23 Heart Failure

Although morbidity and mortality are improving, heart failure continues to present major challenges to healthcare systems. This affliction of mainly the elderly may be falling in incidence but is probably growing in prevalence, in part due to greater longevity resulting from evidence-based drug and device therapy for patients with a low ejection fraction (EF). New guidelines emphasize the combination of an angiotensin-converting enzyme inhibitor and beta-blocker as the cornerstone of therapy, with the addition of either an angiotensin receptor blocker or aldosterone antagonist as the third disease-modifying agent in patients who remain symptomatic. An implanted cardioverter defibrillator should also be added in patients with a persistently low EF and life expectancy of reasonable quality of ≥12 months. In patients in sinus rhythm, New York Heart Association class III–IV, and a QRS duration ≥120ms, cardiac resynchronization therapy has substantial additional morbidity and mortality benefits. This evidence-based care should be delivered through an organized and seamless multidisciplinary framework with a focus on the patients’ and carers’ needs. Apart from transplantation, the place of any other surgical intervention remains uncertain. Unfortunately the success of treatment in low EF heart failure has not been replicated in patients with heart failure and a preserved EF, and treatment of these patients remains empirical. The same is largely true for the treatment of patients with acute heart failure, with no treatment yet shown to be superior to empirical therapy with diuretic, oxygen, and nitrates. In particular the role of inotropes remains uncertain.

Heart failure is the term used to describe a common clinical syndrome arising, in ways that are incompletely understood, as a consequence of reduced cardiac pump function. The term ‘syndrome’ merely describes a constellation of symptoms and signs and, therefore, heart failure is not a diagnosis as such. Unfortunately, the typical symptoms (breathlessness and fatigue) and signs (e.g. oedema) of heart failure are relatively non-specific, making clinical confirmation of the syndrome difficult. The syndrome of heart failure itself can arise as a result of almost any abnormality of the structure, mechanical function, or electrical activity of the heart, each of which may require quite different treatments, emphasizing the importance of appropriate investigation of patients with suspected heart failure.

Many of the typical clinical symptoms and signs of heart failure do not arise directly as a result of the cardiac abnormality but rather from secondary dysfunction of other organs and tissues, e.g. the kidneys, bone marrow, and muscles. These secondary consequences of pump failure are myriad and their causes are not fully elucidated. Dysfunction of tissues and organs remote from the heart cannot, however, be explained solely by reduced perfusion and it is generally believed that other systemic processes (e.g. neurohumoral activation) are involved. In other words, the pathophysiology of heart failure is complex and incompletely understood and, consequently, so is the pathophysiological basis of treatment. Although it has proved difficult to agree a simple definition of heart failure, a pragmatic approach has been advocated ( Table 23.1) [1]. The terms used to describe different types of heart failure can also be confusing. Generally the term ‘heart failure’ is used to describe the symptomatic syndrome (graded according to the

New York Heart Association (NYHA) functional classification), although a patient can be rendered asymptomatic with treatment ( Table 23.2). In our view, a patient who has never exhibited the typical signs or symptoms of heart failure is better described as having asymptomatic left ventricular systolic dysfunction (or whatever the underlying cardiac abnormality is). Recently, however, the American College of Cardiology and American Heart Association have adopted a classification of heart failure that includes asymptomatic patients ( Table 23.2) [2]. Patients who have had heart failure for some time are often said to have ‘chronic heart failure’. If chronic heart failure deteriorates the patient may be described as ‘decompensated’ and this may happen suddenly, i.e. ‘acutely’, usually leading to hospital admission, an event of considerable prognostic importance [3]. New (‘de novo’) heart failure may also present acutely, for example as a consequence of acute myocardial infarction (or in a subacute or acute on chronic fashion, for example in a patient who has had asymptomatic cardiac dysfunction for an often indeterminate period) and may resolve (the patient may become 'compensated') or persist [1, 4]. ‘Congestive heart failure’ is a term still used commonly in the USA and may describe acute or chronic heart failure with evidence of congestion, i.e. sodium and water retention. Congestion, though not some symptoms of heart failure (e.g. fatigue), may resolve with diuretic treatment. Many or all of these terms may be accurately applied to the same patient at different times, depending on what stage of their illness they are in.

The epidemiology of symptomatic heart failure in developed countries is well understood, especially in Europe ( Fig. 23.1) [5–14]. Approximately 2% of the adult population has heart failure, although the syndrome mainly afflicts the elderly, affecting 6–10% of people over the age of 65 years [5–15]. In Europe and North America, the lifetime risk of developing heart failure is approximately one in five for a 40-year-old [16, 17]. The age-adjusted incidence of heart failure appears to have remained stable or even decreased over the past 20 years [18–20]. Prevalence is thought to be increasing, partly because survival is increasing [20, 21]. Approximately two in 1000 of the adult population are discharged from hospital with heart failure each year and heart failure accounts for about 5% of all medical and geriatric admissions and is the single most common cause of such admissions in those >65 years [22–31]. Age at admission (and at death) seems to be increasing, suggesting that preventive treatments, such as antihypertensives, and secondary prevention after myocardial infarction are delaying the development of heart failure [22–31]. Hospital discharges include patients developing heart failure suddenly, de novo, as a consequence of another cardiac event (usually myocardial infarction); patients presenting for the first time with decompensation of previously unrecognized cardiac dysfunction; and patients with established, chronic heart failure who have suffered worsening sufficiently severe to lead to hospital admission (though it is recognized that the ‘threshold’ for admission to hospital may vary substantially between countries). Some of these admissions are unavoidable, reflecting the progressive natural history of heart failure whereas others may be avoidable (e.g. as a result of non-adherence to treatment, failure of prompt recognition, and treatment of early decompensation) [32]. After years of steady increase, age-adjusted rates of admission for heart failure seem to have reached a plateau, or even decreased, in Europe and North America ( Fig. 23.2) though absolute numbers of admissions continue to increase and heart failure is still an enormous burden on health services and a cost to society, accounting for approximately 2% of all healthcare spending [22–31, 33]. Hospital admissions account for the main part of this expenditure, typically about 70%. Even in primary care, heart failure accounts for more consultations than angina ( Fig. 23.3), reflecting the limiting symptoms and reduction in well-being experienced by patients with heart failure [34]. Indeed, quality of life has, consistently, been shown to be reduced more by heart failure than by other chronic illnesses ( Fig. 23.4) [35].

Heart failure is deadly as well as disabling. Community-based surveys indicate that 30–40% of patients die within 1 year of diagnosis and 60–70% die within 5 years, mainly from worsening heart failure or suddenly (probably because of a ventricular arrhythmia) [15, 17, 19, 36, 37]. Thus an adult living to age 40 has a one in five risk of developing heart failure and, once apparent, a one in three chance of dying within a year of diagnosis. Mortality is even higher in patients requiring hospital admission, exceeding that of most cancers ( Fig. 23.5), though a number of recent studies indicate that prognosis may be improving ( Fig. 23.6) [23–31, 38–43].

Left ventricular function has also been measured in a number of population-based echocardiographic studies, notably in Europe, enabling estimation of the prevalence of heart failure with a low and preserved EF, as well as the prevalence of asymptomatic left ventricular systolic and diastolic dysfunction ( Fig. 23.1) [44–54]. Synthesis of these epidemiological surveys suggests that approximately half of patients with symptomatic heart failure in the community have a low EF (and half have a preserved EF) [44]. The epidemiology of symptomatic heart failure with reduced EF differs from that of heart failure with preserved EF in that patients with preserved EF are, on average, older, are more often women, have more comorbidity, and have a better age-adjusted survival ( Fig. 23.7) [31–55]. The causes of heart failure in patients with preserved EF also differ from those with a low EF [31–55].

Because about half of patients with a low left ventricular EF are asymptomatic the argument has been made for screening for symptomless cases although no consensus has been reached on this point [56, 57]. Recent studies have reported very disparate prevalence rates of diastolic dysfunction and proportions of symptomatic and asymptomatic individuals and no conclusions can yet be drawn about the epidemiology of asymptomatic diastolic dysfunction [44–52, 54, 58].

As already mentioned, any structural, mechanical, or electrical abnormality of the heart can cause it to fail ( Table 23.3). Similarly, heart failure can be caused by ischaemic, metabolic, endocrine, immune, inflammatory, infective, genetic, and neoplastic processes, by failure of the heart to develop properly, and even by pregnancy. The potential causes of heart failure are, therefore, legion, vary geographically, and have changed over time. Rheumatic valvular disease remains a common cause in many developing countries whereas this diagnosis is now uncommon in developed countries; in the latter, degenerative valvular disease in the elderly is now more common [42, 59–63] (see Chapter 17). Valve disease may lead to volume and pressure overload of the heart, as described in more detail below. Endocardial disease is very rare in Europe but much less so in parts of Africa, where it can cause what is referred to as a restrictive cardiomyopathy, as described further in the rest of this section [59–63].

In the developed world, ventricular dysfunction is the commonest underlying problem and is caused, mainly, by myocardial infarction (leading to systolic ventricular dysfunction, i.e. failure of normal contraction and emptying of the heart), hypertension (causing systolic dysfunction, diastolic dysfunction, i.e. failure of normal relaxation and filling of the heart or both) or, often, both infarction and hypertension (see Chapter 17). These causes are becoming more important in some parts of the developing world [63]. Whether persisting systolic dysfunction is caused by coronary artery disease in the absence of infarction is uncertain. The converse, i.e. whether treatment of ischaemia and a state known as ‘hibernation’ (where chronic poor perfusion results in non-contracting but viable myocardium) improves systolic function, is also a question of great current interest, addressed by ongoing studies of coronary ‘revascularization’ [64].

In Europe, North America, and Australasia, hypertension was once the principal cause of heart failure whereas now that position is filled by coronary heart disease (or more exactly, myocardial infarction); this is also increasingly the case in many developing countries [63].

While myocardial infarction is a much more important individual risk factor than hypertension, the population-attributable risk due to hypertension is probably still more important [65, 66]. Both causal factors also interact to augment the risk of heart failure, with concomitant hypertension greatly increasing the risk of failure after myocardial infarction [65, 66]. By the time heart failure presents, prior hypertension may no longer be present. Both of these factors probably result in underestimation of the role of hypertension in causing heart failure. Hypertension is a more common aetiology in women than men.

‘Idiopathic’ dilated cardiomyopathy remains the only other cause of systolic dysfunction commonly encountered, perhaps accounting for 15–20% of cases of heart failure with reduced systolic function. These cases probably have multiple causes and an increasing number of genetic causes are being identified [67]. Prior viral infection is a recognized cause and recent research studies suggest that persisting virus infection may be identified (by endomyocardial biopsy) in a high proportion of patients with dilated cardiomyopathy. Whether this is of prognostic significance or requires therapeutic intervention is currently uncertain [68]. Current or previous excessive alcohol consumption may also cause a dilated cardiomyopathy, as can exposure to other toxins, including chemotherapeutic agents used in cancer treatment such as anthracyclines and trastuzumab [69]. These must always be considered when a patient presents with an unexplained dilated cardiomyopathy. If angiography is not carried out to exclude coronary disease, it may also be wrongly concluded that the patient has an ‘idiopathic’ dilated cardiomyopathy. Chagas disease, caused by the protozoan parasite Trypanosoma cruzi, can also cause systolic dysfunction. Though rarely encountered in Europe it is a relatively common cause of heart failure in South America and is now recognized in Central and North America [59, 61].

It is not clear whether diabetes mellitus should be considered an aetiological factor or a comorbidity in heart failure. Its potential role in causation of systolic and diastolic dysfunction is uncertain [70]. Diabetics have a higher prevalence of heart failure. Diabetes accelerates the development of coronary atherosclerosis and is often associated with hypertension. Whether it directly causes a specific cardiomyopathy is, however, uncertain. Diabetes is also associated with a higher risk of developing heart failure in patients with other causes, e.g. acute myocardial infarction. It is believed that diabetes promotes the development of myocardial fibrosis and diastolic dysfunction. Diabetes is also associated with more autonomic dysfunction and worse renal, pulmonary, and endothelial function, as well as worse functional status and a worse prognosis. Conversely, heart failure increases the risk of developing diabetes [70].

Although it is associated with diabetes, hypertension, and coronary artery disease, obesity also seems to be an independent risk factor for developing heart failure [71].

Atrial fibrillation is both an aetiological factor and comorbidity. It can cause heart failure directly as a consequence of the loss of the atrial contribution to cardiac output and reduced diastolic filling as a result of tachycardia [72, 73]. Patients with underlying structural or functional cardiac disease are more likely to develop failure as a consequence of these effects with the onset of atrial fibrillation. There is, however, a growing belief that atrial fibrillation can cause a dilated cardiomyopathy the exact mechanism of which is uncertain, though persistent tachycardia may play a role (i.e. atrial fibrillation may cause a ‘rate-related cardiomyopathy’) [74]. Heart failure also increases the risk of developing atrial fibrillation and this risk increases with the severity of heart failure. Consequently, when a patient presents with left ventricular dilatation, systolic dysfunction, and atrial fibrillation it can be difficult to determine which came first.

It is important to appreciate that heart failure does not occur in isolation. It is caused by an underlying cardiac defect in, usually, elderly individuals frequently treated for other medical problems with multiple medications. Consequently, the patient with heart failure often has comorbidity related to the underlying cardiac problem or its cause (e.g. angina, hypertension, diabetes, smoking-related lung disease) and age (e.g. osteoarthritis), as well as a consequence of heart failure (e.g. arrhythmias) and its treatment (e.g. gout from diuretics) [75]. Some common comorbidities have multiple causes (e.g. renal dysfunction, see Cardiorenal syndrome, p.842), whereas others are not fully explained (e.g. anaemia, depression, disorders of breathing, and cachexia) [75–80]. The existence of multiple comorbidities creates the potential for drug intolerance (e.g. angiotensin-converting enzyme (ACE) inhibitor and renal dysfunction), drug interactions (e.g. non-steroidal anti-inflammatory drugs (NSAIDs) and ACE inhibitors), worsening of heart failure as a specific adverse effect (e.g. thiazolidinediones), and makes the management of heart failure very complex [1, 70, 81–83]. This is especially true of renal dysfunction, the importance of which is increasingly recognized by a growing use of the term ‘cardiorenal syndrome’ [84, 85] to describe concurrent heart and renal failure.

This syndrome arises from multiple interactions between the age-related decline in glomerular filtration, the effects of treatment for heart failure (diuretics, ACE inhibitors, angiotensin receptor blockers (ARBs), and aldosterone antagonists) and other conditions (e.g. NSAIDs for arthritis), comorbidity (e.g. hypertension, diabetes, atherosclerosis), reduced renal blood flow and the actions on the kidneys of the array of neurohumoral pathways activated in heart failure [84, 85]. The prevalence of severe renal dysfunction in heart failure is often underestimated because serum creatinine concentration may not be greatly elevated because of the reduction in skeletal muscle mass in advanced heart failure. Renal dysfunction may contribute to the high prevalence of anaemia in patients with heart failure. The blood concentrations of creatinine, urea (blood urea nitrogen), and estimated glomerular filtration rate are powerful independent predictors of prognosis and deterioration in renal function during an episode of worsening heart failure (e.g. an increase in creatinine concentration of ≥0.3mg/dL or 27µmol/L) is associated with higher morbidity and mortality [86].

Anaemia is another important comorbidity and can be both the cause and, it seems, consequence of heart failure [76, 87–89]. Anaemia is common (especially in more severe heart failure) and is associated with worse symptoms, increased risk of hospital admission, and reduced survival. The causes are unknown but may include haemodilution, renal dysfunction, poor nutrition, inflammation, blood loss related to medication, and reduced production of (or response to) erythropoietin.

Patients with heart failure often exhibit some degree of muscle wasting which is restricted to the lower limbs (disuse atrophy) [77, 90, 91]. This loss of tissue may become more extensive in some patients, usually when their heart failure is more advanced, and may affect all body compartments (muscle, fat, and bone tissue). This general wasting is referred to as cardiac cachexia. The underlying metabolic causes are complex and differ from patient to patient. Three important contributors are, probably, dietary deficiency (exacerbated by anorexia), loss of nutrients through the urinary or digestive tracts (malabsorption), and metabolic dysfunction (including an imbalance of anabolic and catabolic factors and inflammation). The development of cachexia is an ominous sign [77, 90].

It is important to appreciate that it is comorbidity, along with the key pathophysiological processes in heart failure, i.e. left ventricular remodelling and activation of systemic pathways (and age), that are the principal determinants of prognosis.

We have only limited knowledge of the pathophysiology of heart failure, and that of left ventricular systolic dysfunction is best understood. Much of our understanding comes from studies of myocardial infarction ( Fig. 23.8) [92, 93]. Following an initial injury to the myocytes and cytoskeleton, heart failure may develop immediately, in the short term (over days or weeks), over a longer time period (months to years), or not at all. The factors leading to the development of heart failure acutely after myocardial infarction (e.g. size of infarction, concomitant hypertension [66], etc.), the important pathophysiological mechanisms operating (e.g. cardiac remodelling), and the time-course of this complication of infarction are fairly well established. On the other hand, the natural history of asymptomatic left ventricular systolic dysfunction is less well understood, as are the pathophysiological mechanisms causing progression from the asymptomatic to the symptomatic state. Once symptomatic heart failure has developed we believe that the pathophysiology is again better understood, at least in patients with systolic dysfunction. One thing is certain: the syndrome is characterized by progressive worsening of the patient’s symptoms, of cardiac function, and of the function of other tissues (e.g. skeletal muscle, bone marrow) and organs (e.g. the kidneys).

The progression of left ventricular systolic dysfunction (and the heart failure syndrome), because of ‘remodelling’ of the left (and right) ventricle (as a result of the loss of myocytes and maladaptive changes in the surviving myocytes and extracellular matrix), probably occurs in two main ways [92–97]. One is because of intercurrent cardiac events (e.g. myocardial infarction [98]) and the other is as a consequence of the local processes (e.g. the autocrine pathway and molecular adaptations, including, perhaps, apoptosis) and systemic processes (e.g. neurohumoral pathways) that are activated as a result of reduced systolic function ( Fig. 23.9) [92–97]. These systemic processes, which are discussed in detail in later sections, also have detrimental effects on the functioning of the lungs, blood vessels, kidneys, bone marrow, muscles, and probably other organs (e.g. the liver), and contribute to a pathophysiological vicious cycle ( Fig. 23.10). The molecular, structural, and functional changes in the heart and these systemic processes, coupled with electrolyte imbalances, result in electrical as well as mechanical dysfunction of the heart. In addition, cardiac metabolism is altered in the failing heart and some believe that a state of relative energy starvation exists [93, 99].

It is important to remember that atrial function, synchronized contraction of the left ventricle, and normal interaction between the right and left ventricles are also important in preserving stroke volume [93, 100, 101]. Loss of these key mechanical interactions is often secondary to disturbances of conduction arising as a consequence of cardiac fibrosis.

The term ‘cardiomyopathy’ is used to define a myocardial disorder in which the heart muscle is structurally and functionally abnormal, in the absence of coronary artery disease (hence the term ‘ischaemic cardiomyopathy’ should be avoided), hypertension, valvular disease, and congenital heart disease sufficient to cause the observed myocardial abnormality [67, 102]. Myocyte necrosis whether caused by infarction or by other injury has a common consequence. Whether diffuse or focal, myocyte loss leads to replacement fibrosis, hypertrophy of the remaining myocytes and dilatation of the affected cardiac chamber [67, 93, 102–104]. How these molecular and cellular changes affect the left ventricle macroscopically (by causing remodelling) is discussed in more detail in later sections. The resulting anatomical and pathophysiological picture is, however, identified by clinicians as a dilated cardiomyopathy. Certain inherited cardiomyopathies result in a similar phenotype. Poor contraction and emptying of the ventricle are usually referred to as systolic dysfunction. Other cardiomyopathic phenotypes occur ( Fig. 23.11).

Systolic and diastolic dysfunction are terms used to describe whether the principal abnormality of the myocardium is an inability of the ventricle to contract and expel blood or to relax and fill normally, respectively (though in reality these two abnormalities frequently coexist). Systolic dysfunction

is the result of reduced shortening of sarcomeres, which is a consequence of a global or regional reduction of contractility or greatly increased impedance to left ventricular ejection. An increase in preload can provide short-term compensation (via the Frank–Starling mechanism, see p.847) for a reduction in contractility or increases in impedance. However, long-term compensation usually involves myocardial hypertrophy, which is the result of laying down new sarcomeres that increase the width (concentric) or the length (eccentric) of myocytes [93, 105–110]. Remodelling also contributes to reduced sarcomere shortening. All these factors causing reduced fibre shortening and eventually lead to a decrease in the left ventricular ejection fraction (LVEF). Hence, end-systolic volume increases.

Rapid filling during systole is assisted by active, energy-dependent, relaxation of the ventricle [93, 105–110]. Primary myocardial diseases may affect this process. Ventricular relaxation also depends on myocardial mass, collagen content, and extrinsic forces (e.g. the pericardium) [93, 105–110]. The hallmark of diastolic dysfunction is elevation in left ventricular end-diastolic pressure or left arterial pressure in the absence of systolic dysfunction [93, 105, 110].

Infiltrative processes in the myocardium may cause a restrictive cardiomyopathy ( Fig. 23.12) [111, 112] (see Chapter 17). The principal consequence is impaired ventricular filling with normal or decreased diastolic volume of either or both ventricles, akin to the diastolic dysfunction described earlier. Systolic function is typically maintained during the early stages of the disease and wall thickness is normal or increased. The main problem is that increased stiffness of the myocardium causes pressure within the ventricle to rise precipitously with only small increases in volume. Restrictive cardiomyopathy may affect either or both ventricles. Elevated jugular venous pressure, peripheral oedema, and ascites are often prominent features. Pericardial constriction has similar pathophysiological effects and causes a similar clinical picture but can be hard to diagnose. Clinical suspicion may be raised by a history of prior tuberculosis, radiotherapy, cardiothoracic surgery, etc. Special investigations, including cardiac computed tomography (CT) scanning and simultaneous right and left heart pressure measurements at catheterization are often necessary [113, 114].

Arterial hypertension and aortic stenosis cause a sustained increase in systolic wall stress during left ventricular ejection leading to concentric hypertrophy of the left ventricle because of myocyte hypertrophy and extracellular matrix overgrowth [115].

Conversely, mitral and aortic regurgitation result in an increased volume load on the ventricle. The resultant ventricular remodelling is characterized by dilatation, representing, at least in part, lengthening of the cardiac myocytes [116, 117].

These terms are not useful and are reminiscent of the now discarded terminology of ‘backwards’ and ‘forwards’ heart failure. The term right heart failure is often used to describe patients in whom there are prominent signs of peripheral ‘congestion’, e.g. a raised jugular venous pressure, hepatomegaly, and peripheral oedema ( Fig. 23.13), on the basis that these findings reflect right ventricular failure; in fact all of these signs are also found in patients with predominantly left ventricular involvement. The description pulmonary heart disease is used to depict patients who do have isolated right heart failure as a result of primary lung disease and has generally replaced the term ‘cor pulmonale’.

A more useful pathophysiological classification is to distinguish between high- and low-output heart failure, although the former is uncommonly encountered in Western clinical practice. Cardiac index is normally 2.2–3.5L/min/m². Low-output cardiac failure implies that cardiac output fails to rise adequately during exercise or that it is inadequate even at rest. This prototypical form of heart failure is seen in cases of heart failure as a result of left ventricular systolic dysfunction. High-output cardiac failure, on the other hand, implies that although the pumping action of the heart is intact other factors make it difficult for the heart to deliver oxygen commensurate with the needs of the metabolizing tissues, either because of increased tissue demand (e.g. as a result of anaemia, hyperthyroidism, or pregnancy) or reduced oxygen-carrying content of the blood (e.g. anaemia). This type of heart failure can also be caused by arteriovenous shunting.

The Frank–Starling law describes an intrinsic mechanism that helps maintain stroke volume when the heart is acutely injured and may also play a compensatory role in chronic heart failure, though this is less certain [118]. Together with neurohumoral activation (an extrinsic mechanism), the Frank–Starling law is an adaptive phenomenon that comes into play within minutes of cardiac injury. The resultant acute fall in the volume of blood ejected by the ventricle (the stroke volume) leads to a rise in left ventricular end-diastolic volume (and pressure). Through the Frank–Starling mechanism this rise in preload increases the force of contraction, thereby helping restore stroke volume. The mechanism whereby increased stretch of the myocyte causes increased force of contraction is also referred to as the law of heterometric autoregulation. In the chronic setting, sodium retention, water retention, and venoconstriction may represent continuing attempts by the body to use the Frank–Starling mechanism by increasing left ventricular filling pressure and preload.

These adaptations may, however, lead to abnormally high pulmonary capillary and artery pressures, probably contributing to the shortness of breath experienced by patients with heart failure. Furthermore, arterial constriction and stiffening (as a result of sodium and water retention in the vascular wall) increase afterload and will eventually cause the injured left ventricle to fail further because it is especially sensitive to increases in afterload (law of homeometric regulation).

The heart attempts to compensate for increased preload (e.g. because of increased extracellular fluid volume and venous return) and afterload (e.g. because of systemic arterial constriction) in several ways. One is the development of ventricular hypertrophy in an attempt to maintain systolic wall stress within normal limits [92–97, 119, 120]. Pressure overload tends to lead to concentric hypertrophy, whereas volume overload tends to lead to ventricular dilatation [92–97, 119, 120]. Both entities are distinct at the molecular level. Pressure overload is associated with parallel replication of myofibrils and thickening of individual myocytes. Volume overload, on the other hand, leads to replication of sarcomeres in series and elongation of myocytes. The two types of haemodynamic overload are presumed to activate distinct signalling pathways.

Compensated remodelling results in relatively little change in ventricular dimensions, shape, function, and wall thickness. However, these compensatory adaptations only seem to be capable of maintaining pump function over a limited period of time and a ventricle subjected to an elevated load for a prolonged period will ultimately fail. Ventricular dilatation may lead to stretching of the mitral valve ring and cause valvular incompetence ( Fig. 23.14). This may further increase the load on the failing ventricle; this is an example of another ‘vicious cycle’ that develops and which may drive progression of heart failure.

An initial stress-induced increase in sarcomere length yields an optimal overlap between myofilaments [92–97, 119–121]. Severe haemodynamic overload eventually yields depression of myocardial contractility. In patients with mild disease, this depression is manifested by reduced velocity of shortening of the myocardium or by a reduction in the rate of force development during isometric contraction. More severe stages are accompanied by a decline in isometric force development and shortening as well. EF and cardiac output during exercise decline [92-97, 119–121].

Our understanding of the molecular mechanisms behind these changes is still limited and can only be touched upon briefly ( Fig. 23.9). They comprise myocyte loss by necrosis and apoptosis, alterations in excitation–contraction coupling, and alterations in composition of the extracellular matrix [119–120]. Myocyte loss as a result of necrosis is a well-understood process which is localized after myocardial infarction but more diffuse in patients with dilated cardiomyopathy or myocarditis. Apoptosis, or programmed cell death, on the other hand, results from the induction of a genetic programme that leads to degradation of nuclear DNA ( Fig. 23.9) [92–97, 119–122]. Several recent reports have described apoptotic cells in the failing myocardium. It may be relevant that several substances, such as angiotensin II, reactive oxygen species, nitric oxide (NO), and pro-inflammatory cytokines may induce apoptosis experimentally in cardiac myocytes. However, the precise frequency of occurrence and role of apoptosis in the failing myocardium remains unclear [122]. Changes in the extracellular matrix are usually manifested by an increase in collagen content though both degradation and synthesis (and the activity of the enzymes controlling these processes) may be increased [93, 95]. While this change in collagen content may contribute to impaired systolic contraction it may be even more important in reducing ventricular compliance and impaired ventricular filling.

The human body responds to the haemodynamic changes in patients with heart failure in a highly complex way. Many neurohumoral systems appear to be involved to varying degrees and at different stages. It has been suggested that these are initially activated in a manner appropriate to haemorrhage or some other crisis threatening vital organ perfusion [93, 123, 124]. Their sustained activation in heart failure, however, is not only inappropriate but probably detrimental ( Fig. 23.10). Moreover, at some stage during the progression of heart failure, haemodynamic abnormalities may cease to be the main trigger of neurohumoral activation. Instead, a number of other self-sustaining pathophysiological vicious cycles may develop.

The predominant effects of the neurohumoral systems activated in heart failure are to cause vasoconstriction, sodium and water retention, and abnormal cell growth. Two exceptions are the natriuretic peptides [93, 125] (secreted mainly by the failing heart, Fig. 23.15) and adrenomedullin (secreted mainly from blood vessels) [126].

The neurohumoral pathways activated in heart failure are thought to be particularly important because they appear to explain how many of the successful pharmacological treatments for heart failure work (and offer the potential for more such therapeutic interventions). Neurohumoral activation seems to be much less marked in patients with heart failure and preserved EF, compared to those with a low EF [127].

It has long been recognized that an increased activity of the sympathetic nervous system and parasympathetic withdrawal are prototypical characteristics of heart failure [93, 128–130]. Elevated levels of plasma norepinephrine (noradrenaline) are a common finding in patients with heart failure. Increased sympathetic nerve traffic and enhanced spill-over of norepinephrine from the synaptic cleft account for this, and are particularly pronounced in the heart, kidney, and skeletal muscle [93, 128–131]. Raised plasma norepinephrine concentrations predict higher mortality rates [130, 131]. Heart failure is also characterized by reductions in myocardial norepinephrine stores and in myocardial beta-receptor density. These again reflect generalized adrenergic activation [93, 128–131].

It is believed that enhanced sympathetic activity initially increases myocardial contractility and heart rate (leading to an increase in cardiac output). Sympathetic activation also promotes renin release, sodium retention, and vasoconstriction thereby increasing preload and activating the Frank–Starling mechanism (see Frank–Starling mechanism, p.847). These responses are capable of maintaining ventricular performance and cardiac output for a limited period of time. In part this may be because the increase in afterload caused by arterial constriction (to which the failing ventricle is particularly sensitive) leads to a further fall in stroke volume. Sympathetic overdrive probably alters myocardial metabolism and catecholamines may also even be directly toxic to cardiomyocytes [93, 130]. Excessive adrenergic activity (and reduced vagal activity) also increases the electrical instability of the heart. As well as having complex and changing effects on myocardial contractility and structure, activation of the sympathetic nervous system leads to redistribution of regional blood flow and even to changes in the structure of the vasculature [129].

The precise cause of sympathetic activation in heart failure is unknown. Reduced stimulation of stretch-activated baroreceptors in the carotid arteries and the aorta from decreased arterial pressure and stroke volume may contribute (in an analogous way to haemorrhage). Another factor suggested is structural and functional abnormalities of afferent receptors. Other neurohumoral systems may also activate the sympathetic nervous system and augment its actions, i.e. many neurohumoral systems act in concert, synergistically reinforcing each other.

Increased activity of the renin–angiotensin–aldosterone system (RAAS) also produces deleterious effects on the cardiovascular system (and other organs and tissues) and contributes to the poor prognosis in heart failure [93, 123, 124, 132–134]. The plasma components of this system are usually increased in patients with heart failure and a low serum sodium concentration is a marker for particularly excessive activation of the RAAS [134]. The increase in renin release is mediated by decreased stretch of the glomerular afferent arteriole and reduced delivery of chloride to the macula densa. Although the sympathetic nervous system also stimulates renin release, the two systems are independently regulated. One puzzle about heart failure is why the sodium and volume overload that characterizes the syndrome does not suppress renin release, as would normally occur.

The increased secretion of renin leads to an augmented production of angiotensin II, which, it is believed, has mostly deleterious effects in heart failure (although its efferent glomerular arteriolar action may help maintain glomerular filtration). Angiotensin II not only induces vasoconstriction, but also salt and water retention directly and via aldosterone. Moreover, it mediates myocardial cell hypertrophy and fibrosis and these effects may contribute to the maladaptive remodelling and progressive loss of myocardial function in heart failure. Angiotensin II may also cause activation of the sympathetic nervous system, prothrombotic actions, and augment the release of arginine vasopressin.

Plasma aldosterone levels are also increased in heart failure, and release of this hormone is influenced by angiotensin and other stimuli such as potassium and corticotropin. Aldosterone is an independent and harmful component of the RAAS [135]. It causes sodium and water retention and, importantly, potassium wastage. Potassium loss, along with autonomic dysfunction and myocardial fibrosis (both of which are also thought to be caused by aldosterone), may increase the risk of ventricular arrhythmias. Aldosterone may also contribute to vascular fibrosis in heart failure.

There is recent interest in the possibility that renin itself, acting through a specific renin/pro-renin receptor, may have detrimental effects in heart failure [136].

Vasopressin (also known as antidiuretic hormone) is a neurohypophysial peptide involved in the regulation of free water reabsorption, body fluid osmolality, blood volume, blood pressure, cell contraction, cell proliferation, and adrenocorticotropin secretion [137, 138]. Vasopressin binds to three different specific G protein-coupled receptors. These are currently classified as V1 (or V1A)-vascular, V2-renal, and V3 (or V1B)-pituitary subtypes. All subtypes have distinct pharmacological profiles and intracellular second messengers. As well as reducing renal water excretion, vasopressin is one of the most powerful vasoconstricting substances known. It also stimulates blood platelet aggregation, coagulation factor release, and cellular proliferation. This profile of action is clearly unattractive in heart failure.

Elevated circulating levels of vasopressin are often, but not invariably, found in patients with chronic heart failure. It appears that vasopressin release from the posterior pituitary in heart failure is largely non-osmotic, although, normally, increased serum osmolality is the major physiological stimulus for its secretion.

A-type (atrial) natriuretic peptide (ANP) and B-type (brain) natriuretic peptide (BNP) are released in response to atrial and ventricular wall stretch and, as their names suggest, serve to maintain sodium homeostasis by enhancing renal sodium and water excretion ( Fig. 23.15) [93, 125, 139–142]. These peptides also have haemodynamic effects, dilating arteries and, especially, veins. They also suppress the RAAS and, possibly, the sympathetic nervous system. There is some evidence that natriuretic peptides inhibit arginine vasopressin and endothelin-1 release and the biological actions of these peptides. Consequently, the natriuretic peptides are thought to play an important protective role in heart failure, countering the actions of the other vasoconstricting and anti-natriuretic neurohumoral systems which are activated in heart failure.

Circulating levels of both ANP and BNP are greatly increased in heart failure. This is the consequence of an increased synthesis and release of these hormones. In humans, BNP is mostly secreted from the ventricles in both healthy individuals and patients with heart failure. ANP secretion, on the other hand, is mainly from the atria in healthy individuals but from both the atria and the ventricles in patients with heart failure. Therefore, it appears that BNP is the only natriuretic peptide that is specific to the ventricles. Pro-BNP, the precursor of BNP, is stored in granules in myocytes. Pro-BNP is activated by a protease to form its biologically active form, BNP, and N-terminal (NT)-proBNP.

BNP levels vary according to sex and age in healthy subjects [141]. Female patients display higher plasma concentrations than male patients with heart failure. Advancing age and declining renal function are associated with increases in BNP levels.

C-type natriuretic peptide (CNP), which was originally believed to be of endothelial origin, may also be produced by the failing heart. A D-type (Dendroaspis) natriuretic peptide and a renal specific peptide (urodilatin) has also been described recently, though the origin and actions of each in humans are not yet well defined [142].

As well as having important physiological effects, the various A-type and B-type natriuretic peptides and related fragments can be used to aid the diagnosis of heart failure and to provide prognostic information.

Relaxin is a pregnancy-related hormone with vasodilator activity which has been shown to improve haemodynamic indices in heart failure [145, 146].

These three more recently described neurohumoral factors have vasodilator and inotropic properties, as well as other actions [126, 143, 144]. Apelin is unusual in that blood and tissue concentrations are reduced in heart failure [143]. The pathophysiological role, if any, that these factors play in heart failure is presently uncertain as is any therapeutic potential that might exist. [126, 143, 144].

The principal product of the endothelium, NO, plays a central role in vascular homeostasis. Endothelial dysfunction, which is characterized by reduced production and action of NO, occurs as a result of ageing, as well as in a number of chronic conditions related to heart failure, such as hypercholesterolaemia, atherosclerosis, as well as in heart failure itself [147]. Endothelium-dependent dilatation of coronary and peripheral resistance vessels is blunted in patients with heart failure. This probably contributes to the impaired reactive hyperaemia in various vascular beds, an impairment in tissue perfusion and, perhaps, reduced muscular function [148]. There has even been speculation that NO might be a key regulator of lung function and that exercise-induced dyspnoea may be related to impaired pulmonary vasodilatation resulting from reduced NO production compared to healthy subjects. Endothelial dysfunction in heart failure may be partly the result of increased oxidative stress.

Although lack of NO is associated with the development of endothelial dysfunction, its overproduction by the inducible isoform of NO synthase (iNOS) may also be detrimental [148]. Increased iNOS activity is thought to lead to increased free radical formation and to depression of myocardial activity [93]. Increased iNOS expression may result from the actions of inflammatory cytokines.

The endothelins are another important product of the endothelium and endothelin-1 is one of the most powerful vasoconstrictor peptides known [149]. It is also a mitogen and generally shares the potentially detrimental properties of angiotensin II and vasopressin in heart failure. Plasma concentrations of endothelin-1 are increased in heart failure, probably because of increased secretion, both by blood vessels and the failing myocardium. The importance of this is, however, uncertain because specific antagonists have not improved outcome in heart failure.

Oxygen free radicals have a number of potentially detrimental actions in heart failure [93, 150, 151]. They inactivate NO, depress myocardial contractility, and may induce apoptosis [93, 152]. One source of the superoxide anion radical is NADPH oxidase which is activated by angiotensin II and aldosterone [93, 153]. Xanthine oxidase may be another source of increased free oxygen radical load in heart failure and is normally involved in the last step of purine breakdown which yields uric acid. Hyperuricaemia is a consistent finding in patients with heart failure, may reflect impairment of oxidative metabolism, and is a predictor of worse outcome [152]. Oxidative stress may be part of the generalized inflammatory state that characterizes at least some patients with heart failure. The importance of oxidative stress is, however, uncertain because specific antagonists (oxypurinol and vitamin E) did not improve pathophysiological measures or clinical outcomes in heart failure [155, 156].

Inflammation may also be a factor contributing to the progression of heart failure although its role has not been ‘confirmed’ in the same way as neurohumoral activation, i.e. by the demonstration of improved outcomes with blocking agents [93, 157]. Several cytokines have been studied in detail. Different cell types secrete cytokines for the purpose of altering either their own function (autocrine) or that of adjacent cells (paracrine). Some cytokines also act as circulating hormones (i.e. have an endocrine action). Tumour necrosis factor-α (TNF-α), interleukin-1, and interleukin-6 are thought to be the most important pro-inflammatory cytokines that may be implicated in heart failure progression. The cause of cytokine activation in heart failure remains unclear. Pro-inflammatory cytokines may be secreted by mononuclear cells, hypoxic peripheral tissue, or even by the myocardium itself. Catecholamines may augment myocardial cytokine production, one of several possible links between neurohumoral activity and inflammation. It has also been hypothesized that increased bowel wall oedema may lead to translocation of bacterial endotoxin or lipopolysaccharide from the gut which may cause pro-inflammatory cytokine production from blood monocytes and possibly other tissues [157, 158].

Plasma TNF-α and TNF receptor (TNFR) concentrations are increased in some patients with heart failure, especially in those whose disease is severe, and they are independent predictors of poor prognosis. Although TNF-α has several potentially untoward effects that could contribute to the progression of heart failure, studies of TNF antagonists to date have not shown benefit in this syndrome [159, 160].

It is symptoms and signs that usually alert the patient (and physician) to the presence of a cardiac disorder. However, neither the symptoms nor signs commonly recognized as suggesting the presence of heart failure are specific for this syndrome. Therefore confirmation of heart failure requires objective tests to confirm that the patient’s symptoms and the physical findings are the result of abnormal cardiac function and not of another cause (see Simple investigations, p.854) [1, 2].

Fatigue is a key symptom reported by patients with heart failure. Its origins are not clearly understood but probably include low cardiac output and skeletal muscle abnormalities (see Pathophysiology, p.843). Fatigue is very non-specific and is found in the population at large, as well as in many non-cardiovascular disorders.

Dyspnoea or breathlessness is another cardinal symptom of heart failure. Dyspnoea is usually first manifested on exertion and the level of exertion which causes breathlessness is useful in gauging heart failure severity and monitoring the patient’s progress. Though it is more specific than fatigue, dyspnoea is still caused by many other disorders such as pulmonary disease, obesity, and anaemia, which are common in the elderly population and may coexist with heart failure. Even ageing itself is associated with dyspnoea on exercise. The origin of dyspnoea in heart failure is also probably multifactorial. It may be related to elevated pulmonary pressures, abnormalities in pulmonary compliance, respiratory dysfunction, accentuated respiratory drive, increased airway resistance, abnormal chemical and mechanical muscle reflexes [161–166], and even low haemoglobin. It is notable that there is a poor correlation between dyspnoea and left ventricular function at rest [161–166].

Orthopnoea is defined as dyspnoea which occurs in the recumbent position and is usually relieved by sitting upright or by the addition of pillows. In extreme cases, the patient is unable to lie down and may spend the night in the sitting position. Orthopnoea results from the return of venous blood which has pooled in the lower extremities while the patient is ambulatory. The failing heart may be unable to cope with return of this blood from the legs on adoption of the recumbent position and pulmonary oedema may occur.

Paroxysmal nocturnal dyspnoea is characterized by acute episodes of suffocation usually occurring while recumbent at night. It has the same causes as orthopnoea. Paroxysmal nocturnal dyspnoea may manifest as cough or wheezing, possibly because increased pressure in the bronchial arteries (and resultant increase in their diameter), along with interstitial pulmonary oedema, leads to increased airways resistance. Sometimes these patients are described as having ‘cardiac asthma’, which must be differentiated from primary asthma and pulmonary causes of wheezing.

Both orthopnoea and paroxysmal nocturnal dyspnoea are relatively specific for heart failure but are usually only encountered in untreated or advanced heart failure and are uncommon in most patients with mild to moderate heart failure taking diuretics [161–166]. Treated patients developing either symptom should be advised to report this to their physician/nurse as soon as possible. Paroxysmal nocturnal dyspnoea requires urgent treatment.

Cerebral symptoms such as confusion, disorientation, sleep or mood disturbances may be observed in advanced heart failure, particularly in the presence of hyponatraemia. These symptoms can be the first manifestation of heart failure in elderly patients. Sometimes sleep disturbances can be associated with ventilatory abnormalities (including central and obstructive sleep apnoea and Cheyne–Stokes respiration) which occur in advanced heart failure and may be reported by the patient’s spouse or partner [167].

Nausea and abdominal discomfort may occur when there is marked congestion of the liver and gastrointestinal tract (although the former may also be caused by digitalis glycosides). Congestion of the liver and stretching of its capsule may cause pain in the right upper quadrant of the abdomen.

Oliguria is usually present in advanced heart failure as the result of reduced renal perfusion and avid sodium and water retention.

This has been alluded to earlier ( Table 23.2). We prefer to use the NYHA functional classification [168]. Although it is subjective and has a large interobserver variation, it is used worldwide and, most importantly, has been employed as an entry criterion in almost all important clinical trials in heart failure.

At the individual patient level it is useful to measure functional limitation (and monitor progress) by way of the distance the patient can walk on the level, the number of steps that can be climbed, and ordinary activities that can (or cannot) be carried out, e.g. washing, bed-making, vacuum-cleaning, sweeping, shopping, etc. Because there is a poor correlation between symptoms and cardiac dysfunction, it must be emphasized that mild symptoms do not imply minor cardiac dysfunction or a good prognosis. Patients with very different EFs can experience quite similar degrees of functional limitation and treatment with a diuretic may lead to a marked improvement in symptoms without having any effect on cardiac function [169].

The Killip classification may be used to assess the severity of heart failure in the acute context, e.g. in myocardial infarction [170].

‘Quality of life’ assessments have been used to assess the impact of heart failure on patient well-being in a more complete way than just measuring specific symptoms or functional limitations. Various dimensions of quality of life including those reflecting physical, social, sexual, and professional activities can be measured, along with indices of mood, emotions, and mental health. Various questionnaires or visual scales have been proposed but none has been universally accepted. One of the most widely used is the Minnesota Living with Heart Failure questionnaire which includes a list of 21 questions, each answered on a scale of 0 to –5 [171]. More recently, the Kansas City Cardiomyopathy Questionnaire has been used [172]. These measures are more often used in clinical trials than in clinical practice.

Clinical examination, including observation of the patient and palpation and auscultation of the heart, is essential in the assessment of an individual with suspected heart failure. Percussion of the heart is seldom performed although it can provide an accurate assessment of cardiac size; percussion of the lung fields is valuable [165, 166, 173].

Patients with advanced heart failure are sometimes severely dyspnoeic even when speaking and have peripheral oedema, cachexia, or cyanosis. Conversely, the general appearance of a patient presenting with mild to moderate heart failure is often normal.

Systolic blood pressure is usually reduced in heart failure because of left ventricular systolic dysfunction, especially if severe or treated. Sometimes blood pressure may be elevated, especially if hypertension is the cause of heart failure and particularly if systolic function is preserved. Blood pressure can be markedly increased during an episode of acute pulmonary oedema. It is important to distinguish between low blood pressure (hypotension) which may be unimportant per se and hypoperfusion of the vital organs, i.e. where there are symptoms such as dizziness or confusion, renal dysfunction, or myocardial ischaemia, which is always important and requires attention.

Sinus tachycardia is a non-specific sign which is caused by increased sympathetic activity and can be absent in the presence of conduction disturbances (or if the patient is taking beta-adrenergic blocker therapy). Some patients may also have tachycardia because of atrial fibrillation (or another supraventricular arrhythmia) or, rarely, ventricular tachycardia.

Peripheral vasoconstriction with coldness, cyanosis, and pallor of extremities is also caused by increased sympathetic activity.

Peripheral oedema is a key manifestation of heart failure but is non-specific and usually absent in patients already treated with diuretics ( Fig. 23.13). It is related to extracellular volume expansion, is accompanied (and even preceded) by weight gain, and is progressive. It is usually bilateral and symmetrical, painless, pitting, and occurs first in the lower extremities in ambulatory patients, namely the feet and the ankles. In bedridden patients, oedema may instead be found over the sacrum and scrotum. Even oedema to mid-calf may reflect an increase of ≥2L in extracellular fluid volume. Long-standing leg oedema may be associated with indurated and pigmented skin. If untreated, oedema may become generalized (anasarca), with the development of hepatic congestion, ascites, and hydrothorax (pleural effusions). At this stage there is usually clear jugular venous distension ( Fig. 23.13—see Cardiac signs, p.853). Generalized oedema is often accompanied by resistance to oral diuretic treatment. Patients should be warned to be observant for progressive increases in weight, accompanied by ankle swelling and, especially, increasing dyspnoea. Daily weight monitoring is important to identify sodium and water retention episodes and initiate early therapy. A prompt (and often temporary) increase in diuretic therapy may resolve worsening congestion in this situation.

Hepatomegaly is an important but uncommon sign in patients with heart failure. The liver is usually tender except in long-standing heart failure and can pulsate during systole in the presence of tricuspid regurgitation. Firm and continuous compression of the right upper abdominal quadrant for 30s to 1min may exhibit hepato-jugular reflux, i.e. an increase in jugular distension that is sustained during and after compression (see jugular venous distension in Cardiac signs, p.853). Examination should be made on a patient lying comfortably with their head resting on a pillow.

Jugular venous distension, detected by inspection of the internal jugular veins, may identify an elevated right atrial pressure and by inference (and in the absence of tricuspid and pulmonary valve disease) left atrial pressure ( Fig. 23.13). Although, estimates of jugular venous pressure by physical examination correlate poorly with invasive measurement of right atrial pressure and interobserver reproducibility is low among non-specialists, elevation of jugular venous pressure is of prognostic importance [165, 166, 173–177]. Giant ‘V waves’ indicate the presence of tricuspid incompetence.

A third heart sound is usually only heard when there is left ventricular dilatation and systolic dysfunction, but interobserver agreement on this sign is low. A third heart sound is more common in severe heart failure and is associated with poor prognosis [165, 166, 173, 175].

A systolic murmur as because mitral or tricuspid regurgitation can be present, even in the absence of primary valve disease, when the left or right ventricle is markedly enlarged, leading to a dilatation of the mitral or the tricuspid annulus. In the latter case, the tricuspid murmur is selectively increased in loudness following inspiration (Carvallo’s sign). The degree of mitral regurgitation can be dynamic and increase on exertion.

Pulmonary crackles (crepitations or rales) result from the transudation of fluid from the intravascular space into the alveoli. The presence of crackles at the lung bases is suggestive of pulmonary congestion but the positive predictive value of this sign is low and the interobserver variability is high [165, 166, 173]. Moreover, the origin of crackles can be difficult to assess in smokers who may also have chronic pulmonary disease (or in patients at risk of pulmonary disease for some other reason). In acute pulmonary oedema, bubbling crackles may be accompanied by expectoration of frothy sputum which is blood stained. Patients with long-standing heart failure may become resistant to developing pulmonary oedema and only do so at very high left atrial pressures. Pleural effusions can also be detected in patients with heart failure; they are normally bilateral and usually associated with marked dyspnoea and generalized congestion.

Overall, the presence of several of the aforementioned symptoms and signs, particularly in the context of a history of previous cardiac disease, is suggestive of heart failure, if notits precise cause. The declining skill of physicians in clinical examination, the lack of sensitivity and specificity of most signs (and the large interobserver variability in their detection), and the subjective nature of clinical assessment highlight the need for objective assessment of cardiac function. This objective assessment is also essential for diagnosis of the cause of heart failure and, therefore, treatment tailored to aetiology.

A resting 12-lead electrocardiogram (ECG) is one of the most useful investigations in a patient with suspected heart failure and is recommended as a first-line diagnostic test in the European Society of Cardiology (ESC) guidelines ( Fig. 23.16) [1]. The ECG provides diagnostic and prognostic information and helps in choosing treatment. ECG changes are frequent in heart failure and a normal ECG virtually excludes left ventricular systolic dysfunction [1,177]. Various abnormalities may be present, such as abnormal Q waves ( Fig. 23.17), left bundle branch block ( Fig. 23.18) and other conduction disturbances, left atrial or left ventricular hypertrophy ( Fig. 23.19), atrial or ventricular arrhythmias may suggest a specific aetiology or precipitating factor. Some abnormalities provide prognostic information and help in choosing treatment, e.g. bundle branch block is predictive of a worse prognosis (178, 179) in patients with left ventricular systolic dysfunction and may also identify patients who benefit from a specific treatment (cardiac resynchronization therapy). The same is true of atrial fibrillation where there may be an indication for warfarin.

The chest X-ray (or radiograph) is also recommended as a first-line investigation ( Fig. 23.16). It may identify a non-cardiac cause for the patient’s symptoms. It also provides information on the size and shape of the cardiac silhouette and the state of the pulmonary vasculature [180]. This information is, however, of limited value [181]. The absence of cardiomegaly ( Fig. 23.20) does not exclude valve disease or left ventricular systolic dysfunction (heart size is more often normal in patients with left ventricular diastolic dysfunction and in acute compared to chronic heart failure). Even if cardiomegaly is present ( Fig. 23.20), the chest X-ray does not identify the cause of cardiac enlargement. The relationship between central haemodynamic and pulmonary vascular radiological abnormalities is also variable and some patients with long-standing severe heart failure may not show pulmonary venous congestion or oedema ( Fig. 23.20) despite a very high pulmonary capillary pressure [182].

Several laboratory investigations are recommended in the ESC guidelines as part of the routine diagnostic evaluation of patients with suspected heart failure: complete blood count, electrolytes, glucose, urea, creatinine, hepatic enzymes, and urinalysis [1]. Myocardial biomarkers such as troponin T or I are important during an acute episode to rule out myocardial infarction, although some degree of troponin elevation occurs in around a third of patients hospitalized with acute heart failure and is a powerful predictor of outcome [183, 184]. Other tests including serum uric acid, C-reactive protein, and thyroid stimulating hormone are optional. It is important to repeat some of these tests during follow-up and after the initiation of certain treatments (or dose changes), e.g. urea, creatinine, and potassium.

Electrolyte disturbances are uncommon in untreated mild to moderate heart failure.

Hyponatraemia is sometimes found in severe heart failure and its cause is complex and probably multifactorial. Impaired free water excretion, sodium restriction, and diuretic therapy (especially thiazide diuretic treatment or excessive use of loop diuretics) are thought to be important. Hyponatraemia may identify patients with an elevated plasma concentration of arginine vasopressin (AVP) and a particularly activated RAAS [134, 137]. AVP antagonists may have a therapeutic role in patients with hyponatraemia [185].

Potassium concentration is usually normal but may be reduced by the use of kaliuretic agents such as loop diuretics, or increased in patients with end-stage heart failure with markedly reduced glomerular filtration rate, particularly in the presence of concomitant renal disease, treatment with an inhibitor of the RAAS (e.g. some combination of an ACE inhibitor, angiotensin II receptor blocker, or spironolactone) [185]. Nephrotoxic drugs such as NSAIDs may also precipitate hyperkalaemia.

Elevation of creatinine and urea is common in treated heart failure, especially if severe. A significant reduction in glomerular filtration rate may be present despite a normal urea and creatinine, especially in patients with reduced muscle mass. Several causes are recognized: (1) reduced glomerular filtration rate in advanced heart failure; (2) excessive treatment with diuretics alone or in combination with inhibitors of the RAAS (e.g. some combination of an ACE inhibitor, angiotensin II receptor blockerl or spironolactone); (3) ageing; and (4) primary renal disease, including renal artery stenosis.

Impaired renal function and worsening renal function are associated with a poor outcome in chronic heart failure though, paradoxically, drugs which improve prognosis, particularly inhibitors of the RAAS, may cause some deterioration in renal function (although this is usually mild).

Elevation of liver enzymes and other markers, including aspartate aminotransferase (AST) and alanine aminotransferase (ALT) and of serum bilirubin is often observed in heart failure. These changes may be caused by reduced hepatic blood flow as much as by liver congestion and are prognostically important [187, 188].

Anaemia may be the cause or consequence of heart failure. It is common, especially in advanced heart failure, and is associated with a worse prognosis.

Urinalysis is important for the detection of proteinuria or glycosuria, indications of underlying renal disease and diabetes mellitus, respectively.

Thyroid function tests are indicated if either hyper- or hypothyroidism is suspected.

The use of plasma natriuretic peptides as a tool for the diagnosis of heart failure has developed in recent years, particularly in primary care and emergency units ( Fig. 23.21) [189]. Both BNP and N-terminal pro-BNP are available as commercial assays. In clinical practice, both are used as ‘rule out’ tests, i.e. a normal concentration of either peptide in an untreated patient means that it is very unlikely that heart failure is present [189]. Conversely, an elevated concentration identifies a patient who merits comprehensive cardiac examination including echocardiography. Natriuretic peptides may, therefore, offer a cost-effective means of ensuring the efficient use of echocardiography. One important caveat is that prior treatment may reduce natriuretic peptide concentrations to within the normal range.

Age and female gender both increase plasma concentration of natriuretic peptides and must be considered when defining ‘normal’ cut-off values which are also assay-specific [189]. Plasma levels are also increased in other conditions including renal dysfunction, pulmonary embolism, left ventricular hypertrophy, acute ischaemia, and hypertension (but reduced in obesity). Because these various cardiac and non-cardiac disorders lead to a moderate increase in plasma concentrations of natriuretic peptides, a ‘grey zone’ exists and is the reason why natriuretic peptides are used as a ‘rule out’ (rather than a ‘rule in’) test.

In heart failure with preserved EF, plasma levels of BNP, although generally higher than in patients without heart failure, are significantly lower than in patients with left ventricular systolic dysfunction ( Fig. 23.21) [127]. In a general population, the ability of plasma natriuretic peptides to detect left ventricular hypertrophy appears limited [190]. Patients with relaxation abnormalities and mild symptoms or who are asymptomatic may have normal levels of natriuretic peptides. Thus low levels may not exclude a diagnosis of heart failure with preserved EF although those with more severe diastolic dysfunction seem to have higher natriuretic peptide concentrations [106, 127].

Transthoracic Doppler echocardiography (or ultrasound) is recognized by the ESC guidelines as the most important investigation for the patient with suspected heart failure. Echocardiography is a widely available, rapid, non-invasive, and safe technique which provides extensive information on chamber dimensions, wall thickness, and measures of systolic and diastolic function. Disappointingly, despite the value of Doppler echocardiography, surveys in Europe commonly show that cardiac function is only evaluated in approximately 50% of patients treated for suspected heart failure [191, 192]. The recent development of portable echocardiography machines, together with the potential of remote analysis of echocardiograms, may help improve this situation.

Determination of the LVEF is a key measure of left ventricular systolic function and has been used to enrol patients in almost all important clinical trials. Systolic function is usually considered to be reduced when the LVEF is <0.40 and ‘preserved’ if > 0.50. This clearly leaves a ‘grey area’ in the range 0.40–0.50. Furthermore, LVEF is a crude method for the evaluation of systolic function and is dependent not only on the intrinsic inotropic state of the myocardium, but also on the loading conditions of the heart.

The most widely recommended approach to the accurate measurement of LVEF is the modified Simpson’s method, using apical biplane summation of discs ( Fig. 23.22) [193]. It is, however, dependent on reliable endocardial detection. Other methods are less accurate, particularly in the presence of regional hypokinesia or akinesia. Other measures are, however, used including fractional shortening, sphericity index, and left ventricular wall motion index.

Identification of diastolic dysfunction is more difficult and requires evidence of abnormal left ventricular relaxation ( Fig. 23.23) or diastolic distensibility or diastolic stiffness. Which measures to use in daily practice are still debated [106, 194].

The finding of morphological changes such as left atrial dilatation or left ventricular hypertrophy, as well as abnormal measures of left ventricular diastolic function, is particularly helpful in determining whether diastolic dysfunction is clinically important. Abnormal diastolic function has prognostic as well as diagnostic significance [106, 195].

It is important to appreciate that the terms heart failure with preserved EF and heart failure as a result of diastolic dysfunction (‘diastolic heart failure’) describe overlapping but not identical syndromes.

Valve function can also be assessed by Doppler echocardiography, e.g. semi-quantitative evaluation of mitral regurgitation is possible ( Fig. 23.14), as is calculation of pulmonary artery systolic pressure (based on the velocity of tricuspid regurgitation). Doppler echocardiography may also be repeated during the follow-up of patients receiving treatment, to assess changes in cardiac structure and function.

Other new ultrasound techniques, such as three-dimensional echocardiography, are still being evaluated clinically.

Dobutamine (or exercise) stress echocardiography ( Chapter 4) may be used to detect ischaemia as a cause of cardiac dysfunction and to assess myocardial viability in the presence of marked hypokinesia or akinesia. It may be used to identify myocardial stunning and hibernating myocardium [196].

This technique provides information on left and right ventricular volumes and EF but is not widely available. It does not provide information on valve function but is generally more accurate than echocardiography for measuring right ventricular function. The reproducibility of this technique is also better than that of echocardiography.

Cardiac magnetic resonance (CMR) allows comprehensive and reproducible analysis of cardiac anatomy and function, including cardiac volumes and mass, global and regional function, and wall thickening [197]. When combined with contrast agents such as gadolinium, CMR also provides information on myocardial perfusion at rest ( Fig. 23.24) or following pharmacological intervention. CMR is now the gold standard for the assessment of mass volumes and wall motion. It also can give clues to the aetiology of specific cardiomyopathies (e.g. arrhythmogenic right ventricular cardiomyopathy) and specific heart muscle diseases. CMR is expensive and is not as widely available as echocardiography. In addition, CMR cannot be performed in patients with certain metallic implants, including cardiac devices, and patient refusal is relatively common because of claustrophobia and other factors.

Treadmill or bicycle testing can be used to determine the maximum level of exercise which can be achieved. Recent recommendations on testing have been published [198]. Small increments in workload are recommended. As well as exercise duration, gas exchange is often measured. Peak oxygen uptake (vo  ₂) and the anaerobic threshold are useful indicators of the patient’s capacity to exercise. There is a poor correlation between exercise capacity and EF. Peak vo  ₂ has been used also for prognostication and selection of patients for heart transplantation. A peak vo  ₂ >14mL/kg/min is associated with a relatively good prognosis unlikely to be improved by heart transplantation. Patients with a vo  ₂ max of <14mL/kg/min have been shown to have a better survival if transplanted than if treated medically. This latter threshold is part of the criteria used for listing patients for heart transplantation [195]. It should be noted that the survival studies on which this threshold is based were carried out before the advent of modern treatments for heart failure, such as beta-blockers and cardiac resynchronization therapy.

Pulmonary function testing is indicated in patients in whom the origin of dyspnoea is unclear, to determine whether it is of cardiac or pulmonary origin or both. Heart failure itself is associated with abnormalities of pulmonary function [200]. These include reductions in vital capacity, pulmonary diffusion at rest (and on exercise), and pulmonary compliance. Conversely, airway resistance is usually increased.

Pulmonary arterial catheterization is usually only used in acute or emergency situations, especially in patients not responding to appropriate medical therapy and is not routinely indicated [201]. This procedure allows close monitoring of haemodynamic changes following medical intervention. It is also indicated in patients with valvular heart disease who are candidates for valve replacement or repair and in the assessment of patients for transplantation.

Coronary angiography is indicated in patients with heart failure and angina or evidence of myocardial ischaemia, if revascularization is being considered. However, the role of revascularization in the treatment of heart failure, including in patients with hibernating myocardium, remains to be determined [64, 202].

Coronary angiography may also be indicated in acute heart failure with shock that is not responding to initial therapy.

Many also believe that coronary angiography is indicated as a diagnostic test in patients with heart failure or left ventricular systolic dysfunction of unknown origin.

Endomyocardial biopsy of the left or the right ventricle is only indicated when a specific myocarditis or a specific myocardial disease is suspected and during the follow-up of patients after cardiac transplantation to detect graft rejection [203].

Overall, the indication for invasive procedures in heart failure has declined. They are not necessary to establish the presence of heart failure but may be helpful on an individual basis to determine the aetiology and for the monitoring of patients in acute or difficult situations.

Ambulatory monitoring is valuable in the assessment of patients with symptoms suggestive of an arrhythmia (e.g. palpitations and syncope) and in monitoring ventricular rate control in patients with atrial fibrillation.

Asymptomatic ventricular premature beats and non-sustained ventricular tachycardia are frequent in heart failure but do not appear to be predictive of sudden death or of selecting treatment to reduce sudden death.

The following describes a stepwise approach to the diagnosis of the patient with suspected heart failure ( Table 23.4).

This requires the presence of signs and symptoms suggestive of heart failure at rest or on exercise and also objective evidence of a cardiac abnormality, preferably by Doppler echocardiography. Doppler echocardiography may not be necessary in previously untreated patients with a normal plasma natriuretic peptide concentration and these patients should be investigated for another cause of their symptoms or signs ( Fig. 23.16). Other essential first-line diagnostic tests should be performed, including an ECG, chest X-ray, and blood tests. Pulmonary function and exercise testing may help when the diagnosis remains doubtful.

The diagnosis of heart failure with preserved EF can be particularly difficult. Although often described as a diagnosis of exclusion, i.e. other possible causes of the patient’s symptoms and signs must be ruled out, this type of heart failure may co-exist with other relevant comorbidities, e.g. chronic lung disease and anaemia. Figs. 23.25 and 23.26 show an approach to this problem.

This step includes:

Ischaemic heart disease and hypertension are the commonest causes of heart failure in Western countries and their contribution to heart failure is also rapidly expanding in the developing countries. In the Euro Heart Failure Survey, ischaemic heart disease was the commonest cause of heart failure, 40% of patients having myocardial infarction and 51% a history of angina; 53% of the patients also had hypertension [191]. The extent of investigation to identify the underlying cardiac disease should be decided on an individual basis taking into consideration the patient profile (age, severity of heart failure and comorbidities) and the potential reversibility of cardiac dysfunction (such as valvular heart disease and reversible ischaemia).

It is also important to identify precipitating factors: decompensation of heart failure is frequently associated with comorbid factors which:

Air pollution and seasonal temperature changes are more recently described factors associated with risk of decompensation and hospitalization [200, 201].

Among arrhythmias, a new episode of atrial fibrillation can be particularly harmful because it combines both an increase in ventricular rate (with a risk of myocardial ischaemia and reduced time for diastolic filling) and loss of atrial contraction [205, 206].

Prognostic assessment in heart failure remains difficult. Indeed the number of clinical, aetiological, comorbid, biological, haemodynamic, structural, functional, electrical, and neurohumoral variables independently associated with poor outcome is high and suggests that there is no simple method to assess the risk of death or rehospitalization in patients with this syndrome ( Table 23.5) [209, 210]. Most studies have been performed in populations of patients with systolic dysfunction and little is known about prognostic evaluation in patients with preserved systolic function.

The assessment of prognosis has also been conducted differently in acute and in chronic heart failure: in acute heart failure, in-hospital and short-term (3–6 months) mortality or readmission have usually been evaluated whereas in chronic heart failure, long-term (> 1 year) prognosis and rehospitalization rates have normally been considered.

Furthermore, factors which predispose to overall mortality or pump failure mortality do not necessarily apply to sudden death.

An additional problem is that extrapolation of risk assessment based on a small series of selected patients exposed to conventional therapy (including low rate of prescription of ACE inhibitors and beta-blockers) to the overall current heart failure population is difficult. The changing background therapy of heart failure also makes it difficult to provide simple prognostic algorithms. For instance, beta-blockers have more influence on the remodelling process than exercise capacity, so that the relative role of these two independent predictors of mortality may be different in patients treated with a beta-blocker and those not so treated.

The temporal role of the various prognostic factors can be variable: the time course of the activation of the neurohumoral systems after myocardial injury is different. Therefore, the relative weight of elevated plasma levels of neurohumoral factors may be different in the short term compared to the longer term. Moreover, little is known about the relation between the change in plasma concentrations of biochemical markers as a result of treatment and long-term prognosis [211, 212].

The multivariable analyses reported so far usually include only a limited number of parameters and these may lose their predictive power in more comprehensive analyses.

Finally, some factors can provide prognostic information in advanced heart failure but not in mild to moderate heart failure: severely reduced functional capacity, measured by VO₂, is a recognized index for the selection of patients who might benefit from heart transplantation whereas elevated, non-reversible, pulmonary resistance is an index of poor outcome after heart transplantation or implantation of a ventricular assist device [199].

Full diagnosis is a prerequisite for optimal treatment [1]. Often, however, a diuretic may be required before a full diagnostic work-up is completed. The further treatment of heart failure depends on the underlying aetiology (e.g. valve replacement for aortic stenosis), functional status (e.g. spironolactone for severely symptomatic patients), comorbidity (e.g. warfarin if atrial fibrillation), and the results of investigations (e.g. cardiac resynchronization therapy if a broad QRS on the ECG). Each of these may contraindicate treatments as well as indicate them (e.g. caution with ACE inhibitors in aortic stenosis, beta-blockers if atrioventricular block, spironolactone if renal dysfunction etc.).

The goals of treatment for patients with heart failure are relief of symptoms, avoidance of hospital admission, and prevention of premature death.

Drugs are the mainstay of the treatment of patients with heart failure based upon a series of key randomized controlled trials ( Table 23.6). However, devices and surgery have an important and increasing role, particularly in patients with more severe heart failure ( Fig. 23.28). How care is structured and delivered is also important. Although lifestyle measures are also considered important, the evidence base for these interventions is less robust.

Diuretics act by blocking sodium reabsorption at specific sites in the renal tubule, thereby enhancing urinary excretion of sodium and water.

Although not proven to improve mortality and morbidity in large trials, diuretics are required in nearly all patients with symptomatic heart failure to relieve dyspnoea and the signs of sodium and water retention (‘congestion’), i.e. peripheral and pulmonary oedema [213, 214]. No other treatment relieves symptoms and the signs of sodium and water overload as rapidly and effectively. Once a patient needs a diuretic, treatment is usually necessary for the rest of the patient’s life, though the dose and type of diuretic may vary.

The key principle is to prescribe the minimum dose of diuretic needed to maintain an oedema-free state (‘dry weight’) [213]. Excessive use can lead to electrolyte imbalances, such as hyponatraemia, hypokalaemia (with the risk of digitalis toxicity), hyperuricaemia (with the risk of gout), and uraemia. The risk of renal dysfunction is increased by concomitant use of NSAIDs. Diuretic-induced hypovolaemia may also cause symptomatic hypotension and pre-renal uraemia. Restriction of dietary sodium intake may help reduce, but not eliminate, the requirement for diuretics. Diuretic dosing should be flexible with temporary increases for evidence of fluid retention (e.g. increasing symptoms, weight gain, oedema) and decreases for evidence of hypovolaemia (e.g. as consequence of increased electrolyte loss owing to gastroenteritis, decreased fluid intake, or both).

In some patients with milder symptoms of heart failure (NYHA class II), a thiazide diuretic such as bendroflumethiazide may suffice. In more severe heart failure or in patients with concomitant renal dysfunction, a loop diuretic such as furosemide is often needed. Loop diuretics cause a rapid onset of an intense but relatively short-lived diuresis, compared with the longer-lasting but gentler effect of a thiazide diuretic. The timing of administration of a loop diuretic, which need not be taken first thing every morning, can be adjusted according to the patient’s social activities. The dose may be postponed or even temporarily omitted if the patient has to travel or has another activity that might be compromised by the prompt action of the diuretic. In severe heart failure, the effects of long-term administration of a loop diuretic may be diminished by increased sodium reabsorption at the distal tubule. This problem can be offset by use of the combination of a loop diuretic and a thiazide or thiazide-like diuretic (e.g. metolazone), which acts in synergy with a loop diuretic by blocking sodium reabsorption in different segments of the nephron [215]. This combination requires more frequent monitoring of electrolytes and renal function for diuretic-induced hyponatraemia, abnormalities of the serum potassium level, and prerenal uraemia.

A period of intravenous loop diuretic, given either as bolus injections or by continuous infusion, may be required in patients who become resistant to the action of oral diuretics [216]. Why this resistance develops is uncertain, but factors thought to be important include impaired absorption of oral diuretics owing to gut oedema, hypotension, reduced renal blood flow, and adaptive changes in the nephron. In cases of severe resistant volume overload, mechanical removal of fluid using ultra-filtration may be considered [217].

Patients with severe failure (NYHA class III–IV) should usually also be treated with an aldosterone antagonist, such as spironolactone, which increases excretion of sodium but not potassium [218]. Patients receiving a combination of diuretics require careful monitoring of blood chemistry and clinical status. The use of a potassium sparing diuretic or aldosterone antagonist along with an ACE inhibitor or ARB (treatment with all three is not recommended) requires particular care and surveillance for hyperkalaemia [219].

Although highly effective in relieving symptoms and signs, diuretics alone are not a sufficient treatment for heart failure. The addition of other disease-modifying treatments is essential to slow pathophysiological progression, better maintain clinical stability, and reduce the risk of hospital admission and premature death.

These drugs act by inhibiting the enzyme that converts the inactive decapeptide angiotensin I to the active octapeptide angiotensin II. Excessive angiotensin II is thought to exert the myriad of harmful actions described earlier by stimulating the angiotensin II type 1 receptor subtype (AT₁R)—see Pathophysiology, p.843. ACE inhibitors also reduce the breakdown of bradykinin (as ACE is identical to kininase II) and the resultant accumulation of bradykinin is directly or indirectly responsible for two of the specific adverse effects of ACE inhibitors: cough and angio-oedema. Bradykinin may, however, also have beneficial effects (vasodilation, inhibition of adverse cardiovascular remodelling, and antithrombotic actions), though the importance of these bradykinin-mediated actions to the clinical benefits of ACE inhibition is uncertain.

When added to diuretics and digoxin, treatment with an ACE inhibitor decreases left ventricular size, improves systolic function, reduces symptoms and hospital admissions, and prolongs survival ( Table 23.6; Fig. 23.29) [220, 221]. These agents also reduce the risk of developing myocardial infarction and possibly atrial fibrillation [222]. Consequently, treatment with an ACE inhibitor is recommended for all patients with systolic dysfunction, irrespective of symptom severity or aetiology. ACE inhibitors are not a substitute for a diuretic but mitigate diuretic-induced hypokalaemia.

ACE inhibitors should be introduced as early as possible in a patient’s treatment [219]. The only contraindications are current symptomatic hypotension, severe aortic stenosis, and bilateral renal artery stenosis; the latter is often associated with a prompt and marked increase in serum levels of blood urea and creatinine when renal perfusion is reduced precipitously by inhibiting the production and actions of angiotensin. Treatment should be started in a low dose ( Table 23.7) and gradually increased toward that proven to be of benefit in a clinical trial ( Table 23.8). The patient should be evaluated for symptomatic hypotension, uraemia, and hyperkalaemia after each dose increment; these adverse effects are uncommon and can usually be resolved by reduction in the dose of diuretic (if the patient is oedema free) or concomitant hypotensive or nephrotoxic medications (e.g. nitrates, calcium-channel blockers, or NSAIDs). A dry, non-productive cough occurs in approximately 15% of patients treated with an ACE inhibitor and, if troublesome, substitution of an ARB is recommended [223]. In the rare cases of angio-oedema, the ACE inhibitor should be stopped and not used again; an ARB can be cautiously substituted.

Instead of inhibiting the production of angiotensin II through ACE, ARBs block the binding of angiotensin II to the AT₁R. This distinct mechanism of action may be important because angiotensin II is also believed to be produced by other enzymes such as chymase. ARBs do not inhibit kininase II or the breakdown of bradykinin, so they do not cause cough or angio-oedema.

When used as the sole RAAS blocking agent in addition to a diuretic and digoxin, ARBs produce similar benefits to ACE inhibitors and can be substituted for them in patients who have cough or angio-oedema with ACE inhibitors [223, 224]. When used in clinically-effective doses, other adverse effects such as hypotension, renal dysfunction, and hyperkalaemia are encountered as frequently as with an ACE inhibitor. As with an ACE inhibitor, the specific agents, dosing regimens, and target doses that were of demonstrable benefit in clinical trials are recommended ( Table 23.7).

An ARB used in combination with an ACE inhibitor (and beta-blocker) further improves LVEF, relieves symptoms, reduces the risk of hospital admission for worsening heart failure, and can also reduce the risk of cardiovascular death ( Table 23.6) [225–228]. Consequently, the addition of an ARB to both an ACE inhibitor and a beta-blocker should be considered in any patient with persisting symptoms (NYHA class II–IV). There is, however, also strong evidence that adding an aldosterone antagonist to an ACE inhibitor is of benefit in patients with advanced (NYHA class III–IV) heart failure (see Aldosterone antagonists, Clinical benefits, p.874), but the efficacy and safety of the four-drug combination of an ACE inhibitor, beta-blocker, ARB, and aldosterone antagonist are uncertain [218, 229]. Consequently, either an ARB or an aldosterone antagonist, but not both, should be added to an ACE inhibitor and a beta-blocker in such patients.

The approach to initiation, titration, and monitoring of an ARB is similar to an ACE inhibitor ( Table 23.9). The adverse effects, with the exception of cough and angio-oedema, are similar. Use of multiple inhibitors of the RAAS requires even more diligent monitoring, especially in patients at higher risk of uraemia, hypotension, or hyperkalaemia (i.e. patients ≥75 years of age or with a systolic blood pressure <100mmHg, diabetes, or renal impairment) [219, 230, 231].

As with ACE inhibitors, beta-blockers and aldosterone antagonists, treatment with ARBs should be indefinite unless there is intolerance.

Beta-blockers are believed to counteract the many harmful effects of sympathetic nervous system hyperactivity described earlier—see Pathophysiology, p.843.

The long-term addition of a beta-blocker to an ACE inhibitor (and diuretic and digoxin) further improves left ventricular function and symptoms, reduces hospital admissions, and improves survival, strikingly ( Table 23.6; Fig. 23.30). Consequently, a beta-blocker is recommended for all patients with symptomatic systolic dysfunction, irrespective of aetiology and severity. The combination of a beta-blocker with an ACE inhibitor is now the cornerstone of the treatment of symptomatic heart failure ( Fig. 23.28) [232–239].

The major contraindications to using a beta-blocker in heart failure are asthma (though it is important to note that the dyspnoea caused by pulmonary congestion can be confused with reactive airway disease) and second- or third-degree atrioventricular block [219]. Initiation of treatment during an episode of acute decompensated heart failure should also generally be deferred until the patient is stabilized and recovering (but, ideally, treatment should be commenced before discharge). In addition, caution is advised in patients with a heart rate <60 beats per minute (bpm) or a systolic blood pressure <100mmHg. It is recommended that a beta-blocker shown to produce benefits in a randomized trial be used ( Table 23.7).

As with ACE inhibitors, beta-blockers should be introduced as early as possible in a patient’s treatment, started in a low dose ( Table 23.7) and increased gradually toward the target dose used in a clinical trial (the ‘start low–go slow’ approach). The patient should be checked for symptomatic hypotension and excessive bradycardia after each dose increment, but both of these side effects are uncommon, and hypotension can often be resolved by reduction in the dose of other non-essential blood pressure-lowering medications, e.g. nitrates and calcium-channel blockers ( Table 23.10). Bradycardia is more likely in patients who are also taking digoxin, ivabradine, or amiodarone and the necessity for the simultaneous use of these agents should be reviewed if excessive bradycardia occurs. Occasionally, symptomatic worsening and fluid retention (i.e. weight gain or oedema) may occur after initiation of a beta-blocker or during dose up-titration; these side effects usually can be resolved by a temporary increase in the diuretic dose without necessitating discontinuation of the beta-blocker.

Treatment with a beta-blocker should be given for life, though the dose may need to be decreased or temporarily discontinued during episodes of acute decompensation if the patient shows signs of circulatory hypoperfusion or refractory congestion.

These agents block the undesirable actions of aldosterone described earlier and act as potassium sparing diuretics.

The aldosterone antagonist spironolactone ( Table 23.6) improves symptoms, reduces hospital admissions, and increases survival when added to an ACE inhibitor (and diuretics and digoxin) in patients with a reduced LVEF and severely symptomatic heart failure ( Fig. 23.31) [218]. Eplerenone, another aldosterone antagonist, reduces mortality and morbidity when added to both an ACE inhibitor and beta-blocker in patients with a reduced LVEF and heart failure or diabetes after a recent myocardial infarction ( Table 23.7). Consequently, an aldosterone antagonist should be considered in patients who remain in severe heart failure (NYHA class III or IV) despite treatment with a diuretic, ACE inhibitor (or ARB), and beta-blocker. When begun, it should be given indefinitely. The value of an aldosterone antagonist in patients with milder heart failure is uncertain but is under investigation. The combination of an ACE inhibitor, an ARB, and an aldosterone antagonist has not been adequately evaluated and is not recommended  .

Treatment with an aldosterone antagonist should be initiated with a low dose ( Table 23.7) with careful monitoring of serum electrolytes and renal function ( Table 23.11). Hyperkalaemia and uraemia are the adverse effects of greatest concern (as with ACE inhibitors and ARBs) and an aldosterone antagonist should not be given to patients with a serum potassium concentration of >5mmol/L, serum creatinine >221µmol/L (>2.5mg/dL) or other evidence of markedly impaired renal function. The importance of patient selection and dose are underscored by reports of a worrisome incidence of serious hyperkalaemia in community practice settings. Spironolactone can have antiandrogenic effects, especially painful gynaecomastia, in men. Since eplerenone does not block the androgen receptor, it is a reasonable substitute in patients who experience this adverse effect [240].

Digitalis glycosides inhibit the cell membrane sodium–potassium adenosine triphosphatase (ATP) pump, thereby increasing intracellular calcium and myocardial contractility. In addition, digoxin is thought to enhance parasympathetic and reduce sympathetic nervous activity, as well as inhibit renin release [241].

Only one large randomized placebo-controlled trial examined the effects of starting (as opposed to withdrawing) digoxin on mortality and morbidity in patients with heart failure in sinus rhythm [241, 242]. In that trial, digoxin did not reduce mortality but did decrease the risk of admission to hospital for worsening heart failure when added to a diuretic and an ACE inhibitor ( Table 23.6). In patients in sinus rhythm, addition of digoxin is recommended only for patients whose heart failure remains symptomatic despite standard three-drug treatment with a diuretic, ACE inhibitor, beta-blocker, and either an ARB or aldosterone antagonist. In patients with atrial fibrillation, digoxin may be used at an earlier stage if a beta-blocker fails to control the ventricular rate (ideally <80bpm at rest and <110–120bpm during exercise; see Chapter 29). Digoxin can also be used to control the ventricular rate when beta-blocker treatment is being initiated or up-titrated.

Digoxin should be avoided in patients with second-degree or greater atrioventricular block and pre-excitation syndromes; it should be used with caution in patients with sick-sinus syndrome. Hypokalaemia should be corrected before digoxin is administered. A loading dose of digoxin is generally not needed in stable patients in sinus rhythm. A single daily oral maintenance dose of 0.25mg is commonly used in adults with normal renal function. In the elderly and in those with renal impairment, a dose of 0.125mg or 0.0625mg may suffice. If the effect of digoxin is needed urgently, loading with 10–15mcg/kg lean body weight, given in three divided doses, 6 hours apart, may be used. The maintenance dose should be one-third of the loading dose. Smaller maintenance doses (e.g. one-quarter the loading dose and not more than 62.5mcg/day) should be used in the elderly and in patients with reduced renal function, as well as in patients with a low body mass. Monitoring of the serum digoxin concentration is recommended because of the narrow therapeutic window. A steady state is reached 7–10 days after starting treatment; blood should be collected at least 6 hours (and ideally 8–24 hours) after the last dose. The currently recommended therapeutic range is 0.6–1.2ng/mL (approximately 0.77–1.54nmol/L) [1, 2, 243].

Digoxin can cause anorexia, nausea, sinoatrial and atrioventricular block, arrhythmias, confusion, and visual disturbances (including xanthopsia), especially if the serum concentration is >2.0ng/mL. Hypokalaemia increases susceptibility to the adverse effects. The dose of digoxin should be reduced in the elderly and patients with renal dysfunction. Certain drugs increase serum digoxin concentration, including amiodarone, verapamil, and diltiazem.

Hydralazine is a powerful direct acting, arterial vasodilator. Its mechanism of action is not understood, though it may inhibit enzymatic production of superoxide, which neutralizes NO and may induce nitrate tolerance. Nitrates dilate both veins and arteries, thereby reducing preload and afterload by stimulating the NO pathway and increasing cyclic GMP in vascular smooth muscle. Neither drug on its own nor any other direct acting vasodilator has been demonstrated to be beneficial in heart failure.

Although this combination has been known for some time to improve systolic function and probably reduce death in NYHA class II–IV heart failure compared with placebo, a head-to-head comparison showed an ACE inhibitor was superior for improving survival ( Table 23.6) [244, 245]. Nevertheless, based on subgroup analyses suggesting that African Americans responded better to hydralazine and isosorbide dinitrate, a subsequent randomized controlled trial showed that the addition of hydralazine and isosorbide dinitrate in African Americans, most of whom were receiving an ACE inhibitor and beta-blocker and many of whom were on spironolactone, further reduced mortality and hospital admissions for heart failure and improved quality of life [246]. A fixed combination of 37.5mg of hydralazine and 20mg of isosorbide dinitrate was used in the trial; one tablet was given and if tolerated a second was given 12 hours later. One tablet was then prescribed three times daily for 3–5 days, at which point the dose was increased to the target maintenance of two tablets three times daily, i.e. a daily dose of 225mg hydralazine and 120mg isosorbide dinitrate. Because of the limited inclusion criteria of this trial, however, it is uncertain whether this combination of vasodilators is an effective addition in other patient populations.

Other than for African Americans, the main indication for hydralazine and isosorbide dinitrate is as a substitute in patients with intolerance to an ACE inhibitor and an ARB. Hydralazine and isosorbide dinitrate should be used as additional treatment in African Americans and considered, on an empirical basis, for other patients who remain symptomatic on other proven therapies. The main dose-limiting adverse effects with hydralazine and isosorbide dinitrate are headache and dizziness. A rare adverse effect of higher doses of hydralazine, especially in slow acetylators is a systemic lupus erythematosus-like syndrome.

A recent study showed that treatment with 1g of omega-3 polyunsaturated fatty acids (n-3 PUFA) (850–852mg eicosapentaenoic acid and docosahexaenoic acid as ethyl esters in the average ratio of 1:1.2) per day led to a small reduction in cardiovascular morbidity and mortality in patients with heart failure ( Table 23.6) [247]. The exact mechanism of action of this treatment is uncertain, although it may have beneficial, anti-inflammatory and electrophysiological effects (the latter reducing the risk of arrhythmias).

It is important to note that the aforementioned treatments are the only pharmacological agents shown to be of benefit in patients with heart failure and a low LVEF. Many other treatments have been tested in randomized trials and shown to have a neutral (e.g. amlodipine) [248] or uncertain (e.g. alpha-adrenoceptor blockers, bosentan, and etanercept) [249–252] effect on mortality and morbidity. Some also increased mortality and have either been withdrawn from the market or should be avoided in heart failure (e.g. dronedarone, milrinone, flosequinan, vesnarinone, and moxonidine) [253–257].

Other therapies of proven value for cardiovascular conditions underlying or associated with heart failure have not been specifically tested in heart failure (e.g. antiplatelet treatment in patients with coronary heart disease, see Chapter 16) or may not be as clearly beneficial in heart failure (e.g. statins) [258, 259]. In patients with advanced systolic heart failure of ischaemic aetiology, statins did not reduce the primary outcome of cardiovascular death, myocardial infarction, or stroke (or all-cause mortality) but did reduce the number of cardiovascular hospitalizations [258]. A vitamin K antagonist, e.g. warfarin, is indicated in patients with atrial fibrillation provided there is no contraindication to its use ( Chapters 11 and 29). Warfarin may also be used in patients with evidence of intra-cardiac thrombus (e.g. detected during echocardiographic examination) or systemic thromboembolism. Warfarin’s many interactions with other drugs, including with some statins and amiodarone ( Chapter 11), must always be considered when initiating warfarin or another drug in a patient taking warfarin. Low-molecular-weight heparin prophylaxis ( Chapter 37) against deep venous thrombosis is indicated when patients with heart failure are bed-bound, for example during hospital admission, although heparin can cause hyperkalaemia and the dose must be reduced in patients with renal impairment.

Vaccination against influenza and pneumococcal infection is advised in all patients with heart failure because the stress of infection can lead to clinical deterioration [1].

Patients with heart failure, especially if severe, often have renal and hepatic dysfunction, so any drug excreted predominantly by the kidneys or metabolized by the liver may accumulate [75, 83–85, 187, 188]. Similarly, because of their extensive comorbidity, patients with heart failure are inevitably treated with multiple drugs, thereby increasing the risk of drug interactions.

Drugs that should be avoided, if possible, in heart failure include most antiarrhythmic drugs including dronedarone [257] (with the exception of amiodarone and dofetilide), most calcium channel blockers (with the exception of amlodipine), corticosteriods, NSAIDs, COX-2 inhibitors, and many antipsychotics (e.g. clozapine) and antihistamines. Thiazolidinediones (because of the risk of fluid retention) should be used with caution [70, 81–83, 260]. The use of metformin is usually not recommended because of the risk of lactic acidosis although this may be overstated [70]. Some salt substitutes contain substantial amounts of potassium and must be used cautiously. Other dietary constituents (e.g. grapefruit and cranberry juice) and supplements such as St. John’s Wort can interact with drugs taken by patients with heart failure, especially warfarin and digoxin [1, 2].

Several studies have shown that organized, nurse-led, multidisciplinary care can improve outcomes in patients with heart failure, particularly by reducing recurrent hospital admissions [261–263]. The most successful approach seems to involve education of the patients, their families, and caregivers about heart failure and its treatment (including flexible diuretic dosing and reinforcing the importance of adherence), recognizing (and acting upon) early deterioration (dyspnoea, weight gain, oedema), and optimizing proven pharmacological treatments (see General measures, p.913) ( Table 23.12). A home-based rather than clinic-based approach may be best, though trials are needed to compare these types of interventions directly. Even telephone follow-up is of value. New technology enabling non-invasive home telemonitoring of physiological measures (e.g. heart rate and rhythm, blood pressure, temperature, respiratory rate, weight, and estimated body water content) and implanted devices, which collect similar data and may be interrogated remotely, are also being tested as aids to monitoring and management [264, 265].

Education of the patient, family, and caregivers is invaluable ( Table 23.12) [1, 2, 266]. Detection of early signs and symptoms of deterioration provides for earlier intervention. Counselling on the proper use of therapies, with an emphasis on adherence, is critical.

Useful patient and carer orientated material is available from the Heart Failure Association of the European Society of Cardiology in several languages (currently English, French, German, and Spanish) and from the Heart Failure Society of America and other organizations (see Online resources, p.892).

When appropriate, a patient should be taught how to adjust the dose of diuretic within individualized limit—see Diuretics, p.865. The dose should be increased (or a supplementary diuretic added) if there is evidence of fluid retention (symptoms of congestion), and decreased if evidence of hypovolaemia (e.g. increased thirst associated with weight loss or postural dizziness, especially during hot weather or an illness causing decreased fluid intake or sodium and water loss). If hypovolaemia is more marked, the doses of other medications (e.g. ACE inhibitors, spironolactone) also will have to be reduced.

The expected effects, beneficial and adverse, of other drugs should also be explained in detail (e.g. possible association of cough with ACE inhibitor). It is useful to inform patients that improvement with many drugs is gradual and may become fully apparent only after several weeks or even months of treatment. It is also important to explain the need for gradual titration with ACE-inhibitors, ARBs, and beta-blocking drugs to a desired dose level, which again may take weeks or even months to achieve. Patients should be advised not to use NSAIDs without consultation and to be cautious about using herbal or other non-proprietary preparations.

Education and counselling of the patient, caregiver, and family promotes adherence, which is associated with better outcomes [267]. Drug adherence can also be helped by certain pharmacy aids such as dose allocation (‘Dosette’) boxes.

The recent Heart Failure: A Controlled Trial Investigating Outcomes of Exercise TraiNing (HF-ACTION) trial also showed that a programme of tailored, structured, aerobic exercise is safe, improves functional capacity and quality of life, and may also reduce cardiovascular mortality and heart failure hospitalization [268, 269]. The regimen used was, however, labour intensive ( Table 23.13) and, on average, patients did not attain the target level of exercise, especially during the home-maintenance phase.

Most guidelines advocate avoidance of foods containing relatively high salt content in the belief that this may reduce the need for diuretic therapy, although there is little evidence from clinical trials to support this recommendation [1]. Some also believe that excess sodium intake can be a precipitant of clinical decompensation. Certain salt substitutes have a high potassium content, which can lead to hyperkalaemia.

Restriction of fluid intake is indicated only during episodes of decompensation associated with peripheral oedema or hyponatraemia. In these situations, daily intake should be restricted to 1.5–2L to help facilitate elimination of excess extracellular fluid volume and avoid hyponatraemia.

Reducing excessive weight will reduce the work of the heart and may lower blood pressure and is recommended in those who are obese (body mass index >30kg/m²). Conversely, malnutrition is common in severe heart failure, and the development of cardiac cachexia is an ominous sign [77, 90, 91]. Sometimes reduced food intake is caused by nausea (e.g. related to digoxin use or hepatosplenic congestion) or abdominal bloating (e.g. due to ascites). In these cases, small frequent meals and high protein and calorie liquids may be helpful. In severe decompensated heart failure, eating may be difficult because of dyspnoea.

Moderate alcohol intake (up to 10–20g/day, e.g. 1–2 standard glasses of wine is permissible) is not thought to be harmful in heart failure, although excessive intake can cause cardiomyopathy and atrial arrhythmias in susceptible individuals. In patients with suspected alcoholic cardiomyopathy, abstinence from alcohol may improve cardiac function and is recommended.

Smoking causes peripheral vasoconstriction, which is detrimental in heart failure [1, 2]. Nicotine replacement therapy is believed to be safe in heart failure. The safety of bupropion and varenicline in heart failure are uncertain.

Sexual activity need not be restricted in patients with compensated heart failure, though dyspnoea may be limiting [1, 2]. Sublingual nitrate may be used as prophylaxis against chest pain and dyspnoea caused by sexual activity. In men with erectile dysfunction, treatment with a cyclic guanine monophosphate phosphodiesterase type V inhibitor, such as sildenafil, can be useful, but these drugs must not be taken within 24 hours of prior nitrate use, and nitrates must not be restarted for at least 24 hours afterwards [270].

Patients with heart failure can continue to drive provided their condition does not induce undue dyspnoea, fatigue, or other incapacitating symptoms [1, 2]. Patients with recent syncope, cardiac surgery, percutaneous coronary intervention, or device placement may be restricted from driving, at least temporarily, according to local regulations. Patients holding an occupational or commercial licence may be subject to additional restrictions.

Short flights are unlikely to cause problems for a patient with compensated heart failure [1, 2]. Cabin pressure is generally maintained to provide an oxygen level no lower than equivalent to 6000 feet above sea level, which should be well tolerated in patients without severe pulmonary disease or pulmonary hypertension. Longer journeys may cause limb oedema and dehydration, thereby predisposing to venous thrombosis. Adjustment of the dose of diuretics and other treatments should be discussed with the patient wishing to travel to a warm climate or a country where the risk of gastroenteritis is high. It is also advisable for heart failure patients to carry a list of medications and contact information for their healthcare provider.

Beta-blockers are of benefit in both angina and heart failure. Nitrates relieve angina but, on their own, are not of proven value in chronic heart failure. Calcium-channel blockers should generally be avoided in heart failure as they have a negative inotropic action and cause peripheral oedema; only amlodipine has been shown to have no adverse effect on survival [248]. Trimetazine, ranolazine, and nicorandil are available in certain countries but their safety in heart failure is uncertain. Ivabradine is an inhibitor of the I_f current in the sinoatrial node, reduces heart rate, and is an effective antianginal agent. It has been evaluated in patients with coronary heart disease and an EF <40% many of which had heart failure and most of which were treated with an ACE inhibitor/ARB and beta-blocker. In the morBidity-mortality EvAlUaTion of the I_f inhibitor ivabradine in patients with coronary disease and left-ventricULar dysfunction (BEAUTIFUL) trial, ivabradine did not reduce the primary composite endpoint of cardiovascular death, myocardial infarction, or heart failure hospitalisation. However, no safety concerns were identified [271]. Percutaneous and surgical revascularization are also of value in relieving angina in selected patients with heart failure.

The recent Atrial Fibrillation and Congestive Heart Failure (AF-CHF) trial showed that there is no evidence to support a strategy of restoring sinus rhythm over one of controlling the ventricular rate (coupled with thromboembolism prophylaxis) in most patients with heart failure [272]. Exceptions include patients in which new-onset atrial fibrillation has caused myocardial ischaemia, hypotension, or pulmonary oedema and where pharmacological rate control is not rapidly achieved; in these patients, prompt electrical or pharmacological cardioversion may be indicated. In the routine setting, a beta-blocker alone or in combination with digoxin should be used to control the ventricular rate. The patient should be supervised closely after the initiation of these treatments because underlying sinus node dysfunction may raise the risk of bradycardia. Atrioventricular node ablation and pacing may be required to control ventricular rate in resistant cases. There is current interest in catheter ablation to cure atrial fibrillation in patients with heart failure, though this approach remains experimental [273]. There is a strong indication for thromboembolism prophylaxis with warfarin in patients with heart failure and atrial fibrillation.