Acute coronary syndromes (ACS) represent a spectrum of heart disease including unstable angina, non ST-elevation myocardial infarction and ST-elevation myocardial infarction. Within the heart region at risk, necrosis first begins in subendocardium and then progresses on in a “wavefront” toward the epicardium with the increasing occlusion time¹⁾. It was previously believed that transmural infarction is present if the initial electrocardiographic (ECG) finding shows a ST-elevation, and a non-transmural infarction may be present if the ECG shows only transient ST-segment and T-wave changes. However, electrocardiographic-pathologic correlations are known to be far from perfect²⁾. Several techniques for analyzing the extent and degree of myocardial necrosis after an acute ischemic event have been introduced into the clinical setting^3–6). Each of these methods has specific advantages for the diagnosis of myocardial viability, but none of them allows the direct visualization of the transmural extent of viable and non-viable tissues. Recent advances in contrast-enhanced magnetic resonance imaging (ceMRI) offers high spatial resolution, and it allows for the direct visualization of the transmural extent of necrotic and non-necrotic myocardial tissues, as the tissues are defined upon histological examination^7,8).

Myocardial viability in patients with coronary artery disease and left ventricular dysfunction are strong predictors of patient outcome, and the distinction between reversible and irreversible myocardial injury is important^9,10). Ventricular contractile dysfunction is often observed in the acute stage, but it cannot be used to quantify the extent of myocardial salvage and to predict the future outcome because both the necrotic and stunned myocardium will have an impaired contractile function^11,12). Using ceMRI, previous studies have shown the transmural extent of infarction (TEI) to be predictive of recovery of segmental contractile function after myocardial revascularization¹³⁾ and for patients with chronic ischemic heart disease¹⁴⁾.

The aim of this study was to assess the relationship between the transmural extent of myocardial necrosis, as defined by ceMRI, and the initial ECG findings, and also to predict the segmental contractile function according to TEI after reperfusion in patients who had their first myocardial infarction.

We found that the increasing transmural extent of hyperenhancement on MRI was correlated with poor recovery of contractile function after revascularization. This finding would be expected based on the interpretation that hyperenhancement on ceMRI indicates myocardial infarction. Previous experimental studies have demonstrated excellent spatial correlation between the presence of Gd-DTPA hyperenhancement and regional myocardial necrosis on ex vivo imaging⁷⁾. Fieno et al⁸⁾ have reported that the size and shape of the hyperenhanced areas on ceMRI were identical to irreversible injury as defined by pathologic staining. Assuming that hyperenhancement indicates infarction, the greater amount of TEI in ST-elevation AMI patients compared to non ST-elevation AMI patients could be explained by the “wavefront phenomenon” of infarct progression beginning in the endocardium and growing toward the epicardium. The “wavefront phenomenon” could also elucidate our observation that the segment with nearly complete transmural infarction was not observed in non ST-elevation AMI patients. However, we know that electrocardiographic-pathologic correlations are far from perfect²⁾. In an experimental study, considerable animal-to-animal variations for the transmural extent of necrosis were observed, even for those animals with identical occlusion times¹³⁾. In our study, there were considerable person-to-person variations observed by ceMRI for the infarct size. It was interesting that the majority of segments with infarction had only non-transmural involvement and a significant number of segments exhibited a TEI between 25% and 75%. It may be related to the ability of ceMRI to delineate the transmural extent of viable and nonviable myocardium through its high spatial resolution capabilities. The presence of these segments emphasizes the fact that segmental viability, as defined by ceMRI, is not an “all-or-none” phenomenon. The intrinsically high spatial resolution of the technique seems to allow for the direct visualization of the transmural extent of necrosis and the residual viable tissue, which is not possible using any other existing technique¹⁶⁾.

In a previous study, Kim et al¹⁴⁾ have reported that the mean TEI was 10±7% in regions with improved contractility after revascularization, and 41%±14% in regions with no improvement in contractility. They also reported that 90% of the regions with 51% to 75% TEI before revascularization did not improve after revascularization. They suggested that the negative predictive accuracy for prediction of functional improvement was 100%, if a cut-off value of 75% TEI was chosen. Gerber et al¹⁷⁾ have also showed that the magnitude of segmental function recovery decreased with the increased degrees of TEI, and function recovery was completely absent in those segments showing >75% TEI. This finding was similar to the results of our study that showed that 94.7% of the segments with >50% TEI and all the segments with >75% TEI had dysfunctional wall motion. These regions would be considered nonviable according to the wall motion improvement criteria, even though a sizable epicardial rim of viable tissue was still present. Furthermore, Wu et al¹⁸⁾ have reported that nearly 90% of regions with infarction and normal contractility had only subendocardial infarction. This result is also comparable to our results. We found that 94.3% of segments with infarction and normal wall motion had subendocardial infarction and 100% of segments with small infarction (<25% TEI) had normal wall motion. Therefore, we suggest that functionally viable tissues can be identified by ceMRI before coronary revascularization. However, our finding that 4 segments without infarction had dysfunctional wall motion raises the question as to why those segments did not show the normal contractility. It has been shown that the recovery phase for stunned myocardium after myocardial infarction may take several weeks or even months¹⁹⁾. Thus, recovery in those segments may require a longer time period after ischemic injury.

We found a strong correlation between infarct size, as defined by ceMRI, and the peak CK-MB and Troponin-T (Figure 4). Mahmarian et al²⁰⁾ showed the relationship between infarct size as quantified by plasma creative kinase-MB time-activity curves, to the size of the perfusion defects, as quantified by thallium-201 single-photon emission computed tomography (SPECT), in patients with AMI. Although enzymatic and scintigraphic estimates of infarct size have good correlation, the results of other studies^21,22) have suggested that SPECT misses some small infarcts. One reason that infarcts might not be identified by SPECT is the technique’s fairly poor spatial resolution of about 10×10×10 mm full width half maximum^23,24), and infarcts smaller than this size are difficult to detect. Wagner et al²⁵⁾ have reported that of the 181 segments with subendocardial infarction, 85 (47%) were not detected by SPECT. In our results, 6 patients with small infarction (less than 10% LV) were identified by ceMRI, and this technique has high spatial resolution of about 1.9×1.4×8.0 mm-ie, almost 60-fold greater than SPECT.

The MRI sequence that was used in our study is susceptible to artifacts associated with patient movement or imperfect breath-holding, which can be erroneously interpreted as areas of hyperenhancement. In addition, the baseline ceMRI before revascularization was not assessed in the present study. Thus, definite conclusions on the functional recovery of dysfunctional segments deemed viable by ceMRI cannot be drawn from the present study. Others studies have shown that recovery of function after revascularization is related to TEI in patients with ischemic heart disease, with a gradual decrease in functional recovery paralleled by an increasing transmurality of the infarction, and functional recovery was completely absent in segments showing>75% TEI^14,17). This finding is similar to the results of our study. Based of these results, one might speculate that the determination of contractile function after revascularization would be helpful to establish the relative importance of ceMRI to predict functionally viable segments, which is considered to be an important variable of outcome.

The data from this study demonstrate that TEI evaluated by ceMRI is inversely related to the contractility after revascularization in patients with AMI. Functionally nonviable tissue according to wall motion criteria can be identified by assessing the infarct size, as defined by ceMRI, before revascularization in ischemic heart disease with abnormal wall motion, even though there is a sizable epicardial rim of viable tissue present.

Irrespective of the initial ECG findings in patients with AMI, the transmural extent of the necrosis was variable. So TEI, as defined by ceMRI, would be a better predictor than the ECG findings for the prognosis of contractility.