- Email: [email protected]
Beta-blockers and ventricular arrhythmias in dilated cardiomyopathy
Journal of the American College of Cardiology © 2000 by the American College of Cardiology Published by Elsevier Science Inc.
Vol. 35, No. 7, 2000 ISSN 0735-1097/00/$20.00
LETTERS TO THE EDITOR
Beta-Blockers and Ventricular Arrhythmias in Dilated Cardiomyopathy We read with interest the article by Di Lenarda et al. (1) in a recent issue of the Journal. The authors conclude that in patients with dilated cardiomyopathy who are poor responders to metoprolol, carvedilol treatment was associated with a favorable effect on left ventricular function as well as on ventricular arrhythmias. We would like to add some specific comments about ventricular arrhythmias. We are somewhat surprised about the baseline characteristics of the two groups of patients who thereafter received metoprolol or carvedilol. It seems clear that patients in the carvedilol group had a higher prevalence of nonsustained ventricular tachycardia (mean value 10 times higher than that in the metoprolol group) and of ventricular couplets (mean value and SEM ⬎20 times higher than that in the metoprolol group), although a statistical difference was not found between the two groups (Table 1). The statistical method used seems rather sophisticated, but no precise information is given on what variables were used for adjustment in the comparison of prevalence of ventricular arrhythmias, and the adjusted means are not shown. In view of the small number of patients in each group, it is not certain whether a simple nonparametric test would have been more appropriate. Although the inhomogeneity in the baseline characteristics does not seem to exist for ventricular ectopic beats, it is generally accepted that the prognostic value of ventricular arrhythmias is more important for ventricular couplets or nonsustained ventricular tachycardia than for isolated beats. This point has some importance to evaluate the compared effects of metoprolol and carvedilol on ventricular arrhythmias, because it may have been easier to find a favorable effect in the carvedilol group if its baseline characteristics were more severe; similarly a beneficial effect may not have been demonstrated in the metoprolol group because the better baseline characteristics in this group may hardly improve with medical treatment. It is important to separate the effects due to treatment from the effect of regression toward the mean (2). A phenomenon of regression toward the mean in both groups seems possible, because at 12 months, the mean of ventricular couplets remained higher in the carvedilol group (1.9/h) than in the metoprolol group (1.7/h), and the mean of nonsustained ventricular tachycardia also remained higher in the carvedilol group (0.08/h) than in the metoprolol group (0.04/h). There are actually no clear data on the specific effect of carvedilol on sudden death in large studies of patients with congestive heart failure, and the precise level of significance for the reduction of the relative risk of sudden death was not given in the U.S. Carvedilol Study (3). In contrast, the reduction of sudden death with metoprolol in the MERIT-HF study (4) was highly significant (relative risk 0.59, p ⫽ 0.0002). We recognize that the attempt of Di Lenarda to evaluate two different beta-blockers in a selected group of patients with persistent left ventricular dysfunction is of interest, but we consider that the conclusion of a benefit of carvedilol on ventricular arrhythmias remains speculative. Laurent Fauchier, MD
Centre de Recherche Service de Cardiologie B. Centre Hospitalier Universitaire Trousseau 37044 Tours, France E-mail: [email protected] Bruno Giraudeau, PhD Centre de Recherche Clinique Faculte´ de Me´decine de Tours Tours, France 37032 PII S0735-1097(00)00659-8
REFERENCES 1. Di Lenarda A, Sabbadini G, Salvatore L, et al. Long-term effects of carvedilol in idiopathic dilated cardiomyopathy with persistent left ventricular dysfunction despite chronic metoprolol. J Am Coll Cardiol 1999;33:1926 –34. 2. Bland JM, Altman DG. Regression towards the mean. BMJ 1994;308: 1499. 3. Packer M, Bristow MR, Cohn JN, et al. The effect of carvedilol on morbidity and mortality in patients with chronic heart failure. N Engl J Med 1996;334:1349 –55. 4. MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 1999;353:2001–7.
REPLY We are grateful for the opportunity to respond to the comments by Drs. Fauchier and Giraudeau concerning our recent article in the Journal (1). Specifically, they draw attention to the differences in the baseline arrhythmic profile of our patients with dilated cardiomyopathy who thereafter received carvedilol or metoprolol. We agree that the inhomogeneity in patient characteristics is surely an important issue, and, although in general terms, in our article, we have already commented on this point by admitting the existence of some differences in baseline characteristics between the carvedilol and metoprolol groups. With reference to the specific objection made by Drs. Fauchier and Giraudeau, we recognize that patients receiving carvedilol had a higher prevalence of both nonsustained ventricular tachycardia (NSVT) and ventricular couplets (VC), although these differences did not reach statistical significance (in all likelihood because of the small number of patients and the high variability). Because a more severe arrhythmic picture at baseline might have contributed to the finding of a favorable effect of carvedilol on NSVT and VC, Drs. Fauchier and Giraudeau are correct when they claim more precise information on the variables used for adjustment in the comparison of the prevalence of ventricular arrhythmias. In short, we can specify that all differences were tested considering the baseline arrhythmic profile and concurrent amiodarone therapy as potentially confounding factors (Table 1). On the basis of this statistical approach, we believe that the potential impact of patient heterogeneity on our results has been limited, and, moreover, that the effects seen on ventricular arrhythmias are more likely due to beta-blocker therapy than to a phenomenon of regression toward the mean. A simple test (both parametric or nonparametric) would hardly be appropriate in
Vol. 35, No. 7, 2000 ISSN 0735-1097/00/$20.00
LETTERS TO THE EDITOR
Beta-Blockers and Ventricular Arrhythmias in Dilated Cardiomyopathy We read with interest the article by Di Lenarda et al. (1) in a recent issue of the Journal. The authors conclude that in patients with dilated cardiomyopathy who are poor responders to metoprolol, carvedilol treatment was associated with a favorable effect on left ventricular function as well as on ventricular arrhythmias. We would like to add some specific comments about ventricular arrhythmias. We are somewhat surprised about the baseline characteristics of the two groups of patients who thereafter received metoprolol or carvedilol. It seems clear that patients in the carvedilol group had a higher prevalence of nonsustained ventricular tachycardia (mean value 10 times higher than that in the metoprolol group) and of ventricular couplets (mean value and SEM ⬎20 times higher than that in the metoprolol group), although a statistical difference was not found between the two groups (Table 1). The statistical method used seems rather sophisticated, but no precise information is given on what variables were used for adjustment in the comparison of prevalence of ventricular arrhythmias, and the adjusted means are not shown. In view of the small number of patients in each group, it is not certain whether a simple nonparametric test would have been more appropriate. Although the inhomogeneity in the baseline characteristics does not seem to exist for ventricular ectopic beats, it is generally accepted that the prognostic value of ventricular arrhythmias is more important for ventricular couplets or nonsustained ventricular tachycardia than for isolated beats. This point has some importance to evaluate the compared effects of metoprolol and carvedilol on ventricular arrhythmias, because it may have been easier to find a favorable effect in the carvedilol group if its baseline characteristics were more severe; similarly a beneficial effect may not have been demonstrated in the metoprolol group because the better baseline characteristics in this group may hardly improve with medical treatment. It is important to separate the effects due to treatment from the effect of regression toward the mean (2). A phenomenon of regression toward the mean in both groups seems possible, because at 12 months, the mean of ventricular couplets remained higher in the carvedilol group (1.9/h) than in the metoprolol group (1.7/h), and the mean of nonsustained ventricular tachycardia also remained higher in the carvedilol group (0.08/h) than in the metoprolol group (0.04/h). There are actually no clear data on the specific effect of carvedilol on sudden death in large studies of patients with congestive heart failure, and the precise level of significance for the reduction of the relative risk of sudden death was not given in the U.S. Carvedilol Study (3). In contrast, the reduction of sudden death with metoprolol in the MERIT-HF study (4) was highly significant (relative risk 0.59, p ⫽ 0.0002). We recognize that the attempt of Di Lenarda to evaluate two different beta-blockers in a selected group of patients with persistent left ventricular dysfunction is of interest, but we consider that the conclusion of a benefit of carvedilol on ventricular arrhythmias remains speculative. Laurent Fauchier, MD
Centre de Recherche Service de Cardiologie B. Centre Hospitalier Universitaire Trousseau 37044 Tours, France E-mail: [email protected] Bruno Giraudeau, PhD Centre de Recherche Clinique Faculte´ de Me´decine de Tours Tours, France 37032 PII S0735-1097(00)00659-8
REFERENCES 1. Di Lenarda A, Sabbadini G, Salvatore L, et al. Long-term effects of carvedilol in idiopathic dilated cardiomyopathy with persistent left ventricular dysfunction despite chronic metoprolol. J Am Coll Cardiol 1999;33:1926 –34. 2. Bland JM, Altman DG. Regression towards the mean. BMJ 1994;308: 1499. 3. Packer M, Bristow MR, Cohn JN, et al. The effect of carvedilol on morbidity and mortality in patients with chronic heart failure. N Engl J Med 1996;334:1349 –55. 4. MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 1999;353:2001–7.
REPLY We are grateful for the opportunity to respond to the comments by Drs. Fauchier and Giraudeau concerning our recent article in the Journal (1). Specifically, they draw attention to the differences in the baseline arrhythmic profile of our patients with dilated cardiomyopathy who thereafter received carvedilol or metoprolol. We agree that the inhomogeneity in patient characteristics is surely an important issue, and, although in general terms, in our article, we have already commented on this point by admitting the existence of some differences in baseline characteristics between the carvedilol and metoprolol groups. With reference to the specific objection made by Drs. Fauchier and Giraudeau, we recognize that patients receiving carvedilol had a higher prevalence of both nonsustained ventricular tachycardia (NSVT) and ventricular couplets (VC), although these differences did not reach statistical significance (in all likelihood because of the small number of patients and the high variability). Because a more severe arrhythmic picture at baseline might have contributed to the finding of a favorable effect of carvedilol on NSVT and VC, Drs. Fauchier and Giraudeau are correct when they claim more precise information on the variables used for adjustment in the comparison of the prevalence of ventricular arrhythmias. In short, we can specify that all differences were tested considering the baseline arrhythmic profile and concurrent amiodarone therapy as potentially confounding factors (Table 1). On the basis of this statistical approach, we believe that the potential impact of patient heterogeneity on our results has been limited, and, moreover, that the effects seen on ventricular arrhythmias are more likely due to beta-blocker therapy than to a phenomenon of regression toward the mean. A simple test (both parametric or nonparametric) would hardly be appropriate in