Intra-arterial Beta Irradiation Prevents Neointimal Hyperplasia in a Hypercholesterolemic Rabbit Restenosis Model
Background Intra-arterial gamma irradiation has been shown to reduce restenosis after balloon angioplasty. The use of beta emitters should allow similar effects while inducing less undue tissue irradiation radioprotection problems.
Methods and Results Flexible 90-yttrium (90Y) coils inside a centering balloon were used to allow homogeneous intra-arterial dose delivery. One carotid and one iliac artery of 21 hypercholesterolemic rabbits were deendothelialized and then irradiated. Four dose schedules were studied: (1) control (dilated, nonirradiated); (2) 6 Gy; (3) 12 Gy; and (4) 18 Gy. Arterial specimens were histologically evaluated at 8 days and at 6 weeks. For all radiation doses at 8 days compared with controls, there was a significant decrease in bromodeoxyuridine-labeled cells (245±93 cells/cm in control, 42±27 in 6 Gy, 72±107 in 12 Gy, and 2±2 in 18 Gy groups; P<.001) and in total neointimal cells (891±415 cells/cm in control, 79±43 in 6 Gy, 192±264 in 12 Gy and 22±13 in 18 Gy groups; P<.0002). At 6 weeks, computer-derived histological percent area stenosis was reduced from 26±10% in the control group to 1±1.3% in the 18 Gy group (P<.0001), but lower doses had no significant effect.
Conclusions Administration of intra-arterial beta irradiation with a 90Y source is technically feasible and compatible with an ordinary catheterization laboratory environment. A dose of 18 Gy effectively induces long-term inhibition of neointimal hyperplasia.
Despite recent encouraging results,1 2 restenosis after percutaneous transluminal coronary angioplasty remains a largely unresolved problem in contemporary cardiology. Several recent studies have shown the efficacy of intra-arterial gamma irradiation in animal models of postangioplasty restenosis3 4 and in human femoral arteries.5 6 7 We developed a technique of endovascular beta irradiation using pure metallic 90-yttrium (90Y) sources.8 9 10 11 The three major advantages of beta sources in comparison with conventional gamma sources used for brachytherapy are as follows: (1) markedly steeper dose decline in tissue, hence significantly lower undue irradiation of surrounding periarterial structures; (2) much lower half-value layer in water and other tissue-equivalent media eliminating to a large extent the radioprotection problems and making their use compatible with a conventional catheterization laboratory; and (3) possibility of a high focal dose delivery in a short time interval (1 to 5 minutes for 20 Gy) with sources of relatively low activity (40 to 100 mCi).
We used a hypercholesterolemic rabbit model of postangioplasty restenosis to investigate the feasibility of intraluminal 90Y irradiation and the efficacy of different dose schedules in preventing neointimal proliferation.
Radiation Delivery System
Flexible coils with an outer diameter of 0.36 mm (0.014 in) and 24 mm in length were specially manufactured from titanium-coated pure yttrium wire of 0.11-mm diameter (Fig 1A⇓). Both tips of each coil were stretched over a length of 2 mm to form a miniature thread, allowing fixation to a thrust wire. This wire served for advancement, positioning, and withdrawal of the coil through the lumen of a balloon catheter previously placed in the artery.
To provide for a homogeneous intramural radiation dose delivery, a special device (centering balloon) was developed to assure optimal centering of the flexible 90Y source inside the arterial lumen (Schneider AG). This device (Fig 1B⇑) consists of a segmented balloon with four interconnected compartments mounted on an over-the-wire coronary angioplasty balloon shaft with an internal lumen diameter of 0.40 mm (0.016 in). The 10-mm-long flexible tip of a conventional 0.014-in angioplasty guide wire was fixed to the distal outlet of the balloon to ensure complete sealing and to allow steering of the system.
The coils were activated in the neutron flux of a nuclear reactor to yield the pure beta emitter 90Y (half-life, 64.1 hours; maximal energy, 2.284 MeV). Each coil had an approximate activity of 45 mCi after activation. All animal experiments were done within 10 days after source activation in the nuclear reactor.
Dosimetric evaluation in a tissue equivalent phantom showed good axial and circumferential dose homogeneity on the surface of the centering balloon, with a standard deviation of the mean not exceeding 8%.9 10
The dose distribution in tissue for 90Y seeds has been studied previously.12 Our dosimetric study showed a similar curve of the dose falloff as a function of depth in a tissue-equivalent medium for a linear source presentation.8 This curve allows for accurate estimation of depth doses from surface dose measurements: For example, for a radiation dose of 10 Gy at the balloon surface, the dose will be 4 Gy at 1 mm of tissue depth, 1.5 Gy at 2 mm, and 0.8 Gy at 3 mm.
All experiments were performed with the approval of the ethics committee for animal experimentation of the University Hospital of Geneva and the Cantonal Veterinary Office.
Twenty-one New Zealand White rabbits, 19 females and 2 males, 4 to 5 months of age, weighing 3.8±0.3 kg (range 3.2 to 4.5 kg), were used in the study.
Before the procedure and beginning from the age of 1 month, rabbits were fed a commercially available 2% cholesterol diet (Kliba) for 2 weeks followed by a 2-week period on a normal diet. This sequence was repeated 4 times, over a total of 16 weeks.13 After the procedure, a standard diet without cholesterol (Provimi) was fed to the animals.
Under general anesthesia with ketamine (35 mg/kg) and xylazine (5 mg/kg), the left femoral artery was surgically exposed in each animal. A transverse arteriotomy was performed distal to one of the major deep femoral branches. The modified Baumgartner technique of endothelial denudation was used to induce neointimal proliferation.14 A 3.5-mm-diameter centering balloon, 20 mm in length, was inserted into the lumen of the artery and advanced into the abdominal aorta. Under fluoroscopic guidance, one of the carotid arteries (19 right and 2 left) was catheterized, and its middle third was externally marked with a hypodermic needle. The centering balloon then was inflated with contrast medium (Omnipaque 240) with its proximal tip at the marked level and pulled back until its distal tip reached the mark. This was repeated three times, leading to complete deendothelialization of a 40-mm-long arterial segment, as confirmed by our preliminary experiments (unpublished data, 1994).
Afterward, the centering balloon was positioned with its center in the middle of the denuded arterial segment (level of the mark) and left inflated for 360 seconds in all arteries. During inflation, the 90Y source was advanced into the balloon and left in this position (Fig 2⇓) for the time necessary to apply the calculated proposed surface dose. The exposure time increased for each dose as a function of the time interval elapsed since source activation in the reactor. For example, for arteries receiving a dose of 18 Gy, exposure time varied between 105 seconds on day 2 and 322 seconds on day 9 (mean exposure time, 233±87 seconds).
The same procedure was performed in the left iliac artery. The fluoroscopically localized lumbosacral junction was used as an additional landmark of the distal balloon border position during the 360-second inflation.
In control arteries, all sequences of the procedure were performed except insertion of the 90Y source.
Radiation Schedules and Study End Points
The impact of 6 Gy, 12 Gy, and 18 Gy radiation doses administered simultaneously with balloon dilatation was studied. One carotid and one iliac artery were used in each animal, forming four study groups: (1) control group, 11 arteries (5 carotids and 6 iliacs); (2) 6 Gy group, 11 arteries (6 carotids and 5 iliacs); (3) 12 Gy group, 10 arteries (6 carotids and 4 iliacs); and (4) 18 Gy group, 10 arteries (5 carotids and 5 iliacs).
Ten animals were given an overdose of phenobarbital at 8 days and 11 at 6 weeks after intervention. A minimum of 5 vessels (3 carotids and 2 iliacs or vice versa) was obtained for each study group at each study end point.
Application of Bromodeoxyuridine
Bromodeoxyuridine (BrdU) was given to each animal 18 and 12 hours before excision of the vessels. As described by Hanke et al,15 16 100 mg/kg body wt BrdU and 75 mg/kg deoxycytidine (both from Sigma Chemie) were given as a subcutaneous neck depot 18 hours before the animals were killed. In addition to this, intramuscular injections (30 mg/kg BrdU and 25 mg/kg body wt deoxycytidine) were administered 18 and 12 hours before the animals were killed.
One hour before perfusion-fixation, the rabbits were infused intravenously with 60 mg/kg of Evans blue dye in PBS.17 18 After application of a lethal dose of phenobarbital, a thoracotomy was performed and the arteries were fixed in situ with perfusion of 500 mL of 10% neutral buffered formalin solution19 via a catheter inserted into the left ventricle.
After 15 minutes of perfusion-fixation, the vessels of interest were excised and immersion-fixed in the same fixative for 4 to 6 hours, dehydrated in graded ethanol, and embedded in paraffin.19 Two samples from the central deendothelialized region of each artery (stained blue) were taken for cross sectioning. In the arteries studied at 6 weeks, two additional samples were obtained at 10 mm proximal and distal to the center of the blue area (usually corresponding to blue-white boundaries). The two samples from the middle of the deendothelialized region were considered to represent the middle of the denuded and dilated arterial segment and hence the center of the irradiated zone. The proximal and distal samples were considered to correspond to the position of the distal and proximal balloon extremities during the 360-second inflation. Sections were cut 4 μm thick, mounted, and stained with hematoxylin and eosin and Verhoeff–van Gieson (elastic) stain.
A monoclonal antibody against BrdU (DAKO), streptavidin-biotin system (Amersham International), and combined staining with diaminobenzidine and hemalaun were used to identify the cells entering the S-phase of mitosis within 24 hours before the animals were killed.
To allow for quantification of smooth muscle cell (SMC) proliferation in neointima and media of arteries at 8 days, all cells in the intima and media were counted in two adjoining cross sections of each sample. The percentage of DNA synthesis in SMCs in the media and intima was determined as the ratio between BrdU-labeled cells and the total cell number.15 16
All cross sections were morphometrically analyzed with the help of a computer-based Sigma-Scan V 3.90 software (Jandel Scientific). The areas of intima, media, and residual lumen as well as circumferences demarcated by internal and external elastic lamina were determined.
To quantify the degree of neointimal migration and proliferation at 8 days, the following indexes were calculated: (1) number of all neointimal cells per centimeter of internal arterial circumference; (2) number of BrdU-positive neointimal cells per centimeter of internal arterial circumference; and (3) number of BrdU-negative neointimal cells per centimeter of internal arterial circumference. The extent of stenosis at 6 weeks was determined as % area stenosis=(intimal area/intimal area+lumen area)×100%. In addition, the number of intimal cell layers at 6 weeks was determined by counting the number of cell nuclei on a perpendicular line between endothelium and internal elastic membrane in four neointimal regions equidistant from one another.
Data are expressed as mean±SD. A one-way ANOVA was used to test for an overall treatment effect. Leven’s test was used to test for equality of variances. In case of unequal variances, the Welch test and Student’s t test assuming unequal variances were used. All probability values for pairwise comparison after a significant one-way ANOVA were corrected according to Bonferroni. For histological percent area stenosis, an arcsinus transformation was performed. Differences were considered significant at P<.05.
Twenty-one animals underwent interventions on 42 arteries. Two arteries (1 carotid after 18 Gy and 1 iliac from the control group) were found thrombosed at the time of histological examination and were excluded from analysis.
At 8 days, a marked and statistically significant decrease (P<.0002) in total neointimal SMC number percentimeter of inner arterial circumference was seen for all radiation doses as compared with control arteries (Figs 3⇓ and 4⇓). The same was true separately for BrdU-positive neointimal cells (P<.001).
The internal elastic membranes of three control arteries were found to be completely covered by several layers of SMC at 8 days, which was not the case in any of the irradiated arteries. The percentage of neointimal BrdU-labeled SMCs was 54±15% (range, 30% to 72%) in the 6 Gy group and 27±18% (range, 1% to 45%) in the 12 Gy group, which was not statistically different from the control group (33±23%; range, 8% to 66%). In contrast, the same index was 10±6% (range, 3% to 15%) in the 18 Gy group, significantly lower than in the control group (P<.05). The percentage of medial BrdU-labeled SMCs was 7±3% (range, 4% to 12%) in the 6 Gy group, 16±20% (range, 0.6% to 50%) in the 12 Gy group and 4±3% (range, 1% to 7%) in the 18 Gy group, which was not statistically different from the control group (9±11%; range, 3% to 27%).
At 6 weeks, histological percent area stenosis and number of neointimal cell layers (Fig 5⇓) were both significantly reduced (P<.0001 and P<.03, respectively) in the group of arteries exposed to 18 Gy compared with the control group (Fig 6⇓). There was no significant reduction of these indices in the 6 Gy group or in the 12 Gy group.
For arteries irradiated with 18 Gy, the neointimal formation was significantly reduced (P<.004) (Fig 7⇓) in the area corresponding to the middle of the balloon as compared with areas corresponding to proximal and distal balloon borders. Furthermore, the regions corresponding to the distal and proximal extremities of the irradiated zone showed a trend toward a higher degree of percent area stenosis and an increased number of neointimal cell layers in comparison with the middle of the dilated region of control arteries.
Radiation Delivery System
Endoluminal brachytherapy is a promising technique for prevention of restenosis after percutaneous angioplasty. However, the approaches recently reported with 192Ir afterloading techniques in both animal models3 4 and in human femoral arteries5 6 have several readily apparent shortcomings: (1) The amount of unduly irradiated periarterial tissue is far from negligible with 192Ir brachytherapy. Liermann et al5 reported a dose of 12 Gy at 3 mm, 5.5 Gy at 6 mm, and 3 Gy at 10 mm from the source in tissue. This was without adverse effect on magnetic resonance imaging of perivascular tissues, however.6
(2) Appropriate radioprotection is an important obstacle for use of this technology in the setting of a conventional catheterization laboratory. Liermann et al5 describe transporting patients from the catheterization laboratory to the afterloading room situated in another building, adding 45 minutes to the duration of the procedure. The use of low dose rate 192Ir afterloading3 can alleviate this problem to some extent but at the cost of a longer intra-arterial exposure time (about 60 minutes for 20 Gy).
(3) Poor centering of the source within the arterial lumen precludes homogeneous intramural dosage delivery, resulting in areas of both relative underdosage and overdosage with respect to the prescribed dose level.6 This could be of clinical relevance, considering the possibility that modest radiation doses (4 to 8 Gy) might have a stimulatory effect on neointimal proliferation,20 21 22 whereas higher doses (20 to 30 Gy) could lead to late vascular complications.23 24
Although the use of beta irradiation (32P) for neointimal hyperplasia inhibition has been reported recently by others,25 26 to the best of our knowledge, we report11 the first successful in vivo use of endovascular beta afterloading for prevention of postangioplasty restenosis. The described approach allows all three shortcomings of 192Ir endovascular brachytherapy to be circumvented to a large extent.8 9 In addition, the use of yttrium permits fabrication of miniature flexible sources compatible with conventional angioplasty balloons and suitable for high dose rate afterloading in tortuous vessels of small caliber, such as the coronary arteries.
Our study yields some insight into the dynamic process of cellular response of the arterial wall to different doses of homogeneously delivered beta irradiation administered simultaneously with balloon dilatation.
The initial inhibition of neointimal proliferation seen with all radiation schedules at 8 days was not confirmed at 6 weeks after intervention for arteries exposed to 6 and 12 Gy. A possible explanation for this phenomenon could be found in data obtained from in vitro studies of the cellular radiosensitivity of the aorta. It was shown by Fisher-Dzoga et al27 that the proliferative response of aortic medial SMCs was reduced only threefold by a radiation dose of 10 Gy. There is also evidence that partial recovery of irradiated SMCs occurs by the seventh postradiation day.27 The phenomenon of accelerated repopulation,28 well known to radiation oncologists, could theoretically even be responsible for an exaggerated cellular proliferation after low dose schedules. This could explain the putative stimulatory effect observed by others for radiation doses lower than 8 Gy,20 21 22 as well as the absence of long-term SMC inhibition at 6 or 12 Gy in our study.
The pattern of neointimal SMC inhibition reflects a significant decrease in cell migration from media to intima for all doses within 8 postradiation days. However, the percentage of neointimal BrdU-labeled cells after 6 and 12 Gy was not distinguishable from that of control cells. This would suggest that medial cells that succeed in traversing the internal elastic membrane, although reduced in number, have an intact potential for proliferation not distinguishable from that of control cells. In contrast, the 18 Gy dose was associated with a significant decrease in the percentage of BrdU-labeled medial SMCs, suggesting irreversible cell injury with reduced proliferative capacities.
Our study did not show a difference in the percentage of medial BrdU-labeled cells between irradiated arteries and control arteries at 8 days. A possible explanation for this is the observation that the internal elastic membranes of 3 control arteries were found to be completely covered by several neointimal SMC layers. Once it occurs, the maximum of SMC proliferation moves from media to intima, leaving only a very small number of proliferating cells in the media.17 18
The most striking finding in this study was the complete inhibition of neointimal hyperplasia at 6 weeks that was achieved in arteries treated with an 18 Gy radiation dose. The dynamics of neointimal progression in the animal model after arterial injury are now well established and show a maximal neointimal volume increase during the first 2 weeks after intervention, reaching a plateau phase by 4 to 6 weeks.15 18 This implies that an intramural arterial radiation dose of 18 Gy can be considered sufficient for induction of long-term inhibition of neointimal hyperplasia with the intra-arterial beta irradiation technique.
The increase at 6 weeks in the percent area stenosis and number of neointimal cell layers in regions corresponding to the borders of the irradiated zone may be of clinical relevance. Indeed, the intra-arterial longitudinal dose falloff along the long axis of the balloon, although very steep with a 90Y source, creates transitional areas of arterial wall irradiated with doses in an ineffective or even a potentially stimulatory range.20 21 22 In our experimental model, the injured arterial segment extended well beyond the irradiated zone, implying that a potential stimulatory effect of lower radiation doses was exerted over the injured regions of arterial wall. Taking into account that radiation doses of stimulatory range do not appear to induce neointimal proliferation in noninjured arterial segments,4 20 it seems reasonable to conclude from this observation that in clinical practice, the irradiated arterial segment should be at least as long as the dilated one and perhaps even slightly longer.
The modified Baumgartner technique of neointimal hyperplasia induction was used in hypercholesterolemic rabbits. This model allows for accurate quantification of neointimal proliferation at both early and late time points after the procedure, since the internal elastic lamina remains intact and precise measurements of all neointimal parameters are possible. However, the cellular response to any given dose of radiation observed using this model may well differ substantially from that of human arteries with atheromatous plaques. Moreover, the degree to which the hypercholesterolemic rabbit model can predict the rate of short-term and long-term complications after irradiation of human diseased coronary arteries is unknown.
The intra-arterially delivered beta irradiation did not appear to influence the rate of acute and subacute complications (thrombotic occlusions) in comparison with control arteries. Although such events are not well documented in clinical radiooncology, high single radiation doses focally delivered to the arterial wall raise concern of late vascular complications (such as thrombosis and aneurysm formation), potentially life-threatening in coronary arteries. Our study does not provide sufficient data nor was the observation period sufficiently long to address this issue. It has been demonstrated that medial necrosis and arteritis can occur in canine aortas and branch arteries at doses above 20 Gy.23 24 Thrombus formations covering more than 25% of intimal surface and dissecting aneurysms were observed in one half of the animals after 29 Gy and 32.5 Gy.23 24 Although these data derive from experiments involving irradiation of large tissular volumes and may have limited relevance for very localized endoarterial applications, they highlight the necessity to address this problem in future experiments with intra-arterial irradiation.
The Baumgartner model that was used led us to irradiate only part of the denuded arterial segment. This could in theory partly account for the lack of sustained suppression of the proliferative response at 6 weeks in the lower dose schedules. Other groups have found lower doses to be effective in preventing neointimal proliferation in models of overstretch trauma to pig coronary arteries and in human femoral arteries.4 5 However, direct comparison with our own data is difficult because of markedly diverging depth-dose distribution curves for beta and gamma sources.
Intra-arterial beta irradiation with pure metallic 90Y sources is feasible and compatible with the setting of an ordinary catheterization laboratory. Radiation doses between 6 and 18 Gy effectively inhibit neointimal SMC proliferation as assessed at 8 days in a hypercholesterolemic rabbit model of postangioplasty restenosis. This inhibitory effect is lost during the following posttreatment weeks in arteries that received radiation doses of 6 or 12 Gy. A radiation dose of 18 Gy effectively induces long-term inhibition of neointimal hyperplasia. The mild increase in neointimal proliferation in areas proximal and distal to the balloon suggests that the irradiated arterial segment should be at least as long as the injured (dilated) arterial segment and perhaps even slightly longer. The intra-arterially delivered beta irradiation does not appear to influence the rate of acute and subacute complications of the angioplasty procedure.
This study was supported in part by Schneider-Europe AG (Bülach, Switzerland). We are indebted to Jurgen Fingerle, PhD (Hoffmann-La Roche Ltd, Basel, Switzerland), and to Akiko and Robert P. Hof, MD (Sandoz Ltd, Basel, Switzerland) for their advice and assistance in the preparation of the animal experiments.
- Received December 28, 1994.
- Revision received April 24, 1995.
- Accepted May 18, 1995.
- Copyright © 1995 by American Heart Association
Waksman R, Robinson KA, Crocker IR, Gravanis MB, Cipolla GD, King SB III. Endovascular low-dose irradiation inhibits neointima formation after coronary artery balloon injury in swine. Circulation. 1995;91:1533-1539.
Liermann D, Böttcher HD, Kollath J, Schopohl B, Strassmann G, Strecker EP, Breddin KH. Prophylactic endovascular radiotherapy to prevent intimal hyperplasia after stent implantation in femoropopliteal arteries. Cardiovasc Int Radiol. 1994;17:12-16.
Liermann DD, Böttcher HD, Hermann G. Prophylactic endovascular radiotherapy of intimal hyperplasia in peripheral vessels. Cardiovasc Int Radiol. 1994;17(suppl 2):S97. Abstract.
Popowski Y, Verin V, Papirov I, Nouet P, Rouzaud M, Grob E, Schwager M, Urban P, Rutishauser W, Kurtz J. High dose rate brachytherapy for prevention of restenosis after percutaneous transluminal coronary angioplasty: preliminary dosimetric tests of a new source presentation. Int J Radiat Oncol Biol Phys. 1995;33:211-215.
Popowski Y, Verin V, Papirov I, Nouet P, Rouzaud M, Grob E, Schwager M, Urban P, Rutishauser W, Kurtz J. Intra-arterial 90-yttrium brachytherapy: preliminary dosimetric study using a specially modified angioplasty balloon. Int J Radiat Oncol Biol Phys. In press.
Verin V, Popowski Y, Urban P, Nouet P, Rouzaud M, Schwager M, Kurtz JM, Rutishauser W. Intraarterial beta irradiation for prevention of post angioplasty restenosis. J Invasc Cardiol. 1995;7:12A. Abstract.
Verin V, Popowski Y, Urban P, Belenger J, Redard M, Costa M, Widmer M-C, Schwager M, Kurtz JM, Rutishauser W. Intraarterial beta irradiation prevents neointimal hyperplasia in a hypercholesterolemic rabbit restenosis model. J Am Coll Cardiol. 1995;(suppl):2A-3A. Abstract.
Dutreix A, Marinello G, Wambersie A. Dosimétrie des émetteurs bêta. In: Dutreix A, Marinello G, Wambersie A. eds. Dosimétrie en curiethérapie. Paris, France: Masson; 1982:191-218.
Baumgartner HR. Eine neue Methode zur Erzeugung von Thromben durch gezielte Ueberdehnung der Gefaesswand. Z Ges Exp Med. 1963;137:227-230.
Hanke H, Strohscheider T, Oberhoff M, Betz E, Karsch KR. Time course of smooth muscle cell proliferation in the intima and media of arteries following experimental angioplasty. Circ Res. 1990;67:651-659.
Hanke H, Haas KK, Hanke S, Oberhoff M, Hassenstein S, Betz E, Karsch KR. Morphological changes and smooth muscle cell proliferation after experimental eximer laser treatment. Circulation. 1991;83:1380-1389.
Wiedermann JG, Marboe CH, Amols H, Schwartz A, Weinberger J. Intracoronary irradiation: minimal effective dose for prevention of restenosis in swine. Circulation. 1994;90(suppl I):I-59. Abstract.
Mazur W, Ali MN, Dabaghi SF, Crinstead C, Abukhalil J, Paradise P, DeFelice CA, Schulz D, Berner BM, Fajardo LF, French BA, Raizner AE. High dose rate intracoronary radiation suppresses neointimal proliferation in the stented and ballooned model of porcine restenosis. Circulation. 1994;90(suppl I):I-652. Abstract.
Fischell TA, Kharma BK, Fischell DR, Loges PG, Coffey CW, Duggan DM, Naftilan AJ. Low-dose, β-particle emission from ‘stent’ wire results in complete, localized inhibition of smooth muscle cell proliferation. Circulation. 1994;90:2956-2963.
Laird JR, Carter AJ, Kufs WM, Hoopes TG, Farb A, Nott S, Fischell RE, Fischell DR, Virmany R, Fischell TA. Inhibition of neointimal proliferation with a beta particle emitting stent. J Am Coll Cardiol. 1994;(suppl):287A. Abstract.
Hall EJ. Radiobiology for the radiologist. Philadelphia, Pa: JB Lippincott; 1994:219-221.