Two-Year Angiographic Follow-Up of Intracoronary Sr90 Therapy for Restenosis Prevention After Balloon Angioplasty
Background— Postcoronary angioplasty vascular brachytherapy (VBT) has emerged as a successful intervention for restenosis prevention in some clinical scenarios. Longer-term follow-up after VBT in de novo nonstented lesions has not been reported.
Methods and Results— Thirty patients treated with post–percutaneous transluminal coronary angioplasty (PTCA) VBT with Sr90 underwent clinical and angiographic follow-up at 6 and 24 months. Specific vessel segment quantitative coronary angiographic analyses were performed to identify radiation edge effects. Nineteen patients who had not undergone index procedure stenting or target vessel revascularization (TVR) over the 2-year period were analyzed separately. Of the 30 patients, 3 underwent TVR by 6-month follow-up. An additional 4 patients required TVR between 6 and 24 months. In the total cohort of 26 patients undergoing angiographic follow-up at 6 and 24 months, an increase in minimal lumen diameter of the initial target segment was noted at 6 months compared with postprocedure analysis (2.31±0.48 versus 2.04±0.43 mm, P<0.05). At 24 months, this was no longer significant (2.19±0.61 mm). In the proximal segments of the entire cohort and the nonintervened subgroup, the principal late loss occurred over the first 6 months with no additional late loss at 2-year follow-up. The distal segments remained stable over the entire follow-up period.
Conclusions— Although some late failures of post-PTCA VBT are seen between 6 and 24 months, most treated vessels remain stable with no late loss or additional luminal increase beyond the 6-month period. This suggests that late aneurysm formation and significant late edge restenosis are unlikely in VBT after PTCA of de novo lesions for up to 2 years.
Received October 24, 2001; revision received May 4, 2002; accepted May 10, 2002.
Vascular brachytherapy has emerged as a successful intervention for the prevention of restenosis in certain clinical scenarios.1–4⇓⇓⇓ This has been demonstrated at 6 months in several studies both with β and γ emitters. These positive early results must be seen within the perspective of potential late reduction of efficacy and complications, including late occlusion, restenosis, perforation, aneurysm formation, and persistent dissection.5–8⇓⇓⇓ One study has reported the effects at 3 years of a γ emitter in patients with restenotic stented arteries.5 In this study, the 2-year clinical and angiographic outcomes of patients with de novo lesions treated with postangioplasty β radiation are reported.
The Beta Energy Restenosis Trial (BERT) was a 4-center, open-registry study in which patients with de novo coronary artery stenoses that had been successfully treated with balloon angioplasty were then treated with vascular brachytherapy using a 30-mm train of Sr/Y90 seeds to prevent restenosis. The methods have been described previously.9 The protocol was approved by the institutional review board of each participating institution. Written informed consent was obtained from each patient before enrollment and randomization in the trial.
Patients were eligible for inclusion in the study if they were aged between 18 and 80 years, had angina or proven ischemia on laboratory testing, and were scheduled to undergo planned balloon angioplasty of a single coronary lesion in a native coronary artery. The target lesions were required to have between 60% and 99% diameter stenosis. Patients were excluded for evidence of a myocardial infarction within 3 days before the procedure; contraindication to aspirin; left ventricular ejection fraction <40%; serum creatinine >2.0 mg/dL; anticipated difficulty with follow-up; life-threatening coexisting illness; child-bearing potential (women); severe peripheral vascular disease; unprotected left main coronary artery disease; presence of thrombus; past chest radiotherapy; lesion angled >45 degrees; intraprocedural angiographic evidence of thrombus, spasm, or dissection; or unsatisfactory PTCA result requiring stent implantation.
After baseline coronary angiography, PTCA was performed according to standard clinical practice. Fifteen minutes after successful PTCA, as determined by residual stenosis of <50% and an increase of lumen diameter of >20%, patients were randomized to receive 12, 14, or 16 Gy as calculated at 2 mm from the center of the source. A 5F delivery catheter (Beta Cath, Novoste Corp) was positioned at the site of PTCA with 2 markers separated by 30 mm used for positioning of the catheter across the angioplasty site. The guidewire was withdrawn, and a 30-mm-long train of 90Strontium/Yttrium (90Sr/Y) seeds was deployed to the treatment site between the markers for the period determined to deliver the prescribed dose at 2 mm. The specifics of the radiation procedure have been described previously.10
All 30 patients included at the Montreal Heart Institute were requested to return for repeat coronary angiography at 6 months and 2 years. All angiograms were performed after the delivery of intracoronary nitroglycerin. Patients only underwent repeat revascularization if there were recurrent symptoms or evidence of myocardial ischemia on clinically indicated or routine functional testing. All patients at the Montreal Heart Institute remained under either clinical or telephone review at a minimum of 3 monthly intervals.
All procedural, 6-month, and 2-year follow-up angiograms were forwarded to the Angiographic Core Laboratory at the Montreal Heart Institute for interpretation and measurement by independent observers using the Cardiovascular Measurement System (Medis, medical imaging system), a previously validated system11,12⇓ providing diameter of the reference vessel, the minimal lumen diameter (MLD), and mean vessel diameter (MVD) at baseline, after procedure, and at follow-up. MVD is the mean diameter over the specific analyzed segment. These analyses were performed retrospectively after the 2-year follow-up angiography. The analyzed segment of the coronary artery was divided into 3 segments: a central, target segment including the initial baseline lesion, and 2 adjacent segments located proximally and distally. The aim of these analyses, in addition to determining restenosis rates, was to assess the effects of each of the interventions (balloon and radiation) on the vessel wall, both individually and in combination. Because this study was performed before the operators were aware of the significance of geographic miss,13,14⇓ specific care to document balloon and radiation positioning was not performed. We therefore elected to limit the length of the central segment at the site of baseline stenosis to 15 to 17 mm, slightly less than the length of the 20-mm balloon used for dilation, ensuring that the arterial segment under analysis had been submitted to both balloon barotrauma and full-dose radiation. This was confirmed by precise superimposition of vessel, balloon, and radiation source images in each case. Additionally, we analyzed proximal as well as distal subsegments adjacent to the central segment with lengths of 10 to 15 mm each. This resulted in a total analyzed vessel segment length of 35 to 40 mm. These adjacent subsegments represent zones where only radiation or only balloon injury may have occurred. They will also include the radiation source penumbra where low-dose radiation is delivered to vessel segments that potentially have undergone balloon injury, possibly resulting in geographic miss.13,14⇓ In each of the segments, acute gain (in millimeters) was defined as MLD after angioplasty less the MLD before angioplasty. Late loss was defined as MLD after angioplasty less MLD at follow-up. Loss index was defined as late loss divided by the acute gain. Binary restenosis was determined as >50% diameter stenosis at follow-up in the each of the defined vessel segments. A specific analysis was performed on patients who had not undergone acute stent deployment or any other intervention over the 2-year follow-up period.
Target lesion revascularization (TLR) included any angioplasty or surgical bypass performed to the vessel because of a stenosis of >50% at the site of the initial target lesion. Target vessel revascularization (TVR) was defined as any angioplasty or surgical bypass performed to the vessel or one of its branches containing the initial lesion. Non–target vessel revascularization refers to revascularization of another epicardial coronary artery. Finally, an edge lesion was defined as a lesion that was not at the site of the initial stenosis but was within a 5-mm zone axially proximal to the site of the proximal gold seed marker or 5 mm distal to the distal gold seed marker.
All values are provided as proportions or as mean±1 SD. Comparisons between preprocedural, postprocedural, 6-month, and 2-year follow-up results were done using Wilcoxon signed ranks test with 2-tailed test for significance.
Thirty-two patients were recruited between March and June 1997. Because of unsatisfactory angioplasty results requiring immediate stent implantation, 2 patients were not randomized to receive radiation therapy. After successful angioplasty, 30 patients were randomized to receive doses of 12, 14, or 16 Gy and underwent intracoronary radiation therapy. Of this cohort, 4 patients underwent stent deployment to the target lesion within the first week. The baseline clinical and angiographic characteristics of the total cohort have been reported previously.8
Clinical follow-up was obtained in all patients at 6 months and in all living patients at 24 months. Major adverse cardiac events are presented in Table 1. At the 6-month time point, 2 of the 26 patients (7%) without stents had undergone TLR at the site of the initial MLD. One of the 4 patients stented at the baseline procedure underwent TLR for restenosis at the proximal and distal edges of the stent and radiation zone clearly in an area of geographic miss.
At the 2-year time point, a single additional patient of the entire cohort required TLR (cumulative TLR, 3 of 26 or 11.5%). This patient presented with progressive angina at 11 months after index procedure and was found to have a focal restenosis 5 mm proximal to the initial lesion that had been without a significant lesion at the 6-month angiogram. In 2 patients, TVR was performed at a site other than that of the initial lesion (cumulative TVR, 5 of 26 or 19%). In one of these, restenosis occurred in a stent distal to the irradiated site. Non–target vessel revascularization was performed in 5 patients, one in the period between 2 and 6 months, another between 6 and 12 months, and 3 more between 1 and 2 years.
By the 3-year time point, only 2 additional events needed to be reported; one patient presented a myocardial infarction in the territory of the treated artery without any significant angiographic lesion, and another patient needed a TVR but on a lesion more proximal to the area previously instrumented (cumulative TVR, 6 of 26 or 23%).
There was a single death in the cohort. This patient was a heavy smoker and succumbed to carcinoma of the lung 23 months after the index procedure. The diagnosis of this condition was made 12 months after the initial intervention. A 58-year-old woman developed carcinoma of the breast 22 months after the index procedure. There were no episodes of late acute vessel occlusion. The 26 nonstented patients were treated with aspirin alone after the intervention. The 4 stented patients were treated with aspirin and ticlopidine for 1 month and subsequently aspirin only. In a single case, after detection of late stent malposition at 6-month intravascular ultrasound, treatment with long-term clopidogrel was instituted.
The initial angiographic follow-up was performed on all 30 patients at a mean of 5.9±1.1 months (range, 5 to 7 months) after the index procedure. The results have been previously reported, with binary restenosis being found in 3 (10%) of the patients with a late lumen loss of −0.02±0.60 mm.10
For this longer-term review, patients were requested to return for repeat angiograms 2 years after the index procedure. Follow-up angiography was performed in 26 patients (86.66%) at a mean of 25.0±2.3 months (range, 19 to 30 months) after the index procedure. Four patients did not undergo angiographic study because of 1 patient dying, 1 patient undergoing therapy for breast cancer, 1 patient refusing, and 1 patient having undergone coronary artery bypass surgery.
The angiographic core laboratory did not detect any evidence of aneurysm or pseudoaneurysms on any of the 6-month or 2-year follow-up angiograms.
Over the 2-year follow-up, angiographic restenosis in the central segment was detected in 4 patients (15%). In another 4 patients, lesions developed in the same vessel proximally or distally to the target lesion site over the 2-year period. In 3 of these latter patients, the new lesions, although distant from the original target lesion, lay within 5 mm of the vessel segment covered by the source train and were therefore edge lesions. The quantitative angiographic results for the MLDs and the MVDs of the central, proximal, and distal segments are presented in the Figure, A and B, respectively, for the entire cohort of patients undergoing angiography at 2 years. This is a heterogeneous cohort of patients combining all patients irrespective of stenting or subsequent intervention. To better determine the natural history of postangioplasty brachytherapy, we looked separately at the subgroup of patients who did not undergo initial stenting or any other subsequent intervention for up to 2 years (Tables 2 through 4⇓⇓). Both the MVD and the MLD of the central segment increased significantly from immediately after procedure to the 6-month follow-up study with a negative late loss. From 6 months to 2 years after intervention, neither of these parameters showed any significant change indicative either of loss of lumen or dilation (Table 2). The proximal segment analysis revealed a significant reduction of the MLD (2.87±0.82 to 2.51±0.85 mm) and the MVD (3.27±0.66 to 2.99±0.70 mm) from the postprocedure measurements to the 6-month follow-up. This trend, however, arrested and did not continue at the 2-year study, with the MLD and MVD measurements remaining unchanged from the 6-month study (Table 3). The distal segment analysis revealed a nonsignificant reduction of the MVD from immediately after procedure to 6-month follow-up. There was a trend to a reduction of MLD, but this did not achieve statistical significance. At 2 years, there was no significant difference in either MLD or MVD compared with the immediate or 6-month studies (Table 4).
The use of a therapy with long-term implications such as radiation mandates a vigilant long-term follow-up to fully assess the safety and efficacy of the intervention. The absence of significant complications such as aneurysm or pseudoaneurysm formation is encouraging. The presence of intervention for disease progression in other vessels reflects the disease activity in this group of patients. In spite of this, the appearance of restenosis at the lesion site in a single patient at 11 months and 2 additional patients with late clinically significant edge lesions most likely related to the intervention remind us that in a small group of patients, longer-term failures are occurring. This pilot study was performed as a dose-finding trial with relatively low doses and before any awareness of the issue of geographic miss attributable to lack of precise positioning, vessel motion, and unclear full dose margins, all associated with a relatively short source. The stringency of source and balloon position that has developed subsequently when deploying brachytherapy sources was not present, and this may account for the edge stenoses noted. In fact, this is supported by the quantitative coronary angiography results in the nonintervened patients. At 6 months, the central segment, that in which there was complete source coverage, showed an increase in luminal diameters both at the MLD and over the entire segment when compared with postprocedural dimensions. The proximal and distal segments, however, show a loss of diameter over the same 6-month period, representing some subclinical edge effect that is occasionally visible qualitatively. The encouraging finding is that both the 6-month luminal gain at the intervened segment and the luminal loss at the edges remain unaltered from 6 months to 2 years. This seems in contradistinction to the Scripps Coronary Radiation to Inhibit Proliferation Post Stenting (SCRIPPS) trial that demonstrated a 200-μm late loss of mean MLD at 3 years.5 That study differed clearly in 2 ways from the BERT. First, the follow-up at 3 years was 1 year longer than in the results presented here, and although a hypothesis of delayed restenosis has been advanced with late catch-up, some late loss would be expected in the nonintervened cohort. Rather, there seem to be 2 populations of patients, with the nonintervened group representing a more complete control of the restenotic process. Second, the SCRIPPS trial was performed with stents and the BERT was following balloon angioplasty alone. This allows for the possibility of compensatory positive remodeling resulting in a stable lumen diameter over 2 years. An angiographic analysis alone cannot, however, determine such mechanisms. There is no doubt that since these patients were initially treated, great bounds have been made in understanding geographic miss and dosimetric issues that contribute to edge effect, the implications of impaired healing after radiation, and late acute occlusions occurring predominantly in new stent deployments. The natural history of this group of patients gives credence to the theory that when adequately treated, the beneficial outcome of post–balloon angioplasty patients is maintained over the longer term.
This was a pilot study, and, as such, it has a small sample size and no control group for comparison. Furthermore, when the initial procedures were performed, our present heightened awareness of geographic miss had not been developed and, therefore, the extensive documentation of balloon and radiation source positioning was not performed. As a result, the total assessed vessel segments are overestimations of the vessel lengths receiving balloon or radiation injury. The shoulder zones may represent untouched vessel segments, those having undergone either balloon alone or radiation alone or a combination of both.
Postangioplasty intracoronary radiation using a Sr/Y 90 source seems safe with no radiation-related side effects at 2-year follow-up. An increase in late TVR indicates that in some patients there may be a catch-up phenomenon. The late loss at the target side shoulders both proximally and distally that was noted at 6-month follow-up had arrested with no additional progression at the 2-year follow-up in this patient cohort. The central initial target segment seems to remain stable, with no late loss and often late gain to 6 months with no additional luminal dimensional changes at the 2-year follow-up. The proximal and distal shoulders do have evidence of late loss up to 6 months after the procedure, but this process is in the mean arrested by 2-year follow-up. This longer-term follow-up seems to confirm the safety of VBT after angioplasty and, within the limitations of a small, nonrandomized pilot study, suggests that although in isolated cases late catch-up may be noted, in most cases stable luminal dimensions are maintained.
Dr Bonan serves as Medical Director to Novoste Corporation.
- ↵Waksman R, Bhargava B, White L, et al. Intracoronary beta-radiation therapy inhibits recurrence of in-stent restenosis. Circulation. 2000; 101: 1895–1898.
- ↵Teirstein PS, Massullo V, Jani S, et al. Three-year clinical and angiographic follow-up after intracoronary radiation: results of a randomized clinical trial. Circulation. 2000; 101: 360–365.
- ↵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.
- ↵Costa MA, Sabate M, Van der Giessen WJ, et al. Late coronary occlusion after intracoronary brachytherapy. Circulation. 1999; 100: 789–792.
- ↵Meerkin D, Tardif JC, Crocker IR, et al. Effects of intracoronary β-radiation therapy after coronary angioplasty: an intravascular ultrasound study. Circulation. 1999; 99: 1660–1665.
- ↵Reiber JIIC, van der Zwet PMJ, von Land CD, et al. Quantitative coronary arteriography: equipment and technical requirements. In: Reiber JIIC, Serruys PW, eds. Advances in Quantitative Coronary Arteriography. Dordrecht, the Netherlands: Kluwer Academic Publishers; 1993: 75–111.
- ↵Sabate M, Costa M, Kozuma K, et al. Geographic miss: a course of treatment failure in radio-oncology applied to intracoronary radiation therapy. Circulation. 2000; 101: 2467–2471.