Chapter 27. Heart Failure with Preserved Ejection Fraction

• Symptoms and signs of heart failure with preserved left ventricular ejection fraction (LVEF > 50%).
• Presence of an underlying cause of heart failure with preserved ejection fraction (eg, comorbidities such as hypertension, coronary artery disease, diabetes, chronic kidney disease; or underlying valvular heart disease, restrictive cardiomyopathy, or specific myocardial diseases such as amyloidosis).
• Objective evidence of elevated left ventricular filling pressure (at rest or with exercise) on echocardiography or cardiac catheterization.

Heart failure with preserved ejection fraction (HFpEF) is an increasingly common, debilitating syndrome of the elderly, and one that carries a high rate of morbidity and mortality. HFpEF accounts for > 50% of all hospitalizations for heart failure, and while a recent patient-level meta-analysis found that patients with heart failure and reduced ejection fraction (HFrEF) have a worse prognosis compared to HFpEF, two earlier large epidemiologic studies found that patients with HFpEF have a mortality rate that is nearly identical to HFrEF. Regardless of underlying ejection fraction, survival for all heart failure (HF) patients is poor, especially after HF hospitalization.

HFpEF is the preferred term for patients with a normal ejection fraction who have the syndrome of HF, because HFpEF highlights the fact that HF is a syndrome and not a distinct clinical or pathophysiologic entity. Many investigators and experts have used the term “diastolic heart failure” for HFpEF in the past. However, this term is not ideal for two main reasons. First, there is ample evidence that patients with HFpEF have abnormalities in systolic function (as defined by tissue Doppler imaging) despite a normal ejection fraction, and many patients with HFrEF have abnormal diastolic function. Second, in the clinical setting, patients with HF are currently classified into two categories: low ejection fraction (< 50%) and preserved ejection fraction (> 50%). By calling HFpEF “diastolic HF,” clinicians may not consider the entire differential diagnosis of HFpEF (of which pure diastolic dysfunction is only one cause). HFpEF has also previously been called “HF with preserved systolic function” or “HF with normal systolic function.” As stated earlier, it is now clear that many patients with HFpEF have abnormalities in systolic function; therefore, HFpEF is a better term.

Finally, HFpEF has the advantage of being an easy mnemonic for patients to remember. HFpEF sounds like “HUFF-PUFF,” which helps patients understand this syndrome, in which dyspnea and fatigue are two of the most common symptoms.

Since HFpEF is heterogeneous, there is no single mechanism that can explain the pathophysiology of the HFpEF syndrome. In some patients with HFpEF, such as those who have the signs and symptoms of HF due to severe valvular disease or pericardial disease (ie, constrictive pericarditis), pathophysiology is relatively straightforward and well-defined. However, in most patients with HFpEF, pathophysiologic abnormalities cannot be ascribed to a single well-defined mechanism. Instead, these patients typically have one or more of the following underlying pathophysiologic processes: (1) diastolic dysfunction due to impaired left ventricular (LV) relaxation, increased LV diastolic stiffness, or both; (2) LV enlargement with increased intravascular volume, which may be due to extracardiac factors such as renal insufficiency; (3) abnormal ventricular-arterial coupling with increased ventricular systolic stiffness and increased arterial stiffness; (4) right HF due to pulmonary venous hypertension with or without superimposed pulmonary arterial hypertension; and (5) chronotropic incompetence. In addition, LV hypertrophy and coronary artery disease are especially important in the pathophysiology of patients with HFpEF.

Diastolic Dysfunction

Diastolic dysfunction occurs when the ventricle loses its normal ability to suction blood from the left atrium. When the ventricle relaxes abnormally, filling is delayed and left atrial emptying is incomplete. An abnormally stiff ventricle worsens the problem by also impeding left atrial emptying. The end result is abnormally high left atrial and LV diastolic pressures. The LV loses its suction and instead of “pulling” blood from the left atrium and pulmonary veins, it now relies heavily on left atrial contraction so that the LV can fill and distend appropriately and recoil in systole. This is one reason why atrial fibrillation is tolerated so poorly in patients with advanced LV diastolic dysfunction with resultant elevation of left atrial pressure, pulmonary vascular congestion, and poor cardiac output.

In patients with HFpEF who have substantial diastolic dysfunction as a cause of their symptoms, the end-diastolic pressure-volume relationship is shifted up and to the left (Figure 27–1). In these patients, even small increases in central blood volume or vascular (arterial or venous) tone can result in significant increases in left atrial volume and pulmonary venous pressures. Patients with an upward and leftward shift in the LV end-diastolic pressure–volume relationship tend to have a high relative wall thickness (high LV mass/volume ratio); increased fibrosis and scar of the LV myocardium due to ischemia, infarction, infiltrative disease, or radiation; and impaired active relaxation of the myocardium (due to abnormal myocyte calcium homeostasis).

Left Ventricular Enlargement and Increased Intravascular Volume

LV enlargement is a key predictor of HF, regardless of ejection fraction. However, patients with isolated diastolic HF are often thought to have small LV volumes. This apparent discrepancy can be explained by the underlying cause of HFpEF and diastolic HF. It is likely that patients with significant coronary disease or myocardial ischemia (even in the absence of epicardial coronary disease) suffer from increased LV enlargement. In addition, many patients with LV enlargement have increased intravascular volume due to comorbidities such as chronic kidney disease, anemia, and obesity. Thus, LV enlargement and increased intravascular volume cause symptoms of HFpEF by a pathophysiologic mechanism that is distinct from pure LV diastolic dysfunction.

Abnormal Ventricular-Arterial Coupling

Ventricular-arterial coupling describes the interaction between ventricular stiffness and central arterial stiffness. In healthy patients, young and old, arterial and ventricular elastance (stiffness) are matched in order to maintain optimal cardiac efficiency. However, with increasing age, ventricular stiffness is elevated and results in decreased contractile reserve, thereby rendering elderly patients susceptible to HF, blood pressure lability, and decreased exercise tolerance. Some patients with HFpEF appear to be particularly susceptible to abnormal ventricular-arterial coupling. These patients have the age-related increases in ventricular stiffness described earlier, but instead of matched ventricular and arterial stiffness, ventricular stiffness rises out of proportion of arterial stiffness, which results in poor cardiac efficiency. These patients tend to have high pulse pressure, and they tend to be most sensitive to diuretics whereby small changes in blood volume result in large changes in blood pressure (either significantly hypertensive or hypotensive).

Right Heart Failure

Elevated pulmonary artery systolic pressure (PASP) on echocardiography is present in > 75% of patients with HFpEF and is a marker of worse outcomes in HFpEF. In addition, elevated PASP is the echocardiographic finding that has the best test characteristics for differentiating patients with HFpEF from those with systemic hypertension without HF. It is not surprising that elevated PASP and pulmonary hypertension are common in HFpEF since elevated left atrial pressure (at rest and/or with exercise) is present almost universally in these patients. Elevated left atrial pressure results in increased pulmonary venous pressures, thereby causing pulmonary hypertension. Significant systemic hypertension, which also occurs commonly in HFpEF, also contributes to the high PASP. Thus, pulmonary venous hypertension, defined invasively as mean pulmonary artery pressure (mPAP) > 25 mm Hg with a pulmonary capillary wedge pressure (PCWP) > 15 mm Hg, is common in HFpEF and contributes to right ventricular hypertrophy, dysfunction, and eventual right HF. Some patients (most likely < 5–10% of HFpEF) develop superimposed pulmonary arterial hypertension (on top of pulmonary venous hypertension) either as an extension of their severe pulmonary venous hypertension or due to other risk factors such as chronic thromboembolic disease, chronic hypoxemic lung disease, or obstructive sleep apnea. The best invasive marker for differentiating pulmonary venous hypertension from superimposed pulmonary arterial hypertension is not the transpulmonary gradient (mPAP-PCWP) or pulmonary vascular resistance; rather, it is the pulmonary vascular gradient (pulmonary artery diastolic pressure [PADP]-PCWP). If the PADP-PCWP gradient is > 5–10 mm Hg, pulmonary arterial hypertension is present; if the PADP and PCWP are equalized (ie, PADP-PCWP < 5 mm Hg), only pulmonary venous hypertension is present, and there is no significant superimposed pulmonary arterial hypertension.

Chronotropic Incompetence

Up to 20–30% of patients with HFpEF have evidence of chronotropic incompetence. The inability to increase heart rate with exercise causes exercise intolerance. Patients with advanced HFpEF in particular, such as those with restrictive cardiomyopathy, severe LV hypertrophy, or severe coronary artery disease, have a very stiff LV and are unable to augment stroke volume with exercise. Thus, these patients rely upon increasing heart rate to augment cardiac output with exercise. If chronotropic incompetence is present, exercise capacity will be very limited in these types of patients.

Left Ventricular Hypertrophy

LV hypertrophy contributes substantially to the pathophysiology of HFpEF and is an important risk factor. LV hypertrophy limits coronary flow reserve, increases LV diastolic stiffness, and impairs LV relaxation. Patients with LV hypertrophy suffer from an inability to adequately utilize the Frank-Starling mechanism. Therefore, inadequate preload and chronotropic incompetence can lead to decreased cardiac output, with resultant lightheadedness, dizziness, and exercise intolerance. Finally, because of increased LV wall thickness, the subendocardium is especially vulnerable to ischemia in patients with and without epicardial coronary disease due to decreased coronary blood flow during exercise. Subendocardial ischemia can cause both systolic and diastolic dysfunction in these patients and further exacerbate HFpEF.

Coronary Artery Disease

Approximately 50% of patients with HFpEF have concomitant coronary artery disease. In these patients, coronary disease is often severe, involving multiple epicardial coronary arteries. Patients with prior myocardial infarction, ongoing ischemia, or stable chronic coronary disease can all present with HFpEF. Myocardial ischemia causes calcium sequestration in diastole, which results in impaired LV relaxation and increased LV filling pressures. In areas of prior infarction or ongoing ischemia, regional systolic dysfunction and dyssynchrony can further exacerbate abnormal loading conditions and create a mixture of systolic and diastolic dysfunction. Furthermore, patients with chronic coronary artery disease often have LV remodeling with resultant ventricular enlargement, a known risk factor for increased mortality and HF, regardless of ejection fraction. Preservation of ejection fraction can occur in some patients with prior infarction due to hypertrophy and hyperdynamic function of noninfarcted areas.

Patients with coronary disease suffer from a vicious cycle of abnormalities that contribute to HFpEF. As noted earlier, ischemia can cause impaired LV relaxation and increased LV filling pressures. Impaired LV relaxation in turn can also adversely affect coronary blood flow and coronary flow reserve, which exacerbates ischemia. Increased LV filling pressures results in extravascular compression of the small intramyocardial coronary vessels, which can cause subendocardial ischemia. Increased LV end-diastolic pressure can also result in poor epicardial coronary blood flow. Thus, ischemia begets worsening LV diastolic function, which begets more ischemia.

The first step in caring for a patient with HFpEF is to ensure the correct diagnosis (see later section on Differential Diagnosis). Several criteria for the diagnosis of HFpEF exist. All require signs and symptoms of HF and objective evidence of preserved LV ejection fraction (≥ 50%).

Risk Factors

Age

Patients with HFpEF are almost universally elderly, and aging has several effects on cardiovascular structure and function that are pertinent to HFpEF patients. Aging reduces the diastolic filling rate as a result of prolonged relaxation, which results in left atrial overload and pulmonary venous hypertension. Arterial stiffness increases with age, resulting in increased afterload and load-dependent diastolic dysfunction. In addition, stiffening of the central arteries (which is especially common in women) leaves them less capable to handle changes in blood volume, thereby increasing susceptibility to hypotension, lightheadedness, and dizziness. Finally, aging reduces exercise capacity by increasing ventricular end-systolic chamber elastance (stiffness), which results in decreased ability to augment contractility with exercise.

Hypertension

Hypertension is the most important risk factor for HFpEF and is present in most patients with HFpEF. Hypertensive emergency with flash pulmonary edema is a common presentation of HFpEF. Hypertension leads to LV hypertrophy, which causes impaired relaxation, poor coronary flow reserve, and increased diastolic stiffness, all of which exacerbate HFpEF. Hypertension is also a potent risk factor for epicardial coronary disease, which often complicates HFpEF. Ischemia causes both increased LV stiffness and impaired LV relaxation. Many patients with HFpEF have symptoms of chronic angina. Alternatively, recurrent HF may be an anginal equivalent in many patients with concomitant HFpEF and coronary disease.

Obstructive Sleep Apnea

Obstructive sleep apnea is a common comorbidity in patients with HFpEF, and it can result in worsening LV hypertrophy and pulmonary hypertension. In addition, patients with HFpEF may also have sleep-disordered breathing (such as Cheyne-Stokes respirations) due to their HF. Finally, increased upper airway edema due to generalized HF may actually cause obstructive sleep apnea, a finding that has been shown to improve with diuretic therapy. All of the above contribute to nocturnal microarousals and hypoxia, which result in poor sleep quality, which in turn worsens daytime fatigue and exercise intolerance. Therefore, there should be a low threshold to perform overnight polysomnography on the patient with HFpEF.

Other Clinically Important Risk Factors

Other clinically important risk factors for HFpEF include coronary artery disease, diabetes, chronic kidney disease, obesity, atrial fibrillation, anemia, and chronic obstructive pulmonary disease. All of these comorbidities have their own signs and symptoms that can complicate presentations of HFpEF and add to diagnostic, prognostic, and therapeutic complexities.

Symptoms & Signs

Symptoms and signs of HFpEF are identical to those in patients with HFrEF (systolic HF) and include dyspnea, fatigue, peripheral pitting edema, and jugular vein distention (see Chapter 26). Exercise intolerance and acute decompensated HF are two common presentations of HFpEF.

Exercise Intolerance

Exercise intolerance is one of the main symptoms of HFpEF and one of the most debilitating. In patients with HFpEF, there are many reasons for exercise intolerance, including the following:

• Almost all patients with HFpEF have increased LV diastolic or left atrial pressures, or both. These pressure increases are transmitted to the pulmonary veins, which can cause decreased lung compliance, which is exacerbated by exercise.
• Increased LV diastolic pressure during exercise can limit subendocardial blood flow at a time when there are increased myocardial demands, thereby worsening diastolic function. Poor myocardial perfusion is even worse in patients with LV hypertrophy, which is very common in patients with HFpEF.
• Patients with HFpEF have an abnormal stroke volume response to tachycardia with blunted increase in cardiac output with exercise. Inadequate cardiac output can increase lactate production and worsen muscle fatigue.

Acutely Decompensated HFpEF

The most common factor in acute decompensation is uncontrolled, severe hypertension. Other common clinical findings associated with acute decompensated HFpEF include arrhythmias; noncompliance with medications or salt restriction, or both; acute coronary syndrome; renal insufficiency; valvular regurgitation or stenosis; and infection (eg, pneumonia, urinary tract infection). It is important to recognize the clinical factors associated with acute decompensation because preventing hospitalization is one of the most important goals in patients with HFpEF.

Diagnostic Studies

The diagnosis of HFpEF involves two steps: (1) making sure the patient has the HF syndrome (ie, evidence of elevated left-sided filling pressures) and a preserved ejection fraction (ie, LV ejection fraction ≥ 50%); and (2) determining the underlying cause of HFpEF once it is diagnosed. Echocardiography is a key diagnostic test because it allows the determination of LV ejection fraction, cardiac structural and functional abnormalities, and the comprehensive assessment of LV diastolic function, which can help rule in the presence of HF. Table 27–1 lists a standardized battery of diagnostic and prognostic tests for patients being evaluated for HFpEF.

Echocardiography

Echocardiography is the most important tool in diagnosing diastolic dysfunction and evaluating for other etiologies of HFpEF, such as valvular, pericardial, and coronary disease. All patients with possible or confirmed HFpEF should undergo comprehensive Doppler echocardiography with tissue Doppler imaging. Besides assessment of diastolic function (see below), all patients should be evaluated for increased LV mass and increased relative wall thickness (= [septal thickness + posterior wall thickness]/LV end-diastolic dimension > 0.45). Assessment of PASP; right atrial pressure (from size and collapsibility of the inferior vena cava); and right ventricular size, function, and wall thickness is important in evaluation of pulmonary hypertension.

It is critically important to understand that no one abnormality on echocardiography can diagnose diastolic dysfunction, and age must be factored into the diagnosis, since almost all parameters of diastolic function are age-dependent. Only the proper combination of echocardiographic abnormalities can make the diagnosis of diastolic dysfunction. Figure 27–2 displays an algorithm for the diagnosis of diastolic dysfunction, which requires comprehensive assessment of diastolic function, including mitral inflow, tissue Doppler imaging of the mitral annulus, left atrial volume, and pulmonary venous flow. The first step in evaluating diastolic function by echocardiography is to examine mitral inflow. In most patients, the ratio of early (E) to late (A) mitral inflow velocities (E/A ratio) < 0.8 signifies impaired relaxation (grade I diastolic dysfunction), especially if the patient is younger than 70 years and tissue Doppler e′ velocity is < 10 cm/s at the lateral annulus. In these patients, exercise stress echocardiography (see below) can help evaluate the functional significance of impaired LV relaxation. If the E/A ratio is > 1.5 and early mitral deceleration time is < 150 ms in an elderly patient, the diagnosis is grade III diastolic dysfunction. These patients universally should have an e′ velocity < 10 cm/s at the lateral mitral annulus.

If the E/A ratio is 0.8–1.5 or if the E/A ratio is > 1.5 and early mitral deceleration time is > 150 ms, the question of normal versus pseudonormal mitral inflow arises. In these cases, an E/e′ ratio > 15 or e′ velocity < 10 cm/s usually signifies pseudonormal mitral inflow (grade II diastolic dysfunction). In these patients, left atrial volume index should be increased (> 28 mL/m2). Absence of left atrial enlargement should cause reconsideration of the diagnosis of diastolic dysfunction. In patients who do not meet these criteria, further evaluation with Valsalva maneuver, pulmonary venous flow, or flow propagation velocity can all be used to help differentiate normal versus pseudonormal mitral inflow. Increased left atrial volume index and LV mass index can both provide clues to the presence of diastolic dysfunction in these patients. Even with all of these criteria for LV diastolic dysfunction, there is a subset of patients in whom diastolic function is indeterminate. In addition, there are patients in whom diastolic function is difficult or impossible to assess noninvasively. These patients include those with moderate or greater mitral regurgitation, any degree of mitral stenosis, severe mitral annular calcification, a mitral annuloplasty ring, or mitral valve prosthesis. In patients with atrial fibrillation, although diastolic function grade is difficult to determine, an E/e′ ratio > 11, using e′ velocity from the septal mitral annulus, signifies elevated LV filling pressures.

It is important to note that although Doppler echocardiography is a powerful noninvasive tool for the assessment of diastolic function, diastolic dysfunction is not synonymous with diastolic HF. Many asymptomatic patients have abnormal diastolic function, but since they do not have signs and symptoms of HF, they should not be diagnosed incorrectly with diastolic HF. In addition, all Doppler echocardiographic variables for the assessment of diastolic function, except perhaps tissue Doppler e′ velocity, are extremely load sensitive, and therefore convey more information about preload and afterload than true intrinsic diastolic stiffness. Therefore, it is always important to consider all clinical data, including clinical history and physical examination, when making the diagnosis of diastolic HF or HFpEF.

B-Type Natriuretic Peptide (BNP)

BNP may also have a role in the diagnosis of HFpEF. The Breathing Not Properly Study showed that a BNP > 100 pg/mL had a high sensitivity and negative predictive value for the diagnosis of HFpEF. Therefore, the diagnosis of HFpEF is unlikely in patients presenting to the emergency department with dyspnea and a BNP < 100 pg/mL. However, there is considerable overlap of BNP values in patients with and without HFpEF, especially in elderly women, since BNP increases with age, worsening renal function, and in women. To complicate matters, morbid obesity has been associated with low BNP values, which may be due to BNP clearance receptors on adipocytes and decreased BNP production. Furthermore, in a study of HFpEF outpatients, all of whom had elevated PCWP, BNP was normal (ie, < 100 pg/mL) in 29%. Therefore, BNP, like Doppler echocardiography findings, must be considered within the context of the patient, and cannot be used as a stand-alone diagnostic test for HFpEF.

Exercise Testing

Since exercise intolerance is a key symptom in HFpEF, exercise testing is extremely valuable in these patients. Although many patients who are being evaluated for HFpEF may not be able to withstand a Bruce protocol exercise test given their advanced age and multiple comorbidities, low-intensity and bicycle stress protocols are very feasible. There are two exercise tests that are helpful in evaluating patients with HFpEF: cardiopulmonary exercise testing (CPET) and diastolic stress echocardiography. Studies have shown that on CPET, patients with HFpEF have reduced exercise tolerance, low peak workload, and low peak oxygen consumption (Vo2). Exertional dyspnea is prominent in HFpEF, and a Vo2 provides objective evidence of reduced exercise tolerance. In addition, CPET is useful in differentiating cardiac from pulmonary components of dyspnea and decreased exercise tolerance. Often, CPET is coupled with pulmonary function testing, which also helps evaluate for pulmonary dysfunction, which is very common in elderly patients with HFpEF.

Stress echocardiography for the evolution of diastolic function aims to look for increases in LV filling pressures with exercise. The diastolic evaluation stress can be combined with traditional exercise stress echocardiography. Therefore, one test can diagnose coronary disease and exercise-induced LV diastolic dysfunction. Patients are either tested with a treadmill or bicycle stress protocol. Baseline images are obtained in the parasternal long axis, parasternal short axis, and apical two-chamber, four-chamber, and three-chamber (long-axis) views for assessment of wall motion. Baseline images should also include Doppler assessment of early mitral inflow (E) and tissue Doppler imaging of the septal mitral annulus (e′). At peak stress, wall motion analysis should come first, after which the patient should undergo repeat assessment of mitral inflow and mitral annular velocities. In patients with exercise-induced diastolic dysfunction, LV filling pressures remain elevated for several minutes, which is advantageous since the heart rate must come down to below 100 bpm in order to prevent E and A (and e′ and a′) merging on mitral inflow and tissue Doppler imaging, respectively. At peak exercise, an E/e′ ratio > 13 (using e′ at the septal annulus) suggests exercise-induced increase in LV filling pressures and is diagnostic of exercise-induced diastolic dysfunction.

Stress Testing for the Evaluation of Coronary Artery Disease

All patients with HFpEF should undergo evaluation for coronary artery disease. Exercise stress echocardiography is ideal since patients can be evaluated for the presence of coronary artery disease and exercise-induced diastolic dysfunction with one test. However, adenosine or dipyridamole pharmacologic stress imaging is the test of choice in patients who cannot exercise and in institutions where nuclear myocardial perfusion imaging is superior. Dobutamine stress echocardiography can be performed, but in patients with significant LV hypertrophy, this test may have reduced sensitivity for the detection of wall motion abnormalities.

Cardiac Catheterization

In patients with a high pretest probability for coronary artery disease and in patients with abnormal results of stress testing, coronary angiography should be performed. LV diastolic pressures should be measured in all patients to confirm the diagnosis of elevated LV pressures. Simultaneous right- and left-heart catheterization can be extremely valuable in the assessment of patients with HFpEF and is often underutilized. Invasive assessment in the cardiac catheterization laboratory is currently the gold standard for hemodynamic assessment, allows accurate assessment of cardiac output, and can be very helpful in evaluating for restrictive cardiomyopathy, constrictive pericarditis, or pulmonary hypertension. In addition, dynamic maneuvers such as exercise, fluid challenge, nitroprusside challenge, and pulmonary vasodilator testing (with inhaled nitric oxide or intravenous adenosine) can be valuable in specific circumstances.

Cardiac Magnetic Resonance Imaging (MRI)

Cardiac MRI can be extremely helpful in the evaluation of patients with HFpEF. Cardiac MRI is the gold standard for assessment of LV volumes, left atrial volume, and LV mass. In addition, cardiac MRI can evaluate for focal areas of fibrosis (hyperenhancement), aortic enlargement and dissection (which is important since most patients with HFpEF have significant hypertension), infiltrative cardiomyopathies, and pericardial thickness/enhancement.

Endomyocardial Biopsy

In cases where cardiac MRI or echocardiography shows significant LV hypertrophy but the patient has low-voltage QRS complexes on electrocardiogram or the patient does not have a long-standing history of hypertension, it is important to perform endomyocardial biopsy to evaluate for infiltrative cardiomyopathies.

When considering the differential diagnosis in a patient with HFpEF, it is important to first make sure that the diagnosis of HF is correct. Mimickers of HF include pulmonary disease, obesity, and anemia, all of which can cause shortness of breath and exercise intolerance. Edema, whether confined to the lower extremity or more generalized, has a large differential diagnosis beyond HF, and includes venous insufficiency or obstruction (eg, venous thrombosis), liver disease, renal disease (eg, nephrotic syndrome), thyroid disease, and protein-losing enteropathies.

Once the diagnosis of HF is confirmed, it is important to make sure that the LVEF has been accurately measured and that ejection fraction is truly preserved. A multiplanar imaging modality (most commonly two-dimensional echocardiography) with quantitative measurement of ejection fraction (ie, biplane method of discs) is essential for ensuring accurate quantitation of global LV systolic function. An increasingly common group of patients with HF are those with a prior history of severe systolic dysfunction (often with LVEF < 25%) who have recovered ejection fraction with medical or device therapies, but who continue to have mild-to-moderate symptoms of HF. These patients, who may have episodic or reversible LV systolic dysfunction, such as patients with tachycardia-induced, alcoholic, viral, or tako-tsubo cardiomyopathy, appear to be distinct from traditional forms of HFpEF.

Once patients are categorized as true HFpEF, the differential diagnosis is broad. Table 27–2 lists the various causes of HFpEF. It is also important to recognize common comorbidities in patients with HFpEF, which may act in concert to cause signs and symptoms of HF. The most important comorbidities include hypertension, diabetes, coronary artery disease, chronic kidney disease, obesity, anemia, and atrial fibrillation, and it is common for multiple comorbidities to coexist in a single patient with HFpEF. Furthermore, many patients have more than one of the aforementioned etiologies of HFpEF. For example, it is not uncommon for an elderly patient to have HFpEF with hypertension, diabetes, chronic kidney disease, severe coronary artery disease, and moderate mitral regurgitation.

When evaluating the underlying cause of HFpEF in the individual patient, it is important to consider the full differential diagnosis before simply concluding that the patient has “garden-variety” HFpEF (ie, HFpEF due to aging and an amalgamation of hypertension, diabetes, obesity, and/or other comorbidities). All patients should be evaluated by echocardiography for the diagnoses of infiltrative cardiomyopathy and constrictive pericarditis. Clues on echocardiography for infiltrative cardiomyopathy include: (1) severe (grade III) diastolic dysfunction (especially if present in a young patient without end-stage renal disease or severe coronary artery disease); (2) low-voltage QRS on electrocardiography with increased LV wall thickness on echocardiography; and (3) severely reduced longitudinal tissue velocities. Clues for the diagnosis of constrictive pericarditis include: (1) history of radiation to the chest wall, cardiac surgery, connective tissue disease, acute pericarditis, or rheumatic heart disease; (2) markers of diastolic dysfunction on mitral inflow and pulmonary vein flow with preserved or increased tissue Doppler e′ velocities; (3) increased respiratory variation in the mitral inflow Doppler tracing; and (4) diastolic interventricular septal “bounce.”

HFpEF often represents the culmination of several underlying comorbidities such as hypertension, diabetes, coronary artery disease, chronic kidney disease, and obesity. Therefore, it is imperative to aggressively treat these risk factors in patients who may be at risk for HFpEF. Aggressive control of hypertension is probably the most important factor in preventing HFpEF and is a class I American College of Cardiology/American Heart Association recommendation for treatment and prevention of HFpEF.

The Valsartan in Diastolic Dysfunction (VALIDD) trial showed that improvement in diastolic function is related to decreases in blood pressure irrespective of type of antihypertensive therapy, and these findings underscore the importance of aggressive blood pressure control in patients with hypertension in order to improve diastolic function. Many studies have shown that lowering blood pressure reduces HF events, and VALIDD adds to these studies by showing improvement in diastolic function with aggressive control of hypertension. Furthermore, the ALLHAT-HFpEF substudy showed that chlorthalidone outperformed other antihypertensives in reducing incident HFpEF.

Poor control of diabetes has been associated with increased incidence of HF (regardless of type of HF). Although there is a tight association between poor glycemic control and incidence of HF, it is unclear whether tight control of diabetes will reduce HFpEF. In the absence of randomized controlled data, patients with diabetes should be treated aggressively with tight glycemic control, although vigilance is needed to prevent hypoglycemic episodes. Tight glycemic control results in prevention of microvascular complications, particularly diabetic nephropathy, which can result in fluid overload and HF.

Treatment of ischemia is an attractive target for prevention of HFpEF. However, there is little data on whether or not there is a benefit of revascularization. In the Coronary Artery Surgery Study (CASS) database, patients with HFpEF had a higher mortality compared with those with preserved ejection fraction but no HF. Despite these findings, there were no differences in survival between patients with HFpEF in CASS who underwent surgical revascularization and those who underwent medical therapy for multivessel coronary disease. Therefore, patients with symptomatic angina and significant coronary disease should be treated with medications or revascularization, or both, but whether doing so will prevent HFpEF remains to be seen.

Nonpharmacologic Therapy

All patients should keep a diary of daily weight and blood pressure. These two parameters are of extreme importance in evaluating for underdiuresis and overdiuresis. When patients are educated about looking for increased weight gain and increasing blood pressure, they can alert their health care provider in a timely fashion in order to allow for intervention prior to progression of HF, which invariably leads to hospitalization.

Several studies have now shown that patients with HFpEF benefit from exercise training. Thus, in symptomatic HFpEF patients who can exercise, prescribed exercise training should be considered.

An innovative wireless hemodynamic monitor (CardioMEMS), implanted percutaneously in the pulmonary artery, was shown in the CHAMPION trial (N = 550) to reduce HF hospitalizations compared to placebo in HFpEF (16% vs. 33%, P < 0.0001). The CHAMPION trial has been the only clinical trial to show a clear benefit in HFpEF, and although not yet U.S. Food and Drug Administration approved, devices like CardioMEMS may play a big role in HFpEF management in the future.

Finally, catheter-based renal denervation therapy has resulted in significant lowering in blood pressure and regression of LV hypertrophy in patients with resistant hypertension. Although clinical trials of renal denervation in HFpEF are lacking, this therapy may be very important in the future for prevention and treatment of HFpEF.

Pharmacologic Therapy

Treatment of HFpEF remains empiric. Compared to HFrEF, there is a relative paucity of randomized controlled trial data to guide treatment. To date, only few large randomized trials have specifically studied patients with HFpEF. The major trials include DIG-PEF, CHARM-Preserved, I-PRESERVE, PEP-CHF, and SENIORS.

In the DIG trial, digoxin did not decrease mortality, and although there was a trend toward decreased hospitalization, there was also a trend toward increased unstable angina. Digoxin increases systolic energy demand and adds to calcium overload in diastole and may be deleterious to patients with HFpEF; therefore, it is generally not recommended. If digoxin is necessary for rate control in patients with HFpEF who have atrial fibrillation, digoxin concentration should be kept at 0.5–0.9 ng/mL since higher concentrations were associated with increased mortality in the DIG trial.

In the CHARM-Preserved trial of mostly male patients with ejection fraction > 40%, candesartan, an angiotensin receptor blocker, was associated with a slight decrease in hospitalization but no difference in mortality when compared with placebo. The large I-PRESERVE trial (N = 4128) enrolled HFpEF patients who more closely matched with those encountered in clinical practice (60% female with a mean age of 72 years). However, I-PRESERVE, which studied irbesartan versus placebo, also found no improvement in outcomes with angiotensin receptor blocker therapy in HFpEF.

The PEP-CHF trial, which studied perindopril (an angiotensin-converting enzyme inhibitor [ACEI]), and SENIORS, which studied nebivolol (a vasodilating β-blocker), both included patients with HF and preserved or relatively preserved ejection fraction, and both did not show major benefits in terms of improved hard outcomes.

Despite these disappointing results, a few recent clinical trials in patients with HFpEF demonstrate that there may be beneficial therapies on the horizon. In a small randomized controlled trial (N = 44) of sildenafil in patients with HFpEF, pulmonary venous hypertension, and superimposed pulmonary arterial hypertension (ie, PADP-PCWP gradient > 5 mm Hg), sildenafil resulted in improved cardiac structure/function and improved exercise capacity. In PARAMOUNT, a phase 2 randomized controlled trial (N = 301) of a novel class of drug (angiotensin receptor neprilysin inhibitor LCZ696), patients with HFpEF randomized to the LCZ696 had greater reduction, compared to placebo, in N-terminal proBNP and left atrial volume at 3 months. LCZ696 was well tolerated, and there were also less serious adverse outcomes in the LCZ696 arm, although this finding did not reach statistical significance (15% vs. 30%, P = 0.32).

Since extensive randomized controlled trial data for HFpEF are not available, treatment of HFpEF relies on extrapolation of therapies for HFrEF, nonspecific relief of congestion, and ameliorating the underlying disease processes and comorbidities. Table 27–3 lists the most important treatment priorities for patients with HFpEF, as outlined below.

For symptomatic relief, the most important first step is to reduce the congestive state. Salt restriction and vasodilator therapy (ACEI, angiotensin receptor blockers, or hydralazine/nitrates) make up the cornerstone of treatment. Diuretics and other forms of fluid removal (eg, dialysis, ultrafiltration) are often needed, but as the acute congestive episode resolves, it is important to minimize diuretic therapy, since overdiuresis activates a heightened neurohormonal response and aggravates the cardiorenal syndrome.

From an electrophysiologic standpoint, it is important, whenever possible, to maintain atrial contraction and atrioventricular synchrony. Therefore, patients with atrial fibrillation or atrial flutter should undergo cardioversion or ablation. When necessary, patients should undergo pacemaker therapy to ensure atrioventricular synchrony or to treat chronotropic incompetence.

In some patients with HFpEF, it is ideal to promote bradycardia and avoid tachycardia. Tachycardia increases myocardial oxygen demand and decreases coronary perfusion time, which promotes diastolic dysfunction due to ischemia even in the absence of epicardial coronary disease. In addition, the time allotted for LV relaxation is decreased and diastolic filling time is decreased when tachycardia is present. By inducing relative bradycardia (eg, heart rate 50–60 bpm), coronary perfusion is optimized, and LV relaxation and diastolic filling time are both increased. Patients who benefit most from this type of treatment are most likely those who have impaired LV relaxation and prolonged early mitral inflow deceleration times. In patients with more severe, end-stage HFpEF, such as severe restrictive cardiomyopathy, increased heart rate may be the most important factor maintaining cardiac output, since stroke volume is often decreased and fixed. These patients invariably have a very high early mitral inflow velocity and short deceleration time. Although tachycardia should be avoided, heart rates of 80–90 bpm are often required in order to maintain adequate cardiac output. In these patients, overzealous β-blockade or nondihydropyridine calcium channel blocker therapy can result in a precipitous decline in cardiac output.

Most patients with HFpEF will benefit from rate control therapy with medications such as β-blockers and nondihydropyridine calcium channel blockers (verapamil, diltiazem). A good rule of thumb to follow when choosing β-blockers is that metoprolol succinate is a good agent in patients who have problems with rate control (eg, atrial fibrillation) and those with low or normal blood pressure. In patients who have severe hypertension, carvedilol is the agent of choice since it has potent antihypertensive effects due to its α-adrenergic blockade properties. Most patients with HFpEF have significant hypertension; thus carvedilol is typically a first-line drug for most HFpEF patients. In the COHERE registry, which was an observational study, treatment with carvedilol resulted in improved outcomes compared to other β-blockers, regardless of the underlying LV ejection fraction.

As stated earlier, all patients should be evaluated for myocardial ischemia, and when present, ischemia should be treated aggressively with revascularization, β-blockers, nitrates, and dihydropyridine calcium channel blockers. Hypertension should be treated aggressively with goal blood pressure < 130/80 mm Hg. Control of hypertension is the only proven therapy for prevention of HFpEF, and therefore is essential in all patients. HFpEF patients commonly have severe hypertension, and when they are referred, they may be taking four or five or more antihypertensive medications. The number of medications should be kept to a minimum in order to avoid the dangers of polypharmacy and adverse drug–drug interactions. In addition, minimizing medications will often promote increased patient compliance. Patients with HFpEF and severe hypertension are often taking medications such as clonidine and minoxidil, while other medications with proven cardiovascular benefits, such as β-blockers and ACEIs, are not titrated to maximum doses. Therefore, patients with significant hypertension should ideally be treated with a vasodilating β-blocker and maximum dose of an ACEI or angiotensin receptor blocker, unless contraindicated. Most patients will also benefit from a thiazide diuretic such as chlorthalidone. Routine use of more potent thiazides, such as metolazone, should be avoided since these medications often exacerbate the cardiorenal syndrome. Aldosterone antagonists can be very useful in HFpEF patients both for controlling blood pressure and for improving diuresis. However, these patients should be monitored closely for hyperkalemia. The utility of spironolactone in HFpEF is currently being evaluated in the TOPCAT randomized controlled trial.

In patients with severe, resistant hypertension who cannot be treated adequately with a combination of β-blockers, ACEIs, and thiazide diuretics, the following steps should be taken: (1) ensure medication compliance, (2) ensure euvolemia since fluid overload will exacerbate hypertension, and (3) look for causes of secondary hypertension. Using these steps, most patients will have adequately controlled blood pressure with two or three medications. In patients who need an additional agent, the combination of hydralazine and nitrates is a good option given their beneficial effects in HF. If tolerated and not associated with increased lower extremity edema, dihydropyridine calcium channel blockers may also be used to treat severe hypertension. These agents are often useful in patients who have significant chronic kidney disease since they may not be able to take ACEIs, angiotensin receptor blockers, or aldosterone antagonists.

Besides beneficial antihypertensive effects, β-blockers, ACE inhibitors, and angiotensin receptor blockers attenuate neurohormonal activation, which may be beneficial in HFpEF. In addition, ACEIs, angiotensin receptor blockers, and spironolactone may prevent fibrosis and may promote regression of LV hypertrophy. Statins may have pleiotropic benefit in HF, and all HFpEF patients with dyslipidemia or coronary risk factors should be treated with a statin. Interestingly, in a large study of Medicare beneficiaries discharged with a primary diagnosis of HF and documentation of preserved ejection fraction, statins were associated with increased survival irrespective of total cholesterol, coronary disease, diabetes, hypertension, or age.

Categorization of HFpEF Subtype

As stated earlier, HFpEF is a heterogeneous syndrome. Once the HFpEF syndrome is diagnosed, patients can be further categorized clinically into etiologic and pathophysiologic subtypes in order to help guide therapy above and beyond the recommendations listed above. Table 27–4 summarizes these HFpEF subtypes and provides guidance to HFpEF subtype-specific treatment options.

Caveats

Some patients with HFpEF live a delicate balance between symptomatic congestion (due to inadequate diuresis) and poor cardiac output (due to overdiuresis). The latter causes lightheadedness, dizziness, fatigue, and worsening renal dysfunction due to decreased renal perfusion. Patients with HFpEF who have a classic physiologic picture of isolated “diastolic HF” rely on increased LV filling pressures to maintain cardiac output, and they tend to be very sensitive to overdiuresis, with small decreases in LV diastolic pressure resulting in large decreases in stroke volume. Therefore, it is important to start low and go slow with diuretic therapy. Many patients typically require frequent visits in order to find a diuretic regimen that results in optimal symptom control without exacerbating the cardiorenal syndrome.

In patients with hypertrophic cardiomyopathy, nondihydropyridine calcium channel blockers (verapamil, diltiazem) can be beneficial and therefore may be used as first-line therapy before β-blockade. Thiazide diuretics can exacerbate hyperglycemia and hyperuricemia, which are often present in elderly patients with HFpEF. Positive inotropes should be avoided in general because they promote calcium influx into cardiac myocytes, which worsens diastolic function. In addition, many of these patients have hypercontractile ventricles with small LV volumes. Therefore, positive inotropes frequently cause cavity obliteration with resultant obstruction of forward flow and decreased cardiac output. In patients with HFpEF who have non-ST elevation acute coronary syndromes, mortality is increased and patients are often undertreated. Therefore, all efforts should be made to treat this high-risk group (including an early invasive approach) using evidence-based guidelines.

Aside from treatment of hypertension, coronary disease, and atrial fibrillation (as listed earlier), it is important to treat other underlying comorbidities such as diabetes, metabolic syndrome, obesity, chronic kidney disease, and anemia. In addition, many patients with HFpEF have concomitant chronic obstructive pulmonary disease, which should also be treated aggressively in order to improve symptoms of breathlessness.

Drugs to Avoid

In all patients with HFpEF, it is important to avoid polypharmacy at all costs since adverse events increase and compliance decreases with increased numbers of medications. In addition, medications should always be carefully scrutinized as causes of signs and symptoms of HF. For example, calcium channel blockers and thiazolidinediones (eg, pioglitazone) can cause significant edema, nonsteroidal anti-inflammatory drugs (NSAIDs) can cause renal failure, and hydroxychloroquine (which is frequently used in rheumatologic diseases such as systemic lupus erythematosus and rheumatoid arthritis) can cause a restrictive cardiomyopathy. In the elderly cohort of patients with HFpEF, medications for Parkinson disease are common, and these agents have been shown to cause valvular disease. Certain foods and herbal supplements can also be deleterious in patients with HFpEF. Licorice can cause mineralocorticoid excess, ginseng interferes with warfarin, and ginseng also falsely elevates digoxin levels. Treatment of gout is difficult since NSAIDs are contraindicated in patients with HFpEF and colchicine is dangerous because many of these patients are elderly and have abnormal kidney function. In these patients, corticosteroid injection directly into the involved joint, a short pulse of oral corticosteroids, and colchicine 0.6 mg three times a week or allopurinol for maintenance therapy may be the best treatment options.

Once hospitalized for HF, patients with HFpEF have a high mortality. Five-year mortality is high and is similar to HFrEF. Two large epidemiology studies of patients hospitalized with HFpEF have shown that survival is only 30–40% after 5 years. A patient-level meta-analysis, which included both epidemiologic studies and clinical trials and inpatient and outpatient settings, found that survival of patients with HFrEF was worse than HFpEF, although both types of HF were associated with high mortality rates.

Despite the abundance of studies now showing a high rate of mortality in HFpEF, the cause of death remains unclear. Predictors of death in patients hospitalized with HFpEF include increased age, male gender, increased serum creatinine, decreased hemoglobin, decreased blood pressure, increased respiratory rate, and hyponatremia. In addition, diabetes, peripheral arterial disease, cancer, dementia, and end-stage renal disease requiring hemodialysis are all independent predictors of death in HFpEF. In the CHARM trial echocardiographic substudy, evidence of moderate or greater diastolic dysfunction was strongly associated with adverse outcomes. In the DIG trial, worsening HF and arrhythmias were the leading causes of death in all HF patients, regardless of ejection fraction. However, noncardiovascular causes of death were more frequent in patients with HFpEF, due to their advanced age. Several studies, including analyses of the CHARM-Preserved and I-PRESERVE trials, as well as a retrospective analysis of the Duke Databank for Cardiovascular Disease, have found that sudden cardiac death occurs in HFpEF with a frequency ranging from 8–28%.

Cause of death in HFpEF is most likely multifactorial. Although some patients may die of HF and pulmonary edema, most patients do not die of LV pump failure, and as stated earlier, many may die due to arrhythmias or sudden cardiac death. Even more may die due to important age-related comorbidities such as cancer and dementia. Until we learn more about cause of death in patients with HFpEF, it is important for health care providers to understand the importance of comorbidities in the prognosis of these patients. Therefore, although evaluation of treatment of HFpEF is extremely important, attention to comorbidities and a multidisciplinary approach to patients with HFpEF will likely lead to the best possible outcome in these patients.