Pediatr Cardiol (2015) 36:950–959 DOI 10.1007/s00246-015-1105-9
ORIGINAL ARTICLE
Two-Dimensional Speckle-Tracking-Derived Segmental Peak Systolic Longitudinal Strain Identifies Regional Myocardial Involvement in Patients with Myocarditis and Normal Global Left Ventricular Systolic Function Santosh C. Uppu • Amee Shah • Justin Weigand • James C. Nielsen • H. Helen Ko • Ira A. Parness • Shubhika Srivastava Received: 3 June 2014 / Accepted: 9 January 2015 / Published online: 24 January 2015 Ó Springer Science+Business Media New York 2015
Abstract The presence of myocardial late gadolinium enhancement (LGE) by cardiac magnetic resonance (CMR) imaging in concert with electrocardiography and elevated biomarkers helps support the diagnosis of acute myocarditis. Two-dimensional echocardiography is limited to global and qualitative regional function assessment and may not contribute to the diagnosis, especially in the presence of normal LV systolic function. Two-dimensional speckle-tracking (2D-STE)-derived segmental peak systolic (pkS) longitudinal strain (LS) may identify segmental myocardial involvement in myocarditis. We sought to identify an association between segmental pkS, LGE, and troponin levels in patients with myocarditis. Retrospective analysis of myocardial segmental function by 2D-STE segmental strain was compared to the presence of LGE and admission peak troponin levels in patients with acute myocarditis and preserved global LV systolic function. American Heart Association 17-segment model was used
for comparison between imaging modalities. Global function was assessed by m-mode-derived shortening fraction (SF). Descriptive statistics and regression analysis were utilized. Forty-four CMRs performed to evaluate for myocarditis were identified. Of the 44, 10 patients, median age 17.5 years (14–18.5 years) and median SF 35 % (28–44 %), had paired CMR and 2D-STE data for analysis, and 161/170 segments could be analyzed by both methods for comparison. PkS LS was decreased in 51 % of segments that were positive for LGE with average pkS of -14.7 %. Segmental pkS LS abnormalities were present in all but one patient who had abnormal pkS circumferential strain. Global pkS LS was decreased in patients with myocarditis. There is a moderate correlation between decreased pkS LS and the presence of LGE by CMR, 2DSTE for myocardial involvement in acute myocarditis can serve as an useful noninvasive adjunct to the existing tests used for the diagnosis of acute myocarditis and might have a role in prognostication.
S. C. Uppu (&) A. Shah J. Weigand J. C. Nielsen H. H. Ko I. A. Parness S. Srivastava Division of Pediatric Cardiology, Mount Sinai School of Medicine, Box 1201, One Gustave L. Levy Place, New York NY 10029, USA e-mail: [email protected]
Keywords Acute myocarditis Speckle-tracking echocardiography Cardiovascular magnetic resonance imaging
S. C. Uppu J. C. Nielsen Department of Radiology, Mount Sinai School of Medicine, New York, NY, USA
Background
A. Shah Division of Pediatric Cardiology, Columbia University College of Physicians and Surgeons, Morgan Stanley Children’s Hospital of New York-Presbyterian, New York, NY, USA J. C. Nielsen Division of Pediatric Cardiology, Stony Brook Children’s, New York, NY, USA
123
The presentation of acute myocarditis is highly variable, and its diagnosis is challenging, especially when there is normal left ventricular (LV) systolic function [6, 15]. Currently, the diagnosis of myocarditis is based on clinical history, physical exam findings, an electrocardiogram (EKG), and troponin levels, with no single noninvasive imaging test that reliably confirms the diagnosis or predicts prognosis in acute myocarditis. Infectious diseases account
Pediatr Cardiol (2015) 36:950–959
for the majority of myocarditis. After the initial injury, local and systemic immune responses play a major role in the pathogenesis. Cellular infiltration, edema, necrosis, and in later stages fibrotic scars are common pathological features of myocarditis [16, 23]. Accuracy of clinical history, physical examination, EKG, and serology are not always satisfactory [14]. Two-dimensional (2D) echocardiography is limited to global and qualitative regional ventricular function assessment and may not contribute to the diagnosis, especially in the presence of normal systolic function [11]. Endomyocardial biopsy (EMB) is the accepted standard, but is invasive with low yield due to the known patchy involvement of the myocardium, and hence is not recommended for every patient [2]. Cardiovascular magnetic resonance (CMR) imaging has evolved as a noninvasive and valuable clinical tool for the diagnosis of myocarditis [16]. CMR is being increasingly performed in patients presenting with suspected myocarditis. Conventional 2D echocardiography is often the initial imaging test, but diagnostic accuracy is often limited [15]. Newer echocardiographic measures of LV myocardial deformation may have better sensitivity than conventional echocardiography for the detection of subclinical LV dysfunction [10]. Two-dimensional speckle-tracking echocardiography (2D-STE) is based on frame-by-frame tracking of tiny echo-dense speckles within the myocardium and subsequent measurement of LV deformation [17], and 2D-STE can quantitatively measure regional myocardial deformation (strain and strain rate) in longitudinal, radial, and circumferential directions [14]. Recent studies have evaluated LV deformation and have demonstrated abnormalities in longitudinal and circumferential strain in patients with myocarditis and preserved LV systolic function [8, 15]. We hypothesized that 2D-STE-derived segmental peak systolic (pkS) longitudinal and circumferential strain may identify segmental myocardial involvement in myocarditis. We sought to identify an association between segmental strain with areas of late gadolinium enhancement (LGE) by CMR and troponin levels in patients with clinically suspected myocarditis.
951
suggesting myocardial involvement, and the presence of elevated inflammatory markers or cardiac enzymes was also recorded. Subjects with suspected myocarditis by laboratory and EKG findings, who underwent CMR and 2D-STE within 7 days of presentation and normal systolic function, were included in the study. The study protocol was approved by the institutional IRB. We excluded a total of 34 subjects for following reasons (Fig. 1). Missing data including lack of strain imaging data and more than 7 days between the two tests (n = 18), the absence of LGE by CMR (n = 3), reduced left ventricular function (n = 7), follow-up CMR for remote history of myocarditis (n = 2), structural heart disease with no CMR/ laboratory evidence of myocarditis (n = 4) were reasons for exclusion. A total of ten subjects (mean age 17.23 ± 1.34 years) were included for the final analysis. Nineteen age-matched subjects (mean age 15 ± 5.2 years; p value 0.2) with normal systolic function and without myocarditis, who underwent 2D-STE during the study period, were included as controls for cardiac strain comparison. CMR Protocol Studies were performed on either a 1.5-T Siemens scanner (MAGNETOM Avanto, Siemens Medical Solutions, Erlangen, Germany) (n = 3) or a 1.5-T General Electric scanner (GE Signa HD, GE Medical Systems, Waukesha, Wisconsin) (n = 7). After initial scout images, multiple long-axis and short-axis cine steady-state free precession
Methods Study Population Forty-four children with clinical suspicion of myocarditis who underwent CMR at our institute from June 2006 to January 2014 were retrospectively identified. Clinical suspicion of myocarditis was documented by an attending cardiologist based on the presence of new onset typical symptoms including flu-like symptoms, fatigue/malaise, chest pain, dyspnea, and palpitations with EKG changes
Fig. 1 Exclusion criteria
123
952 9
y = 0.5042x + 1.9263 R² = 0.4688 R = 0.68 p=0.028
8 7
LGE Segments
Fig. 2 Pearson’s regression analysis comparing number of late gadolinium enhancement (LGE) segments with number of reduced longitudinal strain segments per patient
Pediatr Cardiol (2015) 36:950–959
6 5 4 3 2 1 0 0
1
2
3
4
5
6
7
8
9
10
Longitudinal Strain Segments
(SSFP) images were obtained from the atrioventricular ring to the apex. The sequence parameters for the SSFP images were as follows: slice thickness 6–9 mm; 0–1-mm interslice gap; repetition time (TR) 3.35–4.0 ms; echo time (TE) 1.4–1.8 ms; flip angle (FA) 40°–45°; receiver bandwidth 125 kHz; acquisition matrix 160–224 9 128–192; field of view (FOV) 360–400 9 240–320 mm; with phase FOV (PFOV) 0.75–1.0, views per segment 12–18; 20 phases/cardiac cycle. LGE images were obtained using a gradient echo inversion recovery technique in the same short- and long-axis views. Sequence parameters were as follows: TR 4.5–5 ms, TE 1.3–1.6 ms, FA 20°–30°, slice thickness 6–8 mm, matrix 256 9 256 and FOV 350 mm, PFOV 80 %. Inversion time (TI) was optimized to a null signal from normal myocardium covering the left ventricle in multiple short-axis, horizontal and vertical long-axis views in end diastole and were obtained 10–15 min after an intravenous bolus of 0.2 mmol/kg MagnevistÒ (Berlex, Montville, New Jersey) with segmented inversion recovery fast gradient echo sequences (TE 1.4–1.8 ms, TR 3.3–4 ms, FA 20°, matrix 256 9 160, FOV 360–400 9 240–320 mm). The optimal TI for late gadolinium enhancement (LGE) images on GE scanner was manually set and varied from 250 to 300 ms. For the Siemens scanner, a modified Look-Locker sequence, which generates multiple images in a single slice location with increasing TIs, identified an ideal TI using a phase-sensitive, inversion recovery sequence (TR 560–800 ms; TE 4–4.3 ms; TI 250–290 ms; FA 25°–30°; FOV 300–340 mm; matrix 208–256 9 192–256; thickness 6–8 mm). In three cases, T2-weighted imaging (TE * 80 ms and TR-2 heart beats) was performed with triple inversion recovery sequences to assess for myocardial edema [9, 11]. Post-processing was performed using the standard
123
Advantage Workstation 4.3 (GE Medical Systems, Waukesha, Wisconsin) and an open source DICOM software OsiriX (OsiriX Foundation, Geneva, Switzerland) [22]. For the purpose of study, LGE was analyzed offline using OsiriX, where the delayed enhancement short-axis, horizontal and vertical long-axis views were placed in multiple windows, so that they can be viewed simultaneously crossreferencing the area of interest in multiple planes. Thicknesses of the apical myocardial segments were crosschecked with the thickness on cine-SSFP images for accuracy. Myocardial edema was detected by comparing a region of interest to adjacent normal myocardium. Patients were considered to have CMR diagnoses of myocarditis if they had subepicardial or intramyocardial late gadolinium enhancement imaging and/or myocardial edema on T2weighted imaging. The presence of LGE in each segment was assessed according to the AHA standardized myocardial segmentation [7] by two independent observers. Echocardiographic Protocol All subjects underwent transthoracic echocardiographic examination according to the recommendations of the American Society of Echocardiography [10]. Myocardial longitudinal and circumferential strain values were measured. Echocardiograms were obtained at rest with either a 4.0-MHz (M4S) or a 5.0-MHz (5S) phased-array transducer using the Vivid 7 BT08 scanner (GE Vingmed Ultrasound AS, N-3190, Horten, Norway). Two-dimensional multiframe gray-scale images were obtained in the apical four, two-chamber, and long-axis views for longitudinal strain, and parasternal short-axis view at the level of the papillary muscles for circumferential strain. Image was optimized to obtain frame rates [60 Hz [20]. Image
Pediatr Cardiol (2015) 36:950–959
data were stored at the same frame rate as the acquisition. Offline speckle-tracking analysis was performed using EchoPAC version 108.1.4 (GE Vingmed Ultrasound AS) software. Endocardial borders of the left ventricle were manually traced and, when necessary, readjusted to cover the desired area. The tracking algorithm then followed the myocardial speckles during the cardiac cycle. Tracking was accepted if both visual inspection and the EchoPAC software were in agreement. The software automatically divided the crosssectional image into seventeen segments. For the study, these segments were renamed and identified according to the statement of the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association [7]. Global pkS longitudinal strain (LS) was obtained after averaging all the regional longitudinal strain values obtained from the apical views. Global pkS circumferential strains were obtained by averaging the regional values obtained in the short-axis view. Nineteen age-matched controls with no cardiac diagnosis, scanned on GE Vivid 7 machine for various indications, with satisfactory image quality for strain analysis were randomly selected from our echocardiographic database. Offline measurements in controls were made using EchoPAC software by two observers. Statistical Analysis Summary statistics were estimated for demographic and echocardiographic parameters. Continuous variables reported as mean ± SD or as median (interquartile or absolute range), and categorical data reported as frequency (percentage). Two-tailed unpaired t tests were used to compare LGE segments, strain segments, and demographic variables. Pearson’s linear regression analysis was performed to evaluate relation between strain and LGE segments, and peak troponin values. Inter-observer variability was examined using the interclass correlation coefficient (ICC). A p value of \0.05 was considered statistically significant. All analyses were carried out using Microsoft ExcelÒ program and QuickCalcsÓ online statistical calculator by GraphPad Software, Inc., San Diego, CA. Reproducibility Inter- and intra-observer agreement for longitudinal strain and LGE segments was assessed in all the subjects by two independent observers. These were re-measured later by a single observer for intra-observer agreement. Reproducibility was analyzed by intraclass correlation coefficients (ICC); the point estimates and the 95 % confidence intervals of the agreement rate were reported.
953
Results A total of ten subjects met the study criteria; (Table 1) mean age of the subjects was 17.23 ± 1.34 years. Nine were male; eight subjects, including a girl, were obese with an overall mean body mass index (BMI) of 32.9 ± 4.59 that was corrected for both age and sex [5]. Troponin was elevated in all the subjects at presentation with a wide range of peak levels (median 9.1 ng/ml, range 1.6–36 ng/ml). A total of 161 among 170 cardiac segments were analyzed for regional strain (161—longitudinal and 58—circumferential strain segments, respectively) (Table 2). For the purpose of this study, segmental pkS LS values more negative than -13.5 % and a circumferential strain more negative than -11 % were defined as reduced, which were two standard deviations below the mean values of our agematched controls so as to improve the sensitivity. Overall global longitudinal strain was reduced at -14.79 ± 2.6; compared to age-controlled normals -17.75 ± 2.08 (p = 0.001). Thirty-five percent of the analyzed segments had reduced pkS LS. Sub-analysis of individual segments revealed that pkS LS was significantly reduced in AHA cardiac segments 4, 10, 12, 14, 15, and 16, respectively (p B 0.05, respectively), in the myocarditis group (Table 3). Average circumferential strain at the level of the papillary muscles was also significantly lower in the myocarditis group (-13.94 ± 3.3 vs.- 15.87 ± 2.25: p = 0.03); 34 % of the analyzed segments had reduced circumferential strain. Regional circumferential strain was noted to be reduced in the inferior and inferoseptal segments at the level of papillary muscles (p B 0.05, respectively) (Table 4). Left ventricular systolic function by echocardiography was normal in all the subjects (Mean shortening fraction 36 ± 4.6 vs. 34 ± 3 %, p = 0.16). The presence of LGE by CMR was noted mainly in the basal and mid-inferior, inferolateral, and anterolateral segments (Figs. 3, 6). Reduced longitudinal strain (\-13.5 %) was found in a total of 57/161 cardiac segments (median 6.5; 0–9), whereas LGE was positive in 48 segments (median 5.5; 1–8). Overlap of segments with abnormal longitudinal strain and LGE was present in 29 segments (51 %); overlap was good for basal and mid-segments, while it was poor for apical segments (13–17) compared to 2D-STE (18 vs. 35 %; p = 0.006). Reduced circumferential strain (\11 %) for segments 7–12 was found in a total of 20/58 segments (median 1.5; 0–4), compared to 16 subepicardial and intramyocardial segments by LGE (median 2; 0–3). Overlap between abnormal circumferential strain and LGE was good for mid-inferior, inferolateral, and anterolateral segments (Figs. 3, 6).
123
954
Pediatr Cardiol (2015) 36:950–959
Table 1 Patient characteristics
Data expressed as median (range). Strain expressed as mean ± standard deviation. p \ 0.05 is significant
Variable
Myocarditis (n = 10)
Controls (n = 19)
Sex (Male)
9 (90 %)
9 (47 %)
Age (years)
17.5 (14–18.6)
14.5 (8.34–22)
0.2
BMI
33.1 (24.1–40.1)
23.4 (16.2–34.1)
0.0001
BSA (m2)
2.23 (1.63–2.42)
1.7 (1.03–1.85)
0.0003
Peak Troponin (ng/ml)
9.1 (1.6–36)
–
LV shortening fraction (%)
35 % (28–44)
33 % (30–40)
LV ejection fraction (%)
66 % (61–72)
64 % (58–73)
0.15
Global longitudinal strain
-14.79 ± 2.62
-17.75 ± 2.08
0.001
Global circumferential strain
-13.94 ± 3.3
-15.35 ± 2.28
0.03
CMR
10
–
Time difference between and 2D-STE and CMR
0 days (-7 to ?1 days)
–
LV end diastolic volume (ml)
189.7 ± 32.49
–
LV ejection fraction (%)
61.4 % (54–73)
–
T2-weighted
3 (30 %)
–
Positive LGE
1 (33 %) 10 (100 %)
– –
Positive
10 (100 %)
–
p value
0.16
Table 2 Echocardiography, strain, and late gadolinium enhancement parameters Subject
Abnormal circumferential strain AHA segments (number)
Peak troponin (ng/ml)
Abnormal longitudinal strain AHA segments (number)
LGE AHA segments (number)
Overlap of longitudinal strain and LGE
SF (%)
EF by CMR (%)
1
2
22.1
9
5
4
35
54
2
1
21.8
5
6
4
35
56
3
1
5.6
7
2
2
37
66
4
4
3.2
5
4
1
41
65
5
1
2.8
1
3
1
44
62
6
1
1.9
0
1
0
34
64
7
3
1.6
8
8
6
28
56
8
4
12.6
8
7
5
33
55
9
0
22.8
6
6
3
32
63
10
3
36
8
6
4
41
73
LGE late gadolinium enhancement, SF shortening fraction by m-mode, EF ejection fraction, CMR cardiac magnetic resonance imaging, AHA American Heart Association
There is moderate correlation between the overall number of LGE segments by CMR and abnormal longitudinal strain per patient (r = 0.68; p = 0.03) (Fig. 2). The correlation was weak when the number of LGE/abnormal longitudinal strain segments compared to peak troponin levels (r = 0.42; p = 0.22 and r = 0.49; p = 0.15, respectively). Global longitudinal strain by 2D-STE imaging in our cohort was reduced in 9/10 patients, and circumferential strain was also reduced in 9/10 patients at mid-ventricular level. The patients with normal longitudinal strain had abnormal circumferential strain, and vice versa (Table 2; Fig. 3).
123
There was good inter-observer agreement for longitudinal strain and LGE segments, ICC of 0.86 (95 % CI 0.83–0.89) and 0.93 (95 % CI 0.91–0.95), respectively. Intra-observer agreement for longitudinal strain and LGE was also very good with ICC 0.93 (95 % CI 0.90–0.95) and 0.96 (95 % CI 0.95–0.97), respectively. Discussion Our study shows decreased global pkS longitudinal and circumferential strain, even with normal left ventricular systolic function among children with acute myocarditis.
Pediatr Cardiol (2015) 36:950–959 Table 3 Longitudinal strain distribution by American Heart Association segments
STDEV standard deviation. n number. p \ 0.05 is significant, AHA American Heart Association AHA segments: Basal segments: 1. basal anterior, 2. basal anteroseptal, 3. basal inferoseptal, 4. basal inferior, 5. basal inferolateral, 6. basal anterolateral; mid-cavity segments: 7. mid-anterior, 8. mid-anteroseptal, 9. midinferoseptal, 10. mid-inferior, 11. mid-inferolateral, and 12. mid-anterolateral; Apical segments: 13. apical anterior, 14. apical septal, 15. apical inferior, 16. apical lateral, and 17. apex
955
AHA segment number
Cases
Controls
Mean
STDEV
1
-15.56
5.51
2
-17.97
1.83
3
-18.96
4
n
p value
Mean
STDEV
n
8
-19.67
5.9
17
0.11
9
-17.53
3.78
19
0.7448
2.96
10
-18.07
2.44
19
0.3931
-14.45
5.83
10
-20.19
3.3
19
0.0021
5
-14.75
5.74
7
-15
6.55
19
0.9299
6
-11.07
6.57
9
-12.78
4.54
19
0.4281
7
-13.35
5.98
8
-16.85
3.59
16
0.0856
8
-16.31
5.67
9
-17.98
4.84
17
0.4376
9
-18.75
2.8
10
-19.32
2.07
19
0.538
10
-15.18
3.03
10
-19.89
3.22
19
0.0007
11 12
-13.77 -10.63
5.35 6.25
9 10
-15.81 -15.31
4.53 3.64
18 19
0.3086 0.0163
13
-15.09
6.51
9
-17.16
6.61
15
0.4631
14
-17.56
4.86
10
-21.16
4.28
18
0.05
15
-16.2
6.11
7
-21.49
3.7
18
0.0138
16
-13.29
6.53
10
-18.88
4.53
19
0.0116
17
-12.5
5.62
8
-15.33
6.69
16
0.316
Table 4 Circumferential strain distribution by segments at mid-ventricular level (papillary muscle) AHA segment number
Cases
Controls
p value
Mean
STDEV
n
Mean
STDEV
n
7
-17.17
5.76
10
-15.05
6.09
16
0.387
8
-20.27
6.19
10
-19.92
4.19
16
0.8645
9
-15.14
6.01
10
-20.44
5.27
16
0.0265
10
-9.65
4.55
9
-15.83
5.92
16
0.0126
11
-7.85
5.28
10
-10.25
9.36
16
0.4681
12
-12.18
6.72
9
-13.49
5.29
15
0.6007
STDEV standard deviation. n number. p \ 0.05 is significant, AHA American Heart Association AHA mid-cavity segments: 7. mid-anterior, 8. mid-anteroseptal, 9. mid-inferoseptal, 10. mid-inferior, 11. mid-inferolateral, and 12. midanterolateral
Our finding of reduced global longitudinal and circumferential strain in acute myocarditis confirms earlier reports by Khoo et al. [15] in adolescents and young adults and by Hsiao et al. [14] and Di Bella et al. [8] in adults. Myocarditis is identified by CMR as areas of hyperenhancement (LGE) preferentially involving subepicardial, intra-myocardial regions of the basal and/or mid-lateral cardiac segments (Fig. 4), and similar areas might show hyper-enhancement suggestive of fluid/edema by T2weighted imaging [6] (Fig. 4). Interestingly, we noted significantly lower regional longitudinal strain predominantly involving basal, mid-and apical inferior and anterolateral segments, thus suggesting a possibility of regional
inflammation altering the regional myocardial function that can be identified easily using strain (2D-STE) imaging, in spite of preserved systolic function. LGE by CMR is prone to subjective bias and is especially difficult to interpret in the apical regions where the spatial resolution is low. This makes LGE by CMR alone a less robust tool to diagnose myocarditis in subjects with no regional wall motion abnormalities and other inconclusive CMR parameters (T2-weighted imaging, first-pass perfusion, and early gadolinium enhancement) (Fig. 5). CMR can be cumbersome and time-consuming especially in children less than 10 years of age with acute myocarditis, who may not be as cooperative during the
123
956 Fig. 3 Graph showing the distribution 18 of abnormal longitudinal, circumferential strain, and late 16 gadolinium enhancement (LGE) 14 segments by American heart 12 association segmentation (AHA segments: Basal segments: 1. basal 10 anterior, 2. basal anteroseptal, 3. basal inferoseptal, 4. basal inferior, 5. basal 8 6 inferolateral, 6. basal anterolateral; mid-cavity segments: 7. mid-anterior, 4 8. mid-anteroseptal, 9. midinferoseptal, 10. mid-inferior, 11. mid- 2 inferolateral, and 12. mid0 anterolateral; Apical segments: 13. apical anterior, 14. apical septal, 15. apical inferior, 16. apical lateral, and 17. apex)
Pediatr Cardiol (2015) 36:950–959
Segemental involvement
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Circumferenal Strain (segments 7-12) Longitudinal Strain MDE
Fig. 5 Same patient as in Fig. 4 with areas of abnormal longitudinal strain by 2D-STE (light red to blue areas). Compare the abnormal longitudinal strain areas with areas of LGE and positive T2-weighted areas on the previous image. Normal longitudinal strain in the base and more areas of strain abnormality in the apical segments (7, 8, 12, 13, 14, 15, 16, and 17). The number in each AHA segment measures peak systolic longitudinal strain
Fig. 4 Patient with acute myocarditis showing areas of late gadolinium enhancement (LGE) (blue arrows) and positive T2-weighted imaging (red arrows). a and g—vertical long axis; b—horizontal long axis; c, d, e—LGE imaging at base, mid-ventricular and apex; f—T2weighted imaging at base. Involved segments by CMR: 4, 5, 11,12,15,16, and 17
CMR scan and thus further reducing the efficacy of this modality alone. We showed overlap among the areas with reduced longitudinal strain and areas of delayed enhancement (LGE) by CMR in 51 % of the segments. This might be due to myocardial injury/inflammation that affects both regional strain and contrast stagnation resulting in a positive LGE. To our surprise, there are more areas of reduced
123
longitudinal strain compared to areas of LGE (57 vs. 48; 29 segments overlap) with moderate correlation between the overall number of LGE segments by CMR and 2D-STE per patient (r = 0.68; p = 0.03) (Fig. 2), although there was no segmental overlap 49 % of times for all patients, suggesting that LGE and 2D-STE technologies likely were identifying different pathological processes. Interestingly, CMR appears to under-diagnose affected myocardium in cardiac apical segments (13–17) compared to 2D-STE (18 vs. 35 %; p = 0.006) (Fig. 6). CMR studies [2, 18] in adults have shown satisfactory accuracy of CMR for acute myocarditis with sensitivity; specificity; and accuracy of 81, 71, and 79 %, respectively. No such studies exist for pediatric population. There is significant practice variation performing CMR’s among
Pediatr Cardiol (2015) 36:950–959
pediatric centers [3] with many centers performing LGE routinely and fewer centers performing first-pass perfusion and early gadolinium enhancement. There is the absence of uniform consensus even among adult centers where LGE is used routinely; some groups advocate inclusion of firstpass perfusion, contrast-enhanced T1-weighted imaging, T1-mapping, T2-weighted imaging, and/or extracellular volume fraction to improve the accuracy [2]. These modalities are prone to image quality and reproducibility issues and are not routinely adopted in the clinical practice. CMR alone to diagnose myocarditis has technical limitations that include trigger problems, artifacts, and poor resolution which might interfere with obtaining diagnostic quality images and accurate interpretation [2]. CMR in younger noncooperative children is even more resource intensive and might require sedation, and/or anesthesia to obtain satisfactory diagnostic images. CMR does have an advantage in obtaining high-quality images even in a patient with poor acoustic windows. Tissue characterization techniques are being increasingly adopted and might have a higher diagnostic yield. Automated post-processing techniques are being developed, which are more objective, thus providing unbiased and consistent results. On the contrary, cardiac strain imaging is less resource intensive, less expensive, portable, easily performed even in a moderately cooperative child by a skilled operator providing diagnostic quality images. 2D-STE is adopted by different vendors making it omnipresent [20]; there is an increased trend using this technology even among children
Fig. 6 AHA 17-segment models—for top two figures, each number within the segment represents number of subjects with involvement of that particular segments comparing late gadolinium enhancement (LGE) and reduced longitudinal strain 2D-STE
957
[15]. Post-processing of 2D-STE is automated requiring minimal interpreter adjustment; however, there is difference among the ‘‘normal’’ values between the vendors [20, 21] that has to be taken into account while interpreting these results. It is known that LV myocardium is mostly composed of myocardial fibers that are circumferentially oriented (*59 %) parallel to the mitral valve orifice; subepicardial fibers that are oriented obliquely and cross the interventricular groove; subendocardial longitudinal fibers (20 %) that insert into the aortic and mitral valves [13]. Myocarditis predominantly affects subepicardial and intramyocardial regions, thus effecting oblique and circumferentially oriented fibers, and involvement of these fibers can be recognized by as abnormal strain on 2D-STE. Left ventricular function is usually preserved as long as the involvement of myocardium is minimal, thus making it difficult to recognize regional wall motion abnormalities in early/minimal myocarditis by traditional echocardiography alone. It is interesting to note that pkS longitudinal strain is globally reduced in spite of subendocardial sparing by CMR in all the subjects, suggesting a global myocardial involvement that is not readily identified by current CMR techniques. Ease of access to 2D-STE imaging also helps to follow these patients serially to evaluate the progression and thus helping with the management plan. We have shown a correlation between the segments of LGE and abnormal 2D-STE; as was reported in previous case reports [1, 12] and although this correlation is modest, it needs to be further evaluated in a prospective study in order to establish its significance. Our correlation might have been affected by the small sample size. It is interesting to note that 9/10 patients were obese, even in comparison with the age-matched controls. Recent evidence shows reduced global longitudinal strain in obese children, in spite of normal left ventricular ejection fraction [4]. Whether the reduced strain in our cohort is related to the obesity, myocarditis, or a combination is hard to determine. It also raises an important question whether obesity predisposes or increases the chance of myocardial damage when exposed to environmental/viral agents. Global longitudinal and circumferential strain values among controls in our laboratory were lower than those of reported by Marcus et al. in 21 subjects aged 15–19 [19] (-17.78 ± 2.19 vs. -20.14 ± 1.6; p = 0.0004 and -15.6 ± 2.2 vs. -23.6 ± 2; p = 0.0001), even though we used similar methods, and vendors for the data collection and analysis; We believe that this difference could be as a result of demographic variation of the sample itself (USA vs. Israel and Netherlands), which brings out an important point on deciding what normal values to use. Due to paucity of normal data for children, we recommend using values obtained in the local laboratory.
123
958
It is not our institutional policy to routinely perform EMB in patients presenting with myocarditis. Histopathological exam when combined with CMR have a higher diagnostic yield of *95 % than compared to CMR or EMB alone (80 and 88 %, respectively) [2]. Baccouche et al. [2] have recommended routinely performing CMR and reserving EMB if the diagnosis is unclear or the symptoms persist. CMR also has a limitation to identify the degree of inflammation or identify special forms of myocarditis (eosinophilic or giant cell etc.). Using peak systolic global longitudinal strain by 2DSTE imaging in our cohort, the sensitivity was 90 %, and 9/10 patients had low global longitudinal and circumferential strain values. The patient with normal longitudinal strain had reduced circumferential strain (segment 11) and a positive LGE in segment 7 with mildly elevated troponin (1.9 ng/ml). Limitations Our study shares the limitations inherent to many other retrospective designs; it was a single-institution study with a small study population, subjects were mostly male, and thus gender-related effects could not be excluded. As previously discussed, majority of our subjects were obese and the effect of obesity on our strain and LGE measures could not be further investigated given the small sample. Our study contributes to the small pool of reported studies comparing CMR in myocarditis with preserved ventricular function and myocardial strain abnormalities [1, 8, 12] and is unique by being exclusively in pediatric subjects. Our study also reports the issues using the current age-matched normals reported in the literature [19] that were collected in a different demographic setting and proposes for a study in a larger population to obtain a normative data that can be widely used. We could not test the influence of regional myocardial edema on regional strain, as T2-weighted imaging was performed in only three subjects. Lack of histopathological correlation is another limitation. The correlation between LGE and 2D-STE could have been stronger if we had performed both tests within a short time period closer to the peak troponin levels, thus correlating imaging to the underlying myocardial involvement. This hypothesis will need further testing in a larger prospectively designed study to better understand the relation between the imaging modalities. As 2D-STE is becoming more universally available and accepted, studies involving various vendors and post-processing software need to be performed on a larger scale in children to obtain vendor-specific normative data and cutoffs. There is an ongoing discussion about the subjective bias involved in identifying areas of LGE by CMR, future
123
Pediatr Cardiol (2015) 36:950–959
protocols like T1 mapping; extracellular volume fraction may prove useful by making analysis more objective, but this needs to be tested prior to their universal acceptance.
Conclusion LV myocardial segmental involvement by acute myocarditis in patients with preserved LV systolic function can be identified by both 2D-STE and the presence of LGE by CMR. 2D-STE is more objective and is better than CMR in identifying abnormalities in the apical regions. 2D-STE assessment for myocardial involvement in acute myocarditis can serve as a useful noninvasive adjunct to the existing tests used for the diagnosis of acute myocarditis and might play a very important role in serial evaluation. Further studies are required to obtain age-specific normative data that can be widely used, and across all vendor platforms. 2D-STE, in addition to CMR, may serve as a clinically useful surrogate for risk stratification and prospective evaluation of patients with acute myocarditis. The findings of this study support need for future studies in a larger sample to establish their clinical utility.
References 1. Afonso L, Hari P, Pidlaoan V, Kondur A, Jacob S, Khetarpal V (2010) Acute myocarditis: can novel echocardiographic techniques assist with diagnosis? Eur J Echocardiogr 11:E5 2. Baccouche H, Mahrholdt H, Meinhardt G, Merher R, Voehringer M, Hill S, Klingel K, Kandolf R, Sechtem U, Yilmaz A (2009) Diagnostic synergy of non-invasive cardiovascular magnetic resonance and invasive endomyocardial biopsy in troponinpositive patients without coronary artery disease. Eur Heart J 30:2869–2879 3. Banka P, Uppu S, Matthew, Hasbani K, Lai W, Richmond M, Fratz S, Jain S, Johnson T, Maskatia S, Lu J, Samyn M, Patton D, Powell A (2014) CMR techniques and findings in children with myocarditis: a multicenter retrospective study. J Cardiovasc Magn Reson. doi:10.1186/1532-429X-16-S1-P119 4. Barbosa JA, Mota CC, Simo˜es E, Silva AC, Nunes MoC, Barbosa MM (2013) Assessing pre-clinical ventricular dysfunction in obese children and adolescents: the value of speckle tracking imaging. Eur Heart J Cardiovasc Imaging 14:882–889 5. Barlow SE (2007) Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report. Pediatrics 120(Suppl 4):S164–S192 6. Camastra GS, Cacciotti L, Marconi F, Sbarbati S, Pironi B, Ansalone G (2007) Late enhancement detected by cardiac magnetic resonance imaging in acute myocarditis mimicking acute myocardial infarction: location patterns and lack of correlation with systolic function. J Cardiovasc Med (Hagerstown) 8:1029–1033 7. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS, Imaging AHAWGoMSaRfC (2002) Standardized myocardial
Pediatr Cardiol (2015) 36:950–959
8.
9.
10.
11.
12.
13.
14.
15.
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. Int J Cardiovasc Imaging 18:539– 542 Di Bella G, Gaeta M, Pingitore A, Oreto G, Zito C, Minutoli F, Anfuso C, Dattilo G, Lamari A, Coglitore S, Carerj S (2010) Myocardial deformation in acute myocarditis with normal left ventricular wall motion–a cardiac magnetic resonance and 2-dimensional strain echocardiographic study. Circ J 74: 1205–1213 Fratz S, Chung T, Greil GF, Samyn MM, Taylor AM, Valsangiacomo Buechel ER, Yoo SJ, Powell AJ (2013) Guidelines and protocols for cardiovascular magnetic resonance in children and adults with congenital heart disease: SCMR expert consensus group on congenital heart disease. J Cardiovasc Magn Reson 15:51 Geyer H, Caracciolo G, Abe H, Wilansky S, Carerj S, Gentile F, Nesser HJ, Khandheria B, Narula J, Sengupta PP (2010) Assessment of myocardial mechanics using speckle tracking echocardiography: fundamentals and clinical applications. J Am Soc Echocardiogr 23:351–369 quiz 453–355 Goitein O, Matetzky S, Beinart R, Di Segni E, Hod H, Bentancur A, Konen E (2009) Acute myocarditis: noninvasive evaluation with cardiac MRI and transthoracic echocardiography. AJR Am J Roentgenol 192:254–258 Ha SJ, Woo JS, Kwon SH, Oh CH, Kim KS, Bae JH, Kim WS (2013) Acute regional myocarditis with normal ventricular wall motion diagnosed by two-dimensional speckle tracking imaging. Korean J Intern Med 28:732–735 Ho SY (2009) Anatomy and myoarchitecture of the left ventricular wall in normal and in disease. Eur J Echocardiogr 10:iii3– iii7 Hsiao JF, Koshino Y, Bonnichsen CR, Yu Y, Miller FA, Pellikka PA, Cooper LT, Villarraga HR (2013) Speckle tracking echocardiography in acute myocarditis. Int J Cardiovasc Imaging 29:275–284 Khoo NS, Smallhorn JF, Atallah J, Kaneko S, Mackie AS, Paterson I (2012) Altered left ventricular tissue velocities,
959
16.
17.
18.
19.
20.
21.
22.
23.
deformation and twist in children and young adults with acute myocarditis and normal ejection fraction. J Am Soc Echocardiogr 25:294–303 Kindermann I, Barth C, Mahfoud F, Ukena C, Lenski M, Yilmaz A, Klingel K, Kandolf R, Sechtem U, Cooper LT, Bo¨hm M (2012) Update on myocarditis. J Am Coll Cardiol 59:779–792 Leitman M, Lysyansky P, Sidenko S, Shir V, Peleg E, Binenbaum M, Kaluski E, Krakover R, Vered Z (2004) Two-dimensional strain-a novel software for real-time quantitative echocardiographic assessment of myocardial function. J Am Soc Echocardiogr 17:1021–1029 Lurz P, Eitel I, Adam J, Steiner J, Grothoff M, Desch S, Fuernau G, de Waha S, Sareban M, Luecke C, Klingel K, Kandolf R, Schuler G, Gutberlet M, Thiele H (2012) Diagnostic performance of CMR imaging compared with EMB in patients with suspected myocarditis. JACC Cardiovasc Imaging 5:513–524 Marcus KA, Mavinkurve-Groothuis AM, Barends M, van Dijk A, Feuth T, de Korte C, Kapusta L (2011) Reference values for myocardial two-dimensional strain echocardiography in a healthy pediatric and young adult cohort. J Am Soc Echocardiogr 24:625–636 Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, Galderisi M, Marwick T, Nagueh SF, Sengupta PP, Sicari R, Smiseth OA, Smulevitz B, Takeuchi M, Thomas JD, Vannan M, Voigt JU, Zamorano JL (2011) Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography. Eur J Echocardiogr 12:167–205 Risum N, Ali S, Olsen NT, Jons C, Khouri MG, Lauridsen TK, Samad Z, Velazquez EJ, Sogaard P, Kisslo J (2012) Variability of global left ventricular deformation analysis using vendor dependent and independent two-dimensional speckle-tracking software in adults. J Am Soc Echocardiogr 25:1195–1203 Rosset A, Spadola L, Ratib O (2004) OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging 17:205–216 Sagar S, Liu PP, Cooper LT (2012) Myocarditis. Lancet 379:738–747
123
ORIGINAL ARTICLE
Two-Dimensional Speckle-Tracking-Derived Segmental Peak Systolic Longitudinal Strain Identifies Regional Myocardial Involvement in Patients with Myocarditis and Normal Global Left Ventricular Systolic Function Santosh C. Uppu • Amee Shah • Justin Weigand • James C. Nielsen • H. Helen Ko • Ira A. Parness • Shubhika Srivastava Received: 3 June 2014 / Accepted: 9 January 2015 / Published online: 24 January 2015 Ó Springer Science+Business Media New York 2015
Abstract The presence of myocardial late gadolinium enhancement (LGE) by cardiac magnetic resonance (CMR) imaging in concert with electrocardiography and elevated biomarkers helps support the diagnosis of acute myocarditis. Two-dimensional echocardiography is limited to global and qualitative regional function assessment and may not contribute to the diagnosis, especially in the presence of normal LV systolic function. Two-dimensional speckle-tracking (2D-STE)-derived segmental peak systolic (pkS) longitudinal strain (LS) may identify segmental myocardial involvement in myocarditis. We sought to identify an association between segmental pkS, LGE, and troponin levels in patients with myocarditis. Retrospective analysis of myocardial segmental function by 2D-STE segmental strain was compared to the presence of LGE and admission peak troponin levels in patients with acute myocarditis and preserved global LV systolic function. American Heart Association 17-segment model was used
for comparison between imaging modalities. Global function was assessed by m-mode-derived shortening fraction (SF). Descriptive statistics and regression analysis were utilized. Forty-four CMRs performed to evaluate for myocarditis were identified. Of the 44, 10 patients, median age 17.5 years (14–18.5 years) and median SF 35 % (28–44 %), had paired CMR and 2D-STE data for analysis, and 161/170 segments could be analyzed by both methods for comparison. PkS LS was decreased in 51 % of segments that were positive for LGE with average pkS of -14.7 %. Segmental pkS LS abnormalities were present in all but one patient who had abnormal pkS circumferential strain. Global pkS LS was decreased in patients with myocarditis. There is a moderate correlation between decreased pkS LS and the presence of LGE by CMR, 2DSTE for myocardial involvement in acute myocarditis can serve as an useful noninvasive adjunct to the existing tests used for the diagnosis of acute myocarditis and might have a role in prognostication.
S. C. Uppu (&) A. Shah J. Weigand J. C. Nielsen H. H. Ko I. A. Parness S. Srivastava Division of Pediatric Cardiology, Mount Sinai School of Medicine, Box 1201, One Gustave L. Levy Place, New York NY 10029, USA e-mail: [email protected]
Keywords Acute myocarditis Speckle-tracking echocardiography Cardiovascular magnetic resonance imaging
S. C. Uppu J. C. Nielsen Department of Radiology, Mount Sinai School of Medicine, New York, NY, USA
Background
A. Shah Division of Pediatric Cardiology, Columbia University College of Physicians and Surgeons, Morgan Stanley Children’s Hospital of New York-Presbyterian, New York, NY, USA J. C. Nielsen Division of Pediatric Cardiology, Stony Brook Children’s, New York, NY, USA
123
The presentation of acute myocarditis is highly variable, and its diagnosis is challenging, especially when there is normal left ventricular (LV) systolic function [6, 15]. Currently, the diagnosis of myocarditis is based on clinical history, physical exam findings, an electrocardiogram (EKG), and troponin levels, with no single noninvasive imaging test that reliably confirms the diagnosis or predicts prognosis in acute myocarditis. Infectious diseases account
Pediatr Cardiol (2015) 36:950–959
for the majority of myocarditis. After the initial injury, local and systemic immune responses play a major role in the pathogenesis. Cellular infiltration, edema, necrosis, and in later stages fibrotic scars are common pathological features of myocarditis [16, 23]. Accuracy of clinical history, physical examination, EKG, and serology are not always satisfactory [14]. Two-dimensional (2D) echocardiography is limited to global and qualitative regional ventricular function assessment and may not contribute to the diagnosis, especially in the presence of normal systolic function [11]. Endomyocardial biopsy (EMB) is the accepted standard, but is invasive with low yield due to the known patchy involvement of the myocardium, and hence is not recommended for every patient [2]. Cardiovascular magnetic resonance (CMR) imaging has evolved as a noninvasive and valuable clinical tool for the diagnosis of myocarditis [16]. CMR is being increasingly performed in patients presenting with suspected myocarditis. Conventional 2D echocardiography is often the initial imaging test, but diagnostic accuracy is often limited [15]. Newer echocardiographic measures of LV myocardial deformation may have better sensitivity than conventional echocardiography for the detection of subclinical LV dysfunction [10]. Two-dimensional speckle-tracking echocardiography (2D-STE) is based on frame-by-frame tracking of tiny echo-dense speckles within the myocardium and subsequent measurement of LV deformation [17], and 2D-STE can quantitatively measure regional myocardial deformation (strain and strain rate) in longitudinal, radial, and circumferential directions [14]. Recent studies have evaluated LV deformation and have demonstrated abnormalities in longitudinal and circumferential strain in patients with myocarditis and preserved LV systolic function [8, 15]. We hypothesized that 2D-STE-derived segmental peak systolic (pkS) longitudinal and circumferential strain may identify segmental myocardial involvement in myocarditis. We sought to identify an association between segmental strain with areas of late gadolinium enhancement (LGE) by CMR and troponin levels in patients with clinically suspected myocarditis.
951
suggesting myocardial involvement, and the presence of elevated inflammatory markers or cardiac enzymes was also recorded. Subjects with suspected myocarditis by laboratory and EKG findings, who underwent CMR and 2D-STE within 7 days of presentation and normal systolic function, were included in the study. The study protocol was approved by the institutional IRB. We excluded a total of 34 subjects for following reasons (Fig. 1). Missing data including lack of strain imaging data and more than 7 days between the two tests (n = 18), the absence of LGE by CMR (n = 3), reduced left ventricular function (n = 7), follow-up CMR for remote history of myocarditis (n = 2), structural heart disease with no CMR/ laboratory evidence of myocarditis (n = 4) were reasons for exclusion. A total of ten subjects (mean age 17.23 ± 1.34 years) were included for the final analysis. Nineteen age-matched subjects (mean age 15 ± 5.2 years; p value 0.2) with normal systolic function and without myocarditis, who underwent 2D-STE during the study period, were included as controls for cardiac strain comparison. CMR Protocol Studies were performed on either a 1.5-T Siemens scanner (MAGNETOM Avanto, Siemens Medical Solutions, Erlangen, Germany) (n = 3) or a 1.5-T General Electric scanner (GE Signa HD, GE Medical Systems, Waukesha, Wisconsin) (n = 7). After initial scout images, multiple long-axis and short-axis cine steady-state free precession
Methods Study Population Forty-four children with clinical suspicion of myocarditis who underwent CMR at our institute from June 2006 to January 2014 were retrospectively identified. Clinical suspicion of myocarditis was documented by an attending cardiologist based on the presence of new onset typical symptoms including flu-like symptoms, fatigue/malaise, chest pain, dyspnea, and palpitations with EKG changes
Fig. 1 Exclusion criteria
123
952 9
y = 0.5042x + 1.9263 R² = 0.4688 R = 0.68 p=0.028
8 7
LGE Segments
Fig. 2 Pearson’s regression analysis comparing number of late gadolinium enhancement (LGE) segments with number of reduced longitudinal strain segments per patient
Pediatr Cardiol (2015) 36:950–959
6 5 4 3 2 1 0 0
1
2
3
4
5
6
7
8
9
10
Longitudinal Strain Segments
(SSFP) images were obtained from the atrioventricular ring to the apex. The sequence parameters for the SSFP images were as follows: slice thickness 6–9 mm; 0–1-mm interslice gap; repetition time (TR) 3.35–4.0 ms; echo time (TE) 1.4–1.8 ms; flip angle (FA) 40°–45°; receiver bandwidth 125 kHz; acquisition matrix 160–224 9 128–192; field of view (FOV) 360–400 9 240–320 mm; with phase FOV (PFOV) 0.75–1.0, views per segment 12–18; 20 phases/cardiac cycle. LGE images were obtained using a gradient echo inversion recovery technique in the same short- and long-axis views. Sequence parameters were as follows: TR 4.5–5 ms, TE 1.3–1.6 ms, FA 20°–30°, slice thickness 6–8 mm, matrix 256 9 256 and FOV 350 mm, PFOV 80 %. Inversion time (TI) was optimized to a null signal from normal myocardium covering the left ventricle in multiple short-axis, horizontal and vertical long-axis views in end diastole and were obtained 10–15 min after an intravenous bolus of 0.2 mmol/kg MagnevistÒ (Berlex, Montville, New Jersey) with segmented inversion recovery fast gradient echo sequences (TE 1.4–1.8 ms, TR 3.3–4 ms, FA 20°, matrix 256 9 160, FOV 360–400 9 240–320 mm). The optimal TI for late gadolinium enhancement (LGE) images on GE scanner was manually set and varied from 250 to 300 ms. For the Siemens scanner, a modified Look-Locker sequence, which generates multiple images in a single slice location with increasing TIs, identified an ideal TI using a phase-sensitive, inversion recovery sequence (TR 560–800 ms; TE 4–4.3 ms; TI 250–290 ms; FA 25°–30°; FOV 300–340 mm; matrix 208–256 9 192–256; thickness 6–8 mm). In three cases, T2-weighted imaging (TE * 80 ms and TR-2 heart beats) was performed with triple inversion recovery sequences to assess for myocardial edema [9, 11]. Post-processing was performed using the standard
123
Advantage Workstation 4.3 (GE Medical Systems, Waukesha, Wisconsin) and an open source DICOM software OsiriX (OsiriX Foundation, Geneva, Switzerland) [22]. For the purpose of study, LGE was analyzed offline using OsiriX, where the delayed enhancement short-axis, horizontal and vertical long-axis views were placed in multiple windows, so that they can be viewed simultaneously crossreferencing the area of interest in multiple planes. Thicknesses of the apical myocardial segments were crosschecked with the thickness on cine-SSFP images for accuracy. Myocardial edema was detected by comparing a region of interest to adjacent normal myocardium. Patients were considered to have CMR diagnoses of myocarditis if they had subepicardial or intramyocardial late gadolinium enhancement imaging and/or myocardial edema on T2weighted imaging. The presence of LGE in each segment was assessed according to the AHA standardized myocardial segmentation [7] by two independent observers. Echocardiographic Protocol All subjects underwent transthoracic echocardiographic examination according to the recommendations of the American Society of Echocardiography [10]. Myocardial longitudinal and circumferential strain values were measured. Echocardiograms were obtained at rest with either a 4.0-MHz (M4S) or a 5.0-MHz (5S) phased-array transducer using the Vivid 7 BT08 scanner (GE Vingmed Ultrasound AS, N-3190, Horten, Norway). Two-dimensional multiframe gray-scale images were obtained in the apical four, two-chamber, and long-axis views for longitudinal strain, and parasternal short-axis view at the level of the papillary muscles for circumferential strain. Image was optimized to obtain frame rates [60 Hz [20]. Image
Pediatr Cardiol (2015) 36:950–959
data were stored at the same frame rate as the acquisition. Offline speckle-tracking analysis was performed using EchoPAC version 108.1.4 (GE Vingmed Ultrasound AS) software. Endocardial borders of the left ventricle were manually traced and, when necessary, readjusted to cover the desired area. The tracking algorithm then followed the myocardial speckles during the cardiac cycle. Tracking was accepted if both visual inspection and the EchoPAC software were in agreement. The software automatically divided the crosssectional image into seventeen segments. For the study, these segments were renamed and identified according to the statement of the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association [7]. Global pkS longitudinal strain (LS) was obtained after averaging all the regional longitudinal strain values obtained from the apical views. Global pkS circumferential strains were obtained by averaging the regional values obtained in the short-axis view. Nineteen age-matched controls with no cardiac diagnosis, scanned on GE Vivid 7 machine for various indications, with satisfactory image quality for strain analysis were randomly selected from our echocardiographic database. Offline measurements in controls were made using EchoPAC software by two observers. Statistical Analysis Summary statistics were estimated for demographic and echocardiographic parameters. Continuous variables reported as mean ± SD or as median (interquartile or absolute range), and categorical data reported as frequency (percentage). Two-tailed unpaired t tests were used to compare LGE segments, strain segments, and demographic variables. Pearson’s linear regression analysis was performed to evaluate relation between strain and LGE segments, and peak troponin values. Inter-observer variability was examined using the interclass correlation coefficient (ICC). A p value of \0.05 was considered statistically significant. All analyses were carried out using Microsoft ExcelÒ program and QuickCalcsÓ online statistical calculator by GraphPad Software, Inc., San Diego, CA. Reproducibility Inter- and intra-observer agreement for longitudinal strain and LGE segments was assessed in all the subjects by two independent observers. These were re-measured later by a single observer for intra-observer agreement. Reproducibility was analyzed by intraclass correlation coefficients (ICC); the point estimates and the 95 % confidence intervals of the agreement rate were reported.
953
Results A total of ten subjects met the study criteria; (Table 1) mean age of the subjects was 17.23 ± 1.34 years. Nine were male; eight subjects, including a girl, were obese with an overall mean body mass index (BMI) of 32.9 ± 4.59 that was corrected for both age and sex [5]. Troponin was elevated in all the subjects at presentation with a wide range of peak levels (median 9.1 ng/ml, range 1.6–36 ng/ml). A total of 161 among 170 cardiac segments were analyzed for regional strain (161—longitudinal and 58—circumferential strain segments, respectively) (Table 2). For the purpose of this study, segmental pkS LS values more negative than -13.5 % and a circumferential strain more negative than -11 % were defined as reduced, which were two standard deviations below the mean values of our agematched controls so as to improve the sensitivity. Overall global longitudinal strain was reduced at -14.79 ± 2.6; compared to age-controlled normals -17.75 ± 2.08 (p = 0.001). Thirty-five percent of the analyzed segments had reduced pkS LS. Sub-analysis of individual segments revealed that pkS LS was significantly reduced in AHA cardiac segments 4, 10, 12, 14, 15, and 16, respectively (p B 0.05, respectively), in the myocarditis group (Table 3). Average circumferential strain at the level of the papillary muscles was also significantly lower in the myocarditis group (-13.94 ± 3.3 vs.- 15.87 ± 2.25: p = 0.03); 34 % of the analyzed segments had reduced circumferential strain. Regional circumferential strain was noted to be reduced in the inferior and inferoseptal segments at the level of papillary muscles (p B 0.05, respectively) (Table 4). Left ventricular systolic function by echocardiography was normal in all the subjects (Mean shortening fraction 36 ± 4.6 vs. 34 ± 3 %, p = 0.16). The presence of LGE by CMR was noted mainly in the basal and mid-inferior, inferolateral, and anterolateral segments (Figs. 3, 6). Reduced longitudinal strain (\-13.5 %) was found in a total of 57/161 cardiac segments (median 6.5; 0–9), whereas LGE was positive in 48 segments (median 5.5; 1–8). Overlap of segments with abnormal longitudinal strain and LGE was present in 29 segments (51 %); overlap was good for basal and mid-segments, while it was poor for apical segments (13–17) compared to 2D-STE (18 vs. 35 %; p = 0.006). Reduced circumferential strain (\11 %) for segments 7–12 was found in a total of 20/58 segments (median 1.5; 0–4), compared to 16 subepicardial and intramyocardial segments by LGE (median 2; 0–3). Overlap between abnormal circumferential strain and LGE was good for mid-inferior, inferolateral, and anterolateral segments (Figs. 3, 6).
123
954
Pediatr Cardiol (2015) 36:950–959
Table 1 Patient characteristics
Data expressed as median (range). Strain expressed as mean ± standard deviation. p \ 0.05 is significant
Variable
Myocarditis (n = 10)
Controls (n = 19)
Sex (Male)
9 (90 %)
9 (47 %)
Age (years)
17.5 (14–18.6)
14.5 (8.34–22)
0.2
BMI
33.1 (24.1–40.1)
23.4 (16.2–34.1)
0.0001
BSA (m2)
2.23 (1.63–2.42)
1.7 (1.03–1.85)
0.0003
Peak Troponin (ng/ml)
9.1 (1.6–36)
–
LV shortening fraction (%)
35 % (28–44)
33 % (30–40)
LV ejection fraction (%)
66 % (61–72)
64 % (58–73)
0.15
Global longitudinal strain
-14.79 ± 2.62
-17.75 ± 2.08
0.001
Global circumferential strain
-13.94 ± 3.3
-15.35 ± 2.28
0.03
CMR
10
–
Time difference between and 2D-STE and CMR
0 days (-7 to ?1 days)
–
LV end diastolic volume (ml)
189.7 ± 32.49
–
LV ejection fraction (%)
61.4 % (54–73)
–
T2-weighted
3 (30 %)
–
Positive LGE
1 (33 %) 10 (100 %)
– –
Positive
10 (100 %)
–
p value
0.16
Table 2 Echocardiography, strain, and late gadolinium enhancement parameters Subject
Abnormal circumferential strain AHA segments (number)
Peak troponin (ng/ml)
Abnormal longitudinal strain AHA segments (number)
LGE AHA segments (number)
Overlap of longitudinal strain and LGE
SF (%)
EF by CMR (%)
1
2
22.1
9
5
4
35
54
2
1
21.8
5
6
4
35
56
3
1
5.6
7
2
2
37
66
4
4
3.2
5
4
1
41
65
5
1
2.8
1
3
1
44
62
6
1
1.9
0
1
0
34
64
7
3
1.6
8
8
6
28
56
8
4
12.6
8
7
5
33
55
9
0
22.8
6
6
3
32
63
10
3
36
8
6
4
41
73
LGE late gadolinium enhancement, SF shortening fraction by m-mode, EF ejection fraction, CMR cardiac magnetic resonance imaging, AHA American Heart Association
There is moderate correlation between the overall number of LGE segments by CMR and abnormal longitudinal strain per patient (r = 0.68; p = 0.03) (Fig. 2). The correlation was weak when the number of LGE/abnormal longitudinal strain segments compared to peak troponin levels (r = 0.42; p = 0.22 and r = 0.49; p = 0.15, respectively). Global longitudinal strain by 2D-STE imaging in our cohort was reduced in 9/10 patients, and circumferential strain was also reduced in 9/10 patients at mid-ventricular level. The patients with normal longitudinal strain had abnormal circumferential strain, and vice versa (Table 2; Fig. 3).
123
There was good inter-observer agreement for longitudinal strain and LGE segments, ICC of 0.86 (95 % CI 0.83–0.89) and 0.93 (95 % CI 0.91–0.95), respectively. Intra-observer agreement for longitudinal strain and LGE was also very good with ICC 0.93 (95 % CI 0.90–0.95) and 0.96 (95 % CI 0.95–0.97), respectively. Discussion Our study shows decreased global pkS longitudinal and circumferential strain, even with normal left ventricular systolic function among children with acute myocarditis.
Pediatr Cardiol (2015) 36:950–959 Table 3 Longitudinal strain distribution by American Heart Association segments
STDEV standard deviation. n number. p \ 0.05 is significant, AHA American Heart Association AHA segments: Basal segments: 1. basal anterior, 2. basal anteroseptal, 3. basal inferoseptal, 4. basal inferior, 5. basal inferolateral, 6. basal anterolateral; mid-cavity segments: 7. mid-anterior, 8. mid-anteroseptal, 9. midinferoseptal, 10. mid-inferior, 11. mid-inferolateral, and 12. mid-anterolateral; Apical segments: 13. apical anterior, 14. apical septal, 15. apical inferior, 16. apical lateral, and 17. apex
955
AHA segment number
Cases
Controls
Mean
STDEV
1
-15.56
5.51
2
-17.97
1.83
3
-18.96
4
n
p value
Mean
STDEV
n
8
-19.67
5.9
17
0.11
9
-17.53
3.78
19
0.7448
2.96
10
-18.07
2.44
19
0.3931
-14.45
5.83
10
-20.19
3.3
19
0.0021
5
-14.75
5.74
7
-15
6.55
19
0.9299
6
-11.07
6.57
9
-12.78
4.54
19
0.4281
7
-13.35
5.98
8
-16.85
3.59
16
0.0856
8
-16.31
5.67
9
-17.98
4.84
17
0.4376
9
-18.75
2.8
10
-19.32
2.07
19
0.538
10
-15.18
3.03
10
-19.89
3.22
19
0.0007
11 12
-13.77 -10.63
5.35 6.25
9 10
-15.81 -15.31
4.53 3.64
18 19
0.3086 0.0163
13
-15.09
6.51
9
-17.16
6.61
15
0.4631
14
-17.56
4.86
10
-21.16
4.28
18
0.05
15
-16.2
6.11
7
-21.49
3.7
18
0.0138
16
-13.29
6.53
10
-18.88
4.53
19
0.0116
17
-12.5
5.62
8
-15.33
6.69
16
0.316
Table 4 Circumferential strain distribution by segments at mid-ventricular level (papillary muscle) AHA segment number
Cases
Controls
p value
Mean
STDEV
n
Mean
STDEV
n
7
-17.17
5.76
10
-15.05
6.09
16
0.387
8
-20.27
6.19
10
-19.92
4.19
16
0.8645
9
-15.14
6.01
10
-20.44
5.27
16
0.0265
10
-9.65
4.55
9
-15.83
5.92
16
0.0126
11
-7.85
5.28
10
-10.25
9.36
16
0.4681
12
-12.18
6.72
9
-13.49
5.29
15
0.6007
STDEV standard deviation. n number. p \ 0.05 is significant, AHA American Heart Association AHA mid-cavity segments: 7. mid-anterior, 8. mid-anteroseptal, 9. mid-inferoseptal, 10. mid-inferior, 11. mid-inferolateral, and 12. midanterolateral
Our finding of reduced global longitudinal and circumferential strain in acute myocarditis confirms earlier reports by Khoo et al. [15] in adolescents and young adults and by Hsiao et al. [14] and Di Bella et al. [8] in adults. Myocarditis is identified by CMR as areas of hyperenhancement (LGE) preferentially involving subepicardial, intra-myocardial regions of the basal and/or mid-lateral cardiac segments (Fig. 4), and similar areas might show hyper-enhancement suggestive of fluid/edema by T2weighted imaging [6] (Fig. 4). Interestingly, we noted significantly lower regional longitudinal strain predominantly involving basal, mid-and apical inferior and anterolateral segments, thus suggesting a possibility of regional
inflammation altering the regional myocardial function that can be identified easily using strain (2D-STE) imaging, in spite of preserved systolic function. LGE by CMR is prone to subjective bias and is especially difficult to interpret in the apical regions where the spatial resolution is low. This makes LGE by CMR alone a less robust tool to diagnose myocarditis in subjects with no regional wall motion abnormalities and other inconclusive CMR parameters (T2-weighted imaging, first-pass perfusion, and early gadolinium enhancement) (Fig. 5). CMR can be cumbersome and time-consuming especially in children less than 10 years of age with acute myocarditis, who may not be as cooperative during the
123
956 Fig. 3 Graph showing the distribution 18 of abnormal longitudinal, circumferential strain, and late 16 gadolinium enhancement (LGE) 14 segments by American heart 12 association segmentation (AHA segments: Basal segments: 1. basal 10 anterior, 2. basal anteroseptal, 3. basal inferoseptal, 4. basal inferior, 5. basal 8 6 inferolateral, 6. basal anterolateral; mid-cavity segments: 7. mid-anterior, 4 8. mid-anteroseptal, 9. midinferoseptal, 10. mid-inferior, 11. mid- 2 inferolateral, and 12. mid0 anterolateral; Apical segments: 13. apical anterior, 14. apical septal, 15. apical inferior, 16. apical lateral, and 17. apex)
Pediatr Cardiol (2015) 36:950–959
Segemental involvement
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Circumferenal Strain (segments 7-12) Longitudinal Strain MDE
Fig. 5 Same patient as in Fig. 4 with areas of abnormal longitudinal strain by 2D-STE (light red to blue areas). Compare the abnormal longitudinal strain areas with areas of LGE and positive T2-weighted areas on the previous image. Normal longitudinal strain in the base and more areas of strain abnormality in the apical segments (7, 8, 12, 13, 14, 15, 16, and 17). The number in each AHA segment measures peak systolic longitudinal strain
Fig. 4 Patient with acute myocarditis showing areas of late gadolinium enhancement (LGE) (blue arrows) and positive T2-weighted imaging (red arrows). a and g—vertical long axis; b—horizontal long axis; c, d, e—LGE imaging at base, mid-ventricular and apex; f—T2weighted imaging at base. Involved segments by CMR: 4, 5, 11,12,15,16, and 17
CMR scan and thus further reducing the efficacy of this modality alone. We showed overlap among the areas with reduced longitudinal strain and areas of delayed enhancement (LGE) by CMR in 51 % of the segments. This might be due to myocardial injury/inflammation that affects both regional strain and contrast stagnation resulting in a positive LGE. To our surprise, there are more areas of reduced
123
longitudinal strain compared to areas of LGE (57 vs. 48; 29 segments overlap) with moderate correlation between the overall number of LGE segments by CMR and 2D-STE per patient (r = 0.68; p = 0.03) (Fig. 2), although there was no segmental overlap 49 % of times for all patients, suggesting that LGE and 2D-STE technologies likely were identifying different pathological processes. Interestingly, CMR appears to under-diagnose affected myocardium in cardiac apical segments (13–17) compared to 2D-STE (18 vs. 35 %; p = 0.006) (Fig. 6). CMR studies [2, 18] in adults have shown satisfactory accuracy of CMR for acute myocarditis with sensitivity; specificity; and accuracy of 81, 71, and 79 %, respectively. No such studies exist for pediatric population. There is significant practice variation performing CMR’s among
Pediatr Cardiol (2015) 36:950–959
pediatric centers [3] with many centers performing LGE routinely and fewer centers performing first-pass perfusion and early gadolinium enhancement. There is the absence of uniform consensus even among adult centers where LGE is used routinely; some groups advocate inclusion of firstpass perfusion, contrast-enhanced T1-weighted imaging, T1-mapping, T2-weighted imaging, and/or extracellular volume fraction to improve the accuracy [2]. These modalities are prone to image quality and reproducibility issues and are not routinely adopted in the clinical practice. CMR alone to diagnose myocarditis has technical limitations that include trigger problems, artifacts, and poor resolution which might interfere with obtaining diagnostic quality images and accurate interpretation [2]. CMR in younger noncooperative children is even more resource intensive and might require sedation, and/or anesthesia to obtain satisfactory diagnostic images. CMR does have an advantage in obtaining high-quality images even in a patient with poor acoustic windows. Tissue characterization techniques are being increasingly adopted and might have a higher diagnostic yield. Automated post-processing techniques are being developed, which are more objective, thus providing unbiased and consistent results. On the contrary, cardiac strain imaging is less resource intensive, less expensive, portable, easily performed even in a moderately cooperative child by a skilled operator providing diagnostic quality images. 2D-STE is adopted by different vendors making it omnipresent [20]; there is an increased trend using this technology even among children
Fig. 6 AHA 17-segment models—for top two figures, each number within the segment represents number of subjects with involvement of that particular segments comparing late gadolinium enhancement (LGE) and reduced longitudinal strain 2D-STE
957
[15]. Post-processing of 2D-STE is automated requiring minimal interpreter adjustment; however, there is difference among the ‘‘normal’’ values between the vendors [20, 21] that has to be taken into account while interpreting these results. It is known that LV myocardium is mostly composed of myocardial fibers that are circumferentially oriented (*59 %) parallel to the mitral valve orifice; subepicardial fibers that are oriented obliquely and cross the interventricular groove; subendocardial longitudinal fibers (20 %) that insert into the aortic and mitral valves [13]. Myocarditis predominantly affects subepicardial and intramyocardial regions, thus effecting oblique and circumferentially oriented fibers, and involvement of these fibers can be recognized by as abnormal strain on 2D-STE. Left ventricular function is usually preserved as long as the involvement of myocardium is minimal, thus making it difficult to recognize regional wall motion abnormalities in early/minimal myocarditis by traditional echocardiography alone. It is interesting to note that pkS longitudinal strain is globally reduced in spite of subendocardial sparing by CMR in all the subjects, suggesting a global myocardial involvement that is not readily identified by current CMR techniques. Ease of access to 2D-STE imaging also helps to follow these patients serially to evaluate the progression and thus helping with the management plan. We have shown a correlation between the segments of LGE and abnormal 2D-STE; as was reported in previous case reports [1, 12] and although this correlation is modest, it needs to be further evaluated in a prospective study in order to establish its significance. Our correlation might have been affected by the small sample size. It is interesting to note that 9/10 patients were obese, even in comparison with the age-matched controls. Recent evidence shows reduced global longitudinal strain in obese children, in spite of normal left ventricular ejection fraction [4]. Whether the reduced strain in our cohort is related to the obesity, myocarditis, or a combination is hard to determine. It also raises an important question whether obesity predisposes or increases the chance of myocardial damage when exposed to environmental/viral agents. Global longitudinal and circumferential strain values among controls in our laboratory were lower than those of reported by Marcus et al. in 21 subjects aged 15–19 [19] (-17.78 ± 2.19 vs. -20.14 ± 1.6; p = 0.0004 and -15.6 ± 2.2 vs. -23.6 ± 2; p = 0.0001), even though we used similar methods, and vendors for the data collection and analysis; We believe that this difference could be as a result of demographic variation of the sample itself (USA vs. Israel and Netherlands), which brings out an important point on deciding what normal values to use. Due to paucity of normal data for children, we recommend using values obtained in the local laboratory.
123
958
It is not our institutional policy to routinely perform EMB in patients presenting with myocarditis. Histopathological exam when combined with CMR have a higher diagnostic yield of *95 % than compared to CMR or EMB alone (80 and 88 %, respectively) [2]. Baccouche et al. [2] have recommended routinely performing CMR and reserving EMB if the diagnosis is unclear or the symptoms persist. CMR also has a limitation to identify the degree of inflammation or identify special forms of myocarditis (eosinophilic or giant cell etc.). Using peak systolic global longitudinal strain by 2DSTE imaging in our cohort, the sensitivity was 90 %, and 9/10 patients had low global longitudinal and circumferential strain values. The patient with normal longitudinal strain had reduced circumferential strain (segment 11) and a positive LGE in segment 7 with mildly elevated troponin (1.9 ng/ml). Limitations Our study shares the limitations inherent to many other retrospective designs; it was a single-institution study with a small study population, subjects were mostly male, and thus gender-related effects could not be excluded. As previously discussed, majority of our subjects were obese and the effect of obesity on our strain and LGE measures could not be further investigated given the small sample. Our study contributes to the small pool of reported studies comparing CMR in myocarditis with preserved ventricular function and myocardial strain abnormalities [1, 8, 12] and is unique by being exclusively in pediatric subjects. Our study also reports the issues using the current age-matched normals reported in the literature [19] that were collected in a different demographic setting and proposes for a study in a larger population to obtain a normative data that can be widely used. We could not test the influence of regional myocardial edema on regional strain, as T2-weighted imaging was performed in only three subjects. Lack of histopathological correlation is another limitation. The correlation between LGE and 2D-STE could have been stronger if we had performed both tests within a short time period closer to the peak troponin levels, thus correlating imaging to the underlying myocardial involvement. This hypothesis will need further testing in a larger prospectively designed study to better understand the relation between the imaging modalities. As 2D-STE is becoming more universally available and accepted, studies involving various vendors and post-processing software need to be performed on a larger scale in children to obtain vendor-specific normative data and cutoffs. There is an ongoing discussion about the subjective bias involved in identifying areas of LGE by CMR, future
123
Pediatr Cardiol (2015) 36:950–959
protocols like T1 mapping; extracellular volume fraction may prove useful by making analysis more objective, but this needs to be tested prior to their universal acceptance.
Conclusion LV myocardial segmental involvement by acute myocarditis in patients with preserved LV systolic function can be identified by both 2D-STE and the presence of LGE by CMR. 2D-STE is more objective and is better than CMR in identifying abnormalities in the apical regions. 2D-STE assessment for myocardial involvement in acute myocarditis can serve as a useful noninvasive adjunct to the existing tests used for the diagnosis of acute myocarditis and might play a very important role in serial evaluation. Further studies are required to obtain age-specific normative data that can be widely used, and across all vendor platforms. 2D-STE, in addition to CMR, may serve as a clinically useful surrogate for risk stratification and prospective evaluation of patients with acute myocarditis. The findings of this study support need for future studies in a larger sample to establish their clinical utility.
References 1. Afonso L, Hari P, Pidlaoan V, Kondur A, Jacob S, Khetarpal V (2010) Acute myocarditis: can novel echocardiographic techniques assist with diagnosis? Eur J Echocardiogr 11:E5 2. Baccouche H, Mahrholdt H, Meinhardt G, Merher R, Voehringer M, Hill S, Klingel K, Kandolf R, Sechtem U, Yilmaz A (2009) Diagnostic synergy of non-invasive cardiovascular magnetic resonance and invasive endomyocardial biopsy in troponinpositive patients without coronary artery disease. Eur Heart J 30:2869–2879 3. Banka P, Uppu S, Matthew, Hasbani K, Lai W, Richmond M, Fratz S, Jain S, Johnson T, Maskatia S, Lu J, Samyn M, Patton D, Powell A (2014) CMR techniques and findings in children with myocarditis: a multicenter retrospective study. J Cardiovasc Magn Reson. doi:10.1186/1532-429X-16-S1-P119 4. Barbosa JA, Mota CC, Simo˜es E, Silva AC, Nunes MoC, Barbosa MM (2013) Assessing pre-clinical ventricular dysfunction in obese children and adolescents: the value of speckle tracking imaging. Eur Heart J Cardiovasc Imaging 14:882–889 5. Barlow SE (2007) Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report. Pediatrics 120(Suppl 4):S164–S192 6. Camastra GS, Cacciotti L, Marconi F, Sbarbati S, Pironi B, Ansalone G (2007) Late enhancement detected by cardiac magnetic resonance imaging in acute myocarditis mimicking acute myocardial infarction: location patterns and lack of correlation with systolic function. J Cardiovasc Med (Hagerstown) 8:1029–1033 7. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, Pennell DJ, Rumberger JA, Ryan T, Verani MS, Imaging AHAWGoMSaRfC (2002) Standardized myocardial
Pediatr Cardiol (2015) 36:950–959
8.
9.
10.
11.
12.
13.
14.
15.
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. Int J Cardiovasc Imaging 18:539– 542 Di Bella G, Gaeta M, Pingitore A, Oreto G, Zito C, Minutoli F, Anfuso C, Dattilo G, Lamari A, Coglitore S, Carerj S (2010) Myocardial deformation in acute myocarditis with normal left ventricular wall motion–a cardiac magnetic resonance and 2-dimensional strain echocardiographic study. Circ J 74: 1205–1213 Fratz S, Chung T, Greil GF, Samyn MM, Taylor AM, Valsangiacomo Buechel ER, Yoo SJ, Powell AJ (2013) Guidelines and protocols for cardiovascular magnetic resonance in children and adults with congenital heart disease: SCMR expert consensus group on congenital heart disease. J Cardiovasc Magn Reson 15:51 Geyer H, Caracciolo G, Abe H, Wilansky S, Carerj S, Gentile F, Nesser HJ, Khandheria B, Narula J, Sengupta PP (2010) Assessment of myocardial mechanics using speckle tracking echocardiography: fundamentals and clinical applications. J Am Soc Echocardiogr 23:351–369 quiz 453–355 Goitein O, Matetzky S, Beinart R, Di Segni E, Hod H, Bentancur A, Konen E (2009) Acute myocarditis: noninvasive evaluation with cardiac MRI and transthoracic echocardiography. AJR Am J Roentgenol 192:254–258 Ha SJ, Woo JS, Kwon SH, Oh CH, Kim KS, Bae JH, Kim WS (2013) Acute regional myocarditis with normal ventricular wall motion diagnosed by two-dimensional speckle tracking imaging. Korean J Intern Med 28:732–735 Ho SY (2009) Anatomy and myoarchitecture of the left ventricular wall in normal and in disease. Eur J Echocardiogr 10:iii3– iii7 Hsiao JF, Koshino Y, Bonnichsen CR, Yu Y, Miller FA, Pellikka PA, Cooper LT, Villarraga HR (2013) Speckle tracking echocardiography in acute myocarditis. Int J Cardiovasc Imaging 29:275–284 Khoo NS, Smallhorn JF, Atallah J, Kaneko S, Mackie AS, Paterson I (2012) Altered left ventricular tissue velocities,
959
16.
17.
18.
19.
20.
21.
22.
23.
deformation and twist in children and young adults with acute myocarditis and normal ejection fraction. J Am Soc Echocardiogr 25:294–303 Kindermann I, Barth C, Mahfoud F, Ukena C, Lenski M, Yilmaz A, Klingel K, Kandolf R, Sechtem U, Cooper LT, Bo¨hm M (2012) Update on myocarditis. J Am Coll Cardiol 59:779–792 Leitman M, Lysyansky P, Sidenko S, Shir V, Peleg E, Binenbaum M, Kaluski E, Krakover R, Vered Z (2004) Two-dimensional strain-a novel software for real-time quantitative echocardiographic assessment of myocardial function. J Am Soc Echocardiogr 17:1021–1029 Lurz P, Eitel I, Adam J, Steiner J, Grothoff M, Desch S, Fuernau G, de Waha S, Sareban M, Luecke C, Klingel K, Kandolf R, Schuler G, Gutberlet M, Thiele H (2012) Diagnostic performance of CMR imaging compared with EMB in patients with suspected myocarditis. JACC Cardiovasc Imaging 5:513–524 Marcus KA, Mavinkurve-Groothuis AM, Barends M, van Dijk A, Feuth T, de Korte C, Kapusta L (2011) Reference values for myocardial two-dimensional strain echocardiography in a healthy pediatric and young adult cohort. J Am Soc Echocardiogr 24:625–636 Mor-Avi V, Lang RM, Badano LP, Belohlavek M, Cardim NM, Derumeaux G, Galderisi M, Marwick T, Nagueh SF, Sengupta PP, Sicari R, Smiseth OA, Smulevitz B, Takeuchi M, Thomas JD, Vannan M, Voigt JU, Zamorano JL (2011) Current and evolving echocardiographic techniques for the quantitative evaluation of cardiac mechanics: ASE/EAE consensus statement on methodology and indications endorsed by the Japanese Society of Echocardiography. Eur J Echocardiogr 12:167–205 Risum N, Ali S, Olsen NT, Jons C, Khouri MG, Lauridsen TK, Samad Z, Velazquez EJ, Sogaard P, Kisslo J (2012) Variability of global left ventricular deformation analysis using vendor dependent and independent two-dimensional speckle-tracking software in adults. J Am Soc Echocardiogr 25:1195–1203 Rosset A, Spadola L, Ratib O (2004) OsiriX: an open-source software for navigating in multidimensional DICOM images. J Digit Imaging 17:205–216 Sagar S, Liu PP, Cooper LT (2012) Myocarditis. Lancet 379:738–747
123
