Classic-Pattern Dyssynchrony and Electrical Activation Delays in Pediatric Dilated Cardiomyopathy


Progressive heart failure leading to transplantation or death is common in pediatric dilated cardiomyopathy (DCM), and treatment options are limited. Select children with DCM have improved after cardiac resynchronization therapy (CRT), but predicting response is challenging. Nonetheless, considering the frequency of death or transplantation in this population, identifying any candidate would be valuable. Classic-pattern dyssynchrony (CPD) identifies mechanical dyssynchrony patterns consistent with underlying electrical activation delays and strongly predicts CRT response in adult DCM but has not been evaluated in pediatric DCM. The aim of this study was to test the hypothesis that CPD is present in a subgroup of patients with pediatric DCM and is associated with activation delays.


Fifty-nine subjects with pediatric DCM (left ventricular end-diastolic diameter Z score > 2 and left ventricular ejection fraction < 40%) who underwent echocardiography with a functional protocol with apical images optimized for two-dimensional speckle-tracking strain analysis (EchoPAC) were retrospectively analyzed for CPD. Electrocardiograms were evaluated for activation delays (prolonged QRS duration and strict criteria for left bundle branch block [LBBB]). Forty control subjects with no cardiac disease and good imaging widows were also analyzed.


The mean age was 5.4 years (range, 1 day to 20 years); idiopathic DCM was most common (57%). Severe cardiomyopathy was present in 75% (end-diastolic diameter Z score > 4.6 and left ventricular ejection fraction < 32%). CPD was identified in seven subjects (12%), and prolonged QRS durations were present in 13 (22%), but only two subjects met strict criteria for LBBB. Six of seven subjects in the CPD group had prolonged QRS durations, and two of seven had LBBB. No control subjects had CPD. The CPD analysis was highly feasible and reproducible.


In this severely affected cohort, the small CPD subgroup is potentially important because their progressive disease may respond to CRT. CPD is associated with activation delays, although not necessarily strict LBBB. This has important potential implications for prospective evaluation of CRT in this disease.

Progressive heart failure is common in pediatric dilated cardiomyopathy (DCM), and treatment options are limited. Cardiac resynchronization therapy (CRT) is now an accepted treatment modality in adult patients with DCM with electrocardiographic activation delays, especially in the form of left bundle branch block (LBBB), but data are limited in pediatric patients with DCM. In contrast to the adult literature, few studies have assessed the frequency and natural history of activation delays and mechanical dyssynchrony in pediatric DCM. Three large retrospective reviews of pediatric CRT subjects across the spectrum of pediatric and congenital cardiac disease, including systemic left ventricle, systemic right ventricle, and single-ventricle anatomy, demonstrated that CRT’s efficacy varied widely (22%–77%) by underlying substrate. One study reported the frequency of CRT response in the pediatric DCM subgroup as 33%, although it combined primary DCM and DCM secondary to dual-chamber pacing for atrioventricular block. Although as a whole, the pediatric DCM population is thought to respond poorly to CRT, clinical uncertainty regarding the use of CRT and the proper criteria for predicting CRT response in these children persists, and we have observed anecdotal cases of marked response to CRT in pediatric patients with DCM.

Recent adult studies from various independent groups have reported on reproducible patterns of mechanical dyssynchrony resulting from activation delays (LBBB) by using echocardiography (e.g., classic-pattern dyssynchrony [CPD], septal flash). These patterns focus on identifying the physiology of LBBB with early termination of the septal contraction opposed by early stretch and late contraction in the delayed lateral wall. CPD was highly predictive of response to CRT (>90% responder rate), which improved on a variety of traditional echocardiographic indices predicting CRT response. CPD was also strongly associated with newly proposed strict criteria for LBBB by electrocardiography in adults. In the present study, we used such newer techniques in adults and applied them to the pediatric population to (1) identify the frequency of CPD in pediatric DCM and (2) investigate whether CPD is associated with electrical activation delay, including LBBB, in these children. We hypothesized that CPD would be found in a proportion of children with DCM in association with LBBB.


Study Subjects

We retrospectively analyzed echocardiograms that had been obtained for clinical indications using a standardized functional protocol, with apical imaging optimized for speckle-tracking. All subjects with pediatric DCM who had available echocardiograms obtained with this protocol on a GE ultrasound system (GE Vingmed Ultrasound AS, Horten, Norway) were included. We also analyzed CPD in a cohort of controls with no cardiac disease who had normal medical histories and echocardiographic results. Subjects were between 1 day and 21 years of age at the Hospital for Sick Children (Toronto, Ontario, Canada) and Duke University Medical Center (Durham, NC). The studies occurred between September 2005 and August 2013.

Studies were performed for clinical indications either during inpatient admissions or at outpatient clinic appointments. Echocardiographic examinations were ordered for control subjects for routine functional assessment or murmur referral with normal ventricular function and size. All subjects with DCM had increased left ventricular (LV) end-diastolic diameters (LVEDDs) ( Z score ≥ 2) and LV ejection fractions (LVEFs) < 40% by echocardiography. The LVEDD Z score represents the number of standard deviations above the average LVEDD indexed for weight and height (body surface area), which allows comparison of ventricular dimensions across the spectrum of pediatric ages. Hence, an indexed LVEDD Z score of +3 reflects a dimension that is 3 standard deviations above the mean LVEDD indexed for body surface area.

Exclusion criteria included congenital anatomic heart disease (except for patent foramen ovale), mitral valve surgery, atrial dysrhythmias, cardiac pacing, and frequent ectopy precluding the evaluation of consecutive sinus beats. Heart failure medications (diuretics, digoxin, β-blockers, calcium channel blockers, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers), demographics, clinical health, and electrocardiographic information were obtained from the medical record. This study was approved by the Toronto Hospital for Sick Children Research Ethics and Duke Institutional Review Boards.


Standard 12-lead electrocardiograms, recorded at 100 to 150 Hz, were identified closest in time to the echocardiograms obtained in the DCM population. They were analyzed for heart rate, QRS duration, and the presence of the strict LBBB criteria of Strauss et al . (QS or rS in leads V 1 or V 2 and mid-QRS notching or slurring in two or more contiguous leads [V 1 , V 2 , V 5 , V 6 , I, and aVL]). Adult QRS duration criteria were modified for pediatrics by using QRS duration > 98th percentile for age and sex. The presence or absence of strict LBBB was determined by two experienced cardiologists. These analyses were performed blinded to other patient data. Electrocardiographic data were not consistently available for the control population.


Echocardiograms were obtained using a functional protocol with grayscale images optimized for longitudinal speckle-tracking strain analysis (60–110 frames/sec) using either GE Vivid 7 or E9 ultrasound scanners. Apical images were optimized for strain analysis by focusing on myocardial border definition while minimizing the depth and sector arc to fill the image with LV myocardium and maximize frame rate. Breath holding was not consistently feasible, because of patient age. LV size was assessed by LVEDD Z score from M-mode measurements using institutional Z scores. LVEF was calculated by using the biplane modified Simpson’s method.

Tissue Doppler Analysis

In the DCM population, offline tissue Doppler analysis was performed on the three apical views with GE EchoPAC PC version BT11. Time to peak velocity was measured for the middle and basal segments, and the 12-segment standard deviation of time to peak velocity (TsSD) was determined according to previously reported methodology.

Two-Dimensional Speckle-Tracking Strain Analysis

Offline strain analysis was performed on four- and three-chamber apical views using GE EchoPAC PC version BT11. The reference point was placed at QRS onset to ensure the inclusion of early systolic events. The aortic valve closure timing relative to the QRS was defined on spectral Doppler from the apical five-chamber view while ensuring that heart rate was similar between acquired strain optimized images. The LV endocardial border was manually traced in end-systole from the medial to the lateral mitral valve annulus for speckle-tracking, and the region of interest was adjusted to exclude the pericardium and papillary-chordal structures. The integrity of the speckle-tracking of each of the six segments was automatically analyzed by EchoPAC and visually confirmed for each view. In case of poor tracking, the tracing was readjusted or the shape of the region of interest changed until adequate by visual assessment. Segments with persistent inadequate tracking were excluded from analysis. A view was defined as adequate for strain analysis if four or more segments were adequately tracked, including at least three middle and basal segments in each view.

Regional Strain Pattern Analysis

LV regional strains were analyzed for the presence or absence of CPD, blinded to the other patient data. Analysis was performed according to the method of Risum et al ., which uses three criteria to define CPD in any of the LV apical views and consists of (1) early stretch with late, postsystolic contraction in at least one delayed lateral wall segment; (2) early terminated contraction in at least one opposing septal wall segment; and (3) early peak contraction that must occur in the first 70% of the systolic ejection phase ( Figure 1 ). Early peaks of contraction or stretch were visually defined by a distinct peak below or above the baseline, not by a minimum peak strain value.

Figure 1

CPD. Strain curves demonstrating CPD that identifies longitudinal strain patterns between opposing walls that are consistent with mechanical dyssynchrony due to underlying electrical activation delays. Strain analysis begins at the QRS onset. A simplified pair of representative septal (blue) and lateral wall (brown) curves along with an electrocardiogram (ECG) illustrate the timing of the CPD criteria (left) . The CPD pattern is present in an infant with DCM with strict LBBB on electrocardiography (right) . The three CPD criteria are labeled numerically in both panels: (1) early stretch in at least one segment with delay activation, (2) early terminated contraction in a septal wall segment with the peak contraction occurring in the first 70% of the systolic ejection phase, and (3) segments with early stretch that must actively contract with a late postsystolic peak strain. AVC , Aortic valve closure; AVO , aortic valve opening.


CRT was performed in one infant with DCM, with LVEF < 35%, LBBB, and severe heart failure symptoms through an uncomplicated sternotomy with placement of right atrial, LV, and right ventricular epicardial leads.

Statistical Analysis

Continuous variables are reported as mean ± SD and were compared using two-sided unpaired t tests. Categorical variables (reported as percentages) were tested for differences using the Fisher exact test. P values < .05 were considered statistically significant. Statistical analyses were performed using Stata version 12 (StataCorp LP, College Station, TX). Interobserver reproducibility for CPD using the same cardiac cycle was performed on all subjects with DCM and controls by assessing agreement and calculating the κ statistic between independent readers, blinded to the results of the other reader. Each observer performed independent strain tracking, analysis, and interpretation de novo. Intraobserver reproducibility for CPD using the same cardiac cycle was performed on the first 10 subjects with DCM by agreement and the κ statistic. This strain interpretation was performed ≥2 weeks apart, blinded to the previous interpretation.


DCM Population

The DCM population comprised 62 pediatric subjects with DCM, all anatomically normal. Fifty-nine were included after three subjects were excluded because of inadequate echocardiographic image quality in both the three- and four-chamber views. Subject demographics are presented in Table 1 . In the DCM cohort, infants were most common, with a relatively equal split between children and adolescents in the remainder of subjects. Idiopathic and familial DCM diagnoses represented the large majority of the subjects with DCM. The mean indexed LVEDD was 5.0 ± 1.4 mm/m 2 (range, 1.8–7.9 mm/m 2 ), with a mean Z score of 6.3 ± 2.5 (range, 2.0–12.5). The mean LVEF was 24 ± 9% (range, 9%–38%). Moderate to severe disease (LVEDD Z score > 4.6, LVEF < 32%, requiring at least two heart failure medications) was found in 75% of subjects.

Table 1


Variable Controls ( n = 40) Patients with DCM ( n = 59)
Age (y) 10.7 ± 5.9 (0–17) 5.4 ± 6.5 (0–20)
Infants 7 (18%) 30 (51%)
Children 16 (40%) 16 (27%)
Adolescents 17 (42%) 13 (22%)
Female 21 (53%) 30 (51%)
≥2 heart failure medications 48 (81%)
DCM diagnosis
Idiopathic 34 (57%)
Familial 15 (25%)
Postchemotherapy 5 (9%)
Postmyocarditis 5 (9%)

Continuous variables are reported as mean ± SD (range).


Most subjects with DCM underwent electrocardiography and echocardiography on the same day (75%) (maximum interval, 4 weeks). Electrocardiographic analysis demonstrated sinus rhythm ranging from bradycardia to tachycardia. The mean heart rate was 122 ± 29 beats/min (range, 59–199 beats/min), with a mean QRS duration of 80 ± 21 msec (range, 42–152). Thirteen subjects (22%) had prolonged QRS durations when adjusted for age and sex. Of these, only two (15%) met the criteria for strict LBBB (3%). Right bundle branch block (RBBB) was seen in three of 13 subjects (23%). The eight subjects with prolonged QRS durations not meeting criteria for bundle branch block had intraventricular conduction delays. Subjects with prolonged QRS durations had lower systolic function than subjects with normal QRS durations (mean LVEF, 19.8 ± 7.7% vs 25.0 ± 8.6%; P = .05).

Echocardiographic Analysis and CPD

CPD was identified in at least one apical view in seven subjects with DCM (12%). CPD was identified in both apical views in one subject, the four-chamber in only five subjects, and the three-chamber view in only one subject. All seven had idiopathic DCM and demonstrated typical CPD. Comparing the CPD and non-CPD subgroups, there were no statistically significant differences in age, sex, TsSD, LVEDD Z score, and LVEF ( Table 2 ). Cardiac transplantation or death occurred in three of seven (43%) in the CPD group and 26 of 52 (50%) in the non-CPD group, with an overall cohort mean follow-up interval of 27 months (range, 1–90 months).

Table 2

CPD subgroup analysis

Characteristic CPD ( n = 7) No CPD ( n = 52) P
Age (y) 2.0 ± 3.0 5.9 ± 6.7 .14
Female 5/7 (71%) 25/52 (48%) .42
LVEDD Z score 7.2 ± 2.6 6.2 ± 2.5 .33
LVEF (%) 23 ± 5 24 ± 9 .88
TsSD (msec) 35 ± 8 34 ± 13 .79
QRS duration > 98th percentile 6/7 (86%) 7/52 (13%) <.001
Strict LBBB 2/7 (29%) 0/52 (0%) <.001

Continuous variables are reported as mean ± SD.

Control Population

Demographics are provided for 40 control subjects in Table 1 . There was a smaller percentage of infants in the control group compared with the DCM population, but all age groups were represented. None of the control subjects had identified cardiac disease, and all had normal LV size and function. No control subjects had CPD.

Feasibility and Reproducibility

In the 59 subjects with DCM, a four-chamber view was adequate for strain analysis in all. A three-chamber view was also adequate for strain analysis in 41 of 59 subjects (70%). The middle and basal segments in both views were adequate tracked in nearly all views (>90%). Tracking of the apical segments was adequate in most views (64%–88%), with the apical free wall segment in the three-chamber view tracking the most poorly.

In the control subjects, chosen for good apical image quality, four-chamber and three-chamber views were adequate for strain analysis in all 40 subjects. The middle and basal segments in both views were adequately tracked in most views (>85%). Tracking of the apical segments was adequate in most views (79%–97%), with the apical septal segment in the three-chamber view tracking the most poorly.

In the subjects with DCM, interobserver reproducibility analysis for the presence of CPD was performed for all studies and demonstrated initial agreement in 57 of 59 (97%; κ = 0.86). After reconciliation, the two subjects for whom disagreement occurred were placed in the non-CPD subgroup because of incomplete or borderline CPD criteria. Intraobserver reproducibility demonstrated agreement in all 10 subjects analyzed (100%; κ = 1.0).

In the control subjects, interobserver reproducibility for the presence of CPD was performed for all 40 studies and demonstrated agreement in all 40 (100%; κ = 1.0).

Correlation between Electrical Delays and CPD

Nearly all of the subjects with CPD had prolonged QRS durations, compared with a minority in the non-CPD group ( Table 2 ). The two subjects with strict LBBB also demonstrated CPD on strain imaging, and the echocardiographic strain curves and electrocardiogram for one of them are shown in Figure 2 . The other subject underwent CRT ( Figure 3 ). Evaluating the five subjects with CPD without strict LBBB, four had prolonged QRS durations with intraventricular conduction delays, and one had a normal QRS duration and morphology. The three subjects with RBBB did not have CPD. In the non-CPD group without RBBB, the four subjects with prolonged QRS durations all demonstrated nonspecific intraventricular conduction delays on electrocardiography.

Figure 2

CPD and strict LBBB in a subject with DCM. This 8-year-old subject with severe idiopathic DCM demonstrated both CPD by strain echocardiography and LBBB on electrocardiography. (Top left) Four-chamber apical two-dimensional image with the LV strain region of interest overlying the myocardium. (Top right) Four-chamber view strain curves that met the criteria for CPD with early contraction in the septum (yellow arrow) opposed by early lateral wall stretch (red arrow) followed by lateral wall contraction once the delayed ventricular activation reaches that region (green arrow) . (Bottom) Three electrocardiographic leads that met strict LBBB criteria with a QRS duration of 138 msec (>90 th percentile) with QRS notching in all leads and an rS wave in lead V 1 . AVC , Aortic valve closure.

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

Stay updated, free articles. Join our Telegram channel

May 31, 2018 | Posted by in CARDIOLOGY | Comments Off on Classic-Pattern Dyssynchrony and Electrical Activation Delays in Pediatric Dilated Cardiomyopathy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access