Correspondence
Bypass Angioplasty Revascularization Investigation
N Engl J Med 1997; 336:136-138January 9, 1997
- Article
To the Editor:
The Bypass Angioplasty Revascularization Investigation (BARI) investigators (July 25 issue)1 reported the five-year clinical results of a trial comparing coronary-artery bypass grafting (CABG) with percutaneous transluminal coronary angioplasty (PTCA) in patients with multivessel disease. The study tested the hypothesis that in patients with multivessel disease and severe angina or ischemia, an initial revascularization strategy involving PTCA does not result in a poorer clinical outcome than CABG during a five-year follow-up period. Although all the patients in the study had clinically severe angina or objective evidence of ischemia, the incidence of these events after the procedures was not reported.
The authors may have assumed that most of the patients with recurrent angina underwent further revascularization, since they did report at least one repeated procedure in 54 percent of the patients initially treated with PTCA. However, not all patients with recurrent angina undergo repeated revascularization, and the BARI investigators do not mention the number of patients who had angina, objective ischemia, or both or used antianginal medication after repeated revascularization.
To draw conclusions about the clinical outcome after CABG or PTCA, one has to consider recurrent angina and objective ischemia as primary end points and include these variables in the event-free survival analysis, as my colleagues and I did in a small, retrospective, nonrandomized study.2
2 ReferencesEric Berreklouw, M.D., Ph.D.
Catharina Hospital, 5602 ZA Eindhoven, the Netherlands1
The Bypass Angioplasty Revascularization Investigation (BARI) Investigators
Full Text | Web of Science | Medline2
Berreklouw E ,Hoogsteen J ,van Wandelen R , et al. Bilateral mammary artery surgery or percutaneous transluminal coronary angioplasty for multivessel coronary artery disease? An analysis of effects and costs. Eur Heart J 1989;10:Suppl H:61-70
Web of Science | Medline
To the Editor:
I wish to comment on the method the BARI investigators used in making the diagnosis of myocardial infarction during the four days after the revascularization procedures. The BARI study used electrocardiography alone for this purpose, in contrast to other revascularization trials, which also measured cardiac-enzyme levels.1 There is no doubt that this approach led to an underestimation of myocardial infarction, as the authors pointed out; however, diagnosing myocardial infarction solely on the basis of the electrocardiographic criterion of Q waves is not an uncommon practice in the literature.
I think that the use of enzymatic confirmation of myocardial necrosis should be a prerequisite in the design of clinical trials, not so much to avoid missing the non–Q-wave myocardial infarctions as to be sure about the firm diagnosis of Q-wave myocardial infarctions. Before delving further in this seemingly paradoxical assertion, let me quote from Dr. Simoons's editorial (July 25 issue)2: “How should we interpret the somewhat better survival in spite of somewhat more frequent myocardial infarctions with CABG in BARI?” It is probable that the patients who underwent CABG, and presumably more complete revascularization than the ones who underwent PTCA, had a comparatively higher rate of new Q waves postoperatively that were due not to a higher rate of intraoperative myocardial infarction but to better perfusion of the wall contralateral to an old myocardial infarction.3 Incidentally, silent myocardial infarction is a common occurrence.
Improvement in myocardial perfusion leads to the unmasking of old myocardial infarctions,3 whereas transient, severe exercise-induced ischemia in the noninfarcted myocardium can have the opposite effect — causing the transient disappearance of the Q waves of an old myocardial infarction.4
4 ReferencesJohn E. Madias, M.D.
Elmhurst Hospital Center, Elmhurst, NY 113731
The EPIC Investigators
Full Text | Web of Science | Medline2
Simoons ML . Myocardial revascularization -- bypass surgery or angioplasty? N Engl J Med 1996;335:275-277
Full Text | Web of Science | Medline3
Bassan MM ,Oatfield R ,Hoffman I ,Matloff J ,Swan HJC . New Q waves after aortocoronary bypass surgery: unmasking of an old infarction. N Engl J Med 1974;290:349-353
Full Text | Web of Science | Medline4
Madias JE ,Mahjoub M ,Valance J . The paradox of negative exercise stress ECG/positive thallium scintigram: ischemic ST-segment counterpoise as the underlying mechanism. J Electrocardiol 1996;29:243-248
CrossRef | Web of Science | Medline
To the Editor:
The BARI trial, which randomly assigned 914 patients to undergo CABG and 915 patients to PTCA, “was designed to provide an overall comparison between coronary bypass surgery and angioplasty as initial treatment strategies in patients with multivessel coronary disease.” This statement implies that either treatment is equally suitable for the average patient with multivessel coronary disease. However, no information is given as to the proportion of patients with multivessel disease for whom either PTCA or CABG was equally suitable. It is only in the third-to-last sentence of this eight-page article (and in the last line of Simoons's accompanying editorial) that we are informed that the trial conclusions apply to only 12 percent of patients who require coronary revascularization.
As in other trials comparing PTCA with CABG for multivessel coronary disease, the fact that less than 10 percent of such patients are suitable for PTCA1 is heavily disguised in the BARI trial. Although such facts can be deduced by experts prepared to read the small print, they will invariably be missed by generalists and patients.
1 ReferencesDavid P. Taggart, M.D.
Oxford Heart Centre, Oxford OX3 9DU, United Kingdom1
Taggart DP . Angioplasty versus bypass surgery. Lancet 1996;347:271-272
CrossRef | Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: We agree with Dr. Berreklouw on the importance of functional outcomes after coronary revascularization, and we plan to report these results separately.
We are surprised by Dr. Taggart's assertion that we “heavily disguised” information on the patient population for whom the BARI results are applicable. We referenced a detailed report on patient selection in the BARI trial1 as well as a separate survey that we conducted of coronary revascularization in the United States.2 On the basis of this survey, we estimated 12 percent of all candidates for coronary revascularization would fit BARI inclusion criteria. It is probably more relevant to know that one third of patients with multivessel disease who undergo a first coronary revascularization would be eligible for BARI, since the choice between PTCA and CABG for the treatment of single-vessel or left main coronary artery disease is more settled and patients who have undergone prior procedures would be difficult to randomize.
We concur with Dr. Madias that enzyme confirmation of myocardial necrosis should be routinely used in the diagnostic algorithm for myocardial infarction. Chest pain, electrocardiography, and enzymatic data were used to determine Q-wave and non–Q-wave myocardial infarctions in the trial and are the subject of a future report. When the protocol was designed, it was recognized from the outset that it was difficult to diagnose non–Q-wave infarction in the immediate postoperative phase, since most patients have elevated creatine kinase MB concentrations after CABG, as compared with the average of 5 to 20 percent of patients reported in trials of PTCA alone.3 The specificity of minor increases in creatine kinase MB soon after CABG for myocardial necrosis is suboptimal, and a comparison of the rate of non–Q-wave myocardial-infarct events immediately after CABG and PTCA would be inappropriate.
The hypothesis that the higher rate of postoperative new Q waves after CABG is not due to a higher rate of intraoperative infarction but is rather a result of better perfusion of the wall contralateral to an old myocardial infarction is interesting but not tenable.4 In BARI, the development of new pathologic Q waves in patients after the procedure (a worsening by two grades) according to the Minnesota code was associated with a five-year mortality rate twice as high as that in patients without such Q waves (P<0.05), regardless of treatment (20 percent vs. 10 percent after CABG; 27 percent vs. 13 percent after PTCA).
In his editorial, Dr. Simoons pointed out that our results should not be interpreted to mean that the two procedures were equivalent simply because statistical significance of the survival advantage with CABG (89.3 percent, vs. 86.3 percent for PTCA) was not achieved (P = 0.19). Our interpretation agrees with his assessment. Patients with medically treated diabetes derived a very significant benefit from CABG, whereas among the other 1440 patients, the five-year survival rates were nearly identical (91.3 percent for CABG vs. 91.9 percent for PTCA).
4 ReferencesBernard Chaitman, M.D.
St. Louis University, St. Louis, MO 63110Allan D. Rosen, M.S.
University of Pittsburgh, Pittsburgh, PA 15261George Sopko, M.D., M.P.H.
National Heart, Lung, and Blood Institute, Bethesda, MD 20892Katherine M. Detre, M.D., Dr.P.H.
University of Pittsburgh, Pittsburgh, PA 15261Robert L. Frye, M.D.
Mayo Clinic Foundation, Rochester, MN 559051
Bourassa MG ,Roubin GS ,Detre KM , et al. Bypass Angioplasty Revascularization Investigation: patient screening, selection, and recruitment. Am J Cardiol 1995;75:3C-8C
CrossRef | Web of Science | Medline2
Detre KM ,Rosen AD ,Bost JE , et al. Contemporary practice of coronary revascularization in U.S. hospitals and hospitals participating in the Bypass Angioplasty Revascularization Investigation (BARI). J Am Coll Cardiol 1996;28:609-615
CrossRef | Web of Science | Medline3
Chaitman BR ,Jaffe AS . What is the true periprocedure myocardial infarction rate? Does anyone know for sure? The need for clarification. Circulation 1995;91:1609-1610
Web of Science | Medline4
Bassan MM ,Oatfield R ,Hoffman I ,Matloff J ,Swan HJC . New Q waves after aortocoronary bypass surgery: unmasking of an old infarction. N Engl J Med 1974;290:349-353
Full Text | Web of Science | Medline
