Correspondence
Drug-Eluting Stents vs. Coronary-Artery Bypass Grafting
N Engl J Med 2008; 358:2641-2644June 12, 2008
- Article
To the Editor:
Hannan et al. (Jan. 24 issue)1 are to be commended for appropriately selecting patients with multivessel coronary disease for stenting or bypass surgery. Their clinical judgment ensured that unadjusted survival rates were equal in the two groups, even among patients with diabetes.
The authors attempt to equalize the two groups by evaluating risk-adjusted survival. However, they endeavor to adjust an unadjustable characteristic — the judgment of the treating physician, which is uncorrectable by adjusting for simple clinical variables. Although propensity analyses are valuable in assessing differences between groups, propensity-score matching might have been more enlightening.2
Meta-analyses of randomized, controlled trials showed no difference in survival between percutaneous and surgical intervention.3 Further meta-analyses have shown favorable survival among patients treated with drug-eluting stents as compared with those treated with bare-metal stents.4
Given these contemporary findings, the overall conclusion reached by the authors that “CABG [coronary-artery bypass grafting] continues to be associated with lower mortality rates than does treatment with drug-eluting stents” is misleading. The results of ongoing randomized, controlled trials are eagerly awaited to appropriately evaluate the relative merits of both strategies.
4 ReferencesJoost Daemen, M.D.
Neville Kukreja, M.A., M.R.C.P.
Patrick W. Serruys, M.D., Ph.D.
Erasmus Medical Center, Rotterdam 3015, the Netherlands1
Hannan EL ,Wu C ,Walford G , et al. Drug-eluting stents vs. coronary-artery bypass grafting in multivessel coronary disease. N Engl J Med 2008;358:331-341
Full Text | Web of Science | Medline2
D'Agostino RB Jr . Propensity scores in cardiovascular research. Circulation 2007;115:2340-2343
CrossRef | Web of Science | Medline3
Bravata DM ,Gienger AL ,McDonald KM , et al. The comparative effectiveness of percutaneous coronary interventions and coronary artery bypass graft surgery. Ann Intern Med 2007;147:703-716
Web of Science | Medline4
Stettler C ,Wandel S ,Allemann S , et al. Outcomes associated with drug-eluting and bare-metal stents: a collaborative network meta-analysis. Lancet 2007;370:937-948
CrossRef | Web of Science | Medline
To the Editor:
The conclusions set forth by Hannan et al. are potentially misleading. Survival was virtually identical in the surgical and interventional cohorts (93.7% and 93.4%, respectively). The “adjusted” survival suggested a 1.3% difference. The propensity-score analysis, however, should be viewed as “an additional tool to estimate the effects of treatment strategies”1 and should not be used as the only tool. Independent of the “fitness” of the model, a propensity-score analysis has the potential to give a better estimate but also leaves space for incomplete or inappropriate adjustments. In this case, the authors fail to include a thorough assessment of coronary anatomy for appropriateness for CABG versus percutaneous coronary intervention (PCI) (i.e., target vessels), coexisting conditions, or treatment preference.
In conclusion, while randomized trials are pending, these data show that consecutive patients with multivessel coronary artery disease who were nonrandomly selected by the treating physicians for treatment with CABG or PCI were equally likely to be alive at 18 months. These conclusions are very similar to those obtained from data from the Bypass Angioplasty Revascularization Investigation registry, which showed a 7-year survival rate of 85.8% for CABG and 86.1% for PCI.2
2 ReferencesAntonio Abbate, M.D.
Virginia Commonwealth University, Richmond, VA 23298
aabbate@vcu.edu Michael J. Lipinski, M.D.
University of Virginia, Charlottesville, VA 229021
D'Agostino RB Jr . Propensity score methods for bias reduction in the comparison of a treatment to a non-randomized control group. Stat Med 1998;17:2265-2281
CrossRef | Web of Science | Medline2
Feit F ,Mori Brooks M ,Sopko G , et al. Long-term clinical outcome in the Bypass Angioplasty Revascularization Investigation Registry: comparison with the randomized trial. Circulation 2000;101:2795-2802
Web of Science | Medline
To the Editor:
Hannan et al. report that CABG was associated with lower rates of death, death or myocardial infarction, and repeat revascularization than was treatment with drug-eluting stents. The stent group received drug-eluting stents “with or without other devices.” What proportion of patients was treated with drug-eluting stents alone rather than a combination of drug-eluting stents and bare-metal stents? Did the authors compare outcomes among patients treated exclusively with drug-eluting stents and those who underwent CABG? Furthermore, most repeat procedures in the stent group were planned, staged revascularization procedures, performed on arteries not treated in the first procedure. Did the authors compare target-vessel revascularization in the stent group with that in the CABG group? In addition, unadjusted survival curves for death and death or myocardial infarction indicate remarkably similar outcomes with the two treatment strategies, even among patients who had three-vessel disease. It appears that when the treating physicians selected stenting as the revascularization strategy, they were able to obtain 18-month outcomes that were similar to the outcomes after CABG. These findings serve as testimony to the excellent clinical judgment exercised by the treating physicians, who picked the appropriate revascularization method with knowledge of their patients' coexisting conditions.
Kishore J. Harjai, M.D.
Guthrie Clinic, Sayre, PA 18840
harjai_kishore@guthrie.org To the Editor:
Hannan et al. report that their evaluation of registry data for patients with multivessel coronary disease shows the superiority of CABG over PCI with drug-eluting stents. However, the relevance to clinical practice is severely limited because of important procedural and clinical omissions.
First, the extent of revascularization in each patient population and its relationship to the number of diseased vessels were not described. The equivalence of revascularization offered by CABG and PCI is essential in randomized studies comparing CABG and PCI.1 If all hemodynamically significant lesions were effectively revascularized by both PCI and CABG, then the variations in outcome could be more confidently attributed to the revascularization procedure. However, if patients with anatomy that is unsuitable for complete revascularization by CABG more frequently underwent incomplete revascularization with PCI, then this would bias the results against PCI.
Second, concurrent medical therapy was not reported. Statins, angiotensin-converting–enzyme inhibitors, beta-blockers, and antiplatelet agents all confer a survival benefit in this patient population, and systematic bias in their administration after PCI and CABG must be excluded.
1 ReferencesAustin C.C. Ng, M.B., B.S., B.Sc.(Med.)
Leonard Kritharides, M.B., B.S., Ph.D.
Concord Repatriation General Hospital, Concord, NSW 2139, Australia
lenk@med.usyd. edu. au 1
Mercado N ,Wijns W ,Serruys PW , et al. One-year outcomes of coronary artery bypass graft surgery versus percutaneous coronary intervention with multiple stenting for multisystem disease: a meta-analysis of individual patient data from randomized clinical trials. J Thorac Cardiovasc Surg 2005;130:512-519
CrossRef | Web of Science | Medline
To the Editor:
Hannan et al. provide yet another analysis of multivessel coronary-artery revascularization. A total of 76,714 patients had risk-adjusted survival that was significantly better with surgical revascularization than with bare-metal or drug-eluting stents.1 In addition, “there were no significant differences between the two groups [patients treated with drug-eluting stents and those treated with CABG] in the risk-adjusted rates of in-hospital or 30-day mortality,” which further illuminates the advanced state of surgical treatment. The “up-front” risk of surgery is approaching the risk of a percutaneous procedure. Furthermore, divergence of survival at 18 months is apparent, and the salutary effect of surgery may intensify with longer follow-up. The argument regarding poor vein patency may be irrelevant with increased use of arterial conduits and the application of neointimal hyperplasia inhibitors to vein grafts.2,3
Unlike the situation in randomized trials, New Yorkers are not confined to study criteria and stent use is not limited to Food and Drug Administration guidelines. Population studies, powered by an excellent clinical database, remain a valid observation of “real world” patients, and these observed significant survival benefits must be shared with our patients.4
4 ReferencesJoshua H. Burack, M.D.
New York Society for Thoracic Surgery, Brooklyn, NY 10801
jburack@downstate.edu 1
Hannan EL ,Racz MJ ,Walford G , et al. Long-term outcomes for coronary artery bypass surgery and stenting for the treatment of multivessel disease. N Engl J Med 2005;352:2174-2183
Full Text | Web of Science | Medline2
Schachner T . Pharmacologic inhibition of vein graft neointimal hyperplasia. J Thorac Cardiovasc Surg 2006;131:1065-1072
CrossRef | Web of Science | Medline3
Rankin JS ,Tuttle RH ,Wechsler AS ,Teichman TL ,Glower DD ,Califf RM . Techniques and benefits of multiple internal mammary artery bypass at 20 years of follow-up. Ann Thorac Surg 2007;83:1008-1015
CrossRef | Web of Science | Medline4
The Society of Thoracic Surgeons. Web page. (Accessed May 23, 2008, at http://www.sts.org/sections/patientinformation/articles/superioritycabg.html.)
Author/Editor Response
A common theme of the first three letters is that the unadjusted outcomes for patients treated with CABG surgery and those treated with drug-eluting stents were very similar, which is an indication that cardiologists and surgeons are exhibiting excellent clinical judgment in selecting patients. My colleagues and I would agree with this conclusion if the patient mixes were similar for the two groups, but patients treated with CABG had significantly worse ventricular function, were more likely to have had a recent myocardial infarction, and had a significantly higher prevalence of several coexisting conditions. Since all these factors are predictive of higher rates of death and myocardial infarction, we adjusted for them and other factors with the use of both multivariable methods and propensity analyses. Both methods led to the same conclusion. Propensity-score matching, another use of propensity methods, yielded similar relative outcomes.
Also, we do not agree that randomized, controlled trials will “appropriately evaluate the relative merits” of the two treatments. The results of randomized, controlled trials do not apply to many patients because such trials are limited primarily to low-risk patients and because they have artificially higher rates of complete revascularization among patients treated with stents and more frequent patient follow-up than in the real world. These factors all introduce a bias in favor of stenting. In our view, a combination of randomized, controlled trials and observational studies is required to provide a complete picture.
In response to Harjai, we did not think it was appropriate to exclude patients with both drug-eluting stents and bare-metal stents, because stenting was chosen as an option for them, and the comparison was between CABG and stenting. Excluding them would only have created more dissimilarities between the patients treated with surgery and those treated with PCI.
Ng and Kritharides imply that our comparison is not fair because the patients treated with stents were not as likely to have been effectively (completely) revascularized. We agree that patients with stents in whom complete revascularization is not attempted fare significantly worse than patients in whom complete revascularization is attempted.1 However, the purpose of our study was to compare the two procedures as they are used in practice, not in some ideal setting that does not reflect what patients can expect to encounter.
We agree with Ng and Kritharides that the use of medical therapy affects outcomes, and differential use according to procedure could be a source of bias. Also, a shorter duration of use of clopidogrel in the early years of drug-eluting stents may have created a bias against stenting.
1 ReferencesEdward L. Hannan, Ph.D., M.S.
University at Albany School of Public Health, Rensselaer, NY 12144
elh03@health.state. ny. us 1
Hannan EL ,Racz M ,Holmes DR , et al. The impact of completeness of percutaneous coronary intervention revascularization on long-term outcomes in the stent era. Circulation 2006;113:2406-2412
CrossRef | Web of Science | Medline
- Citing Articles (5)
Citing Articles
1
Warren M. Rozen, Xuan Ye, Pedro L. Guio-Aguilar, Alberto Alonso-Burgos, John Goldblatt, Mark W. Ashton, Iain S. Whitaker. (2012) Autologous microsurgical breast reconstruction and coronary artery bypass grafting: an anatomical study and clinical implications. Breast Cancer Research and Treatment
CrossRef2
Stuart J Head, Ad JJC Bogers, A Pieter Kappetein. (2012) Drug-eluting stent implantation for coronary artery disease: current stents and a comparison with bypass surgery. Current Opinion in Pharmacology
CrossRef3
Scot Garg, Giovanna Sarno, Juan-Luis Gutiérrez-Chico, Hector Garcia-Garcia, Josep Gomez-Lara, Patrick Serruys. (2011) Five-year outcomes of percutaneous coronary intervention compared to bypass surgery in patients with multivessel disease involving the proximal left anterior descending artery: an ARTS-II sub-study. EuroIntervention 6:9, 1060-1067
CrossRef4
Scot Garg, Patrick W. Serruys. (2010) Coronary Stents. Journal of the American College of Cardiology 56:10, S1-S42
CrossRef5
Yoshinobu Onuma, Neville Kukreja, Nicolo Piazza, Jannet Eindhoven, Chrysafios Girasis, Lisanne Schenkeveld, Ron van Domburg, Patrick W. Serruys. (2009) The Everolimus-Eluting Stent in Real-World Patients. Journal of the American College of Cardiology 54:3, 269-276
CrossRef
