Article Text

Download PDFPDF

Original article
Long-term prognostic value of myocardial salvage assessed by cardiovascular magnetic resonance in acute reperfused myocardial infarction
  1. Ingo Eitel1,2,
  2. Steffen Desch1,2,
  3. Suzanne de Waha1,2,
  4. Georg Fuernau1,2,
  5. Matthias Gutberlet3,
  6. Gerhard Schuler1,2,
  7. Holger Thiele1,2
  1. 1Department of Internal Medicine, University of Leipzig, Heart Center, Leipzig, Germany
  2. 2Department of Cardiology, University of Leipzig, Heart Center, Leipzig, Germany
  3. 3Department of Diagnostic and Interventional Radiology, University of Leipzig, Heart Center, Leipzig, Germany
  1. Correspondence to Dr Ingo Eitel, University of Leipzig, Heart Center, Department of Internal Medicine and Cardiology, Strümpellstr. 39, Leipzig 04289, Germany; ingoeitel{at}


Objective In acute myocardial infarction, cardiovascular magnetic resonance (CMR) allows for quantifying the extent of salvaged myocardium after reperfusion as a potential strong end point for clinical trials. The aim of this study was to investigate whether the early prognostic significance of myocardial salvage assessed by CMR is sustained at long-term clinical follow-up in patients with ST-elevation myocardial infarction (STEMI) undergoing primary angioplasty.

Design, setting, patients We analysed 208 consecutive patients with STEMI undergoing primary angioplasty <12 h after symptom onset. T2-weighted and contrast-enhanced CMR was used to calculate the myocardial salvage index (MSI). Patients were categorised into two groups defined by the median MSI. The primary end point was the occurrence of major adverse cardiovascular events defined as death, reinfarction and new congestive heart failure at long-term follow-up.

Results The median MSI was 48 (IQR 27 to 73). Long-term follow-up was available in 202 patients (97%) at a median of 18.5 months (IQR 13.8 to 20.8). Major adverse cardiovascular events occurred in 33 patients (16%), with a significantly lower event rate in the MSI ≥ median group (7 vs 26 events, p<0.001). Mortality was significantly reduced in the MSI ≥ median group (2 vs 12 deaths, p=0.001). MSI was a significant independent predictor for a favourable long-term survival on multivariable Cox regression analysis after adjustment for established prognostic markers.

Conclusions MSI assessed by CMR predicts long-term clinical outcome in acute reperfused STEMI. Therefore, our data support the use of myocardial salvage as an end point for clinical trials investigating novel reperfusion strategies.

Trial registration number NCT00952224.

View Full Text

Statistics from


Survival of patients presenting with ST-segment elevation myocardial infarction (STEMI) is enhanced by rapid, complete and sustained reperfusion of the infarct-related artery.1 2 Time is important in this setting because if not timely reperfused, a wave front of irreversible injury gradually replaces the myocardium at risk.3–5 Reestablishing reperfusion of the acutely ischaemic, but reversibly injured, myocardium has the potential for salvaging viable myocytes. The assessment of myocardial salvage is important clinically because significant salvage results in the long-term reestablishment of myocardial contractile function and is associated with improved prognosis and outcome.6–8

Non-invasive imaging methods that can delineate the area at risk and salvageable myocardium are crucial in assessing the efficacy of reperfusion treatments. In recent years, T2-weighted cardiac magnetic resonance (CMR) imaging of myocardial oedema has been established first in animal models9 10 and then in humans11 12 as a robust measure of myocardium at risk in the setting of acute infarction. Therefore, the salvaged myocardium at risk can be measured as the difference between the area at risk and final infarct size.

We have recently demonstrated that the myocardial salvage index (MSI) assessed by CMR is a strong indicator of 6-month clinical outcome after STEMI.6 Notably, receiver operating characteristic (ROC) curve analyses demonstrated that myocardial salvage remained the strongest predictor of major adverse cardiac events (MACE), outperforming infarct size and microvascular obstruction (MO). However, prognostic data of myocardial salvage assessment by CMR are currently limited, and especially long-term follow-up data are lacking.

The aim of this study was to investigate whether the early outcome benefits in patients with significant myocardial salvage are sustained at long-term clinical follow-up in patients with STEMI reperfused by primary percutaneous coronary intervention (PPCI).


The detailed design of this prospective study has been previously described.6 In brief, patients with STEMI undergoing PPCI were eligible for this trial if symptoms lasted <12 h and if ST-segment elevation of at least 0.1 mV in ≥2 extremity leads or at least 0.2 mV in ≥2 precordial leads was present. To ensure that CMR imaging findings reflected an acute myocardial injury, patients were not enrolled if they had previous infarction. The study was approved by the local ethics committee, and all patients gave written informed consent.

Primary percutaneous coronary intervention and secondary prevention

PPCI was performed as described previously.6 The use of bare-metal or drug-eluting stents was left to the discretion of the interventional cardiologist. All patients were treated with aspirin, unfractionated heparin and clopidogrel. The use of glycoprotein IIb/IIIa inhibitors, ACE inhibitors, β-blockers and statins was strongly recommended according to guidelines.

Secondary prevention after infarction included lifestyle modification (motivation to diet, physical activity, weight reduction, smoking cessation), blood pressure control (target <140/90 mm Hg) and lipid control (target low-density lipoprotein cholesterol <100 mg/dl). In diabetic patients, the aim was to achieve hemoglobin A1c levels ≤6.5%.

Angiographic, electrocardiographic and enzymatic analysis

Coronary angiography of the target lesion was performed as described previously.6 In short, angiographic analysis included initial and final flow of the culprit vessel. Visual assessments were performed offline in the angiographic core laboratory by two blinded observers. For electrocardiographic interpretation, cumulative ST-segment resolution approximately 90 min after PCI expressed as percentage was calculated by two blinded observers as defined previously. Plasma samples for creatine kinase (CK) and the CK-myocardial band (MB) fraction were collected on admission and subsequently during the hospitalisation every 6 h for 2 days.

Cardiac magnetic resonance

All CMR examinations were performed on a 1.5-tesla scanner (Intera CV, Philips Medical Systems, Best, the Netherlands) with a dedicated five-channel phased-array surface coil in supine position as previously described.6 Left ventricular (LV) function was assessed by a standard steady-state free precession technique. For area at risk determination, short-axis slices covering the whole ventricle using a T2-weighted fat-suppressed triple inversion recovery breath-hold pulse sequence were obtained using a body coil. Late enhancement images covering the whole ventricle were acquired approximately 15 min after intravenous administration of 0.2 mmol/kg body weight of gadobutrol (Gadovist, Bayer Schering Pharma, Berlin, Germany). A three-dimensional inversion recovery turbo gradient echo sequence was used for image acquisition.

Image analysis

Offline image analysis was performed on an independent workstation with dedicated software (View-Forum release 5.2, Philips Medical Systems) by fully blinded observers as described elsewhere. The area of abnormal signal intensity was measured in the T2-weighted images and in the corresponding late enhancement images by manual delineation in each of the short-axis images (figures 1 and 2). A central core of hypointense signal within the area of increased T2 signal intensity, which is deemed to be haemorrhagic infarction, was included in the area at risk assessment.13 Care was taken to exclude increased signal intensity from the blood pool adjacent to the endocardium due to slow flow (figure 1, asterisk).14 Myocardial oedema was defined as a mean signal intensity >2 SDs of remote myocardium in T2-weighted images. Myocardium was considered infarcted if the signal intensity was >5 SDs above remote myocardium in late gadolinium enhancement (LGE) images.15 In patients with MO, these dark areas were included for infarct size analysis, and the area of MO was assessed separately. Salvaged myocardium was quantified as the difference between the volume of increased T2 signal (area at risk) and the volume of LGE (infarct size) as previously described (figures 1 and 2).6 11 The CMR core laboratory has good reproducibility and low interobserver and intraobserver variability for infarct size and myocardial salvage assessment.16 17 For myocardial salvage assessment, the coefficient of repeatability for interobserver variability was 5.4, and the coefficient for intraobserver variability was 4.8.

Figure 1

Myocardial salvage assessment in a patient with a myocardial salvage index (MSI) ≥ median 2 days after acute reperfused myocardial infarction (mid-ventricular, short-axis views). A, B: Image from T2-weighted CMR showing myocardial oedema of the anteroseptal segment (B: purple contour). Asterisks refer to inadequately suppressed blood signal (B: ‘slow flow artefact’). C, D: Corresponding late enhancement image showing a non-transmural, minor necrosis (D: purple contour) indicating major myocardial salvage.

Figure 2

Myocardial salvage assessment in a patient with a myocardial salvage index (MSI) < median (mid-ventricular, short-axis views) 2 days after infarction. A, B: T2-weighted images showing myocardial oedema of the anteroseptal segment (B: purple contour). C, D: Contrast-enhanced image showing high signal intensity on the same myocardial segments (D: purple contour). The comparison of oedema (myocardium at risk) with necrosis shows significant overlap, indicating only minor myocardial salvage.

Clinical end points

The primary end point of this study was the occurrence of MACE, defined as a composite of death, reinfarction and new congestive heart failure. In a secondary analysis, the individual components of the primary end point were analysed. Long-term follow-up was conducted via telephone contact using a standard questionnaire. In case of any events, these were verified by hospital charts, direct contact with the treating physician or contact with the local government registration. The follow-up interviewer was not aware of the results of the MRI studies.

The diagnosis of reinfarction during the index hospitalisation was based on clinical symptoms, new ST-segment changes, and an increase in the CK-MB levels above the reference limits in patients with normalised values or an increase of >20% from the last non-normalised measurement.18 At follow-up, any new ischaemic symptoms leading to hospital admission accompanied by elevated troponin were defined as recurrent myocardial infarction. New heart failure was defined as any congestive heart failure (rales, dyspnoea New York Heart Association class III–IV) requiring medical attention and treatment with diuretics occurring >24 h after the index event. If a patient experienced more than one event during follow-up, then only the first was recorded, and no additional data were used for the composite MACE analysis. If two events occurred simultaneously, then the more severe event was recorded (death > recurrent infarction > heart failure).

Statistical analysis

Patients were grouped by the median MSI into a < median MSI and a ≥ median MSI group. Each categorical variable is expressed as the number and percentage of patients. Most continuous variables had non-normal distribution and are therefore presented as medians together with the IQR. Differences between groups were assessed by the Fisher exact or the χ2 test for categorical variables and by the Student t test for continuous data with normal distribution. Otherwise, the nonparametric Wilcoxon rank-sum test was used. Correlation analyses were done by using Pearson or Spearman tests, as indicated.

For the combined clinical end point, the Kaplan–Meier method was applied, and differences were assessed by the log-rank test. Simple Cox regression analysis was used to identify predictors of MACE during long-term follow-up. HRs with their corresponding 95% CIs are reported. All variables that appeared to be associated with MACE at the p<0.05 level in univariable analysis were then tested for their independent prognostic value in a multivariable Cox regression analysis based on a stepwise algorithm with the p value set at 0.05 for entering and 0.1 for exclusion.

For additional comparison of the prognostic value of the MSI, infarct size and late MO with regard to MACE as well as mortality, ROC curves were generated, and the areas under the curves (AUCs) were calculated and compared using the c-statistics. C-statistic results were compared using the nonparametric method previously described by De Long et al.19

All statistical tests were performed with SPSS software, V.15.0. A two-tailed p value <0.05 was considered statistically significant.


Of 267 consecutive patients with STEMI, this prospective CMR study included 208 patients. Long-term follow-up was available in 202 patients (97%) at a median of 18.5 months (IQR 13.8 to 20.8). The study profile is illustrated in figure 3.

Baseline characteristics

Demographic and clinical characteristics are shown in table 1. The baseline characteristics (age, sex, risk factors) were similar between groups. Patients with an MSI ≥ median had a significantly higher frequency of anterior myocardial infarction and lower Killip class at presentation. The time from symptom onset to reperfusion was significantly shorter in the MSI ≥ median group (p<0.003). There was a significant inverse correlation of symptom duration and MSI (p<0.001). Patients with anterior myocardial infarction had a significantly higher MSI compared with patients with non-anterior myocardial infarction (58.6 (IQR 33.2 to 73.9) vs 39.9 (IQR 25.4 to 70.8), p=0.01).

Table 1

Main patient characteristics

Myocardial salvage was similar between women and men (MSI females 46.8 (28.6 to 70.5) vs MSI males 48.0 (30.9 to 73.5), p=0.76). No relationship between age and the MSI could be observed (r=−0.082, p=0.24).

Angiographic analysis, ST-segment resolution and enzymatic analysis

As previously reported, patients with an MSI ≥ median had a significantly higher frequency of Thrombolysis In Myocardial Infarction (TIMI) flow grade 3 before PCI (p<0.005). Consequently, complete reperfusion after PCI as assessed by TIMI flow grade 3 after PCI correlated significantly with the MSI (r=0.160, p=0.03).

ST-segment resolution as a continuous variable was significantly better in the MSI ≥ median group (p=0.007), and patients with an MSI ≥ median also had a significantly lower peak CK and CK-MB level (p<0.001).

Cardiac magnetic resonance

The median time between the index event and CMR was 3 days (IQR 2 to 4) for both groups. The main CMR results are presented in table 2. The median amount of oedema was 35.5%LV (IQR 29.2 to 44.9), and the median infarct size was 16.3% LV (IQR 10.0 to 26.4). The median calculated MSI was 48.3 (IQR 27.7 to 73.2). Late MO was identified in 134 (64%) patients, with a significantly higher occurrence in the MSI < median group (85 vs 49 patients, p<0.001). Infarct size and the extent of late MO were also significantly larger in the MSI < median group, whereas LV ejection fraction was significantly smaller (table 2).

Table 2

Cardiovascular magnetic resonance results

Clinical outcome

At long term follow-up, there were 12 cardiac deaths (11.8%) in the < median MSI group and 2 (2.0%) in the MSI ≥ median group (HR 5.96 (CI 1.33 to 26.64), p=0.006) (figure 4A). Likewise, nonfatal reinfarctions (7 (7%) vs 3 (3%), p=0.05) and congestive heart failure (7 (7%) vs 2 (2%), p=0.03) occurred significantly more often in the < median MSI group. Consequently, at long-term follow-up, MACE were significantly lower in the MSI ≥ median group (26 (25.5%) vs 7 (7%), p<0.001, HR 4.66 (CI 1.92 to 11.32)) (figure 4B).

Figure 4

Kaplan–Meier curve of the cumulative incidence of death, reinfarction and new congestive heart failure (A) and incidence of mortality (B) during long-term follow-up.

Several established markers of increased patient risk were associated with increased MACE and mortality at long-term follow-up by simple Cox regression analysis (tables 3 and 4). Using stepwise multiple Cox regression analysis, infarct size remained an independent predictor of the combined clinical end point, whereas myocardial salvage was excluded from the model (table 3). However, using multiple stepwise Cox regression analysis for prediction of long-term mortality, myocardial salvage together with age, diabetes mellitus and TIMI flow grade after PPCI emerged as independent predictors of mortality, and infarct size was excluded from the model (table 4).

Table 3

Predictors of major adverse cardiac events in univariable and stepwise multivariable cox regression analysis

Table 4

Predictors of mortality in univariable and stepwise multivariable cox regression analysis

ROC curve analyses further illustrated that the MSI is a strong indicator of MACE and mortality at long-term follow-up (figure 5A,B). The MSI was the strongest indicator for mortality, with an AUC of 0.796 (CI 0.672 to 0.925), as compared with late MO (AUC 0.783, CI 0.688 to 0.885) and infarct size (AUC 0.724, CI 0.638 to 0.860). The c-statistics, however, revealed only a trend for a significant difference between the MSI and infarct size (p=0.09). Regarding MACE, the calculated AUC for infarct size was 0.786 (CI 0.704 to 0.868) and thereby numerically greater than that for MSI (AUC 0.769, CI 0.689 to 0.850) and late MO (AUC 0.752, CI 0.676 to 0.829). However, c-statistics were not significantly different between infarct size and MSI (p=0.39).

Figure 5

Receiver operating characteristic curves showing the prognostic sensitivity and specificity of the myocardial salvage index (MSI), infarct size (IS) and late microvascular obstruction (MO) with regard to major adverse cardiovascular events (A) and mortality (B) at long-term follow-up. AUC = area under the curve. MSI, MO and infarct size were included as continuous variables.


The present study represents the largest series of patients with STEMI treated by PPCI in whom the long-term impact of myocardial salvage assessed by CMR on hard clinical outcome has been reported. Our results confirm and extend our previous findings, as we could demonstrate that the MSI is a strong predictor of clinical events 6 months after infarction. In our comprehensive long term follow-up analysis, the prognostic significance of the MSI was maintained. Therefore, our data support the use of myocardial salvage as a strong end point for clinical trials investigating novel reperfusion strategies.

Prognostic value of myocardial salvage

Although cell damage is induced early after coronary artery occlusion, reperfusion can limit infarct size and promote salvage of the ischaemic myocardium, which would otherwise become necrotic. Thus, to assess the efficacy of reperfusion treatment, it is necessary to determine how much myocardium is salvaged by measuring the final infarct size in relation to the initial myocardium at risk.

Single photon emission computed tomography has been extensively used in previous clinical trials to assess myocardium at risk20–23 and has shown that myocardial salvage predicts outcome.7 The use of this technique, however, is limited by low spatial resolution, radiation exposure, the need for two scans and the requirement for radioisotope administration at the unpredictable times of patient presentation.24 25 CMR overcomes many of these limitations and can retrospectively assess myocardial salvage within 1 week after acute coronary occlusion without interfering with patient care, requires no radiation and has high spatial resolution.

Previous studies have validated T2-weighted CMR histopathologically for area at risk and myocardial salvage assessment in reperfused and non-reperfused infarcts.9 10 Furthermore, trials demonstrated that myocardial salvage assessment by CMR is a reproducible tool17 that identifies and quantifies myocardial salvage in excellent agreement with single photon emission computed tomography12 and angiographic scores of myocardial salvage.26 27 However, prognostic myocardial salvage data using CMR are currently limited.

Our group was the first to demonstrate that the MSI assessed by CMR predicts outcome in acute reperfused STEMI.6 Although in our previous report several clinical and imaging variables were predictors of outcome, MSI was the strongest independent predictor on multivariable analyses. However, one of the limitations of this study was the relatively short follow-up time of 6 months. Whether the amount of myocardial salvage represents long-lasting clinical benefits was unknown. The present long-term follow-up analysis confirms our previous findings by demonstrating a persistent survival benefit and reduced MACE rate in patients with significant myocardial salvage.

In line with our 6-month results, MSI was a better indicator of mortality as compared to overall MACE. Interestingly, in our stepwise multivariate Cox regression analysis for prediction of long-term MACE, infarct size remained as an independent predictor in our model, whereas myocardial salvage was excluded. Of note, myocardial salvage and infarct size have a strong inverse correlation (r=−0.86, p<0.001), and if one of the variables is in the model, the other will be excluded if it does not yield important new prognostic information. ROC curve analysis further demonstrated that infarct size had a slightly better predictive value for clinical events than the MSI at long-term follow-up. Likewise, a recent study observed that infarct size was a stronger predictor of LV dysfunction than myocardial salvage.28 The researchers concluded that infarct size provides incremental value for predicting heart failure after acute reperfused STEMI beyond traditional risk factors, MO and myocardial salvage. However, this study had a small sample size (108 patients), did not compare the prognostic value of myocardial salvage versus infarct size on hard clinical outcomes and acquired only three short-axis slices for area at risk determination.

In contrast, another study has shown that the MSI was a major and independent determinant of adverse LV remodelling, and this result remained unchanged at multivariate analysis after correction for important parameters, such as the occurrence of MO, infarct transmurality and baseline LV ejection fraction.8 Adverse LV remodelling is an even more important predictor of survival than LV dysfunction after infarction.29 The strong correlation of adverse remodelling with myocardial salvage might explain that the MSI was the strongest correlate of mortality in the present study.8

Myocardial salvage as end point in clinical trials

Infarct size and myocardial salvage have been commonly used as surrogate end points for mortality in the setting of myocardial infarction.21–23 25 30 There are advantages of measuring the salvaged area at risk over infarct size as an indicator of therapeutic efficacy in clinical trials. Studies have demonstrated that the amount of myocardium at risk can vary substantially, even for a coronary artery occlusion in the same territory.31–33 Two patients may both have final infarct size of 30%, but in one patient, the area at risk may have been 60%, whereas in the other patient, it may have been 30%. Measuring myocardial salvage provides knowledge of treatment efficacy in these two patients that would not be available from measurement of infarct size alone.34 Consequently, infarct size reduction is suboptimal to assess the efficacy of reperfusion treatment, as final infarct size depends on multiple factors such as extent of area at risk, collateral blood flow, duration of myocardial ischaemia and metabolic demand.35 Notably, in animal studies, small variations in the occluded vessel resulted in a significant change in infarct size, and the myocardium at risk accounted for >70% of the variability in the extent of myocardial necrosis.32 33 Thus, myocardial salvage with an adjustment for the area at risk may be a better surrogate end point in clinical trials.

The prognostic relevance of myocardial salvage might also help to select patients who may benefit from intensification of medical or device treatment. However, further studies are needed to prove the clinical relevance of myocardial-salvage-guided treatment decisions.


Our study represents a single-centre experience with a limited number of patients, despite being the largest CMR study to date assessing the prognostic significance of myocardial salvage. We used the body coil for T2-weighted image acquisition to have a uniform signal reception within the field of view, but at the cost of a lower signal-to-noise ratio. Furthermore, we used visual assessment and subsequent manual delineation of enhanced regions. Although analyses were performed by experienced readers, manual tracing might be partially subjective. However, semiautomatic methods require some subjective input as well, and currently, there is no proof of superiority to simple visual assessment and manual delineation.

Finally, early imaging of infarct size after STEMI may overestimate true infarct size because of either peri-infarct oedema or partial volume effects and may therefore underestimate the salvaged are at risk.36 37 However, a recent study has demonstrated that early assessment of LGE after infarction is a strong independent predictor of late heart failure and adverse events and therefore is likely biologically significant.28 Furthermore, for myocardial salvage measurement, early assessment of the area at risk is crucial as myocardial oedema is maximal and constant over the first week after infarction, reducing thereafter.36 Thus, a stable window for the retrospective evaluation of the area at risk and thereby myocardial salvage is only present within the first week after infarction, with a risk of underestimation after that time.


This study demonstrates that myocardial salvage assessed by CMR predicts long-term clinical outcome in acute reperfused STEMI. The prognostic value of myocardial salvage enables its use as a primary end point in clinical trials assessing the efficacy of reperfusion treatment in patients with acute coronary occlusion.


View Abstract


  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics committee of the University of Leipzig.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.