Carvedilol Produces Dose-Related Improvements in Left Ventricular Function and Survival in Subjects With Chronic Heart Failure

In an attempt to stabilize compromised pump performance, the failing human heart is subjected to increased adrenergic stimulation.¹²³⁴⁵ Although this compensatory mechanism may initially support and maintain cardiac output, model systems⁶⁷ and human studies^89101112 provide considerable evidence that continuous adrenergic stimulation may be harmful. Perhaps the most compelling evidence that chronic adrenergic stimulation is harmful to the failing heart derives from experience with β-adrenergic blocking agents, which consistently improve LV function in subjects with chronic heart failure^{13141516171819202122} by reversing intrinsic systolic dysfunction¹⁵¹⁶²⁰ via a time-dependent “biological” effect on the myocardium.²³

Carvedilol is a “third-generation”²⁴ β-blocking agent that at therapeutic target doses blocks all three adrenergic receptors that mediate a positive inotropic response in human cardiac myocytes, with a rank order of potency of β₁>α₁>β₂.²⁵ Because of its α-blocking properties,²⁶²⁷ carvedilol is a moderate vasodilator on acute administration,²¹²⁸ but with long-term treatment the vasodilator activity is no longer prominent.¹⁷ However, the vasodilator action of the compound contributes to its relatively good initial tolerability, because, in contrast to nonvasodilator β-blockers,²⁹³⁰ acute administration of carvedilol does not typically result in profound myocardial depression and clinically important reductions in cardiac output.²¹²⁸

One important aspect of investigating drug action is to define both efficacy and adverse event dose-response characteristics to calculate a therapeutic index and predict a risk:benefit ratio for individual patients. Although previous single-center smaller trials conducted with a single target dose of 25 to 50 mg BID have indicated that carvedilol improves LV function and may improve heart failure symptoms and submaximal exercise,¹⁷¹⁸²¹ the influence of dose on these and other important parameters in longer-term follow-up is unknown. For these reasons, we undertook an evaluation of the dose-response characteristics of carvedilol over a 6-month period of maintenance treatment in a design that included evaluation of functional capacity (submaximal exercise), LV function, heart failure symptoms, heart failure morbidity, and survival.

Methods

Protocol Description

The MOCHA trial was a 6-month, double-blind, placebo-controlled, multicenter dose-response evaluation of this third-generation β-blocking agent in subjects with stable, symptomatic, chronic heart failure. Four treatment groups were evaluated: placebo, low-dose carvedilol (6.25 mg BID), medium-dose carvedilol (12.5 mg BID), and high-dose carvedilol (25 mg BID).

The primary objective of this trial was to compare the efficacy of three increasing doses of carvedilol with that of placebo for improvement in submaximal exercise. Two submaximal exercise tests were used: the 6-minute walk test³¹ and the 9-minute self-powered treadmill test.³² The secondary objectives of this trial were to assess changes in QOL with the Minnesota Living With Heart Failure questionnaire,³³ changes in NYHA functional class, changes in EF, need for hospitalization due to heart failure and other cardiovascular causes, and changes in signs and symptoms of heart failure. The safety assessment included an evaluation of the occurrence of adverse clinical experiences, changes in laboratory values, deaths, and ECG values in the four treatment arms.

Patients initially underwent a 3-week screening phase, during which time eligibility for the trial was determined, including a baseline radionuclide LVEF. Qualified patients signed informed written consent, followed by performance of additional baseline measurements and then institution of a 2-week challenge phase of open-label carvedilol. The carvedilol challenge consisted of an initial dose of 6.25 mg BID, which could be reduced to 3.125 mg BID if the patient developed symptoms related to hypotension or worsening heart failure. In patients reduced to 3.125 mg BID, the challenge dose was increased to 6.25 mg BID in the second week, and successful placement on this dose was required to establish tolerability. Patients tolerating challenge were then randomized to one of the four treatment groups. Study medication was increased on a weekly basis in the 12.5-mg BID and 25-mg BID carvedilol groups, such that at the end of 2 weeks of up-titration all patients would have reached their target or maximally tolerated dose. All patients were up-titrated in a fashion that maintained the double-blind feature. Patients having difficulty with up-titration could return to the previous level and had 4 weeks to reach the target dose. After up-titration, a 6-month maintenance period ensued, during which time patients received placebo or one of the three different doses of carvedilol and were followed up at frequent intervals by the investigators and study nurses. After 6 months of maintenance treatment, efficacy tests performed at baseline were repeated. In addition, submaximal exercise tests and QOL measurements were performed at the end of 2 and 4 months of maintenance treatment, and NYHA functional class and global assessments were determined on a monthly basis during maintenance treatment.

Entry criteria for this study included male or female patients between 18 and 85 years of age who had an EF ≤35% and symptomatic heart failure from ischemic or nonischemic dilated cardiomyopathy. Symptoms had to be present for at least 3 months, the 6-minute walk test had to be between 150 and 425 m (revised upward to 150 to 450 m by protocol amendment 6 months into the study), and stability was defined as no change in NYHA class or absence of hospitalization for heart failure during the 1 month before baseline testing. Patients had to be on stable doses of diuretics for 2 weeks before baseline testing and a stable dose of ACE inhibitor for at least 1 month before baseline testing. Patients who were intolerant of ACE inhibitors were allowed in the study, but they could not be rechallenged with an ACE inhibitor during the course of the trial. Digoxin use was optional, but if patients were taking digoxin, they had to have been started at least 2 months before baseline testing and to have been on a stable dose for at least 1 month. The use of hydralazine and nitrates was also optional, but if patients were on these medications they had to have been started at least 2 months before baseline testing and to have been on a stable dose for at least 1 month. Additional entry criteria included a resting heart rate in the sitting position of ≥68 bpm.

Exclusion criteria included the presence of uncorrected valvular disease, hypertrophic cardiomyopathy, or postpartum cardiomyopathy. Additional exclusion criteria included documented uncontrolled symptomatic or sustained ventricular tachycardia. Acute myocardial infarction could not have occurred within 3 months before screening, and a percutaneous transluminal coronary angioplasty, coronary artery bypass graft surgery, or heart transplant could not be planned or be likely within the 6 months after screening. Other exclusions included the presence of sick sinus syndrome, second- or third-degree heart block not treated with a pacemaker, symptomatic peripheral vascular disease limiting exercise testing, a sitting systolic blood pressure of <85 mm Hg or >160 mm Hg, a cerebrovascular accident within the previous 3 months, cor pulmonale, obstructive pulmonary disease requiring oral bronchodilator or steroid therapy, serum creatinine >2.5 mg/dL, serum SGOT or SGPT >3 times the upper limits of normal, a chronic biliary disorder, limitation of exercise other than that due to heart failure, a systemic or terminal disease that would limit physical function or survival during the trial, an untreated endocrine disorder such as hyperthyroidism, brittle insulin-dependent diabetes requiring frequent hospitalizations, an alcohol intake of >100 g/d, unwillingness to cooperate or give written informed consent, pregnant or lactating women, platelet count <100 000 mm³ or white blood cell count <3000 mm³, use of an investigational drug within 30 days of entry into the challenge phase, and a history of drug sensitivity or adverse reactions to α- or β-blockers. The following medications were excluded for concomitant use within 2 weeks of baseline testing: monamine oxidase inhibitors, calcium channel blockers, flosequinan, α- or β-blockers, disopyramide, flecainide, encainide, moricizine, propafenone, or sotalol. Amiodarone could not have been used within 3 months of baseline testing.

The 6-minute corridor walk test was conducted as previously described by Guyatt et al.³¹ A 9-minute self-activated treadmill test³² was performed in a standardized fashion with a Tunturi Jogger II mechanical treadmill equipped with odometers that were standardized in all centers. Only subjects whose baseline or screening exercise tests were limited by dyspnea or fatigue were continued in the study. When exercise tests were performed on the same day, the 6-minute walk test was performed first, and at least a 1-hour rest period was allowed before the second test. Subjects underwent at least four baseline exercise tests before initial drug challenge: two during the screening phase to familiarize subjects with the tests and at least two during baseline testing. Results of the fourth 6-minute walk test were used as the baseline value provided that the distance was within ±10% of the previous test. If this was not the case, the subject returned for an additional test and was discontinued from the study if the distance was not within ±10% of the previous test. Whenever a 6-minute walk test was performed, a 9-minute treadmill test was also obtained.

LVEF was measured by radionuclide ventriculography. Hospitalizations due to heart failure or other cardiovascular causes were reported prospectively by the investigators, and mortality was classified by the US Carvedilol Heart Failure Trials Program Steering Committee according to procedures used in the PROMISE¹¹ and SAVE³⁴ trials. A data and safety monitoring board prospectively monitored the serious adverse events, including deaths, in this and the three other US carvedilol trials that composed the trial program.

Statistical Analysis

On the basis of a review of the literature and preliminary data available from pilot studies,¹⁷¹⁸ it was projected that a sample size of 300 patients (75 per treatment group) would provide 90% power at the P=.05 level of significance to detect a dose-response effect for both the 6-minute corridor and 9-minute self-powered treadmill walk tests. The effect sizes used for the power calculations for the 6-minute corridor walk test were 7 m,14 m, 28 m, and 56 m for the placebo group and the 6.25-mg-BID, 12.5-mg-BID, and 25-mg-BID carvedilol groups, respectively. All analyses were by intent to treat; patients who did not tolerate their target dose and were maintained at a lower dose were analyzed with the treatment group to which they were originally randomized. For submaximal exercise tests, heart failure global assessments, QOL measurements, and estimates of NYHA functional class, the end-point value used was the last available measurement. Per protocol specification, the secondary end point of hospitalizations for heart failure or other cardiovascular reasons was tabulated during the maintenance period for patients completing at least 2 months of maintenance. The analysis of deaths was inclusive of the entire up-titration and maintenance periods.

For the primary efficacy evaluations (change from baseline in distance covered for each of the two walk tests), both multivariate and univariate ANOVA procedures that included the effects of study center, treatment group, and the interaction between them were used. The effectiveness of carvedilol was assessed with a two-tailed test for linear trend, and if it was significant, pairwise comparisons with placebo were performed. In addition, 95% CIs on the difference between each carvedilol group and the placebo mean change were estimated. Supplemental analyses of the primary efficacy variables using two-way ANOVA on rank-transformed data were carried out to determine the effects of early termination or withdrawal from the study. The change from baseline for all continuous secondary and safety variables was analyzed by use of the same model as for the primary efficacy analysis; categorical variables were analyzed by contingency table analyses using the Cochran-Mantel-Haenszel procedure when necessary. Once these predetermined assessments were completed, patients were stratified for subgroup analyses by baseline variables that could have influenced outcomes. ANOVA procedures were used to evaluate the significance of drug-stratification variable interactions. All-cause mortality was tested for linear trend in two ways: (1) by examination of the linear component of the total χ² (using the Mantel-Haenszel option) to assess crude rates and (2) by use of a Cox proportional-hazards regression model with treatment as covariant to assess the linear relationship among the four treatment groups. Other categorical variables were analyzed as per crude mortality rates by contingency table analysis. In these categorical analyses, when the linear trend test was significant (P<.05), the significance of individual carvedilol groups compared with placebo was determined by 2×2 contingency table analysis, without adjustment for multiple groups. Finally, additional analyses of mortality and other important categorical variables were performed by combining all the carvedilol groups into one active treatment group and comparing the results with those of the placebo groups. For mortality, Kaplan-Meier curves were constructed by use of a Cox regression model, and crude mortality rates and other categorical variables were compared by contingency table analysis as described above.

Results

Subject Demographics and Baseline Cardiac and Exercise Function

As can be observed in Table 1, there were no differences in subject descriptors among the four treatment groups. The majority of subjects were Caucasian men ≈60 years old who were evenly divided between NYHA class II and class III heart failure, who had moderate to severe LV dysfunction with a mean EF of 23%, and who had moderate impairment of submaximal exercise, with baseline 6-minute walk tests of ≈360 m. Table 2 shows the percentages of subjects on digoxin, diuretics, and ACE inhibitors on entry into and during the maintenance period of the trial, because background treatment could not be altered by adding or dropping these medications.

Results of Open-Label Challenge

Fig 1 gives the carvedilol challenge data. Three hundred seventy-six subjects signed informed consent forms and received a 6.25-mg-BID×2-week challenge of open-label carvedilol. As shown in Figs 1 and 2, 92% of subjects tolerated this challenge and were randomized to one of the four groups. As shown in Fig 2, in the majority of subjects who did not tolerate challenge, this was because of adverse events, primarily symptoms related to orthostatic hypotension or myocardial depression. One subject died during the 2- to 4-week challenge period.

Target Doses Achieved, Heart Rate, and Blood Pressure Effects

Table 3 gives the maintenance doses of study medication achieved in the four groups. As can be seen, the maintenance doses in the three carvedilol groups were similar to the target doses. Heart rate data obtained from the ventricular rate of ECGs taken at baseline and at 2, 4, and 6 months of maintenance treatment are also given in Table 3. There is a reduction in heart rate at all three doses of carvedilol that is minimally dose-related. As shown in Table 3, systolic and diastolic blood pressures taken in the sitting position were not affected by chronic carvedilol treatment.

Median Follow-up, Adverse Events, and Discontinuation From Study Medication

This trial consisted of an initial up-titration period of 2 to 4 weeks, followed by 6 months on maintenance treatment. The intent-to-treat median values for exposure to study medication were 190, 194, 196, and 196 days on the placebo, 6.25-mg-BID, 12.5-mg-BID, and 25-mg-BID groups, respectively. A list of adverse events encountered on study medication is given in Table 4. As can be seen, by linear trend test the only adverse events more prevalent in carvedilol-treated subjects were dizziness and bradycardia. As shown in Table 5, however, relatively few carvedilol-treated subjects were withdrawn because of adverse effects—in fact, a lower percentage than in the placebo group. It can also be observed in Table 5 that the overall discontinuation from the study rate was lower in the carvedilol groups than in the placebo group.

Submaximal Exercise

Fig 3 gives the 6-minute walk results, one of two components of the primary end point. Fig 4 gives the results of the second component of the primary end point, the 9-minute self-activated treadmill test. As can be seen in Figs 3 and 4, compared with the placebo group, carvedilol had no effect on either component.

LV Function

Fig 5 gives the results on LV function as assessed by radionuclide ventriculography expressed as the change from baseline in EF units (%) in the four groups. Carvedilol treatment resulted in a dose-related improvement in LVEF, with each group being statistically significantly different from placebo and the dose response assessed by the linear-trend test being highly statistically significant (P<.01). The effect of carvedilol was apparent for both ischemic and nonischemic cardiomyopathies (Fig 6). As Fig 6 shows, there was a tendency for the improvement in LV function to be more dose-related in the nonischemic group. However, there was no statistically significant difference between the dose-response relations by the linear-trend test (P=.068).

Hospitalizations

Fig 7 and Table 6 give the cardiovascular hospitalization rate per subject data, obtained as per protocol during the 2- to 6-month maintenance phase of the trial. As can be seen in Fig 7, carvedilol treatment was associated with a reduction in cardiovascular hospitalization rate that was statistically significant in all individual groups and for linear trend. As can be observed in Table 6, the reduction in hospitalization rate in the carvedilol groups was not at the expense of length of hospital stay.

Although it was not a protocol-specified end point, we also analyzed total hospitalizations from time of randomization. As a percentage of the number of patients randomized, the results were 23.8%, 13.2%, 18.0%, and 13.5% in the placebo and three increasing carvedilol doses, respectively (P=.16 by linear trend). The mean numbers of hospitalizations per patient were 0.37, 0.16, 0.21, and 0.18, respectively (P=.07 by linear trend). The mean number of hospital days per patient was reduced by carvedilol (3.1, 1.1, 1.5, and 1.3, respectively, P=.01). In all hospitalization categories, if the carvedilol groups were combined, there was a significant reduction (P<.05) in favor of carvedilol treatment.

QOL and Global Assessment Measurements

Table 7 presents the results of the Minnesota Living With Heart Failure Questionnaire, indicating that there are no significant differences among the four groups in this index expressed as the total score incorporating both physical and emotional dimensions.³³ There were also no significant changes in the individual components of the Minnesota Questionnaire (physical dimension, respective mean changes from baseline in the placebo and low-, medium-, and high-dose carvedilol groups of −3.1, −3.7, −3.2, and −2.0 [P=.360] and emotional dimension, −1.4, −1.5, −1.3, and −0.5 [P=.25]). Results of patient and physician global assessments are given in Table 8, and for both types of assessment groups, carvedilol treatment was associated with a trend toward significant improvement. NYHA functional class ranking was also unaffected by carvedilol (P=.64 for linear trend in improvement, data not shown).

Survival

Fig 8 gives the crude mortality rate as a percentage of randomized subjects in the four treatment groups. The placebo-treated group had 13 deaths, for a 15.5% crude mortality over the ≥6 months of the study. As can be observed in Fig 8, there was a dose-related, statistically significant reduction in mortality in the carvedilol-treated groups, with respective mortality rates of 6.0% (log-rank analysis: relative risk, 0.356 with 95% CI of 0.127 to 0.998, P<.05), 6.7% (relative risk, 0.416 and 95% CI, 0.158 to 1.097, P=.07), and 1.1% (relative risk, 0.067 and 95% CI, 0.009 to 0.512, P<.001) for the carvedilol doses of 6.25 mg BID, 12.5 mg BID, and 25 mg BID, respectively. As shown in Fig 8, the reduction in mortality by carvedilol was highly statistically significant (P<.001) by the linear trend test.

Fig 9 gives actuarial survival curves for the three carvedilol groups combined and compared with the placebo-treated group. As can be observed, by log-rank analysis there was a highly statistically significant reduction in mortality (by 73%; relative risk, 0.272 and CI, 0.124 to 0.597, P<.001) in carvedilol-treated subjects. Moreover, the reduction in mortality shown in Fig 9 appeared to occur in both ischemic and nonischemic cardiomyopathies, with respective relative risk reductions in the crude mortality rate of 73% and 63% (Table 9). In addition, carvedilol appeared to favorably affect deaths classified as sudden as well as those due to progressive pump dysfunction, as shown in Table 9.

Discussion

In failing human ventricular myocardium, β-blocking agents produce effects unique among heart failure medications. Initially, β-blocking agents depress myocardial function through the pharmacological action of withdrawal of adrenergic support to the myocardium, but after several months of treatment they improve intrinsic myocardial function through a time-dependent and as yet undetermined salutary “biological” effect on failing heart muscle.²³³⁵ The third-generation β-blocker/vasodilator carvedilol differs from first-generation compounds such as propranolol or timolol in that it produces less initial hemodynamic compromise, since on acute dosing, cardiac output is usually maintained or even slightly increased because of the vasodilator properties of the compound.²¹²⁸ Indeed, in phase II studies in moderate or severe heart failure, carvedilol was very well tolerated,¹⁷¹⁸²¹ in contrast to propranolol.³⁶ Carvedilol differs from second-generation, β₁-selective compounds in that (1) at higher doses it blocks β₂- and α₁- as well as β₁-receptors; (2) it reduces cardiac adrenergic activity³⁷ ; and (3) it prevents β-receptor upregulation and restoration of β-adrenergic signal transduction,²⁵³⁷ which typically occurs with metoprolol.²⁹³⁷³⁸ The result is that at target pharmacological doses, such as those used in phase II heart failure trials,¹⁷¹⁸²¹ carvedilol produces a complete and comprehensive antiadrenergic action in the failing human heart.

In this trial, carvedilol produced a dose-related improvement in LV function as assessed by radionuclide-determined EF similar to previously reported results with bucindolol.²² The increase in LVEF by the highest dose (25 mg BID) of carvedilol was similar to results in three previously reported phase II trials with carvedilol.¹⁷¹⁸²¹ The increase in LVEF occurred in both ischemic and nonischemic cardiomyopathies, which is consistent with previous carvedilol data¹⁷¹⁸²¹ and with previous results with the third-generation nonselective β-blocker bucindolol.²² In what appeared to be a mirror-image pattern to the improvement in LV function, carvedilol was associated with a dose-related reduction in mortality. Moreover, this reduction in mortality appeared to be present in both ischemic and nonischemic cardiomyopathies and was observed for both sudden and progressive pump dysfunction classifications of death. This differs from results reported for the second-generation compound bisoprolol, which in the CIBIS trial³⁹ did not reduce mortality in the ischemic cardiomyopathy population and did not reduce the incidence of sudden death. Although in MOCHA the number of deaths was small (a total of 25), the dose-relatedness of the apparent survival benefit and the robust significance levels (P<.001 both for linear dose response and for the combined carvedilol groups, relative risk reduction of 73% for the combined carvedilol groups with confidence limits not including 1.0) support the interpretation that the observed mortality-reducing effect was not due to chance. In addition, results in the combined US Carvedilol Heart Failure Trials Program indicate a similar, highly statistically significant reduction in mortality (by 65%) that was consistent across all trials.⁴⁰ Although the current results on mortality reduction by carvedilol exceed the quantitative estimates from the CIBIS (by 21%)³⁹ and MDC (by 35% for the combined morbidity-mortality end point)¹⁹ trials, differences in the degree of antiadrenergic action³⁷ or important ancillary properties such as antioxidant effects⁴¹ provide a potential pharmacological basis for the greater efficacy of carvedilol.

The “primary” end point of this study, improvement in submaximal exercise, was not altered by treatment with carvedilol. This is in contrast to results in two phase II trials, in which, compared with placebo, carvedilol either improved the 6-minute walk distance¹⁸ or, compared with baseline, improved the duration of sustained submaximal exercise maintained at a fixed percentage of the workload.¹⁷ The intent was to measure submaximal exercise by two techniques, in view of the lack of a generally accepted method that can be used to establish efficacy of heart failure treatment. Unfortunately, information available after the completion of the trial indicates that the 9-minute self-powered treadmill test, although quite reproducible, measures predominantly maximal rather than submaximal exercise capacity.⁴² In the present study, no positive trends in favor of carvedilol were noted in either the 6-minute corridor walk or the 9-minute self-powered treadmill test. However, on the basis of the MOCHA trial data, it can be stated that carvedilol has no adverse effect on submaximal or maximal exercise, which is important for the use of this type of agent in chronic heart failure. The reason that carvedilol did not improve submaximal exercise may relate to a true lack of a treatment effect or to methodological difficulties in measuring submaximal exercise in multicenter trials.

Carvedilol was associated with a reduction in cardiovascular hospitalization rate by 58% to 64%. This important index of morbidity, QOL, and pharmacoeconomics was reduced by all three doses of carvedilol. Although the reduction in hospitalization rate obeyed a dose-response pattern statistically, within the carvedilol-treated groups there was no obvious relationship between increasing dose and decreasing rate. QOL as assessed by the Minnesota instrument³³ was not statistically improved by carvedilol at the end of the study. However, there was a trend toward improvement by carvedilol in the global assessment instrument for both the patient and physician assessments. On balance, the lack of a measurable effect on improvement in QOL is similar to what was observed in the bucindolol multicenter trial, in which the same instrument did not detect differences between active drug and placebo after 12 weeks of treatment. As with carvedilol,¹⁷¹⁸ bucindolol¹³¹⁴ had been associated with improved QOL in smaller single-center studies that used different assessment techniques. It is likely that the inconsistency in these results is due to difficulties in measuring QOL, the unique effects of β-blocker therapy in a heart failure population in which benefit may be variably preceded by drug-related worsening of symptoms and QOL, or an inherent lack of a favorable QOL effect with this type of treatment. However, the latter would seem unlikely in view of the aforementioned previous positive QOL results. The most likely explanation may be the relatively short duration of follow-up. This explanation is supported by QOL data from the MDC trial, in which metoprolol was associated with an improved QOL at 12 or 18 months.

This trial was not designed to evaluate an efficacy effect on mortality, because in the entire Carvedilol US Trials program, deaths were being assessed primarily as a safety end point. In addition, the median follow-up of ≈6 months was less than is usually associated with mortality trials. However, it should be noted that the number of subjects enrolled in the MOCHA trial was 36% greater and the median follow-up was longer than in CONSENSUS-I,⁴³ a trial that is widely considered to have unambiguously demonstrated a mortality reduction by the ACE inhibitor enalapril. Although the CONSENSUS trial had more end points (a total of 118 deaths), the present trial has the advantage of having demonstrated a highly significant dose-related reduction in mortality that correlated with a progressive reduction in estimated catecholamine β-receptor occupancy to values that are probably below or close to the pharmacological thresholds of norepinephrine- or epinephrine-mediated responses.⁴⁴

Finally, the results of this study strongly support the hypothesis²³³⁵ that the improvement in LV function observed for β-blocking agents is associated with improved survival. Unlike that for positive inotropic agents or vasodilators, improvement in LV function by antiadrenergic treatment is via a time-dependent “biological” improvement in intrinsic systolic function,²³³⁵ resulting in a partial reversal of the underlying abnormalities afflicting the failing human heart. It is perhaps not surprising that such a favorable effect on the intrinsic myocardial function would be associated with an improvement in the natural history of heart failure, but until now the data supporting this concept were only suggestive.¹⁹³⁹ The fact that carvedilol is associated with an apparently greater effect on survival than are second-generation compounds is also not surprising, in view of the more complete antiadrenergic properties of carvedilol²⁵³⁷ and/or the existence of additional favorable ancillary properties.⁴¹

In summary, in this trial carvedilol produced a dose-related reduction in mortality in a chronic heart failure population. In addition, MOCHA is the first individual clinical trial to demonstrate a mortality benefit with β-blocker treatment in chronic heart failure. Although the number of deaths in the trial was small (n=25), the dose-relatedness of the findings and the relationship to improved LV systolic function suggests that the observations are not due to chance. Further clinical trials will be necessary to confirm and extend these findings as well as to position this treatment into its proper place in the treatment of chronic heart failure.

Selected Abbreviations and Acronyms

Appendix

The following investigators (and their performance sites) composed the MOCHA Study Group.

L. Kuo, Albuquerque, NM (Lovelace Scientific Resources); R. Karlsburg, Beverly Hills, Calif; K. Adams, Chapel Hill, NC (University of North Carolina); E. Eichhorn, Dallas, Tex (Dallas Veterans Administration Medical Center); C.W. Yancy, Dallas, Tex (University of Texas Southwestern Medical Center); M. Bristow, W. Abraham, Denver, Colo (University of Colorado); J. Young, Houston, Tex (Baylor College of Medicine); S.E. El Hafi, Houston, Tex; G. Schroth, Houston, Tex (University of Texas Medical School); J. O'Connell, Jackson, Miss (University of Mississippi Medical Center); A. Miller, Jacksonville, Fla (University of Florida); J.A. Bowers, Las Vegas, Nev (Heart Institute of Nevada); S. Krueger, Lincoln, Neb (Nebraska Heart Institute); V. DeQuattro, Los Angeles, Calif (University of Southern California School of Medicine); P.S. Rahko, Madison, Wis (University of Wisconsin); K.B. Ramanathan, Memphis, Tenn (University of Tennessee, Memphis); E. deMarchena, Miami, Fla (University of Miami); S. Kubo, J. Cohn, Minneapolis, Minn (University of Minnesota Medical School); U. Thadani, Oklahoma City, Okla (University of Oklahoma Health Sciences Center); R.P. Sorkin, Park Ridge, Ill (Lutheran General Hospital); J.V. Felicetta, Phoenix, Ariz (Carl T. Hayden VA Medical Center); R. Hershberger, Portland, Ore (Oregon Health Sciences); L.J. Olson, Rochester, Minn (Mayo Medical School); E.M. Gilbert, Salt Lake City, Utah (University of Utah); L. Yellen, San Diego, Calif (Cardiology Associates Medical Group of East San Diego, Inc); H. Ingersoll, San Diego, Calif (Sharp Rees-Stealy Medical Group Center); S. Woodley, San Francisco, Calif (California Pacific Medical Center); B. Massie, San Francisco, Calif (Veterans Administration Medical Center); M. Fowler, Stanford, Calif (Stanford University Cardiovascular Medicine); L.W. Miller, S.H. Jennison, St Louis, Mo (St Louis University); A.J. Lonigro, H. Stratmann, St Louis, Mo (St Louis University Medical Center); K.A. Narahara, Torrance, Calif (Harbor-UCLA Medical Center); S. Butman, Tucson, Ariz (University Medical Center); F. Kahl, Winston-Salem, NC (Bowman-Gray School of Medicine).

US Multicenter Carvedilol Trials Program Steering Committee: M.R. Bristow, J.N. Cohn, W.S. Colucci, M.B. Fowler, E.M. Gilbert, M. Packer. Data and Safety Monitoring Board: A.M. Katz (chair), T. Bashore, C.E. Davis, P. Kowey. Biostatistics: J. Hosking (University of North Carolina), S.T. Young (SmithKline Beecham Pharmaceuticals). Study Operations/Monitoring: N.H. Shusterman, M.A. Lukas, A. Flagg, T. Holcslaw, L.G. Parchman (SmithKline Beecham Pharmaceuticals).

Reprint requests to Michael R. Bristow, MD, PhD, Head, Division of Cardiology, University of Colorado HSC, 4200 E 9th Ave, B139, Denver, CO 80262. E-mail [email protected]

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint