Heart failure may be classified symptomatically on the basis of its clinical severity (Table 58-2). Despite the varied causes and classifications of heart failure, the manifestations all reflect intravascular volume overload, inadequate tissue perfusion, or their combination. A thorough explanation of all these forms of heart failure is beyond the scope of this chapter. The following discussion, therefore, focuses primarily on left ventricular systolic failure in both the acute and chronic setting.

Table 58–2 Clinical Classifications of Heart Failure

New York Heart Association Classification
Class I	Symptoms only with greater than usual activity
Class II	Asymptomatic at rest but with symptoms during normal activities
Class III	Asymptomatic at rest but with symptoms during minimal exertion
Class IV	Symptoms at rest
American College of Cardiology/American Heart Association Classification
Stage A	Patients with structurally normal hearts, asymptomatic, but at risk for the development of heart failure due to the presence of risk factors (e.g., hypertension, coronary artery disease, diabetes)
Stage B	Patients with structurally abnormal hearts (e.g., left ventricular systolic dysfunction, left ventricular hypertrophy, valvular dysfunction, prior myocardial infarction) but without symptoms of heart failure
Stage C	Patients with structurally abnormal hearts and current or prior symptoms of heart failure
Stage D	Patients with end-stage heart failure symptoms not responsive to standard therapy

Acute Heart Failure

The response of the cardiovascular system to the onset of myocardial dysfunction and the pathophysiologic mechanisms underlying the subsequent progression to heart failure depends in large part on the acuity of the dysfunction. An acute cardiac insult results in a series of hemodynamic alterations that account for the clinical manifestations of left ventricular failure. This occurs irrespective of whether the initial insult depresses myocardial contractility (systolic dysfunction) or impairs ventricular filling (diastolic dysfunction). The cascade begins with a rise in left ventricular end-diastolic pressure. This elevated pressure is transmitted to the left atrium and subsequently to the pulmonary venous and capillary system. The increased intravascular pressure results in transudation of fluid into the pulmonary interstitium, where it interferes with gas exchange, resulting in hypoxemia and dyspnea. In addition, there is often an associated reduction in cardiac output resulting in inadequate delivery of blood to the arterial system with resultant organ hypoperfusion.

The heart’s response to these hemodynamic alterations is the activation of several compensatory mechanisms (Table 58-3). A rapid, generalized activation of the adrenergic system occurs and is associated with a withdrawal of parasympathetic tone. Direct sympathetic stimulation of the heart and β-adrenergically induced release of epinephrine and norepinephrine from the adrenal glands result in tachycardia and an increase in myocardial contractility, both of which serve to augment cardiac output. Catecholamine-induced peripheral arterial vasoconstriction redirects the available cardiac output away from relatively nonessential organs (i.e., skin, skeletal muscle, gut, kidney) and helps maintain sufficient blood pressure to ensure adequate perfusion of more vital organs (i.e., heart and brain). Furthermore, β-adrenergic stimulation of the juxtaglomerular apparatus in the kidneys results in the release of renin and activation of the renin-angiotensin system. The angiotensin II thus produced is a potent vasoconstrictor and acts in concert with the direct α-adrenergic stimulation of the vasculature to maintain blood pressure. The reduced renal blood flow from depressed cardiac output and redistribution of blood volume results in activation of renal baroreceptors. This further stimulates renin release and augments sympathetic activation, thereby contributing to vasoconstriction.

Table 58–3 Compensatory Mechanisms in Heart Failure

In addition to producing these hemodynamic alterations, acute heart failure is characterized by marked sodium and water retention. This occurs through a variety of mechanisms. Angiotensin II directly promotes the reabsorption of sodium in the proximal nephron and indirectly promotes the reabsorption of sodium from the distal nephron, this latter effect being mediated through angiotensin II–induced release of aldosterone from the adrenal cortex. Furthermore, angiotensin II and norepinephrine stimulate hypothalamic release of arginine vasopressin, resulting in further vasoconstriction and free water reabsorption. These changes produce an expansion of intravascular volume and augmentation of venous return, thereby increasing ventricular end-diastolic volume (preload). The increased preload results in an increase in stroke volume by the Frank-Starling mechanism (Fig. 58-1) and thereby helps support the cardiac output.

Figure 58–1 Frank-Starling curve with normal systolic function and with heart failure and the hemodynamic effect of pharmacologic therapy. In the normal setting, an increase in preload results in an increase in stroke volume. In heart failure, this relationship is blunted, and at any given stroke volume, the preload must be higher (horizontal dotted line). Similarly, for any given preload, the stroke volume is lower (vertical dotted line). Diuretic therapy (D) reduces preload without a significant effect on stroke volume, whereas inotropic therapy (I) augments stroke volume without an appreciable effect on preload. Vasodilator therapy (V) has moderate beneficial effects on both preload and stroke volume; however, the greatest effects are seen with combination therapy (I + V, I + V + D).

Chronic Heart Failure

In combination, the aforementioned mechanisms serve a compensatory role in acute heart failure, helping to maintain cardiac output and blood pressure to allow adequate perfusion of vital organs. These compensatory mechanisms may initially be adequate to allow clinical stability, and the patient may subsequently go through a stage of asymptomatic ventricular dysfunction, maintained in part by chronic stimulation of the adrenergic and renin-angiotensin systems. As heart failure progresses, the hemodynamic overload induces changes in the shape and size of the ventricle, a process known as ventricular remodeling. The specific changes that occur depend in part on the hemodynamic stressors facing the ventricle. In predominantly pressure-overloaded conditions (e.g., hypertension, aortic stenosis), there is a rise in systolic wall stress that results in left ventricular hypertrophy. If this hypertrophy is insufficient to normalize the wall stress, dilation occurs. Under conditions of volume overload (e.g., aortic or mitral insufficiency), there is a rise in diastolic wall stress that induces ventricular dilation. This dilation in turn results in increased systolic wall stress (by the Laplace relationship) and subsequent hypertrophy. These hypertrophic changes help maintain systolic wall stress within a normal range and help preserve ventricular contractile function. However, with continued hemodynamic overload, there is progressive ventricular dilation, eventuating in the development of a dilated, spherical heart. This altered ventricular morphology produces less efficient ventricular contraction, may induce mitral regurgitation due to annular dilation and malcoaptation of the valve leaflets, and is associated with an adverse prognosis.

The stimuli that induce ventricular remodeling are varied and the mechanisms underlying the remodeling process are complex (Table 58-4). It appears that increased wall stress (resulting from ventricular dilation and increased afterload) and neurohormones (i.e., β-adrenergic and renin-angiotensin systems), vasoactive peptides (e.g., endothelin), and cytokines (e.g., tumor necrosis factor α) mediate remodeling. These factors may have direct effects on the cardiac myocytes or may act indirectly through stimulation of second-messenger systems and thereby induce a variety of changes in myocyte structure and function. On a cellular level, myocyte hypertrophy results from the replication of sarcomeres either in parallel (producing ventricular hypertrophy) or in series (producing ventricular dilation). Alterations in the expression of various contractile proteins occurs with reexpression of fetal genes and reduced expression of adult contractile genes, resulting in abnormalities of calcium handling and excitation-contraction coupling. Chronic stimulation of the sympathetic system is accompanied by a reduction in the density of β-adrenergic receptors in the myocardium and an uncoupling of the receptors from their intracellular mediators. This results in a blunted response of the failing myocardium to either endogenous (e.g., exercise) or exogenous (e.g., dopamine or dobutamine) adrenergic stimulation.

Table 58–4 Factors Involved in Ventricular Remodeling

Stimulants of Ventricular Remodeling	Molecular and Cellular Events That Mediate Ventricular Remodeling
Altered hemodynamic load (increased preload, afterload, wall stress) β-Adrenergic stimulation Activation of the renin-angiotensin system Inflammatory cytokines (e.g., tumor necrosis factor α, interleukins 1 and 6) Vasoactive peptides (e.g., endothelin) Oxidative stress	Myocyte hypertrophy Myocyte loss (necrosis and apoptosis) Myocyte “slippage” Transition to fetal myocyte phenotype Neurohormonal activation Fibroblast proliferation Alterations in the interstitial matrix Alterations in excitation-coupling

In addition to these alterations of the contractile apparatus within the myocytes, there is a progressive reduction in the number of myocytes, in part the result of apoptosis induced by the various stimuli of remodeling. Furthermore, changes occur in the extracellular matrix related to fibroblast proliferation, interstitial fibrosis, and increased expression of degradative enzymes such as matrix metalloproteinases. The last factor results in the loss of mechanical coupling of myocytes and may contribute to the remodeling process by facilitating “myocyte slippage” and, thereby, ventricular dilation.

As the remodeling process develops, the neurohormonal effects of the β-adrenergic and renin-angiotensin systems on the peripheral vasculature and renal salt and water handling continue. The intense vasoconstriction, while maintaining flow to vital organs, contributes to hypoperfusion of the kidneys and progressive renal dysfunction. The augmented preload and cardiac output resulting from sodium and water retention help maintain circulating blood volume and tissue perfusion. However, the associated increase in end-diastolic pressure and increased ventricular wall stress contribute to progressive ventricular remodeling and result in pulmonary and systemic venous hypertension and precipitation of congestive symptoms. Thus, these initially compensatory changes become deleterious in the chronic setting. The inhibition of these processes offers not only a mechanism for the treatment of heart failure but also the potential to reverse the adverse remodeling seen in the chronic state (see Table 58-3).

PHARMACOLOGIC AGENTS USED IN THE TREATMENT OF HEART FAILURE

The majority of pharmacologic agents used for the treatment of heart failure effect benefit either directly by interfering with the hemodynamic alterations described before or through inhibition of the neurohormonal activation underlying these alterations (see Fig. 58-1). These medications fall into several categories: diuretics, vasodilators, inotropic agents, and neurohormonal inhibitors. Diuretics act to reduce preload (leftward shift on the Frank-Starling curve), resulting in decreased filling pressures and improved congestive symptoms. Although this fall in preload may be associated with a reduction in stroke volume, this effect is minimal in patients with elevated filling pressures. Pure venodilators similarly reduce filling pressures and congestive symptoms and have minimal effect on stroke volume. Arterial vasodilators and inotropic agents predominantly augment cardiac output and thereby improve organ perfusion; arterial vasodilators act indirectly through a reduction in vascular resistance, whereas inotropic agents directly increase contractility and stroke volume. Although the increased cardiac output seen with these agents may result in a fall in filling pressures, the effect may be relatively modest. As a result of the differential effects of these agents, many patients with heart failure attain the greatest benefit from combination therapy. Neurohormonal inhibitors may have mixed hemodynamic effects. The blockade of adrenergic tone (beta-blockers) and inhibition of the renin-angiotensin system (angiotensin-converting enzyme [ACE] inhibitors, angiotensin receptor blockers [ARBs]) results in vasodilation and augmented cardiac output. These agents may also decrease preload and reduce filling pressures through a reduction in neurohormonally mediated renal sodium and water retention.

Specific Agents

Diuretics

Diuretics have long played an important role in the symptomatic treatment of heart failure (Table 58-5). The induced diuresis and natriuresis reduce extracellular volume and ventricular filling pressures, thereby ameliorating congestive symptoms. This effect occurs without a significant decrease in cardiac output or systemic blood pressure unless excessive diuresis and intravascular volume depletion occur. Whereas diuretics are beneficial in controlling symptoms and improving exercise capacity in patients with heart failure, with the exception of spironolactone, eplerenone diuretic use has not resulted in a decrease in mortality.

Table 58–5 Diuretics Commonly Used in the Treatment of Heart Failure

Loop diuretics act in the thick ascending limb of the loop of Henle, where they inhibit the Na⁺-K⁺-2Cl⁻ transporter, resulting in increased delivery of sodium and water to the distal nephron. They also decrease the tonicity of the medullary interstitium and thereby limit the osmotic reabsorption of free water from the collecting tubules. Currently available loop diuretics include furosemide, bumetanide, torsemide, and ethacrynic acid. There is an increased risk of ototoxicity with ethacrynic acid; therefore, this agent should be reserved for patients who are allergic to or intolerant of other agents.

Furosemide is the loop diuretic most commonly used for the treatment of heart failure. For patients with mild to moderate congestive symptoms, it can be given orally at initial doses of 20 to 40 mg daily. Its bioavailability ranges from 40% to 70%, and gradual dose titration is frequently required. Furosemide has a relatively short half-life. Once renal tubular levels of the drug decline, avid sodium reabsorption occurs throughout the nephron, potentially limiting or preventing effective natriuresis. A twice-daily dosing regimen may therefore be required to produce adequate salt and water loss. In patients with more severe volume retention or decompensated heart failure, intravenous administration of furosemide (20 to 100 mg) may produce a more rapid and effective diuresis. The maximum intravenous dose is 300 mg; however, the risk of ototoxicity increases at such high doses. For patients who require frequent high doses of intravenous furosemide, a continuous infusion may produce a more effective diuresis and require a lower total daily dose. The infusion is usually started at 5 to 10 mg/hr and titrated as needed to obtain the desired effect. Bumetanide and torsemide have greater bioavailability than does furosemide (~80%) but have not demonstrated better efficacy and are significantly more expensive.

Thiazide diuretics act in the distal convoluted tubule, where they inhibit the Na⁺-Cl⁻ cotransporter. Their efficacy is dependent on the delivery of sodium to the distal nephron; therefore, their diuretic effect is limited by sodium reabsorption from more proximal regions, as occurs during intravascular volume depletion or in low-flow states. In addition, they are ineffective when glomerular filtration rates fall below 30 mL/min. Thiazides may be useful as the sole diuretic in treatment of mild congestive symptoms; however, their predominant role in the management of more advanced heart failure is as an adjunct to other diuretic therapy in patients who exhibit diuretic resistance. Although several thiazides are available, the most frequently used are hydrochlorothiazide (12.5 to 50 mg daily) and metolazone (2.5 to 10 mg daily). These agents exhibit synergism with loop diuretics and should be given approximately 30 minutes before administration of furosemide, bumetanide, or torsemide.

Spironolactone is a competitive inhibitor of aldosterone in the distal convoluted tubule, whereas eplerenone is a selective aldosterone blocker. These agents stimulate a mild natriuresis and potassium reabsorption and may be most effective in patients with advanced heart failure, in whom marked activation of the renin-angiotensin-aldosterone system results in aldosterone levels as high as 20 times normal. In the Randomized Aldactone Evaluation Study (RALES), patients with moderate to severe congestive heart failure (NYHA class III-IV) were treated with spironolactone (25 to 50 mg daily) and had improved symptoms, reduced rates of hospitalization for heart failure, and 30% reduction in mortality. In the Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS), patients with acute myocardial infarction, left ventricular ejection fraction of 40% or less, and evidence of heart failure were treated with eplerenone (25 to 50 mg daily) and had a 15% reduction in mortality and significantly fewer episodes of heart failure. The mechanism of benefit of these agents is unlikely to be related to a diuretic effect as they are relatively weak diuretics. Rather, it probably reflects inhibition of aldosterone-induced myocardial fibrosis and ventricular remodeling. The major side effect of spironolactone is gynecomastia, which is not seen with eplerenone because of its selective mineralocorticoid blockade. The dose of these agents should not exceed 50 mg daily because of the risk of hyperkalemia, especially in patients with a serum creatinine concentration of 2.5 mg/dL or higher or when they are used in conjunction with an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker for the treatment of heart failure.

Amiloride and triamterene inhibit the reabsorption of sodium in the distal convoluted tubule and proximal collecting duct, resulting in a mild natriuresis and reduction of the ionic gradient required for potassium secretion into the urine. These agents produce a mild diuresis without the potassium wasting seen with loop diuretics and thiazides and may be effective for the control of mild congestive symptoms. However, when they are given alone, they are not effective in maintaining a negative fluid balance in patients with advanced heart failure. In such patients, these agents may have benefit as part of a combination diuretic regimen, especially given their potassium-sparing properties.

Patients treated with diuretics require close monitoring of their renal function and serum electrolytes. Loop diuretics and thiazides can lead to profound hypokalemia and hypomagnesemia, especially when they are used in combination; spironolactone may result in hyperkalemia. In contrast to loop diuretics, thiazides do not alter the tonicity of the renal medullary interstitium and may produce significant hyponatremia due to the reabsorption of free water from the distal convoluted tubule in the face of a preserved interstitial gradient. In addition to these metabolic effects, thiazides may adversely affect serum lipid levels, and spironolactone may induce gynecomastia.

Vasodilators

Nitrovasodilators

Nitric oxide is formed by normal endothelial and smooth muscle cells throughout the vasculature and functions in both a paracrine and an autocrine fashion. Its primary mechanism of action involves an induced increase in intracellular cyclic guanosine monophosphate, which results in vascular smooth muscle relaxation. Nitrovasodilators such as sodium nitroprusside and organic nitrates (i.e., nitroglycerin) are metabolized to nitric oxide within the vasculature. They are potent vasodilators and, as such, are useful in the management of heart failure.

Nitroprusside is the sodium salt of nitric oxide and ferricyanide. It is a balanced vasodilator and produces both vasodilation and venodilation in both the systemic and pulmonary systems. These effects result in favorable hemodynamic changes, including a decrease in right atrial and pulmonary capillary wedge pressures (i.e., decreased preload), a reduction in pulmonary and systemic vascular resistance (i.e., decreased afterload), and an increase in stroke volume and cardiac output/index (Fig. 58-2). In contrast to other arterial vasodilators, nitroprusside does not cause a significant increase in heart rate, and its use is usually associated with a decrease in myocardial oxygen demand. Nitroprusside is useful for the management of heart failure associated with elevated filling pressures, low cardiac output, and high vascular resistance, as occurs in patients with decompensated systolic failure. It is also ideally suited for the management of heart failure associated with profound hypertension, acute mitral regurgitation, acute aortic insufficiency, or acute ventricular septal defect.

Figure 58–2 Comparative hemodynamic effects of maximally tolerated doses of nitroprusside, dobutamine, and milrinone in patients with severe heart failure. CI,cardiac index; dP/dt, change in pressure per change in time; HR, heart rate; LVEDP, left ventricular end-diastolic pressure; MAP, mean arterial pressure; RAP, right atrial pressure; SVR, systemic vascular resistance.

(Modified from Colucci WS, Wright RF, Jaski BE, et al. Milrinone and dobutamine in severe heart failure: differing hemodynamic effects and individual patient responsiveness. Circulation 1986;73:III175.)

Nitroprusside is administered as a continuous infusion. Its onset of action is rapid (within 30 seconds), and it reaches peak effect within 2 minutes. Similarly, its effects completely resolve within 3 minutes of discontinuation of its infusion. Because of these rapid changes in hemodynamics, it is best administered under the guidance of pulmonary (i.e., Swan-Ganz catheter) and systemic arterial monitoring. The usual starting dose is 0.1 to 0.25 μg/kg/min. The dose may be titrated up by 0.25 μg/kg/min every 5 to 10 minutes until the desired effect is achieved or the maximum dose is reached (10 μg/kg/min). The use of nitroprusside may be limited by the development of hypotension, especially in patients with normal left ventricular systolic function or low filling pressures. Rapid cessation of nitroprusside may result in rebound hypertension, probably reflecting neurohormonal activation. Nitroprusside is metabolized in the vasculature to nitric oxide and cyanide; cyanide is further metabolized in the liver to thiocyanate, which is excreted by the kidneys. Accumulation of these toxic metabolites is more likely to occur when nitroprusside is infused at higher doses or for prolonged periods, especially in the setting of hepatic and renal dysfunction. Cyanide toxicity may be manifested as abdominal pain, confusion, or seizure and is usually preceded by lactic acidosis. Thiocyanate toxicity usually manifests as nausea, confusion, fatigue, psychosis, and, rarely, coma. If toxicity is suspected, the infusion should be discontinued and serum levels of the metabolites should be measured. Cyanide toxicity can be treated with sodium nitrite (300 mg) or sodium thiosulfate (12.5 g), but thiocyanate toxicity may require hemodialysis.

Nitroglycerin, like nitroprusside, is a potent vasodilator; however, it has dose-dependent effects in the arterial and venous systems. At low doses, it is a relatively selective venodilator, resulting in increased venous capacitance and decreased left and right ventricular filling pressures. At higher doses, it is also an arterial dilator and results in a fall in pulmonary and systemic vascular resistance, although less predictably and to a lesser extent than nitroprusside. Intravenous nitroglycerin is an effective agent in the management of acute decompensated heart failure characterized by increased filling pressures and elevated vascular resistance. In addition, nitroglycerin has a significant vasodilative effect on epicardial coronary arteries and may indirectly improve left ventricular function by improving blood flow to ischemic myocardium. It is thus an agent of choice in managing heart failure associated with acute myocardial ischemia or infarction.

Intravenous nitroglycerin is usually initiated at 20 μg/min and titrated up by 10 to 20 μg/min every 5 to 10 minutes until the desired hemodynamic effect is achieved or the maximum dose is reached (400 μg/min). Its effects are immediate and resolve rapidly after discontinuation of the infusion. Nitroglycerin may result in hypotension, especially at high doses and in patients with low filling pressures. Its use is commonly associated with a headache, occasionally requiring down-titration or discontinuance of the infusion. Nitrate tolerance frequently develops but can usually be overcome by increasing the infusion rate.