Author + information
- Received August 11, 2006
- Revision received September 20, 2006
- Accepted October 1, 2006
- Published online February 6, 2007.
- Alban Dibra, MD⁎,
- Adnan Kastrati, MD⁎,⁎ (, )
- Fernando Alfonso, MD†,
- Melchior Seyfarth, MD⁎,
- Maria-José Pérez-Vizcayno, MD†,
- Julinda Mehilli, MD⁎ and
- Albert Schömig, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Adnan Kastrati, Deutsches Herzzentrum, Lazarettstr. 36, 80636 Munich, Germany
Objectives We sought to synthesize the available evidence on the effectiveness of drug-eluting stents for bare-metal in-stent restenosis.
Background Although there is clinical evidence that drug-eluting stents are associated with better results than other treatments for in-stent restenosis, they are not yet approved for this indication. Meta-analysis of randomized trials may yield more precise estimates of treatment effects and enable a rapid adoption of effective treatments in clinical practice.
Methods Data sources included PubMed and conference proceedings. Prespecified criteria were met by 4 randomized studies comparing sirolimus- or paclitaxel-eluting stents versus balloon angioplasty or vascular brachytherapy in 1,230 patients with bare-metal in-stent restenosis. Studies reported the clinical outcomes of efficacy and safety during a minimum of 9 months. The primary outcome was target lesion revascularization.
Results No significant heterogeneity was found across trials, thus showing a similar effect size regardless of the use of balloon angioplasty or vascular brachytherapy as comparators. The risk of target lesion revascularization (odds ratio 0.35, 95% confidence interval [CI] 0.25 to 0.49; p < 0.001) and that of angiographic restenosis (odds ratio 0.36, 95% CI 0.27 to 0.49; p = 0.001) were markedly lower in patients treated with drug-eluting stents. There were no differences between patients treated with drug-eluting stents and those treated with other techniques with respect to the composite of death or myocardial infarction (odds ratio 1.04, 95% CI 0.54 to 2.03; p = 0.55).
Conclusions Drug-eluting stents are markedly superior to conventional techniques (balloon angioplasty and vascular brachytherapy) and should be considered as first-line treatment for patients with bare-metal in-stent restenosis.
The use of bare-metal stents (BMS) has lead to improved outcomes in patients undergoing percutaneous revascularization, yet its success has been limited by the development of in-stent restenosis (1). Treatment of bare-metal in-stent restenosis, which presents as an acute coronary syndrome in a considerable proportion of patients (2) and may negatively influence their long-term survival (3), has always been considered as problematic (4). Although several treatment modalities have been used, they have generally been associated with unsatisfactory long-term results (4). Studies have shown that a large proportion of restenotic lesions require repeat revascularization procedures (5).
The introduction of drug-eluting stents (DES) marks the most successful effort in reducing the risk of restenosis after percutaneous coronary interventions. After the initial success in patients and lesions with low to intermediate risk (6–8), the use of DES was extended to high-risk patient and lesion subsets (9–13). Recently, the effectiveness of DES has been evaluated for treatment of in-stent restenosis in patients previously treated with BMS. Studies have shown that sirolimus- and paclitaxel-eluting stents (SES and PES, respectively), the most studied DES, are superior to conventional treatment modalities for in-stent restenosis, such as ballon angioplasty (BA) and vascular brachytherapy (14–18). Nevertheless, the use of DES for treatment of in-stent restenosis has not yet received approval by the U.S. Food and Drug Administration (19). Also, recent guidelines do not distinguish between the relative efficacy of DES and vascular brachytherapy (19,20).
Meta-analysis of randomized trials may yield more precise and reliable estimates of treatment effect (21). It can be useful during the development of guidelines (22,23) and the regulatory process for drugs and devices (24,25), contributing to the rapid adoption of effective treatment modalities in clinical practice (26). Therefore, we performed a meta-analysis of randomized trials that have evaluated the effectiveness of DES in patients with bare-metal in-stent restenosis.
For inclusion, studies comparing DES with other conventional treatment modalities (BA or vascular brachytherapy) in patients with bare-metal in-stent restenosis had to have a randomized design and report the clinical outcomes of efficacy and safety during a minimum follow-up period of 9 months. All studies meeting the requirements, regardless of the language or form of publication, were considered to be eligible for this meta-analysis.
Data sources included PubMed, reference lists of retrieved articles, pertinent reviews and editorials from leading medical journals, abstracts from scientific meetings of major cardiology societies (American College of Cardiology, American Heart Association, and European Society of Cardiology), as well as Internet-based sources of information on clinical trials in cardiology (www.tctmd.com, www.theheart.org, and www.clinicaltrials.gov). The last search was performed in July 2006.
Outcomes and data abstraction
We evaluated the efficacy and safety outcomes. The primary outcome of interest was ischemia-driven target lesion revascularization (TLR), which was defined as any revascularization procedure, either percutaneous or surgical, involving the target lesion. Stent thrombosis, death, and the composite of death or myocardial infarction were secondary clinical outcomes. In addition, we evaluated binary angiographic restenosis measured by quantitative coronary angiography at follow-up. It was defined as stenosis of at least 50% diameter in the segment including the stented area, as well as the proximal and distal margins within 5 mm from the stent edges.
Two investigators (A.D., J.M.) independently performed study selection and data abstraction. Differences were resolved by discussion. In addition to pertinent data on the outcomes of interest, we gathered information on trial names, first author, year of publication, and number of patients enrolled. The information was collected regarding main clinical characteristics (age, distribution of gender, and proportion of patients with diabetes mellitus) as well as angiographic characteristics (number of treated lesions, vessel size, and lesion length). We recorded the length and rate of clinical and angiographic follow-up in each study included in this analysis. Additionally, selected studies were evaluated for performance of the analysis according to the intention-to-treat principle and blind assessment of the outcomes of interest. If not reported in the existing presentation format, pertinent information was asked from the primary investigators of the studies. Based on collected information, there were no differences between treatment arms with respect to the length and rate of clinical or angiographic follow-up in all of the studies. Patients were analyzed according to the intention-to-treat principle. Outcomes were evaluated by blind assessors.
The Mantel-Haenszel method for fixed effects (27) and the DerSimonian and Laird method for random effects (28) were used to investigate the combined results of individual studies. Odds ratios (ORs) with 95% confidence intervals (CIs) were computed as summary statistics. To quantify the heterogeneity between trials, we used the Cochran test based on the pooled OR by Mantel Haenszel. We assessed publication bias with respect to the primary outcome of interest, TLR, using the Begg adjusted rank correlation test according to the method of Begg and Mazumdar (29) and regression asymmetry test by Egger et al. (30). A sensitivity analysis was performed by assessing the contribution of individual studies to the summary effect estimate with respect to the primary outcome. This was done by excluding each trial one at a time and computing meta-analysis estimates for the remaining studies. Results were considered statistically significant at p < 0.05. Analyses were performed with the Stata version 9 statistical package (Stata Corp., College Station, Texas).
Selected studies and their characteristics
Four randomized studies of patients with bare-metal in-stent restenosis treated with DES versus other treatment modalities were selected for this meta-analysis. A total of 1,230 patients were included in these studies: 730 of them were treated with DES, either SES or PES, and 500 were treated with plain BA or vascular brachytherapy. More specifically, patients treated with DES received SES in the SISR (Sirolimus-Eluting Stents Versus Vascular Brachytherapy for In-Stent Restenosis Within Bare-Metal Stents) (16) and RIBS-II (Restenosis Intrastent: Balloon Angioplasty Versus Elective Sirolimus-Eluting Stenting II) (18) trials, PES in the TAXUS-V ISR (Prospective Randomized Trial Evaluating Slow-Release Formulation Taxus Paclitaxel-Eluting Coronary Stent in the Treatment of In-Stent Restenosis) trial (17) and either SES or PES in the ISAR-DESIRE (Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for In-Stent Restenosis) trial (15). Patients treated with conventional therapeutic modalities underwent plain BA in the ISAR-DESIRE and RIBS-II trials and vascular brachytherapy in the SISR and TAXUS-V ISR trials. In the ISAR-DESIRE and RIBS-II trials the primary end point was angiographic restenosis. The composite of death, myocardial infarction, or target vessel revascularization was the primary end point in the SISR trial, whereas the main outcome of interest in the TAXUS-V ISR trial was ischemia-driven target vessel revascularization. All patients were preprocedurally treated with aspirin and thienopyridines (a loading dose of 300 to 600 mg of clopidogrel or 500 mg of ticlopidine). Post-intervention anti-platelet therapy consisted of aspirin and either 75 mg/day clopidogrel or 250 mg ticlopidine twice daily for at least 6 months. Length of clinical follow-up varied between 9 and 12 months, and length of angiographic follow-up varied between 6 and 9 months after the index procedure. Overall clinical follow-up rate was 97.4% (98.1% among patients treated with SES or PES and 96.4% among patients treated with plain BA or vascular brachytherapy). The overall angiographic follow-up rate was 88.6% (89.3% among patients treated with SES or PES and 87.6% among patients treated with plain BA or vascular brachytherapy).
Men comprised 70.4% of included patients, and 32.2% had diabetes mellitus. There were no significant differences between patients assigned to either DES or conventional treatment with regard to age, proportion of diabetics, lesion length, or vessel size (Tables 1 and 2).⇓
Odds ratios and their 95% CIs for clinical and angiographic outcomes are displayed in Figures 1 to 4.⇓⇓⇓Only fixed-effects model results are shown because of the lack of heterogeneity across trials included in this meta-analysis. With respect to the primary outcome of interest of this study, patients who underwent implantation of DES presented a marked reduction in the need for TLR. Thus, 9.5% of the patients treated with SES or PES required a repeat revascularization procedure compared with 21.2% of the patients treated with plain BA or vascular brachytherapy. As shown in Figure 1, the pooled OR was 0.35 (95% CI 0.25 to 0.49; p < 0.001), consistent with a 65% reduction in the odds for TLR with DES. There was no heterogeneity between trials (p = 0.86). There was no significant publication bias using the Begg adjusted correlation test (p > 0.99) and regression asymmetry test by Egger (p > 0.99). Omission of individual trials from the analysis did not have any relevant influence on the overall results of the analysis. Two additional analyses on TLR were also performed, one with respect to the form of conventional treatment (BA or brachytherapy) and the other with respect to the type of DES (SES or PES). When we grouped the trials according to the conventional control treatment used (BA in the ISAR-DESIRE and RIBS-II trials and brachytherapy in the SISR and TAXUS-V ISR trials), pooled ORs were 0.31 (95% CI 0.19 to 0.51) when comparing DES with BA and 0.44 (95% CI 0.28 to 0.69) when comparing DES with brachytherapy, without significant heterogeneity between them (p = 0.33). When we grouped the trials according to the type of DES used (SES in 1 of the 3 arms of the ISAR-DESIRE trial and in the RIBS-II and SISR trials, PES in 1 of the 3 arms of the ISAR-DESIRE trial and in the TAXUS-V ISR trial), pooled ORs were 0.27 (95% CI 0.18 to 0.41) when comparing the SES with conventional treatment (BA or brachytherapy) and 0.47 (95% CI 0.30 to 0.75) when comparing the PES with conventional treatment, without significant heterogeneity between them (p = 0.08).
With respect to the other clinical outcomes, no differences were observed between patients treated with DES compared with those treated with conventional treatment modalities for bare-metal in-stent restenosis. The incidence of stent thrombosis was 0.8% with DES and 1.2% with plain BA or vascular brachytherapy (OR 0.85, 95% CI 0.28 to 2.64; p = 0.79) (Fig. 2). The mortality rate was 0.8% among patients who received DES and 1.4% among patients who received conventional treatment (OR 0.63, 95% CI 0.22 to 1.85; p = 0.41) (Fig. 3). The incidence of the composite outcome of death or myocardial infarction was also comparable between patients who underwent SES or PES implantation and those who underwent plain BA or vascular brachytherapy (3.0% vs. 3.2%, OR 1.04, 95% CI 0.54 to 2.03; p = 0.90) (Fig. 4). The test for heterogeneity between trials was not significant regarding any of these study outcomes.
With respect to the angiographic outcome of interest in this study, angiographic restenosis, DES were markedly superior to conventional treatment of bare-metal in-stent restenosis. Patients treated with SES or PES had an angiographic restenosis rate significantly lower than patients treated with plain BA or vascular brachytherapy (17.0% vs. 34.9%, OR 0.36, 95% CI 0.27 to 0.49; p = 0.001) (Fig. 5).There was no significant heterogeneity between trials with respect to this outcome (p = 0.11).
A large body of evidence has established the remarkable anti-restenotic effectiveness of DES in patients and lesions with various risk profiles. Recently, there have been reports that DES are superior to conventional treatment modalities for patients with in-stent restenosis, which has been traditionally considered to be a major challenge in percutaneous coronary interventions. No DES, however, has yet been approved for this indication. By increasing statistical power, meta-analyses can increase the precision of treatment effect estimates. They are used during the development of guidelines as evidence in support of recommendations and may be useful to the regulatory process of drugs and devices. Therefore, they may contribute to accelerating the introduction of effective treatment modalities into clinical practice. We performed a meta-analysis of randomized trials evaluating the performance of SES and PES compared with plain BA and vascular brachytherapy in this group of patients. Based on the results of this meta-analysis, which demonstrated the superior efficacy of the SES and PES, it can be concluded that these DES should be considered the treatment of choice for patients with bare-metal in-stent restenosis.
In-stent restenosis occurs in up to one-half of patients treated with BMS (5). Although it has been generally considered to be a benign process, recent evidence shows that it presents with acute myocardial infarction or unstable angina in more than one-third of cases (2). In addition, there is evidence that in-stent restenosis has a negative impact on the long-term survival of patients treated with BMS (3). It has been estimated that about 60% of restenotic lesions require treatment with a repeat revascularization procedure (5), and this highlights the importance of finding effective therapeutic solutions for in-stent restenosis.
Several treatment modalities have been proposed and used for bare-metal in-stent restenosis: plain/cutting BA, repeat stenting, vascular brachytherapy, rotational directional atherectomy, pullback atherectomy, and excimer laser-based angioplasty (19). Although plain BA has been the most widely used treatment for in-stent restenosis, it is associated with high recurrent restenosis rates (31). Use of the cutting balloon technique has not provided any additional benefit (32). Repeat stent implantation, despite its improved immediate results compared with BA, has not been associated with better angiographic and clinical outcomes (33). On the other hand, there are no data to support the long-term advantage of ablative techniques (19). Therefore, BA with or without brachytherapy has been the most frequently used treatment for patients with in-stent restenosis.
Drug-eluting stents have been established as the most successful treatment for the reduction of restenosis risk in several patient and lesion subsets. Recently, some registry and randomized studies showed that SES and PES were more effective than plain BA and vascular brachytherapy in reducing recurrent restenosis rates after treatment of in-stent restenotic lesions (14–18,34). The present meta-analysis, which included all currently performed randomized studies, with a total population of 1,230 patients treated with either SES or PES versus plain BA or beta/gamma vascular brachytherapy, significantly strengthens the evidence of the superiority of DES over conventional treatment modalities for restenosis in BMS (15–18).
There was a marked reduction in the risk of restenosis with DES, as demonstrated by the 65% reduction in the odds of TLR and 64% reduction in the odds of angiographic restenosis with SES and PES compared with conventional treatment. This is consistent with the significantly greater amount of acute lumen gain and inhibition of neointimal formation with DES demonstrated in the studies included in the present meta-analysis. Further insights into the mechanisms explaining the better outcome with DES are provided from the intravascular ultrasound substudy of the RIBS-II trial (18). This study showed that patients treated with SES had a larger post-procedural lumen volume as well as a significant reduction in the volume of neointimal proliferation with SES than with plain BA at follow-up angiography (18).
The incidence of stent thrombosis was not different between the 2 treatment arms evaluated in this meta-analysis. Pooled 9- to 12-month stent thrombosis rates were 0.8% with DES and 1.2% with plain BA or vascular brachytherapy. None of the individual trials included in this meta-analysis reported any difference between DES and the conventional treatments. Similarly, patients treated with SES or PES and plain BA or vascular brachytherapy had a comparable incidence of death or the composite of death or myocardial infarction. These observations suggest a similar safety profile for these treatment strategies. On the other hand, although meta-analysis may increase statistical power and even has the potential to answer questions that might not have been posed at the time the study was conceived, the rare occurrence of these outcomes may explain the observed lack of difference in safety clinical outcomes between the 2 treatment arms in the meta-analysis. Studies including larger numbers of patients and with a longer follow-up are needed to answer questions with respect to the relative clinical safety of SES or PES and plain BA or vascular brachytherapy.
The individual trials selected in the present meta-analysis included only 1 conventional treatment to be compared with DES: either BA or vascular brachytherapy. In this all-inclusive analysis, there was no significant heterogeneity across trials; however, the ORs tended to be lower when DES were compared with BA than with brachytherapy. This is in line with the results of earlier clinical trials showing that brachytherapy is superior to plain BA in terms of reduction of recurrences in patients with in-stent restenosis (35), especially in those with long restenotic lesions (36). In addition, the ORs tended to be lower when SES instead of PES was compared with conventional treatment. This is in line with the results of a recent meta-analysis that showed the superiority of SES in terms of restenosis (13).
It should be stressed that the conclusions of the present meta-analysis are confined only to SES and PES. The effectiveness of other DES for patients with bare-metal in-stent restenosis remains to be established. On the other hand, the results of this meta-analysis, which studied patients with in-stent restenosis after implantation of BMS, cannot be extrapolated to patients with in-stent restenosis after DES placement. The optimal treatment for patients incurring restenosis after DES still remains to be elucidated by specifically designed studies (37).
Patients who develop in-stent restenosis after implantation of BMS carry a high risk for recurrence. Use of SES and PES effectively reduces the risk of recurrence and is associated with superior results compared with plain BA and vascular brachytherapy. These DES should be recommended as the treatment of choice for patients with restenosis of BMS.
- Abbreviations and Acronyms
- balloon angioplasty
- bare-metal stent(s)
- drug-eluting stent(s)
- paclitaxel-eluting stent(s)
- sirolimus-eluting stent(s)
- target lesion revascularization
- Received August 11, 2006.
- Revision received September 20, 2006.
- Accepted October 1, 2006.
- American College of Cardiology Foundation
- Radke P.W.,
- Kaiser A.,
- Frost C.,
- Sigwart U.
- Hoffmann R.,
- Mintz G.S.
- Sabate M.,
- Jimenez-Quevedo P.,
- Angiolillo D.J.,
- et al.
- Neumann F.J.,
- Desmet W.,
- Grube E.,
- et al.
- Alfonso F.,
- Perez-Vizcayno M.J.,
- Hernandez R.,
- et al.
- Smith S.C. Jr..,
- Feldman T.E.,
- Hirshfeld J.W. Jr..,
- et al.
- Silber S.,
- Albertsson P.,
- Aviles F.F.,
- et al.
- Egger M.,
- Ebrahim S.,
- Smith G.D.
- Egger M.,
- Smith G.D.
- Mantel N.,
- Haenszel W.
- Egger M.,
- Davey Smith G.,
- Schneider M.,
- Minder C.
- Albiero R.,
- Silber S.,
- Di Mario C.,
- et al.
- Alfonso F.,
- Zueco J.,
- Cequier A.,
- et al.
- Iofina E.,
- Radke P.W.,
- Skurzewski P.,
- et al.
- Waksman R.,
- Cheneau E.,
- Ajani A.E.,
- et al.