Definition of Heart Failure, 418
Epidemiology of Heart Failure, 418
Objectives of Cardiac Imaging in Heart Failure, 419
Cost of Imaging Tests, 419
Heart Failure With Reduced Ejection Fraction Versus Heart Failure With Preserved Ejection Fraction, 419
Evaluation of Left Ventricular Diastolic Dysfunction, 419
Multiple Modality Cardiac Imaging, 420
Assessment of Left Ventricular Function by Echocardiography, 421
Myocardial Strain and Strain Rate, 421
Nuclear Cardiology: Radionuclide SPECT and PET, 423
Assessment of Physiologic Ischemia, 424
Assessment of Right and Left Ventricular Volumes and Function, 424
Imaging Autonomic Dysfunction, 425
Computed Tomography, 425
Cardiac Magnetic Resonance Imaging, 425
Valvular Heart Disease and Heart Failure, 426
Mitral Valve Regurgitation, 426
Aortic Valve Disease as a Cause of Heart Failure, 428
Aortic Stenosis, 428
Cardiovascular Magnetic Resonance Assessment of Valvular Heart Disease, 429
Myocardial Viability, 430
Stress Echocardiography, 430
Radiotracer-Based Assessment of Myocardial Viability and Remodeling, 430
Cardiovascular Magnetic Resonance Stress, 430
Detection of Myocardial Ischemia with Cardiovascular Magnetic Resonance, 430
Detection of Infarction with Cardiovascular Magnetic Resonance I, 431
Transition From Myocardial Infarction to Heart Failure With Reduced Ejection Fraction, 431
Cardiac Magnetic Resonance Infarction II, 432
Cardiac Resynchronization Therapy, 432
Nuclear Imaging for Cardiac Resynchronization Therapy, 433
Cardiovascular Magnetic Resonance for Resynchronization Therapy, 433
Idiopathic Dilated Cardiomyopathy, 433
Miscellaneous Causes of Heart Failure 434
Assessment of Right Ventricular Function, 435
Complications of Heart Failure, 436
Future Directions in Cardiac Imaging, 436
Nuclear Imaging, 436
Injury and Inflammation, 436
Apoptosis and Cell Death, 436
Imaging Inflammation, 437
Imaging Inflammatory Cell Activity, 439
Ventricular Remodeling, 440
Metabolism and Ischemic Memory, 443
Cardiac Magnetic Resonance, 445
Four-Dimensional Flow for Cardiac Resynchronization Therapy Optimization, 445
Creatine Chemical Exchange Saturation Transfer Imaging, 445
Cardiovascular Magnetic Resonance Assessment of Myocardial Fiber Orientation, 445
Cardiovascular Magnetic Resonance Molecular Imaging of Apoptosis, 447
Definition of Heart Failure
Heart failure can be defined as a clinical syndrome caused by an abnormality of cardiac structure or function that results in failure to deliver oxygen at a rate commensurate with the needs of the body tissues (systolic failure), or failure to receive blood at normal filling pressures (diastolic failure). The diagnosis of heart failure may be difficult to establish clinically, especially during its early stages because the symptoms and physical signs are nonspecific, consisting of dyspnea, effort intolerance, fatigue, elevated jugular venous pressure, and lower extremity edema ( see also Chapter 31 ). As a result, cardiac imaging has assumed a pivotal role in supporting early diagnosis and in guiding optimal management of patients with heart failure from acquired and congenital heart disease.
Epidemiology of Heart Failure (see also Chapter 18 )
Heart failure currently affects approximately 30 million people worldwide, nearly six million of whom reside in the United States. One in nine deaths in 2009 were attributable to heart failure, and about half of the patients with heart failure die within 5 years . There are an additional 650,000 new cases of heart failure diagnosed each year in the United States, which is, in part, due to improved noninvasive cardiac imaging methods such as transthoracic two-dimensional (2D) echocardiography (TTE). The incidence of heart failure increases with advancing age, so that approximately 10% of all males and females over 70 years of age have heart failure. In addition, heart failure is the most frequent hospital discharge diagnosis in patients >65 years old. Heart failure costs the nation an estimated $30.7 billion each year, including costs related to health care services, diagnostic imaging, medications, and missed days of work. Furthermore, the increase in life expectancy over the last three decades predicts that by the year 2035 there will be approximately 70 million subjects in the United States over 75 years of age, of whom 7 million (10%) will have heart failure. For this reason, heart failure has been targeted as a major health care initiative.
Objectives of Cardiac Imaging in Heart Failure
There are three major objectives of cardiac imaging in the setting of heart failure. The primary objective of noninvasive and invasive cardiac imaging is to establish the definitive cardiac diagnosis de novo , or to confirm the clinically suspected diagnosis. The secondary objective is to acquire reproducible high-quality, high-resolution images that enable accurate quantitative assessment of cardiac chamber size, architecture, global, and regional left ventricular (LV) function. The tertiary objective is to relate metrics of cardiac chamber size and function to risk stratification and long-term clinical outcome.
The aims of this chapter are to discuss the optimal and appropriate use of the panoply of multi-modal cardiovascular imaging techniques that are currently used routinely in patients with heart failure. We do not wish to compare and contrast the individual strengths and weaknesses of each of the various imaging modalities in each clinical situation, but rather to describe the most efficacious contemporary use of imaging modalities for a range of specific etiologies of heart failure. Special attention is given to a number of different pathoetiologies of heart failure that include (1) systolic and diastolic heart failure, (2) valvular heart disease, (3) myocardial viability and ventricular remodeling postinfarction, (4) detection of myocardial fibrosis, (5) selection of patients for device deployment, (6) dilated cardiomyopathy (DCM), and (7) right heart failure .
Choosing the cardiovascular imaging modality best suited to resolve the clinical differential diagnoses and to safely guide therapy can be challenging. This is because many of the cardiovascular symptoms in heart failure are nonspecific and correlate poorly with the degree of ventricular dysfunction.
Cost of Imaging Tests
Cardiac imaging is a frequent costly component of cardiovascular health care. Thus, when choosing a cardiovascular imaging modality it is important to be cognizant of the cost effectiveness of each modality relative to the clinical question. For example, echocardiography is the method of choice in patients with suspected HF for reasons of accuracy, availability (including portability), safety, and cost. However, echocardiography is frequently complemented by other modalities based on the specific clinical questions. Adherence to the American College of Cardiology Foundation/American Heart Association (AHA) guidelines for the diagnosis and treatment of heart failure should dovetail with the appropriate use of multimodal imaging. The ultimate aims of cardiovascular imaging are to maximize diagnostic accuracy and to improve patient care and clinical outcomes cost effectively, using evidence-based medicine to guide the appropriate use of our limited health care system resources.
Heart Failure With Reduced Ejection Fraction Versus Heart Failure With Preserved Ejection Fraction
Recent studies place the prevalence of heart failure with preserved ejection fraction (HFpEF) at around 54% with a range from 40% to 71% ( see also Chapter 39 ). Clinical presentation of patients presenting with the new onset of heart failure and reduced LV ejection fraction (HFrEF/systolic heart failure) cannot always be reliably distinguished clinically from HFpEF patients. Exam findings, such as lateral displacement of the LV apical impulse, as a clinical clue to LV dilatation often cannot be appreciated in HFrEF due to body habitus, chronic lung disease, or a diffuse and weak apical impulse. LV dilatation can be easily confirmed by chest x-ray (CXR) or TTE. LV dilatation and decreased ejection fraction (EF) are crucial for the diagnosis of HFrEF. There is a preponderance of elderly women with systemic hypertension in HFpEF as compared with HFrEF. More than two thirds of patients with HFrEF have an ischemic etiology for their LV dysfunction. The diagnosis of HFpEF can strictly only be made by the combination of clinical demographics and computational image analysis, which provides an estimation of EF and an assessment of the severity of LV diastolic dysfunction with TTE. HFpEF has only recently been acknowledged as a discrete pathophysiologic entity because formerly it was considered to be a benign condition for which there is still no specific treatment. However, studies have demonstrated that HFpEF is a discrete entity with a significant annual morbidity and mortality, and a readmission rate for acute exacerbations of heart failure. Doppler echocardiographic imaging can distinguish HFpEF from HFrEF because in HFrEF there is obligatory LV dilatation to maintain a normal stroke volume as LVEF declines. In contrast, LV cavity size in HFpEF is normal or small and there is usually mild to moderate concentric hypertrophic remodeling ( Fig. 32.1 A and B) induced by concomitant hypertension (HTN), with preserved left ventricular ejection fraction (LVEF) (≥50%).
Evaluation of Left Ventricular Diastolic Dysfunction
Diastolic dysfunction is the underlying pathophysiologic abnormality that typifies HFpEF. Diastolic LV dysfunction as assessed by echocardiography includes an abnormal transmitral peak E wave, a peak A wave that ranges from delayed relaxation to irreversible restrictive physiology, and severely elevated filling pressures ( Fig. 32.2 ). (E/e′), reduced e′ myocardial velocities by tissue Doppler imaging in the septum and lateral wall, are consistent with delayed relaxation and an enlarged left atrium. In addition, there is consistently abnormal pulmonary venous flow (see Fig. 32.2 ). Diastolic ventricular function also can be assessed with nuclear techniques from ventricular volume curves in terms of LV peak filling rate (PFR), time to PFR, and filling fractions. However, there is no single echo, Doppler, or nuclear parameter that is sufficiently robust to diagnose the presence of LV diastolic dysfunction.
The reason for distinguishing HFpEF from HFrEF patients is because they have different LV morphology, epidemiology, and mechanisms of heart failure systolic and diastolic LV function—all of which are evident by 2D and 3D transthoracic echocardiographic imaging . Furthermore, patients with HFpEF do not derive a clear benefit from traditional heart failure therapy, which includes β-adrenergic receptor blockers, angiotensin converting enzyme inhibitors, angiotensin receptor blockers, and mineralocorticoids inhibitors. Heart size in HFpEF remains stable although LV hypertrophy may increase, and diastolic dysfunction may worsen over time. In contrast, heart size in HFrEF is increased at the time of diagnosis and increases progressively thereafter ( Fig. 32.3 ). The prognosis of HFpEF is better than in HFrEF, although the rate of hospital re-admission for recurrent heart failure is similar. Echocardiographic changes in LV size, geometry, hypertrophy, and function have been reported prospectively in multiple studies of systolic heart failure/HFrEF for more than three decades. By comparison there is a paucity of Doppler echocardiographic information on HFpEF available for assessment of diastolic LV function over time, except for studies from the Framingham database.
Multiple Modality Cardiac Imaging
Cardiac diagnostic imaging techniques are used routinely in heart failure and range from simple assessment of overall heart size, cardiac silhouette, and the presence of pulmonary congestion by CXR to real-time 3D-Doppler echocardiographic reconstruction of the heart and the great vessels, or CMR perfusion showing delayed gadolinium enhancement due to myocardial fibrosis. Although computed tomography (CT) provides spectacular high-resolution cardiac imaging, the doses of ionizing radiation and of iodinated contrast agents precludes serial studies that are required in patients with heart failure undergoing LV remodeling.
CXR is obtained in all heart failure patients for the detection of cardiomegaly, individual chamber and great vessel enlargement, pulmonary venous congestion, and pericardial and pleural effusions that are frequent accompaniments of decompensated heart failure. Even minor hemodynamically irrelevant pericardial effusions in heart failure are associated with increased risk of cardiac mortality.
Echocardiography is the imaging modality of choice in all causes of heart failure because it is portable, safe, and provides accurate quantitative information at the bedside, which includes LV volumes, chamber architecture, regional and global systolic and diastolic function, valve function, and pulmonary artery systolic pressure, all of which correlate with clinical outcomes. In addition, Doppler measures of intracardiac blood flow velocities and myocardial velocities permit the calculation of myocardial strain and torsion/rotation that enable the complete assessment of global and regional myocardial mechanics. A great advantage of 2D and 3D echocardiography is that all of this information regarding myocardial mechanics is immediately available at the bedside and, furthermore, can be repeated safely as needed. However, misgivings have been expressed recently regarding the use of 2D echocardiography to assess LV volumes, EF, and LV mass in serial studies because of the poor test and retest reproducibility and the magnitude of the standard deviations derived from a meta-analysis, which involved a large number of studies. However, these findings are discordant with a number of recent studies in which serial echoes were performed and consistently demonstrated important reproducible changes in LV size, mass, and function following intervention.
Assessment of Left Ventricular Function by Echocardiography
M-Mode echocardiography has been used for measurements of LV size, mass, and loading conditions (end-systolic meridional and circumferential wall stress), and myocardial function (fractional and mid-wall shortening and velocity of circumferential fiber shortening peak). However, M-Mode echo assessments of LV function are limited because it is assumed that there is uniform wall thickness and normal concentric wall motion, whereas the majority of patients with HFrEF have coronary artery disease (CAD) and ischemic cardiomyopathy in which the hallmarks of CAD are LV wall motion abnormalities and variations in LV wall thickness. Thus M-Mode echo measurements of LV size and function are not admissible in over two-thirds of the HFrEF population. LV linear dimensions are still used in randomized clinical trials in hypertension that require serial measurements with or without an intervention as the primary outcome measure.
TTE is the most frequently used, and clinically the most important, diagnostic imaging modality in HFrEF and in HFpEF. LV end-systolic, end-diastolic volumes and stroke volume can be calculated from biplane orthogonal images of the LV in the apical 4-chamber and the apical 2-chamber planes using Simpson’s method of discs. LV mass (LVM) also can be estimated from measurements of end-diastolic wall thickness, LV cavity diameter/2 and LV length (5/6 short-axis area × length). The important fundamental relation between LV volume and mass can be examined. Estimates of LV volumes, LVM, and EF by 2D echo provide insight into the structural, geometric, and functional changes in the left ventricle as the heart remodels and the severity of heart failure progresses. 2D Doppler echocardiography has played a major role in elucidating our current understanding of the different natural histories and etiological mechanisms involved in HFpEF and HFrEF.
In addition to TTE determination of LV volumes, mass (LVH as increased relative wall thickness), LV shape, and EF have proved to be powerful predictors of clinical outcome in patients with heart failure. LV stroke volume can be calculated from LV volumes (EDV – ESV = SV) by 2D and 3D TTE, and by Doppler measurement of intracardiac blood flow velocities in the LV outflow tract (LVOT). Blood volume flow in unit time can be assessed as the product of the time velocity integral (TVI) and the cross-sectional area (CSA) of the flow stream (TVI × CSA). When recorded from the LVOT, stroke volume correlates closely with stroke volume estimated from LV volumes. Recently, the LVOT has been shown to be elliptical rather than circular by transesophageal echocardiography (TEE), CMR, and CT, especially in the presence of LVH that protrudes into the LVOT in patients with hemodynamically important aortic stenosis (AS) and severe systemic hypertension. The influence of an abnormal LV outflow tract cross-sectional shape can be minimized by planimetry analysis of the cross-section directly.
Myocardial Strain and Strain Rate
Measurement of myocardial strain is a relatively new concept that describes global and regional ventricular systolic function using speckle-tracking echocardiography or magnetic resonance with myocardial tagging. Speckle tracking echocardiography depends upon the temporal and spatial tracking of naturally occurring intramyocardial reflectors of ultrasound (speckles) within the 2D echocardiographic images of the LV walls. Displacement of these speckles is due to myocardial deformation from which myocardial strain is calculated. Strain is defined as the change in myocardial segment length (ΔL) divided by resting segment length (L 0 ): S = ΔL / L 0 . The insonating beam is directed parallel to the LV long axis. Myocardial strain is assessed in three planes: longitudinal, circumferential, and radial.
Longitudinal strain is calculated from the LV long axis, and the radial and circumferential strains from the LV short-axis images obtained at the LV mid-cavity level ( Fig. 32.4 ). Estimates of myocardial strains by speckle tracking echocardiography have been validated in man by CMR with myocardial tagging and in animals by sonomicrometry. Global systolic strains can be assessed in addition to simultaneous assessment of myocardial strains in each segment using the 16- or 17-segment model of the LV. Myocardial strains can be recorded simultaneously from the interventricular septum and from the lateral LV wall in each of three myocardial segments (apical, mid, and proximal) from the apical 4-chamber, 2-chamber, and apical long-axis views (see Fig. 32.4 ). Strain analysis can be quantified after acquisition of the echo images. Measurement of the time period from onset of QRS to peak strain for each myocardial segment provides insight as to the coordination of contraction and the degree of dyssynchrony. Strain can detect mild perturbations in LV function before any change is detectable in LV volumes or EF. The strain rate is the rate of change of strain, which has not always proved as robust or reproducible as the measurement of deformation due to a number of confounding factors.
3D echocardiography: Numerous studies have demonstrated the correlations between LV volumes, mass, and EF estimated by 2D and CMR. Still closer correlations have been demonstrated between real time 3D echocardiography and CMR with less variability about the mean.
Real time 3D echocardiographic assessment of LV volumes, mass, and LVEF ( Fig. 32.5 ) correlates more closely to CMR than 2D Echo, with less variability than with 2D. However, the greater precision of measurement of LV mass and EF by CMR is the reason that CMR has become the standard of reference for quantification of LV mass and LV volumes.
A proportion of echocardiograms in heart failure patients are technically limited because endocardial definition is incomplete, resulting in poor image quality, necessitating interpolation of extensive regions of the LV endocardium. This occurs especially in patients with emphysema or morbid obesity that may even preclude quantitative analysis. However, endocardial definition is improved with harmonic imaging and can be further enhanced with intravenous echo contrast so that a proportion of poor-quality studies can be recovered for quantitative analysis.
TEE: When image quality is poor, switching to TEE or to an alternative imaging modality, such as CMR, for better image quality may be a better strategy. However, there is a trade-off for the exquisite image quality attained by TEE, and that is that quantitation of LV volumes from biplane TEE images consistently underestimates volumes calculated by TTE. This occurs because of unavoidable foreshortening of the LV long axis in the apical imaging planes with TEE. This foreshortening artifact results in the underestimation of LV volumes, EF, and longitudinal strain.
TEE is not indicated in the routine assessment of patients with heart failure except in special circumstances, which include poor image quality, suspected vegetative endocarditis, assessment of magnetic resonance due to papillary muscle infarction, and occasionally to measure left atrial size.
An additional important role performed by 2D echocardiography is the detection of LV cavity thrombus ( Fig. 32.6 ) adherent to severely hypokinetic or akinetic myocardial segments. Thrombus formation also occurs in the left atrial cavity and/or appendage, especially in patients with left atrial enlargement and atrial fibrillation. In patients with unexplained worsening symptoms, 2D echocardiograms should be performed to rule out significant pericardial effusion, pleural effusion, or the onset of atrial fibrillation, which is a common dysrhythmia in heart failure.
Nuclear Cardiology: Radionuclide SPECT and PET
Single photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging have become standard approaches for quantitative physiologic imaging in patients with HF. Radiotracer techniques have been widely used to evaluate regional and global ventricular function, and to detect myocardial ischemia, hibernation, infarction, and ventricular remodeling. Nuclear imaging techniques also are well suited for in vivo molecular imaging because of their high sensitivity, spatial resolution, and availability of new hybrid instrumentation and molecular targeted probes. Molecular imaging offers a novel approach for detecting molecular or cellular changes in vivo before the development of any physiological or anatomic changes. Recently dedicated hybrid imaging systems combining nuclear detector systems with high resolution structural imaging modalities such as x-ray CT or magnetic resonance imaging have been introduced into routine clinical practice. In addition to aiding the interpretation of the SPECT/PET findings, anatomical information in hybrid imaging facilitates correction for attenuation, scatter, and partial volume effects, resulting in enhanced image quality, dose reduction, and radiotracer quantification. Beyond this, hybrid systems have the potential to provide independent and real-time synergistic data that improve disease characterization and patient care, as recently described for PET/magnetic resonance hybrid scanners.
Assessment of Physiologic Ischemia
In the evaluation of patients with heart failure it is important to rule out ischemic heart disease as the etiology of the LV dysfunction because of the potential to impact ventricular dysfunction by revascularization ( see also Chapter 19 ). This can be accomplished by an assessment of stress-induced changes in regional perfusion or function. Rest and stress myocardial perfusion imaging can be accomplished with either SPECT or PET perfusion imaging. Importantly, radionuclide myocardial perfusion imaging effectively visualizes regional changes in myocardial blood flow, which is a principal target of many therapies in patients with CAD. Rest and stress radiotracer studies continue to play a major role in the diagnosis of CAD, which is the major cause of HF. Thus, stress/rest SPECT perfusion imaging can effectively separate ischemic from nonischemic cardiomyopathy.
The care of patients with acute myocardial infarction (MI) is directed at establishing early coronary reperfusion, since aborting ischemia may result in myocardial salvage. Myocardial salvage results in preservation of LV function, which is the most important predictor of long-term survival postinfarction. Radiotracer imaging has proved effective in estimation of infarct size and salvage of cardiomyocytes after MI. The radiotracer technique also provides a reliable estimation of residual LV function. There is a well-established relationship between survival and global RV and LV function in patients undergoing reperfusion with thrombolytic therapy or percutaneous coronary intervention (PCI).
Assessment of Right and Left Ventricular Volumes and Function
The traditional radiotracer approaches to assess ventricular function included first-pass radionuclide angiocardiography, and equilibrium radionuclide angiocardiography (ERNA), which use imaging of Tc99m-labeled red blood cells (RBCs). The first-pass technique permits assessment of global right and left ventricular size and function, but is inadequate for the evaluation of regional LV function, because only a single projection of the ventricle is obtained. To assess RV and LV functional reserve by the first-pass technique, separate injections of the radiotracer are made at rest and during peak exercise. The first-pass technique has been replaced by ERNA, which, in turn, has been superseded by gated SPECT blood pool angiography and gated SPECT perfusion imaging. Analysis of temporal changes in count density from 4D SPECT images provides an index of regional LV wall thickening. New 3D radiotracer-based imaging approaches offer a more comprehensive evaluation of regional and global LV function. Serial equilibrium blood pool imaging can be performed at rest and during various levels of exercise or after pharmacologic perturbations to evaluate ventricular functional reserve. Gated SPECT perfusion imaging is primarily restricted to the evaluation of resting function.
LV end-diastolic (LVED) and end-systolic volumes can be evaluated serially using nuclear approaches in patients with HF and can be used to track LV remodeling and to monitor therapy. Volumes from gated blood pool images are calculated based on radiotracer count density and are therefore relatively independent of alterations in regional geometry. Simple count-based techniques allow volume measurements to be made without the confounding technical issues of accurate measurement of attenuation. These estimates of volumes can be improved by applying 3D imaging techniques.
Assessment of diastolic function is also important in the evaluation of patients with HFpEF and with HFrEF in the elderly with hypertension and/or CAD. ERNA results have high reproducibility because there are no geometric assumptions and there is much less operator dependence in image acquisition. Diastolic parameters, such as filling rate, time to peak filling, and filling fractions, can be readily assessed from the ventricular volume curve. Studies have demonstrated good correlation between echocardiography and ERNA for reliable determination of diastolic parameters. In patients with diastolic dysfunction, there will be a prolongation of isovolumic relaxation time, a delay in the onset of rapid filling, a decrease in slope of rapid filling phase, and an exaggerated atrial kick. The lower limit of normal PFR is 2.50 end-diastolic volumes per second (EDV/s). In addition, time to PFR can be expressed in milliseconds and is expected to be less than 180 milliseconds in normal subjects. The PFR and the atrial filling rate have been shown to correspond to the E and A waves of Doppler echocardiographic mitral velocity wave forms. When ERNA is used, such measurements should be routine in the assessment of heart failure in the presence or absence of CAD.
Imaging Autonomic Dysfunction
Cardiac autonomic dysfunction is associated with an increased risk of ventricular arrhythmia and sudden cardiac death in heart failure ( see also Chapter 42 ). Alterations in pre- and postsynaptic cardiac sympathetic function can be assessed noninvasively using both SPECT and PET radiotracers. The most widely used SPECT radiotracers for imaging of presynaptic function is 123 I-meta-iodobenzylguanidine ( 123 I-MIBG), which shares many cellular uptake and storage properties with norepinephrine. Many studies have demonstrated the clinical value of 123 I-MIBG imaging for both diagnostic and prognostic purposes in patients with heart failure. In these patients, 123 I-MIBG scans typically show a reduced heart-to-mediastinal uptake ratio (HMR), heterogeneous distribution within the myocardium, and increased 123 I-MIBG wash-out from the heart. HMR is a marker of specific sympathetic nerve terminal tracer retention and has prognostic value in heart failure. The wash-out ratio of 123 I-MIBG predicts sudden cardiac death, independent of LVEF. A large prospective study of 123 I-MIBG imaging demonstrated a significant relationship between the heart failure related events and the HMR, which was independent of LVEF and BNP. This clinical study also showed an association between myocardial sympathetic neuronal dysfunction and the risk for subsequent cardiac death. Moreover, the size of the MIBG defect on delayed SPECT imaging also predicts ventricular arrhythmias ( Fig. 32.7 ).
Presympathetic function also can be assessed by PET imaging with 11 C-meta-hydroxyephedrine ( 11 C-HED). In nontransmural MIs, the HED imaging defect can exceed the perfusion defect, but it is not clear whether this is associated with higher ventricular arrhythmia risk. In patients with nonischemic cardiomyopathy, there is decreased HED uptake, and in a recent retrospective study, global HED uptake was an independent predictor of adverse outcomes in patients with New York Heart Association (NYHA) class II and III heart failure ( Fig. 32.8 ). This PET sympathetic imaging approach is under evaluation in patients with CAD and depressed LVEF. A recently completed single site prospective clinical study (PAREPET) showed that quantitative 11 C-HED PET imaging was one of the best predictors of sudden cardiac death, independent of LVEF and infarction size.
CT and CMR imaging both produce exquisite image quality with and without contrast enhancement and allow comprehensive quantitative assessment of LV and RV architecture and function as well as delineation of the coronary artery anatomy. The disadvantage of CT is the exposure to both ionizing radiation and iodinated contrast agents. Recently, attempts have been made to minimize radiation exposure during CT and have had a modicum of success.
Cardiac Magnetic Resonance Imaging
The acquisition of such high-fidelity image quality is the reason that CMR has become the standard of reference for quantification of LV volumes, LVEF, and LV mass ( Fig. 32.9A ). The accuracy and reproducibility of CMR makes it the ideal tool for serial assessment of ventricular size and function, and at the same time it reduces the sample sizes necessary in clinical trials. However, a substantial proportion of heart failure patients have ICDs, permanent pacemakers, and there is an increasing number of CRT device implants in whom CMR imaging is currently contraindicated. The numerical data generated in healthy normal subjects and in acute and chronic heart failure by CMR are commonly used to risk stratify patients with heart failure. Serial imaging of the heart in patients with heart failure is frequently used to assess the rate of disease progression, and the response to pharmacologic agents, devices, and surgical and cellular therapies. Serial evaluation of large patient cohorts has also been used to assess the efficacy of novel pharmacologic agents and device therapies for heart failure in large, randomized, clinical trials. Gadolinium-containing CMR contrast agents have been used safely in many millions of patients. A rare disorder known as nephrogenic sclerosing fibrosis was reported as a potential side effect following high-dose contrast use in patients with severe renal impairment (glomerular filtration rate <30 mL/min). The use of a cyclic chelate-based gadolinium agent along with abstaining from the use of gadolinium in patients with severe renal impairment have virtually eliminated this problem. These agents can be used with caution in patients with milder forms of renal impairment, due to the greater risks of iodinated contrast agents used in cardiac CT.
CMR avoids errors imposed by geometric assumptions during the acquisition of 3D stacked sets of contiguous cine slices, usually in the short-axis plane (see Fig. 32.9B ). End-diastolic and end-systolic volumes and mass for both ventricles are determined by the planimetry of each slice and then summed for the entire ventricle. Measures of ventricular performance, including stroke volume, EF, and cardiac output, may be accurately quantified using this methodology. Similarly, calculation of abnormal hemodynamic states resulting from coronary artery, valvular, and congenital heart disease causing left- or right-sided heart failure may be performed routinely. The high quality of the data allows indexation to important variables such as body surface area, gender, and age. These data demonstrate that indexation is important for the confident diagnosis of conditions in their early stages; for example, dilated or hypertrophic cardiomyopathy. LV diastolic function also can be assessed by CMR. However, echocardiography is used routinely, is readily available, and has been the preferred imaging modality of choice in large, randomized clinical trials.
Valvular Heart Disease and Heart Failure (see also chapter 26 )
Mitral Valve Regurgitation
Mitral regurgitation (MR) is more frequently associated with chronic HFrEF than with HFpEF, occurring in approximately 50% of patients, ranging in severity from mild to severe. MR is readily detected by Doppler because color flow velocity mapping is exquisitely sensitive even to trivial MR. The underlying mechanism of MR in HFrEF can be appreciated by 2D and 3D TTE imaging, which demonstrates distortion of the geometry of the mitral subvalve apparatus, stretching of the mitral annulus preventing adequate coaptation of the mitral valve leaflets caused by progressive LV cavity enlargement, which results in MR ( Fig. 32.10 ; see Fig. 32.3 ). Compared to HFrEF that undergoes eccentric remodeling, LV cavity size in HFpEF is not prone to dilatation but instead undergoes concentric hypertrophic remodeling that only rarely results in severe MR.
Evaluation of the presence and severity of MR in heart failure is hemodynamically important because the excessive volume handling from MR increases LV loading conditions, causing further deterioration in LV function as predicted by the inverse relationship between load and ejection phase indices. This shows that the greater the increase in LV load, the greater the reduction in LVEF. The corollary of this relationship provides the rationale for the chronic use of vasodilator therapy to reduce load and improve refractory heart failure. Another cause of MR, which is also more frequent in HFrEF than HFpEF, is due to recurrent episodes of myocardial ischemia triggered by the worsening of regional myocardial blood flow that may cause severe sudden onset MR and flash pulmonary edema/acute heart failure. This can be documented by stress echo, radionuclide stress, or CMR perfusion studies.
The clinical question that often needs resolution occurs when HFrEF and MR coexist, is whether MR is the primary cause of the heart failure or whether the heart failure is the cause of the secondary MR. Since the majority of HFrEF have an ischemic etiology, the development of MR is the result of postinfarction remodeling. When the MR is moderate to severe following remote MI and the mitral valve leaflets and mitral annulus are intrinsically normal, it is more likely that the MR is secondary due to alteration of LV architecture by cavity dilatation and remodeling, which causes the failure of normal coaptation of the mitral leaflets in heart failure. In contrast, when the mitral valve leaflets are rheumatically thickened with commissural fusion or when the leaflets are myxomatous and degenerative, it is more than likely that the MR is primary and the heart failure is secondary to hemodynamically significant MR. The mitral valve can be imaged in exquisite detail by TTE and TEE, such that there is little doubt as to its normality or abnormality. In a small minority of patients, it is not possible to determine which is the primary event. However, care should be taken to inspect the mitral valve leaflets, commissures, and subvalve tensor apparatus for undisclosed ruptured chordae tendineae from excessive traction. The hemodynamic severity of MR should be quantified noninvasively using the proximal iso-velocity surface area (PISA) and not by visual estimation alone.
When MR is severe, it may spuriously increase LVEF causing delay in urgently needed surgical mitral valve repair or replacement. The presence of even trivial MR can be demonstrated by Doppler color flow velocity mapping. However, color flow mapping describes velocity distribution and not regurgitant volume flow. The severity of MR should be quantified by PISA, which relies upon the principle that blood flow velocity increases as it approaches a restrictive orifice forming a series of hemispheric iso-velocity shells. The finite orifice is the mitral regurgitant orifice with the velocity shells forming on the ventricular side of the regurgitant orifice. The conservation of mass dictates that the flow rates at each of the hemispherical iso-velocity shells are equal to the flow rates at the mitral regurgitant orifice. The flow rate can be calculated as the surface area of a hemisphere 2π r 2 × aliasing velocity, where r = the radius of the hemisphere which is the only Doppler parameter that needs to be measured. The severity of MR can be evaluated comprehensively by a number of echo parameters derived from PISA, including: effective regurgitant orifice area which equals flow rate/peak mitral regurgitant jet velocity; regurgitant volume which is the product of effective mitral regurgitant orifice area and the TVI of the mitral regurgitant jet; and regurgitant fraction, which is the ratio of regurgitant volume to stroke volume.
Aortic Valve Disease as a Cause of Heart Failure
Subjects with congenital AS present in their late teens to early 20s with chest pain, syncope, or symptoms of heart failure for valvuloplasty or open-chest valve replacement. Adult onset/acquired severe “senile” calcific AS occurs from the sixth to ninth decade. Bicuspid valves require surgical intervention for valve replacement on average a decade earlier than stenotic trileaflet aortic valves. TTE is the diagnostic imaging modality of choice for aortic valve stenosis and enables the precise assessment of the severity of AS in terms of the aortic valve effective orifice area, and the peak and mean trans-aortic valve gradients and LVEF ( Fig. 32.11 ). A proportion of adults with severe or critical AS develop heart failure, which is associated with a poor prognosis. If critical AS is not recognized and the LV is allowed to dilate, the trans-aortic gradient falls as LV function deteriorates so the aortic valve orifice area must be calculated by 2D or 3D echo. There is an increasingly recognized subgroup of symptomatic patients with severe AS who have low transvalvular gradients due to LV systolic dysfunction, known as low-flow AS. It is important to identify these elderly patients because they do well clinically with trans-cutaneous intervention. Hemodynamically, chronic decompensated calcific AS associated with LV dilatation and decreased LVEF must be differentiated from idiopathic dilated cardiomyopathy with concomitant age-related calcification of the aortic valve. This can be achieved with 2D Doppler echo of the apical long axis and the apical 5-chamber view (see Fig. 32.11 ). 2D, 3D, or TEE are the imaging modalities of choice for the diagnosis and quantification of severity of AS. Bicuspid aortic valves (BAVs) are associated with ascending aortic aneurysm formation that is independent of the hydraulics across the aortic valve. The extent and severity of the aortopathy should ideally be assessed by either CT angiography (CTA) or CMR with contrast, and if the maximal dimension distal to the sinuses of Valsalva at the sinotubular junction is >4.5 cm, surgical repair or endovascular stenting should be considered, irrespective of symptoms. The anatomy of the commissures can be determined best by 2D, TEE, or 3D echo. Once the diagnosis of hemodynamically significant AS is established, management should follow the American College of Cardiology (ACC)/AHA Guidelines. Ascending aortic aneurysms need to be carefully monitored by serial measurements of maximal diameter and growth rate that determine the timing of surgery or endovascular stenting. Dilatation of the aortic root with effacement of the sinuses of Valsalva may disrupt the architecture of the aortic valve leaflets resulting in secondary rather than primary aortic regurgitation (AR).
Chronic AR has become a less common cause of congestive heart failure in the Western hemisphere because of the decline in incidence of rheumatic heart disease. The diagnosis of AR is easily made by documenting the regurgitant diastolic blood flow reversal in the LVOT flow by Doppler color flow velocity mapping, which is extremely sensitive even to trivial AR. The etiology of AR is due either to abnormalities of the aortic valve leaflets or to abnormalities of the aortic root geometry. The aortic valve is bicuspid in 1% of all live births and undergoes calcification, becoming stenotic and/or regurgitant. BAV may be associated with ascending aortopathy, which is independent of the hemodynamic severity of the aortic valve pathology. The valve leaflet morphology may vary from rheumatic-like thickening with calcified commissural fusion to floppy myxomatous leaflets that fail to coapt. Moreover, the geometry of the aortic root may be altered by aneurysmal dilation or effacement of the sinuses of Valsalva, as seen in Marfan syndrome, resulting in secondary AR.
In heart failure due to chronic severe AR, the LV is markedly dilated with moderate to severe eccentric hypertrophy, and though there may be preserved function early in the course, systolic function gradually deteriorates ( Fig. 32.12 ). One-third of patients with moderate to severe AR do not develop cardiovascular symptoms, and become symptomatic late in the natural history of their disease when irreversible LV dysfunction develops. This emphasizes an important role for imaging with Doppler echocardiography in AR. The severity of the AR can be accurately quantified by Doppler echo assessment of the deceleration rate of the regurgitant aortic jet, by the diameter of the vena contracta , regurgitant volume, regurgitant fraction, and the effective regurgitant orifice area. Clinical decision-making to recommend aortic valve replacement (AVR) is based largely on echo measurements of LV cavity size (end-systolic and end-diastolic dimensions) and function (see Fig. 32.12 ) in conjunction with symptoms as recommended in the ACC/AHA 2014 Guidelines for the management of patients with valvular heart disease. Thus, TTE imaging provides the requisite information regarding LV size, geometry, and hemodynamic assessment necessary to confidently advocate AVR to remove the regurgitant volume, which is important in the appropriate management of patients with chronic AR and congestive heart failure.
Cardiovascular Magnetic Resonance Assessment of Valvular Heart Disease
CMR may be a useful alternative to echocardiography in patients with valvular disease with limited image quality, or those with conflicting or inconclusive results. Tissue characterization may help physicians to determine the etiology and potential therapy of valvular heart disease (e.g., ischemic vs. nonischemic MR). While regurgitant and stenotic jets are evident in cine images, a more complete valvular disease assessment requires a quantitative evaluation of flow and flow velocity data that have been previously validated.
The regurgitant volume and regurgitant fraction in MR may be measured by comparing the LV stroke volume with the systolic aortic flow. More recent applications have used quantitative assessment of MR jets in adjudicating the severity of MR, which has informed therapeutic decision making and impacted outcomes. CMR was able to better predict patients with severe MR than TTE or TEE, especially in those patients with highly eccentric jets.
In AR, the regurgitant fraction is measured by integrating the diastolic flow across the aortic valve, and comparing it to the systolic forward flow ( Fig. 32.13 ). CMR can accurately quantify the effective cardiac output and provide an important objective parameter of heart failure related to low output states.
In stenotic valvular disease, flow quantification can be used to identify increased transvalvular flow velocity and to calculate pressure gradients. By visualizing the stenotic orifice, CMR offers reliable aortic valve area quantification and avoids limitations related to oblique or turbulent jets and unreliable pressure gradients, particularly in low output states.
HFrEF patients who have an ischemic etiology may present with angina, shortness of breath, reduced exercise capacity, and a low LVEF due to the presence of noncontracting but viable myocardium ( see also Chapter 19 ). There are three noninvasive imaging methods commonly used for assessing myocardial viability: stress echocardiography, nuclear imaging, and CMR perfusion studies.
Viable noncontracting myocardium in heart failure can be detected by exercise and/or pharmacologic (e.g., low-dose dobutamine) stress echocardiography. The purpose of identifying viable myocardium in selected heart failure patients is because restoration of myocardial perfusion targeted to the ischemic regions can potentially translate into increased regional and global (LVEF) function. There are three nonoverlapping phasic responses to dobutamine stress echocardiography. First, at low levels of stress there is recruitment of myocardial shortening that is lost as the level of stress (dobutamine dosage) increases, described as a “biphasic response.” Second is the continued improvement in regional and global function known as the “contractile reserve.” Third is a plateau or worsening global LV function. Technical limitations in image acquisition and image analysis due to lung disease, enhanced respiratory excursion, or abnormal body habitus (obesity) can be minimized by augmenting endocardial definition with intravenous echo contrast agents. The advantages of stress echocardiography for detecting viable, noncontracting myocardium are that stress echo correlates well with nuclear imaging, yet it does not involve radiation exposure, and stress echo is relatively inexpensive. Patients with a biphasic response to stress benefit from revascularization more than those patients with contractile reserve. In contrast, those with a flat or negative response to stress usually do not benefit from revascularization and are at high risk for adverse clinical cardiovascular outcomes. In chronic ischemic heart disease there may be a history of remote transmural MI that can be identified as a hyper-echoic (highly echo reflective) region of the free LV wall or septum that is thinner than normal wall thickness (<6 mm) and does not contract or change in thickness with increasing work load during stress echo testing.
A prerequisite for the diagnosis of viability and ischemic myocardium with noninvasive imaging tools is the acquisition of high-quality images at rest but especially during stress testing. CMR studies provide complete and continuous endocardial surfaces. By contrast, the 2D and 3D TTE images often need to be interpolated, although the need for major interpolation has diminished since the introduction of harmonic imaging and the use of intravenous echo contrast for the express purpose of enhancing endocardial definition. Myocardial viability also can be detected by radionuclide angiography and CMR perfusion studies.
Radiotracer-Based Assessment of Myocardial Viability and Remodeling
Serial SPECT perfusion imaging can be used for the assessment of myocardial viability and salvage following reperfusion, and later following PCI. These indices of myocardial salvage derived from nuclear imaging are predictive of the long-term outcome.
Nuclear imaging techniques also have been used specifically to predict LV remodeling. 201 Tl and 99m Tc-labeled perfusion images also have been used to assess infarct size, and this radiotracer-based estimation of myocardial viability has correlated with subsequent LV remodeling. 99m Tc-sestamibi imaging has been used to assess infarct size early post-MI in patients and these perfusion-based estimates of infarct size have correlated with LV volumes at 1 year. Another study suggested that infarct severity on 99m Tc-sestamibi imaging maybe a better predictor of remodeling than infarct size alone. Radionuclide assessment of viability was also shown to predict LV remodeling in a study of patients with ischemic cardiomyopathy. In this study, viability was assessed with serial nitrate enhanced 201 Tl and 99m Tc-sestamibi imaging. Remodeling was prevented in patients with at least five viable segments who underwent revascularization at the 21-month follow-up. The direct relationship between the extent of 201 Tl defects and collagen content in severe ischemic cardiomyopathy was assessed in a study of hearts excised after cardiac transplantation. There was greater collagen content in segments with irreversible 201 Tl defects compared to segments with reversible defects or normal 201 Tl perfusion. Furthermore, noninfarcted segments in the hearts with cardiomyopathy also had elevated collagen content compared to control hearts.
Cardiovascular Magnetic Resonance Stress
CMR can also be used to predict post-revascularization recovery of LV function, using late gadolinium enhancement (LGE) and low-dose dobutamine stress testing. The degree of transmural LGE predicts regional functional recovery, and improvement in LVEF frequently occurs when there is hibernation of >20% of the global LV myocardium. On a regional basis, LGE has an 80% positive predictive value for recovery in segments demonstrating no infarction, and a 90% negative predictive valve for no improvement in segments with >50% transmural scar. Other work suggests that low-dose dobutamine stress CMR (DSMR) may be more sensitive for viability assessment, and additional clinical experience is accumulating. LGE demonstrates myocardium with nontransmural scar that is destined not to recover function despite revascularization ( Fig. 32.14 ). Failure of functional improvement also may be due to incomplete revascularization, persistent hibernation early post-revascularization, or abnormal wall mechanics, which may be recruitable with dobutamine or exercise stress. CMR also predicts the response to beta-blockers in heart failure patients, and demonstrates lower response rates in patients with larger MIs.