CHAPTER 24 Acute Heart Failure and Pulmonary Edema

Theo E. Meyer, Rajan Krishnamani, William H. Gaasch

Clinical Presentation of Acute Heart Failure

Diagnosis and Evaluation of Acute Heart Failure

ACUTE HEART failure (AHF) is a clinical syndrome of new or worsening signs and symptoms of heart failure (decompensated), often leading to hospitalization or a visit to the emergency department. Patients with AHF represent a heterogeneous population with high postdischarge readmission rates.¹^–⁵ Most hospitalized patients have significant volume overload, and congestive symptoms predominate. Fewer patients present with hypotension and symptoms of reduced organ perfusion.¹^–⁵ A few patients who come to the emergency department with AHF have “flash pulmonary edema.”¹ A few patients have onset of symptoms within 4 hours of arrival; most patients have a slow progression of disease, resulting from cardiac ischemia, medication noncompliance, dietary indiscretion, or exacerbation of hypertension. The average patient has had symptoms for almost 5 days before seeking medical attention.¹^–⁵

AHF is the most common cause of hospital admission in patients older than 65 years, accounting for 1 million admissions annually.⁶ AHF represents a period of high risk for patients, with a 20% to 30% mortality rate within 6 months after admission.⁷^–⁹ Early care of AHF and time to treatment are linked to outcome. These patients are generally cared for in telemetry units in the United States. Only 10% to 20% of these patients are admitted to intensive care units (ICUs).^2,³

Pathophysiologic Considerations

Integral to the understanding of the pathogenesis and treatment of AHF and pulmonary edema is a basic understanding of the forces involved in fluid retention, capillary-interstitial fluid exchange (Starling relationship), the determinants of myocardial pump performance, and the pathophysiology of AHF.

Chronic Progressive Fluid and Water Retention

Renal sodium and water excretion normally parallels sodium and water intake, so that an increase in plasma and blood volume is associated with increased renal sodium and water excretion. In patients with heart failure, sodium and water are retained, despite an increase in intravascular fluid volume. Renal sodium and water retention in these patients may be regulated not by the total blood volume, but by the degree of filling of the arterial compartment—the so-called effective blood volume. The dynamic equilibrium of the arterial circulation, as determined by cardiac output and peripheral vascular resistance or compliance, is the predominant determinant of renal sodium and water excretion.

Arterial underfilling is sensed by mechanoreceptors in the left ventricle, carotid sinus, aortic arch, and renal afferent arterioles.¹⁰ Decreased activation of these receptors because of a decrease in systemic arterial pressure, stroke volume, renal perfusion, or peripheral vascular resistance leads to an increase in sympathetic outflow from the central nervous system, activation of the renin-angiotensin-aldosterone system, and the nonosmotic release of arginine vasopressin, as well as the stimulation of thirst.¹⁰ These factors together with increased release of endothelin and vasopressin and resistance to natriuretic peptides contribute to sodium and water retention leading to decompensation of chronic heart failure.

Pulmonary Edema

The flux of fluid out of any vascular bed results from the sum of forces promoting extravasation of fluid from the capillary lumen versus forces acting to retain intravascular fluid. This concept of a dynamic equilibration between opposing forces in the lung is given mathematical expression in the Starling equation:¹¹

where Qf is net transvascular fluid flow across the pulmonary capillary endothelium; Kf is filtration coefficient of the microvascular endothelium (hydraulic conductivity of the capillary wall × surface area); Pv is hydrostatic pressure in the pulmonary capillaries; Pint is hydrostatic pressure in the pulmonary interstitium; pV is plasma protein oncotic pressure; and pint is protein oncotic pressure within the interstitial space.

Under normal conditions, the sum of the forces is slightly positive, producing a small vascular fluid flux into the precapillary interstitium of the lung that is drained as lymph into the systemic veins. The capillary coefficient (Kf) determines the effectiveness of the endothelial barrier to protein permeability, and so determines the effectiveness of the oncotic gradient. Because the intravascular pressure in the pulmonary capillaries is always higher than plasma osmotic pressure, transcapillary fluid flux out of the pulmonary capillary is continuous. When the interstitial fluid exceeds the interstitial space capacity, fluid floods into the alveoli.¹² The interstitial space is drained by a rich bed of lymphatics. It is estimated that pulmonary lymph flow may increase threefold before fluid extravasates into the alveolar airspaces. In patients in the ICU, the two most common forms of pulmonary edema are pulmonary edema initiated by an imbalance of Starling forces and pulmonary edema initiated by disruption of one or more components of the alveolar-capillary membrane. Table 24-1 lists the causes of pulmonary edema based on the initiating mechanism.

Table 24–1 Classification of Acute Pulmonary Edema

Cardiogenic Pulmonary Edema

A. Acute Increase in Pulmonary Capillary Pressure

1. Increased LA pressure with normal LV diastolic pressure

a. Thrombosed prosthetic mitral valve

b. Obstructive left atrial myxoma

2. Increased LA pressure owing to elevated LV diastolic pressure

a. Acute increases in myocardial stiffness or impaired relaxation

i. Myocardial ischemia

ii. Acute myocardial infarction

iii. Hypertrophic heart disease complicated by tachycardia or ischemia

b. Acute volume load

i. Acute mitral or aortic regurgitation

ii. Ischemic septal rupture

c. Acute pressure load

i. Hypertensive crisis

ii. Thrombosed prosthetic aortic valve

B. Exacerbation of Chronically Elevated Pulmonary Capillary Pressures

1. Increase in elevated LA pressure with normal LV diastolic pressure

a. Mitral stenosis

b. Left atrial myxoma

2. Increase in elevated LA pressure owing to a further increase in LV diastolic pressure

a. Further increases in myocardial stiffness or impaired relaxation

i. Cardiomyopathy complicated by myocardial ischemia or infarction

ii. Hypertrophic heart disease complicated by tachycardia or ischemia

b. Volume load imposed on preexisting LV diastolic dysfunction

i. Worsening mitral regurgitation

ii. Vigorous postoperative fluid administration

iii. Dietary indiscretion

c. Pressure load imposed on preexisting LV systolic dysfunction

i. Accelerated hypertension

Noncardiogenic Pulmonary Edema

A. Altered Alveolar Capillary Membrane Permeability (Adult Respiratory Distress Syndrome)

1. Infectious or aspiration pneumonia

2. Septicemia

3. Acute radiation or hypersensitivity pneumonitis

4. Disseminated intravascular coagulopathy

5. Shock lung

6. Hemorrhagic pancreatitis

7. Inhaled and circulating toxins

8. Massive trauma

B. Acute Decrease in Interstitial Pressure of the Lung

1. Rapid removal of unilateral pleural effusion

C. Unknown Mechanisms

1. High-altitude pulmonary edema

2. Neurogenic pulmonary edema

3. Narcotic overdose

4. Pulmonary embolism

5. After cardioversion

6. After anesthesia or cardiopulmonary bypass

LA, left atrial; LV, left ventricular.

It has been shown experimentally that pulmonary edema occurs if the pulmonary capillary pressure exceeds the plasma colloid osmotic pressure, which is approximately 28 mm Hg in humans. The normal pulmonary capillary wedge pressure is approximately 8 mm Hg, which allows a margin of safety of about 20 mm Hg in the development of pulmonary edema.¹³ Although pulmonary capillary pressure must be abnormally high to increase the flow of the interstitial fluid, these pressures may not correlate with the severity of pulmonary edema when edema is clearly present.¹⁴ These pressures may have returned to normal when there is still considerable pulmonary edema because time is required for removal of interstitial and pulmonary edema.

The rate of increase in lung fluid at any given elevation of pulmonary capillary pressure is related to the functional capacity of the lymphatics, which may vary from patient to patient, and to variations in osmotic and hydrostatic pressures. Chronic elevations in left atrial pressures are associated with hypertrophy in the lymphatics.¹⁵ These lymphatics clear greater quantities of capillary filtrate during acute increases in pulmonary capillary pressure. Clinical experience with patients who have chronically elevated atrial pressures suggests that these patients show minimal or no evidence of interstitial lung edema. The mechanisms by which pulmonary capillary pressure increases when the pumping ability of the ventricle is suddenly impaired are discussed later.

When the alveolar-capillary membrane is injured, proteins leak from the capillary into the interstitium, reducing the oncotic counterpressure tendency to oppose capillary filtration. Interstitial edema and the consequent alveolar edema formation can occur in the presence of low hydrostatic pressures.

Determinants of Left Ventricular Pump Performance

Cardiac output equals the heart rate multiplied by the stroke volume. The stroke volume is the volume of blood ejected with each heartbeat; the magnitude of the stroke volume is governed by ventricular loading conditions and the contractile state of the myocardium. The factors involved in regulating cardiac output are discussed in Chapter 6. The following discussion briefly reviews the determinants of left ventricular (LV) pump performance as it pertains to a patient with AHF.

Preload

Preload is the force or load acting to stretch the LV fibers at the end of diastole, determining the resting length of the sarcomeres.¹⁶ Preload expressed in terms of LV end-diastolic pressure or volume (i.e., dimension) is common, but neglects the confounding influence of the complex geometry of the left ventricle.¹⁷ Preload is probably best measured as end-diastolic wall stress because this reflects the interactions between transmural diastolic pressures, ventricular dimensions, and wall thickness. The level of ventricular preload is influenced by multiple variables, including chamber compliance, overall LV performance, intravascular blood volume and venous return, atrial contribution to ventricular filling, intrapericardial and intrathoracic pressures, right ventricular-LV interaction, and pericardial restraint.¹⁸

Because preload determines the end-diastolic fiber length, it is a major determinant of systolic performance. This phenomenon is generally referred to as Starling’s law of the heart. The ability of the heart to increase performance using the Starling mechanism (i.e., preload reserve) is limited by the inherent stiffness of the myocardium. In compliant, normal hearts, the preload reserve is an effective mechanism to increase the stroke volume, but in hypertrophied hearts or in hearts that are already at the limits of their preload reserve (i.e., end-stage heart failure), stroke volume cannot be substantially increased by this mechanism. When the heart operates on the steep portion of the diastolic pressure-volume curve, small increments in LV volumes are associated with large changes in filling pressures. In other words, large increases in filling pressures are required for a small increase in end-diastolic volume.

Myocardial relaxation is also an important determinant of the fiber length at end-diastole. Impaired relaxation shortens the time available for diastolic filling, and may interfere with fiber stretch when relaxation is markedly prolonged or when the heart rate is rapid. As a result, the force acting to stretch the myocardial fiber is opposed by incomplete relaxation at end-diastole.

Afterload

Afterload is the force opposing fiber shortening after the onset of ejection.¹⁶ This force or stress is not constant because arterial and ventricular pressure changes during ejection. Even if the pressure were constant, the systolic force that the ventricle develops would vary in accordance with a complex relationship between the wall force, ventricular pressure, and ventricular dimension.¹⁹ According to the Laplace equation, wall stress (s) is directly proportional to ventricular systolic pressure (P) and radius (R), and is inversely related to wall thickness (h):

An increase in wall stress because of an increase in ventricular pressure, an increase in ventricular dimension, inadequate hypertrophy, or any combination of these results in decreased myocardial fiber shortening. Because there has been insufficient time for compensatory hypertrophy to develop in patients with AHF, ventricular pump performance is particularly sensitive to changes in afterload.

Preload and afterload are intimately related. When LV preload is increased in a normal heart, systolic LV pressures generally increase, and as a result systolic wall stress (afterload) increases. Likewise, a decrease in afterload promotes LV emptying, which leads to a decrease in preload.

Contractility

Contractility, or inotropic state, is defined as the property of myocardial fibers that determines the extent of shortening independent of loading conditions. Compared with the control state, a positive inotropic intervention (e.g., epinephrine) increases myocardial fiber shortening and stroke volume at any given preload and afterload. A negative inotropic intervention (e.g., β-blocker) has the opposite effect. For any given increment in afterload, shortening and stroke volume decrease to a greater extent when the contractile state is depressed. Conversely, afterload reduction has a small effect on increasing shortening and stroke volume during a positive inotropic intervention.

Heart Rate

Heart rate affects cardiac output under resting conditions and during exercise. When the heart rate increases, the time available for filling during diastole is substantially abbreviated. When LV diastolic compliance is decreased, tachycardia can produce a decrease in ventricular preload that leads to reduced LV stroke volume despite normal systolic function. In hearts with impaired relaxation (i.e., hypertrophied hearts), a significant increase in heart rate would cause LV filling to occur at the time when the heart is not completely relaxed, which results in increased LV diastolic pressures. In patients with abnormal hearts, tachycardia may cause an exaggerated decrease in stroke volume and an increase in filling pressures.

The normal heart responds to an increase in heart rate by a positive inotropic effect, also known as the treppe or Bowditch phenomenon.²⁰ In a failing or hypertrophied heart, an increase in heart rate may produce a reduction in, or even a reversal of, the normally positive force-frequency effect.^21,²² Tachycardia is also likely to increase myocardial oxygen consumption, which may lead to subendocardial ischemia in a heart with a compromised coronary circulation.

Left Ventricular Pressure-Volume Relationships

To appreciate fully the factors that contribute to AHF, it is appropriate to review briefly the pressure-volume relationships of normal and diseased hearts. The relationship between pressure and volume throughout the cardiac cycle can be presented as a pressure-volume loop (Fig. 24-1). The pressure-volume loop encapsulates the systolic and diastolic functions of the heart. Because these loops also circumscribe end-systolic and end-diastolic volumes, the stroke volume and ejection fraction can be derived. The bottom limb of the loop, called the diastolic pressure-volume curve, describes LV diastolic compliance. The pressure-volume loop can provide a simple but comprehensive description of LV pump function.

Figure 24-1 Schematic representation of the left ventricular (LV) pressure-volume loop. The aortic valve opens at b and closes at c. The mitral valve opens at d and closes at a. The slope of the broken line through c represents the end-systolic pressure-volume relationship (E_max), and the broken line through d and a represents the end-diastolic-volume relationship. As contractile force develops in the left ventricle (LV) in systole, the pressure rapidly increases in the ventricular chamber (a → b) without changing its volume (i.e., isovolumic phase of systole). When the pressure exceeds the diastolic aortic pressure, the aortic valve opens, and the ventricle ejects its contents into the arterial circulation (b → c, the ejection phase of systole). At the end of ejection (point c), LV pressure decreases, and the aortic valve closes. Pressure rapidly declines at a constant volume (c → d, isovolumic relaxation) to levels below that of the left atrium. At this point, the mitral valve opens (point d), and the relaxing LV fills along the segment d → a. The trajectory a → b → c represents the contractile or inotropic function of the LV at any given end-diastolic volume, whereas the trajectory c → d → a represents the lusitropic function (relaxation and filling) of the heart at any given end-systolic pressure. The area within the loop graphically depicts the external work (i.e., stroke work) of the ventricle.

Progressive increases in systolic pressure produce a nearly linear increase in end-systolic volume. By matching the end-systolic pressure and volume coordinates from multiple, variably loaded beats, a near-linear relationship is established. The slope of this relationship (E_max), determined by altering load, reflects LV contractility (see Fig. 24-1).²³ A positive inotropic intervention is associated with an increased end-systolic pressure and stroke volume and a decreased end-diastolic volume; this results in an increased E_max and a shift of the pressure-volume relationship to the left (Fig. 24-2A). Conversely, a negative inotropic intervention decreases end-systolic pressure and stroke volume and increases end-diastolic volume; this results in a decrease in E_max and a shift of the pressure-volume relationship to the right (Fig. 24-2B).

Figure 24-2 Schematic diagrams illustrating the effects of inotropic interventions on the pressure-volume loop. A, With a positive inotropic intervention, the pressure-volume loop (broken line) is shifted to the left, and the slope of the end-systolic pressure-volume line is increased. B, With a negative inotropic intervention, the pressure-volume loop is shifted to the right, and the slope of the end-systolic pressure-volume line is decreased.

In the intact human heart, an increase in systolic pressure is associated with an increase in end-systolic volume, and if preload does not increase, stroke volume decreases (Fig. 24-3A). An increase in preload is accompanied by an increase in stroke volume and a modest increase in end-systolic pressure (Fig. 24-3B). Acute and chronic changes in the pressure-volume relationship in the failing heart depend on the underlying myocardial structure and function, the type and extent of injury (e.g., infarction or myocarditis, regional or global myocardial depression), and the severity and nature of the hemodynamic load (pressure versus volume).

Figure 24-3 Schematic diagrams illustrating the effects of changing loading conditions on the pressure-volume loop in the intact heart. A, An increase in afterload shifts the pressure-volume loop (broken line) to the right, increasing the end-systolic and end-diastolic volumes and the end-systolic and end-diastolic pressures, while decreasing the stroke volume. The slope of the end-systolic pressure-volume line is usually not affected by a pure change in afterload. B, An increase in preload also shifts the pressure-volume loop (broken line) to the right, increasing the end-diastolic volume and end-diastolic pressure. The increase in preload may be associated further with a small increase in end-systolic volume and a modest increase in end-systolic pressure, but in contrast to the case with an increase in afterload, the stroke volume increases. Similar to an increase in afterload, the slope of the end-systolic pressure-volume line is not affected by a change in preload.

Chamber Stiffness

Chamber stiffness is determined by analyzing the curvilinear diastolic pressure-volume relationships (Fig. 24-4). The slope of the tangent (dP/dV) to this curvilinear relationship defines the chamber stiffness at a given filling pressure. An increase in dP/dV because of an increase in volume, shown in Figure 24-4 (A → B), has been called a preload-dependent change in stiffness. When the pressure-volume relationship shifts to the left (A → C), the tangent is steeper at the same diastolic pressure. The latter may be caused by an increase in myocardial mass or intrinsic myocardial stiffness or by changes in several extramyocardial factors. Chamber stiffness of the left ventricle is determined by static factors (e.g., chamber volume, wall mass, stiffness of the wall) and dynamic factors (e.g., pericardium, right ventricle, myocardial relaxation, erectile effects of the coronary vasculature).^18,²⁴ Most acute alterations in LV chamber stiffness result from a preload-dependent increase in chamber stiffness, a shift to a different pressure-volume curve, or a combination of the two. All can result in elevated left atrial pressures, pulmonary venous hypertension, and the signs and symptoms of acute heart failure.