Ischemic heart failure (HF) is the leading cause of HF and an epidemic disease worldwide with growing prevalence and high mortality rate.1^,2 In 2011, 1 in 9 death certificates in the United States listed HF.1 In 2015 in Brazil, 27,434 deaths occurred due to HF.3 Medical treatment, cardiac rehabilitation, revascularization and the increased understanding of its pathophysiology have improved the overall prognosis and survival of patients with HF over the last years, but, despite that, around 50% of the patients diagnosed with HF will die 5 years after the initial diagnosis.2

Accumulated evidence of the past years has suggested that individualized-target therapy with viability imaging assessment may improve outcome.4^,14^,15 This review will focus on the understanding of the viability concept and current evidence.

A simplistic way to describe viable myocardium is all tissue that is not scar/fibrosis (non-viable myocardium). Naturally, normal myocardium is viable. Dysfunctional myocardium that is viable has the potential to recover from an injury.4^,14^,15 Meanwhile, two concepts under the umbrella of “viable myocardium” can be often misunderstood. “Stunned” and “hibernating” myocardium are conditions in which function is impaired but is potentially reversible. Stunned myocardium is characterized by the persistent dysfunction that follows an episode of ischemia. Hence, there is normal rest flow and impaired function. The severity and duration of the stunning (post-ischemic dysfunction) depend on duration, extent and severity of the preceding ischemic insult. So long as there is no infarction during such ischemia, full recovery is expected, the timing of which also depends on the duration, extent and severity of the preceding ischemia. If stunning occurs repeatedly, the myocardium must adapt to the repetitive injury. It does so by reducing contractile function and flow in response to these events.15 Repetitive stunning is believed to be the precursor to hibernating myocardium, where both measured perfusion and function are reduced but restorable in whole or in part if blood flow can be adequately restored before irreversible injury occurs. This is the area of focus for viability imaging (Table 1).4^,14^,15

Several imaging modalities can be used to assess hibernating myocardium, and each has different metabolic/cellular targets and findings to detect viable and hibernating myocardium. Cardiac positron emission tomography (PET) with ¹⁸Fluorodeoxyglucose (¹⁸FDG) uses a glucose analogue to measure myocardial glucose uptake. Single-photon emission computed tomography (SPECT) with thallium-201 (²⁰¹Tl), a potassium analogue, has the sarcolemma membrane integrity as its target (sodium/potassium ATPase pump activity).16 SPECT with technetium-99m (^99mTc)-based tracers test the mitochondrial membrane integrity.17^,18 Dobutamine echocardiogram (ECHO) and dobutamine magnetic resonance imaging (MRI) measure myocardial contractile reserve. Delayed enhancement MRI and computed tomography target the amount of fibrotic tissue, and myocardial contrast ECHO targets the microvascular integrity.19^,20

In a meta-analysis by Schinkel et al.5 reviewing 24 studies (756 patients) comparing all available imaging modalities, ¹⁸FDGPET was shown to be the most sensitive to predict regional function recovery, and dobutamine ECHO was the most specific (92%, 63%, 74% and 87% and 80%, 78%, 75% and 83% of sensitivity, specificity, positive and negative predictive value for PET and ECHO, respectively).5 Cardiac MRI, which was underrepresented in this meta-analysis, had sensitivity, specificity, positive and negative predictive values of 74%, 82%, 78% and 78% for dobutamine stress MRI and 84%, 63%, 72% and 78% for delayed enhancement MRI.5

In this same meta-analysis, a total of 721 patients underwent ^99mTc-tracer-based SPECT and 1,119 had ²⁰¹Tl SPECT to assess viability. ²⁰¹Tl was more sensitive and ^99mTc-tracer-based SPECT more specific to predict recovery, with sensitivity, specificity, positive and negative predictive values of 87%, 54%, 67% and 79% and 83%, 65%, 74% and 76% for ²⁰¹Tl and ^99m Tc, respectively.5 Comparisons between nuclear techniques suggest ¹⁸FDG PET is the superior technique to detect the amount of hibernating myocardium,21^-25 except for one study directly comparing ²⁰¹Tl and ¹⁸FDG PET, which suggested similar viability detection for both methods.26

More recent data analyzing MRI performance in detecting viable myocardium have supported its high sensitivity.27^-30 Romero et al.27 have conducted a meta-analysis of MRI prospective trials including 24 studies (698 patients) and found a sensitivity of 95% for predicting functional recovery for MRI with delayed enhancement. Dobutamine MRI was the most specific (91%) when compared to delayed enhancement and end-diastolic wall thickness techniques.27 Kühl et al.29 have studied 29 patients with chronic ischemic HF and mean ejection fraction of 32% who had both MRI and PET/SPECT (¹⁸FDG for metabolism and ^99mTc SPECT for perfusion) performed at baseline and at 6-month follow-up after revascularization.29 The group found MRI to have higher sensitivity and PET/SPECT to be more specific (97% versus 87% sensitivity and 68% versus 76% specificity for MRI and PET/SPECT, respectively).29 A more recent study has analyzed the feasibility of PET/MRI scanners in evaluating segment functional recovery in 28 patients post-acute myocardial infarction (MI) and percutaneous revascularization.30 All patients underwent PET/MRI with contrast for delayed enhancement and ¹⁸FDG injection for uptake assessment 5-7 days after the acute event and had a follow-up MRI for contractility assessment at 6 months.30 The study has concluded that simultaneous assessment of glucose metabolism and scar assessment using a hybrid PET/MRI scanner is feasible. Moreover, the agreement between the techniques was high (82% of the segments were either non-viable or viable for both PET and MRI, k = 0.65). In only 18% of the segments was there disagreement, and, in all of them, PET suggested non-viability while MRI suggested viability.30 The recovery was higher in the segments in which there was agreement between the techniques (78% versus 41% for PET viable/MRI viable and PET non-viable/MRI viable, respectively). Recovery was similar between PET non-viable/MRI viable and PET non-viable/MRI non-viable segments, suggesting PET better dichotomized the degree of recovery between viable and non-viable myocardium. In the PET non-viable/MRI viable segments, there was some recovery (41%), suggesting a lower threshold for % FDG uptake cutoff (40-45% instead of 50%) may have detected some viable segments identified by MRI. Overall the techniques appear complementary. Their combined use as PET/MR may offer comprehensive tissue characterization of metabolism, scar and function and may refine our ability to define viable myocardium. Further studies are warranted (Figures 1 and 2).30

Figure 1

Figure 2

AMI: acute myocardial infarction; LGE: late gadolinium enhancement; PET: positron emission tomography; MR: magnetic resonance imaging.

Several non-randomized studies have reported data that suggest a benefit of viability imaging in patients with ischemic HF.9^,31^-34 Allman et al.9 have conducted a meta-analysis with 24 studies, and their analysis has shown the benefit of revascularization only in patients with viable myocardium as opposed to scar.9 More recently, a meta-analysis including 29 studies by Inaba et al. has documented the benefit of revascularization over medical therapy in patients with dysfunctional viable myocardium.31

To date, there have been two major prospective randomized trials comparing outcome in patients with ischemic HF who underwent viability assessment: PARR-2 (Positron emission tomography And Recovery following Revascularization phase 2)6 and STICH (Surgical Treatment for Ischemic Heart Failure) viability substudy35 trials.

PARR-2 has randomized 430 patients from 9 centers, to have either viability assessment with ¹⁸FDG PET or standard care without ¹⁸FDG PET, before decisions regarding revascularization.6 A trend toward benefit for the primary outcome (cardiac death, MI and cardiac hospitalization at 1 year) has been observed in the arm that underwent FDG PET to assist with clinical decision-making [36% of events in the standard care arm and 30% in the PET arm, relative risk 0.82; p = 0.16 and hazard ratio (HR) 0.78; p = 0.15].6 However, not all patients in the study followed the imaging recommendation. When analyzing only the patients who adhered to the recommendations from the imaging report, a significant reduction in outcome was observed in the PET arm versus standard care (HR 0.62; p = 0.019), indicating that management adhered to the imaging recommendations may have an impact on patient outcome.6

A PARR-2 substudy has supported the importance of adherence to PET findings and that of teamwork of: i) revascularization (surgeons, interventional cardiology); ii) HF; and iii) imaging specialists.8 This along with iv) access to FDG and v) the cardiac PET imaging experience of a centre has the potential to impact outcome. The Ottawa-FIVE (i.e. i-v above) study has had 111 patients from an experienced center in which PET was easily available and physicians were comfortable with the technology and its interpretation. In this scenario, patients in the FDG PET arm had clear benefit when compared to standard care (19% of cumulative proportion of events in the PET arm versus 41% in the standard care group) and multivariable analysis showed benefit (HR 0.34; 95% confidence interval 0.16-0.72; p = 0.005).8

In long-term (5 years) follow-up, the PARR-2 population in which PET recommendations were followed had improved primary outcome (HR 0.73, 95% confidence interval 0.54-0.99, p = 0.042) (Figure 3).11

Figure 3

In addition, PARR-2 has shown that the amount of hibernating myocardium also plays an important role in patient outcome.7 With increasing extent of mismatch (hibernating myocardium), the likelihood of benefit with revascularization also increases. In this substudy of the PARR-2 trial involving 182 patients in the PET arm, a cutoff of 7% was able to distinguish between patients who would or would not benefit from revascularization, which is in accordance with previous values reported by Di Carli el at.10 (5%), Lee et al.12 (7.6%) and Ling et al.36 (10%) (Figure 4).

Figure 4

CI: confidence interval; HR: hazard ratio.

It is important to understand the differences between PARR-2 and STICH in order to appreciate their respective significance.6^,35^,40^,43 First, in STICH, patients had to be acceptable for revascularization. While patients were randomized to coronary artery bypass graft surgery versus optimal medical therapy, imaging was not randomized nor did it direct the therapy decision. Conversely, in PARR-2, patients in whom decisions regarding revascularization was uncertain were randomized to FDG PET viability imaging versus standard care with no FDG PET imaging. The tests for viability assessment were also different: ¹⁸FDG PET in PARR-2 and SPECT or dobutamine ECHO in STICH. Compared to the STICH population, PARR-2 patients had more renal dysfunction (7.5% versus 34%), had more prior coronary artery bypass graft surgery (3% versus 19%), more multivessel coronary artery disease (75% versus 90%) and less viable myocardium (81% versus 22%), suggesting these patient cohorts were not the same (Table 2).40^,43 From those studies, it is safe to conclude that viability imaging is not needed in all patients with ischemic heart disease and left ventricular dysfunction who are being considered for revascularization. However, there may be high-risk patients whose decisions are particularly difficult where viability imaging has a role.40^,43

Current evidence and guidelines support the use of viability imaging to assist decision-making in patients with ischemic HF (Table 3).44^-50 The imaging modality of choice for viability assessment needs to be individualized according to each clinical scenario, technology availability and institution expertise.14^,40^,41^,49^-52

In our experience, viability imaging is appropriate in patients with known or strongly suspected ischemic HF, New York Heart Association (NYHA) ≥ II, moderate to severe left ventricular dysfunction (left ventricular ejection fraction < 40%), moderate to large perfusion defects and no significant ischemia, significant comorbities and/or poor vessel targets (Figure 5).49^,51^,52 On the other hand, viability is not (or less) useful in patients with predominantly angina CCS > II, those with normal or mild left ventricular dysfunction, critical left main coronary artery disease, patients with good revascularization targets, those with already-demonstrated moderate to severe ischemia and those with minimal or no comorbidities.51^,52 Figure 6 illustrates two examples of viability imaging.

Figure 5

CABG: coronary artery bypass grafting; CAD: coronary artery disease; CRT: cardiac resynchronization therapy; EF: ejection fraction; ICD: implantable cardiac defibrillator; LV: left ventricle; MRI: magnetic resonance imaging; NYHA: New York Heart Association; PCI: percutaneous coronary intervention; SPECT: single-photon emission computed tomography. †Denotes the required indications for 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) imaging in Ontario.

The IMAGE HF (Imaging Modalities to Assist with Guiding therapy in the Evaluation of patients with Heart Failure) project includes a group of clinical trials, one of which is the AIMI-HF trial (Alternative Imaging Modalities in Ischemic Heart Failure) (NCT01288560)54 (Figure 7). AIMI-HF is a multicenter randomized trial and registry study involving centers from Canada, United States, Finland, Brazil and Argentina. It compares the impact of standard of care investigation (SPECT) versus advanced imaging (PET and MRI) for viability and ischemia assessment. Composite outcomes are cardiac death, resuscitated cardiac arrest, MI and cardiac hospitalization. In cases where the patient is not randomized to one or the other arm, they are included in a clinical registry.54 This study will help us understand the impact of the advanced cardiac imaging modalities for the viability assessment and their impact on patient outcome.

Figure 7

CMR: cardiac magnetic resonance; LGE: late gadolinium enhancement; LVEF: left ventricular ejection fraction; SPECT: single-photon emission computed tomography; PET: positron emission tomography; R: randomization.

Although the value of viability imaging may have been called into question by the STICH trial, several studies have reinforced the relationship between the extent of hibernating myocardium and improvement in patient outcome, left ventricular ejection fraction and quality of life if nutrient flow can be restored with revascularization. In general, there is accepted utility in using viability imaging in patient populations where decisions for revascularization are most difficult. Ongoing trials will further enable the identification of which patients most benefit from viability imaging and by which methods. Can biomarkers be used to guide revascularization or at least to guide imaging to guide revascularization? Further research is needed here. In the meantime, clinicians, surgeons, interventional cardiologists and imaging specialists must work together as a team to enable the best decisions for each individualized patient in order to optimize the patient’s desired outcomes.