Author + information
- Received January 7, 2005
- Revision received April 25, 2005
- Accepted May 9, 2005
- Published online September 6, 2005.
- Fernando Alfonso, MD⁎,⁎ (, )
- José M. Augé, MD†,§§,
- Javier Zueco, MD‡,
- Armando Bethencourt, MD§,
- José R. López-Mínguez, MD∥,
- José M. Hernández, MD¶,
- Juan A. Bullones, MD#,
- Isabel Calvo, MD⁎⁎,
- Enrique Esplugas, MD††,
- María J. Pérez-Vizcayno, MD⁎,
- Raul Moreno, MD⁎,
- Cristina Fernández, MD⁎,
- Rosana Hernández, MD⁎,
- Vasco Gama-Ribeiro, MD‡‡,
- RIBS Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Fernando Alfonso, Cardiología Intervencionista, Hospital Universitario “San Carlos,” Plaza de Cristo Rey, Madrid 28040, Spain
Objectives We sought to analyze the very late outcomes of patients treated for in-stent restenosis (ISR) according to treatment allocation and 10 prespecified variables.
Background Long-term results (>2 years) of patients with ISR undergoing repeat coronary interventions are not well established.
Methods The Restenosis Intrastent: Balloon angioplasty versus elective Stenting (RIBS) randomized study compared these two strategies in 450 patients with ISR. A detailed systematic protocol was used for late clinical follow-up.
Results At one-year follow-up (100% of patients), the event-free survival was similar in the two groups (77% stent implantation [ST] arm, 71% balloon angioplasty [BA] arm, log-rank p = 0.19). Additional long-term clinical follow-up (median 4.3 years, range 3 to 5 years) was obtained in 98.6% of patients. During this time 22 additional patients died (9 ST arm, 13 BA arm), 7 suffered a myocardial infarction (3 ST arm, 4 BA arm), 23 required coronary surgery (11 ST arm, 12 BA arm), and 9 underwent repeat coronary interventions (4 ST arm, 5 BA arm) (nonexclusive events). At four years the event-free survival was 69% in the ST arm and 64% in the BA arm (log-rank p = 0.21). Among the 10 prespecified variables, vessel size ≥3 mm had a major influence on the clinical outcome at four years, with better results in the ST group (hazard ratio 0.51, 95% confidence interval 0.3 to 0.89, p = 0.016).
Conclusions Patients with ISR undergoing repeat interventions have a significant event rate at late follow-up. Continued medical surveillance should be continued after one year. Patients with large vessels have a better outcome after repeat stenting.
Coronary stenting currently represents the default strategy during percutaneous coronary interventions (PCI) (1,2). In-stent restenosis (ISR), however, remains the major limitation of this therapy (1–10). Although several mechanical strategies have been proposed in patients with ISR (3–9), several randomized trials have failed to confirm initial expectations (7–9). Brachytherapy is highly effective in these patients, but technical and logistic drawbacks limit its use (11). Drug-eluting stents represent a major therapeutic breakthrough (2,12,13), but larger studies are required before the systematic use of these new devices is justified in all patients with ISR.
Accordingly, there is still a need to gain further insights on the long-term outcome of patients with ISR undergoing repeat PCI. In the Restenosis Intrastent: Balloon angioplasty versus elective Stenting (RIBS) randomized study, 450 patients with ISR were randomized. The acute and mid-term results of this study have been previously reported (8). The present report describes the long-term (three to five years) clinical outcome of these patients. In addition, we sought to evaluate the influence of 10 “predefined” relevant variables on long-term prognosis.
Patients and protocol
The protocol and primary end point of the RIBS trial have been reported (8). Patients with myocardial ischemia and ISR lesions <32 mm in length in vessels >2.5 mm were eligible. Twenty-four sites from Spain and Portugal participated in the trial. All patients received aspirin, and those undergoing repeat stenting also received ticlopidine for one month. Patients were followed-up at 1, 7, and 12 months. Late angiographic follow-up was obtained in 96% of eligible patients (8). When complete one-year results were available, all participating centers agreed to maintain a close clinical follow-up of these patients. Accordingly, the same detailed standardized forms were forwarded to the coordinating center yearly thereafter. Queries about missing or inconsistent data were sent back to the sites. Summaries of clinical records from all patients with potential events were reviewed. Eventually, all events were classified and adjudicated by an independent clinical events committee unaware of the assigned treatment.
Death was considered as cardiac unless a noncardiac cause could be demonstrated. Differentiation between cardiac/noncardiac death was predefined in the protocol. Myocardial infarction (MI) required two of the following: prolonged (>30 min) chest pain, new Q waves, and creatine kinase levels ≥2 of the upper normal value (with MB fraction >10% of total). The composite of death, MI, and target vessel revascularization was a prespecified secondary study end point.
Kaplan-Meier curves were constructed to estimate event-free survival. Event-rates were compared with the log-rank test. Cox proportional hazard analyses were used to determine long-term outcome. Hazard ratios were calculated for the 10 relevant prespecifiedvariables (8). All analyses were performed according to the intention-to-treat principle, using the SPSS package (version 12.0, SPSS Inc., Chicago, Illinois). A p value <0.05 was considered statistically significant.
One year clinical follow-up was obtained in all 450 patients (100%). A clinical follow-up >1 year was obtained in 445 patients (98.8%); a follow-up >3 years was obtained in 444 patients (98.6%) (223 stent implantation [ST] arm, 221 balloon angioplasty [BA] arm). The last clinical follow-up was obtained at 4.3 years (mean 1,537 ± 260 days; median 1,569 days [interquartile range 1,362 to 1,736 days]). Table 1summarizes clinical events (as total counts) during the complete follow-up period. A total of 52 patients (12%) experienced at least one adverse event after the first year of follow-up (including 36 patients that did not have an adverse event during the first year). When all events occurring after the first year were analyzed (nonexclusive events), 22 patients died (9 ST arm, 13 BA arm), 7 suffered a MI (3 ST arm, 4 BA arm), 23 required coronary surgery (11 ST arm, 12 BA arm), and 9 underwent repeat PCI (4 ST arm, 5 BA arm). Of the 32 patients requiring “late” target vessel revascularization, 11 patients had documented restenosis at 6 months (initially left untreated due to a lack of symptoms), 10 developed “late restenosis,” and 11 had repeated procedures. Ranked by hierarchy (exclusive events), 22 patients died (9 ST arm, 13 BA arm), 5 suffered a MI (2 ST arm, 3 BA arm), 18 required surgical revascularization (10 ST arm, 8 BA arm), and 7 required repeat PCI (3 ST arm, 4 BA arm) after the first year.
At last clinical follow-up, a total of 37 patients had died (17 ST arm, 20 BA arm), 16 had suffered a MI (6 ST arm, 10 BA arm), 38 had required surgery (20 ST arm, 18 BA arm), and 62 patients had required repeat PCI (27 ST arm, 35 BA arm) (hierarchical ranking). Death was cardiac in 20 cases (6 ST arm, 14 BA arm, p = 0.07), and 10 of these patients died after the first year (1 ST arm, 9 BA arm, p = 0.02). Only two cardiac deaths were sudden (one patient in each group). Angiographically demonstrated abrupt vessel closure associated with an acute coronary syndrome was only seen in one patient in each group (both during the first year of follow-up).
Figure 1displays the event-free survival according to treatment allocation. A nonsignificant trend favoring repeat stenting was found for the combined clinical end point. In addition, a significant reduction in the occurrence of cardiac death/MI (end point not prespecified) was demonstrated after restenting. Cox analysis of the long-term clinical outcome according to the 10 prespecified variables is shown in Figure 2.Most subgroups showed results consistent with the main outcome measure except for patients with large vessels that presented a significantly better outcome after repeat stenting (hazard ratio 0.51, 95% confidence interval 0.30 to 0.89, p = 0.016). Patients with diffuse ISR (Cox, p = 0.07) and those with lesions in the right or circumflex coronary arteries (Cox, p = 0.02) also had a better prognosis after repeat stenting (Fig. 2). However, an interaction was found between vessel size and these two variables. The reference vessel diameter was significantly larger in patients with diffuse restenosis (2.94 ± 0.5 mm vs. 2.79 ± 0.5 mm, p < 0.001) and in non-left anterior descending coronary artery lesions (2.97 ± 0.5 mm vs. 2.74 ± 0.4 mm, p < 0.001).
Event-free survival for patients with large vessels is depicted in Figure 3.The better long-term results obtained with repeat stenting were largely driven by a lesser requirement for target vessel revascularization (16% vs. 35%; Cox, p = 0.016).
This study constitutes the longest and most exhaustive clinical follow-up of a large cohort of patients with ISR undergoing repeat PCI. Our follow-up protocol enabled us to obtain a uniquely high rate of late clinical information. In the RIBS study, most adverse events clustered during the first year of follow-up. Although the systematic late angiographic follow-up could have influenced the rate of subsequent target vessel revascularization (14), by protocol all repeated revascularizations needed to be clinically driven (symptoms or evidence of ischemia). Accordingly, in some patients with recurrent ISR, the procedure was deferred until ischemia could be demonstrated. This could explain the relatively high attrition rate noticed from the sixth month to the end of the first year of follow-up. After this time the rate of new events was markedly reduced, but still a significant number of patients suffer from new events.
No previous study has systematically analyzed the late clinical outcome of patients with ISR undergoing repeated mechanical PCI after two years of follow-up. Reimers et al. (3) reported an event-free survival of 81% at 24 months (mean follow-up 27 ± 15 months) in 124 patients with ISR. Bossi et al. (5) followed-up (median 459 days) 234 ISR patients treated with balloon angioplasty and found an event-free survival of 75% at 24 months. Finally, in the meta-analysis of Radke et al. (10) (3,012 ISR patients from 28 studies), the estimated probability of experiencing a major adverse event (mean follow-up 9 ± 4 months) was 30%. The rate of adverse events was comparable among different treatment modalities, and the postprocedural diameter stenosis was the only independent predictor of long-term outcome. Recent data from three large randomized trials confirm these findings demonstrating comparable clinical results at one year, irrespective of the mechanical strategy selected (balloon, rotational atherectomy, cutting balloon, or restenting) (7–9).
Conversely, the late outcome of selected patients with ISR undergoing brachytherapy has been examined in detail. The concerns of a potential late “catch-up” phenomenon (delaying rather than preventing restenosis) stimulated longer follow-up studies in small patient subsets. A delayed mitigation of efficacy, associated with a need for late target vessel revascularization, was demonstrated in some irradiated patients (11). The lifespan of this technology, however, is currently seriously threatened in the era of drug-eluting stents.
Our findings suggest that most patients treated for ISR have a good long-term outcome, but 34% of patients experience at least one adverse event at 4.3 years of follow-up. In addition, our data also indicate the importance of maintaining an adequate clinical follow-up after the first year, considering that some patients (up to 12%) will experience very late adverse events. It should be kept in mind that most patients included in our trial had complex and advanced coronary artery disease (27% diabetics, 43% previous MI, 45% multivessel disease) (8). These factors and an overall recurrent restenosis rate of 38% (8) could explain the relatively high rates of coronary surgery and mortality at late follow-up.
Interestingly, we found a trend, which did not reach statistical significance, toward a better outcome after repeat stenting in all analyzed clinical outcome measures that emerged in the early period. Although cardiac death and the combined end point of cardiac death/MI at late follow-up were significantly lower after repeat stenting, these findings should be interpreted with caution because they come from post-hoc analyses, and the study design was underpowered to address this issue.
Our findings are particularly relevant for patients undergoing repeat stenting because long-term data in this scenario is very limited (8). Furthermore, this information is also of interest considering that currently most centers are using drug-eluting stents as the workhorse strategy for patients with ISR. Once the benefit of the drug coating is obtained, our results, demonstrating that the presence of two layers of metal on the vessel wall is not associated with any adverse long-term implication, are particularly reassuring.
Another important finding of the current study is that patients with ISR in large vessels do much better after repeat stenting than after balloon angioplasty, and this improved clinical outcome is maintained after 4.3 years. Therefore, the present report confirms our initial findings (8) in this important prespecified patient subset. The better outcome after restenting for patients with diffuse ISR and ISR in locations other than the left anterior descending coronary artery are also noteworthy. However, these two anatomic scenarios were associated with larger vessels, and their clinical implications remain difficult to establish. Finally, some patient subsets (small vessels, focal, left anterior descending coronary artery, and early ISR) do not appear to gain additional clinical benefit from repeat stenting.
Only bare-metal stents were used in the RIBS study. The mid-term results of drug-eluting stents in this setting are excellent (12,13), but long-term information is eagerly awaited. Because intravascular ultrasound was not systematically performed, its potential implications on long-term outcome were not analyzed. Medical treatment at one year (statins 48%, beta-blockers 63%, and angiotensin-converting enzyme inhibitors 44%) was well balanced in the two groups. However, the potential influence of medical treatment on clinical outcome after this time cannot be established from our study.
Patients with ISR undergoing repeat PCI have a significant rate of adverse events at late follow-up. Although most events tend to concentrate during the initial year, a low but steady rate of adverse events occur later on. Accordingly, continued clinical surveillance appears warranted in these patients. Finally, the long-term clinical outcome of patients with ISR in large vessels treated with repeat stenting is favorable.
For a list of the RIBS Investigators and institutions that participated in this study, please see the online version of this article.
Supplementary data associated with this article can be found, in the online version, at doi: 10.1016/j.jacc.2005.05.050.
- Abbreviations and Acronyms
- balloon angioplasty
- in-stent restenosis
- myocardial infarction
- percutaneous coronary intervention
- Restenosis Intrastent: Balloon angioplasty versus elective Stenting randomized study
- stent implantation
- Received January 7, 2005.
- Revision received April 25, 2005.
- Accepted May 9, 2005.
- American College of Cardiology Foundation
- Reimers B.,
- Moussa I.,
- Akiyama T.,
- et al.
- Bossi I.,
- Klersy C.,
- Black A.J.,
- et al.
- Mehran R.,
- Dangas G.,
- Abizaid A.S.,
- et al.
- vom Dahl J.,
- Dietz U.,
- Haager P.K.,
- et al.
- Alfonso F.,
- Zueco J.,
- Cequier A.,
- et al.
- Albiero R.,
- Silber S.,
- Di Mario C.,
- et al.
- Radke P.W.,
- Kaiser A.,
- Frost C.,
- Sigwart U.
- Waksman R.,
- Ajani A.E.,
- White L.,
- et al.
- Sousa J.E.,
- Costa M.,
- Abizaid A.,
- et al.
- Tanabe K.,
- Serruys P.W.,
- Grube E.,
- et al.