CHAPTER 68: PATHOPHYSIOLOGYOF HEART FAILURE

Heart failure (HF) is a complex clinical syndrome associated with increased mortality and morbidity.¹ It is also common and becoming more prevalent with the aging population and improvement in survival of patients with cardiovascular diseases such as coronary artery disease. Although coronary artery disease is the most common cause of HF, patients can develop HF from any number of different cardiovascular diseases (Table 68–1); HF appears to be the final common pathway of ventricular dysfunction that share some important clinical characteristics and pathophysiology. Although the molecular biology and integrated physiology behind HF remain incompletely understood, several fundamental concepts and principles have evolved and are described here. In addition, many of the same proteins found in key HF pathophysiologic pathways have been evaluated as potential biomarkers in HF (some are well established in their clinical utility for diagnostic and prognostic purposes) and are important targets of therapeutic applications and exploration. An important caveat is that many of the proven concepts in HF were derived from HF models with reduced ejection fraction (HFrEF), and relatively little is known about the mechanisms involved in the development of HF with preserved ejection fraction (HFpEF). A special section on the pathophysiology of HFpEF is presented near the end of this chapter. Again, some of the concepts are shared between HFrEF and HFpEF.

A key feature of HF is the impaired ability of the heart to move blood forward to meet the needs of the body, whether it is a result of a dysfunction in its ability to fill, contract, or coordinate with the rest of the body. However, many of the body's responses to cardiovascular abnormalities, which are initially adaptive and maintain short-term circulatory function, eventually become maladaptive; these changes ultimately result in further cardiovascular injury and contribute to the progression of HF. Adaptations, in response to HF, occur not only within the heart, but also in the peripheral circulation, kidneys, skeletal muscle, and almost all organs of the body. There is some suggestion that maladaptive response of the extracardiac structures may contribute to HF development and progression, as may happen in HFpEF. A better understanding of how these changes occur may provide important insight into the complex pathophysiology of the HF syndrome and may help to identify specific therapeutic targets.

HF is generally defined as a complex clinical syndrome in which an abnormality of cardiac function and connected systems is responsible for the failure of the heart to move blood forward at a rate commensurate with the requirements of the metabolizing tissues.²

Starting in 2001, the American College of Cardiology (ACC)/American Heart Association (AHA) working group³ introduced and continued to define a broadened concept of HF to include two additional groups of patients before they develop functional or structural heart disease and symptoms: patients at high risk for developing HF with a structurally normal heart and patients with an abnormal cardiovascular system who have not developed clinical symptoms. A broadened definition, such as this, more accurately describes the spectrum of HF and can potentially lead to earlier interventions based on the specific molecular and hemodynamic abnormalities present before symptoms develop. As a matter of fact, a recent meta-analysis showed that the presence of asymptomatic left ventricular (LV) systolic or diastolic dysfunction was associated with a substantial risk for incident HF.⁴

Once the diagnosis of HF has been determined, the following descriptive categorizations are available with significant overlap and coexistence within the groups: (1) acute versus chronic; (2) left, right, or biventricular; and (3) reduced or preserved ejection fraction (also commonly described as systolic or diastolic dysfunction).

Acute Versus Chronic Heart Failure

Acute decompensated HF is defined as a relatively rapid change in HF symptoms and signs resulting in a need for urgent therapy (Fig. 68–1).⁵ These symptoms can develop in a progressive manner over a short period of time (hours, days, or weeks), sometimes with a defining event (eg, during acute myocardial ischemia, acute myocardial infarction, onset of atrial fibrillation, or a sudden development of mitral regurgitation caused by rupture of a papillary muscle or chordae tendineae) and sometimes with a more nonspecific event such as dietary or medication noncompliance. Regardless of cause, the transition to acute decompensated state is typically preceded by a rise in filling pressures several days to weeks beforehand even in the absence of significant weight changes (Fig. 68–2).^6,7 Acute shift of blood volume from the systemic to the pulmonary circulation (sometimes referred to as flash pulmonary edema) can occur before significant salt or water retention has ensued. Acute HF is distinguished from chronic HF, which refers to a relatively more stable but symptomatic condition, in many cases being considered as “compensated.” In chronic HF, fatigue can occur because of a limited cardiac output and neurologic signals from underperfused skeletal muscle. Fluid accumulation can occur, leading to pulmonary congestion and peripheral edema (ie, “congestive” failure).

Left, Right, or Biventricular Failure

HF can also be described by the predominant ventricle affected, although this is in part an artificial division because the ventricles are functionally and structurally related to each other; left-sided dysfunction can contribute to the right-side dysfunction and vice versa, and the two often coexist. HF with LV dysfunction is classically associated with signs and symptoms of pulmonary congestion, whereas HF with right ventricular dysfunction is typically associated with signs and symptoms of venous congestion, such as abdominal bloating and lower extremity edema. Again, a typical patient presents with both types of symptoms.

Heart Failure With Reduced or Preserved Ejection Fraction

Approximately half of HF patients have HFrEF and half have HFpEF. This is typically based on the ejection fraction of the left ventricle on cardiac imaging. The cutoff LV ejection fraction (LVEF) by which to distinguish HFrEF from HFpEF is not well established, but clinical trials (the same trials that have either supported or refuted a specific HF therapy for use) have used LVEF ≤ 35% to 40% to define HFrEF^8,9 and LVEF ≥ 40% to 50% to define HFpEF.^{10,11,12,13,14,15} It is unclear how exactly to categorize patients with LVEF between 40% and 50%. The only study to include a significant number of such patients, as a part of LVEF ≥ 40% criteria, was the Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM)-Preserved trial,¹² which showed a small and almost significant reduction in the incidence of the primary end point of cardiovascular death or hospitalization for HF with candesartan for the overall study. With any cardiomyopathies, especially for those with LVEF in the lower range between 40% and 45%, it may not be unreasonable to treat patients as having HFrEF, but some have called for these patients to be defined as HFpEF instead, because development of symptomatic HF symptoms at a relatively normal LVEF may suggest a pathophysiology closer to HFpEF. However, definitive evidence is lacking. More recent trials of HFpEF are using higher cutoff values of ≥ 45% to 50% in order to have a more streamlined cohort of HFpEF, but perhaps even more so than with HFrEF patients, patients with HFpEF appear to be a heterogeneous group of HF patients, and a better understanding of different HFpEF phenotypes may be needed.^10,14

Distinction between HFpEF and HFrEF is important because therapies that have a proven mortality and morbidity benefit in patients with HFrEF do not appear to be effective in patients with HFpEF,^{10,12,14,16,17} and to some extent, these conditions may have fundamentally different pathophysiologic mechanisms and phenotypes. However, some believe that negative or neutral results from these trials are more reflective of the heterogeneity of the HFpEF population (some of whom may respond to therapy) more than lack of benefit. For example, the Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist (TOPCAT) trial randomized 3445 chronic HF patients with LVEF ≥ 45% to spironolactone or placebo over 3 years. Although the study was overall negative for the primary composite end point (death from cardiovascular causes, aborted cardiac arrest, or hospitalization for the management HF), there was a significant reduction in the primary composite end point in the subgroup of 981 patients (28.5%) who were enrolled based on the criterion of B-type natriuretic peptide (BNP) ≥ 100 pg/mL or N-terminal pro-BNP (NT-proBNP) ≥ 360 pg/mL rather than a history of HF hospitalization within 12 months (hazard ratio of 0.65 with spironolactone compared with placebo, P = .003). These patients were significantly different from those enrolled based on the hospitalization criterion and tended to be older, were less likely to be current smokers, had worse renal function, had lower potassium levels, and were less likely enrolled in Russia and Georgia. Results of the TOPCAT trial bring to light the fact that many of these negative trials have failed to comprehensively phenotype the HFpEF cohort, and the negative result may in fact be more reflective of the failure to appropriately phenotype HFpEF patients rather than actual lack of efficacy.

Patients with HFpEF have commonly been described as having diastolic HF. Although diastolic ventricular dysfunction is an important contributor to the pathophysiology of HFpEF in many cases, not all cases of HFpEF are driven by diastolic ventricular dysfunction. In addition, diastolic dysfunction is often present in patients with HFrEF.

There is also a group of patients with HFrEF who have partially or fully recovered their LVEF into the normal range with or without treatment. Because the original etiology of the cardiomyopathy is caused by a pathophysiologic process resulting in HFrEF, together with evidence of reversal of benefits to some extent on imaging and biomarker studies when these medical therapies for HFrEF are withdrawn, it is reasonable to continue to treat these patients as having HFrEF unless a new pathophysiologic process occurs.

HF can be caused by a number of different cardiovascular disease processes (see Table 68–1), and the interplay between the causative injury and development and progression of HF is complex. However, several important concepts have emerged with the latest developments in genetics and molecular biology. In addition, these index cardiac or extracardiac injuries can stimulate responses that can be roughly grouped into categories of neurohormonal activation for cellular and molecular responses and remodeling for hemodynamic responses with much overlap and redundancy (Fig. 68–3). Although initially adaptive, continued neurohormonal activation and adverse remodeling can, in turn, result in further cardiac and extracardiac injury (eg, vascular, pulmonary, renal) in a vicious cycle. Relatively little is known regarding what exactly precipitates the transition from a chronic, compensated state to acute decompensated state in many cases. Interestingly, some patients can spontaneously resolve their HF through mechanisms that are not well understood and remain a focus of active exploration.

Mechanism of Cardiovascular Injury and Progression

Altered Cellular Proteins

Alterations are found in the failing heart in numerous contractile proteins, especially in heredity-based dilated cardiomyopathies. In the latter situation, these alterations can interact with additional injuries or abnormal loading conditions to result in HF. Such alterations found in the contractile proteins (myosin and actin), regulatory proteins (troponins and tropomyosin), and cytoskeletal proteins (myosin-binding protein C and titin) are likely to contribute to diminished myocardial performance. In the failing human heart, many changes in gene expression at the messenger ribonucleic acid (mRNA) or protein level have been found in hearts explanted at the time of cardiac transplantation. However, interpretation of end-stage HF findings may be complicated by many factors (eg, receiving multiple inotropic drugs) that can obscure the initial pathogenesis.

Sarcomere Proteins

The fact that more than half of the volume of cardiac myocytes is comprised of contractile proteins speaks volumes about the importance of contractile proteins in cardiac function. Briefly, a myocardial cell contains bundles of myofibril, which is made of functional units of sarcomere. A sarcomere contains myosin (thick) and actin (thin) filaments (Fig. 68–4).¹⁸ Hydrolysis of adenosine triphosphate (ATP) causes myosin and actin to slide past each other, shortening myofibrils and ultimately producing myocyte contraction; this is the foundation of the power of myocardium. Myosin contains both heavy and light chains. Two myosin heavy-chain (MHC) isoforms are present in mammalian heart, α- and β-MHC. The α-MHC is cardiac specific and is more enzymatically active. The less active β-MHC is present in the heart and in slow-twitch skeletal muscle. The distribution of α- and β-MHC is developmentally and hormonally regulated. Mechanical stress, such as pressure overload, induces an α- to β-MHC transition in the ventricles of experimental animals, thus imparting a slower but more economical type of work for the overloaded heart.

There is a general agreement that myofibril function is decreased in the failing human heart, but its causative role remains controversial. Downregulation of α-MHC and upregulation of β-MHC using mRNA measurements from right ventricular endomyocardial biopsies from non-HF and HF patients has been demonstrated, and their sequential changes can be observed with recovery of cardiac performance.^19,20 This alteration, if translated into protein expression, would decrease myosin adenosine triphosphatase (ATPase) enzyme velocity and slow the speed of contraction. Although such adaptive changes may provide a survival advantage in the face of increased load, slower contraction and relaxation could also contribute to abnormal myocardial relaxation (also known as lusitropy). In addition, isoform changes involving both the heavy and the light chains, as have been suggested, can play a role in HF. However, in contrast to smaller mammalian species such as mouse, rat, and rabbit, in which many of mechanistic experiments have been performed, the human ventricle contains a large amount of slow β-MHC to begin with (at least 90%). An isoform shift to increase the large amount of β-MHC already present in the cell may not translate to a significant difference in its function. In addition to myosin heavy- and light-chain isoform switches, other well-understood pathologic events also contribute to altered cardiac function in various forms of HF. Point mutations of virtually all of the sarcomere proteins cause hypertrophic cardiomyopathy. Similarly, mutations in the cytoskeletal proteins that provide the molecular scaffolding for the sarcomere have been found in both dilated and hypertrophic cardiomyopathies. Echocardiographic studies have demonstrated that 20% of first-degree relatives of patients with idiopathic dilated cardiomyopathy have enlarged LV cavities. In addition, some of the circulating autoantibodies specific to the heart and its contractile proteins have been linked to development of HF²¹; a high incidence of circulating anti-heart antibodies is observed among first-degree relatives with dilated cardiomyopathy and can precede development of cardiomyopathy.^22,23

Cell Membrane Ion Channels and Intracellular Calcium Kinetic Proteins

Plasma membrane ion channels initiate excitation-contraction by generating and propagating the action potentials that depolarize the myocardium (Fig. 68–5).²⁴ These ion channels contain several subunits that surround the ion-selective pore. The intracellular calcium (Ca²⁺) release channels are found in the sarcoplasmic reticulum (SR) and are quite different from those of the plasma membrane. The SR Ca²⁺ release channels are referred to as ryanodine receptors (RyRs) and interact with the ligand inositol triphosphate (IP₃). The Ca²⁺ pump ATPases are found in both the plasma membrane and SR. Both calcium pumps are activated by cytosolic Ca²⁺. The sodium (Na⁺)-Ca²⁺ exchanger transports Ca²⁺ out of the cytosol into the extracellular space, using the osmotic energy of the Na⁺ gradient across the plasma membrane to generate active transport. The Na⁺-potassium (K⁺)-ATPase pump uses energy derived from ATP hydrolysis to exchange Na⁺ that enters the cell for K⁺ lost from the cytosol during repolarization. Also, Ca²⁺- binding storage proteins (eg, calsequestrin) maintain a Ca²⁺ store that can be readily used during excitation-contraction coupling. The heart's voltage-gated ion channels (especially inward Na⁺ channels) are altered in the failing heart, as are outwardly rectifying K⁺ channels.^25,26 A common feature of both animal and human models of HF is prolongation of the action potential. Both decreased Na⁺ influx and K⁺ efflux are contributory and are mediated by reduced activity of the myocardial membrane (also known as sarcolemma) Na⁺ and K⁺ channels, respectively. This aberrant channel behavior contributes to the arrhythmias, which are the second most common cause of death of patients with HF.

Excitation-Contraction Coupling Proteins

Excitation-contraction coupling links plasma membrane depolarization to the release of Ca²⁺ into the cytosol, where it binds to troponin C, permitting the force-generating interaction between myosin and actin and ultimately contraction (see Fig. 68–5). Relaxation is also an energy-dependent process, but is not simply a reversal of the steps in excitation-contraction coupling. During relaxation, Ca²⁺ is actively transported out of the cytosol by entirely different molecular pumps. The basic mechanism of cardiac excitation-contraction coupling involves Ca²⁺ entry from the extracellular fluid by means of the voltage-dependent L-type Ca²⁺ channel to produce a trigger in increasing [Ca²⁺]_i and opening of the intracellular SR Ca²⁺ release channel or RyR. Key defects in SR Ca²⁺ uptake and release are present in HF, including dysregulation of RyR, RyR2/Ca²⁺ release channel macromolecular complexes, and the SR Ca²⁺ trans­port proteins, especially the sarcoplasmic-endoplasmic reticulum Ca²⁺ ATPase (SERCA) and phospholamban, a reversible inhibitor of cardiac SR Ca²⁺-ATPase activity.²⁷ Other Ca²⁺-cycling proteins, such as Na⁺-Ca²⁺ exchanger proteins, can also be altered in HF.²⁸ Hyperphosphorylation of phosphokinase A, oxidation, or nitrosylation of RyR2 can lead to SR Ca²⁺ depletion, increased risk of arrhythmias, and impaired contractility. HF is characterized by reduced myosin-actin myofibril activation and decreased Ca²⁺ available for activation as well as heightened cytosolic Ca²⁺ levels in diastole. Some studies have shown increased myosin-actin myofibril Ca²⁺ sensitivity and altered Ca²⁺ kinetics.²⁹ These abnormalities of Ca²⁺ metabolism can affect HF manifestation to varying degrees. Several potential therapeutic strategies to improve utilization of Ca²⁺ are being tested in humans, including those targeting SERCA2a and RyR2, and many more are being tested in animal models.^30,31

Metabolic Adaptations and Maladaptations

The healthy heart works incredibly hard in maintaining forward blood flow every day; it pumps more than 7000 L/d and extracts energy from more than 6000 g of ATP daily.³² Not surprisingly, many abnormalities and inefficiencies in myocardial energy metabolism have been shown to occur in HF; these include altered energy substrate with an increased dependence on glucose, decreased oxidative phosphorylation, and high-energy phosphate and mitochondrial dysfunction.³³ This is a potentially valuable area of therapeutic exploration because decreased myocardial metabolism and resulting dysfunction can result in adverse compensatory remodeling in the rest of the heart. Improving myocardial metabolism, in particular, of stunned or hibernating myocardium (as happens with chronic oxygen deprivation from coronary artery disease) may break this cycle of adverse remodeling.

Substrate Use

The primary substrate of energy in the normal myocardium is fatty acid, with a small contribution by glucose, lactate, amino acids, and ketones.³⁴ Although the myocardial uptake of fatty acids and glucose appears to be preserved in HF, there is a shift from primarily fatty acid metabolism to glucose metabolism through changes in enzymatic regulation. Interestingly, this increased dependence on glucose metabolism is similar to the fetal heart's use of glucose as the primary energy substrate in the relatively hypoxic environment.³⁵ This is likely an initially adaptive mechanism as energy extraction from glucose is much more efficient than from fatty acids; in addition, abnormalities of glucose transport in this setting have been shown to be harmful to the myocardium.³⁶ However, insulin resistance is common in HF and may limit the benefits of a shift in metabolism to glucose.

Energy Production and Storage

Oxidative phosphorylation is the mitochondrial process by which ATP is produced in myocardial cells. In HF, there is a reduction in oxidative phosphorylation capacity, as evidenced by reduced high-energy phosphate levels, such as phosphocreatine and ATP, and an altered phosphocreatine-to-ATP ratio.³⁷ This reduction in dilated cardiomyopathy has been closely associated with future mortality.³⁸

Mitochondrial ATP production is linked to myocardial need by highly efficient shuttling of high-energy phosphates via creatine kinase, adenosine monophosphate (AMP)–activated protein kinase and adenylate kinase, and glycolysis. Dysregulation of any of these enzymatic pathways is associated with myocardial dysfunction; for example, creatine kinase activities are reduced in HF, which results in a decrease in shuttling of high-energy phosphate.

Mitochondrial Dysfunction

Mitochondria are the driving power source of the cell; the majority of ATPs are produced by the mitochondria. There is evidence that suggests that abnormalities in mitochondria may contribute to the development and progression of HF, but many of the areas are incompletely evaluated.³⁹ Genetic mutations affecting key mitochondrial structures, such as mitochondrial ATP synthase and acylglycerol kinase, have been shown to result in myocardial dysfunction.^40,41 Some of the proven HF medications, such as angiotensin-converting enzyme (ACE) inhibitors, are associated with an improvement in mitochondrial function, as evidenced by an increase in high-energy phosphate such as ATP and creatine phosphate, the mitochondrial oxygen consumption rate,⁴² and viable but metabolically stunned myocytes.³³ Dysfunction of the mitochondria includes abnormalities in the production of new mitochondria, enhanced generation of reactive oxidative species in the mitochondria, and dysregulation of mitochondrial iron handling.

Neurohormonal Activation

The neurohormonal system is one of the first activated responses to myocardial injury or to alternations in cardiac loading. Decreased cardiac output (any combination of decreased blood pressure, pulse pressure, and perfusion) is sensed by various mechanoreceptors throughout the body including the left ventricle, carotid sinus, aortic arch, and afferent renal arterioles. When there is diminished activation of these receptors, as in HF, the neurohormonal system is activated in an attempt to maintain cardiac output and vital organ perfusion.

The sympathetic nervous system is activated early, followed by the renin-angiotensin-aldosterone system (RAAS). Nonosmotic release of AVP (also known as vasopressin or antidiuretic hormone)⁴³ and underfilling of renal arterial bed result in Na⁺ and water retention. Heightened peripheral vasoconstriction, increased myocardial contracti­lity and heart rate, and increased blood volume restore cardiac output and arterial pressure in the short term.

However, sustained and unopposed neurohormonal activation has many important adverse consequences at the cellular level, including facilitation of myocyte hypertrophy, collagen synthesis, and fibrosis, as well as promoting apoptosis and return to fetal isoforms of contractile proteins.^{44,45,46,47,48} Cardiac myocyte necrosis also occurs in response to pathophysiologic levels of endogenous and low-dose exogenous angiotensin II infusion.⁴⁹ Activation of the sympathetic nervous system contributes to tachycardia and arrhythmias and can be directly toxic to the myocardium.^50,51 With increased sympathetic tone in HF, increased phosphorylation of the ryanodine component of the L-type Ca²⁺ channel can lead to abnormalities of Ca²⁺ activation.

Although neurohormonal activation is typically in balance with vasodilatory and natriuretic effects of counter-regulatory pathways, such as natriuretic peptides, nitric oxide, prostaglandins, and bradykinin, cardiosensory activity is impaired as HF progresses, and fails to keep unopposed neurohormonal activation in check.⁵²

For unclear reasons, cardiac afferent activity to the central nervous system is reduced, leading to unhindered, efferent excitatory responses from the brain to the periphery. Reflex vasoconstrictor responses to unloading the heart are paradoxically blunted.⁵³ HF parasympathetic (vagal) tone is decreased, and heart rate variability is markedly reduced. Furthermore, decreased heart rate variability can provide independent prognostic value in the identification of patients at risk for cardiac death.⁵⁴ Beat-to-beat QT interval variability on surface electrocardiograms (given relatively stable heart rate) is thought to be closely related to the subtle variations in ventricular repolarization duration as well as sympathetic nervous system activation, most likely through SR Ca²⁺ release.⁵⁵ Increased QT interval variability has been associated with the presence of coronary artery disease, LV hypertrophy, and LV systolic dysfunction, as well as increased risk of death, sudden cardiac death, and ventricular tachycardia or fibrillation shocks in patients with implantable cardiac defibrillators. However, these data are largely based on retrospective analysis, and prospective studies are needed in a well-defined population before routine use can be recommended.

A large number of these neurohormones and their inert counterparts circulate in increased concentrations in HF or with altered activity levels and have been evaluated as potential biomarkers (Table 68–2). Because some of these biomarkers are quite intimately linked with key pathophysiologic processes in the development and progression of HF, they constitute a fertile field for potential diagnostic, prognostic, and therapeutic targets.

As a matter of fact, neurohormonal mechanisms are the targets of several important and successful therapeutic interventions in HF and hypertension, and have a key role in determining prognosis.^56,57 ACE inhibitors, β-adrenergic blockers, and aldosterone antagonists now have a prominent role in the treatment of HF, and new, more innovative neurohormonal-blocking agents (eg, neprilysin inhibitors) are being rapidly developed, adding strong support to the neurohormonal hypothesis.^58,59,60

Autonomic Nervous System Imbalance

Normally, the balance between the sympathetic and parasympathetic system is tightly regulated and maintained. In HF, one of the earliest activated pathways is the sympathetic nervous system, and the withdrawal of parasympathetic tone results in a predominantly sympathetic milieu.⁶¹ Vascular constriction, tachycardia, increased myocardial contractility, diaphoresis, and oliguria are manifestations of increased sympathetic drive, but often at the expense of increased myocardial oxygen demand.

In addition, HF is characterized by abnormal reflex control mechanisms. Peripheral vascular resistance (systemic vascular resistance) is increased; this is likely caused by a combination of locally active heightened vasoconstrictors (norepinephrine [NE], angiotensin II, endothelin, vasopressin, neuropeptide Y) and by structural changes in blood vessels from fluid retention and reduced endothelial-dependent vasodilatation. These later changes are closely associated with limitations of exercise in HF.⁶² There is also defective cardiac parasympathetic control and altered baroreceptor function and reduced cardiac sympathetic activity in response to a variety of stimuli.^{53,63,64,65,66} β-Blockers, shown to improve clinical outcomes in HF, reduce β₁-receptor–related catecholamine activity and oppose sympathetic activation.⁶⁷ In animal models, renal sympathetic denervation, which aims to reduce renal sympathetic activation, was shown to improve survival and LV function as well as decrease myocardial fibrosis.⁶⁸

Norepinephrine

Sympathetic nervous system activation and subsequent increase in plasma NE levels are noted in patients with HF.⁶⁹ Plasma NE levels correlate with functional class and extent of hemodynamic dysfunction.^70,71,72 Although early studies of plasma NE level showed promise as a useful research and prognostic marker for patients with HF, contemporary analyses failed to show incremental benefit of NE use for prognosis beyond other established biomarkers and clinical models such as the Seattle Heart Failure Model.⁷³ Moxonidine, an imidazoline agonist that decreases NE levels in HF patients, failed to improve HF clinical outcomes and, in fact, was associated with increased risk of death and HF hospitalizations.^74,75 Interestingly, myocardial stores of NE were found to be depleted in advanced HF patients, and severe New York Heart Association (NYHA) class IV HF patients may be quite dependent on catecholamine support (ie, continuous dobutamine infusion) to maintain suitable organ perfusion. However, there is no question that high circulating levels of NE are toxic to the myocardium and are partly responsible for progressive adverse LV remodeling.⁷⁶ NE is thus a double-edged sword in HF.

Myocardial Receptor Dysfunction

The failing heart commonly demonstrates a decreased responsiveness to inotropic stimuli. A reduction in myocardial β-adrenergic receptors and the subsequent second messenger cyclic adenosine monophosphate (cAMP) plays an important role.⁷⁷ β-Adrenergic stimulation contributes importantly to the cardiac response to exercise, and β-adrenergic desensitization and uncoupling can be at least partially responsible for the reduced chronotropic and inotropic response to peak exercise commonly found in patients with HF. The β-adrenergic receptor abnormalities in HF appear to be caused by desensitization and uncoupling of the β₁-receptor produced by local rather than systemic alterations in catecholamines. In severe HF, the NE stores in sympathetic nerve endings are depleted. In a sense, the failing myocardium becomes functionally denervated. cAMP responses are reduced by approximately 30% to 35%, leading to further contractile dysfunction. Despite downregulation of the β₁-receptor, a relatively high proportion of β₂-receptors remains to mediate chronotropic and inotropic responses.⁷⁸ However, there is some uncoupling of the β₂-receptor from its G protein and a modest upregulation of the Gα_i subunit, further contributing to a depressed response to chronotropic and inotropic stimuli.⁷⁹ There is also a marked decrease in cardiac β-adrenergic responsiveness with aging, which has clinical implications because HF is heavily concentrated in the elderly population.⁸⁰

The desensitization and uncoupling of β-adrenergic receptors that occurs early with mild-to-moderate ventricular dysfunction is related to the degree of HF and is associated with a reduced response to β-adrenergic stimulation with drugs such as dobutamine. Long-term stimulation of β-adrenergic receptors can enhance myocardial β-adrenergic receptor kinase activity, leading to further desensitization and uncoupling of the β-adrenergic receptor.⁸¹

Of great therapeutic interest, β-adrenergic blockade with metoprolol and bisoprolol, relatively cardioselective β₁-blockers, upregulates the β₁-receptor, but carvedilol, a nonselective β₁^- and β₂-blocker with additional α₁-blocking activity, does not increase β₁-receptor density.^82,83 These drugs improve LV function substantially in approximately two-thirds of patients. The ventricular improvement seen with chronic β-blocker use may not be caused by upregulation of β-adrenergic receptors alone, and the beneficial effects of β-adrenergic receptor blockade in HF remains incompletely explained. Moreover, high plasma NE levels do not predict benefit from carvedilol,⁸⁴ suggesting that there is not a simple relationship between activation of the sympathetic nervous system and response to β-adrenergic blocking drugs in patients with HF.

The Renin-Angiotensin-Aldosterone System

The RAAS plays a pivotal role in the pathogenesis of HF (see Fig. 68–4),⁵⁹ and consistent benefit has been derived from ACE inhibitor, angiotensin II receptor blocker (ARB), and aldosterone antagonist therapy in HF patients.^58,85 Mechanisms responsible for the release of renin from the renal cortex have been exhaustively studied and include increased sympathetic drive in the kidneys, β-adrenergic receptor activation, and actual or perceived hypovolemia with hyponatremia or hypochloremia at the level of renal macular densa or arteriole.^86,87,88 Renin proteolytic enzyme has little biologic activity, but it interacts with angiotensinogen to split off two amino acids to form angiotensin I. This is then cleaved by ACE (distributed widely in the vascular system, especially the lungs) to produce angiotensin II, a peptide with a vast range of biologic activities. Angiotensin II in turn stimulates release of aldosterone from the adrenal cortex, which also has an array of biologic effects, including K⁺ and water retention, renal salt excretion through increased urinary K⁺ loss by affecting the Na⁺-K⁺ exchange, and enhanced collagen turnover and organ remodeling.

There are at least four recognized angiotensin II receptors, but much of the activity is promoted by the AT₁ receptor. AT₁ receptor actions include arterial vasoconstriction, hypertrophy, apoptosis in myocytes, polydipsia, NE release, sensitization of blood vessels to NE, AVP release, and aldosterone release. The AT₂ receptor appears to have somewhat counter-regulatory effects, including antigrowth and antiremodeling, antiapoptosis, vasodilatation, and activation of the kinin–nitric oxide–cyclic 3′,5′-guanosine monophosphate system.⁸⁹ Because AT₁ receptor–blocking drugs (ARBs) increase angiotensin II levels, they can indirectly activate unoccupied AT₂ receptor activity. Angiotensin II levels tend to escape the pharmacologic effects of chronic ACE inhibition, irrespective of dosage, and can stimulate AT₂ receptor activity.⁹⁰ Although all of these components of the RAAS can be measured peripherally, local tissue activities may have a greater impact on the RAAS and reflect true biological activity of these components. Both ACE inhibitors and ARBs reduce cardiovascular events, and the role of the RAAS in the pathogenesis of heart and vascular disease, including progressive HF, is quite significant and remains an active area of investigation for therapeutic options.

Arginine Vasopressin

HF is frequently characterized by water retention in excess of Na⁺ retention, leading to hyponatremia. The hyponatremia is caused in part by the nonosmotic release of AVP from the neurosecretory cells located in the hypothalamus; AVP acts on the kidneys to reduce free water clearance and to promote vasoconstriction. Release of AVP in HF probably occurs by means of activation of carotid baroreceptors.⁹¹ Plasma AVP levels are often, but not always, increased in patients with LV dysfunction and HF.^92,93 AVP acts on the V₂ receptors in the collecting duct of the kidney via adenylate cyclase to translocate aquaporin-2 water channels from cytoplasmic vesicles to the apical surface of the collecting duct. AVP also increases aquaporin-2 synthesis. Activation of V₁ receptors in vascular tissue contributes to heightened vascular resistance and myocardial dysfunction in HF.⁹⁴ Hyponatremia is a powerful predictor of poor outcome in HF, and elevated AVP levels in HF patients are associated with severe HF. Reliable measurement of circulating AVP is challenging because of a short half-life, but the C-terminal portion of the precursor of pro-vasopressin, copeptin, has been shown to be a strong predictor of cardiovascular outcomes in HF.⁹⁵ Despite its association with HF prognosis, therapeutic effects from modulation of AVP may be challenging. For example, tolvaptan, an oral AVP antagonist that improves hyponatremia, failed to improve HF mortality.^96,97

Natriuretic Peptides

A family of natriuretic peptides, including atrial natriuretic peptide (ANP), BNP, and C-type natriuretic peptide (CNP), is encoded by separate genes, each with a tissue-specific distribution, regulation, and biologic activity.⁹⁸ ANP and BNP are often increased in patients with HF, and this has been leveraged to assist in the diagnosis of acute HF.⁹⁹ ANP is a 28-amino-acid peptide that is normally synthesized and stored in the atria and to some extent in the ventricles. It is released into the circulation during atrial distension. BNP is synthesized mainly by the ventricles and is released in LV dysfunction or early HF after cleavage (presumably by a protease, corin or furin) from the propeptide to BNP and NT-proBNP (Fig. 68–6).¹⁰⁰ For the most part, these peptides act via guanylate cyclase receptors to promote vasodilatation (ANP, BNP, CNP) and natriuresis (ANP, BNP). The natriuretic peptides, ANP and BNP in particular, are considered counter-regulatory because they tend to reduce right atrial pressure, systemic vascular resistance, aldosterone secretion, sympathetic nerve stimulation, RAAS activity, and hypertrophy of cells and can enhance Na⁺ excretion when infused.⁵²

BNP is removed from circulation by a number of mechanisms, including passive clearance by high-flow organs, such as kidneys; membrane-bound, receptor-mediated clearance; and enzymatic processes including neutral endopeptidases, meprin-A, and dipeptidylpeptidase-4.^101,102

Interestingly, data suggest that patients with HF may in fact have low circulating levels of biologically active BNP despite elevated detected levels of BNP and NT-proBNP; this is because commercially available BNP and NT-proBNP assays detect a mixture of the intended target BNP or NT-proBNP, pro-BNP, and degraded BNP fragments in the overall measured level.¹⁰³ Regardless, both BNP and NT-proBNP have proven diagnostic and prognostic value in HF as biomarkers (ACC/AHA HF guideline recommendation Class I), and their potential role in guiding HF therapy is being evaluated.^{99,104,105,106}

Recently, drugs designed to inhibit degradation of natriuretic peptides (neutral endopeptidase or neprilysin inhibitor), among other effects, have been combined with ARBs (sacubitril/valsartan) and have shown mortality and morbidity benefit in HF.⁵⁸

Endothelin

Endothelins are a family of vasoconstrictor peptides produced by vascular endothelial cells and may play a role in HF and pulmonary hypertension. Although blood levels are increased in patients with HF,¹⁰⁷ endothelin-1 (ET-1) is more of a paracrine than an endocrine hormone.¹⁰⁸ Myocardial tissue ET-1 levels may be increased in HF more as a result of decreased clearance rates than as a result of increased synthesis. Endothelial cells synthesize ET-1 rapidly and convert so-called big ET-1 into ET-1 by an endothelin-converting enzyme. Both elevated ET-1 and big ET-1 have been shown to be associated with adverse clinical outcomes in chronic HF patients,¹⁰⁹ but the incremental role of adding ET to established markers of prognosis remains unclear. The synthesis of ET-1 is enhanced by angiotensin II, NE, growth factors, insulin, hypoxia, oxidized low-density lipoproteins, shear stress, and thrombin.¹¹⁰ Its synthesis is antagonized by ANP and prostaglandins.

Endothelin acts on at least two types of G protein–coupled receptors, A and B. The ET-A receptor increases smooth muscle vasoconstriction, cell proliferation, and hypertrophy and mainly resides on vascular smooth muscle cells. The ET-B receptor, which is mainly endothelial, promotes vasodilatation that is probably mediated by a variety of mechanisms, including increased production of nitric oxide (NO) and prostaglandins and activation of K⁺ channels. ET-1 can also act on the heart to cause hypertrophy, on the adrenal gland to release aldosterone, and on the kidney to promote Na⁺ and water retention. However, blocking the effects of ET-1 has not been shown to improve symptoms or prognosis in HF in contrast to its beneficial effects in patients with pulmonary arterial hypertension.^111,112

Inflammatory Responses and Remodeling

Cellular Responses and Cytokines

It is now widely recognized that there is a strong association between inflammation and cardiovascular diseases. Many of the biomarkers of inflammation, such as erythrocyte sedimentation rate, C-reactive protein, tumor necrosis factor-α (TNF-α), Fas, and interleukin (IL) 1, 6, and 18, are recruited after myocardial injury, such as myocardial infarction, and are thought to be involved in the initiation of the repair process and remodeling as well as cardiac cachexia in certain subgroups of patients with injury.^{113,114,115,116,117} These same biomarkers are also elevated in HF and associated with poor HF prognosis independent of traditional risk factors and baseline characteristics.¹¹⁸ There is increasing evidence that a systemic inflammatory state caused by comorbidities may be intricately involved, especially in HFpEF, but more information is needed before a causative statement can be made.¹¹⁹ Practical application of these biomarkers is challenging because of the nonspecific nature of inflammation and difficulty in adjusting for potential covariables that may affect them.

Elevated circulating levels of TNF-α can mediate cell growth, negative inotropy, and apoptosis. In cachectic HF patients, TNF-α has been associated with a marked activation of the RAAS.¹¹⁵ Disappointingly, inhibitors of circulating TNF-α have not been shown to alter the outcome of HF in humans despite the beneficial effects in animal studies.¹²⁰ Another potential target of interest is the inflammasome. The inflammasome is a macromolecular structure that plays a central role in the inflammatory response to injury by first sensing the injury and then amplifying the inflammatory response by activating powerful cytokines. This activation of the inflammasome in the heart in the setting of injury may promote adverse ventricular remodeling and HF.¹²¹ In a mouse acute myocardial infarction model, infusion with Nod-like receptor P₃ (NLRP₃) inhibitor resulted in > 40% reduction in infarct size and > 70% reduction in troponin I levels (Fig. 68–7), but no clinical trials have been done to date.¹²²

Nitric Oxide, Endothelial Dysfunction, and Oxidative Stress

Reactive oxygen species (ROS) are produced as a byproduct of normal aerobic metabolism and are involved in a variety of key cellular signaling and immune responses, such as myocardial excitation-contraction coupling and myocardial growth. Typically, free ROS are scavenged and neutralized by antioxidants, but when this balance between ROS and antioxidants is thrown off, these species can promote progressive cellular injury or oxidative stress.¹²³ Prolonged exposure of ROS can lead to activation of specific pathways of oxidant stress with available substrates, such as NO, leading to enzyme activation and deactivation, DNA breakage, and lipid peroxidation, which are ultimately responsible for the subsequent myocyte apoptosis and pathologic remodeling seen in HF. Growing evidence has supported the role of heightened oxidative stress in the pathophysiology of HF.¹²⁴

Biomarkers of oxidative stress such as myeloperoxidase, albeit nonspecific, are often elevated in patients with chronic HF.^125,126 Markers of oxidative stress were positively correlated to HF severity irrespective of its underlying cause.¹²⁴ However, the precise mechanisms contributing to oxidative stress in HF are multifactorial. Catecholamines and angiotensin II promote pro-oxidant activities within the myocardium, including enhanced superoxide (O₂^–) formation and generation of NO-derived oxidants, as a necessary consequence of increased myocardial oxygen consumption. Several studies have also observed an association between insulin resistance and HF in the absence of overt diabetes.^127,128 Animal and human studies indicate that endothelium-dependent vasodilatation is abnormal in a number of disease states, including atherosclerosis, hypertension, HF, hyperhomocysteinemia, insulin resistance, and hypercholesterolemia. Endothelial dysfunction has been demonstrated in patients with HF.¹²⁹ As already discussed briefly, such endothelial dysfunction in HF (ie, failure to vasodilate in response to a specific endothelial-dependent vasodilator or exertion) can be caused by a reduced release of NO during stimulation. The basal release of NO can be preserved or even enhanced in HF and can be compensatory by antagonizing neuroendocrine vasoconstrictor forces.¹³⁰ However, impairment of endothelium-dependent peripheral vasodilatation can be a factor contributing to exercise intolerance in patients with chronic HF, perhaps by limiting nutritive skeletal muscle flow during exercise.¹³¹ This dysfunction of the endothelium can be related to deconditioning in later stages of HF and, with training, it can be largely reversible. The roles of NO and NO synthase (NOS) in the failing heart are much more complex, and new data are emerging.^132,133 NO has been considered to be present as a freely diffusible molecule, which inhibits the positive inotropic response to β-adrenergic stimulation in the failing heart.¹³⁴ Smaller physiologic amounts of NO produced by constitutive NOS (cNOS or NOS3) are necessary for normal function and have an antioxidant effect that can protect cells. The inducible isoform of NOS (iNOS or NOS2) is overexpressed in human HF and, therefore, can contribute to worsening HF because high levels of NO in the heart can induce proapoptotic and cytotoxic effects.¹³⁵ NO can act in subcellular signaling compartments or modules. Furthermore, covalent modification of cysteine thiol moieties of proteins (S-nitrosylation) can be an important second messenger signaling process, working in parallel with ROS and nitrogen species.¹³⁶ Therefore, disruption of this cyclic guanosine monophosphate signaling process can lead to nitroso-redox imbalance by either increased formation of ROS or decreased production of reactive nitrogen species.¹³⁷

As HF progresses, there are important changes in the peripheral circulation, particularly downregulation of the NO system. Under normal conditions, NO permits peripheral arteriole dilatation and increased peripheral blood flow in response to exercise. As this system downregulates, peripheral dilatation does not occur with exercise; the lack of appropriate blood flow limits exercise and decreased functional capacity. With exercise training, this system can be restored and clinical status improved.

Extracellular Matrix Changes

The extracellular matrix in the heart provides the scaffolding within which contractile cardiomyocytes are housed; it contains a basement membrane, collagen network, proteoglycans, and glycosaminoglycans. Of the different types of collagens, type I and III collagens are the predominant forms found in fibrils deposited in scar tissue after myocardial injury, more specifically demonstrated in myocardial infarction models. These collagens are initially synthesized by cardiac fibroblasts as procollagen precursors before both the N-terminal and the C-terminal are cleaved by proteinases, and then the resulting tropocollagen is assembled into mature fibrils. Markers of collagen turnover, such as serum N-terminal type III collagen peptide (PIIINP), have been associated with increased death and hospitalization rates, and procollagen type I and PIIINP levels appeared to decrease following aldosterone antagonist therapy in chronic HF patients.¹³⁸ In the 967 Framingham subjects without HF, PIIINP levels were not independently associated with LV mass, fractional shortening, end-diastolic dimensions, or left atrial size.¹³⁹ The role of PIIINP in the pathophysiology of HFpEF has not been evaluated to date.

The extracellular matrix is a rather dynamic system that is constantly turned over. In the setting of cardiac or extracardiac injury, regulation of extracellular matrix likely plays an important role in ventricular remodeling and fibrosis. For example, bone morphogenetic protein 1, a C-proteinase, plays a crucial role in the processing of extracellular matrix proteins and collagen deposition and regulation of excessive collagen deposition in fibrosis after tissue injury.¹⁴⁰ Recent studies have found that gene expression of tissue inhibitor of metalloproteinases 1 (TIMP-1) and matrix metallopeptidase 9 (MMP-9) was significantly increased in the border zone of myocardial infarct models as well as ischemic HF models in rats, and that treatment with antifibrotic therapy can prevent the upregulation of MMP-9, ultimately leading to suppression of collagen deposition.^141,142 Interestingly, concentrations of TIMP-1 appeared to correlate with diastolic LV dysfunction,¹⁴³ In a multimarker analysis of HF patients, a panel that included TIMP-1 as well as NT-proBNP, hs-TnT, growth differentiation factor 15, and insulin-like growth factor-binding protein 4 had the best performance in predicting all-cause mortality at 3-year follow-up.¹⁴⁴

Soluble ST2

Soluble ST2 (sST2) initially gained attention as a marker of inflammation, cell proliferation, and autoimmune diseases involving T-helper type 2 lymphocyte responses. Since then, the close relationship between the failing heart and sST2 has come to light. In response to the biomechanical stretch of cardiomyocytes and cardiac fibroblasts, the ST2 gene is strongly upregulated, and circulating sST2 levels increase. sST2 interacts with circulating IL-33, a protein with antifibrosis and antiremodeling effects when it binds to the membrane-bound ligand receptor (ST2L), and appears to mediate the heart's ability to adapt to biomechanical stress. In experimental models, interruption of ST2L expression or infusion of high concentrations of sST2 leads to unchecked ventricular hypertrophy, fibrosis, remodeling, and higher risk for death.¹⁴⁵ Elevated concentrations of sST2 have been closely associated with the phenotype of cardiac decompensation and remodeling and very strongly linked to adverse clinical outcomes in HF. sST2 testing for risk prediction in both acute and chronic HF recently received a Class IIb recommendation in the recent ACC/AHA HF guidelines.³ Studies have shown that sST2 adds independent and additive information to clinical assessment and natriuretic peptides.¹⁴⁶

Galectin-3

A member of the lectin family closely involved in immune modulation, galectin-3 is found in a variety of cells and tissues, including the heart. Galectin-3 is involved in the initiation of the inflammatory cascade following cardiac insult and contributes to ventricular remodeling by the way of tissue repair, myofibroblast proliferation, and fibrogenesis. The galectin-3 gene is induced in animal HF models, and when galectin-3 is instilled in the pericardium, there is a considerable increase in collagen deposition.¹⁴⁷ Murine gene knockout models showed that lack of galectin-3 offered partial protection from LV pressure and volume overload with a slower progression to ventricular dysfunction and HF. Circulating values of galectin-3 are elevated in acute and chronic HF, and data suggest that they may identify a high-risk HF phenotype resistant to conventional HF management.^148,149,150 Galectin-3 testing also received a Class IIb recommendation in the ACC/AHA HF guidelines for risk prediction in HF,⁸⁵ but in a head-to-head analysis, galectin-3 testing failed to show independent and additive value to sST2 in a multivariable model; when sST2 was added to a baseline model that included age, sex, LVEF, estimated glomerular filtration rate, NYHA class, diabetes mellitus, ischemic etiology HF, hemoglobin, Na⁺, β-blocker treatment, ACE inhibitors or ARB, and NT-proBNP for all-cause mortality at 5 years, there was a significant improvement in the C-statistic from 0.757 to 0.770 (P = .004), whereas addition of galectin-3 did not significantly change the C-statistic in a discrimination analysis (C-statistic 0.760, P = .14).¹⁵¹

Myocyte Regeneration and Apoptosis

A normal human heart weighs approximately 300 g and contains approximately 2 × 10¹⁰ myocytes. Until recently, the accepted understanding was that myocyte cell division in the human heart ceased a few weeks after birth.¹⁵² Thereafter, enlargement of the heart was a result of cell hypertrophy or the laying down of collagen in the extracellular space. That view was based on the observations that cancer in the heart is extremely rare, DNA turnover is almost undetectable except in pathologic states, and pathologists disputed whether mitotic figures of myocytes had ever been seen. Approximately 20% of myocytes in the human heart have two nuclei, so that cell separation, rather than mitosis, could bring about a small increase in the total cell number. The precise mechanism for the suppression of cell division in the human heart is unknown. What is established is that damage to human heart muscle results in fibrosis and any repair process, if it were to exist, is insufficient to overcome the rate of loss of cells. In some species, such as the salamander and zebra fish, damage to myocardium is repaired by the generation of new myocardial cells, and the process is under genetic control. If the major problem in the worsening of HF is the continuing loss of myocytes, then a new approach is to either inhibit the loss of cells through the processes of apoptosis or necrosis or to promote the growth of new cells.

Apoptosis is increased in pressure overload LV remodeling and post–myocardial infarction animal models as well as in humans (over 200-fold increase in explanted hearts).^153,154 Rodent models appeared to show some promise in apoptosis as a potential target for therapy.¹⁵⁵ However, the extent to which apoptosis contributes to progressive LV dysfunction in HF patients is unclear because the absolute apoptosis rates are quite low at 0.08% to 0.25%, compared with 0.001% to 0.002% in controls.¹⁵⁶

Growth Differentiation Factor-15

A member of the transforming growth factor-β family, growth differentiation factor-15 (GDF-15) is involved in regulating response to injury in a broad array of tissues, including the heart. In the setting of ischemia or pressure overload, the gene for GDF-15 is strongly induced, and the resulting protein appears to modulate myocardial strain, remodeling, and apoptosis.¹⁵⁷ Circulating GDF-15 levels are increased in HF patients and correlate with prognosis, but their ability to add independent information over existing biomarkers and clinical variables was minimal.¹⁵⁸

Stem Cell and Gene Therapy

For the past few decades, much effort has been spent in an effort to increase the number of myocytes in human myocardium either by introducing other cells (eg, cells derived from bone marrow, induced pluripotent stem cells), which are intended to transform into cardiac myocytes, or by the stimulation of resident primitive cells within cardiac muscle. Most of the randomized controlled stem cell trials to date have been limited by their small size. Recent meta-analyses evaluating stem cell trials in post–myocardial infarction and HF patients showed a small improvement in ejection fraction in post–myocardial infarction patients, but overall failed to show significant improvement in mortality or morbidity (reinfarction, hospitalization, restenosis, and target vessel revascularization) with stem cell therapy.^159,160,161 However, results are limited by the heterogeneity of individual trials, including stem cell types, cell modification, and dose, frequency, and method of delivery. Several studies published after these meta-analyses appear to show promise. A recent promising study demonstrated individual and synergistic benefit of bone marrow–derived mesenchymal stem cells and c-kit–positive cardiac stem cells in a swine ischemia/reperfusion injury model. Both cardiosphere-derived cells via intracoronary injection and mesenchymal stem cells via transendocardial stem cell injection also showed regenerative effects in patients with ischemic cardiomyopathy.^162,163

Experts continue to encourage further exploration in stem cell therapy because the potential benefit from targeting specific mechanisms contributing to progressive LV dysfunction is large. In addition, the primary benefit of stem cell therapy may not be from the adult stem cell differentiating and regenerating into the target tissue, but rather from their paracrine activities, such as secretion of numerous cytokines and growth factors, including those that modulate inflammation, adverse LV remodeling, collagen deposition, angiogenesis, apoptosis, mitochondrial dysfunction, and microvascular dysfunction.^{164,165,166,167,168,169,170,171}

In the largest gene transfer study to date in the HF population, intracoronary injection of adeno-associated virus 1 (AAV1)/SERCA2a failed to improve clinical outcomes in patients with HFrEF despite promising previous studies. However, the gene therapy was not associated with serious safety issues and may assist in future gene therapy trial designs.¹⁷²

Cardiac Remodeling

Cardiac remodeling is defined as cardiac structural changes that happen in response to cardiovascular injury, neurohormonal activation, or abnormal hemodynamic loading conditions. Although initially adaptive, when sustained, remodeling can contribute to the development and progression of HF.

Physiologic Adaptations and Maladaptations

Oxygen deprivation, which is most often caused by coronary artery disease, results in impaired relaxation and weakened contraction, as can be seen in transient angina pectoris. This is readily reversible. With prolonged ischemia, decreased contraction (hypokinesis or akinesis) can persist for hours beyond return of blood flow (stunning). If coronary blood flow is chronically reduced, the myocardium can fail to contract normally even if there is no myocardial necrosis (hibernation). With a more serious loss of flow, infarction can occur; the time to infarction can be quite variable as a result of factors such as collateral coronary artery blood flow and myocardial oxygen requirement (also called myocardial demand). All of these stages can produce significant myocardial dysfunction for which the unaffected myocardium will sustain this load. The result is myocardial hypertrophy (increase in myocardial cell size with or without an increase in myocardial mass) of the nonaffected portion of the ventricle; if this is inadequate, an increase in ventricular volume and pressure occurs using the Frank-Starling mechanism to sustain stroke volume (Fig. 68–8).

In HF, myocardial demand may be increased for a variety of reasons, including increased myocardial wall tension caused by myocardial hypertrophy and inefficient contractile energy metabolism. This increase can result in extraction of a greater amount of oxygen from each unit of coronary blood flow and widened coronary arteriovenous oxygen differential. Although significant reduction in coronary perfusion as the primary etiology of HF (in the absence of obstructive coronary artery disease) has not been documented, coronary microvascular flow is often impaired in acute dilated cardiomyopathy.¹⁷³

Even when myocardial perfusion is normal at rest, exercise can bring out significant impairment in HF patients including widened coronary arteriovenous oxygen difference caused by exercise-induced LV dilatation, impaired coronary vascular bed dilation during reactive hyperemia in the setting of LV hypertrophy and elevated filling pressures (which can normally increase up to five times), and tachycardia-induced reduction in diastolic filling time for coronary arteries.¹⁷⁴

Force-Frequency Response in Heart Failure

Normally, an increase in the frequency of stimulation and heart rate is accompanied by an increased rate of force development, a decreased duration of contraction, and an enhanced rate of relaxation (Bowditch effect). This tends to preserve or increase contractile force while preserving diastolic filling time. In the setting of HF, an increase in stimulation and heart rate is accompanied by a decrease in the rate of myocardial performance; some impairment of systolic function can be related to impaired LV filling as well as negative inotropic effect as a result of alterations in intracellular Ca²⁺ handling. Further reduction in contractile force and inability to decrease duration of contraction can worsen myocardial performance during exercise.^175,176

Hemodynamic Perturbations: The Hemodynamic Hypothesis

In the early stages of HF, the ventricular end-diastolic pressure (EDP) and cardiac output can be normal at rest. However, with exercise or stress (and the accompanying increase in cardiac output and afterload caused by increased blood pressure and heart rate), there is an increase in EDP. The ability to increase cardiac output in response to the increase in oxygen consumption is also reduced. As HF progresses, the resting EDP increases and end-diastolic and end-systolic volumes increase. This results in reduced elastic recoil of the ventricle during diastolic relaxation and is reflected in loss of rapid early diastolic ventricular filling (as revealed by a reduced E wave of the echocardiogram). This helps to further increase the mean diastolic pressure. The elevated ventricular diastolic pressure increases pulmonary venous and capillary pressures and decreases pulmonary compliance. These changes, in combination with increased volume status, can result in pulmonary edema and clinical symptoms of dyspnea.

HF can develop from pressure or volume overload. In pressure overload, myocytes hypertrophy to overcome the increased demands of the excess load. Hypertrophied cells contract and relax more slowly and can be subject to metabolic limitations. In addition, hypertrophied myocardial cells can have a shortened life span. This is of considerable prognostic importance because cardiac myocytes appear to have a reduced capacity to proliferate. When age-related myocyte loss is added to the picture, particularly in association with a decrease in myocyte contractile activity, diastolic dysfunction can occur. As the process continues, ventricular dilatation can occur with systolic dysfunction as well. Loss of myocytes—whether segmental, as in acute myocardial infarction, or diffuse, as in dilated cardiomyopathy—sets up a vicious cycle that leads to reactive hypertrophy in the remaining myocytes. As compensatory hypertrophy becomes more marked in some disease states, the contractility unit of the myocardium often declines because of molecular changes in the heart's contractile proteins and activation system. This is especially likely to occur in response to pressure overload, as in systemic arterial hypertension or aortic stenosis, but also ensues when myocytes are lost from any mechanism.

The Frank-Starling Law and Heart Failure Hemodynamics

The volume of blood that the heart moves forward can be described as cardiac output (usually described as liters per minute), which depends on both the stroke volume (SV; the base unit of the volume of blood moving forward per heartbeat and usually described as liters per beat or milliliters per beat) and heart rate (beats per minute). Although many factors contribute to SV for a given heart, in general, SV depends on three major factors: preload, contractility, and afterload. Preload describes the end-diastolic ventricular volume (EDV) and is determined by the volume of the blood circulating as well as the ability of the ventricle to relax and fill. Contractility describes the force generated by shortening of the sarcomere or the ability to pump blood forward. Afterload is the opposing force to flow, such as peripheral arterial pressure, that the moving blood encounters once it is pumped out of the heart. LV hemodynamics are described next, but several of the concepts can be applied to the right ventricle with some important differences; for example, the right ventricle is not meant to withstand a high-pressure system, and its ability to compensate for acute changes in pressure is much more limited.

The Frank-Starling law describes the increase in SV when EDV is increased and is critical for the response to LV systolic dysfunction (see Fig. 68–8). Inadequate emptying of the left ventricle leads to increased EDV (also referred to as increased preload), sarcomere lengthening as a result of LV stretch from increased EDV, and an increase in SV during the next contraction. For any given amount of Ca²⁺ released into the myocyte, there is increased cross-bridge formation and enhanced sensitivity of the myofilament to Ca²⁺ as the sarcomeres lengthen.

In systolic LV dysfunction, the extent of shortening for a given diastolic fiber length and afterload is reduced. The left ventricle can initially maintain a normal or near-normal SV with an increased EDV and thus maintain end-diastolic fiber length. Eventually, the filling pressure increases inordinately, limiting this compensatory mechanism. The dilated left ventricle requires more tension in the LV walls to generate the same pressure, as per the Laplace relationship, which describes that the pressure inside the ventricle is inversely related to the radius of the ventricle and directly related to the tension in the LV wall. Furthermore, the dilated left ventricle can lose elasticity, like an overstretched elastic band, and EDV can increase somewhat without an increase in LV EDP, reflecting a shift in the passive pressure-volume curve to the right. An obligatory reduction in ejection fraction occurs when SV is maintained in the face of a large EDV (ejection fraction = SV/EDV). Eventually, further increases in EDP produce little change in EDV, thus flattening the SV-EDP curve. There is no true descending limb to the Frank-Starling curve because increasing preload indefinitely will ultimately lead to mitral regurgitation, displacing the increased pressure. As the heart dilates, the increase in wall stress according to the Laplace relationship will also increase afterload, which can account for any observed reduction in SV as the heart dilates further (ie, the perception of a descending limb). It is important to keep in mind that LV performance depends not only on systolic pump function, but also on active relaxation, diastolic elastic recoil, passive diastolic properties, and vascular loading conditions. It is likely that at high LV EDP, valvular incompetence (mitral regurgitation) is a major cause of a decrease in cardiac output. Thus, in end-stage HF in the intact circulation, the Frank-Starling curve flattens.

Loading Conditions and the Concept of the Laplace Relationship

A characteristic feature of the dilated, failing heart is that it gradually becomes less sensitive to preload (EDV and fiber length) and more sensitive to afterload stress. At very high LV filling pressures (> 30 mm Hg), when the sarcomeres are fully stretched and the preload reserve is exhausted, the SV becomes exquisitely sensitive to alterations in the afterload. The impedance to ejection includes blood viscosity, vascular resistance, vascular distensibility, and myocardial wall tension. Much of the afterload is made up of ventricular myocardial wall tension. In the ventricle, the tension on the walls increases as ventricular chamber volume increases, even if intraventricular pressure remains constant. As the ventricle empties, tension is reduced, even as pressure increases. Calculations of myocardial wall tension are defined by the Laplace equation and are expressed in terms of tension, T, per unit of cross-sectional area (dynes per centimeter [dyn/cm]).

Within a cylinder, the law of Laplace states that wall tension is equal to the pressure within a thick-walled cylinder times the radius of curvature of the wall:

where T is wall tension (dyn/cm), P is pressure (dyn/cm²), R is the radius (cm), and h is wall thickness. Wall tension is proportional to the radius. Because the heart has thick ventricular walls, wall tension is distributed over a large number of muscle fibers, thereby reducing tension on each.

Because the geometry of the ventricles is more complex than that of a cylinder, ventricular wall tension cannot be measured with precision. Wall stress, the force distributed across an area, is actually more correct, but is seldom measured.

Two fundamental principles stem from the relationship between the geometry of the ventricular cavity and the tension on its muscular walls: (1) dilatation of the ventricles leads directly to an increase in tension on each muscle fiber, and (2) an increase in wall thickness reduces the tension on any individual muscle fiber. Therefore, ventricular hypertrophy reduces afterload by distributing tension among more muscle fibers.

The wall tension is highest in the inner surface of the heart. The endocardial surfaces must do more work and, therefore, are more vulnerable to reductions in coronary blood flow. Dilatation of the heart decreases cardiac efficiency as measured by myocardial oxygen consumption, unless hypertrophy is sufficient to normalize wall stress. In HF, wall tension (or stress) is high, and thus, afterload is increased. The energetic consequences of the law of Laplace can have some role in progressive deterioration of energy-starved cardiac myocytes in the failing heart.

Another major disadvantage of the dilated ventricle is the inability to decrease the average radius during contraction. In the normal heart, wall tension falls during ventricular ejection as the volume decreases, even though pressure is rising. In HF, given the dilated heart with reduced ejection, the average tension in the myocardial fibers actually can continue to increase from the beginning of the ejection until peak systolic pressure is reached, adding additional afterload during ejection. The rate of myocardial fiber shortening is reduced, further contributing to diminished myocardial performance. It is difficult to overstate the importance of the law of Laplace when considering the syndrome of HF. This contrast is apparent in mitral insufficiency. With preserved contractility and a relatively small EDV, mitral insufficiency leads to rapid unloading of volume and reduced tension. When ventricular dilatation occurs with decreased ventricular contractility, ejection is reduced, and tension remains high during systole, leading to an unsteady state that cannot be maintained for long.

Overstretched myocardial cells can induce programmed cell death (apoptosis), thereby contributing to further disease progression.^177,178 The plasticity of progressive dilatation is now more apparent, with remarkable reversal of dilatation observed in response to ACE inhibitors and β-blockers, cessation of alcohol use in patients with alcoholic cardiomyopathy, and spontaneous improvement in patients with inflammatory myocarditis.

Myocyte Response to Altered Loading Conditions

In response to increased load, whether created by increased pressure or loss of myocytes, hypertrophy occurs and tends to normalize the load per cell. With an increased volume load, myocytes elongate and, to a small extent, may undergo division.¹⁷⁹ Hyperplasia and apoptosis of myocytes occur with abnormal loading, but involve less than 1% of the cardiac myocytes. Reprogramming of the cardiac myocytes occurs, resulting in a more fetal-like state. More BNP is synthesized. Metabolism begins to favor glucose over free fatty acids. The myocytes enlarge, presumably rendering a short-term structural and functional advantage.¹⁸⁰ The reprogramming requires altered signals, both mechanical and chemical, to reach the nucleus of the cardiac myocyte to set into motion new gene transcription.⁴⁶ Contractile proteins are altered and improve energetic efficiency. Ultimately, there is a transition from hypertrophy to HF,¹⁸¹ which has been recognized for a long time, but is still not well understood. In a sense, myocyte hypertrophy leads to the structural changes of LV remodeling, thus creating a large, dilated, and poorly functioning heart. The processes of cellular remodeling and subsequent architectural changes in cell and chamber size and shape are highly complex and include many components other than myocardial cell hypertrophy.⁴⁶ Myocardial fibrosis and cell dropout occur, and perhaps myocyte slippage can occur, increasing dilatation. As cardiac output falls, multiple neurohormones, including renin and NE, are released in an attempt to preserve blood pressure and organ perfusion,¹⁸² and atavistic counter-regulatory natriuretic peptides (such as BNP) are released in an attempt to offset vasoconstriction, hypertrophy, and volume conservation.⁵² Multiple molecular mechanisms,^46,183 some of which primarily affect the cardiac interstitium and others the cardiac myocytes, are involved in the process.

Maladaptive remodeling of cardiac myocyte size and shape begins long before clinical HF begins.^184,185 Alterations in myocyte proteins and mitochondria size, as well as changes in myocardial interstitium and collagen content and architecture, are seen in response to a variety of injuries, including pressure overload, volume overload, and myocardial ischemia.^185,186 Additional phenotypic changes in HF include apoptosis.^153,187 It is important to recognize that much of the neuroendocrine activation that occurs in a primordial attempt to conserve organ perfusion appears to facilitate this myriad of pathologic changes in the heart at the cellular level, thereby possibly contributing to the success of neuroendocrine blockers as therapy for HF. Lastly, there is no single phenotypic change, protein expression, or signal transduction pathway that is dominant. Rather, there is extraordinary redundancy in these mechanisms. This observation has important implications for therapy. For example, blocking one neuroendocrine system can lead to enhanced overactivity of other neuroendocrine systems.

There are many etiologies where a patient may have the clinical syndrome of HF and preserved ejection fraction (Table 68–3). For this section, we will concentrate on the first category of HF where there are no clear etiologic drivers, such as aortic stenosis, and the diagnosis is currently primarily a diagnosis of exclusion. Not surprisingly, a growing body of evidence suggests that this category of HFpEF may in fact be a rather heterogeneous collection of different HFpEF phenotypes (Fig. 68–9). Many trials did not phenotype the patient cohort well, and this may, in part, explain the lack of demonstrated therapeutic benefit in large phase III trials of nontargeted therapies to date.¹⁸⁸ Incomplete understanding of the pathophysiology of HFpEF and lack of clear diagnostic criteria further add to the challenges surrounding HFpEF, and better understanding is direly needed. Nonetheless, some of the known and shared pathophysiologic features are discussed. Two of the most important, and best studied, aspects of HFpEF are diastolic LV dysfunction and LV stiffness, but other contributing factors include LV systolic dysfunction, extracardiac vascular dysfunction, incoordination of the LV and the vascular system, and impaired reserve system and volume status.

Diastolic Left Ventricular Dysfunction

The term diastolic HF was initially used to describe HFpEF, but not all HFpEF has etiologic origin in diastolic LV dysfunction. Having said that, diastolic LV dysfunction is quite common in HFpEF. Studies of diastolic function in HFpEF showed that it may be the primary factor responsible for hemodynamic abnormalities and symptoms in HFpEF, primarily in hypertrophic cardiomyopathy, marked LV hypertrophy, and restrictive cardiomyopathy.

The ventricle fills during diastole, and diastole can be divided into isovolemic relaxation and ventricular filling phases (Fig. 68–10). The ventricular filling phase is further made up of early rapid filling, diastasis, and atrial systole. The early rapid filling makes up the majority of the ventricular filling normally (70%-80%) and is driven by the pressure gradient between left ventricle and left atrium. Other contributing factors to the pressure gradient includes myocardial relaxation, ventricular stiffness, ventricular elasticity, systolic function, left atrial (LA) pressure, ventricular interaction, pericardium, LA stiffness, pulmonary vein status, and mitral valve. As a person gets older, the early rapid filling does not make up as much of the ventricular filling. Atrial systole further contributes 15% to 25% of ventricular filling followed by diastasis at another 5% in normal individuals without HF (Fig. 68–11).

Left Ventricular Relaxation

LV relaxation, or lusitropy, begins during end systole and isovolemic relaxation and through the rapid filling phase. It is one of the most important determinants of diastolic LV dysfunction in HFpEF.¹⁸⁹ The process of LV relaxation can be assessed by either an invasive or noninvasive measurement. By using a special manometer-tipped LV catheter to measure LV pressures at various stages of LV relaxation, one can derive tau, the time constant of relaxation that describes the rate of LV pressure changes during isovolemic relaxation.

where P₀ = LV pressure at end ejection; t = time, and τ = tau.

Normally, tau is less than 40 milliseconds, and the larger the tau, the more impaired the relaxation. More commonly, LV relaxation is estimated from echocardiographic data. Depending on LA pressure, LV pressure, and the gradient between LA and LV pressures, the transmitral Doppler flow velocity pattern during diastole is reflective of various degrees of diastolic dysfunction (Fig. 68–12). Mitral annular velocity during early diastole (e′) with tissue Doppler also provides more preload-independent information regarding LV relaxation. E/e′ ratio correlates reasonably with filling pressures; high E/e′ ratio is suggestive of elevated filling pressures. Transmitral flow velocity pattern, mitral annular tissue velocity values, E/e′ ratio and E/A reversal with Valsalva maneuver, pulmonary venous inflow velocity patterns, LA dilation, and elevated pulmonary pressures can help to assess diastolic dysfunction. An isolated finding has limited diagnostic performance, and there is no gold standard definition of diastolic dysfunction; therefore, the final determination of diastolic dysfunction is ultimately based on gathering as much information as possible and making the most likely diagnosis.

Several factors impact LV relaxation, with a number of potential mechanisms contributing to the pathophysiology of HFpEF, including afterload, the rate and extent of detachment of actin-myosin cross-bridges, and the timing and distribution of intraventricular pressure (Table 68–4).

For the myofiber to lengthen during diastole, the actin-myosin cross-bridges have to detach in an active, ATP-dependent process. This process involves close and rapid regulation of cytosolic Ca²⁺. In animal models of HFpEF, there appears to be reduced Ca²⁺ reuptake into the SR via decreased SERCA activity, impaired β-adrenergic signaling leading to reduced protein kinase A activity, increased protein kinase C via protein phosphatases, and elevated catecholamine activation.^190,191 Myocardial injury, such as from myocardial ischemia, can directly impact ATP metabolism via reduced ability to cross-bridge detachment and ultimately myocardial relaxation.

These cellular abnormalities in LV relaxation contribute to hemodynamic abnormalities by increasing filling pressures, and these changes become even more apparent during exercise. In the normal heart, there is enhanced relaxation during exercise mediated by an increase in the LA-LV pressure gradient without increasing LA pressures. In the setting of impaired LV relaxation, there is evidence of chronotropic incompetence and impaired myocardial reserve during exercise as well as an increase in LV diastolic pressures when heart rate and blood pressure increase in response to exercise.^192,193

Because patients with HFpEF and abnormal LV relaxation are increasingly dependent on atrial contraction for filling, atrial fibrillation along with its tachycardia can contribute to elevated filling pressure and onset of acute HF.¹⁹⁴

Left Ventricular Diastolic Stiffness

LV diastolic stiffness is also one of the biggest contributors to diastolic dysfunction in HFpEF patients.^{189,195,196,197} With increased stiffness of the left ventricle or impaired elastance, there is an increase in LA pressures, pulmonary venous pressure, and congestion. Elastance is the ability of the left ventricle to stretch and recoil with the distending intraventricular force on the myocardium. Estimation of LV diastolic stiffness requires formation of pressure-volume curves while varying preload invasively or by Doppler evidence of elevated filling pressures in the setting of normal LV dimension, decreased deceleration in early diastolic transmitral flow velocity in the setting of impaired relaxation, and rapid equalization of LA and LV pressures during early diastolic filling.

Extracellular matrix surrounding myofibers contains fibrillar proteins such as collagen type I and III, elastin, proteoglycans, and the basement membrane proteins laminin and fibonectin, and contributes significantly to LV diastolic stiffness.¹⁹⁸ Extracellular matrix, while acting as the scaffolding of the heart, is a rather dynamic system with close interdependence and regulation by myocardial injury, neurohormonal activation, and remodeling. There are close ties between extracellular matrix and fibrosis, with a close association between interventions that affect extracellular matrix formation/removal and LV diastolic stiffness and relaxation.

Another factor affecting LV diastolic stiffness is titin, a large protein that acts as a spring in the heart's scaffolding. Change in titin isoforms or decreased phosphorylation of titin can raise resting tension of biopsied human myocytes.^199,200

Inflammation

Patients with HFpEF have increased frequency of comorbidities including hypertension, obesity, diabetes mellitus, chronic obstructive pulmonary disease, and iron deficiency. Although a definitive link is lacking, studies to date suggest that these comorbidities may be interconnected with the pathophysiology of HFpEF through increased inflammation found in patients with these comorbidities.²⁰¹ Human models of HFpEF have shown increased biomarkers of inflammation, such as C-reactive protein, sST2, and tissue inhibitor of metalloproteinase.¹⁹⁷ One of the working hypotheses is that the inflammatory state may result in altered coronary microvascular endothelial function and increased ROS; reduced NO bioavailability, cyclic guanosine monophosphate content, and protein kinase G activity; hypophosphorylation of titin; and increased interstitial fibrosis, ultimately resulting in HFpEF (Fig. 68-13).¹¹⁹ A study of mineralocorticoid receptor antagonism appeared to improve obesity-associated cardiac diastolic dysfunction, as documented by echocardiography in mice, with a reduction in cardiac fibrosis and restoration of endothelium-dependent vasodilation of isolated coronary arterioles via an NO-independent mechanism.²⁰² A promising therapy, the phosphodiesterase-5 inhibitor sildenafil appeared to benefit systemic vasculature and endothelium, but was found to be deleterious on ventricular function in HFpEF patients.²⁰³

Exercise Tolerance in Heart Failure With Preserved Ejection Fraction

Normally cardiac output can increase dramatically in response to exercise, and this increase is accomplished by an increase in heart rate and SV, decreased peripheral vascular resistance, increased LV contractility, and increased filling rate. However, to increase cardiac output during exercise, both the force-frequency relationship and the Frank-Starling mechanism have to be intact to compensate for the shortening of diastolic filling time during exercise. In the setting of HFpEF, there is an impaired force-frequency relationship and a prerequisite of a normal LV elasticity in order for the Frank-Starling mechanism to result in an increase in EDV without increasing filling pressures. Cardiopulmonary exercise testing suggests that reduced cardiac output responses, such as impaired chronotropic responses and reduced peak SV likely influenced by LV diastolic dysfunction, and pulmonary vascular dysfunction mainly limit exercise capacity in HFpEF patients, rather than limitations in peripheral oxygen extraction.^192,204 Factors beyond the heart, such as impaired skeletal muscle vasodilation during exercise, likely contribute to exercise intolerance in HFpEF.²⁰⁵ Fascinatingly, exercise training and even caloric restriction in obese HFpEF patients appeared to improve exercise tolerance in small randomized trials, with the greatest effect when combined.²⁰⁶ Despite multiple therapeutic trials, the only intervention shown to improve exercise tolerance and quality of life in HFpEF patients is exercise training.²⁰⁷ To date, no medications have been shown to improve hard clinical outcomes such as mortality or HF hospitalizations in HFpEF patients.

It is important to note that LVEF does not correlate well with exercise tolerance, and impaired cardiac output alone is not the dominant reason for exercise intolerance. Many factors limit increased cardiac output during exercise in patients with HF, including a reduced myocardial force-frequency response, an inability to fully use the Starling effect, and chronotropic incompetence. These limitations of cardiac performance are combined with endothelial dysfunction in the peripheral arterioles, leading to impaired vasodilatation and reduced nutritive skeletal muscle blood flow. Disuse atrophy of skeletal muscles also occurs. The latter effects appear to be excessively dependent on a glycolytic metabolism, partly because of reduced metabolic efficiency in performing external work.²⁰⁸ Additional mechanisms include changes in muscle fiber recruitment, selective atrophy of oxidative fibers, and physical deconditioning. The mitochondrial content of skeletal muscle is reduced in HF.²⁰⁹ Increased expression of the inducible isoform of NOS in skeletal muscle is correlated with reduced mitochondrial creatine kinase expression and exercise intolerance.²¹⁰ Apoptosis is frequently found in skeletal muscle of patients with HF and is also associated with exercise impairment.²¹¹ This crosstalk between the heart and the peripheral organs is gaining increasing attention as new evidence comes to light.

Inflammation

Myocardial injury, such as from myocardial ischemia, or LV volume or pressure overload, results in recruitment of inflammatory cytokines such as IL-1β, IL-6, IL-18, and TNF-α. However prolonged inflammatory pathway activation can cause these factors to circulate in elevated concentrations.²¹² These changes may be an attempt to repair injury or increased exposure to endotoxin as a result of altered gut permeability in patients with edema.²¹³ When elevated, these cytokines are associated with further injury as well as poor clinical outcomes in HF. TNF-α, in particular, has garnered attention because it is elevated in patients with HF, the degree of elevation closely mirrors NYHA class, and elevated values are associated with worse clinical outcomes.^115,212 In murine models, overexpression of TNF-α resulted in dilated cardiomyopathy with biventricular fibrosis, suggesting an etiologic role in the development of HF.²¹⁴ However, treatment with anti–TNF-α did not alter clinical outcomes in HF.

Furthermore prolonged exposure to elevated inflammatory cytokines may even result in peripheral organ damage. Inflammation and the nucleotide-binding oligomerization domain-like receptors with pyrin domain (NLRP₃) inflammasome have been implicated in playing a central role in the connection between the heart and the peripheral organs in HF.²⁰¹

Cardiometabolic Connections: Cardiac Cachexia

Obesity is a known risk factor for cardiovascular disease and for the development of chronic HF. However, recent data suggest that increased body mass index is associated with a more favorable prognosis once HF is established.^215,216,217 Negative energy balance and unintentional, nonedematous weight loss, or cachexia, occur in an estimated 13% to 35% of patients with HF.²¹⁸ Cachexia is also associated with increased risk of adverse clinical outcomes.^219,220 Studies suggest that there may be multiple factors involved in the development of cachexia in HF, such as altered nutritional absorption, neurohormonal activation, and immunologic activation.²²¹ Both human and animal models of cachexia have revealed that elevated levels of TNF-α correlate with sarcopenia and fat, although the mechanism is not well understood.^115,221 Adipose tissue secretes a number of cytokines, including adiponectin. Elevated adiponectin level is associated with increased mortality and is probably a marker of wasting.²²² Moreover, low rather than high cholesterol levels are associated with a poor outcome in patients with chronic HF. In HF, whole-body protein synthesis seems to be suppressed, and breakdown of myofibrillar protein, principally leg skeletal muscle, is increased. A significant decrease in LV mass also occurs in patients with cardiac cachexia.²²³

Cardiorenal Connections

The interaction between the heart and the kidney is bidirectional, with the heart affecting the renal function and the kidneys affecting the cardiac function.²²⁴ A number of mechanisms may be involved including neurohormonal activation, reduced renal perfusion, and increased renal venous pressures. Activation of the neurohormonal system, including the activation of the sympathetic nervous system and RAAS system, release of vasopressin, and ET-1 lead to excessive salt and water retention and vasoconstriction and overwhelm the counter-regulatory systems. However, such vasoconstriction also results in increased afterload and reduces cardiac output. Decreased cardiac output can further contribute to reduced renal perfusion in a vicious cycle. However, the relationship between decreased cardiac output and decreased renal perfusion may not be a simple one; evidence suggests that invasive measure of cardiac output does not correlate well with estimated glomerular filtration rate at baseline, and a change in cardiac output did not correlate with a change in renal function at moderate cardiac dysfunction, but may correlate in severely reduced cardiac output states.²²⁵ More importantly, both animal and human models have shown that increased renal venous pressures resulting in a decreased perfusion gradient across the glomerular capillary network in accordance with the Poiseuille law may cause a reduction in renal function.^{226,227,228,229,230} Right ventricular dilation and systolic dysfunction have also been suggested as other mechanisms of cardiorenal syndrome through elevated central venous pressure as well as interventricular interaction impeding LV filling caused by increased right-sided pressures.²³¹

HF is a heterogeneous clinical syndrome composed of several phenotypes. The pathophysiology is complex, and our understanding of it is still emerging. Rather than the heart as an isolated organ, the intricate relationship between the heart and the rest of the body needs to be better understood. Treatment for HFrEF has improved over the years, but is not curative. The prevalence of HFpEF is rapidly increasing, and no medical therapy has been shown to benefit the clinical status of affected patients. In addition, there is no question that prevention of HF must be given a higher priority. We now know that most coronary heart disease, the number one cause of HF, can be prevented by control of coronary risk factors. Because 60% to 65% of all HF in Western civilization is associated with advanced coronary disease, it is clear that controlling blood pressure, reducing plasma lipids, maintaining an ideal weight (and thus reducing the chance of developing diabetes mellitus), staying physically fit, and avoiding cigarettes can have a substantial prevention effect on the development of HF. More precise understanding of varying pathophysiologic mechanisms is needed in order to identify a more precise, targeted therapy for HF phenotypes, thus allowing for a rational, aggressive, and cost-effective prevention strategy before patients reach the late stages of the syndrome.

We would like to thank Dr. Gary S. Francis, Dr. W.H. Wilson Tang, and Dr. Richard A. Walsh for their work on the previous version of this chapter.