Study Population

We retrospectively identified 329 consecutive individuals with LBBB, who underwent 2D echocardiography at Seoul National University Bundang Hospital between July 2009 and August 2015. We excluded those who had undergone heart surgery before the index echocardiographic examination ( n = 14) and those with atrial fibrillation ( n = 31). We also excluded those with poor image quality for strain measurements ( n = 15). Finally, we included 269 patients with LBBB and with adequate image quality for post hoc analysis of strain measurements, including GLS, global circumferential strain (GCS), and twist. Additionally, we selected 77 age- and sex-matched healthy control subjects from a previously established cohort of the Normal Echocardiographic Measurements in Korean Population study.

The study was carried out according to the principles of the Declaration of Helsinki and approved by the Clinical Research Institute of Seoul National University Bundang Hospital (L-2016-978).

Clinical Data

Clinical data on comorbidities were obtained from hospital records. Hypertension was defined as systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or current use of antihypertensive medications. Diabetes mellitus (DM) was defined as fasting blood glucose ≥ 126 mg/dL, glycated hemoglobin ≥ 6.5%, or use of antidiabetic medications. Chronic kidney disease (CKD) was defined as an estimated glomerular filtration rate < 60 mL/min/1.7 3 m ² , using the 2009 CKD Epidemiology Collaboration creatinine equation.

Echocardiography

All images were obtained with a standard ultrasound machine (Vivid E9; GE Medical Systems, Milwaukee, WI) with a 2.5-MHz probe. Standard techniques were used to obtain M-mode, 2D, and Doppler measurements in accordance with American Society of Echocardiography guidelines. Left atrial dimension was measured in the parasternal long-axis view, and left atrial volume index was measured using a biplane area-length method from the apical four- and two-chamber views and indexed on the basis of body surface area. LV diameter was measured using M-mode imaging. LV end-systolic and end-diastolic volumes and LVEF were calculated using the biplane Simpson method from apical four- and two-chamber views. Mitral E and A peak velocities (cm/s) and deceleration time (msec) were measured. Tissue Doppler–derived peak systolic (s′), early (e′), and late diastolic (a′) velocities were derived from the septal mitral annulus, using color Doppler. Pulmonary artery systolic pressure was estimated by summing the peak systolic transtricuspid pressure gradient calculated from the peak velocity of tricuspid regurgitation and right atrial pressure estimated by the diameter and inspiratory collapsibility of the inferior vena cava (IVC). According to current guidelines, the right atrial pressure was estimated as 3 mm Hg in patients with IVC diameter < 2.1 cm that collapsed >50% with sniffing, 8 mm Hg in those with IVC diameter > 2.1 cm that collapsed > 50% and in those with IVC diameter < 2.1 cm that collapsed <50%, and 15 mm Hg in those with IVC diameter > 2.1 cm that collapsed <50%. Among the total study population, pulmonary artery systolic pressure values were available in 210 patients (78.1%).

Follow-up echocardiography was performed in 124 patients at a ≥6-month interval to obtain LV diameter using M-mode imaging and LV end-systolic and end-diastolic volumes with LVEF using the biplane Simpson method.

Strain Analysis

We used 2D echocardiographic images of study participants for post hoc strain analysis. All echocardiograms were analyzed in a blinded fashion by an independent core laboratory (Seoul National University Bundang Hospital) according to current guidelines and recommendations. As a routine protocol followed by our laboratory, echocardiographic images for deformation analyses were recorded at a frame rate of 60 frames/sec (with individual adjustment between 50 and 70 frames/sec) for two cardiac cycles. Recordings were processed with acoustic-tracking software (EchoPAC BT12; GE Medical Systems, Milwaukee, WI), allowing offline semiautomated speckle-based strain analyses.

For global 2D strain analysis, a digital loop was acquired from three parasternal short-axis views (at the apical, midpapillary, and basal level) and three apical views (four-, two-, and three-chamber views). All images were transferred to network-attached storage and analyzed offline. We traced along the LV endocardial border at the end-systolic frame. The strain curve was extracted from grayscale images using dedicated software (EchoPAC BT12). We calculated the global strain values by averaging the values computed at the segmental level, as recommended in the consensus document of the American Society of Echocardiography and the European Association of Cardiovascular Imaging, and obtained GLS and GCS from three standard apical views and short-axis view at midpapillary level, respectively ( Figure 1 ). For measurement of LV twist, we used the difference in rotation between cardiac apex and base: the basal level was marked as one showing the tips of mitral valve leaflets and the apical level as just proximal to the level with LV luminal obliteration at the end-systolic period.

Representative figures of global strain and twist measurements using 2D STE. (A) LV GLS was calculated from three standard apical views with corresponding time-strain curves from 18 segments. (B) LV GCS was obtained from short-axis view at the midpapillary level, and (C) twist value was calculated as the net difference between LV peak rotation angles from basal and apical parasternal short-axis views ( white arrow ).

Two specialists (G.-Y.C. and I.-C.H.) reviewed the echocardiographic images for measurement of LV global strain values. We randomly selected 20 patients from our study, and intraobserver and interobserver variabilities were analyzed from the randomly selected 20 patients, using Pearson’s correlation coefficient ( r ), the intraclass correlation coefficient, and Bland-Altman statistics ( Table 1 , Supplementary Figure S1 , available at www.onlinejase.com ).

Outcome Measure

Our study population was followed up until August 2016. The study end point was a composite of the occurrence of cardiovascular mortality and hospitalization for HF during follow-up. Cardiovascular mortality was defined as death due to HF, acute coronary syndrome, fatal arrhythmia, stroke, or sudden cardiac death and was verified from hospital records or death certificates from primary doctors. Hospitalization for HF was defined as admission for worsening signs or symptoms of HF requiring intravenous diuretics or vasodilators and was verified from hospital records. If a patient experienced multiple events among the composite, only the first event was counted. If a patient underwent CRT, the patient was censored at the time of admission for device implantation and was neither deleted from the study population nor counted as having had an event.

Statistical Analysis

Categorical variables are presented as frequencies and percentages and continuous variables as mean ± SD or median (interquartile range). For the normality test, we used the Kolmogorov-Smirnov test: variables with P values < .50 were considered as having nonparametric distribution and variables with P values ≥ .50 as having parametric (normal) distribution. Group comparisons were performed using Student’s t test, the matched-pair t test, or the Mann-Whitney U test. The χ ² test or the Fisher exact test was used for categorical variables. Multicollinearity was assessed by computing the variance inflation factor. The cutoff values for GLS, GCS, and twist were set according to receiver operating characteristic (ROC) curve analysis, and the value showing the maximum likelihood ratio on the curve was established as the cutoff point. The cutoff value for LVEF was determined primarily from current clinical guidelines, as well as the results of ROC curve analysis. For survival analysis, the hazard ratios (HRs) for the composite end point were calculated among the baseline demographic factors, echocardiographic parameters, and LV global strain measurements, using the Kaplan-Meier method with the log-rank test and the univariate Cox proportional hazard model for comparison of time to the event. Multivariate Cox proportional hazard model analysis was used for univariate markers with P values < .10. Because of the relatively small number of events, we used stepwise backward elimination methods to select factors for inclusion in the multivariate analysis (inclusion criterion, P < .05; exclusion criterion, P > .10). Additionally, the adjusted HRs for composite end points were calculated across the subgroups stratified by the levels of strain measurements and LVEF (35% and 40%).

All statistical analyses were performed using SAS version 9.3 (SAS Institute, Cary, NC) and SPSS version 20.0 (SPSS, Chicago, IL), and P value < .05 was considered to indicate statistical significance.

Baseline Characteristics

Table 2 shows the baseline characteristics of the study population, divided according to the occurrence of the composite end point. The mean age was 69.5 ± 10.9 years, and 46.8% were men. During a median of 27.5 months (interquartile range, 12.8–43.9 months) of follow-up, 55 patients (20.4%) had composite end points, including cardiovascular mortality ( n = 12 [4.5%]) and hospitalization for HF ( n = 51 [19.0%]). Twenty-nine patients (10.8%) underwent CRT during follow-up, and these patients were censored at the time of the procedure and regarded as not having had clinical events. Presence of DM, CKD, ischemic etiology of LBBB, and use of cardiovascular preventive medications were significantly more prevalent among those who had had composite end points. Mean QRS duration was not different between the subgroups.

Echocardiographic Parameters and Strain Measurements

Echocardiographic parameters showed significant differences between patients without clinical events and those who had clinical events ( Table 3 ). Among the group with poor prognosis, LV cavity size was larger, LVEF was lower (48.4 ± 13.8% vs 29.5 ± 13.9%, P < .001), and LV diastolic function was more severely impaired (E/e′ ratio 14.3 ± 7.1 vs 22.3 ± 11.6, P = .001). Of note, 85.5% of patients (47 of 55) who had had clinical events showed LVEFs < 40%.

LV strain parameters were measured by 2D speckle-tracking. Intraobserver and interobserver variabilities for our laboratory, obtained from 20 randomly selected patients among our study population, are summarized in Table 1 and Supplementary Figure S1 (available at www.onlinejase.com ). The intraclass correlation coefficients for GLS, GCS, and twist measurements were 0.987, 0.948, and 0.917 for intraobserver variation and 0.978, 0.947, and 0.920 for interobserver variation. Bland-Altman plots demonstrated the following limits of agreement (LOA) across a broad range of GLS: the bias for intraobserver measurements of GLS was 0.5% with a range in strain values of −1.6% to +2.6% (95% LOA), and the bias for interobserver measurements of GLS was 0.5% with a range in strain values of −2.1% to +3.1% (95% LOA). The bias for intraobserver measurements of GCS was 1.1% (95% LOA, −2.9% to +5.1%), and the bias for interobserver measurements of GCS was −1.0% (95% LOA, −4.8% to +2.8%). The bias for twist measurements was 2.1% (95% LOA, −3.5% to 7.7%) for intraobserver measurements and was −1.5% (95% LOA, −6.6% to 3.6%) for interobserver measurements.

In our study population, the mean values of GLS, GCS, and twist were −13.5 ± 4.6%, −12.8 ± 4.8%, and 10.1 ± 6.9°, respectively. GLS was significantly impaired among those who had clinical events during follow-up (−14.8 ± 4.0% vs −8.7 ± 3.9%, P < .001). Similarly, patients with composite end points had significantly impaired GCS (−13.7 ± 4.6% vs −9.0 ± 3.9%, P < .001) and twist (10.8 ± 7.1° vs 7.1 ± 5.0°, P < .001).

Compared with age- and sex-matched healthy control subjects, the study population had significantly larger LV size and poorer LV function ( Supplementary Table S1 , available at www.onlinejase.com ): the mean LVEF was 44.5 ± 15.8% in patients with LBBB and 63.1 ± 4.0% in healthy control subjects ( P < .001), and the mean GLS was −13.5 ± 4.6% in patients with LBBB and −20.7 ± 2.4% in healthy control subjects ( P < .001).

Association of Global Function with Cardiovascular Events in Patients with LBBB

ROC analyses showed the cutoff values of strain measurements for the composite end point: −12.2% for GLS, −11.8% for GCS, and 6.5° of twist ( Figure 2 ). The cutoff value of LVEF by ROC curve analysis was 39.8% and was substituted with 40% and 35% considering current clinical guidelines. Impaired GLS (>−12.2%) showed significant association with the occurrence of cardiovascular mortality and hospitalization for HF (adjusted HR, 4.697; 95% CI, 1.344–16.413; P = .015), even after adjusting for univariate risk factors such as presence of DM, CKD, ischemic etiology of LBBB, and echocardiographic parameters such as left atrial volume index, LV cavity size, and LVEF, using the stepwise backward elimination method ( Figure 3 A, Tables 4 and 5 ). However, impaired GCS (>−11.8%) and twist (<6.5°) were not significantly associated with the composite end point ( Figures 3 B and 3C, Table 5 ).

ROC analysis with area under the curve, 95% CI, and P values shown for GLS ( red ), GCS ( orange ), twist ( green ), and LVEF ( blue ). The cutoff points were determined at each value showing the maximum likelihood ratio in the curve.

Risk-adjusted event-free survival curves according to global strain and twist. Cumulative incidence curves are shown for the composite end point of all-cause mortality and hospitalization for heart failure, according to GLS (A) , GCS (B) , and twist (C) . P values were obtained from multivariate Cox proportional-hazards model analysis with the significant univariate predictors among the clinical risk factors and echocardiographic parameters, using the stepwise backward elimination method.

Table 5

Multivariate analyses

	Multivariate analysis using GLS ∗			Multivariate analysis using GCS †
	HR	95% CI	P	HR	95% CI	P
Categorical variables
CKD	1.865	1.000–3.477	.050	1.795	0.902–3.571	.096
Ischemic etiology	1.596	0.862–2.955	.137	1.381	0.694–2.750	.358
LVEF < 40%	3.303	0.965–11.299	.057	7.733	2.187–27.341	.002
GLS > −12.2%	4.697	1.344–16.413	.015	—	—	—
GCS > −11.8%	—	—	—	1.645	0.499–5.429	.414
Continuous variables
CKD	2.248	1.241–4.071	.008	2.095	1.086–4.042	.027
Ischemic etiology	1.281	0.697–2.352	.425	1.177	0.602–2.300	.634
LVEF (per 1% decrease)	1.022	0.991–1.053	.169	1.058	1.017–1.100	.005
GLS (per 1% impairment)	1.201	1.078–1.339	.001	—	—	—
GCS (per 1% impairment)	—	—	—	1.030	0.905–1.172	.653

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. Recommendations for Multimodality Cardiac Imaging in Patients with Chagas Disease: A Report from the American Society of Echocardiography in Collaboration With the InterAmerican Association of Echocardiography (ECOSIAC) and the Cardiovascular Imaging Department of the Brazilian Society of Cardiology (DIC-SBC)
2. Abnormal Coronary Flow Velocity Reserve and Decreased Myocardial Contractile Reserve Are Main Factors in Relation to Physical Exercise Capacity in Cardiac Amyloidosis
3. Authors’ Reply
4. Correction

Stay updated, free articles. Join our Telegram channel

Join