The aim of this study was to test the hypothesis that dyssynchronous contraction of functional single ventricles occurs in Fontan patients and is related to indices of myocardial deformation and global ventricular function.
Twenty patients with tricuspid atresia (mean age, 23.5 ± 7.1 years) were studied 17.8 ± 3.8 years after undergoing the Fontan procedure. Three-dimensional echocardiographic data were acquired for determination of left ventricular (LV) volumes and systolic dyssynchrony indices. LV myocardial deformation was determined using speckle-tracking echocardiography. Calibrated integrated backscatter intensity was measured as an index of myocardial fibrosis. The results were compared with those in 20 controls.
Compared with controls, patients had significantly greater systolic dyssynchrony indices (6.13 ± 1.32% vs 4.06 ± 0.84%, P < .001). The prevalence of LV mechanical dyssynchrony (systolic dyssynchrony index > 5.74%) in patients was 55% (95% confidence interval, 32%–77%). LV global systolic longitudinal, radial, and circumferential strain ( P < .001 for all), longitudinal systolic ( P < .001) and early diastolic ( P < .001) strain rate, and circumferential systolic ( P < .001) and early diastolic ( P = .009) strain rate were significantly lower in patients than in controls, while the average calibrated integrated backscatter was higher ( P < .001). Patients with LV dyssynchrony ( n = 11) had lower global LV longitudinal strain ( P = .02), reduced LV ejection fractions ( P = .002), and higher average calibrated integrated backscatter ( P = .03) compared with those without LV dyssynchrony ( n = 9).
A high proportion of patients with tricuspid atresia after the Fontan operation exhibit LV mechanical dyssynchrony, which may in part be related to myocardial fibrosis and has implications for myocardial deformation and global ventricular function.
Ventricular dysfunction remains one of the most important long-term complications in patients with functionally univentricular hearts palliated by the Fontan procedure. Conventional two-dimensional and Doppler echocardiographic assessments have revealed evidence of progressive systolic and diastolic ventricular dysfunction in these patients. Nonetheless, the cause of ventricular dysfunction remains unclear.
The role of mechanical dyssynchrony in the development of ventricular dysfunction and heart failure is increasingly recognized. In Fontan patients who have functional single ventricles, abnormal orientation of myofibers, myocardial fibrosis, and absence of ventricular-ventricular interaction may provide substrates for the development of dyssynchronous myocardial contraction. Additionally, these unfavorable factors may potentially impair the magnitude and rate of deformation of the systemic ventricular myocardium in different dimensions. However, data on the mechanics of the systemic ventricle in patients after the Fontan procedure are limited.
Advances in echocardiography have enabled the quantification of three-dimensional systemic ventricular mechanical dyssynchrony, taking into account differences in the timing of volume changes of all ventricular segments during systole. Additionally, the new angle-independent speckle-tracking technique permits the evaluation of myocardial deformation of both ventricles in a biventricular circulation and functional single ventricles. The recently introduced three-dimensional speckle-tracking technique has further enabled comprehensive and simultaneous evaluation of deformation of different myocardial segments. Nonetheless, the three-dimensional mechanics of functional single ventricles in Fontan patients has hitherto not been explored.
The aim of the present study was to test the hypothesis that dyssynchronous contraction of functional single ventricles occurs in Fontan patients and is related to indices of myocardial deformation and global ventricular function. To test the hypothesis, we compared the index of three-dimensional left ventricular (LV) systolic dyssynchrony in patients with tricuspid atresia after the Fontan procedure with that of controls. We further determined relationships between LV systolic dyssynchrony and indices of myocardial deformation, fibrosis, and global function.
Twenty-five of the 32 Fontan patients with tricuspid atresia identified from the congenital heart disease database agreed to participate in this study. Five patients who required pacing or were not in sinus rhythm at the time of study were excluded. The following data were retrieved from the case records: cardiac diagnosis, need for systemic-to-pulmonary shunt insertion or pulmonary arterial banding, type of Fontan procedure, age at operation, and duration of follow-up since Fontan procedure. Twenty healthy subjects matched for age and sex were recruited as controls. These were subjects with functional cardiac murmurs, those with nonspecific chest pain or palpitation but without identifiable organic causes, and healthy siblings of patients. Body weight and height were measured, and body surface area was calculated accordingly. The QRS durations in patients were also measured. The institutional review board approved the study, and all subjects and parents of minors gave informed consent.
Conventional Echocardiographic Assessment
All echocardiographic assessments were performed using the Vivid 7 ultrasound system (GE Vingmed Ultrasound AS, Horten, Norway). The recordings were stored on DVDs for offline analyses by a single investigator (P.-K.H.). Echocardiographic measurements from three cardiac cycles were determined, and the average was used for analysis.
Color Doppler mapping was performed to assess semiquantitatively the degree of atrioventricular valvar regurgitation. From the apical four-chamber view, pulsed tissue Doppler assessment with positioning of the sample volume at the LV free wall–atrioventricular valve annular junction was performed. The peak myocardial velocities at systole (s), early diastole (e), and late diastole (a) and myocardial acceleration during isovolumic contraction were determined as reported previously.
Calibrated Integrated Backscatter (cIB)
Myocardial calibrated integrated backscatter (cIB) has been found to correlate with myocardial collagen content and has been used as an index of myocardial fibrosis. End-diastolic myocardial integrated backscatter of the posterior LV wall and interventricular septum at the midventricular level of the parasternal short-axis view was determined using EchoPAC (GE Vingmed Ultrasound AS). Calibrated integrated backscatter of the septum and posterior wall was calculated as the difference between the integrated backscatter measured at the respective region of interest and that of the pericardium. The average of three readings was used for analysis. Care was taken to avoid the bright epicardial and endocardial specular reflections, and output power settings were adjusted to ensure that signals were not saturated in both the pericardium and myocardium. The sample volume was manually tracked in each frame to ensure its location within the same region throughout the cardiac cycle.
Two-dimensional images were acquired from the apical four-chamber view for assessment of global LV longitudinal strain and strain rate (SR), while those from the midventricular short-axis view were acquired for determination of circumferential strain and SR and radial strain. The frame rate of cine loops acquired for analysis was between 60 and 80 frames/sec. The LV free wall and ventricular septum were divided respectively into three segments (basal, mid, and apical) for determination of regional longitudinal strain, while the LV short-axis plane was divided into six segments (anterior, anteroseptal, inferoseptal, inferior, posterior, and lateral) for determination of regional circumferential and radial strain. Global longitudinal and circumferential strain was calculated on the basis of the entire traced LV contour in the apical four-chamber view and short-axis plane, respectively. Global radial strain was calculated as the average of radial strain of six regions. Strain and SR were analyzed offline using EchoPAC. Our group has previously reported low intraobserver variability for measurement of global strain and SR.
Real-time three-dimensional echocardiography for the determination of LV volumes, ejection fraction, and dyssynchrony was performed as reported previously. Briefly, a full-volume LV data set was acquired from the apical four-chamber view and analyzed offline using 4D LV-function software (TomTec Imaging Systems, Unterschleissheim, Germany). After semiautomatic tracing of the LV endocardial border, an LV cast was created for the derivation of LV end-diastolic and end-systolic volumes and ejection fractions. The LV volumes were indexed to body surface area for analysis. To quantify global LV mechanical dyssynchrony, the systolic dyssynchrony index (SDI), defined as the standard deviation of time taken to reach minimum regional volume for each of the 16 LV segments as a percentage of the cardiac cycle, was determined.
All data are expressed as mean ± SD. The absolute values of strain and SR were used to facilitate interpretation and analyses. Intraobserver and interobserver variability was reported as the coefficient of variation, calculated by dividing the standard deviation of the differences between measurements by the mean, and expressed as a percentage. One reader (P.-K.H.) performed repeated measurements in the same offline analysis session to assess intraobserver reproducibility, while two readers (P.-K.H. and C.T.M.L.) performed blinded repeated measurements to determine interobserver variability. Differences in demographic data and echocardiographic parameters between patients and controls, as well as between patients with and without dyssynchrony, were compared using unpaired Student’s t tests and Fisher’s exact tests as appropriate. Pearson’s correlation analysis was used to study the relationships between SDI and parameters of myocardial deformation, cIB intensity, and LV ejection fraction. P values < .05 were considered statistically significant. All statistical analyses were performed using SPSS (SPSS, Inc., Chicago, IL).
A total of 20 patients (seven men, 13 women; mean age, 23.5 ± 7.1 years) were studied. All but one patient had situs solitus with absent right atrioventricular connection, while the remaining patient had situs inversus totalis. The ventriculoarterial connection was concordant in 13 patients, double-outlet from the left ventricle in five, and discordant in two. Surgical interventions before the Fontan procedure included systemic-to-pulmonary arterial shunt insertion in nine patients, pulmonary arterial banding in two, and ligation of persistent arterial duct in one. The Fontan procedure was performed at a mean of 5.8 ± 5.1 years of age, with 17 patients having atriopulmonary connections and three with total cavopulmonary connections. The mean duration of follow-up after the Fontan procedure was 17.8 ± 3.8 years. The mean QRS duration in patients was 105 ± 17 msec, with five patients having QRS durations > 120 msec. The 20 controls (eight men, 12 women; P = .74) were aged 23.1 ± 6.1 years ( P = .84). Compared with controls, patients had similar body weights (54.8 ± 13.4 vs 56.4 ± 10.6 kg, P = .66) and body surface areas (1.6 ± 0.2 vs 1.6 ± 0.2 m 2 , P = .26).
Conventional Echocardiographic Parameters
Table 1 summarizes the echocardiographic findings in patients and controls. Compared with controls, patients had significantly reduced mitral annular s ( P < .001) and e ( P = .001) velocities. On the other hand, mitral annular e/a ratio, a velocity, and myocardial acceleration during isovolumic contraction were similar between patients and controls ( P > .05 for all). Mitral regurgitation was absent in six patients, mild in 11, moderate in two, and severe in one.
( n = 20)
( n = 20)
|Mitral annular myocardial tissue velocities|
|s (cm/sec)||4.9 ± 1.9||6.8 ± 1.3||<.001 ∗|
|e (cm/sec)||7.7 ± 3.5||11.0 ± 2.3||.001 ∗|
|a (cm/sec)||4.2 ± 3.0||4.6 ± 1.3||.54|
|e/a ratio||2.5 ± 1.9||2.6 ± 0.8||.83|
|Myocardial acceleration during isovolumic contraction (cm/sec 2 )||1.1 ± 0.7||1.2 ± 0.5||.45|
|Indexed LV end-diastolic volume (mL/m 2 )||68.3 ± 16.6||55.9 ± 11.9||.009 ∗|
|Indexed LV end-systolic volume (mL/m 2 )||37.6 ± 11.6||23.8 ± 7.7||<.001 ∗|
|LV ejection fraction (%)||46.4 ± 8.1||57.1 ± 6.3||<.001 ∗|
|SDI (%)||6.1 ± 1.3||4.1 ± 0.8||<.001 ∗|
Calibrated Integrated Backscatter
The average cIB were significantly greater in patients than controls (−18.9 ± 4.7 vs −25.0 ± 2.9 dB, P < .001). In patients, the average cIB was not correlated with age at operation ( P = .32).
Myocardial Strain and SR
The coefficients of variation for intraobserver and interobserver measurements of global longitudinal strain were 3.8% and 4.8%, respectively, while those of global circumferential strain were 3.9% and 5.1%, respectively.
Table 2 summarizes the parameters of global myocardial deformation. Compared with controls, patients had significantly reduced global systolic strain in the longitudinal, circumferential, and radial dimensions ( P < .001 for all). Similarly, the longitudinal systolic ( P < .001), early diastolic ( P < .001), and late diastolic ( P = .002) SRs were significantly lower in patients than controls. The circumferential systolic ( P < .001) and early diastolic ( P = .009) SRs were also lower in patients than controls.
( n = 20)
( n = 20)
|Systolic strain (%)||15.4 ± 4.5||23.8 ± 4.5||<.001 ∗|
|Systolic SR (sec −1 )||0.9 ± 0.2||1.4 ± 0.3||<.001 ∗|
|Early diastolic SR (sec −1 )||1.1 ± 0.3||2.1 ± 0.6||<.001 ∗|
|Late diastolic SR (sec −1 )||0.6 ± 0.3||0.8 ± 0.3||.002 ∗|
|Systolic strain (%)||15.7 ± 2.8||22.2 ± 3.7||<.001 ∗|
|Systolic SR (sec −1 )||0.9 ± 0.2||1.3 ± 0.3||<.001 ∗|
|Early diastolic SR (sec −1 )||1.2 ± 0.4||1.6 ± 0.6||.009 ∗|
|Late diastolic SR (sec −1 )||0.6 ± 0.4||0.5 ± 0.2||.19|
|Systolic strain (%)||19.2 ± 7.6||41.2 ± 12.5||<.001 ∗|
The regional systolic strain parameters are shown in Figure 1 . Regional longitudinal and radial strain in all six segments was significantly lower in patients than controls ( P < .001 for all). Regional circumferential strain was significantly lower in all ( P < .001) but the inferior segment ( P = .57) in patients than in controls.
Three-Dimensional Echocardiographic Parameters
Compared with controls, patients had significantly greater indexed LV end-diastolic volumes ( P = .009) and indexed LV end-systolic volumes ( P < .001) and significantly lower LV ejection fractions ( P < .001) ( Table 1 ).
The LV SDI was significantly greater in patients than controls (6.13 ± 1.32% vs 4.06 ± 0.84%, P < .001; Figure 2 A). On the basis of control data, global LV mechanical dyssynchrony was defined as LV SDI > 5.74% (mean + 2 SDs for controls). The prevalence of LV systolic mechanical dyssynchrony in patients was 55% (95% confidence interval, 32%–77%). Figure 2 B shows examples of time-volume curves demonstrating synchronous and dyssynchronous contraction of the 16 LV segments during systole in a control and a patient, respectively.
For the whole cohort, LV SDI was correlated negatively with global longitudinal systolic strain ( r = −0.66, P < .001), longitudinal systolic SR ( r = −0.48, P = .002), longitudinal early diastolic SR ( r = −0.57, P < .001), circumferential systolic strain ( r = −0.46, P = .003), circumferential systolic SR ( r = −0.36, P = .02), radial systolic strain ( r = −0.64, P < .001), and LV ejection fraction ( r = −0.76, P < .001) ( Figure 3 A). On the other hand, LV SDI was correlated positively with average cIB ( r = 0.66, P < .001) ( Figure 3 B).
Compared with patients without LV dyssynchrony, patients with LV dyssynchrony had significantly lower global LV longitudinal systolic strain ( P = .02), reduced LV ejection fractions ( P = .002), and higher average cIB ( P = .03) ( Table 3 ). Within the patient group, LV SDI was correlated with LV ejection fraction ( r = −0.71, P < .001) and cIB ( r = 0.55, P = .01) ( Figure 3 ) but not with QRS duration ( P = .68), age at Fontan procedure ( P = .88), or duration of follow-up since operation ( P = .70).
|Variable||Presence of dyssynchrony |
( n = 11)
|Absence of dyssynchrony |
( n = 9)
|Demographic and clinical variables|
|Age at study (y)||23.5 ± 6.7||23.6 ± 8.2||.97|
|Age at Fontan procedure (y)||5.9 ± 5.5||5.7 ± 4.8||.94|
|Follow-up since Fontan procedure (y)||17.6 ± 2.9||17.9 ± 4.9||.86|
|Type of Fontan procedure (APC/TCPC)||10/1||7/2||.57|
|QRS duration (msec)||104.7 ± 16.6||105.4 ± 19.5||.93|
|Systolic strain (%)||13.3 ± 4.4||18.0 ± 3.3||.02 ∗|
|Systolic SR (sec −1 )||0.9 ± 0.2||1.0 ± 0.2||.29|
|Early diastolic SR (sec −1 )||1.0 ± 0.4||1.2 ± 0.3||.15|
|Systolic strain (%)||15.1 ± 2.6||16.5 ± 3.1||.26|
|Systolic SR (sec −1 )||0.9 ± 0.2||0.9 ± 0.2||.63|
|Early diastolic SR (sec −1 )||1.1 ± 0.4||1.3 ± 0.4||.31|
|Systolic strain (%)||16.8 ± 6.5||22.1 ± 8.2||.13|
|Three-dimensional echocardiographic parameter|
|Indexed LV end-diastolic volume (mL/m 2 )||71.1 ± 16.4||65.0 ± 17.2||.43|
|Indexed LV end-systolic volume (mL/m 2 )||41.5 ± 10.4||32.8 ± 11.7||.09|
|LV ejection fraction (%)||41.5 ± 4.3||52.5 ± 7.6||.002 ∗|
|Average cIB (dB)||−16.8 ± 4.0||−21.4 ± 4.5||.03 ∗|