The accurate assessment of right ventricular (RV) function and dimensions has important prognostic implications in patients with repaired tetralogy of Fallot (ToF). Three-dimensional imaging is the preferred methodology to evaluate RV function. Novel postprocessing software applications to evaluate three-dimensional data have provided insight into RV function and dimensions by analyzing the various RV components (inlet, apical trabecular, outlet). The aim of this study was to characterize regional RV function and dimensions with real-time three-dimensional echocardiography (RT3DE) in patients with repaired ToF.
Forty-one patients with repaired ToF (age range, 8–18 years) and 20 control subjects were enrolled. Global and segmental RV volumes and ejection fraction (EF) were assessed with RT3DE and compared between patients with repaired ToF and controls.
Segmental analysis on RT3DE demonstrated that the apical trabecular region was the most remodeled RV component in patients with repaired ToF, with significantly increased end-diastolic volume and end-systolic volume compared with controls (59 ± 19 vs 41 ± 16 mL and 36 ± 13 vs 24 ± 8 mL, respectively; P = .001 for both). However, EF was preserved at that region. In contrast, EF was reduced at the RV inlet (53 ± 6% vs 58 ± 7%, P = .003) and outlet (44 ± 16% vs 52 ± 10% ( P = .032).
Patients with repaired ToF show characteristic RV remodeling as assessed with RT3DE. At the apical trabecular region, the largest volumes were observed compared with control patients, whereas EF at the inlet and outlet components was significantly impaired. RT3DE may facilitate future studies of segmental RV volumes and function in patients with repaired ToF.
Right ventricular (RV) dilatation and subsequent RV failure in patients with repaired tetralogy of Fallot (ToF) are associated with poor clinical outcomes. The clinical management of these patients is highly determined by RV dimensions and function. The evaluation of RV volumes using conventional two-dimensional echocardiography is hampered by the complex geometry of the right ventricle. Recently, real-time three-dimensional echocardiography (RT3DE) has overcome some of the limitations of two-dimensional echocardiography to assess RV volumes and function. The feasibility and accuracy of RT3DE for RV volume assessment have been demonstrated in several studies including patients with repaired ToF.
From an embryologic and anatomic perspective, the right ventricle can be divided into three sections: the inlet, the apical trabecular region, and the outlet. Recent studies in patients with repaired ToF using magnetic resonance imaging have shown that the three different components of the right ventricle show different volumetric and functional responses to volume overload. The performance of RT3DE to assess the various RV components has not been evaluated so far. Accordingly, the objective of the current study was to evaluate the feasibility of a novel semiautomated software algorithm based on RT3DE to assess the dimensions and function of the various components of the right ventricle. Furthermore, the contributions of the various RV components to global RV volume and function were evaluated in patients with repaired ToF and compared with a group of control subjects.
Study Population and Protocol
A total of 41 consecutive patients with repaired ToF (mean age, 13.1 ± 2.8 years; 56% men) and 20 healthy controls, matched for age, gender, and body surface area, were prospectively enrolled in the present study. All patients and controls were evaluated with standard two-dimensional echocardiography and RT3DE. Particularly, global and segmental RV volumes and ejection fraction (EF) were assessed with RT3DE. Using dedicated software, real-time three-dimensional echocardiographic RV data were segmented offline in three different functional regions: inlet, apical trabecular, and outlet. The contribution of each functional component to global RV volume and function was compared between patients with repaired ToF and healthy controls. The study protocol was approved by the institutional review board, and all subjects gave written informed consent.
Real-Time Three-Dimensional Echocardiographic Image Acquisition and Analysis
Transthoracic real-time three-dimensional echocardiographic images of the right ventricle were acquired by a single experienced sonographer using a commercially available system equipped with a 3V phased-array transducer (Vivid 7.0.0; GE Vingmed Ultrasound AS, Horten, Norway). Patients and control subjects were imaged in the left lateral decubitus position. To encompass the full RV volume into the data set, six electrocardiographically gated subvolumes were acquired to form a larger pyramidal volume including the entire right ventricle. Images were acquired during a single breath-hold in end-expiration to avoid translational motion. Compression and gain settings were adjusted to optimize image quality and subsequent endocardial border visualization. Real-time three-dimensional echocardiographic RV data sets were stored digitally, and quantitative analysis of the three-dimensional RV volumes was performed offline using commercially available semiautomated software (TomTec Imaging Systems, Unterschleissheim, Germany).
Global Assessment of RV Volumes and Function
From the real-time three-dimensional echocardiographic RV full-volume data sets, first, the software algorithm automatically displayed the RV volume in three imaging planes (four-chamber, sagittal, and coronal; Figure 1 ). In these planes, manual identification of three anatomic landmarks (center of the tricuspid annulus, center of the mitral valve annulus, and left ventricular apex) was performed. Subsequently, the endocardial RV border was manually traced at end-systole and end-diastole in the four-chamber, sagittal, and coronal planes. During contour tracing, the apical trabeculae were included in the chamber volume. The automated algorithm tracked the endocardial borders frame by frame throughout the cardiac cycle. After automated endocardial border tracking, the algorithm displayed the frame-by-frame endocardial contour position in a cine loop. The dynamic change of the endocardial border and the apical trabeculae could be appreciated in this cine loop, and the contours were adjusted when necessary. Finally, the software automatically calculated RV end-diastolic volume (EDV), RV end-systolic volume (ESV), and RV ejection fraction.
Regional Assessment of RV Volumes and Function
The algorithm for segmental analysis of the right ventricle is displayed in Figure 2 . After manual tracing of the endocardial borders of the full RV volume in the four-chamber view, sagittal RV view, and coronal RV view, the software identified three anatomic landmarks (lateral wall of the tricuspid annulus, lateral wall of the pulmonary annulus, and apex). On the basis of these anatomic landmarks defined by the observer, three surface landmarks were subsequently identified mathematically by the automated software (landmark A to landmark C, Figure 2 ). Landmark A was defined as the region at 50% of the distance between the tricuspid valve border (lateral wall of tricuspid annulus) and the apex. Landmark B was defined as the region at 50% of the distance between the pulmonary valve border (lateral wall of pulmonary valve annulus) and the apex. Landmark C was defined as the as the region at 50% of the distance between the tricuspid valve border and the pulmonary valve border. From these surface landmarks, the three RV regions (inlet, apical trabecular, and outlet) were automatically identified. Subsequently, the software provided volume computations for the three subvolumes in every time frame, from which EDV (largest volume), ESV (smallest volume), and EF ([ESV/EDV] × 100%) were derived for each RV functional component.
Continuous variables are expressed as mean ± SD. Categorical variables are presented as numbers and percentages. Differences between patients with repaired ToF and controls were analyzed using unpaired t tests. Categorical data were analyzed using χ 2 tests. To assess the interobserver agreement of RT3DE, Bland-Altman analysis was performed. Furthermore, the coefficient of variation was calculated (the absolute difference between the observers in percentage of the study population average). Data were analyzed using SPSS version 17.0 (SPSS, Inc., Chicago, IL). P values < .05 were considered statistically significant.
Table 1 summarizes the clinical and echocardiographic characteristics of patients and controls and the surgical details of the patients with repaired ToF. In all patients, the surgical repair was performed via a trans-atrial approach. An additional commissurotomy or valvotomy of the pulmonary valve was performed in two patients who received RV outflow tract or pulmonary artery patches. Transannular patches were placed in 65% of patients. None of the patients showed hemodynamically significant residual ventricular septal defects at the time of inclusion. By definition, there were no significant differences between patients and controls in terms of age, gender, and body surface area. QRS duration was significantly increased in patients with repaired ToF (134 ± 19 vs 94 ± 8 ms, P < .001). In patients with repaired ToF, global RV volumes as assessed with RT3DE were significantly increased (RV EDV, 164 ± 48 vs 133 ± 50 mL, P = .026; RV ESV, 89 ± 26 vs 64 ± 24 mL, P = .001), and global RV EF was significantly reduced (46 ± 8% vs 52 ± 5 %, P = .007) compared with control subjects. None of the included patients with repaired ToF showed an aneurysmatic RV outflow tract.
|Variable||Patients with repaired ToF ( n = 41)||Controls ( n = 20)||P|
|Age (y)||13.1 ± 2.8||13.8 ± 2.5||.202|
|Male/female||23 (56%)/18 (44%)||13 (65%)/7 (35%)||.507|
|BSA (mL/m 2 )||1.4 ± 0.3||1.5 ± 0.3||.129|
|QRS duration (ms)||134 ± 19||94 ± 8||<.001|
|Type of surgery ∗|
|Transannular patch||26 (65%)||–|
|RVOT or PA patch||6 (15%)||–|
|RV EDV (mL)||164 ± 48||133 ± 50||.026|
|RV ESV (mL)||89 ± 26||64 ± 24||.001|
|RV EF (%)||46 ± 8||52 ± 5||.007|
Segmental Analysis of RV Volumes with RT3DE
Table 2 outlines the segmental analysis of the right ventricle in patients with repaired ToF and control subjects. At the RV inlet, the EDV and ESV were not different between patients with repaired ToF and controls. In contrast, EF at the RV inlet was significantly reduced in patients with repaired ToF compared with controls (53 ± 6% vs 58 ± 7%, respectively, P = .003).
|Variable||Patients with repaired ToF ( n = 41)||Controls ( n = 20)||P|
|EDV (mL)||67 ± 25||63 ± 27||.546|
|ESV (mL)||31 ± 10||26 ±12||.098|
|EF (%)||53 ± 6||58 ± 7||.003|
|RV apical trabecular|
|EDV (mL)||59 ± 19||41 ± 16||.001|
|ESV (mL)||36 ± 13||24 ± 8||.001|
|EF (%)||39 ± 12||39 ± 12||.916|
|EDV (mL)||38 ± 13||36 ± 21||.728|
|ESV (mL)||21 ± 9||17 ± 9||.099|
|EF (%)||44 ± 16||52 ± 10||.032|