The purpose of this investigation was to test the hypothesis that flow patterns in the right ventricle are abnormal in patients with repaired tetralogy of Fallot (TOF). High-resolution echocardiographic contrast particle imaging velocimetry was used to investigate rotation intensity and kinetic energy dissipation of right ventricular (RV) flow in patients with TOF compared with normal controls.
Forty-one subjects (16 with repaired TOF and varying degrees of RV dilation and 25 normal controls) underwent prospective contrast imaging using the lipid-encapsulated microbubble (Definity) on Sequoia systems. A mechanical index of 0.4, three-beat high–frame rate (>60 Hz) captures, and harmonic frequencies were used. Rotation intensity and kinetic energy dissipation of flow in the right and left ventricles were studied (Hyperflow). Ventricular volumes and ejection fractions in all subjects were derived from same-day cardiac magnetic resonance (CMR).
Measurable planar maps were obtained for the left ventricle in 14 patients and the right ventricle in 10 patients among those with TOF and for the left ventricle in 23 controls and the right ventricle in 21 controls. Compared with controls, the TOF group had higher RV indexed end-diastolic volumes (117.8 ± 25.5 vs 88 ± 15.4 mL/m 2 , P < .001) and lower RV ejection fractions (44.6 ± 3.6% vs 51.8 ± 3.6%, P < .001). Steady-streaming (heartbeat-averaged) flow rotation intensities were higher in patients with TOF for the left ventricle (0.4 ± 0.13 vs 0.29 ± 0.08, P = .012) and the right ventricle (0.53 ± 0.15 vs 0.26 ± 0.12, P < .001), whereas kinetic energy dissipation in TOF ventricles was lower (for the left ventricle, 0.51 ± 0.29 vs 1.52 ± 0.69, P < .001; for the right ventricle, 0.4 ± 0.24 vs 1.65 ± 0.91, P < .001).
It is feasible to characterize RV and left ventricular flow parameters and planar maps in adolescents and adults with repaired TOF using echocardiographic contrast particle imaging velocimetry. Intraventricular flow patterns in the abnormal and/or enlarged right ventricle in patients with TOF differ from those in normal young adults. The rotation intensity and energy dissipation trends in this investigation suggest that they may be quantitative markers of RV and left ventricular compliance abnormalities in patients with repaired TOF. This hypothesis merits further investigation.
Echocardiographic particle imaging velocimetry (PIV) is an ultrasound technique for the quantification of multidirectional blood flow within cardiac chambers by the application of speckle-tracking to contrast-enhanced two-dimensional echocardiographic images. Left ventricular (LV) and left atrial flows have been previously characterized using this technique. In the LV, vortical flow structures form below the mitral valve leaflets in early diastole and support redirection of blood from the LV inflow to the outflow. It has been shown in a number of investigations that intraventricular flow patterns in the left ventricle are altered with changes in the shape or contractility of the chamber. Flow in the right ventricle has not been previously characterized using echocardiographic PIV. Concepts developed from the left ventricle cannot be directly translated to the right ventricle, because of fundamental differences in the anatomy, myofiber arrangement, and systolic and diastolic performance between the two.
Tetralogy of Fallot (TOF) is the most common cyanotic congenital heart defect (CHD). Large numbers of patients with TOF repaired as infants and children are living into late adulthood. The right ventricle in these patients bears most of the hemodynamic burden after repair; RV dilation and functional abnormalities are commonly seen. The adaptive responses of the right ventricle to long-term load stresses in this lesion are incompletely understood, so assessment of intraventricular flow fields may provide information about the circulatory performance of the right ventricle. The purpose of this investigation was to test the hypothesis that flow patterns in the right ventricle measured by echocardiographic PIV are abnormal in patients with repaired TOF. The specific aims were (1) to investigate intraventricular flow rotation intensity and energy dissipation in patients with TOF in comparison with normal controls using echocardiographic PIV and (2) to evaluate the differences and similarities in quantitative flow parameters between the right and left ventricles in patients with TOF.
This was a single-center, prospective, clinical study. The institutional review board approved the study protocol. Inclusion criteria consisted of (1) repaired TOF, (2) age ≥ 13 years, (3) absence of any intracardiac shunt on previous imaging studies, and (4) sinus rhythm. Specific exclusion criteria were (1) greater than mild tricuspid valve regurgitation, (2) greater than mild pulmonary valve regurgitation, (3) contraindications to ultrasound contrast, and (4) lack of consent to participate in the study protocol. Young adult controls were recruited in response to an advertisement inviting participation, which was approved by the institutional review board and placed in the institutional employee newsletter. Informed, written consent was obtained from all patients or legal guardians, as well as the recruited controls.
All examinations were performed using a Sequoia 512 system (Siemens Medical Solutions USA Inc, Mountain View, CA), equipped with low–mechanical index real-time pulse sequence schemes (1.7 MHz; contrast pulse sequencing). Before the administration of ultrasound contrast, all patients underwent complete diagnostic imaging, including spectral and color Doppler evaluation of the ventricular inflow, outflow, and valves according to the standard institutional practice for CHD evaluation. For echocardiographic PIV, the lipid-encapsulated microbubble Definity (Lantheus Medical Imaging, North Billerica, MA) was infused as a 3% dilution (4–6 mL/min). A mechanical index of 0.4 and settings to achieve the highest possible frame rate (consistently >60 Hz) were used. Two to three three-beat captures of two-dimensional images were obtained for the right ventricle (apical four-chamber view), and separate captures were obtained for the left ventricle (apical four-chamber view). The contrast images were exported to a dedicated echocardiographic PIV software program (Hyperflow; AMID srl, Sulmona, Italy) for analysis, as described previously. Figures 1 and 2 demonstrate echocardiographic PIV tracking and steady-streaming (heartbeat-averaged) color maps of flow fields in normal controls and in patients with TOF for the left and right ventricles, respectively. The software evaluated blood velocities and, from these, generated parameters of kinetic energy dissipation and vortex properties in the flow field. The kinetic energy dissipation is a measure of the amount of energy loss by friction within the right and left ventricles; it increases when the flow is more irregular (turbulent) and mechanical performance is reduced. The lower the energy dissipation in blood flow through the ventricular chamber, the higher the ventricular efficiency. The kinetic energy dissipation can be computed from the appropriate combination of spatial variation of the velocity vector field. It is normalized with the average amount of kinetic energy in the heartbeat to be a dimensionless parameter not directly influenced by chamber size, and it characterizes the amount of kinetic energy dissipated with respect to that available. The vortex is automatically delineated in the steady-streaming (heartbeat-averaged) field, including the rotation from inlet to outlet, and its area, as a percentage of chamber area, is evaluated. The flow rotation intensity is the amount of swirl calculated inside this vortex (its circulation) normalized with the total swirl in the whole chamber. The diastolic rotation is the unnormalized circulation (summing up the clockwise and counterclockwise values) in the whole chamber in the diastolic period. These rotation properties characterize the circulatory component of intraventricular flow.
Cardiac Magnetic Resonance (CMR)
Quantitative RV volumes were obtained from CMR performed within 2 hours of the echocardiographic PIV study. Contiguous echocardiographic PIV and CMR examinations were performed in controls in the same manner as in patients with TOF. A 1.5-T scanner (Intera R version 188.8.131.52; Philips Medical Imaging, Best, The Netherlands) with a five-channel cardiac coil (Philips Medical Imaging) was used. Cardiac synchronization was performed with vector electrocardiography. Ventricular dimensions and function were assessed using steady-state free precession cine (repetition time, 2.8–3.2 msec; echo time, 1.4–1.6 msec; field of view, 380 × 380 mm; matrix size, 160 × 130 to 228 × 216) during brief periods of breath holding in the following planes: ventricular two-chamber, four-chamber, LV and right ventricular (RV) outflow tract, and short axis with 12 to 14 equidistant slices (slice thickness, 6–8 mm; interslice spacing, 0–2 mm). Measurements of LV and RV end-diastolic and end-systolic volumes were obtained from short-axis cine stack by manual tracing of endocardial contours. A single observer (L.L.) performed all measurements using commercially available software packages (Medis Medical Imaging, Leiden, The Netherlands). Ventricular end-diastolic volumes and mass were adjusted to body surface area using the Haycock formula. Each normal control underwent quantitative CMR assessment from steady-state free precession cine images acquired at rest in the two-chamber, four-chamber, and short-axis planes as above.
Data are presented as mean ± SD; categorical variables are reported as frequencies and percentages. A comparison of continuous variables was accomplished with linear regression analysis. Echocardiographic PIV data were compared between the TOF group and normal controls a two-tailed unpaired Student’s t test. To assess intraobserver and interobserver agreement, measurements of flow rotation intensity and energy dissipation in the left and right ventricles were repeated in 10 randomly chosen TOF studies by the primary observer and by a second blinded observer. Bland-Altman plots were derived to identify possible bias (mean divergence) and the limits of agreement (2 SDs of the divergence). Mean percentage error was calculated as the absolute difference between the two sets of observations, divided by the mean of the observations: [Absolute ( X 1 − X 2 )/Mean ( X 1 , X 2 )] × 100, where X 1 − X 2 is the absolute value of the difference between observer 1 and observer 2. In addition, intraclass correlation coefficients were calculated according to standard methodology. P values < .05 were considered significant. Statistical analyses were performed using SPSS version 17.0.2 (SPSS, Inc, Chicago, IL) and Minitab version 16.1 (Minitab Inc, State College, PA).
The study population consisted of 41 subjects: 16 patients with TOF (10 men, six women) and 25 normal adult controls (15 men, 10 women). In the TOF group, 11 patients (69%) had primary repair, and five (31%) had previous shunt palliation before definite repair. Nine patients (57%) were repaired without transannular patches, five (31%) were repaired by the use of transannular patches, and the details of repair were unknown in two (12%). For the TOF group, planar maps suitable for intraventricular flow analysis were obtained for the left ventricle in 14 patients and the right ventricle in 10 patients. For controls, intraventricular flow analysis was feasible for the left ventricle in 23 and the right ventricle in 21. Those who were excluded from analysis were excluded either for suboptimal image quality in the apical four-chamber view (often from RV dilation) or because of bubble density that was too high (saturation), prohibiting analysis of bubble velocity. None of the patients with TOF had evidence of antegrade diastolic flow in the pulmonary artery, coincident with atrial systole (restrictive RV physiology).
Demographic data and CMR measurements in patients with TOF and control subjects who had measurable planar maps are shown in Table 1 . There was no significant difference between the TOF group and controls for body weight, body surface area, heart rate, LV indexed end-diastolic volume, and LV ejection fraction. Compared with controls, the TOF group had higher RV indexed end-diastolic volume (117.8 ± 25.5 vs 88 ± 15.4 mL/m 2 , P < .001) and lower RV ejection fraction (44.6 ± 3.6% vs 51.8 ± 3.6%, P < .001). All but one patient with TOF had CMR-measured RV end-diastolic volume < 150 mL/m 2 . The single subject with relatively higher RV end-diastolic volume (177 mL/m 2 ) had only mild tricuspid regurgitation and pulmonary regurgitation (by echocardiography and magnetic resonance imaging) and qualitatively mild RV dilatation on echocardiography. Measurements of pulsed Doppler of the tricuspid inflow and tissue Doppler of the lateral tricuspid annulus showed mild elevation of the E/e′ ratio in the TOF group (6.8 ± 2.4 vs 3.8 ± 1.4 in controls, P < .001), indicating abnormal diastolic filling.
|Variable||Controls ( n = 23)||Patients with TOF ( n = 14)|
|Mean ± SD||Range||Mean ± SD||Range|
|Age (y)||34.4 ± 8.4||23.0–43.0||22.3 ± 8.9||13.0–42.6||<.001|
|Weight (kg)||79.8 ± 15.3||58.5–112.5||77.7 ± 20.1||40.9–108.2||.752|
|Body surface area (m 2 )||2.0 ± 0.2||1.66–2.45||2.0 ± 0.3||1.4–2.4||.676|
|Heart rate (beats/min)||71.8 ± 9.8||53.0–92.1||72.0 ± 14.2||52.1–115.4||.928|
|LV indexed EDV (mL/m 2 )||83.6 ± 13.1||59.7–104.4||86.1 ± 12.1||63.0–109.0||.501|
|LV EF (%)||58.9 ± 4.4||54.4–66.4||56.6 ± 2.2||52.0–60.0||.181|
|RV indexed EDV (mL/m 2 )||88 ± 15.4||62.5–102.3||117.8 ± 25.5||97.0–177.0||.016|
|RV EF (%)||51.8 ± 3.6||47.3–62.8||44.6 ± 3.6||40.0–50.4||<.001|
Table 2 shows intraventricular flow parameters compared between the groups. Flow rotation intensity in the left ventricle was higher in patients with TOF compared with controls (0.40 ± 0.13 in patients with TOF vs 0.29 ± 0.08 in controls, P = .012). The diastolic rotation in the left ventricle was also lower in patients with TOF. Dissipation of kinetic energy within the left ventricle was lower in patients with TOF (0.51 ± 0.29 in patients with TOF vs 1.52 ± 0.69 in controls, P < .001). For the right ventricle in patients with TOF, flow rotation intensity was significantly higher (0.53 ± 0.15 vs 0.26 ± 0.12, P < .001), while kinetic energy dissipation was lower (0.4 ± 0.24 vs 1.65 ± 0.91, P < .001). When measures of flow rotation intensity and energy dissipation were compared with patient age, RV end-diastolic volume, RV ejection fraction, and RV E/e′ ratio, no correlations were found ( Table 3 ). Bland-Altman analysis showed good intraobserver and interobserver agreement for flow rotation intensity and energy dissipation measurements for the left and right ventricles in patients with TOF, as represented in Figures 3 and 4 . The mean percentage error between observers for flow rotation intensity was 21% for the left ventricle and 23% for the right ventricle. For kinetic energy dissipation, the mean percentage error was 14% for the left ventricle and 16% for the right ventricle. The results of intraclass correlations are shown in Table 4 . For both the left and right ventricles, intraclass correlations for kinetic energy dissipation were greater than those for flow rotation intensity.
|Variable||Controls||Patients with TOF||P value|
|Mean ± SD||Range||Mean ± SD||Range|
|Left ventricle||( n = 23)||( n = 14)|
|Flow rotation intensity (absolute)||0.29 ± 0.08||0.13–0.42||0.4 ± 0.13||0.16–0.7||.012|
|Kinetic energy dissipation||1.52 ± 0.69||0.47–3.26||0.51 ± 0.29||0.11–0.97||<.001|
|Vortex area||0.19 ± 0.05||0.08–0.26||0.26 ± 0.10||0.11–0.41||.019|
|Diastolic rotation (cm 2 /sec)||203.3 ± 84.7||90.4–380.8||87.6 ± 44.4||29.4–170.6||<.001|
|Right ventricle||( n = 21)||( n = 10)|
|Flow rotation intensity (absolute)||0.26 ± 0.12||0.04–0.58||0.53 ± 0.15||0.29–0.78||<.001|
|Kinetic energy dissipation||1.65 ± 0.91||0.15–3.33||0.40 ± 0.24||0.19–0.84||<.001|