Interaction of Left Ventricular Remodeling and Regional Dyssynchrony on Long-Term Prognosis after Cardiac Resynchronization Therapy


Left ventricular (LV) remodeling in heart failure (HF) manifested by chamber dilatation is associated with worse clinical outcomes. However, the impact of LV dilatation on the association of measures of dyssynchrony with long-term prognosis and resynchronization potential after cardiac resynchronization therapy (CRT) remains unclear.


Two hundred sixty CRT patients in New York Heart Association classes II to IV, with ejection fractions ≤ 35% and QRS intervals ≥ 120 msec, were prospectively studied. Quantitative echocardiographic assessment of LV volumes and mechanical dyssynchrony by radial strain was conducted at both baseline and 6-month follow-up. Primary outcome events were predefined as death or HF hospitalization, and secondary outcome events were defined as all-cause death over the 4 years after CRT.


Patients were divided into two groups using the median of the baseline indexed LV end-diastolic volume (EDVI). Patients with less dilated left ventricles (EDVI < 90 mL/m 2 ) had improved prognosis compared to those with severely dilated left ventricles (EDVI ≥ 90 mL/m 2 ) for both primary (adjusted hazard ratio [HR], 2.20; 95% CI, 1.44–3.38; P < .01) and secondary (adjusted HR, 1.94; 95% CI, 1.21–3.11; P < .01) events. Similarly, reduction in baseline dyssynchrony was associated with good prognosis for both the primary (HR, 0.39; 95% CI, 0.23–0.68; P = .001) and secondary (HR, 0.41; 95% CI, 0.22–0.75; P = .004) events. A linear association was found between each 10% reduction in dyssynchrony and events ( P < .01). Notably, patients with less dilated left ventricles had nearly fourfold more frequent improvement in dyssynchrony compared to those with severely dilated left ventricles (odds ratio, 4.10; 95% CI, 1.81–9.28; P < .01). No other baseline prognostic marker was associated with the resynchronization ability of CRT.


Patients with severe LV remodeling (EDVI ≥ 90 mL/m 2 ) have a poor prognosis following CRT device implantation. This is most likely due to impaired resynchronization efficacy.


  • Reduction in baseline mechanical dyssynchrony after CRT in patients with HF is associated with significant reduction in mortality and HF hospitalization.

  • Patients with extensive remodeling has poor prognosis after CRT because of a lack of mechanical resynchronization after CRT.

  • There is a linear association between the amount of reduction in mechanical dyssynchrony after CRT device implantation and survival. Devising a strategy to follow patients treated with CRT devices by the assessment of dyssynchrony could be beneficial.

Left ventricular (LV) dilatation in patients with heart failure (HF) is associated with worse outcomes. Likewise, some studies have suggested that patients with extensive LV remodeling do not derive much benefit from cardiac resynchronization therapy (CRT). It is well known that mechanical dyssynchrony predicts a favorable response to CRT, but the interaction between mechanical dyssynchrony and extensive LV remodeling before CRT device implantation is poorly elucidated. Although it has been reported that dyssynchrony by Doppler tissue imaging (DTI) lacks significant additive value in predicting LV reverse remodeling after CRT in patients with extensively dilated left ventricles, DTI measurements of velocity can provide inaccurate assessment because of technical limitations, particularly in a dilated left ventricle with spherical deformation. Importantly, though DTI-based methods have shown promising results, most robust data on mortality reduction are obtained using two-dimensional speckle-tracking methods.

Furthermore, a number of studies have reported that echocardiographic strain can be used to quantify the synchronization ability of CRT and provide crucial prognostic information. However, it is uncertain whether the resynchronization potential of CRT is independent of baseline LV volume. In the present study we used deformation analysis to investigate (1) the interaction of baseline LV remodeling and dyssynchrony with long-term prognosis after CRT and (2) the resynchronization efficacy of CRT and its association with long-term prognosis in patients with severely dilated left ventricle compared with those with less dilated left ventricles.


Patients implanted with CRT devices at a single center were enrolled in a prospective manner. The inclusion criteria for the study were presence of HF symptoms (New York Heart Association classes II–IV), QRS duration ≥ 120 msec, and LV ejection fraction ≤ 35%. All patients received optimal medical therapy before CRT device implantation. The exclusion criteria were recent myocardial infarction (≤3 months), failure to perform echocardiography at our institution before implantation of the device, poor image quality of echocardiograms, loss to follow-up, and failure to obtain informed consent for participation in the study. A total of 260 patients were enrolled in the study initially. Ischemic cardiomyopathy was defined as a documented history of myocardial infarction, prior revascularization, or presence of significant coronary artery disease (≥70% stenosis in at least one major coronary artery). All implanted CRT devices had defibrillators with right atrial leads, right ventricular apical leads, and LV leads through the coronary sinus implanted in the LV free wall. The institutional review board of the University of Pittsburgh approved the study protocol, and the study complied with the Declaration of Helsinki.


All patients underwent echocardiography (Vivid 7; GE Vingmed Ultrasound AS, Horten, Norway) at baseline, and a subgroup of patients underwent 6-month follow-up studies. The analyses were performed in a blinded manner by the echocardiography core laboratory, consisting of the coauthors. LV volumes and ejection fraction were measured using the biplane Simpson rule. LV volumes were indexed to body surface area. Offline analyses were performed using EchoPAC PC versions BT08 and B11 (GE Vingmed Ultrasound AS).

Radial Dyssynchrony

Dyssynchrony was evaluated by speckle-tracking radial strain, as described in detail previously. Briefly, speckle-tracking radial strain was assessed in the midventricular short-axis view, and regions of interest were adjusted manually for the endocardium and epicardium. The frame rates of the short-axis images for speckle-tracking were 60 to 90 Hz. The times to peak of six segments were calculated from the start of the QRS complex. A delay of ≥130 msec between the anteroseptal segment and the posterior wall segment was considered to indicate the presence of dyssynchrony ( Figure 1 ). Inter- and intraobserver variability has been reported previously from our laboratory, with excellent agreement between the observers. Radial strain dyssynchrony assessment was performed on both the baseline and the available 6-month follow-up echocardiograms. Reduced dyssynchrony was defined as a reduction in baseline dyssynchrony (radial strain delay ≥ 130 msec) to no dyssynchrony (radial strain delay < 130 msec) at 6-month follow-up after CRT. Absolute improvement in radial strain delay was calculated, which was converted to percentage change in dyssynchrony of baseline radial dyssynchrony. We further used this percentage change in baseline dyssynchrony as a continuous variable and calculated its association with the study end points.

Figure 1

Dyssynchrony assessment using two-dimensional speckle-tracking radial strain. ( Left ) Short-axis image with region of interest. ( Right ) Strain curves with presence of time to peak radial dyssynchrony, with a time difference of 130 msec between the peak of the anteroseptum and posterior segment.

Study End Points

The primary end point of the study was death or hospitalization for HF within the first 4 years after CRT device implantation. We further examined a secondary end point of all-cause death within the first 4 years after CRT device implantation. The data for end points were collected through chart review and the US Social Security Death Index. Patients were screened at the first event of HF hospitalization for the primary end points.

Statistical Analysis

Patients were divided into two groups on the basis of the median value of baseline indexed LV end-diastolic volume (EDVI). Continuous variables are expressed as mean ± SD and were compared using Student’s t test, and proportional variables are expressed as absolute numbers and percentages and were compared using χ 2 tests or Fisher exact tests as appropriate. Time to first event for combined end points between groups was graphically illustrated by Kaplan-Meier analysis and compared using the log-rank test for the two groups. Univariate Cox regression analysis was used to calculate hazard ratios (HRs) and their 95% CIs for different variables. A multivariate Cox regression analysis was performed for risk for events, including age, gender, ischemia, left bundle branch block, QRS duration ≥ 150 msec, baseline radial dyssynchrony, and EDVI (≥90 mL/m 2 ) in a backward selection model. A multivariate analysis was performed a second time, replacing baseline radial dyssynchrony with reduced dyssynchrony in the model. Binary logistic regression was performed for the outcome as reduced dyssynchrony after CRT, including the variables ischemia, left bundle branch block, QRS duration, and baseline EDVI. Odds ratios and their 95% CIs are presented. A two-tailed P value < .05 was considered to indicate statistical significance. All statistical analyses were performed in SPSS Statistics version 22 (IBM, Armonk, NY).


From the initial 260 patients, 20 (8%) were excluded because of poor image quality, and three patients (1%) were subsequently excluded because of loss to follow-up. Patients were divided into two groups using the median EDVI (90 mL/m 2 ): patients with less dilated left ventricles (EDVI < 90 mL/m 2 ) and those with severely dilated left ventricles (EDVI ≥ 90 mL/m 2 ). The baseline characteristics of the population are summarized in Table 1 .

Table 1

Baseline characteristics of the population

Variable Total population ( n = 237) EDVI < 90 mL/m 2 ( n = 135) EDVI ≥ 90 mL/m 2 ( n = 102) P
Age (y) 65 ± 11 68 ± 11 63 ± 12 .001
Gender (male) 174 (73%) 81 (69%) 93 (78%) .19
NYHA class II or III 196 (83%) 100 (87%) 96 (81%) .22
Ischemia 144 (61%) 69 (59%) 75 (63%) .89
Diabetes 83 (35%) 40 (34%) 43 (36%) .58
QRS duration (msec) 159 ± 23 161 ± 24 157 ± 23 .20
LBBB 108 (46%) 49 (42%) 59 (49%) .30
β-blocker 207 (87%) 101 (86%) 106 (88%) .70
ACE inhibitor/ARB 204 (86%) 100 (86%) 104 (87%) .85
LVEF (%) 24 ± 6 27 ± 5 22 ± 5 <.001
LV EDVI (mL/m 2 ) 97 ± 35 71 ± 12 123 ± 32 <.001
Radial dyssynchrony ≥ 130 msec 168 (71%) 69 (59%) 99 (83%) <.001

ACE , Angiotensin-converting enzyme; ARB , angiotensin II receptor blocker; LBBB , left bundle branch block; LVEF , left ventricular ejection fraction; NYHA , New York Heart Association.

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

P < .05.

Association of LV Volume with Clinical Outcomes

There were 100 unique combined events (42%) of HF hospitalization and death in the first 4 years of follow-up after CRT. Forty-one patients (41%) were hospitalized for HF, and 59 patients (59%) died during these 4 years, with no prior HF hospitalizations. Among the 41 patients who were hospitalized for HF symptoms, a further 21 patients died within the time frame of 4 years, leading to a total number of deaths among the population of 80 (34%).The median durations to primary and secondary events were 0.81 years (interquartile range, 0.28–2.15 years) and 1.38 years (interquartile range, 0.55–2.68 years), respectively.

Patients with EDVI ≥ 90 mL/m 2 at baseline had an increased risk for primary events (HR, 1.85; 95% CI, 1.24–2.78; P = .003) ( Figure 2 ). The risk for secondary events showed a trend toward significance in a univariate analysis (HR, 1.54; 95% CI, 0.99–2.40; P = .06) ( Figure 2 ). However, in a multivariate analysis, it had a significant association with both primary (HR, 2.20; 95% CI, 1.44–3.38; P = .0003) and secondary (HR, 1.94; 95% CI, 1.21–3.11; P = .006) events. Other covariates associated with events were left bundle branch block and QRS duration ≥ 150 msec (only with primary events), while the presence of baseline radial dyssynchrony was associated with both the primary and secondary events. The univariate and multivariate analyses are summarized in Table 2 .

Figure 2

Association of LV volume with long-term outcome after CRT. Kaplan-Meier curve showing that patients with less dilated left ventricles (EDVI < 90 mL/m 2 ) have improved prognosis after CRT compared with those with dilated left ventricles. Although the association with all-cause death was borderline significant in univariate analysis, it became significant after adjusting for other baseline variables.

Table 2

Univariate and multivariate Cox regression analysis

Variable Death and HF hospitalization Death
HR (95% CI) P HR (95% CI) P
Univariate analysis
Age (per year) 1.00 (0.99–1.02) .45 1.02 (0.99–1.04) .07
Gender (male) 1.40 (0.89–2.31) .14 1.45 (0.85–2.48) .17
Ischemia 1.77 (1.15–2.73) .01 1.76 (1.09–2.86) .02
LBBB 0.58 (0.38–0.87) .009 0.63 (0.40–1.00) .048
QRS ≥ 150 msec 0.53 (0.36–0.79) .002 0.64 (0.41–1.003) .05
EDVI ≥ 90 mL/m 2 1.85 (1.24–2.78) .003 1.54 (0.99–2.40) .06
Radial strain ≥ 130 msec 0.64 (0.42–0.97) .03 0.67 (0.42–1.06) .09
Reduced dyssynchrony 0.40 (0.23–0.68) .001 0.41 (0.22–0.75) .004
Multivariate analysis
LBBB 0.63 (0.41–0.96) .03 NS
QRS ≥ 150 msec 0.58 (0.39–0.87) .008 0.64 (0.41–1.01) .06
EDVI ≥ 90 mL/m 2 2.20 (1.44–3.38) .0003 1.94 (1.21–3.11) .006
Radial strain ≥ 130 msec 0.60 (0.38–0.94) .03 0.60 (0.37–0.97) .04
Reduced dyssynchrony 0.39 (0.23–0.68) .001 0.41 (0.22–0.75) .004

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

Stay updated, free articles. Join our Telegram channel

Apr 15, 2018 | Posted by in CARDIOLOGY | Comments Off on Interaction of Left Ventricular Remodeling and Regional Dyssynchrony on Long-Term Prognosis after Cardiac Resynchronization Therapy

Full access? Get Clinical Tree

Get Clinical Tree app for offline access