Summary
Background
Evaluation of the right ventricle (RV) using transthoracic echocardiography is challenging in patients with repaired tetralogy of Fallot (rTOF).
Aims
To evaluate the accuracy of conventional echocardiographic variables and real-time three-dimensional echocardiography (RT3DE) in assessing right ventricular (RV) volumes and function compared with magnetic resonance imaging (MRI), in adult patients with rTOF and referred for pulmonary valve replacement (PVR).
Methods
Complete echocardiography was performed on 26 consecutive patients referred for PVR, before and 1 year after surgery. All variables were compared with MRI.
Results
Correlations between conventional variables and MRI were absent or poor when assessing RV ejection fraction (RVEF), except for fractional area of change (FAC; r = 0.70, P < 0.01 before PVR; r = 0.68, P < 0.01 after PVR) and RT3DE ( r = 0.96, P < 0.01 before PVR; r = 0.98, P < 0.01 after PVR). The RV volume correlation between RT3DE and MRI was excellent before and after surgery for RV end-diastolic volume ( r = 0.88, P < 0.01 and r = 0.91, P < 0.01, respectively) and RV end-systolic volume ( r = 0.92, P < 0.01 and r = 0.95, P < 0.01, respectively). The accuracy of these indices, as a diagnostic test for impaired RV (<45%), was good: Youden’s indexes varied from 0.47 to 0.89; areas under the receiver operating characteristic curve before and after PVR were 0.86 and 0.81 for FAC and 0.98 and 0.97 for RT3DE, respectively.
Conclusion
Commonly used echocardiography variables, such as tricuspid annular plane systolic excursion and tricuspid annular peak systolic velocity, did not sensitively evaluate global RVEF. A global approach, that includes the whole RV and integration of its different components, was more reliable in patients with rTOF.
Résumé
Contexte
L’évaluation du ventricule droit (VD) par l’échocardiographie trans-thoracique est un véritable défi chez les patients avec une tétralogie de Fallot corrigée.
Objectifs
Nous avons ainsi évalué la précision des paramètres échographiques conventionnels et ceux de l’échocardiographie en 3D pour analyser les volumes et la fonction du VD en comparaison à l’évaluation par IRM, chez les patients adultes avec tétralogie de Fallot corrigée adressés pour une chirurgie de remplacement valvulaire pulmonaire (RVP).
Méthodes
Une échocardiographie complète a été réalisée chez 26 patients adressés pour RVP, avant chirurgie et un an après chirurgie. Les paramètres obtenus ont été comparés à ceux issus de l’IRM.
Résultats
La corrélation entre les paramètres conventionnels et l’IRM était absente ou faible en termes d’analyse de la fonction du VD excepté pour la fraction de raccourcissement de surface FRS ( r = 0,70, p < 0,01 et r = 0,68, p < 0,01, avant et après RVP, respectivement) et l’échocardiographie 3D ( r = 0,96, p < 0,01 et r = 0,98, p < 0,01 avant et après RVP, respectivement). La corrélation des volumes ventriculaires entre l’échocardiographie 3D et l’IRM était excellente avant et après RVP pour le volume télé-diastolique du VD ( r = 0,88, p < 0,01 et r = 0,91, p < 0,01, respectivement), et pour le volume télé-systolique ( r = 0,92, p < 0,01 et r = 0,95, p < 0,01, respectivement). La précision de ces mêmes indices en tant que test diagnostique de l’altération de la fonction VD (< 45 %) était bonne : les indices de Youden variaient de 0,47 à 0,89 et les aires sous la courbe de 0,86 et 0,81 pour la FRS, 0,98 et 0,97 pour l’échocardiographie 3D, respectivement avant et après RVP.
Conclusion
Les paramètres échocardiographiques habituellement utilisés, comme le TAPSE et le S’VD, ne sont pas sensibles à évaluer la fonction du VD. Une approche globale, qui prend en compte l’ensemble du VD et l’intégration de ses différents composants, est plus adaptée chez les patients avec tétralogie de Fallot corrigée.
Introduction
Tetralogy of Fallot (TOF) is the most common cause of cyanotic congenital heart disease, and is associated with a high prevalence of pulmonary regurgitation following repair, which often requires later pulmonary valve replacement (PVR) . Monitoring right ventricular (RV) volume and function is mandatory when managing these patients.
Echocardiography is the most widely available and cost-effective imaging modality, but its use in assessing the right ventricle (RV) after repaired TOF (rTOF) remains challenging. American and European guidelines for chamber quantification recommend the use of different variables to assess RV function, such as tricuspid annular plane systolic excursion (TAPSE) and tricuspid annular peak systolic velocity (TAPSV) using tissue Doppler imaging, myocardial performance index (MPI) and fractional area of change (FAC). These variables have been validated in patients with coronary artery disease and cardiomyopathy. Other advanced variables, such as real-time three-dimensional echocardiography (RT3DE), are promising in assessing RV function . The normal RV contraction pattern is different from that of the left ventricle (LV) because of different muscle fibre organization and low vascular bed impedance . Longitudinal shortening is the main component of RV systolic function, and measurement of simple variables, such as TAPSE or TAPSV, has enabled assessment of RV function in a simple, repeatable and reproducible way in normal subjects and in patients with ischaemic heart disease . However, there are few data on the reliability of these conventional or advanced variables in patients with rTOF associated with a severely dilated RV related to chronic pulmonary regurgitation.
The objective of this study was to evaluate the accuracy of echocardiographic variables, including conventional variables (TAPSE, MPI, FAC, TAPSV) and RT3DE, to assess RV volume and function compared with magnetic resonance imaging (MRI), in patients with rTOF referred for PVR, as MRI is widely accepted as the gold standard for RV assessment in patients with congenital heart disease . Patients were re-evaluated 1 year after surgery to determine the effects of surgery and any change to these variables.
Methods
Subjects
We enrolled 26 consecutive patients referred for PVR between 2010 and 2011 in the Adult Congenital Heart Disease Clinic at the University Hospital of Bordeaux. All patients were evaluated before and 1 year after surgery. Our institutional review board approved the study and all subjects gave their informed consent. Transthoracic echocardiography and MRI at rest were performed on the same day or on a subsequent day, according to the following protocols.
Echocardiography
Standard echocardiography
Transthoracic echocardiography was performed using Vivid 7 ® (GE Vingmed Ultrasound A.S., Horten, Norway) by two experienced cardiologists (J.-B.S. and X.I.). All echocardiographic recordings were stored digitally for offline analysis. Measurements were made in three cardiac cycles, and average values were used for statistical analyses.
TAPSE was assessed in M-mode in the apical four-chamber view, and RV end-diastolic and end-systolic areas were assessed in the same view. From these measurements, we calculated the FAC. TAPSV at the junction of the RV free wall and the tricuspid annulus was assessed using pulsed tissue Doppler echocardiography. The RV MPI – defined as (isovolumic contraction + isovolumic relaxation time)/ejection time – was also assessed by Doppler echocardiography.
Three-dimensional echocardiography
Data acquisition
Three-dimensional echocardiography was performed using a 3 V matrix-array transducer (1–4 MHz). The entire echocardiography dataset was acquired from a single apical transducer location that was slightly modified from the traditional apical four-chamber view: the right-sided structures were maximized and clearly visualized, and appeared in the centre of the field of view. Data acquisition required electrocardiogram gating, so output was not truly in real-time, but was actually reconstructed from four subvolumes. The entire reconstructed three-dimensional (3D) dataset was first inspected for whole-body motion artefacts that might have occurred during data acquisition. The reconstructed data were then reviewed as a loop with a temporal resolution of 55–65 ms (15–18 vol/s).
Automated border detection and volume-computation algorithm
Analyses of original raw data were performed using dedicated RV analysis software (TomTec Imaging Systems GmbH, Munich, Germany) using the usual protocol for this software, as described in our previous study . The 3D dataset could be manipulated offline by a series of translational, rotational and pivoting manoeuvres to best visualize RV inflow and outflow tracts, and to display the reference planes. End-diastolic and end-systolic phases were first defined, then contours were manually drawn in end-diastolic and end-systolic images for three selected images (four-chamber, coronal and sagittal views) and adjusted as closely as possible to the endocardial border ( Fig. 1 ). Heavy parietal trabeculations were included in the RV chamber, as performed using MRI. These contours served to initiate the semiautomatic algorithm. Using this method and blinded to the MRI results, the software analysed RV volumes and function.
Magnetic resonance imaging
MRI was performed on a 1.5 T system (Sonata; Siemens, Erlangen, Germany) with a phased-array radiofrequency receiver coil placed on the chest. All images were gated to the electrocardiogram. Double oblique long-axis and four-chamber scouts were acquired to obtain a true short-axis reference. Steady-state free-precession prospectively electrocardiogram-gated breath-hold images, encompassing the whole RV, were then acquired in the short-axis orientation, with no gaps between the slices (TrueFISP sequence: slice thickness, 7 mm; TE, 1.53 ms; TR, 33.6 ms [depending on the R–R interval]; matrix, 256 × 256 mm; field of view, 38 cm). RV end-systolic volume (RVESV), RV end-diastolic volume (RVEDV) and RV ejection fraction (RVEF) were measured on a postprocessing workstation (Leonardo; Siemens), using commercially available software (Syngo Argus; Siemens), by a radiologist blinded to the results of the echocardiography ( Fig. 2 ).
Statistical analyses
Relationships between the echocardiographic variables of RV function and MRI of RVEF were evaluated using Pearson’s correlation coefficient and linear regression. The same method was used to compare the RV volumes obtained by RT3DE with MRI RV volumes. To estimate echocardiographic semiquantitative impaired RV systolic function, the data were presented as means ± standard deviations. Stratified correlation, according to whether MRI RVEF was normal or not, was done, and a Fisher’s z-transformation was used to compare the correlations. A threshold value of 45% for MRI RVEF was chosen according to the reference radionucleotide and the MRI results described in the review of RV function by Haddad et al. . A P -value <0.05 was considered statistically significant. Statistical analyses were performed using SAS software (version 9.1; SAS Institute, Cary, NC, USA) and receiver operating characteristic (ROC) curve analyses were obtained using R (version 2.15.2; R Foundation for Statistical Computing, Vienna, Austria).
Results
Study population
Twenty-six patients with rTOF were included in this study. All patients underwent surgical PVR for severe pulmonary regurgitation, defined as MRI pulmonary regurgitation fraction >40% and peak velocity across the RV outflow tract (RVOT) <2.5 m/s. The patients’ characteristics are presented in Table 1 : their mean age was 27 ± 12 years.
| Age (years), mean ± S.D. | 27 ± 12 |
| Ratio men: women, n/n | 16/10 |
| Palliative surgery, n | 14 |
| Complete repair (years), mean | 4 |
| Transannular patch, n | 16 |
| Delay between complete repair and PVR (years), mean | 24 |
| NYHA classification, n | |
| I | 9 |
| II | 10 |
| III | 6 |
| IV | 1 |
| PVR indication, n | |
| Severe PR, RVEDV >150 mL/m 2 , symptomatic | 14 |
| Severe PR, RVEDV <150 mL/m 2 , symptomatic | 6 |
| Severe PR, RVEDV >150 mL/m 2 , asymptomatic | 6 |
Magnetic resonance imaging results
The MRI results are shown in Table 2 . In our study, 17 (65%) patients had an RVEDV of >150 mL/m 2 . One year after PVR, RVEDV had decreased significantly from 152.1 ± 38.5 to 111.7 ± 31.7 mL/m 2 ( P < 0.01) and RVESV had decreased from 91.6 ± 32.5 to 66.2 ± 27.6 mL/m 2 ( P < 0.01). There was no significant change in RVEF when measured by MRI.
| MRI variable | Before surgery | After surgery | P a |
|---|---|---|---|
| RVEDV (mL/m 2 ) | 152.1 ± 38.5 | 111.7 ± 31.7 | <0.01 |
| RVESV (mL/m 2 ) | 91.6 ± 32.5 | 66.2 ± 27.6 | <0.01 |
| RVEF (%) | 41.1 ± 12.5 | 44.4 ± 10.3 | 0.30 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
