OUP user menu

Measuring myocardial salvage

Hans Erik Bøtker, Anne Kjer Kaltoft, Steen Fjord Pedersen, Won Yong Kim
DOI: http://dx.doi.org/10.1093/cvr/cvs081 266-275 First published online: 5 February 2012


The efficacy of cardioprotective strategies can be quantified by myocardial salvage as an indicator of therapeutic benefit. Salvage is calculated as the difference between the area at risk (AAR) and the final infarct size (FIS). AAR has been quantified by angiographic assessment followed by quantification of FIS by biochemical ischaemic markers or imaging modalities such as cardiovascular magnetic resonance (CMR). Angiographical methods may overestimate AAR and since methodological differences may exist between different modalities, the use of different modalities for estimating AAR and FIS may not be recommended. 99mTechnetium (Tc)-Sestamibi single-photon emission tomography (SPECT) allows quantification of AAR and FIS by tracer injection prior to revascularization and after 1 month, respectively. SPECT provides the most validated measure of myocardial salvage and has been utilized in multiple randomized clinical trials. However, SPECT is logistically challenging, expensive, and includes radiation exposure. More recently, a large number of studies have suggested that CMR can determine salvage in a single examination by combining measures of myocardial oedema in the AAR exposed to ischaemia reperfusion with FIS quantification by late gadolinium enhancement. The T1- and T2-weighted CMR approaches for quantification of AAR utilize non-contrast, early and late gadolinium enhancement techniques. The technical progress, high spatial resolution and the potential for retrospective quantification of the AAR makes CMR the most appropriate technique for assessment of myocardial salvage. However, the optimum CMR technique for assessment of myocardial AAR remains to be defined. Consequently, we recommend a comprehensive CMR protocol to ensure reliable assessment of myocardial salvage.

  • Cardioprotection
  • Myocardial salvage
  • Acute myocardial infarction
  • Area at risk
  • Cardiovascular magnetic resonance

1. Introduction

Salvage of threatened myocardium is the principal mechanism by which patients with acute myocardial infarction (AMI) benefit from reperfusion. The combination of interventional and medical treatment has reduced mortality and morbidity significantly. Consequently, large patient numbers are required in clinical trials that aim at demonstrating improved survival of a new treatment. To diminish the need for large population sizes, numerous trials have used surrogate endpoints, including imaging measures of infarct size after coronary revascularization and indirect estimates of tissue damage by release of biomarkers, resolution of ST-segment elevation and myocardial blush grade.1,2 Although these endpoints have predictive power in large cohorts, the usefulness in the individual patient and in small-sized study populations is moderate.3 More recent clinical studies have used FIS and myocardial salvage as the primary endpoint (Table 1).47 Techniques that allow quantification of myocardial salvage provide advantages compared with those measuring only FIS: (i) quantification of salvage rather than FIS eliminates inter-individual variability and reduces requirement to the study population size, (ii) they permit proof-of-concept studies to test the potential efficacy of new therapeutic approaches in smaller study groups, and importantly (iii) they may provide further insight into the pathophysiology of ischaemia-reperfusion injury. The electrocardiogram may be used to estimate area at risk (AAR), but not FIS.8 Indirect measures of salvage can be calculated from angiographic estimates of AAR combined with estimates of FIS by imaging techniques or biochemical markers. Echocardiography is the preferred clinical method for evaluation of left ventricular (LV) function in patients with AMI. However, there are currently no specific echocardiographic measures of myocardial salvage. 99mTechnetium (Tc)-Sestamibi single-photon emission tomography (SPECT) and cardiovascular magnetic resonance (CMR) provide the assessment of myocardial salvage, using paired quantification of AAR and FIS either from two separate studies before and after revascularization, or retrospectively in one step when the patient has been treated and stabilized. The aim of the present review is to provide an overview of the current methods used for quantification of myocardial salvage in preclinical studies and clinical trials assessing the efficacy of reperfusion therapy and cardioprotection in AMI.

View this table:
Table 1

Capability, merits, disadvantages, and utility of methods used for assessment of myocardial salvage

TechniqueArea at riskFinal infarct sizeAdvantageDisadvantageSuitability
Radionuclide myocardial perfusion imaging SPECTGold standardAchievable
  • Well validated

  • Paired imaging

  • Can be performed in almost every patient

  • Assessment of AAR prior to intervention

  • Visualizes perfusion regardless of anatomy

  • Logistically demanding set-up as tracer has to be available 24/7

  • Radiation exposure

  • Tissue attenuation

  • Low spatial resolution

  • No distinction between new and old perfusion defects

  • Requires two examinations

  • Radiation exposure

Well validated in several clinical trials and in proof-of-concept studies
Radionuclide myocardial perfusion and metabolism imaging PETAchievableAchievableFIS is consistent with histopathology
  • Low spatial resolution

  • Logistical challenges due to tracer production that requires cyclotron facilities

Magnetic resonance
 T2-weighted CMRAchievableGold standard
  • Logistically easy

  • AAR and FIS in one examination

  • No radiation

  • No attenuation

  • High resolution

  • Able to detect small defects

  • Additional information (motion, flow, tissue characteristics) within the same examination

  • Limitations in evaluation of patients with pacemakers, internal defibrillators, claustrophobia or haemodynamic and electrical instability.

  • Gadolinium-based contrast agents should be avoided in patients with renal failure in patients with renal failure due to the risk of nephrogenic systemic fibrosis.

  • Imaging and quantification remain in evolution

  • T2-weighted methods are currently the best validated CMR method for quantification of salvage in clinical trials and in proof-of-concept studies, but other methods have shown potential

  • Comprehensive protocols necessary to secure reliable assessment

 Precontrast T1-weighted CMRAchievable
 Early gadolinium enhancementAchievable
 Late gadolinium enhancementAchievable
 VentriculogramAchievableNot achievableEasily achievable
  • Estimates only AAR. FIS has to be estimated by different methods

  • Inaccuracy of salvage due to diverging alignment between methods

Used for AAR in combination with other imaging modalites and biochemical ischaemia markers
 AngiogramAchievableNot achievableReadily available
 Cardiac CTTheoretically achievableTheoretically achievableNot knownRadiation exposureNot validated
  • Low specificity and sensitivity

  • Semiquantitative

Not validated
 EchocardiographyTheoretically achievableTheoretically achievableAvailability
  • SPECT, single-photon emission computerized tomography; PET, positron emission tomography; CMR, cardiac magnetic resonance; CT, computerized tomography; AAR, area at risk; FIS, final infarct size.

2. Pathophysiological events underlying the use of imaging techniques

Myocardial ischaemia-reperfusion injury is caused by a dynamic sequence of processes initiated by abrupt cessation of blood flow. A detailed description of the cellular and subcellular pathophysiological mechanisms has recently been reviewed.9 Imaging techniques use specific events in the cascade following ischaemia reperfusion for delineation of the affected structure, including impaired flow (perfusion imaging), impaired motion (strain), oedema (T2-Imaging), reversible and irreversible mitochondrial damage (99Tc-Sestamibi SPECT), cell membrane break-down (T1-Imaging with contrast agents), and finally necrosis (T1-Imaging with necrosis specific contrast agents).

3. Salvage as endpoint

Myocardial salvage is quantified by the difference between AAR and FIS, and myocardial salvage index can be calculated as the proportion of salvaged AAR (AAR-FIS/AAR). The salvage index is a measure of treatment efficacy, which allows comparisons among infarcts of different sizes. A salvage index of 1.0 indicates maximum treatment success, whereas a salvage index of 0 means no benefit of treatment. The ratio is sensitive to measurement errors when AAR is small, since the diagnostic precision of measuring perfusion defects of <3% of the left ventricle is reduced due to the limited resolution of the technique.10

We have measured AAR, FIS, salvage, and salvage index by SPECT in a large cohort of patients undergoing successful primary percutaneous coronary intervention (PCI) for STEMI (90% obtained TIMI 3 flow).11 The study was undertaken to compare post-infarction salvage and LV function in early (<12 h) and late presenters (12–72 h). Like others,5 we were unable to justify a specific limit of 12 h delay for offering STEMI patients primary PCI treatment. Scintigraphic measures of tissue damage and LV function were associated with the presentation delay, but with remarkably weak correlations (r2 < 0.05 for all). LV ejection fraction correlated positively with salvage index (r2 = 0.27) and inversely to FIS (r2 = 0.46). AAR was a strong predictor of FIS. We found no significant dependency of salvage index on AAR.11 Importantly, post-infarction shrinkage of AAR during the 30 days until imaging of FIS does not seem to affect the measured salvage.

The salvage index measured by SPECT12 or by CMR13,14 is associated with mortality. FIS has superior predictive value, since LV dysfunction after STEMI predicts major adverse cardiac events and mortality regardless of the size of the salvage index. If an AAR is reperfused optimally with a large salvage index and a small FIS as a consequence, the major adverse cardiac events and mortality would be similar to those of a small AAR and consequently, a small FIS even with a small salvage index.

4. Myocardial perfusion imaging: 99mTc-Sestamibi SPECT

At present the best validated method for imaging FIS and AAR is myocardial perfusion imaging (MPI) using 99Tc-sestamibi single-photon emission computerized tomography SPECT15 (Figure 1).

Figure 1

Area at risk (AAR) and final infarct size (FIS). Upper pictures show SPECT with myocardial AAR as blue and black (A) and FIS as blue (B). Myocardial salvage index (MSI) is calculated as the difference between AAR and FIS divided by the AAR. In this example, MSI = 60 (% of LV volume)–23 (% of LV volume)]/60(% of LV volume) = 0.61, where LV, left ventricle. Lower panel shows CMR with T2-weighted assessment of AAR (C) and FIS measured by the late gadolinium enhancement technique (D). In this example, MSI = 52 (g)–21 (g)]/52 (g) = 0.6. E represents the color scaling of SPECT where yellow represents normal perfusion, blue reduced and black absent perfusion.

99mTc-Sestamibi (a cationic, lipophilic compound) is one of the most applicable agents for clinical perfusion studies. The 140 keV photon energy peak of 99mTc-Sestamibi is optimal for gamma camera imaging. Its relatively short half-life (6 h) provides favourable patient dosimetry allowing administration of large doses to produce high-quality images. Experimental animal studies of the fundamental cellular uptake mechanisms have demonstrated that 90% of 99mTc-Sestamibi is concentrated within the myocyte mitochondria. Uptake of 99mTc-Sestamibi in viable myocardium is proportional to regional blood flow at rates up to 2 mL/min/g with very close correlations (r = 0.92–0.97) to radioactive microspheres uptake.16,17 At higher flow rates, 99mTc-Sestamibi underestimates myocardial blood flow as the peak extraction of the tracer is diminished.

4.1 Assessment of AAR

After initial myocardial uptake, subsequent 99mTc-Sestamibi clearance is slow without redistribution once bound to viable myocardium.18 Hence, any myocardial perfusion study will reflect perfusion at the time of tracer injection. Therefore, it is possible to assess AAR by tracer injection before coronary intervention and, due to the half-life time, postpone the subsequent SPECT imaging up to 8 h after angioplasty. SPECT visualization yields information about perfusion defects regardless of anatomy. The method inherently accounts for perfusion through collaterals that may reduce the hypoperfused ischaemic area. However, collateralization may change in relation to revascularization and lead to inaccurate assessment of salvage. Perfusion defects by an infarct-related artery (IRA) may be erroneously increased in the presence of other perfusion defects from co-existing non-IRA coronary arterial lesions, or underestimated in multivessel disease.

Metabolic derangements simulating severe ischaemia or hypoxia, producing membrane polarization changes, may diminish 99mTc-Sestamibi uptake independent of flow.19 We studied myocardial perfusion in STEMI patients treated with PCI in several studies. In two studies, 99mTc-Sestamibi was injected before PCI delineating AAR.6,20 In another patient group, 99mTc-Sestamibi was injected immediately after successful PCI in patients with initial total occlusion of the IRA.3 The perfusion defects after the PCI procedure were of the same size as the defects before PCI, indicating that myocardial perfusion was either not normalized by the PCI procedure despite achievement of normal flow in the IRA, or the extraction and fixation of 99mTc-Sestamibi in the myocytes were reduced due to loss of cell viability. Consequently, we evaluated the reliability of 99mTc-Sestamibi as a tracer of myocardial perfusion in a pig-model of myocardial infarction and reperfusion, conducted as 45 min of occlusion of the left anterior descending artery followed by 120 min of reperfusion. Distributions of 99mTc-Sestamibi and microspheres administered after 120 min of reperfusion were compared with histochemical staining for delineation of AAR and FIS. The 99mTc-Sestamibi defects were not significantly smaller than AAR. After functional impairment of myocardial tissue, interstitial and cellular oedema are the next events following ischaemia reperfusion, whereas mitochondrial dysfunction accompanying sustained ischaemia is considered a relatively late phenomenon that leads to irreversible damage to the cell.21 While oedematous tissue would be unaffected by 99Tc-Sestamibi given after reperfusion, the tracer dependency of mitochondrial dysfunction might lead to underestimation of the real AAR. However, our findings indicate that 99mTc-Sestamibi SPECT is a valid method to detect AAR in STEMI patients up to 120 min after reperfusion, either spontaneously or following PCI.22 Mitochondrial dysfunction commences rapidly and is reversible in an area representing AAR subjected to myocardial ischaemia-reperfusion injury, regardless of vessel patency at the time of 99mTc-Sestamibi injection. The duration of the ischaemic memory of 99mTc-Sestamibi beyond 120 min is unknown.

4.2 Assessment of FIS

Assessment of FIS by 99mTc-Sestamibi SPECT requires repeated imaging in a stable post-infarction state. FIS imaging must be postponed at least 120 h as assessment of 99mTc-Sestamibi uptake performed at 48–72 h overestimates FIS. A further reduction in infarct size can be seen during the following days to weeks.23 Optimal timing is probably after several weeks.

FIS measured by 99mTc-Sestamibi SPECT is consistent with histopathological estimates,24,25 but 99mTc-Sestamibi MPI allows no distinction between previous and acute infarction. FIS estimated with 99mTc-Sestamibi SPECT is significantly associated with mortality.12,26 A cut-off value of <50% MIBI-uptake yields the best combination of sensitivity (89%) and specificity (96%) with a positive predictive accuracy of 81% for FIS. However, in the inferoseptal and diaphragmatic regions, accuracy is only 68%.27 In the study of 1164 patients from the Core trial, 6-month mortality was 1% for patients with the smallest quartile of FIS (<12% of the left ventricle) compared with 5% in the quartile with the largest FIS (>35% of the left ventricle). The usefulness of 99mTc-Sestamibi SPECT infarct size is well validated, as recently reviewed by Gibbons.28

4.3 Myocardial salvage

Several randomized single-centre trials have used myocardial salvage by SPECT as end point.4,6,7,20,2933 The trials are proof-of-concept studies testing the potential efficacy of new therapeutic approaches in smaller study groups. Subsequent larger trials are required to clarify whether the results are associated with clinical outcome. We have demonstrated that increased salvage is associated with improved LV function after remote conditioning during transportation to primary PCI in STEMI patients.12,34 Delineation of AAR prior to reperfusion therapy requires tracer availability in the catheterization laboratory on a 24 h basis and technical support for imaging within the following few hours. Because of the demanding set-up, efforts are made to develop methods for reliably estimating initial myocardium at risk after reperfusion therapy. Sciagra et al. recently published a small study of 36 AMI patients successfully treated with primary PCI.35 Comparing function and perfusion 5 days after revascularization, they found a close correlation between salvage index by the functional wall ‘thickening’ and the conventionally performed ‘perfusion’ salvage index—data supporting a previous BRAVE-2 sub study.36 These data are preliminary37 and need further investigation addressing quantification and optimized timing.

4.4 Analysis of imaging data

We recommend that at least two experienced readers of nuclear cardiology studies analyse data independently. Different methods and quantification tools can be used for evaluation of perfusion data. A complete evaluation of this technical topic is beyond the scope of this review. We have used the commercially available automatic quantitative program QPS (Cedars-Sinai Medical Center),38 which has been extensively described and validated.10,3941 In case of failure of the automatic quantification algorithm, we have manually applied tools for masking extra cardiac activity and/or defining the valve plane and the apex of the left ventricle. The sizes of the final infarct and of the initial AAR can be calculated as the areas of the left ventricle containing counts lower than a mean normal limit for pixels, using a gender-specific reference database.

4.5 Conclusion

Myocardial salvage estimated by 99mTc-Sestamibi SPECT is a validated measure for comparing the efficacy of different treatment modalities. Translation of increased myocardial salvage into a clinical benefit should be considered in light of the fact that a large salvage is usually associated with a small FIS. The resultant LV dysfunction determines subsequent major adverse cardiac events and mortality after STEMI.

5. Cardiovascular magnetic resonance

CMR integrates in a single examination an assessment of cardiac anatomy, function, and blood flow without radiation and with high spatial and temporal resolution. A unique feature of CMR is the capability to visualize myocardial tissue morphology to quantify AAR, FIS, and myocardial salvage. Compared with 99mTc-Sestamibi SPECT, myocardial salvage assessment by CMR is less validated as an endpoint in trials testing novel reperfusion strategies. While CMR with contrast-enhanced imaging of irreversible injury using gadolinium contrast (‘late gadolinium enhancement’, LGE) has become the gold standard for the assessment of FIS, several alternative CMR methods for the assessment of AAR have been proposed. AAR by CMR is determined retrospectively, which is a fundamental difference from SPECT, in which the AAR is usually determined by injection of the perfusion tracer before revascularization. Furthermore, CMR using advanced strain encodes imaging techniques (SENC) enables objective quantification of regional myocardial functions in patients with AMI, 42 which may aid in defining the AAR. There are several contraindications to CMR, including pacemakers, metallic implants or devices and claustrophobia. The use of gadolinium contrast agents should be avoided in patients with severe renal failure due to the risk of developing nephrogenic systemic fibrosis. Image quality may be degraded in patients with irregular heartbeats such as in atrial fibrillation. It has been estimated that 5–10% of patients who have been included in the study cohort will eventually not undergo CMR.43,44

5.1 Terminology of CMR

By CMR, imaging is determined by the sequential relaxation properties of protons after excitation by radiofrequency pulses. Accordingly, the characterization of myocardial tissue morphology by CMR is based on differences in T1 and T2 relaxation time between different tissues. T1 relaxation time (the spin-latice relaxation time) is a rate constant that describes the time required for the longitudinal magnetization to realign to its original value after excitation by radiofrequency pulses. T1-weighted images can differentiate between tissues with differences in T1 relaxation time. Hence, T1-weighted imaging can be used to visualize the difference of tissue distribution of gadolinium in normal vs. infarcted myocardium by LGE because tissue T1 is shortened by gadolinium contrast. Other pathologies with short T1 relaxation time such as the accumulation of methaemoglobin may also be depicted by T1-weighted imaging to delineate intramyocardial haemorrhage.

T2 relaxation time (spin–spin relaxation time) is the rate constant that describes how long protons remain synchronous or ‘in-phase’ after being tipped perpendicular to the main magnetic field, i.e. the time required for the transverse magnetization to fall by 63% of its original value. T2-weighted imaging generally depicts fluid with hyper-intense signal intensity while solid tissue such as myocardium appears with intermediate signal intensity. Consequently, an increase in free water content within the myocardium appears bright on T2-weighted images.

5.2 CMR assessment of area at risk

Since all current CMR AAR methods need final validation, the superiority of any of these approaches needs to be determined. Consequently, several CMR methods for assessment of AAR within the same patient may be required to improve diagnostic accuracy.

5.2.1 T2-weighted CMR for assessment of AAR

Myocardial oedema is a fundamental reaction to ischaemia and reperfusion. Visualization of myocardial oedema as a measure of AAR has been most intensively studied by T2-weighted CMR.45 Currently, most of the clinical validation in visualizing myocardial oedema has been reported for T2-weighted CMR using a short-TI triple-inversion recovery (STIR) prepared fast spin echo sequence. Dual suppression of the signal from fat and flowing blood establishes a contrast between regional oedema and normal myocardium (Figure 2).46 While prolonged T2 relaxation times of water-bound protons generate a water-specific brightness, the detection of myocardial oedema depends on subtle changes in T2 values between normal myocardium (45–50 ms) and acutely infarcted myocardium, which has T2 values ranging between 60 and 65 ms. Despite acceptable reproducibility,47 T2-weighted techniques are fundamentally subject to a low contrast-to-noise ratio between injured and normal myocardium.

Figure 2

Short-axis CMR images and corresponding pathology obtained four days following ischaemic reperfusion injury in a porcine heart. (A) Pathology shows intramyocardial haemorrhage in the antero-septal myocardium, which corresponds to a hypo-intense region on the T2-weigthed AAR image (B). On the T1-weighted image, intramyocardial haemorrhage is depicted by a hyper-intense region (C). The LGE image, which reflects the FIS, shows that microvascular obstruction is present within the infarct core (D).

A difficulty with T2-weighted techniques includes poor definition of endocardial borders in particular at the apex, where slow flowing blood in the ventricle may mimic myocardial oedema. A ‘bright blood’ technique, which is a hybrid between turbo spin-echo and steady-state-free precession, improves distinction between endocardial borders and seems to improve diagnostic accuracy.48 T2-weighted CMR may severely underestimate AAR by evaluation of large infarcts with severe reperfusion injury because the typical presentation of T2-weighted scans in such cases is the occurrence of hypointense cores that may reflect either the absence of oedema or the presence of intramyocardial haemorrhage. To overcome this challenge, T2*-weighted gradient echo sequences have been proposed for specific visualization of intramyocardial haemorrhage by exploiting the T2* shortening effect due to the elevated myocardial densities of paramagnetic haemoglobin degradation products (deoxyhaemoglobin, methaemoglobin) or blood degradation products (ferritin and haemosiderin).49 Our own preliminary studies in an experimental pig model with large anterior infarctions show that T1-weighted CMR using an inversion recovery sequence can accurately determine the size and presence of intramyocardial haemorrhage at 1 week after infarction (Figure 2). The combination of T2-weighted and T1-weighted scans may yield unique and synergistic information about the size and morphology of the AAR that may provide incremental prognostic information and allow for a more comprehensive evaluation of strategies to reduce reperfusion injury.

5.2.2 Precontrast T1-weighted CMR for assessment of AAR

Both T1 and T2 values in the myocardium change during the first hours after AMI consistent with development of myocardial oedema.45 Thus, T1- and T2-weighted imaging relies on the same pathophysiological processes associated with myocardial oedema. However, T1-weighted imaging for the assessment of AAR is currently investigational and needs more validation before implementation in clinical trials. While both T1- and T2-weighted imaging are non-contrast-enhanced techniques, gadolinium enhancement may also delineate AAR.

5.2.3 Early gadolinium enhancement for assessment of AAR

Early gadolinium enhancement (EGE) has been suggested as a measure of AAR.50 T1-weighted images acquired as early as 2 min after contrast injection showed enhancement in agreement with corresponding T2-weighted scans.51 The timing of acquisition seems to determine the enhancement area and the distinction between AAR and FIS. Further studies are required to establish the temporal distribution of gadolinium contrast agents in the ischaemic myocardium to ensure optimal timing of both AAR and FIS estimation.

5.2.4 Late gadolinium enhancement for assessment of AAR

LGE has also been suggested as an indirect measure of AAR in myocardial infarctions. The pathophysiological rationale behind this approach refers to the original work by Reimer and Jennings, who described the wave front of myocardial necrosis that occurs in myocardial infarctions.52 The transmural extent of infarction is considered inversely related to the extent of salvaged myocardium. The LGE AAR measurement is therefore based on the endocardial extent of FIS and salvage is reflected by the difference in transmurality between AAR and FIS. The endocardial surface area (ESA) method is the simplest CMR method. However, there is consistent evidence that the circumferential extent of LGE underestimates AAR and hence myocardial salvage.5355

5.3 CMR assessment of final infarct size

As recently reviewed, the LGE technique is validated in numerous experimental and clinical studies to establish CMR as a reference method for imaging irreversible myocardial injury and FIS.56 LGE is considerably more sensitive than perfusion SPECT for detecting subendocardial infarcts due to a higher spatial resolution.57 However, both CMR and SPECT have suboptimal spatial resolution for detailed assessment of FIS and AAR, which often comprises a mixture of non-ischaemic and necrotic myocardium. Increasing the spatial resolution of CMR by 3D acquisition is expected to more accurately visualize these complex border zones. The pathophysiological basis for imaging FIS by LGE CMR relies on the distribution of gadolinium contrast agents into the extracellular space after intravenous injection. In chronic infarctions, gadolinium enhances in the large extracellular space of collagenous scar tissue absent of cells. In acutely infarcted myocardial tissue, gadolinium enhances mainly due to rupture of the cell membranes, allowing a greater fraction of the contrast agent to enter than in viable myocardium. T2-weighted CMR, which is used in acute infarction to identify cellular oedema, also allows for differentiation of acute from chronic infarction since only recent infarcts contains oedema. This is a distinct advantage of CMR compared with SPECT, which may include older infarcts resulting in erroneous underestimation of myocardial salvage.

5.4 Myocardial salvage

For the assessment of salvage in AMI, it is important to consider the timing of the scanning. During the first week after the AMI, LGE CMR tends to overestimate the size of FIS presumable due to contrast enhancement of oedematous but viable myocardial tissue. In contrast at 4 to 6 weeks after infarction, infarct volume may shrink due to replacement of necrotic muscle by scar. Therefore, it is critical to standardize the timing of the FIS measurement in the study design. The optimal time to evaluate myocardial salvage within one cardiac examination seems to be at 1 to 2 weeks post-infarction, since the size of FIS by LGE CMR remains almost constant after 1 week58 and AAR remains stable for at least 7 days.58

Since essentially all CMR methods for assessment of AAR relies on detection of myocardial oedema or gadolinium enhancement in regions with disruption of cell membranes, there is concern that an intervention itself may affect the formation of oedema and the size of the measured AAR. In the presence of an increase in salvage compared with standard treatment, postconditioning and administration of exenatide, a glucagon-like-peptide-1, at the time of reperfusion in patients with STEMI AAR was unaffected by treatment.58 A few randomized single-centre trials have used myocardial salvage by CMR as an endpoint.5961 These trials are of proof-of-concept studies testing the potential efficacy of new therapeutic approaches and need confirmation in large-scale clinical outcome trials.

5.5 Analysis of imaging data

The greatest challenge in analysing CMR data for assessing myocardial salvage involves the segmentation of myocardial oedema on T2-weighted scans to depict AAR. Threshold techniques based on relative differences in signal intensity between oedematous and normal myocardium have mostly been used. However, segmentation based solely on threshold is inaccurate when there is more complex myocardial morphology such as in infarcts with microvascular obstruction or intramyocardial haemorrhage, which tends to create hypo-intense signal within the core of the infarcted area.49 More recently acquisition of T1 maps56 and T2 maps62 to quantitate absolute values of T1 and T2 relaxation has been suggested. However, it remains to be shown whether such an approach will improve accuracy and avoid misinterpretation of the true size of AAR. A more simplified approach to segmentation of AAR involves delineation of the infarct-ESA on LGE images, presuming that the wave front of myocardial injury in AMI travels strictly along the endocardial borders. This may not be correct in patients with large salvage56 where AAR and hence salvage are underestimated by ESA.

5.6 Direct comparison of AAR between CMR and SPECT

The immediate benefits of CMR compared with SPECT include the absence of radioactive tracers and a minor logistical challenge because CMR can potentially determine the AAR retrospectively from 24 h to several days after the infarction.

Compared with SPECT, LGE CMR has higher spatial resolution, allowing delineation of the transmural extent of myocardial infarction. In contrast to SPECT, the CMR technology enables homogenous tissue signal from the entire field of view alleviating tissue attenuation as a limitation. Consequently, both anterior and inferior myocardial infarctions are equally well depicted.

There are relatively few direct comparisons between CMR and SPECT for quantification of AAR. In selected STEMI patients with a totally occluded coronary artery at arrival to the catheterization laboratory, a reasonable agreement was found between T2-weighted CMR and SPECT for the assessment of AAR63 and also between contrast-enhanced CMR and SPECT.49 Our own data in consecutive STEMI patients suggest that CMR in comparison with SPECT shows a systematic overestimation of AAR, which means that the two techniques should not be used interchangeable in studies. This discrepancy in AAR estimation may reflect both fundamental differences between the two imaging modalities and also the fact that the tracer injection was given before primary PCI, while the CMR detection of myocardial oedema was done several days after the infarction.

5.7 Conclusion

There is no current consensus about the most appropriate CMR technique for assessment of myocardial AAR. We recommend a comprehensive CMR protocol, including acquisition of precontrast T1-weighted and T2-weighted images as well as EGE and LGE CMR in order to ensure reliable assessment of myocardial salvage.

6. Angiographic assessment of anatomical AAR combined with other estimates of FIS

Ventriculographic assessment of AAR relies on abnormally contracting segments,64 while angiographic assessment relies on the anatomical distribution of the coronary arteries.52,65 Three models have been proposed to be clinically useful to estimate the anatomic territory at risk of infarction. The most simple model is the Duke Jeopardy Score,66 by which the coronary tree is divided into six segments. All segments distal to ≥70% stenosis are considered at risk and assigned 2 points.

The BARI-score grades all terminating arteries by vessel length and calibre according to specific criteria. All scores associated with the culprit lesion are summed and divided by the global score of the entire left ventricle. The jeopardized myocardium is calculated as a percentage of myocardial volume. Quantification of AAR by the BARI-score has been validated by comparison with T2-weighted CMR 67 and CMR infarct-ESA.55,68 The APPROACH-score is based on a model by which the left ventricle is divided into regions determined by the relative proportion of myocardium perfused by each coronary artery in previous pathological studies of humans, taking the location of the culprit lesion as proximal, mid, or distal into consideration.69 Compared with T2-weighted oedema by CMR, the APPROACH-score seem to overestimate the AAR.55,70

6.1 Analysis of data and calculation of myocardial salvage

Calculation of salvage using angiographic estimates of AAR requires quantification of FIS by other imaging modalities, most frequently CMR71,72 or by measurement of biochemical ischaemic markers.64 Although inaccuracies may be introduced due to diverging alignment between angiography and CMR, salvage may still be calculated as a difference between AAR and FIS relative to LV mass. Specific cardiac troponins or creatine kinases may be useful for assessment of FIS. Peak troponins correlate with scintigraphic infarct size, and troponin T concentration 3 days after acute ST-elevation myocardial infarction predicts infarct size. Repeated measurements and calculation of area under the curve is preferred.64 However, when biochemical ischaemic markers are used for FIS, the effect of any infarct reducing intervention relies on similar baseline AARs in the compared study groups or more appropriately by regression analysis of the well-known correlation between AAR and FIS illustrated graphically supplemented by a comparison between slope to detect a difference with and without intervention64 (Figure 3).

Figure 3

Area under the curve (AUC) for the release of serum creatine kinase is expressed as a function of the circumferential extent of abnormally contracting segments (ACS), which is used as an estimate of the area at risk. Data points for the cyclosporine group lie below the regression line for the control group, indicating that, for any given area at risk, cyclosporine administration was associated with a reduction in the resulting infarct size as measured by creatine kinase release (adapted from Ref.64).

6.2 Conclusion

Angiographical assessment of myocardium at risk may overestimate AAR compared with T2-weighted CMR and CMR infarct-ESA. Because the angiographical models are frequently combined with FIS by CMR, caution should be raised as the methods cannot be used interchangeable without running a risk of overestimating salvage. Angiographic assessment of AAR may be combined with an FIS estimate obtained with biochemical ischaemia markers.

7. Other methods

Positron emission tomography using fatty acid tracers or 13F-fluordeoxyglucose relies on metabolic disturbances that estimate the ischaemic AAR retrospectively. The method is limited by low spatial resolution and logistical challenges due to tracer production that requires cyclotron facilities. Cardiac CT has not yet reached a stage of maturity for the assessment of AAR.73

8. Overall conclusion and present recommendations

Myocardial salvage is a valid surrogate endpoint for comparison of the efficacy of cardioprotective strategies. 99m Tc-Sestamibi SPECT for quantification of AAR and FIS remains the best validated method at present but the method is logistically challenging and expensive. CMR may represent an applicable future alternative. However, as the most appropriate CMR technique for assessment of myocardial AAR remains to be settled, we recommend a comprehensive CMR protocol including acquisition of precontrast T1-weighted and T2-weighted images as well as EGE and LGE CMR in order to ensure reliable assessment of myocardial salvage. Interchangeable use of different modalities may introduce inaccuracies because primarily AAR may not show agreement between methods.

Conflict of interest: none declared.


This work was supported by Fondation Leducq (06 CVD), The Danish Council for Independent Research (11–108354), The Danish Council for Strategic Research (11-115818) and The Novo Nordic Foundation.


  • This article is part of the Spotlight Issue on: Reducing the Impact of Myocardial Ischaemia/Reperfusion Injury


View Abstract