- Email: [email protected]
Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials
IJCA-14301; No of Pages 6 International Journal of Cardiology xxx (2012) xxx–xxx
Contents lists available at SciVerse ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials Ryota Sakurai a, b,⁎, Tsukasa Inajima b, Hideaki Kaneda c, Ryozo Nagai b, Hideki Hashimoto d a
Department of Planning, Information and Management, The University of Tokyo Hospital, Tokyo, Japan Department of Cardiovascular Medicine, The University of Tokyo Hospital, Tokyo, Japan c Okinaka Memorial Institute for Medical Research, Tokyo, Japan d Department of Health Economics and Epidemiology Research, School of Public Health, The University of Tokyo, Tokyo, Japan b
a r t i c l e
i n f o
Article history: Received 12 August 2011 Received in revised form 1 November 2011 Accepted 17 December 2011 Available online xxxx Keywords: Sirolimus-eluting stent ST-segment elevation myocardial infarction Randomized controlled trial Long-term follow up Meta-analysis
a b s t r a c t Background: Sirolimus-eluting stents (SES) have demonstrated more favorable outcomes compared with bare metal stents (BMS) for ST-segment elevation myocardial infarction (STEMI) within medium term follow up in randomized controlled trials (RCT). However, long-term outcomes remain unknown. Methods: We conducted a meta-analysis of RCT comparing SES with BMS in STEMI patients at long-term follow up, defined as 2 years or more after primary percutaneous coronary intervention (PCI). The clinical end points of our interest were death, recurrent myocardial infarction (MI), definite stent thrombosis (ST), and target lesion revascularization (TLR). We calculated the pooled estimate based on a fixed-effects model using Peto odds ratio (OR) for rare events. If heterogeneity was observed across an individual RCT, an analysis based on a random-effects model was performed. Results: Four RCT were included in this study, involving 1304 patients (656 patients randomized to SES and 648 patients to BMS). Up to 4 years, SES showed a significant reduction in not only TLR (OR: 0.44, 95% confidence interval (CI): 0.31 to 0.62, p b 0.001) but also mortality (OR: 0.62, 95% CI: 0.39 to 1.00, p = 0.049) compared with BMS. In contrast, the proportions of recurrent MI (OR: 0.82, 95% CI: 0.52 to 1.28, p = 0.378) and definite ST (OR: 1.13, 95% CI: 0.56 to 2.27, p = 0.740) were comparable between the 2 groups. Conclusions: In this meta-analysis of long-term RCT, primary PCI for STEMI patients with SES was associated with a decrease in mortality as well as TLR without an increase in recurrent MI or definite ST compared with BMS. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Sirolimus-eluting stents (SES) have demonstrated to be more effective and equivalently safe compared with bare metal stents (BMS) in various lesions or patients with coronary artery disease in randomized controlled trials (RCT) during a wide range of follow-up periods [1–12]. In patients with ST-segment elevation myocardial infarction (STEMI), considered to be one of the most complicated patient subsets and therefore have been excluded from the above RCT, more favorable outcomes of SES than BMS have also been reported in several RCT within medium term follow up after primary percutaneous coronary interventions (PCI) [13–17]. However, long-term efficacy and safety remain unknown. Therefore, we conducted a metaanalysis of RCT to investigate clinical outcomes of SES as compared ⁎ Corresponding author at: The Department of Planning, Information and Management, The University of Tokyo Hospital, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan. Tel.: + 81 3 5800 8685; fax: + 81 3 3813 7238. E-mail address: [email protected] (R. Sakurai).
to BMS in STEMI patients at long-term follow up, defined as 2 years or more after primary PCI. The clinical end points of our interest regarding efficacy and safety were death, recurrent myocardial infarction (MI), definite stent thrombosis (ST) according to the Academic Research Consortium definition [18], and target lesion revascularization (TLR). 2. Materials and methods We performed an electronic search for articles in any language using MEDLINE (January 2004 to August 2011), EMBASE (January 2004 to August 2011), the Cochrane Library (January 2004 to August 2011), and Web of Science (January 2004 to August 2011). Search terms included “myocardial”, “infarction”, “sirolimus”, “eluting”, “bare”, “metal”, “stent”, and “randomized”. The same terms or relevant trials were also searched on the website including the U.S. National Institute of Health (including clinicaltrials.gov), TCTMD.com, escardio.org, and pcronline.com. The final search was run in August, 2011. Eligibility criteria of studies for this meta-analysis were as follows: 1) RCT comparing SES with BMS in STEMI patients undergoing primary PCI during 2year follow up or more; and 2) reporting the number of death, recurrent MI, definite ST, and TLR. We included only RCT in this meta-analysis to increase internal validity. When multiple follow-up outcomes after 2 years were reported in the same study,
0167-5273/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2011.12.054
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
2
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
those from the latest follow up were abstracted. Two reviewers (R.S. and T.I.) performed eligibility assessment for studies to include this meta-analysis independently, and disagreements were resolved by discussion between the 2 authors. If no agreement could be reached, a third author rendered a decision. Two authors (R.S. and T.I.) extracted the data independently from each eligible RCT on study characteristics, baseline demographics of participants, procedural information, and outcome measures. From this information, the adequacy of randomization, double-blinding and the reasons for lost to follow-up were checked, and Jadad score was calculated by 2 authors (R.S. and T.I.) independently to assess the quality of each RCT [19]. Furthermore, to investigate outcome changes between medium and long term, we extracted the 1-year outcomes of each RCT. Disagreements were resolved by discussion between the 2 authors, and if no agreement could be reached, a third author rendered a decision. We calculated the pooled estimate by averaging the odds ratio (OR) of each study based on a fixed-effects model using Peto OR for rare events [20,21]. To assess heterogeneity across an individual RCT, the I2 statistics and the Cochran's Q test were performed [22]. If these results rejected the null hypothesis of homogeneity, an analysis based on a random-effects model would be performed. If the pooled estimate for the outcomes was statistically significant, we also calculated the pooled risk difference to estimate the number needed to treat, defined as 1/absolute risk difference. In addition, a cumulative meta-analysis was also performed to examine whether the pooled estimate of effect size had changed over time, as new studies had been published. To compare the long-term outcomes with those of medium term, the pooled estimate at 1 year was also calculated. The possibility of publication bias was assessed visually by a funnel plot for asymmetry plotting of the standard error of log OR against the log OR. In addition, Begg's rank correlation test [23] and Egger's linear regression test [24] were conducted as formal statistical tests for publication bias. If publication bias seemed to be present through the above process, the Duval and Tweedie's trim and fill procedure was planned to be performed to estimate the possible impact of unpublished studies on the pooled estimate. Two-sided p values of b 0.05 were considered to be statistically significant. In the tests for assessment of heterogeneity or publication bias, however, the threshold for statistical significance was defined as a p value of 0.10 [25]. Analyses were conducted using STATA 11.2 (Stata Corp., College Station, Texas) and validated using R statistical software (http://www.r-project.org). This study was conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) [26].
3. Results Fig. 1 shows a flow diagram of study selection in this metaanalysis. We excluded one study, which allocated patients to SES plus tirofiban infusion or BMS plus abciximab [27], since the effects of SES versus BMS might be masked by tirofiban and/or abciximab. Six studies (including abstracts of scientific sessions) were excluded since other reports at longer-term follow up existed in the same studies, respectively. One study reported outcomes of both SES and other types of drug-eluting stents (DES) together, precluding specific SES data retrieval [28]. Another study was not an RCT [29]. During additional web searches of the U.S. National Institute of Health, TCTMD. com, escardio.org, and pcronline.com, a study of 3-year follow up presented at European Society of Cardiology Congress 2010 was excluded since the total number of patients who had been completely followed was uncertain in some groups [30]. As a result, the latest long-term reports in 4 RCT (TYPHOON, PASEO, MISSION!, and SESAMI) were included in this study, involving 1304 patients (656 patients randomized to SES and 648 patients to BMS) [31–34]. The proportion of patients with Thrombolysis in Myocardial Infarction flow b2 was 68% in TYPHOON, 77% in PASEO, and 71% in MISSION!, respectively. Mean left ventricular ejection fraction (LVEF) of the patients was 53% in TYPHOON, and the proportion of patients with LVEF b40% was 36% in PASEO. The proportion of patients with cardiogenic shock was 15% in PASEO. Patients with overt acute heart failure and a previously documented LVEF of b30% were excluded from TYPHOON, those requiring mechanical ventilation were excluded from MISSION!, and those with cardiogenic shock were excluded from SESAMI. Table 1 shows the characteristics of each RCT. The duration of follow up was 3 to 4 years. Although the results at several follow-up periods were reported in all RCT, only the latest follow-up results were selected for comparison. All RCT were conducted in Europe or Australia, and only TYPHOON was conducted as a multicenter trial.
Fig. 1. Flow chart of study selection. DES = drug-eluting stents; RCT = randomized controlled trial(s); SES = sirolimus-eluting stent.
The method of randomization was not described in MISSION!, and none of the RCT was performed in double-blinded fashion. Reasons for patient withdrawals or dropouts were not described in SESAMI. Accordingly, Jadad score ranged from 2 to 3, indicating that the overall quality of RCT was moderate. Data analysis was conducted according to the intention-to-treat principle in each RCT, which guaranteed the appropriateness of randomization and decision-making strategy in real world clinical settings. The proportion of follow-up in TYPHOON was approximately 70% since the trial was once closed at 1-year follow up and a new informed consent had to be obtained for long-term follow up. In TYPHOON, even though most of the baseline clinical and procedural characteristics were similar between patients with and without complete follow up, lower ejection fraction and smaller in stent minimal luminal diameter were observed in patients without complete follow up [31]. On the other hand, since the proportion of follow up other than TYPHOON was relatively high, the overall proportion of follow up in all 4 RCT was 85.7% (1304 out of 1522). In TYPHOON, the proportion of patients receiving dual antiplatelet therapy (DAPT) at 4 years was comparable between SES and BMS (46/242 versus 37/236, p = 0.40). Although the actual duration of aspirin therapy was not described, the duration of clopidogrel administration was similar between SES and BMS in PASEO (181 ± 7 versus 184 ± 9 days, p = 0.14). In MISSION!, there was also no detailed information regarding aspirin, and the proportion of patients taking clopidogrel up to 12 months was comparable between SES and BMS (147/158 versus 146/152, p = 0.24). In SESAMI, the duration of DAPT was similar between SES and BMS (375 ± 12 versus 369 ± 35 days, p = not significant). Fig. 2 shows the long-term clinical outcomes of 4 RCT up to 4 years. In the pooled analysis, SES showed a significant reduction in proportion of TLR compared with BMS (OR: 0.44, 95% confidence interval (CI): 0.31 to 0.62, p b 0.001). Moreover, SES also showed a significantly reduced proportion of mortality as compared to BMS (OR: 0.62, 95% CI: 0.39 to 1.00, p = 0.049). On the other hand, the proportions of recurrent MI (OR: 0.82, 95% CI: 0.52 to 1.28, p = 0.378) and definite ST (OR: 1.13, 95% CI: 0.56 to 2.27, p = 0.740) were comparable between the 2 stents. An analysis based on a random-effects model was not performed since there was no evidence of heterogeneity across the RCT in each outcome (I 2 = 0.0%, p > 0.10 for each outcome, respectively, Fig. 2). The pooled risk difference was − 0.08 (95% CI: − 0.11 to − 0.05) in TLR and − 0.03 (95% CI: − 0.05 to − 0.00) in mortality; that is, the
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
3
Table 1 Characteristics of RCT.
Number of patients Type of BMS
TYPHOON [31]⁎
PASEO [32]⁎
MISSION! [33]⁎
SES:BMS = 251:250 Any commercially available BMS 4 48
SES:BMS = 90:90 Any commercially available BMS 4 1
SES:BMS = 158:152 MULTI-LINK VISION (Abbott Vascular, Santa Clara, California) 3 1
SES:BMS = 157:156 BX VELOCITY (Cordis Corporation, Miami Lakes, Florida) 3 1
62b
59b
63b
77.7 (241/310) 9.7 (30/310) 28.1 (87/310) 54.5 (169/310) 54.8 (170/310)
80.0 20.3 57.8 54.4 49.7
≦12
b9
≦ 12
6
12
12
SES:BMS = 5:19 SES:BMS = 7:11 SES:BMS = 8:12 SES:BMS = 1:2 Yes
SES:BMS = 10:19 SES:BMS = 7:10 SES:BMS = 12:17 SES:BMS = 4:1 Yes
SES:BMS = 11:21 SES:BMS = 5:8 SES:BMS = 4:4 SES:BMS = 3:2 Yes
Yes Yes No Yes
Not described Yes No Yes
Yes Yes No Yes
Yes 100 3
Yes 100 2
Without reasons 97.8 2
Follow-up period (years) Number of centers Mean age 59b (years) Male (%) 78.6 (394/501) Diabetes (%) 15.2 (76/501) Hypertension (%) 41.9 (210/501) Smoking (%) 51.1 (241/472) LAD culprit 43.5 (218/501) (%) Ischemia time b 12 (hours) Duration of >6 DAPT (months) TLR SES:BMS = 19:37 Death SES:BMS = 10:16 Recurrent MI SES:BMS = 13:11 Definite ST SES:BMS = 9:10 Described as Yes randomized Appropriate randomization Yes Intention-to-treat Yes Described as double-blinded No Described blinding or independence of Yes outcome assessments Describing withdrawals and dropouts Yes Follow-up proportion (%) 70.4 Jadad score 3
71.1 26.7 26.1 25.6 51.7
(128/180) (48/180)b (47/180)b (46/180)b (93/180)b
b
a
SESAMI [34]⁎
c
(256/320)c (65/320)c (185/320)c (174/320)c (159/320)c
BMS = bare metal stent; DAPT = dual antiplatelet therapy; LAD = left anterior descending artery; MI = myocardial infarction; RCT = randomized controlled trials; SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization; TVF = target vessel failure; TVR = target vessel revascularization. ⁎ Some data were also abstracted from the previous reports [13,14,16,44]. a Data of 310 patients in whom all clinical event data were not available. b Numbers were estimated based on the table in each study. c Data of baseline 320 patients.
Fig. 2. Forest plot. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. BMS = bare metal stent; CI = confidence intervals; MI = myocardial infarction; OR = odds ratio(s); SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
4
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
Fig. 3. Cumulative meta-analysis. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. BMS = bare metal stent; CI = confidence intervals; MI = myocardial infarction; OR = odds ratio(s); SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization.
number needed to treat was 13 (95% CI: 9 to 20) for TLR and 33 (95% CI: 20 to ∞) for mortality. In a cumulative meta-analysis, the pooled OR for TLR had been increasing but 95% CI had been narrowing as
new studies had been published. The pooled OR for death had been almost the same over time, but 95% CI had been narrowing as new studies had been published. On the other hand, the pooled OR for
Fig. 4. Funnel plot. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. MI = myocardial infarction; or = odds ratio; s.e. = standard error; ST = stent thrombosis; TLR = target lesion revascularization.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
recurrent MI and definite ST had been fluctuating, as new studies had been published (Fig. 3). At 1 year, SES also showed a significant reduction in TLR as compared to BMS (OR: 0.35, 95% CI: 0.24 to 0.50, p b 0.001). In contrast, SES demonstrated no difference in death (OR: 0.63, 95% CI: 0.33 to 1.22, p = 0.172), recurrent MI (OR: 0.67, 95% CI: 0.38 to 1.18, p = 0.166), or definite ST (OR: 0.64, 95% CI: 0.25 to 1.61, p = 0.341) compared with BMS. A funnel plot for each outcome seemed not asymmetric, even by plotting a small number of RCT included in this meta-analysis (Fig. 4). In addition, neither Begg's rank correlation test nor Egger's linear regression test suggested the presence of publication bias for death (p= 0.333 and p = 0.766, respectively), recurrent MI (p= 0.750 and p = 0.808, respectively), or definite ST (p= 0.750 and p = 0.706, respectively). In contrast, Begg's rank correlation test, but not Egger's linear regression test, indicated the possibility of publication bias for TLR (p= 0.083 and p = 0.368, respectively). However, there was no additional plot for TLR by the Duval and Tweedie's trim and fill procedure. 4. Discussion Through this meta-analysis of 4 RCT at long-term follow up after primary PCI in 1304 STEMI patients (SES:BMS = 656:648), our findings suggest that SES patient outcomes were associated with the decreased proportions of TLR and mortality compared with BMS patients. On the other hand, the proportions of recurrent MI and definite ST were comparable between the 2 stents. These findings were derived from RCT that were relatively of moderate quality, homogeneous, and possibly free from publication bias. A meta-analysis of RCT and registries demonstrated reduced target vessel revascularization (TVR) in patients treated with DES (both SES and paclitaxel-eluting stents) compared with BMS patients within 2-year follow up [35]. The pooled risk difference of TLR at 4 years in our study was similar to that of TVR at 2 years in the RCT included in the previous study (−7.0%, 95% CI: − 9.2% to −4.7%). Furthermore, in addition to the small number needed to treat, the pooled OR for TLR at 4 years changed only slightly from 1 year in our study. Therefore, we may validate and extend the efficacy of SES observed during medium term in STEMI patients up to 4 years after primary PCI by the meta-analysis of strictly selected RCT. Our study also suggests the possibility of a reduction in mortality of SES patients compared to BMS patients. Some previous registries have reported no difference in mortality between SES and BMS [36,37]. In these registries, however, in addition to the relatively small number of included patients (369 and 716 patients, respectively), the cohort in which patients were implanted with BMS was different from that of SES. On the other hand, another registry, which included a larger number of patients (1926 patients) and was conducted in 1 cohort, showed a significant reduction in mortality in DES compared with BMS [38]. However, in addition to the fact that observational studies are likely to be influenced by several unobserved biases, the stable pooled OR with the narrowing 95% CI in our cumulative meta-analysis may support the reliable effects of SES on mortality. With regard to recurrent MI and definite ST, the pooled estimates worsened for SES versus BMS from 1 year to 4 years, respectively, even though the overall proportion of both events was comparable. These deteriorations did not influence the change of pooled estimate for mortality between 1 and 4 years. Follow-up angiography was mandatory in all patients in MISSION! at 9 months and in SESAMI at 1 year, and in a part of patients in TYPHOON at 8 months. The use of routine angiography may increase the proportion of angiography-related outcomes, such as TLR and definite ST [39]. However, the proportion of TLR at 1 year was comparable between the RCT with and without routine follow-up angiography (MISSION! versus SESAMI versus TYPHOON versus PASEO = 3.2% versus 4.5% versus 4.2% versus 3.3% in SES and 11.2% versus 11.8% versus
5
10.9% versus 14.4% in BMS). A similar tendency was observed with regard to definite ST as well (0.6% versus 1.3% versus 1.1% versus 0% in SES and 0.7% versus 0.7% versus 2.2% versus 1.1% in BMS). Furthermore, routine angiography at medium term follow up may also not influence the long-term proportions (TLR: MISSION! versus SESAMI versus TYPHOON versus PASEO = 6.3% versus 7.0% versus 7.6% versus 5.6% in SES and 12.5% versus 13.5% versus 14.8% versus 21.1% in BMS; definite ST: 2.5% versus 1.9% versus 3.6% versus 1.1% in SES and 0.7% versus 1.3% versus 4.0% versus 2.2% in BMS). Thus, the findings of our study were irrespective of mandated angiographic follow up. Although the presence of publication bias was not suggested for most of the end points by either Begg's rank correlation test or Egger's linear regression test and for TLR by either Egger's linear regression test or the Duval and Tweedie's trim and fill procedure, the number of RCT included in this meta-analysis was relatively small. Other than the excluded studies described above, long-term outcomes were not reported in 1 RCT registered in clinicaltrials.gov, even though the proportions of random allocation between SES and BMS versus between with and without abciximab were uncertain [40]. Furthermore, no long-term results were reported in another RCT in which medium term outcomes were presented previously [15]. Therefore, our findings may change if long-term results are reported from all these RCT. Several limitations are considered to exist in this study. First of all, the number of RCT and patients included in this study was relatively small in terms of meta-analysis. Although each RCT met the requirements for this study, the quality varied across individual RCT. The proportion of patients lost to follow-up was higher in TYPHOON, which included the largest number of patients and was the only multicenter RCT conducted in various countries. This fact may lead to an inaccurate pooled estimate by selection bias since a higher proportion of dropouts seems to cause an increase in number of unwitnessed events. In fact, OR for death decreased from 1 year to longer-term follow up only in TYPHOON. None of the RCT was double-blinded, which may lead to selection bias, response bias, etc. However, while trials with open assessment of the outcomes were related to overestimation of treatment effects [41], all clinical events were adjudicated by the reviewers who were blinded to the treatment assignment or who were members of an independent clinical events committee in each RCT. In addition, it may not be critical with regard to “hard” end points such as death. Detailed data regarding ischemia time or duration of DAPT were not available in all RCT. The patient population was not the same across the studies, which is the nature of metaanalysis. However, heterogeneity across RCT was not statistically proven regarding each outcome. Our findings may not be generalizable for patients in the real world due to the limited eligibilities of included patients in each RCT. A major potential limitation of DES implantation in patients with acute MI seems that the clinicians do not always have time to thoroughly rule out contraindications to long-term DAPT that is needed after DES implantation. This could potentially expose these patients to the risk of ST if they subsequently are unable to receive DAPT. The ability to detect patients unlikely to take DAPT is probably improved in RCT and may potentially be missed in real world patients undergoing emergent PCI. Finally, findings from pooled-analysis with study-level data may be more biased and misleading than those with patient-level data. For instance, conclusions from analyses with study-level data and with patient-level data differed from each other in previous reports [42,43]. In conclusion, from this meta-analysis of RCT featuring long-term follow up, primary PCI with SES for STEMI patients may be associated with a decrease in not only TLR but also mortality without an increase in recurrent MI or definite ST compared to those treated with BMS. Further longer-term results from ongoing RCT may be warranted to validate the effectiveness and safety issues of SES in STEMI patients.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
6
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
Acknowledgments The authors thank Heidi N. Bonneau, RN, MS, CAA for her editorial review of the manuscript. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [45]. References [1] Morice MC, Serruys PW, Sousa JE, et al. A randomized comparison of a sirolimuseluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–80. [2] Morice MC, Serruys PW, Barragan P, et al. Long-term clinical outcomes with sirolimus-eluting coronary stents: five-year results of the RAVEL trial. J Am Coll Cardiol 2007;50:1299–304. [3] Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–23. [4] Weisz G, Leon MB, Holmes Jr DR, et al. Five-year follow-up after sirolimus-eluting stent implantation results of the SIRIUS (Sirolimus-Eluting Stent in De-Novo Native Coronary Lesions) trial. J Am Coll Cardiol 2009;53:1488–97. [5] Schofer J, Schluter M, Gershlick AH, et al. Sirolimus-eluting stents for treatment of patients with long atherosclerotic lesions in small coronary arteries: double-blind, randomised controlled trial (E-SIRIUS). Lancet 2003;362:1093–9. [6] Schampaert E, Cohen EA, Schluter M, et al. The Canadian study of the sirolimuseluting stent in the treatment of patients with long de novo lesions in small native coronary arteries (C-SIRIUS). J Am Coll Cardiol 2004;43:1110–5. [7] Suttorp MJ, Laarman GJ, Rahel BM, et al. Primary Stenting of Totally Occluded Native Coronary Arteries II (PRISON II): a randomized comparison of bare metal stent implantation with sirolimus-eluting stent implantation for the treatment of total coronary occlusions. Circulation 2006;114:921–8. [8] Rahel BM, Laarman GJ, Kelder JC, Ten Berg JM, Suttorp MJ. Three-year clinical outcome after primary stenting of totally occluded native coronary arteries: a randomized comparison of bare-metal stent implantation with sirolimus-eluting stent implantation for the treatment of total coronary occlusions (Primary Stenting of Totally Occluded Native Coronary Arteries [PRISON] II study). Am Heart J 2009;157: 149–55. [9] Baumgart D, Klauss V, Baer F, et al. One-year results of the SCORPIUS study: a German multicenter investigation on the effectiveness of sirolimus-eluting stents in diabetic patients. J Am Coll Cardiol 2007;50:1627–34. [10] Jimenez-Quevedo P, Sabate M, Angiolillo DJ, et al. Long-term clinical benefit of sirolimus-eluting stent implantation in diabetic patients with de novo coronary stenoses: long-term results of the DIABETES trial. Eur Heart J 2007;28:1946–52. [11] Kelbaek H, Thuesen L, Helqvist S, et al. The Stenting Coronary Arteries in Nonstress/Benestent Disease (SCANDSTENT) trial. J Am Coll Cardiol 2006;47:449–55. [12] Kelbaek H, Klovgaard L, Helqvist S, et al. Long-term outcome in patients treated with sirolimus-eluting stents in complex coronary artery lesions: 3-year results of the SCANDSTENT (Stenting Coronary Arteries in Non-Stress/Benestent Disease) trial. J Am Coll Cardiol 2008;51:2011–6. [13] Spaulding C, Henry P, Teiger E, et al. Sirolimus-eluting versus uncoated stents in acute myocardial infarction. N Engl J Med 2006;355:1093–104. [14] Menichelli M, Parma A, Pucci E, et al. Randomized trial of Sirolimus-Eluting Stent Versus Bare-Metal Stent in Acute Myocardial Infarction (SESAMI). J Am Coll Cardiol 2007;49:1924–30. [15] Diaz de la Llera LS, Ballesteros S, Nevado J, et al. Sirolimus-eluting stents compared with standard stents in the treatment of patients with primary angioplasty. Am Heart J 2007;154:164.e1–6. [16] van der Hoeven BL, Liem SS, Jukema JW, et al. Sirolimus-eluting stents versus bare-metal stents in patients with ST-segment elevation myocardial infarction: 9-month angiographic and intravascular ultrasound results and 12-month clinical outcome results from the MISSION! Intervention study. J Am Coll Cardiol 2008;51:618–26. [17] Valgimigli M, Campo G, Percoco G, et al. Comparison of angioplasty with infusion of tirofiban or abciximab and with implantation of sirolimus-eluting or uncoated stents for acute myocardial infarction: the MULTISTRATEGY randomized trial. JAMA 2008;299:1788–99. [18] Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation 2007;115:2344–51. [19] Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1–12. [20] Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004;23: 1351–75. [21] Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med 2007;26:53–77.
[22] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [23] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101. [24] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [25] Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 2000;53:1119–29. [26] Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151:W65–94. [27] Valgimigli M, Campo G, Arcozzi C, et al. Two-year clinical follow-up after sirolimus-eluting versus bare-metal stent implantation assisted by systematic glycoprotein IIb/IIIa Inhibitor Infusion in patients with myocardial infarction: results from the STRATEGY study. J Am Coll Cardiol 2007;50:138–45. [28] Kelbaek H, Thuesen L, Helqvist S, et al. Drug-eluting versus bare metal stents in patients with st-segment-elevation myocardial infarction: eight-month followup in the Drug Elution and Distal Protection in Acute Myocardial Infarction (DEDICATION) trial. Circulation 2008;118:1155–62. [29] Yoshikawa D, Isobe S, Umeda H, et al. Three-year prognosis of Japanese patients with ST-elevation myocardial infarction treated with sirolimus-eluting stents. Coron Artery Dis 2009;20:422–7. [30] Valgimigli M. 3-year follow-up of the multicentre evaluation of single high-dose bolus tirofiban versus abciximab with sirolimus eluting stent or bare metal stent in acute myocardial infarction study. Paper presented at: European Society of Cardiology Congress 2010. Sweden: Stockholm; August 29, 2010. [31] Spaulding C, Teiger E, Commeau P, et al. Four-year follow-up of TYPHOON (trial to assess the use of the CYPHer sirolimus-eluting coronary stent in acute myocardial infarction treated with BallOON angioplasty). JACC Cardiovasc Interv 2011;4: 14–23. [32] Di Lorenzo E, Sauro R, Varricchio A, et al. Benefits of drug-eluting stents as compared to bare metal stent in ST-segment elevation myocardial infarction: four year results of the PaclitAxel or Sirolimus-Eluting stent vs bare metal stent in primary angiOplasty (PASEO) randomized trial. Am Heart J 2009;158:e43–50. [33] Atary JZ, van der Hoeven BL, Liem SS, et al. Three-year outcome of sirolimus-eluting versus bare-metal stents for the treatment of ST-segment elevation myocardial infarction (from the MISSION! Intervention study). Am J Cardiol 2010;106:4–12. [34] Violini R, Musto C, De Felice F, et al. Maintenance of long-term clinical benefit with sirolimus-eluting stents in patients with ST-segment elevation myocardial infarction 3-year results of the SESAMI (sirolimus-eluting stent versus baremetal stent in acute myocardial infarction) trial. J Am Coll Cardiol 2010;55:810–4. [35] Brar SS, Leon MB, Stone GW, et al. Use of drug-eluting stents in acute myocardial infarction: a systematic review and meta-analysis. J Am Coll Cardiol 2009;53: 1677–89. [36] Daemen J, Tanimoto S, Garcia-Garcia HM, et al. Comparison of three-year clinical outcome of sirolimus- and paclitaxel-eluting stents versus bare metal stents in patients with ST-segment elevation myocardial infarction (from the RESEARCH and T-SEARCH Registries). Am J Cardiol 2007;99:1027–32. [37] Kukreja N, Onuma Y, Garcia-Garcia H, Daemen J, van Domburg R, Serruys PW. Primary percutaneous coronary intervention for acute myocardial infarction: long-term outcome after bare metal and drug-eluting stent implantation. Circ Cardiovasc Interv 2008;1:103–10. [38] Hannan EL, Racz M, Walford G, et al. Drug-eluting versus bare-metal stents in the treatment of patients with ST-segment elevation myocardial infarction. JACC Cardiovasc Interv 2008;1:129–35. [39] Pinto DS, Stone GW, Ellis SG, et al. Impact of routine angiographic follow-up on the clinical benefits of paclitaxel-eluting stents: results from the TAXUS-IV trial. J Am Coll Cardiol 2006;48:32–6. [40] Michels R, Wijnbergen I. A comparison of drug eluting stents and bare metal stents with or without abciximab in ST-segment elevation myocardial Infarction, The Eindhoven Reperfusion study. Paper presented at: TCT09; September 25, 2009; San Francisco, CA; 2009. [41] Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA 1999;282:1054–60. [42] Nordmann AJ, Briel M, Bucher HC. Mortality in randomized controlled trials comparing drug-eluting vs. bare metal stents in coronary artery disease: a meta-analysis. Eur Heart J 2006;27:2784–814. [43] Holmes Jr DR, Moses JW, Schofer J, Morice MC, Schampaert E, Leon MB. Cause of death with bare metal and sirolimus-eluting stents. Eur Heart J 2006;27:2815–22. [44] Di Lorenzo E, De Luca G, Sauro R, et al. The PASEO (PaclitAxel or Sirolimus-Eluting Stent Versus Bare Metal Stent in Primary Angioplasty) randomized trial. JACC Cardiovasc Interv 2009;2:515–23. [45] Shewan LG, Coats AJ. Ethics in the authorship and publishing of scientific articles. Int J Cardiol 2010;144:1–2.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
Contents lists available at SciVerse ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials Ryota Sakurai a, b,⁎, Tsukasa Inajima b, Hideaki Kaneda c, Ryozo Nagai b, Hideki Hashimoto d a
Department of Planning, Information and Management, The University of Tokyo Hospital, Tokyo, Japan Department of Cardiovascular Medicine, The University of Tokyo Hospital, Tokyo, Japan c Okinaka Memorial Institute for Medical Research, Tokyo, Japan d Department of Health Economics and Epidemiology Research, School of Public Health, The University of Tokyo, Tokyo, Japan b
a r t i c l e
i n f o
Article history: Received 12 August 2011 Received in revised form 1 November 2011 Accepted 17 December 2011 Available online xxxx Keywords: Sirolimus-eluting stent ST-segment elevation myocardial infarction Randomized controlled trial Long-term follow up Meta-analysis
a b s t r a c t Background: Sirolimus-eluting stents (SES) have demonstrated more favorable outcomes compared with bare metal stents (BMS) for ST-segment elevation myocardial infarction (STEMI) within medium term follow up in randomized controlled trials (RCT). However, long-term outcomes remain unknown. Methods: We conducted a meta-analysis of RCT comparing SES with BMS in STEMI patients at long-term follow up, defined as 2 years or more after primary percutaneous coronary intervention (PCI). The clinical end points of our interest were death, recurrent myocardial infarction (MI), definite stent thrombosis (ST), and target lesion revascularization (TLR). We calculated the pooled estimate based on a fixed-effects model using Peto odds ratio (OR) for rare events. If heterogeneity was observed across an individual RCT, an analysis based on a random-effects model was performed. Results: Four RCT were included in this study, involving 1304 patients (656 patients randomized to SES and 648 patients to BMS). Up to 4 years, SES showed a significant reduction in not only TLR (OR: 0.44, 95% confidence interval (CI): 0.31 to 0.62, p b 0.001) but also mortality (OR: 0.62, 95% CI: 0.39 to 1.00, p = 0.049) compared with BMS. In contrast, the proportions of recurrent MI (OR: 0.82, 95% CI: 0.52 to 1.28, p = 0.378) and definite ST (OR: 1.13, 95% CI: 0.56 to 2.27, p = 0.740) were comparable between the 2 groups. Conclusions: In this meta-analysis of long-term RCT, primary PCI for STEMI patients with SES was associated with a decrease in mortality as well as TLR without an increase in recurrent MI or definite ST compared with BMS. © 2011 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Sirolimus-eluting stents (SES) have demonstrated to be more effective and equivalently safe compared with bare metal stents (BMS) in various lesions or patients with coronary artery disease in randomized controlled trials (RCT) during a wide range of follow-up periods [1–12]. In patients with ST-segment elevation myocardial infarction (STEMI), considered to be one of the most complicated patient subsets and therefore have been excluded from the above RCT, more favorable outcomes of SES than BMS have also been reported in several RCT within medium term follow up after primary percutaneous coronary interventions (PCI) [13–17]. However, long-term efficacy and safety remain unknown. Therefore, we conducted a metaanalysis of RCT to investigate clinical outcomes of SES as compared ⁎ Corresponding author at: The Department of Planning, Information and Management, The University of Tokyo Hospital, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-8655, Japan. Tel.: + 81 3 5800 8685; fax: + 81 3 3813 7238. E-mail address: [email protected] (R. Sakurai).
to BMS in STEMI patients at long-term follow up, defined as 2 years or more after primary PCI. The clinical end points of our interest regarding efficacy and safety were death, recurrent myocardial infarction (MI), definite stent thrombosis (ST) according to the Academic Research Consortium definition [18], and target lesion revascularization (TLR). 2. Materials and methods We performed an electronic search for articles in any language using MEDLINE (January 2004 to August 2011), EMBASE (January 2004 to August 2011), the Cochrane Library (January 2004 to August 2011), and Web of Science (January 2004 to August 2011). Search terms included “myocardial”, “infarction”, “sirolimus”, “eluting”, “bare”, “metal”, “stent”, and “randomized”. The same terms or relevant trials were also searched on the website including the U.S. National Institute of Health (including clinicaltrials.gov), TCTMD.com, escardio.org, and pcronline.com. The final search was run in August, 2011. Eligibility criteria of studies for this meta-analysis were as follows: 1) RCT comparing SES with BMS in STEMI patients undergoing primary PCI during 2year follow up or more; and 2) reporting the number of death, recurrent MI, definite ST, and TLR. We included only RCT in this meta-analysis to increase internal validity. When multiple follow-up outcomes after 2 years were reported in the same study,
0167-5273/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijcard.2011.12.054
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
2
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
those from the latest follow up were abstracted. Two reviewers (R.S. and T.I.) performed eligibility assessment for studies to include this meta-analysis independently, and disagreements were resolved by discussion between the 2 authors. If no agreement could be reached, a third author rendered a decision. Two authors (R.S. and T.I.) extracted the data independently from each eligible RCT on study characteristics, baseline demographics of participants, procedural information, and outcome measures. From this information, the adequacy of randomization, double-blinding and the reasons for lost to follow-up were checked, and Jadad score was calculated by 2 authors (R.S. and T.I.) independently to assess the quality of each RCT [19]. Furthermore, to investigate outcome changes between medium and long term, we extracted the 1-year outcomes of each RCT. Disagreements were resolved by discussion between the 2 authors, and if no agreement could be reached, a third author rendered a decision. We calculated the pooled estimate by averaging the odds ratio (OR) of each study based on a fixed-effects model using Peto OR for rare events [20,21]. To assess heterogeneity across an individual RCT, the I2 statistics and the Cochran's Q test were performed [22]. If these results rejected the null hypothesis of homogeneity, an analysis based on a random-effects model would be performed. If the pooled estimate for the outcomes was statistically significant, we also calculated the pooled risk difference to estimate the number needed to treat, defined as 1/absolute risk difference. In addition, a cumulative meta-analysis was also performed to examine whether the pooled estimate of effect size had changed over time, as new studies had been published. To compare the long-term outcomes with those of medium term, the pooled estimate at 1 year was also calculated. The possibility of publication bias was assessed visually by a funnel plot for asymmetry plotting of the standard error of log OR against the log OR. In addition, Begg's rank correlation test [23] and Egger's linear regression test [24] were conducted as formal statistical tests for publication bias. If publication bias seemed to be present through the above process, the Duval and Tweedie's trim and fill procedure was planned to be performed to estimate the possible impact of unpublished studies on the pooled estimate. Two-sided p values of b 0.05 were considered to be statistically significant. In the tests for assessment of heterogeneity or publication bias, however, the threshold for statistical significance was defined as a p value of 0.10 [25]. Analyses were conducted using STATA 11.2 (Stata Corp., College Station, Texas) and validated using R statistical software (http://www.r-project.org). This study was conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) [26].
3. Results Fig. 1 shows a flow diagram of study selection in this metaanalysis. We excluded one study, which allocated patients to SES plus tirofiban infusion or BMS plus abciximab [27], since the effects of SES versus BMS might be masked by tirofiban and/or abciximab. Six studies (including abstracts of scientific sessions) were excluded since other reports at longer-term follow up existed in the same studies, respectively. One study reported outcomes of both SES and other types of drug-eluting stents (DES) together, precluding specific SES data retrieval [28]. Another study was not an RCT [29]. During additional web searches of the U.S. National Institute of Health, TCTMD. com, escardio.org, and pcronline.com, a study of 3-year follow up presented at European Society of Cardiology Congress 2010 was excluded since the total number of patients who had been completely followed was uncertain in some groups [30]. As a result, the latest long-term reports in 4 RCT (TYPHOON, PASEO, MISSION!, and SESAMI) were included in this study, involving 1304 patients (656 patients randomized to SES and 648 patients to BMS) [31–34]. The proportion of patients with Thrombolysis in Myocardial Infarction flow b2 was 68% in TYPHOON, 77% in PASEO, and 71% in MISSION!, respectively. Mean left ventricular ejection fraction (LVEF) of the patients was 53% in TYPHOON, and the proportion of patients with LVEF b40% was 36% in PASEO. The proportion of patients with cardiogenic shock was 15% in PASEO. Patients with overt acute heart failure and a previously documented LVEF of b30% were excluded from TYPHOON, those requiring mechanical ventilation were excluded from MISSION!, and those with cardiogenic shock were excluded from SESAMI. Table 1 shows the characteristics of each RCT. The duration of follow up was 3 to 4 years. Although the results at several follow-up periods were reported in all RCT, only the latest follow-up results were selected for comparison. All RCT were conducted in Europe or Australia, and only TYPHOON was conducted as a multicenter trial.
Fig. 1. Flow chart of study selection. DES = drug-eluting stents; RCT = randomized controlled trial(s); SES = sirolimus-eluting stent.
The method of randomization was not described in MISSION!, and none of the RCT was performed in double-blinded fashion. Reasons for patient withdrawals or dropouts were not described in SESAMI. Accordingly, Jadad score ranged from 2 to 3, indicating that the overall quality of RCT was moderate. Data analysis was conducted according to the intention-to-treat principle in each RCT, which guaranteed the appropriateness of randomization and decision-making strategy in real world clinical settings. The proportion of follow-up in TYPHOON was approximately 70% since the trial was once closed at 1-year follow up and a new informed consent had to be obtained for long-term follow up. In TYPHOON, even though most of the baseline clinical and procedural characteristics were similar between patients with and without complete follow up, lower ejection fraction and smaller in stent minimal luminal diameter were observed in patients without complete follow up [31]. On the other hand, since the proportion of follow up other than TYPHOON was relatively high, the overall proportion of follow up in all 4 RCT was 85.7% (1304 out of 1522). In TYPHOON, the proportion of patients receiving dual antiplatelet therapy (DAPT) at 4 years was comparable between SES and BMS (46/242 versus 37/236, p = 0.40). Although the actual duration of aspirin therapy was not described, the duration of clopidogrel administration was similar between SES and BMS in PASEO (181 ± 7 versus 184 ± 9 days, p = 0.14). In MISSION!, there was also no detailed information regarding aspirin, and the proportion of patients taking clopidogrel up to 12 months was comparable between SES and BMS (147/158 versus 146/152, p = 0.24). In SESAMI, the duration of DAPT was similar between SES and BMS (375 ± 12 versus 369 ± 35 days, p = not significant). Fig. 2 shows the long-term clinical outcomes of 4 RCT up to 4 years. In the pooled analysis, SES showed a significant reduction in proportion of TLR compared with BMS (OR: 0.44, 95% confidence interval (CI): 0.31 to 0.62, p b 0.001). Moreover, SES also showed a significantly reduced proportion of mortality as compared to BMS (OR: 0.62, 95% CI: 0.39 to 1.00, p = 0.049). On the other hand, the proportions of recurrent MI (OR: 0.82, 95% CI: 0.52 to 1.28, p = 0.378) and definite ST (OR: 1.13, 95% CI: 0.56 to 2.27, p = 0.740) were comparable between the 2 stents. An analysis based on a random-effects model was not performed since there was no evidence of heterogeneity across the RCT in each outcome (I 2 = 0.0%, p > 0.10 for each outcome, respectively, Fig. 2). The pooled risk difference was − 0.08 (95% CI: − 0.11 to − 0.05) in TLR and − 0.03 (95% CI: − 0.05 to − 0.00) in mortality; that is, the
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
3
Table 1 Characteristics of RCT.
Number of patients Type of BMS
TYPHOON [31]⁎
PASEO [32]⁎
MISSION! [33]⁎
SES:BMS = 251:250 Any commercially available BMS 4 48
SES:BMS = 90:90 Any commercially available BMS 4 1
SES:BMS = 158:152 MULTI-LINK VISION (Abbott Vascular, Santa Clara, California) 3 1
SES:BMS = 157:156 BX VELOCITY (Cordis Corporation, Miami Lakes, Florida) 3 1
62b
59b
63b
77.7 (241/310) 9.7 (30/310) 28.1 (87/310) 54.5 (169/310) 54.8 (170/310)
80.0 20.3 57.8 54.4 49.7
≦12
b9
≦ 12
6
12
12
SES:BMS = 5:19 SES:BMS = 7:11 SES:BMS = 8:12 SES:BMS = 1:2 Yes
SES:BMS = 10:19 SES:BMS = 7:10 SES:BMS = 12:17 SES:BMS = 4:1 Yes
SES:BMS = 11:21 SES:BMS = 5:8 SES:BMS = 4:4 SES:BMS = 3:2 Yes
Yes Yes No Yes
Not described Yes No Yes
Yes Yes No Yes
Yes 100 3
Yes 100 2
Without reasons 97.8 2
Follow-up period (years) Number of centers Mean age 59b (years) Male (%) 78.6 (394/501) Diabetes (%) 15.2 (76/501) Hypertension (%) 41.9 (210/501) Smoking (%) 51.1 (241/472) LAD culprit 43.5 (218/501) (%) Ischemia time b 12 (hours) Duration of >6 DAPT (months) TLR SES:BMS = 19:37 Death SES:BMS = 10:16 Recurrent MI SES:BMS = 13:11 Definite ST SES:BMS = 9:10 Described as Yes randomized Appropriate randomization Yes Intention-to-treat Yes Described as double-blinded No Described blinding or independence of Yes outcome assessments Describing withdrawals and dropouts Yes Follow-up proportion (%) 70.4 Jadad score 3
71.1 26.7 26.1 25.6 51.7
(128/180) (48/180)b (47/180)b (46/180)b (93/180)b
b
a
SESAMI [34]⁎
c
(256/320)c (65/320)c (185/320)c (174/320)c (159/320)c
BMS = bare metal stent; DAPT = dual antiplatelet therapy; LAD = left anterior descending artery; MI = myocardial infarction; RCT = randomized controlled trials; SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization; TVF = target vessel failure; TVR = target vessel revascularization. ⁎ Some data were also abstracted from the previous reports [13,14,16,44]. a Data of 310 patients in whom all clinical event data were not available. b Numbers were estimated based on the table in each study. c Data of baseline 320 patients.
Fig. 2. Forest plot. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. BMS = bare metal stent; CI = confidence intervals; MI = myocardial infarction; OR = odds ratio(s); SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
4
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
Fig. 3. Cumulative meta-analysis. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. BMS = bare metal stent; CI = confidence intervals; MI = myocardial infarction; OR = odds ratio(s); SES = sirolimus-eluting stent; ST = stent thrombosis; TLR = target lesion revascularization.
number needed to treat was 13 (95% CI: 9 to 20) for TLR and 33 (95% CI: 20 to ∞) for mortality. In a cumulative meta-analysis, the pooled OR for TLR had been increasing but 95% CI had been narrowing as
new studies had been published. The pooled OR for death had been almost the same over time, but 95% CI had been narrowing as new studies had been published. On the other hand, the pooled OR for
Fig. 4. Funnel plot. (A) TLR; (B) mortality; (C) recurrent MI; and (D) definite ST. MI = myocardial infarction; or = odds ratio; s.e. = standard error; ST = stent thrombosis; TLR = target lesion revascularization.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
recurrent MI and definite ST had been fluctuating, as new studies had been published (Fig. 3). At 1 year, SES also showed a significant reduction in TLR as compared to BMS (OR: 0.35, 95% CI: 0.24 to 0.50, p b 0.001). In contrast, SES demonstrated no difference in death (OR: 0.63, 95% CI: 0.33 to 1.22, p = 0.172), recurrent MI (OR: 0.67, 95% CI: 0.38 to 1.18, p = 0.166), or definite ST (OR: 0.64, 95% CI: 0.25 to 1.61, p = 0.341) compared with BMS. A funnel plot for each outcome seemed not asymmetric, even by plotting a small number of RCT included in this meta-analysis (Fig. 4). In addition, neither Begg's rank correlation test nor Egger's linear regression test suggested the presence of publication bias for death (p= 0.333 and p = 0.766, respectively), recurrent MI (p= 0.750 and p = 0.808, respectively), or definite ST (p= 0.750 and p = 0.706, respectively). In contrast, Begg's rank correlation test, but not Egger's linear regression test, indicated the possibility of publication bias for TLR (p= 0.083 and p = 0.368, respectively). However, there was no additional plot for TLR by the Duval and Tweedie's trim and fill procedure. 4. Discussion Through this meta-analysis of 4 RCT at long-term follow up after primary PCI in 1304 STEMI patients (SES:BMS = 656:648), our findings suggest that SES patient outcomes were associated with the decreased proportions of TLR and mortality compared with BMS patients. On the other hand, the proportions of recurrent MI and definite ST were comparable between the 2 stents. These findings were derived from RCT that were relatively of moderate quality, homogeneous, and possibly free from publication bias. A meta-analysis of RCT and registries demonstrated reduced target vessel revascularization (TVR) in patients treated with DES (both SES and paclitaxel-eluting stents) compared with BMS patients within 2-year follow up [35]. The pooled risk difference of TLR at 4 years in our study was similar to that of TVR at 2 years in the RCT included in the previous study (−7.0%, 95% CI: − 9.2% to −4.7%). Furthermore, in addition to the small number needed to treat, the pooled OR for TLR at 4 years changed only slightly from 1 year in our study. Therefore, we may validate and extend the efficacy of SES observed during medium term in STEMI patients up to 4 years after primary PCI by the meta-analysis of strictly selected RCT. Our study also suggests the possibility of a reduction in mortality of SES patients compared to BMS patients. Some previous registries have reported no difference in mortality between SES and BMS [36,37]. In these registries, however, in addition to the relatively small number of included patients (369 and 716 patients, respectively), the cohort in which patients were implanted with BMS was different from that of SES. On the other hand, another registry, which included a larger number of patients (1926 patients) and was conducted in 1 cohort, showed a significant reduction in mortality in DES compared with BMS [38]. However, in addition to the fact that observational studies are likely to be influenced by several unobserved biases, the stable pooled OR with the narrowing 95% CI in our cumulative meta-analysis may support the reliable effects of SES on mortality. With regard to recurrent MI and definite ST, the pooled estimates worsened for SES versus BMS from 1 year to 4 years, respectively, even though the overall proportion of both events was comparable. These deteriorations did not influence the change of pooled estimate for mortality between 1 and 4 years. Follow-up angiography was mandatory in all patients in MISSION! at 9 months and in SESAMI at 1 year, and in a part of patients in TYPHOON at 8 months. The use of routine angiography may increase the proportion of angiography-related outcomes, such as TLR and definite ST [39]. However, the proportion of TLR at 1 year was comparable between the RCT with and without routine follow-up angiography (MISSION! versus SESAMI versus TYPHOON versus PASEO = 3.2% versus 4.5% versus 4.2% versus 3.3% in SES and 11.2% versus 11.8% versus
5
10.9% versus 14.4% in BMS). A similar tendency was observed with regard to definite ST as well (0.6% versus 1.3% versus 1.1% versus 0% in SES and 0.7% versus 0.7% versus 2.2% versus 1.1% in BMS). Furthermore, routine angiography at medium term follow up may also not influence the long-term proportions (TLR: MISSION! versus SESAMI versus TYPHOON versus PASEO = 6.3% versus 7.0% versus 7.6% versus 5.6% in SES and 12.5% versus 13.5% versus 14.8% versus 21.1% in BMS; definite ST: 2.5% versus 1.9% versus 3.6% versus 1.1% in SES and 0.7% versus 1.3% versus 4.0% versus 2.2% in BMS). Thus, the findings of our study were irrespective of mandated angiographic follow up. Although the presence of publication bias was not suggested for most of the end points by either Begg's rank correlation test or Egger's linear regression test and for TLR by either Egger's linear regression test or the Duval and Tweedie's trim and fill procedure, the number of RCT included in this meta-analysis was relatively small. Other than the excluded studies described above, long-term outcomes were not reported in 1 RCT registered in clinicaltrials.gov, even though the proportions of random allocation between SES and BMS versus between with and without abciximab were uncertain [40]. Furthermore, no long-term results were reported in another RCT in which medium term outcomes were presented previously [15]. Therefore, our findings may change if long-term results are reported from all these RCT. Several limitations are considered to exist in this study. First of all, the number of RCT and patients included in this study was relatively small in terms of meta-analysis. Although each RCT met the requirements for this study, the quality varied across individual RCT. The proportion of patients lost to follow-up was higher in TYPHOON, which included the largest number of patients and was the only multicenter RCT conducted in various countries. This fact may lead to an inaccurate pooled estimate by selection bias since a higher proportion of dropouts seems to cause an increase in number of unwitnessed events. In fact, OR for death decreased from 1 year to longer-term follow up only in TYPHOON. None of the RCT was double-blinded, which may lead to selection bias, response bias, etc. However, while trials with open assessment of the outcomes were related to overestimation of treatment effects [41], all clinical events were adjudicated by the reviewers who were blinded to the treatment assignment or who were members of an independent clinical events committee in each RCT. In addition, it may not be critical with regard to “hard” end points such as death. Detailed data regarding ischemia time or duration of DAPT were not available in all RCT. The patient population was not the same across the studies, which is the nature of metaanalysis. However, heterogeneity across RCT was not statistically proven regarding each outcome. Our findings may not be generalizable for patients in the real world due to the limited eligibilities of included patients in each RCT. A major potential limitation of DES implantation in patients with acute MI seems that the clinicians do not always have time to thoroughly rule out contraindications to long-term DAPT that is needed after DES implantation. This could potentially expose these patients to the risk of ST if they subsequently are unable to receive DAPT. The ability to detect patients unlikely to take DAPT is probably improved in RCT and may potentially be missed in real world patients undergoing emergent PCI. Finally, findings from pooled-analysis with study-level data may be more biased and misleading than those with patient-level data. For instance, conclusions from analyses with study-level data and with patient-level data differed from each other in previous reports [42,43]. In conclusion, from this meta-analysis of RCT featuring long-term follow up, primary PCI with SES for STEMI patients may be associated with a decrease in not only TLR but also mortality without an increase in recurrent MI or definite ST compared to those treated with BMS. Further longer-term results from ongoing RCT may be warranted to validate the effectiveness and safety issues of SES in STEMI patients.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054
6
R. Sakurai et al. / International Journal of Cardiology xxx (2012) xxx–xxx
Acknowledgments The authors thank Heidi N. Bonneau, RN, MS, CAA for her editorial review of the manuscript. The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [45]. References [1] Morice MC, Serruys PW, Sousa JE, et al. A randomized comparison of a sirolimuseluting stent with a standard stent for coronary revascularization. N Engl J Med 2002;346:1773–80. [2] Morice MC, Serruys PW, Barragan P, et al. Long-term clinical outcomes with sirolimus-eluting coronary stents: five-year results of the RAVEL trial. J Am Coll Cardiol 2007;50:1299–304. [3] Moses JW, Leon MB, Popma JJ, et al. Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery. N Engl J Med 2003;349:1315–23. [4] Weisz G, Leon MB, Holmes Jr DR, et al. Five-year follow-up after sirolimus-eluting stent implantation results of the SIRIUS (Sirolimus-Eluting Stent in De-Novo Native Coronary Lesions) trial. J Am Coll Cardiol 2009;53:1488–97. [5] Schofer J, Schluter M, Gershlick AH, et al. Sirolimus-eluting stents for treatment of patients with long atherosclerotic lesions in small coronary arteries: double-blind, randomised controlled trial (E-SIRIUS). Lancet 2003;362:1093–9. [6] Schampaert E, Cohen EA, Schluter M, et al. The Canadian study of the sirolimuseluting stent in the treatment of patients with long de novo lesions in small native coronary arteries (C-SIRIUS). J Am Coll Cardiol 2004;43:1110–5. [7] Suttorp MJ, Laarman GJ, Rahel BM, et al. Primary Stenting of Totally Occluded Native Coronary Arteries II (PRISON II): a randomized comparison of bare metal stent implantation with sirolimus-eluting stent implantation for the treatment of total coronary occlusions. Circulation 2006;114:921–8. [8] Rahel BM, Laarman GJ, Kelder JC, Ten Berg JM, Suttorp MJ. Three-year clinical outcome after primary stenting of totally occluded native coronary arteries: a randomized comparison of bare-metal stent implantation with sirolimus-eluting stent implantation for the treatment of total coronary occlusions (Primary Stenting of Totally Occluded Native Coronary Arteries [PRISON] II study). Am Heart J 2009;157: 149–55. [9] Baumgart D, Klauss V, Baer F, et al. One-year results of the SCORPIUS study: a German multicenter investigation on the effectiveness of sirolimus-eluting stents in diabetic patients. J Am Coll Cardiol 2007;50:1627–34. [10] Jimenez-Quevedo P, Sabate M, Angiolillo DJ, et al. Long-term clinical benefit of sirolimus-eluting stent implantation in diabetic patients with de novo coronary stenoses: long-term results of the DIABETES trial. Eur Heart J 2007;28:1946–52. [11] Kelbaek H, Thuesen L, Helqvist S, et al. The Stenting Coronary Arteries in Nonstress/Benestent Disease (SCANDSTENT) trial. J Am Coll Cardiol 2006;47:449–55. [12] Kelbaek H, Klovgaard L, Helqvist S, et al. Long-term outcome in patients treated with sirolimus-eluting stents in complex coronary artery lesions: 3-year results of the SCANDSTENT (Stenting Coronary Arteries in Non-Stress/Benestent Disease) trial. J Am Coll Cardiol 2008;51:2011–6. [13] Spaulding C, Henry P, Teiger E, et al. Sirolimus-eluting versus uncoated stents in acute myocardial infarction. N Engl J Med 2006;355:1093–104. [14] Menichelli M, Parma A, Pucci E, et al. Randomized trial of Sirolimus-Eluting Stent Versus Bare-Metal Stent in Acute Myocardial Infarction (SESAMI). J Am Coll Cardiol 2007;49:1924–30. [15] Diaz de la Llera LS, Ballesteros S, Nevado J, et al. Sirolimus-eluting stents compared with standard stents in the treatment of patients with primary angioplasty. Am Heart J 2007;154:164.e1–6. [16] van der Hoeven BL, Liem SS, Jukema JW, et al. Sirolimus-eluting stents versus bare-metal stents in patients with ST-segment elevation myocardial infarction: 9-month angiographic and intravascular ultrasound results and 12-month clinical outcome results from the MISSION! Intervention study. J Am Coll Cardiol 2008;51:618–26. [17] Valgimigli M, Campo G, Percoco G, et al. Comparison of angioplasty with infusion of tirofiban or abciximab and with implantation of sirolimus-eluting or uncoated stents for acute myocardial infarction: the MULTISTRATEGY randomized trial. JAMA 2008;299:1788–99. [18] Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation 2007;115:2344–51. [19] Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1–12. [20] Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance of continuity corrections in meta-analysis of sparse data. Stat Med 2004;23: 1351–75. [21] Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. Stat Med 2007;26:53–77.
[22] Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58. [23] Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088–101. [24] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [25] Sterne JA, Gavaghan D, Egger M. Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature. J Clin Epidemiol 2000;53:1119–29. [26] Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. Ann Intern Med 2009;151:W65–94. [27] Valgimigli M, Campo G, Arcozzi C, et al. Two-year clinical follow-up after sirolimus-eluting versus bare-metal stent implantation assisted by systematic glycoprotein IIb/IIIa Inhibitor Infusion in patients with myocardial infarction: results from the STRATEGY study. J Am Coll Cardiol 2007;50:138–45. [28] Kelbaek H, Thuesen L, Helqvist S, et al. Drug-eluting versus bare metal stents in patients with st-segment-elevation myocardial infarction: eight-month followup in the Drug Elution and Distal Protection in Acute Myocardial Infarction (DEDICATION) trial. Circulation 2008;118:1155–62. [29] Yoshikawa D, Isobe S, Umeda H, et al. Three-year prognosis of Japanese patients with ST-elevation myocardial infarction treated with sirolimus-eluting stents. Coron Artery Dis 2009;20:422–7. [30] Valgimigli M. 3-year follow-up of the multicentre evaluation of single high-dose bolus tirofiban versus abciximab with sirolimus eluting stent or bare metal stent in acute myocardial infarction study. Paper presented at: European Society of Cardiology Congress 2010. Sweden: Stockholm; August 29, 2010. [31] Spaulding C, Teiger E, Commeau P, et al. Four-year follow-up of TYPHOON (trial to assess the use of the CYPHer sirolimus-eluting coronary stent in acute myocardial infarction treated with BallOON angioplasty). JACC Cardiovasc Interv 2011;4: 14–23. [32] Di Lorenzo E, Sauro R, Varricchio A, et al. Benefits of drug-eluting stents as compared to bare metal stent in ST-segment elevation myocardial infarction: four year results of the PaclitAxel or Sirolimus-Eluting stent vs bare metal stent in primary angiOplasty (PASEO) randomized trial. Am Heart J 2009;158:e43–50. [33] Atary JZ, van der Hoeven BL, Liem SS, et al. Three-year outcome of sirolimus-eluting versus bare-metal stents for the treatment of ST-segment elevation myocardial infarction (from the MISSION! Intervention study). Am J Cardiol 2010;106:4–12. [34] Violini R, Musto C, De Felice F, et al. Maintenance of long-term clinical benefit with sirolimus-eluting stents in patients with ST-segment elevation myocardial infarction 3-year results of the SESAMI (sirolimus-eluting stent versus baremetal stent in acute myocardial infarction) trial. J Am Coll Cardiol 2010;55:810–4. [35] Brar SS, Leon MB, Stone GW, et al. Use of drug-eluting stents in acute myocardial infarction: a systematic review and meta-analysis. J Am Coll Cardiol 2009;53: 1677–89. [36] Daemen J, Tanimoto S, Garcia-Garcia HM, et al. Comparison of three-year clinical outcome of sirolimus- and paclitaxel-eluting stents versus bare metal stents in patients with ST-segment elevation myocardial infarction (from the RESEARCH and T-SEARCH Registries). Am J Cardiol 2007;99:1027–32. [37] Kukreja N, Onuma Y, Garcia-Garcia H, Daemen J, van Domburg R, Serruys PW. Primary percutaneous coronary intervention for acute myocardial infarction: long-term outcome after bare metal and drug-eluting stent implantation. Circ Cardiovasc Interv 2008;1:103–10. [38] Hannan EL, Racz M, Walford G, et al. Drug-eluting versus bare-metal stents in the treatment of patients with ST-segment elevation myocardial infarction. JACC Cardiovasc Interv 2008;1:129–35. [39] Pinto DS, Stone GW, Ellis SG, et al. Impact of routine angiographic follow-up on the clinical benefits of paclitaxel-eluting stents: results from the TAXUS-IV trial. J Am Coll Cardiol 2006;48:32–6. [40] Michels R, Wijnbergen I. A comparison of drug eluting stents and bare metal stents with or without abciximab in ST-segment elevation myocardial Infarction, The Eindhoven Reperfusion study. Paper presented at: TCT09; September 25, 2009; San Francisco, CA; 2009. [41] Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA 1999;282:1054–60. [42] Nordmann AJ, Briel M, Bucher HC. Mortality in randomized controlled trials comparing drug-eluting vs. bare metal stents in coronary artery disease: a meta-analysis. Eur Heart J 2006;27:2784–814. [43] Holmes Jr DR, Moses JW, Schofer J, Morice MC, Schampaert E, Leon MB. Cause of death with bare metal and sirolimus-eluting stents. Eur Heart J 2006;27:2815–22. [44] Di Lorenzo E, De Luca G, Sauro R, et al. The PASEO (PaclitAxel or Sirolimus-Eluting Stent Versus Bare Metal Stent in Primary Angioplasty) randomized trial. JACC Cardiovasc Interv 2009;2:515–23. [45] Shewan LG, Coats AJ. Ethics in the authorship and publishing of scientific articles. Int J Cardiol 2010;144:1–2.
Please cite this article as: Sakurai R, et al, Sirolimus-eluting stents reduce long-term mortality compared with bare metal stents in ST-segment elevation myocardial infarction: A meta-analysis of randomized controlled trials, Int J Cardiol (2012), doi:10.1016/j.ijcard.2011.12.054