Continuing Progress in the Treatment of Severe Congestive Heart Failure

Despite advances in its medical management, severe congestive heart failure remains associated with substantial morbidity and mortality. Prognosis has improved only slightly during the past decade among unselected community patients with congestive heart failure.¹ It appears that the benefits documented in large randomized clinical trials have not been fully realized. Two articles in this issue of THE JOURNAL describe encouraging findings regarding 2 relatively novel therapeutic approaches for congestive heart failure: β-blockers and cardiac resynchronization therapy.

During the last 5 years, randomized clinical trials have consistently demonstrated that several β-blocking agents (bisoprolol, carvedilol, and metoprolol) substantially reduce morbidity and mortality in a wide spectrum of patients with heart failure and depressed left ventricular systolic function.² Thus, current practice guidelines recommend β-blockers as first-line therapy in all patients with symptomatic left ventricular systolic dysfunction.³ β-Blocker utilization, however, is not yet widespread, especially among primary care physicians. A recent European survey showed that despite physicians' awareness of the benefits of β-blocker therapy, only 34% of a representative sample of patients with heart failure were being treated with these agents, and only 20% with angiotensin-converting enzyme (ACE) inhibitors and β-blockers combined.⁴ β-Blocker use was even lower—20%—among patients with documented left ventricular systolic dysfunction.⁴

Physician reluctance to prescribe β-blockers to patients with congestive heart failure and depressed left ventricular function is understandable. Pharmacology textbooks emphasize β-blockers' acute negative inotropic effects.⁵ Classic teaching and noncontrolled evidence (including data from run-in open-label periods in early randomized trials) suggest that withdrawal of sympathetic support in stable patients with congestive heart failure and severe left ventricular dysfunction can result in acute decompensation. Initiation of β-blockers in such patients is thought to carry a high risk of complications. The early risk-benefit ratio is thought to be unfavorable because the clinical benefit of β-blockers is assumed to be delayed and in parallel to their beneficial effects on ventricular structure and function.⁶

Findings from double-blind randomized clinical trials challenge these beliefs. In this issue of THE JOURNAL, Krum et al⁷ analyze the clinical outcomes during the initial 8-week double-blind titration phase in the Carvedilol Prospective Randomized Cumulative Survival (COPERNICUS) study, which enrolled 2289 clinically euvolemic patients with severe congestive heart failure and left ventricular ejection fraction less than 25%. The primary results of the study showed a 35% reduction in mortality with carvedilol compared with placebo after a mean follow-up of only 10.4 months.⁸ In the present analysis, patients in the carvedilol group did not experience an early increase in events (death, hospitalization, or permanent withdrawal from the study). Instead, they exhibited a trend toward early benefit for all end points of magnitude similar to that seen during the entire study. Beneficial effects of carvedilol were evident as early as 2 to 3 weeks after randomization. An early salutary effect of carvedilol was particularly apparent in an a priori defined high-risk subgroup of patients with recent or recurrent decompensation or very depressed systolic function. Although patients treated with carvedilol had more bradycardia, hypotension, edema, and dizziness, these adverse effects were generally mild and did not result in discontinuation of study medication.

These findings are in general agreement with those of the Metoprolol CR/XL Randomized Intervention Trial in Congestive Heart Failure (MERIT-HF), in which initiation of metoprolol was also well tolerated.⁹ In MERIT-HF, the curves for the end point of mortality or hospital admission started to diverge in favor of β-blockade somewhat later (after 60 days). During the first 90 days, a slightly higher proportion of patients in the New York Heart Association functional class III/IV discontinued metoprolol CR/XL (8.1%) than did those receiving placebo (5.9%), but after adjusting for multiple comparisons, the difference was not statistically significant. Whether the small differences seen in these 2 trials result from chance, different inclusion criteria and titration schedules, or some additional salutary effect due to the ancillary properties of carvedilol is not clear.

Both trials included careful standardized dose titration, frequent patient visits, and care provided by experienced physicians and nurses, often in the setting of dedicated heart failure clinics. Delivery of similar high-quality care by educated and committed practitioners in the community is feasible.¹⁰ The message is that fear of early decompensation should not deter from careful initiation of β-blockers in patients with severe heart failure. Even milder adverse effects during long-term β-blockade are not as prevalent as common wisdom suggests. For example, a recently published meta-analysis of 15 randomized trials of β-blockers in 35 000 patients with either hypertension, postmyocardial infarction, or congestive heart failure showed no significant increased risk of depressive symptoms and only small increased risks of fatigue and sexual dysfunction.¹¹

A wide QRS complex, most often with a pattern of left bundle-branch block, is common in patients with ischemic or idiopathic dilated cardiomyopathy, and predicts a worse prognosis.¹² Delayed electrical activation of the left free wall results in mechanical dyssynchrony and negatively effects pumping function. Cardiac resynchronization therapy counteracts this by retiming ventricular activation.¹³ Clinically, resynchronization is achieved by implantable devices capable of stimulating not only the right ventricle (as in conventional cardiac pacing) but also the left ventricle, most often via a special lead inserted transvenously in a coronary vein. Initial randomized studies showed a striking early improvement in functional capacity and long-term favorable changes in cardiac structure and function during resynchronization therapy.^{14 - 15} These effects were seen in patients already receiving optimized medical therapy. A beneficial effect on hard end points of mortality and morbidity, however, was not always documented, because of low statistical power resulting from small sample sizes. Despite the observed clinical benefits, many have considered that demonstration of a survival advantage was crucial before an invasive and expensive therapy such as cardiac resynchronization could be widely recommended.¹⁶ Unfortunately, industry-sponsored, large, conclusive mortality randomized trials of implantable cardiac devices are a difficult endeavor because of their expense and the potential for rapid technology obsolescence.¹⁷

In this context, a carefully performed meta-analysis can provide otherwise unavailable insight into the effects of a therapy. Bradley et al¹⁸ report such a meta-analysis in this issue of THE JOURNAL. A comprehensive search yielded 4 eligible controlled studies of cardiac resynchronization therapy including 1634 patients. After a brief follow-up of 3 to 6 months, cardiac resynchronization resulted in a 51% reduction in death from progressive heart failure and a 29% reduction in hospitalizations for heart failure, both statistically significant. Furthermore, there was also a statistically nonsignificant 23% reduction in total mortality. The relatively small sample size and short follow-up are most likely responsible for the lack of statistical significance in the reduction in total mortality. A deleterious effect of cardiac resynchronization in death from causes other than heart failure appears unlikely. Not included in the meta-analysis are the results of the 1600-patient Comparison of Medical Therapy, Pacing, and Defibrillation in Chronic Heart Failure (COMPANION) trial, which was very recently halted prematurely due to a statistically significant benefit of cardiac resynchronization in the primary combined end point of all-cause mortality and hospitalization.¹⁹ The final results from COMPANION should become available later this year. Their incorporation to the meta-analysis (ie, incremental meta-analysis) should further clarify the impact of resynchronization in overall mortality.

Of the 4 studies included in the meta-analysis, only 2 have been published in the peer-reviewed literature. The other 2 have been presented at national meetings and, more important, have made their detailed results publicly available on the US Food and Drug Administration (FDA) Web site as part of the device manufacturers' submission for premarket approval. Approval of new technologies demands intense clinical and statistical scrutiny of the supporting data at the FDA level. This fact highlights the current multiplication of legitimate sources of medical information.

Congestive heart failure will continue to be an increasing public health problem for the foreseeable future: the lifetime risk of developing heart failure has been recently estimated at 1 in 5.²⁰ Although therapeutic progress has accelerated in the last few years, its implementation in the community has been slower. Both β-blockers and cardiac resynchronization improve well-being and prolong life in patients with severe congestive heart failure due to left ventricular systolic dysfunction. Hopefully, dissemination of the findings of the 2 studies in this issue will lead to more appropriate use of these therapies. β-Blockade is now a mature therapy and should be prescribed to every patient without a strong contraindication. Cardiac resynchronization is still in its infancy and has a smaller target population. Ongoing and future research will focus on more refined identification of candidates for resynchronization, optimization of therapy delivery (including improved leads and more versatile devices), and the importance of back-up defibrillation capability.^{13 ,16}