Summary

Heart failure is a major cause of cardiovascular morbidity and mortality and its incidence is on the increase. The pathophysiology of heart failure is multi-factorial but recent studies suggest that aldosterone plays an important and independent role in its progression. Emerging evidence now suggests that aldosterone exerts renal-independent effects. It binds to its mineralocorticoid receptor to produce direct effects on the myocardium and vasculature, leading to damaging processes such as hypertrophy, necrosis, fibrosis and endothelial dysfunction, factors known to contribute to the pathophysiology of heart failure. Mineralocorticoid receptor antagonists have thus emerged as a new paradigm for the treatment of heart failure. The benefits of these agents on both morbidity and mortality when used in patients with chronic symptomatic heart failure have been demonstrated by recent trials.

Introduction

Heart failure is a major cause of mortality and morbidity, resulting in more than 200,000 hospital admissions per year (5% of all adult medical admissions) in the UK. Annually, between one and five new cases of heart failure per 1000 population (0.15-0.5%) are diagnosed in the UK. The incidence doubles for every decade of life after the age of 45 and reaches 3% in those aged 85-94 years.^[1] The incidence of heart failure is on the rise and this may be attributable to numerous factors such as the ageing population or the improving treatment of acute myocardial infarction, which has led to more patients surviving but consequently being left with impaired myocardial function.

Pathophysiology of Heart Failure

Chronic heart failure does not have a single standard universally accepted definition; this to a large extent illustrates the complex multi-factorial nature of this syndrome. The European Society of Cardiology has defined chronic heart failure as a syndrome where characteristic symptoms are present (typically breathlessness or fatigue, either at rest or during exertion, or ankle swelling) together with objective evidence of cardiac dysfunction at rest.^[2] A clinical response to treatment directed at heart failure alone is not sufficient evidence for diagnosis, although the patient should generally demonstrate some improvement in symptoms and/or signs in response to those treatments where a relatively fast symptomatic improvement would be anticipated (e.g. diuretic administration).

A seminal event such as myocardial infarction leads to an immediate depression of contractility, or it can occur more gradually for example in the context of valvular heart disease, hypertension or dilated cardiomyopathy. The ventricular function Starling curve is depressed, resulting in a state whereby adequate perfusion can only be maintained with an increase in filling pressure and hence elevated end diastolic pressure.

In the setting of decreased left ventricular function, adaptive mechanisms come into play. One of these, namely the neurohormonal system, will be discussed in detail, with particular emphasis on the renin-angiotensin-aldosterone system (RAAS).

Neurohormonal Adaptive Responses

During the acute stage of heart failure there is activation of the neurohormonal system (Figure 1). The first response in the face of a failing heart is to activate the sympathetic nervous system, which results in both positive inotropic and chronotropic effects, resulting in an increase in cardiac output. Constriction of the venous system further elevates cardiac output. This is associated with a decrease in the flow of blood to the kidneys, and this stimulates activation of the RAAS resulting in an increase in blood volume. This increased blood volume aids maintenance of cardiac output by increasing the filling of the heart (preload). Additionally sympathetic activation in itself is capable of activating the RAAS.

Figure 1. 

Mechanisms by which heart failure leads to the activation of neurohormonal vasoconstriction responses and renal sodium and water retention [Adapted from Schrier and Abraham^[24] with permission.]

In evolutionary terms such a mechanism would cater for survival in the natural environment, for example in situations when there is sudden blood or fluid loss stemming from bleeding, diarrhoea or vomiting. Thus, activation of these neurohormonal pathways could facilitate haemodynamic stability for a time in the acute setting. However, in the longer term, these mechanisms are detrimental, and lead to a progressive decline in cardiac function, resulting in a state in which the heart is unable to generate sufficient cardiac output to maintain adequate tissue perfusion, referred to as 'decompensated' heart failure. It appears, therefore, that the acute compensatory mechanisms play a paradoxical maladaptive and pathological role in the progression of heart failure in the long term (chronic stage). The activation of the sympathetic system results in the constriction of arterioles, which causes a resultant increase in blood pressure. This increase in afterload forces the heart to work harder and thus consume a greater amount of oxygen, an effect that results in deterioration of cardiac function over time. The activation of the RAAS results in an increase in blood volume as well as vasoconstriction, thus again leading to increased blood pressure, oedema and an increased cardiac workload.

Recently, it has become increasingly apparent that the natriuretic peptides play an important modulatory role in the pathophysiology of heart failure which, by contrast to the sympathetic and RAAS activation described above, is beneficial.^[3,4] B-type natriuretic peptide (BNP) is produced by the ventricles in response to pressure and volume load, and similarly atrial natriuretic peptide (ANP) is synthesised in response to atrial stretch. Both peptides reduce the preload and afterload in normal and failing hearts. They reduce blood volume over the short term by sequestering plasma and over the longer term by promoting renal salt and water excretion and by antagonising the RAAS at many levels. They also antagonise the cardiac hypertrophic action of angiotensin II. Endogenous BNP levels are significantly elevated in patients with acute heart failure, and this may contribute to volume contraction and vasorelaxation, as well as causing an improvement in renal haemodynamics and tubular function, thereby opposing the effects of high sympathetic and RAAS tone. Infusion of nesiritide, a recombinant form of endogenous human BNP, increases circulating BNP levels by several-fold, and clinical trials in patients with acute decompensated heart failure have shown that nesiritide rapidly reduces clinical symptoms and improves mortality.

The RAAS in Heart Failure

Heart failure generates a pathophysiological state in which the internal organs compete for the reduced systemic blood flow. Such competition is heightened by any extra demand for supply placed on the already stretched circulation. This is exemplified by the situation that arises during exercise, when vasodilatation of skeletal muscle arterioles diverts blood flow away from the kidneys, which in the normal healthy adult accounts for 25% of cardiac output.^[5]

The RAAS plays a key role in the maintenance of salt and water homeostasis by the kidney. Renin is released by the juxtaglomerular cells lining afferent renal arterioles and neighbouring macular densa cells of the distal tubule. Renin catalyses angiotensin I production from angiotensinogen (produced in the liver) via the cleavage of four amino acids. The biologically inert decapeptide angiotensin I is converted into the active product angiotensin II by the action of angiotensin converting enzyme (ACE), located on the plasma membrane of endothelial cells. Additionally, it is now clear that angiotensin II can also be produced through other (non-ACE-mediated) mechanisms.

Angiotensin II has many important physiological actions that are essential for the maintenance of circulatory homeostasis. It binds to both angiotensin II type 1 (AT₁) and angiotensin II type 2 (AT₂) receptors, the AT₁-mediated actions being predominant. It promotes constriction of arterioles within the circulation and encourages the absorption of sodium in the proximal segments of the nephron. A further major effect of angiotensin II is stimulation of aldosterone production from the adrenal cortex, which promotes the distal segment of the nephron to absorb sodium.

The secretion of renin is inhibited when sodium and water are at normal levels. Conversely, renin is released when there is a loss of sodium and water.^[6] Renin release is also stimulated by changes in renal perfusion, renin release increasing with decreased renal perfusion. Angiotensin II also has a key role to play in regulating the RAAS, by inhibiting renin release, thereby in effect providing the RAAS with negative feedback control.

However, when elevations of angiotensin II and aldosterone occur in the absence of sodium/water loss, sustained activation of the RAAS is inappropriate and pathological. This is the case in heart failure, where the potent activation of the RAAS overwhelms the ability of natriuretic peptides released by the distended heart (due to hypervolaemia) to maintain euvolaemia and compensation.^[6] However, the pathological actions of RAAS activation do not stop there, and many other factors leading to heart failure progression have been attributed to both angiotensin II and aldosterone action, independent of their effects on circulating volume, and these will be discussed later.

Conventional Therapy of Heart Failure

For the reasons outlined above, it is clear that overstimulation of RAAS contributes to the pathological state seen in heart failure. Therefore, many agents have been devised to antagonise the RAAS and hence ameliorate its damaging actions. Some of these 'classic' heart failure therapy agents will be discussed below, before a more detailed discussion on the emerging role of aldosterone.

Angiotensin Converting-Enzyme Inhibitors and Angiotensin Receptor Blockers

Angiotensin converting enzyme inhibitors (ACE-I) constitute a class of medication that have conclusively been shown to prolong life when given to heart failure patients. These are potent vasodilators and act through decreasing the production of angiotensin II. However, the Survival and Ventricular Enlargement trial showed that the benefits on mortality associated with ACE-I declines over 1 year post-myocardial infarction in heart failure patients, to levels comparable with placebo.^[7] Thus it is clear that whilst these agents still form the mainstay of heart failure therapy, there is still room for improvement in terms of prolonging life and improving symptoms.

A major drawback of ACE-I is that they are responsible for the development of chronic cough in a minority yet significant group of users. This is believed to be the result of inhibition of bradykinin metabolism. This has led to the advent of angiotensin II receptor blocker (ARB) drugs. These agents inhibit the action of angiotensin II via competitive blockade of the AT₁ receptors, and thus do not interfere with the metabolism of bradykinin. Hence these agents do not produce the cough associated with ACE-I use. The Candesartan in Heart failure Assessment of Reduction in Mortality and morbidity (CHARM) study showed that angiotensin receptor blockade was an effective alternative to ACE inhibition in the treatment of heart failure.^[8]

The controversial role of combined therapy with ACE-I and ARBs is illustrated by two recent clinical trials. The addition of valsartan to patients with chronic heart failure with background treatment of ACE-I in the ValHeFT Trial^[9] was shown to significantly reduce morbidity (combined end point of total mortality or heart failure hospitalisation) without reducing total mortality alone. Post hoc observation showed an increased morbidity and mortality in the subgroup receiving triple therapy with valsartan, an ACE-I, and a β-blocker; the clinical significance of this remains uncertain, and in particular, as this was a post hoc finding, it is not clear if this was a statistical 'quirk'. The CHARM-added trial^[10] showed a significant (15%) reduction in the combined end point of cardiovascular death or heart failure hospitalisation with the addition of candesartan therapy to patients with heart failure. There was also no increase in morbidity or mortality when candesartan was given to patients already on ACE-I and β-blocker therapy. Thus, there is an increasing acceptance that combined ACE inhibition and angiotensin blockade may be more effective than ACE inhibition alone; however, what remains the subject of debate is whether this may be explained by submaximal ACE inhibition or angiotensin blockade occurring when either drug type is used alone, at the doses used in these clinical trials.

Renin inhibitors, soon to be released for treatment of hypertension, are currently undergoing trials for heart failure, but as yet no data are available in this indication.

β-Blockers

The use of β-blockers in heart failure has a chequered history, as they have not always been thought to be beneficial in the heart failure setting. However, in recent years, increasing numbers of studies have shown that they improve haemodynamics and decrease mortality in patients with heart failure (^[11,12]), and this is likely to be caused by inhibition of the deleterious effects of long-term sympathetic activation.

Aldosterone and Heart Failure

Aldosterone

Simpson and Tait were the first to discover aldosterone (or electrocortine as they originally named it), elucidate its adrenal origin and deduce its steroidal structure over 50 years ago.^[13] Subsequent discovery of its role in reabsorption of sodium and concomitant excretion of potassium by epithelial cells in the kidneys, intestine, and sweat and salivary glands led to its designation as a mineralocorticoid. The essential physiological role that aldosterone plays in maintaining sodium and water homeostasis during periods of dietary sodium deprivation is now clear. It is secreted in response to numerous stimuli, particularly in response to decreased renal perfusion associated with haemodynamic compromise. Upon binding to intracellular mineralocorticoid receptors located in the distal nephron, aldosterone translocates to the nucleus, where it exerts various effects, mainly promoting gene transcription. The resulting protein synthesis ultimately mediates the cellular responses; namely increase in sodium transport through activation and de novo synthesis of epithelial sodium channels. The subsequent increased concentrations of intracellular sodium activate Na⁺/K⁺ ATPase molecules mainly at the basolateral membrane which extrude sodium back into the bloodstream.^[14]

Although the contribution of aldosterone to retention of sodium in heart failure patients was established long ago, its role was perceived to be diminished in patients treated with ACE-I and ARB. In recent years, however, it has become evident that aldosterone can exert deleterious effects in heart failure, even in patients on full-dose ACE-I and/or ARB therapy.

Deleterious Actions of Aldosterone

The prognosis of patients with heart failure is believed to be worsened by the neurohormonal actions of aldosterone in a variety of ways. The classic mechanism is through inappropriate retention of sodium and water, thus causing hypervolaemia. Additionally, aldosterone also gives rise to a range of other deleterious actions, and raised aldosterone levels are associated with increased mortality.^[15] Aldosterone modulates parasympathetic tone resulting in reduced heart-rate variability, which in heart failure has been shown to be associated with increased mortality.^[16] The compliance of the aorta and its major branches has been shown to be inversely related to aldosterone levels in patients with heart failure, and this has detrimental effects on myocardial function.^[16] Aldosterone has also been shown to exhibit pro-dysrhythmogenic properties.^[17] Finally, chronically elevated levels of aldosterone lead to myocyte necrosis and macroscopic scarring.^[18] The evidence for this will be expanded upon later. A summary of the renal and cardiovascular actions of aldosterone in heart failure is given in Figure 2.

Figure 2. 

Renal and cardiovascular actions of aldosterone in heart failure [Adapted from Bauersachs^[39] with permission.]

Although it was long considered that angiotensin II was the major RAAS-related mediator of cardiovascular damage,^[19] it is now apparent that aldosterone also has a significant and independent role in the progression of heart failure.^[20] It was previously believed that such actions of aldosterone should be abrogated if the pathway upstream to its production were inhibited through the therapeutic use of ACE-I and/or ARB; and there is evidence that this is indeed the case in the acute setting.^[21,22] However, long-term therapy with such agents has been suggested to result in 'aldosterone escape'.

The 'Aldosterone Escape' Phenomenon

Angiotensin converting enzyme inhibitors and ARB reduce the action of angiotensin II and theoretically should therefore also block production of aldosterone. However, it has been observed that chronic (months) therapy with these drugs is associated with an increase in aldosterone levels; referred to as 'aldosterone escape'. Various mechanisms have been postulated to account for this phenomenon, but the three major mechanisms are as follows:

The standard doses of ACE-I do not fully abrogate angiotensin II-regulated adrenal production of aldosterone, as the use of high doses of ACE-I is limited by the occurrence of hypotension and renal insufficiency.^[5] Angiotensin-independent production of aldosterone also occurs, chiefly in circumstances of hyperkalaemia, and this production appears to be integral to intravascular volume regulation.^[23] It is stimulated by extra-vascular volume depletion and hyponatraemia. Additionally, reduced hepatic clearance of aldosterone occurs in heart failure, resulting in increased levels of the mineralocorticoid, with heart failure patients exhibiting reductions in aldosterone metabolism of 25-50%, and this is largely explained by reduction in hepatic blood flow.^[24] It has also been postulated that extra-adrenal production of aldosterone occurs in numerous tissues, including the myocardium and blood vessels, and this will be discussed later.

It is instructive to note the differences in aldosterone effects between patients with heart failure and those suffering from primary hyperaldosteronism. Normal subjects respond to high doses of mineralocorticoid by initially increasing renal sodium retention so that the volume of extracellular fluid is increased by up to 1.5-2 l. Subsequently, however, there is cessation of retention resulting in the re-establishment of sodium balance and the lack of detectable oedema. This 'escape' from mineralocorticoid-mediated sodium retention explains why oedema is not a characteristic feature of primary hyperaldosteronism.^[24] It is believed that this escape is at least partially dependent on an increased delivery of sodium to the site of action of aldosterone in the collecting ducts.^[25] However, in heart failure, this escape from the sodium-retaining action of aldosterone does not occur; hence sodium continues to be retained in response to aldosterone. Thus commencement of therapy with spironolactone, a mineralocorticoid receptor antagonist (MRA), results in substantial natriuresis.^[26]

The reason for this disparity appears to be as follows. Heart failure patients are believed to have a reduced salt load arriving at the collecting ducts, because α-adrenergic stimulation and angiotensin II, both known to be elevated in such patients, lead to an increase in sodium transport in the proximal tubule. However, no such elevation in α-adrenergic stimulation and angiotensin II occurs in patients with primary hyperaldosteronism. This decreased sodium delivery to the collecting duct has been put forward as a likely important explanation for the persistent aldosterone-mediated sodium retention and the absence of the escape phenomenon in heart failure patients.^[24]

Cardiac Aldosterone Production

Extra-adrenal aldosterone production has been identified by some groups, and one important such site is the myocardium.^[27] In rodents,^[28] aldosterone produced in the myocardium does not contribute to circulating concentrations of aldosterone, therefore serves an autocrine or paracrine function. Aldosterone mRNA in the heart was shown not to be related to plasma levels of angiotensin II, further supporting this hypothesis. Other groups have gone on to document extra-adrenal production of aldosterone mRNA in ventricular fibroblasts.^[29]

The work of Mizuno et al.^[30] in particular has provided direct evidence in humans of the deleterious effects of aldosterone on cardiac structure and function. This group utilised simultaneous blood samples from the aortic root, anterior cardiac vein and coronary sinus in three distinct clinical cohorts of patients, from which the aldosterone concentrations were measured; these results showed that, in those patients with chronic heart failure, there was a significant positive gradient in aldosterone concentration from the aorta to the coronary sinus. This was not observed in subjects without heart failure. This study demonstrated, therefore, that the heart produces aldosterone under conditions of chronic heart failure. This work was supported by the subsequent findings of Yoshimura et al.,^[31] who showed that aldosterone synthase expression is upregulated in the hearts of patients with chronic heart failure when compared with those of patients without heart failure.

The Pathophysiological Role of Aldosterone in Heart Failure

It is clear from the above that there is a good deal of emerging evidence supporting the idea that aldosterone plays a pathophysiological role in heart failure. However, the mechanism by which it does this is uncertain, and the evidence for some of these putative mechanisms will now be explored.

Myocardial Hypertrophy and Necrosis

There is now compelling evidence from animal models and clinical studies that aldosterone exerts harmful remodelling independent of its effects on blood pressure. Hayashi et al.^[32] studied the effects of transcardiac aldosterone extraction on the progression of left ventricular remodelling after first anterior wall myocardial infarction. Measurement of aldosterone, cardiac size and function was performed at baseline and 1 month after initiation of post-myocardial infarction therapy (including ACE-I and β-blockers). A higher concentration of aldosterone was found in the aorta compared with the coronary sinus at both measurements, indicating that aldosterone was being persistently extracted from the coronary circulation after myocardial infarction. Further analysis showed that transcardiac aldosterone extraction at baseline directly correlated with both 1 month left ventricular end diastolic volume index and coronary sinus-measured procollagen III N-terminal peptide (PIIINP), a marker of cardiac collagen turnover. It was also found that coronary sinus PIIINP at 1 month positively correlated with left ventricular end-diastolic volume index, hence implicating collagen turnover as a prerequisite for cardiac remodelling. The authors proposed that, post-myocardial infarction, despite ACE-I therapy, there is persistent cardiac extraction of aldosterone which in turn accounts for stimulation of collagen turnover leading to ventricular remodelling.

Rocha et al.^[33] showed that rats administered angiotensin II along with the nitric oxide synthase inhibitor l-NAME showed myocardial necrosis and fibrinoid necrosis of the coronary vasculature, which were both greatly reduced by therapy with eplerenone (an MRA) without any change in blood pressure, or by adrenalectomy, but recurred after further aldosterone infusion. This cardiomyocyte necrosis has recently been shown not to be secondary to angiotensin II-induced increase in plasminogen activator inhibitor,^[34] nor to be prevented by oral potassium loading.^[35] Thus, there is clear evidence implicating aldosterone as a direct mediator of cardiac tissue injury.

Studies in rats have provided more direct evidence that aldosterone is responsible for cardiac injury. De Angelis et al.^[36] showed that 24 h aldosterone infusion at a dose of 1 mg/kg in normotensive rats stimulated an increase in cardiomyocyte apoptosis, which can often lead to secondary necrosis. Eplerenone was shown to suppress aldosterone infusion-induced myocardial and coronary lesions by Rocha et al..^[37] Necrotic myocytes are known to stimulate replacement fibrosis in the myocardium. This type of cardiac insult thus promotes further cardiac dysfunction, sequentially triggering further neurohormonal activation, hence spiralling into adverse remodelling.

Some groups have suggested that aldosterone is responsible for mediating angiotensin II-induced cardiac injury, thus accounting for the amelioration of cardiomyocyte death, cardiac hypertrophy, inflammation and extracellular matrix deposition with MRA. Numerous mechanisms have been suggested to account for these actions of aldosterone, with induction of myocardial fibrosis being of particular importance. In patients with heart failure without underlying coronary heart disease, treatment with an MRA (spironolactone) significantly decreased left ventricular mass and volume, to a greater extent than that seen in patients receiving only an ACE-I.^[38]

The results of recent work by Bauersachs,^[39] utilising rats with severe left ventricular dysfunction post-myocardial infarction, are summarised in Figure 3. Chronic treatment with the MRA eplerenone ameliorated left ventricular filling pressure and volume, and improved isovolumic relaxation. The surviving left ventricular myocardium was later shown to have reduced collagen type I mRNA expression, collagen deposition and matrix metalloproteinase 13 protein expression.^[40]

Figure 3. 

Attenuation of left ventricular remodelling by selective aldosterone receptor blockade in experimental myocardial infarction. Shown are representative left ventricular sections (upper panel) and left ventricular end-diastolic volume measured in vivo in sham-operated rats (sham) and in rats with extensive myocardial infarction treated either with placebo, eplerenone, trandolapril, or combined eplerenone and trandolapril [Adapted from Bauersachs^[39] with permission.]

Left ventricular remodelling was also found to be beneficially affected by the action of eplerenone, with the effect being attributed to prevention of pathological hypertrophy via reduction of expression of fetal genes such as β-myosin heavy chain and ANP.^[40] In addition, it was shown that combination therapy with MRA and ACE-I was more effective in preventing left ventricular remodelling and molecular alterations than monotherapy with either drug. In humans, this was placed into context by the work of Hayashi et al.^[41] who demonstrated that, in patients with first anterior myocardial infarction, spironolactone therapy for 1 month commencing immediately post-coronary reperfusion improved left ventricular remodelling and function.

Myocardial Fibrosis

Animal models of hypertension and hyperaldosteronism exhibit myocardial fibrosis in the hypertrophied left ventricle as well as in the non-hypertrophied right ventricle.^[42] Additionally, concomitant administration to rats of aldosterone with a high sodium diet produces fibrosis in both left and right atria.^[43] The fibrosis of the ventricles was shown to be prevented by spironolactone, a non-selective aldosterone receptor antagonist, an effect that was not abrogated by the development of hypertension or left ventricular hypertrophy.^[42] The authors therefore suggested that aldosterone mediates a dose-dependent increase in cardiac fibroblast collagen synthesis in the presence of high-salt diet. The essential requirement of the high-salt diet in this model is underscored by the fact that the effect is not observed in the presence of a low-salt diet.

The use of the selective aldosterone receptor blocker eplerenone early post-myocardial infarction was shown by Delyani et al.^[44] to not affect the reparative process of collagen deposition in infarcted areas, whilst at the same time to not worsen the expansion of the infarcted myocardial wall. This group also showed that eplerenone reduced reactive fibrosis in the viable myocardium of rats with moderate myocardial infarction. In a dog model of moderate heart failure, long-term monotherapy with eplerenone decreased interstitial fibrosis and activation of matrix metalloproteinases, concomitantly improving ventricular function.^[45]

The role of aldosterone in promoting myocardial fibrosis has also been shown in humans. Heart failure patients were shown to have increased expression of myocardial aldosterone synthase mRNA that was associated with greater myocardial fibrosis and severity of left ventricular hypertrophy, hence implicating a possible role for locally synthesised aldosterone.^[46] Furthermore, chronic heart failure patients exhibit increased collagen turnover (PIIINP) which is decreased by the addition of spironolactone to standard drug therapy.^[47] In this study, serum concentration of collagen at baseline correlated with risk of death and hospitalisation. The survival benefit among patients receiving spironolactone was associated with a reduction in serum propeptide concentrations. These findings therefore show that morbidity is linked with collagen turnover and fibrosis, and in heart failure patients that spironolactone may attenuate such structural remodelling. There is however some uncertainty as to why myocardial fibrosis is variable among patients with heart failure, unrelated to the severity of the initial insult.^[48]

It still remains to be ascertained exactly how aldosterone produces myocardial fibrosis. It has been suggested that local aldosterone production directly stimulates collagen production by fibroblasts; alternatively it may cause hypertrophic effects with consecutive fibrosis, or possibly reparative fibrosis post-myocardial injury secondary to its pro-inflammatory actions, coronary endothelial dysfunction, vascular damage and platelet activation.^[48] These possibilities remain to be clarified.

Endothelial Dysfunction and Arterial Stiffening

Aldosterone can cause endothelial dysfunction, and this is postulated to involve inhibition of nitric oxide release by aldosterone.^[49] Aldosterone has also been shown to stimulate a disproportionate increase in vascular collagen content and fibrosis.^[50] These effects lead to reduced arterial compliance and impaired baroreceptor activity and, consequently, to impaired autonomic control of cardiac and vascular function in patients with heart failure.^[22]

Animal models have also provided experimental evidence that aldosterone is capable of inducing endothelial dysfunction. Aortic rings and isolated renal artery segments have been shown to exhibit endothelial dysfunction upon exposure to high mineralocorticoid levels, and spironolactone restored normal endothelial function.^[51] Further evidence in support of the ability of aldosterone to cause endothelial dysfunction comes from rabbits fed a pro-atherosclerotic diet; in these animals, endothelial function as well as superoxide formation were improved by spironolactone therapy.^[52] Some work has also been carried out in humans demonstrating the propensity for aldosterone to produce endothelial dysfunction acutely.^[53]

Endothelial dysfunction is itself the end result of a number of pathophysiological mechanisms. Decreased production of nitric oxide is one of the major causes of endothelial dysfunction, and this has been suggested to be one of the pathways by which aldosterone produces endothelial dysfunction.^[54] It has been shown, in a rat model of heart failure, that combination therapy with ACE-I and spironolactone reverses the impaired endothelial-dependent vasodilator response, and this is associated with upregulation of endothelial nitric oxide synthase expression and reduction of both vascular superoxide formation and NADPH oxidase subunit expression.^[55]

Virdis et al.^[56] showed that at least part of the angiotensin II-induced increase in superoxide formation in the vasculature is attributable to aldosterone. By showing that vascular superoxide generation can be normalised by spironolactone therapy, these workers underlined the role of aldosterone in stimulating vascular superoxide formation in heart failure, supporting similar findings reported in atherosclerotic animal models.^[51]

Is Aldosterone Linked to Heart Failure Progression? The Clinical Evidence

Cooperative North Scandinavian Enalapril Survival Study (CONSENSUS)

In the clinical setting, elevated serum levels of aldosterone have been shown to be associated with higher mortality, as shown by the neurohormonal substudy of the CONSENSUS. In this study, patients with a baseline aldosterone concentration above the median at 6 months had significantly greater mortality than those below the median (55% vs. 32%, p < 0.001). Moreover, ACE-I therapy with enalapril conferred no mortality benefit compared with placebo in patients with baseline aldosterone concentrations below the median, whereas in those with above median values of aldosterone the 6-month mortality was markedly reduced by enalapril from 55% to 20% (p < 0.01).^[15]

Randomized Aldactone Evaluation Study (RALES)

This double-blind, placebo-controlled study was conducted in more than 15 countries on five continents encompassing 195 centres.^[57] A cohort of 1663 patients with symptomatic NYHA class IV heart failure and systolic dysfunction (ejection fraction ≤35%) were assigned randomly to receive spironolactone or placebo, in combination with standard care. The study was terminated early by the data monitoring board as the overall risk of death due to progressive heart failure and sudden death from cardiac causes was reduced by approximately 30% among spironolactone-treated patients. An additional benefit was the finding that those on spironolactone had a significant improvement in functional status compared with controls.

The impact of spironolactone was not attributable to lowering systemic blood pressure as no effect was seen on blood pressure; nor could it be explained simply by an increase in diuresis. This trial thus indicated that the beneficial effects of MRA in patients with heart failure are over and above those of ACE-I. As a result of this landmark trial, patients with class IV heart failure now routinely receive spironolactone in conjunction with other therapy.

The trial was criticised for recruiting a study population which included only 10% of patients on β-blocker therapy, as these agents may utilise similar mechanisms for their beneficial effects. β-blockers reduce the sympathetic arm of the neurohormonal response to heart failure, and also reduce the secretion of renin and hence aldosterone production. It remains to be proved, therefore, that spironolactone is of benefit in patients with severe heart failure who are already on β-blocker therapy.

Ten per cent of men in the spironolactone group complained of gynaecomastia or breast pain compared with 1% of men in the placebo group (p < 0.001). As a consequence, a greater proportion of patients in the spironolactone group discontinued treatment (10 vs. 1, p = 0.006). Both are well described side effects caused by the affinity of spironolactone for the androgen and progesterone receptors. Reassuringly, no differences were seen in the occurrence of hyperkalaemia in patients in the treatment (14 patients, 2%) when compared with the placebo (10 patients, 1%) groups. Despite the lack of hyperkalaemia seen in RALES, there has been some clinical concern about this issue. A strong correlation has been observed between increases in spironolactone prescription post-publication of the results from RALES and the rate of hyperkalaemia-induced hospital admissions,^[48] as shown in Figure 4. This has been ascribed largely to the use of higher doses of spironolactone and the treatment of patients outside the inclusion criteria used in the RALES trial. In RALES, the dose of spironolactone was low (25 mg daily), and significant renal dysfunction was excluded; in clinical practice, hyperkalaemia may therefore be an important issue, especially in patients with any degree of renal impairment, and in particular where other potassium-elevating drugs (especially ACE-Is and/or ARBs) are co-prescribed.

Figure 4. 

Apparent effect of the release of the results from the Randomized Aldactone Evaluation Study on spironolactone prescriptions and on hospital admissions for hyperkalaemia. (a) The rate of prescriptions for spironolactone among patients recently hospitalised for heart failure who were receiving angiotensin converting enzyme inhibitors. (b) Rate of hospitalisation for hyperkalaemia [From Ertl^[48] with permission.]

Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS)

The EPHESUS trial^[58] studied the more selective MRA eplerenone, in a different group of heart failure patients. In EPHESUS, patients had acute-onset heart failure post-myocardial infarction, higher baseline left ventricular ejection fraction and greater β-blocker usage. Thus, the EPHESUS patient population may represent a healthier population than that in RALES. EPHESUS was a multi-centre, randomised, double-blind, placebo-controlled study that randomised 6642 patients with documented left ventricular function of <40% on echocardiography to receive either placebo or eplerenone within 3-14 days of an acute myocardial infarction, excluding patients in whom spironolactone was indicated.

The trial showed that eplerenone treatment resulted in a significant 15% relative risk reduction in all-cause mortality compared with placebo over a follow-up period of 24 months (absolute mortality 14.4% vs. 16.7% eplerenone vs. placebo respectively). Eplerenone was also shown to reduce chronic heart failure admissions by 23% and recurrent all cause cardiovascular hospitalisations by 15%.

No gynaecomastia or breast pain was apparent in EPHESUS, as eplerenone is a selective MRA and does not appreciably bind to androgen and progesterone receptors unlike spironolactone. On the other hand, a significant increase in hyperkalaemia (≥6.0 mmol/l) was observed in subjects randomised to eplerenone therapy compared with placebo (5.5% vs. 3.9%, p = 0.002). Fifteen patients (12 in the eplerenone group and three in the placebo group) were hospitalised for serious hyperkalaemia, and one death in the placebo group was attributed to it. Nevertheless, the risk of hypokalaemia was more than twice as high as the risk of serious hyperkalaemia, and eplerenone was shown to significantly reduce this risk (8.4% vs. 13.1%, p < 0.001).

Effects of Eplerenone, Enalapril and Eplerenone/Enalapril (4E)

As left ventricular hypertrophy is an established risk factor for the development of chronic heart failure, Pitt et al. conducted the 4E study on subjects with primary hypertension and left ventricular hypertrophy.^[59] This was a randomised, active controlled, three arm trial carried out over 9 months in a population of 202 patients. The primary end point in the trial was a change in left ventricular mass as assessed by magnetic resonance imaging. Patients were treated with either enalapril 40 mg/day, eplerenone 200 mg/day, or both (enalapril 10 mg/day and eplerenone 200 mg/day), following an initial 2 week washout period.

The results from serial magnetic resonance imaging evaluation in 153 subjects demonstrated a significant reduction over a 9-month period in left ventricular mass in the group treated with eplerenone/enalapril compared with the eplerenone monotherapy group. This could not be attributed to a disparity in blood pressure reduction between the groups, as all three trial arms had equivalent reductions in blood pressure. These results led the authors to propose that combination therapy with an MRA and ACE-I can give rise to a greater reduction in left ventricular mass than ACE-I therapy alone.

Conclusions

Heart failure is a multifactorial disease with multiple aetiologies as well as complex pathophysiological mechanisms of disease progression. Therapy must therefore be aimed at targeting these factors to produce maximal benefit on morbidity and mortality.

As a result of evolutionary progress, humans in common with other mammals have developed ways of coping with life-threatening deprivation of salt and water. One important such coping mechanism is the RAAS. This neurohormonal system is essential for survival in the acute setting of haemodynamic instability, but clearly its role in the setting of chronic heart failure is not advantageous and in fact in most cases is maladaptive. This understanding has led to the therapeutic use of agents (ACE-I, ARB, β-blockers) that block this system, in the context of heart failure.

It is now evident that aldosterone in its own right exerts numerous undesirable long-term effects in heart failure, not only via stimulation of salt and water retention but also via direct effects on the myocardium and vasculature, producing aberrant remodelling, necrosis, hypertrophy and fibrosis - actions known to be detrimental in heart failure. Much research therefore has focused on elucidating the mechanisms by which aldosterone mediates these actions on the cardiovascular system, and much has been learnt, especially with regard to fibrosis and the role of collagen turnover. Additionally, this understanding of the role of aldosterone in heart failure progression has led to an increase in the use of MRA in this setting. Results of trials thus far show an improvement in life expectancy as well as in quality of life, with the use of MRA therapy in patients with heart failure. At present, trial evidence exists for the use of spironolactone in patients with class IV chronic heart failure, and for the use of eplerenone in patients post-acute myocardial infarction with documented left ventricular dysfunction. Whether MRA therapy will be adopted more generally in heart failure remains to be seen, and will depend on the results of future trials.

CLICK HERE for subscription information about this journal.

Dr Albert Ferro, Department of Clinical Pharmacology, GKT School of Medicine, 2.38A New Hunts House, King's College London, Guy's Hospital Campus, London Bridge, London SE1 1UL, UK Tel.: + 44 20 7848 6233 Fax: + 44 20 7848 6220 Email: albert.ferro@kcl.ac.uk