Abstract
Background:
The purpose of this study was to assess the prevalence of right ventricular (RV) systolic dysfunction in adults with anatomic repair of d-transposition of great arteries (d-TGA), and to determine its relationship to clinical outcomes across multiple domains (functional status, peak oxygen consumption, NT-proBNP, and heart failure hospitalization).
Methods:
Adults with anatomic repair for d-TGA, and echocardiographic images for strain analysis were divided into 2 groups: (1) d-TGA status post arterial switch operation (d-TGA-ASO group), (2) d-TGA status post Rastelli operation (d-TGA-Rastelli group). RV systolic function was assessed using RV global longitudinal strain (RVGLS), and RV systolic dysfunction was defined as RVGLS >−18%.
Results:
We identified 151 patients (median age 21 [19–28] years; d-TGA-ASO group 89 [59%], and d-TGA-Rastelli group 62 [41%]). The mean RVGLS was −22±4%, and 47 (31%) had RV systolic dysfunction. Compared to d-TGA-ASO group, the d-TGA-Rastelli group had lower (less negative) RVGLS (−19±3% versus −25±3%, p<0.001), and higher prevalence of RV systolic dysfunction (48% versus 19%, p<0.001). RVGLS (absolute value) was associated with peak oxygen consumption (r=0.58, p<0.001; adjusted R2=0.28), log-NT-proBNP (r=−0.41, p=0.004; adjusted R2 = 0.21), New York Heart Association class III/IV (odds ratio 2.29, 1.56–3.19, p=0.01), and heart failure hospitalization (hazard ratio 0.93, 0.88–0.98, p=0.009).
Conclusions:
RV systolic dysfunction was common in adults with d-TGA and anatomic repair, and was associated with clinical outcomes. Longitudinal studies are required to determine the risk factors for progressive RV systolic dysfunction, and to identify strategies for preventing RV systolic dysfunction in this population.
Keywords: Transposition of great arteries, Right ventricular systolic function, Prognostication
INTRODUCTION
D-Transposition of great arteries (d-TGA) is characterized by ventriculoarterial discordance leading to cyanosis at birth.1 Neonatal anatomic repair is the standard of care for the management of d-TGA because it restores the morphologic left ventricle (LV) as the systemic ventricle, thereby overcoming the limitations of physiologic repair (atrial switch operation).2, 3 Of the different techniques for anatomic repair, the arterial switch operation (ASO) is the most common, while the Rastelli operation is typically reserved for the patients with d-TGA and associated pulmonic stenosis/LV outflow tract obstruction.2–4 The long-term complications after anatomic repair of d-TGA include neo-aortic regurgitation and neo-aortic root aneurysm, coronary ostial stenosis, and right ventricular (RV) outflow tract and supravalvular pulmonic stenosis.5–8
Several studies have assessed the impact of myocardial ischemia from coronary ostial stenosis and volume overload from neo-aortic regurgitation on the systemic LV in patients with d-TGA and anatomic repair.7, 9, 10 However, there are limited data about RV adaptation to chronic pressure overload resulting from RV outflow tract and supravalvular pulmonic stenosis in this population.8, 11 The purpose of this study was to assess the prevalence of RV systolic dysfunction in adults with anatomic repair of d-TGA, and to determine its relationship to clinical outcomes across multiple domains.
METHODS
Study Population
This is a retrospective study of adults (age >18 years) with d-TGA and neonatal anatomic repair that received care at Mayo Clinic, Rochester, MN from January 1, 2003, and December 31, 2021. The patients without adequate images for the offline assessment of RV strain were exclude, and the patients that met the inclusion criteria were divided into 2 groups based on the type of anatomic repair: (1) d-TGA status post ASO (d-TGA-ASO group), and (2) d-TGA status post Rastelli operation (d-TGA-Rastelli group). Simple and complex d-TGA were defined as the absence or presence of associated structural lesions such as ventricular septal defect, pulmonic stenosis or coarctation of aorta, respectively.
Data Collection
The medical records, including clinic notes, echocardiograms, cross-sectional imaging, surgical notes, and cardiopulmonary exercise tests, were reviewed. The first clinical encounter in the adult congenital heart disease clinic after January 1, 2003, was used as the baseline encounter, and the clinical and imaging data obtained within 12 months from the baseline encounter were used to define the baseline characteristics of the cohort.
Echocardiography
We used the first echocardiogram performed within the study period as the baseline echocardiogram, and offline image analyses were performed in all patients. RV function was assessed using speckle tracking strain imaging obtained using Vivid E9 and E95 (General Electric Co, Fairfield, Connecticut) with M5S and M5Sc-D transducers (1.5–4.6 MHz) at frame rate of 40 to 80 Hz. These images were exported (DICOM) and then analyzed offline using TomTec (TomTec Imaging Systems, Unterschleissheim, Germany). Three-beat cine-loop clips were obtained from RV-focused apical four-chamber views with the reference point placed at the beginning of the QRS complex. The endocardial border was traced automatically by the software after setting the reference points at the septal and lateral borders of the tricuspid annulus and the apex. The automatic tracings were adjusted manually to ensure optimal tracking throughout the cardiac cycle.12 The RV global longitudinal strain (RVGLS) was used as the primary metric of RV systolic function, and was calculated as the average value from the 3 RV free wall segments and the ventricular septum. RV systolic dysfunction was defined as RVGLS >−18% (less negative than −18%).13 Previous studies from our group have demonstrated good intra-observer and inter-observer agreements for RVGLS (intraclass correlation coefficient 0.86–0.90 and 0.81–0.82, respectively), using similar techniques.14, 15
In addition to RVGLS, we also assessed the following parameters as secondary indices of RV systolic function: RV fractional area change (RVFAC); RV tissue Doppler systolic velocity (RV s’); and tricuspid annular plane systolic excursion (TAPSE). Based on these indices, we defined RV systolic dysfunction as RVFAC <35%, RV s’ <10 cm/s, and TAPSE <16 mm.16 In addition to RV systolic function, we also assessed chamber structure, function, and hemodynamics based on standard techniques according to contemporary guidelines.
Clinical Outcomes
The following outcomes were assessed: (1) New York Heart Association (NYHA) functional class, and we defined impaired functional status as NYHA class III/IV. (2) Peak oxygen consumption as a measure of aerobic capacity, and the peak oxygen consumption was derived from a maximum effort cardiopulmonary exercise test with respiratory exchange ratio >1.1 performed within 12 months from the baseline encounter. (3) N-terminal pro-brain natriuretic peptide (NT-proBNP) as a measure of neurohormonal activation. (4) Heart failure hospitalization, defined hospitalization for volume overload requiring intravenous diuretics.
Statistical Analysis
Data were presented as mean ± standard deviation, median (interquartile range [IQR]), and count (%). Normality was assessed using Shapiro-Wilk test of normality. Between-group comparisons were performed using Fisher’s exact test, unpaired t-test, and Wilcoxon rank sum test, as appropriate. The relationships between RVGLS and outcomes (peak oxygen consumption, NT-proBNP, NYHA III/IV, and heart failure hospitalization) were assessed using linear regression, logistic regression, and Cox regression as appropriate. The regression models were adjusted for age, sex, LV global longitudinal strain (LVGLS), and valve function (tricuspid regurgitation, pulmonary valve systolic mean gradient, pulmonary regurgitation, and aortic regurgitation). Both RVGLS and LVGLS were modeled as absolute (i.e., positive) values, so that higher numerical value signified better systolic function. The covariates in the final model were chosen using stepwise backwards selection with a p<0.1 required for a covariate to remain in the model. All statistical analyses were performed with BlueSky Statistics software (version. 7.10; BlueSky Statistics LLC, Chicago, IL, USA), and SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). A p value <0.05 was considered statistically significant for all analyses
RESULTS
Baseline Characteristics
Of 166 patients with d-TGA that underwent anatomic repair, offline analysis of RV strain was feasible in 151 (91%) patients, and these patients comprised our study group. The median age at the time of baseline encounter was 21 (IQR19–28) years, 93 (62%) were males, and 99 (66%) had complex d-TGA (Table 1).
Table 1:
Baseline Characteristics
| All (n=151) |
d-TGA-ASO (n=89, 59%) |
d-TGA-Rastelli (n=62, 41%) |
p | |
|---|---|---|---|---|
|
| ||||
| Age, years | 21 (19–28) | 20 (18–22) | 27 (19–33) | <0.001 |
| Male sex | 93 (62%) | 54 (61%) | 39 (63%) | 0.2 |
| Body mass index, kg/m2 | 26±6 | 25±5 | 26±5 | 0.8 |
| Associated structural lesions | ||||
| Isolated d-TGA | 52 (34%) | 52 (58%) | 0 | --- |
| Complex d-TGA | 99 (66%) | 37 (42%) | 62 (100%) | --- |
| Ventricular septal defect | 95 (63%) | 33 (37%) | 62 (100%) | --- |
| Pulmonic stenosis | 63 (42%) | 1 (1%) | 62 (100%) | --- |
| Coarctation of aorta | 8 (5%) | 7 (8%) | 1 (2%) | --- |
| Cardiac procedures prior to repair | ||||
| Balloon atrial septostomy | 74 (49%) | 53 (60%) | 21 (34%) | <0.001 |
| Blalock-Taussig shunt | 10 (7%) | 1 (1%) | 9 (15%) | 0.002 |
| Pulmonary artery banding | 16 (11%) | 13 (15%) | 3 (5%) | 0.06 |
| Surgical repair of coarctation of aorta | 8 (5%) | 7 (8%) | 1 (2%) | 0.09 |
| Age at complete repair, weeks | 1 (1–3) | 1 (0–2) | 1 (1–3) | 0.1 |
| Lecompte maneuver | --- | 54 (62%) | --- | --- |
| Cardiac procedures after repair | ||||
| PVR/RV-PA conduit replacement | 69 (46%) | 7 (8%) | 62 (100%) | <0.001 |
| PA plasty | 9 (6%) | 1 (1%) | 8 (13%) | 0.006 |
| Branch PA stenting | 5 (3%) | 4 (5%) | 1 (2%) | 0.3 |
| Coarctation of aorta reintervention | 2 (1%) | 2 (2%) | 0 | --- |
| Aortic valve/root surgery | 17 (11%) | 13 (15%) | 4 (7%) | 0.1 |
| Aortic valve replacement | 14 (9%) | 10 (11%) | 4 (7%) | 0.2 |
| Aortic valve repair | 3 (2%) | 3 (3%) | 0 | --- |
| Aortic root repair/replacement | 12 (8%) | 10 (11%) | 2 (3%) | 0.3 |
| CIED implantation | ||||
| Pacemakers | 3 (2%) | 1 (1%) | 2 (3%) | 0.2 |
| ICD | 5 (3%) | 1 (1%) | 4 (7%) | 0.2 |
| Tricuspid valve repair | 3 (2%) | 0 | 3 (4%) | --- |
| Tricuspid valve replacement | 6 (4%) | 2 (2%) | 4 (7%) | 0.4 |
| Comorbidities | ||||
| Hypertension | 8 (5%) | 5 (6%) | 3 (5%) | 0.9 |
| Diabetes | 6 (4%) | 1 (1%) | 5 (8%) | 0.7 |
| Medications | ||||
| Beta blockers | 24 (16%) | 9 (10%) | 15 (24%) | 0.02 |
| ACEI/ARB | 26 (17%) | 18 (20%) | 8 (13%) | 0.1 |
| Mineralocorticoid antagonist | 1 (0.6%) | -- | 1 (2%) | -- |
| Loop diuretics | 12 (8%) | 2 (2%) | 10 (16%) | 0.002 |
| Laboratory data | ||||
| Estimated GFR, ml/min/1.73m2 | 80±26 | 89±17 | 67±21 | 0.008 |
| NT-proBNP, pg/ml [N=104] | 216 (89–801) | 128 (77–234) | 499 (145–914) | 0.006 |
TGA: transposition of great arteries; ASO: arterial switch operation; PVR: pulmonary valve replacement; RV-PA: right ventricular to pulmonary artery; ACEI: angiotensin converting enzyme inhibitor; ARB: angiotensin-II receptor blocker; GFR: glomerular filtration rate; NT proBNP: N-terminal pro hormone brain natriuretic peptide; CIED: cardiac implantable electronic devices; ICD: internal cardioverter defibrillator; [N] denotes number of patients with available data.
Cardiac Interventions
Cardiac Interventions Prior to Complete Repair
The median age at the time of complete repair was 1 (1–3) weeks, and 89 (59%) underwent ASO (d-TGA-ASO group) while 62 (41%) patients underwent Rastelli operation (d-TGA-Rastelli group). Of the 151 patients, 89 (59%) had palliative procedures prior to complete repair, and the palliative procedures were balloon atrial septostomy (n=74, 49%), Blalock-Taussig shunt (n=10, 7%), and pulmonary artery banding (16, 11%). Table 1 show a comparison of the baseline characteristics and cardiac interventions between the 2 groups. Compared to the d-TGA-ASO group, the patients in the d-TGA-Rastelli group were older and more likely to have had Blalock-Taussig shunt prior to complete repair.
Cardiac Interventions After Complete Repair but Prior to Baseline Encounter
Of the 151 patients, 86 (57%) had at least one cardiac re-intervention between the time of complete repair and baseline encounter in the adult congenital heart disease clinic. The most common re-interventions were pulmonary valve replacement/RV-pulmonary artery conduit replacement (n=69, 46%) and aortic valve/root surgery (n=17, 11%).
Cardiac Interventions After Baseline Encounter
In addition to the 86 patients that had cardiac re-intervention prior to the baseline encounter, 13 (9%) patients underwent cardiac re-interventions between the baseline encounter and the end of the study period. The cardiac re-interventions were aortic valve/root replacement (n=3), valve sparing aortic root replacement (n=1), surgical pulmonary valve replacement (n=3), transcatheter pulmonary valve replacement (n=4), and surgical tricuspid valve repair (n=3).
RV Systolic Function
The mean RVGLS was −22±4%, and 47 (31%) had RV systolic dysfunction (RVGLS <−18%), Figure 1. Of the secondary indices of RV systolic function assessed, the mean RVFAC, TAPSE, RV s’ were 43±13%, 19±5 mm, and 13±3 cm/s, respectively. The prevalence of RV systolic dysfunction was 21% based on RVFAC (RVFAC <35%: 32/151), 20% based on TAPSE (TAPSE <16 mm: 28/142), and 23% based on RV s’ (RV s’ <10 cm/s: 33/141). Table 2 shows a comparison of the imaging data between the 2 groups. All the RV function indices were significantly worse in the d-TGA-Rastelli group as compared to the d-TGA-ASO group.
Figure 1:
Prevalence and prognostic implications of right ventricular systolic dysfunction (RVSD). (A) Box and whisker plot showing absolute values (without the negative sign) of RV global longitudinal strain (RVGLS) in all patients (red), patients with arterial switch operation (ASO) (black), and patients with Rastelli operation (grey). (B) Bar graph showing the prevalence of RVSD defined as RVGLS >−18% in all patients (red), patients with ASO (black), and patients with Rastelli operation (grey). P values signify between-group comparisons for ASO versus Rastelli groups. (C) Associations between RVGLS and outcomes.
VO2: oxygen consumptions; NT-proBNP: N-terminal pro-brain natriuretic peptide; NYHA: New York Heart Association; HF: heart failure; OR: odds ratio; HR: hazard ratio, r: correlation coefficient
Table 2:
Imaging Data
| All (n=151) |
d-TGA-ASO (n=89, 59%) |
d-TGA-Rastelli (n=62, 41%) |
p | |
|---|---|---|---|---|
|
| ||||
| Right Heart Indices | ||||
| RA volume index, ml/m2 | 26±14 | 23±11 | 32±8 | 0.008 |
| Estimated RA pressure, mmHg | 5 (3–8) | 3 (3–5) | 5 (5–8) | <0.001 |
| Estimated RV systolic pressure, mmHg | 41 (32–61) | 36 (30–47) | 54 (39–75) | <0.001 |
| RV global longitudinal strain, % | -22±4 | -25±3 | -19±3 | <0.001 |
| RV s’, cm/s | 13±3 | 15±3 | 11±3 | 0.009 |
| TAPSE, mm | 19±5 | 22±5 | 17±4 | 0.006 |
| RV fractional area change, % | 43±13 | 49±10 | 38±9 | <0.001 |
| RV end-diastolic area, cm2 | 25±8 | 23±5 | 32±6 | 0.008 |
| RV end-systolic area, cm2 | 16±4 | 11±3 | 21±4 | 0.006 |
| ≥Moderate tricuspid regurgitation | 19 (13%) | 3 (3%) | 16 (26%) | <0.001 |
| ≥Moderate pulmonary regurgitation | 18 (12%) | 7 (8%) | 11 (18%) | 0.06 |
| Pulmonary valve mean gradient | 14 (11–27%) | 11 (8–17) | 21 (16–29) | <0.001 |
| Left Heart Indices | ||||
| LA volume index, ml/m2 | 28±11 | 26±9 | 30±10 | 0.1 |
| LV longitudinal strain, % | -21±4 | -22±4 | 21±3 | 0.6 |
| LV end-diastolic volume index, ml/m2 | 53±15 | 55±11 | 51±12 | 0.1 |
| LV end-systolic volume index, ml/m2 | 34±7 | 35±8 | 32±7 | 0.1 |
| LV ejection fraction, % | 57±9 | 58±8 | 56±7 | 0.3 |
| ≥Moderate aortic regurgitation | 12 (8%) | 8 (9%) | 4 (7%) | 0.7 |
| CMRI Volumetric Indices [N=57] | ||||
| RV end-diastolic volume index, ml/m2 | 103±19 | 96±16 | 112±17 | 0.07 |
| RV end-systolic volume index, ml/m2 | 56±14 | 47±12 | 68±16 | 0.008 |
| RV ejection fraction, % | 48±16 | 54±12 | 41±14 | 0.003 |
| LV end-diastolic volume index, ml/m2 | 98±15 | 101±13 | 96±12 | 0.3 |
| LV end-systolic volume index, ml/m2 | 41±11 | 42±9 | 40±10 | 0.6 |
| LV ejection fraction, % | 57±13 | 59±14 | 55±12 | 0.4 |
| Aortic dimensions | ||||
| Echo-derived aortic root, mm | 47±5 | 49±45 | 44±4 | 0.08 |
| Echo-derived mid-ascending aorta, mm | 44±5 | 45±4 | 43±5 | 0.2 |
| CMRI/CT-derived aortic root, mm | 48±5 | 51±6 | 45±4 | 0.04 |
| CMRI/CT-derived mid-ascending aorta, mm | 45±4 | 46±5 | 44±4 | 0.3 |
d-TGA: dextro transposition of great arteries; ASO: arterial switch operation; LA: left atrium; RV: right ventricle; RA: right atrium; LV: left ventricle; TAPSE: tricuspid annular plane systolic excursion; s’: tissue Doppler systolic velocity; CMRI: cardiac magnetic resonance imaging; CT: computer tomography
Cardiac magnetic resonance imaging (CMRI) data was available in only 57 (38%) patients, and the mean RV ejection faction based on CMRI was 48±16%. There was a correlation between the RVGLS and CMRI-derived RVEF (r=0.61, p<0.001).
Outcomes
Aerobic Capacity
Of the 151 patients, 93 (62%) patients underwent cardiopulmonary exercise test, and the mean peak oxygen consumption was 28±9 ml/kg/min, corresponding to 67±13 percent of predicted. There was a correlation between RVGLS and peak VO2 (r=0.58, p<0.001), and this correlation remained significant after adjustment for age, sex, LVGLS, valve dysfunction (tricuspid regurgitation, pulmonary valve mean gradient, pulmonary regurgitation, and aortic regurgitation) (adjusted R2 = 0.28).
Neurohormonal Activation
NT-proBNP was measured in 104 (69%) patients, and median NT-proBNP was 216 (89–801) pg/ml. There was a correlation between RVGLS and log-NT-proBNP (r=−0.41, p=0.004), and this correlation remained significant after adjustment for age, sex, LVGLS, and valve dysfunction (tricuspid regurgitation, pulmonary valve mean gradient, pulmonary regurgitation, and aortic regurgitation) (adjusted R2 = 0.21).
Functional Status
Of the 151 patients, 26 (17%) were in NYHA class III/IV at the time of baseline encounter. RVGLS (odds ratio 0.94, 95% confidence interval 0.91–0.97, p=0.006) and moderate tricuspid regurgitation (odds ratio 2.29, 95% confidence interval 1.56–3.19, p=0.01) were associated with NYHA III/IV functional class, after adjustment for age, sex, LVGLS, and valve dysfunction (tricuspid regurgitation, pulmonary valve mean gradient, pulmonary regurgitation, and aortic regurgitation) (Table 3).
Table 3:
Multivariable Logistic Regression Model Showing the Clinical and Hemodynamic Correlates of NYHA III/IV
| OR (95%CI) | p | |
|---|---|---|
|
| ||
| RVGLS, % | 0.94 (0.91–0.97) | 0.006 |
| Age, years | --- | --- |
| Male sex | --- | --- |
| LVGLS, % | --- | --- |
| ≥Moderate tricuspid regurgitation | 2.29 (1.56–3.19) | 0.01 |
| ≥Moderate pulmonary regurgitation | --- | -- |
| Pulmonary valve mean gradient, mmHg | --- | --- |
| Aortic regurgitation | --- | --- |
RVGLS: right ventricular global longitudinal strain; LVGLS: left ventricular global longitudinal strain; OR: odds ratio; CI: confidence interval; NYHA: New York Heart Association
Note that both RVGLS and LVGLS were modeled as absolute values, i.e., ignoring the negative (-) sign so that higher numerical value signified better systolic function. The selection of covariates was based on stepwise backwards selection, and a p<0.1 was used at the threshold for a covariate to remain in the model. --- signifies covariates with p ≥0.1.
Heart Failure Hospitalization
Of the 151 patients, 9 (6%) were hospitalized for heart failure during a median follow-up of 5.6 (3.1–8.9) years. Eight of the 9 hospitalizations for heart failure occurred in the setting of atrial arrhythmias. RVGLS (hazard ratio 0.93, 95% confidence interval 0.88–0.98, p=0.009), older age (hazard ratio 1.06, 95% confidence interval 1.02–1.10, p=0.007), and moderate tricuspid regurgitation (hazard ratio 1.74, 95% confidence interval 1.26–2.08, p=0.02) were associated with heart failure hospitalization after adjustment for sex, LVGLS, RV systolic pressure, and valve dysfunction (tricuspid regurgitation, pulmonary valve mean gradient, pulmonary regurgitation, and aortic regurgitation (Table 4).
Table 4:
Multivariable Cox Regression Model Showing the Clinical and Hemodynamic Correlates of Heart Failure Hospitalization
| HR (95%CI) | p | |
|---|---|---|
|
| ||
| RVGLS, % | 0.93 (0.88–0.98) | 0.009 |
| Age, years | 1.06 (1.02–1.10) | 0.007 |
| Male sex | --- | --- |
| LVGLS, % | --- | --- |
| ≥Moderate tricuspid regurgitation | 1.74 (1.26–2.08) | 0.02 |
| ≥Moderate pulmonary regurgitation | --- | -- |
| Pulmonary valve mean gradient, mmHg | --- | --- |
| Aortic regurgitation | --- | --- |
| Cardiac intervention | 0.96 (0.91–1.01) | 0.08 |
RVGLS: right ventricular global longitudinal strain; LVGLS: left ventricular global longitudinal strain; HR: hazard ratio; CI: confidence interval
Note that both RVGLS and LVGLS were modeled as absolute values, i.e., ignoring the negative (-) sign so that higher numerical value signified better systolic function. The selection of covariates was based on stepwise backwards selection, and a p<0.1 was used at the threshold for a covariate to remain in the model. --- signifies covariates with p ≥0.1. Cardiac intervention was modeled as a time-dependent covariate.
Subgroup Analysis
d-TGA-ASO group
The mean RVGLS was −25±3%, and 17 (19%) had RV systolic dysfunction (RVGLS >−18%), Figure 1. Based on the secondary indices of RV systolic function, the prevalence of RV systolic dysfunction was 12% for RVFAC (RVFAC <35%:11/89), 10% for TAPSE (TAPSE <16 mm: 8/83), and 11% for RV s’ (RV s’ <10 cm/s: 9/82). Table 5 shows the clinical and hemodynamic correlates of RVGLS at baseline echocardiogram in the d-TGA-ASO group. Older age (β±SE −0.06±0.04 per year, p=0.01) and Doppler derived RV systolic pressure (β±SE −0.21±0.16 per 1 mmHg, p<0.001) were associated with RVGLS after multivariable adjustments.
Table 5:
Multivariable Linear Regression Model Showing the Clinical and Hemodynamic Correlates of RVGLS in Patients with d-TGA-ASO
| β±SE | p | β±SE | p | |
|---|---|---|---|---|
|
| ||||
| Demographic indices | ||||
| Age, years | -0.08±0.04 | 0.002 | -0.06±0.04 | 0.01 |
| Male sex | -0.03±0.09 | 0.3 | ||
| Anatomic/surgical history | ||||
| Complex TGA | -1.65±1.84 | 0.3 | ||
| Prior pulmonary artery banding | -2.06±1.18 | 0.07 | --- | --- |
| Prior palliative shunt | 1.03±2.15 | 0.6 | ||
| Lecompte maneuver | 0.06±0.14 | 0.2 | ||
| PVR/RV-PA conduit replacement | -4.17±2.41 | <0.001 | --- | --- |
| Number of cardiac surgeries | -0.48±1.82 | 0.4 | ||
| Echocardiographic indices | ||||
| Estimated RV systolic pressure, mmHg | -0.29±0.08 | <0.001 | -0.21±0.16 | <0.001 |
| ≥Moderate tricuspid regurgitation | -3.28±1.05 | <0.001 | --- | --- |
| ≥Moderate pulmonary regurgitation | -1.74±1.98 | 0.4 | ||
| Pulmonary valve mean gradient, mmHg | -0.24±0.09 | 0.006 | --- | --- |
| LVGLS, % | 0.29±0.12 | 0.03 | --- | --- |
TGA: transposition of great arteries; ASO: arterial switch operation; RV: right ventricle; PA: pulmonary artery; PVR: pulmonary replacement; LVGLS: left ventricular global longitudinal strain; RVGLS: right ventricular global longitudinal strain
Note that both RVGLS and LVGLS were modeled as absolute values, i.e., ignoring the negative sign so that higher numerical value signified better systolic function.
d-TGA-Rastelli group
The mean RVGLS was −19±3%, and 30 (48%) had RV systolic dysfunction (RVGLS >−18%), Figure 1. Based on the secondary indices of RV systolic function, the prevalence of RV systolic dysfunction was 34% for RVFAC (RVFAC <35%:21/62), 34% for TAPSE (TAPSE <16 mm: 20/59), and 49% for RV s’ (RV s’ <10 cm/s: 29/59). Table 6 shows the clinical and hemodynamic correlates of RVGLS at baseline echocardiogram. Older age (β±SE −0.11±0.08 per year, p=0.02), Doppler derived RV systolic pressure (β±SE −0.15±0.06 per 1 mmHg, p<0.001), and moderate tricuspid regurgitation (β±SE −1.19±0.73, p=0.004), were associated with RVGLS after multivariable adjustments.
Table 6:
Multivariable Linear Regression Model Showing the Clinical and Hemodynamic Correlates of RVGLS in Patients with d-TGA-Rastelli
| β±SE | p | β±SE | p | |
|---|---|---|---|---|
|
| ||||
| Demographic indices | ||||
| Age, years | -0.06±0.02 | 0.006 | -0.11±0.08 | 0.02 |
| Male sex | -0.02±0.1 | 0.4 | ||
| Anatomic/surgical history | ||||
| Prior palliative shunt | -1.14±1.21 | 0.3 | ||
| Number of cardiac surgeries | -0.48±1.82 | 0.4 | ||
| Echocardiographic indices | ||||
| Estimated RV systolic pressure, mmHg | -0.22±0.03 | <0.001 | -0.15±0.06 | <0.001 |
| ≥Moderate tricuspid regurgitation | -2.11±0.84 | <0.001 | -1.19±0.73 | 0.004 |
| ≥Moderate pulmonary regurgitation | 1.21±1.55 | 0.3 | ||
| Pulmonary valve mean gradient, mmHg | -0.03±0.65 | 0.5 | ||
| LVGLS, % | 0.17±0.1- | 0.03 | --- | --- |
TGA: transposition of great arteries; RV: right ventricle; LVGLS: left ventricular global longitudinal strain; RVGLS: right ventricular global longitudinal strain
Note that both RVGLS and LVGLS were modeled as absolute values, i.e., ignoring the negative sign so that higher numerical value signified better systolic function.
Compared to the d-TGA-ASO group, the d-TGA-Rastelli group had a higher prevalence of RV systolic dysfunction as measured by RVGLS, RVFAC, TAPSE, and RV s’ (p<0.05 for all), higher prevalence of functional impairment (NYHA class III/IV (26% [16/62] versus 11% [10/89], p=0.02), higher NT-proBNP (499 [145–914] versus 128 [77–234], pg/ml, p=0.006), lower peak oxygen consumptions (61±8 versus 72±10 percent of predicated, p=0.008), and higher incidence of heart failure hospitalization (11% [7/62] versus 2% [2/89], p=0.02).
DISCUSSION
In this study, we assessed the prevalence, correlates, and prognostic implications of RV systolic dysfunction in adults with d-TGA and anatomic repair. The main findings were: (1) The prevalence of RV systolic dysfunction was 31% based on RVGLS and ranged from 20–23% based on other echocardiographic indices of RV systolic function. (2) RVGLS was associated with clinical outcomes across multiple domains, and there was an inverse correlation between RVGLS and RV systolic pressure, a surrogate of RV afterload. (3) Compared to the d-TGA-ASO group, the d-TGA-Rastelli group had worse RV systolic function, and worse clinical outcomes (functional status, aerobic capacity, neurohormonal activation, and heart failure hospitalization).
Atrial switch operation (physiologic repair) was associated suboptimal long-term outcome because of systemic RV failure, atrial baffle dysfunction, and a high prevalence of atrial arrhythmias.17, 18 In contrast, anatomic repair provides superior long-term outcomes as compared to physiologic repair because it restored the morphologic LV as the systemic ventricle, thereby avoiding the complications and limitations of the atrial switch operation.2–4, 18 Notwithstanding, adults with prior anatomic repair remain vulnerable to certain long-term complications such as supravalvular pulmonic stenosis and RV outflow tract/conduit dysfunction, and these complications expose the RV to chronic pressure and volume overload.5–8 Supravalvular pulmonic stenosis and RV outflow tract/conduit dysfunction are relatively common after anatomic repair, and about 20–28% of patients would require surgical or transcatheter interventions because of these complications within the first 2 decades of life after an ASO.8, 11 These concerns are even more pertinent in adults with prior Rastelli operation because of the need for pulmonary conduit re-interventions.19, 20
Prevalence of RV Systolic Dysfunction
There are limited data about the RV remodeling and dysfunction in this population. In a retrospective study of 220 patients with prior ASO, Shepard et al demonstrated that 5% of their cohort had RV systolic dysfunction, defined as RV ejection fraction <50% based on CMRI.7 In a different study, Pettersen et al showed that patients with d-TGA that underwent ASO had lower RVGLS as compared to age matched controls.21 The prevalence of RV systolic dysfunction observed in the current study was significantly higher than that of previous studies.7, 21–23 There are 3 plausible explanations for the observed differences in the prevalence of RV systolic dysfunction between the current study and prior studies. First, our study was based on a mixed cohort of patients that underwent ASO and those that had Rastelli operation. Since the patients with prior Rastelli operation had RV outflow disease at birth and were subjected to the hemodynamic burden of RV outflow conduit dysfunction from early childhood, we would expect more RV systolic dysfunction in this group. Second, the patients in our cohort were relatively older (median age of 20 years) as compared to the patients reported in previous studies (median age of 12–16 years), and hence potentially had a longer duration of exposure to RV pressure overload.7, 21–23 Third, we defined RV systolic function based on RVGLS, which is a more sensitive measure of RV systolic function, as shown in previous studies demonstrating superior diagnostic performance of RVGLS for detecting subclinical RV systolic dysfunction that may not be apparent using other indices of RV systolic function.13, 24
Prognostic Implications of RV Systolic Dysfunction
Another important observation from the current study was that RV systolic function (rather than LV systolic function) was associated with clinical outcomes across multiple domains (functional status, aerobic capacity, neurohormonal activation, and heart failure hospitalization). Most of the hemodynamic and outcomes studies in this population have focused mostly on LV systolic function, and on the determinants of LV systolic function such as myocardial ischemia from coronary ostial stenosis and volume overload from aortic regurgitation.7, 9, 10 The current study suggests that the RV, rather than, the LV, may be the more important determinant of outcomes in this population.
Determinants of RV Systolic Dysfunction
We identified RV systolic pressure, tricuspid regurgitation, and older age as correlates of RVGLS suggesting that chronic RV pressure and/or volume overload might contribute to the pathogenesis of RV systolic dysfunction in this population. However, we cannot infer causality or temporality from a cross-sectional analysis. There is, therefore, a need for longitudinal studies to assess the risk factors for progressive RV systolic dysfunction, and potential interventions to prevent or reverse RV systolic dysfunction in this population.
Limitations
This is a retrospective single center study, and hence it is prone to selection, and ascertainment bias. The clinical and hemodynamic correlates of RVGLS reported in this study were based on a cross-sectional analysis, and hence do not provide mechanistic insight into the etiology and pathogenesis of RV systolic dysfunction in this population Cardiac magnetic resonance imaging, which is considered as the gold standard for RV assessment, was only available in one-third of the patent, and this limits the robustness of our results. Finally, we adjusted for a limited number of covariates in the multivariable models because of small sample size and limited number of events (there were only 9 patients that were hospitalized for heart failure).
Conclusions
RV systolic dysfunction, as measured by RVGLS, was common in adults with d-TGA and prior anatomic repair, and RVGLS was associated with clinical outcomes. The clinical and hemodynamic correlates of RVGLS were RV systolic pressure, tricuspid regurgitation, and older age, suggesting that chronic RV pressure and/or volume overload might contribute to the pathogenesis of RV systolic dysfunction in this population. The patients with prior Rastelli operation had worse right heart indices and clinical outcome suggesting a more aggressive natural history. Longitudinal studies are required to determine the risk factors for progressive RV systolic dysfunction, and to identify strategies for preventing RV systolic dysfunction in this population.
Funding:
Dr. Egbe is supported by National Heart, Lung, and Blood Institute (NHLBI) grants (R01 HL158517 and R01 HL160761). The MACHD Registry is supported by the Al-Bahar Research grant.
Abbreviations:
- TGA
transposition of great arteries
- LV
left ventricle
- RV
right ventricle
- ASO
arterial switch operation
- RVGLS
right ventricular global longitudinal strain
- FAC
fractional area change
- s’
tissue Doppler systolic velocity
- TAPSE
tricuspid annular plane systolic excursion
- CMRI
Cardiac magnetic resonance imaging
- NT-proBNP
N-terminal pro-brain natriuretic peptide
- IQR
interquartile range
- LVGLS
left ventricular global longitudinal strain
- NYHA
New York Heart Association
Footnotes
Conflict of Interest: none
Disclosures: none
REFERENCES
- 1.Warnes CA. Transposition of the great arteries. Circulation. 2006;114:2699–709. [DOI] [PubMed] [Google Scholar]
- 2.Colan SD, Boutin C, Castaneda AR and Wernovsky G. Status of the left ventricle after arterial switch operation for transposition of the great arteries. Hemodynamic and echocardiographic evaluation. The Journal of thoracic and cardiovascular surgery. 1995;109:311–21. [DOI] [PubMed] [Google Scholar]
- 3.Wernovsky G, Hougen TJ, Walsh EP, Sholler GF, Colan SD, Sanders SP, Parness IA, Keane JF, Mayer JE, Jonas RA and et al. Midterm results after the arterial switch operation for transposition of the great arteries with intact ventricular septum: clinical, hemodynamic, echocardiographic, and electrophysiologic data. Circulation. 1988;77:1333–44. [DOI] [PubMed] [Google Scholar]
- 4.Rastelli GC, Wallace RB and Ongley PA. Complete repair of transposition of the great arteries with pulmonary stenosis. A review and report of a case corrected by using a new surgical technique. Circulation. 1969;39:83–95. [DOI] [PubMed] [Google Scholar]
- 5.van der Palen RLF, Blom NA, Kuipers IM, Rammeloo LAJ, Jongbloed MRM, Konings TC, Bouma BJ, Koolbergen DR and Hazekamp MG. Long-term outcome after the arterial switch operation: 43 years of experience. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery. 2021;59:968–977. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Angeli E, Formigari R, Pace Napoleone C, Oppido G, Ragni L, Picchio FM and Gargiulo G. Long-term coronary artery outcome after arterial switch operation for transposition of the great arteries. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery. 2010;38:714–20. [DOI] [PubMed] [Google Scholar]
- 7.Shepard CW, Germanakis I, White MT, Powell AJ, Co-Vu J and Geva T. Cardiovascular Magnetic Resonance Findings Late After the Arterial Switch Operation. Circulation Cardiovascular imaging. 2016;9. [DOI] [PubMed] [Google Scholar]
- 8.Cleuziou J, Vitanova K, Pabst von Ohain J, Ono M, Tanase D, Burri M and Lange R. Incidence and Risk Factors for Right Ventricular Outflow Tract Obstruction after the Arterial Switch Operation. The Thoracic and cardiovascular surgeon. 2019;67:37–43. [DOI] [PubMed] [Google Scholar]
- 9.Hui L, Chau AK, Leung MP, Chiu CS and Cheung YF. Assessment of left ventricular function long term after arterial switch operation for transposition of the great arteries by dobutamine stress echocardiography. Heart. 2005;91:68–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang C, Li VW, So EK and Cheung YF. Left Ventricular Stiffness in Adolescents and Young Adults After Arterial Switch Operation for Complete Transposition of the Great Arteries. Pediatr Cardiol. 2020;41:747–754. [DOI] [PubMed] [Google Scholar]
- 11.Nellis JR, Turek JW, Aldoss OT, Atkins DL and Ng BY. Intervention for Supravalvar Pulmonary Stenosis After the Arterial Switch Operation. The Annals of thoracic surgery. 2016;102:154–62. [DOI] [PubMed] [Google Scholar]
- 12.Badano LP, Kolias TJ, Muraru D, Abraham TP, Aurigemma G, Edvardsen T, D’Hooge J, Donal E, Fraser AG, Marwick T, Mertens L, Popescu BA, Sengupta PP, Lancellotti P, Thomas JD, Voigt JU, Industry r and Reviewers: This document was reviewed by members of the ESDC. Standardization of left atrial, right ventricular, and right atrial deformation imaging using two-dimensional speckle tracking echocardiography: a consensus document of the EACVI/ASE/Industry Task Force to standardize deformation imaging. European heart journal cardiovascular Imaging. 2018;19:591–600. [DOI] [PubMed] [Google Scholar]
- 13.Morris DA, Krisper M, Nakatani S, Kohncke C, Otsuji Y, Belyavskiy E, Radha Krishnan AK, Kropf M, Osmanoglou E, Boldt LH, Blaschke F, Edelmann F, Haverkamp W, Tschope C, Pieske-Kraigher E, Pieske B and Takeuchi M. Normal range and usefulness of right ventricular systolic strain to detect subtle right ventricular systolic abnormalities in patients with heart failure: a multicentre study. European heart journal cardiovascular Imaging. 2017;18:212–223. [DOI] [PubMed] [Google Scholar]
- 14.Egbe A, Miranda W, Connolly H and Dearani J. Haemodynamic determinants of improved aerobic capacity after tricuspid valve surgery in Ebstein anomaly. Heart. 2021;107:1138–1144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Egbe AC, Miranda WR and Connolly HM. Role of Echocardiography for Assessment of Cardiac Remodeling in Congenitally Corrected Transposition of Great Arteries. Circulation Cardiovascular imaging. 2022;15:e013477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, Solomon SD, Louie EK and Schiller NB. Guidelines for the echocardiographic assessment of the right heart in adults: a report from the American Society of Echocardiography endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. Journal of the American Society of Echocardiography : official publication of the American Society of Echocardiography. 2010;23:685–713; quiz 786–8. [DOI] [PubMed] [Google Scholar]
- 17.Vejlstrup N, Sorensen K, Mattsson E, Thilen U, Kvidal P, Johansson B, Iversen K, Sondergaard L, Dellborg M and Eriksson P. Long-Term Outcome of Mustard/Senning Correction for Transposition of the Great Arteries in Sweden and Denmark. Circulation. 2015;132:633–8. [DOI] [PubMed] [Google Scholar]
- 18.Junge C, Westhoff-Bleck M, Schoof S, Danne F, Buchhorn R, Seabrook JA, Geyer S, Ziemer G, Wessel A and Norozi K. Comparison of late results of arterial switch versus atrial switch (mustard procedure) operation for transposition of the great arteries. The American journal of cardiology. 2013;111:1505–9. [DOI] [PubMed] [Google Scholar]
- 19.Barron DJ, Mehsood DK, Kutty RS, Stickley J, Botha P, Khan NE, Jones TJ and Brawn WJ. Fate of the Left Ventricular Outflow Tract After Rastelli With Selective Infundibular Muscle Resection. The Annals of thoracic surgery. 2019;107:1226–1231. [DOI] [PubMed] [Google Scholar]
- 20.Lee JR, Lim HG, Kim YJ, Rho JR, Bae EJ, Noh CI, Yun YS and Ahn C. Repair of transposition of the great arteries, ventricular septal defect and left ventricular outflow tract obstruction. European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery. 2004;25:735–41. [DOI] [PubMed] [Google Scholar]
- 21.Pettersen E, Fredriksen PM, Urheim S, Thaulow E, Smith HJ, Smevik B, Smiseth O and Andersen K. Ventricular function in patients with transposition of the great arteries operated with arterial switch. The American journal of cardiology. 2009;104:583–9. [DOI] [PubMed] [Google Scholar]
- 22.Schuwerk R, Freitag-Wolf S, Krupickova S, Gabbert DD, Uebing A, Langguth P and Voges I. Ventricular and atrial function and deformation is largely preserved after arterial switch operation. Heart. 2021;107:1644–1650. [DOI] [PubMed] [Google Scholar]
- 23.Grotenhuis HB, Kroft LJ, van Elderen SG, Westenberg JJ, Doornbos J, Hazekamp MG, Vliegen HW, Ottenkamp J and de Roos A. Right ventricular hypertrophy and diastolic dysfunction in arterial switch patients without pulmonary artery stenosis. Heart. 2007;93:1604–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tjahjadi C, Fortuni F, Stassen J, Debonnaire P, Lustosa RP, Marsan NA, Delgado V and Bax JJ. Prognostic Implications of Right Ventricular Systolic Dysfunction in Cardiac Amyloidosis. The American journal of cardiology. 2022;173:120–127. [DOI] [PubMed] [Google Scholar]