1 What is heart failure?

It is a commonplace in writings about heart failure (HF) that it has become an ‘epidemic’ in Western societies in particular. In truth, the incidence of HF is not rising, but the prevalence is. HF is thus not a true epidemic, which properly is a rise in the age-specific incidence. The major causes for its increasing prevalence are threefold: although the incidence of acute myocardial infarction may be falling, more patients survive acute coronary disease and go on to develop chronic HF; treatment of chronic HF has dramatically improved, and so many more patients survive for much longer; and the population generally is ageing—and HF is a disease of older people.

Although HF is a modern blight, it has been known for thousands of years. There is some suggestion from the Ebers papyrus (dated around 1500 bce) that the ancient Egyptians recognized it (‘When there is inundation of the heart, the saliva is in excess, and therefore the body is weak’), and Hippocrates (460–370 bce) gave a much quoted description of cardiac cachexia: ‘The flesh is consumed and becomes water … the abdomen fills with water; the feet and legs swell; the shoulders, clavicles, chest, and thigh melt away.’¹

It was not until after Harvey described the circulation of the blood that the HF syndrome truly began to be related to the heart, with Richard Lower perhaps giving the first textbook discussion of HF in 1669.² Treatment for HF with venesection, perhaps one of the few instances in which the procedure might be helpful, was formally described in 1696.³ William Withering described the formal use of Digitalis extracts, giving birth to clinical pharmacology,⁴ although cardiac glycosides had undoubtedly been used for hundreds, and perhaps thousands,⁵ of years previously.

The modern era of HF treatment truly began with the discovery of mercurial,⁶ and then thiazide and subsequently loop, diuretics in the late 1950s and early 1960s. Perhaps the most important single trial in HF therapy demonstrating the beneficial effects of angiotensin converting enzyme (ACE) inhibitors was published in 1987.⁷

Neither the epidemiology of a condition not its treatment can properly be understood unless properly defined. The term ‘heart failure’ is usually used freely between clinicians to describe what is wrong with individual patients, yet despite the fact that HF is so very common, it is very difficult to define it satisfactorily (Box 1.1).⁸ Some difficulties arise because of the effects of modern treatment: although it may be reasonable to define acute HF in terms of some haemodynamic variable, the situation becomes very different in chronic treated HF.

Older general definitions of HF centred on haemodynamic changes, and were phrased in terms of inadequacy of cardiac output in response to normal filling pressure of the heart, with the inadequacy of the output thought of in terms of being inadequate to meet the requirements of the metabolizing tissues.⁹ These sorts of definition are of some value in thinking about the pathophysiology of patients being admitted acutely with salt and water retention or pulmonary oedema, but less so in thinking about patients with chronic HF.

Patients with chronic HF, particularly when adequately treated, have normal resting cardiac output and normal left ventricular filling pressure. Their metabolizing tissues are well enough perfused that they are usually asymptomatic at rest: chronic treated HF is a condition of exercise limitation. Even so, for many patients, cardiac output and oxygen consumption go up as normal during modest exercise, only falling below normal towards peak exercise.

Ultimately, HF is a clinical syndrome characterized by a constellation of symptoms and signs, and not a discrete diagnosis. Much epidemiological work has defined HF in terms of those symptoms and signs, but simply defining the syndrome by its symptoms and signs may mistakenly include many patients without cardiac pathology.^10,11 The situation is even worse if simply considering treatment for HF as being adequate to define the presence of HF in epidemiological studies: such an approach may lead to gross over-diagnosis.¹²

The key combination is to recognize that HF is accompanied by a recognizable constellation of symptoms and signs, coupled with objective evidence that there is an abnormality of the heart consistent with the diagnosis. This is the line now taken by the European Society of Cardiology,¹³ with the added clause that in doubtful cases, a response to treatment directed at HF sustains the diagnosis.

Such an approach is pragmatic, at least, and is rooted in clinical life. Some problems do arise in borderline cases. In an elderly patient, for example, breathlessness is a very common symptom, and peripheral oedema is a very common physical sign: if an echocardiogram shows left ventricular hypertrophy, can the patient truly be defined as having HF? The missing part of the equation is some objective test, independent of cardiac imaging, which allows the clinician to be sure that the cardiac abnormality is the cause of the patient’s symptoms.

The natriuretic peptides may offer at least a partial solution in this regard. These hormones and their derivatives, released from the heart in response to cardiac stretch, should be raised in patients with HF. The next step in defining HF is likely to include natriuretic peptide level. In an untreated patient, if the natriuretic peptide level is normal, then there will be an alternative cause for the patient’s symptoms.

Why does HF present clinically as it does? This seems an odd question, as clinicians are so familiar with the clinical syndrome, but it is not immediately obvious why a patient whose heart function declines should start to retain fluid and develop neurohormonal activation. Harris emphasized the importance of blood pressure in the evolution of terrestrial animals.¹⁴ In order to perfuse a large body unsupported by water; in order to allow rapid movement of that body; and in order to excrete the high level of waste products incurred by having a large, rapidly moving body, high blood pressure is fundamental—certainly compared with the blood pressure needed to service a fish.

An array of very powerful defensive mechanisms has evolved to maintain that high blood pressure at more-or-less all costs. The responses of the body to a fall in blood pressure induced by, say, haemorrhage, are very similar to those induced by HF. Vasoconstriction, salt and water retention and neurohormonal activation are the responses to both conditions. The clinical pattern of HF can thus be viewed as a consequence of mammalian evolution and the vital importance of maintaining high blood pressure.

Older textbooks of cardiology abound in paired descriptions of HF: forward versus backward; right versus left; high versus low output; systolic versus diastolic; acute versus chronic. Some of the terms are now largely redundant but are worth considering in brief.

Forward HF refers to the notion that there is primarily failure of forward pump function leading to inadequate perfusion of peripheral tissues, particularly skeletal muscle, causing fatigue and exercise intolerance. Conversely, backward failure is thought to arise from the need to maintain cardiac output via increased left ventricular filling pressure, which results in left atrial hypertension and thus lung congestion and breathlessness.

Right HF suggest that the HF is predominantly due to failure of the right ventricle with consequent systemic venous congestion and ‘backward’ failure, whereas left HF leads to pulmonary venous hypertension and pulmonary oedema together with the consequences of reduced pump function. Such a classification is not very helpful: the commonest cause of right HF is left HF, and the two rarely occur as separate entities.

High-output cardiac failure is a rarity caused by excessive vasodilation together with salt and water retention; it should not be thought of as being primarily a cardiac condition. It is more correctly thought of as circulatory failure. Diastolic versus systolic HF remains a controversial distinction: some investigators report that up to half of patients with HF have impaired ventricular relaxation as their primary pathophysiological problem. In consequence, there is decreased stroke volume and the syndrome of HF. The implication is that had the heart been able to fill more completely, then there would be no HF.

The distinction between acute and chronic HF is clinically helpful, as long as the terms are understood correctly. The word ‘acute’ is often taken, wrongly, to mean ‘severe’, and should be used to mean ‘presenting suddenly’. In very broad-brush terms, acute HF refers to patients presenting as emergencies to hospital, usually with either pulmonary oedema or with fluid retention. Such patients are often presenting for the first time, but may be patients having an exacerbation of their chronic, previously stable, HF. They have acutely abnormal haemodynamics.

In contrast, most patients with chronic HF have been treated medically and will usually have few if any symptoms or signs at rest. The term ‘congestive’ HF, often used to describe patients in this condition (particularly in North America), is inappropriate: patients with treated chronic HF should not be congested.¹⁵

The prognosis of both acute and chronic HF is bleak, although improved dramatically by modern therapy, with an average life expectancy from diagnosis of around 3 years (depending on the population studied),¹⁶ and a prognosis worse than for many kinds of cancer.¹⁷ Such statistics disguise the fact that for the individual patient, the course of HF can be highly variable, and is much less predictable than the course of other malignant diseases (Fig. 1.1).

The initial presentation of HF is usually acute. The commonest cause of HF is coronary heart disease, and so an acute myocardial infarction is a common initial precipitant. With treatment,

a number of outcomes is then possible: the patient might return to normal with impaired left ventricular function; the patient might reach a plateau of impaired function; or the patient might decline relentlessly toward death or transplantation.

Following an initial event and recovery or stabilization, a patient may continue unchanged for several years, or may have repeated episodes of decompensation of chronic HF. Each time, it is less likely that there will be complete recovery of the myocardium, and progressively left ventricular function worsens in a stuttering, stepwise course. As a general rule, such a trajectory often follows the pattern of a flat stone skimming across water: decompensation episodes become longer and the intervals between episodes shorten.

For some patients, the decline in left ventricular function is more gradual than punctuated: in this scenario, a patient may enter a ‘vicious cycle’ of decline (see below). An alarming feature of HF is that at any time in its clinical course, patients are at risk of sudden death.

A less common way for HF to present is with a less abrupt onset and gradually progressive symptoms of breathlessness, fatigue and peripheral oedema. The typical patient presenting this way may have had a remote myocardial infarct or have underlying valvular heart disease or dilated cardiomyopathy. Such a patient will usually present through primary care, and the diagnosis can be delayed in consequence—the range of causes of breathlessness is very broad.

Occasional patients appear completely to recover from an episode of HF. Such a recovery may happen in patients with a discrete episode of illness, such as acute myocarditis or postpartum cardiomyopathy. Some patients with dilated cardiomyopathy may apparently return to having normal left ventricular systolic function with medical therapy, and it can be difficult to judge in such circumstances whether medication should stop or continue indefinitely.¹⁸

Much of the thinking about the clinical course of HF has focused on potential vicious circles of decline (better thought of as spirals—the starting point is not regained). With all of the potential spirals, an abnormality induced by HF results in further deterioration in heart function, thereby worsening the HF. These models are helpful in thinking about the pathophysiology of HF, and in suggesting avenues for therapeutic development.

The haemodynamic model of HF decline is the traditional way of looking at the pathophysiology of HF (Fig. 1.2). Initial damage to the heart is detected by body systems (particularly via a fall in blood pressure and in renal perfusion) and causes consequent haemodynamic changes to maintain tissue perfusion. Salt and water retention help to maintain output via the Frank–Starling relation by increasing preload; and vasoconstriction maintains blood pressure, but at a cost of increasing afterload. The increases in preload and afterload, however, exacerbate the heart’s problems, leading to further decline.

Treatments based on the haemodynamic understanding of HF have not proved very successful: positive inotropic drugs have almost uniformly proved unsuccessful, and abrupt changes in haemodynamics (as with vasodilators or even heart transplantation) do not lead to immediate improvements in exercise function.

The neurohormonal model¹⁹ has been particularly fruitful in guiding new treatments for HF. Note that the effectors in this model are the sympathetic nervous system and the renin–angiotensin system. These hormones have much more widespread effects than just their haemodynamic actions, causing direct harm to the heart, for example, by inducing programmed cell death and fibrosis. Thus neurohormonal activation leads to worsening HF.

As a guide to therapeutic advance, the neurohormonal model has been particularly helpful, underlying the development of modern therapy with ACE inhibitors, β-blockers, and aldosterone antagonists.

The peripheral model (Fig. 1.3)²⁰ draws attention to the changes that happen in the periphery as a consequence of HF, particularly to skeletal muscle. Perhaps in part due to poor perfusion, perhaps due to lack of fitness, and perhaps due to neurohormonal and cytokine activation, a skeletal myopathy develops. The myopathy is a major cause of symptoms, particularly fatigue and breathlessness, but also causes sympathetic activation, leading to further damage to the heart. The peripheral model suggests that intervention to preserve skeletal muscle function or even reverse the myopathy may be helpful in managing HF.

A problem in thinking about the pathophysiology of decline in terms of vicious cycles or spirals is the implication that there is a continuing rapid downhill trajectory as HF inexorably declines. Untreated HF may behave in this way, but treated HF is typically much more stable—a punctuated equilibrium—presumably as a consequence of treatment.