|
|
Background Several studies have compared the treatment effects of coronary stenting and coronary-artery bypass grafting (CABG). However, there are limited data regarding the long-term outcomes of these two interventions for patients with unprotected left main coronary artery disease.
Methods We evaluated 1102 patients with unprotected left main coronary artery disease who underwent stent implantation and 1138 patients who underwent CABG in Korea between January 2000 and June 2006. We compared adverse outcomes (death; a composite outcome of death, Q-wave myocardial infarction, or stroke; and target-vessel revascularization) with the use of propensity-score matching in the overall cohort and in separate subgroups according to type of stent.
Results In the overall matched cohort, there was no significant difference between the stenting and CABG groups in the risk of death (hazard ratio for the stenting group, 1.18; 95% confidence interval [CI], 0.77 to 1.80) or the risk of the composite outcome (hazard ratio for the stenting group, 1.10; 95% CI, 0.75 to 1.62). The rates of target-vessel revascularization were significantly higher in the group that received stents than in the group that underwent CABG (hazard ratio, 4.76; 95% CI, 2.80 to 8.11). Comparisons of the group that received bare-metal stents with the group that underwent CABG and of the group that received drug-eluting stents with the group that underwent CABG produced similar results, although there was a trend toward higher rates of death and the composite end point in the group that received drug-eluting stents.
Conclusions In a cohort of patients with unprotected left main coronary artery disease, we found no significant difference in rates of death or of the composite end point of death, Q-wave myocardial infarction, or stroke between patients receiving stents and those undergoing CABG. However, stenting, even with drug-eluting stents, was associated with higher rates of target-vessel revascularization than was CABG.
The development of coronary-artery stenting has led to a reevaluation of the role of PCI for left main coronary artery disease, and several studies have reported on the midterm safety and feasibility of stenting.7,8,9 Interest in stenting of the left main coronary artery has increased further with the availability of drug-eluting stents that significantly reduce the rates of restenosis and repeat revascularization.10,11,12 However, there are limited data on the long-term safety and effectiveness of PCI with bare-metal or drug-eluting stents, as compared with CABG, in patients with unprotected left main coronary artery disease. We therefore compared the long-term outcomes of the implantation of coronary stents and CABG among patients with unprotected left main coronary artery disease in Korea, where stenting of the left main coronary artery has been a more common practice than in Western countries, as recorded in the MAIN-COMPARE (Revascularization for Unprotected Left Main Coronary Artery Stenosis: Comparison of Percutaneous Coronary Angioplasty versus Surgical Revascularization) registry.
Methods
Study Population
The MAIN-COMPARE registry holds data on consecutive patients from 12 major cardiac centers in Korea that performed PCI or CABG for unprotected left main coronary artery disease (defined as stenosis of more than 50%) between January 2000 and June 2006. We excluded patients who had undergone previous CABG, those who underwent concomitant valvular or aortic surgery, and those who had myocardial infarction with ST-segment elevation or presented with cardiogenic shock. The registry is sponsored by the Korean Society of Interventional Cardiology. There was no industry involvement in the design, conduct, or analysis of the study. The local ethics committee at each hospital approved the use of clinical data for this study, and all patients provided written informed consent.
Patients underwent PCI, instead of CABG, because of either the patient's or physician's preference or the high risk associated with CABG (see the Supplementary Appendix, available with the full text of this article at www.nejm.org). From January 2000 through May 2003, coronary stenting was performed exclusively with bare-metal stents, whereas from May 2003 through June 2006, exclusively drug-eluting stents were used. Methods of stent implantation for patients with left main coronary artery disease have been described previously10,13,14 (see the Supplementary Appendix). Interventions for other clinically important types of coronary artery disease were performed according to current practice guidelines.6 All patients undergoing PCI were prescribed aspirin plus clopidogrel (loading dose, 300 mg or 600 mg) or ticlopidine (loading dose, 500 mg) before or during the coronary intervention. After the procedure, aspirin was continued indefinitely. Patients treated with bare-metal stents were prescribed clopidogrel or ticlopidine for at least 1 month, and patients treated with drug-eluting stents were prescribed clopidogrel for at least 6 months. Surgical revascularization was performed with the use of standard bypass techniques.5 Whenever possible, the internal thoracic artery was used preferentially for revascularization of the left anterior descending artery.
Follow-up and End Points
Clinical, angiographic, procedural or operative, and outcome data were collected with the use of a dedicated Internet-based reporting system. All outcomes of interest were confirmed by source documentation collected at each hospital and were centrally adjudicated by the local events committee at the University of Ulsan College of Medicine, Asan Medical Center, Seoul. For validation of complete follow-up data, information about vital status was obtained through July 15, 2007, from the National Population Registry of the Korea National Statistical Office with the use of a unique personal identification number.
Clinical follow-up after PCI and after CABG was recommended at 1 month, 6 months, and 1 year and then annually thereafter. Routine angiographic follow-up for all patients treated with PCI was recommended 6 to 10 months after the procedure. However, patients who were at high risk for procedural complications of angiography and had no symptoms or signs of ischemia, as well as patients who declined the recommendation, did not undergo routine follow-up angiography. For patients who underwent CABG, angiographic follow-up was recommended only if there were ischemic symptoms or signs during follow-up.
The end points of the study were death; the composite of death, Q-wave myocardial infarction, or stroke; and target-vessel revascularization. All events were based on clinical diagnoses assigned by the patient's physician and were centrally adjudicated by an independent group of clinicians. Death was defined as death from any cause. Q-wave myocardial infarction was defined as documentation of a new abnormal Q wave after the index treatment. Stroke, as indicated by neurologic deficits, was confirmed by a neurologist on the basis of imaging studies. Target-vessel revascularization was defined as repeat revascularization of the treated vessel, including any segments of the left anterior descending artery and the left circumflex artery.15
Statistical Analysis
Among patients who had unprotected left main coronary artery disease, we compared long-term outcomes for those who underwent PCI, irrespective of stent type, with the outcomes for those who underwent CABG. In addition, we compared the long-term outcomes of patients who received bare-metal or drug-eluting stents with patients who underwent CABG. Patients who received bare-metal stents were compared with patients who underwent CABG between January 2000 and May 2003 (Wave 1 of the registry). Patients who received drug-eluting stents were compared with patients who underwent CABG between May 2003 and June 2006 (Wave 2 of the registry).
To reduce the effect of treatment-selection bias and potential confounding in this observational study, we performed rigorous adjustment for significant differences in the baseline characteristics of patients with the use of propensity-score matching.16,17 For each comparison (the entire cohort, Wave 1, and Wave 2), a separate propensity score for PCI versus CABG was derived. The details of the propensity-score method, with the resulting models and their predictive characteristics, are described in the Supplementary Appendix.
After all the propensity-score matches were performed, we compared the baseline covariates between the two intervention groups. Continuous variables were compared with the use of the paired t-test or the Wilcoxon signed-rank test, as appropriate, and categorical variables were compared with the use of McNemar's test. Statistical significance and the effect of treatment on outcomes were estimated with the use of appropriate statistical methods for matched data.18,19 In the propensity-score–matched cohort, the risks of each outcome were compared with the use of Cox regression models, with robust standard errors that accounted for the clustering of matched pairs. Survival curves were constructed with Kaplan–Meier estimates and compared with the use of methods described by Klein and Moeschberger.19
All reported P values are two-sided, and P values of less than 0.05 were considered to indicate statistical significance. SAS software, version 9.1, and the R programming language were used for statistical analyses.
Results
Characteristics of the Study Population
Between January 2000 and June 2006, a total of 2240 patients with unprotected left main coronary artery disease met the criteria for inclusion. Of these, 1102 were treated with PCI and 1138 with CABG. Of the patients who underwent PCI, 1073 (97.4%) had clinical and angiographic conditions that made them eligible for either PCI or CABG, but they underwent PCI because of the patient's or physician's preference. The remaining 29 patients (2.6%) had underlying conditions that made them poor candidates for surgery (8 were 
80 years of age and had a poor performance status, 3 had limited life expectancy, 2 had a current malignant condition, 12 had a concomitant severe medical illness, and 4 had no suitable bypass conduits).
In the PCI group, 318 patients (28.9%) received bare-metal stents and 784 (71.1%) received drug-eluting stents. Of the patients treated with drug-eluting stents, 607 (77.4%) received sirolimus-eluting stents and 177 (22.6%) received paclitaxel-eluting stents. The mean (±SD) number of stents implanted in a patient's left main coronary lesions was 1.2±0.5, the mean total length of the stents was 28.0±20.7 mm, and the mean stent diameter was 3.5±0.4 mm. The mean total number of stents implanted in a patient (including left main and other vessels) was 1.9±1.1.
In the CABG group, 478 patients (42.0%) underwent off-pump surgery and 1120 (98.4%) received at least one arterial conduit that, in 1096 patients (97.9%), was used in revascularization of the left anterior descending artery. The mean number of grafts used was 2.9±1.0 (2.2±0.9 arterial grafts and 0.7±0.8 venous grafts).
The baseline characteristics of the study patients according to the revascularization procedure are shown in Table 1. Patients undergoing CABG were significantly older and had a higher prevalence of diabetes, hyperlipidemia, smoking, and history of myocardial infarction or peripheral vascular disease than those receiving stents. Patients undergoing CABG were also more likely to have lower ejection fractions, to present with unstable angina, and to have distal-bifurcation stenosis, three-vessel disease, and involvement of the right coronary artery. More patients in the PCI group than in the CABG group had undergone previous coronary angioplasty and had restenotic left main coronary lesions.
|
The median follow-up was 1017 days (interquartile range, 688 to 1451) in the PCI group and 1152 days (interquartile range, 681 to 1590) in the CABG group. Complete follow-up data for major clinical events were obtained in 98.9% of the overall cohort. During follow-up, 187 patients (8.3%) died, of whom 127 (67.9%, or 5.7% of the overall cohort) died of a cardiovascular cause. Twenty patients (0.9%) had a Q-wave myocardial infarction, and 35 (1.6%) had a stroke. Target-vessel revascularization was performed in 141 patients (6.3%).
Characteristics of Patients Matched for Propensity Scores
Details of the propensity-score analysis are provided in the Supplementary Appendix. After propensity-score matching was performed for the entire population, there were 542 matched pairs of patients (Table 2). In this matched cohort, 25.5% of the PCI group received bare-metal stents and 74.5% received drug-eluting stents. After propensity-score matching was performed separately in the Wave 1 and Wave 2 cohorts, there were 207 matched pairs of patients who received bare-metal stents and concurrent control patients who underwent CABG, and 396 matched pairs of patients who received drug-eluting stents and concurrent control patients who underwent CABG. In the matched cohorts, there was no longer any significant difference between the PCI group and the CABG group for any covariate, according to the use of statistical methods appropriate for matched data (Table 2).
|
Figure 1 and Table 3 show the long-term rates of clinical outcomes according to the treatment approach in the overall matched cohort. For the 542 matched pairs, there was no significant difference between the PCI and CABG groups in the risk of death during the 3-year follow-up period. The rates of major adverse cardiovascular events (death, Q-wave myocardial infarction, or stroke) were also similar. However, the rate of target-vessel revascularization was significantly higher in the PCI group than in the CABG group. Of the patients with initial PCI treatment who underwent target-vessel revascularization during the study period, 82.1% underwent repeat PCI (stenting or balloon angioplasty) and 17.9% underwent CABG. Of the patients with initial CABG treatment who underwent target-vessel revascularization, all underwent PCI, not repeat CABG. The risk of target-lesion revascularization specifically for restenosis of the left main coronary artery was also significantly higher in the PCI group than in the CABG group (hazard ratio, 2.72; 95% confidence interval [CI], 1.51 to 4.91).
|
|
|
|
Event-free survival in the unmatched cohort was consistent with the lower-risk characteristics of the patients who underwent PCI and the higher-risk characteristics of the patients who underwent CABG (see the Supplementary Appendix). Among the patients not included in the overall match, the 3-year rates of death and the composite of death, Q-wave myocardial infarction, or stroke were significantly higher in the CABG group, whereas the rate of target-vessel revascularization was higher in the PCI group. Among unmatched patients in Wave 1, the risks of death and the composite of death, Q-wave myocardial infarction, or stroke did not differ significantly between the two groups. However, the risk of target-vessel revascularization was consistently higher among patients who received bare-metal stents. Among unmatched patients in Wave 2, CABG was associated with higher risks of death and the composite of death, Q-wave myocardial infarction, or stroke than was PCI with drug-eluting stents, whereas the risk of revascularization was higher among patients who received stents.
Acute Complications
Acute complications occurred in 2.7% of patients undergoing PCI. The details of specific acute complications are provided in the Supplementary Appendix.
Discussion
We compared the long-term outcomes of coronary stenting and CABG in a large cohort of patients with unprotected left main coronary artery disease. Our observational study showed that the risks of death and a composite of serious outcomes (death, Q-wave myocardial infarction, or stroke) were similar in the PCI and the CABG groups. These results were consistent when bare-metal stents or drug-eluting stents were compared with CABG, although there was a nonsignificant trend toward higher risk with drug-eluting stents. In contrast, the rate of target-vessel revascularization was significantly lower in the CABG group than in the PCI group, although hazard ratios varied depending on the type of stent.
Another large observational study, published before the development of drug-eluting stents, suggested that patients with left main coronary artery disease did significantly better with CABG than with PCI.20 Although this was a risk-adjusted analysis, patient-selection factors probably contributed to the results. More recent observational studies have shown similar mortality rates and similar risks of major adverse cardiac and cerebrovascular events between patients receiving drug-eluting stents and those undergoing CABG, although patients undergoing CABG have typically had significantly lower rates of repeat revascularization.15,21,22,23 These studies have all been limited by small numbers of patients, a limited duration of follow-up, and selective performance of PCI in patients considered to be poor candidates for CABG.
In our study, despite the lower rates of repeat revascularization with drug-eluting stents than with bare-metal stents, CABG was still more effective than drug-eluting stents in reducing the need for target-vessel revascularization. However, there was a significantly higher rate of follow-up angiography in the PCI group than in the CABG group (73.0% vs. 14.6%, P<0.001). Therefore, the rate of asymptomatic graft stenosis or occlusion may have been underestimated in the CABG group relative to the PCI group.
The major limitation of this study is that we evaluated observational data, and therefore the treatment strategy was not based on randomized assignment. The choice of revascularization was at the discretion of the treating physician or the patient. In the Supplementary Appendix we have attempted to provide as clear a description as possible of the factors that were likely to have influenced the selection of a procedure for individual patients. We acknowledge, however, that the particulars of clinical practice in the institutions in this trial, as well as the specific expertise of the interventional cardiologists and cardiac surgeons who performed the procedures, may differ from those of other institutions and practitioners, potentially limiting the reproducibility of these results in other settings.
From an analytical standpoint, our findings are subject to selection bias and confounding with respect to the relative severity of preprocedural risks among patients who underwent PCI and those who underwent CABG. To minimize these biases, we used propensity-score matching.16,17 Previous research has suggested that matching according to the propensity score eliminates a greater proportion of baseline differences between two treatments than does stratification or covariate adjustment.24 Nevertheless, hidden bias may remain because of the influence of unmeasured confounders. Given these issues and the findings of our study, we believe that a randomized trial of PCI with drug-eluting stents as compared with CABG is warranted in patients with unprotected left main coronary artery disease who are candidates for revascularization.
A final caveat is that our analysis was underpowered to detect significant differences in mortality, especially in the comparison of drug-eluting stents with CABG. More than 5500 patients would have been needed for such an analysis. Nonsignificant trends toward higher event rates were seen in the group that received drug-eluting stents; these trends might have been significant with a larger cohort of patients.
In conclusion, we found that in matched cohorts of patients with unprotected left main coronary artery disease, PCI with stenting and CABG were associated with similar long-term rates of death and the composite end point of death, Q-wave myocardial infarction, or stroke. Rates of target-vessel revascularization were higher among patients who underwent PCI than among those who underwent CABG.
Supported by the Korean Society of Interventional Cardiology and the Cardiovascular Research Foundation, Seoul.
Drs. Seung, Hong, and Chung report receiving lecture fees from Cordis and Boston Scientific; Dr. Y.-H. Kim, receiving lecture fees from Cordis; Dr. C.W. Lee, receiving lecture fees from Medtronic; Dr. Jeong, receiving lecture fees from Cordis, Medtronic, and Boston Scientific; Dr. Tahk, receiving lecture fees from Boston Scientific; Dr. S.-W. Park, receiving research grant support from Medtronic; Dr. Seong, receiving grant support from Boston Scientific; Dr. Gwon, receiving lecture fees from Boston Scientific and Medtronic and research grant support from Medtronic; Dr. H.-S. Kim, receiving consulting fees from Abbott Vascular; and Dr. S.-J. Park. receiving consulting fees from Cordis, lecture fees from Cordis, Medtronic, and Boston Scientific, and research grant support from Cordis and Medtronic. No other potential conflict of interest relevant to this article was reported.
We thank the staff members of the MAIN-COMPARE registry, the other members of the cardiac catheterization laboratories and the cardiac surgery departments of the participating centers, and the study coordinators for their efforts to collect the clinical data that made this study possible and to ensure the accuracy and completeness of the registry data.
Source Information
From the Catholic University of Korea, St. Mary's Hospital, Gangnam (K.B.S., P.J.K.) and Yeoido (W.-S.C.), Seoul; the Department of Cardiology (D.-W.P., Y.-H.K., S.-W.L., C.W.L., M.-K.H., S.-W.P., S.-J.P.) and Division of Biostatistics (S.-C.Y.), Center for Medical Research and Information, University of Ulsan College of Medicine, Asan Medical Center, Seoul; Samsung Medical Center, Seoul (H.-C.G.); Chonnam National University Hospital, Gwangju (M.-H.J.); Yonsei University Severance Hospital, Seoul (Y.J.); Seoul National University Hospital, Seoul (H.-S.K.) and Bundang (I.-H.C.); Chungnam National University Hospital, Daejeon (I.-W.S.); Kyung Pook National University Hospital, Daegu (H.S.P.); Gachon University Gil Medical Center, Incheon (T.A.); and Ajou University Medical Center, Suwon (S.-J.T.) — all in Korea.
Drs. K.B. Seung and D.-W. Park contributed equally to this article.
This article (10.1056/NEJMoa0801441) was published at www.nejm.org on March 31, 2008.
Address reprint requests to Dr. Seung-Jung Park at the Department of Cardiology, University of Ulsan College of Medicine, Cardiac Center, Asan Medical Center, 388-1 Poongnap-dong, Songpa-gu, Seoul, 138-736, Korea, or at sjpark{at}amc.seoul.kr.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
