Abstract

Background

Myocardial stunning is an important sequela of acute coronary syndromes and its determination might affect decisions on defibrillator implantation and assist devices after myocardial infarction (AMI). The aim of the study was to evaluate and compare the sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of cardiac magnetic resonance imaging (CMR) assessing myocardial stunning after acute myocardial infarction using low-dose dobutamine (LDD), end-diastolic wall thickness, and contrast delayed enhancement (DE).

Methods and results

A systematic review of Medline, Embase, and Cochrane for all prospective trials assessing myocardial stunning by CMR following AMI was performed using a standard approach for meta-analysis for diagnostic test and a bivariate analysis. Search results revealed 9384 studies, out of which 17 met criteria. A total of 634 patients (mean age 59 years, 85% male, mean left ventricular ejection fraction: 52%) were included. DE-CMR had a weighted sensitivity of 87% and specificity of 68% to detect myocardial stunning using 50% transmurality as a cut-off, with a PPV and NPV of 83 and 72%, respectively. With an overall diagnostic accuracy of 82%, LDD-CMR had a sensitivity of 67% and a specificity of 81%, with a PPV and NPV of 82 and 63%, respectively. LDD showed an overall accuracy of 74%.

Conclusion

DE-CMR has a higher sensitivity, whereas LDD-CMR has a higher specificity for the detection of viable stunned myocardium following myocardial infarction. Whether the combination of DE and LDD may improve the prediction of myocardial recovery remains to be determined.

Introduction

Myocardial stunning refers to reversible myocyte dysfunction seen after blood flow has normalized post-reperfusion following an ischaemic event. Numerous scenarios can cause myocardial stunning, including coronary artery disease, global ischaemia during coronary artery bypass grafting (CABG), and plaque disruption with distal embolization during percutaneous coronary intervention (PCI).1 Stunned myocardium is considered a potentially reversible acute injury that can lead to left ventricular (LV) dysfunction and heart failure if unnoticed and/or untreated making the detection of dysfunctional myocardium post-reperfusion therapy clinically relevant.2 Myocardial stunning when coupled with haemodynamic instability prompts the utilization of several treatments. These include, among others inotropic agents, intra-aortic balloon pump (IABP) and temporary ventricular assist device support. It is important to mention that the utility of IABP support remains controversial based on the IABP-SHOCK II trial results.3 Furthermore, that the feasibility of cardiac magnetic resonance imaging (CMR) in haemodynamically unstable patients for the detection of myocardial stunning is limited.3 Furthermore, the detection of myocardial stunning with CMR while a patient is haemodynamically unstable is currently unfeasible.3 Similarly, the potential for recovery of a severely reduced left ventricular ejection fraction (LVEF) in this setting should be taken into account and the need for a cardiac defibrillator following standard therapeutic strategies could be utilized earlier based on the fact that the highest incidence of sudden cardiac death has clearly been demonstrated to be in the first 30 days after myocardial infarction (AMI).4 Nevertheless, this potential therapeutic modification has yet to be investigated as a consequence of detected myocardial stunning.5–7 Moreover, repeated bouts of ischaemia leading to episodes of stunning can lead to chronic LV dysfunction due to hibernating myocardium.8

There are few non-invasive modalities available to determine viability and stunned myocardium in the ischaemic heart, including echocardiography, nuclear perfusion imaging, and CMR. Dobutamine stress echocardiography detects contractile reserve, and positron emission tomography with fluorine-18-deoxyglucose shows viability via intact metabolism.9 Single photon emission computed tomography (SPECT) uses the radiotracers thallium-201, technetium-sestamibi-99 m, and technetium tetrofosmin-99 m to assess viability via myocardial metabolism and perfusion.10

CMR is one of the most versatile techniques to assess viability, with contrast delayed enhancement (DE)-CMR and low-dose dobutamine (LDD)-CMR being the main modalities used. DE-CMR uses gadolinium to determine the extent of fibrosis post-AMI in the LV wall, while LDD-CMR uses dobutamine in doses of 5–10 μg/kg/min range to assess contractile reserve by measuring wall thickness during contraction.11 A third method based on the determination of end-diastolic wall thickness (EDWT) has also been shown to be useful in patients with chronic ischaemic heart disease.12 Advantages of CMR are non-reliance on body habitus or acoustic windows, avoidance of ionizing radiation or angiography, and rapid data acquisition.11 Techniques in CMR, such as respiratory motion suppression, segment inversion recovery, and improved electrocardiographic gating, have allowed for increased quality in cross-sectional images of the heart, as well as more accuracy in determining the percentage of viable myocardial tissue.11,13–15

To the best of our knowledge, no meta-analysis has been conducted to evaluate the accuracy of the different CMR techniques to detect myocardial stunning.

Methods

Search strategy

The objective of this study was to evaluate prospective studies where cine-CMR was used to assess LV regional and global function in patients admitted with ST-elevation myocardial infarction who underwent percutaneous coronary angioplasty revascularization or CABG. A search was performed using PubMed, Cochrane Central Register of Clinical Trials, and Embase using the term: (acute myocardial infarction OR myocardial infarction OR stunned myocardium OR stunning myocardium OR revascularization) AND (MRI OR magnetic resonance imaging OR magnetic resonance OR cardiac magnetic resonance OR cardiovascular magnetic resonance OR contrast-enhanced MRI). The only set limit of the search was to adult humans in peer-reviewed journals from 1966 to June 2012. Trials in abstract form without a peer-reviewed manuscript were excluded. References were also reviewed and original articles were included.

Selection criteria

Trials included in the analysis had to meet following criteria: (i) prospective study involving patients with acute myocardial infarction undergoing revascularization (i.e. PCI, pharmacologic thrombolysis or CABG); (ii) use of CMR in evaluating LV function recovery in the short (i.e. <4 weeks) and long terms (>2 months) following revascularization; (iii) allowed for sensitivity, specificity, PPV, and NPV calculations based on viability segment analysis; and (iv) use of standardized cut-offs to predict accuracy.

Data extraction

J.R, J.K., and O.W. extracted the data independently using standardized protocol and reporting form. Characteristics of each trial, initial and follow-up CMR scan, interval between revascularization, baseline demographics and number of viable segments found in each scan were extracted. In cases where data were not readily available, the main investigator of that trial was approached to supply the relevant information.

Quality assessment

The assessment of quality of each study was done by evaluating 14 items considered relevant to the review topic, based on the Quality Assessment of Diagnostic Accuracy Studies instrument (QUADAS).16

Statistical analysis

A bivariate random effects model was used to calculate weighted sensitivities, and specificities, PPV, and NPV were calculated.17 This eliminates the problem with standard summary receiver-operating curve (ROC) model, which are able to generate diagnostic odds ratios but cannot differentiate between a test that can detect viable segments (sensitivity) from a test that is better at detecting non-viable segments (specificity).18,19

We assessed between-study heterogeneity by first plotting inverse logistic sensitivity and specificity into a summary ROC curve. By doing this, we used a bivariate approach, allowing us to estimate heterogeneity, obtain PPV and NPV, compare sensitivities and specificities between different methods, and increase the selection of summary ROC curves available.20–22 Figures and final analyses were created by using R (R 2.12.2 2001. The R foundation for Statistical Computing) and SAS (SAS 9.2, 2002–2008 SAS Institute, Cary, NC, USA), respectively. QUADAS and data entry was achieved by using Review Manager Revman, version 5.1 (Copenhagen, Nordic Cochrane Center, The Cochrane Collaboration, 2011).

Sensitivity analysis

Both CMR imaging methods were further evaluated depending on techniques and patient characteristics. For DE-CMR, standard deviation cut-offs, time between initial and follow-up CMR, age and per cent male were examined. For LDD-CMR, percent male, time between initial and follow-up CMR, and average age were examined. For both techniques, no factor appeared to affect the sensitivity of either DE or LDD-CMR.

Results

Study selection

We identified 9384 articles, out of which 6301 abstracts were retrieved and reviewed for inclusion (Figure 1). Seventeen studies (Tables 1 and 2) with 634 patients (mean age 59 ± 3 years, 85% male, mean LVEF: 52% ± 3) with a total of 7076 LV segments were included in the study, using the standard American Heart Association (AHA) 17-segment model.23 Nineteen studies were excluded from final analysis because they did not meet inclusion criteria (Figure 1). Resting EDWT was not analysed because only one study met inclusion criteria.24

Table 1

Baseline characteristics of DE-CMR studies

First author (ref. no.) Study design n Male (%) Age (years) LVEF (%) Revascularization Initial MRI (days) Follow MRI (weeks) Technique to assess LVEF Time after Gadolinium Admin (min) Hyper-enhancement (SD above normal intensity) Cut-off for viability (%) 
Beek et al.30 Prospective 30 90 59 51 PCI 7 ± 3 13 MRI 10–15 >6 <50 
Bodi et al.25 Prospective 40 92 57 49 PCI 4 ± 1 24 MRI 10 >2 <50 
Choi et al.33 Prospective 24 90 54 N/A PCI <7 10 MRI 10 >2 <50 
Engblom et al.26 Prospective 22 99 62 47 PCI 7 ± 6 28 MRI 20 N/A <50 
Gerbaud et al.29 Prospective 72 80 58 N/A PCI 5 ± 2 12 MRI 15 N/A <50 
Gerber et al.31 Prospective 20 70 61 N/A PCI 28 MRI 10 >2 <50 
Janssens et al.28 Prospective 67 82 57 56 PCI 4 ± 1 12 MRI 10–20 N/A <50 
Kitagawa et al.34 Prospective 18 60 60 N/A PCI 5 ± 1 39 MRI 15 N/A <50 
Motoyasu et al.32 Prospective 23 90 63 51 PCI 25 ± 6 25 MRI 10 >2 <50 
Natale et al.35 Prospective 46 N/A 65 N/A PCI 5 ± 2 20 MRI 10–15 N/A <50 
Shapiro et al.27 Prospective 17 80 60 55 PCI 6 ± 4 26 MRI N/A N/A <50 
First author (ref. no.) Study design n Male (%) Age (years) LVEF (%) Revascularization Initial MRI (days) Follow MRI (weeks) Technique to assess LVEF Time after Gadolinium Admin (min) Hyper-enhancement (SD above normal intensity) Cut-off for viability (%) 
Beek et al.30 Prospective 30 90 59 51 PCI 7 ± 3 13 MRI 10–15 >6 <50 
Bodi et al.25 Prospective 40 92 57 49 PCI 4 ± 1 24 MRI 10 >2 <50 
Choi et al.33 Prospective 24 90 54 N/A PCI <7 10 MRI 10 >2 <50 
Engblom et al.26 Prospective 22 99 62 47 PCI 7 ± 6 28 MRI 20 N/A <50 
Gerbaud et al.29 Prospective 72 80 58 N/A PCI 5 ± 2 12 MRI 15 N/A <50 
Gerber et al.31 Prospective 20 70 61 N/A PCI 28 MRI 10 >2 <50 
Janssens et al.28 Prospective 67 82 57 56 PCI 4 ± 1 12 MRI 10–20 N/A <50 
Kitagawa et al.34 Prospective 18 60 60 N/A PCI 5 ± 1 39 MRI 15 N/A <50 
Motoyasu et al.32 Prospective 23 90 63 51 PCI 25 ± 6 25 MRI 10 >2 <50 
Natale et al.35 Prospective 46 N/A 65 N/A PCI 5 ± 2 20 MRI 10–15 N/A <50 
Shapiro et al.27 Prospective 17 80 60 55 PCI 6 ± 4 26 MRI N/A N/A <50 

PCI, percutaneous coronary intervention; MRI, magnetic resonance imaging; N/A, none available; n, number of participants in each trial.

Table 2

Baseline characteristics of LDD-CMR studies.

First author (ref. no.) Study design n Male (%) Age (years) LVEF (%) Revascularization Initial MRI (days) Follow MRI (weeks) Technique to assess LVEF Dobutamine dose (μg/kg/min) Cut-off viability (mm) 
Barmeyer et al.36 Prospective 50 84 56 53 PCI 3.5 ± 14 32 MRI 
Barmeyer et al.36 Prospective 50 84 56 53 PCI 3.5 ± 14 32 MRI 10  
Bodi et al.25 Prospective 40 92 57 49 PCI 4 ± 1 24 MRI 10 
Dendale et al.37 Prospective 20 90 57 N/A PCI <7 16 MRI 
Gerbaud et al.29 Prospective 72 80 58 N/A PCI 5 ± 2 12 MRI 10 
Motoyasu et al.32 Prospective 23 90 64 51 PCI 25 ± 6 28 MRI 10 
First author (ref. no.) Study design n Male (%) Age (years) LVEF (%) Revascularization Initial MRI (days) Follow MRI (weeks) Technique to assess LVEF Dobutamine dose (μg/kg/min) Cut-off viability (mm) 
Barmeyer et al.36 Prospective 50 84 56 53 PCI 3.5 ± 14 32 MRI 
Barmeyer et al.36 Prospective 50 84 56 53 PCI 3.5 ± 14 32 MRI 10  
Bodi et al.25 Prospective 40 92 57 49 PCI 4 ± 1 24 MRI 10 
Dendale et al.37 Prospective 20 90 57 N/A PCI <7 16 MRI 
Gerbaud et al.29 Prospective 72 80 58 N/A PCI 5 ± 2 12 MRI 10 
Motoyasu et al.32 Prospective 23 90 64 51 PCI 25 ± 6 28 MRI 10 

All study designs were prospective, n = number of participants in each trial. PCI, percutaneous coronary intervention; N/A, none available; MRI, magnetic resonance imaging.

Figure 1

Selection of studies. Nineteen studies were also excluded from the final analysis because they did not meet inclusion criteria. AMI, acute myocardial infarction; MRI, magnetic resonance imaging; DE, contrast delayed enhancement; LDD: low-dose dobutamine.

Figure 1

Selection of studies. Nineteen studies were also excluded from the final analysis because they did not meet inclusion criteria. AMI, acute myocardial infarction; MRI, magnetic resonance imaging; DE, contrast delayed enhancement; LDD: low-dose dobutamine.

Baseline characteristics

Of the 16 studies, 11 studies25–35 enrolling 379 patients (mean age 60 years ± 3; 83% male) and analysing 5699 LV segments evaluated stunning myocardium by DE-CMR. All 11 studies used cine-CMR for follow-up. Correspondingly, for the LDD method, six studies25,29,32,36,37 with 255 patients (mean age 58 ± 3 years; 87% male) and 1377 LV segments were evaluated for stunned myocardium, again with all using cine-MRI for follow-up (Tables 1 and 2).

Quality assessment

Using the recommended 14-item list for evaluating imaging studies using QUADAS items, 11–14 were scored poorly or were considered unclear: item 11: (‘Reference standard results blinded?’), item 12: (‘Relevant clinical information?’), item 13: (‘Uninterpretable results reported?’), and item 14: (‘Withdrawals explained?’). Item 11 refers to blinding and may affect diagnostic accuracy, otherwise known as review bias. Item 12 refers to the baseline characteristics of the patients, and may lead to selection bias. Item 13 refers to any segments that may have been uninterpretable, and can, therefore, lead to false elevations of test accuracy and item 14 may lead to test performance bias, as patients unfit are removed to improve accuracy. One article did not explain the baseline characteristics of the patients beyond per cent male and age which may have led to selection bias.29 Otherwise, all studies showed high-quality scoring on the remaining 10 items (Figures 2 and 3).

Figure 2

Methodological quality summary. QUADAS, quality assessment of diagnostic accuracy studies.

Figure 2

Methodological quality summary. QUADAS, quality assessment of diagnostic accuracy studies.

Figure 3

Methodological quality graph.

Figure 3

Methodological quality graph.

Publication bias

There was no publication bias using Egger's test (P = 0.421). Similarly, no indication of publication bias was found using Peter's test (P = 0.111).

Delayed-enhancement CMR

A total of 11 studies evaluated myocardial stunning using DE with initial CMR performed during the first week after PCI, or thrombolysis and the second CMR being performed between weeks 5 and 39 with a mean of 22 ± 9 weeks. The standard deviation used to define hyper-enhancement ranged from >2 to >6 SD and did not appear to influence the test performance characteristics. Gadolinium was the contrast agent used in all studies and images were obtained 5–20 min after administration.

All studies used a 50% wall thickness cut-off to define the presence (<50%) or absence (>50%) of viability. Seven studies also reported the results when using 0%, <25 and <75% cut-offs, allowing us to compare the performance characteristics using variable definitions. One study included the 50% cut-off in a wider range (1–75%) in the follow-up CMR and, therefore, could not be included in final analysis.38 Also, one study involved bone marrow stem cell transplantation after revascularization, therefore, only control cases were included.28 The weighted mean sensitivity and specificity for the <50% cut-off was 87% (95% CI: 86–88) and 68% (95% CI: 66–70), respectively. The PPV was found to be 83 (95% CI: 82–84) and NPV was 72 (95% CI: 70–74; Tables 3 and 4). This technique had an overall weighted accuracy of 82% (95% CI: 81–83; Table 3).

Table 3

Sensitivities/specificities and predictive values of DE-CMR

Author Sensitivity Specificity PPV NPV Accuracy 
Beek et al.30 81 (221/273) 63 (143/227) 73 (221/305) 73 (143/195) 70 (364/500) 
Bodi et al.25 68 (40/59) 93 (87/94) 85 (40/47) 82 (87/106) 80 (127/153) 
Choi et al.33 92 (292/316) 48 (100/208) 73 (292/400) 81 (100/124) 70 (392/524) 
Engblom et al.26 79 (168/214) 39 (75/195) 58 (168/288) 62 (75/121) 60 (243/409) 
Gerbaud et al.29 91 (908/1003) 61 (104/170) 93 (908/974) 52 (104/199) 90 (1012/1173) 
Gerber et al.31 82 (179/219) 64 (109/170) 75 (179/240) 73 (109/149) 70 (288/389) 
Janssens et al.28 69 (42/61) 71 (93/132) 52 (42/81) 83 (93/112) 70 (135/193) 
Kitagawa et al.34 89 (141/158) 66 (38/58) 88 (141/161) 69 (38/55) 80 (179/216) 
Motoyasu et al.32 83 (146/175) 72 (74/103) 83 (146/175) 72 (74/103) 80 (220/278) 
Natale et al.35 92 (1065/1152) 75 (310/412) 91 (1065/1167) 78 (310/397) 90 (1375/1564) 
Shapiro et al.27 53 (80/150) 87 (130/150) 80 (80/100) 65 (130/200) 70 (210/300) 
Weighted mean 87 68 83 72 82 
Author Sensitivity Specificity PPV NPV Accuracy 
Beek et al.30 81 (221/273) 63 (143/227) 73 (221/305) 73 (143/195) 70 (364/500) 
Bodi et al.25 68 (40/59) 93 (87/94) 85 (40/47) 82 (87/106) 80 (127/153) 
Choi et al.33 92 (292/316) 48 (100/208) 73 (292/400) 81 (100/124) 70 (392/524) 
Engblom et al.26 79 (168/214) 39 (75/195) 58 (168/288) 62 (75/121) 60 (243/409) 
Gerbaud et al.29 91 (908/1003) 61 (104/170) 93 (908/974) 52 (104/199) 90 (1012/1173) 
Gerber et al.31 82 (179/219) 64 (109/170) 75 (179/240) 73 (109/149) 70 (288/389) 
Janssens et al.28 69 (42/61) 71 (93/132) 52 (42/81) 83 (93/112) 70 (135/193) 
Kitagawa et al.34 89 (141/158) 66 (38/58) 88 (141/161) 69 (38/55) 80 (179/216) 
Motoyasu et al.32 83 (146/175) 72 (74/103) 83 (146/175) 72 (74/103) 80 (220/278) 
Natale et al.35 92 (1065/1152) 75 (310/412) 91 (1065/1167) 78 (310/397) 90 (1375/1564) 
Shapiro et al.27 53 (80/150) 87 (130/150) 80 (80/100) 65 (130/200) 70 (210/300) 
Weighted mean 87 68 83 72 82 
Table 4

Sensitivities/specificities and predictive values of dobutamine stress CMR

Author Sensitivity (%) segments Specificity (%) segments PPV (%) segments NPV (%) segments Accuracy 
Barmeyer et al.36 49 (78/158) 71 (64/90) 75 (78/104) 44 (64/144) 60 (142/233) 
Barmeyer et al.36 56 (87/155) 71 (55/78) 79 (87/110) 48 (55/123) 60 (142/248) 
Bodi et al.25 42 (25/59) 93 (87/94) 78 (25/32) 72 (87/121) 70 (112/153) 
Dendale et al.37 57 (43/76) 88 (66/75) 82 (43/52) 67 (66/99) 70 (109/151) 
Gerbaud et al.29 72 (119/165) 86 (128/149) 85 (119/140) 74 (128/174) 80 (247/314) 
Motoyasu et al.32 88 (156/177) 81 (82/101) 89 (156/175) 80 (82/103) 90 (238/278) 
Weighted mean 67 81 82 63 74 
Author Sensitivity (%) segments Specificity (%) segments PPV (%) segments NPV (%) segments Accuracy 
Barmeyer et al.36 49 (78/158) 71 (64/90) 75 (78/104) 44 (64/144) 60 (142/233) 
Barmeyer et al.36 56 (87/155) 71 (55/78) 79 (87/110) 48 (55/123) 60 (142/248) 
Bodi et al.25 42 (25/59) 93 (87/94) 78 (25/32) 72 (87/121) 70 (112/153) 
Dendale et al.37 57 (43/76) 88 (66/75) 82 (43/52) 67 (66/99) 70 (109/151) 
Gerbaud et al.29 72 (119/165) 86 (128/149) 85 (119/140) 74 (128/174) 80 (247/314) 
Motoyasu et al.32 88 (156/177) 81 (82/101) 89 (156/175) 80 (82/103) 90 (238/278) 
Weighted mean 67 81 82 63 74 

Delayed-enhancement CMR using quartiles

The seven studies that reported results using additional cut-offs included a total number of 201 patients with 5558 segments analysed.

  • Cut-off 0%:30,33,34 The weighted sensitivity and specificity were 62% (95% CI: 58–65) and 81% (95% CI: 78–85, while the PPV and NPV were 81% (95% CI: 77–84) and 55% (95% CI: 52–59), respectively.

  • Cut-off <25%:26,28,30,32–34 The weighted sensitivity and specificity were 68% (95% CI: 65–70) and 72% (95% CI: 69–75), while the PPV and NPV were 74 (95% CI: 72–77) and 62% (95% CI: 59–65), respectively.

  • Cut-off <75%:26–28,30,33,34 The weighted sensitivity and specificity were 93% (95% CI: 91–94) and 46% (95% CI: 43–50), while the PPV and NPV were 65 (95% CI: 63–68) and 80% (95% CI: 77–84), respectively.

Dobutamine stress CMR

Stunned myocardium was assessed in six studies. All studies defined viability as a 2 mm increase in LV wall systolic wall thickening during a variable infusion rate of 5–10 μg/kg/min. One study separated results for infusion rates of 5, 10, and 20 μg/kg/min, and these results were reported separately while the 20 μg/kg/min dose was excluded.36 The average age of patients was 58 ± 3 years, with 87% male, and average LVEF was 52 ± 2% (Table 2). Initial CMR was performed in the first week except for one study which performed an initial CMR at week 4 following PCI.32 Follow-up CMR was performed between weeks 8 and 39, the average being week 24 ± 8, which was not statistically significant (Table 2). Using this technique, the mean weighted sensitivity and specificity were 67% (95% CI: 64–70) and 81% (95% CI: 78–85). The PPV and NPV were 82% (95% CI: 80–86) and 63% (95% CI: 67–71), respectively. The weighted overall accuracy was 74% (95% CI: 72–77) (Table 4).

The bivariate model showing summary diagnostic accuracies and comparing both methods are shown in Table 5. Forest plots and ROC curves are displayed in Figures 4A–E and 5A and B, respectively.

Table 5

Summary estimates for weighted mean sensitivity, specificity, positive predictive value, negative predictive value, and overall accuracy using bivariate model

 Overall (%) Delayed-enhancement CMR (%) LDD-CMR (%) P-value 
Mean weighted sensitivity 84.65 (83.70–85.61) 87.41 (86.40–88.42) 66.96 (63.99–69.92) 0.02 
Mean weighted specificity 72. 9 (70.71–74.07) 68.46 (66.48–70.43) 81.45 (78.26–84.65) 0.05 
Mean weighted PPV 83.20 (82.14–84.25) 83.25 (82.12–84.38) 82.83 (79.92–85.75) 0.54 
Mean weighted NPV 69.11 (67.39–70.83) 71.65 (69.62–73.68) 63.26 (60.01–66.51) 0.22 
Mean weighted acccuracy 80.95 (80.08–81.83) 82.17 (81.22–83.12) 74.49 (72.30–76.68) 0.46 
 Overall (%) Delayed-enhancement CMR (%) LDD-CMR (%) P-value 
Mean weighted sensitivity 84.65 (83.70–85.61) 87.41 (86.40–88.42) 66.96 (63.99–69.92) 0.02 
Mean weighted specificity 72. 9 (70.71–74.07) 68.46 (66.48–70.43) 81.45 (78.26–84.65) 0.05 
Mean weighted PPV 83.20 (82.14–84.25) 83.25 (82.12–84.38) 82.83 (79.92–85.75) 0.54 
Mean weighted NPV 69.11 (67.39–70.83) 71.65 (69.62–73.68) 63.26 (60.01–66.51) 0.22 
Mean weighted acccuracy 80.95 (80.08–81.83) 82.17 (81.22–83.12) 74.49 (72.30–76.68) 0.46 
Figure 4

Forest plots of diagnostic accuracies of DE-CMR and LDD-CMR. The size of the square-plotting symbol is proportional to the same size for each study. Horizontal lines are the 95% confidence intervals, and the summary sensitivity, specificity, and predictive values are calculated based on a random effects model.

Figure 4

Forest plots of diagnostic accuracies of DE-CMR and LDD-CMR. The size of the square-plotting symbol is proportional to the same size for each study. Horizontal lines are the 95% confidence intervals, and the summary sensitivity, specificity, and predictive values are calculated based on a random effects model.

Figure 5

Hierarchical summary receiver operating characteristic (HSROC) curves. HSROC plots of (A) DE-CMR and (B) LDD-CMR. Based on combined sensitivity and specificity weighted for sample size of each data set reflected by the size of the circles, showing average sensitivity and specificity estimate of the study results (solid square) and 95% confidence region around it.

Figure 5

Hierarchical summary receiver operating characteristic (HSROC) curves. HSROC plots of (A) DE-CMR and (B) LDD-CMR. Based on combined sensitivity and specificity weighted for sample size of each data set reflected by the size of the circles, showing average sensitivity and specificity estimate of the study results (solid square) and 95% confidence region around it.

Discussion

Our meta-analysis showed that DE-CMR provides the highest sensitivity and NPV for detecting stunned myocardium, while LDD-CMR provides the highest specificity and PPV. As expected, increasing the enhancement cut-off value for DE-CMR led to a higher sensitivity but a lower specificity. Alternatively, decreasing the enhancement cut-off value led to a higher specificity but a lower sensitivity. Differences in performance between the same CMR techniques for detecting myocardial stunning versus hibernating may be secondary to the differences in timing of follow-up MRI to detect stunning, a lack of a standard definition on the number of weeks needed to classify stunned myocardium in LV segments with late gadolinium enhancement and a decrease in scar volume over time which may be expected weeks to month AMI.39

Nevertheless, the distinction between reversible and irreversible myocardial damage and what exactly defines stunned myocardium is difficult to make with alternative imaging techniques. The European Society of Cardiology in their guidelines for acute heart failure defines stunned myocardium as myocardial dysfunction that persists in the short-term following restoration of blood flow and recommends the use of IABP if myocardial stunning is detected, as this is a reversible cause of severe heart failure after an AMI.40 Of note, there is no strict definition of imaging technique or time course to define stunning. These recommendations were based on consensus as well as the study by Waksman et al.41 which examined a subset of 16 patients post-revascularization complicated by cardiogenic shock, showing that IABPs were effective in treating cardiogenic shock, although this recommendation may change after the results of IABP-SHOCK II trial.3 However, whether selective use of the IABP support may be useful in those patients with stunned myocardium remains to be determined. The AHA and the American College of Cardiology have no specific recommendations in their guidelines for the detection and management of myocardial stunning.42,43

A comprehensive meta-analysis comparing CMR to other imaging techniques was not performed secondary to the limited number of studies found. Regardless, comparing our results with isolated studies using different imaging modalities, the superior results are evident. Ward et al.44 looked at gated SPECT versus transthoracic echocardiography to evaluate post-ischaemic stunning and found a specificity of 41% and PPV of 8% for segments in the moderate-severe perfusion defect range. In a recent study using resting redistribution 99-mTc-sestamibi SPECT, Javadi et al.45 found similar specificities ranging from 49 to 52%, although sensitivities were higher, ranging from 80 to 93% depending on the follow-up time used. CMR has been shown to be superior in evaluating LV function to echocardiography in previous studies, as well as a studies included in this meta-analysis.10,46–48 The improved accuracy of CMR in comparison with other imaging modalities evaluating stunning can be attributed to the fact that tomographic evaluation which is feasible with CMR, in which acoustic windowing optimization is negligible, with optimal assessment of the endocardial border definition and with the added value of late gadolinium enhancement. This difference in performance may be due to the fact that while echocardiography is adequate for evaluating myocardial contractility and nuclear scintigraphy for evaluating myocardial perfusion, CMR is the only method capable of evaluating the two criteria required to define myocardial stunning simultaneously.

It is worth pointing out that the DE-CMR studies included in this meta-analysis did not take into consideration microvascular obstruction (MVO) as an additional factor when assessing for myocardial stunning. MVO has clearly demonstrated to be an important factor for long-term prognosis after AMI regardless of infarct size.49

Data regarding potential treatments targeting stunned myocardium are rapidly accumulating. To date, ionotropic agents, such as dobutamine, dopamine, and milrinone and calcium sensitizers such as levosimendan are used to manage symptoms while cardiovascular shock recovers.7 Interestingly, Glucagon like peptide-1 inhibitors and ATP-potassium channel openers have been shown to improve myocardial metabolism and appear to be a promising lead for the treatment of myocardial stunning.50,51

Clinical implications

DE-CMR has a higher sensitivity than LDD-CMR for the detection of viable stunned myocardium following myocardial infarction, while LDD-CMR has a higher specificity. Whether the combination of DE and LDD may improve the prediction of myocardial recovery remains to be determined. This can be achieved by having patients undergo DE-CMR first to assess any areas of hyper-enhancement, followed by LDD-CMR to assess areas of wall motion abnormalities. The latter of course would be unnecessary if a patient has no hyper-enhancement in the LV wall indicating a higher probability of functional recovery. Likewise, if the area of DE is >50% of the LV wall, we could assume based of the high sensitivity and NPV of this technique that the LVEF will probably remain low over time. In this case, although unproven, it is feasible to consider the patients functional status, such as ejection fraction, set, and, therefore, the clinician may consider advanced therapies such as AICDs in an earlier phase where the incidence of sudden cardiac death is the highest.4 This approach might probably gain some acceptance if the results of VEST trial demonstrate that wearable defibrillator decrease mortality in this early period after AMI (Vest Prevention of Early Sudden Death Trial and VEST Registry, ClinicalTrials.gov Identifier: NCT01446965 VEST trial). It is crucial to recognize that myocardial stunning similarly to myocardial viability assessed by DE-CMR is not an all or none phenomenon given the fact that its diagnostic accuracies significantly depends on the amount of delayed enhancement as demonstrated in this meta-analysis.

Limitations

One limitation of our study was that DE-CMR and LDD-CMR diagnostic accuracies could not be statistically combined in order to evaluate how adding these methods might improve sensitivity and specificity upon each technique separately, given the fact that only two studies implemented both DE and LDD techniques on the same patients and patient level data were not provided in these articles. Finally, EDTW was not included as no studies were found in order to complete a comprehensive meta-analysis.

Conclusion

DE-CMR has a higher sensitivity while LDD-CMR has a higher specificity for the detection of viable stunned myocardium following myocardial infarction. Whether the combination of DE and LDD may improve the prediction of myocardial recovery remains to be determined.

Conflict of interest: none declared.

References

1
Vitale
J
Qiu
H
Depre
C
Pre-emptive conditioning of the ischemic heart
Cardiovasc Hematol Agents Med Chem
 , 
2008
, vol. 
6
 (pg. 
92
-
104
)
2
Barnes
E
Myocardial stunning in man
Heart Fail Rev
 , 
2003
, vol. 
8
 (pg. 
155
-
60
)
3
Thiele
H
Neumann
FJ
Ferenc
M
Olbrich
HG
Hausleiter
J
Richardt
G
, et al.  . 
IABP-SHOCK II Trial Investigators
. Intraaortic balloon support for myocardial infarction with cardiogenic shock
N Engl J Med
 , 
2012
, vol. 
367
 (pg. 
1287
-
96
)
4
Solomon
SD
McMurray
JJ
Finn
PV
Velazquez
E
Erti
G
Harsanyi
A
, et al.  . 
Sudden death in patients with myocardial infarction and left ventricular dysfunction, heart failure, or both
N Engl J Med
 , 
2005
, vol. 
352
 (pg. 
2581
-
8
)
5
Hida
S
Tanaka
H
Igarashi
Y
Shiba
C
Hatano
T
Usui
Y
, et al.  . 
Postischemic myocardial stunning is superior to transient ischemic dilation for detecting multivessel coronary artery disease
Circulation
 , 
2011
 
[Epub ahead of print December 15, 2011].
6
Braunwald
E
The stunned myocardium: prolonged postischemic ventricular dysfunction
Circulation
 , 
1982
, vol. 
66
 (pg. 
1146
-
9
)
7
Epstein
AE
Ellenbogen
KA
Mark Estes
NA
III
Freedman
RA
Gettes
LS
Gillinov
M
, et al.  . 
ACC/AHA/HRS 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the ACC/AHA/NASPE 2002 Guideline update for implantation of cardiac pacemakers and antiarrhythmia devices) developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons
J Am Coll Cardiol
 , 
2008
, vol. 
51
 (pg. 
e1
-
e62
)
8
Depre
C
Cardioprotection in stunned and hibernating myocardium
Heart Fail Rev
 , 
2007
, vol. 
12
 (pg. 
307
-
17
)
9
Armin
AE
The cardiac magnetic resonance (CMR) approach to assessing myocardial viability
J Nucl Cardiol
 , 
2011
, vol. 
18
 (pg. 
1
-
8
)
10
Dilsizian
V
Current diagnostic techniques of assessing myocardial viability in patients with hibernating and stunned myocardium
Circulation
 , 
1993
, vol. 
87
 (pg. 
1
-
20
)
11
KIM
RJ
Rafael
A
Chen
EL
Parker
MA
Simonetti
O
Klocke
FJ
, et al.  . 
The use of contrast enhanced magnetic resonance imaging to identify reversible myocardial dysfunction
N Engl J Med
 , 
2000
, vol. 
343
 (pg. 
1445
-
53
)
12
Klow
NE
Gullestad
L
Seem
E
Endresen
K
Outcome of bypass surgery in patients with chronic ischemic left ventricular dysfunction. Predictive value of MR imaging
Acta Radiol
 , 
1997
, vol. 
38
 (pg. 
76
-
82
)
13
Hundley
WG
Willett
DL
Sayad
DE
Lange
RA
Willard
JE
Landau
C
, et al.  . 
Comparison of quantiation of left ventricular volume, ejection fraction and cardiac output in patients with atrial fibrillation by cine magnetic resonance imaging versus invasive measurements
Am J Cardiol
 , 
1996
, vol. 
78
 (pg. 
1119
-
23
)
14
Fieno
DS
Chen
EL
Lomasney
JW
Klocke
FJ
Judd
RM
Contrast-enhanced magnetic resonance imaging of myocardium at risk
J Am Coll Cardiol
 , 
2000
, vol. 
36
 (pg. 
1985
-
91
)
15
Bruder
O
Schneider
S
Nothnagel
D
Pilz
G
Lombardi
M
Sinha
A
, et al.  . 
Acute adverse reactions to gadolinium-based contrast agents in CMR: multicenter experience with 17 767 patients from the EuroCMR registry
JACC Cardiovasc Imaging
 , 
2011
, vol. 
4
 (pg. 
1171
-
6
)
16
Whiting
P
REitsma
JB
Bossuyt
PMM
Kleijnen
J
The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews
BMC Med Res Methodol
 , 
2003
, vol. 
3
 (pg. 
25
-
38
)
17
Reitsma
JB
Rutjes
AW
Scholten
RJ
Bossuyt
PM
Zwinderman
AH
Bivariate analysis for sensitivity and specificity produces informative summary measures in diagnostic reviews
J Clin Epidemiol
 , 
2005
, vol. 
58
 (pg. 
982
-
90
)
18
Moses
LE
Littenberg
B
Combining independent studies of a diagnostic test into a summary ROC curve: data-analytic approaches and some additional considerations
Stat Med
 , 
1993
, vol. 
12
 (pg. 
293
-
316
)
19
Littenberg
B
Estimating diagnostic accuracy from multiple conflicting reports: a new meta-analytic method
Med Decis Mak
 , 
1993
, vol. 
13
 (pg. 
13
-
21
)
20
Arends
LR
van Houwelingen
JC
Heijenbrok-Kal
MH
Hunink
MG
Stijnen
T
Bivariate random effects meta-analysis of ROC curves
Med Decis Mak
 , 
2008
, vol. 
28
 (pg. 
621
-
38
)
21
Chu
H
Cole
SR
Poole
C
Meta-analysis of diagnostic accuracy studies accounting for disease prevalence: alternative parameterizations and model selection
Stat Med
 , 
2009
, vol. 
28
 (pg. 
2384
-
99
)
22
Harbord
RM
Egger
M
Whiting
P
Sterne
JA
A unification of models for meta-analysis of diagnostic accuracy studies
Biostatistics
 , 
2007
, vol. 
8
 (pg. 
239
-
51
)
23
Cerqueira
MD
Dilsizian
V
Jacobs
AK
Kaul
S
Laskey
WK
Pennell
DJ
, et al.  . 
AHA writing group on myocardial segmentation and registration for cardiac imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the cardiac imaging committee of the council on clinical cardiology of the American Heart Association
Circulation
 , 
2002
, vol. 
105
 (pg. 
539
-
42
)
24
Hillenbrand
HB
Lipke
C
Kostler
H
Pabst
T
Werner
E
Ertl
G
, et al.  . 
Detection of myocardial viability in acute infarction using contrast-enhanced (1)h magnetic resonance imaging
MAGMA
 , 
2003
, vol. 
16
 (pg. 
129
-
34
)
25
Bodi
V
Lopez-Lereu
MP
Losada
A
Nunez
J
Pellicer
M
Bertomeu
V
, et al.  . 
Usefulness of a comprehensive cardiovascular magnetic resonance imaging assessment for predicting recovery of left ventricular wall motion in the setting of myocardial stunning
J Am Coll Cardiol
 , 
2005
, vol. 
46
 (pg. 
1747
-
52
)
26
Engblom
H
Heiberg
E
Wagner
GS
Pahlm
O
Arheden
H
Rapid initial reduction of hyperenhanced myocardium after re-perfused first myocardial infarction suggests recovery of the peri-infarction zone: one year follow-up by MRI
Circulation
 , 
2008
, vol. 
2
 (pg. 
47
-
55
)
27
Shapiro
MD
Nasir
K
Nomura
CH
Sarwar
A
Ferencik
M
Abbara
S
, et al.  . 
Utility of cardiovascular magnetic resonance to predict left ventricular recovery after primary percutaneous coronary intervention for patients presenting with acute ST-segment elevation myocardial infarction
Am J Cardiol
 , 
2007
, vol. 
100
 (pg. 
211
-
6
)
28
Janssens
S
Bogaert
J
Theunissen
K
Deroose
C
Desmet
W
Kalantzi
M
, et al.  . 
Autologous bone marrow-derived stem-cell transfer in patients with ST-segment elevation myocardial infarction: double-blind, randomised control trial
Lancet
 , 
2006
, vol. 
367
 (pg. 
113
-
21
)
29
Gerbaud
E
Coste
P
Erickson
M
Corneloup
O
Dos Santos
P
Durrieu-Jais
C
, et al.  . 
Comparative analysis of cardiac magnetic resonance viability indexes to predict functional recovery after successful percutaneous coronary intervention in acute myocardial infarction
Am J Cardiol
 , 
2009
, vol. 
105
 (pg. 
598
-
604
)
30
Beek
AM
Bondarenko
O
Twisk
JWR
Hofman
MBM
van Dockum
WG
Visser
CA
, et al.  . 
Delayed contrast-enhanced magnetic resonance imaging for the prediction of regional functional improvement after acute myocardial infarction
J Am Coll Cardiol
 , 
2003
, vol. 
42
 (pg. 
895
-
901
)
31
Gerber
BL
Bluemke
DA
Wu
KC
Lima
JAC
Accuracy of contrast-enhaned magnetic resonance imaging in predicting improvement of regional myocardial function in patients after acute myocardial infarction
Circulation
 , 
2002
, vol. 
106
 (pg. 
1083
-
9
)
32
Motoyasu
M
Ichikawa
Y
Ishida
N
Uemura
S
Okinaka
T
Isaka
N
, et al.  . 
Prediction of regional functional recovery after acute myocardial infraction with low dose dobutamine stress cine MR imaging and contrast enhanced MR imaging
J Cardiovasc Magn Reson
 , 
2003
, vol. 
5
 (pg. 
563
-
74
)
33
Choi
KM
Gubernikoff
G
Vargas
JD
Parker
M
Judd
RM
Transmural extent of acute myocardial infarction predicts long-term improvement in contractile function
Circulation
 , 
2001
, vol. 
104
 (pg. 
1101
-
7
)
34
Kitagawa
K
Hirano
T
Makino
K
Kobayashi
S
Takeda
K
Sakuma
H
Diagnostic value of late gadolinium-enhanced MRI and first-pass dynamic MRI for predicting functional recovery in patients after acute myocardial infarction
Radiat Med
 , 
2007
, vol. 
25
 (pg. 
263
-
71
)
35
Natale
L
Bernardini
A
Meduri
A
Marano
R
Lombardo
A
Crea
F
, et al.  . 
Role of first pass and delayed enhancement in assessment of segmental functional recovery after acute myocardial infarction
Radiol Med
 , 
2011
 
[Epub ahead of print March 5, 2012]
36
Barmeyer
AA
Bansmann
M
Muellerleile
K
Bavastro
M
Heuer
M
Adam
G
, et al.  . 
Myocardial contractile response to increasing doses of dobutamine in patients with reperfused acute myocardial infraction by cardiac magnetic resonance imaging
Cardiology
 , 
2008
, vol. 
110
 (pg. 
153
-
9
)
37
Dendale
P
Holman
E
Avenarius
J
van der Wall
EE
de Roos
A
Validation of low-dose dobutamine magnetic resonance imaging for the assessment of myocardial viability after infarction by serial imaging
Am J Cardiol
 , 
1998
, vol. 
82
 (pg. 
982
-
90
)
38
Dall'Armellina
E
Lindsay
AC
Karamitsos
TD
Ferreira
V
Robson
MD
Kellman
P
, et al.  . 
Dynamic changes of edema and late gadolinium enhancement after acute myocardial infarction and their relationship to functional recovery and salvage index
Circ Imaging
 , 
2010
, vol. 
4
 (pg. 
228
-
36
)
39
Romero
J
Gonzalez
W
Garcia
MJ
Magnetic resonance imaging assessing viability in patients with chronic ventricular dysfunction due to coronary artery disease: a meta-analysis of prospective trials
J Am Coll Cardiol Cardiovasc Imaging
 , 
2012
, vol. 
5
 (pg. 
494
-
508
)
40
Nieminen
MS
Bohm
M
Cowie
MR
Drexler
H
Filippatos
GS
Jondeau
G
, et al.  . 
Executive summary of the guidelines on the diagnosis and treatment of acute heart failure: the task force on acute heart failure of the European Society of Cardiology
Eur Heart J
 , 
2005
, vol. 
26
 (pg. 
384
-
416
)
41
Waksman
R
Gotsman
MS
Hasin
Y
Intra-aortic balloon counterpulsation improves survival in cardiogenic shock complicating acute myocardial infarction
Eur Heart J
 , 
1993
, vol. 
14
 (pg. 
71
-
74
)
42
Antman
EM
Armstrong
PW
Bates
ER
Green
LA
Hand
M
Hochman
JS
, et al.  . 
ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction—executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1999 guidelines for management of patients with acute myocardial infarction
Circulation
 , 
2004
, vol. 
110
 (pg. 
588
-
636
)
43
Hunt
SA
Chin
MH
Feldman
AM
Francis
GS
Ganiats
TG
Jessup
M
, et al.  . 
ACC/AHA 2005 guideline update for diagnosis and management of chronic heart failure in adult: a report of American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to update the 2001 guidelines for the evaluation and management of heart failure): developed in collaboration with the American College of Chest Physicians and the International Society for Heart and Lung Transplantation: endorsed by the Heart Rhythm Society
Circulation
 , 
2005
, vol. 
112
 (pg. 
e154
-
e235
)
44
Ward
RP
Lang
RM
Spencer
KT
Williams
KA
Overestimation of postischemic myocardial stunning on gated SPECT imaging: correlation with echocardiography
J Nucl Cardiol
 , 
2006
, vol. 
4
 (pg. 
514
-
20
)
45
Javadi
H
Semnani
S
Jallalat
S
Yavari
P
Mogharrabi
M
Hooman
A
, et al.  . 
Scintigraphic parameters with emphasis on perfusion appraisal in rest 99mTc-sestamibi SPECT in the recovery of myocardial function after thrombolytic therapy in patients with ST elevation myocardial infarction (STEMI)
Perfusion
 , 
2011
, vol. 
26
 (pg. 
409
-
21
)
46
Bellenger
NG
Ray
SG
Lahiri
A
Coats
AJ
Cleland
JG
Pennel
DJ
Comparison of left ventricular ejection fraction and volumes in heart failure by echocardiography, radionuclide ventriculography and cardiovascular magnetic resonance: are they interchangeable?
Eur Heart J
 , 
2000
, vol. 
16
 (pg. 
1387
-
96
)
47
Baer
FM
LaRosee
K
Schneider
CA
Theissen
P
Deutsch
HJ
Schicha
H
, et al.  . 
Comparison of dobutamine transesophageal echocardiography and dobutamine magnetic resonance imaging for detection of residual myocardial viability
Am J Cardiol
 , 
1996
, vol. 
78
 (pg. 
415
-
9
)
48
Choi
EY
Park
S
Kim
HJ
Ahn
JA
Ko
YG
Choi
BW
, et al.  . 
Prediction of transmural extent of infarction with contrast echocardiographically derived index of myocardial blood flow and myocardial blood volume fraction: comparison with contrast-enhanced magnetic resonance imaging
J Am Soc Echocardiogr
 , 
2006
, vol. 
19
 (pg. 
1211
-
9
)
49
Wu
KC
Judd
RM
Lugo-Olivieri
CH
Barouch
LA
Schulman
SP
Blumenthal
RS
, et al.  . 
Prognostic significance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction
Circulation
 , 
1998
, vol. 
97
 (pg. 
765
-
72
)
50
Read
PA
White
PA
Khan
FZ
O'Sullivan
M
West
NEJ
Dutka
DP
A pilot study to assess whether glucagon-like peptide-1 protects the heart from ischemic dysfunction and attenuates stunning after coronary balloon occlusion in humans
Circ Cardiovasc Interv
 , 
2011
, vol. 
4
 (pg. 
266
-
72
)
51
Niwano
S
Niwano
H
Sasaki
S
Masuda
R
Sato
K
Masuda
T
, et al.  . 
Cardioprotective effects of sarcolemmal and mitochondrial K-ATP channel openers in an experimental model of autoimmune myocarditis. Role of the reduction in calcium overload during acute heart failure
Int Heart J
 , 
2012
, vol. 
53
 (pg. 
139
-
45
)

Comments

0 Comments