INTRODUCTION
The critical role of neurohormonal factors in regulating circulation and volume status has been known for over a century. The importance of neurohormonal activation in heart failure pathophysiology was identified in the middle and latter part of the 20th century and formed the basis for the neurohormonal model of heart failure. Major contributions to this understanding came from observational data in large, well-characterized populations, often participating in studies of systolic heart failure treatments. In these studies, plasma levels of vasoactive hormones involved in regulation of circulation and renal function—such as vasopressin, renin, angiotensin II (A-II), aldosterone, the cardiac natriuretic peptides, and norepinephrine (NE)’were not only shown to be significantly elevated compared with normal controls, but also clearly provided prognostic information. These seminal observations have since been confirmed by multiple studies in a variety of settings. Early attention was focused on the pathophysiological roles of the sympathetic nervous system (SNS), the renin-angiotensin system (RAS), and aldosterone. A large body of evidence, much of it emanating from studies using targeted blockade of neurohormones, attests to the pivotal role of these systems in established heart failure. The role of more recently discovered systems such as those involving cardiac natriuretic peptides, endothelin (ET), and adrenomedullin (AM) has now become a major focus of study. The discovery of these and additional neurohormonal systems and clarification of their roles in heart failure has opened up exciting new opportunities to better define the pathophysiology of heart failure and to develop new therapies.
The importance of the neurohormonal model of heart failure cannot be understated when considering the recent advances in heart failure treatment. Awareness that activation of the RAS, aldosterone, and the SNS leads to vasoconstriction, sodium retention, and adverse effects on cardiac and vascular remodeling, thereby contributing to morbidity and mortality in systolic heart failure, has led to the development of successful therapies through blockade of these systems. , Whether blockade of additional neurohormonal systems can be tolerated and offers additional benefit is unclear but remains the focus of several current studies. Likewise, it remains to be seen whether administration of vasodilator hormones such as the natriuretic peptides, urocortin, and AM or synthetic compounds acting as agonists on their specific receptors produce beneficial effects on cardiovascular and renal functions. In the case of B-type natriuretic peptide (BNP), this has led to a potential therapeutic application, an option that might also be possible for other peptides.
Recent interest has focused on the potential role of neurohormones as markers of cardiac dysfunction and heart failure. In particular, BNP has been validated as a diagnostic marker for heart failure, and its role in monitoring and guiding therapy is being studied actively.
A significant limitation of the current understanding of neurohormonal factors in heart failure is that there is relatively scant data describing the pattern and role of neurohormonal activation for patients in whom left ventricular ejection fraction (LVEF) is preserved. The available neurohormonal model of heart failure is based largely on observations from patients with systolic heart failure or in whom left ventricular (LV) function was not well described. Based on the known differences in the epidemiology of heart failure with preserved versus impaired systolic function, it is likely that neurohormonal profiles will also differ. Available data suggest that this is indeed the case. A greater understanding of neurohormonal profiles can potentially lead to new and more effective therapy for heart failure in the setting of preserved LVEF.
PATHOPHYSIOLOGY
Neurohormonal Regulation of the Circulation: Normal Versus Heart Failure
Under normal conditions, tightly counterbalanced neurohormonal systems provide dynamic regulation of the circulation and of volume status. These finely tuned systems respond with various time frames to perturbations in body posture and volume status—for example, to maintain optimal perfusion of vital organs. On the one hand, arginine vasopressin (AVP), the SNS, and the RAS respond rapidly to stimuli such as falls in arterial pressure, reduced delivery of chloride and sodium to the macular densa, and altered plasma osmolality to maintain homeostasis by increasing vascular tone and inducing renal retention of sodium and water. In contrast, increased cardiac or vascular wall stress stimulates secretion of natriuretic peptides, AM, urocortin, and other endothelium-derived peptides that produce vasodilatation and promote diuresis. Within this complex counterpoised system, there are multiple levels of feedback and control. Vasoconstrictor systems can respond rapidly through near instantaneous changes in sympathetic outflow or rather more gradually through the renin-angiotensin pathway and aldosterone. Equally, release of stored atrial natriuretic peptide (ANP) from granules in the atria allows rapid vasodilator and natriuretic responses, whereas more gradual effects occur through changes in constitutive BNP synthesis. Multiple feedback interactions between vasodilator and vasoconstrictor systems at several levels modulate the overall balance. For example, A-II attenuates the natriuretic effect of BNP while simultaneously contributing to activation of ANP and BNP synthesis through increased wall stress in cardiac myocytes. In contrast, the natriuretic peptides inhibit renin release and the aldosterone-stimulating actions of A-II.
In established heart failure, the normal balance of neurohormonal regulation is disturbed ( Fig. 27-1 ). Marked activation of both vasodilator and vasoconstrictor systems is seen. A key mechanism in this activation appears to be perceived end-organ hypoperfusion and a reduced ability of the heart to maintain adequate arterial filling and cardiac output. Such activation can be detected as an increase in circulating plasma neurohormone levels, but it occurs first at the tissue level through upregulated gene expression in response to various stimuli, including cardiac injury, reduced organ or tissue perfusion, and fluid overload. Neurohormonal activation may precede the onset of frank clinical heart failure and is a feature of asymptomatic LV dysfunction. A defining characteristic of the neurohormonal response to cardiac injury or heart failure is the predominance of vasoconstrictor over vasodilator systems. This loss of counterpoise leads to attenuation of the beneficial effects of vasodilator systems and contributes to or mediates many of the major adverse pathophysiological changes in heart failure, including progressive renal dysfunction and LV remodeling.
Neurohormones and the Transition from Hypertrophy to Heart Failure
Adverse cardiac remodeling characterized by myocyte hypertrophy and interstitial or perimysial fibrosis are hallmarks of heart failure with preserved ejection fraction (EF). While hemodynamic loading such as in hypertension or valvular disease is a critical stimulus for adverse remodeling, neurohormonal factors appear to be key mediators of cardiac and vascular remodeling processes that lead to heart failure. Increased myocardial levels of A-II, ET, and aldosterone have been demonstrated in models of hypertensive heart disease. Each has been linked to the development of fibrosis, cardiomyocyte hypertrophy, ventricular remodeling, and progression to the heart failure phenotype. Increased levels of these and other hormones, including NE, can be directly toxic to the myocardium. Experimentally, blockade of these hormones attenuates the development of hypertrophy and fibrosis while delaying the onset of heart failure. In contrast, other hormones, such as AM and the cardiac natriuretic peptides, are also activated in hypertension and heart failure but appear to attenuate adverse remodeling.
Neurohormonal Factors Contributing to Heart Failure Pathophysiology
Sympathetic Nervous System
The SNS is a crucial regulator of arterial pressure and vital organ perfusion during the myriad of daily activities such as exercise, eating, and changing body posture. In heart failure, the SNS is activated at an early stage in response to reduced carotid and aortic baroreceptor sensitivity and altered arterial compliance. Changes in SNS activity in heart failure produce a dramatic reduction in regional blood flow to the skin, gut, and kidney while maintaining vital flow to coronary, cerebral, and skeletal muscle circulations. The SNS also contributes to the pathophysiology of heart failure through its stimulation of renin release. Increased sympathetic activation in heart failure can be documented by elevated plasma or urinary catecholamines; by increased NE spillover from the heart, kidney, brain, and skeletal muscle; or by increased common peroneal nerve sympathetic nerve traffic from microneurographic recordings. However it is documented, this activation correlates with the severity of LV dysfunction and predicts survival. While increased SNS activity in established systolic heart failure acts to maintain blood flow to vital organs, long-term and intense overactivation can have harmful effects through direct toxicity to myocardium, an adverse increase in LV afterloading (as a consequence of peripheral vasoconstriction), and LV hypertrophy (LVH) and subsequent dilatation.
The pathophysiological importance of SNS activation and the benefit of β-adrenergic receptor blockade in systolic heart failure is clearly established. SNS activation is generally considered to be also important in primary (essential) hypertension and reflects increased sympathetic neuronal firing and decreased NE uptake. SNS activation is further augmented in hypertensive patients with heart failure, possibly through impaired baroreceptor function. There are, however, limited data regarding the degree of SNS activation in heart failure with preserved LVEF. Nevertheless, plasma NE levels do appear to rise in diastolic heart failure, but not to the extent seen in systolic heart failure. In animal models, the SNS appears to have a significant role in the development of LVH, LV remodeling, and heart failure due to pressure overload. Dopamine beta-hydroxylase knockout mice do not develop LVH during aortic banding, suggesting an important requirement for the SNS in activation of signaling pathways and the development of hypertrophy and heart failure, at least in this model.
Modulation of sympathetic activity may be one important mechanism of effective antihypertensive therapy and prevention of heart failure. Therapy with angiotensin-converting-enzyme (ACE) inhibitors and angiotensin receptor blockers, both of which reverse LVH and reduce heart failure events in high-risk or hypertensive subjects, also suppresses cardiac efferent sympathetic activity. The role of β-adrenergic receptor blockade as first-line therapy for hypertension has, however, been questioned recently. First-line treatment with the alpha-adrenergic blocker doxazosin was reportedly associated with increased heart failure events compared with treatment with chlorthalidone, lisinopril, or amlodipine, although the accuracy of the diagnosis of heart failure has been questioned. Whereas the therapeutic efficacy of selected β-adrenergic receptor blockers in patients with all grades of systolic heart failure is not in question, the place of these drugs in the treatment of established diastolic heart failure remains unclear. Treatment with carvedilol has been shown to have beneficial effects on LV diastolic function in the setting of heart failure with preserved EF, but the effect on clinical outcomes is also uncertain.
Renin-Angiotensin System
Secretion of renin from renal juxtaglomerular cells increases in response to many stimuli, including increased SNS activity, low systemic and renal arterial perfusion pressure (renal baroreceptor mechanism), and reduced delivery of chloride and sodium to the distal renal tubule (macula densa mechanism). Vasopressin, prostaglandins, nitric oxide levels, the natriuretic peptides, AM, and A-II—by negative feedback—all modulate renin secretion. Circulating renin acts on angiotensinogen to produce angiotensin I, which is converted to A-II by ACE. A-II generated through non-ACE pathways, including tissue and serum proteases (chymases), may contribute to increased tissue and circulating A-II and return of plasma aldosterone to baseline, pre-ACE inhibitor treatment levels.
The major pathophysiological actions of A-II are mediated via A-II type 1 receptors (AT 1 R). A potent vasoconstrictor, A-II also stimulates aldosterone secretion and AVP release and thirst, augments SNS activity, and antagonizes the actions of the natriuretic peptides. A-II has critical effects on renal function and, experimentally, stimulates hypertrophy of vascular smooth muscle cells (VSMCs) and cardiomyocytes, thereby contributing to vascular as well as cardiac remodeling. The pivotal role of the RAS in systolic heart failure and the beneficial effect of either ACE inhibition or AT 1 R blockade is well established.
A large body of basic science and clinical research implicates the RAS in the pathophysiology of heart failure with normal systolic function. , In animal models and cell culture studies, A-II has been shown to stimulate stretch-induced cardiomyocyte hypertrophy and promote fibrosis through increased collagen deposition and impaired metalloproteinase function. A-II also augments diastolic wall stress during volume overload, leading to increased expression of fetal genes associated with remodeling. Furthermore, A-II stimulates increases in intra-cellular calcium levels, which may contribute to abnormal diastolic function and wall stress. A-II also increases activity of nicotinamide adenine dinucleotide phosphate oxidase (NADPH), causing superoxide production that may trigger vascular nitric oxide synthase uncoupling, leading to impaired nitric oxide signaling, endothelial dysfunction, and vascular remodeling.
Evidence from animal models points to a contributory role for A-II in hypertensive heart failure. Dahl salt-sensitive rats fed a high salt diet developed hypertension and subsequent hypertensive cardiomyopathy and heart failure by age 19 weeks. In this model, expression of A-II mRNA increases more than fourfold at the onset of heart failure. Administration of low-dose AT 1 R blocker at subvasodepressor levels prevents the onset of heart failure in this model, suggesting an important effect of A-II that is independent of arterial pressure. In the same model, AT 1 R blockade expression was upregulated despite increased A-II expression, suggesting abnormal receptor/agonist balance. Further evidence for A-II causing hypertrophy and subsequent heart failure comes from a transgenic mouse model of localized excess angiotensinogen production. In this model, excess localized cardiomyocyte A-II levels are associated with myocyte and ventricular hypertrophy even in the absence of increased arterial pressure. Hypertrophy is reversed with effective blockade of A-II production by ACE inhibition or by AT 1 R blockade at hemodynamically neutral doses.
Studies of hypertension in humans indicate a pivotal role for the RAS. In particular, treatment of hypertension with ACE inhibitors or AT 1 R blockers is associated with greater reductions in LVH than are seen with beta blockers and calcium channel blockers, achieving similar blood pressure lowering.
In LVH secondary to aortic stenosis, intracoronary administration of the ACE inhibitor enalaprilat, at a level that did not alter measured ACE activity or plasma levels of renin and ANP, produced decreases in LV end diastolic pressure accompanied by improvements in diastolic distensibility and isovolumic relaxation, suggesting that the actions of intracardiac A-II contribute to abnormal diastolic wall stress and filling patterns in LVH. Treatment with AT 1 R blockade or ACE inhibition has been shown to reduce fibrosis and indices of cardiac stiffness and diastolic dysfunction in subjects with hypertension. Importantly, ACE inhibition or AT 1 R blockade in high-risk subjects or after myocardial infarction has been shown to reduce cardiovascular events, including heart failure. The largest randomized treatment study in heart failure with preserved systolic function demonstrated that AT 1 R blockade with candesartan reduced heart failure hospitalization.
Aldosterone
Aldosterone secretion from the zona glomerulosa of the adrenal cortex is stimulated by a range of secretagogues, of which A-II is the most potent, although potassium and adrenocorticotropic hormone (ACTH) (which augment aldosterone production) and the cardiac natriuretic peptides (which are inhibitory) may be important in heart failure. ACE inhibitors may initially reduce aldosterone formation, but levels often rise later, possibly reflecting incomplete ACE inhibition, non-ACE-generated A-II, and the action of other aldosterone secretagogues.
Aldosterone has an important role in established systolic heart failure. Early studies in patients with heart failure demonstrated that aldosterone levels were often (though not invariably) elevated; but perhaps more importantly, the kidney failed to “escape” from the antinatriuretic actions of aldosterone. Furthermore, the kidney in heart failure is more sensitive than normal to the sodium-retaining action of aldosterone. There is evidence for an equally important role for aldosterone in mediating the cardiac and vascular remodeling that occurs as hypertension progresses to heart failure. Local synthesis of aldosterone and expression of mineralocorticoid receptors (MCRs) have been demonstrated within the cardiac interstitium and cardiomyocytes, although there is dispute regarding the capability of the heart to secrete aldosterone. Whether of adrenal origin or secreted within the heart, aldosterone appears to increase collagen deposition and fibrosis and cause hypertrophy of cardiomyocytes. Recent cohort studies demonstrate correlations between plasma aldosterone levels and the degree of concentric LVH, particularly in women. The relationship of aldosterone to LVH may be independent of hemodynamic loading, indicating specific effects on cardiomyocyte hypertrophy and interstitial fibrosis. Aldosterone may also impair arterial compliance, an index of vascular remodeling in hypertensive heart failure.
Specific aldosterone blockade appears to reduce collagen formation and fibrosis. More recently, aldosterone blockade was shown to improve myocardial function in hypertensive patients with heart failure symptoms and echocardiographic evidence of LVH with diastolic dysfunction. Compared with placebo, 6 months of treatment with spironolactone resulted in significant improvements in longitudinal systolic LV strain and strain rate, a decrease in left atrial (LA) area, and improvements in conventional Doppler estimates of LV stiffness and end diastolic pressure. There was also a significant improvement in arterial compliance with spironolactone. These effects support the hypothesis that aldosterone contributes to adverse myocardial and vascular remodeling in hypertensive heart disease.
Arginine Vasopressin
AVP secretion is mediated by multiple stimuli, including high plasma osmolality, low intracardiac and arterial pressure, circulating A-II, ANP, and adrenergic and other neurohormonal factors. AVP increases water uptake in the collecting ducts via vasopressin (V)2 receptors and produces vasoconstriction and impaired cardiac contractility via V1 receptors. Elevated plasma AVP levels are often seen in heart failure even when cardiac filling pressures are high, suggesting that carotid baroreceptor activation may outweigh inhibitory cardiac stretch reflexes. Experimentally, AVP receptor antagonists partially correct hyponatremia in this setting, suggesting that AVP along with the RAS and other systems play a mechanistic role in the hyponatremia of heart failure.
AVP levels are often elevated in asymptomatic LV dysfunction. In addition, levels are elevated in heart failure when LVEF is preserved, suggesting an important pathophysiological role in this context. AVP has recently become a therapeutic target in heart failure. V2 blockade increases free water excretion without altering renal hemodynamics or function and appeared beneficial in initial studies: Larger morbidity and mortality studies are under way. Given the elevation of AVP levels in heart failure with preserved LVEF, V2 blockade could potentially be a target for therapy in this context, but this possibility remains to be tested in this setting.
Endothelin-1
ET-1 is a potent vasoconstrictor peptide secreted mainly from vascular endothelial cells in response to falls in vascular shear stress and stimulation by neurohormones (NE, A-II, and AVP), tissue growth factor-β, and cytokines (such as tumor necrosis factor [TNF]-α). Nitric oxide, the natriuretic peptides, and prostacyclin inhibit ET-1 synthesis. ET-1 is synthesized also in cardiac myocytes, VSMCs, and renal tubular and glomerular mesangial cells.
Pre-pro-ET is cleaved to form “big endothelin” (Big ET), which in turn is converted to vasoactive ET-1 by endothelin-converting enzyme. ET-1 acts via ET-A and ET-B receptors. Binding with ET-A increases mobilization of intracellular calcium to produce vasoconstriction and positive inotropy. ET-A receptors also mediate the hypertrophic action of ET-1 in VSMCs, cardiomyocytes, and glomerular mesangial cells via protein kinase C and mitogen-activated protein kinase.
Plasma ET-1 levels are higher in venous than arterial blood, reflecting vascular secretion. Under normal conditions, ET-1 contributes to basal vascular tone and cardiac function through paracrine and autocrine actions. In heart failure, plasma ET levels rise in proportion to the severity of cardiac dysfunction and are powerful independent predictors of outcome. Myocardial ET-1 levels and ET-A receptor density also increase in heart failure, in which ET-1 initially increases cardiac contractility at the cost of impaired myocardial energy balance. Under experimental conditions, ET-1 induces vasoconstriction, hypertrophy of cardiac myocytes, and cellular injury through direct toxic effects. Chronic activation of tissue and plasma ET-1 in experimental models is associated with cardiac and vascular remodeling and a decline in LV function. In contrast, myocardial ET-1 expression falls with unloading of the left ventricle. A number of studies implicate a synergistic interaction of ET-1 with A-II in the development of LVH and diastolic heart failure. Myocardial expression of A-II occurs prior to ET-1 in models of hypertensive heart failure and appears to stimulate myocardial ET-1 expression. ET-A receptor blockade attenuates the development of LVH and heart failure in hypertension models but is associated with augmented RAS activity. Unsurprisingly then, combined ET receptor blockade and ACE inhibition appears more effective than either agent alone at reversing LVH in heart failure models.
While a short-term ET-A or a combined ET-A/ET-B receptor blockade in systolic heart failure demonstrated beneficial effects on endothelial function and hemodynamics, results of mortality and morbidity studies with dual receptor blockade have been disappointing. The relative merit of selective ET-A blockade and whether earlier blockade of ET in hypertension can prevent LVH and diastolic heart failure is the focus of current research.
Tissue Necrosis Factor-Alpha
TNF-α is a 76-amino-acid peptide secreted in response to a range of stimuli. TNF-α circulates in the blood (and is also present as a transmembrane form) and acts via specific receptors to induce pleomorphic effects in many cells. TNF-α promotes the inflammatory response, stimulates growth factors, and is directly cytotoxic to endothelial cells. Plasma TNF-α levels are elevated and may have a pathophysiological role in malignancy, septic shock, rheumatoid arthritis, and transplant rejection.
TNF-α appears to have a role in the pathophysiology of end-stage heart failure, where plasma levels are elevated and correlate with symptomatic status. It has a direct effect on cardiac muscle, including negative inotropism (through inhibition of calcium regulation in the cytoplasmic reticulum and inactivation of β-adrenergic receptors), activation of matrix metalloproteases, and promotion of cardiomyocyte hypertrophy. Overexpression of TNF-α produces a dilated cardiomyopathy phenotype with heart failure and premature death. TNF-α may mediate these effects through expression of fetal gene programs in cardiac myocytes, activation of pro-apoptotic pathways, and stimulation of other cytokines. Despite promising indications that expression of the cardiomyopathic phenotype could be attenuated by anticytokine strategies, subsequent studies in advanced heart failure have not demonstrated a clinical benefit from TNF-α blockade.
Cardiotrophin-1
Cardiotrophin (CT)-1 is a 201-amino-acid peptide from the interleukin (IL)-6 family of cytokines. It appears to have a key role in mediating myocyte hypertrophy. Absence of CT-1 leads to hypoplastic development of the heart, whereas increased CT-1 is associated with myocyte hypertrophy that can lead to eccentric LVH and chamber dilatation. CT-1 expression is induced by mechanical stretch and is augmented by sympathetic stimulation. Increased expression is also seen in hypoxia, indicating a potential protective and reparative role in myocardial ischemia. CT-1 expression increases in experimental heart failure and may precede BNP activation. Circulating CT-1 levels appear to rise in human heart failure in relation to the severity of LV dysfunction. Increased myocardial CT-1 expression is associated with downregulation of its major receptor, glycoprotein 130, but whether or not this contributes to impaired contractility is unclear. The role of CT-1 in diastolic heart failure is less clear. However in experimental models, transplantation of CT-1-expressing myoblasts into hypertensive rat hearts attenuated the progression to heart failure, indicating a potential protective role.
Adrenomedullin
AM is a 52-amino-acid peptide from the calcitonin gene related peptide (CGRP) family that acts via calcitonin receptorlike receptors (CRLRs) modified by receptor activator modifying proteins (RAMPs) 2 and 3, with cAMP as second messenger. Secreted mainly from vascular endothelial and smooth muscle cells, mRNA and peptide immunoreactivity for AM and its receptors have been demonstrated in cardiac myocytes and fibroblasts with increased expression in models of cardiac hypertrophy and heart failure. Recent data suggest that AM may have protective actions during development of hypertensive heart failure. Exogenous AM appears to reduce fibrosis in vivo and in cultured cardiac fibroblasts. AM also appears to specifically inhibit A-II-stimulated hypertrophic responses and upregulation of ANP and BNP in cardiac myocytes. During development of hypertrophy in the salt-sensitive hypertensive rat model, myocardial expression and peptide levels of AM increased early, congruent with natriuretic peptide levels and before RAS activation occurred. Chronic administration of low-dose AM in this model significantly lowered LV diastolic pressure, increased cardiac output, and lowered end systolic elastance. These effects were accompanied by a significant attenuation of RAS activation and prolongation of the time to onset of heart failure or death. In human subjects, plasma levels of AM increase in relation to the severity of heart failure. Short-term infusion in human hypertensive or heart failure subjects to achieve AM levels within the pathophysiological range significantly lowered blood pressure and increased cardiac output and diuresis, while attenuating aldosterone secretion. These findings suggest that AM may have an important protective role in attenuating the progression to hypertrophy and heart failure. Whether augmentation of endogenous levels or exogenous treatment could be used therapeutically has not been evaluated.
Urocortin
The urocortin (UCN) peptides 1, 2, and 3 belong to the corticotrophin-releasing factor (CRF) family that acts via CRF receptor subtypes with cAMP as second messenger. These peptides have multiple actions, including stress and inflammatory responses. Plasma levels are increased in human heart failure but have not been carefully examined in hypertension. Infusion studies in animal models and human subjects demonstrate that UCN-1 and UCN-2 have differential cardiovascular actions, with the latter causing more profound lowering of blood pressure and increased cardiac output associated with attenuation of RAS activation and maintenance of renal function. The role of UCN-1 and −2 in the development of cardiac hypertrophy and progression to heart failure has not been clearly elucidated.
Cardiac Natriuretic Peptides
The cardiac natriuretic peptides are a group of related hormones with structural homology and similar bioactivity. These peptides share a highly conserved ring structure that is responsible for bioactivity ( Fig. 27-2 ). Each is secreted in a 1 : 1 ratio with the amino-terminal portion of its pro-hormone. Normally secreted in small amounts—giving picomolar levels in the circulation—synthesis increases with pressure and volume overload as fetal gene expression is reactivated.
The ANP and BNP peptides are synthesized primarily by cardiac myocytes in response to mechanical stretch. Synthesis is regulated by multiple factors, including activity of other vasoactive hormones, such as A-II. Their major actions are mediated through natriuretic peptide receptor (NPR)-A, which is widespread, including in the vasculature and renal tubules. These peptides produce vasodilatation and natriuresis and also suppress thirst and inhibit RAS and aldosterone secretion, fibrosis, and proliferative responses to cardiac or vascular injury. ANP and BNP are actively cleared by competitive uptake at NPR-C receptors or by cleavage of the ring structure by neutral endopeptidase, which is found in high density within the renal tubules. Plasma levels of atrial and B-type peptides are higher in women, increase with age and renal dysfunction, and fall with increasing body mass index (BMI). In contrast, C-type natriuretic peptide (CNP) is secreted primarily from vascular endothelial cells and acts predominantly as a vasodilator. Its metabolism and clearance are less well characterized.
Although circulating levels of all the natriuretic peptides show relationships with indices of systolic and diastolic LV function and to LVH, the strongest correlations are seen in general for BNP.
Atrial Natriuretic Peptide (ANP/NT-proANP)
This peptide is secreted mainly from the cardiac atria and to a lesser extent from the ventricles. Increased wall stretch stimulates its release from storage granules while also augmenting its transcription and synthesis. ANP contributes to control of basal vascular tone and blood pressure. In ANP gene knockout mice, absence of ANP was associated with mildly elevated basal blood pressure compared with wild-type littermates. NPR-A receptor knockout mice also developed hypertension by a mechanism independent of salt intake.
In healthy human volunteers, low-dose ANP and BNP infusions that produce plasma levels within the normal range induced vasodilatation, natriuresis, and inhibition of renin and aldosterone. Falls in blood pressure reflect a reduction in cardiac preload and direct arterial vasodilatation. Long-term (4-5 day) infusions of low-dose ANP induce sustained falls in arterial pressure, peripheral vascular resistance, plasma volume, and central filling pressure, without activating the RAS, aldosterone, or SNS. Similar effects have been seen during BNP infusion, although fewer studies have been performed.
The cardiac peptides produce natriuresis and diuresis via renal glomerular and tubular actions. The natriuretic action of ANP and BNP is highly dependent on renal perfusion pressure. ANP increases glomerular filtration rate by simultaneously dilating glomerular afferent arterioles, constricting efferent arterioles, and relaxing glomerular mesangial cells. It also blocks sodium reabsorption in the distal collecting ducts, antagonizes AVP-mediated water uptake in collecting ducts, and inhibits A-II-mediated sodium and water uptake in proximal tubules. Administration of a competitive antagonist for the NPR-A receptor blocks the natriuretic action of the natriuretic peptides in normal and heart-failed animals.
Under most circumstances, ANP and BNP inhibit the RAS, aldosterone, and SNS. Experimental administration of the NPR blocker HS142 produces an elevation in plasma levels of renin activity, aldosterone, and catecholamines. Conversely, low-dose infusion of ANP to produce physiological or mildly elevated plasma levels results in suppression of aldosterone secretion and reduced renin and SNS activity. ANP (and BNP) inhibits renin release from the juxtaglomerular apparatus and reduces peripheral sympathetic tone through effects on baroreceptors, suppression of catecholamine release from autonomic nerve endings, and inhibition of central sympathetic outflow.
C-type Natriuretic Peptide (CNP/NT-CNP)
Produced mainly from vascular endothelial cells, CNP is synthesized as a precursor and cleaved into a biologically active carboxy-terminal peptide, CNP, and an apparently inactive amino-terminal peptide, NT-proCNP. In normal humans, circulating levels of CNP are very low and at the limits of detection measurement by immunoassays. The greater size and presumed longer plasma half-life of NT-proCNP may allow more accurate measurement and a more reliable indication of CNP production. Contradictory reports have suggested the presence or absence of elevated plasma CNP concentrations in congestive heart failure. In one large cohort, NT-proCNP levels were clearly shown to be elevated in heart failure. Levels of NT-proCNP are independently related to gender, age, and LV systolic function. Levels rise with age, are higher in men than women, and are inversely related to creatinine clearance. Plasma NT-proCNP also appears to identify heart failure with modest incremental diagnostic value over NT-proBNP and independent of age, gender, and renal function. Although CNP is reportedly a potent venodilator, it has little effect on arterial pressure when infused intravenously into healthy volunteers. Its actions in heart failure and hypertension are not yet clearly defined.
B-Type Natriuretic Peptides
BNPs/NT-proBNPs are synthesized and secreted mainly from LV cardiomyocytes. The primary stimulus is mechanical stretch due to increased wall stress. Diastolic wall stress appears to be the most important stimulus ( Fig. 27-3 ), which in part explains higher levels in systolic dysfunction where end diastolic volumes are larger. Significant atrial and right ventricular (RV) contributions to total BNP/NT-proBNP secretion are seen in advanced heart failure. Cleavage of the precursor peptide (proBNP, amino acids 1-108) produces the 32-amino-acid BNP (77-108) and its corresponding amino-terminal component, NT-proBNP (1–76), which are secreted in 1 : 1 ratio. Plasma levels of these peptides correlate strongly with each other. BNP is bioactive and has a shorter half-life due to active clearance, hence levels are lower than for the more stable NT-proBNP by a factor of 5-10-fold. Levels of both peptides increase in parallel with LV pressure or volume loading and reflect the severity of LV dysfunction, correlating inversely with LVEF and positively with increasing LV mass, indices of LV filling pressure, LV stiffness, and measures of LA size.
Wide interindividual variation in BNP and NT-proBNP levels in stable symptomatic heart failure reflects many factors. In addition to age and gender, key determinants include renal dysfunction and atrial fibrillation, both of which cause higher levels. Higher levels in renal dysfunction appear to reflect reduced clearance, as the transcardiac gradient, reflecting cardiac secretion, appears unaltered in this setting. Nevertheless, increased cardiac production is probable in many patients with end-stage renal failure, since heart failure, LVH, and coronary artery disease are common and leading causes of death in this setting. Body mass is an important determinant of BNP levels, possibly through increased NPR-C receptors in adipose tissue. NT-proBNP levels are less affected by body mass, possibly owing to a different clearance mechanism. RV systolic function and mitral regurgitation are key determinants in more advanced heart failure, reflecting LA and RV production. Most of the inter-individual variation in peptide levels is explained by LV systolic and diastolic functions, RV dysfunction, renal function, gender, age, and mitral regurgitation. Hereditary factors or molecular heterogeneity may be responsible for much of the residual variation in BNP levels. BNP levels therefore act as a global marker of cardiac as well as end-organ dysfunction rather than as an index of a single cardiac index, such as LA pressure. Levels should be understood within the context of these multiple determining factors.
In severe heart failure, BNP levels and gene expression are strongly related to changes in multiple genes that play a part in LV remodeling, such as matrix metalloproteinases. Mehra et al. looked at gene expression in myocardium from donor hearts at the time of transplantation and demonstrated that upregulation of BNP was associated with upregulation of more than 25 specific genes associated with cellular remodeling, vascular injury and repair, and alloimmune inflammatory interactions. Additional genes involved in stem cell mobilization pathways and apoptosis were also identified. These findings suggest that BNP may be a marker of active remodeling.
A variety of immunoassays are available for the measurement of BNP and NT-proBNP. Each has different assay characteristics with specific detection limits and analytical coefficients of variation. There are significant correlations between different assay measurements taken from the same samples. However, assays with lower analytical variation are more likely to detect clinically important changes in serial samples. There has been recent interest in the biological variation in serial samples from clinically stable patients. Coefficients of biological variability in circulating levels of BNP or NT-proBNP vary between 30% and 55%. In healthy subjects in whom peptide levels are very low, a doubling in levels—for example, from 8 to 16 pg/ml—would be necessary for a change to confidently exceed usual biological variation. In unstable heart failure patients, a change of 30% is likely to exceed background variation. The clinical importance of the usual variation seen in peptide levels requires some clarification. These changes almost certainly reflect subclinical changes in hemodynamic indices, dynamic myocardial ischemia, and the complex neurohormonal milieu that regulates BNP secretion. Serial BNP or NT-proBNP levels have greater prognostic value than a single value, suggesting that there may be a role for repeated measurements to identify response to therapy and to stratify risk.
Activation of the cardiac natriuretic peptide system is a feature of hypertension. Plasma levels increase in relation to hemodynamic load and to the degree of LVH. In animal models such as the spontaneously hypertensive rat, BNP gene expression has been demonstrated before the development of LVH. In human hypertensive subjects, plasma BNP levels are elevated compared with age-matched controls and reflect LV wall thickness, the severity of LVH, and LA dilatation. Levels of BNP are higher again in subjects with symptomatic heart failure due to diastolic dysfunction when compared with matched controls with hypertension and LVH.
It seems unlikely that BNP levels can be used to accurately determine invasively measured LV pressures, particularly in the setting of a normal LVEF. While cross-sectional observational studies demonstrate statistically significant positive correlations for BNP with LV end diastolic pressure and preatrial contraction (pre-A) pressure, the associations are relatively weak. In patients with more advanced heart failure during hemodynamically guided treatment with pulmonary artery catheter monitoring of LV filling pressures, there are concordant changes in simultaneously measured BNP levels and pulmonary capillary wedge pressures, but the changes do not correlate strongly. Maximal changes in BNP levels lag hemodynamic changes by up to 24 hours, presumably reflecting relatively sluggish changes in constitutive synthesis and secretion of these peptides. Patients in whom BNP levels do not fall, despite an improvement in LA pressure estimates, are, however, at highest risk for adverse events. Data from patients with implantable hemodynamic monitoring devices indicate that while absolute levels of BNP or NT-proBNP reflect multiple factors specific to that individual patient, changes in BNP from baseline for that patient are highly correlated with changes in estimated LV filling pressures. Whether this is true in the setting of a normal EF is still the subject of ongoing study.
Prostaglandins
Hormones in the prostaglandin family are derived from arachidonic acid and are synthesized throughout the body. Prostaglandin I 2 and E 2 are potent vasodilators. Their levels increase in heart failure in plasma and at tissue level, particularly in the kidneys. The prostaglandins modulate the secretion and effects of ANP, renin, and A-II, and prostaglandin release is in turn increased by A-II, NE, and AVP. Prostaglandins play an important role in renal homeostasis during heart failure. The adverse renal glomerular and tubular effects of nonsteroidal anti-inflammatory drugs in heart failure reflect their inhibition of prostaglandins.
Pattern of Neurohormonal Activation in Heart Failure with Normal Left Ventricular Ejection Fraction
The pattern of neurohormonal activation in systolic heart failure has been well characterized. In contrast there are fewer data in the context of heart failure when LVEF is preserved. While the clinical presentation may be similar, the demographic features of heart failure with preserved EF differ from those of systolic heart failure. Patients with preserved EF are generally older, more likely to be female, more likely to have antecedent hypertension and diabetes, and less likely to have had prior myocardial infarction than counterparts with systolic heart failure. There may be differences in ethnic or racial profiles as well. Each of these differences is likely to affect the neurohormonal profile seen in heart failure with preserved EF.
Intuitively, it seems likely that there may be a lesser degree of hemodynamic impairment in the context of heart failure with preserved EF, particularly with regard to arterial underfilling, a key stimulus for neurohormonal activation. Hence it seems likely that the degree of neurohormonal activity detected in plasma may be less than that seen in systolic heart failure. In one sense, the differentiation of systolic from diastolic heart failure on the basis of arbitrary cutpoints in EF could be regarded as artificial and an oversimplification of the underlying pathophysiology. This is particularly so as the degree of neurohormonal activation reflects indices of cardiac function such as LVEF, wall stress, and end diastolic pressure in a relatively continuous manner.
Data from the Vasodilator in Heart Failure Trials (V-HeFT) study indicate that lesser degrees of systolic impairment are associated with less neurohormonal activation. The V-HeFT study group assessed patients with milder versus more severe (LVEF <35%) systolic dysfunction. Patients with a higher LVEF (>35%) had less activation of NE and fewer adverse clinical outcomes.
A number of studies have attempted to define the neurohormonal profile of heart failure with preserved systolic function. Cohort sizes are relatively small, but the study groups were well characterized and afford some insight. In general, these studies demonstrate that there is neurohormonal activation in heart failure where LV ejection is preserved but that this activation is mild and significantly less than that seen in systolic heart failure. A group of patients from the Studies of Left Ventricular Dysfunction (SOLVD) registry with radiological pulmonary congestion and either preserved LVEF (>45%, n = 41) or systolic impairment (LVEF <45%, n = 89) were compared with matched controls. Plasma levels of NE, vasopressin, ANP, and renin activity were significantly elevated in systolic heart failure subjects compared with controls and, with the exception of vasopressin, were higher than in heart failure patients with a normal EF. These differences remained significant when adjusted for clinical and treatment variables. Subjects with preserved EF had small but significant increases in plasma renin activity and vasopressin levels compared with controls ( Fig. 27-4 ). The impact of drug treatment, including diuretics and ACE inhibitors, may have contributed to differences in hormones between the groups.
In a second, more recent study, Kitzman et al. studied 147 clinically stable subjects aged at least 60 years. Fifty-nine subjects had stable heart failure with an LVEF of at least 50% and no evidence of valvular disease, ischemic heart disease, or pulmonary disease. A further 60 subjects had systolic heart failure with LVEF under 35%. These two groups were compared with 28 age-matched control subjects. All subjects underwent comprehensive echocardiography, symptom-limited exercise testing, and blood sampling for neurohormones. Compared with both control and systolic heart failure groups, the diastolic heart failure group differed significantly, with more women, greater mean BMI, a higher frequency of hypertension, higher recorded blood pressures, and less frequent treatment with diuretics, ACE inhibitors, nitrates, and beta blockers. There were also trends to more frequent diabetes history and greater use of calcium channel blockers. Diastolic heart failure subjects had similar LV volumes and EFs to controls, but greater LV mass. Systolic heart failure subjects had substantially larger LV volumes, lower EFs, and greater LV mass than both diastolic heart failure and control subjects. Diastolic indices were similar in all heart failure subjects. LA size was not reported. Exercise performance was impaired to a similar degree in both systolic and diastolic heart failure groups. Mean plasma NE, ANP, and BNP levels were significantly higher than controls in both diastolic and systolic heart failure groups. Whereas NE levels were similar in the two heart failure groups, both ANP and BNP were significantly higher in patients with systolic versus diastolic heart failure.
In a prospective study of 556 patients presenting with heart failure in the Olmsted County community, Bursi et al. found that plasma BNP levels (median, 257; interquartile range [IQR], 115-211 pg/ml) were significantly higher than those of asymptomatic subjects in the same community (17-58 pg/ml). BNP levels in the 248 subjects with an LVEF below 50% (388 [164–251] pg/l) were significantly higher than levels in the 308 subjects with preserved EF (183 [88–351] pg/ml; p < 0.001). BNP levels also increased significantly with more severe diastolic dysfunction in the group of patients with reduced systolic function and in the group with preserved systolic function. BNP levels were independently associated with age, LVEF, and severity of diastolic dysfunction. These findings indicate that both systolic and diastolic dysfunctions are important determinants of BNP levels.
The pattern of neurohormonal activation in acute heart failure with preserved systolic function is less clear. In one small study, 14 of 30 patients presenting with acute heart failure had an LVEF greater than 50%. Plasma NE and BNP levels were measured within 48 hours of presentation and at follow-up. NE levels were significantly elevated in both diastolic and systolic heart failure groups to a similar degree both at admission with acute heart failure and at follow-up. In contrast, BNP levels were elevated above normal controls in the diastolic heart failure group but on average were less than half the level seen in systolic heart failure. These findings suggest a similar level of compensatory sympathetic activation whether heart failure is due to systolic or primarily diastolic abnormality. Higher BNP levels in systolic heart failure most likely reflect the effect of LV geometry on wall stress, the primary stimulus for BNP secretion. Iwanaga et al. assessed 62 subjects with diastolic heart failure and 98 with systolic heart failure and found a strong correlation in both groups between plasma BNP levels and end diastolic wall stress, both of which were higher in the systolic heart failure group, reflecting in part larger LV volumes ( Fig. 27-5 ).
The relative levels of BNP in acute heart failure with reduced versus preserved EF have been confirmed in larger studies. In the Breathing Not Properly (“BNP”) multinational study, of 1582 patients presenting with acute dyspnea, a subgroup of 452 patients with confirmed heart failure subsequently underwent echocardiography within 30 days of admission. One hundred and sixty five of these (35%) had preserved LV systolic function with LVEF greater than 45%. Within this group, median BNP levels of 413 pg/ml at presentation were higher than for non-heart failure patients (34 pg/ml) but significantly lower than for the 283 subjects with systolic heart failure (821 pg/ml). There were, however, significant differences in demographic and clinical factors between the two heart failure groups, rendering interpretation difficult. BNP levels for the systolic heart failure group may have been underestimated due to an upper limit of only 1300 pg/ml for the assay. No other neurohormones were reported from this series.
Plasma AM levels were assessed by Yu et al. in a study of 77 patients with heart failure and either normal or impaired LVEF. AM levels were higher in subjects with heart failure (47.5 ± 6.5 pmol/L) than those without (6.9 ± 1.2; p < 0.01). There was a gradient in AM levels, which were significantly higher in those with both systolic heart failure and a restrictive filling pattern (92 ± 21) than in those with either systolic impairment and nonrestrictive filling (38 ± 7) or preserved systolic function (34 ± 6).
CLINICAL RELEVANCE
Neurohormones as Treatment Targets for Hypertension and Diastolic Heart Failure
Neurohormonal systems are the target for a number of the effective antihypertensive therapies. Treatment of hypertension with ACE inhibitors, AT 1 R antagonists, or aldosterone receptor blockade is associated with regression of LVH and improved diastolic function. More importantly, these agents reduce the incidence of new heart failure. Furthermore, in combination with β-adrenergic receptor blockers, they are the cornerstone therapy for systolic heart failure, , but only AT 1 R antagonists have been tested in a large randomized study in the context of heart failure with preserved systolic function. Smaller studies suggest beneficial effects on diastolic function and neurohormonal activation from aldosterone, β-adrenergic receptors, and ACE inhibitors. However, adequate mortality studies have not been performed in the setting of heart failure with preserved systolic function.
Natriuretic Peptides and Screening for Cardiac Dysfunction
Because plasma BNP and NT-proBNP levels reflect LV structural and functional abnormalities, potentially they could be used to screen for cardiac dysfunction. This approach could identify subclinical abnormalities or facilitate more appropriate referral for echocardiography. Several studies have demonstrated potential utility for screening with BNP or NT-proBNP ( Table 27-1 ). Accuracy of these peptides for detecting cardiac dysfunction appears to depend on the clinical context, the abnormality in question, and its prevalence within that population.
| B-TYPE NATRIURETIC PEPTIDE AS A MEASURE OF VENTRICULAR FUNCTION IN SUBJECTS REFERRED FOR ECHOCARDIOGRAPHY | ||||||||
|---|---|---|---|---|---|---|---|---|
| Study | Cohort Type | End Point | Prevalence | AUC | SENS * | SPEC * | PPV * | NPV * |
| Yamamoto et al | Echocardiography referrals; n = 466 | EF < 45% | 11% | 0.79 | 79% | 64% | 21% | 96% |
| Maisel et al | Echocardiography referrals; n = 200 | EF < 50% or diastolic dysfunction | 47% | 0.96 | 86% | 98% | 98% | 89% |
| Lubien et al | Echocardiography referrals; n = 294 | Diastolic dysfunction | 40% | 0.92 | 74% | 98% | 96% | 85% |
| B-TYPE NATRIURETIC PEPTIDE FOR SCREENING COMMUNITY AND CLINIC POPULATIONS | ||||||||
|---|---|---|---|---|---|---|---|---|
| Study | Cohort Type | End Point | Prevalence | AUC | SENS * | SPEC * | PPV * | NPV * |
| Vasan et al | Community volunteers, n = 3,177; age 55 ± 10 yrs | EF ≤ 50% | 5.6% | M 0.72; W 0.56 | 42% | 90% | 30% | 94% |
| EF ≤ 40% | 2.2% | M 0.79; W 0.85 | 55% | 89% | 18% | 98% | ||
| LV mass ↑ | 4.8% | M 0.72; W 0.57 | 41% | 90% | 32% | 93% | ||
| Redfield et al | Community population based, n = 1,997; age 62 ± 11 yrs | EF ≤ 50% | 6.0% | M 0.69; W 0.70 | 64% | 68% | NA | NA |
| EF ≤ 40% | 1.1% | M 0.90; W 0.92 | 90% | 76% | 4% | 99% | ||
| Moderate-severe DD % | 6.9% | M 0.79; W 0.77 | 75% | 69% | 15% | 97% | ||
| Costello-Boerrigter | Community population based, n = 1,869; | |||||||
| Age < 65 yrs | EF < 40 | 1% | M 0.97; W 0.97 | 90.9% | 90.9% | NA | NA | |
| Age > 65 yrs | EF < 40 | 3.5% | M 0.93; W 0.91 | 88.5% | 88.4% | NA | NA | |
| McDonagh et al | Community population based; n = 1,252; age 51 ± 14 yrs | EF ≤ 30% | 3.2% | 0.88 | 76% | 87% | 16% | 98% |
| Nakamura et al | Community volunteers, n = 1,098; age 56 yrs | “Heart disease” (atrial fibrillation-flutter, MI, valvular or hypertensive, cardiomyopathy, ASD, cor pulmonale) | 3.6% | 0.97 | 90% | 96% | 44% | 99% |
| Niinuma et al | Health screening clinic; n = 481 | “Heart disease” | 2.7% | 0.94 | 85% | 92% | NA | NA |
* Calculated at the B-type natriuretic peptide discriminatory value that provides best overall test accuracy; i.e., provides the highest combination of sensitivity and specificity.
In patients referred for echocardiography, several studies with moderate-sized cohorts have demonstrated that BNP has excellent sensitivity and specificity for detecting systolic (LVEF <40%-50%) or diastolic dysfunction. Greater negative predictive, or “rule out,” values are seen when prevalence is lower, while specificity and positive predictive values are higher with higher prevalence. Krishnaswamy et al. studied 400 patients referred for echocardiography and demonstrated excellent detection of any ventricular dysfunction by BNP, with a level of at least 75 pg/ml having an accuracy of 90% for detecting either systolic (LVEF <50%, n = 225) or diastolic dysfunction (impaired relaxation or worse on transmitral filling; n = 98). Among patients with isolated diastolic abnormality, BNP levels were higher in those with restrictive filling patterns than those with impaired relaxation. Highest levels were seen with systolic impairment and restrictive diastolic filling. Because of overlap between groups, BNP levels did not differentiate systolic from primary diastolic dysfunction with preserved EF.
Lubien et al. assessed the utility of BNP in detecting diastolic dysfunction in the setting of normal systolic function (LVEF >50%; LV internal diastolic diameter <5.5 cm) in 294 patients referred for echocardiography. Clinical heart failure was present in 5% of the cohort. Diastolic function was defined as normal, impaired relaxation, pseudonormal, or restrictive based on transmitral filling patterns without reference to tissue Doppler or other diastolic indices. BNP levels (Biosite® assay) increased with greater severity of diastolic dysfunction and were higher still if clinical heart failure was present ( Fig. 27-6 ). A BNP level of 65 pg/ml had a sensitivity of 85%, a specificity of 83%, and an overall accuracy of 84% for detecting any diastolic abnormality when systolic function was normal. Accuracy was higher when only restrictive filling was considered, with an area under the receiver operating characteristic (ROC) curve of 0.98. BNP levels also differentiated the presence of an LA abnormality and increased LV mass.
