|Home | About | Journals | Submit | Contact Us | Français|
We compared the efficacy and safety of drug-eluting stents with that of bare-metal stents in patients who experienced acute ST-segment–elevation myocardial infarction (STEMI) and underwent primary percutaneous coronary intervention. To do this, we performed a meta-analysis of 13 randomized controlled trials in which drug-eluting stents were compared with bare-metal stents in STEMI patients. The trials involved 6,769 patients (4,246 received drug-eluting stents and 2,523 received bare-metal stents) and follow-up periods of 6 to 48 months. In comparison with bare-metal stents, drug-eluting stents significantly reduced the incidence of major adverse cardiac events, with a risk ratio (RR) of 0.59 (95% confidence interval [CI], 0.47–0.73; P < 0.00001). Drug-eluting stents were not associated with a significant reduction in overall death (RR = 0.94; 95% CI, 0.74–1.20; P = 0.64), but were associated with significant reductions in recurrent myocardial infarction (RR = 0.76; 95% CI, 0.58–0.98; P = 0.03), target-vessel revascularization (RR = 0.47; 95% CI, 0.39–0.56; P <0.00001), and in-stent restenosis (RR = 0.32; 95% CI, 0.25–0.39; P < 0.00001). Moreover, no significant difference was found in the comparative risk of stent thrombosis (RR = 0.85; 95% CI, 0.63–1.14; P = 0.27).
On the basis of risk ratio, we conclude that using drug-eluting stents in STEMI patients who undergo primary percutaneous coronary intervention is safe with regard to stent thrombosis within 48 months, and that drug-eluting stents improve clinical outcomes by reducing the risks of major adverse cardiac events, recurrent myocardial infarction, reintervention, and in-stent restenosis, compared with bare-metal stents. However, in order to investigate possible very late stent thrombosis, follow-up of these trials beyond 48 months is warranted.
Primary percutaneous coronary intervention (PCI) is the preferred treatment for acute ST-segment–elevation myocardial infarction (STEMI), and this method is performed in increasing numbers of patients.1,2 Several randomized and observational trials have shown that, in c-omparison with bare-metal stents (BMSs), drug-eluting stents (DESs) can be safely used in cases of acute myocardial infarction (AMI) and are associated with significantly lower rates of in-stent restenosis observed on angiography and fewer repeat revascularization procedures.3–19 However, several clinical studies of AMI patients have indicated that DESs may be associated with no significantly lower risk of major adverse cardiac events (MACE) during short- or long-term follow-up, or even with an increased risk.20–26 Moreover, until recently, evidence in regard to the use of DESs in STEMI patients who underwent primary PCI has been sparse, and DESs have not yet been recommended for use in primary PCI by the National Institute for Health and Clinical Excellence.27 We conducted a meta-analysis by pooling available randomized controlled trials in order to quantify the efficacy and safety of DESs versus that of BMSs in STEMI patients who underwent primary PCI. Because clinical questions are often not resolved by individual studies, meta-analysis is commonly used as an analytical tool, due to its advantages of increased statistical power, large sample size, wide population coverage, and low cost.
Studies were included in our meta-analysis if they met all 3 of the following criteria: they compared DESs with BMSs in STEMI patients who underwent primary PCI, they were randomized controlled trials, and their results were reported or made available by the trial investigators for a mean follow-up period of at least 6 months. Nonclinical trials and studies that did not meet these criteria were excluded from analysis. When 2 publications reported the same study, only the publication with more complete information was included in the analysis.
A systematic MEDLINE® and PubMed® database search (up to 23 May 2010) was performed with the following key words: ([drug-eluting stents] OR [sirolimus] OR [paclitaxel]) AND (myocardial infarction) AND (randomized). Furthermore, the websites of the American College of Cardiology (www.cardiosource.org/acc), the American Heart Association (www.americanheart.org), the European Society of Cardiology (www.escardio.org), Transcatheter Cardiovascular Therapeutics (www.tctmd.com), and EuroPCR (www.europcr.com) were searched for pertinent abstracts and expert presentations. No language restriction was applied. We also checked the citations of existing reviews and of all studies that were identified by the above methods. For additional information, we contacted the authors of the identified literature, as well as other relevant specialists.
Two of our authors (PPH and YGC) searched for and reviewed articles independently of each other, and those that met the inclusion criteria were selected for further analysis.
The initially selected studies were evaluated for quality in accordance with a well-established, validated scale developed by Jadad and colleagues.28 A Jadad score was calculated using 7 elements of consideration. The first 5 indicated good quality, and each counted as 1 point toward an overall quality score. The final 2 indicated poor quality, and 1 point was subtracted from the quality score for each one.
The possible range of scores was 0 (weakest) to 5 (strongest). Any study with a Jadad score below 3 was considered to be of poor quality and was excluded from consideration in our meta-analysis.
Two authors (PPH and YGC) independently evaluated trial quality. Using a standardized protocol and reporting form, they extracted data on the subjects' characteristics at baseline and data on the clinical outcomes. Absolute numbers were recalculated when percentages were reported. Any disagreement was resolved through discussion.
RevMan 5.0.23 software (The Cochrane Collaboration) was used for meta-analysis. The heterogeneity between selected articles was tested with use of the χ2 and I2 tests. Statistical significance was set at 2-tailed P <0.05. When there was no statistically significant difference among the results, the fixed-effects model was applied; when there was a significant difference, the random-effects model was applied. Finally, we evaluated publication bias by using funnel plots and the fail-safe number (Nfs), with the significance set at 0.05 for the MACE and the stent-thrombosis comparisons. Any calculated Nfs value smaller than the number of observed studies indicated publication bias that might influence the meta-analysis results. We calculated the Nfs0.05 according to the formula Nfs0.05 = (ΣZ/1.64)2 − k, with k defined as the number of studies that were included in the meta-analysis.
The initial electronic and manual search yielded 476 studies. At an interrater agreement of κ = 0.92, the authors identified 13 trials that met the inclusion criteria—11 full articles and 2 meeting presentations, all in English. These trials included an aggregate of 6,769 patients (Fig. 1), 4,246 of whom had received DESs and 2,523 of whom had received BMSs.5–16,20 Table I shows the chief characteristics of the trials. All 13 studies were of acceptable quality (Jadad score, ≥3). The mean ages of the participants in individual trials ranged from 58.5 to 64.5 years, and mean follow-up periods ranged from 6 to 48 months. No data on type of DES were found in the DEDICATION trial.14
During analysis of the MACE rate, heterogeneity was found among the 13 trials (χ2 = 26.07, P = 0.01, I2 = 54%). Therefore, the random-effects model was selected to analyze the difference in this rate between the 2 stent groups. However, there was no significant evidence of heterogeneity in relation to the other 5 endpoints:
Accordingly, for these endpoints, the fixed-effects model was applied to analyze the differences in the respective outcomes of DES and BMS treatment.
In comparison with BMSs, DESs significantly reduced the incidence of MACEs (9.3% vs 15.54%), with a risk ratio (RR) of 0.59 (95% confidence interval [CI], 0.47–0.73; P <0.00001) (Fig. 2). Table II shows the subgroup analyses with regard to the endpoint of MACE. The effect of sirolimus on the risk of MACE (RR = 0.5; 95% CI, 0.4–0.62) was stronger than that of paclitaxel (RR = 0.67; 95% CI, 0.45–1.00). When the duration of thienopyridine therapy (at 3, 6, 9, and 12 mo) was analyzed, the strongest effect of DESs on MACEs was at 9 months (RR = 0.22; 95% CI, 0.12–0.42; P <0.00001). The mean follow-up period (<1, ≥1, and ≥2 yr) was also analyzed, and the effect of DESs was strongest at less than 1 year (RR = 0.34; 95% CI, 0.16–0.75; P = 0.007).
Although DESs were not associated with a significant reduction in overall death in comparison with BMSs (3.98% vs 4.32%; RR = 0.94; 95% CI, 0.74–1.20; P = 0.64) (Fig. 3), DESs were associated with significant reductions in recurrent MI (3.38% vs 4.01%; RR = 0.76; 95% CI, 0.58–0.98; P = 0.03) (Fig. 4), TVR (5.78% vs 13.14%; RR = 0.47; 95% CI, 0.39–0.56; P <0.00001) (Fig. 5), and in-stent restenosis (7.28% vs 22.01%; RR = 0.32; 95% CI, 0.25–0.39; P <0.00001) (Fig. 6).
In regard to the risk of stent thrombosis, the use of DESs in STEMI patients was safer than the use of BMSs, but the benefit was not significant (2.71% vs 3.09%; RR = 0.85; 95% CI, 0.63–1.14; P = 0.27) (Fig. 7). However, in all 13 trials, data on possible very late stent thrombosis (>2 yr) were deficient.
Figure 8A shows the funnel plot of the publication bias of the 13 trials. The graphic analysis of the funnel plot showed a degree of asymmetry that was possibly consistent with small-study bias. We calculated the comparative Nfs0.05 value for MACE as 227 and the value for stent thrombosis as 2. After adjustment for DES type (Fig. 8B) or the duration of thienopyridine therapy (Fig. 8C), the asymmetry decreased.
We found that treatment with DESs significantly reduced the incidence of recurrent MI, TVR, and in-stent restenosis in comparison with BMSs. On the other hand, we found no significant difference in the risk of all-cause death or stent thrombosis, which is similar to the results of previous meta-analyses.29–31 We examined more trials and found that DESs were significantly associated with a decreased MI recurrence rate and a decreased in-stent restenosis rate. Our study, which included more trials than the others, more widely supported the use of DESs.
In STEMI, primary PCI has been accepted as the default reperfusion approach if it can be performed rapidly in an experienced center.32 However, 1 study showed that patients whose infarct-related arteries reoccluded after primary PCI experienced higher 1-year mortality rates.33 A major concern surrounding the deployment of DESs in STEMI patients who undergo primary PCI is an increased risk of stent thrombosis, both acute and subacute.34,35 Platelet activation increases in AMI, and primary PCI is associated with more intense platelet activation.36 Greater platelet activation increases the risk of DES thrombosis in patients who have experienced AMI. Furthermore, due to impaired endothelial function, angiographic stent thrombosis may be higher in DESs than in BMSs.37 Injury secondary to drug toxicity or inflammation caused by polymers has been observed after DES implantation. Granulomatous and hypersensitivity reactions have been observed after the deployment of Cypher® stents (Cordis Corporation, a Johnson & Johnson company; Miami Lakes, Fla). Taxus® stents (Boston Scientific Corporation; Natick, Mass) are associated with medial necrosis, positive remodeling, and excessive fibrin deposition.38 Other factors that increase risk are bifurcation stenting, malapposition, and penetration of the necrotic core. Our results, however, showed no increased risk of stent thrombosis with DESs in comparison with BMSs in STEMI patients. These findings support the safety of using DESs. However, the difference was not significant in the results, and our results should be interpreted with caution. Further studies on possible very late stent thrombosis (>2 yr) should be performed.
The results of this meta-analysis show that the use of DESs in patients who undergo primary PCI for STEMI is safe within 48 months with regard to stent thrombosis, and that the use of DESs improves clinical outcomes by reducing the risks of MACE, recurrent MI, TVR, and in-stent restenosis in comparison with BMSs. Follow-up of these 13 trials beyond 48 months is warranted.
Address for reprints: Yu-Guo Chen, MD, Key Laboratory of Cardiovascular Remodeling & Function Research, Chinese Ministry of Education & Chinese Ministry of Public Health, Qilu Hospital, Shandong University, No. 107, Wen Hua Xi Rd., Jinan 250012, PRC
This research was supported by the Major State Basic Research Development Program of China (grant no. 2010CB732605)