MRI for Area at Risk, Myocardial Infarction, and Myocardial Salvage

Abstract

Almost all published pre-clinical studies of cardioprotective agents include a measurement of area at risk, infarct size, and thus allow determination of myocardial salvage as an indicator of therapeutic benefit. Until recently, SPECT imaging with injection of sestamibi prior to intervention was the only clinical method suitable for making similar assessments in patients. Over the past 5 years, a large number of papers have documented that MRI can non-invasively determine area at risk, infarct size, and myocardial salvage. While T2 weighted imaging has been the method used most commonly, pre-contrast T1 weighted images, and early gadolinium enhancement (EGE) images can also determine the size of the area at risk. All three of these MRI methods detect the area at risk based on myocardial edema resulting from ischemia. Late gadolinium enhancement (LGE) images provide a well -accepted reference standard for infarct size in all of those methods. Finally, late gadolinium enhancement images can also provide a single modality measure of myocardial salvage based on the “wavefront” of myocardial injury associated with the progressively more severe damage associated with acute MI. Essentially, the late gadolinium enhanced images can provide an endocardial based snap shot of infarct size and salvaged myocardium is estimated as the viable myocardium within the circumferential extent of the infarct. Thus, the purpose of this review is to provide an overview of how MRI can determine area at risk, infarct size, and thus measure myocardial salvage.

Introduction

An assessment of area at risk, myocardial infarction, and myocardial salvage are fundamental parameters needed in understanding the pathophysiology of ischemia reperfusion injury as well as for assessing the efficacy of treatments aimed at improving myocardial salvage. Seminal work in defining the wavefront of injury associated with myocardial infarction determined that the duration of ischemia was a major factor in determining what fraction of the area at risk went on to be infarcted.^{1, 2} These same physiological principles underlie the MRI methods used to image myocardial infarct size and myocardial salvage. While area at risk is a major determinant of infarct size other factors such as the severity of the perfusion defect modulate this relationship.³

Magnetic resonance imaging (MRI) has many characteristics that make it well suited to measuring area at risk, myocardial infarction, and myocardial salvage in pre-clinical studies and in clinical trials. The purpose of this review is to provide an overview of how MRI can determine area at risk, infarct size, and thus measure myocardial salvage.

Imaging provides one way to determine efficacy of cardioprotective therapy in terms of myocardial salvage using a metric that parallels many pre-clinical studies. The following diagram (Figure 1) was adapted from Kloner and Jennings⁴ to illustrate how the area at risk, the infarcted myocardium, and the rim of salvaged myocardium can be determined with imaging techniques such as MRI (Figure 2) that have high enough spatial resolution to resolve the transmural extent of infarction.⁵

Figure 1.

Schematic Diagram of Area at Risk, Infarct Size, and Myocardial Salvage

Figure 2.

Example of estimating area at risk with T2-prepared SSFP MRI (left panel), infarct size with late gadolinium enhancement (middle panel), and myocardial salvage as the blue zone in the schematic diagram.

MRI Terminology Relevant to Imaging Infarct Size and Area at Risk

MRI methods or “sequences” characterize tissue based on specific nuclear magnetic properties including T1 and T2. Simplistically, T1 relaxation time (spin-lattice relaxation time) is the rate constant which describes how quickly protons realign with the main magnetic field. T1-weighted images can differentiate tissues if they have suitable differences in intrinsic or pathological T1. T1-weighted images are commonly used to image the tissue distribution of gadolinium contrast which shortens tissue T1 (for example when imaging myocardial infarction).

T2 relaxation time (spin-spin relaxation time) is the rate constant that describes how long protons remain synchronously or “in-phase” after being tipped perpendicular to the main magnetic field (i.e. in the transverse plane). T2-weighted images generally show fluids as having high or bright signal intensity while solid tissue like myocardium has intermediate signal intensity. An increase in free water content of tissue increases the signal intensity on T2 weighted images.²⁴ Thus in the case of myocardial edema associated with acute myocardial infarction, the area at risk appears slightly brighter than remote myocardium when imaged with T2-weighted methods. While this summary cannot cover the subtleties of MRI, it provides hints of why some tissue may be bright or dark depending on either intrinsic characteristics of tissue, pathologic changes the change the tissue magnetic properties, or differential accumulation of gadolinium contrast.

Pathophysiological Basis of Imaging Myocardial infarction by MRI

Although gadolinium contrast enhanced imaging of myocardial infarctions had been reported as early as 1984⁶, the development of inversion recovery late gadolinium enhancement (LGE) MRI in the late 1990’s ⁷ and extensive pre-clinical and clinical validation studies established MRI as a reference standard method for imaging myocardial infarction and viability.

Kim et al published a seminal canine infarct study that documented that gadolinium enhanced MRI depicted infarcted myocardium with remarkable fidelity.⁸ That work was supported by extensive validations of gadolinium concentrations in acutely infarcted myocardium.⁹ Careful studies resolved infarct from the area at risk and remote myocardium to determine that gadolinium enhances acute and chronic myocardial infarction.¹⁰

Gadolinium enhancement of acute and chronic myocardial infarctions occurs for related reasons. Clinically available gadolinium contrast agents are injected into the intravascular space but rapidly distribute into the extracellular space as well. Due to size and charge, these contrast agents are excluded from the intracellular space. Acute myocardial infarction enhances because cell rupture allows gadolinium into a higher fraction of the tissue than in viable myocardium. Chronic infarcts enhance because the collagenous scar that forms is relatively acellular and thus has a high effective extracellular space.

Clinical validation studies established that gadolinium differentiated viable myocardium from infarcted myocardium in patients. Kim et al showed that the transmural extent of infarction predicted the recovery of contractile function after revascularization⁵, a study confirmed by Selvanyagan et al.¹¹ Three early studies showed that MRI infarct size correlated with serum biomarkers and predicted recovery of function in patients shortly after acute myocardial infarction.^12–14

One of the main advantages of MRI as a measure of infarct size is that it has the spatial resolution needed to determine the transmural extent of infarction. SPECT and PET validation studies confirmed that gadolinium reported the presence of infarction with as good or better sensitivity than these established clinical methods.¹⁵ In a combined canine and human study of myocardial infarctions by MRI and SPECT, SPECT clearly failed to detect small subendocardial infarctions that were seen on MRI and pathology.¹⁶ The high resolution of MRI allowed detection of microinfarcts after coronary angioplasty or stenting that were previously undetectable.¹⁷

The method was proven to be reproducible¹⁸ and was robust enough to perform with high sensitivity and specificity in a multicenter clinical trial.¹⁹ Thus, gadolinium late enhancement MRI has been accepted as an excellent method for assessing myocardial viability in both acute and chronic myocardial infarctions.

Pathophysiological Basis of Imaging Area at Risk and Myocardial Salvage by MRI

From an imaging perspective, many techniques can determine infarct size but fewer tests can image both area at risk and infarct size adequately enough to measure myocardial salvage for clinical trials. The two clinically relevant methods that are closest to meeting these requirements are single photon emission tomography (SPECT) and MRI. MRI can image the area at risk in at least 6 different ways, each with certain advantages and disadvantages relative to each other.

T2-weighted MRI has been studied most extensively and depends in general terms on myocardial edema delineating the area at risk.^{20, 21} Recent reports suggest pre-contrast T1-weighted MRI should also be able to determine the area at risk based on similar reliance on myocardial edema. Work in press shows that early gadolinium enhancement (EGE) images taken about 2–3 minutes after injection of gadolinium can be used to determine the area at risk.[in press] LGE images can be used to estimate the area at risk based on the assumption that LGE images the circumferential extent of infarction with very high sensitivity and therefore the transmural extent of infarction is a reciprocal measure of the area at risk. As an intracellular agent, manganese can depict the area at risk as a perfusion defect with similar timing considerations as sestamibi studies but this contrast agent is not ready for testing in humans.²² It is also possible to image a perfusion defect by first pass perfusion imaging but that is impractical in people due to the need to move the patient to the MRI scanner during the course of the acute myocardial infarction.²³ The four clinically relevant MRI methods will be reviewed in detail as metrics of the area at risk.

T2-Weighted MRI as a Measure of Area at Risk

Higgins and colleagues were the first investigators to study and quantify myocardial T2 in experimental acute myocardial infarctions.²⁴ Although the study was performed at very low field strength and thus absolute measurements cannot be directly interpreted within the context of modern higher field MRI scanners, many of the important concepts are well delineated in the paper. In particular, myocardial T2 and T1 both change in setting of acute myocardial infarction. Furthermore, the observed changes were theoretically consistent with myocardial edema and correlated with measurements of myocardial water content estimated by wet-weight to dry-weight ratios. Early work focusing on the relationship between myocardial T2 and myocardial infarction aimed to develop non-contrast enhanced methods for diagnosing myocardial infarction but showed heterogeneous results.²⁵ While some investigators found good correlations with infarct size, others found that T2-weighted images tended to overestimate infarct size. It was more than 2 decades after the first papers on myocardial T2 in acute infarction before in vivo MRI was clearly shown to correspond to the area at risk in acute myocardial infarction.²⁰

T2-weighted MRI has several characteristics that make it suitable and convenient with regard to imaging area at risk. T2 weighted MRI can determine the area at risk retrospectively – conveniently 48 hours after the infarct²⁰ That study used a canine model with 90 minute coronary occlusions to produce mostly subendocardial infarcts to highlight differences vs the transmural T2-weighted signal intensity changes and validated against microsphere measurements of area at risk. Also consistent with imaging subendocardial infarcts, quantitative analysis of regional strain²⁶ showed a partially reversible recovery of function in chronic follow-up. T2 weighted images correlated with the area at risk in occluded vessels in a model without reperfusion injury that might have altered water content.²⁵ Thus, T2-weighted MRI was able to measure the area at risk in both reperfused and non-reperfused canine infarct models when compared with microsphere reference standards.

Myocardial edema is a fundamental reaction to myocardial ischemia and reperfusion.^{27, 28} Abdel Aty et al documented that myocardial edema actually starts to occur early in the course of myocardial ischemia.²⁹. In fact, myocardial edema may be a mechanism that contributes to post-ischemic myocardial stunning as intracellular edema might alter the efficiency of the contractile apparatus.³⁰ In an editorial, Klem et al³¹ questioned whether edema occurs as result of ischemia based on work by Jennings et al³² but the quoted experiments actually showed statistically significant increases in measured total tissue water at 0.5, 3, and 15 minutes after reperfusion of a completely reversible coronary occlusion in dogs (15 minute). Whether edema forms prior to reperfusion is more controversial as the study by Jennings et al³² did not detect increases in total tissue water prior to reperfusion. However, Ugander et al reported that T1 increases during coronary occlusion at least confirming that another MRI parameter changes in ways consistent with development of myocardial edema.³³

One must recognize that the changes in water content in edematous myocardium are small relative to the differences in signal intensity between myocardium and blood. To put things in perspective, the T2 of normal myocardium is approximately 45–50 ms, the T2 of acutely infarcted myocardium is about 60–65 ms, while the T2 of blood is about 150–200 ms. Thus, unless T2-weighted images are displayed in ways to optimally discriminate the very subtle differences in T2 between normal myocardium and the area at risk, it can be easy to completely miss the diagnostic finding. MR images of the heart can now be obtained that quantify myocardial T2 in ways the may help solve some of the initial problems related to subtle differences in signal intensity between normal myocardium and the area at risk.^{34, 35}

Pre-contrast T1-Weighted MRI as a Measure of Area at Risk

As described by Higgins et al in 1983, myocardial T1 and myocardial T2 both change during the first few hours after acute myocardial infarction in ways that are consistent with myocardial edema.²⁴ Ugander et al presented a validation study correlating T1-weighted MRI and microsphere measures of area at risk at the combined 2011 Scientific Sessions of the Society of Cardiovascular Magnetic Resonance and the EuroCMR.³⁶ The pre-contrast T1-weighted approach to imaging the area at risk relies on similar pathophysiologic processes associated with edema formation in the myocardium as described in the T2 section. Thus, the T1-weighted approach to imaging the area at risk is likely to have similar pathophysiologic strengths and weaknesses as the T2-weighted methods. Since the imaging methods for T1-weighted and T2-weighted sequences are quite distinct from an MR physics perspective, the artifacts are likely to be different. Methods to quantitatively image T1 in the heart are now widely available and could allow objective approaches to determining what portions of the heart were ischemic. This area of research is relatively new and will require additional validation studies to understand what role it may play in imaging area at risk.

Early Gadolinium Enhancement (EGE) as a Measure of Area at Risk

Matsumoto et al from Kyoto University in Japan have realized that gadolinium enhanced images taken about 2 minutes after contrast administration enhance the area at risk associated with an acute myocardial infarction (in press). Images obtained in the first few minutes after contrast enhancement are usually termed early gadolinium enhancement (EGE) to distinguish them from typical late gadolinium enhancement images (LGE) which have been validated to correspond to the infarcted myocardium. In a study of 34 patients with acute myocardial infarction, Matsumoto et al found that EGE images consistently enhanced a more transmural distribution extent of the infarct related wall than the LGE images. The EGE images correlated with T2-weighted determined area at risk. If confirmed, this method has the benefit of being able to determine both area at risk and infarct size in the same examination and even the same dose of contrast. Significant work will be needed to understand the kinetics of contrast enhancement to ensure optimal timing of both the area at risk measurement and the infarct sizing determination.

Late Gadolinium Enhancement (LGE) as a Reciprocal Measure of Myocardial Salvage

Of all MRI methods proposed to determine myocardial salvage, the simplest method relies only on the transmural extent of the infarct on late gadolinium enhancement images.³⁷ The pathophysiological basis of this method stems from the work of Reimer and Jennings that helped describe the wavefront of myocardial injury that occurs in acute myocardial infarction.^{1, 2} As the endocardium is the part of the heart that is most vulnerable to ischemia, it will become infarcted after shorter ischemic episodes. Longer durations of ischemia generally produce more transmural infarcts as opposed to infarcts with increasing circumferential extent. Thus, the transmural extent of infarction is inversely related to the extent of salvaged myocardium.

O’Regan et al analyzed the transmural and circumferential relationships between T2 abnormalities and LGE.³⁸ They found that T2 abnormalities always encompassed the infarct and that the majority of the difference between the two methods was seen in the transmural direction of the infarct. Since the circumferential extent of the infarct averaged 82% of the T2-weighted area at risk, an estimate of area at risk based on the circumferential extent of infarction may underestimate the amount of myocardial salvage but the differences are only moderate in size. Ubachs et al found that area at risk based on the endocardial extent of LGE underestimated area at risk from T2 weighted images (23% vs 34% of myocardium) and consequently underestimated myocardial salvage.³⁹

Additional Insights from Preclinical and Clinical Validation Studies

Clinical studies support the concept that T2-weighted images characterize important aspects of acute myocardial infarction in patients. For example, T2-weighted images can differentiate acute from chronic myocardial infarction with high sensitivity and specificity.⁴⁴ Friedrich et al documented that T2-weighted abnormalities in acute MI are more transmural than LGE consistent with the concept that T2 depicts area at risk and LGE highlights the infarcted myocardium.⁴⁵ Comparisons between SPECT imaging during acute MI provided the most direct validation that T2 abnormalities represent the area at risk.⁴⁶ Invasive coronary angiography also provided an independent metric of area at risk to validate that MRI can measure the area at risk.^{47, 48} Carlsson et al found that the size of the T2 determined area at risk was stable between day 1 and day 7 post-MI but decreased in size over the next 6 weeks to 6 months.⁴⁶

Clinical prognosis studies of the area at risk have had heterogeneous results. Raman et al found that patients with non-ST elevation acute coronary syndrome T2 abnormalities in the heart were much more likely to require coronary revascularization than those without T2 abnormalities.⁴⁹ Masci et al reported that the size of the area at risk was independently associated with adverse left ventricular remodeling and early ST segment resolution.⁵⁰ On the other hand, Larose et al found that LGE was the strongest predictor of poor outcomes after acute myocardial infarction but that area at risk modulated the likelihood of functional recovery at all levels seen on early post-infarct imaging.⁵¹ While salvaged myocardium was a significant predictor of outcomes in univariable analysis, it was not significant in multivariable analysis. Long term outcome was best predicted by the permanent injury (i.e. late gadolinium enhancement). Larger studies will be required to understand the relative prognostic value of function, infarct size, and salvage.

Even though the use of MRI for imaging the area at risk is relatively new, the technique has already been used successfully in a small clinical trial of therapeutic benefits of hypothermia.⁵² A small sample size can show a therapeutic benefit if the methods are precise and reproducible. Pre-clinical studies with small sample sizes and precise laboratory measures have easily detected therapeutic benefits. Thus, one potential benefit of imaging area at risk, infarct size, and myocardial salvage in humans may be more rapid verification of a therapeutic benefit through smaller clinical trials that could be followed by pivotal large trials.

Comparisons between MRI and SPECT and Unmet Needs by MRI

Table 1 summarizes many of the relative advantages and disadvantages to using SPECT or MRI in clinical trials for assessing myocardial salvage. Overall, SPECT has major advantages in that it is has been tested in clinical trials and has few contraindications but SPECT measurements of area at risk were not feasible in AMISTAD and AMISTAD II. On the other hand, SPECT is limited by enough serious considerations that MRI should be studied carefully as an alternative methodology for assessing myocardial salvage. The two most substantial reasons to pursue MRI relate to the ability to make area at risk measurements after reperfusion and stabilization of the patient and the high image resolution which allows direct measurement of the transmural extent of infarction.

Table 1.

Relative Merits and Disadvantages of SPECT and MRI for Assessing Myocardial Salvage

Advantages	SPECT	MRI
	The sestamibi area at risk method has worked in multiple clinical trials.	High enough image resolution to determine the transmural extent of infarction and salvage.
	Clinicians understand the simplicity of an image of the perfusion defect.	Can be measured during a single MRI scan conveniently 2–7 days post-infarct.
	Can be performed in almost any patient.	At least 4 clinically relevant methods are available. The T2-weighted MRI area at risk method has worked in a few small clinical trials.
Disadvantages	SPECT	MRI
	Difficult to manage nuclear tracers in an Emergency Department.	No multicenter clinical trial experience with this method to date.
	Expensive to maintain staff and nuclear tracers on a 24 hour per day basis.	The T2-weighted methods are currently the best validated MRI methods but it is not clear which method will be most practical in clinical trials
	The sestamibi area at risk was not practical in a major multicenter trial.	Area at risk is reported based on myocardial edema associated with the ischemia.
	Image resolution is low enough that the transmural extent of infarction or area at risk cannot be visualized.	Theoretically, cardioprotective agents may modulate ischemia sufficiently that myocardial edema is not detectable.
	Requires a second nuclear exam about one week post-infarct.	Imaging and quantification methods remain in evolution.
	Uses significant doses of ionizing radiation	Gadolinium for infarct imaging is not recommended in patients with severe kidney disease.
		MRI has contraindications for some patients.
		Claustrophobia can limit scanning in some patients.

Significant unmet challenges exist with regard to using MRI as a tool to assess cardioprotective therapies. T2 weighted image quality remains an issue at many MRI centers. Cardiac MRI is not as widely distributed as the U.S. centers capable of enrolling patients into clinical trials of acute myocardial infarction. Cardiac MRI availability is better in Europe than in the US and is starting to be available in other parts of the world. There are important differences between scanning the heart by MRI vs other parts of the body. Thus, equipment availability is not equivalent to having physicians and technologists with the expertise needed to perform high quality scans, particularly in the setting of studies needing research measurements.

There is another layer of patient related issues that can influence the feasibility of using MRI to scan patients. Newer methods can make MRI more feasible even in patients that cannot hold their breath during imaging^{53,54, 55} Claustrophobia occurs in about 10% of the population and prevents successful MRI in about 5% of people and needs to be considered during sample size calculations. Patient size can limit image quality in almost all cardiac modalities. For MRI scanners with a ~70 cm bore it is possible to perform high quality cardiac MRI in patients up to about 400 pounds but scanners with a narrower bore might not accommodate patients over about 250 pounds depending on body habitus. Implanted devices such as pacemakers or defibrillators are generally considered contraindications to MRI. Although there is now one pacemaker that has been approved for diagnostic MRI scans, the US approval does not include MRI scans of the chest or heart.

There remain questions about how and when to interpret gadolinium and myocardial edema in the setting of acute myocardial infarction. Despite the validation studies, concerns remain whether cardioprotective therapies might also alter ischemia-related edema. If so, one might have to rely on the transmural extent of infarction as a reciprocal measure of salvage.

Finally, one can ask whether clinical trialists will accept any information short of randomized controlled clinical trials with hard outcomes such as death and MI. Considering the cost of large clinical trials, it makes sense to consider pilot studies to verify whether a proposed therapy is capable of reducing infarct size. Alternatively, imaging area at risk and infarct size could be performed in large enough subgroups to learn whether infarct size reduction is a plausible explanation for clinical benefits or failures in larger trials. Although much still needs to be learned about MRI in acute myocardial infarction, the multifaceted information that can be obtained by MRI warrants further study.

Recommendations

When designing a clinical trial of a cardioprotective strategy, it would be wise to incorporate several MRI measurements. Cine MRI would assess left ventricular size and function. Pre-contrast T2 mapping or T1 mapping could provide a volumetric assessment of area at risk. Early gadolinium enhancement could provide a second measure of area at risk (if timing issues are resolved and further validations are published) and would assess intracardiac thrombus. Late gadolinium enhancement (preferably at least 10 minutes and possibly 20 minutes after gadolinium injection) would provide a measure of infarct size. In a case where the primary measure of area at risk was not available or not assessable, the area at risk could be approximated by the reciprocal of the transmural extent of infarction. Recognizing there are many unresolved issues regarding quantification and serial changes in the degree of myocardial swelling post-MI, quantification should be based on the best available methods at the time of the study.