Assessment of myocardial viability by cardiac magnetic resonance imaging

Abstract

Purpose of review:

Left ventricular systolic dysfunction due to coronary artery disease is common and ascertaining which patients will benefit from revascularization can be challenging. Viability testing is an accepted means by which to base this decision, with multiple noninvasive imaging modalities available for this purpose. This review aims to highlight the key role of cardiac magnetic resonance in myocardial viability assessment, with a focus on its unique strengths over other imaging modalities.

Recent findings:

Transmural extent of hyperenhancement with late gadolinium imaging has been shown to be greater acutely in ST elevation myocardial infarction patients undergoing primary percutaneous coronary intervention and regress at follow up studies. An explanation for this reported phenomenon and an argument against redefining CMR viability criteria in the acute setting will be offered.

Summary:

Although not universally available, cardiac magnetic resonance is an exceptionally powerful and safe imaging modality that should be considered when viability testing will influence patient management. While observational outcomes data suggest a promising prognostic role for viability, randomized studies in this area are needed.

Introduction

Hibernating myocardium refers to hypocontractile myocardial tissue that remains alive and metabolically active despite chronic hypoperfusion. Hibernating myocardium has the potential to recover contractile function when coronary flow is restored either percutaneously or surgically. Conversely, when myocardial function is impaired acutely and transiently after an ischemic insult — such as the period following reperfusion therapy for an acute myocardial infarct — it is deemed stunned. The concept of myocardial viability refers to living myocardial cells and encompasses both hibernating and stunned myocardium. The identification of viable but hypocontractile tissue is of major clinical interest in patients with ischemic left ventricular (LV) systolic dysfunction as it has prognostic implications, and furthermore may influence patient selection for revascularization.

While the prevailing benchmark for success of revascularization remains functional recovery — often defined by improvements in global LV ejection fraction or regional wall thickening — the notion that its absence implies nonviability may need to be nuanced a bit further. Incomplete revascularization,¹ tethering of viable myocardium to regions with extensive scar,² and perioperative necrosis in previously viable myocardium are all potential reasons for lack of functional improvement after revascularization.³ Moreover, a single evaluation of ventricular function post-revascularization may be inadequate or premature. Regional variations exist in the time required for viable myocardium to manifest recovery, highlighting the challenge of optimal timing of functional assessment after revascularization. In isolation, functional recovery is therefore at best an incomplete “standard of truth” with respect to viability assessment.

Accurate noninvasive assessment of living, salvageable myocardium in ischemic heart failure has been incorporated into major guideline recommendations concerning the management of both chronic heart failure patients and those with ST elevation myocardial infarcts (STEMI), though the evidence for mortality benefit with such testing is lacking. The 2013 American College of Cardiology heart failure guidelines support viability testing when planning revascularization in patients with coronary artery disease (CAD) and heart failure (class IIa recommendation, level of evidence B).⁴ In patients presenting with STEMI, the European Society of Cardiology 2017 guidelines state that ischemia and viability testing with any available modality can be considered prior to discharge (class IIb recommendation, level of evidence C).⁵ Recommendations on the specific imaging modality of choice are absent as the decision depends on patient-specific factors, local resource availability, and physician preference. While each modality has its advantages and shortcomings with respect to viability assessment, the expectation for an undisputed frontrunner for all patients is neither realistic nor practical. Nevertheless, this review serves to highlight the pivotal role of cardiac magnetic resonance (CMR) in the pursuit of viable myocardium.

Basis of Viability Testing Using CMR

The general advantages of CMR over other imaging modalities are well established; it is relatively unhindered by body habitus, is completely unhindered by imaging windows since imaging planes are self-prescribed and theoretically infinite in orientations, has a large field of view, lacks ionizing radiation, and has excellent spatial and temporal resolution. The two methods of viability testing by CMR are contractile reserve assessment using dobutamine stress and late gadolinium enhancement (LGE) imaging using gadolinium-based contrast agents (GBCA), with the latter being the more common and preferred technique.

GBCAs are paramagnetic metal compounds that, when administered intravenously, cannot penetrate intact myocardial sarcolemma and accumulate extracellularly in the intravascular blood pool and within myocardial interstitium. With LGE imaging, GBCAs are used to index cell membrane integrity, as living myocardial cells exclude GBCA when steady-state concentrations are reached. The volume of distribution of GBCA in the extracellular space is thus inversely related to the proportion of densely packed, viable myocardial cells. In an acute myocardial infarction, GBCA passively diffuses intracellularly through ruptured cell membranes and extracellularly in surrounding necrotic tissue, whereas in chronic infarcts GBCA concentrates in collagenous scar that has replaced necrotic tissue.

With respect to viability testing, CMR offers the unique ability to directly image infarcted myocardium — or scar — simultaneously with normal myocardium. LGE imaging techniques have been described in detail elsewhere,⁶ and are typically performed 5–15 minutes after intravenous GBCA administration. This allows normal myocardium, which appears dark, to be distinguished from myocardial scar, which appears bright. The direct visualization of hyperenhanced scar is the sine qua non of CMR viability assessment. The exquisitely high correlation in both scar pattern and infarct size between CMR and ex-vivo myocardial histology specimens in a canine model was previously shown by Kim et al.⁷ Although several methods exist to qualitatively or quantitatively assess the degree of hyperenhancement, it is most often determined visually by comparing the ratio of the thickness of infarcted myocardium to that of total wall thickness within a given segment (Figure 1).

Figure 1. Examples of five different groups of visual quantification of delayed enhancement used in routine clinical cardiac magnetic resonance reporting.

Cartoon illustrations and examples of respective late gadolinium enhancement inversion recovery images in short axis slices are shown. Infarct transmurality occurs across a continuum rather than as a dichotomous variable, with four subdivisions of hyperenhanced scar expressed as a percentage of total wall thickness used for ease of clinical reporting.

Source: Souto AL, Souto RM, Teixeira IC, Nacif MS. Myocardial viability on cardiac magnetic resonance. Arq Bras Cardiol 2017;108(5):458–469.

Reproduced under CC BY license from reference 30

Validation of CMR for Viability Testing

In a landmark study by Kim et al., the transmural extent of myocardial infarction (TEI) correlated best with the reversibility of myocardial dysfunction post revascularization, independent and across the spectrum of wall motion abnormalities.⁸ In this study, 41 patients with stable CAD scheduled to undergo coronary revascularization underwent LGE imaging prior to, and cine-CMR before and after revascularization. The likelihood of improvement in contractility declined as the transmural extent of hyperenhancement increased in the 804 dysfunctional myocardial segments analyzed (Figure 2). Importantly, this relationship was progressive and stepwise rather than abrupt, highlighting the concept that viability and likelihood of recovery should be interpreted within the context of a probabilistic continuum rather than a dichotomous variable. This continuum is in juxtaposition with the decision to revascularize an individual patient, which is binary and thus gives rise to a temptation to seek a universal cutoff of transmural hyperenhancement beyond which revascularization is unlikely to be fruitful.

Figure 2. Relation between the transmural extent of hyperenhancement before revascularization and the likelihood of increased contractility after revascularization.

Among 41 patients analyzed, data is shown for all 804 dysfunctional segments and separately for 462 segments with at least severe hypokinesia and 160 segments with akinesia or dyskinesia before revascularization. For all three analyses, there was an inverse relation between the transmural extent of hyperenhancement and the likelihood of improvement in contractility.

Source: Kim RJ, Wu E, Rafael A, et al. The use of contrast-enhanced magnetic resonance imaging to identify reversible myocardial dysfunction. N Engl J Med 2000;343:1445–1453.

Reproduced with permission from reference 8

Another focus of interest has been regionally thinned myocardium and whether it has the capacity to recover both contractility and wall thickness.⁹ In 1055 consecutive patients with CAD undergoing CMR viability assessment, 201 (19%) were found to have regional wall thinning (LV end diastolic wall thickness ≤ 5.5 mm). In these patients, limited scar burden (≤ 50% TEI) within the thinned region was found in 18% of patients. Scar extent was inversely proportional to both regional and global improvement in contractility, and patients with limited scar extent undergoing revascularization saw a significant improvement in regional wall thickening and furthermore demonstrated a resolution of wall thinning. Moreover, multivariate analysis determined scar extent in thinned segments to be most predictive of contractile improvement after revascularization. These results contradicted the common assumption that thinned myocardial segments are synonymous with nonviable scar. Rather, it is the transmural extent of scar that should be emphasized when predicting viability, despite the presence of wall thinning (Figure 3).

Figure 3. Examples of nonviable and viable anterior wall segments with wall motion abnormalities before and two months after coronary revascularization.

Patient C has an akinetic anterior wall and 4.5 mm scar thickness out of a total wall thickness of 8 mm, representing > 50% transmural extent of infarct that neither recovers contractility nor wall thickness after revascularization. Conversely, patient D only has 1.5 mm scar thickness out of a total wall thickness of 5 mm, representing < 50% transmural extent of infarct. Despite a dyskinetic anterior wall, the patient recovers both contractility and wall thickness (increased to 9 mm) after revascularization.

Source: Shah DJ, Kim RJ. Fundamental concepts in myocardial viability assessment revisited: when knowing how much is “alive” is not enough. Heart 2004;90:137–140.

Reproduced with permission from reference 31

Low-dose dobutamine may be used to assess contractile reserve in dysfunctional myocardial segments. In both acute and chronic ischemic myocardial dysfunction, separate meta-analyses have shown that low dose dobutamine enhances the specificity of CMR viability assessment.^10–11 However, the potential value of dobutamine stress CMR is additive but not substitutive to LGE imaging, as the absence of contractile reserve does not necessarily exclude viability. Severely dysfunctional myocardium that has exhausted its coronary flow reserve even at rest, for example, may demonstrate ischemia at low doses of inotropic stress but viability with LGE.

Prognostic Implications of CMR Viability Testing

While improvement of systolic function after revascularization intuitively suggests favorable clinical outcomes, the prognostic impact of viability testing using CMR is less well established than its ability to predict functional myocardial recovery. Nevertheless, a few points are worth noting. In a prospective cohort study, Gerber et al. showed that in patients with CAD and severe LV systolic dysfunction, patients with viable myocardium by CMR who underwent complete revascularization had a survival benefit compared to their medically treated counterparts.¹² In addition, patients with nonviable myocardium had similar mortality outcomes whether they were revascularized or not. Interestingly, among medically treated patients, those with nonviable myocardium fared better. The explanation posited for this counterintuitive result relates to repeated ischemia-reperfusion injury in hibernating tissue that can predispose to electrical instability, arrhythmogenicity, and sudden cardiac death. Randomized data in the post-STICH era to corroborate these results is still needed.

CMR has also helped validate the prognostically unfavorable entity of microvascular obstruction (MVO), which is characterized by damage and dysfunction of the myocardial microvasculature resulting in a “no-reflow” phenomenon whereby erythrocytes cannot penetrate beyond the myocardial capillary bed. CMR is currently considered the gold standard for the detection of MVO, which has characteristic imaging features and has been shown to occur in more than half of STEMI patients successfully reperfused with primary PCI.¹³ Its prognostic significance as an ominous predictor of adverse left ventricular remodeling, major adverse cardiac events, and cardiovascular death is now firmly accepted.^13–14

Advantages of CMR Over Other Imaging Techniques

In single photon emission computed tomography (SPECT) or positron emission tomography (PET) imaging, the presence of scar is inferred by the lack of uptake of myocardial perfusion tracers, whereas CMR affords the luxury of direct visualization of scar and normal myocardium within the same image. This reduces the likelihood of falsely labeling viable segments as nonviable due to relatively low perfusion tracer counts, especially in thinned walls where tracer counts will inherently be lower. Nuclear perfusion techniques also lack the excellent spatial resolution of CMR (1.5 mm vs 10 mm for nuclear) and suffer from ionizing radiation exposure. Wagner et al. showed that SPECT is inadequately sensitive in the detection of subendocardial scar (<50% TEI) compared to CMR in both human patients and a canine model with histopathologic correlation. Nearly one half of subendocardial infarcts were missed by SPECT in human subjects when CMR was used as the reference standard.¹⁵ Both modalities fared identically in the detection of near transmural infarcts (>75% TEI), but nearly one quarter of infarcts with 50–75% TEI went undetected by SPECT. This may result in the converse labeling of nonviable myocardium as viable.

Dobutamine stress echocardiography may be an attractive option for viability assessment because it is widely available, free of ionizing radiation, and cheaper. This method relies upon the demonstration of contractile reserve when myocardial function is augmented with the use of inotropic stimulation. A biphasic response of dysfunctional myocardium, whereby contractility improves with low dose dobutamine but deteriorates with higher doses due to inducible ischemia, increases the specificity for viability.¹⁶ Profoundly ischemic tissue, by comparison, demonstrates worsening contractility with low dose dobutamine that is sustained or accentuated with higher doses – this is sometimes referred to as a uniphasic response and is also thought to be a marker of viability. Myocardial wall thinning has traditionally been emphasized as a marker of scar and nonviability using this technique. However, as discussed earlier, a thinned myocardial wall may in fact have a limited scar burden and be mistakenly deemed nonviable. High rates of interobserver variability and limited acoustic windows are also significant limitations.^17–18

Current Applications of Viability Testing Using CMR

The most common clinical scenario concerning viability revolves around patients with obstructive coronary artery disease (CAD) and chronic LV systolic dysfunction, so-called “ischemic cardiomyopathy.” While the mortality benefits of surgical revascularization over guideline-directed medical therapy (GDMA) alone or in addition to percutaneous coronary intervention (PCI) have been well established in select patient populations with mild systolic dysfunction,^19–20 controversy still exists over such benefit in those with more significantly reduced LV ejection fractions. The STICH trial did not suggest mortality benefit of coronary artery bypass grafting over medical therapy in this patient population, nor did it support viability testing to help determine who may benefit from revascularization.²¹ However, its viability substudy did not cause a fundamental shift in the management of such patients due to some of the methodological limitations that have been discussed at length elsewhere.²² Nevertheless, the mortality and morbidity benefit was noted with longer term follow up from the same cohort,²³ improvement in accuracy of viability assessment in the decade that followed the original STICH trial, and the need for randomized controlled trials in this area have all sustained interest in this topic.

It has long been known that myocardial infarct size, when quantitated histologically using a canine model, is larger acutely than when assessed several weeks later.²⁴ This observation deserves special attention as appreciating its pathophysiologic basis and radiologic correlation will prevent misunderstanding this phenomenon as an “overestimate” of LGE by CMR in the peri-infarct period and thus as an inherent limitation of the modality. Early after a myocardial infarction, tissue edema, hemorrhage, and inflammation are present due to coagulation necrosis, cell membrane rupture, and spillage of cytosolic contents. The combined effect of these processes results in an expansion of extracellular space and increase in total myocardial wall thickness. During the healing process, however, the edema and inflammation gradually resolve while replacement fibrosis and scar formation take place. This may result in a relative reduction in wall thickness compared to the immediate post-infarction period, manifesting with LGE as a relative reduction in the ratio of hyperenhanced to normal myocardium. It is therefore not surprising that two thirds of patients in a recent STEMI cohort undergoing primary PCI²⁵ had myocardial segments reclassified as viable (TEI ≤ 50%) from nonviable (TEI 51–75%) when CMR was performed both acutely and several months after their infarcts. This reclassification, however, was not due to an exaggeration of LGE in the acute setting but rather a consequence of the natural evolution of myocardial cell death and subsequent healing. It can thus be argued that there is potential for recovery in myocardial segments with higher grade TEI in the acute setting, a finding that has previously been corroborated by Choi et al.²⁶ The recent proposal to redefine viability criteria when CMR is performed acutely after an infarct,²⁵ while deemed unnecessary by the authors in light of the above, highlights the unresolved question of optimal timing of viability testing post-myocardial infarct.

In the EXPLORE study, concurrent PCI of chronic totally occluded (CTO) arteries in 150 STEMI patients did not result in global LV systolic improvement compared with medical management.²⁷ A CMR substudy of this trial, however, showed regional functional improvement in CTO territory segments as assessed by segmental wall thickening at 4 months,²⁸ a finding that only held true when TEI was < 50%. Kirschbaum et al. had previously reported that CTOs may take up to 3 years to manifest functional recovery after revascularization, with a statistically significant increase in segmental wall thickening as assessed by CMR at 3 years compared with 5 months post PCI.²⁹ This finding was sustained in both subgroups with intermediate scar burden (TEI < 25% and 25–75%). The results of EXPLORE are therefore not surprising but suggest that serial and delayed evaluation of ventricular function may have altered the results, particularly in segments with TEI >50%.

Conclusion

In patients with CAD and significant LV systolic dysfunction, clinicians are faced daily with the decision of whether revascularization would be beneficial. Through its ability to directly visualize scar and evaluate its transmural extent with LGE, CMR offers a unique advantage over its rivals as a central player in viability assessment and should be considered a first line technique in this regard (Figure 4). Randomized trial data related to outcomes after CMR viability testing is still lacking.

Figure 4. Proposed algorithm for viability assessment at our center.

We propose that cardiac magnetic resonance (CMR) be the first line test of choice for viability assessment, followed by positron emission tomography (PET) if contraindications to CMR exist. GDMT: guideline-directed medical therapy; ICD: implantable cardioverter-defibrillator; CRT-D: cardiac resynchronization therapy-defibrillator; PCI: percutaneous coronary intervention; CABG: coronary artery bypass grafting; AKI: acute kidney injury; GFR: glomerular filtration rate; CIED: cardiovascular implantable electronic device.

* At some centers, including our own, CMR is performed on patients with CIED with institutional safety protocols in place.

Key Points.

• In patients with ischemic heart disease and reduced left ventricular ejection fractions, cardiac magnetic resonance is an attractive option for the assessment of myocardial viability because it is unencumbered by body habitus or imaging planes, has excellent temporal and spatial resolution, lacks ionizing radiation exposure, and has relatively few contraindications.

• Late gadolinium enhancement using gadolinium-based contrast agents forms the fundamental basis for CMR viability assessment, with contractile reserve using low dose dobutamine potentially providing additive information.

• The transmural extent of infarction (TEI) of the left ventricular wall is how a myocardial segment is deemed viable or not and is inversely proportional to the likelihood of contractile improvement along a spectrum; a TEI value of ≤ 50% is generally considered reasonably likely to recover contractility after revascularization.

• While randomized data remains elusive, observational data support a mortality benefit when viable myocardium is revascularized either surgically or percutaneously compared to medical therapy in patients with coronary artery disease and severe left ventricular systolic dysfunction.

• The optimal timing of viability testing with CMR after an acute myocardial infarction as well as ventricular functional assessment post revascularization remains unclear.