38 b-Adrenoreceptor antagonists and heart failure

Sir James Black’s quest for a substance that would block the potentially harmful effects of adrenaline on the ischaemic heart led to his invention of the β-adrenoceptor (AR) antagonists pronethalol and propranalol between 1959 and 1962.¹ The ubiquitous nature of adrenaline ensured that these β-AR antagonists, commonly referred to in clinical practice as β-blockers, found a role in multiple aspects of cardiovascular therapeutics.² Sir James could not have predicted the colossal impact his discovery of β-blockers subsequently would have in improving the lives and preventing the deaths of countless patients over the subsequent 50 years. Regarded by basic scientists as the father of analytical pharmacology, he was also a committed exponent of what we now regard as translational medical research. James Black’s achievement in discovering β-blockers followed by histamine type 2 antagonists (H₂ blockers), the first effective treatment for peptic ulcer, was recognized by the award of the Nobel Prize in 1986.

Black’s original hypothesis was that patients with stable angina pectoris might benefit by reducing the work of the heart, rather than increasing the blood flow. So the first clinical study with the initial β-blocker pronethalol was carried out in angina pectoris by Pritchard.³ In a careful dose–response study he showed that there was clinical improvement in 16 of 17 patients. Pritchard also noted that during the dose–response period there was also a fall in blood pressure. This serendipitous observation was confirmed in a subsequent study with propranolol showing a significant fall in blood pressure in 17 of 18 patients. Subsequent clinical studies showed that propranolol had antiarrhythmic properties, and much later reduced mortality in patients with myocardial infarction (MI).⁴ Most recently β-blockers have conclusively been shown to be of great benefit in heart failure (HF), clearly demonstrating the extensive role of sympathetic nervous activity in cardiovascular disease.^5,6 The pharmacological properties of β-blockers that have been used in heart failure studies are summarized in Table 38.1.

Using β-blockers in HF seems counter-intuitive given that the increased sympathetic activity that occurs in patients with HF is required to support the failing heart. Early experimental studies of propranolol in experimental HF in calves^7,8 fuelled a natural concern at inhibiting the very system on which the heart appeared to depend for support. Based on such clinical considerations, there was a universally agreed view that the use of the nonselective β-blockers available at that time was contraindicated in HF.⁹

Despite these experimental findings and clinical anecdotage, the early published studies of β-blockade in patients with HF in the United Kingdom and Sweden did point tantalizingly to potential benefit in selected patients, but it was not until the 1970s that the somewhat counter-intuitive notion of using β-blockers to treat patients with HF began to be taken seriously. The initial clinical studies of the effects of β-blockade in HF were performed both by Gibson et al. at the National Heart Institute, London⁹ and by Waagstein and colleagues in Gothenburg.^10,11 Both groups studied the β₁ cardioselective blocker practolol (ICI50172), and the Swedish workers also used alprenolol, which has partial agonist activity, in patients with idiopathic dilated cardiomyopathy. The initial studies from both groups involved the parenteral administration of practolol to patients in severe HF with supraventricular tachycardia. The marked reduction in heart rate was accompanied by significant clinical improvement without side effects. A subsequent long-term study with the β₁-selective blocker metoprolol in patients with dilated cardiomyopathy showed that after 3 years of treatment, the survival in the β-blocker-treated group was 52%, compared to only 10% for matched controls.¹²

The idea that β-blockers might be of value in HF was consistent with the developing ‘neuroendocrine hypothesis’ of HF. This was prompted when the CONSENSUS study showed a huge 60% reduction in mortality by the angiotensin converting enzyyme (ACE) inhibitor enalapril.¹³ Subsequently it was suggested that several neurohormonal systems, principally the renin–angiotensin–aldosterone system (RAAS) and the sympathetic nervous system, could be responsible for the apparently inexorable deterioration of cardiac function and high mortality in HF.

Thereafter, inhibition of this neuroendocrine response by ACE inhibitors,^13,14 selective inhibition of angiotensin II receptors by angiotensin receptor blockers (ARBs),¹⁵ and of aldosterone receptors¹⁶ together with β-blockers has become the standard evidence-based approach to the medical treatment of HF. This combination of treatments markedly improves survival by reducing both sudden cardiac death and death due to worsening HF. It also leads to improved ventricular function due to amelioration of ventricular remodelling consequent upon myocardial cell loss from MI or other heart muscle disorders.

There can be few medicines whose efficacy in HF has been so clearly demonstrated as β-blockers insofar as the two major outcomes of survival and need for hospital admission are concerned. Following the early pioneering and hypothesis-generating studies of the potential value of β-blockers in MI and HF performed in the United Kingdom and Sweden with practolol, alprenolol, and metoprolol, there are now data from over 15 000 patients from randomized placebo controlled trials that, collectively, demonstrate impressive statistically and clinically important reductions in death and unscheduled admissions to hospital for worsening HF. A summary of the major trials can be found in Table 38.2.

Additional interest in the use of β-blockers in chronic HF dramatically increased more than 15 years ago when the results of the US Carvedilol Heart Failure Trial Programme on survival in just over 1000 patients with NYHA class II–IV HF were disclosed.¹⁷ This was a series of double-blind randomized controlled trials in HF which individually investigated dose, haemodynamics, exercise capacity, and symptoms while collectively assessing the impact of β-blockade on mortality. The reduction in mortality reported in this programme greatly strengthened the hypothesis that β-blockers could indeed improve survival to a similar extent as had been demonstrated in the definitive ACE inhibitor studies in HF of the previous decade. It also lent even greater credence to the neuroendocrine hypothesis of HF promulgated in the wake of the early ACE inhibitor studies—that activation of a number of neural and endocrine systems, notably the RAAS and sympathetic nervous system, was not simply a reflection of the presence and severity of HF, but also directly contributed to poor outcomes in patients with HF through deleterious effects on the myocardium, blood vessels, and kidney.

Published in 1996, the carvedilol HF programme was an ambitious enterprise of great significance to the development of a comprehensive evidence base for the use of β-blockers together with ACE inhibitors in patients with HF due to left ventricular systolic dysfunction (LVSD). Nevertheless the survival benefit was based on a relatively small number of deaths, 31 (7.8%) on placebo versus 22 (3.2%) on carvedilol, representing absolute and relative reductions in mortality of 4.6% and 65% respectively. The overall mortality in the carvedilol programme was relatively low and it was clear that the results of further ongoing and new studies in larger and more diverse patient groups were required to establish as securely as with ACE inhibitors a strong evidence base for the safety and efficacy of β-blockers to improve clinical status and outcome in HF.

In 1998 the results of the CIBIS II trial of bisoprolol (β₁ selective) in over 2000 patients, essentially the first large single randomized controlled trial of β-blockade in HF, showed a survival advantage for patients in the bisoprolol arm with an increase in survival of 5.5% in a trial where the mortality rate in the placebo group was 17.3%, the highest studied thus far in a large clinical trial.¹⁸ The trial patients included patients in NYHA classes II–IV. CIBIS II also drew attention to the remarkable relative reduction of 54% in the sudden death rate and the important reductions not only on hospital admission for worsening HF but also on all-cause hospital admissions.

Later that year the MERIT trial of metoprolol in an even larger study of almost 4000 patients essentially completed the evidence base for the impressive improvement in outcome in terms of survival and admission to hospital afforded by chronic treatment with β-blockers.¹⁹ It also confirmed the marked reduction in sudden cardiac death demonstrated by the preceding trials. Thus in the short space of 3 years the cumulative evidence for the efficacy of β-blockers raised the profile of β-blockers as having at least as salutary an additional beneficial effect on morbidity and mortality as ACE inhibitors when given in conjunction with them.

Unfavourable changes to the geometry, structure, and function of the left (or right) ventricles accompany several disease processes affecting the myocardium, among which MI is the most typical and frequent cause. Myocyte loss or dysfunction can set in train a series of cellular processes including myocyte hypertrophy and renewal, as well as interstitial fibrosis and dilatation of the left ventricle, causing the chamber to become more spherical than elliptical. This results in increased wall tension, reduced contractility, impaired filling, and ultimately the clinical syndrome of HF.²⁰

In 1997 the Australia and New Zealand (ANZ) trial in 415 patients convincingly demonstrated that carvedilol improved left ventricular systolic function although exercise capacity did not change.²¹ The ANZ trial also contributed to the evidence that clinical outcomes improved on β-blockers. Subsequently the CAPRICORN trial of carvedilol showed that similar benefits can occur with β-blockers in conjunction with ACE inhibitors in patients with post-MI left ventricular dysfunction or HF.^22,23 The totality of data from these and other mechanistic trials indicate that β-blockers have a clinically important beneficial effect on remodelling in HF.

It is said that it was the sudden unexpected death of the young James Black’s father following a very stressful day at work that caused him to consider the possibility that high levels of adrenaline from excess sympathetic nervous activity could have been responsible and that blocking these effects of adrenaline on the heart might prevent such tragedies. Perspicacious in the extreme, perhaps, but there is no doubt that a major feature of β-blocker therapy is the reduction in sudden death observed both after MI and in HF. Given the proarrhythmic effects of catecholamines, the major antiarrhythmic effects of β-blockers is scarcely surprising. This is a singular benefit of β-blockers, for no other medicine currently available in the entire pharmacopoeia has conclusively been shown to reduce sudden death. The precise mechanism by which β-blockers achieve this effect is unclear, but substantial reductions in serious ventricular arrhythmias including ventricular tachycardia and ventricular fibrillation were noted during ambulatory ECG monitoring in CAPRICORN, a trial in which a marked reduction in sudden death was a feature in the carvedilol-treated group.^22,23

Of considerable importance also is the suppression of supraventricular arrhythmias, especially atrial fibrillation and atrial flutter, which in patients with HF often precipitate acute HF and decompensation in patients with known chronic HF.

The down side of the antiarrhythmic properties of β-blockers is bradycardia, which may limit the maximum tolerated dose in individual patients and lead to various degrees of heart block, most commonly first degree but occasionally second degree and complete heart block. In the latter it is usually in the context of pre-existing conduction disturbance and in such cases permanent pacing may be required.

Despite the impressive results of the US Carvedilol programme and the CIBIS and MERIT trials, many questions still remained to be answered about β-blocking therapy, including its safety and efficacy in very severe HF. For, although patients with NYHA classes III and IV were included in the preceding trials, they were relatively few in number and had to have been clinically stable for some time before entry to these trials. This question has best been addressed in the COPERNICUS trial of carvedilol in over 2000 patients with NYHA class III or IV HF in which there was a total of over 500 deaths, the largest yet recorded in a HF trial. Carvedilol reduced all-cause mortality,²⁴ but, overall, significant survival benefit was also derived by patients in a prespecified subgroup at even higher risk due to, among other criteria, recent decompensation or a severely depressed ejection fraction of 14% or less. In this subgroup the annualized mortality rate was 24%. COPERNICUS indicated that, far from being hazardous, it is in the highest-risk groups of patients that β-blockers can exert their greatest effects. It must be stressed, however, that these benefits of β-blockers were obtained in the context of a standard of patient care as near optimal as possible.

The powerful effects of β-blockers on survival in patients with cardiovascular disease was first demonstrated in the β-blocker trials in MI in the previous decade. The first of these trials, and still the most impressive in terms of outcome benefits, the Norwegian multicentre trial of timolol published 15 years previously in 1980, was followed by several other trials with metoprolol, propranolol, practolol, oxprenolol, sotalol, and alprenolol in which the results on survival and recurrent MI, though more variable, were directionally similar to those of the timolol trial. Subsequent meta-analyses confirmed the powerful beneficial effects of β-blockers in the post-MI patient on survival and recurrent MI.⁴ Patients with HF were included in some of these studies but generally they were excluded because of safety concerns. Consequently it soon became apparent that, in clinical practice, the prescription rate of β-blockers in patients whose MI was complicated by HF remained low. This was the reason for carrying out a specific trial in post-MI patients with significant left ventricular dysfunction and/or HF, the CAPRICORN trial with carvedilol in approximately 2000 patients.²⁴ The impact on survival reflected in a relative risk reduction of 23% (absolute reduction 3%) was identical to that calculated in meta-analyses of the previous MI trials, and the effects on sudden death and recurrent MI were also statistically and clinically highly significant. The CAPRICORN trial therefore completed the spectrum of post-MI trials by indicating that patients with HF or severe left ventricular dysfunction post MI should definitely and specifically be considered for β-blocker therapy. An important practical contribution was the experience gained in initiating β-blocker treatment safely in patients with acute HF guided by specific clinical indicators which could be applied for the safe use of β-blockers in patients with HF however caused.

Over many years the main experimental and clinical focus of interest in HF encompassing particularly the neuroendocrine hypothesis for the development and progression of HF has been on LVSD. It is now recognized that while patients in this category who are characterized by a reduced ejection fraction (REF), a significant proportion of patients who present with all the typical clinical features of HF have a ‘preserved’ or normal ejection fraction (PEF or NEF). These patients more often are older, with a history of hypertension and evidence of left ventricular hypertrophy.²⁵ Although they are not rare there is debate on whether they are as numerous as those with a reduced ejection fraction. This has led to the introduction of numerous acronyms by which patients with HF may be described, HeFREF and HeFNEF being, currently, among the most frequently employed. Most of the ‘landmark’ clinical trials of β-blockers in HF have been in patients with HeFREF while such data in patients with HeFNEF are few. Theoretically the risk/benefit ratio could be more favourable with less myocardial depression and improved ventricular filling from a slower heart rate.

Although no large trial specifically in HeFNEF has been reported, the SENIORS trial studied nebivolol, a β₁-selective β-blocker with a nitric oxide based vasodilator activity, in a large trial of patients aged 70 or older. A secure clinical diagnosis of HF but no specific value of ejection fraction was required.²⁶ The composite 1-year endpoint of all-cause mortality or cardiovascular hospitalizations was significantly reduced by nebivolol regardless of ejection fraction. Nebivolol is licensed for the treatment of HF in some countries but it has not been approved in the United States. Thus there is insufficient evidence to say confidently that nebivolol is more effective in HF or that other β-blockers would have a similar effect in the SENIORS patient group. Nevertheless it is common practice to prescribe β-blockers in clinically stable patients with HeFNEF according to the same safety specifications applied to patients with HeFREF.

It is always a significant clinical disappointment when contraindications prevent the use of life-enhancing and life-saving medicines. With respect to β-blockers, one of the most common is reversible airways obstruction. In patients who have a firm diagnosis of asthma based on appropriate investigations β-blockers, including those that are said to be β₁ selective, are firmly contraindicated. In those with confirmed chronic obstructive pulmonary disease (COPD) without significant reversibility, clinical trials and meta-analyses suggest that the risk–benefit ratio remains favourable.²⁷ In the presence of diagnostic doubt it is advisable to seek a formal opinion from a respiratory specialist as to the correct underlying diagnosis. Ultimately the decision lies with the physician who, aware of the balance of risks of exacerbating bronchoconstriction and of withholding a most effective treatment for HF and preventing sudden cardiac death, can give an informed opinion.

Men and women do not always respond equally to medicines. but as far as β-blockers are concerned a meta-analysis involving the major trials (USCTP, CIBIS II, MERIT-HF, and COPERNICUS) confirms equal benefit in terms of major outcomes.²⁸

The human heart contains both β₁ and β₂ ARs in a ratio of approximately 4:1. It is commonly believed that the harmful effects of increased sympathetic activity are mediated by the β₁ receptor through G-protein-coupled stimulation of cyclic AMP (cAMP) leading to activation of a number of downstream signalling pathways associated with ventricular and vascular remodelling.

Nevertheless, there has been prolonged debate as to the relative merits clinically of β₁ ‘selective’ and ‘nonselective’ β₁ and β₂ AR antagonists in cardiovascular disease generally. Insofar as HF is concerned this issue seemed to be resolved when the β₁-selective β-blockers bisoprolol and metoprolol and the nonselective β-blocker carvedilol were all shown to reduce all-cause death substantially and to the similar extent of about one-third in, respectively, the CIBIS II, MERIT-HF, and COPERNICUS trials. But interest in potential differences, initiated by the remarkable two-thirds reduction in mortality in the US carvedilol trials, was dramatically heightened by the results of the COMET trial, in which the nonselective β-blocker carvedilol was associated with a statistically and clinically significant lower mortality than the β₁-selective β-blocker metoprolol.²⁹ Although carvedilol has other potentially valuable ancillary properties, including α₁ AR blockade, there is no supportive outcome data from large clinical trials in HF for benefit of α₁ blockade³⁰ or antioxidant activity.

For these reasons, the question of dose was questioned in COMET in which the aim was to effect comparable reductions in heart rate between the two groups since the molecular effects in the myocardium are similar when equipotent doses of carvedilol and metoprolol, in terms of inhibition of exercise-induced tachycardia, are compared.³¹

The resting heart-rate reduction of 13 beats/min in COMET with carvedilol 50 mg/day was very similar to that achieved with the same dose in the US carvedilol studies on which the dose was based. The heart-rate reduction with the dose of 100 mg/day metoprolol tartrate, however, was only 11.7 beats/min compared with 15 beats/min achieved with 150 mg/day in the Metoprolol in Dilated Cardiomyopathy trial (MDC) on which the dose of metoprolol in COMET was based. Moreover, the major study of metoprolol in HF that addressed outcome was the MERIT-HF study in which the preparation of metoprolol used was metoprolol succinate, a controlled/extended release formulation (metoprolol CR/XL) in a target dose of 200 mg/day. The mean dose actually taken and the mean reduction in heart rate achieved were 106 mg and 14 beats/min, both very similar to that found in the MDC trial. Thus the greater benefit in outcomes with carvedilol over metoprolol in the COMET trial might suggest that equivalent blockade of cardiac β₁ receptors may not have been achieved in the metoprolol arm.

Recent experimental data on β-blocker pharmacology (Fig. 38.1) challenge the conventional wisdom regarding the primacy of the β₁ receptor in health and disease and HF in particular. First, receptor-binding studies using cultured human cells have questioned the validity of previous clinical classifications of β-blockers as ‘selective’ or ‘nonselective’.³² For example, carvedilol in human tissue is a more potent blocker of β₂ than of β₁ receptors, while the β₂ effects of drugs formerly classified as ‘selective’ β₁ antagonists such as bisoprolol and metoprolol could be clinically more significant than previously appreciated.³² Secondly, in health, the β₁ ARs are located on the cell crests, ensuring their wide distribution over the entire cell surface (Fig. 38.2).³³ This facilitates, following their stimulation, wide diffusion of the Gs-protein-coupled production of the second messenger cAMP throughout the cytoplasm where it increases the strength and frequency of myocyte contraction. The effects of cAMP following stimulation of the β₂ ARs, residing in the base of the T tubules of the cell membrane, are more localized. Recent studies show, however, that this compartmentalization of β-AR function is disrupted, leading to relocation of β₂ ARs to the cell surface where their subsequent stimulation causes a pattern of cAMP release similar to that of β₁ receptor stimulation. Thirdly, β₁ but not β₂ receptors are down-regulated in HF,^34,35 thereby changing the effective ratio of functioning β₁ and β₂ activity from 4:1 to 3:2.

It has also been hypothesized that carvedilol may have a protective effect by stimulating the β-arrestin signalling pathway in the presence of chronic catecholamine stimulation leading to inhibition of mycocyte apoptosis, as is the case in HF. In contrast, G-protein-dependent signalling may be cardiotoxic under these same conditions.³⁶ Thus there remains a place for further investigation of the relative roles of β₁ and β₂ receptor function in HF.

Several β-adrenoceptor antagonists have partial agonist activity, which it was hoped might protect against β-blocker-induced myocardial depression while also ameliorating or preventing some common side effects including excessive bradycardia due to either β₁ or β₂ antagonism, cold hands, and bronchoconstriction due to β₂ antagonism. Partial agonism has never been shown conclusively to have beneficial effects, however, and in terms of outcome no improvement has been seen in post-MI trials.

The role of partial agonism in HF was investigated through the clinical development programme of the β₁ selective partial agonist xamoterol (CORWIN).^37,38 It was hypothesized that xamoterol would be useful in protecting the heart from the adverse effects of increased sympathetic stimulation during daytime activities while providing support for the heart through expression of partial agonism during periods of low sympathetic traffic during rest. Ambulatory ECG monitoring showed a clear reduction of the heart rate during the day and a substantial increase nocturnally, in keeping with its significant degree of β₁ agonism. The clinical trials that led to the licensing of xamoterol for mild to moderate HF demonstrated clear evidence of benefit by improving quality of life and increasing exercise capacity.³⁷ Unfortunately the large trial set up to investigate its efficacy and safety in severe HF was stopped by the data monitoring committee because of excess mortality in the xamoterol arm.³⁸ There was no pattern to the mode of death and no plausible explanation for the adverse outcome could be discerned from all the data collected. Despite the ambulatory monitoring studies no definite association between the heart rate findings and mortality could be determined. It remains tantalizing to speculate that had xamoterol been introduced at a very small dose and titrated slowly to the target dose, in keeping with modern β-blocking therapy in HF, the outcome of that trial might have been very different.

The only other β-blocker with partial agonism to be studied in a large randomized controlled trial in HF is bucindolol. Unlike all the other recent landmark trials of β-blockers, the BEST trial of bucindolol failed to show a convincing increase in survival^39,40 by being detrimental in black Americans but beneficial in white subjects. It cannot be concluded that partial β-receptor partial agonism was culpable, but it has been suggested that genetic polymorphisms among the subjects could have influenced the result. Bucindolol is known to be strongly sympatholytic and in BEST, as in the MOXCON study of moxonidine in heart failure, sympatholysis was associated with poor survival. Polymorphism of two genes concerned with β₁ AR function, the glycine moiety of the β-arginine/glycine 389 and the preadrenergic junctional α₂c-Del 3222–3225 AR, were linked to sympatholysis and poorer survival.^41,42

β-Blockers are a class 1 (A) recommendation in the major guidelines on HF treatment because of the overwhelming evidence of benefit, especially in patients with LVSD (HeFREF).^43,44 The evidence for benefit in patients with a preserved ejection fraction (HeFNEF) is much less complete at present, due not to adverse results but to a relative dearth of trials similar to those on which the secure evidence base for β-blockers in HeFREF has been built. The data in SENIORS,²⁶ the only large trial to address this issue, is consistent with the trials in HeFREF and their meta-analyses. In SENIORS, the most recent of the large trials on β-blockers, a pragmatic, more modern, approach was taken by including patients with ‘heart failure’ or ‘HeF’ in the acronymic parlance of the present times, as determined by a clinical diagnosis made in a specialist setting. The results revealed no heterogeneity of effect between those with impaired or significantly unimpaired left ventricular systolic function. Since the overarching view of the progression of HF and its pharmacological treatment is based on antagonizing the putative deleterious effects of neuroendocrine systems, especially the RAAS and the sympathetic nervous system,⁴⁵ it is not unreasonable to believe that inhibitors of these systems should be beneficial in HF per se, both HeFREF and HeFNEF, to an extent depending on the degree of activation of these systems. There are also other reasons for prescribing β-blockers due to the protean nature of their indications, including hypertension and angina pectoris, conditions which are present in many patients either as a comorbidity or a cause of their HF. The antiarrhythmic effects of β-blockers are particularly relevant in HF for preventing or gaining rate control of atrial fibrillation and for preventing sudden cardiac death from serious ventricular arrhythmia. In the specific case of HeFPEF there may also be pathophysiological justification for β-blockade including the promotion of better filling by reducing the heart rate.⁴⁶ For all these reasons β-blockers should be considered in all patients with HF in the absence of a specific disease related or other standard contraindication.

Research and especially clinical trials of medications has been almost completely dominated by studies in LVSD, although more recently complemented by studies of angiotensin receptor antagonists in patients with preserved systolic function. Very little is known about the effects of the major ‘heart failure’ medications including β-blockers in patients with other forms of specific heart disease such as valvular heart diseases and specific heart muscle disorders, apart from their indication for rate control in atrial fibrillation. In those circumstances clinical judgement is required to identify those patients with characteristics that might suggest benefit from β-blockers. in whom all of these issues consistently arise in individual patients.

In both Europe and the United States bisoprolol, carvedilol, and metoprolol are approved by the relevant regulatory bodies for the treatment of heart failure due to LVSD because of the strong evidence for benefit for all three medicines. In the United Kingdom but not in the United States nebivolol is also licensed for the treatment of HF in patients over the age of 70 years. Currently it has not been approved for HF by either the European Medicines Agency or the US Food and Drug Administration. No other β-blocker has been licensed in the HF indication because of the absence of large trials confirming safety and efficacy. In these circumstances it is recommended that, where possible, licensed medicines should be prescribed at a dose and in a manner that was tested in the relevant clinical trials.

Recent pharmacoepidemiological data^47,48 provide a perspective on outcomes of different classes of β-blockers from managed care databases—DECIDE and the North Carolina Medicaid/Medicare patients. The objective was to determine the outcomes of patients using either β-blockers approved by the FDA, namely carvedilol, metoprolol succinate, and bisoprolol fumarate, which are considered to have evidence-based treatment for HF, or atenolol. In DECIDE, there was no difference between atenolol and carvedilol in either mortality or rehospitalization rates, whereas there appeared to be an increased risk using short-acting metoprolol tartrate.⁴⁷ The Medicare/Medicaid study showed that there was no difference between outcomes in either approved or nonapproved β-blockers, and both showed a substantial benefit compared with no β-blocker treatment.⁴⁸

Pharmacoepidemiology data is available for atenolol—one of the most widely prescribed β-blockers in the world—for many other indications, which suggests that long-term outcomes were similar in patients on licensed β-blockers and atenolol. This is by no means equivalent to data on safety and efficacy but does give some comfort to patients and clinicians practising in economically challenged countries throughout the world.

The data from the large trials were obtained in patients in with NYHA class II–IV HF who were clinically stable at the time of initiation of treatment. The earlier trials recruited largely ambulant outpatients of milder severity and greater stability, while studies such as CAPRICORN²² and COPERNICUS²⁴ included inpatients recovering from an acute MI or recent decompensation of chronic HF respectively. These trials demonstrated the safety of inpatient prescribing in an environment where the availability of specialist expertise in HF and adherence to a clear protocol for initiation and up-titration allowed the safe introduction and continuation of β-blockade. Subsequently the additional benefit obtained from initiating treatment during an acute admission has been confirmed in clinical practice.⁴⁹

A detailed exposition of initiation, titration, and chronic dosing is beyond the scope of this chapter but it is crucially important that patients are clinically stable at the time of initiation even though this may be only a few days after an episode of decompensation or of an acute MI. Beta blockers are normally commenced following ACE inhibition. There is difference in outcomes between the two approaches of ACE inhibitor first versus beta-blocker first.⁵⁰ Recognition of ‘stability’ is a matter of clinical judgement but cardinal signs include absence of fluid retention and of other signs of circulatory failure such as hypotension (systolic blood pressure ≥90 mmHg), peripheral hypoperfusion, and oliguria. These criteria are a matter for local implementation guidelines ideally in the context of a multidisciplinary heart failure programme within which the safe and effective use of β-blockers is best achieved (Chapter 54).

The clinical trials all have had detailed initial dose titration schedules and a final target dose and these are a practical guide to management.

Adherence to medication and doses achieved in clinical trials are generally much higher than in ‘real life’ clinical practice for several very good reasons related to the selection criteria of the former circumstance and the clinical status and frailty of unselected patients in the latter. In clinical audits of HF prescription rates vary from around 80% in the US ‘ADHERE’ Registry⁵¹ to 49% in the EuroHeart Study⁵² and 40–50% in various audits in the United Kingdom.⁵³ These data may represent an apparently wide variation in practice but, more likely, they are simply reflecting a direct relationship between the degree of selectivity of patients enrolled in the individual studies and the rate of prescription and doses achieved. Nevertheless, in clinical practice it is possible to achieve target doses in a high percentage of patients depending on the skill, patience and time available to the health professionals concerned. Medical therapy in heart failure is complex for many reasons and should be conducted in a multidisciplinary setting such as in the clinical trials within which the data on safety and efficacy were obtained.

‘Withdrawal’ effects due to excess sympathetic stimulation have been noted after abrupt cessation of β-blockers in circumstances other than heart failure. In hypertension this has led to loss of control of blood pressure and in angina pectoris to acute exacerbations.⁵⁴ There has therefore been concern over the possible consequences of sudden withdrawal of β-blockers in patients with HF, especially in the situation of acute decompensation. In order to avoid this potential hazard many have simply reduced the dose to that of initiation with up-titration when stabilization has occurred. Recently no detriment has been reported by continuing with the maintenance dose.⁵⁵ Nevertheless, no hard and fast rule need apply and there will be clinical situations in which acute, partial, or no withdrawal will be appropriate according to the individual circumstances.

Reduction or withdrawal of β-blockers can be necessary in very advanced HF associated with hypotension and deteriorating renal function as part of a general review of the overall medicines prescription since all ‘evidence-based medicines’ for HF can lower the blood pressure to an extent that renal function is compromised. This is often temporary, but in the state of terminal care it is reasonable to continue only those medicines that will contribute positively to the palliative care of the patient.

The use of β-blockers in HF has been both a revelation and a revolution. Sir James Black, the inventor of by far the most successful class of medicines for cardiovascular disease, died earlier this year at the age of 82. Modest to the end about his immense contribution to cardiovascular science and medicine, his quests for new mechanisms and medicines had never ceased. Without prejudice to his many colleagues with whom he was still working, it may be said that the β-blocker odyssey of discovery has not ended.