the ab-sorbed dose value is almost flat the first 4 mm around thestent surface.

To simulate Kerma from the Gd-tube as well as otherradiation contributions, Gd-tubes with two different diam-eters are placed inside a phantom at different depths. Rela-tive Kerma as the function of distance from a 3 cm longGd-foil with 5 mm diameter at 3 cm depth in the phantom ispresented in Fig. 7�a�. The same stent is simulated at 5 cmdepth and at 10 cm depth, illustrated in Figs. 7�b� and 7�c�.The relative Kerma distribution as the function of distancefrom a 3 cm long Gd-foil with 3 mm in diameter at 3 cmdepth is shown in Fig. 7�d�.

Kerma calculations in Figs. 7�a�–7�d� are simulated withMCNP4B and have a statistical uncertainty below 5%.Thickness of the Gd-tubes is 25 �m.

Dose rate �Gy/min� as the function of distance from stentsurface is presented in Fig. 8. The stent is placed at 3 cmdepth and has a thickness of 25 �m.

IV. DISCUSSION

The idea to use neutron activation of 157Gd has been con-sidered before but mainly in the context of tumor therapy.3–7

In this paper the emphasis has been on the application toIVBT, a new concept with many prospective advantages,

ong Gd-tube with 5 mm diameter placed at 3 cm depth in the phantom, �b�c� a 3 cm long Gd-tube with 5 mm diameter placed at 10 cm depth, and �d�All simulations are performed with MCNP.

cm lom, �tom.

which need extensive investigation.

50 Enger et al.: Gadolinium neutron capture brachytherapy 50

Two major irradiation principles for IVBT can be identi-fied, where one applies radiation beta or gamma sources,e.g., 90Sr, 125I, 192Ir, 32P, and 103Pd, for a short while duringthe stenting procedure. The other is to permanently implantradioactive stents. Comparison of different techniques aswell as gamma and beta sources in terms of dose rate, sourcecentring, penetration and scatter in the target region are dis-cussed in several papers.15,20,21

Compared to present techniques for IVBT, GdNCB offersseveral advantages. Due to the build-up factor caused by thehigh-energy photons involved the absorbed dose is almostflat the first 4 mm around the stent. Based on clinical datathe biological window of therapeutic efficiency without nor-mal tissue toxicity is in the range of 8–40 Gy to the surfaceof the vessel wall. The dose fall off as the function of dis-tance from the potential source should not exceed 40:8 or 5:1over a distance of 4 mm.13 The GdNCB-system satisfies thiscondition. For low-energy gammas, where absorption by tis-sue is greater than scatter, the dose drops faster than theinverse square of distance from the source. The handling ofnonradioactive material during the stent implantation alsoreduces radiation hazards both to the patient and the person-nel.

ISR and stent edge renarrowing are adverse effects asso-ciated with radioactive stents after conventional stent im-plantation. The cause of these effects is not completely un-derstood. Possible explanations are the stimulation of tissueproliferation by low-dose irradiation, by mechanical injury,or by a combination of both.22–24 Inadequate radiation cov-erage of the injured stent edge may be another cause of theedge effect phenomenon. Animal studies have shown thatextending the radiation margins from each end of the stentminimized the edge effect.25 Clinical trials treating in-stentand stent edge restenosis by gamma radiation �192Ir� havebeen successful.26–30 In GdNCB-system dose distributionfrom gamma and charge particle radiation at the edges andinside the stent contributes to a nonuniform dose distribution.This will lead to higher doses to the surrounding tissue andmay prevent stent edge and in-stent restenosis.

The stent can be loaded with gadolinium in differentways, either incorporated as an alloy in the stent or as a verythin, 5 �m foil �Fig. 2� between two stents. The dose from

gadolinium absorption reaction is high the first millimeters

Medical Physics, Vol. 33, No. 1, January 2006

from the Gd-foil. Photon dose from 1H�n ,��2H and gamma-contamination in the beam are the major contributors to thetotal dose the first centimeters of the phantom. Gd-tubesplaced at different depths �3, 5, and 10 cm� in the phantomhave a slight different Kerma distribution �Figs. 7�a�–7�c��.The relation between maximum Kerma and the entranceKerma increases from 0.47 at 3 cm depth to 0.56 at 5 cmdepth to 0.85 at 10 cm depth. Since thermal neutron peak intissue for the present beam standard is at 3 cm depth, theGd-tube placed at this position will yield higher amounts ofGd-associated gamma. At 5 and 10 cm depth the Gd-associated gamma are fewer but also have a longer pathwayto the surface and give a relatively higher contribution to theentrance Kerma.

NCT is simple in theory but in practise it demands a hightechnical quality of involved components and competence ofpersonal. Available facilities for NCT based on reactors cantoday deliver the wanted epithermal neutrons clean ofgamma and high energy neutrons contamination. An ab-sorbed dose rate of 1.33 Gy/min, 0.25 mm from the stentsurface is possible for 600 kW nominal reactor power whilethe dose to normal tissue is in order of 0.22 Gy/min �Fig. 8�,i.e., a factor 6 lower. To spare normal tissue further fraction-ation of the dose is also possible.

One limiting factor is the availability of neutron sourcesfor NCT clinical trials. World over, only a few reactors are atpresent used as NCT facilities. Most of these are separatedfrom hospitals, which mean practical problems to use themin clinical trials. Some research regarding the installation ac-celerator based NCT facilities at hospitals are made.31,32

Compared to reactors, the dose rate will be lower but theirradiation time will still be acceptable especially for the usein restenosis where the total dose is low compared to otherapplications.

V. CONCLUSION

GdNCB offers a number of advantages. The PTCA pro-cedure is to a high extent complicated by vascular brachy-therapy methods involving the handling with radioactivestents or with after-load technique with gamma emittingsources. With the Gd-system the stenting procedure is made

FIG. 8. Dose rate �Gy/min� as a function of distancefrom stent surface for 600 kW nominal reactor powersimulated with MCNP.

according to the normal protocol but with a nonradioactive

51 Enger et al.: Gadolinium neutron capture brachytherapy 51

Gd-loaded stent. After the PTCA-procedure the placing ofthe stent can be verified the next day using CT, and doseplanning can be made without stress. The irradiation can thenbe performed a couple of days after the catherization. Thereare also some more speculative radiobiological advantageswith the Gd-system. It is known that it takes some days afterthe incision before the cells start to proliferate. Proliferatingcells should be more radiosensitive than nonproliferating. Byadministering the irradiation a couple of days after the stentimplantation, the cell sensitivity may be considered. The Gd-system also allows fractionation. This will spare the healthytissue while increasing the chance of damaging proliferatingcells. Over all, these advantages may compensate for thesomewhat larger irradiated volume.

a�Author to whom correspondence should be addressed. Telephone: �46-18-4713829; fax: �46-18-4713432. Electronic mail: shirin@enger.se

1G. L. Locher, “Biological effects and therapeutic possibilities of neu-trons,” Am. J. Roentgenol. Radium Ther. 36, 1–13 �1936�.

2J. A. Coderre and G. M. Morris “The radiation biology of Boron neutroncapture therapy,” Radiat. Res. 151, 1–18 �1999�.

3D. P. Gierga, J. C. Yanch, and R. E. Shefer, “An investigation of thefeasibility for neutron capture synovectomy,” Med. Phys. 27, 1685–1692�2000�.

4J. T. Masiakowski, J. L. Horton, and L. J. Peters. “Gadolinium neutroncapture therapy for brain tumors: A computer study,” Med. Phys. 19,1277–1284 �1992�.

5J. A. Shih and R. M. Brugger, “Gadolinium as a neutron capture therapyagent,” Med. Phys. 19, 733–744 �1992�.

6T. Goorley and H. Nikjoo, “A comparison of three gadolinium basedapproaches to cancer therapy,” Radiat. Res. 154, 556–563 �2002�.

7T. Goorley, W. S. Kiger, and R. G. Zamenhof, “Reference dosimetrycalculations for neutron capture therapy with comparison of analyticaland voxel models,” Med. Phys. 29, 145–156 �2002�.

8A. Lövqvist, H. Emanuelsson, J. Nilsson, H. Lundqvist, and J. Carlsson,“Pathophysiological mechanisms for restenosis following coronary angio-plasty: possible preventive alternatives,” J. Intern Med. 233, 215–226�1993�.

9C. Janicki, D. M. Duggan, and A. Gonzales, “Dose model for a beta-emitting stent in a realistic artery consisting of soft tissue and plaque,”Med. Phys. 26, 2451–2460 �1999�.

10N. Reynaert, F. Verhaegen, Y. Taeymans, M. V. Eijkeren, and H. Thierens,“Monte Carlo calculations for dose distribution around 32P and 198Austents for intravascular brachytherapy,” Med. Phys. 26, 1484–1491�1999�.

11H. I. Amols, M. Zaider, J. Weinberger, R. Ennis, P. B. Schiff, and L. E.Reinstein, “Dosimetric considerations for catheter-based beta and gammaemitters in the therapy of neointimal hyperplasia in human arteries,” Int.J. Radiat. Oncol., Biol., Phys. 36, 913–921 �1996�.

12N. S. Patel, S. T. Chiu-Tsao, Y. Ho, T. Duckworth, J. S. Shih, H. S. Tsao,H. Quon, and L. B. Harrison, “High beta and electron dose from 192Ir:Implications for “gamma” intravascular brachytherapy,” Int. J. Radiat.Oncol., Biol., Phys. 54, 972–980 �2002�.

13H. I. Amols, “Review of endovascular brachytherapy, Physics for preven-tion of restenosis,” Cardiovasc. Radiat. Med. 1, 64–71 �1999�.

14D. A. Rahdert, W. L. Sweet, F. O. Tio, C. Janicki, and D. M. Duggan,“Measurement of density and calcium in human atherosclerotic plaqueand implications for arterial brachytherapy,” Cardiovasc. Radiat. Med. 1,

358–367 �1999�.

Medical Physics, Vol. 33, No. 1, January 2006

15R. Nath, H. Amols, C. Coffey, D. Duggan, S. Jani, Z. Li, M. Schell, C.Soares, J. Whiting, P. E. Cole, I. Crocker, and R. Schwartz, “Intravascularbrachytherapy physics: Report of the AAPM Radiation Therapy Commit-tee Task Group No. 60,” Med. Phys. 26, 119–152 �1999�.

16V. Giusti, P. M. Munck af Rosenschold, K. Skold, B. Montagnini, and J.Capala, “Monte Carlo model of the Studsvik BNCT clinical beam: de-scription and validation,” Med. Phys. 30, 3107–3117 �2003�.

17J. Stepanek, “Emission spectra of gadolinium-158,” Med. Phys. 30,41–43 �2003�.

18X-5 Monte Carlo Team, MCNP—A General Monte Carlo N-Particle Trans-port Code, Version 5, LA-UR-03-1987, Los Alamos National laboratory�2003�.

19S. Agostinelli et al., “GEANT4—a simulation toolkit,” Nucl. Instrum.Methods Phys. Res. A 506, 250–303 �2003�.

20X. A. Li, R. Wang, C. Yu, and M. Suntharlingham, “Beta versus gammafor catheter-based intravascular brachytherapy: Dosimetric perspectivesin the presence of metallic stents and calcified plaques,” Int. J. Radiat.Oncol., Biol., Phys. 46, 1043–1049 �2000�.

21X. A. Li and R. Shih, “Dose effects of guide wire for catheter-basedintravascular brachytherapy,” Int. J. Radiat. Oncol., Biol., Phys. 51,1103–1110 �2001�.

22P. W. Serruys and I. P. Kay, “I like candy, I hate the wrapper The 32Pradioactive stent” Circulation 101, 3–7 �2000�.

23W. J. Van der Giessen et al., ““Edge effect” of 32P Radioactive stent iscaused by the combination of chronic stent injury and radioactive dosefalloff,” Circulation 104, 2236–2241 �2001�.

24R. Albiero, M. Adamian, N. Kobayashi, A. Amato, M. Vaghetti, C. DiMario, and A. Colombo, “Short- and intermediate-term results of 32Pradioactive �-emmiting stent implantation in patients with coronary ar-tery disease,” Circulation 101, 18–26 �2000�.

25E. Cheneau, R. Waksman, H. Yazdi, R. Chan, J. Fourdnadjiev, C. Berz-ingi, V. Shah, A. E. Ajani, L. Leborgne, and F. O. Tio, “How to fix theedge effect of catheter-based radiation therapy in stented arteries,” Circu-lation 106, 2271–2277 �2002�.

26P. S. Teirstein, V. Massullo, S. Jani, and P. Tripuraneni, “Initial studieswith gamma radiotherapy to inhibit coronary restenosis,” Cardiovasc. Ra-diat. Med. 1, 3–7 �1999�.

27T. Hagenaars, I. F. A. Po, M. R. van Sambeek, V. L. Coen, R. B. VanTongeren, F. M. Gescher, C. H. Wittens, R. U. Boelhouwer, P. M. Pat-tynama, and E. J. Gussenhoven, “Gamma radiation induces positive vas-cular remodeling after balloon angioplasty: A prospective, randomizedintravascular ultrasound scan study,” J. Vasc. Surg. 36, 318–324 �2002�.

28R. Waksman, B. Bhargava, R. C. Chan, W. Sherman, J. Pisch, G. S.Mintz, A. J. Lansky, J. Ahmed, N. A. Ricci, and S. F. Liprie, “Intracoro-nary radiation with gamma wire inhibits recurrent in-stent restenosis,”Cardiovasc. Radiat. Med. 2, 63–68 �2001�.

29R. Waksman, A. E. Ajani, R. L. White, R. C. Chan, L. F. Satler, K. M.Kent, A. D. Pichard, E. E. Pinnow, A. B. Bui, S. Ramee, P. Teirstein, andJ. Lindsay, “Intravascular gamma radiation for in-stent restenosis insaphenous-vein bypass grafts,” N. Engl. J. Med. 346, 1194–1199 �2002�.

30M. A. Grise, V. Massullo, S. Jani, J. J. Popma, R. J. Russo, R. A. Schatz,E. M. Guarneri, S. Steuterman, D. A. Cloutier, M. B. Leon, P.Tripuraneni, and P. S. Teirstein, “Five-year clinical follow-up after intra-coronary radiation: Results of a randomized clinical trial,” Circulation105, 2737–2740 �2002�.

31D. J. Lee, C. Y. Han, S. H. Park, and J. K. Kim, ”An accelerator-basedepithermal neutron beam design for BNCT and dosimetric evaluationusing a voxel head phantom,” Radiat. Prot. Dosim. 110, 655–60 �2004�.

32A. E. Hawk, T. E. Blue, and J. E. Woollard, ” A shielding design for anaccelerator-based neutron source for boron neutron capture therapy,”

Appl. Radiat. Isot. 61, 1027–1031 �2004�.