Study Population

This retrospective study consisted of 70 subjects (mean age, 62.4 ± 13.3 years; 49 men) who underwent layer-specific 2D STE and LGE CMR at our institution.

We excluded patients in whom the interval between 2D STE and LGE CMR exceeded 6 months. Most of the examinations were done with a time interval of slightly less than 3 months between 2D STE and LGE CMR, and the median time between the examinations was 73 days.

Forty-nine patients were affected by chronic ischemic heart disease, while 21 control subjects with normal values on 2D STE and LGE CMR and without any history of ischemia were selected on the basis of normal LGE CMR reports collected at our institution. LGE CMR examination of control subjects was prescribed for different clinical reasons but mainly for research of the foci of fibrosis or dysplasia in healthy subjects with unexplained ventricular extrasystole or arrhythmia.

Absence of sinus rhythm as well as myocardial infarction or percutaneous coronary revascularization between 2D STE and LGE CMR or <3 months before the first examination were exclusion criteria. We also excluded patients with suboptimal echocardiographic views.

Echocardiography

Echocardiography was performed using a Vivid E9 System (GE Vingmed Ultrasound AS, Horten, Norway) equipped with a 5–4.6 MHz transducer active-matrix single-crystal phased array (M5S). Standard parasternal long-axis and short-axis views at the basal, midventricular, and apical levels and in the apical four-, three-, and two-chamber views were acquired with a frame rate of 56 to 92 frames/sec using tissue harmonic imaging.

Speckle-Tracking Analysis

Echocardiographic images from the apical and parasternal short-axis views (basal, midventricular, and apical levels) were stored in digital cine loop format. Peak negative end-systolic LS and CS from the LSEN/CSEN, LSEP/CSEP, and full wall thickness were assessed using an offline software package (EchoPAC version BT11; GE Vingmed Ultrasound AS). The analysis of myocardial deformation according to strain was evaluated on a frame-by-frame basis by automatic tracking of acoustic markers (speckles) that are identified and followed within consecutive 2D echocardiographic images and are expected to change their positions in accordance with the surrounding tissue motion. Loops of three consecutive cardiac cycles were acquired, and deformation parameters were determined as an average of the three loops.

The system automatically determines tracking quality, and segments with suboptimal tracking quality were excluded from the analysis. The endocardial and epicardial borders were automatically traced in the end-systolic frame of the 2D images from the four-, three-, and two-chamber apical views for the analysis of LS and from the parasternal short-axis views for the analysis of CS. In cases of disagreement by the observer regarding the automatically defined borders, manual correction was applied for endocardial-like epicardial borders. Segments with insufficient quality despite manual correction were excluded.

The system allowed separate analysis of CS and LS peak end-systolic and peak postsystolic strains of the LSEN/CSEN, LSEP/CSEP, and full wall thickness outlined by the tracked borders of each LV segment.

The measurements presented for the LSEN/CSEN and LSEP/CSEP layers are the values at the inner- and outer-tracked region-of-interest lines, which represent the endocardial and epicardial border, respectively. Thus, according to the definition of endocardial and epicardial contours, the layers were automatically defined by the system.

End-systole was defined as aortic valve closure in the apical axis view and transferred to all other views. Peak end-systolic strain is the peak before or at the time of aortic valve closure, while peak postsystolic strain is peak systolic strain independent of aortic valve closure.

All views were analyzed using the system to obtain quantitative function parameters for each segment in an 18-segment LV model (six segments at each LV level). Segmental speckle-tracking echocardiographic data were matched with CMR data using the same 18-segment model.

Late Gadolinium-Enhanced Cardiac Magnetic Resonance

All patients were examined in the supine position using a 3-T scanner (Magnetom Skyra; Siemens Healthcare, Erlangen, Germany) with a 32-channel phased-array coil while holding their breath.

After localization, cine sequences were acquired using balanced steady-state free precession sequences with long- and short-axis images and projections comparable with those used in cardiac ultrasonography. Short-axis acquisitions were performed using multiple slices to cover both ventricles entirely from the base to the apex (field of view, 350 × 295; matrix size, 216 × 256; acceleration factor, 2 generalized autocalibrating partial parallel acquisition; slice thickness, 8 mm; interslice gap, 20%; number of phases, 25). To evaluate myocardial scars, late enhancement acquisitions (2D phase-sensitive segmented, inversion recovery gradient echo) were performed 10 min after an intravenous bolus injection of 0.2 mmol/kg gadolinium-based contrast medium. Positive hyperenhancement was visually defined as >2 SDs of the signal intensity of the nonenhanced myocardium.

Short- and long-axis slices (8 mm) were acquired on a prospective electrocardiographically gated gradient-echo sequence with inversion prepulse using the same slices of cine sequences (field of view, 330 × 330; matrix size, 192 × 256; acceleration factor, 2 generalized autocalibrating partial parallel acquisition; slice thickness, 8 mm; interslice gap, 20%; inversion time set according to scout images). To evaluate the scar extent, the same 18-segment LV model used for the echocardiographic analysis was applied. To align the echocardiographic and LGE CMR segments, anatomic landmarks (i.e., the papillary muscles and LV outflow tract orientation) were used. Each myocardial segment was evaluated for the presence of hyperenhancement; 0% hyperenhancement was considered to be normal, 1% to 50% hyperenhancement was considered to reflect SES, and 51% to 100% hyperenhancement was considered to reflect TMS.

We differentiated TMS from SES using a 50% cutoff value for the extent of TMS because the recovery of scar segments after revascularization above this percentage is poor.

Statistical Analysis

Computations were performed using Stata version 14.1 (StataCorp, College Station, TX); a two-tailed P value < .05 was considered to indicate statistical significance. Data are expressed as mean ± SD if continuous and as counts and percentages if categorical. Comparisons between myocardial scar groups were performed using linear regression models with Huber-White robust standard errors to account for heteroscedasticity. The unit of analysis was the patient or the segment. In the latter case, regression models included calculation of Huber-White robust standard errors while clustering on the patient to account for intrapatient measurement correlations. The mean difference and 95% CI were derived from the model. The following groups were compared: (1) no scar versus any scar, followed by post hoc comparisons of (1a) no scar versus SES and (1b) no scar versus TMS, and (2) SES versus TMS. Receiver operating characteristic (ROC) curve analysis was used to identify the optimal cutoff for the type of scar that maximizes sensitivity and specificity. The area under the ROC curve (AUC) was used as an overall measure of discrimination. The closer the AUC to 1, the greater the discrimination; an AUC of 0.50 corresponds to no discrimination.

Patient Characteristics

The study population consisted of 70 subjects (mean age, 62.4 ± 13.3 years; 49 men; 21 healthy control subjects). The mean ejection fraction defined by biplane echocardiography was 51.2 ± 13.1%, and the mean ejection fraction defined by CMR was 50.8 ± 14.6%. All patients were in sinus rhythm at the time of the examinations (mean heart rate, 66.6 ± 13 beats/min). Scar was visible on LGE CMR images in 49 patients (SES in 24 [55.9%], TMS in 19 [44.2%], and both SES and TMS in six [8.6%]). A total of 21 patients without LGE CMR findings of scar and histories of ischemic heart disease as described in our database were considered the control group. The mean age of normal control subjects was 54.5 ± 12.1 years, 55.0% were men, 30.0% had histories of hypertension, and none had histories of diabetes, renal failure, or myocardial infarction or had undergone percutaneous coronary intervention. The patients affected by chronic ischemic heart disease had a mean age of 65.4 ± 9.7 years, 81.6% were men, 16.3% had diabetes, 79.6% had histories of hypertension, and none had histories of renal failure. The baseline clinical, echocardiographic, and LGE CMR data of the 70 subjects are shown in Table 1 .

Cardiac Magnetic Resonance

In total, 1,260 segments were analyzed: 1,036 normal, 116 with SES (1%–50% hyperenhancement), and 108 with TMS (51%–100% hyperenhancement). The distribution of segments with scar (SES and TMS) was as follows: 79 apical (37 SES, 42 TMS), 80 midventricular (40 SES, 40 TMS), and 65 basal (39 SES, 26 TMS).

Normal Myocardial versus Scar Segments

The analysis included 1,020 normal and 219 scar (113 SES, 106 TMS) myocardial segments for LS and 830 normal and 171 scar (91 SES, 80 TMS) segments for CS. Twenty-one segments (2%) for LS and 259 segments (21%) for CS were excluded because of suboptimal echocardiographic tracking quality.

In Tables 2 and 3 , peak end-systolic ( Table 2 ) and peak postsystolic ( Table 3 ) LS and CS values of normal segments are compared with those of scar segments (SES and TMS).

Table 3

Peak postsystolic strain measurements related to hyperenhancement by LGE CMR in normal versus scar (SES and TMS)

Parameter	Level	No scar, mean ± SD	Scar (SES and TMS), mean ± SD	Difference	P	Cutoff	AUC	Sensitivity	Specificity
CS full wall thickness		−19.7 ± 8.5%	−10.3 ± 6.2%	8.9	<.001	−14.9%	0.73	0.77	0.70
	Basal	−17.0 ± 7.4%	−10.9 ± 5.2%	5.9	<.001	−13.2%	0.72	0.77	0.68
	Midventricular	−18.6 ± 7.7%	−10.3 ± 5.9%	7.6	<.001	−15.7%	0.73	0.81	0.64
	Apical	−24.4 ± 9.0%	−10.4 ± 7.8%	14.4	<.001	−16.8%	0.82	0.81	0.82
CSEP		−14.5 ± 6.9%	−7.42 ± 4.7%	6.5	<.001	−10.6%	0.74	0.79	0.69
	Basal	−12.4 ± 6.2%	−7.1 ± 4.4%	4.2	<.001	−10.0%	0.72	0.77	0.66
	Midventricular	−13.5 ± 6.1%	−7.5 ± 4.3%	5.5	<.001	−12.1%	0.72	0.88	0.56
	Apical	−18.3 ± 7.1%	−7.7 ± 5.6%	11.0	<.001	−10.7%	0.82	0.77	0.87
CSEN		−27.2 ± 11.3%	−14.1 ± 9.6%	12.7	<.001	−16.2%	0.74	0.65	0.83
	Basal	−23.2 ± 9.4%	−13.2 ± 9.1%	9.2	<.001	−15.9%	0.73	0.70	0.76
	Midventricular	−25.6 ± 9.8%	−14.4 ± 8.67%	10.6	<.001	−16.0%	0.73	0.63	0.83
	Apical	−34.1 ± 12.3%	−14.7 ± 11.6%	19.9	<.001	−25.9%	0.81	0.84	0.77
LS full wall thickness		−17.1 ± 6.2%	−10.8 ± 5.9%	7.0	<.001	−14.4%	0.72	0.76	0.70
	Basal	−13.7 ± 4.8%	−8.7 ± 5.7%	6.2	<.001	−13.2%	0.71	0.84	0.57
	Midventricular	−15.9 ± 4.9%	−10.8 ± 5.5%	5.1	<.001	−14.5%	0.73	0.79	0.67
	Apical	−21.0 ± 6.9%	−10.47 ± 6.5%	10.8	<.001	−15.8%	0.81	0.83	0.79
LSEP		−14.6 ± 5.3%	−9.2 ± 5.2%	5.8	<.001	−12.8%	0.73	0.79	0.67
	Basal	−13.7 ± 4.8%	−10.1 ± 5.1%	4.1	<.001	−12.8%	0.68	0.75	0.61
	Midventricular	−14.3 ± 4.6%	−9.6 ± 5.0%	4.7	<.001	−12.35%	0.71	0.72	0.71
	Apical	−15.7 ± 6.1%	−7.9 ± 5.2%	8.0	<.001	−10.8%	0.79	0.75	0.83
LSEN		−20.1 ± 8.9%	−12.8 ± 7.4%	8.2	<.001	−14.81%	0.67	0.62	0.71
	Basal	−15.2 ± 4.9%	−12.2 ± 6.2%	4.4	<.001	−13.3%	0.62	0.59	0.65
	Midventricular	−17.5 ± 5.5%	−11.0 ± 6.2%	5.6	<.001	−14.8%	0.70	0.68	0.73
	Apical	−27.8 ± 9.9%	−14.2 ± 8.9%	13.9	<.001	−21.01%	0.79	0.81	0.78

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Ascending Aortic Dimension Measurements: Apples, Oranges, and Lemons
2. Diagnostic Accuracy of Transthoracic Echocardiography for Infective Endocarditis Findings Using Transesophageal Echocardiography as the Reference Standard: A Meta-Analysis
3. The Effect of the Superior Cavopulmonary Anastomosis on Ventricular Remodeling in Infants with Single Ventricle
4. Evaluating Appropriate Use of Pediatric Echocardiograms for Chest Pain in Outpatient Clinics

Stay updated, free articles. Join our Telegram channel

Join