Despite many advances in therapy, chronic heart failure remains a prevalent condition with a high mortality rate.1 The successful treatment of heart failure patients requires establishing an accurate diagnosis, identifying potentially reversible etiologies, determining the optimal therapy, and reliable risk assessment for stratification of patients at high risk for worsening. Several of these aspects of heart failure care can be gainfully evaluated via radionuclide imaging. This chapter will review established applications of radionuclide imaging in heart failure.
The clinician has several goals when evaluating a heart failure patient. One potential work flow sequence for the evaluation of newly diagnosed heart failure is shown in Figure 18-1. Once a clinical diagnosis of the syndrome of heart failure is made, an initial step is often the determination of left ventricular (LV) systolic function. Approximately one-half of patients will have heart failure with preserved ejection fraction (HFpEF, EF ≥40%), while the remainder will have LV systolic dysfunction (heart failure with reduced ejection fraction HFrEF, EF <40%).2 Radionuclide imaging methods including single-photon emission computed tomography (SPECT), radionuclide ventriculography (RVG), and positron emission tomography (PET) can all provide highly accurate and repeatable measures of LV systolic and diastolic function (Chapter 11). Despite being as prevalent as HFrRF, the treatment of HFpEF remains largely symptom based and empirical, with very little supportive data from clinical trials. For patients with systolic dysfunction, HFrEF, a critical next step is the determination of etiology. Etiology evaluation can include identifying specific and potentially remediable causes such as valvular disease, coronary artery disease (CAD), specific cardiomyopathies, and pericardial disease. When extensive CAD is found, testing for ischemia and viability is helpful to determine benefit from coronary revascularization. Radionuclide imaging has critical roles in the determination of heart failure etiology (see below), identifying patients for coronary revascularization (Chapter 21), and the evaluation for specific cardiomyopathies such as amyloidosis and sarcoidosis (Chapter 24). For patients with nonischemic cardiomyopathy (NICM), and those with persistent LV systolic dysfunction after specific intervention, a combination of guideline-directed medical therapy (GDMT), and device therapy in selected patients (implantable cardioverter defibrillator [ICD], and cardiac resynchronization therapy [CRT]), form the cornerstone of current recommendations. Evolving applications such as myocardial sympathetic neuronal function (Chapter 23) and dyssynchrony imaging (Chapter 11) may have relevance to the selection of patients for ICD and CRT. Furthermore, PET/CT imaging with F-18 fluorodeoxyglucose (FDG) has established utility in the challenging area of diagnosing device infections (Chapter 24). A minority of patients will receive advanced heart failure therapies, including left and right ventricular assist devices (LVAD and RVAD) and cardiac transplantation. In posttransplant patients, radionuclide imaging has important prognostic value, which may influence therapeutic options in patients with suspected allograft vasculopathy (see below).
Scheme for the evaluation of patients with heart failure. Arrows indicate steps where radionuclide imaging has application. HF, heart failure; LV, left Ventricle; HFrEF, heart failure with reduced ejection fraction; HFpEF, heart failure with preserved ejection fraction; CAD, coronary artery disease; ICM, ischemic cardiomyopathy; NICM, nonischemic cardiomyopathy; GDMT, guideline-directed medical therapy; ICD, implantable cardioverter defibrillator; CRT, cardiac resynchronization therapy; VAD, ventricular assist device.
Determination of Heart Failure Etiology
The etiology of heart failure varies considerably depending on the population studied.3 Based on clinical trial data on patients with established heart failure, CAD is the attributed etiology for 60% to 70% of heart failure in the United States.4 However, the mere presence of CAD in the setting of a cardiomyopathy does not imply an ischemic etiology to the LV dysfunction. What is traditionally referred to as significant CAD in the literature, that is, ≥50% luminal stenosis, may be encountered in 15% to 30% of patients with a dilated cardiomyopathy, and thus may not be sufficiently sensitive for accurate risk stratification of the heart failure population. Felker et al. addressed this question, and tested a more stringent definition of ischemic cardiomyopathy for characterization of heart failure patients.5 They defined ischemic cardiomyopathy as LV dysfunction with one or more of the following angiographic criteria: significant left main or proximal left anterior descending coronary artery stenosis, at least two-vessel disease with ≥70% stenosis, or single-vessel disease with prior myocardial infarction, or prior coronary revascularization. For example, a patient with LV dysfunction and a 70% stenosis of one major epicardial vessel without antecedent myocardial infarction or revascularization would be adjudicated to the nonischemic cardiomyopathy group (with coexisting, but not causally related CAD). Using these more restrictive criteria, patients with LV dysfunction and single-vessel CAD had a prognosis comparable to those with nonischemic cardiomyopathy.5 Patients with true CAD-related heart failure have a worse prognosis than those with nonischemic cardiomyopathy, but the former may improve cardiac function dramatically with revascularization, highlighting the critical importance of an accurate diagnosis. The literature regarding the use of SPECT for the diagnosis of underlying CAD in LV dysfunction has primarily focused upon patients with chronic heart failure, with scant data addressing the diagnosis of CAD in new-onset heart failure.
In the setting of newly diagnosed LV systolic dysfunction, the identification of underlying CAD and potential "at risk" dysfunctional myocardium that might recover with coronary revascularization is critical. Although current practice guidelines specifically mandate coronary angiography only in heart failure patients with angina, chest pain is often absent in patients with ischemic cardiomyopathy, even those with significant amounts of viable myocardium.2,6
The Investigation of Myocardial Gated SPECT Imaging (IMAGING) in Heart Failure trial specifically addressed the utility of gated SPECT as an initial diagnostic modality in the de novo acute heart failure setting.7 Two hundred and one patients hospitalized with new-onset heart failure were prospectively enrolled, and underwent exercise or pharmacologic SPECT during the index hospitalization. At the physician's discretion, approximately one-third of the patients underwent coronary angiography. Using a summed stress score (SSS) >3 to define an abnormal study, SPECT had a sensitivity of 96% and a negative predictive value of 96% for the diagnosis of ischemic cardiomyopathy using the criteria proposed by Felker, but was less accurate in detecting limited-extent CAD (Table 18-1). Thus, this study provides proof of concept of the utility of myocardial SPECT for the initial characterization of patients presenting with severe new-onset heart failure. Such patients who have normal stress myocardial SPECT are very unlikely to have underlying extensive CAD that is etiologically related to their heart failure (Fig. 18-2).
Categorization of heart failure etiology using technetium-99m sestamibi MPI. (A) Left ventricular (LV) dilation (abnormal LV systolic function by gated SPECT not shown) with large, fixed perfusion defects in the septum, anterior wall, apex and inferior wall suggestive of CAD-related ("ischemic") cardiomyopathy. (B) Normal stress–rest perfusion and LV size (normal LV EF on gated SPECT not shown) indicative of heart failure likely related to diastolic mechanisms. (C) LV dilation (with abnormal LV systolic function on gated SPECT, not shown) and normal perfusion suggestive of non-CAD related ("nonischemic"). (Reproduced with permission from Soman P, Lahiri A, Mieres JH, et al. Etiology and pathophysiology of new-onset heart failure: evaluation by myocardial perfusion imaging. J Nucl Cardiol. 2009;16(1):82–91.)
Table 18-1Performance Characteristics of Gated SPECT Tc-99m Sestamibi for CAD Diagnosis in Patients with New-Onset Heart Failure from the IMAGING in Heart Failure Study ||Download (.pdf) Table 18-1 Performance Characteristics of Gated SPECT Tc-99m Sestamibi for CAD Diagnosis in Patients with New-Onset Heart Failure from the IMAGING in Heart Failure Study
|CAD Definition ||Any CAD: ≥70% Stenosis in Any Coronary Artery ||Extensive CAD: Stenosis ≥70% in the LM or Proximal LAD, ≥70% in ≥2 Major Epicardial Coronary Arteries or Any Stenosis ≥70% with a Prior MI or Coronary Revascularization |
|CAD prevalence by angiography ||51% (n = 38) ||36% (n = 27) |
|Sensitivity (95% CI) ||82% (66–92) ||96% (81–99) |
|Specificity (95% CI) ||57% (40–72) ||56% (41–71) |
|PPV ||67% ||55% |
|NPV ||75% ||96% |
Several previous studies have established the utility of myocardial perfusion imaging (MPI) for the diagnosis of CAD in chronic heart failure.8 Although many of these studies predated contemporary MPI, they uniformly demonstrated a very high negative predictive value for excluding CAD. Using more contemporary imaging with gated Tc-99m SPECT, Danias et al. reported an SSS >8 as 87% sensitive for detection of CAD, and that incorporating the SDS with findings of ischemia and regional wall motion abnormalities increased this to 94%.9 Thus, in the setting of both new-onset and established heart failure, and global dysfunction, a normal stress myocardial perfusion scan virtually excludes a diagnosis of ischemic LV dysfunction.
One important question remains, and that is whether SPECT MPI can replace coronary angiography as the diagnostic test for important underlying CAD in patients with new-onset heart failure. A major concern is that of balanced ischemia due to extensive CAD which might be missed or underestimated due to the fact that the MPI assessment of regional myocardial perfusion is relative. While it is unlikely that a patient with severe and extensive CAD will have no angina and a normal and rest/stress ECG and MPI, given the critical importance of excluding CAD in this population and the lack of substantive clinical trial data with MPI, patients with new-onset heart failure continue to undergo diagnostic coronary angiography for this purpose. Therefore, although evolving data increasingly suggest that this might be the case, only a large prospective clinical trial can definitively answer this question.
From a practical perspective, new-onset heart failure patients with angina and/or an intermediate to high probability of CAD (based on age, symptoms, and risk factors) should undergo diagnostic coronary angiography. Heart failure patients with a low probability of CAD, with clinical circumstances suggestive of nonischemic LV dysfunction can safely have a rest/stress MPI as the initial diagnostic test. In patients with known CAD being evaluated for new-onset or established heart failure, a rest/stress MPI provides invaluable information on ischemia, viability and quantitative LV function which can be used to drive important management decisions such as the choice between targeted percutaneous or surgical revascularization.
It is important to recognize that mild perfusion defects are common in nonischemic cardiomyopathy, and may reflect true physiological phenomena such as myocardial fibrosis or abnormal coronary vasodilator reserve, and have prognostic significance.10–12 Inferior defects may also be caused by diaphragmatic attenuation and attenuation from LV dilatation. Attenuation correction is helpful in identifying soft tissue artifacts in SPECT imaging, but its effect on diagnostic accuracy for CAD has not been specifically tested in the heart failure population.
Selecting Patients for Coronary Revascularization
In selected heart failure patients with LV systolic dysfunction, coronary revascularization may offer a unique opportunity for "cure." The selection of patients for coronary revascularization requires a consideration of its potential benefits against perioperative mortality and morbidity. Patients with severe LV systolic dysfunction are at the highest risk, but might also derive the most benefit. While the concept of preserved myocardial viability (Chapter 21) and its impact on prognosis appears physiologically sound, the lack of conclusive evidence from the randomized trials has resulted in uncertainty surrounding it. Two large randomized trials that addressed this issue, STICH and PPAR-2 did not support the selection of patients for revascularization based on the presence of myocardial viability, but could be critiqued for the nonrandomized design (a deviation from the originally intended study design) of the STICH viability substudy, and suboptimal adherence to the protocol-defined treatment arm in PPAR-2. Thus, the major guidelines for the role of viability testing have not been impacted by these trials (revascularization in patients with one- or two-vessel CAD without proximal LAD CAD, but with a large area of viable myocardium and high-risk criteria on noninvasive testing: Class IB).2 One practical approach might be to revascularize patients with extensive CAD, good coronary target vessels, and average surgical risk without routine prior viability testing. For patients at high surgical risk, the demonstration specifically of hibernating or stunned myocardium establishes the etiology of LV systolic dysfunction and informs the risk–benefit ratio of surgery favorably. A more recent analysis of the STICH cohort revealed better outcomes after coronary artery bypass grafting in patients with preserved functional capacity, multi-vessel CAD, lower ejection fraction, and higher end-systolic volume.13
Radionuclide Approaches to Risk Stratification in Heart Failure
The role of myocardial sympathetic neuronal imaging for risk stratification in heart failure patients is discussed in Chapter 23. I-123 metaiodobenzylguanidine (mIBG) is a SPECT agent that was approved by the Food and Drug Administration for this purpose in 2013.14 C-11 labeled PET agents have also been used for sympathetic neuronal imaging, but the requirement for a cyclotron in close proximity makes clinical use logistically difficult.15 An F-18 labeled agent, LMI1195 is currently undergoing phase 1 studies.16
Other radionuclide approaches have also been proven valuable for risk stratification in heart failure. A preserved myocardial flow reserve on Rb-82 PET MPI is indicative of a more benign prognosis in patients with ischemic and nonischemic cardiomyopathy compared to patients with a low myocardial flow reserve, as demonstrated elegantly in a study of 510 patients followed up for 8 months.17 The use of PET-derived myocardial flow reserve to identify low-risk patients is a very promising approach to risk stratification, and is addressed again in the section on cardiac transplantation.
The cellular and interstitial changes that underlie the phenomenon of LV remodeling are accompanied by a transformation of the normally ellipsoid LV into a more spherical shape. LV shape indices, such as the sphericity index derived from echocardiography, have established utility in predicting prognosis and response to therapy heart failure patients.18–20 Analogous measurements derived from gated SPECT have shown similar prognostic value (Fig. 18-3).21
Illustration of two cases with comparable left ventricular ejection fraction (LVEF) but different left ventricular shape indices (LVSI). The patient with a normal ellipsoid-shaped LV and preserved LVSI (A) had no symptoms whereas the patient with a remodeled, spherical LV and abnormal LVSI (B) had severe symptoms. (Reproduced with permission from Abidov A, Slomka PJ, Nishina H, et al. Left ventricular shape index assessed by gated stress myocardial perfusion SPECT: initial description of a new variable. J Nucl Cardiol. 2006;13(5):652–659.)
Role of Radionuclide Imaging in Device and Advanced Heart Failure Therapies
The potential role of sympathetic neuronal imaging for selecting patients for ICD therapy, and of dyssynchrony imaging for CRT are discussed in Chapters 23 and 11, respectively. The evolving role of PET/CT for the diagnosis of device infections is discussed in Chapter 24.
Another area of great clinical challenge is the surveillance of patients after orthotopic heart transplantation for the detection of allograft vasculopathy. Currently, posttransplant patients undergo annual surveillance coronary angiography. This disease is characterized by diffuse arterial hyperplasia rather than focal obstruction, and therefore, may be missed by conventional coronary "luminography," particularly in the early stages. Intriguing features include its development in young patients without traditional risk factors for atherosclerosis, and the selective involvement of allograft vessels with sparing of the host's native arterial system.22 Once developed, there is no therapy proven to definitively reverse the process, and the clinical course is usually one of progressive ischemic LV failure. Survival rates are low (~20% at 1 year) after a clinical ischemic event.23
The role of MPI for posttransplant follow-up has been evaluated. Both SPECT perfusion imaging24 and PET perfusion with myocardial flow reserve estimation25 have been found to have good prognostic utility for this purpose. Single-center studies indicate that normal perfusion on SPECT or myocardial flow reserve on PET predicts an excellent prognosis in the intermediate term (2–5 years). The ability to risk stratify using noninvasive approaches would be an important clinical advantage for posttransplant patients who are already burdened with a substantial load of testing. Also, these data attest to the fact that, analogous to atherosclerotic CAD, functional testing provides important prognostic information that may not be available from anatomy-based diagnostic approaches.26
Molecular Imaging in Heart Failure
Molecular mechanisms of heart failure are operative at the preclinical stage (Stage A of the ACC/AHA classification), and molecular imaging approaches have greatly enhanced our understanding of heart failure pathophysiology. It is hoped that the clinical translation of molecular imaging approaches will identify specific processes that may predominate in individual patients or patient groups, and explain the heterogeneity in response to therapy, for example, beta-blockers, thus facilitating personalized medicine. Such approaches include the imaging cellular mechanisms such as apoptosis (annexin-V)27,28 and the renin–angiotensin system (F-18 captopril, F-18 lisinopril),29 myocardial sympathetic innervation (I-123 mIBG, C-11 agents, and F-18 LMI1195), and myocardial metabolism (C-11 palmitate, I-123 BMIPP, F-18 FDG).30 Molecular imaging techniques have also been applied with success to the monitoring of regenerative cell therapy.31