Background
Right ventricular (RV) volume overload results in RV dilatation and dysfunction in patients with pulmonary regurgitation after tetralogy of Fallot (ToF) repair, affecting left ventricular (LV) function because of ventricular interaction. The aim of this study was to assess the effect of chronic RV volume loading on LV myocardial mechanics in patients after ToF repair by combining cardiac magnetic resonance imaging with echocardiography.
Methods
Seventy-five subjects were studied: 50 patients after ToF repair and 25 age-matched controls. All patients with ToF and controls underwent echocardiography at the time of clinically indicated cardiac magnetic resonance imaging. Myocardial deformation, including LV torsion, was analyzed using speckle-tracking echocardiography.
Results
RV free wall global and segmental longitudinal strain and strain rate were significantly lower in patients with ToF compared with controls ( P < .001). All LV longitudinal, strain, and torsion parameters were significantly reduced in patients with ToF ( P < .01). Basal rotation was particularly abnormal, with 38% of patients with ToF having reversed basal rotation. In contrast, apical rotation was reduced but not reversed. On multivariate regression analysis, the only significant predictor of counterclockwise basal rotation was RV strain, suggesting that RV function more than dilatation influences abnormal LV torsion.
Conclusion
Patients with ToF have abnormal LV myocardial mechanics, as demonstrated by speckle-tracking echocardiography. The most striking changes were noted in LV torsion, especially related to abnormal LV basal rotation. RV dysfunction seems the most important determinant of abnormal LV rotation. The clinical significance and potential prognostic implications of these observations remain to be determined.
Tetralogy of Fallot (ToF) is the most common cyanotic congenital heart defect, with good long-term survival after surgical correction. Postoperative chronic pulmonary insufficiency (PI) results in progressive right ventricular (RV) dilatation and dysfunction, both being predictors of adverse clinical outcomes. A close relationship between the RV and left ventricular (LV) ejection fractions (EFs) has been shown in patients after ToF repair, and impaired LV function has been proved to be an important predictor of adverse clinical outcome. More recently, myocardial deformation techniques have been used to assess cardiac function in this population, showing decreases in multiple components of RV and LV myocardial deformation. Moreover, reduced LV deformation was shown to be an important independent predictor of the occurrence of arrhythmia and death in adults after ToF repair. However, the mechanisms driving LV dysfunction in this population are likely related to RV dysfunction but remain inadequately defined. Likewise, although the development of overt ventricular dysfunction occurs relatively late in the clinical course, these are likely to have their origins much earlier in the clinical course in childhood. Accordingly, the objective of this study was to investigate the influence of RV function and dilatation on LV mechanics in children after ToF repair using simultaneous evaluation by echocardiography and cardiac magnetic resonance imaging (MRI).
Methods
This was a single-center prospective observational clinical study. The study was approved by the institutional research ethics board. Informed consent was obtained from all participants and/or their legal guardians before enrollment. The study population consisted of two groups: group 1 included patients after ToF repair and group 2 included normal controls. All patients were recruited at the time of a clinically indicated cardiac MRI study. Group 2 patients were asymptomatic relatives of patients with arrhythmogenic RV dysplasia who underwent screening cardiac MRI and had completely normal results. All patients underwent limited research echocardiographic examinations just before or immediately after cardiac MRI. Exclusion criteria were a lack of informed consent, an inability to cooperate, known or detected arrhythmia interfering with image acquisition, more than mild tricuspid regurgitation, and more than mild RV outflow tract obstruction.
Echocardiography
All echocardiographic studies were performed using a GE Vivid 7 or Vivid E9 machine (GE Medical Systems, Milwaukee, WI). Images were acquired from standard parasternal and apical views. Raw Digital Imaging and Communications in Medicine imaging data were digitally stored and analyzed offline. RV measurements were performed according to recently published guidelines. RV and LV lengths were measured in the apical four-chamber view, from the center of the valve to ventricular apex. RV fractional area of change was calculated by measuring RV end-diastolic and end-systolic areas from the apical four-chamber view, as (RV end-diastolic area – RV end-systolic area)/RV end-systolic area × 100. Pulsed tissue Doppler traces were obtained in the basal segment of the RV free wall, the interventricular septum (IVS), and the basal segment of the LV free wall. Peak systolic tissue Doppler velocity (S′) was measured on the tracings. Myocardial performance index was calculated for each ventricle on tissue Doppler tracings of the RV and LV free walls by measuring isovolumic relaxation time and isovolumic contraction time as well as systolic S-wave duration. Myocardial performance index was calculated as (isovolumic relaxation time + isovolumic contraction time)/S-wave duration. Annular peak systolic excursion was measured from the apical four-chamber view using anatomic M-mode imaging for the RV and LV free walls and the IVS. Two-dimensional speckle-tracking (EchoPAC version 110.1.3; GE Medical Systems, Milwaukee, WI) was used for deformation analysis. To analyze RV longitudinal strain (ε) and strain rate, an apical four-chamber view centered on the right ventricle was used. The LV algorithm was applied to the right ventricle. Briefly, the endocardial border was manually traced at end-systole (starting from the tricuspid annulus in the RV free wall from the apical four-chamber view). Tracking was automatically performed, and the analysis was accepted after visual inspection and when the software indicated adequate tracking. If tracking was suboptimal, the endocardial border was retraced. If satisfactory tracking was not accomplished, the nontracking segments were excluded from analysis. Values of the best tracked cardiac cycle of three were considered for analysis. Peak systolic ε was calculated for the RV lateral wall, IVS, and LV lateral wall as means of the basal, middle, and apical segments for each wall. LV rotation, rotation rate, global torsion, and untwist rate were calculated by analysis of the basal and apical short-axis views of the left ventricle and corrected for LV length as previously described. Specific attention was paid at the time of image acquisition to make sure both basal and apical cuts were appropriate for tracking. This included avoiding lung artifacts and making sure that with the translation of the heart, the basal cut contained LV myocardium throughout the cardiac cycle. LV radial and circumferential ε was calculated at the midventricular level. Briefly, the endocardial border of the left ventricle was traced from parasternal short-axis views, and tracking was automatically performed. LV torsion was calculated as the peak instantaneous end-systolic difference between apical and basal rotation ( Figure 1 ). Images were obtained at frame rates between 50 and 100 frames/sec. One observer performed all data analysis, and measurement reproducibility was assessed in 10 randomly selected patients with blinded repeated analysis after 1 month.
RV Volumes
RV volumes, EF, and pulmonary regurgitation were obtained from clinically performed MRI (1.5-T Avanto; Siemens Medical Systems, Erlangen, Germany). A short-axis cine stack was acquired during breath-hold, to allow 20 true reconstructed phases per cardiac cycle, with slice thickness of 5 to 6 mm, 10 to 12 slices, and the gap adjusted to cover both ventricles, including the RV outflow tract. The MRI data were analyzed using commercially available software packages (Mass Analysis and CV Flow; Medis Medical Imaging Systems, Leiden, The Netherlands). Right and left cardiac indexes was calculated using the formula cardiac index (L/min/m 2 ) = heart rate × stroke volume/body surface area. Flow measurements were performed in patients with ToF to determine PI volume and fraction.
Statistical Analysis
Results are expressed as mean ± SD for continuous variables and numbers and percentages for categorical data. Two-tailed Student’s t tests were used for data analysis and comparisons between study groups. Pearson’s correlation analysis was used to study relationships between continuous variables. The association between RV parameters and LV function was assessed using multivariate regression analysis. Receiver operating characteristic analysis and two-way contingency tables, expressed as areas under the curve and odds ratios (ORs) with 95% confidence intervals (CIs), using Fisher’s exact test to assess statistical significance, were used to assess predictors of counterclockwise basal rotation. Measurement variability was assessed in a subset of patients using Bland-Altman analysis, calculating the bias (mean difference) and 95% limits of agreement and by calculation of the interclass correlation coefficient. Paired t tests were used to analyze the significance of the bias between subsequent analyses performed by the same observer at approximately 1 month. P values < .05 were considered statistically significant. The analysis was performed using SPSS Statistics version 20 (IBM, Armonk, NY).
Results
In total, 75 subjects were included in the study (50 patients after ToF repair and 25 age- and gender-matched controls). Demographic data are presented in Table 1 . Mean age of repair was 0.9 ± 0.7 years, with valve-sparing procedures in 10 patients, transannular patches in 34, and valved conduits in six. There were no significant differences between the groups in body dimensions and heart rate. However, patients with ToF had significantly longer QRS durations. Table 2 summarizes the MRI volumetric data, calculated ventricular EF, and indexed output. PI was not quantified in the control group, because none of the subjects had more than mild PI. Of the patients with ToF, there was severe PI in 31, moderate PI in 18, and mild PI in one. RV dimensions and volumes were significantly increased in patients with ToF, whereas RV EF and echocardiographic RV functional parameters were reduced. In patients with ToF, RV indexed output was almost double LV indexed output because of the significant amount of PI in all patients with ToF. There were no significant differences in LV end-systolic and end-diastolic volumes and LV EF between patients with ToF and controls. Table 3 summarizes the echocardiographic parameters measured. Significant differences were observed between patients with ToF and controls, including reduced septal and LV lateral longitudinal annular excursion and systolic tissue Doppler mitral annular velocities in patients with ToF.
| Variable | Patients with ToF ( n = 50) | Controls ( n = 25) | P |
|---|---|---|---|
| Age (y) | 13.62 ± 2.7 | 12.5 ± 3 | .14 |
| Sex (male/female) | 23/27 | 12/13 | .87 |
| Body surface area (m 2 ) | 1.44 ± 0.36 | 1.45 ± 0.29 | .89 |
| Heart rate (beats/min) | 75.1 ± 16.3 | 74.5 ± 12.9 | .85 |
| QRS duration (msec) | 142.9 ± 23.1 | 88.2 ± 9.6 | <.0001 |
| Variable | Patients with ToF ( n = 50) | Controls ( n = 25) | P |
|---|---|---|---|
| RV EDVi (mL/m 2 ) | 180 ± 40.1 | 99.1 ± 14.5 | <.001 |
| RV end-systolic volume index (mL/m 2 ) | 99.2 ± 32.6 | 47.2 ± 9.7 | <.001 |
| RV stroke volume index (mL/m 2 ) | 80.6 ± 18.7 | 53 ± 9.3 | <.001 |
| RV EF (%) | 45.8 ± 8 | 52.6 ± 4.5 | <.001 |
| RV indexed output (L/min/m 2 ) | 6 ± 1.7 | 3.9 ± 0.86 | <.001 |
| PI volume index (mL/m 2 ) | 31.4 ± 15.2 | — | |
| PI fraction (%) | 40.8 ± 10.2 | — | |
| LV EDVi (mL/m 2 ) | 85.5 ± 14.4 | 81.5 ± 18.5 | .34 |
| LV end-systolic volume index (mL/m 2 ) | 36.7 ± 9.2 | 32.9 ± 8.9 | .09 |
| LV stroke volume index (mL/m 2 ) | 48.77 ± 9.5 | 48.6 ± 12.4 | .95 |
| LV EF (%) | 57.2 ± 6.8 | 60 ± 7 | .10 |
| LV indexed output (L/min/m 2 ) | 3.6 ± 0.8 | 3.7 ± 1.1 | .90 |
| Variable | Patients with ToF ( n = 50) | Controls ( n = 25) | P |
|---|---|---|---|
| TAPSE (cm) | 1.49 ± 0.44 | 2.32 ± 4.2 | <.001 |
| SAPSE (cm) | 1.1 ± 0.2 | 1.24 ± 0.2 | .011 |
| MAPSE (cm) | 1.54 ± 0.35 | 1.36 ± 0.2 | .007 |
| DTI RV S′ (cm/sec) | 9.6 ± 2 | 14.7 ± 2.3 | <.001 |
| DTI IVS S′ (cm/sec) | 8.1 ± 1.6 | 9.6 ± 1.2 | <.001 |
| DTI LV S′ (cm/sec) | 9.7 ± 2.1 | 12.6 ± 2.3 | <.001 |
| RA end-systolic area (A4C view) (cm 2 ) | 16.5 ± 6.4 | 13.3 ± 2.9 | .004 |
| RV end-diastolic area (A4C view) (cm 2 ) | 38.5 ± 9.9 | 22.4 ± 4.3 | <.001 |
| RV end-systolic area (A4C view) (cm 2 ) | 23.3 ± 8.4 | 11.2 ± 2.6 | <.001 |
| RV fractional area of change (%) | 40.3 ± 9.7 | 49.7 ± 6.1 | <.001 |
RV longitudinal deformation parameters are presented in Table 4 . Only one apical segment (0.67% of all RV segments) could not be tracked. RV global as well as segmental longitudinal ε and strain rate were significantly reduced in patients with ToF compared with controls. The apical segment was affected most, with a progressive decrease in segmental deformation from base to apex ( P < .005).
| Variable | Patients with ToF ( n = 50) | Controls ( n = 25) | P |
|---|---|---|---|
| Global RV longitudinal ε (%) | −22.9 ± 4.6 | −32.3 ± 3.8 | <.001 |
| Basal | −27.2 ± 5.3 | −32.8 ± 6.4 | <.001 |
| Middle | −22 ± 5.2 | −31.8 ± 6.7 | <.001 |
| Apical | −19.4 ± 6.7 | −32.6 ± 5.6 | <.001 |
| Global RV longitudinal systolic strain rate (sec −1 ) | −1.4 ± 0.3 | −2.3 ± 0.5 | <.001 |
| Basal | −1.8 ± 0.5 | −2.5 ± 0.7 | <.001 |
| Middle | −1.3 ± 0.3 | −2.1 ± 0.5 | <.001 |
| Apical | −1.2 ± 0.4 | −2.2 ± 0.6 | <.001 |
Table 5 summarizes LV deformation parameters. Overall, 17 LV segments could not be reliably tracked for longitudinal ε. Of these, three septal (apical) segments (2% of septal segments) and 14 LV lateral (eight apical, five middle, and one basal) segments (9.3% of LV lateral wall segments) did not track, but there were never more than two nontracked segmentsper patient. Mean longitudinal ε in the IVS and LV lateral wall was reduced in the ToF group, with a uniform reduction in all lateral segments, whereas only the septal apical segment was significantly reduced compared with controls. There were no correlations between LV size or volume and LV ε. LV radial and circumferential ε, measured at the midventricular level, was significantly lower in the ToF population, as were all parameters of LV rotation and torsion. The most affected parameter was basal rotation, with 19 of the 50 patients with ToF demonstrating reversed (i.e., counterclockwise) basal rotation at the time of peak torsion. In contrast, apical rotation was reduced but not reversed in patients with ToF ( Figure 2 ). We further compared patients with ToF with basal clockwise versus counterclockwise rotation. Those with reversed (counterclockwise) basal rotation had higher apical rotation (9.8 ± 5.2° vs 7.1 ± 3.9°, P = .047) and higher septal longitudinal ε, mainly because of higher apical segment ε (−21.1 ± 5.6% vs −17.2 ± 4.8%, P = .01). LV lateral wall longitudinal ε and LV radial and circumferential ε were not different between these two groups. Patients with counterclockwise basal rotation also had a mildly higher pulmonary regurgitation volume compared with those with preserved clockwise rotation (36.6 ± 11.9 vs 28.2 ± 16.2 mL/m 2 , P = .058). There were no significant differences between these groups in age, type of surgery, time from surgery, QRS duration, RV end-diastolic volume index (EDVi), RV EF, and global RV longitudinal ε, while RV apical longitudinal ε was lower in those with counterclockwise basal rotation (−17.6% vs −21.6%, P = .02).
| Variable | Patients with ToF ( n = 50) | Controls ( n = 25) | P |
|---|---|---|---|
| Global LV lateral longitudinal ε (%) | −18.4 ± 8.6 | −22.1 ± 2.9 | .007 |
| Basal | −18.4 ± 10.9 | −22.8 ± 4.4 | .015 |
| Middle | −18.4 ± 7.7 | −20.9 ± 3.2 | .056 |
| Apical | −19.8 ± 4.3 | −22.7 ± 4.6 | .017 |
| Global IVS longitudinal ε (%) | −19.2 ± 3.8 | −21.4 ± 2.6 | .005 |
| Basal | −19.2 ± 4.4 | −19.9 ± 2.6 | .355 |
| Middle | −19.8 ± 3.9 | −20.8 ± 2.6 | .183 |
| Apical | −18.7 ± 5.4 | −23.6 ± 4.4 | <.001 |
| LV radial ε (%) | 44.3 ± 16.1 | 65.2 ± 15.6 | <.001 |
| LV circumferential ε (%) | −19.1 ± 2.7 | −21.6 ± 2.2 | <.001 |
| LV peak torsion (°) | 8.7 ± 4.6 | 14.7 ± 5.7 | <.001 |
| Apical rotation (°) | 8.1 ± 4.6 | 10.9 ± 4.9 | .02 |
| Basal rotation (°) | −0.9 ± 4.1 | −4.9 ± 1.9 | <.001 |
| LV torsion rate (°/sec) | 74.4 ± 34.3 | 122.8 ± 46.4 | <.001 |
| LV peak untwist rate (°/sec) | −103.6 ± 66.2 | −170.4 ± 60.3 | <.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
