Author + information
- Received June 12, 2008
- Revision received August 11, 2008
- Accepted August 19, 2008
- Published online November 11, 2008.
- Fernando Alfonso, MD⁎,⁎ (, )
- María-José Pérez-Vizcayno, MD⁎,
- Rosana Hernández, MD⁎,
- Armando Bethencourt, MD†,
- Vicens Martí, MD‡,
- José R. López-Mínguez, MD§,
- Juan Angel, MD∥,
- Andrés Iñiguez, MD¶,
- César Morís, MD#,
- Angel Cequier, MD⁎⁎,
- Manel Sabaté, MD‡,
- Javier Escaned, MD⁎,
- Pilar Jiménez-Quevedo, MD⁎,
- Camino Bañuelos, MD⁎,
- Alfonso Suárez, MD⁎,
- Carlos Macaya, MD⁎,
- RIBS-II Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Fernando Alfonso, Cardiología Intervencionista, Instituto Cardiovascular, Hospital Universitario “San Carlos,” Ciudad Universitaria, Plaza de Cristo Rey, Madrid 28040 Spain
Objectives We sought to assess the long-term effectiveness and safety of sirolimus-eluting stents (SES) in patients with in-stent restenosis (ISR).
Background Treatment of patients with ISR remains a challenge. The long-term outcome of patients with ISR treated with SES remains unknown.
Methods The RIBS-II (Restenosis Intra-stent: Balloon angioplasty vs. elective sirolimus-eluting Stenting) study was a randomized trial conducted in 150 patients with ISR (76 SES, 74 balloon angioplasty [BA]). The long-term (>1 year) clinical outcome and pre-specified subgroup analyses were pre-defined secondary study end points.
Results At 1 year, the event-free survival (death, myocardial infarction, target vessel revascularization [TVR]) was better in the SES group (88% vs. 69%, p < 0.005). Additional long-term (>3 years) clinical follow-up was obtained in 97% of patients (median 3.3 years). After the first year, 3 patients died (1 SES, 2 BA), 5 suffered myocardial infarction (4 SES, 1 BA), and 7 required TVR (4 SES, 3 BA). At last follow-up, definitive/probable/possible stent thrombosis was similar in both groups (2/2/1 SES vs. 1/0/3 BA, p = NS). At 4 years, the event-free survival was 76% in the SES arm and 65% in the BA arm (p = 0.019). On multivariate analysis, SES implantation was an independent predictor of event-free survival. Subgroup analyses were consistent with the main outcome measure.
Conclusions In patients with ISR, SES implantation remains effective and safe at very long-term clinical follow-up.
Drug-eluting stents (DES) are highly effective in patients with in-stent restenosis (ISR) (1–4). In this challenging setting both sirolimus-eluting stents (SES) and paclitaxel-eluting stents have been shown to reduce the restenosis rate compared with balloon angioplasty (BA) (1,2). Furthermore, in these patients DES are also superior to brachytherapy (3,4). Therefore, currently most patients with bare-metal ISR are treated with DES.
The efficacy of DES has been demonstrated in several adverse scenarios (1–5). More recently, however, safety issues have been raised suggesting an increased risk of late thrombosis, primarily when DES have been used with “off-label” indications in “real world” clinical practice (6,7).
All randomized studies confirming the efficacy of DES in patients with ISR had a relatively short clinical follow-up, ranging from 9 months to 1 year (1–4). Therefore, the long-term safety and efficacy of SES for patients with ISR remains to be established. The present report describes the very long-term clinical outcome (3 to 4 years) of these patients.
Patient selection and study design
The RIBS-II (Restenosis Intra-stent: Balloon angioplasty vs. elective sirolimus-eluting Stenting) study was a prospective, multicenter, randomized trial designed to compare these 2 therapeutic strategies in patients with ISR (see Online Appendix). Inclusion/exclusion criteria have been previously defined in detail (2). Briefly, patients with ischemia secondary to bare-metal ISR were eligible. Patients with ISR on small vessels (<2.5 mm) or very diffuse ISR (>32 mm) were excluded (2). The study was an investigator-driven initiative, and 8 Spanish University Hospitals participated in the trial.
Careful lesion pre-dilation (undersized and short balloons) followed by full lesion coverage was recommended in the SES arm. Relatively high pressures (>12 atm) were used in both arms (2). After intervention, patients received aspirin indefinitely and clopidogrel (75 mg/day) for at least 9 months. Thereafter, the duration of extended clopidogrel use was left to the attending cardiologist's discretion.
Follow-up and definitions
Initially, patients were followed-up at 1 month, 9 months, and 1 year. Subsequently, all patients were prospectively followed on a yearly basis with a detailed, standardized, structured questionnaire. Noninvasive tests to detect ischemia were recommended in these visits. Detailed information on antiplatelet therapy was collected. Telephone contact was obtained in patients not seen at the clinics. Consistency checks were performed at the coordinating center, and whenever needed, queries were sent back to sites. Clinical records of all patients with potential events were reviewed. All major events were verified against source documentation. Clinical events (death, myocardial infarction [MI], target vessel revascularization [TVR]), as previously defined (2), were adjudicated by an independent Clinical Events Committee blinded to the assigned treatment. The protocol emphasized that reinterventions at follow-up required previous demonstration of myocardial ischemia.
The original protocol analyzed angiographic SES thrombosis. However, because this definition is now considered too restrictive, the Academic Research Consortium (ARC) hierarchical definition was used to readjudicate cases of SES thrombosis (8).
Initial results of the RIBS-II study have been previously reported (2). Patients assigned to SES had a lower restenosis rate (11% vs. 39%, p < 0.001), a reduced neointimal proliferation on intravascular ultrasound, and a better 1-year outcome compared with the BA group (2).
The objective of the present study was to assess the very long-term (>1 year) clinical outcome in these patients. The composite of death, MI, and TVR at late follow-up was a pre-specified secondary end point. We also evaluated the role of 10 pre-defined variables and extended dual antiplatelet therapy on late clinical events. Finally, we sought to assess predictors of clinical outcome.
Data are presented as values and percentages or mean ± SD. Median and interquartile range were used when data were not normally distributed (Kolmogorov-Smirnov test). Event-free survival was estimated by Kaplan-Meier analysis and compared with the Breslow's exact test. Cox proportional hazard analyses were used to assess long-term clinical outcome (hazard ratios [HRs] and 95% confidence intervals [CIs]) and the influence of 10 pre-specified relevant baseline variables. Interactions were also studied. Multivariate Cox analyses were used to study independent predictors of major events. Unless otherwise specified, analyses were performed according to the intention-to-treat principle with the SPSS package, version 12.0 (SPSS Inc., Chicago, Illinois). A p value <0.05 was considered statistically significant.
One-year follow-up was obtained in all 150 randomized patients. A clinical follow-up >1 year was obtained in 146 patients (97.3%) and >3 years in 145 patients (96.7%) (97% SES, 96% BA). The last clinical follow-up was at 3.3 years (median: 1,195 days; interquartile range: 1,123 to 1,274 days; mean: 1,151 ± 262 days). Table 1 summarizes adverse events. Overall, clinical events at long-term follow-up were more frequent in the BA group (HR: 1.92; 95% CI: 1.03 to 3.59), largely as the result of a higher requirement for TVR (HR: 2.35; 95% CI: 1.17 to 4.70). The early benefit achieved with SES during the first year was maintained at long-term follow-up. Figure 1 displays survival estimates at late follow-up. Figure 2 shows late clinical outcome according to 10 pre-specified variables. No significant interactions were found with the main outcome measure.
Clinical events after the first year
In the SES group 1 patient (without clopidogrel) died after a non–Q-wave MI (possible stent thrombosis). In the BA group, 2 patients died, 1 suddenly (on clopidogrel), 2 weeks after revascularization of a nontarget vessel (possible stent thrombosis), and another from a major bleeding and a non–Q-wave MI that complicated an urotelioma resection (clopidogrel previously withdrawn).
Three patients in the SES arm suffered from large Q-wave MIs after 1 year (1 definitive, 2 probable SES thromboses). The first patient (on clopidogrel) suffered a definitive SES thrombosis treated with thrombectomy and new SES implantation; the second (on clopidogrel) suffered thrombosis in an SES immediately distal to the target SES that eventually needed bare-metal stenting; the third patient (without clopidogrel) required thrombolysis for an ST-segment elevation MI, but 2 days later angiography revealed a thrombus-free SES. In the BA group, only the previously described patient with urotelioma surgery developed a non–Q-wave MI after the first year.
In the SES group 3 patients required late TVR for symptomatic recurrent ISR: 2 had “real” late restenosis (no restenosis at 9 months), and 1 had asymptomatic restenosis at 9 months. Three patients in the BA arm required reintervention (1 surgical, 2 percutaneous) for late restenosis (not present at 9 months).
Table 2 summarizes episodes of stent thrombosis according to the ARC definition, time of occurrence, and clopidogrel status. During the first year, 2 patients suffered angiographic (definitive) stent thrombosis and developed large Q-wave MI (1 in each arm). The patient in the SES arm had hyperhomocysteinemia and questionable adherence to the dual antiplatelet regimen. The patient in the BA arm suffered thrombosis (on clopidogrel) 1 month after an SES was implanted to treat an early recurrent ISR. During the first year, 2 BA patients died suddenly (1 with and 1 without clopidogrel) and were classified as possible stent thrombosis. After the first year, 1 patient in the SES group experienced definitive very late stent thrombosis and 2 experienced probable very late stent thrombosis (all previously described as Q-wave MIs). Finally, 1 patient in each arm suffered possible very late stent thrombosis (2 unexpected deaths previously described).
Notably, in the BA arm the patient with definitive stent thrombosis and 1 of the 3 patients with possible thrombosis suffered this event only after SES implantation was required for recurrent ISR at the target lesion. Therefore, in a “per final treatment analysis,” 7 of the 9 episodes of definitive/probable/possible stent thrombosis occurred in patients who were initially treated with or eventually required SES implantation. In fact, only 2 patients treated with BA “alone” experienced this problem during follow-up.
Predictors of adverse clinical events
Univariate and multivariate predictors of adverse events and TVR are presented in Table 3. On Cox analyses the use of SES was an independent predictor of event free survival.
All patients in the SES group received clopidogrel for 12 months (median 17 ± 13 months). After 1 year, clopidogrel therapy could not be determined with certainty or was unknown in 5 of 71 patients. In the remaining 66 patients, clopidogrel was maintained for >1 year in 22 patients (33%), >2 years in 16 (24%), and was used indefinitely in 15 patients (23%). Clopidogrel status at the time of stent thrombosis is detailed in Table 2. Of the 4 patients with definitive/probably/possible very late SES thrombosis, 2 were receiving clopidogrel therapy. Adverse events, at any time or after the first year of follow-up, were not related to clopidogrel status.
This study demonstrates that in patients with ISR the beneficial clinical effects of SES are maintained up to 4 years after intervention. The lower event rate seen in the SES group was mainly driven by a reduced need for target lesion revascularization. Late events were rare and equivalent in both groups. Notably, after 1 year, the clinical requirement for TVR was very low. This dissipates potential concerns of a potential late “catch-up” phenomenon elicited by SES that might diminish their long-term efficacy. Moreover, pre-specified analyses of relevant subgroups were consistent with this main outcome measure. In addition, the incidence of stent thrombosis after 1 year was also very low. Although very late stent thrombosis was numerically more frequent in the SES arm, the difference was not statistically significant. Therefore, our study with systematic, prospective, and complete clinical follow-up provides important reassurance regarding long-term safety concerns after SES implantation in this challenging setting.
Long-term results after DES implantation
Initial concerns suggesting an increased rate of major adverse clinical events (death, MI) associated with DES (6,7) have been recently dissipated. Meta-analysis with independent adjudication and exhaustive assessment of late events—with patient-level data—failed to confirm an increased event rate after DES (5). Furthermore, although some initial real world registries also suggested the possibility of a higher event rate after DES (7), recent larger observational studies failed to confirm any risk increase and even suggested that DES might be beneficial in this regard (9). Conversely, the issue of very late thrombosis after DES implantation still remains of concern (10). The delayed endothelization process detected in some patients seems to be associated with a marginal but definitive higher risk for late thrombosis (10). Therefore, it is conceivable that in particularly adverse scenarios, such as ISR, DES implantation might be associated with a higher thrombotic risk.
Long-term follow-up in patients with ISR
Until very recently brachytherapy was considered the standard of care for patients with ISR (3,4). Recent studies, however, demonstrated a reduction in efficacy overtime, suggesting a delayed endothelization process (increasing late thrombosis risk) and a late restenotic response (11). These reasons, together with emerging data suggesting the superiority of DES (3,4), led to the widespread abandonment of this form of therapy. However, the 2 large randomized studies demonstrating the superiority of DES over brachytherapy (3,4) had a limited (9 months) clinical follow-up. This is of relevance, because in the SISR (Sirolimus-Eluting Stents Versus Vascular Brachytherapy for In-Stent Restenosis) trial (3), 2 SES patients suffered late stent thrombosis and 6 developed MIs, but none of these events occurred during follow-up in the brachytherapy arm. More recently, in a retrospective, single-center, observational study, Lee et al. (12) compared the long-term outcome of SES versus brachytherapy in patients with diffuse ISR. The event-free survival was significantly better (92.5% vs. 84.2%) in the 120 patients treated with SES (follow-up 35 ± 8 months). However, during this period, 2 SES patients suffered angiographic stent thrombosis, 1 had an MI, and 4 died. Nevertheless, these findings are difficult to compare with those of the 2 landmark randomized studies (3,4), because cutting-balloon or rotational atherectomy were systematically used and rhenium-188 balloons were selected as the beta-source (12).
The ISAR-DESIRE (Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for In-Stent Restenosis) randomized trial (1) demonstrated that both paclitaxel-eluting stents and SES were superior to BA in reducing TVR at 12 months. During this first year no early angiographic occlusions were detected, but late total occlusions were observed in 1 SES patient, 3 patients treated with paclitaxel-eluting stents, and in 2 BA patients (1). However, a longer clinical follow-up was not available. Yet, a recent observational study (13) assessing the 2-year clinical follow-up of 138 ISR patients treated with SES provided worrisome results: major events occurred in 5.8%, 14.3%, and 25% of patients at 6, 12, and 24 months, respectively. Therefore, the assessment of the very long-term safety and efficacy of SES in patients with ISR, especially as compared with simpler strategies such as BA, remains of paramount clinical importance.
The long-term follow-up of this randomized study demonstrates the superiority of SES over BA in patients with ISR. SES maintained their effectiveness in reducing clinical events during this uniquely long period of follow-up. Furthermore, SES proved to be safe in these patients who had a very low frequency of late events such as TVR and stent thrombosis. Notably, on multivariate analysis, SES implantation emerged as an independent predictor of very long-term event-free survival. In our series, time to restenosis and the Mehran and American College of Cardiology/American Heart Association angiographic classifications were also independent predictors of adverse clinical events (14).
The present study is the first to use the ARC definition to systematically assess thrombosis rates in patients treated for ISR (8). The dilemma as to whether to attribute thrombosis to the initial strategy or to the final intervention used to treat ISR has been previously highlighted (8). In our series, the higher requirement of TVR in the BA group might help to explain some cases of late thrombosis after subsequent reinterventions with SES. In fact, ISR restenosis is not always a benign clinical entity, and recently, repeated procedures have been associated with a higher risk of MI and stent thrombosis (15). In our intention-to-treat analysis no differences were detected regarding this complication in the 2 arms. However, when thrombotic episodes of patients in the BA arm—but eventually requiring an SES implantation for ISR recurrences—were computed as belonging to the SES strategy (worst case scenario), a trend favoring BA was detected. Further studies including larger series of patients and longer periods of follow-up are still warranted to address this issue.
Finally, the potential role of extended use of clopidogrel beyond 12 months in complex patients such as those with ISR remains to be defined. In our study, late discontinuation of dual antiplatelet therapy was not associated with an increased frequency of major events. However, a minority of patients treated with SES extended the dual antiplatelet for >1 year. Of the 4 episodes of definitive/probable/possible very late thrombosis in the SES group, 2 occurred in patients receiving clopidogrel.
First, only a relatively small number of patients were analyzed. Therefore, the power to detect small differences in clinical event rates after 1 year was limited. This problem particularly affects our subgroup analyses. Likewise, the low number of late events limits the value of the multivariate analysis (over-fitting) to identify event predictors. Further studies including a larger number of patients with ISR are required to confirm the very long-term safety and efficacy of SES in this setting. Second, the implications of an extended dual antiplatelet duration could not be definitively elucidated from our series, considering that the use of clopidogrel >1 year was left to the criteria of the attending physician (not randomized) and only retrospectively ascertained during clinical visits. Whether a very prolonged dual antiplatelet therapy might offer clinical benefit in selected patients receiving SES for ISR remains to be elucidated.
The very long-term clinical follow-up of this randomized clinical trial demonstrates that in patients with ISR the use of SES is safe and highly effective, providing long-term clinical results superior to BA.
For the organization of the trial, coinvestigators, and participating institutions, please see the online version of this article.
- Abbreviations and Acronyms
- balloon angioplasty
- confidence interval
- drug-eluting stent(s)
- hazard ratio
- in-stent restenosis
- myocardial infarction
- sirolimus-eluting stent(s)
- target vessel revascularization
- Received June 12, 2008.
- Revision received August 11, 2008.
- Accepted August 19, 2008.
- American College of Cardiology Foundation
- Alfonso F.,
- Perez-Vizcayno M.J.,
- Hernandez R.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Tsimikas S.
- Waksman R.,
- Ajani A.E.,
- White L.,
- et al.
- Stone G.W.,
- Ellis S.G.,
- Colombo A.,
- et al.