Clinical Pharmacology, 11e

Arterial hypertension, angina pectoris, myocardial infarction and heart failure

Kevin M. O'Shaughnessy

Synopsis

Hypertension and coronary heart disease (CHD) are of great importance. Hypertension affects more than 20% of the total population of the USA, with its major impact on those aged over 50 years. CHD is the cause of death in 30% of males and 22% of females in England and Wales. Management requires attention to detail, both clinical and pharmacological.

The way in which drugs act in these diseases is outlined and the drugs are described according to class.

• Hypertension and angina pectoris: how drugs act.

• Drugs used in both hypertension and angina.

   Diuretics

   Vasodilators: organic nitrates, calcium channel blockers, ACE inhibitors, angiotensin II receptor antagonists

   Adrenoceptor-blocking drugs, α and β

   Peripheral sympathetic nerve terminal

   Autonomic ganglion-blocking drugs

   Central nervous system

   Treatment of angina pectoris.

• Acute coronary syndromes and myocardial infarction.

• Arterial hypertension.

• Sexual function and cardiovascular drugs.

• Phaeochromocytoma.

There is also now a better understanding of the mechanisms that sustain the failing heart. Carefully selected and monitored drugs can have a major impact on morbidity and mortality. However, much of the risk that patients with heart failure encounter is due to ventricular arrhythmias, which are minimised with implantable cardioverter defibrillators (ICDs) and cardiac resynchronisation therapy (CRT) rather than drugs. In view of the current complex range of choices for individuals with these issues, specialist referral should be considered in all cases.

• Specific treatments, including those for cardiac arrest.

• Drugs for cardiac failure.

Hypertension: how drugs act

Consider the following relationship:

This being true, drugs can lower blood pressure by:

• Dilating arteriolar resistance vessels; achieved through direct relaxation of vascular smooth muscle cells, indirect relaxation by stimulating nitric oxide (NO) production, or by blocking the production or action of endogenous vasconstrictors, such as noradrenaline/norepinephrine and angiotensin.

• Dilating venous capacitance vessels; reduced venous return to the heart (preload) leads to reduced cardiac output, especially in the upright position.

• Reduction of cardiac contractility and heart rate.

• Depletion of body sodium. This reduces plasma volume (transiently), and reduces arteriolar response to noradrenaline/norepinephrine.

Angina pectoris: how drugs act

Angina can be viewed as a problem of supply and demand. So the drugs used in angina pectoris either increase supply of oxygen and nutrients, or reduce the demand for them, or both.

The supply of myocardial oxygen can be increased by:

• dilating coronary arteries

• slowing the heart (coronary flow, uniquely, occurs in diastole, which lengthens as heart rate falls).

Demand can be decreased by:

• reducing afterload (i.e. peripheral resistance), so reducing the work of the heart in perfusing the tissues

• reducing preload (i.e. venous filling pressure); according to Starling's law of the heart, workload and therefore oxygen demand varies with stretch of cardiac muscle fibres

• slowing the heart rate.

Drugs used in hypertension and angina

Two groups of drugs, β-adrenergic blockers and calcium channel blockers, are used in both hypertension and angina. Several drugs for hypertension are also used in the treatment of heart failure.

Diuretics (see also Ch. 27)

Diuretics, particularly the thiazides, are useful antihypertensives. They cause an initial loss of sodium with a parallel contraction of the blood and extracellular fluid volume. The effect may reach 10% of total body sodium, but it is not maintained. After several months of treatment, the main blood pressure-lowering effect appears to reflect a reduced responsiveness of resistance vessels to endogenous vasoconstrictors, principally noradrenaline/norepinephrine. Although this hyposensitivity may be a consequence of the sodium depletion, thiazides are generally more effective antihypertensive agents than loop diuretics, despite causing less salt loss, and evidence suggests an independent action of thiazides on an unidentified ion channel on vascular smooth muscle cell membranes. Maximal effect on blood pressure is delayed for several weeks and other drugs are best added after this time.

Adverse metabolic effects of thiazides on serum potassium, blood lipids, glucose tolerance and uric acid metabolism led to suggestions that they should be replaced by newer agents without these effects. It is now recognised that unnecessarily high doses of thiazides were used in the past and that with low doses, e.g. bendroflumethiazide 1.25–2.5 mg/day or less (or hydrochlorothiazide 12.5–25 mg), thiazides are both effective and well tolerated. Moreover, they are by far the least costly antihypertensive agents available worldwide and have proved to be the most effective in several outcome trials in preventing the major complications of hypertension, myocardial infarction and stroke. The characteristic reduction in renal calcium excretion induced by thiazides may, in long-term therapy, also reduce the occurrence of hip fractures in older patients and benefit women with postmenopausal osteoporosis.

Vasodilators

Organic nitrates

Organic nitrates (and nitrite) were introduced into medicine in the 19th century.¹ De-nitration in the smooth muscle cell releases nitric oxide (NO), which is the main physiological vasodilator, normally produced by endothelial cells. Nitrodilators (a generic term for drugs that release or mimic the action of NO) activate the soluble guanylate cyclase in vascular smooth muscle cells and cause an increase in intracellular cyclic guanosine monophosphate (GMP) concentrations. This is the second messenger which alters calcium fluxes in the cell, decreases stored calcium and induces relaxation. The result is a generalised dilatation of venules (capacitance vessels) and to a lesser extent of arterioles (resistance vessels), causing a fall of blood pressure that is postural at first; the larger coronary arteries especially dilate. Whereas some vasodilators can ‘steal’ blood away from atheromatous arteries, with their fixed stenoses, to other, healthier arteries, nitrates probably have the reverse effect as a result of their supplementing the endogenous NO. Atheroma is associated with impaired endothelial function, resulting in reduced release of NO and, possibly, its accelerated destruction by the oxidised low-density lipoprotein (LDL) in atheroma (see Ch. 26).

The venous dilatation causes a reduction in venous return and a fall in left ventricular filling pressure with reduced stroke volume, but cardiac output is sustained by the reflex tachycardia induced by the fall in blood pressure.

Pharmacokinetics

The nitrates are generally well absorbed across skin and the mucosal surface of the mouth or gut wall. Nitrates absorbed from the gut are subject to extensive first-pass metabolism in the liver, as shown by the substantially higher doses required by that route compared with sublingual application (and explains why swallowing a sublingual tablet of glyceryl trinitrate terminates its effect). They are first de-nitrated and then conjugated with glucuronic acid. The t_½ periods vary (see below), but for glyceryl trinitrate (GTN) it is 1–4 min. The de-nitration of GTN is in fact genetically determined as the enzyme responsible, a mitochondrial alcohol dehydrogenase, ALDH2, is polymorphic and in subjects carrying a common coding variant (E504K) sublingual GTN has reduced efficacy.²

Tolerance

to the characteristic vasodilator headache comes and goes quickly (hours).³ Ensuring that a continuous steady-state plasma concentration is avoided prevents tolerance. This is easy with occasional use of GTN, but with nitrates having longer t_½ (see below) and sustained-release formulations it is necessary to plan the dosing to allow a low plasma concentration for 4–8 h, e.g. overnight; alternatively, transdermal patches may be removed for a few hours if tolerance is suspected.

Uses

Nitrates are chiefly used to relieve angina pectoris and sometimes left ventricular failure. An excessive fall in blood pressure will reduce coronary flow as well as cause fainting due to reduced cerebral blood flow, so it is important to avoid accidental overdosing. Patients with angina should be instructed on the signs of overdose – palpitations, dizziness, blurred vision, headache and flushing followed by pallor – and what to do about it (see below).

The discovery that coronary artery occlusion by thrombosis is itself ‘stuttering’ – developing gradually over hours – and associated with vasospasm in other parts of the coronary tree has made the use of isosorbide dinitrate (Isoket) by continuous intravenous infusion adjusted to the degree of pain, a logical, and effective, form of analgesia for unstable angina.

Transient relief of pain due to spasm of other smooth muscle (colic) can sometimes be obtained, so that relief of chest pain by nitrates does not prove the diagnosis of angina pectoris.

Nitrates are contraindicated in angina due to anaemia.

Adverse effects

Collapse due to fall in blood pressure resulting from overdose is the commonest side-effect. The patient should remain supine with the legs raised above the head to restore venous return to the heart. The patient should also spit out or swallow the remainder of the tablet.

Nitrate headache, which may be severe, is probably due to the stretching of pain-sensitive tissues around the meningeal arteries resulting from the increased pulsation that accompanies the local vasodilatation. If headache is severe the dose should be halved. Methaemoglobinaemia can occur with heavy dosage.

Interactions

An important footnote to the use of nitrates (and NO dilators generally) has been the marked potentiation of their vasodilator effects observed in patients taking phosphodiesterase (PDE) inhibitors, such as sildenafil (Viagra) and tadalafil (Cialis). These agents target an isoform of PDE (PDE-5) expressed in the blood vessel wall. Other methylaxanthine PDE inhibitors, such as theophylline, do not cause a similar interaction because they are rather weak inhibitors of PDE-5, even at the doses effective in asthma. A number of pericoital deaths reported in patients taking sildenafil have been attributed to the substantial fall in blood pressure that occurs when used with a nitrate. This is an ironic twist for an agent in first-line use in erectile dysfunction that was originally developed as a drug to treat angina.⁴

Glyceryl trinitrate (see also above)

Glyceryl trinitrate (1879) (trinitrin, nitroglycerin, GTN) (t_½ 3 min) is an oily, non-flammable liquid that explodes on concussion with a force greater than that of gunpowder. Physicians meet it mixed with inert substances and made into a tablet, in which form it is both innocuous and fairly stable. But tablets more than 8 weeks old or exposed to heat or air will have lost potency by evaporation and should be discarded. Patients should also be warned to expect the tablet to cause a burning sensation under the tongue if it still contains active GTN. An alternative is to use a nitroglycerin spray (see below), which, formulated as a pressurised liquid GTN, has a shelf-life of at least 3 years.

GTN is the drug of choice in the treatment of an attack of angina pectoris. The tablets should be chewed and dissolved under the tongue, or placed in the buccal sulcus, where absorption is rapid and reliable. Time spent ensuring that patients understand the way to take the tablets, and that the feeling of fullness in the head is harmless, is time well spent. The action begins in 2 min and lasts for up to 30 min. The dose in the standard tablet is 300 micrograms, and 500- or 600-microgram strengths are also available; patients may use up to 6 mg daily in total, but those who require more than two or three tablets per week should take a long-acting nitrate preparation. GTN is taken at the onset of pain and as a prophylactic immediately before any exertion likely to precipitate the pain. Sustained-release buccal tablets are available (Suscard), 1–5 mg. Absorption from the gastrointestinal tract is good, but extensive hepatic first-pass metabolism renders the sublingual or buccal route preferable; an oral metered aerosol that is sprayed under the tongue (nitrolingual spray) is an alternative.

For prophylaxis,

GTN can be given as an oral (buccal, or to swallow, Sustac) sustained-release formulation or via the skin as a patch (or ointment); these formulations can be useful for sufferers from nocturnal angina.⁵

Venepuncture

The ointment can assist difficult venepuncture and a transdermal patch adjacent to an intravenous infusion site can prevent extravasation and phlebitis, and prolong infusion survival.

Isosorbide dinitrate

(Cedocard) (t_½ 20 min) is used for prophylaxis of angina pectoris and for congestive heart failure (tablets sublingual, and to swallow). An intravenous formulation, 500 micrograms/mL (Isoket), is available for use in left ventricular failure and unstable angina.

Isosorbide mononitrate

(Elantan) (t_½ 4 h) is used for prophylaxis of angina (tablets to swallow). Hepatic first-pass metabolism is much less than for the dinitrate so that systemic bio-availability is more reliable.

Pentaerithrityl tetranitrate (Peritrate) (t_½ 8 h) is less efficacious than its metabolite pentaerithrityl trinitrate (t_½ 11 h).

Calcium channel blockers

Calcium is involved in the initiation of smooth muscle and cardiac cell contraction, and in the propagation of the cardiac impulse. Actions on cardiac pacemaker cells and conducting tissue are described in Chapter 25.

Vascular smooth muscle cells

Contraction of these cells requires an influx of calcium across the cell membrane. This occurs through voltage-operated ion channels (VOCs) and this influx is able to trigger further release of calcium from intracellular stores in the sarcoplasmic reticulum. The VOCs have relatively long opening times and carry large fluxes; hence they are usually referred to as L-type channels.⁶ The rise in intracellular free calcium results in activation of the contractile proteins, myosin and actin, with shortening of the myofibril and contraction of smooth muscle. During relaxation calcium is released from the myofibril and either pumped back into the sarcoplasm or lost through Na/Ca exchange at the cell surface.

There are three structurally distinct classes of calcium channel blocker:

• Dihydropyridines (the most numerous).

• Phenylalkylamines (principally verapamil).

• Benzothiazepine (diltiazem).

The differences between their clinical effects can be explained in part by their binding to different parts of the L-type calcium channel. All members of the group are vasodilators, and some have negative inotropic and negative chronotropic effects on the heart via effects on pacemaker cells in the conducting tissue. The attributes of individual drugs are described below.

The therapeutic benefit of the calcium blockers in hypertension and angina is due mainly to their action as vasodilators. Their action on the heart gives non-dihydropyridines an additional role as class 4 antiarrhythmics.

Pharmacokinetics

Calcium channel blockers in general are well absorbed from the gastrointestinal tract and their systemic bio-availability depends on the extent of first-pass metabolism in the gut wall and liver, which varies between the drugs. All undergo metabolism to less active products, predominantly by cytochrome P450 CYP3A4, which is the source of interactions with other drugs by enzyme induction and inhibition. As their action is terminated by metabolism, dose adjustments for patients with impaired renal function are therefore either minor or unnecessary.

Indications for use

• Hypertension: amlodipine, isradipine, nicardipine, nifedipine, verapamil.

• Angina: amlodipine, diltiazem, nicardipine, nifedipine, verapamil.

• Cardiac arrhythmia: verapamil.

• Raynaud's disease: nifedipine.

• Prevention of ischaemic neurological damage following subarachnoid haemorrhage: nimodipine.

Adverse effects

Headache, flushing, dizziness, palpitations and hypotension may occur during the first few hours after dosing, as the plasma concentration is increasing, particularly if the initial dose is too high or increased too rapidly. Ankle oedema may also develop. This is probably due to a rise in intracapillary pressure as a result of the selective dilatation by calcium blockers of the precapillary arterioles. Thus the oedema is not a sign of sodium retention. It is not relieved by a diuretic but disappears after lying flat, e.g. overnight. Bradycardia and arrhythmia may occur, especially with the non-dihydropyridines. Gastrointestinal effects include constipation, nausea and vomiting; palpitation and lethargy may be experienced.

There has been some concern that the shorter-acting calcium channel blockers may adversely affect the risk of myocardial infarction and cardiac death. The evidence is based on case–control studies which cannot escape the possibility that sicker patients, i.e. with worse hypertension or angina, received calcium channel blockade. The safety and efficacy of the class has been strengthened by the recent findings of two prospective comparisons with other antihypertensives.⁷

Interactions

are numerous. Generally, the drugs in this group are extensively metabolised, and there is risk of decreased effect with enzyme inducers, e.g. rifampicin, and increased effect with enzyme inhibitors, e.g. ketoconazole or cimetidine. Conversely, calcium channel blockers decrease the plasma clearance of several other drugs by mechanisms that include delaying their metabolic breakdown. The consequence, for example, is that diltiazem and verapamil cause increased exposure to carbamazepine, quinidine, statins, ciclosporin, metoprolol, theophylline and (HIV) protease inhibitors. Verapamil increases plasma concentration of digoxin, possibly by interfering with its biliary excretion. β-Adrenoceptor blockers may exacerbate atrioventricular (AV) block and cardiac failure. Grapefruit juice raises the plasma concentration of dihydropyridines (except amlodipine) and verapamil, while St John's wort, as an inducer of CYP3A4, can reduce bio-availability of verapamil and dihydropyridines.

Individual calcium blockers

Nifedipine

(t_½ 2 h) is the prototype dihydropyridine. It selectively dilates arteries with little effect on veins; its negative myocardial inotropic and chronotropic effects are much less than those of verapamil. There are sustained-release formulations of nifedipine that permit once-daily dosing, minimising peaks and troughs in plasma concentration so that adverse effects due to rapid fluctuation of concentrations are lessened. Various methods have been used to prolong, and smooth, drug delivery, and bio-equivalence between these formulations cannot be assumed; prescribers should specify the brand to be dispensed. The adverse effects of calcium blockers with a short duration of action may include the hazards of activating the sympathetic system each time a dose is taken. The dose range for nifedipine is 30–90 mg daily. In addition to the adverse effects listed above, gum hypertrophy may occur. Nifedipine can be taken ‘sublingually’, by biting a capsule and squeezing the contents under the tongue. In point of fact, absorption is still largely from the stomach after this manoeuvre, and it should not be used in a hypertensive emergency because the blood pressure reduction is unpredictable and sometimes large enough to cause cerebral ischaemia (see p. 417).

Amlodipine

has a t_½ (40 h) sufficient to permit the same benefits as the longest-acting formulations of nifedipine without requiring a special formulation. Its slow association with L-channels and long duration of action render it unsuitable for emergency reduction of blood pressure where frequent dose adjustment is needed. On the other hand, an occasional missed dose is of little consequence. Amlodipine differs from all other dihydropyridines listed in this chapter in being safe to use in patients with cardiac failure (the PRAISE study).⁸

Verapamil

(t_½ 4 h) is an arterial vasodilator with some venodilator effect; it also has marked negative myocardial inotropic and chronotropic actions. It is given thrice daily as a conventional tablet or daily as a sustained-release formulation. Because of its negative effects on myocardial conducting and contracting cells it should not be given to patients with bradycardia, second- or third-degree heart block, or patients with Wolff–Parkinson–White syndrome to relieve atrial flutter or fibrillation. Amiodarone and digoxin increase the AV block. Verapamil increases plasma quinidine concentration and this interaction may cause dangerous hypotension.

Diltiazem

(t_½ 5 h) is given thrice daily, or once or twice daily if a slow-release formulation is prescribed. It causes less myocardial depression and prolongation of AV conduction than does verapamil but should not be used where there is bradycardia, second- or third-degree heart block or sick sinus syndrome.

Nimodipine

has a moderate cerebral vasodilating action. Cerebral ischaemia after subarachnoid haemorrhage may be partly due to vasospasm; clinical trial evidence indicates that nimodipine given after subarachnoid haemorrhage reduces cerebral infarction (incidence and extent). Although the benefit is small, the absence of any more effective options has led to the routine administration of nimodipine (60 mg every 4 h) to all patients for the first few days after subarachnoid haemorrhage. No benefit has been found in similar trials following other forms of stroke.

Other members

include felodipine, isradipine, lacidipine, lercanidipine, nisoldipine.

Angiotensin-converting enzyme (ACE) inhibitors, angiotensin (AT) II receptor blockers (ARBs) and renin inhibitors

Renin is an enzyme produced by the kidney in response to a number of factors, but principally adrenergic (β₁ receptor) activity and sodium depletion. Renin converts a circulating glycoprotein (angiotensinogen) into the biologically inert angiotensin I, which is then changed by angiotensin-converting enzyme (ACE or kininase II) into the highly potent vasoconstrictor angiotensin II. ACE is located on the luminal surface of capillary endothelial cells, particularly in the lungs; and there are also renin–angiotensin systems in many organs, e.g. brain, heart, the relevance of which is uncertain.

Angiotensin II acts on two G-protein-coupled receptors, of which the angiotensin ‘AT₁’ subtype accounts for all the classic actions of angiotensin. As well as vasoconstriction these include stimulation of aldosterone (the sodium-retaining hormone) production by the adrenal cortex. It is evident that angiotensin II can have an important effect on blood pressure. In addition, it stimulates cardiac and vascular smooth muscle cell growth, probably contributing to the progressive amplification in hypertension once the process is initiated. The AT₂-receptor subtype is coupled to inhibition of muscle growth or proliferation, but appears of minor importance in the adult cardiovascular system. The recognition that the AT₁-receptor subtype is the important target for drugs that antagonise angiotensin II has led, a little confusingly, to alternative nomenclatures for these drugs: angiotensin II blockers, AT₁-receptor blockers or the acronym, ARB. The latter abbreviation is used here for consistency.

Bradykinin (an endogenous vasodilator found in blood vessel walls) is also a substrate for ACE. Potentiation of bradykinin contributes to the blood pressure-lowering action of ACE inhibitors in patients with low-renin causes of hypertension. Either bradykinin or one of the neurokinin substrates of ACE (such as substance P) may stimulate cough (below). The ARBs differ from the ACE inhibitors in having no effect on bradykinin and they do not cause cough. ARBs are slightly more effective than ACE inhibitors at preventing angiotensin II vasoconstriction, because angiotensin II can be generated from angiotensin I by non-ACE enzymes such as chymase. ACE inhibitors are more effective at suppressing aldosterone production in patients with normal or low plasma renin levels.

Uses

Hypertension

The antihypertensive effect of ACE inhibitors, ARBs and renin inhibitors results primarily from vasodilatation (reduction of peripheral resistance) with little change in cardiac output or rate; renal blood flow may increase (desirable). A fall in aldosterone production may also contribute to the blood-pressure-lowering action of ACE inhibitors. Both classes slow progression of glomerulopathy. Whether the long-term benefit of these drugs in hypertension exceeds that to be expected from blood pressure reduction alone remains controversial.

ACE inhibitors, ARBs and renin inhibitors are most useful in hypertension when the raised blood pressure results from excess renin production, e.g. renovascular hypertension, or where concurrent use of another drug (diuretic or calcium blocker) renders the blood pressure renin dependent. The fall in blood pressure can be rapid, especially with short-acting ACE inhibitors, and low initial doses of these should be used in patients at risk: those with impaired renal function or suspected cerebrovascular disease. These patients may be advised to omit any concurrent diuretic treatment for a few days before the first dose. The antihypertensive effect increases progressively over weeks with continued administration (as with other antihypertensives) and the dose may be increased at intervals of 2 weeks.

Cardiac failure

(see p. 406). ACE inhibitors have a useful vasodilator and diuretic-sparing (but not diuretic-substitute) action that is critical to the treatment of all grades of heart failure. Mortality reduction here may result from their being the only vasodilator that does not reflexly activate the sympathetic system.

The ARBs are at least as effective as ACE inhibitors in patients with heart failure and they can be substituted if patients are intolerant of an ACE inhibitor. Based on the Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM) trial, they may also benefit patients with heart failure and a low ejection fraction when added to treatment with a β-blocker and ACE inhibitor.⁹

Diabetic nephropathy

In patients with type I (insulin-dependent) diabetes, hypertension often accompanies the diagnosis of frank nephropathy, and aggressive blood pressure control is essential to slow the otherwise inexorable decline in renal function that follows. ACE inhibitors appear to have a specific renoprotective effect, probably because of the role of angiotensin II in driving the underlying glomerular hyperfiltration.¹⁰ These drugs are now first-line treatment for hypertensive type I diabetics, although most patients will need a second or third agent to reach the rigorous blood pressure targets for this condition (see below). Their role in preventing the progression of the earliest manifestation of renal damage, microalbuminuria, is more complicated. Here the evidence suggests that ACE inhibitors do not slow the incidence of microalbuminuria in type I diabetics and an ARB may actually substantially increase it.¹⁰ In contrast, an ACE inhibitor halves the incidence of microalbuminuria in type 2 diabetics with hypertension and normal renal function on follow-up. A parallel group on verapamil did not show any protection confirming that inhibition of the renin–angiotensin–aldosterone (RAAS) axis is required for this effect, not simply lowering the blood pressure.¹¹ For hypertensive type 2 diabetics with established nephropathy, both ARBs and ACE inhibitors protect against a decline in renal function and reduce macroproteinuria.¹⁰ The evidence suggests they are interchangeable in this respect. Whether combining the two classes of drugs (‘dual block’) confers further protection of renal function is not yet resolved, although ‘dual block’ does produce substantially better urine protein sparing than either agent alone.¹⁰

Myocardial infarction (MI)

Following a myocardial infarction, the left ventricle may fail acutely from the loss of functional tissue or in the long term from a process of ‘remodelling’ due to thinning and enlargement of the scarred ventricular wall (see p. 425). Angiotensin II plays a key role in both of these processes and an ACE inhibitor given after MI markedly reduces the incidence of heart failure. The effect is seen even in patients without overt signs of cardiac failure, but who have low left ventricular ejection fractions (< 40%) during the convalescent phase (3–10 days) following the MI. Such patients receiving captopril in the SAVE trial,¹² had a 37% reduction in progressive heart failure over the 60-month follow-up period compared with placebo. The benefits of ACE inhibition after MI are additional to those conferred by thrombolysis, aspirin and β-blockers. ARBs also prevent remodelling and heart failure in post-MI patients, but there is no additional benefit from ‘dual blockade’.¹³

Cautions

Certain constraints apply to the use of ACE inhibitors:

• Heart failure: severe hypotension may result in patients taking diuretics, or who are hypovolaemic, hyponatraemic, elderly, have renal impairment or with systolic blood pressure of less than 100 mmHg. A test dose of captopril 6.25 mg by mouth may be given because its effect lasts for only 4–6 h. If tolerated, the preferred long-acting ACE inhibitor may then be initiated in low dose.

• Renal artery stenosis (RAS, whether unilateral, bilateral renal or suspected from the presence of generalised atherosclerosis): an ACE inhibitor may cause renal failure and is contraindicated. ARBs are not necessarily any safer in this situation, because angiotensin II-mediated constriction of the efferent arteriole is thought to be crucial to the maintenance of glomerular perfusion in RAS.

• Aortic stenosis/left ventricular outflow tract obstruction: an ACE inhibitor may cause severe, sudden hypotension and, depending on severity, is relatively or absolutely contraindicated.

• Pregnancy represents an absolute contraindication (see below).

• Angioedema may result (see below).

Adverse effects

ACE inhibitors:

• Persistent dry cough occurs in 10–15% of patients.

• Urticaria and angioedema (less than 1 in 100 patients) are much rarer, occurring usually in the first weeks of treatment. The angioedema varies from mild swelling of the tongue to life-threatening tracheal obstruction, when subcutaneous adrenaline/epinephrine should be given. The basis of the reaction is probably pharmacological rather than allergic, due to reduced breakdown of bradykinin.

• Impaired renal function may result from reduced glomerular filling pressure, systemic hypotension or glomerulonephritis, and plasma creatinine levels should be checked before and during treatment.

• Hyponatraemia may develop, especially where a diuretic is also given; clinically significant hyperkalaemia (see effect on aldosterone above) is confined to patients with impaired renal function.

• ACE inhibitors cause major malformations in the first trimester and are fetotoxic in the second trimester, causing reduced renal perfusion, hypotension, oligohydramnios and fetal death (see Pregnancy hypertension, p. 417).

• Neutropenia and other blood dyscrasias occur. Other reported reactions include rashes, taste disturbance (dysguesia), musculoskeletal pain, proteinuria, liver injury and pancreatitis.

ARBs

are contraindicated in pregnancy as are ACE inhibitors, but avoid the other complications of these drugs – especially the cough and angioedema. They are, in fact, the only antihypertensive drug class for which there is no ‘typical’ side-effect.

Interactions

Hyperkalaemia can result from use with potassium-sparing diuretics. Renal clearance of lithium is reduced and toxic concentrations of plasma lithium may follow. Severe hypotension can occur with diuretics (above), and with chlorpromazine, and possibly other phenothiazines.

Individual drugs

Captopril

(Capoten) has a t_½ of 2 h and is partly metabolised and partly excreted unchanged; adverse effects are more common when renal function is impaired; it is given twice or thrice daily. Captopril is the shortest acting of the ACE inhibitors, one of the few that is itself active by mouth, not requiring de-esterification after absorption.

Enalapril

(Innovace) is a prodrug (t_½ 35 h) that is converted to the active enalaprilat (t_½ 10 h). Effective 24-h control of blood pressure probably requires twice-daily administration.

Other members

include cilazapril, fosinopril, imidapril, lisinopril, moexipril, perindopril, quinapril, ramipril and trandolapril. Of these, lisinopril has a marginally longer t_½ than enalapril (it is the lysine analogue of enalaprilat), probably justifying its popularity as a once-daily ACE inhibitor. Some of the others are longer acting, with quinapril and ramipril also having a higher degree of binding to ACE in vascular tissue. The clinical significance of these differences is disputed. In the Heart Outcomes Prevention Evaluation (HOPE) study of 9297 patients, ramipril reduced, by 20–30%, the rates of death, myocardial infarction and stroke in a broad range of high-risk patients who were not known to have a low ejection fraction or heart failure.¹⁴ The authors considered (probably erroneously) that the results could not be explained entirely by blood pressure reduction.

Losartan

was the first ARB to be licensed in the UK. It is a competitive blocker with a non-competitive active metabolite. The drug has a short t_½ (2 h) but the metabolite is much longer lived (t_½ 10 h), permitting once-daily dosing.

Other ARBs

in clinical use include candesartan, eprosartan, irbesartan, telmisartan, valsartan and olmesartan. Some of these may be marginally more effective than losartan at lowering blood pressure, but few if any comparisons have been performed at maximal dose of each drug. Losartan is generally used in combination with hydrochlorothiazide. In a landmark study this combination was 25% more effective than atenolol plus hydrochlorothiazide in preventing stroke.¹⁵

This class of drug is very well tolerated; in clinical trials the side-effect profiles are indistinguishable or even better than those of placebo. Unlike the ACE inhibitors they do not produce cough, and are a valuable alternative for the 10–15% of patients who thereby discontinue their ACE inhibitor. ARBs are used to treat hypertension, left ventricular (LV) failure after MI and established heart failure. With the possible exception of chronic heart failure, they do not appear to be superior to ACE inhibitors.

The cautions listed for the use of ACE inhibitors (above) apply also to AT₁-receptor blockers.

Renin inhibitors

share the benefit of ARBs over ACE inhibitors in terms of cough and angioedema. By implication with other agents targeting the RAAS they should not be used in pregnancy.

Interactions

Hyperkalaemia can result from use with potassium-sparing diuretics. Severe hypotension can occur on first dosing if given after agents that increase circulating renin levels such as diuretics (especially loop diuretics) and potent vasodilators.

Individual drugs

Aliskiren

is the only orally active non-peptide renin inhibitor licensed (t_½ 40 h). The agent is well tolerated apart from dose-dependent diarrhoea; it is not clear if this is a class side-effect. It produces additive effects on blood pressure with ACE inhibitors, ARBs, calcium channel blockers and thiazide diuretics. There are currently no outcome data in terms of preventing hypertension-related cardiovascular events, so it should be reserved for inhibiting the RAAS where an ACE inhibitor or ARB is not tolerated.¹⁶

Other vasodilators

Several older drugs are powerfully vasodilating, but precluded from routine use in hypertension by their adverse effects. Minoxidil and nitroprusside still have special indications.

Minoxidil

is a vasodilator selective for arterioles rather than for veins, similar to diazoxide and hydralazine. Like the former, it acts through its sulphate metabolite as an adenosine triphosphate (ATP)-dependent potassium channel opener. It is highly effective in severe hypertension, but in common with all potent arterial vasodilators its hypotensive action is accompanied by a compensatory baroreceptor-mediated sympathetic discharge, causing tachycardia and increased cardiac output. There is also renin release with secondary salt and water retention, which antagonises the hypotensive effect (so-called ‘tolerance’ on long-term use). Therefore, it is used in combination with a β-blocker and loop diuretic (as is hydralazine; see below). Hypertrichosis is perhaps the most notorious side-effect of minoxidil. The hair growth is generalised when taken orally and, although a cosmetic problem in women, it has been exploited as a 2–5% topical solution for the treatment of male-pattern baldness (Regaine).

Sodium nitroprusside

is a highly effective antihypertensive agent when given intravenously. Its effect is almost immediate and lasts for 1–5 min. Therefore it must be given by a precisely controllable infusion. It dilates both arterioles and veins, which would cause collapse were the patient to stand up, e.g. for toilet purposes. There is a compensatory sympathetic discharge with tachycardia and tachyphylaxis to the drug.

The action of nitroprusside is terminated by metabolism within erythrocytes. Specifically, electron transfer from haemoglobin iron to nitroprusside yields methaemoglobin and an unstable nitroprusside radical. This breaks down, liberating cyanide radicals capable of inhibiting cytochrome oxidase (and thus cellular respiration). Fortunately, most of the cyanide remains bound within erythrocytes but a small fraction does diffuse out into the plasma and is converted to thiocyanate. Hence, monitoring plasma thiocyanate concentrations during prolonged (days) nitroprusside infusion is a useful marker of impending systemic cyanide toxicity. Poisoning may be obvious as a progressive metabolic acidosis, or may manifest as delirium or psychotic symptoms. Intoxicated subjects are also reputed to have the characteristic bitter almond smell of hydrogen cyanide. Clearly nitroprusside infusion must be used with caution, and outside specialist units it may be safer overall to choose another more familiar drug.

Sodium nitroprusside is used in hypertensive emergencies, refractory heart failure and for controlled hypotension in surgery. An infusion¹⁷ may begin at 0.3–1.0 micrograms/kg/min, and control of blood pressure is likely to be established at 0.5–6.0 micrograms/kg/min; close monitoring of blood pressure is mandatory, usually by direct arterial monitoring; rate changes of infusion may be made every 5–10 min.

Diazoxide

mimics the actions of other ATP-dependent potassium channel openers or KCOs. The t_½ is 36 h. Diazoxide was used as an intravenous bolus for the emergency treatment of severe hypertension, but this use is now obsolete.

During its use as an antihypertensive it was noted that it caused hyperglycaemia because, unlike other KCOs, it can activate the sulphonylurea-sensitive form of the ATP-potassium channel in the pancreatic islet cells switching off insulin release. Hence it is used in patients with chronic hypoglycaemia from excess endogenous insulin secretion, either from an islet cell tumour or islet cell hyperplasia. Long-term use can cause the same problems of hair growth seen with minoxidil, albeit less consistently.

Hydralazine

is now little used for hypertension except for that related to pregnancy (owing to its established lack of teratogenicity), but it may have a role as a vasodilator (plus nitrates) in heart failure. It reduces peripheral resistance by directly relaxing arterioles, with negligible effect on veins; the mechanism of vasorelaxation is unclear. The t_½ is 1 h.

In most hypertensive emergencies (except for dissecting aneurysm) hydralazine 5–20 mg i.v. may be given over 20 min, when the maximum effect will be seen in 10–80 min; it can be repeated according to need and the patient transferred to oral therapy within 1–2 days.

Prolonged use of hydralazine at doses above 50 mg/day may cause a systemic lupus-like syndrome, more commonly in white than in black races, and in those with the slow acetylator phenotype.

Three other vasodilators find a role outside hypertension:

Nicorandil

is effective through two actions: it acts as a nitrate by activating cyclic GMP (see above) but also opens the ATP-dependent potassium channel to allow potassium efflux and hyperpolarisation of the membrane, which reduces calcium ion entry and induces muscular relaxation. It is indicated for use in angina, where it has similar efficacy to β-blockade, nitrates or calcium channel blockade. It is administered orally and is an alternative to nitrates when tolerance is a problem, or to the other classes when these are contraindicated by asthma or cardiac failure. Adverse effects to nicorandil are similar to those of nitrates, with headache reported in 35% of patients. It is the only antianginal drug for which at least one trial has demonstrated a beneficial influence on outcome.¹⁸

Papaverine

is an alkaloid present in opium, but is structurally unrelated to morphine. It inhibits phosphodiesterase and its principal action is to relax smooth muscle throughout the body, especially in the vascular system. It is occasionally injected into an area where local vasodilatation is desired, especially into and around arteries and veins to relieve spasm during vascular surgery and when setting up intravenous infusions. It is also used to treat male erectile dysfunction (see p. 465).

Alprostadil

is a stable form of prostaglandin E₁. It is effective in psychogenic and neuropathic penile erectile dysfunction by direct intracorporeal injection (see p. 465) and is used intravenously to maintain patency of the ductus arteriosus in the newborn with congenital heart disease.

Vasodilators in heart failure

See page 394.

Vasodilators in peripheral vascular disease

The aim has been to produce peripheral arteriolar vasodilatation without a concurrent significant drop in blood pressure, so that an increased blood flow in the limbs will result. Drugs are naturally more useful in patients in whom the decreased flow of blood is due to spasm of the vessels (Raynaud's phenomenon) than where it is due to organic obstructive changes that may make dilatation in response to drugs impossible (arteriosclerosis, intermittent claudication, Buerger's disease).

Intermittent claudication

Patients should ‘stop smoking and keep walking’, i.e. take frequent exercise within their capacity. Other risk factors should be treated vigorously, especially hypertension and hyperlipidaemia. Patients should also receive low-dose aspirin (75 mg daily) as an antiplatelet agent. Most patients with intermittent claudication succumb to ischaemic or cerebrovascular disease, and therefore a major objective of treatment should be prevention of such outcomes. Vasodilators such as naftidrofuryl (Praxilene) and pentoxifylline (Trental) increase blood flow to skin rather than muscle; they have been used successfully in the treatment of venous leg ulcers (varicose and traumatic). A trial of these drugs for intermittent claudication is worthwhile but they should be withdrawn if there is no benefit within a few weeks.

Naftidrofuryl has several actions. It is classed as a metabolic enhancer because it activates the enzyme succinate dehydrogenase, increasing the supply of ATP and reducing lactate concentrations in muscle. It also blocks 5HT₂receptors and inhibits serotonin-induced vasoconstriction and platelet aggregation.

Pentoxifylline is thought to improve oxygen supply to ischaemic tissue by improving erythrocyte deformability and reducing blood viscosity, in part by reducing plasma fibrinogen. Neither of these drugs is a direct vasodilator, as is the third drug used for intermittent claudication, inositol nicotinate. The evidence in favour of any benefit is stronger for the first two, for which meta-analyses provide some evidence of efficacy (increase in walking distance). Most vasodilators act selectively on healthy blood vessels, causing a diversion (‘steal’) of blood from atheromatous vessels.

Night cramps occur in the disease and quinine has a somewhat controversial reputation in their prevention. Nevertheless, meta-analysis of six double-blind trials of nocturnal cramps (not necessarily associated with peripheral vascular disease) shows that the number, but not severity or duration of episodes, is reduced by a night-time dose.¹⁹ The benefit may not be seen for 4 weeks.

Raynaud's phenomenon

may be helped by nifedipine, reserpine (effectively an α-adrenoceptor blocker, in low doses) and also by topical glyceryl trinitrate; indeed any vasodilator is worth trying in resistant cases; enalapril (an ACE inhibitor) seems to lack efficacy. In severe cases, especially patients with ulceration, intermittent infusions over several hours of the endogenous vasodilator, epoprostenol (prostacyclin), achieve long-lasting improvements in symptoms.

β-Adrenoceptor blockers exacerbate peripheral vascular disease and Raynaud's phenomenon by reducing perfusion of a circulation that is already compromised. Switching to a β₁-selective blocker is unhelpful, because the adverse effect is due to reduced cardiac output rather than unopposed α-receptor-induced vasoconstriction.

Adrenoceptor-blocking drugs

Adrenoceptor-blocking drugs occupy the adrenoceptor in competition with adrenaline/epinephrine and noradrenaline/norepinephrine (and other sympathomimetic amines) whether released from stores in nerve terminals or injected. There are two principal classes of adrenoceptor, α and β: for details of receptor effects see Table 23.1.

α-Adrenoceptor-blocking drugs

There are two main subtypes of α adrenoceptor:

• ‘Classic’ α₁ adrenoceptors, on the effector organ (post-synaptic), mediate vasoconstriction.

• α₂ Adrenoceptors are present both on some effector tissues (post-synaptic) and on the nerve ending (pre-synaptic). The pre-synaptic receptors (or autoreceptors) inhibit release of chemotransmitter (noradrenaline/norepinephrine), i.e. they provide negative feedback control of transmitter release. They are also present in the CNS.

The first generation of α-adrenoceptor blockers were imidazolines (e.g. phentolamine), which blocked both α₁ and α₂ receptors. When subjects taking such a drug stand from the lying position or take exercise, the sympathetic system is physiologically activated (via baroreceptors). The normal vasoconstrictive (α₁) effect (to maintain blood pressure) is blocked by the drug and the failure of this response causes further sympathetic activation and the release of additional transmitter. This would normally be restrained by negative feedback through α₂ autoreceptors, but these are blocked too.

The β adrenoceptors, however, are not blocked and the excess transmitter released at adrenergic endings is free to act on them, causing a tachycardia that may be unpleasant. Hence, non-selective α-adrenoceptor blockers are not used on their own in hypertension.

An α₁-adrenoceptor blocker that spares the α₂ receptor, so that negative feedback inhibition of noradrenaline/norepinephrine release is maintained, is more useful in hypertension (less tachycardia and postural and exercise hypotension); prazosin is such a drug (see below).

For use in prostatic hypertrophy, see page 619.

Uses of α-adrenoceptor-blocking drugs

• Hypertension

   essential: doxazosin, labetalol

   phaeochromocytoma: phenoxybenzamine; phentolamine (for crises).

• Peripheral vascular disease.

• Benign prostatic hypertrophy (to relax capsular smooth muscle that may contribute to urinary obstruction).

Adverse effects

The converse of the benefit in the treatment of prostatism is the adverse effect of urinary incontinence in women. Other adverse effects of α-adrenoceptor blockade are postural hypotension, nasal stuffiness, red sclerae and, in the male, failure of ejaculation. They may also exacerbate symptoms of angina.²⁰ Effects peculiar to each drug are mentioned below.

Notes on individual drugs

Prazosin

blocks postsynaptic α₁ receptors but not presynaptic α₂ autoreceptors. It has a curious adverse ‘first-dose effect’: within 2 h of the first dose (rarely after another) there may be a brisk fall in blood pressure sufficient to cause loss of consciousness. Hence the first dose should be small (0.5 mg) and given before going to bed. This unwanted effect, together with a rather short duration of action (t_½ 3 h) has led to a sharp decline in its use.

Doxazosin

(t_½ 8 h) was the first α-adrenoceptor blocker suitable for once-daily prescribing. The first-dose effect is also much less marked, although it is still advisable to start patients at a lower dose than is intended for maintenance. It is convenient, for instance, to prescribe 1 mg daily, increasing after 1 week to double this dose without repeating the blood pressure measurement at this stage. A slow-release formulation, doxazosin XL, can be started at the maintenance dose of 4 mg daily.

Other α-adrenoceptor blockers used for prostatic symptoms are alfuzosin and terazosin.

Indoramin

is an older α₁-blocker, which is a less useful antihypertensive but still used for prostatic symptoms. It is taken twice or thrice daily.

Phentolamine

is a non-selective α-adrenoceptor blocker. It is given intravenously for brief effect in adrenergic hypertensive crises, e.g. phaeochromocytoma or the monoamine oxidase inhibitor–sympathomimetic interaction (‘cheese reaction’). In addition to α-receptor block it has direct vasodilator and cardiac inotropic actions. The dose for hypertensive crisis is 2–5 mg i.v. repeated as necessary (in minutes to hours). This is not a reliable diagnostic test for phaeochromocytoma!

Phenoxybenzamine

is an irreversible non-selective α-adrenoceptor-blocking drug whose effects may last for 2 days or longer. The daily dose must therefore be increased slowly. It is impossible to reverse the circulatory effects by secreting noradrenaline/norepinephrine or other sympathomimetic drugs because its effects are insurmountable. This makes it the preferred α-blocker for treating phaeochromocytoma (see p. 419).

It is wise to observe the effects of a single test dose closely before starting regular administration.

Indigestion and nausea can occur with oral therapy, which is best given with food.

Moxisylyte

(thymoxamine) is a non-selective α-blocker for which Raynaud's phenomenon is the only extant indication.

Labetalol

has both α- and β-receptor-blocking actions that are due to different isomers (see β-adrenoceptor block, below). Its parenteral preparation is valuable in the treatment of hypertension emergencies (see p. 416).

Chlorpromazine and amitriptyline,

among their other actions, can both produce clinically significant α-blockade, which may be sufficient to cause postural hypotension and falls in the elderly.

β-Adrenoceptor-blocking drugs

Actions

These drugs selectively block the β-adrenoceptor effects of noradrenaline/norepinephrine and adrenaline/epinephrine. They may be pure antagonists or may have some agonist activity in addition (when they are described as partial agonists).

Intrinsic heart rate

Sympathetic activity (through β₁ adrenoceptors) accelerates, and parasympathetic activity (through muscarinic M₂ receptors) slows, the heart. If the sympathetic and the parasympathetic drives to the heart are simultaneously and adequately blocked by a β-adrenoceptor blocker plus atropine, the heart will beat at its ‘intrinsic’ rate. The intrinsic rate at rest is usually about 100 beats/min, as opposed to the usual rate of 80 beats/min, i.e. normally there is parasympathetic vagal ‘tone’, which decreases with age.

The cardiovascular effects of β-adrenoceptor block depend on the amount of sympathetic tone present. The chief effects result from reduction of sympathetic drive:

• Reduced automaticity (heart rate).

• Reduced myocardial contractility (rate of rise of pressure in the ventricle).

• Reduced renin secretion from the juxtaglomerular apparatus in the renal cortex.

With reduced rate the cardiac output per minute is reduced and the overall cardiac oxygen consumption falls. The results are more evident on the response to exercise than at rest. With acute administration of a pure β-adrenoceptor blocker, i.e. one with no intrinsic sympathomimetic activity (ISA), peripheral vascular resistance tends to rise. This is probably a reflex response to the reduced cardiac output, but also occurs because the β-adrenoceptor (vasoconstrictor) effects are no longer partially opposed by β₂-adrenoceptor (dilator) effects; peripheral flow is reduced. With chronic use peripheral resistance returns to about pre-treatment levels or a little below, varying according to presence or absence of ISA. But peripheral blood flow remains reduced. The cold extremities that accompany chronic therapy are probably due chiefly to reduced cardiac output with reduced peripheral blood flow, rather than to the blocking of peripheral (β₂) dilator receptors.

Hepatic blood flow may be reduced by as much as 30%, prolonging the t_½ of the lipid-soluble drugs whose metabolism is limited by hepatic blood flow, i.e. whose first-pass metabolism is so extensive that it is actually limited by the rate of blood delivery to the liver; these include propranolol, verapamil and lidocaine, which may be used concomitantly for cardiac arrhythmias.

Effects

Within hours of starting treatment with a β-blocker, blood pressure starts to fall. This reflects the acute effect on cardiac output (heart rate and contractility) but this is not sustained and on chronic administration the blockade of renin secretion appears to be the main cause of blood pressure reduction. An additional contributor may be the two- to three-fold increase in natriuretic peptide secretion caused by β-blockade.

A substantial advantage of β-blockade in hypertension is that physiological stresses such as exercise, upright posture and high environmental temperature are not accompanied by hypotension, as they are with agents that interfere with α-adrenoceptor-mediated homeostatic mechanisms. With β-blockade these necessary adaptive α-receptor constrictor mechanisms remain intact.

At first sight the cardiac effects might seem likely to be disadvantageous rather than advantageous, and indeed maximum exercise capacity is reduced. But the heart has substantial functional reserves so that use may be made of the desired properties in the diseases listed below, e.g. angina, without inducing heart failure. Indeed, β-blockade is now routine practice in patients with established mild to moderate heart failure and usually safe provided uptitration is slow. But heart failure can occur in patients with seriously diminished cardiac reserve.

For the effect on plasma potassium concentration, see page 426.

β-Adrenoceptor selectivity

Some β-adrenoceptor blockers have higher affinity for cardiac β₁ receptors than for cardiac and peripheral β₂ receptors (Table 24.1). The ratio of the amount of drug required to block the two receptor subtypes is a measure of the selectivity of the drug. (See note to Table 23.1, p. 384, regarding the use of the terms ‘β₁ selective’ and ‘cardioselective’.) The question is whether the differences between selective and non-selective β-blockers confer clinical advantages. In theory β₁-blockers are less likely to cause bronchoconstriction, but in practice few available β₁-blockers are sufficiently selective to be safely recommended in asthma. Bisoprolol and nebivolol may be exceptions that can be tried at low doses in patients with mild asthma and a strong indication for β-blockade. There are unlikely ever to be satisfactory safety data to support such use. The main practical use of β₁-selective blockade is in diabetics, where β₂ receptors mediate both the symptoms of hypoglycaemia and the counter-regulatory metabolic responses that reverse the hypoglycaemia.

Table 24.1 β-Adrenoceptor blocking drugs: properties at therapeutic doses

Some β-blockers (antagonists) also have agonist action or ISA, i.e. they are partial agonists. These agents cause less fall in resting heart rate than do the pure antagonists and may thus be less effective in severe angina pectoris where reduction of heart rate is particularly important. The fall in cardiac output may be less, and fewer patients may experience unpleasantly cold extremities. Intermittent claudication may be worsened by β-blockade whether or not there is partial agonist effect. Both classes of drug can precipitate heart failure, and indeed no important difference is to be expected because patients with heart failure already have high sympathetic drive (but note that β-blockade can be used to treat cardiac failure, p. 406).

Abrupt withdrawal may be less likely to lead to a rebound effect if there is some partial agonist action, as there may be less up-regulation of receptors, such as occurs with prolonged receptor block.

Some β-blockers have a membrane-stabilising (quinidine-like or local anaesthetic) effect, a property that is unimportant at clinical doses but relevant in overdose (see below). Additionally, agents having this effect will anaesthetise the eye (undesirable) if applied topically for glaucoma (timolol is used in the eye and does not have this action).

The ankle jerk relaxation time is prolonged by β₂-adrenoceptor block, which may be misleading if the reflex is being relied on in diagnosis and management of hypothyroidism.

Pharmacokinetics

The plasma concentration of a β-adrenoceptor blocker may have a complex relationship with its effect, for several reasons. First-order kinetics usually apply to elimination of drug from plasma, but the decline in receptor block is zero order. The practical application is important: within 4 h of giving propranolol 20 mg i.v. the plasma concentration falls by 50%, but the receptor block (as measured by exercise-induced tachycardia) falls by only 35%. The relationship between the concentration of the parent drug in plasma and its effect is further obscured if pharmacologically active metabolites are also present. Additionally, for some of the lipid-soluble β-blockers, especially timolol, plasma t_½ may not reflect the duration of β-blockade, because the drug remains bound to the tissues near the receptor when the plasma concentration is negligible.

Most β-adrenoceptor blockers can be given orally once daily in either ordinary or sustained-release formulations because the t_½ of the pharmacodynamic effect exceeds the elimination t_½ of the parent substance in the blood.

Lipid-soluble

agents are extensively metabolised (hydroxylated, conjugated) to water-soluble substances that can be eliminated by the kidney. Plasma concentrations of drugs subject to extensive hepatic first-pass metabolism vary greatly between subjects (up to 20-fold) because the process itself is dependent on two highly variable factors: speed of absorption and hepatic blood flow, with the latter being the rate-limiting factor.

Lipid-soluble agents readily cross cell membranes and so have a high apparent volume of distribution. They also readily enter the central nervous system (CNS), e.g. propranolol reaches concentrations in the brain 20 times those of the water-soluble atenolol.

Water-soluble

agents show more predictable plasma concentrations because they are less subject to liver metabolism, being excreted unchanged by the kidney; thus their half-lives are greatly prolonged in renal failure, e.g. atenolol t_½ is increased from 7 to 24 h. Drugs (of any kind) having a long t_½ and an action terminated by renal elimination are best avoided in patients with renal disease. Water-soluble agents are less widely distributed and may have a lower incidence of effects attributed to penetration of the CNS, e.g. nightmares.

• The most lipid-soluble agents are propranolol, metoprolol, oxprenolol and labetalol.

• The least lipid-soluble (and most water-soluble) agents are atenolol, sotalol and nadolol.

• Others are intermediate.

Classification of β-adrenoceptor-blocking drugs

• Pharmacokinetic: lipid soluble, water soluble, see above.

• Pharmacodynamic (see Table 24.1). The associated properties (partial agonist action and membrane-stabilising action) have only minor clinical importance with current drugs at doses ordinarily used and may be insignificant in most cases. But it is desirable that they be known, for they can sometimes matter and they may foreshadow future developments.

β-Adrenoceptor blockers not listed in Table 24.1 include:

• Non-selective: carteolol, bufuralol.

• β₁-receptor selective: betaxolol, esmolol (ultra-short acting: minutes).

• β- and α-receptor block: bucindolol.

Uses of β-adrenoceptor-blocking drugs

Cardiovascular uses:

Angina pectoris: β-blockade reduces cardiac work and oxygen consumption.

Hypertension: β-blockade reduces renin secretion and cardiac output; there is little interference with homeostatic reflexes.

Cardiac tachyarrhythmias: β-blockade reduces drive to cardiac pacemakers: subsidiary properties (see Table 25.1, p. 430) may also be relevant.

Myocardial infarction and β-adrenoceptor blockers: there are two modes of use that reduce acute mortality and prevent recurrence: the so-called ‘cardioprotective’ effect.

• Early use within 6 h (or at most 12 h) of onset (intravenously for 24 h then orally for 3–4 weeks). Benefit has been demonstrated only for atenolol. Cardiac work is reduced, resulting in a reduction in infarct size by up to 25% and protection against cardiac rupture. Surprisingly, tachyarrhythmias are not less frequent, perhaps because the cardiac β₂ receptor is not blocked by atenolol. Maximum benefit is in the first 24–36 h, but mortality remains lower for up to 1 year. Contraindications to early use include bradycardia (< 55 beats/min), hypotension (systolic < 90 mmHg) and left ventricular failure. A patient already taking a β-blocker may be given additional doses.

• Late use for secondary prevention of another myocardial infarction. The drug is started between 4 days and 4 weeks after the onset of the infarct and is continued for at least 2 years.²¹

• Choice of drug. The agent should be a pure antagonist, i.e. without ISA.

Aortic dissection and after subarachnoid haemorrhage: by reducing force and speed of systolic ejection (contractility) and blood pressure.

Obstruction of ventricular outflow where sympathetic activity occurs in the presence of anatomical abnormalities, e.g. Fallot's tetralogy (cyanotic attacks): hypertrophic subaortic stenosis (angina); some cases of mitral valve disease.

Hepatic portal hypertension and oesophageal variceal bleeding: reduction of portal pressure (see p. 548).

Cardiac failure (see also Ch. 25): there is now clear evidence from prospective trials that β-blockade is beneficial in terms of mortality for patients with all grades of moderate heart failure. Data support the use of both non-selective (carvedilol, α-blocker as well) and β₁-selective (metoprolol and bisoprolol) agents. Survival benefit exceeds that provided by ACE inhibitors over placebo. The negative inotropic effects can still be significant, so the starting dose is low (e.g. bisoprolol 1.25 mg orally daily in the morning or carvedilol 3.125 mg twice daily, with food) and may be tolerated only with additional antifailure therapy, e.g. diuretic.

Endocrine uses

Hyperthyroidism: β-blockade reduces unpleasant symptoms of sympathetic overactivity; there may also be an effect on metabolism of thyroxine (peripheral de-iodination from T₄ to T₃). A non-selective agent (propranolol) is preferred to counteract both the cardiac (β₁ and β₂) effects, and tremor (β₂).

Phaeochromocytoma: blockade of β-agonist effects of circulating catecholamines always in combination with adequate α-adrenoceptor block. Only small doses of a β-blocker are required.

Other uses:

• Central nervous system:

   anxiety with somatic symptoms (non-selective β-blockade may be more effective than β₁-selective)

   migraine prophylaxis (see p. 292)

   essential tremor, some cases

   alcohol and opioid acute withdrawal symptoms.

• Eyes:

   glaucoma: carteolol, betaxolol, levobunolol and timolol eye drops act by altering production and outflow of aqueous humour.

Adverse reactions due to β-adrenoceptor blockade

Bronchoconstriction (β₂ receptor) occurs as expected, especially in patients with asthma²² (in whom even eye drops are dangerous²³). In elderly chronic bronchitics there may be gradually increasing bronchoconstriction over weeks (even with eye drops). Plainly, risk is greater with non-selective agents, but β₁-receptor-selective members can still have significant β₂-receptor occupancy and may precipitate asthma.

Cardiac failure may arise if cardiac output is dependent on high sympathetic drive (but β-blockade can be introduced at very low dose to treat cardiac failure; see above). The degree of heart block may be made dangerously worse.

Incapacity for vigorous exercise due to failure of the cardiovascular system to respond to sympathetic drive.

Hypotension when the drug is given after myocardial infarction.

Hypertension may occur whenever blockade of β receptors allows pre-existing α effects to be unopposed, e.g. phaeochromocytoma.

Reduced peripheral blood flow, especially with non-selective members, leading to cold extremities which, rarely, can be severe enough to cause necrosis; intermittent claudication may be worsened.

Reduced blood flow to liver and kidneys, reducing metabolism and biliary and renal elimination of drugs, is liable to be important if there is hepatic or renal disease.

Hypoglycaemia: β₂ receptors mediate both the symptoms of hypoglycaemia and the counter-regulatory metabolic responses that restore blood glucose. Non-selective β-blockers, by blocking β₂ receptors, impair the normal sympathetic-mediated homeostatic mechanism for maintaining blood glucose levels, and recovery from hypoglycaemia is delayed; this is important in diabetes and after substantial exercise. Further, as α adrenoceptors are not blocked, hypertension (which may be severe) can occur as the sympathetic system discharges in an ‘attempt’ to reverse the hypoglycaemia. The symptoms of hypoglycaemia, in so far as they are mediated by the sympathetic nervous system (anxiety, palpitations), will not occur, except (cholinergic) sweating, and the patient may miss the warning symptoms of hypoglycaemia and slip into coma. β₁-Selective drugs are preferred in diabetes.

Plasma lipoproteins: high-density lipoprotein (HDL) cholesterol falls and triglycerides rise during chronic β-blockade with non-selective agents. β₁-Selective agents have much less impact overall. Patients with hyperlipidaemia needing a β-blocker should generally receive a β₁-selective one.

Sexual function: interference is unusual and generally not supported in placebo-controlled trials.

Abrupt withdrawal of therapy can be dangerous in angina pectoris and after myocardial infarction, and withdrawal should be gradual, e.g. reduce to a low dose and continue this for a few days. The existence and cause of a β-blocker withdrawal phenomenon is debated, but probably occurs due to up-regulation of β₂ receptors. It is particularly inadvisable to initiate an α-blocker at the same time as withdrawing a β-blocker in patients with ischaemic heart disease, because the β-blocker causes reflex activation of the sympathetic system. The β-blocker withdrawal phenomenon appears to be least common with partial agonists and most common with β₁-selective antagonists. Rebound hypertension is insignificant.

Adverse reactions not certainly due to β-adrenoceptor blockade

These include loss of general well-being, tired legs, fatigue, depression, sleep disturbances including insomnia, dreaming, feelings of weakness, gut upsets, rashes.

Oculomucocutaneous syndrome occurred with chronic use of practolol (now obsolete) and even occasionally after cessation of use.²⁴ Other members either do not cause it, or so rarely do so that they are under suspicion only and, properly prescribed, the benefits of their use far outweigh such a very low risk. The mechanism of the syndrome is uncertain but appears immunological.

Overdose

Overdose, including self-poisoning, causes bradycardia, heart block, hypotension and low-output cardiac failure that can proceed to cardiogenic shock; death is more likely with agents that have a membrane-stabilising action (see Table 24.1). Bronchoconstriction can be severe, even fatal, in patients subject to any bronchospastic disease; loss of consciousness may occur with lipid-soluble agents that penetrate the CNS. Receptor blockade will outlast the persistence of the drug in the plasma.

Rational treatment includes:

• Atropine (1–2 mg i.v. as one or two bolus doses) to eliminate the unopposed vagal activity that contributes to bradycardia. Most patients also require direct cardiac pacing.

• Glucagon, which has cardiac inotropic and chronotropic actions independent of the β-adrenoceptor (dose 50–150 micrograms/kg in glucose 5% i.v., repeated if necessary) to be used at the outset in severe cases (an unlicensed indication).

• If there is no response, intravenous injection or infusion of a β-adrenoceptor agonist is an alternative, e.g. isoprenaline (4 micrograms/min, increasing at 1–3-min intervals until the heart rate is 50–70 beats/min).

• Other sympathomimetics may be used as judgement counsels, according to the desired receptor agonist actions (β₁, β₂, α) required by the clinical condition, e.g. dobutamine, dopamine, dopexamine, noradrenaline/norepinephrine, adrenaline/epinephrine.

• For bronchoconstriction, salbutamol may be used; aminophylline has non-adrenergic cardiac inotropic and bronchodilator actions and should be given intravenously very slowly to avoid precipitating hypotension.

• A cardiac pacemaker may be used to increase the heart rate.

Treatment may be needed for days. With prompt treatment, death is unusual.

Interactions

Pharmacokinetic

β-blockers that are metabolised in the liver exhibit higher plasma concentrations when co-administered with drugs that inhibit hepatic metabolism, e.g. cimetidine. Enzyme inducers enhance the metabolism of this class of β-blockers. β-Adrenoceptor blockers themselves reduce hepatic blood flow (with fall in cardiac output) and reduce the metabolism of β-blockers and other drugs whose metabolic elimination is dependent on the rate of delivery to the liver, e.g. lidocaine, chlorpromazine.

Pharmacodynamic

The effect on the blood pressure of sympathomimetics having both α- and β-receptor agonist actions is increased by block of β receptors, leaving the α-receptor vasoconstriction unopposed (even adrenaline/epinephrine added to local anaesthetics may cause hypertension); the pressor effect of abrupt clonidine withdrawal is enhanced, probably by this action. Other cardiac antiarrhythmic drugs are potentiated, e.g. hypotension, bradycardia, heart block. Combination with verapamil (i.v.) is hazardous in the presence of atrioventricular nodal or left ventricular dysfunction because the latter has stronger negative inotropic and chronotropic effects than do other calcium channel blockers.

Most non-steroidal anti-inflammatory drugs (NSAIDs) attenuate the antihypertensive effect of β-blockers (but not perhaps of atenolol), presumably due to inhibition of formation of renal vasodilator prostaglandins, leading to sodium retention.

β-Adrenoceptor blockers potentiate the effect of other antihypertensives, particularly when an increase in heart rate is part of the homeostatic response (calcium channel blockers and α-adrenoceptor blockers).

Non-selective β-receptor blockers potentiate hypoglycaemia of insulin and sulphonylureas.

Pregnancy

β-Adrenoceptor-blocking agents are used in pregnancy-related hypertension, including pre-eclampsia. Both lipid- and water-soluble members enter the fetus and may cause neonatal bradycardia and hypoglycaemia. They are not teratogenic in pregnancy, but some studies have suggested they cause intrauterine growth retardation.

Notes on some individual β-adrenoceptor blockers

(For general pharmacokinetics, see p. 79.)

Propranolol

is available in standard (twice or three times daily) and sustained-release (once daily) formulations. When given i.v. (1 mg/min over 1 min, repeated every 2 min up to 10 mg) for cardiac arrhythmia or thyrotoxicosis it should be preceded by atropine (1–2 mg i.v.) to prevent excessive bradycardia; hypotension may occur.

Atenolol

has a β₁:β₂ selectivity of 1:15. It is widely used for angina pectoris and hypertension, in a dose of 25–100 mg orally once a day. The tendency in the past has been to use higher than necessary doses. When introduced, atenolol was considered not to need dose ranging, unlike propranolol, but this was in part because the initial dose was already at the top of the dose–response curve. Some 90% of absorbed drug is excreted by the kidney and the dose should be reduced when renal function is impaired, e.g. to 50 mg/day when the glomerular filtration rate (GFR) is 15–35 mL/min. It is best avoided in patients with GFR of less than 10 mL/min. The t_½ is 7 h.

Bisoprolol

is more β₁ selective than atenolol (ratio 1:50). Although a relatively lipid-soluble agent, its t_½ (11 h) is one of the longest and there is not the wide range of dose requirement seen with propranolol. It is worth starting at a low dose (5 mg), to avoid causing unnecessary tiredness and obtain the maximum benefit of its selectivity. There is no need to alter doses when renal or hepatic function is reduced.

Nebivolol

resembles bisoprolol in terms of lipophilicity and t_½ (10 h) but is more β₁ selective (ratio 1:300). Its unique feature is a direct vasodilator action (due to the D-isomer of the racemate, the L-isomer being the β₁ antagonist). The mechanism appears to be through direct activation of nitric oxide production by vascular endothelium.

Combined β₁- and α-adrenoceptor-blocking drug

Labetalol

is a racemic mixture: one isomer is a β-adrenoceptor blocker (non-selective), another blocks α-adrenoceptors. Its dual effect on blood vessels minimises the vasoconstriction characteristic of non-selective β-blockade so that, for practical purposes, the outcome is similar to that of a β₁-selective β-blocker (see Table 24.1). It is less effective than drugs such as atenolol or bisoprolol for the routine treatment of hypertension, but is useful for some specific indications.

The β-blockade is 4 to 10 times greater than the α-blockade, varying with dose and route of administration. Labetalol is useful as a parenterally administered drug in the emergency reduction of blood pressure. Ordinary β-blockers may lower blood pressure too slowly, in part because reflex stimulation of unblocked α receptors opposes the fall in blood pressure. In most patients, even those with severe hypertension, a gradual reduction in blood pressure is desirable to avoid the risk of cerebral or renal hypoperfusion, but in the presence of a great vessel dissection or of fits, a more rapid effect is required (below).

Postural hypotension (characteristic of α-receptor blockade) is liable to occur at the outset of therapy and if the dose is increased too rapidly. But with chronic therapy when the β-receptor component is largely responsible for the antihypertensive effect, it is not a problem.

Labetalol reduces the hypertensive response to orgasm in women.

The t_½ is 4 h; it is extensively metabolised in the hepatic first-pass. The drug is given twice daily in a dose of 100–800 mg.

For emergency control of severe hypertension (including pregnancy), the most convenient regimen is to initiate infusion at 1 mg/min, and titrate upwards at half-hourly intervals as required. If bradycardia is a problem, then intravenous atropine should be given (as 600-microgram boluses). The labetalol infusion is stopped as blood pressure control is achieved (up to 200 mg may be required), and re-initiated as frequently as required until regular oral therapy has been successfully introduced.

Peripheral sympathetic nerve terminal

Adrenergic neurone-blocking drugs

Adrenergic neurone-blocking drugs are taken up into adrenergic nerve endings by the active noradrenaline/norepinephrine reuptake mechanism (uptake 1) (see Fig. 23.1). They are relatively ineffective in reducing blood pressure except in the erect position, and their use to control hypertension is now obsolete. Guanethidine is still licensed in the UK as an option for the rapid control of blood pressure, and may also be used for regional intravenous sympathetic blockade in patients with intractable Raynaud's disease.

Meta-iodobenzylguanidine (MIBG) is used diagnostically as a radio-iodinated tracer, to locate or confirm chromaffin tumours (phaeochromocytoma and neuroblastoma), which accumulate with drugs in this class (see p. 419).

Depletion of stored transmitter (noradrenaline/norepinephrine)

Reserpine is an alkaloid from plants of the genus Rauwolfia, used in medicine since ancient times for insanity. Reserpine depletes adrenergic nerves of noradrenaline/norepinephrine, primarily by blocking the transport of noradrenaline/norepinephrine into storage vesicles (see Fig. 23.1). Its antihypertensive action is chiefly a peripheral action, but it enters the CNS and depletes central catecholamine stores; this explains the sedation, depression and parkinsonian side-effects that can accompany its use. Reserpine is rarely used now that its low cost is matched by many superior classes.

Inhibition of synthesis of transmitter

Metirosine (α-methyl-p-tyrosine) is a competitive inhibitor of the enzyme tyrosine hydroxylase, which converts tyrosine to DOPA. It is occasionally used as an adjuvant (with phenoxybenzamine) to treat phaeochromocytomas that cannot be removed surgically. Catecholamine synthesis is reduced by up to 80% over 3 days. It also enters the CNS and depletes brain noradrenaline/norepinephrine and dopamine, causing reserpine-like adverse effects (see above). Hence, if life expectancy is threatened more by tumour invasion than hypertension, the need for the drug should be weighed carefully.

Autonomic ganglion-blocking drugs

Hexamethonium

was the first orally active drug to treat hypertension, but the blockade of both sympathetic and parasympathetic systems caused severe side-effects. Its use in hypertension is long obsolete.

Trimetaphan,

a short-acting agent, is given intravenously for the emergency control of hypertension; pressure may be adjusted by tilting the body to provide ‘minute to minute’ control, when the lack of selectivity is useful.

Central nervous system

α₂-Adrenoceptor agonists

Clonidine

is an imidazoline that is an agonist to α₂-adrenoceptors (postsynaptic) in the brain, stimulation of which suppresses sympathetic outflow and reduces blood pressure. Drugs of this type are said to be selective for an imidazoline receptor (I₁), rather than the α₂ receptor. In fact, no such receptor has been identified at the molecular level, and genetic knockout experiments have shown that it is the α₂ receptor that is required for the blood pressure-lowering action of imidazoline drugs. At high doses clonidine activates peripheral α₂-adrenoceptors (presynaptic autoreceptors) on the adrenergic nerve ending; these mediate negative feedback suppression of noradrenaline/norepinephrine release.

Clonidine was discovered to be hypotensive, not by the pharmacologists who tested it in the laboratory but by a physician who used it on himself as nose drops for a common cold. The t_½ is 6 h. Clonidine reduces blood pressure with little postural or exercise-related drop.

Its most serious handicap is that abrupt or even gradual withdrawal causes rebound hypertension. This is characterised by plasma catecholamine concentrations as high as those seen in hypertensive attacks of phaeochromocytoma. The onset may be rapid (a few hours) or delayed for as long as 2 days; it subsides over 2–3 days. The treatment is either to reinstitute clonidine, intramuscularly if necessary, or to treat as for a phaeochromocytoma (see below). Clonidine should never be used with a β-adrenoceptor blocker that exacerbates withdrawal hypertension (see phaeochromocytoma, p. 419). Other common adverse effects include sedation and dry mouth.

Tricyclic antidepressants antagonise the antihypertensive action and increase the rebound hypertension of abrupt withdrawal. Low-dose clonidine (Dixarit 50–75 mg twice daily) also has a minor role in migraine prophylaxis, menopausal flushing and choreas.

Rebound hypertension is a less important problem with longer-acting imidazolines, e.g. moxonidine, and a single dose can be omitted without rebound.

False transmitter

Chemotransmitters and receptors in the CNS are similar to those in the periphery, and the drug in this section also has peripheral actions, as is to be expected.

Methyldopa

acts in the brainstem vasomotor centres. It is a substrate (in the same manner as L-dopa) for the enzymes that synthesise noradrenaline/norepinephrine. The synthesis of α-methylnoradrenaline results in tonic stimulation of CNS α₂receptors because α-methylnoradrenaline cannot be metabolised by monoamine oxidase, and selectively stimulates the α₂-adrenoceptor, i.e. methyldopa acts in the same way as clonidine. Methyldopa is reliably absorbed from the gastrointestinal tract and readily enters the CNS. The t_½ is 1.5 h.

Adverse effects include: sedation (frequent), nightmares and depression. Less common is a positive Coombs’ test with haemolytic anaemia, and a rare but life-threatening adverse event is hepatitis.

For these reasons, methyldopa has long been dropped from routine management of hypertension, but remains popular with obstetricians for the hypertension of pregnancy because of its apparent safety for the fetus.

Drug treatment of angina, myocardial infarction and hypertension

Angina pectoris²⁵

An attack of angina pectoris²⁶ occurs when myocardial demand for oxygen exceeds supply from the coronary circulation. The principal forms relevant to choice of drug therapy are angina of exercise (commonest) and its worsening form, unstable (preinfarction or crescendo) angina (see below), which occurs at rest. Variant (Prinzmetal) angina (very uncommon) results from spasm of a large coronary artery.

Antiangina drugs

act as follows:

• Organic nitrates reduce preload and afterload and dilate the main coronary arteries (rather than the arterioles).

• β-Adrenoceptor-blocking drugs reduce myocardial contractility and slow the heart rate. They may increase coronary artery spasm in variant angina.

• Calcium channel-blocking drugs reduce cardiac contractility, dilate the coronary arteries (where there is evidence of spasm) and reduce afterload (dilate peripheral arterioles).

These classes of drug complement one another and can be used together. The combined nitrate and potassium channel activator nicorandil is an alternative when any of the other drugs is contraindicated.

Summary of treatment

• Any contributory cause is treated when possible, e.g. anaemia, arrhythmia.

• Lifestyle is changed so as to reduce the number of attacks. Weight reduction can be very helpful; stop smoking.

• For immediate pre-exertional prophylaxis: glyceryl trinitrate sublingually or nifedipine (bite the capsule and hold the liquid in the mouth or swallow it).

• For an acute attack: glyceryl trinitrate (sublingual) or nifedipine (bite capsule, as above).

For long-term prophylaxis:

• A β₁-adrenoceptor-blocking drug, e.g. bisoprolol, is given regularly (not merely when an attack is expected). Dosage is adjusted by response. Some put an arbitrary upper limit to dose, but others recommend that, if complete relief is not obtained, the dose should be raised to the maximum tolerated, provided the resting heart rate is not reduced below 55 beats/min; or raise the dose to a level at which an increase causes no further inhibition of exercise tachycardia. In severe angina a pure antagonist, i.e. an agent lacking partial agonist activity, is preferred, as the latter may not slow the heart sufficiently. Warn the patient of the risk of abrupt withdrawal.

• A calcium channel-blocking drug, e.g. nifedipine or diltiazem, is an alternative to a β-adrenoceptor blocker; use especially if coronary spasm is suspected or if the patient has myocardial insufficiency or any reversible airflow obstruction. It can also be used with a β-blocker, or

• A long-acting nitrate, isosorbide dinitrate or mononitrate: use so as to avoid tolerance (see p. 394).

• Nicorandil, a long-acting potassium channel activator, does not cause tolerance like the nitrates.

• Drug therapy may be adapted to the time of attacks, e.g. nocturnal (transdermal glyceryl trinitrate, or isosorbide mononitrate orally at night).

• Antiplatelet therapy (aspirin or clopidogrel) reduces the incidence of fatal and non-fatal myocardial infarction in patients with unstable angina, used alone or with low-dose heparin.

• Revascularisation in selected cases (largely by percutaneous coronary intervention (PCI) and stenting).

Newer therapies

Ivabradine

causes bradycardia and hence a reduction in myocardial work by slowing the sinus node. It does this by blocking the so-called ‘pacemaker’ or ‘funny’ current in sinoatrial (SA) nodal cells. It offers an advantage over β-blockers in being safe to use in asthmatics, but its use is limited by the curious visual disturbance it causes by blocking the same funny current in the retina. Ranolazine blocks the late sodium current and prevents calcium overload in ischaemic cardiac tissue. It also has antiarrhythmic properties and by an unknown mechanism reduces fasting blood glucose and HbA1c levels in diabetics. Nevertheless, the exact role of this agent remains to be determined.

In treating angina, it is important to remember not only the objective of reducing symptoms but also that of preventing complications, particularly myocardial infarction and sudden death. This requires vigorous treatment of all risk factors (hypertension, hyperlipidaemia, diabetes mellitus) and, of course, cessation of smoking. There is little evidence that the symptomatic treatments, medical or surgical, themselves affect outcome except in patients with stenosis of the main stem of the left coronary artery, who require surgical intervention. Although aspirin has not been studied specifically in patients with stable angina, it is now reasonable therapy, by extrapolation from the studies of aspirin in other patient groups.

Myocardial infarction (MI)

(See also Ch. 29.)

An overview

The acute coronary syndromes (ACSs) are now classified on the basis of the ECG and plasma troponin measurements into: (1) patients with ST elevation myocardial infarction (STEMI); (2) non-ST elevation myocardial infarction (non-STEMI, by ECG and a positive troponin test); (3) unstable angina (by ECG and negative troponin test). The present account recognises that this is a rapidly evolving field, but therapeutic strategies are likely to evolve according to these forms of ACS.

A general practitioner or paramedic can administer the initial treatment appropriately before a definite diagnosis is established or the patient reaches hospital, namely:

• Morphine or diamorphine (2.5 or 5 mg i.v. because of the certainty of haematoma formation when intramuscular injections are followed by thrombolytic therapy).

• Aspirin 150–300 mg orally.

• Sixty per cent oxygen.

The immediate objectives are relief of pain and initiation of treatment demonstrated to reduce mortality. Subsequent management of proven MI is concerned with treatment of complications, arrhythmias, heart failure and thromboemboli, and then prevention of further infarctions.

When STEMI is diagnosed, instituting myocardial reperfusion as early as possible provides the greatest benefit. Previously this was achieved pharmacologically, although in many centres primary coronary angioplasty (PCI), with or without stenting, is the preferred option. If thrombolysis is used it is initiated after arrival at hospital and provided there are no contraindications to thrombolysis (see below). Patients with non-STEMI may still benefit, especially those with left bundle branch block, but several trials have shown that patients without ECG changes (especially ST elevation) and patients with unstable angina benefit only slightly, if at all, from thrombolytic therapy.

The choice of thrombolytic is in most places dictated first by a wealth of comparative outcome data from well-designed trials, and second by relative costs. So, for a first MI, patients should receive streptokinase 1 500 000 units infused over 1 h, unless they are in cardiogenic shock. For subsequent infarcts, the presence of antistreptokinase antibodies dictates the use of the recombinant tissue plasminogen activator (rtPA), alteplase (or reteplase). Both alteplase and streptokinase bind plasminogen and convert it to plasmin, which lyses fibrin. Alteplase has a much higher affinity for plasminogen bound to fibrin than in the circulation. This selectivity does not confer any therapeutic advantage as was originally anticipated, as severe haemorrhage following thrombolysis is almost always due to lysis of an appropriate clot at previous sites of bleeding or trauma. Indeed, the tendency for some lysis of circulating fibrinogen as well as fibrin gives streptokinase anticoagulant activity, which is lacking with alteplase, use of which needs to be accompanied and followed by heparin (for further details of thrombolytics, see p. 490).

For a discussion about the role of aspirin, see p. 246.

A third treatment reduces mortality in MI, namely β-blockade. In the ISIS-1 study,²⁷ atenolol 5 mg was given intravenously, followed by 50 mg orally. The reduction in mortality is due mainly to prevention of cardiac rupture, which appears interestingly to remain the only complication of MI that is not reduced by thrombolysis. The usual contraindications to β-blockade apply, but most patients with a first MI should be able to receive this treatment.

Other antiplatelet agents

The final common pathway to platelet aggregation and thrombus formation involves the expression of the glycoprotein IIb/IIIa receptor at the cell surface. This receptor binds fibrinogen with high affinity and can be blocked using either a specific monoclonal antibody (abciximab) or one of a rapidly expanding class of specific antagonists, e.g. eptifibatide and tirofiban. Another agent, clopidogrel, acts by inhibiting ADP-dependent platelet aggregation. It is more effective than aspirin for the prevention of ischaemic stroke, cardiovascular death in patients at high risk (see p. 492) or following a STEMI or non-STEMI event.

These drugs are useful adjuncts for the treatment of unstable angina, and in the prevention of thrombosis following percutaneous revascularisation procedures such as angioplasty and coronary artery stenting (especially with coated stents).

Principal contraindications to thrombolysis

• Haemorrhagic diathesis.

• Pregnancy.

• Recent symptoms of peptic ulcer, or gastrointestinal bleeding.

• Recent stroke (previous 3 months).

• Recent surgery (previous 10–14 days), especially neurosurgery.

• Prolonged cardiopulmonary resuscitation (during current presentation).

• Proliferative diabetic retinopathy.

• Severe, uncontrolled hypertension (diastolic blood pressure > 120 mmHg).

Unstable angina

necessitates admission to hospital, the objectives of therapy being to relieve pain, and avert progression to MI and sudden death. Initial management is with aspirin 150–300 mg chewed or dispersed in water, followed by heparin or one of the low molecular weight preparations, e.g. dalteparin or enoxaparin. Nitrate is given preferably as isosorbide dinitrate by intravenous infusion until the patient has been pain-free for 24 h. A β-adrenoceptor blocker, e.g. metoprolol, should be added orally or intravenously unless it is contraindicated, when a calcium channel blocker is substituted, e.g. diltiazem or verapamil. Patients perceived to be at high risk may also receive a glycoprotein IIb/IIIa inhibitor, e.g. eptifibatide or tirofiban.

Secondary prevention

(See also Ch. 29.)

The best predictor of the risk of MI is to have had previous infarction. After the measures instituted in the first few hours, the principal objective of treatment therefore becomes prevention of future infarcts. Patients should receive advice about exercise and diet before discharge, and most enter a formal rehabilitation programme after leaving hospital. In particular, patients need to reduce saturated fat intake, and there is increasing evidence of the benefit of increased intake of fish and olive oil.

Drugs for secondary prevention

All patients should receive aspirin (see Ch. 4, Fig. 4.3), an ACE inhibitor and a β-blocker for at least 2 years, unless contraindicated. The commonest contraindications to β-blockade after MI are transientheart failure, which should now be uncommon after a first MI, and various degrees of heart block or bradyarrythmias. These are, however, usually transient, so the β-blocker can be introduced during convalescence.

Any of these agents, aspirin, a β-blocker or an ACE inhibitor,²⁸ will reduce the incidence of reinfarction by 20–25%, although their benefit has not been shown to be additive.

In addition to these drugs, most patients should receive a statin, regardless of their plasma cholesterol concentration. Long-term benefit from LDL reduction after MI has been shown for simvastatin (20–40 mg/day) and pravastatin (40 mg/day).²⁹

There is no place for routine antiarrhythmic prophylaxis, and long-term anticoagulation is similarly out of place, except when indicated by arrhythmias or poor left ventricular function.

Arterial hypertension

Clinical evaluation of antihypertensive drugs seeks to answer two types of question:

1. Whether long-term reduction of blood pressure benefits the patient by preventing complications and prolonging life; these studies take years, require enormous numbers of patients and are extremely costly.

2. Whether a drug is capable of effective, safe and comfortable control of blood pressure for about 1 year. There is now sufficient evidence of the benefit of reducing raised blood pressure that regulatory authorities do not demand trials of the first kind for all new drugs. Shorter studies are therefore deemed sufficient to allow the introduction of a new drug.

Aims of treatment

The long-term aim is the prevention of stroke, MI and (especially in older patients) heart failure; prevention also requires attention to other risk factors such as smoking and plasma cholesterol. The more immediate aim of treatment is to reduce the blood pressure as near to normal as possible without causing symptomatic hypotension or otherwise impairing well-being (quality of life).

When this aim is achieved in severe cases there is great symptomatic improvement: retinopathy clears and vision improves; headaches are abolished. A variable amount of irreversible damage has often been done by the high blood pressure before treatment is started; so renal failure may progress despite treatment, left ventricular hypertrophy may not fully reverse, and arterial damage leads to ischaemic events (stroke and MI).

It is desirable to start treatment before irreversible changes occur, and in mild and moderately severe cases this usually means advising treatment for symptom-free people whose hypertension was revealed by screening.

Threshold and targets for treatment

The joint NICE/British Hypertension Society guidelines³⁰ require that antihypertensive drug therapy be initiated:

• when sustained blood pressure exceeds 160/100 mmHg, or

• when blood pressure is in the range 140–159/90–99 mmHg and there is evidence of target organ damage, cardiovascular disease or a 10-year cardiovascular risk greater than 20%

• for diabetics when blood pressure exceeds 140/90 mmHg.

The optimal target is to lower blood pressure to 140/85 mmHg or less in all patients except those with renal impairment, in diabetics or in established cardiovascular disease where there is a lower target of less than 130/80 mmHg.

Effective treatment reduces the risk of all complications: strokes and MI, but also heart failure, renal failure and possibly dementia. It is easier in individual trials to demonstrate the benefits of treatment in preventing stroke, because the curve relating risk of stroke to blood pressure is almost twice as steep as that for MI. This raises issues of relative and absolute risk.

Relative risk refers to the increased likelihood of a patient having a complication, compared with a normotensive patient of the same age and sex. Absolute risk refers to the number of patients out of 100, with the same age, sex and blood pressure, predicted to have a complication over the next 10 years (see p. 50). So, the relative risk of MI due to hypertension is fixed, but substantial reduction in the absolute risk of MI is possible by reducing the level of cholesterol and blood pressure, i.e. both factors contribute independently to the risk of MI whereas hypertension is a more important risk factor for stroke than hypercholesterolaemia.

Treatment will almost always be lifelong for essential hypertension, because discontinuation of therapy leads to prompt restoration of pre-treatment blood pressures. If it does not, one should suspect the original diagnosis of hypertension, which should not be made unless blood pressure is increased on at least three occasions over 3 months.

The relative risks of hypertension and the benefits of treating the condition in the elderly are less than in those aged under 65 years, but the absolute risks and benefits are greater. Given the large choice of treatments available, doctors cannot cite improved quality of life as an excuse for not treating hypertension in the elderly. Starting doses should often be halved and, pending further evidence, less challenging targets for blood pressure reduction may be acceptable.

It is obvious that adverse effects of therapy are important in that very large numbers of patients must be treated with antihypertensives so that a few may gain (in terms of numbers needed to treat this is several hundreds); this is a salient feature of the use of drugs to prevent disease.

Principles of antihypertensive therapy

General measures may be sufficient to control mild cases as follows:

• Obesity: reduce it.

• Alcohol: stay within recommended limits, e.g. 14 units/week for women, 21 units/week for men.

• Smoking: stop it.

• Diet: of proven value for the short-term reduction in blood pressure is reduction in fat content, and increase in fruit, vegetables and fibre. There is additional benefit from reducing intake of salt (< 6 g/day): avoidance of highly salted foods, and omission of added salt from freshly prepared food.³¹

• Relaxation therapy: worth considering for highly motivated borderline patients.

Drug therapy

Blood pressure may be reduced by any one or more of the actions listed at the beginning of this chapter (see p. 393). The large number of different drug classes for hypertension reduces, paradoxically, the likelihood of a randomly selected drug being the best for an individual patient. Patients and drugs can be divided broadly into two groups depending on their renin status and drug effect on this (Fig. 24.1).

• Type 1, or high-renin patients, are the younger, non-black patients (aged < 55 years); they respond better to an ACE inhibitor, ARB, or β-blocker.

• Type 2, or low-renin patients, in whom diuretics or calcium blockers are more likely to be effective as single agents.

Fig. 24.1 Effects of drugs on the renin–angiotensin system. ARB, angiotensin II receptor blocker; AI/II, angiotensin I/II.

As each drug acts on only one or two of the blood pressure control mechanisms, the factors that are uninfluenced by monotherapy are liable to adapt (homeostatic mechanism), to oppose the useful effect and to restore the previous state. There are two principal mechanisms of such adaptation or tolerance:

1. Increase in blood volume: this occurs with any drug that reduces peripheral resistance (increases intravascular volume) or cardiac output (reduces glomerular flow) due to activation of the renin–angiotensin system. The result is that cardiac output and blood pressure rise. Adding a diuretic in combination with the other drug can prevent this compensatory effect.

2. Baroreceptor reflexes: a fall in blood pressure evokes reflex activity of the sympathetic system, causing increased peripheral resistance and cardiac activity (rate and contractility).

Therefore, whenever high blood pressure is proving difficult to control and whenever a number of antihypertensives are used in combination, the drugs chosen should between them act on all three main determinants of blood pressure, namely:

• Blood volume.

• Peripheral resistance.

• Cardiac output.

Such combinations will:

• Maximise antihypertensive efficacy by exerting actions at three different points in the cardiovascular system.

• Minimise the opposing homeostatic effects by blocking the compensatory changes in blood volume, vascular tone and cardiac function.

• Minimise adverse effects by permitting smaller doses of each drug each acting at a different site and having different unwanted effects.

First-dose hypotension is now uncommon and occurs mainly with drugs having an action on veins (α-adrenoceptor blockers, ACE inhibitors) when baroreflex activation is impaired, e.g. old age or with a contracted intravascular volume following diuretics.

Treating hypertension

A simple stepped regimen in keeping with the 2006 revision of the National Institute for Health and Clinical Excellence (NICE)/British Hypertension Society guidelines³² is the ‘A/CD’ schema illustrated in Figure 24.2.³³

1. If the patient is young (age < 55 years) or non-black, use either an ACE inhibitor (or ARB) (A). For older patients start with either a Calcium channel blocker or thiazide Diuretic as first-line therapy (C or D). If a drug is effective but not tolerated, switch to the other member of the pair.

2. If the blood pressure is not controlled at 4 weeks, a second agent should be added, using the opposite pair to the first drug, e.g. if the patient is on an ACE inhibitor add a calcium channel blocker or thiazide diuretic (A + C or A +D), as both vasodilatation or diuresis will stimulate the renin–angiotensin system and turn non-renin-dependent hypertension into renin-dependent hypertension.

3. If blood pressure control is still inadequate on dual therapy, A + C + D is the ideal triple regimen.

4. Patients whose blood pressure remains substantially above target on triple therapy should have either escalated diuretic therapy, e.g. by addition of spironolactone; or addition of α- or β-blockade. There is some evidence that a trial-and-error approach at this point can be avoided by measurement of plasma renin, with low plasma renin indicating the need for further diuretic.

Fig. 24.2 ACD scheme for escalation of antihypertensive therapy. A, ACE inhibitor; C, calcium channel blocker; D, diuretic (see text).

(From Williams B W, Poulter N R, Brown M J et al 2004 British Hypertension Society guidelines for hypertension management 2004 (BHS-1 V): summary. British Medical Journal 328:634–640).

Treatment and severity

A single drug may adequately treat mild hypertension, and a few patients with hypertension due to pure vasoconstriction or pure sodium excess. In most patients, the target systolic blood pressure of < 140 mmHg recommended by most guidelines requires two or more drugs. While convention until recently has been to ‘step-care’ from single to combination therapy, the risks of adverse effects from treatment, and of early complications from uncontrolled hypertension, may be diminished by starting with combination therapy.³⁴

Monitoring

The blood pressure must be monitored by a doctor or specialist nurse (particularly important in the elderly) and also sometimes by the patient. A 24-h ambulatory blood pressure monitoring (ABPM) system is the ‘gold standard’, but the devices are too expensive to be recommended for most patients. Nevertheless, the 24-h blood pressure profile does predict outcome better than clinic blood pressure and can indicate whether a difficult or high-risk patient does need additional medication. Home monitoring is a cheaper alternative, provided the sphygmomanometer is validated. The easy-to-use wrist monitors are unfortunately unreliable in patients receiving drug treatment.

Diuretics and potassium

The potassium-losing (kaliuretic) diuretics used in hypertension can deplete body potassium by up to 10–15%. However, at low doses, e.g. bendroflumethiazide 2.5 mg/day, significant hypokalaemia is unusual, and should raise suspicion of Conn's syndrome (see p. 420). Vulnerable patients, e.g. the elderly, should be monitored for potassium loss at 3 months and thereafter every 6–12 months. If required, for correction of hypokalaemia, a potassium-sparing diuretic (amiloride) in a fixed-dose combination with a thiazide (co-amilozide) is preferred over the use of fixed-dose diuretic/potassium chloride formulations (most supplements, typically 8 mmol KCl, are in any case inadequate).

A potassium-sparing diuretic should be used with caution in patients who have reduced renal function or diabetes, especially if they are also receiving an ACE inhibitor or ARB.

Compliance

Multi-drug therapy poses a substantial problem of compliance. As treatment will be lifelong, it is worthwhile finding the most convenient regimen for each individual and transferring to a fixed-dose combination where possible. A single daily dose is available for most antihypertensive drugs, where necessary by using sustained-release formulations.

Treatment of hypertensive emergencies

It is important to distinguish three circumstances that may exist separately or together; see the Venn diagram (Fig. 24.3)³⁵ which emphasises the following:

• Severe hypertension is not on its own an indication for urgent (or large) reductions in blood pressure.

• Blood pressure can occasionally require urgent (emergency) reduction even when the hypertension is not severe, especially where the blood pressure has risen rapidly.

• Accelerated phase (malignant) hypertension rarely requires urgent reduction, and should instead be regarded as an indication for slow reduction in blood pressure during the first few days.

Fig. 24.3 Venn diagram illustrating intersections of three overlapping clinical states defined in the text.

The indications for emergency reduction of blood pressure are rare. They are:

• Hypertensive encephalopathy (including eclampsia).

• Acute left ventricular failure (due to hypertension).

• Dissecting aortic aneurysm.

In these conditions, blood pressure should be reduced over the course of an hour. In patients with a dissecting aneurysm, where the blood pressure may have been completely normal prior to dissection, the target is a blood pressure of about 110/70 mmHg. Otherwise even small reductions will usually remove the emergency.

Accelerated phase hypertension

was previously called ‘malignant’ hypertension because the lack of treatment heralded death within a year of diagnosis. It is characterised pathologically by fibrinoid necrosis of the small arteries. An important consequence is the loss of autoregulation of the cerebral and renal circulation, so that any reduction in blood pressure causes a proportional fall in perfusion of these organs. It is therefore vital not to reduce diastolic blood pressure by more than 20 mmHg on the first day of treatment. To ignore this is to risk cerebral infarction.

Treatment

Unless contraindicated, the best treatment for all circles in the Venn diagram is β-blockade, e.g. atenolol 25 or 50 mg orally. In emergencies, a vasodilator should be given intravenously in addition.

A theoretically preferable, but often impractical, alternative is intravenous infusion of the vasodilator, nitroprusside (see p. 400). In dissecting aneurysm, vasodilators should not be used unless patients are first β-blocked, because any increase in the rate of rise of the pulse stroke is undesirable. Labetalol provides a convenient method of treating all patients within the three circles (except asthmatics), using either oral or parenteral therapy as appropriate. That said, it is not the most effective therapy and should be combined with a long-acting formulation of nifedipine, orally, where further blood pressure reduction is required.

Low doses of all drugs should be used if other antihypertensive drugs have recently been taken or renal function is impaired.

Oral maintenance treatment for severe hypertension should be started at once if possible; parenteral therapy is seldom necessary for more than 48 h.

Pregnancy hypertension

Effective treatment of pregnancy-induced hypertension improves fetal and perinatal survival. There is a lack of good clinical trial evidence on which to base recommendations. Instead, drug selection reflects longevity of use without obvious harm to the fetus and methyldopa is still the drug of choice for many obstetricians. Calcium channel blockers (especially nifedipine) are common second-line drugs; parenteral hydralazine is reserved for emergency reduction of blood pressure in late pregnancy, preferably in combination with a β-blocker to avoid unpleasant tachycardia.

β-Blockers (labetalol and atenolol) are often effective and are probably the drugs of choice in the third trimester; there is anecdotal evidence to suggest growth retardation with β-blockade used in the first and second trimesters. Diuretics reduce the chance of developing pre-eclampsia, but are avoided in pre-eclampsia itself because these patients already have a contracted circulatory volume.

ACE inhibitors and angiotensin II receptor blockers (ARBs) are absolutely contraindicated during pregnancy. They cause major malformations after first-trimester exposure³⁶ and fetal death, typically mid-trimester. For this reason they are probably best avoided in women of child-bearing age, especially where there is no effective contraception, since it is not uncommon for women to discover their pregnancy late into its first trimester. If they are used, women should be counselled to stop an ACE or ARB as soon as they suspect they are pregnant.

Raised blood pressure and proteinuria (pre-eclampsia) complicates 2–8% of pregnancies and may proceed to fitting (eclampsia), a major cause of mortality in mother and child. Magnesium sulphate halves the risk of progress to eclampsia (typically 4 g i.v. over 5–10 min followed by 1 g/h by i.v. infusion for 24 h after the last seizure).³⁷ Additionally, if a woman has one fit (treat with diazepam), then the magnesium regimen is superior to diazepam or phenytoin in preventing further fits.³⁸

Aspirin, in low dose, was reported in early studies to reduce the incidence of pre-eclampsia in at-risk patients, but a more recent meta-analysis has not supported this. Consequently, it is not routinely recommended.

Unwanted interactions with antihypertensive drugs

Specific interactions are described in the accounts of individual drugs. The following are general examples for this diverse group of drugs.

Alcohol intake

is the commonest contributing factor, or even cause, of hypertension, and should always be considered as a cause of erratic or failed responses to treatment (measurement of the γ-glutamyl transpeptidase and red cell mean corpuscular volume may be useful).

Prostaglandin synthesis

NSAIDs, e.g. indometacin, attenuate the antihypertensive effect of β-adrenoceptor blockers and of diuretics, perhaps by inhibiting the synthesis of vasodilator renal prostaglandins. This effect can also be important when a diuretic is used for severe left ventricular failure.

Enzyme inhibition

Ciprofloxacin and cimetidine inhibit hepatic metabolism of lipid-soluble β-adrenoceptor blockers, e.g. metoprolol, labetalol, propranolol, increasing their effect. Methyldopa plus an monoamine oxidase (MAO) inhibitor may cause excitement and hallucinations.

Pharmacological antagonism

Sympathomimetics, e.g. amfetamine, phenylpropanolamine (present in anorectics and cold and cough remedies), may lead to loss of antihypertensive effect, and indeed to a hypertensive reaction when taken by a patient already on a β-adrenoceptor blocker, due to unopposed α-adrenergic stimulation.

Surgical anaesthesia

may lead to a brisk fall in blood pressure in patients taking antihypertensives. Antihypertensive therapy should not be routinely altered before surgery, although it obviously can complicate care both during and after the operation. Anaesthetists must be informed.

Sexual function and cardiovascular drugs

All drugs that interfere with sympathetic autonomic activity can potentially interfere with male sexual function, expressed as a failure of ejaculation or difficulty in sustaining an erection. Nevertheless, placebo-controlled trials have emphasised how common a symptom this is in the untreated male population (sometimes approaching 20–30%). It is also likely that hypertension itself is associated with an increased risk of sexual dysfunction, since loss of nitric oxide production by the vascular endothelium is an early feature of the pathophysiology of this disease.

Laying the blame on antihypertensive medication is incorrect in most instances. Calcium channel blockers, ACE inhibitors and angiotensin II (AT₁) receptor blockers (ARBs) all have reported rates of sexual dysfunction that do not differ from placebo. If symptoms persist with these drugs other causes should be sought. Sildenafil (Viagra) can be used safely in patients receiving any of the commonly used antihypertensive drugs.

As well as the concerns about sexual performance in treated hypertensives, there may be concerns about fitness per se to attempt intercourse. The real possibility that it is hazardous is compounded often by their age and concurrent coronary artery disease.

Sexual intercourse and the cardiovascular system

Normal sexual intercourse with orgasm is accompanied by transient but brisk physiological changes, e.g. tachycardia up to 180 beats/min, with increases of 100 beats/min over less than 1 min. Systolic blood pressure may rise by 120 mmHg and diastolic by 50 mmHg. Orgasm may be accompanied by transient pressure of 230/130 mmHg even in normotensive individuals. Electrocardiographic abnormalities may occur in healthy men and women. Respiratory rate may rise to 60 beats/min.

Such changes in the healthy might bode ill for the unhealthy (with hypertension, angina pectoris or after myocardial infarction). Sudden deaths do occur during or shortly after sexual intercourse (ventricular fibrillation or subarachnoid haemorrhage), usually in clandestine circumstances such as the bordello or the mistress's bed, especially when an older man and a younger woman are involved – although this may just reflect reporting bias in the press. In one series, 0.6% of all sudden deaths were (reportedly) attributable to sexual intercourse and in about half of these cardiac disease was present. Clearly, the older patient with coronary heart disease should aspire cautiously to the haemodynamic heights attainable in youth.

There are few, if any, records of sudden cardiovascular death among women under these circumstances.

If there is substantial concern about cardiovascular stress (hypertension or arrhythmia) during sexual intercourse in either sex, a dose of labetalol about 2 h before the event may well be justified (taking account of other therapy already in use). But patients taking a β-blocker long term for angina prophylaxis have shown reductions in peak heart rate during coitus from 122 to 82 beats/min.

Patients subject to angina pectoris should also use glyceryl trinitrate or isosorbide dinitrate as usual for pre-exertional prophylaxis 10 min before intercourse. But they should be aware of the potentially fatal interaction of sildenafil (Viagra) and other PDE5 inhibitors with nitrates (see p. 394).

Summary

• The treatment of both hypertension and angina requires drugs that reduce the work of the heart either directly or by lowering peripheral vascular resistance.

• β-Blockade, which acts mainly through reduced cardiac output, and calcium channel blockade, acting by selective arterial dilatation, may be used in either condition.

• Other vasodilators are suited preferentially to hypertension (ACE inhibitors, angiotensin (AT₁) receptor blockers (ARBs) and α-adrenoceptor blockers) or to angina (nitrates).

• The treatment of myocardial infarction requires thrombolysis, aspirin and β-adrenoceptor blockade acutely, with the latter two continued for at least 2 years as secondary prevention of a further infarction.

• Other important steps in secondary prevention include ACE inhibitors for cardiac failure and statins for hypercholesterolaemia in selected patients.

Pulmonary hypertension

Therapy is determined by the underlying cause. When the condition is secondary to hypoxia accompanying chronic obstructive pulmonary disease, long-term oxygen therapy improves symptoms and prognosis; anticoagulation is essential when the cause is multiple pulmonary emboli.

Idiopathic (primary) pulmonary arterial hypertension (PAH)

Verapamil may give symptomatic benefit. Prostanoid formulations used to treat PAH include intravenous epoprostenol (prostacyclin) and inhaled iloprost.³⁹ The prostanoid formulations have the limitations of a short half-life and a heterogeneous response to therapy. Evidence suggests that endothelin, a powerful endogenous vasoconstrictor, may play a pathogenic role; antagonists, bosentan, ambrisentan and sitoxsentan, improve symptoms and haemodyamic measurements, but dose is limited by hepatic toxicity. The PDE5 inhibitors, sildenafil, tadalafil, and vardenafil, are an alternative. Only tadalafil has been shown to improve survival, although in a comparison of the three drugs only sildenafil increased exercise tolerance. Heart and lung transplantation is recommended for younger patients.

Treatment of secondary hypertension

In about 5% of patients it is possible to identify a single cause of the hypertension, and some of these can be cured by a surgical or endovascular intervention. In some of these cases, either the diagnostic work-up, preparation for surgery, or long-term management requires some clinical pharmacology different from the management of essential hypertension. This is particularly true of adrenal causes of hypertension.

Phaeochromocytoma

This tumour of chromaffin tissue, usually arising in the adrenal medulla, secretes principally noradrenaline/norepinephrine, but often also variable amounts of adrenaline/epinephrine. Symptoms are related to this. Hypertension may be sustained or intermittent. If the tumour secretes only noradrenaline/norepinephrine, which stimulates α- and β₁-adrenoceptors, rises in blood pressure are accompanied by reflex bradycardia due to vagal activation; this is sufficient to overcome the chronotropic effect of β₁-receptor stimulation. The recognition of bradycardia at the time of catecholamine-induced symptoms (e.g. anxiety, tremor or sweating) is useful in alerting the physician to the possibility of this rare syndrome, as physiological sympathetic nervous activation is coupled to vagal withdrawal and causes tachycardia. If the tumour also secretes adrenaline/epinephrine, which stimulates α-, β₁- and β₂-adrenoceptors, blood pressure and heart rate change in parallel. This is because stimulation of the vasodilator β₂ receptor in resistance arteries attenuates the rise in diastolic pressure, and vagal activation is insufficient then to oppose the chronotropic effect of combined β₁ and β₂ receptor stimulation in the heart.

Diagnosis

is made by measurement of the O-methylated metabolite of catecholamines, namely normetanephrine (from noradrenaline/norepinephrine) and metanephrine (from adrenaline/epinephrine) in blood or 24-h urine. Because the enzyme catechol-O-methyltransferase is present in phaeochromocytoma but not sympathetic nerve endings, the measurement of ‘fractionated metanephrines’ (i.e. normetanephrine and metanephrine separated from each other) provides > 90% specificity and sensitivity – higher than older diagnostic tests – and even modest elevations should not be ignored. Biochemical evidence for a phaeochromocytoma should usually precede the imaging hunt for a tumour. The finding of elevated metaphrine secretion (as well as normetanephrine) indicates an adrenal location since only the adrenal tumours are exposed to the high concentration of cortisol required to induce phenylethanolamine N-methyltransferase (PNMT) – the enzyme which catalyses methylation of noradrenaline/norepinephrine to adrenaline/epinephrine. However, the portocapillary circulation from cortex to medulla is progressively disrupted as a tumour grows, so that very large adrenal tumours may cease to secrete adrenaline/epinephrine.

In cases of borderline biochemistry, pharmacological suppression tests are useful. Either the ganglion-blocking drug pentolinium or centrally acting α₂-agonist clonidine suppresses physiological elevations of metaphrines, but not autonomous secretion from a tumour.^40,41 Provocation tests should not be deliberately employed; but the initial search for phaeochromocytoma may be prompted by a history of hypertensive crisis induced by dopamine antagonists (e.g. metoclopramide) or any drug that releases histamine (opioids, curare, trimetaphan).

Control of blood pressure

before surgery or when the tumour cannot be removed is achieved by α-adrenoceptor blockade, which reverses peripheral vasoconstriction. An important function of the α-blockade is not just blood pressure control, but expansion of intravascular volume. Phaeochromocytoma is the best example of pure vasoconstrictor hypertension, with compensatory pressure natriuresis. Consequently patients are usually volume depleted, often severely, and this must be corrected before it is safe to proceed to surgery.

Not infrequently, the pressure natriuresis completely compensates for vasoconstriction, and patients present with features other than hypertension – even hypotension after an episode of fluid depletion. Such patients need α-blockade prior to surgery in order to volume expand, being at risk of postoperative hypotension if not adequately prepared. β-Blockade may also be required to control tachycardia in patients with adrenaline/epinephrine-secreting tumours. Since adrenaline/epinephrine secretion, as explained above, tends to fall as tumours enlarge, tachycardia is not usually a major problem. Initiation of α-blocker treatment can unmask tachycardia, because there is no longer baroreceptor-induced vagal activation to oppose β-receptor stimulation of the heart. A β-receptor blocker should never be given alone, because abolition of the peripheral vasodilator effects of adrenaline/epinephrine leaves the powerful α effects unopposed. A low dose of a β₁-selective agent (e.g. bisoprolol 5 mg) is safe in the presence of α-blockade. Occasionally, non-selective β-blockade is required, once α-blockade is established, in order to treat β₂-effects (tremor, tachycardia) of an atypical adrenaline/epinephrine-secreting phaeochromoctyoma.

For phaeochromocytoma the preferred α-blocker is not one of the selective α₁-blockers, as in essential hypertension, but the irreversible α-blocker, phenoxybenzamine 10–80 mg daily, whose blockade cannot be overcome by a catecholamine surge, e.g. during tumour manipulation at surgery. Titration of the dose requires inspection of the jugular venous pressure, as index of volume replacement, as well as measurement of blood pressure in the supine and erect position.

During surgical removal – which is usually by laparoscopic adrenalectomy – phentolamine (or sodium nitroprusside) should be at hand to control rises in blood pressure when the tumour is handled. When the adrenal veins have been clamped, volume expansion is often required to maintain blood pressure even after adequate preoperative α-blockade. If a pressor infusion is still needed, isoprenaline is more use than the usual α agonists, to which the patient will be insensitive due to existing α-receptor blockade.

Metirosine (α-methyltyrosine) has been used with some success to block catecholamine synthesis in malignant phaeochromocytomas.

Meta-iodobenzylguanidine (MIBG, an analogue of guanethidine) is actively taken up by adrenergic tissue and is concentrated in phaeochromocytomas. Radio-iodinated MIBG ([¹²³I]MIBG) allows localisation of tumours and detection of metastases, and selective therapeutic irradiation of functioning metastases or other tumours of chromaffin tissue, e.g. carcinoid.

Conn's syndrome

This refers to benign adenomas of the adrenal cortex, which secrete the sodium-retaining hormone aldosterone, and are present in 2–3% of patients with hypertension. Synonyms include primary hyperaldosteronism, although this term can extend to a larger number of patients with elevated plasma aldosterone to renin ratios but no lateralisation of aldosterone secretion. Conn's adenomas are diagnosed by finding a suppressed plasma renin, without suppression of aldosterone, and an adrenal adenoma on CT or MRI. Since 5% of adults have incidental non-functional adrenal adenomas, the key step in diagnosis is lateralisation: the demonstration that the adenoma is responsible for excess aldosterone secretion. Conventionally this is done by adrenal venous sampling. An alternative, in specialist centres, is a PET-CT using a tracer dose of the anaesthetic drug metomidate labelled with F18, which has high affinity binding to the steroid synthases (Fig. 24.4).

  

Fig. 24.4 A PET-CT of a 0.5 cm right adrenal adenoma. The radio-tracer is 11 C-metomidate

(From Burton TJ, MacKenzie IS, Balan K et al 2012. Evaluation of the Sensitivity and Specificity of 11C-Metomidate Positron Emission Tomography (PET)-CT for Lateralizing Aldosterone Secretion by Conn's Adenomas. J Clin Endocrinol Metab 97:100–109, with permission).

Definitive treatment of a proven aldosteronoma is by laparoscopic adrenalectomy. This is recommended in younger patients, who are the ones most likely to have their hypertension cured, and in older patients uncontrolled by, or intolerant of, multiple drugs. Prior to surgery, or longer-term when surgery is not selected, the hypertension and hypokalaemia should be treated by the mineralocorticoid receptor antagonists spironolactone or eplerenone, or by the diuretic amiloride, which inhibits Na⁺ transport through the epithelial Na⁺ channel stimulated by aldosterone. Spironolactone 25–100 mg daily is the most effective, but causes gynaecomastia on chronic dosing. A useful strategy is to combine eplerenone or low-dose spironolactone with amiloride 5–10 mg daily, although regular electrolyte monitoring is necessary to avoid hyperkalaemia and hyponatraemia.

Heart failure and its treatment

Some physiology and pathophysiology

Cardiac output (CO) depends on the rate of contraction of the heart (HR) and the volume of blood that is ejected with each beat, the stroke volume (SV); it is expressed by the relationship:

The three factors that regulate the stroke volume are preload, afterload and contractility:

• Preload is the load on the heart created by the volume of blood received into the left ventricle from the left atrium (at the end of ventricular diastole) and that it must eject with each contraction. It can also be viewed as the amount of stretch to which the left ventricle is subject. As the preload rises so also do the degree of stretch and the length of cardiac muscle fibres. Preload is thus a volume load and can be excessive, e.g. when there is valvular incompetence.

• Afterload refers to the load on the contracting ventricle created by the resistance to the blood projected by the ventricle into the arterial system, i.e. the total peripheral resistance. Afterload is thus a pressure load and is excessive, e.g. in arterial hypertension.

• Contractility refers to the capacity of the myocardium to generate the force necessary to respond to preload and to overcome afterload.

Definition of chronic heart failure

As the population ages and the treatment of acute myocardial infarction improves, this condition is becoming increasingly common and there is talk of an ‘epidemic’ of heart failure. Chronic heart failure is present when the heart cannot provide all organs with the blood supply appropriate to demand. This definition incorporates two elements: firstly, cardiac output may be normal at rest, but secondly, when demand is increased, perfusion of the vital organs (brain and kidneys) continues at the expense of other tissues, especially skeletal muscle. Overall, systemic arterial pressure is sustained until the late stages. These responses follow neuroendocrine activation when the heart begins to fail.

The therapeutic importance of recognising this pathophysiology is that many of the neuroendocrine abnormalities of heart failure – particularly the increased renin output and sympathetic activity – can be a consequence of drug treatment, as well as the disease. Renal perfusion is normal in early heart failure, whereas diuretics and vasodilators stimulate renin and noradrenaline/norepinephrine production through actions at the juxtaglomerular apparatus in the kidney and on the arterial baroreflex, respectively. The earliest endocrine abnormality in almost all types of cardiac disease is increased release of the heart's own hormones, the natriuretic peptides ANP and BNP (A for atrial, B for brain, where it was first discovered). The concentration in plasma of BNP provides a strong prognostic indicator for patients with all stages of heart failure. These peptides normally suppress renin and aldosterone production, but heart failure overrides this control, and measurement of BNP now aids the diagnosis of heart failure, with a raised plasma concentration being a sensitive indicator of the disease.⁴²

The Starling curve and heart failure

The Starling⁴³ curve originally described increased contractility of cardiac muscle fibres in response to increased stretch but, applied to the whole ventricle, it can explain the normal relationship between filling pressure and cardiac output (Fig. 24.5). Most patients with heart failure present in phase ‘A’ of the relationship, and before the ‘decompensated’ phase (B), in which there is gross dilatation of the ventricle. Diuretic therapy improves the congestive symptoms of heart failure, which are due to the increased filling pressure (preload), but actually reduces cardiac output in most patients. Depending on whether their predominant symptom is dyspnoea (due to pulmonary venous congestion) or fatigue (due to reduced cardiac output), patients feel better or worse. It is likely that a principal benefit of using angiotensin-converting enzyme (ACE) inhibitors in heart failure is their diuretic sparing effect.

Fig. 24.5 The Starling curve of the relationship between cardiac filling pressure and cardiac output. In phase A, the curve shows that lowering the blood volume (by diuretics) will reduce the filling pressure but the cardiac output will fall. In phase B, lowering the blood volume will reduce the filling pressure but the cardiac output will increase (see text).

Natural history of chronic heart failure

Injury to the heart, e.g. myocardial infarction, hypertension, leads to adaptive (‘compensatory’) molecular, cellular and interstitial changes that alter its size, shape and function. Myocardial hypertophy and ‘remodelling’ takes place over weeks or months in response to haemodynamic load, neurohormonal activation and other factors, and the resulting pattern differs according to whether the stimulus is a pressure or volume overload. With the passage of time, and with maladaption, the heart ‘decompensates’ and heart failure worsens. The process is outlined in Figure 24.6.

Fig. 24.6 Progression of ventricular disease from hypertrophy through to failure following the initial insult. The process is generically referred to as ‘cardiac remodelling’ and the aim of much drug therapy in heart failure treatment is to try to place a brake on this process.

The degree of activity that the patient can undertake without becoming dyspnoeic provides one useful classification of the severity of heart failure. The New York Heart Association (NYHA) classification⁴⁴ offers also an approximate prognosis, with that of the worst grade (Class IV) being as bad as most cancers. Many patients with heart failure die from an arrhythmia, rather than from terminal decompensation, and drugs that avoid increasing the heart's exposure to increased catecholamine concentrations, as do some vasodilators (but see below), appear best for improving prognosis.

Objectives of treatment

As for cardiac arrhythmias, these are to reduce morbidity (relief of symptoms, avoid hospital admission) and mortality.

There is some tension between these two objectives in that the condition is both disabling and deadly, and the action of diuretic and some vasodilator drugs, which temporarily improve symptoms, can jeopardise survival. There is a further tension between the needs of treating the features of forwards failure, or low output, and backwards failure, or the congestive features. The principal symptom of a low cardiac output, fatigue, is difficult to quantify, and patients have tended to have their treatment tailored more to the consequences of venous congestion.

Haemodynamic aims of drug therapy

Acute or chronic failure of the heart usually results from disease of the myocardium itself, mainly ischaemic, or an excessive load imposed on it by arterial hypertension, valvular disease or an arteriovenous shunt. The management of chronic heart failure requires both the relief of any treatable underlying or aggravating cause, and therapy directed at the failure itself.

Distinguishing between the capacity of the myocardium to pump blood and the load against which the heart must work is useful in therapy. The failing myocardium is so strongly stimulated to contract by increased sympathetic drive that therapeutic efforts to induce it to function yet more vigorously are in themselves alone unlikely to be of benefit. Despite numerous candidate drugs introduced over recent years, digoxin remains the only inotropic drug suitable for chronic oral use.

By contrast, agents that reduce preload or afterload can be very effective, especially where the left ventricular volume is increased (less predictably so for failure of the right ventricle). The main hazard of their use is a drastic fall in cardiac output in those occasional patients whose output is dependent on a high left ventricular filling pressure, e.g. who are volume depleted by diuretic use or have severe mitral stenosis.

Classification of drugs

Reduction of preload

Diuretics

increase salt and water loss, reduce blood volume and lower excessive venous filling pressure (see Ch. 27). They are almost invariably required to relieve the congestive features of oedema, in the lungs and the periphery; when the heart is grossly enlarged, cardiac output will also increase (see discussion of Starling curve, above). They are used flexibly, starting with a low dose; the usual sequence would be to begin with a thiazide, then move to furosemide, and in the most extreme cases then judiciously add metolazone.

Nitrates

(see also Ch. 24) dilate the smooth muscle in venous capacitance vessels, increase the volume of the venous vascular bed (which normally may comprise 80% of the whole vascular system), reduce ventricular filling pressure, thus decreasing heart wall stretch, and reduce myocardial oxygen requirements. Their arteriolar dilating action is relatively slight. Glyceryl trinitrate provides benefit in acute left ventricular failure sublingually or by intravenous infusion. For chronic left ventricular failure nitrates have fallen out of favour, because the ACE inhibitors are more effective.

Reduction of afterload

Hydralazine (see also Ch. 24) relaxes arterial smooth muscle and reduces peripheral vascular resistance. Reflex tachycardia limits its usefulness and lupus erythematosus is a risk usually only if the dose exceeds 100 mg per day.

Reduction of preload and afterload

ACE inhibitors and angiotensin receptor II blockers (ARBs)

(see also Ch. 24) act by:

• reduction of afterload, by preventing the conversion of angiotensin I to the active form, angiotensin II, or by blocking the effects of angiotensin II, which is a powerful arterioconstrictor and is present in the plasma in high concentration in heart failure

• reduction of preload, because the formation of aldosterone, and thus retention of salt and water (increased blood volume), is prevented by reducing the effects of angiotensin II.

ACE inhibitors are the only drugs that reduce peripheral resistance (afterload) without causing a reflex activation of the sympathetic system. The landmark CONSENSUS study compared enalapril with placebo in patients with NYHA class IV heart failure; after 6 months 26% of the enalapril group had died, compared with 44% in the control group. The reduction in mortality occurred among patients with progressive heart failure.⁴⁵ There is now strong evidence from long-term studies that ACE inhibitors⁴⁶ and ARBs⁴⁷ improve survival in and reduce hospital admissions for heart failure.

A common practice has been to give a test dose of a short-acting ACE inhibitor (e.g. ramipril 1.25 mg by mouth) to patients who are in heart failure or on diuretic therapy for another reason, e.g. hypertension. Maintenance of blood pressure in such individuals may depend greatly on an activated renin–angiotensin–aldosterone system, and a standard dose of an ACE inhibitor or ARB can cause a sudden fall in blood pressure. That said, some of the many ACE inhibitors now available (see p. 399) have a sufficiently prolonged action that the initial doses have a cumulative effect on blood pressure over several days. Long-acting ACE inhibitors such as lisinopril (t_½ 12 h) and perindopril (t_½ 31 h) avoid the risk of sudden falls in blood pressure or renal function (glomerular filtration) after the first dose. Such drugs can be initiated outside hospital in patients who are unlikely to have a high plasma renin (absence of gross oedema or widespread atherosclerotic disease), although it is prudent to arrange for the first dose to be taken just before going to bed. Therapy begins with an ACE inhibitor, and an ARB is substituted if there is intolerance, or added if symptoms continue.

β-Adrenoceptor blockers

The realisation that activation of the renin–angiotensin–aldosterone and sympathetic nervous systems can adversely affect the course of chronic heart failure led to exploration of the possible benefits in heart failure from blockade of β-adrenoceptors. Clinical trials have, indeed, shown that bisoprolol, carvedilol or metoprolol lower mortality and decrease hospitalisation when added to conventional treatment that is likely to include diuretics, digoxin and an ACE inhibitor (see below).

Spironolactone

Plasma aldosterone levels are raised in heart failure. Spironolactone acts as a diuretic by competitively blocking the aldosterone receptor, but in addition it has a powerful effect on outcome in heart failure (see below). Eplerenone, an alternative mineralocorticoid antagonist, also has beneficial effects including in mild to moderate heart failure.

Stimulation of the myocardium

Digoxin

improves myocardial contractility (positive inotropic effect) most effectively in the dilated, failing heart and, in the longer term, after an episode of heart failure has been brought under control. This effect occurs in patients in sinus rhythm and is distinct from its (negative chronotropic) action of reducing ventricular rate and thus improving ventricular filling in atrial fibrillation. Over 200 years after the first use of digitalis for dropsy, the DIG trial provided relief for doctors seeking evidence of long-term benefit.⁴⁸ Unlike all other positive inotropes, digoxin does not increase overall mortality or arrhythmias.

The phosphodiesterase inhibitors enoximone and milrinone have positive inotropic effects due to selective myocardial enzyme inhibition and may be used for short-term treatment of severe congestive heart failure. Evidence from longer-term use indicates that these drugs reduce survival.

Drug management of heart failure

Chronic heart failure

A scheme for the stepwise drug management of chronic heart failure appears in Figure 24.7. Points to emphasise in this scheme are that all patients, even those with mild heart failure, should receive an ACE inhibitor as first-line therapy. Several long-term studies have demonstrated improved survival even when cardiac failure is mild.⁴⁹

Fig. 24.7 Treatment algorithm for patients with heart failure and reduced left ventricular systolic function. *These treatments generally supplement existing diuretic drugs, with flexible dosing to maintain dry weight; consider hydralazine plus isosorbide dinitrate for black patients, as clinical trials indicate a favourable response in this subgroup. † The safety and efficacy of the combination of an ACE inhibitor, an ARB, and an aldosterone antagonist are unknown. NYHA, New York Heart Association.

(From McMurray J J, Pfeffer M A 2005 Heart failure. Lancet 365:1877–1889 with permission from The Lancet.)

Black patients have a less activated renin system than other ethnic groups. In a landmark study, following subgroup analyses of earlier trials, 1050 black patients who had New York Heart Association class III or IV heart failure with dilated ventricles were randomly assigned to receive a combination of isosorbide dinitrate 120 mg daily plus hydralazine 225 mg daily or placebo in addition to standard therapy for heart failure. The study was terminated early owing to a significantly higher mortality rate in the placebo group, 10.2%, than in the group receiving the active combination, 6.2%, P = 0.02.⁵⁰

Diuretic therapy is very useful for symptom management but has no impact on survival. For most patients the choice will be a loop diuretic, e.g. furosemide starting at 20–40 mg/day. Because of the potassium-sparing effect of ACE inhibition, amiloride (also potassium sparing) is often not required, at least with low doses of a loop diuretic.

There is now overwhelming evidence for the benefit of β-blockers in chronic heart failure, despite the long-held belief that their negative inotropic effect was a contraindication. Early trials were underpowered but a meta-analysis did suggest a 31% reduction in the mortality rate. Subsequently, the CIBIS-2 and MERIT-HF trials, have independently confirmed that chronic β-blockade has a survival effect of this size in moderate to severe (NHYA III/IV) heart failure.⁵¹ Both studies confirmed the one-third reduction in mortality. In MERIT-HF a life was saved for just 27 patient-years of treatment, i.e. it was unusually cost effective – more so than ACE inhibitor therapy. The action is probably a class effect of β-blockade, given the divergent pharmacology of the drugs used to date.

The reduction in mortality is additive to ACE inhibition and the survival benefit is largely through a decrease in sudden deaths as opposed to a reduction in progressive pump failure seen with ACE inhibitors. The only cautionary note is that patients must be β-blocked very gradually from low starting doses (e.g. bisoprolol 1.25 mg/day or carvedilol 3.125 mg twice daily) with regular optimisation of the dose of other drugs, especially the loop diuretic, to prevent decompensation of heart failure control.

The use of spironolactone has received considerable support from the RALES trial,⁵² which implies that ACE inhibition even at high dose does not effectively suppress hyperaldosteronism in heart failure. The benefit occurs at a surprisingly low dose of spironolactone (25 mg/day); it probably reflects both improved potassium and magnesium conservation (both are antiarrhythmic) and reversal of fibrosis in the myocardium by aldosterone.

None of the available oral phosphodiesterase inhibitors is established in routine therapy, because the short-term benefit of the increased contractility is offset by an increased mortality rate (presumably due to arrhythmias) on chronic dosing. Their use is restricted to short-term symptom control prior to, for example, transplantation.

Acute left ventricular failure

This is a common medical emergency, despite possible lessening in frequency with the advent of thrombolysis for myocardial infarction. The approach should be to reassure the anxious patient, who should sit upright with their legs dependent to reduce systemic venous return. A loop diuretic, e.g. furosemide 40–80 mg i.v., is the mainstay of therapy and provides benefit both by a rapid and powerful venodilator effect, reducing preload, and by the subsequent diuresis. Oxygen should be given, if the patient can tolerate a face mask, and diamorphine or morphine i.v. which, in addition to relieving anxiety and pain, have valuable venodilator effects.

Although there may be a case for short-term use of inotropic drugs (see Ch. 25) for heart failure where low output is a dominating feature, most such drugs substantially increase the risk of arrhythmias when the heart is hypoxic. The pharmacokinetics of digoxin does not favour emergency use. The possibility of assisted ventilation should be considered; where pulmonary oedema is the main problem, ventilation is likely to be both safer and more effective than inotropic drugs.

Surgery for heart failure

Although these options lie outside the scope of clinical pharmacology, an important element in meeting the objectives of treatment (see p. 422) is to recognise when further drug treatment is unlikely to improve symptoms or prognosis. Then, the physician must consider the possibility of a surgical intervention. Increasingly this may involve procedures short of transplantation itself, e.g. bypass grafting or stenting where stenosed vessels contribute to the heart failure or even a left ventricular assist device (LVAD) or totally artificial heart. On occasion, it can help to make the patient aware that failure of both the heart and the drugs is not necessarily the end of the road.

Summary

• Heart failure is present when the heart cannot provide all organs with the blood supply appropriate to demand.

• Stroke volume is regulated by preload, afterload and contractility.

• In chronic heart failure, diuretics and nitrates reduce preload and provide symptomatic relief without affecting outcome.

• ACE inhibitors reduce both preload and afterload, and reduce morbidity and mortality by about one-third in all patients.

• β-Adrenoceptor blockers, gradually introduced, have an effect equivalent to that of ACE inhibitors in patients with moderate or severe heart failure (NYHA III or IV).

• Spironolactone, in low dose, adds further benefit.

• Digoxin improves myocardial contractility most effectively in the dilated, failing heart but also in the longer term, including in patients in sinus rhythm.

• The principal agents for treating acute left ventricular failure are furosemide, diamorphine and oxygen.

Guide to further reading

Armstrong P.W. Aldosterone antagonists – last man standing? N. Engl. J. Med.. 2011;364:79–80.

Ashrafian H., Williams L., Frenneaux M.P. The pathophysiology of heart failure: a tale of two paradigms revisited. Clin. Med. (Northfield Il). 2008;8(2):192–197.

Braunwald E. Biomarkers in heart failure. N. Engl. J. Med.. 2009;358(20):2148–2159.

Brown M.J. Hypertension and ethnic group. Br. Med. J.. 2006;332:833–836.

Brown M.J. Renin: friend or foe? Heart. 2007;93:1026–1033.

Brown M.J. Heterogeneity of blood pressure response to therapy. Am. J. Hypertens.. 2010;23:926–928.

Brown M.J., Secondary hypertension: Warrell. D., Cox T., Firth. J. Oxford Textbook of Medicine, fifth ed, Oxford: Oxford University Press, 2010. (Chapter 16.17.3)

Brown M.J. Aliskiren. Circulation. 2008;118:773–784.

Brown M.J., Cruickshank J.K., Macdonald T.M. Navigating the shoals in hypertension: discovery and guidance. Br. Med. J.. 2012;344:23–26.

Camm A.J., Kirchhof P., Lip G.Y., et al. Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC). Eur. Heart J.. 2010;31:2369–2429.

Crystal E., Connolly S.J. Role of oral anticoagulation in management of atrial fibrillation. Heart. 2004;90:813–817.

Delacretaz E. Clinical practice. Supraventricular tachycardia. N. Engl. J. Med.. 2006;354:1039–1051.

Dobrev D., Nattel S. New antiarrhythmic drugs for treatment of atrial fibrillation. Lancet. 2010;375:1212–1223.

Duley L., Meher S., Abalos A. Clinical review: management of pre-eclampsia. Br. Med. J.. 2006;332:463–468. Available online at: http://www.bmj.com/cgi/content/extract/332/7539/463 (accessed 2 August 2010)

Dworkin L.D., Cooper C.J. Clinical practice: renal artery stenosis. N. Engl. J. Med.. 2009;361:1972–1978. Available online at: http://www.nejm.org/doi/pdf/10.1056/NEJMcp0809200 (accessed 16 November 2011)

Gaziano T.A., Opie L.H., Weinstein M.S., et al. Cardiovascular disease prevention with a multidrug regimen in the developing world: a cost–effectiveness analysis. Lancet. 2006;368:679–686.

Hansson G.K. Inflammation, atherosclerosis, and coronary artery disease. N. Engl. J. Med.. 2005;352:1685–1695.

Hillis L.D., Lange R.A. Optimal management of acute coronary syndromes. N. Engl. J. Med.. 2009;360:2237–2239. Available online at: http://www.nejm.org/doi/pdf/10.1056/NEJMe0902632 (accessed 2 August 2010)

Huikuri H.V., Castellanos A., Myerburg R.J., et al. Sudden death due to cardiac arrhythmias. N. Engl. J. Med.. 2001;345(20):1473–1482.

Jarcho J.A. Resynchronizing ventricular contraction in heart failure. N. Engl. J. Med.. 2005;352:1594–1597.

JBS 2. Joint British Societies’ guidelines on prevention of cardiovascular disease in clinical practice. Heart. 91(Suppl. 5), 2005. Available online at: http://www.bcs.com/download/651/JBS2final.pdf (accessed 2 August 2010)

Kaplan N.M., Opie L.H. Controversies in hypertension. Lancet. 2006;367:168–176.

Krum H., Abraham W.T. Heart failure. Lancet. 2009;373:41–955.

Lip G.Y., Halperin J.L. Improving stroke risk stratification in atrial fibrillation. Am. J. Med.. 2010;123:484–488.

McMurray J.J. Systolic heart failure. N. Engl. J. Med.. 2010;362:228–238.

Messerli F.H. This day 50 years ago. N. Engl. J. Med.. 1995;332(15):1038–1039. [an account of the hypertension and stroke suffered by US President F D Roosevelt]

Morady F. Catheter ablation of supraventricular arrhythmias: state of the art. J. Cardiovasc. Electrophysiol.. 2004;15(1):124–139.

Neubauer S. The failing heart – an engine out of fuel. N. Engl. J. Med.. 2007;356:1140–1151.

Page R.L. Newly diagnosed atrial fibrillation. N. Engl. J. Med.. 2004;351(23):2408–2416.

Page R.L., Roden D.M. Drug therapy for atrial fibrillation: where do we go from here? Nat. Rev. Drug. Discov.. 2005;4(11):899–910.

Pickering T.G., Shimbo D., Haas D., et al. Ambulatory blood-pressure monitoring. N. Engl. J. Med.. 2006;354:2368–2374.

Schmieder R.E., Hilgers K.F., Schlaich M.P., Schmidt B.M.W. Renin-angiotensin system and cardiovascular risk. Lancet. 2007;369:1208–1219.

Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC7). Hypertension. 2003;42:1206. Available online at: http://www.nhlbi.nih.gov/guidelines/hypertension/ (accessed 2 August 2010)

Staessen J.A., Li Y., Richart T. Oral renin inhibitors. Lancet. 2006;368:1449–1456.

Torp-Pedersen C., Pedersen O.D., Kober L. Antiarrhythmic drugs: safety first. J. Am. Coll. Cardiol.. 2010;55:1577–1579.

Turnbull F. Effects of different blood-pressure-lowering regimens on major cardiovascular events: results of prospectively-designed overviews of randomised trials. Lancet. 2003;362:1527–1535.

Vaughan C.J., Delanty N. Hypertensive emergencies. Lancet. 2000;356:411–417.

Williams B., Poulter N.R., Brown M.J., et al. British Hypertension Society guidelines for hypertension management 2004 (BHS-IV): summary. Br. Med. J.. 2004;328:634–640.

Zimetbaum P. Amiodarone for atrial fibrillation. N. Engl. J. Med.. 2007;356:935–941.

¹ Murrell W 1879 Nitroglycerin as a remedy for angina pectoris. Lancet i:80–81. Nitroglycerin was actually first synthesised by Sobrero in 1847 who noted that, when he applied it to his tongue, it caused a severe headache.

² Journal of Clinical Investigation 2006; 116:506–511. This same coding variant confers flushing to alcohol challenge and in parts of South-East Asia has a prevalence of almost 50%.

³ Explosives factory workers exposed to a nitrate-contaminated environment lost it over a weekend and some chose to maintain their intake by using nitrate-impregnated headbands (transdermal absorption) rather than have to accept the headaches and re-acquire tolerance so frequently. A recent study has also reported that patients with angina who develop a headache with GTN are less likely to have obstructive coronary artery disease (His D H, Roshandel A, Singh N, Szombathy T, Meszaros Z S 2005 Headache response to glyceryl trinitrate in patients with and without obstructive coronary artery disease. Heart 91:1164–1166).

⁴ It has been argued that deaths on sildenafil largely reflect the fact that it is used by patients at high cardiovascular risk. But post-marketing data show that death is 50 times more likely after sildenafil taken for erectile failure than alprostadil, the previous first-line agent (Mitka M 2000 Some men who take Viagra die – why? Journal of the American Medical Association 283:590–593).

⁵ Useful, but not always safe. Defibrillator paddles and nitrate patches make an explosive combination, and it is not always in the patient's interest to have the patch as unobtrusive as possible (see Canadian Medical Association Journal 1993; 148:790).

⁶ Several calcium-selective channels have been described in different tissues, e.g. the N (present in neuronal tissue) and T (transient, found in brain, neuronal and cardiovascular pacemaker tissue); the drugs discussed here selectively target the L-channel for its cardiovascular importance.

⁷ Both the NORDIL and INSIGHT trials confirmed that a calcium channel blocker (diltiazem and nifedipine respectively) had the same efficacy as older therapies (diuretics and/or β-blockers) in hypertension, with no evidence of increased sudden death (Hansson L, Hedner T, Lund-Johansen P et al 2000 Randomised trial of effects of calcium antagonists compared with diuretics and beta-blockers on cardiovascular morbidity and mortality in hypertension: the Nordic Diltiazem [NORDIL] study. 356:359–365; Brown M J, Palmer C R, Castaigne A et al 2000 Morbidity and mortality in patients randomised to double-blind treatment with a long-acting calcium-channel blocker or diuretic in the International Nifedipine GITS study: Intervention as a Goal in Hypertension Treatment [INSIGHT]. Lancet 356:355–372).

⁸ PRAISE – Prospective Randomised Amlodipine Survival Evaluation (see Packer M, O'Connor C M, Ghali J K et al 1996 The effect of amlodipine on morbidity and mortality in severe chronic heart failure. New England Journal of Medicine 335:1107–1114).

⁹ Demers C, McMurray J J V, Swedberg K et al for the CHARM investigators 2005 Impact of candesartan on nonfatal myocardial infarction and cardiovascular death in patients with heart failure. Journal of the American Medical Association 294:1794–1798.

¹⁰ For a review see: Ruggenenti P, Cravedi P, Remuzzi G 2010 The RAAS in the pathogenesis and treatment of diabetic nephropathy. Nature Reviews Nephrology 6:319–330. Available online at: http://www.nature.com/nrneph/journal/v6/n6/full/nrneph.2010.58.html (accessed 3 August 2010).

¹¹ Ruggenenti P, Fassi A, Ilieva A P et al 2004 Preventing microalbuminuria in type 2 diabetes. New England Journal of Medicine 351:1941–1951. This was the BENEDICT trial comparing type 2 diabetics randomised to trandolopril, verapamil and the combination versus placebo with a 3.6 year follow-up.

¹² Pfeffer M A, Braunwald E, Moye L A et al 1992 Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. Results of the survival and ventricular enlargement trial. The SAVE Investigators. New England Journal of Medicine 327:669–677.

¹³ Pfeffer M A, McMurray J V C, Velaquez E J et al for the Valsartan in Acute Myocardial Infarction Trial Investigators 2004 Valsartan, captopril, or both in myocardial infarction complicated by heart failure, left ventricular dysfunction, or both. New England Journal of Medicine 349:1893–1906.

¹⁴ Yusuf S, Sleight P, Pogue J et al 2000 Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. New England Journal of Medicine 342:145–153.

¹⁵ Dahlof B, Devereux R B, Kjeldsen S E et al 2002 Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 359:995–1010.

¹⁶ As we went to press, an outcome study (‘Altitude’) was stopped early because the hypothesised benefit of adding aliskiren to ACE inhibitors or ARBs in patients with diabetes showed no benefit. A borderline significant excess of non-fatal strokes led to a provisional warning against the combination in patients with diabetes.

¹⁷ Light causes sodium nitroprusside in solution to decompose; hence solutions should be made fresh and immediately protected by an opaque cover, e.g. aluminium foil. The fresh solution has a faint brown colour; if the colour intensifies it should be discarded.

¹⁸ The Impact Of Nicorandil in Angina (IONA) study was a double-blind, randomised, placebo-controlled trial conducted in the UK, in which high-risk patients with stable angina were assigned placebo or nicorandil 10–20 mg. Over a mean follow-up of 1.6 years significantly more placebo-treated patients suffered an acute coronary syndrome or coronary death (15.5% versus 13.1%, P = 0.01) (IONA Study Group 2002 Effect of nicorandil on coronary events in patients with stable angina: the Impact Of Nicorandil in Angina (IONA) randomised trial. Lancet 359:1269–1275).

¹⁹ Man-Son-Hing M, Wells G 1995 Meta-analysis of efficacy of quinine for treatment of nocturnal cramps in elderly people. British Medical Journal 310:13–17.

²⁰ It can be the reflex sympathetic activation, as much as hypotension itself, that causes problems. Many cardiologists have had their efforts at controlling angina in elderly patients sabotaged when the patient visits a urologist for his prostatic symptoms, and is treated with a powerful α₁-blocker.

²¹ In the first major study, sudden death occurred in 13.9% of placebo-treated and 7.7% of timolol-treated patients (Norwegian Multicentre Study Group 1981 Timolol-induced reduction in mortality and reinfarction in patients surviving myocardial infarction. New England Journal of Medicine 304:801–807).

²² A 36-year-old patient with asthma collected, from a pharmacy, chlorphenamine for herself and oxprenolol for a friend. She took a tablet of oxprenolol by mistake. Wheezing began in 1 h and worsened rapidly; she experienced a convulsion, respiratory arrest and ventricular fibrillation. She was treated with positive-pressure ventilation (for 11 h) and intravenous salbutamol, aminophylline and hydrocortisone, and survived (Williams I P, Millard F J 1980 Severe asthma after inadvertent ingestion of oxprenolol. Thorax 35:160). There is a logical – or rather pharmacological – link between the use of timolol as eye drops and the risk of asthma. For local administration, a drug needs high potency, so that a high degree of receptor blockade is achieved using a physically small (and therefore locally administrable) dose of drug. Nevertheless, timolol is used topically as a 0.25–0.5% solution, which means the initial concentration of timolol in the tear film is up to 5 mg/mL (or > 10 mmol/L). As the majority of this will be swallowed and a few milligrams orally will block systemic β₂ receptors, it is apparent why one drop of timolol down the lachrymal duct (of the wrong patient) is hazardous.

²³ Müller M E, van der Velde N, Krulder J W M, van der Cammen T J M 2006 Syncope and falls due to timolol eye drops. British Medical Journal 332:960–961.

²⁴ Practolol was developed to the highest current scientific standards; it was marketed in 1970 as the first cardioselective β-blocker, and only after independent review by the UK drug regulatory body. All seemed to go well for about 4 years, by which time there had accumulated about 200 000 patient-years of experience with the drug. It then became apparent that a small proportion of patients taking practolol could develop a bizarre syndrome that included conjunctival scarring, nasal and mucosal ulceration, fibrous peritonitis, pleurisy and cochlear damage (oculomucocutaneous syndrome). The condition was first recognised by an alert ophthalmologist who ran a special clinic for external eye diseases. (See Wright P 1975 Untoward effects associated with practolol administration: oculomucocutaneous syndrome. British Medical Journal i:595–589.)

²⁵ Angina pectoris: angina, a strangling; pectoris, of the chest.

²⁶ For a personal account by a physician of his experiences of angina pectoris, coronary bypass surgery, ventricular fibrillation and recovery, see Swyer G I M 1986 Personal view. British Medical Journal 292:337. Compelling and essential reading.

²⁷ First International Study of Infarct Survival Collaborative Group 1986 Randomised trial of intravenous atenolol among 16027 cases of suspected acute myocardial infarction: ISIS-1. Lancet ii:57–66.

²⁸ In the SAVE (Survival and Ventricular Enlargement) study, captopril 50 mg three times daily or placebo was started 3–16 days after MI in 2231 patients without overt cardiac failure but with a left ventricular ejection fraction of less than 40%. The captopril group had a lower incidence of recurrent myocardial infarction (133) and death (228) than the placebo group (170 and 275 respectively) (Rutherford J D, Moye L A, Pfeffer M A et al for the SAVE investigators 1994 Effects of captopril on ischemic events after myocardial infarction. Results of the Survival and Ventricular Enlargement trial. SAVE Investigators. Circulation 90:1731–1738). Several other trials of ACE inhibitors have provided similar results.

²⁹ In the Heart Protection Study of 20 536 high-risk patients (one-third had previous MI), those randomly assigned to simvastatin 40 mg daily (compared with placebo) had a 12% reduction in all-cause mortality, and 24% reduction in strokes and coronary heart disease. The authors estimated that 5 years of statin treatment will prevent 100 major vascular events in every 1000 patients with previous MI, or 70 to 80 events in patients with other forms of coronary heart disease or diabetes. There was no upper age limit to this benefit, and no lower limit to the level of LDL at which benefit was seen (Heart Protection Study Collaborative Group 2002 MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20536 high-risk individuals. Lancet 360:7–22).

³⁰ The British Hypertension Society Guidelines (BHS-IV) are summarised in British Medical Journal 2004; 328:634–640 or online at http://www.bhsoc.org. Joint guidance with the National Institute of Clinical Excellence (NICE) was issued in 2006. The update of 2011 is flawed and the Editors advise following the 2006 recommendations (see Brown MJ et al 2012, in Guide to Further Reading).

³¹ DASH-Sodium Collaborative Research Group 2001 Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. New England Journal of Medicine 344:3–10.

³² Available online at: http://www.nice.org.uk/CG034

³³ The original schema included β-blockers, i.e. four drug groups AB/CD (Dickerson J E C, Hingorani A D, Ashby M J, Palmer C R, Brown M J 1999 Optimisation of anti-hypertensive treatment by crossover rotation of four major classes. Lancet 353:2008–2013). This is revised in the light of clinical trial evidence that β-blockers are usually less effective than other antihypertensives at reducing major cardiovascular events, particularly stroke, and are associated with an unacceptably high risk of diabetes especially in combination with diuretics. Hence they are no longer recommended either as monotherapy (B in the original schema) or in combination with diuretics (B + D) unless there is a second indication for prescribing the β-blocker, e.g. angina, or in women planning to have children.

³⁴ Brown M J, McInnes G T, Papst C C et al 2011 Aliskiren and the calcium channel blocker amlodipine combination as an initial treatment strategy for hypertension control (ACCELERATE): a randomised, parallel-group trial. Lancet. 377:312–320.

³⁵ J Venn (1834–1923), an English logician who ‘adopted the diagrammatic method of illustrating propositions by inclusive and exclusive circles’ (Dictionary of National Biography). A medical pilgrimage to Cambridge, where Venn worked, should take in Gonville and Caius College (named after its founder, Dr Caius, physician to the Tudor court and early president of the London College of Physicians in the 16th century); as well as stained glass windows celebrating Venn's circles, the visitor can see a portrait of the most famous medical Caian, William Harvey.

³⁶ Cooper W O, Hernandez-Diaz S, Arbogast P G et al 2006 Major congenital malformations after first-trimester exposure to ACE inhibitors. New England Journal of Medicine 354:2443–2451.

³⁷ Magpie Trial Collaborative Group 2002 Do women with pre-eclampsia, and their babies, benefit from magnesium sulphate? The Magpie Trial: a randomised placebo-controlled trial. Lancet 359:1877–1890.

³⁸ Eclampsia Trial Collaborative Group 1995 Which anticonvulsant for women with eclampsia? Evidence from the Collaborative Eclampsia Trial. Lancet 345:1455–1463.

³⁹ Barst R J, Rubin L J, Long W A et al 1996 A comparison of continuous intravenous epoprostenol (prostacyclin) with conventional therapy for primary pulmonary hypertension. New England Journal of Medicine 334:296–302. In a trial of 81 patients with severe PAH, a 12-week infusion of epoprostenol improved quality of life, mean pulmonary arterial pressure (− 8 versus + 3 %), pulmonary vascular resistance (− 21 versus + 9 %), and exercise capacity, as measured by a 6-min walk test (+ 47 versus − 66 meters). Eight patients died during the trial, all of whom were in the standard therapy group.

⁴⁰ Brown M J, Allison D J, Jenner D A et al 1981 Increased sensitivity and accuracy of phaeochromocytoma diagnosis achieved by plasma adrenaline estimations and a pentolinium suppression test. Lancet ii:174–177.

⁴¹ Bravo E L, Tarazi R C, Fouad F M et al 1981 Clonidine-suppression test: a useful aid in the diagnosis of pheochromocytoma. New England Journal of Medicine 305:623–626.

⁴² Braunwald E 2008 Biomarkers in heart failure. New England Journal of Medicine 358:2148–2159.

⁴³ Ernest Henry Starling, 1866–1927, Professor of Physiology, University College, London. He also coined the word ‘hormone’.

⁴⁴ NYHA Class I, minimal dyspnoea (except after moderate exercise); Class II, dyspnoea while walking on the flat; Class III, dyspnoea on getting in/out of bed; Class IV, dyspnoea while lying in bed.

⁴⁵ Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS) Trial Study Group 1987 Effects of enalapril on mortality in severe congestive heart failure. New England Journal of Medicine 316:1430–1435.

⁴⁶ Flather M D, Yusuf S, Kober L et al for the ACE-Inhibitor Myocardial Infarction Collaborative Group 2000 Long-term ACE-inhibitor therapy in patients with heart failure or left-ventricular dysfunction: a systematic overview of data from individual patients. Lancet 355:1575–1587.

⁴⁷ Demers C, McMurray J J V, Swedberg K et al for the CHARM investigators 2005 Impact of candesartan on nonfatal myocardial infarction and cardiovascular death in patients with heart failure. Journal of the American Medical Association 294:1794–1798.

⁴⁸ This prospective randomised trial compared digoxin with placebo in 7788 patients in NYHA Class II–III heart failure and sinus rhythm, most of whom also received an ACE inhibitor and a diuretic. Overall mortality did not differ between the groups but patients who took digoxin had fewer episodes of hospitalisation for worsening heart failure (Digitalis Investigation Group 1997 The effect of digoxin on mortality and morbidity in patients with heart failure. New England Journal of Medicine 336:525–532).

⁴⁹ In the Studies of Left Ventricular Dysfunction (SOLVD, enalapril was compared with placebo in patients with either clinical features of heart failure or reduced left ventricular function in the absence of symptoms. Treatment reduced serious events (myocardial infarction and unstable angina) by approximately 20% and hospital admissions with progressive heart failure by up to 40%. (SOLVD Investigators 1991 Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. New England Journal of Medicine 325:293–302).

⁵⁰ Taylor A L, Ziesche S, Yancy C et al 2004 Combination of isosorbide dinitrate and hydralazine in blacks with heart failure. New England Journal of Medicine 351:2049–2057.

⁵¹ Until 1997, 24 trials of β-blockade in heart failure provided just 3141 patients. MERIT-HF (1999 Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 353:2001–2007) alone contained 3991 patients and CIBIS-2 (1999 The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet 353:9–13) provided a further 2467.

⁵² The RALES trial randomised 1663 patients with stable heart failure to either placebo or spironolactone. All patients maintained their ‘optimised’ therapy, which included ACE inhibitors. After 2 years of follow-up, the trial terminated prematurely following the demonstration of a 30% reduction in the mortality rate of spironolactone-treated patients, from sudden death as well as progressive pump failure. Gynaecomastia or breast discomfort occurred in 10% of patients receiving spironolactone (1% in controls), but significant hyperkalaemia occurred in surprisingly few patients. RALES was not adequately powered to decide whether the action of spironolactone was additive to that of a β-blocker (Pitt B, Zannad F, Remme W J et al 1999 The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators. New England Journal of Medicine 341(10):709–717).