Subjects

Twenty-six adults with repaired TOF and with no significant RV outflow obstruction, as defined by Doppler-derived pressure gradients < 40 mm Hg, and absence of significant residual left-to-right shunts were recruited consecutively from the adult congenital heart clinic. Of the 26 patients, 20 (76.9%) with quality RV 3D echocardiographic data were included in the final analysis. Their clinical data were collected from the case records: demographics, age at operation, types of previous procedures, and duration of follow-up since surgical repair of TOF. Twenty-two healthy subjects were recruited as controls. These included healthy siblings of patients and adult healthy staff volunteers. The target recruitment of about 20 subjects in each group was based on a power estimation of 90% and two-sided α error of .05 to detect a difference as small as 2% in global strain with an estimated standard deviation of 2% per Kleijn et al . The body weight and height of all subjects were measured, and the body surface area and body mass index were calculated accordingly. In patients, QRS duration was determined on 12-lead surface resting electrocardiography. The institutional review board approved the study, and all subjects gave informed written consent.

3D Speckle-Tracking Echocardiographic Assessment

Transthoracic echocardiography was performed using the Artida Ultrasound System (Toshiba Medical Systems, Tokyo, Japan). Three-dimensional echocardiographic imaging was performed at the cardiac apex using a matrix-array transducer, with subjects lying in the left lateral decubitus position. Full-volume acquisition, in which four adjacent subvolumes were captured over four consecutive cardiac cycles, was performed during a breath hold. A wide sector of 90° × 90° was used to ensure inclusion of the entire RV inflow and body cavity in the 3D full-volume data set. The RV outflow could not, however, be included completely in this 3D data set. The frame rate ranged from 20 to 25 volumes/sec. All echocardiographic recordings were stored on an external hard disk for offline analysis using 3D tracking software (Advanced Cardiology Package; Toshiba Medical Systems).

On the basis of the 3D volume data-set, a five-plane evaluation that included the apical four-chamber view, two-chamber orthogonal view, and three short-axis views near the apex, at the midlevel, and at the base of the right ventricle were displayed. On the basis of these five views, the RV endocardium was traced, and its movements during cardiac systole and diastole were tracked automatically for the calculation of global and regional area strain. Area strain is defined as the percentage change in endocardial surface area during the cardiac cycle relative to end-diastolic endocardial area. Although designed for assessment of LV mechanics, the software also allows division of the right ventricle into 16 segments for assessing differences in timing of regional deformation. On the basis of the global and regional time-strain curves, the RV global peak area strain and standard deviation of the 16 segmental times to peak area strain were determined. A systolic dyssynchrony index (SDI) was calculated as the standard deviation of the 16 segmental times to peak area strain as a percentage of the RR interval as described for the left ventricle. In contrast to previously reported dedicated RV analysis software, the RV outflow was not included in the current 3D data set. The volumes of the entire right ventricle, therefore, could not be determined and compared against those obtained using CMR.

To determine the reproducibility of 3D STE–derived RV area strain and SDI, 20 subjects (10 patients and 10 controls) were selected randomly and their data sets analyzed offline at different times by the same observer (H.Y.) and two different ones (H.Y., S.L.). For each of the preselected 3D volume data sets, the same or different observers were blinded to previous measurements, and each of the data sets was measured twice.

CMR

CMR was performed in patients using a 1.5-T superconducting whole-body imager (GE Signa Horizon Echospeed; GE Medical Systems, Milwaukee, WI) with a phase-array torso coil. The time intervals between echocardiographic and CMR studies were <3 months. For baseline images, electrocardiographically triggered fast spin-echo double inversion recovery axial (automatic repetition time; echo time, 42 msec) was performed. Analysis of RV function was obtained by fast card spoiled gradient recalled cine on axial and short-axis planes, respectively. Flow analysis of the pulmonary artery was obtained using fast 2D phase contrast (flip angle, 15°; velocity encoding, 150 cm/sec) and taking a plane perpendicular to the flow direction of the RV outflow tract just above the pulmonary valve but before bifurcation. Ventricular function and flow analyses were performed using the software in Advantage Window version 4.2 (GE Medical Systems).

Statistical Analysis

All data are expressed as mean ± SD. Ventricular volumes were indexed to body surface area. Absolute values of global area strain were used to facilitate presentation and interpretation. Intraobserver and interobserver variability for the measurement of RV area strain and SDI were reported as the coefficients of variation, calculated by dividing the standard deviation of the differences between measurements by the mean and expressed as a percentage. Reproducibility was further determined by calculation of absolute difference between two measurements divided by the mean and expressed as percentages and intraclass correlation coefficient. An intraclass correlation coefficient ≥ 0.75 indicates good reproducibility, 0.40 to <0.75 moderate reproducibility, and <0.40 poor reproducibility. Differences in demographic and echocardiographic parameters between patients and controls were compared using unpaired Student’s t tests. Pearson’s correlation analysis was used to assess the relationships between global area strain parameters and CMR indices. The area under the receiver operating characteristic curve was calculated to determine capability of global area strain to detect RV ejection fraction < 50%. Multiple linear regression analysis was performed to assess significant correlates of RV area strain. Two-tailed P values < .05 were considered statistically significant. All statistical analyses were performed using SPSS version 16.0 (SPSS, Inc, Chicago, IL).

Subjects

The 20 patients (12 men) were aged 24.7 ± 8.6 years at the time of study. They underwent surgical repair of TOF at 4.5 ± 3.7 years. Transannular patch repair was required in 16 of the 20 patients. All patients were free of cardiac symptoms at the time of study, and none had cardiac arrhythmias. Their mean QRS duration was 152 ± 25 msec, with 17 patients (85%) having QRS durations > 120 msec. Their indexed RV end-systolic volume was 89 ± 63 mL/m ² , RV end-diastolic volume 149 ± 79 mL/m ² , RV ejection fraction 52 ± 11%, and pulmonary regurgitant fraction 34 ± 23%. The RV ejection fraction was >50% in 70% of the patients ( n = 14). The 22 controls (11 men) were aged 20.6 ± 7.0 years ( P = .10). Compared with controls, patients had similar body weights (53.3 ± 14.7 vs 56.5 ± 12.6 kg, P = .45) and heights (162 ± 14 vs 166 ± 9 cm, P = .19).

Observer Variability

The measurements of reproducibility for RV global area strain and SDI measurements are summarized in Table 1 .

RV Mechanics

Figures 1 and 2 show representative 3D speckle-tracking echocardiographic analysis of RV area strain curves in a control subject and a patient, respectively. Compared with controls, the peak RV global area strain was significantly lower in patients (28.1 ± 4.4% vs 31.9 ± 3.8%, P = .005; Figure 3 A).

Representative color-coded area strain images and curves in a control subject. Relatively homogenous color-coded area strain is observed in the control subject ( top ). Segmental area strain curves show synchronicity in endocardial area deformation ( bottom ).

Representative color-coded area strain images and curves in a patient. The color-coded area strain showed heterogeneity among different ventricular segments ( top ). Segmental area strain curves show reduced magnitude of and dyssynchrony in endocardial area deformation in the patient ( bottom ).

(A) Global area strain and (B) SDI of the right ventricles in patients and controls.

The SDI was significantly larger in patients than in controls (9.1 ± 2.8% vs 6.9 ± 2.2%, P = .008; Figure 3 B). On the basis of control data, RV mechanical dyssynchrony could be defined as RV SDI > 11.3% (control mean + 2 SDs). Using this definition, the prevalence of RV mechanical dyssynchrony in patients was 30% ( n = 6; 95% confidence interval, 14%–52%).

Correlates of RV 3D Area Strain and SDI

For the whole cohort and in patients, RV area strain correlated negatively with SDI ( r = −0.56, P < .001, and r = −0.42, P = .04, respectively; Figure 4 A).

(A) Scatterplot showing the relationship between RV global area strain and SDI in patients ( red ) and controls ( blue ) and (B) receiver operating characteristic curve for RV global area strain to detect ejection fraction < 50%.

In patients, RV global peak area strain correlated positively with RV ejection fraction ( r = 0.51, P = .02). The area under the receiver operating characteristic curve for using 3D global area strain to detect RV ejection fraction < 50% was 0.78 (95% confidence interval, 0.58–0.99; Figure 4 B). The best cutoff point for area strain was 24.4%, which had sensitivity of 92.3% and specificity of 71.4%. On the other hand, RV global area strain correlated negatively with RV end-diastolic ( r = −0.48, P = .032) and end-systolic ( r = −0.48, P = .031) volumes and QRS duration ( r = −0.47, P = .039).

RV SDI in patients correlated positively with RV end-systolic volume ( r = 0.55, P = .012), RV ejection fraction ( r = −0.62, P = .004), pulmonary regurgitation fraction ( r = 0.54, P = .031), and QRS duration ( r = 0.51, P = .022) but not with RV end-diastolic volume ( P > .05).

Multiple linear regression analysis identified RV ejection fraction as the only significant correlate of global peak area strain (β = 0.22, P = .048) after adjustment for age at study, sex, RV end-diastolic and end-systolic volumes, RV ejection fraction, pulmonary regurgitant fraction, SDI, and QRS duration.