Review of the U.S. Carvedilol Heart Failure Study

Carvedilol vs placebo in patients with systolic heart failure

N Engl J Med 1996;334:1349-1355

Background Before 1990, the prevailing idea held that the negative inotropy of beta-blockers would harm patients with impaired systolic function. Yet part of the progression of systolic heart failure involved over stimulation of the sympathetic nervous system. Norepinephrine can exert adverse effects on the circulation, both directly and indirectly. Smaller trials of beta-blockers in systolic heart failure found trends for benefit with beta-blockers, however, a mortality benefit had not yet been proven. The U.S. Carvedilol Heart Failure Study aimed to study mortality in patients with heart failure with a reduced ejection fraction.

Patients The study enrolled 1094 patients with chronic heart failure symptoms for at least 3 months, LVEF ≤ 0.35%, at least 2 months of treatment with diuretics and an angiotensin-converting enzyme (ACE) inhibitor (if tolerated). Treatment with digoxin, hydralazine, or nitrates was permitted but not required.

Exclusion criteria were extensive and important to understand. These included any recent major cardiac events or surgery within the previous 3 months, uncorrected valvular disease, active myocarditis, sustained VT or higher degrees of AV block not controlled by pacing, systolic blood pressure of more than 160 or less than 85 mm Hg or diastolic blood pressure of more than 100 mm Hg, clinically significant kidney or liver disease or use of calcium-channel blockers, adrenergic agonists/antagonists, or class IC/III antiarrhythmic agents. Patients receiving β-adrenergic agonists or antagonists (presumably for another indication) were not enrolled.

Baseline Characteristics The results of this and other beta-blocker trials in heart failure will be clear. One of the most important points for translating this evidence to patients will be the baseline characteristics. It is vital to understand who these patients were.

The mean age was 58 years and approximately 76% were male. Most patients had mild to moderate heart failure, with 53% in NYHA Class II, 44% in Class III, and only 3% in Class IV. The etiology of heart failure was nearly evenly split between coronary artery disease (47%) and nonischemic cardiomyopathy (53%).

Patients had significantly impaired cardiac function with a mean LVEF of 0.23. The mean six-minute walk distance ranged from 386 to 390 meters. Hemodynamic parameters were relatively stable, with mean systolic blood pressure of 116 mmHg, and mean heart rate of 83-84 beats per minute. Most patients were receiving standard heart failure therapy at baseline, including digitalis (90-91%), loop diuretics (95%), and ACE inhibitors (95%), while approximately one-third (32%) were on direct-acting vasodilators.

Trial Procedures Patients were assessed for eligibility during a 3-week screening period during which exercise capacity was assessed with a 6-minute walk test. Notable was that these were outpatients able to complete a 6-minute walk test.

Enrollment was stratified to one of four treatment protocols on the basis of the patients' performance on the exercise test: patients able to walk between 426 and 550 m when tested were assigned to the mild-heart-failure protocol; those able to walk between 150 and 425 m were assigned either to the moderate-heart-failure protocol or to a dose-ranging protocol, depending on the location of the study center; and those able to walk only less than 150 m were assigned to the severe-heart-failure protocol.

After this base-line testing, all patients received 6.25mg of carvedilol twice daily for two weeks in an open-label run-in period. Those who tolerated this initial dose were then randomized to receive either placebo (n=398) or carvedilol (n=696) on a double-blind basis, in addition to their usual medications.

The allocation ratio (carvedilol:placebo) was 2:1 in the mild and severe heart failure protocols and 1:1 in the moderate heart failure protocol. The dose was gradually increased to target levels of 25-50mg twice daily over 2-10 weeks, followed by maintenance therapy for an additional 6 months (12 months for mild heart failure).

Endpoints At the time of trial planning, the original intent was safety. That is, to show that carvedilol did not increase mortality. The original intent was to enroll 1100 patients. As smaller trials on beta-blockers were published, the statistical plan included the possibility of beta-blocker benefit. The trialists therefore planned two sided statistical analysis.

Cumulative survival curves were constructed as time-to-first-event plots by Kaplan–Meier survivorship methods and differences between the curves were tested for significance by the log-rank statistic with use of a Cox proportional-hazards regression model (which included the protocol as a covariate).

Results Median follow-up was only 6.5 months due to early termination for benefit. The patients mean total daily dose of carvedilol was 45±27 mg.

Overall mortality was 7.8% in the placebo group vs. 3.2% in carvedilol group. The relative risk reduction from carvedilol vs placebo was 65% (95% CI, 39-80%; p<0.001).

The authors also noted similar reduction in mortality across all subgroups regardless of age, sex, cause of heart failure, ejection fraction, exercise tolerance, blood pressure, or heart rate. Progressive heart failure deaths (3.3% vs. 0.7%) and sudden deaths (3.8% vs. 1.7%) were also reduced by carvedilol.

Carvedilol also improved morbidity. There was a 27% reduction in risk of hospitalization for cardiovascular causes (19.6% vs. 14.1%, p=0.036) and a 38% reduction in combined risk of hospitalization or death (24.6% vs. 15.8%, p<0.001).

Safety signals were reassuring. While heart rate decreased significantly in the carvedilol vs placebo group (12.6±12.8 vs. 1.4±12.2 beats/min, p<0.001). there were no significant changes in systolic or diastolic blood pressure. Discontinuation rates were similar: 7.8% in placebo vs. 5.7% in carvedilol group.

Conclusion This was the first trial to demonstrate that a beta-blocker could reduce mortality (and morbidity) in patients with symptomatic heart failure and reduced systolic function. Not only did it establish carvedilol as a foundational therapy in systolic heart failure, but it also strongly supported the notion that sympathetic activation was a detrimental process in progressive heart failure.

We at Cardiology Trials cannot emphasize enough the trial procedures and baseline characteristics of these patients. These were young, mostly male, ambulatory outpatients who were shown to be tolerant of carvedilol. Patients on beta-blocker for another indication were not randomized. Patients with a recent major adverse cardiac event were not randomized. What’s more, the mean daily dose of carvedilol was nearly 50mg daily.

We believe that patients in the US Carvedilol trial resemble those who would be in a chronic heart failure clinic. We caution against translating these results to older patients with multiple comorbid conditions and less robust hemodynamics.