Int J Cardiovasc Imaging DOI 10.1007/s10554-014-0555-0
ORIGINAL PAPER
Differences in quantitative assessment of myocardial scar and gray zone by LGE-CMR imaging using established gray zone protocols Olurotimi Mesubi • Kelechi Ego-Osuala • Jean Jeudy • James Purtilo Stephen Synowski • Ameer Abutaleb • Michelle Niekoop • Mohammed Abdulghani • Ramazan Asoglu • Vincent See • Anastasios Saliaris • Stephen Shorofsky • Timm Dickfeld
•
Received: 2 July 2014 / Accepted: 15 October 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) imaging is the gold standard for myocardial scar evaluation. Heterogeneous areas of scar (‘gray zone’), may serve as arrhythmogenic substrate. Various gray zone protocols have been correlated to clinical outcomes and ventricular tachycardia channels. This study assessed the quantitative differences in gray zone and scar core sizes as defined by previously validated signal intensity (SI) threshold algorithms. High quality LGECMR images performed in 41 cardiomyopathy patients
Electronic supplementary material The online version of this article (doi:10.1007/s10554-014-0555-0) contains supplementary material, which is available to authorized users. O. Mesubi K. Ego-Osuala J. Jeudy A. Abutaleb M. Niekoop M. Abdulghani R. Asoglu V. See A. Saliaris S. Shorofsky T. Dickfeld Maryland Arrhythmia and Cardiology Imaging Group (MACIG), University of Maryland, Baltimore, MD, USA O. Mesubi K. Ego-Osuala S. Synowski M. Abdulghani R. Asoglu V. See A. Saliaris S. Shorofsky T. Dickfeld (&) Division of Cardiology, University of Maryland School of Medicine, 22 S Greene St, Baltimore, MD 21201, USA e-mail: [email protected] O. Mesubi Division of Cardiovascular Medicine, University of Iowa, Iowa City, IA, USA J. Jeudy Division of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, MD, USA J. Purtilo Department of Computer Science, University of Maryland, College Park, MD, USA
[ischemic (33) or non-ischemic (8)] were analyzed using previously validated SI threshold methods [Full Width at Half Maximum (FWHM), n-standard deviation (NSD) and modified-FWHM]. Myocardial scar was defined as scar core and gray zone using SI thresholds based on these methods. Scar core, gray zone and total scar sizes were then computed and compared among these models. The median gray zone mass was 2–3 times larger with FWHM (15 g, IQR: 8–26 g) compared to NSD or modified-FWHM (5 g, IQR: 3–9 g; and 8 g. IQR: 6–12 g respectively, p \ 0.001). Conversely, infarct core mass was 2.3 times larger with NSD (30 g, IQR: 17–53 g) versus FWHM and modified-FWHM (13 g, IQR: 7–23 g, p \ 0.001). The gray zone extent (percentage of total scar that was gray zone) also varied significantly among the three methods, 51 % (IQR: 42–61 %), 17 % (IQR: 11–21 %) versus 38 % (IQR: 33–43 %) for FWHM, NSD and modified-FWHM respectively (p \ 0.001). Considerable variability exists among the current methods for MRI defined gray zone and scar core. Infarct core and total myocardial scar mass also differ using these methods. Further evaluation of the most accurate quantification method is needed. Keywords Cardiac MRI Gray zone Myocardial scar tissue characterization Abbreviations FWHM Full width at half maximum CMR Cardiac magnetic resonance LGE-CMR Late gadolinium enhancement cardiac magnetic resonance LV Left ventricle LVEDV Left ventricular end-diastolic volume LVESV Left ventricular end-systolic volume m-FWHM Modified full width at half maximum
123
Int J Cardiovasc Imaging
MRI MI NSD ROI SCMR SI VT
Magnetic resonance imaging Myocardial infarction N-standard deviation Region of interest Society for cardiovascular resonance Signal intensity Ventricular tachycardia
magnetic
Introduction Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) imaging is the current gold standard for non-invasive imaging and assessment of myocardial fibrosis [1, 2]. Thus, magnetic resonance imaging (MRI) has emerged as an essential tool in our armamentarium for substrate assessment in the management of ventricular tachycardia (VT). The current mechanistic conceptual framework of reentrant VT is based on the understanding that the myocardial substrate consists of areas of surviving myocardium interspersed in myocardial scar. These form slow conduction pathways needed for the propagation and maintenance of reentrant VT circuits [3, 4]. Prior studies have demonstrated a strong and consistent correlation between LGE-CMR identified myocardial fibrosis and the incidence and occurrence of ventricular arrhythmias in ischemic and non-ischemic cardiomyopathy patients [5–7]. The Society for Cardiovascular Magnetic Resonance (SCMR) recommends that scar definition on LGE-CMR imaging should be signal intensity (SI) [2 SD above the average of normal or remote myocardium, based on initial validation studies [8]. From the original binary approach, based on LGE-CMR, where ‘‘white’’ regions of hyperenhancement identified areas of myocardial fibrosis, and ‘‘black’’ areas without enhancement correlated with viable myocardium, a new concept of ‘‘gray zone’’ has emerged. The gray zone is characteristically thought of as myocardial tissue that consists of an admixture of fibrotic areas (‘‘white’’) with surviving myocardium (‘‘black’’) resulting in a zone of intermediate SI with a net ‘‘gray’’ intensity appearance. Partial volume effects due to less spatial resolution in vivo compared to ex vivo myocardial imaging and cardiac motion are also thought to contribute to the appearance of the gray zone [1, 2, 9, 10]. Several techniques have been developed, primarily as research tools, for LGE myocardial scar tissue characterization [11–17]. There is no consensus on the optimal technique for these purposes. A recent SCMR position statement [18] acknowledges the existence of these several methods, but
123
refrained from making specific recommendations given the paucity of evidence. Using three different gray zone definitions, several landmark studies have demonstrated that the amount of gray zone in ischemic cardiomyopathy patients is an independent predictor of post myocardial infarction (MI) all-cause mortality [12], inducibility of ventricular arrhythmias [13] and appropriate implantable cardioverter-defibrillator (ICD) therapy [14] (‘‘gray zone for sudden cardiac death risk stratification’’). Additionally, gray zone reconstruction is now being evaluated as a potential complementary tool for MRI-guided anatomic VT ablation in animal and human studies [19–21]. This study sought to assess the quantitative differences in gray zone and scar core sizes as defined by three previously validated SI threshold algorithms and thus help define their role for diagnostic and future therapeutic applications.
Methods Study population This is a retrospective study of forty-one consecutive patients who presented for late gadolinium enhanced cardiac MRI from 2009 to 2011 at the University of Maryland Medical Center in Baltimore, Maryland. The institutional review board approved the study. Patients with cardiac device therapies such as ICDs or pacemakers that would result in significant artifacts were excluded from the study. MRI protocol A 1.5-T MRI Scanner (Siemens Medical Systems; Avanto, Erlangen, Germany) with an eight-element phased-array receiver body coil placed anterior and/or posterior to the chest was used. Images were acquired using vector ECGgating. The heart was imaged to cover the whole left ventricle (LV) depending on LV size for the cine images. Typical parameters were as follows: slice thickness of 8 mm; slice gap of 1 mm; average in-plane spatial resolution between 1.4 9 1.4 mm; temporal resolution 46–50 ms; and field of view (FOV) of 340 9 340 mm2. Delayed enhancement images were acquired 10 min after intravenous administration of gadolinium-DTPA (diethylenetriaminepentacetate) (0.15–0.20 mmol/kg, MultiHance; Bracco SpA; Milan, Italy). Great care was taken to optimize the nulling of the inversion time (usually between 200 and 300 ms) to assess for LGE. Views per segment and trigger delay were adjusted according to the patients’ heart rate to minimize image blurring. Typical parameters were as follows: FOV 360 9 360 mm2; slice thickness of 8 mm; no gap; average inplane spatial resolution between 1.4 9 1.4 mm, flip angle 25°, echo time 3.4 ms and repetition time 4–8 ms.
Int J Cardiovasc Imaging Fig. 1 Short axis LGE-CMR images from ischemic patient with inferolateral infarct. Epicardial and endocardial borders are outlined. a Myocardium without contours. b Full width half maximum (FWHM) method. c n-standard deviation (NSD) method. d Modified-FWHM method. The area demarcated in red is the scar core and the area in yellow is the gray zone. Note scar core and gray zone variations in all three techniques
MRI image analysis All MRI data analysis was performed on Digital Imaging and Communications in Medicine images with specialized software (QMass MR 7.5, Medis Medical Imaging Systems BV, Leiden, The Netherlands). For cine analysis, the endocardial and epicardial contours were drawn manually at the left ventricular end-systolic (LVES) and left ventricular end-diastolic (LVED) phases of the short axis cine images (Fig. 1). The papillary muscles were regarded as part of the ventricular cavity and the epicardial fat was excluded. The LVED and LVES volumes and masses were automatically calculated by the software. The software also automatically calculated the left ventricular ejection fraction (LVEF), stroke volume and LV mass (Fig. 2). Contrast enhanced MRI images were analyzed to determine infarct core, gray zone, and total infarct sizes. LV endocardial and epicardial contours were manually outlined on short axis contrast enhanced images. The papillary muscles were incorporated into the LV cavity. Subsequently, a CMR image slice amongst the series, determined by a trained observer to have a well-defined area of hyperenhancement with the most SI within the epicardial and endocardial borders, was selected and a Delayed Signal Intensity (DSI) analysis was performed. The DSI analysis automatically selected the gray zone and scar core area within the epicardial and endocardial borders, and also determined a remote region of interest (ROI) in the normal myocardium without artifact or
hyperenhancement [11, 12]. The QMass software automatically determined the scar core, gray zone and normal myocardium according to the three previously published algorithms [12–14]. In brief, the full-width at half maximum method (FWHM) [11, 13] assigns 100 % to the maximal SI value within the endocardial and epicardial contours. Any voxel that contains SI values between 50 and 100 % is defined as scar core [11, 13]. The gray zone is defined as voxels with SI higher than the peak SI of the remote ROI but \50 % of the maximal SI [13, 22]. The n-standard deviation method (NSD) determines the mean and standard deviation (SD) of the SI of a ROI in normal myocardium without artifact or hyperenhancement Infarct core was defined as voxels with SI [3 SD above the mean SI of the ROI while the gray zone was defined as areas with 2–3 SD above the mean SI of the ROI in the normal myocardial [12, 22]. Finally, the modified-FWHM defined the scar core in the same manner as the traditional FWHM, but defined the gray zone as areas with SI [35 % of the maximal SI but \50 % of the maximal SI [14, 22]. Statistical analysis Data is summarized as mean ± SD for continuous variables that are normally distributed and as median with interquartile range (IQR) for variables without normal distribution. Comparison across groups with repeated measures was assessed with the Friedman test [a nonparametric equivalent of the one-way Analysis of Variance
123
Int J Cardiovasc Imaging Fig. 2 Short axis LGE-CMR images from non-ischemic patient with an anteroseptal scar. Epicardial and endocardial borders are outlined. a Myocardium without contours. b Full width half maximum (FWHM) method. c n-standard deviation (NSD) method. d Modified-FWHM method. Red indicates scar core and yellow indicates the gray zone. Note scar core and gray zone variations in all 3 techniques
(ANOVA) test]. Post-hoc pairwise comparisons between groups were made with Dunn’s test of multiple comparisons using rank sums based on the Wilcoxon Signed-Rank test. Where appropriate, comparison between groups (without multiple comparisons) was done using the Wilcoxon Signed-Rank test. All tests were two-sided, and p values \0.05 were considered statistically significant. Correlations between gray zone mass and total myocardial scar were tested using Spearman’s correlation coefficient. Bland–Altman analysis was used to determine interobserver variability and interclass correlation coefficient (ICC) with 95 % confidence interval (CI) was computed. This was performed on a subset of 14 ischemic patients who had LGE analyzed by two independent readers (OM and KE). All analyses was done using STATA version 12.1 software package (StataCorp LP, College Station, Texas).
Results Study patients Table 1 summarizes the clinical characteristics and CMR findings of the study population. Indications for CMR were arrhythmias (20 %), recent MI (20 %), ischemic heart disease (20 %), myocardial scar evaluation (17 %), assessment of myocardial viability (10 %), cardiomyopathy (10 %), myocarditis (7 %), and ventricular aneurysm (2 %). Some patients had more than one indication for CMR. Of the 41 patients, there were 33 (80 %) ischemic
123
and 8 (20 %) non-ischemic patients. Online Resource 1 (Supplementary Table 1) summarizes the baseline characteristics of the groups (ischemic and non-ischemic). MRI characteristics All the patients demonstrated delayed cardiac enhancement on MRI. Study results are shown in Table 2. Significant differences between total scar, scar core, and gray zone were seen between the different gray zone algorithms (p \ 0.001 in all three categories). The median gray zone masses were 15 g (IQR: 8–26 g), 5 g (IQR: 3–9 g), and 8 g (IQR: 5–12 g) for FWHM, NSD, and modified-FWHM, respectively (Fig. 3a). The FWHM resulted in a two to threefold larger gray zone mass than the NSD (p \ 0.001) or the modified-FWHM method (p \ 0.001). There was no significant difference between NSD and modified-FWHM (p = 0.053). The observed median scar core masses were [2-fold larger with the NSD versus FWHM [30 g (IQR: 17–53 g) vs. 13 g (IQR: 7–23 g); p \ 0.001] or modified FWHM (13 g (IQR: 7–22 g; p \ 0.001) respectively (Fig. 3b). Given the identical definition of scar core in the FWHM and modified FWHM algorithms, no differences were seen between these algorithms. The median total LGE mass was smallest when using the modified FWHM [25 g (IQR: 12–32 g)] compared to NSD [39 g (IQR: 20–63 g); p \ 0.001] and was also smaller than the FWHM [34 g (IQR: 17–46 g); p \ 0.001] (Fig. 3c). The gray zone extent (percentage of LGE identified as gray zone) differed between the three algorithms (Fig. 3d)
Int J Cardiovasc Imaging Table 1 Baseline characteristics and CMR findings Variables
All patients (n = 41)
Age (years)
53 ± 14
Male gender n (%)
36 (88)
LVEF (%)
38 ± 13
Etiology of cardiomyopathy Ischemic n (%)
33 (80)
Non-ischemic n (%)
8 (20)
Co-morbidities Hypertension n (%)
20 (49)
Diabetes n (%)
11 (27)
Hypercholesterolemia, n (%)
15 (37)
Arrhythmia n (%) Scar location
11 (27)
Septal n (%)
20 (27)
Inferior n (%)
21 (28)
Inferolateral n (%)
14 (19)
Anterior n (%)
10 (14)
Anterolateral n (%)
9 (12)
CMR measurements LV Mass (g)
124 ± 37
LVEF (%)
41 ± 16
LVEDV (ml/m2)
212 ± 63
LVESV (ml/m2)
132 ± 72
LVSV (ml/m2)
80 ± 28
Values are n (%) or mean ± SD Some patients had multiple co-morbidities and LGE locations based on the AHA Bull’s-eye plot nomenclature LV left ventricle, LVEDV left ventricular end-diastolic volume, LVEF left ventricular ejection fraction, LVESV left ventricular end-systolic volume, LVSV left ventricular stroke volume
and was highest using the FWHM [51 % (IQR: 42–61 %)] compared with the modified FWHM [38 % (IQR: 33–43 %); p = 0.009], and lowest using the NSD method Table 2 Comparison of MRI defined myocardial scar variables
[17 % (IQR: 11–21 %)] compared to FWHM (p \ 0.001) and modified FWHM (p \ 0.001). There was a strong correlation between gray zone mass and total myocardial scar, irrespective of the SI threshold method (r = 0.95, p \ 0.001 for FWHM, r = 0.89, p \ 0.001 for NSD, and r = 0.97, p \ 0.001 for modified FWHM). Sub-group analysis (Fig. 4) comparing gray zone, scar core and total LGE masses in ischemic and non-ischemic patients demonstrated similar findings in both patient groups. The differences in gray zone and total scar masses among the three protocols were consistent in both the ischemic and non-ischemic subjects. The median scar core mass was [2 fold larger in the NSD versus FWHM and modified FWHM [33 g (IQR: 20–54 g) versus 13 g (IQR: 10–25 g); p \ 0.001 and 15 g (IQR: 9–23 g); p \ 0.001] respectively in the ischemic patients and in the nonischemic patients 17 g (IQR: 10–45 g) versus 7 g (IQR: 4–16 g) (p = 0.018) and 7 g (IQR: 4–15 g) (p = 0.004) respectively (Fig. 4a). Analogously, the median gray zone masses, in ischemic patients, were *2-fold larger when comparing FWHM versus NSD [16 g (IQR: 10–26 g) vs. 6 g (IQR: 4–10 g); p \ 0.001) and modified FWHM [8 g (IQR: 6–13 g); p \ 0.001). A two to threefold proportion was observed in the non-ischemic group [FWHM vs. NSD, 10 g (IQR: 6–25 g) vs. 3 g (IQR: 3–6 g); p = 0.001 and modified FWHM, 6 g (IQR: 2–9 g); p = 0.037] (Fig. 4b). Finally, the observation that the total LGE mass was smallest in the modified FWHM compared to NSD and FWHM applied equally in the ischemic [26 g (IQR: 16–34 g) vs. 41 g (IQR: 23–66 g); p \ 0.001 and 38 g (IQR: 19–47 g); p = 0.012 respectively], and non-ischemic populations [14 g (IQR: 6–24 g) vs. 21 g (IQR: 12–51 g); p \ 0.001 and 17 g (IQR: 9–40 g); p = 0.073 respectively] (Fig. 4c). No statistically significant differences were found comparing the results in ischemic versus non-ischemic patients. A summary of the myocardial
Variables
Method 1a
Method 2b
Method 3c
p values (Dunn’s test)
p-values (Friedman test)
Total late gadolinium Enhancement (g)
34
39
25
M1/M2 \ 0.001
\0.001
(17–46)
(20–63)
(12–33)
M2/M3 \ 0.001
Scar core (g)
13
30
13
M1/M3 = 0.010 M1/M2 \ 0.001
(7–23)
(17–53)
(7–22)
M2/M3 \ 0.001
15
5
8
M1/M2 \ 0.001
(8–26)
(3–9)
(6–12)
M2/M3 = 0.053
51
17
38
M1/M2 \ 0.001
(42–61)
(11–21)
(33–43)
M2/M3 \ 0.001
M1/M3 [ 0.999
Values are median (IQR) g Gram, M method
Gray zone (g)
a
Method 1 FWHM by Schmidt et al. [9]
b
Method 2 NSD by Yan et al. [8]
c
Method 3 modified-FWHM by Roes et al. [10]
\0.001
\0.001
M1/M3 \ 0.001 Gray zone extent (%)
\0.001
M1/M3 = 0.009
123
Int J Cardiovasc Imaging
A
Median Gray Zone Mass
B
30
p < 0.001
Mass ((g)
20
10
C
NSD
Median Total LGE mass
FWHM
D
40 20
FWHM
NSD
P < 0.001
40
0
m-FWHM
FWHM
B
Scar Core Mass - Ischemic vs Non-ischemic
30
p = 0.12
40 p = 0.07
p = 0.08
20
0
Mass (g g)
Mass (g g)
Median GZ extent
20
0
NSD
m-FWHM
Gray Zone mass Ischemic vs. Non-ischemic p = 0.39
20 p = 0.11 p = 0.14
10
0 FWHM
80
m-FWHM
60
%
Mass ((g)
60
C
NSD
80
p < 0.001
60
20
m-FWHM
80
A
40
0
0 FWHM
Fig. 4 Sub-group analysis: ischemic versus Non-ischemic patients. a, b, c Show size of scar (grams) according to scar characteristics (median and IQR) for all patients in the study. Black Bars Ischemic patients, Gray Bars Nonischemic patients. m-FWHM modified-FWHM
Median Scar Core Mass 60
p < 0.001
Mass ((g)
Fig. 3 Scar characterization and quantification by the different protocols. a, b, c Show size of scar (grams) according to scar characteristics (median and IQR) for all patients in the study. d Shows gray zone extent. Error bars represent IQR. FWHM, Full-width half maximum [11], m-FWHM (modified-FWHM) [12], NSD n-standard deviation [10]
NSD
m-FWHM
FWHM
NSD
m-FWHM
Total LGE Mass Ischemic vs Non-ischemic p = 0.14
Mass (g)
60
p = 0.21 p = 0.09 0 09
40 20 0
FWHM
NSD
distribution of LGE among the non-ischemic patients is provided in Online Resource 1 (Supplementary Table 2). Among ischemic patients, inter-observer variability assessed by Bland–Altman plots and ICC for gray zone and
123
m-FWHM
total LGE mass by the three methods is shown in Fig. 5. Gray zone mass assessed by modified FWHM and NSD had the least bias (-4.16 ± 4.33 and -4.14 ± 7.93 g respectively). Total LGE mass assessed by modified
Int J Cardiovasc Imaging y NSD Gray Zone mass by
Gray Zone mass by y FWHM 50
30
ICC = 0.63
y m-FWHM Gray Zone mass by
ICC = 0.57
-50
20
Differenc ce (g)
Differenc ce (g)
Differenc ce (g)
0
10 0 -10 -20
-100 50
100
150
0
Average Gray Zone Size (g)
ICC = 0.70
20
30
40
-20 -40
150
10
20
30
40
Total LGE mass by m-FWHM 40
ICC = 0.92
20 0 -20 -40
-80
Average Total LGE Size (g)
0
Average Gray Zone Size (g)
-60 100
-20
50
Diffe erence (g)
Diffe erence (g)
Differenc ce (g)
10
40
0
50
-10
Total LGE mass by NSD
20
0
0
Average Gray Zone Size (g)
Total LGE mass by FWHM 40
10
-30 30
-30
0
ICC = 0.63
30
20
0 62 ICC = 0.62
20 0 -20
-40 -60
0
50
100
150
Average Total LGE Size (g)
200
0
20
40
60
80
Average Total LGE Size (g)
Fig. 5 Inter-observer variability—Bland–Altman plots and ICC. Top row, gray zone mass by the three methods. Bottom row, total LGE mass by the three methods. FWHM full-width half maximum [11], modified-FWHM [12], NSD n-standard deviation [9], m-FWHM modified FWHM
FWHM had the least bias (-5.01 ± 13.57 g). Scar core mass measurement by FWHM and modified FWHM had the least bias (data not shown).
Discussion The findings of this study are that (1) important quantitative differences exist among the three currently applied algorithms; (2) gray zone size is largest when using the maximum intensity based FWHM method; (3) scar core size is largest when applying the n-standard deviation based NSD method; (4) modified-FWHM results in the smallest amount of total LGE; and (5) these findings appear to apply equally to ischemic and non-ischemic patients. CMR is a non-invasive tool to assess cardiac LV volumes [23], myocardial fibrosis and the extent of MI [1]. CMR is able to determine necrotic myocardium at all stages of infarct healing [1, 11], and thus differentiate between areas with viable and non-viable myocardium. Recently, it has been demonstrated that the histological observation that myocardial fibrosis interspersed with surviving myocardium creates a proarrhythmic milieu [3], which could be potentially translated into the field of cardiac imaging as MRI-defined gray zone [11]. First, the amount of gray zone has been investigated as an independent risk factor for cardiovascular mortality and
arrhythmia. Yan et al. [12] assessed the gray zone in a study of ischemic patients with abnormal myocardial LGE consistent with MI. During the 2.4 years of follow-up, there was an increased risk of all-cause mortality in patients with higher gray zone burdens (28 vs. 13 %). Schmidt et al. [13] subsequently showed that the size of the gray zone is an independent predictor of inducibility of monomorphic VT in post-MI patients. Another study by Roes et al. [14] in ischemic cardiomyopathy patients, demonstrated that the gray zone was the strongest predictor of appropriate ICD therapy during a median follow-up of 8.5 months. Second, MRI-defined gray zone is being evaluated as the possible imaging equivalent of VT-related conducting channels for purposes of therapeutic anatomic MRI-guided VT ablation. A recent study in a swine infarct model demonstrated that successful VT ablation included at least one ablation lesion in the gray zone in all animals, and that residual inducibility correlated with the persistence of the gray zone [19]. Andreu et al. [20] found that CMR techniques for gray zone quantification using 60 % of maximal SI threshold for both scar core and GZ best matched voltage-defined scar areas and identified 69 % of conducting channels identified on electroanatomic maps. It has also been observed that gray zone channels are more frequently observed in patients with clinical sustained monomorphic VT than in control patients, and that
123
Int J Cardiovasc Imaging
these channels were mainly located in the endocardium [21]. It is important to note that different definitions of gray zone and scar core, based on SI thresholds, have been proposed and were used in these earlier studies. Yan et al. used the more traditional definition based on the n-standard deviation values derived from normal myocardium, while the FWHM algorithm used by Schmidt et al. is based on the maximal SI of the hyperenhanced region to define thresholds for scar core and gray zone. A modified version of the FWHM as described by Roes et al. [14] also relies on the maximal SI of the hyperenhanced region to define these regions of the scar, but effectively increases the lower limit of the gray zone scale, thereby reducing the gray zone size. Perez-David et al. and Esther et al. utilized the NSD method while the study by Andreu et al. incorporated the FWHM to determine the scar core and gray zone but used three different cutoff values to define the best match between MRI scar and voltage (50, 60 and 70 % maximal SI) [19–21]. Our study demonstrates that these different algorithms cannot be used interchangeably and they significantly affect gray zone and scar core sizes and extent. This may have a significant impact on the application of gray zone for risk stratification or possible therapeutic ablation guidance. The median gray zone size was about two to threefold larger using the FWHM compared to the other two algorithms, and the NSD method resulted in about a 2.3-fold larger median scar core size. Other studies have reported differences in gray zone algorithms. In 33 patients with hypertrophic cardiomyopathy, S´piewak et al [24]. found a significant difference in gray zone mass in patients with similar LGE size while using the NSD and FWHM methods (21.2 and 43.5 g respectively). However, the modified-FWHM was not assessed and the patients had preserved LV function. De Haan et al. [22] found in 55 patients with ischemic cardiomyopathy that there was a significant difference in gray zone mass amongst the modified-FWHM, FWHM, and NSD methods (6.0 ± 4.2, 9.3 ± 8.0 and 1.8 ± 1.5 g respectively). The NSD method had the largest scar core size when compared to the FWHM and modified-FWHM methods. These findings are similar to our findings in this study. Interestingly, quantification of the gray zone did not improve the prediction of appropriate ICD shocks over the total LGE size nor was any of the three gray zone algorithms superior in predicting future ICD therapy. Our study suggests that the MRI concept of gray zone and scar core definition can be applied equally to the ischemic and non-ischemic cardiomyopathy population. Previous studies all investigated its use in ischemic patients [12–14] or hypertrophic cardiomyopathy patients with preserved LV function [24]. While our sample size is
123
small, our analysis shows similar proportions for gray zone and scar core among the different algorithms for both ischemic and non-ischemic patients. Several technical considerations have to be mentioned that can impact measurements of the gray zone. First, the selection of scar area and remote myocardium on a chosen reference slice to define the standard deviation and SI limits is a semi-automated process in which the observer determines the ideal slice. The computed sizes could be different if a different equally suited slice is chosen as the reference slice. Secondly, precise contouring of the endocardium can be challenging with the possible inclusion of blood pool voxels, which changes exact infarct boundaries [25] and the reference ranges for SI. Importantly, the percentage of gray zone SI voxels is resolution dependent (volume-averaging effect) and doubled when changing the isotropic voxel size from 51 9 51 9 50 to 408 9 408 9 1600 lm. Additionally, respiratory artifacts or motion can lead to spatial misalignment of slices [26] increasing voxels with intermediate gray zone intensities. Finally, Schuleri et al. [27] showed in a swine model that the gray zone decreased by 56 % from day 10 to day 90. The NSD method has been previously criticized for a possible overestimation of the infarct size [11, 15], which may be partly due to suboptimal signal suppression of remote myocardium or image artifacts [28].
Limitations First, the sample size in this study is small and heterogeneous. However, it is within the range of other previous gray zone studies [13] and the results are consistent across the total population and ischemic and non-ischemic subgroups. Second, the radiologist performed minor adjustments in MR imaging parameters to achieve the highest diagnostic yield from each CMR. While some differences in the quantitative results may occur as a result of these variations in protocol, all the parameters used in this study represent frequently applied settings for fibrosis imaging in clinical practice. The inherent technical limitations of CMR imaging explained above apply similarly to this patient cohort. Standard clinical precautions were used to attempt scanning in the same expiratory state and none of the patients had a recent MI in the preceding 90 days. Third, there are no clinical outcome data such as mortality, incidence of ventricular arrhythmias or inducibility of VT available for these patients as imaging was done for various indications. These would have been useful to evaluate the clinical significance of the observed differences. Lastly, the gray zone is an imaging-based concept to describe the size of the transition zone from the healthy myocardium to mostly homogenous scar, which differs with each of the
Int J Cardiovasc Imaging
different histogram-based definitions and has not been correlated to a specific histological area due to the lack of a well-defined pathological gray zone correlate.
Conclusion In conclusion, the scar core, gray zone, and total LGE quantified sizes change significantly depending on which of the currently employed algorithms is used. This strongly suggests that these currently employed algorithms cannot be used interchangeably and results obtained in one method may not be easily replicated with other methods Given the ongoing assessment of this promising technology for diagnostic and therapeutic electrophysiological decision making, a careful approach with further histological confirmation of these algorithms and correlation with clinical outcomes seems justified and clinically warranted. Acknowledgments The authors would like to thank Medis Medical Imaging Systems for the strong research support and Martine EtienneMesubi, DrPH, for her assistance with the statistical analysis. We sincerely appreciate the administrative and research assistance of Travis L. Mann, MPH, BSRT, and Robert Altom. Conflict of interest
None.
References 1. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J et al (1999) Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 100:1992–2002 2. Schelbert EB, Hsu LY, Anderson SA, Mohanty BD, Karim SM, Kellman P, Aletras AH et al (2010) Late gadolinium-enhancement cardiac magnetic resonance identifies postinfarction myocardial fibrosis and the border zone at the near cellular level in ex vivo rat heart. Circ Cardiovasc Imaging 3:743–752 3. de Bakker JM, van Capelle FJ, Janse MJ, Wilde AA, Coronel R, Becker AE, Dingemans KP et al (1988) Reentry as a cause of ventricular tachycardia in patients with chronic ischemic heart disease: electrophysiologic and anatomic correlation. Circulation 77:589–606 4. Stevenson WG, Khan H, Sager P, Saxon LA, Middlekauff HR, Natterson PD, Wiener I et al (1993) Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction. Circulation 88:1647–1670 5. Bello D, Fieno DS, Kim RJ, Pereles FS, Passman R, Song G, Kadish AH et al (2005) Infarct morphology identifies patients with substrate for sustained ventricular tachycardia. J Am Coll Cardiol 45:1104–1108 6. Kwong RY, Chan AK, Brown KA, Chan CW, Reynolds HG, Tsang S, Davis RB (2006) Impact of unrecognized myocardial scar detected by cardiac magnetic resonance imaging on eventfree survival in patients presenting with signs or symptoms of coronary artery disease. Circulation 113:2733–2743
7. Assomull RG, Prasad SK, Lyne J, Smith G, Burman ED, Khan M, Sheppard MN et al (2006) Cardiovascular magnetic resonance, fibrosis, and prognosis in dilated cardiomyopathy. J Am Coll Cardiol 48:1977–1985 8. Hundley WG, Bluemke D, Bogaert JG, Friedrich MG, Higgins CB, Lawson MA, McConnell MV et al (2009) Society for Cardiovascular Magnetic Resonance guidelines for reporting cardiovascular magnetic resonance examinations. J Cardiovasc Magn Reson 11:5. doi:10.1186/1532-429X-11-5 9. Kim HW, Farzaneh-Far A, Kim RJ (2010) Cardiovascular Magnetic Resonance in patients with myocardial infarction: current and emerging applications. JACC 55:1–16 10. Heiberg E, Ugander M, Englom H, Gotberg M, Olivercrona GK, Erlinger D, Arheden H (2008) Automated quantification of myocardial infarction from MR images by accounting for partial volume effects: animal, phantom, and human study. Radiology 246:581–588 11. Amado LC, Gerber BL, Gupta SN, Rettmann DW, Szarf G, Schock R, Nasir K et al (2004) Accurate and objective infarct sizing by contrast-enhanced magnetic resonance imaging in a canine myocardial infarction model. J Am Coll Cardiol 44:2383–2389 12. Yan AT, Shayne AJ, Brown KA, Gupta SN, Chan CW, Luu TM, Di Carli MF et al (2006) Characterization of the peri-infarct zone by contrast-enhanced cardiac magnetic resonance imaging is a powerful predictor of post-myocardial infarction mortality. Circulation 114:32–39 13. Schmidt A, Azevedo CF, Cheng A, Gupta SN, Bluemke DA, Foo TK, Gerstenblith G et al (2007) Infarct tissue heterogeneity by magnetic resonance imaging identifies enhanced cardiac arrhythmia susceptibility in patients with left ventricular dysfunction. Circulation 115:2006–2014 14. Roes SD, Borleffs CJ, van der Geest RJ, Westenberg JJ, Marsan NA, Kaandorp TA, Reiber JH et al (2009) Infarct tissue heterogeneity assessed with contrast-enhanced MRI predicts spontaneous ventricular arrhythmia in patients with ischemic cardiomyopathy and implantable cardioverter-defibrillator. Circ Cardiovasc Imaging 2:183–190 15. Bondarenko O, Beek AM, Hofman MB, Kuhl HP, Twisk JW, van Dockum WG, Visser CA et al (2005) Standardizing the definition of hyperenhancement in the quantitative assessment of infarct size and myocardial viability using delayed contrast-enhanced CMR. J Cardiovasc Magn Reson 7:481–485 16. Hsu LY, Natanzon A, Kellman P, Hirsch GA, Aletras AH, Arai AE (2006) Quantitative myocardial infarction on delayed enhancement MRI. Part I: animal validation of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging 23:298–308 17. Hsu LY, Ingkanisorn WP, Kellman P, Aletras AH, Arai AE (2006) Quantitative myocardial infarction on delayed enhancement MRI. Part II: clinical application of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging 23:309–314 18. Schulz-Menger J, Bluemke DA, Bremerich J, Flamm SD, Fogel MA, Friedrich MG, Kim RJ et al (2013) Standardized image interpretation and post processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance (SCMR) board of trustees task force on standardized post processing. J Cardiovasc Magn Reson 15:35. doi:10.1186/1532429X-15-35 19. Estner HL, Zviman MM, Herzka D, Miller F, Castro V, Nazarian S, Ashikaga H et al (2011) The critical isthmus sites of ischemic ventricular tachycardia are in zones of tissue heterogeneity, visualized by magnetic resonance imaging. Heart Rhythm 8:1942–1949 20. Andreu D, Berruezo A, Ortiz-Pe´rez JT, Silva E, Mont L, Borras R, de Caralt TM et al (2011) Integration of 3D electroanatomic
123
Int J Cardiovasc Imaging
21.
22.
23.
24.
maps and magnetic resonance scar characterization into the navigation system to guide ventricular tachycardia ablation. Circ Arrhythm Electrophysiol 4:674–683 Perez-David E, Arenal A, Rubio-Guivernau JL, del Castillo R, Atea L, Arbelo E, Caballero E et al (2011) Noninvasive identification of ventricular tachycardia-related conducting channels using contrast-enhanced magnetic resonance imaging in patients with chronic myocardial infarction: comparison of signal intensity scar mapping and endocardial voltage mapping. J Am Coll Cardiol 57:184–194 de Haan S, Meijers TA, Knaapen P, Beek AM, van Rossum AC, Allaart CP (2011) Scar size and characteristics assessed by CMR predict ventricular arrhythmias in ischaemic cardiomyopathy: comparison of previously validated models. Heart 97:1951–1956 Pujadas S, Reddy GP, Weber O, Lee JJ, Higgins CB (2004) MR imaging assessment of cardiac function. J Magn Reson Imaging 19:789–799 S´piewak M, Małek Ł, Chojnowska L, Misko J, Petryka J, Klopotowski M, Milosz B et al (2010) Late gadolinium enhancement gray zone in patients with hypertrophic cardiomyopathy. Comparison of different gray zone definitions. Int J Cardiovasc Imaging 26:693–699
123
25. Witschey WR, Zsido GA, Koomalsingh K, Kondo N, Minakawa M, Shuto T, McGarvey JR et al (2012) In vivo chronic myocardial infarction characterization by spin locked cardiovascular magnetic resonance. J Cardiovasc Magn Reson 15:14–37 26. Makowski MR, Wiethoff AJ, Jansen CH, Uribe S, Parish V, Schuster A, Botnar RM et al (2012) Single breath-hold assessment of cardiac function using an accelerated 3D single breathhold acquisition technique—comparison of an intravascular and extravascular contrast agent. J Cardiovasc Magn Reson 14:53. doi:10.1186/1532-429X-14-53 27. Schuleri KH, Centola M, Evers KS, Zviman A, Evers R, Lima JA, Lardo AC (2012) Cardiovascular magnetic resonance characterization of peri-infarct zone remodeling following myocardial infarction. J Cardiovasc Magn Reson 14:24. doi:10.1186/1532429X-14-24 28. Scott PA, Morgan JM, Carroll N, Murday DC, Roberts PR, Peebles CR, Harden SP et al (2011) The extent of left ventricular scar quantified by late gadolinium enhancement MRI is associated with spontaneous ventricular arrhythmias in patients with coronary artery disease and implantable cardioverter-defibrillators. Circ Arrhythm Electrophysiol 4:324–330
ORIGINAL PAPER
Differences in quantitative assessment of myocardial scar and gray zone by LGE-CMR imaging using established gray zone protocols Olurotimi Mesubi • Kelechi Ego-Osuala • Jean Jeudy • James Purtilo Stephen Synowski • Ameer Abutaleb • Michelle Niekoop • Mohammed Abdulghani • Ramazan Asoglu • Vincent See • Anastasios Saliaris • Stephen Shorofsky • Timm Dickfeld
•
Received: 2 July 2014 / Accepted: 15 October 2014 Ó Springer Science+Business Media Dordrecht 2014
Abstract Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) imaging is the gold standard for myocardial scar evaluation. Heterogeneous areas of scar (‘gray zone’), may serve as arrhythmogenic substrate. Various gray zone protocols have been correlated to clinical outcomes and ventricular tachycardia channels. This study assessed the quantitative differences in gray zone and scar core sizes as defined by previously validated signal intensity (SI) threshold algorithms. High quality LGECMR images performed in 41 cardiomyopathy patients
Electronic supplementary material The online version of this article (doi:10.1007/s10554-014-0555-0) contains supplementary material, which is available to authorized users. O. Mesubi K. Ego-Osuala J. Jeudy A. Abutaleb M. Niekoop M. Abdulghani R. Asoglu V. See A. Saliaris S. Shorofsky T. Dickfeld Maryland Arrhythmia and Cardiology Imaging Group (MACIG), University of Maryland, Baltimore, MD, USA O. Mesubi K. Ego-Osuala S. Synowski M. Abdulghani R. Asoglu V. See A. Saliaris S. Shorofsky T. Dickfeld (&) Division of Cardiology, University of Maryland School of Medicine, 22 S Greene St, Baltimore, MD 21201, USA e-mail: [email protected] O. Mesubi Division of Cardiovascular Medicine, University of Iowa, Iowa City, IA, USA J. Jeudy Division of Diagnostic Radiology and Nuclear Medicine, University of Maryland School of Medicine, Baltimore, MD, USA J. Purtilo Department of Computer Science, University of Maryland, College Park, MD, USA
[ischemic (33) or non-ischemic (8)] were analyzed using previously validated SI threshold methods [Full Width at Half Maximum (FWHM), n-standard deviation (NSD) and modified-FWHM]. Myocardial scar was defined as scar core and gray zone using SI thresholds based on these methods. Scar core, gray zone and total scar sizes were then computed and compared among these models. The median gray zone mass was 2–3 times larger with FWHM (15 g, IQR: 8–26 g) compared to NSD or modified-FWHM (5 g, IQR: 3–9 g; and 8 g. IQR: 6–12 g respectively, p \ 0.001). Conversely, infarct core mass was 2.3 times larger with NSD (30 g, IQR: 17–53 g) versus FWHM and modified-FWHM (13 g, IQR: 7–23 g, p \ 0.001). The gray zone extent (percentage of total scar that was gray zone) also varied significantly among the three methods, 51 % (IQR: 42–61 %), 17 % (IQR: 11–21 %) versus 38 % (IQR: 33–43 %) for FWHM, NSD and modified-FWHM respectively (p \ 0.001). Considerable variability exists among the current methods for MRI defined gray zone and scar core. Infarct core and total myocardial scar mass also differ using these methods. Further evaluation of the most accurate quantification method is needed. Keywords Cardiac MRI Gray zone Myocardial scar tissue characterization Abbreviations FWHM Full width at half maximum CMR Cardiac magnetic resonance LGE-CMR Late gadolinium enhancement cardiac magnetic resonance LV Left ventricle LVEDV Left ventricular end-diastolic volume LVESV Left ventricular end-systolic volume m-FWHM Modified full width at half maximum
123
Int J Cardiovasc Imaging
MRI MI NSD ROI SCMR SI VT
Magnetic resonance imaging Myocardial infarction N-standard deviation Region of interest Society for cardiovascular resonance Signal intensity Ventricular tachycardia
magnetic
Introduction Late gadolinium enhancement cardiac magnetic resonance (LGE-CMR) imaging is the current gold standard for non-invasive imaging and assessment of myocardial fibrosis [1, 2]. Thus, magnetic resonance imaging (MRI) has emerged as an essential tool in our armamentarium for substrate assessment in the management of ventricular tachycardia (VT). The current mechanistic conceptual framework of reentrant VT is based on the understanding that the myocardial substrate consists of areas of surviving myocardium interspersed in myocardial scar. These form slow conduction pathways needed for the propagation and maintenance of reentrant VT circuits [3, 4]. Prior studies have demonstrated a strong and consistent correlation between LGE-CMR identified myocardial fibrosis and the incidence and occurrence of ventricular arrhythmias in ischemic and non-ischemic cardiomyopathy patients [5–7]. The Society for Cardiovascular Magnetic Resonance (SCMR) recommends that scar definition on LGE-CMR imaging should be signal intensity (SI) [2 SD above the average of normal or remote myocardium, based on initial validation studies [8]. From the original binary approach, based on LGE-CMR, where ‘‘white’’ regions of hyperenhancement identified areas of myocardial fibrosis, and ‘‘black’’ areas without enhancement correlated with viable myocardium, a new concept of ‘‘gray zone’’ has emerged. The gray zone is characteristically thought of as myocardial tissue that consists of an admixture of fibrotic areas (‘‘white’’) with surviving myocardium (‘‘black’’) resulting in a zone of intermediate SI with a net ‘‘gray’’ intensity appearance. Partial volume effects due to less spatial resolution in vivo compared to ex vivo myocardial imaging and cardiac motion are also thought to contribute to the appearance of the gray zone [1, 2, 9, 10]. Several techniques have been developed, primarily as research tools, for LGE myocardial scar tissue characterization [11–17]. There is no consensus on the optimal technique for these purposes. A recent SCMR position statement [18] acknowledges the existence of these several methods, but
123
refrained from making specific recommendations given the paucity of evidence. Using three different gray zone definitions, several landmark studies have demonstrated that the amount of gray zone in ischemic cardiomyopathy patients is an independent predictor of post myocardial infarction (MI) all-cause mortality [12], inducibility of ventricular arrhythmias [13] and appropriate implantable cardioverter-defibrillator (ICD) therapy [14] (‘‘gray zone for sudden cardiac death risk stratification’’). Additionally, gray zone reconstruction is now being evaluated as a potential complementary tool for MRI-guided anatomic VT ablation in animal and human studies [19–21]. This study sought to assess the quantitative differences in gray zone and scar core sizes as defined by three previously validated SI threshold algorithms and thus help define their role for diagnostic and future therapeutic applications.
Methods Study population This is a retrospective study of forty-one consecutive patients who presented for late gadolinium enhanced cardiac MRI from 2009 to 2011 at the University of Maryland Medical Center in Baltimore, Maryland. The institutional review board approved the study. Patients with cardiac device therapies such as ICDs or pacemakers that would result in significant artifacts were excluded from the study. MRI protocol A 1.5-T MRI Scanner (Siemens Medical Systems; Avanto, Erlangen, Germany) with an eight-element phased-array receiver body coil placed anterior and/or posterior to the chest was used. Images were acquired using vector ECGgating. The heart was imaged to cover the whole left ventricle (LV) depending on LV size for the cine images. Typical parameters were as follows: slice thickness of 8 mm; slice gap of 1 mm; average in-plane spatial resolution between 1.4 9 1.4 mm; temporal resolution 46–50 ms; and field of view (FOV) of 340 9 340 mm2. Delayed enhancement images were acquired 10 min after intravenous administration of gadolinium-DTPA (diethylenetriaminepentacetate) (0.15–0.20 mmol/kg, MultiHance; Bracco SpA; Milan, Italy). Great care was taken to optimize the nulling of the inversion time (usually between 200 and 300 ms) to assess for LGE. Views per segment and trigger delay were adjusted according to the patients’ heart rate to minimize image blurring. Typical parameters were as follows: FOV 360 9 360 mm2; slice thickness of 8 mm; no gap; average inplane spatial resolution between 1.4 9 1.4 mm, flip angle 25°, echo time 3.4 ms and repetition time 4–8 ms.
Int J Cardiovasc Imaging Fig. 1 Short axis LGE-CMR images from ischemic patient with inferolateral infarct. Epicardial and endocardial borders are outlined. a Myocardium without contours. b Full width half maximum (FWHM) method. c n-standard deviation (NSD) method. d Modified-FWHM method. The area demarcated in red is the scar core and the area in yellow is the gray zone. Note scar core and gray zone variations in all three techniques
MRI image analysis All MRI data analysis was performed on Digital Imaging and Communications in Medicine images with specialized software (QMass MR 7.5, Medis Medical Imaging Systems BV, Leiden, The Netherlands). For cine analysis, the endocardial and epicardial contours were drawn manually at the left ventricular end-systolic (LVES) and left ventricular end-diastolic (LVED) phases of the short axis cine images (Fig. 1). The papillary muscles were regarded as part of the ventricular cavity and the epicardial fat was excluded. The LVED and LVES volumes and masses were automatically calculated by the software. The software also automatically calculated the left ventricular ejection fraction (LVEF), stroke volume and LV mass (Fig. 2). Contrast enhanced MRI images were analyzed to determine infarct core, gray zone, and total infarct sizes. LV endocardial and epicardial contours were manually outlined on short axis contrast enhanced images. The papillary muscles were incorporated into the LV cavity. Subsequently, a CMR image slice amongst the series, determined by a trained observer to have a well-defined area of hyperenhancement with the most SI within the epicardial and endocardial borders, was selected and a Delayed Signal Intensity (DSI) analysis was performed. The DSI analysis automatically selected the gray zone and scar core area within the epicardial and endocardial borders, and also determined a remote region of interest (ROI) in the normal myocardium without artifact or
hyperenhancement [11, 12]. The QMass software automatically determined the scar core, gray zone and normal myocardium according to the three previously published algorithms [12–14]. In brief, the full-width at half maximum method (FWHM) [11, 13] assigns 100 % to the maximal SI value within the endocardial and epicardial contours. Any voxel that contains SI values between 50 and 100 % is defined as scar core [11, 13]. The gray zone is defined as voxels with SI higher than the peak SI of the remote ROI but \50 % of the maximal SI [13, 22]. The n-standard deviation method (NSD) determines the mean and standard deviation (SD) of the SI of a ROI in normal myocardium without artifact or hyperenhancement Infarct core was defined as voxels with SI [3 SD above the mean SI of the ROI while the gray zone was defined as areas with 2–3 SD above the mean SI of the ROI in the normal myocardial [12, 22]. Finally, the modified-FWHM defined the scar core in the same manner as the traditional FWHM, but defined the gray zone as areas with SI [35 % of the maximal SI but \50 % of the maximal SI [14, 22]. Statistical analysis Data is summarized as mean ± SD for continuous variables that are normally distributed and as median with interquartile range (IQR) for variables without normal distribution. Comparison across groups with repeated measures was assessed with the Friedman test [a nonparametric equivalent of the one-way Analysis of Variance
123
Int J Cardiovasc Imaging Fig. 2 Short axis LGE-CMR images from non-ischemic patient with an anteroseptal scar. Epicardial and endocardial borders are outlined. a Myocardium without contours. b Full width half maximum (FWHM) method. c n-standard deviation (NSD) method. d Modified-FWHM method. Red indicates scar core and yellow indicates the gray zone. Note scar core and gray zone variations in all 3 techniques
(ANOVA) test]. Post-hoc pairwise comparisons between groups were made with Dunn’s test of multiple comparisons using rank sums based on the Wilcoxon Signed-Rank test. Where appropriate, comparison between groups (without multiple comparisons) was done using the Wilcoxon Signed-Rank test. All tests were two-sided, and p values \0.05 were considered statistically significant. Correlations between gray zone mass and total myocardial scar were tested using Spearman’s correlation coefficient. Bland–Altman analysis was used to determine interobserver variability and interclass correlation coefficient (ICC) with 95 % confidence interval (CI) was computed. This was performed on a subset of 14 ischemic patients who had LGE analyzed by two independent readers (OM and KE). All analyses was done using STATA version 12.1 software package (StataCorp LP, College Station, Texas).
Results Study patients Table 1 summarizes the clinical characteristics and CMR findings of the study population. Indications for CMR were arrhythmias (20 %), recent MI (20 %), ischemic heart disease (20 %), myocardial scar evaluation (17 %), assessment of myocardial viability (10 %), cardiomyopathy (10 %), myocarditis (7 %), and ventricular aneurysm (2 %). Some patients had more than one indication for CMR. Of the 41 patients, there were 33 (80 %) ischemic
123
and 8 (20 %) non-ischemic patients. Online Resource 1 (Supplementary Table 1) summarizes the baseline characteristics of the groups (ischemic and non-ischemic). MRI characteristics All the patients demonstrated delayed cardiac enhancement on MRI. Study results are shown in Table 2. Significant differences between total scar, scar core, and gray zone were seen between the different gray zone algorithms (p \ 0.001 in all three categories). The median gray zone masses were 15 g (IQR: 8–26 g), 5 g (IQR: 3–9 g), and 8 g (IQR: 5–12 g) for FWHM, NSD, and modified-FWHM, respectively (Fig. 3a). The FWHM resulted in a two to threefold larger gray zone mass than the NSD (p \ 0.001) or the modified-FWHM method (p \ 0.001). There was no significant difference between NSD and modified-FWHM (p = 0.053). The observed median scar core masses were [2-fold larger with the NSD versus FWHM [30 g (IQR: 17–53 g) vs. 13 g (IQR: 7–23 g); p \ 0.001] or modified FWHM (13 g (IQR: 7–22 g; p \ 0.001) respectively (Fig. 3b). Given the identical definition of scar core in the FWHM and modified FWHM algorithms, no differences were seen between these algorithms. The median total LGE mass was smallest when using the modified FWHM [25 g (IQR: 12–32 g)] compared to NSD [39 g (IQR: 20–63 g); p \ 0.001] and was also smaller than the FWHM [34 g (IQR: 17–46 g); p \ 0.001] (Fig. 3c). The gray zone extent (percentage of LGE identified as gray zone) differed between the three algorithms (Fig. 3d)
Int J Cardiovasc Imaging Table 1 Baseline characteristics and CMR findings Variables
All patients (n = 41)
Age (years)
53 ± 14
Male gender n (%)
36 (88)
LVEF (%)
38 ± 13
Etiology of cardiomyopathy Ischemic n (%)
33 (80)
Non-ischemic n (%)
8 (20)
Co-morbidities Hypertension n (%)
20 (49)
Diabetes n (%)
11 (27)
Hypercholesterolemia, n (%)
15 (37)
Arrhythmia n (%) Scar location
11 (27)
Septal n (%)
20 (27)
Inferior n (%)
21 (28)
Inferolateral n (%)
14 (19)
Anterior n (%)
10 (14)
Anterolateral n (%)
9 (12)
CMR measurements LV Mass (g)
124 ± 37
LVEF (%)
41 ± 16
LVEDV (ml/m2)
212 ± 63
LVESV (ml/m2)
132 ± 72
LVSV (ml/m2)
80 ± 28
Values are n (%) or mean ± SD Some patients had multiple co-morbidities and LGE locations based on the AHA Bull’s-eye plot nomenclature LV left ventricle, LVEDV left ventricular end-diastolic volume, LVEF left ventricular ejection fraction, LVESV left ventricular end-systolic volume, LVSV left ventricular stroke volume
and was highest using the FWHM [51 % (IQR: 42–61 %)] compared with the modified FWHM [38 % (IQR: 33–43 %); p = 0.009], and lowest using the NSD method Table 2 Comparison of MRI defined myocardial scar variables
[17 % (IQR: 11–21 %)] compared to FWHM (p \ 0.001) and modified FWHM (p \ 0.001). There was a strong correlation between gray zone mass and total myocardial scar, irrespective of the SI threshold method (r = 0.95, p \ 0.001 for FWHM, r = 0.89, p \ 0.001 for NSD, and r = 0.97, p \ 0.001 for modified FWHM). Sub-group analysis (Fig. 4) comparing gray zone, scar core and total LGE masses in ischemic and non-ischemic patients demonstrated similar findings in both patient groups. The differences in gray zone and total scar masses among the three protocols were consistent in both the ischemic and non-ischemic subjects. The median scar core mass was [2 fold larger in the NSD versus FWHM and modified FWHM [33 g (IQR: 20–54 g) versus 13 g (IQR: 10–25 g); p \ 0.001 and 15 g (IQR: 9–23 g); p \ 0.001] respectively in the ischemic patients and in the nonischemic patients 17 g (IQR: 10–45 g) versus 7 g (IQR: 4–16 g) (p = 0.018) and 7 g (IQR: 4–15 g) (p = 0.004) respectively (Fig. 4a). Analogously, the median gray zone masses, in ischemic patients, were *2-fold larger when comparing FWHM versus NSD [16 g (IQR: 10–26 g) vs. 6 g (IQR: 4–10 g); p \ 0.001) and modified FWHM [8 g (IQR: 6–13 g); p \ 0.001). A two to threefold proportion was observed in the non-ischemic group [FWHM vs. NSD, 10 g (IQR: 6–25 g) vs. 3 g (IQR: 3–6 g); p = 0.001 and modified FWHM, 6 g (IQR: 2–9 g); p = 0.037] (Fig. 4b). Finally, the observation that the total LGE mass was smallest in the modified FWHM compared to NSD and FWHM applied equally in the ischemic [26 g (IQR: 16–34 g) vs. 41 g (IQR: 23–66 g); p \ 0.001 and 38 g (IQR: 19–47 g); p = 0.012 respectively], and non-ischemic populations [14 g (IQR: 6–24 g) vs. 21 g (IQR: 12–51 g); p \ 0.001 and 17 g (IQR: 9–40 g); p = 0.073 respectively] (Fig. 4c). No statistically significant differences were found comparing the results in ischemic versus non-ischemic patients. A summary of the myocardial
Variables
Method 1a
Method 2b
Method 3c
p values (Dunn’s test)
p-values (Friedman test)
Total late gadolinium Enhancement (g)
34
39
25
M1/M2 \ 0.001
\0.001
(17–46)
(20–63)
(12–33)
M2/M3 \ 0.001
Scar core (g)
13
30
13
M1/M3 = 0.010 M1/M2 \ 0.001
(7–23)
(17–53)
(7–22)
M2/M3 \ 0.001
15
5
8
M1/M2 \ 0.001
(8–26)
(3–9)
(6–12)
M2/M3 = 0.053
51
17
38
M1/M2 \ 0.001
(42–61)
(11–21)
(33–43)
M2/M3 \ 0.001
M1/M3 [ 0.999
Values are median (IQR) g Gram, M method
Gray zone (g)
a
Method 1 FWHM by Schmidt et al. [9]
b
Method 2 NSD by Yan et al. [8]
c
Method 3 modified-FWHM by Roes et al. [10]
\0.001
\0.001
M1/M3 \ 0.001 Gray zone extent (%)
\0.001
M1/M3 = 0.009
123
Int J Cardiovasc Imaging
A
Median Gray Zone Mass
B
30
p < 0.001
Mass ((g)
20
10
C
NSD
Median Total LGE mass
FWHM
D
40 20
FWHM
NSD
P < 0.001
40
0
m-FWHM
FWHM
B
Scar Core Mass - Ischemic vs Non-ischemic
30
p = 0.12
40 p = 0.07
p = 0.08
20
0
Mass (g g)
Mass (g g)
Median GZ extent
20
0
NSD
m-FWHM
Gray Zone mass Ischemic vs. Non-ischemic p = 0.39
20 p = 0.11 p = 0.14
10
0 FWHM
80
m-FWHM
60
%
Mass ((g)
60
C
NSD
80
p < 0.001
60
20
m-FWHM
80
A
40
0
0 FWHM
Fig. 4 Sub-group analysis: ischemic versus Non-ischemic patients. a, b, c Show size of scar (grams) according to scar characteristics (median and IQR) for all patients in the study. Black Bars Ischemic patients, Gray Bars Nonischemic patients. m-FWHM modified-FWHM
Median Scar Core Mass 60
p < 0.001
Mass ((g)
Fig. 3 Scar characterization and quantification by the different protocols. a, b, c Show size of scar (grams) according to scar characteristics (median and IQR) for all patients in the study. d Shows gray zone extent. Error bars represent IQR. FWHM, Full-width half maximum [11], m-FWHM (modified-FWHM) [12], NSD n-standard deviation [10]
NSD
m-FWHM
FWHM
NSD
m-FWHM
Total LGE Mass Ischemic vs Non-ischemic p = 0.14
Mass (g)
60
p = 0.21 p = 0.09 0 09
40 20 0
FWHM
NSD
distribution of LGE among the non-ischemic patients is provided in Online Resource 1 (Supplementary Table 2). Among ischemic patients, inter-observer variability assessed by Bland–Altman plots and ICC for gray zone and
123
m-FWHM
total LGE mass by the three methods is shown in Fig. 5. Gray zone mass assessed by modified FWHM and NSD had the least bias (-4.16 ± 4.33 and -4.14 ± 7.93 g respectively). Total LGE mass assessed by modified
Int J Cardiovasc Imaging y NSD Gray Zone mass by
Gray Zone mass by y FWHM 50
30
ICC = 0.63
y m-FWHM Gray Zone mass by
ICC = 0.57
-50
20
Differenc ce (g)
Differenc ce (g)
Differenc ce (g)
0
10 0 -10 -20
-100 50
100
150
0
Average Gray Zone Size (g)
ICC = 0.70
20
30
40
-20 -40
150
10
20
30
40
Total LGE mass by m-FWHM 40
ICC = 0.92
20 0 -20 -40
-80
Average Total LGE Size (g)
0
Average Gray Zone Size (g)
-60 100
-20
50
Diffe erence (g)
Diffe erence (g)
Differenc ce (g)
10
40
0
50
-10
Total LGE mass by NSD
20
0
0
Average Gray Zone Size (g)
Total LGE mass by FWHM 40
10
-30 30
-30
0
ICC = 0.63
30
20
0 62 ICC = 0.62
20 0 -20
-40 -60
0
50
100
150
Average Total LGE Size (g)
200
0
20
40
60
80
Average Total LGE Size (g)
Fig. 5 Inter-observer variability—Bland–Altman plots and ICC. Top row, gray zone mass by the three methods. Bottom row, total LGE mass by the three methods. FWHM full-width half maximum [11], modified-FWHM [12], NSD n-standard deviation [9], m-FWHM modified FWHM
FWHM had the least bias (-5.01 ± 13.57 g). Scar core mass measurement by FWHM and modified FWHM had the least bias (data not shown).
Discussion The findings of this study are that (1) important quantitative differences exist among the three currently applied algorithms; (2) gray zone size is largest when using the maximum intensity based FWHM method; (3) scar core size is largest when applying the n-standard deviation based NSD method; (4) modified-FWHM results in the smallest amount of total LGE; and (5) these findings appear to apply equally to ischemic and non-ischemic patients. CMR is a non-invasive tool to assess cardiac LV volumes [23], myocardial fibrosis and the extent of MI [1]. CMR is able to determine necrotic myocardium at all stages of infarct healing [1, 11], and thus differentiate between areas with viable and non-viable myocardium. Recently, it has been demonstrated that the histological observation that myocardial fibrosis interspersed with surviving myocardium creates a proarrhythmic milieu [3], which could be potentially translated into the field of cardiac imaging as MRI-defined gray zone [11]. First, the amount of gray zone has been investigated as an independent risk factor for cardiovascular mortality and
arrhythmia. Yan et al. [12] assessed the gray zone in a study of ischemic patients with abnormal myocardial LGE consistent with MI. During the 2.4 years of follow-up, there was an increased risk of all-cause mortality in patients with higher gray zone burdens (28 vs. 13 %). Schmidt et al. [13] subsequently showed that the size of the gray zone is an independent predictor of inducibility of monomorphic VT in post-MI patients. Another study by Roes et al. [14] in ischemic cardiomyopathy patients, demonstrated that the gray zone was the strongest predictor of appropriate ICD therapy during a median follow-up of 8.5 months. Second, MRI-defined gray zone is being evaluated as the possible imaging equivalent of VT-related conducting channels for purposes of therapeutic anatomic MRI-guided VT ablation. A recent study in a swine infarct model demonstrated that successful VT ablation included at least one ablation lesion in the gray zone in all animals, and that residual inducibility correlated with the persistence of the gray zone [19]. Andreu et al. [20] found that CMR techniques for gray zone quantification using 60 % of maximal SI threshold for both scar core and GZ best matched voltage-defined scar areas and identified 69 % of conducting channels identified on electroanatomic maps. It has also been observed that gray zone channels are more frequently observed in patients with clinical sustained monomorphic VT than in control patients, and that
123
Int J Cardiovasc Imaging
these channels were mainly located in the endocardium [21]. It is important to note that different definitions of gray zone and scar core, based on SI thresholds, have been proposed and were used in these earlier studies. Yan et al. used the more traditional definition based on the n-standard deviation values derived from normal myocardium, while the FWHM algorithm used by Schmidt et al. is based on the maximal SI of the hyperenhanced region to define thresholds for scar core and gray zone. A modified version of the FWHM as described by Roes et al. [14] also relies on the maximal SI of the hyperenhanced region to define these regions of the scar, but effectively increases the lower limit of the gray zone scale, thereby reducing the gray zone size. Perez-David et al. and Esther et al. utilized the NSD method while the study by Andreu et al. incorporated the FWHM to determine the scar core and gray zone but used three different cutoff values to define the best match between MRI scar and voltage (50, 60 and 70 % maximal SI) [19–21]. Our study demonstrates that these different algorithms cannot be used interchangeably and they significantly affect gray zone and scar core sizes and extent. This may have a significant impact on the application of gray zone for risk stratification or possible therapeutic ablation guidance. The median gray zone size was about two to threefold larger using the FWHM compared to the other two algorithms, and the NSD method resulted in about a 2.3-fold larger median scar core size. Other studies have reported differences in gray zone algorithms. In 33 patients with hypertrophic cardiomyopathy, S´piewak et al [24]. found a significant difference in gray zone mass in patients with similar LGE size while using the NSD and FWHM methods (21.2 and 43.5 g respectively). However, the modified-FWHM was not assessed and the patients had preserved LV function. De Haan et al. [22] found in 55 patients with ischemic cardiomyopathy that there was a significant difference in gray zone mass amongst the modified-FWHM, FWHM, and NSD methods (6.0 ± 4.2, 9.3 ± 8.0 and 1.8 ± 1.5 g respectively). The NSD method had the largest scar core size when compared to the FWHM and modified-FWHM methods. These findings are similar to our findings in this study. Interestingly, quantification of the gray zone did not improve the prediction of appropriate ICD shocks over the total LGE size nor was any of the three gray zone algorithms superior in predicting future ICD therapy. Our study suggests that the MRI concept of gray zone and scar core definition can be applied equally to the ischemic and non-ischemic cardiomyopathy population. Previous studies all investigated its use in ischemic patients [12–14] or hypertrophic cardiomyopathy patients with preserved LV function [24]. While our sample size is
123
small, our analysis shows similar proportions for gray zone and scar core among the different algorithms for both ischemic and non-ischemic patients. Several technical considerations have to be mentioned that can impact measurements of the gray zone. First, the selection of scar area and remote myocardium on a chosen reference slice to define the standard deviation and SI limits is a semi-automated process in which the observer determines the ideal slice. The computed sizes could be different if a different equally suited slice is chosen as the reference slice. Secondly, precise contouring of the endocardium can be challenging with the possible inclusion of blood pool voxels, which changes exact infarct boundaries [25] and the reference ranges for SI. Importantly, the percentage of gray zone SI voxels is resolution dependent (volume-averaging effect) and doubled when changing the isotropic voxel size from 51 9 51 9 50 to 408 9 408 9 1600 lm. Additionally, respiratory artifacts or motion can lead to spatial misalignment of slices [26] increasing voxels with intermediate gray zone intensities. Finally, Schuleri et al. [27] showed in a swine model that the gray zone decreased by 56 % from day 10 to day 90. The NSD method has been previously criticized for a possible overestimation of the infarct size [11, 15], which may be partly due to suboptimal signal suppression of remote myocardium or image artifacts [28].
Limitations First, the sample size in this study is small and heterogeneous. However, it is within the range of other previous gray zone studies [13] and the results are consistent across the total population and ischemic and non-ischemic subgroups. Second, the radiologist performed minor adjustments in MR imaging parameters to achieve the highest diagnostic yield from each CMR. While some differences in the quantitative results may occur as a result of these variations in protocol, all the parameters used in this study represent frequently applied settings for fibrosis imaging in clinical practice. The inherent technical limitations of CMR imaging explained above apply similarly to this patient cohort. Standard clinical precautions were used to attempt scanning in the same expiratory state and none of the patients had a recent MI in the preceding 90 days. Third, there are no clinical outcome data such as mortality, incidence of ventricular arrhythmias or inducibility of VT available for these patients as imaging was done for various indications. These would have been useful to evaluate the clinical significance of the observed differences. Lastly, the gray zone is an imaging-based concept to describe the size of the transition zone from the healthy myocardium to mostly homogenous scar, which differs with each of the
Int J Cardiovasc Imaging
different histogram-based definitions and has not been correlated to a specific histological area due to the lack of a well-defined pathological gray zone correlate.
Conclusion In conclusion, the scar core, gray zone, and total LGE quantified sizes change significantly depending on which of the currently employed algorithms is used. This strongly suggests that these currently employed algorithms cannot be used interchangeably and results obtained in one method may not be easily replicated with other methods Given the ongoing assessment of this promising technology for diagnostic and therapeutic electrophysiological decision making, a careful approach with further histological confirmation of these algorithms and correlation with clinical outcomes seems justified and clinically warranted. Acknowledgments The authors would like to thank Medis Medical Imaging Systems for the strong research support and Martine EtienneMesubi, DrPH, for her assistance with the statistical analysis. We sincerely appreciate the administrative and research assistance of Travis L. Mann, MPH, BSRT, and Robert Altom. Conflict of interest
None.
References 1. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J et al (1999) Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 100:1992–2002 2. Schelbert EB, Hsu LY, Anderson SA, Mohanty BD, Karim SM, Kellman P, Aletras AH et al (2010) Late gadolinium-enhancement cardiac magnetic resonance identifies postinfarction myocardial fibrosis and the border zone at the near cellular level in ex vivo rat heart. Circ Cardiovasc Imaging 3:743–752 3. de Bakker JM, van Capelle FJ, Janse MJ, Wilde AA, Coronel R, Becker AE, Dingemans KP et al (1988) Reentry as a cause of ventricular tachycardia in patients with chronic ischemic heart disease: electrophysiologic and anatomic correlation. Circulation 77:589–606 4. Stevenson WG, Khan H, Sager P, Saxon LA, Middlekauff HR, Natterson PD, Wiener I et al (1993) Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction. Circulation 88:1647–1670 5. Bello D, Fieno DS, Kim RJ, Pereles FS, Passman R, Song G, Kadish AH et al (2005) Infarct morphology identifies patients with substrate for sustained ventricular tachycardia. J Am Coll Cardiol 45:1104–1108 6. Kwong RY, Chan AK, Brown KA, Chan CW, Reynolds HG, Tsang S, Davis RB (2006) Impact of unrecognized myocardial scar detected by cardiac magnetic resonance imaging on eventfree survival in patients presenting with signs or symptoms of coronary artery disease. Circulation 113:2733–2743
7. Assomull RG, Prasad SK, Lyne J, Smith G, Burman ED, Khan M, Sheppard MN et al (2006) Cardiovascular magnetic resonance, fibrosis, and prognosis in dilated cardiomyopathy. J Am Coll Cardiol 48:1977–1985 8. Hundley WG, Bluemke D, Bogaert JG, Friedrich MG, Higgins CB, Lawson MA, McConnell MV et al (2009) Society for Cardiovascular Magnetic Resonance guidelines for reporting cardiovascular magnetic resonance examinations. J Cardiovasc Magn Reson 11:5. doi:10.1186/1532-429X-11-5 9. Kim HW, Farzaneh-Far A, Kim RJ (2010) Cardiovascular Magnetic Resonance in patients with myocardial infarction: current and emerging applications. JACC 55:1–16 10. Heiberg E, Ugander M, Englom H, Gotberg M, Olivercrona GK, Erlinger D, Arheden H (2008) Automated quantification of myocardial infarction from MR images by accounting for partial volume effects: animal, phantom, and human study. Radiology 246:581–588 11. Amado LC, Gerber BL, Gupta SN, Rettmann DW, Szarf G, Schock R, Nasir K et al (2004) Accurate and objective infarct sizing by contrast-enhanced magnetic resonance imaging in a canine myocardial infarction model. J Am Coll Cardiol 44:2383–2389 12. Yan AT, Shayne AJ, Brown KA, Gupta SN, Chan CW, Luu TM, Di Carli MF et al (2006) Characterization of the peri-infarct zone by contrast-enhanced cardiac magnetic resonance imaging is a powerful predictor of post-myocardial infarction mortality. Circulation 114:32–39 13. Schmidt A, Azevedo CF, Cheng A, Gupta SN, Bluemke DA, Foo TK, Gerstenblith G et al (2007) Infarct tissue heterogeneity by magnetic resonance imaging identifies enhanced cardiac arrhythmia susceptibility in patients with left ventricular dysfunction. Circulation 115:2006–2014 14. Roes SD, Borleffs CJ, van der Geest RJ, Westenberg JJ, Marsan NA, Kaandorp TA, Reiber JH et al (2009) Infarct tissue heterogeneity assessed with contrast-enhanced MRI predicts spontaneous ventricular arrhythmia in patients with ischemic cardiomyopathy and implantable cardioverter-defibrillator. Circ Cardiovasc Imaging 2:183–190 15. Bondarenko O, Beek AM, Hofman MB, Kuhl HP, Twisk JW, van Dockum WG, Visser CA et al (2005) Standardizing the definition of hyperenhancement in the quantitative assessment of infarct size and myocardial viability using delayed contrast-enhanced CMR. J Cardiovasc Magn Reson 7:481–485 16. Hsu LY, Natanzon A, Kellman P, Hirsch GA, Aletras AH, Arai AE (2006) Quantitative myocardial infarction on delayed enhancement MRI. Part I: animal validation of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging 23:298–308 17. Hsu LY, Ingkanisorn WP, Kellman P, Aletras AH, Arai AE (2006) Quantitative myocardial infarction on delayed enhancement MRI. Part II: clinical application of an automated feature analysis and combined thresholding infarct sizing algorithm. J Magn Reson Imaging 23:309–314 18. Schulz-Menger J, Bluemke DA, Bremerich J, Flamm SD, Fogel MA, Friedrich MG, Kim RJ et al (2013) Standardized image interpretation and post processing in cardiovascular magnetic resonance: Society for Cardiovascular Magnetic Resonance (SCMR) board of trustees task force on standardized post processing. J Cardiovasc Magn Reson 15:35. doi:10.1186/1532429X-15-35 19. Estner HL, Zviman MM, Herzka D, Miller F, Castro V, Nazarian S, Ashikaga H et al (2011) The critical isthmus sites of ischemic ventricular tachycardia are in zones of tissue heterogeneity, visualized by magnetic resonance imaging. Heart Rhythm 8:1942–1949 20. Andreu D, Berruezo A, Ortiz-Pe´rez JT, Silva E, Mont L, Borras R, de Caralt TM et al (2011) Integration of 3D electroanatomic
123
Int J Cardiovasc Imaging
21.
22.
23.
24.
maps and magnetic resonance scar characterization into the navigation system to guide ventricular tachycardia ablation. Circ Arrhythm Electrophysiol 4:674–683 Perez-David E, Arenal A, Rubio-Guivernau JL, del Castillo R, Atea L, Arbelo E, Caballero E et al (2011) Noninvasive identification of ventricular tachycardia-related conducting channels using contrast-enhanced magnetic resonance imaging in patients with chronic myocardial infarction: comparison of signal intensity scar mapping and endocardial voltage mapping. J Am Coll Cardiol 57:184–194 de Haan S, Meijers TA, Knaapen P, Beek AM, van Rossum AC, Allaart CP (2011) Scar size and characteristics assessed by CMR predict ventricular arrhythmias in ischaemic cardiomyopathy: comparison of previously validated models. Heart 97:1951–1956 Pujadas S, Reddy GP, Weber O, Lee JJ, Higgins CB (2004) MR imaging assessment of cardiac function. J Magn Reson Imaging 19:789–799 S´piewak M, Małek Ł, Chojnowska L, Misko J, Petryka J, Klopotowski M, Milosz B et al (2010) Late gadolinium enhancement gray zone in patients with hypertrophic cardiomyopathy. Comparison of different gray zone definitions. Int J Cardiovasc Imaging 26:693–699
123
25. Witschey WR, Zsido GA, Koomalsingh K, Kondo N, Minakawa M, Shuto T, McGarvey JR et al (2012) In vivo chronic myocardial infarction characterization by spin locked cardiovascular magnetic resonance. J Cardiovasc Magn Reson 15:14–37 26. Makowski MR, Wiethoff AJ, Jansen CH, Uribe S, Parish V, Schuster A, Botnar RM et al (2012) Single breath-hold assessment of cardiac function using an accelerated 3D single breathhold acquisition technique—comparison of an intravascular and extravascular contrast agent. J Cardiovasc Magn Reson 14:53. doi:10.1186/1532-429X-14-53 27. Schuleri KH, Centola M, Evers KS, Zviman A, Evers R, Lima JA, Lardo AC (2012) Cardiovascular magnetic resonance characterization of peri-infarct zone remodeling following myocardial infarction. J Cardiovasc Magn Reson 14:24. doi:10.1186/1532429X-14-24 28. Scott PA, Morgan JM, Carroll N, Murday DC, Roberts PR, Peebles CR, Harden SP et al (2011) The extent of left ventricular scar quantified by late gadolinium enhancement MRI is associated with spontaneous ventricular arrhythmias in patients with coronary artery disease and implantable cardioverter-defibrillators. Circ Arrhythm Electrophysiol 4:324–330
