β-Radiation to Reduce Restenosis

With almost 1 million procedures undertaken worldwide each year, coronary angioplasty is flourishing. This elegant technique, however, continues to be plagued by a restenosis rate of 30% to 60%.¹ Efforts to reduce restenosis have been herculean. Literally, scores of pharmaceutical agents have been tested in clinical trials.²³⁴ Despite initial promising results in experiments with animals, each clinical trial has been a disappointment. Coronary stents stand alone as the only intervention that has been proved to decrease restenosis. In the STRESS and BENESTENT trials, the implantation of a single Palmaz-Schatz coronary stent was associated with a 30% reduction in restenosis rates.⁵⁶ The impact of stents on restenosis, however, is purely mechanical.⁷⁸ Stent implantation expands the vessel lumen further than the use of balloon angioplasty alone. This larger lumen creates more space for the still ubiquitous intimal proliferation. Stents do not diminish the cellular response to injury. In fact, the proliferative response, as measured on the basis of late loss after the procedure, is actually increased by the use of stents. Stents decrease restenosis by simply increasing the capacity of the artery to tolerate intimal proliferation. Despite extensive clinical testing, no agent has been shown to inhibit the proliferative component of restenosis.

Radiotherapy is the latest in a long line of potential antiproliferative agents to be enthusiastically tested as an adjunct to angioplasty. There is much to recommend the use of radiotherapy in the fight against restenosis. In more than 100 years of clinical experience, radiotherapy has proved to be highly effective in inhibiting cellular proliferation, in both malignant and benign disease. Examples of benign hyperplastic entities that have been effectively treated with radiotherapy include the exuberant fibroblastic activity of keloid scar formation, heterotopic ossification, desmoid/aggressive fibromatosis, Peyronie's disease, and pterygium.^910111213 In these benign proliferative disorders, doses of 700 to 1000 cGy in one treatment <72 hours after the stimulus have proved to be effective in inhibiting fibroblastic activity without significantly interfering with the normal healing process. Wiedermann et al¹⁴ and Waksman et al¹⁵ were the first to demonstrate significant reduction in intimal proliferation using radiotherapy in the swine model of restenosis. Subsequently, numerous groups demonstrated the efficacy of both γ- and β-radiation in various animal models of restenosis. Others have successfully inhibited neointimal proliferation with the use of β-emitting radioactive stents. Importantly, these animal models demonstrated efficacy without evidence of necrosis, significant fibrosis, or aneurysm formation. Inspired by these early preclinical trials and the encouraging clinical work of Bottcher et al,¹⁶ who applied radiotherapy to peripheral vessels, several clinical trials have been initiated to test this new treatment in the coronary circulation.

In this issue of Circulation, Verin and colleagues¹⁷ describe the first human coronary application of β-radiation to reduce restenosis. The trial is a small feasibility study and lacks a control group. The findings are disappointing. With the use of a standard definition of ≥50% diameter stenosis at 6-month follow-up, 6 of the 15 patients treated sustained restenosis. This 40% restenosis rate is very similar to historical restenosis rates without the use of radiotherapy. A more sensitive measurement of the impact of radiotherapy on preservation of the postprocedural luminal diameter is the “late loss index,” which is defined as late loss (in mm) divided by the acute gain (in mm). The late loss index is the “tax rate” associated with coronary intervention and refers to the percentage of increase in luminal diameter achieved at the initial procedure that is given up at follow-up. Historically, the late loss index usually ranges from 0.40 to 0.50.⁸ With the data presented in this trial (acute gain of 1.6 mm and late loss of 0.8 mm), we can estimate an unimpressive late loss index of 0.50; this is not very encouraging.

The radioactive source used in this trial was ⁹⁰Y, a β-emitter. The benefits of the use of a β-emitter include its ease of use and practical integration into current catheterization laboratory policies and procedures. Radiation energy from ⁹⁰Y, like other β sources, diminishes rapidly with distance. This isotope penetrates poorly and therefore does not require any special radiation shielding in the catheterization laboratory. The virtues of β-emitters that make them practical for use in the catheterization laboratory, however, also create challenges to the delivery of an effective dose to the coronary wall. Investigators have localized the cellular proliferation that follows balloon angioplasty to the media and even adventitia of the artery.¹⁸ The arterial wall of a diseased human coronary artery can easily be ≥2 mm thick in an artery containing eccentric plaque. The Figure illustrates an estimate of the dose-depth relation of a ⁹⁰Y linear source, with the assumption that the desired target dose is 18 Gy (the dose that was used in the present study and has been proven to be effective in numerous animal models of restenosis). Dosimetric evaluation very close to radioactive sources is difficult to perform. This figure is adapted from the preclinical work of Popowski et al,¹⁹ who are coauthors of the present study. When 18 Gy is prescribed to be administered to the luminal surface of the vessel wall, the dose delivered at a depth of 1 mm is only ≈5.4 Gy. At a depth of 2 mm, the dose is only ≈2.7 Gy. This represents only 15% of the prescribed dose. Therefore, despite the authors' use of a centering device, it is very difficult to determine the actual dose of radiation delivered to the region of the arterial wall containing actively proliferating cells. The authors' negative findings must therefore be viewed in light of the strong possibility that they administered a lower-than-intended radiation dose to the actively proliferating tissue elements. Furthermore, Waksman et al demonstrated potentiation of the effects of radiotherapy if treatment is given 48 hours after balloon injury. Another possible explanation for the authors' negative results is that the radiotherapy was given too soon after balloon angioplasty. Of course, any requirement for a repeat invasive procedure would severely limit the practical application of this technique.

Although the authors found no discernible impact on restenosis, they are to be commended for developing an elegant delivery system that effectively centers the radioactive source within the lumen. The study procedures were executed successfully and without complications. The authors' use of a segmented balloon-centering device raises numerous questions regarding the design of centering devices. One problem with centering is knowing where the center is. In a diseased coronary artery, plaque formation is usually eccentric. On the basis of histologic and intravascular ultrasound examination, the media and adventitia are usually disrupted and replaced by a heterogeneous collection of cholesterol, calcium, and fibrous tissue elements. The arterial wall is not round but instead contains numerous lumps, bumps, and bulges. This is quite different from the uniform, circular lumen found in animal models of restenosis. A catheter centered in the lumen, therefore, is not necessarily centered with respect to the arterial wall.

Another critical design question concerns the need for coronary perfusion during radiation delivery. The device used in this study does not provide perfusion. Given the high specific activity of the ⁹⁰Y emitter that was used, the prescribed dose was delivered in just 391±206 seconds. Despite this very short dwell time, the radioactive source had to be transiently withdrawn and the centering balloon deflated in 27% of patients due to the presence of intolerable ischemia. A centering device that allows perfusion would solve this problem and allow even longer dwell times. A second theoretical concern is the potential reduction in efficacy when radiotherapy is used in an ischemic environment. Tissue ischemia is associated with diminished radiotherapy kill rates when applied to other disease entities. In theory, radiotherapy is most effective when treating well-oxygenated, actively dividing cells.

Clearly, a dose-finding study with β-radiotherapy is required, as well as continued studies using other emitters. One potentially effective β-emitter is a radioactive, β-emitting coronary stent. Metal stents have been made radioactive by ion implantation of ³¹P beneath the metal surface followed by exposure of them to neutron irradiation to convert ³¹P into ³²P²⁰ or by activation of stainless steel stents in a cyclotron.²¹ The β-emitting radioactive stent is applied directly to the vessel wall, which may provide more favorable dosimetry. Other radiotherapy sources such as ¹⁹²Ir, a γ-emitter that provides more homogeneous dosimetry but is more complicated to use in the catheterization laboratory, are also under active investigation. The results of a double-blind, randomized clinical trial of the use of ¹⁹²Ir plus coronary stents for restenotic lesions (the SCRIPPS trial) have been reported.²²

Although the results of the present study raise questions about the effectiveness of the use of radiotherapy to reduce restenosis, we should not be pessimistic. With 100 years of clinical experience to draw on, we know that in radiotherapy, dose and timing are critical to success. The negative results of this trial are quite possibly the result of improper dosimetry and/or timing. In our quest for a vascular antiproliferative agent, radiotherapy remains high on the list of potentially efficacious treatments.

The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.

• Download figure
• Download PowerPoint

Footnotes