Treatment of Long, Diffuse, In-Stent Restenotic Lesions with Beta Radiation Using Strontium 90 and Sequential Positioning “Pullb

Diffuse, in-stent restenosis, characterized by lesions greater than 10 mm in length, is associated with revascularization rates ranging from 34% to more than 80% following conventional percutaneous coronary intervention (PCI).1 Lesion lengths extending beyond 30 mm are especially problematic and constitute a group of patients that benefit most from vascular brachytherapy.2 Five randomized, placebo-controlled trials have established that beta- and gamma-based vascular brachytherapy reduce the incidence of restenosis and clinical event rates following PCI for the treatment of in-stent restenosis.3–7 The treatment effect of any kind of radiation was over 30% for any analyzed parameter with a similar benefit for both beta emitters (Sr-90, P-32) and gamma emitters (Ir-192). The Novoste™ Beta-Cath™ System (Novoste Corporation, Norcross, Georgia) with the 30 or 40 mm source train is the only commercially available beta radiation system available to most institutions performing vascular brachytherapy in the United States. These source trains may not fully cover the vessel injury created during treatment of long, diffuse, in-stent restenosis. The safety and efficacy of manual stepping using a beta source was first reported at the American College of Cardiology in March 2000.8 The BETA Washington Radiation for In-stent Restenosis Trial (Beta WRIST) included 16 out of a total of 50 patients with diffuse in-stent restenosis in native coronary arteries who were treated with pullback of the catheter to lengthen the treatment zone. Target vessel revascularization at 6 months was reported to be 31.3% with no perforations or aneurysms noted at the junction. Between August 1999 and July 2000, twenty-three patients were treated with radiation using pullback of the available catheter for long, diffuse, in-stent restenotic lesions at the Montreal Heart Institute (MHI), and are the object of this report. The following is a description of the pullback technique as practiced at the MHI, a dosimetric evaluation of this technique and the clinical outcomes in this group of patients. METHODS Treatment device. The Beta-Cath™ System, which was approved for use in the United States on November 3, 2000 by the United States Food and Drug Administration, consists of two components.9 The first component is the Transfer Device, which hydraulically transports a 30 or 40 mm Sr-90/Y-90 beta radiation source train to and from the treatment zone. The second component is a 5 French, triple-lumen coronary delivery catheter designed to contain the radiation source train during the intra-coronary treatment. The radiation source train is comprised of 12, 16 or 24 individual sources measuring 2.5 mm in length and 0.64 mm in diameter. The 12 and 16 seed source trains are now available in the United States and Canada. At each end of the radiation source train, there is one radiopaque non-radioactive source to identify the position of the source train under fluoroscopy. Each source has an average individual source activity of 3.5 mCi and a typical dose rate of 8.5 cGy/second at 2 mm from the centerline axis of the source in water. The Delivery Catheter is inserted over a standard 0.014´´ guidewire and is positioned in the designated treatment area by the Interventional Cardiologist. The Radiation Oncologist injects sterile water through the transfer device to send the radioactive sources to the end of the catheter and to return the radioactive sources after the appropriate dwell time. Treatment prescription. The recommended dose prescription for a single application of radiation following PCI for the Beta-Cath™ System is 18.4 Gy at 2 mm for vessels whose reference vessel diameter (RVD) is Pullback technique. The extension of radiation treatment length beyond the length of the Beta-Cath™ System Radiation Source Train (RST) requires sequential positioning or “pullback” of the catheter. In an ideal situation, the proximal seed of the distal source train would be exactly juxtaposed to the distal seed of the proximal source train. Because precise positioning of a source train is not practical in a moving target, the recommended technique is to attempt to achieve a “one active source to one active source” overlap to ensure that an adequate dose of radiation is delivered to the entire treated segment. Figures 1–3 illustrate a typical example of such a procedure. The pullback technique is as follows: 1. It is recognized that accurate positioning of the RST may be difficult due to longitudinal movement of the Delivery Catheter, cardiac cycling and difficult visualization of anatomic landmarks. For greater accuracy in the RST positioning using the sequential positioning technique, a “Marker Wire” (Cordis Corporation, Miami Lakes, Florida) may be used (Figure 2). 2. Following PCI, position the distal marker of the marker wire to a referenced anatomical landmark within the coronary artery, ensuring that the distal marker of the marker wire extends > 10 mm beyond the distal end of the interventional injury length (Figure 2). 3. Evaluate the total interventional injury length by reviewing all recorded interventions. 4. Identify the most distal and proximal points of the injury according to any landmarks to determine the total interventional injury length in mm. The use of a marker guidewire allows for accurate evaluation of the injury length. a. For an injury length 40 mm and = 7.5 mm (>= 3 active sources) beyond the distal injury margin. Note that the area of “overlap” (proximal end of the RST in the distal position that overlaps the distal end of the RST in the proximal position) should be within an existing stent. 6. The RST is sent to the distal end of the Delivery Catheter and left in position the length of time required to deliver the desired dose of radiation (Figure 2A). Using contrast, record the position of the proximal RST marker relative to anatomic landmarks or a marker wire. At the end of the designated dwell time, return the RST to the Transfer Device. 7. The Delivery Catheter is withdrawn over the guidewire and positioned in the proximal portion of the coronary artery to be treated. Care is taken to position the catheter so that there will be a one source “overlap” in the junction between the distal and proximal RST position. Make sure that the overlap is within the existing stent and that the proximal radiation margin extends >= 7.5 mm beyond the proximal injury margin. The marker wire aids in positioning the distal seed of the proximal source train such that there is an overlap with the first proximal seed of the distal source train. (Figure 2B). When treating patients using a pullback method, the overlap area should remain constant (one active source to one active source overlap), and the margins can vary in length (7.5 mm or greater) (Figure 2C). 8. The RST is sent to the distal end of the Delivery Catheter and left to dwell in place to deliver the intended dose of radiation (Figure 2B). At the end of the designated dwell time, the RST is returned to the Transfer Device. 9. Following completion of treatment, the Delivery Catheter is removed and the treated area assessed (Figure 3A). The 6-month angiographic follow-up shows the persistence of the post-procedure result (Figure 3B). Dosimetry methods. iPlan™ is a PC-based vascular treatment planning system that has been previously validated.10 The calculation of dose surrounding the source train is based upon the recommendations of the AAPM TG-60 and is based on TG43 formalism. This software allows quantitative and qualitative evaluation of the dose distribution with a variety of radiation source trains (P-32, Ir-192, and Sr/Y-90). All dosimetric analyses were carried out with this software. The dose distributions were calculated for the following sequential radiation source train positions: 1) 2 mm overlap (Figure 4A); 2) 4 mm overlap (Figure 4B); 3) 2 mm gap (Figure 4C); and 4) 4 mm gap (Figure 4D). The dose was normalized at 2 mm from the centerline axis of the source train in water. The recommended dose prescription for the Novoste system is 18.4 Gy at 2 mm for smaller vessels (Clinical materials. From August 1999 through July 2000, twenty-three consecutive patients were treated for long, in-stent restenotic lesions using pullback at the Montreal Heart Institute in the Compassionate Use Registry. Specific institutional informed consent was obtained from all patients. RESULTS Effect of gap and overlap on dose distribution. Figures 4A–4D show the effect of increasing overlap and separation of the source trains on the dose distribution at 2 mm from the centerline of the source. For the 2 mm overlap position, the relative dose at 2 mm is 1.5 times the prescription dose. This increases to almost 1.8 times the prescribed dose with a 4 mm overlap. With even a 1 mm gap there is small volume, which receives almost 70% of the prescribed dose. This drops to approximately 10% of the prescribed dose with a 5 mm gap. Figure 5 shows the dose delivered to depths of 2 and 3 mm within an idealized smaller vessel (3 mm RVD) or a larger vessel (3.5 mm RVD) according to the Novoste and Guidant dose prescriptions with no overlap. The effect of an overlap of 1.25 or 2.5 mm in a 3 mm diameter vessel is shown in Figure 6. Longitudinal catheter movement. Longitudinal movement of the catheter tends to blur out the effect of hot spots at the junction.11 Source movement of 1 and 2 mm will result in 5% and 18% reduction of the dose enhancement factor (DEF) for a one source overlap. Source movement of 3 mm reduces the DEF by 24%. Thus, longitudinal source movement will reduce the peak dose. Clinical and angiographic outcomes. Mean age of the patients was 59 years. Fifteen patients (65%) were male; seven (30.4%) were diabetics. Mean lesion length was 49.4 ± 19.8 mm with a mean reference vessel diameter of 3.47 ± 0.5 mm. Seven patients (30%) received additional stents to achieve an optimal interventional result. Single pullbacks (2 irradiations) were performed in 21 patients and double pullbacks (3 irradiations) were performed in 2 patients. Out of the lesions treated, nine were in the left anterior descending coronary artery, four were in the circumflex artery, nine were in the right coronary artery, and one was a saphenous vein graft. Clinical follow-up information was available on all 23 patients at a mean follow-up of 7.0 ± 2.9 months. One patient died suddenly 75 days post-intervention. Another patient died from a non-cardiac cause (lung cancer) having no restenosis at the 6-month follow-up. Fourteen of the 23 patients were asymptomatic; three patients (3/23; 13%) underwent target vessel revascularization. Angiographic follow-up was obtained in 18 of the 23 patients at 6 months, at which point 7 patients (7/18; 38.8%) showed restenosis (> 50% luminal re-narrowing). There were no aneurysms or zones of ectasia in the treated arterial segments. On the baseline angiogram, sixteen of 25 junctions were found to have an overlap and 6 junctions had a gap. One junction was not assessable. The mean overlap was 5.61 ± 3.02 mm and the mean gap was 2.95 ± 2.32 mm. Two patients were found to have perfect alignment, defined as one active source to one gold marker overlap. None of the follow-up angiograms exhibited any thrombus, aneurysm formation or unhealed dissections at the site of overlap or gap. DISCUSSION This study examines a technique to extend the prescription dose length of a fixed length radiation source to treat vessel injury lengths that occur with the treatment of long, diffuse, in-stent restenosis. The technique appears to be safe from both a dosimetric and clinical point of view. Treatment with the Novoste system entails a dose prescription which is about half of what is currently recommended with the P-32 system from Guidant. If the entire treated segment can receive twice as much dose as is given with the Novoste device and show no evidence of negative clinical or angiographic outcomes as in the INHIBIT trial, then it is axiomatic that treatment of a small segment of vessel where the sources overlap and receive 1.8 times the prescription dose should also be well tolerated.7 In addition, in the INHIBIT trial, approximately 40% of the patients were treated with pullback without apparent deleterious effects.5 Waksman et al. reported treatment of long in-stent restenotic lesions from a clinical point of view in the Long WRIST trial.2 This trial was a randomized, placebo-controlled study in which long lesions were treated with gamma radiation. As seen in Figure 8, the average lesion length in these patients was 30.2 mm and at 6 months the target vessel revascularization rate was 60.7% in the control group and 38.3% in the treated group. The results of treatment using pullback at the MHI compare very favorably to the active treatment group in the Long WRIST trial (Figure 7). The recommended protocol for pullback using a marker wire was arrived at after confronting the difficulty in visually assessing the sequential positioning of the source train. A marker wire is particularly valuable in small vessels where diminished contrast flow impairs the ability to be guided by the position of sidebranches. This study is limited by virtue of the fact that there is no randomized control group, a small sample size, not all patients received angiographic follow-up, and the length of follow-up is limited. At this point in time, it would be difficult to mount a randomized study given the profound effect of radiation on both angiographic and clinical outcomes as observed in the Long WRIST trial and in this single-center experience. It is possible that further follow-up of these patients might reveal late effects not evident at the 6-month endpoint. Given the degree of benefit already observed and the lack of early adverse events, it is unlikely that late effects will negate the benefit of treating longer lesions using pullback. Based on the dosimetric, angiographic and clinical evaluation, pullback seems to be a safe and effective technique for the treatment of long, diffuse, in-stent restenotic lesions.

1. Mehran R, Dangas G, Abizaid AS, et al. Angiographic patterns of in-stent restenosis. Classification and implications for long-term outcome. Circulation 1999;100:1872–1878. 2. Waksman R, Elsayyad S, Mehran R, et al. High-dose intracoronary gamma radiation for patients with diffuse in-stent restenosis (The Long WRIST High Dose Trial) (Abstr). Circulation 2000;102:II-667. 3. Teirstein PS, Massullo V, Jani S, et al. Catheter-based radiotherapy to inhibit restenosis after coronary stenting. N Engl J Med 1997;336:1697–1703. 4. Waksman R, White RL, Chan RC, et al. Intracoronary g-radiation therapy after angioplasty inhibits recurrence in patients with in-stent restenosis. Circulation 2000;101:2165–2171. 5. Leon MB, Teirstein PS, Moses JW, et al. Localized intracoronary gamma-radiation therapy to inhibit the recurrence of restenosis after stenting. N Engl J Med 2001;344:250–256. 6. Popma J, Suntharalingam M, Lansky A, et al. A randomized trial of 90-Strontium/90-Yttrium beta radiation versus placebo control for the treatment of in-stent restenosis. Oral Presentation, ACCIS 2000, Anaheim, California. 7. Waksman R, Raizner A, Chiu K, et al. Beta radiation to inhibit recurrence of in-stent restenosis: Clinical and angiographic results of the multicenter, randomized double blind study. Circulation (online) 2000;102:e9046. 8. Waksman R, Ghargava B, Chan R, et al. Safety and efficacy of manual stepping and overlapping of a beta-emitter for diffuse in-stent restenosis lesions. J Am Coll Cardiol 2001;7(Suppl):1063–1113. 9. Sapirstein W, Zuckerman B, Dillard J. FDA approval of coronary artery brachytherapy. N Engl J Med 2001;344:297–299. 10. Fox T, Soares C, Crocker I, et al. Calculated dose distributions of beta-particle sources used for intravascular brachytherapy. Intern J Radiat Oncol 1997;39(Suppl):344. 11. Giap HB, Bendre DD, Huppe GB, et al. Source displacement during the cardiac cycle in coronary endovascular brachytherapy. Intern J Radiat Oncol Biol Phys 2001;49:273–277.