Cardiac resynchronization therapy (CRT) is an established therapeutic option for patients with advanced heart failure (HF). Previous studies have demonstrated that CRT reduces HF symptoms and improves acute and chronic hemodynamics, exercise capacity, the quality of life, survival, and left ventricular (LV) systolic function. Randomized clinical trials have demonstrated, however, that up to one third of patients with HF do not favorably respond to CRT with standard clinical selection criteria, including electrocardiographic QRS width. Accordingly, the quantification of LV mechanical dyssynchrony by echocardiography has emerged as an important option for predicting response to CRT. Although LV mechanical dyssynchrony has been associated with response to CRT, some patients with LV mechanical dyssynchrony assessed by echocardiography do not respond to CRT. Other factors, such as LV lead position or scar burden, may also influence response to CRT, regardless of the presence of dyssynchrony. Previous investigators have reported that myocardial viability or scar tissue is an important marker for response to CRT. However, the impact of regional myocardial systolic function assessed by echocardiography in the posterolateral wall, where the LV pacing lead is positioned for response to CRT, is not known. Automated function imaging (AFI), based on two-dimensional speckle (2D) tracking imaging, can be used for the assessment of regional longitudinal strain of the 17-segment left ventricle as a bull’s-eye map.

Our objective was thus to test the impact of posterolateral myocardial systolic function assessed by AFI on the response to CRT. Furthermore, we evaluated the hypothesis that the addition of posterolateral myocardial systolic function to LV dyssynchrony assessed by echocardiography would further improve the ability to predict response to CRT.

Methods

Study Population

We studied 47 consecutive patients with HF who underwent CRT. The selection criteria for CRT included chronic severe HF (New York Heart Association functional class III or IV), LV ejection fraction (EF) ≤ 35%, and QRS duration ≥ 120 ms. Seven patients were excluded from subsequent analysis because their echocardiographic images were technically unsatisfactory. Accordingly, the patient study group consisted of 40 patients with HF, 28 (70%) in New York Heart Association functional class III and 12 (30%) in class IV at the initial evaluation. The group’s mean age was 67 ± 12 years, 10 (25 %) were women, the mean EF was 25 ± 8%, the mean QRS duration was 163 ± 25 ms, and 9 patients (23%) had ischemic cardiomyopathy ( Table 1 ). Twenty-four patients were diagnosed with sinus rhythm and 6 with atrial fibrillation, while 10 had previously undergone the implantation of permanent right ventricular pacemakers ≥1 year before enrollment, and they were predominantly right ventricular paced, which was defined as ≥90% paced when the device was interrogated at the time of enrollment. All patients were on optimal pharmacologic therapy, if tolerated. Written informed consent to participate in the study was obtained from all patients.

Table 1

Baseline characteristics of patients and their responses to CRT

	All patients	Responders	Nonresponders
Variable	(n = 40)	(n = 30)	(n = 10)
Age (y)	67 ± 12	68 ± 12	66 ± 10
Men/women	30/10	22/8	8/2
NYHA class (III/IV)	28/12	22/8	6/4
QRS duration (ms)	163 ± 25	164 ± 25	161 ± 26
Left bundle branch block	22 (55%)	19 (63%)	3 (30%)
Rhythm (SR/AF/paced)	24/6/10	19/3/8	5/3/2
HR (beats/min)	73 ± 13	73 ± 13	74 ± 11
PR duration (ms)	188 ± 29	187 ± 28	191 ± 35
Posterior wall thickness (mm)	10 ± 2	10 ± 2	10 ± 2
Mitral regurgitation grade (0/1/2/3)	27/12/1/0	21/9/0/0	6/3/1/0
Etiology (ischemic/nonischemic)	9/31	6/24	3/7
Medication
Diuretics	34 (85%)	24 (80%)	10 (100%)
ACE inhibitors/ARBs	37 (93%)	28 (93%)	9 (90%)
β-blockers	33 (83%)	25 (83%)	8 (80%)
Ejection fraction (%)
Baseline	25 ± 8	25 ± 9	26 ± 6
Follow-up	35 ± 14	39 ± 14 ∗	24 ± 5
End-diastolic volume (mL)
Baseline	170 ± 77	154 ± 56	217 ± 109
Follow-up	139 ± 80	113 ± 51 ∗	216 ± 104
End-systolic volume (mL)
Baseline	131 ± 67	119 ± 53	166 ± 95
Follow-up	95 ± 69	72 ± 44 ∗	165 ± 82
LV dyssynchrony (ms)
Baseline	258 ± 154	272 ± 154	216 ± 151
Follow-up	81 ± 68	53 ± 39 ∗	168 ± 66

 

ACE , Angiotensin-converting enzyme; AF , atrial fibrillation; ARB , angiotensin type 1 receptor blocker; NYHA , New York Heart Association; SR , sinus rhythm.

Data are presented as mean ± SD or as number (percentage).

∗ P < .05 vs baseline.

Echocardiography

All echocardiographic studies were performed with a commercially available echocardiographic system (Vivid 7; GE-Vingmed Ultrasound AS, Horten, Norway). Patients were studied before and 4 ± 2 months after CRT. Echocardiographic images were obtained in the left lateral decubitus position with a 3.5-MHz transducer. The mean frame rate was 65 ± 15 frames/s for grayscale imaging used for speckle-tracking analysis. Sector width was optimized for complete myocardial visualization while maximizing the frame rate. Gain settings were adjusted for routine grayscale 2D imaging to optimize endocardial definition. Pulsed-wave Doppler of the LV outflow tract was obtained to determine LV ejection phase. The end-diastolic wall thickness of the posterior wall was measured using routine grayscale 2D imaging. LV end-diastolic volume, LV end-systolic volume, and EF were calculated from the apical 2-chamber and 4-chamber images using the biplane Simpson’s technique. For patients with atrial fibrillation, measurements of standard echocardiographic and speckle-tracking parameters were obtained as the averages of ≥3 consecutive cardiac cycles. Response to CRT was defined as reverse remodeling detected by a relative decrease in end-systolic volume ≥ 15% from baseline. Patients were defined as nonresponders if they did not show the above prespecified echocardiographic changes. Mitral regurgitation was visually assessed on the basis of the ratio of regurgitant jet area to left atrial area as none (grade 0), mild (grade 1), moderate (grade 2), or severe (grade 3).

LV Dyssynchrony Analysis

Speckle tracking of routine grayscale mid-LV short-axis images was performed as previously described to assess LV dyssynchrony. The measurements were performed offline using dedicated software (EchoPAC version BTO6; GE-Vingmed Ultrasound AS). Briefly, an end-systolic circular region of interest was traced on the endocardial cavity using a point-and-click approach, with special care taken to adjust tracking of all endocardial segments. A second and larger concentric circle was then automatically generated and manually adjusted near the epicardium. Speckle tracking automatically analyzed frame-by-frame movement of the stable patterns of natural acoustic markers, or speckles, over the cardiac cycle. Significant LV dyssynchrony was defined as a time difference > 130 ms between the anteroseptal and posterior wall peak strain ( Figure 1 ).

An example of a 2D midventricular short-axis image demonstrating radial time strain curves in a patient with HF with left bundle branch block. Dyssynchrony is shown as the time difference (arrow) between the peak strain in the anterior septum (yellow curve) and in the posterior wall (purple curve) .

Regional Myocardial Functional Analysis

Regional myocardial function was quantified using AFI. The measurements were performed offline using dedicated software (EchoPAC version BTO6). This technique was based on 2D speckle-tracking imaging that can be used for the assessment of regional longitudinal strain of the left ventricle and was performed as previously described in detail. Briefly, the mitral annulus and LV apex with 3 index points in the standard 3 apical views were defined at end-systole. The LV end-systolic frame was defined in the apical long-axis view, and the closure of the aortic valve was marked. The time interval between the R wave and aortic valve closure was used as a reference for the 4-chamber and 2-chamber view loops. The software then automatically detected the endocardium and tracked myocardial motion during the entire cardiac cycle. The left ventricle was automatically divided into 6 segments for each apical view, and the peak systolic longitudinal strain was displayed for each plane, after which the results for all 3 images were combined into a single bull’s-eye summary using a 17-segment model ( Figure 2 ), which was automatically translated from the 18-segment model ( Figure 3 A ). Myocardial lengthening (positive value) is color-coded red and shortening (negative value) blue ( Figure 2 ). Inadequately tracked segments determined by a tracking score obtained with the speckle-tracking algorithm were automatically excluded from analysis. If the tracking by AFI was poor, the region of interest was manually fine-tuned using visual assessment during the cine loop play to ensure that all segmental wall motions were included throughout the cardiac cycle. In this study, two different types of regional myocardial functional indices were determined by AFI: ϵ-global was calculated as the average of all 17 segments and ϵ-pl as the average of 4 posterior and lateral segments in which the LV pacing lead was positioned ( Figure 3 B). In normal subjects, longitudinal strain values are generally negative, with a larger negative value indicating greater longitudinal strain. The peak longitudinal strain was used for regional myocardial functional analysis by AFI. In cases of segments with positive or biphasic strain curves, a peak-positive or a larger peak strain was used, respectively.

An example of a bull’s-eye map generated by AFI from 3 standard apical views. Myocardial lengthening is color-coded red and shortening blue .

(A) Method for translation from the 18-segment model to the 17-segment model showing apical LV short-axis images. (GE Healthcare, Waukesha, WI) (B) The LV pacing lead was positioned in the lateral or posterolateral vein, corresponding to segments 1 to 4 in the LV 17-segment model.

Pacemaker Implantation and LV Pacing Lead Position

The LV pacing lead was inserted transvenously via the subclavian route in 38 patients, and the epicardial surgical approach was used for remaining 2 patients. The LV pacing lead was positioned in the lateral or posterolateral vein, corresponding to segments 1 to 4 in the 17-segment LV model ( Figure 3 B). Device implantation was successful in all patients, without major complications. After implantation, the atrioventricular interval was optimized for maximal diastolic filling using Doppler echocardiography.

Statistical Analysis

All parametric data are expressed as mean ± SD. Group comparisons between before and after CRT were performed using the paired t test and group comparisons between responders and nonresponders using the unpaired t test. Proportional differences were evaluated using Fisher’s exact test or the χ ² test as appropriate. Correlation analysis was performed using linear regression, and results are expressed as Pearson’s correlation coefficients. Analysis of the receiver operating characteristic curve was used to assess the optimal cutoff values to predict response to CRT. For all tests, P values < .05 were considered statistically significant.