Chronic obstructive pulmonary disease

In England, the Department of Health estimates that 3.2 million people have COPD and 40% of these patients also have heart disease,2 especially heart failure.3 ,4 People with COPD and heart disease have a particularly high risk of death from heart disease and stroke (2–5 times higher than those with heart disease alone5 ,6). As β-blockade is highly desirable in this high-risk group, it is therefore very important to sort out the thorny issue of whether any, some or all of these patients can be given a β-blocker and how this can be safely achieved in everyday clinical practice and not just within a carefully monitored study population.

The definition of COPD includes ‘airflow limitation that is not fully reversible’.7 Theoretically, patients with COPD have significant fixed, irreversible airways obstruction and therefore little reversibility and so β-blockade should pose no risk. Global initiative for chronic obstructive lung disease (GOLD)7 recommend that ideally the obstruction should be measured postbronchodilator (in order to truly assess the degree of ‘fixed’ obstruction), although for practical reasons this is not always followed in the real world. The problem is that possibly half of patients with a diagnosis of COPD have some significant reversibility and up to 50% of patients change from being ‘irreversible’ to ‘reversible’ between visits, making it difficult for the clinician to be certain of a real ‘fixed’ component.8–10

Traditionally reversibility has been defined as >12% improvement, or 200 mL increase, in FEV₁. However, the American Thoracic Society/European Respiratory Society (ATS/ERS) Task Force concluded that for COPD, a 100 mL change in FEV₁ was clinically significant and a 5%–10% change from baseline clinically important.11 Furthermore, β2-agonists are the mainstay of treatment in COPD and have been proven to improve lung function, breathlessness and quality of life, and to reduce the exacerbation rate (many studies including12–14).

β-Blockers

β-Blockers were developed by Sir James Black in the 1960s for the treatment of hypertension and angina15 and there are now many different β-blockers available for use in patients. β-Blockers primarily have their beneficial effects by blocking the β1-adrenoceptors in the heart14 and thus preventing the action of endogenous catecholamines adrenaline and noradrenaline. Therefore, β-blockade results in a reduction in both the rate and force of contraction and thus a reduction in myocardial demand and, beyond doubt, reduce mortality in heart disease (table 1).16–20

In patients with heart failure, randomised placebo-controlled trials have demonstrated a 34%–35% reduction in mortality (metoprolol,21 bisoprolol,22 carvedilol23). A further β-blocker, nebivolol, was found useful in elderly patients.24 However, β-blockers are not a single entity and differ, among other properties, in their receptor selectivity (between β1-adrenoceptors and β2-adrenoceptors) and in the extent of intrinsic sympathomimetic activity (ISA, or partial agonism). In heart failure, the two studies that examined β-blockers with ISA, were found to be not beneficial (bucindolol25) or actually increased mortality (xamoterol26). In ischaemic heart disease, a variety of β-blockers without ISA reduced mortality during the acute MI period (usually taken as 2 years), for example, timolol by 39%,27 propranolol by 26%28 and metoprolol by 36%,29 whereas trials with β-blockers with ISA (eg, oxprenolol and pindolol) did not show beneficial effects.30–33

Although β1-adrenoceptors predominate, the human heart also has a small proportion of β2-adrenoceptors.34 ,35 Animal studies have suggested that long-term activation of β1-adrenoceptors has more marked deleterious effects than activation of β2-adrenoceptors,36 and the beneficial effects of β-blockers are therefore attributed to β1-blockade.16

The β-adrenoceptor present in the lungs is the β2-adrenoceptor. Stimulation of these β2-adrenoceptors results in bronchodilation and this has been the mainstay of treatment for both short-term rescue and long-term maintenance in patients with asthma and COPD for decades.37 Current β-blockers are poorly selective for the cardiac β1-receptors over the respiratory β2-receptors.38–41 Even those considered ‘cardioselective’ are actually poorly β1-selective, with a β1-selectivity of only 13-fold to 19-fold (eg, bisoprolol38–40). To put this in context, the β2-agonist salmeterol has a β2-selectivity of 1000–3000 over the β1-adrenoceptor, and commonly used laboratory antagonists are about 500-fold selective for their respective β1-adrenoceptor or β2-adrenoceptor subtypes. Thus, current β-blockers available for clinical use, even those considered to be the most cardioselective, can cause the β2-mediated respiratory side effects of bronchospasm or a fall in FEV₁ in susceptible individuals.42

Theoretical and clinical study concerns of β-blockers in COPD

There are several theoretical concerns over giving β-blockers to patients with COPD. First, a β-blocker may reduce the patient's basal (stable) FEV₁, and in severe disease, even a small decrease in lung function could have a major impact on symptoms. Second, a β-blocker would diminish the response to rescue β2-agonist therapy. Thus, during an acute exacerbation a β-blocked patient may have a lack or diminished response to important and potential life-saving β2-agonist rescue therapy. This would occur even if the β-blocker is well tolerated while the patient was stable. Third, there is concern that the long-term co-administration of a β-blocker would block or reduce the long-term beneficial effects reported with long-acting β2-agonist therapy (table 2).

Several clinical studies have shown that β-blockers do reduce FEV₁. In a recent large general population study, non-selective β-blockers reduced FEV₁ by 198 mL and ‘cardioselective’ β-blockers by 118 mL,43 that is, above the ATS/ERS Task Force ceiling of 100 mL change in FEV₁ being clinically significant in COPD. This population included people with asthma and COPD. If these people (ie, those at high risk of bronchospasm) were excluded, the reduction in FEV₁ was still over 100 mL, even with ‘cardioselective’ β-blockers. Furthermore in the general population, FEV₁ increased in those who stopped taking β-blockers during the study. Other studies in patients with COPD report a fall in lung function and wheeze with propranolol and metoprolol.44 ,45 A study directly examining β-blocker tolerability in patients with concomitant heart disease and carefully documented airways disease, found that β-blockers were tolerated by 84% of the COPD patients (but only 50% with asthma).46

A complication of examining historical β-blocker studies for safety in obstructive airways disease is that many of the older studies used β-blockers that are no longer recommended. For example, β-blockers with ISA (ie, partial agonists41 ,47 ,48) were thought to be beneficial because they would block the potentially harmful effect of high levels of catecholamines while providing a background increased sympathetic activation to maintain cardiac output and prevent bronchospasm.49 Clinical studies have since shown that β-blockers with agonist activity are not beneficial, and even detrimental, in cardiovascular therapy and so are no longer recommended. A Cochrane review concluded that β-blockers were safe in patients with COPD.50 However, 11 of the original 20 studies cited were one day, single-dose studies in patients with stable COPD. The remainder were of short duration (2 days to 12 weeks), and many studies used β-blockers with ISA that are not now recommended. Furthermore, the limitations listed by the authors included small sample size studies, 80% were male patients, randomisation was not always clear, some studies were only single blinded and some had no placebo controls.

To assess the safety of β-blockers in people with COPD, and particularly to assess their safety during exacerbations, long-term clinical studies are needed. Dorow et al51 studied the effects of bisoprolol and atenolol in 40 patients with COPD and angina in a double-blind cross-over study. They reported a 460 mL/min reduction in peak expiratory flow rate (PEFR) following 6 months bisoprolol and a 270 mL reduction in PEFR with atenolol. They also showed that the decrease in FEV₁ improved during the washout period after the β-blocker was stopped, suggesting the decline was treatment related rather than disease progression. Hawkins et al52 studied bisoprolol in patients with heart failure and COPD and found a 70 mL reduction in FEV₁ with bisoprolol (14 patients), while a 120 mL increase in FEV₁ with placebo (13 patients). They also assessed symptoms and concluded that ‘despite the fall in FEV₁, symptoms and quality of life were not impaired’. Another cross-over study looking at lung function in heart failure patients taking carvedilol, metoprolol and bisoprolol noted that the biggest reduction in lung function was in those taking the least β1-selective β-blocker carvedilol.53

Thus, β-blockers are associated with a reduction of lung function (FEV₁) in the general population and in those with COPD, and the β1-selectivity versus β2-selectivity of the β-blocker used appears to play a part.

A recent study examined the effect of β-blockers on mortality in patients with severe, oxygen-dependent COPD.54 In this prospective study, which attempted to account for the immortal time bias55 and immeasurable time bias56 suggested to be present in many other COPD and β-blocker studies,57 Ekström et al reported an increased mortality in patients with severe COPD taking β-blockers.

Theoretical and clinical study benefits of β-blockers in COPD

There are several theoretical reasons why β-blockers could be protective in the COPD population (table 2). COPD and heart disease share environmental risk factors (smoking and pollution) and both share pathophysiological factors—high adrenergic activation and increased systemic inflammation.20 ,58 ,59 A deterioration or exacerbation of either disease will increase both the adrenergic and inflammatory state, which could potentially worsen the other condition. Furthermore, β-adrenergic activation and immunomodulation are linked.60–62 β-blockers have been shown to reduce cytokines in patients with heart failure63 and improve outcomes in sepsis.64 So, first, β-blockers may reduce the background risk of exacerbation of both COPD and heart disease by lowering both the background adrenergic and inflammatory states (which may include both systemic inflammation and organ-specific inflammation).

Second, during an exacerbation of COPD, there are yet higher circulating catecholamines. This increases the risk of myocardial ischaemia, tachycardias, deterioration in heart failure and hypertension (and therefore stroke risk). β-Blockers could counteract all of these endogenous exacerbation-associated cardiovascular risks.

Third, during an exacerbation of COPD, patients are treated with β-agonists (often nebulised salbutamol), which in themselves cause significant tachycardia. Salbutamol (especially when nebulised) is significantly systemically absorbed but this compound is also a poorly selective ligand41 and will cause activation of cardiac β1-adrenoceptors. Treatment with β2-agonists has been shown to precipitate MI and increase adverse cardiovascular events.65–68 β-Blockade might also prevent this drug-induced cardiovascular risk.

There are several observational studies suggesting that β-blockers are effective in heart disease in people with COPD, the benefit outweighing the risk in terms of improved overall survival. Following an MI, Gottlieb et al69 noted a 40% reduction in mortality with β-blockers (in keeping with the later randomised controlled trials), for those with and those without COPD. Lower all-cause mortality was also observed in COPD patients post-MI by Quint et al70 (average follow-up 2.9 years) and Andell et al71 (average follow-up 2.8 years). Quint et al used data from the Myocardial Ischaemia National Audit Project and highlighted the ongoing uncertainty of prescribing β-blockers to patients with COPD, for only 38% were taking one of seven β-blockers on discharge, despite a very high uptake of other drugs for secondary prophylaxis. In another study in people with heart failure and COPD, patients taking β-blockers had less exacerbations and emergency room visits (although exacerbation here was taken as any change from the stable state, be that cardiovascular, respiratory or anything else).72

With regard to the general COPD population, observational studies have reported lower exacerbation rates and improved survival in patients with COPD and cardiovascular disease treated with β-blockers.73 ,74 Others also reported a decrease in COPD exacerbations, including in people with severe COPD and those on home oxygen (in contrast to the study by Ekström et al54), and that this benefit was only seen with β-blockers, not with ACE inhibitors or calcium channel blockers, suggesting there is something special about β-blockers that reduced exacerbations.75 This issue is to be further investigated in a placebo-controlled trial with metoprolol,76 which has a β1-selectivity versus β2-selectivity of twofold to sixfold.38–40

Giving β-blockers during exacerbations has also been reported as safe. Dransfield et al77 observed a lower mortality in inpatients with an acute exacerbation of COPD taking β-blockers than those not. Rutten et al78 noted an improved mortality in patients with ‘acute bronchitis’ (taken as a surrogate for COPD) taking ‘cardioselective’ β-blockers and calcium antagonists (but not for non-cardioselective β-blockers, ACE inhibitors or angiotension receptor blockers). In a study involving 35 000 patients, Stefan et al79 observed no difference in mortality in those taking β-blockers admitted for an exacerbation of COPD (ie, β-blockers did not kill patients during the exacerbation), but neither did they observe the reduced mortality reported by Dransfield et al77 nor Rutten et al,78 and non-selective β-blocker use was associated with an increase in readmissions.

However, before recommending prescribing β-blockers to all with a diagnosis of COPD and heart disease, it is important to remember that the above are all observational studies and therefore open to significant bias.57 The severity of, or specific diagnostic criteria for COPD, is not always known, β-blockers used are not always specified, prior or new events of angina and MI and specific causes of death are not always reported. There are no long-term randomised studies assessing cardiovascular prophylaxis with β-blockers in patients with a history of either angina or MI and COPD, which is where a large potential benefit lies. Thus, although there is plenty of observational evidence of a potential benefit in people with COPD and cardiovascular disease, conclusive evidence is lacking. Despite this crucial deficit, some argue that it would be unethical to withhold β-blockers, for example, after an MI, a view seemingly endorsed by the National Institute for Health and Care Excellence.70 ,80 ,81 However, we believe it is time to move on with randomised controlled trials to generate definite answers, and to define which groups benefit from which treatments and how long the treatment should be continued (particularly in the post-MI COPD cohort).

Current recommendations and moving forward

There are an estimated >1.2 million people in England alone with COPD and heart disease.2 Furthermore, given the smoking rates in Asia (eg, two-thirds of men in China smoke82), this is a growing global problem that needs to be addressed. Even if a small decrease in FEV₁ does not impair quality of life, clinician anxiety and the need to ensure that absolutely no harm is caused (including the need for effective β2-agonist rescue therapy during and acute exacerbation) is paramount in achieving better β-blockade for those who have COPD and concomitant heart disease. Furthermore, the difficulty in separating ‘pure’ COPD patients from those with asthma/COPD is not always easy and it is now recognised that there is significant overlap.83

GOLD84 recommends treating those with heart failure or ischaemic heart disease and COPD with β-blockers. Clearly, however, it is crucial to observe patients for worsening breathlessness and deterioration of spirometry despite this guidance, and review the patient and stop β-blocker treatment if needed. The current recommendation is therefore to use the most ‘cardioselective’ β-blockers currently available (eg, bisoprolol), starting at a low dose, whenever possible in patients with COPD in a ‘try it and see’ approach.20 ,85 The Scottish guidelines (SIGN 147, also published this year) makes similar recommendations for those with ‘COPD without significant reversible airways obstruction,’ where there is ‘no suitable alternative’, and also recommends close monitoring for adverse effects (see Annex 4 of SIGN 147.).86 Likewise, limiting systemic exposure to β-agonists (eg, using 2.5 mg nebulised salbutamol instead of 5 mg, and stopping nebulisers as soon as possible after the acute phase of an exacerbation) in those with COPD and heart disease would reduce the cardiovascular risk as much as possible.85 These recommendations, however, are not founded on well-conducted prospective randomised trials and does not extrapolate to patients with asthma, where β-blockers remain contraindicated.

Moving forward, to understand whether our current β-blockers are safe or indeed beneficial, we need large randomised placebo-controlled trials to truly explore whether the benefit for cardiovascular mortality and morbidity (and potential improvement in COPD morbidity) does indeed outweigh the potential mortality and morbidity from bronchospasm or reduction in benefit of both long-acting and short-acting β-agonists in COPD.

An alternative may be the further development and assessment of a truly β1-selective β-blocker, one that has such poor affinity for the β2-adrenoceptor that it could not cause any β2-mediated bronchospasm. Such molecules have recently been synthesised (with β1-selectivity vs β2-selectivity of 500-fold to 5000-fold87) and have been proven to be orally bioavailable and highly β1-selective in animals. Such a drug would circumnavigate the entire bronchospasm issue, yet still provide the well-proven cardiovascular β-blocker benefit. In addition to providing a safe way forward for those with COPD and heart disease, peripheral vascular disease or other conditions mentioned above, it would potentially extend the safe use of β-blockers to those who have heart disease and asthma.

Another alternative is to avoid β1-adrenoceptor antagonists altogether. If their beneficial prophylactic effect is predominantly due to a reduction in heart rate,88 ,89 then other drugs with this effect might fulfil this role. Ivabradine is such a contender. This recently licensed drug reduces heart rate by blocking I_f sodium-potassium channels, expressed particularly in the sinoatrial node, without reducing the power of myocardial contractility, unlike β-blockers. In a large randomised controlled trial in patients with stable coronary artery disease and left ventricular systolic dysfunction <40%, ivabradine (in addition to standard therapy, ie, for many patients in addition to β-blockers) failed to reduce the primary end point of cardiovascular death, admission for new or worsening heart failure or MI, but it did establish its tolerability and a suggested benefit in those with an entry heart rate above 70 bpm.90 The SHIFT trial, in patients with stable systolic heart failure (left ventricular ejection fraction ≤35%) and basal sinus rhythm >70 bpm, suggested an improvement in heart failure (due to a reduction in heart rate) in those taking ivabridine in addition to standard therapy.91 Unfortunately, this benefit was not confirmed in a subsequent trial of patients with stable coronary artery disease without heart failure, in whom the addition of ivabridine reduced heart rate but not improve clinical outcomes.92

In the COPD arena, ivabradine has been demonstrated to lower heart rate without impairing lung function,93 and to reduce tachycardia after salbutamol inhalation.94 In a substudy of patients with COPD in the SHIFT (heart failure) trial, those randomised to ivabradine (in addition to routine β-blockers) derived similar relative reductions in cardiovascular events to patients with no COPD, albeit without statistical confirmation.95 So the role of ivabridine in safe and effective cardiovascular prophylaxis in patients with COPD is, like β-blockers, not yet established either alone or in combination.