Author
Wilson S Colucci, MD
Section Editor
Stephen S Gottlieb, MD
Deputy Editor
Susan B Yeon, MD, JD, FACC



INTRODUCTION — In patients with heart failure (HF) due to systolic dysfunction, clinical trials have demonstrated that beta blocker use slows HF progression and improves survival. The data supporting the efficacy of beta blockers in patients with HF due to systolic dysfunction are presented here. Beta blockers are started at low doses, and titrated up to effective doses.

The clinical use of beta blockers, an overview of medical therapy in patients with heart failure due to systolic dysfunction, and the possible role of beta blockers in asymptomatic left ventricular (LV) dysfunction and in patients with diastolic dysfunction are discussed separately.. (See "Use of beta blockers in heart failure due to systolic dysfunction" and "Overview of the therapy of heart failure due to systolic dysfunction" and "Evaluation and management of asymptomatic left ventricular systolic dysfunction" and "Treatment and prognosis of diastolic heart failure".)

RATIONALE FOR BETA BLOCKER THERAPY — Symptomatic HF results in activation of neurohumoral mechanisms, including the sympathetic nervous system [1]. An elevated plasma norepinephrine concentration is a marker for poor survival in these patients [2]. (See "Pathophysiology of heart failure: Neurohumoral adaptations" and "Predictors of survival in heart failure due to systolic dysfunction", section on Neurohumoral activation.)

The increase in sympathetic activity may have deleterious effects and a number of trials have shown that blockade of beta adrenergic receptors leads to symptomatic improvement and enhanced survival in many patients with HF [3,4]. As a result of this compelling evidence, beta blockers are now an important component of standard therapy in HF. This is an important shift, since beta blockers were previously considered contraindicated in patients with HF because of their negative inotropic activity.

Possible mechanisms of benefit — It seems likely that a number of factors contribute to the beneficial effect of beta blockers in HF [5,6].

* Long-term exposure to catecholamines is directly detrimental to the myocardium in both animals and humans [7-9]. The effect of catecholamines on the heart and the likelihood of developing HF may be amplified by gene polymorphisms in adrenergic receptors that enhance cardiac sympathetic activity. In a report in which 159 patients with HF were compared to 189 controls, blacks who were homozygous for a particular alpha-2C receptor polymorphism had a marked increase in risk of HF (adjusted odds ratio 5.7); the risk was further increased in blacks who were also homozygous for a beta-1 receptor polymorphism (odds ratio 10.1) [10].
* Chronic stimulation of beta receptors reduces the responsiveness to beta adrenergic agonists due to downregulation and desensitization of the beta receptor and its coupled signaling pathways [1,11-13]. Beta blockade upregulates myocardial beta-1-receptor density in patients with HF [11], helping to restore the inotropic and chronotropic responsiveness of the myocardium, thereby improving contractile function.
* Beta blockers reduce the circulating level of vasoconstrictors, including norepinephrine [14,15], renin [16,17], and endothelin [18]. Vasoconstriction induced by these hormones increases afterload, thereby promoting the rate of progression of cardiac dysfunction.
* Beta blockade may reduce myocardial gene production of some of the inflammatory cytokines that occurs during the development of HF [19-21]. (See "Nitric oxide, other hormones, cytokines, and chemokines in heart failure".)

* Beta blockers have a beneficial effect on LV remodeling and can decrease LV end-systolic and end-diastolic volume [22-24]. The improvement in LV geometry can diminish functional mitral regurgitation [25,26]. (See "Cardiac remodeling: Basic aspects" and "Functional mitral regurgitation".)

* In ischemic cardiomyopathy, beta blockers may improve function in regions of hibernating myocardium by reducing myocardial oxygen consumption and increasing diastolic perfusion [27]. (See "Clinical syndromes of stunned or hibernating myocardium".)

* Functional improvement with beta blockers is associated with normalization in the expression of a number of myocardial genes [28]. The changes that occur would be expected to enhance contractility and reduce pathologic hypertrophy, but it is not clear if they were responsible for or a result of the clinical improvement.

In addition to possible hemodynamic benefits, beta blockers also decrease the frequency of ventricular premature beats and the incidence of sudden cardiac death (SCD), especially after a myocardial infarction. (See "Beta blockers in the management of acute coronary syndrome".)

EFFECTS OF BETA BLOCKADE IN HF — Early physiology studies demonstrated that beta blockade results in volume expansion [29-31]. For these and other reasons, beta blockers were long considered at least relatively contraindicated in patients with HF. However, evidence has now established that cautious use of beta blockers, initiated in stable patients, results in important benefits, including reduced mortality.

What follows is a brief review of the studies that evaluated the effects of beta blockade on hemodynamics, LVEF, and exercise capacity in patients with HF. This will be followed by a discussion of the major trials that have evaluated the efficacy of beta blockers on patient survival [6].

Ventricular performance — The first observation suggesting that beta blockade might be beneficial in HF was made in 1975 [32]. The administration of beta blocker to seven patients with congestive cardiomyopathy resulted in improvements in LVEF and overall clinical status.

Subsequent reports consistently confirmed these findings, demonstrating that over a three to six month period of follow-up, beta blockers led to improvements in LVEF and resting hemodynamics [14,17,23,33-37]. One study, for example, performed a double-blind, placebo-controlled comparison of bucindolol plus standard therapy versus placebo plus standard therapy in 24 patients with HF due to idiopathic dilated cardiomyopathy [14]. Bucindolol (12.5 mg twice daily), a nonselective beta blocker with direct vasodilatory activity, was given acutely and was tolerated in 23 of the 24 patients. Long-term therapy resulted in improvements in LVEF, cardiac index, and LV stroke work, while mean pulmonary capillary wedge pressure (PCWP) and heart rate decreased. NYHA functional class also improved in the bucindolol group (p<0.01).

Similar improvements have been demonstrated with metoprolol and bisoprolol, which are beta-1 selective agents [23,34], and carvedilol, a nonselective beta blocker that also blocks the alpha-1 receptor [35-37]. The improvement in LVEF (graph 1) and the reduction in symptoms observed with carvedilol occurred in patients with mild (class II) and more severe disease (class III and IV); dose-dependence was more prominent in patients with idiopathic dilated cardiomyopathy (graph 2) [38].

Improvement in contractile function with beta blocker therapy occurs in regions of dysfunctional but viable myocardium, but not in regions of extensive scarring. This is illustrated by the following observations:

* In one report, 45 patients with cardiomyopathy (28 ischemic and 17 nonischemic) underwent cardiovascular magnetic resonance imaging (CMR) with gadolinium enhancement, followed by six months of beta blocker therapy [39]. Contractility improved in 56 percent of regions with no scarring but in only 3 percent of regions with more than 75 percent scarring. (See "Clinical utility of cardiovascular magnetic resonance imaging", section on 'Infarct detection and sizing'.)

* In another study, 41 patients in the BEST trial with NYHA class III to IV HF underwent dobutamine stress echocardiography before treatment with bucindolol [40]. There was a significant correlation between contractile reserve assessed by dobutamine echo and the improvement in LVEF after three months of therapy.

Exercise capacity — Several small studies have demonstrated an increase in maximum exercise duration following the administration of a beta blocker in HF due to either idiopathic dilated cardiomyopathy or ischemic heart disease [14,41,42]. This improvement in exercise duration may be associated with a reduction in maximal VO2 (oxygen utilization), an effect that is probably due in part to an attenuation in the heart rate response to exercise. In addition, the nonselective beta blockers propranolol and carvedilol do not prevent patients from deriving benefit from exercise training [43].

Cardiac remodeling — Chronic beta blocker therapy can have a favorable impact on cardiac remodeling. This issue is discussed in detail separately. (See "Cardiac remodeling: Clinical assessment and therapy", section on 'Beta blockers'.)

Reduced risk of atrial fibrillation — There is evidence that beta blockers may also reduce the likelihood of the development of AF in patients with HF due to systolic dysfunction. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Clinical trials", section on 'Beta blockers'.)

CLINICAL TRIALS — After initial observations suggesting a survival benefit with beta blockers in HF [44], a number of major trials confirmed this benefit with several beta blockers (sustained release metoprolol succinate, carvedilol, and bisoprolol) in selected patients with HF. Before reviewing the data from the specific trials, the range of benefit can be illustrated by findings from a meta-analysis that included 22 trials involving more than 10,000 patients, almost all of whom had NYHA class II or III HF and were also treated with ACE inhibitors [3]. The following findings were noted:

* Beta blockers significantly reduced mortality at one year (odds ratio 0.65, 95% CI 0.53-0.80) and two years (odds ratio 0.72, 95% CI 0.61-0.84). Assuming a mortality rate of 12 percent in the placebo group at one year (derived from the three largest and most recent trials), beta blocker therapy saved 3.8 lives in the first year per 100 patients treated.
* Beta blockers also reduced hospitalization for HF (odds ratio 0.64, 95% CI 0.53-0.79) with an absolute benefit of four fewer hospitalizations in the first year per 100 patients treated [3].

Similar findings were noted in a meta-analysis limited to the large randomized trials described below [4].

Metoprolol trials — The efficacy of metoprolol has been evaluated in several randomized trials, including MDC, MERIT-HF, and RESOLVD.

MDC trial — The Metoprolol in Dilated Cardiomyopathy (MDC) trial randomly assigned 383 patients to placebo or metoprolol tartrate (beginning at a dose of 10 mg and increasing slowly to a maximum of 150 mg/day) [45]. Eligible patients were required to have dilated cardiomyopathy, symptomatic HF, an LVEF <40 percent, and a systolic pressure above 90 mmHg. Potential participants were excluded if they had significant coronary disease, active myocarditis, other life-threatening disease, or they deteriorated after a test dose of 5 mg of metoprolol twice daily for two to seven days. Digitalis, diuretics, and angiotensin converting enzyme (ACE) inhibitors were used as necessary.

Of the patients who entered the study, the following significant benefits were noted in the metoprolol group when compared to therapy with placebo at 12 to 18 months:

* A reduced likelihood of progressing to cardiac transplantation (1 versus 10 percent)
* A greater increase in LVEF (12 versus 6 percent) and exercise tolerance
* A subjective improvement in the quality of life

There was no significant difference in mortality — 12 versus 10 percent; however, the use of cardiac transplantation in patients who otherwise might have died may have prevented an improvement in mortality from being observed.

MERIT-HF trial — In comparison to the nonsignificant benefit in the MDC trial, a mortality benefit with sustained release metoprolol succinate was demonstrated in the much larger MERIT-HF trial [46,47]. In the MERIT-HF trial, 3991 patients with class II to IV HF and an LVEF ≤40 percent who were receiving digoxin, an ACE inhibitor, and a diuretic, were randomly assigned to therapy with extended-release metoprolol, beginning with 12.5 or 25 mg daily and titrated up to 200 mg/day, or placebo [46,47]. The mean dose was 159 mg/day, with 64 percent of patients receiving the target dose; the discontinuation rate for patients taking active drug was 14 percent at one year. The study was prematurely terminated when the following significant benefits were noted in the metoprolol group:

* A 34 percent relative risk reduction in all-cause mortality at 12 months (7.2 versus 11 percent for placebo) (graph 3); there was also a reduction in the combined end point of death or need for transplant (7.5 versus 10.3 percent)
* A reduction in hospitalization for cardiovascular causes (20 versus 25 percent) or for HF (10 versus 15 percent) (graph 4)

* Improved NYHA functional class (table 1) and quality of life measures

When analyzed by mode of death, there were significantly fewer SCDs (3.9 versus 6.6 percent) and fewer deaths from worsening of HF (1.5 versus 2.9 percent) in the metoprolol group. The proportion of SCDs decreased and those due to HF increased with increasing severity of HF.

Carvedilol trials — The United States Carvedilol Heart Failure Program, which consisted of four component trials including MOCHA and PRECISE, was initially designed to evaluate nonfatal end points in patients with mild to moderate HF [35,36,38,48]. Mortality, which was not a prospectively designated primary end point, was also measured to assess safety and potential benefit. The results were combined to form a double-blind controlled study of almost 1100 patients with NYHA class II to III HF (table 1) [48]. Patients were maintained on digoxin, diuretics, and ACE inhibitors and then randomly assigned to treatment with carvedilol or placebo.

Marked improvement in mortality in patients treated with carvedilol (initial dose 6.25 mg twice daily and then gradually increased to a maximum of 25 mg twice daily) led to early termination of the study after 25 months of enrollment:

* There was a significant decrease in total mortality (3.2 versus 7.8 percent) (graph 5). This benefit was due to reductions in death due to progressive HF or SCD; it was independent of age, sex, HF etiology, LVEF, exercise tolerance, systolic blood pressure, or heart rate.
* There were also significant reductions in the need for hospitalization (14.1 versus 19.6 percent) and the number of intensive care unit days and length of hospital stay [49], and a 38 percent increase in event-free survival (15.8 versus 24.6 percent) (graph 6).

These benefits translated into an estimated reduction of in-patient costs for cardiovascular and HF admissions by 57 and 81 percent, respectively [49]. The most common adverse reaction was worsening HF, which was less frequent with carvedilol than with placebo (1.6 versus 2.3 percent).

Despite these apparent benefits, this study was criticized for the following reasons [50]:

* Very short average follow-up (6.5 months)
* Lack of mortality as a prospective primary end point
* A six week run-in period to demonstrate patient tolerability to open-label carvedilol prior to randomization

Carvedilol is also useful in patients with more severe HF (symptoms with minimal exertion at rest) as demonstrated in the COPERNICUS trial described below [51,52] and in patients with a dilated cardiomyopathy and HF who are on chronic hemodialysis [53]. (See 'Class IV HF' below and "Myocardial dysfunction in end-stage renal disease".)

Bisoprolol (CIBIS and CIBIS II trials) — After demonstration that bisoprolol had both hemodynamic benefits and a trend toward improved survival in 641 patients in CIBIS [34,54], the efficacy of bisoprolol was more fully evaluated in the much larger CIBIS II trial [55,56]. CIBIS II randomly assigned 2647 patients with class III or IV HF and an LVEF <35 percent to bisoprolol or placebo; the patients also received standard therapy with diuretics and ACE inhibitors. After an average follow-up of 1.4 years, the trial was prematurely terminated when the following benefits were observed in the bisoprolol group:

* A significant reduction in total all-cause mortality (11.8 versus 17.3 percent) that was independent of the severity or cause of HF (graph 7). This benefit was primarily due to a reduction in SCD (3.6 versus 6.3 percent, p<0.001), with a nonsignificant trend toward fewer deaths from HF.
* A significant 15 percent reduction in hospital admissions for any cause and a 30 percent reduction in admissions for HF (p<0.0001). These benefits resulted in a 5 to 10 percent reduction in the cost of care for these patients [57].
* Although mortality and hospitalization for HF were significantly related to a higher baseline heart rate, the improvement in outcome with bisoprolol was similar at any level of resting heart rate or amount of change in heart rate at two months [58]. However, bisoprolol had no effect on outcome in patients in atrial fibrillation. (See "Atrial fibrillation in heart failure and cardiomyopathy".)

Bucindolol (BEST trial) — The BEST trial evaluated the role of bucindolol (3 mg twice daily titrated up to 100 mg twice daily) in 2708 patients with class III (92 percent of the patients) or IV HF [59]. After a mean follow-up of two years, the study was prematurely terminated. At the time of termination, bucindolol did not reduce overall mortality compared to placebo (30 versus 33 percent). It did reduce cardiovascular mortality by 14 percent, but there was no difference in HF death, sudden death, death from myocardial infarction, or noncardiovascular mortality.

On prespecified subgroup analysis, there was a significant survival benefit with bucindolol in nonblack patients. There was a trend toward improved survival in patients with class III HF, an LVEF above 20 percent, and nonischemic cardiomyopathy. There was no survival benefit with bucindolol in patients with an LVEF equal to or less than 20 percent or in blacks (see 'Use in blacks' below.

Possible explanations for the lack of overall mortality benefit with bucindolol include differences in pharmacologic properties of bucindolol as compared to other studied beta blockers or differences in studied patients populations (such as racial composition). Although bucindolol is similar to carvedilol as a nonselective beta-blocking agent with weak alpha 1 blocking properties, unlike carvedilol it has inverse agonist and intrinsic sympathomimetic activity [60,61].

Comparison of beta blockers

Comparison between metoprolol and carvedilol — As described above, both sustained release metoprolol succinate and carvedilol reduce mortality in patients with HF. These two drugs have different pharmacologic properties: metoprolol has a high degree of specificity for the beta-1 adrenergic receptor, while carvedilol acutely blocks beta-1, beta-2, and alpha-1 adrenergic receptors. Whether the broader adrenergic blocking effects of carvedilol confer an additional advantage over beta-1 selective agents is not known [62]. The alpha-1 adrenergic blocking effect of carvedilol may not be significant in patients with HF on chronic beta blocker therapy [63,64].

The effects of carvedilol and short-acting metoprolol (metoprolol tartrate) were directly compared in the COMET trial; carvedilol was more effective than metoprolol in reducing mortality [65] and vascular events [66]. This study enrolled 3029 patients with class II to IV HF, at least one admission for a cardiovascular diagnosis in the preceding two years, and an LVEF ≤35 percent. Subjects were randomly assigned to either carvedilol at a target dose of 25 mg twice daily or metoprolol tartrate at a target dose of 50 mg twice daily.

At a mean follow-up of five years, the carvedilol arm had a significantly lower rate of all-cause mortality (34 versus 40 percent for metoprolol) that was entirely due to a reduction in cardiovascular mortality (29 versus 35 percent). Extrapolation of the survival curves gave a significantly longer estimate of median survival for the carvedilol group (8.0 versus 6.6 years).

A potentially important limitation to these findings is that the degree of beta blockade may not have been equivalent in the two arms of COMET [67].

* The carvedilol arm had significantly greater reductions in heart rate (13.3 versus 11.7 beats per minute with metoprolol) and systolic blood pressure (3.8 versus 2.0 mmHg). These differences are consistent with carvedilol being given in a dose similar to that in the Carvedilol Trials, while the metoprolol goal dose was substantially lower than in the MDC and MERIT-HF trials (100 versus 150 and 200 mg/day) [45-48].
* Another potential concern is that short-acting metoprolol (metoprolol tartrate) was used in COMET, whereas the extended-release form (metoprolol succinate) was used in MERIT-HF.

Data directly comparing outcomes on carvedilol versus extended release metoprolol (metoprolol succinate) are not available. A meta-analysis of 15 placebo controlled trials (nine using carvedilol and six using metoprolol of which four used the short-acting form and two used extended-release drug), suggested that carvedilol may produce a greater improvement in LVEF (placebo corrected increases of 6.5 versus 3.8 percent for the placebo-controlled trials) [68]. A similar difference in LVEF change was noted upon analysis of four trials directly comparing carvedilol and short-acting metoprolol. However, the relevance of differences in LVEF is uncertain. LVEF may be affected by afterload so its improvement does not prove an increase in contractility.

A clinically important difference among beta blockers may be their effects on blood pressure. Patients with low blood pressure may be less likely to tolerate carvedilol because of its vasodilator activity. Conversely, those with higher blood pressure may have a greater lowering of blood pressure with carvedilol. In MERIT-HF, metoprolol resulted in a higher blood pressure than placebo, presumably because of improved cardiac function.

Comparison with other beta blockers — Limited data are available comparing other beta blockers with carvedilol or metoprolol for treatment of HF.

A meta-analysis compared the effects of vasodilating beta blockers, primarily carvedilol, with non-vasodilating beta blocker (largely bisoprolol) using indirect evidence provided by trials comparing beta blocker to placebo [69]. Vasodilating beta blockers produced a greater survival benefit than nonvasodilating agents (45 versus 27 percent). This difference was primarily seen in patients with nonischemic cardiomyopathy.

The beta blocker bucindolol reduced all-cause mortality in whites but not blacks, while other beta blockers appear to be equally effective in whites and blacks (see 'Use in blacks' below.

Observational studies suggest that the beneficial effects of beta blockers on mortality in patients with HF may not be limited to those beta blockers with proven efficacy in randomized trials. However, such retrospective studies are difficult to interpret.

* Among 11,326 privately insured adults hospitalized for HF, 70 percent received beta blockers [70]. The mortality at one year varied with beta blocker use (atenolol, 20.1; metoprolol tartrate, 22.8; carvedilol,17.7; and no beta blockers, 37 percent). After adjustment for confounders and the propensity to receive carvedilol, the risk of death at one year compared with atenolol was higher for metoprolol tartrate and no beta blockers but was not significantly different for carvedilol.
* Among 11,959 Medicare/Medicaid beneficiaries hospitalized with HF, 59 percent received no beta blockers, 23 percent received beta blockers with randomized controlled trial evidence of efficacy in treatment of HF (evidence based beta blocker [EBBB]: carvedilol, metoprolol succinate or bisoprolol), and 18 percent received other beta blockers (non-EBBB) [71]. Patients receiving no beta blockers had significant higher adjusted one year mortality than patients receiving either non-EBBB or EBBB (28.3 versus 22.8 and 24.2 percent). There was no difference in adjusted mortality between patients receiving non-EBBBs and EBBBs. Patients receiving EBBBs had a higher adjusted rehospitalization rate than patients not receiving beta blockers or patients taking non-EBBBs.

Taken together, these retrospective data suggest a possible beneficial effect of non-EBBB, such as atenolol, but do NOT support a benefit for short-acting metoprolol, a conclusion that is also consistent with the effects of short-acting metoprolol in the COMET trial (see 'Comparison between metoprolol and carvedilol' above. However, important limitations of these studies include lack of information on HF severity, left ventricular ejection fraction, and indications for beta blocker use (treatment of hypertension or HF). Therefore, these studies do not provide a sufficient basis for recommendation of non-EBBB for HF treatment.

Use with other HF drugs — Only three other drugs, ACE inhibitors, angiotensin II receptor blockers (ARBs), and aldosterone antagonists, also improve survival in patients with HF. The benefit from all of these drugs appears to be additive to that of beta blockers. (See "Overview of the therapy of heart failure due to systolic dysfunction".)

ACE inhibitors — Most patients in the major trials evaluating metoprolol [45-47], carvedilol [48], and bisoprolol [55] were also treated with an ACE inhibitor (as well as diuretics and digoxin as necessary). Combination therapy with an ACE inhibitor and beta blocker is more effective than an ACE inhibitor or beta blocker alone (graph 1 and graph 2) [72].

ACE gene polymorphisms may affect the response to combination therapy. The DD genotype of the ACE gene has been associated with increases in ACE activity and mortality and a reduction in transplant-free survival in patients with HF [73]. This difference may be abolished with beta blocker therapy as transplant-free survival is equivalent in patients with the DD, ID, and II genotypes. (See "Actions of angiotensin II on the heart", section on 'ACE gene polymorphism'.)

ARBs — Combination therapy with an ARB and a beta blocker also appears to be more effective than either agent alone. In the CHARM-Alternative trial of patients with HF who were treated with candesartan because they could not tolerate an ACE inhibitor (most often due to cough), a reduction in cardiovascular death or hospitalization for HF was seen whether or not the patient was taking a beta blocker [74]. (See "Angiotensin II receptor blockers in heart failure due to systolic dysfunction: Therapeutic use".)

There are conflicting data on the value of adding an ARB to patients already treated with an ACE inhibitor and beta blocker. In the CHARM-ADDED trial, outcomes were improved with the addition of candesartan, a benefit that was seen regardless of whether or not patients were also treated with a beta blocker [75]. In contrast, adding an ARB to an ACE inhibitor in patients treated with a beta blocker appeared to be associated with increased mortality in a post hoc subgroup analysis from the Val-HeFT trial [76].

We regard the CHARM-Added findings as more definitive and suggest the addition of an ARB, if tolerated to a regimen including ACE inhibitors and beta blockers, in patients with persistent class II to III HF if the patient has normal renal function and low to normal serum potassium. (See "Angiotensin II receptor blockers in heart failure due to systolic dysfunction: Therapeutic use".)

Aldosterone antagonists — Two major trials have demonstrated benefit from aldosterone antagonists in patients with HF; almost all patients in both trials were treated with an ACE inhibitor or ARB. The RALES trial evaluated spironolactone in patients with recent or current class IV HF [77]. Although the benefit from spironolactone was similar in all subgroups, only 10 percent were taking a beta blocker.

More convincing evidence for benefit in association with beta blockers comes from EPHESUS in which 75 percent were taking a beta blocker [78]. The benefit from eplerenone was best seen in patients treated with both a beta blocker and an ACE inhibitor or ARB.

Similarly, patients taking an aldosterone antagonist appear to benefit from a beta blocker. This was suggested in a post-hoc analysis from the COPERNICUS trial in which the benefits of carvedilol (reduced mortality and HF admissions) were similar in the subsets of patients who were (19 percent) and were not (81 percent) taking spironolactone [79]. (See "Use of diuretics in heart failure".)

Use with aspirin — Concern has been raised that aspirin (ASA) may offset some of the benefit of ACE inhibitors in patients with HF, although most of the evidence does not support the presence of such an interaction. (See "ACE inhibitors in heart failure due to systolic dysfunction: Therapeutic use", section on 'Use with aspirin'.)

Whether there is an interaction between ASA and beta blockers was addressed in a retrospective analysis of 293 patients enrolled in a randomized, placebo-controlled trial of carvedilol [80]. Among patients receiving carvedilol, those using ASA had less improvement in LVEF compared to ASA nonusers (6 versus 9.5 EF units). This effect was dose related; for each 81 mg/day increase in dose, there was a decrease of 0.5 EF units. The mechanism for this possible interaction is unknown.

Given the demonstrated value of ASA and the confounding features of a retrospective analysis, it would be premature to modify recommendations for use of ASA until further data are available. (See "Drugs that should be avoided or used with caution in patients with heart failure", section on 'Aspirin'.)

DEGREE AND DURATION OF BETA BLOCKADE

Degree of beta blockade — Individualized dose titration leading to effective beta blockade was supported by a post-hoc subgroup analysis from the MERIT-HF trial in which the patients were divided into a high-dose (mean metoprolol dose 192 mg/day) and low-dose group (mean 76 mg/day); a low heart rate was the most common cause for dose-limitation [81]. The two groups had similar heart rates at baseline (83 and 81 beats/min) and during therapy (67 beats/min); the similar fall in heart rate in the low-dose group suggested increased sensitivity to beta blocker therapy. The two groups, compared to placebo, had a similar reduction in mortality (38 percent) and other cardiovascular end points.

Aiming for a particular resting heart rate or a particular reduction in heart rate is not of proven value [58,82]. This was illustrated in a post-hoc subgroup analysis from the MERIT-HF trial [82]. There was no relationship between the benefit from controlled release/extended release metoprolol and the baseline heart rate, achieved heart rate, or change in heart rate.

Duration of therapy — Although data about the duration of beta blocker therapy in HF are lacking, the available limited data favor continuation of beta blocker therapy in patients with left ventricular systolic dysfunction and history of HF [83-85]. This approach is supported by a report of 13 patients with a dilated cardiomyopathy who were successfully managed with metoprolol for ≥30 months [85]. Within four months of tapering the dose and then discontinuing the drug, SCD or death from progressive HF occurred in four and clinical deterioration in three.

SPECIFIC PATIENT SUBSETS

General approach — Less evidence is available to support beta blocker use in specific patients subsets. As recommended in the 2005 ACC/AHA guidelines, in the absence of specific evidence to the contrary subgroups should receive therapy identical to that applied to the broader population [86].

Use in ischemic and nonischemic cardiomyopathy — Clinical trials have demonstrated that improvements in exercise duration, stabilization of LV function, and mortality are similar in ischemic and nonischemic cardiomyopathies [48,69,87].

Class IV HF — The benefit of beta blockade appears to extend to patients with severe class III and stable class IV HF (table 1). In a subgroup analysis of 795 patients with severe HF (NYHA functional class III/IV) in the MERIT-HF trial, the benefit of metoprolol was similar to that seen in the entire study population [88]. Metoprolol reduced total mortality (11.3 versus 18.2 percent for placebo), SCD (5.5 versus 9.8 percent), death from HF (3.3 versus 7 percent), and the number of hospitalizations for HF (15 versus 26 percent).

Similar results were seen with carvedilol in the COPERNICUS trial, which specifically assessed the efficacy of beta blockade in 2289 patients with severe class III or stable class IV HF (symptoms with minimal exertion or at rest) and an LVEF less than 25 percent [51]. The trial was terminated one year prematurely (mean follow-up 10.4 months) because of a significant mortality benefit from carvedilol compared to placebo (annual mortality rate 11.4 versus 18.5 percent) (graph 8). Carvedilol was associated with an equivalent reduction in the total number of hospitalizations, hospitalizations for cardiovascular reasons, and those for HF. These benefits were seen in all subgroups, not being affected by age, LVEF, or the cause of the cardiomyopathy (graph 9). The patients treated with carvedilol also spent fewer days in the hospital per admission and were less likely to develop serious adverse events such as sudden death, ventricular tachycardia, or cardiogenic shock [52].

Based upon the results from COPERNICUS, the FDA has approved the use of carvedilol in patients with severe HF.

These benefits were confirmed in a pooled analysis of 3836 patients with NYHA functional class III/IV HF and an LVEF ≤25 percent enrolled in COPERNICUS, MERIT-HF, and CIBIS II [88]. Beta blockade was associated with a significant reduction in total mortality (13 versus 18 percent with placebo, relative risk 0.72) and a 45 percent reduction in the number of hospitalizations. However, initial worsening of symptoms may be more common in patients with severe disease [89]. (See "Use of beta blockers in heart failure due to systolic dysfunction", section on 'Worsening of HF symptoms'.)

Combination with PD inhibitor — A potential approach to the therapy of severe HF, particularly in patients who cannot tolerate a beta blocker alone, is combination therapy with a phosphodiesterase (PD) inhibitor (a positive inotrope). It has been proposed that the beta blocker would prevent the deleterious effects on survival associated with long-term treatment with a PD inhibitor, while the PD inhibitor would permit the successful initiation and upward titration of beta blockade [90]. This hypothesis was not supported by the ESSENTIAL and EMPOWER studies which have been presented, but not published. (See "Inotropic agents in heart failure due to systolic dysfunction".)

Combination with dobutamine — Patients with advanced HF who decompensate may be treated with dobutamine. In this setting, it may matter which beta blocker has been used. In a randomized, short-term study, pretreatment with carvedilol markedly reduced the inotropic response to dobutamine, whereas metoprolol had only a slight inhibitory effect [91]. Metoprolol is a relatively selective beta-1-receptor blocker that results in beta-1 receptor upregulation and does not inhibit the dobutamine effect on the beta-2 receptors. In contrast, carvedilol does not cause upregulation of the beta-1 receptors and blocks the beta-2 receptors, inhibiting both inotropic mechanisms of dobutamine. In contrast, the hemodynamic responses to enoximone were maintained or enhanced by both drugs. (See "Inotropic agents in heart failure due to systolic dysfunction".)

Use in women — Women have been underrepresented in the beta blocker trials, but appear to benefit to the same degree as men. A post hoc analysis of 898 women enrolled in the MERIT-HF trial found that the beneficial effects of metoprolol extended to women [92]. Treatment with metoprolol reduced all-cause mortality or hospitalizations by 21 percent, cardiovascular hospitalizations by 29 percent, and hospitalizations for worsening of HF by 42 percent.

A pooled analysis from MERIT-HF, COPERNICUS, CIBIS II, and the United States Carvedilol Heart Failure trials found that the mortality benefit from beta blocker therapy was the same in men and women (relative risk 0.66 and 0.63, respectively) [93].

Use in blacks — There are conflicting data on the efficacy of beta blockers in black patients. In the carvedilol trials, the benefit of beta blockade was of similar magnitude in blacks and nonblack patients [94]. In comparison, it appeared that blacks derived no benefit with bucindolol in the BEST trial [59].

A meta-analysis of beta blocker trials in HF confirmed this distinction, finding different results depending upon whether or not the BEST data were included [93]. In COPERNICUS, MERIT-HF, and United States Carvedilol Heart Failure trials, the reduction in all-cause mortality with beta blockers was the same for blacks and whites (relative risk 0.67 and 0.63 respectively). With inclusion of the data from BEST, the benefit of beta blockers remained significant for whites but was no longer significant in blacks (relative risk 0.69 and 0.97, respectively).

These observations suggest that bucindolol, a beta blocker with inverse agonism and partial beta agonist activity [61], is not effective in reducing mortality in blacks. The reasons for this difference are not clear, but (as speculated by the authors of BEST) may include race-specific differences in the beta adrenergic pathway, although the difference could be due to chance or socioeconomic differences.

A potential genetic explanation for variation in beta blocker therapy response was provided by a report of genetic variants of G protein-coupled receptor kinases (GRKs) [95]. GRKs desensitize beta adrenergic receptors and may therefore modify the effect of beta blocker therapy. (See "Peptide hormone signal transduction and regulation", section on 'Desensitization',A GRK5-Leu41 polymorphism was identified as frequent (about 40 percent) among black subjects and rare (about 2 percent) among white subjects [95]. GRK5-Leu41 uncoupled isoproterenol-stimulated responses more effectively than GRK5-Gln41 (the most common variant) in transfected cells and transgenic mice and protected against experimental catecholamine-induced cardiomyopathy.

Human association studies showed that the presence of GRK5-Leu41 polymorphism was associated with a lower rate of mortality or cardiac transplantation in 375 black patients with HF or cardiac ischemia. These findings suggest that GRK5-Leu41 may produce a beneficial "genetic beta blockade" and may explain the conflicting results of beta blocker trials in blacks [95].

Use in diabetes — The above meta-analysis of beta blocker trials in HF included 1883 diabetics and 7042 nondiabetics [93]. The survival benefit with beta blocker therapy was significant for both those with diabetes and for those without (relative risk 0.77 and 0.65, respectively) [93]. The difference in risk reduction between diabetics and nondiabetics was not significant. (See "Heart failure in diabetes mellitus", section on Drug therapy.)

Use in patients with COPD — Beta-1 selective beta blockers (eg, atenolol or metoprolol) appear to be safe in patients with COPD, even when there is a bronchospastic component. In a meta-analysis of major trials, no changes in FEV1, respiratory symptoms, or the use of inhaled beta agonists were seen [96,97]. More limited data suggest that combined alpha and beta blockade (eg, carvedilol) may be safe in the setting of COPD as well, but less well tolerated in patients with asthma [98].

Use in elderly adults — All of the major randomized trials included many elderly patients who appeared to derive similar benefit as younger patients. In MERIT-HF, for example, the one-third of patients ≥70 years of age had the same benefit from metoprolol as younger patients [46]. The risk reduction was also similar in patients above age 75 [99]. Similar findings were noted in the carvedilol trials; the mean age was 58 and the benefit of carvedilol was the same in patients ≥59 years of age as in younger patients [48].

The benefit of beta blockers in elderly adults was directly demonstrated in the SENIORS trial of 2128 patients ≥70 years of age (mean 76 years) [100]. Inclusion in this trial was based on a history of HF (hospital admission for HF within the previous year or known LVEF ≤35 percent). About 35 percent of patients had an LVEF >35 percent. The patients were randomly assigned to nebivolol, or placebo. Nebivolol is a beta 1 blocker, a beta 3 agonist (which attenuates contraction [101]) and a vasodilator. An ACE inhibitor was used in 82 percent, a diuretic in 86 percent, digoxin in 40 percent, and aspirin in 42 percent. At a mean follow-up of 21 months, nebivolol therapy was associated with a significant reduction in the primary end point of all-cause mortality or cardiovascular hospital admission (31.1 versus 35.3 percent, hazard ratio 0.86, 95% CI 0.74-0.99) and a trend toward lower all-cause mortality (15.8 versus 18.1 percent, hazard ratio 0.88, 95% CI 0.71-1.08). The effect of nebivolol was similar in those with preserved and impaired LVEF [102].

These benefits are smaller than those noted in other trials. Possible explanations include the drug used, a higher rate of discontinuation during the study (35 percent compared to 14 percent overall and 20 percent in patients ≥65 years of age in MERIT-HF [46,99]), the age of the patients, and the inclusion of patients with asymptomatic left ventricular dysfunction [103]. While the outcome benefits were similar in patients with an LVEF above or below 35 percent, reverse remodeling only occurred in the patients with an LVEF ≤35 percent. There was no evidence of improvement in diastolic function in those with an LVEF >35 percent [104].

It is possible that some elderly patients with comorbidities would have been excluded from these trials, potentially altering the clinical applicability of the findings. This issue was addressed in a cohort study of almost 12,000 elderly patients (mean age 79); during follow-up, 3539 were hospitalized for HF and 1162 received a beta blocker [105]. Beta blocker therapy was associated with substantial reductions in all-cause and HF mortality (adjusted hazard ratio 0.72 and 0.65); this benefit was seen in all subgroups, even those with a comorbidity score ≥2. Although lower doses were associated with improved outcomes, the benefit was greater if the dose was titrated up to that used in the clinical trials.

USE OUTSIDE OF CLINICAL TRIALS — The subjects enrolled in randomized trials represent a highly selected patient subset who tend to be healthier, younger, more compliant and more closely followed than those in clinical practice. The use of beta blockers in a nontrial setting was examined in a cohort of 1041 patients followed in a tertiary care HF clinic (median age 69 years; mean ejection fraction 33 percent; 51 percent with class III or IV HF) (table 1) [106]. The following observations were made:

* A total of 475 patients (46 percent of the study group) were prescribed beta blockers. The use of beta blockers was significantly associated with class I or II symptoms and with age younger than 69 years.
* The doses used in clinical trials were achieved in only 18 percent of the study group. The mean doses achieved were 27 mg/day for carvedilol and 81 mg/day for metoprolol tartrate.
* During 32 months of follow-up, beta blocker use was associated with significantly reduced mortality after adjusting for age, sex, ejection fraction, NYHA class, and concomitant medication use (hazard ratio 0.63).

Observational data has also been analyzed to compare outcomes among various types of beta blockers. (See 'Comparison with other beta blockers' above.)

Adherence to beta blocker use is discussed in detail separately. (See "Guideline adherence and outcomes in coronary heart disease and heart failure".)

INFORMATION FOR PATIENTS — Educational materials on this topic are available for patients. (See "Patient information: Heart failure causes, symptoms, and diagnosis" and "Patient information: Heart failure treatments".) We encourage you to print or e-mail these topic reviews, or to refer patients to our public web site, www.uptodate.com/patients, which includes these and other topics.

SUMMARY AND RECOMMENDATIONS

* Use of beta blockers in patients with HF due to systolic dysfunction, particularly carvedilol, extended release metoprolol succinate, and bisoprolol, improves indices of LV function, delays progression of myocardial dysfunction, and improves survival.
* Beta blocker therapy improves exercise duration, stabilizes LV function, and improves mortality to a similar degree in patients with ischemic and nonischemic cardiomyopathies.
* Available evidence also supports beta blocker use for treatment of HF due to systolic dysfunction in subgroups including patients with ischemic and nonischemic cardiomyopathies, stable class IV HF, women, blacks, patients with diabetes, and elderly adults.
* Beta blockers should be initiated at the doses started in the controlled trials, and the doses should be titrated up to the target doses of these studies as tolerated.
* Although retrospective data suggest a possible beneficial effect in HF of some beta blockers without proven benefit in randomized controlled trials (eg, atenolol, but NOT short-acting metoprolol tartrate), these studies do not provide a sufficient basis for recommendation of their use for HF treatment.
* Beta blocker therapy provides incremental benefit in patients with HF treated with an ACEI inhibitor or an ARB or an aldosterone antagonist. (See 'Use with other HF drugs' above.)

Reference:
http://www.uptodate.com/home/index.html