Multi‐modality imaging of the systemic right ventricle in congenital heart disease

A comprehensive and structured imaging approach in the evaluation of the systemic right ventricle (sRV) in patients with complete transposition of the great arteries (TGA) after atrial switch procedure and congenitally corrected transposition of the great arteries (ccTGA) is a key for their optimal lifelong surveillance. Despite the improvements in cardiovascular imaging of adults with congenital heart disease (ACHD), the imaging of sRV remains an ongoing challenge due to its complex morphology and the difficulty in applying the existing knowledge for the systemic left ventricle. While cardiac magnetic resonance (CMR) is considered the gold standard imaging method, echocardiographic evaluation is primarily preferred in everyday clinical setting. Although qualitative assessment of its systolic function is primarily used, the introduction of advanced echocardiographic techniques, such as speckle tracking echocardiography (STE) and three‐dimensional echocardiography (3DE), has provided new insights into the optimal assessment of the sRV. Standardized quantitative parameters remain to be elucidated, and morphometric and mechanistic studies are warranted to validate reference ranges for the sRV. This review highlights the challenges in the optimal evaluation of sRV and summarizes the available imaging tools.

• Advanced echocardiographic techniques (STE and 3DE) provide optimal sRV assessment.
• Reference ranges for the sRV indices are warranted to be validated.

INTRODUCTION
In the setting of a biventricular circulation, a systemic right ventricle (sRV) is encountered in patients with complete transposition of the great arteries (TGA) operated with atrial switch by a Mustard or Senning procedure and in congenitally corrected TGA (ccTGA) patients. 1der these anatomical and haemodynamic conditions, chronic pressure overload on the right ventricle (RV), which normally supports the low-pressure pulmonary circulation, 2 leads to progressive deterioration of the sRV function and various long-term complications, including heart failure, arrhythmias, and sudden cardiac death. 1 Subsequently, the routine surveillance of patients with sRV is of paramount importance for predicting early ventricular dysfunction and guiding heart failure management.
Due to the complex geometry and the retrosternal position of the sRV, its imaging remains a challenge. 2Extrapolating measurements and normal value ranges from the systemic left ventricle (LV) to the anatomically and functionally sRV is inappropriate due to the different geometric conditions and morphology.Conventional two-dimensional, Doppler and three-dimensional echocardiography (3DE), cardiac magnetic resonance (CMR), and cardiovascular computed tomography (CCT) have been used to evaluate sRV structure and function noninvasively.Although CMR constitutes the most accurate imaging tool in assessing sRV dimensions and function, transthoracic echocardiography is the first-line diagnostic modality in clinical practice. 1To date, scarce data exist on imaging parameters that should be used for the evaluation of sRV, as well as their diagnostic and prognostic values.
The present review aims to summarize the current literature on the imaging evaluation of the sRV in patients with TGA postatrial switch repair and patients with ccTGA.

Conventional variables
The two-dimensional assessment of the sRV ejection fraction (sRVEF) seems problematic, as the geometric formulae of Simpson's method used for LV volume cannot be applied in sRV volume estimation due to the complexity of the sRV geometry. 3Notably, the sRV contraction patterns differ between patients with TGA and those with ccTGA.In TGA, the anteroposterior component predominates in the total sRVEF, implying that the longitudinal and radial shortening of the sRV is diminished.In contrast, in patients with ccTGA the three components are equally responsible for the pump function of the sRV. 4 Wall motion abnormalities are present in patients with sRV 5 and further complicate the functional echocardiographic evalua-tion.In TGA patients, such abnormalities have been described even before atrial switch, meaning that myocardial injury already exists preoperatively. 6These wall motion abnormalities have been associated with systolic dysfunction 7 but not with diastolic myocardial velocities of the sRV. 5 Visual assessment (or the so-called "eye-balling" method) constitutes the most widely used echocardiographic method for evaluating the sRV systolic function in everyday routine practice.The sRV function is typically described by qualitative terms as normal, borderline, or impaired.In contrast to LV, the sRV has complex structural geometry and unique adaptation to each stimulus.
For these reasons, the quantitative sRV assessment is more complex, and physicians mainly rely on eyeball estimates.Nevertheless, this assessment method is based on each operator's individualized interpretation of echocardiographic measurements and therefore lacks interobserver reliability. 8In a recent study of patients with TGA after atrial switch and ccTGA, the interobserver agreement among six independent observers for the qualitative evaluation of the sRV systolic function was found poor (Kappa < 0.25).The intraobserver agreement, though, was reported good to excellent (Kendall tau b = 0.408-0.843). 8 Subsequently, quantitative parameters for the echocardiographic evaluation of sRV size and systolic function are fundamental, especially for echocardiographers working in non-high-volume adult congenital heart disease (ACHD) centers.

Tricuspid annular plane systolic excursion (TAPSE) and tissue
Doppler-derived right ventricular systolic excursion velocity (RV S') constitute measures of longitudinal contraction of the RV (Figure 1, Panels A and B, and Table 1).These echocardiographic parameters are highly reproducible and easy to acquire.A threshold of 14 mm for TAPSE has been related to sRV systolic dysfunction. 9,10[13] Moreover, previous studies which associated TAPSE with N-terminal probrain natriuretic peptide (NT-proBNP) have been contradictory. 4,14threshold of 10 cm/s for RV S' has been reported by Khattab et al. for adults with sRV. 9 In a study of 35 patients with TGA and atrial switch, sRV S' was moderately correlated with CMR-derived sRVEF.12 TAPSE and RV S' TDI assemble specific similar limitations.Firstly, both are load-dependent and can be affected by systemic atrioventricular valve (tricuspid) regurgitation (SAVVR).15 Furthermore, they cannot discriminate the active from the passive motion of the myocardium (tethering phenomenon), occasionally resulting in unreliable measures.16 Besides, both may be affected by pericardial adhesions due to a prior cardiac surgery. I a normally structured heart, RV contraction has a predominantly longitudinal component; however, in sRV, especially as it dilates, the radial component becomes important.The questionable prognostic value of these parameters for patients with sRV could be attributed to the inconsistency between the measured systolic No correlation with CMR-derived sRVEF 10 .

RV S' Reproducible and easy to acquire
Similar with TAPSE <10 cm/s 9 Moderately correlated with CMR-derived sRVEF 12 .

FAC High sensitivity and accuracy
Requires good endocardial definition which is often difficult due to the heavily trabeculated free wall of the sRV and the challenges to acquire a stable high-quality four chamber view.

dp/dt Independent of geometric assumptions
Questionable prognostic value in patients with SAVVR and no well-defined isovolumetric contraction phase.

IVA Relatively preload and afterload independent
High measurement variability 1 m/s 5,10 Not consistently correlated with CMR-derived sRVEF in previous studies 22 .

MPI or Tei Index Independent of geometric assumptions
Difficulty in the assessment of the required intervals in TDI images, Invalid with irregular rates, such as AF and elevated atrial pressure.

NA
No correlation with CMR-derived sRVEF.longitudinal excursion and the predominant contraction pattern of sRV, which tends to be circumferential. 11actional area change (FAC) is obtained by manually tracing the RV endocardium at both end-diastole, and end-systole in a four-chamber or RV-focused apical view, without including the trabeculation in the wall (Figure 1, Panel C and Table 1).FAC was shown to have high sensitivity for discerning between normal and impaired sRV function. 9It is strongly correlated with the CMR-estimated sRVEF and is considered among the most accurate indexes for assessing sRV. 17Its primary limitation is the requirement of good endocardial definition, which is often intricate, particularly in the heavily trabeculated free wall and the challenges to acquire a stable high quality four chamber view in the setting of a sRV. 18Nonetheless, FAC calculation includes the circumferential contraction pattern of the RV free wall, while it does not take into account the contribution of the RV outflow tract to overall systolic function. 16In patients with sRV, a cut-off FAC value of 29.5%-33% has been previously proposed. 4,9,10,13e rate of systolic RV pressure increase from the SAVVR continuous wave signal (dp/dt) should also be included in the comprehensive sRV assessment.The difference, in comparison to the nonsystemic tricuspid regurgitation (TR), is that the value should now be calculated from the slope of the line between 1 and 3 m/s (4 to 36 mmHg) of the TR spectral display (Figure 2, Panel A and Table 1).Data about its correlation with sRVEF is conflicting. 9,13In two cohorts of patients with TGA after the atrial switch procedure, dp/dt was not confoundedly associated with systolic ventricular function. 10,13In contrast, in a study of 37 adults with sRV, dp/dt was significantly correlated with CMRderived sRVEF. 9Additionally, Khattab et al. correlated the threshold of dp/dt < 1000 mm Hg/s with RVEF values < 50%. 9Dp/dt may be a useful index in the serial follow-up of patients with sRV dysfunction, but more and larger studies are warranted to establish its clinical value.
Myocardial acceleration during isovolumic contraction (IVA) is a sensitive index of TDI ventricular contractile function, irrespectively of preload or afterload changes. 19IVA, better than dp/dt, represents the rate of change of contractile force during isovolumic contraction (Figure 2, Panel B and Table 1) and is, therefore, a significant, relatively load-independent index for the sRV assessment. 19IVA has been found to be diminished in patients with sRV compared to the normal subpulmonary RV.In TGA patients following atrial switch, IVA correlates well with RV S' , regional strain and systolic strain rate of the basal segment. 20Furthermore, the accuracy of this parameter has been compared to valid invasive indexes; a parallel increase in IVA and endsystolic elastance has been remarked in 12 TGA patients with atrial switch during dobutamine infusion. 5ocardial performance index (MPI or Tei Index) is an index of global right ventricular performance, reflecting both RV systolic and diastolic function (Table 1).Despite the plethora of studies focused on assessing MPI in ACHD patients with systemic LV, it has not been studied in patients with sRV.MPI may theoretically emerge as a promising measure of sRV function since it is not restricted by geometric assumptions. 21However, not all evidence points in the same direction, as no correlation between MPI and CMR-derived sRVEF has been described in a recent study. 22The presumable inaccuracy of its measurements could be explained by the intricacy in the assessment of the required intervals in TDI images. 22Another disadvantage is that the method is not valid with irregular rates, such as atrial fibrillation and elevated atrial pressure.This could be a major limitation of the parameter, especially in patients with TGA after atrial switch procedure, in whom significant filling abnormalities through surgical baffles may exist. 23,24sessment of the SAVVR is also a crucial component of the echocardiographic evaluation of patients with sRV, as it has been shown to interplay with sRV dysfunction. 10In TGA patients, SAVVR is attributed to annular dilatation and sRV dysfunction; therefore, replacement of the SAVV is usually not required. 25Conversely, in ccTGA, there may be intrinsic abnormalities of the SAVV (e.g., Ebsteintype tricuspid valve) in the majority of patients.These abnormalities result in SAVVR, which may be further deteriorated due to the subsequent sRV dilatation and decreased contractility. 25Subsequently, prompt echocardiographic monitoring of the SAVVR is of great importance in order to consider timely surgical replacement of the SAVV prior to the emergence of significant sRV dysfunction. 25 adverse ventricular-ventricular interaction in patients with sRV has been reported, suggesting that systemic and subpulmonary myocardial functions are interrelated.Interestingly, the magnitude of this interaction was more prominent in patients with ccTGA and pulmonary stenosis. 26The pathophysiology behind that interaction might be related to the shared myocardial fibers and electromechanical asynchrony. 26hocardiographic evaluation of the diastolic function of the sRV remains a challenge.Although the diastolic function is usually impaired in patients with TGA after atrial switch, 27 its association with clinical outcomes has not been elucidated yet. 28Interestingly, the irregular filling of the ventricles in patients with TGA after atrial repair is provoked not only by the abnormal diastolic properties of the ventricles but also principally by the abnormal properties of the atrial baffles. 29

Novel variables
Speckle tracking echocardiography (STE) is superior to tissue Dopplerderived strain and strain rate, as the analysis is two-dimensional, angle-independent, 16 less load-dependent and without the limitation of signal-to-noise ratio.These features are of utmost importance in patients with sRV, given the high prevalence of the SAVVR. 30stemic right ventricular global longitudinal strain (sRVGLS), sRV free wall longitudinal strain (sRVFWLS), and regional strain and strain rate of the six ventricular wall segments can be included in the sRV evaluation by STE.Global right ventricular longitudinal strain (RVGLS) (Figure 3, Panel A and Table 1) and free wall RV longitudinal strain (RVFWS) (Figure 3, Panel B) have been advocated as sensitive tools to evaluate RV function and predict prognosis in patients with several conditions.There is still debate about whether strain analysis should look at the RV free wall alone or whether the strain in the interventricular septum should be included as a measure of global longitudinal RV strain.However, as the interventricular septum is an integral part of the LV as well, RVGLS might be influenced by LV dysfunction.Given the sRV contraction pattern, there is no such dilemma in patients with sRV where the estimation of global strain is preferred.sRVGLS is strongly correlated with sRVEF and is considered a reliable predictor of clinical outcomes. 13,26In fact, corroborating evidence on the predictive role of sRVGLS indicates that it is the most powerful index of sRV function 22 and the only index allied with the functional capacity and detection of early subclinical sRV myocardial damage. 31Additionally, it has been proved as a quantitative assessment parameter of contractile reserve during dobutamine stress echocardiography in TGA patients after Sen-ning atrial switch. 32An sRVGLS cutoff value of −14.2% to −16.3% has been associated with preserved sRV systolic function. 13,17The main limitations of the method include the low temporal resolution (frame rate) and the image quality.
Real-time 3DE has not been yet sufficiently examined in patients with sRV (Figure 4 and Table 1).As this modality is not dependent on sRV anatomy and geometry, quantification of ventricular parameters by 3DE should be preferred over two-dimensional quantifications. 3reover, 3DE takes into consideration not only the longitudinal and radial sRV contraction directions but also the anteroposterior direction and their contributions to total RVEF.This is of great importance, especially in TGA after atrial switch procedure, where the anteroposterior component has been reported to be dominant and significantly different from longitudinal and radial components. 4However, the difficulty in encompassing the entire RV in the pyramidal sector size raises concerns regarding the reliability of the RV volume evaluation.In the case of RV dilation, a situation commonly encountered in patients with sRV, 3DE could possibly lead to underestimation of sRV volumes.Therefore, it is more advisable to perform 3DE to assess volumes and EF, in case that the image quality is good. 3,10,14correlation of sRV volumes and sRVEF between 3DE and CMR has been previously described. 14Additionally, a recent study of 38 patients with sRV (24 with TGA after atrial switch and 14 with ccTGA) demonstrated a strong correlation between 3DE sRVEF and BNP plasma concentration.Intra-and interobserver variability and reliability for sRV volumes and sRVEF were also acceptable. 4 A three-dimensional sRVEF cutoff value of 45% has been proposed. 10o-dimensional echocardiography with three-dimensional knowledge-based reconstruction (2DE-KBR) is a relatively novel imaging tool that enables three-dimensional reconstruction of RV without requiring endocardial border delineation.Its feasibility in sRV volumes and function evaluation has been validated by Kutty et al. in   adolescents and young adults. 33However, this needs to be confirmed by larger studies.

CMR
CMR is considered the gold standard modality for evaluating sRV volumes and sRVEF. 34In contrast with echocardiography, CMR is not confined by acoustic windows, ventricular sizes or geometric assumptions.Nevertheless, in clinical practice, its usage may be limited as it is relatively expensive, not everywhere accessible, and cannot be performed in patients with claustrophobia or CMR nonconditional devices which may be present in situ in these patients (implantable cardioverter defibrillators/ cardiac pacemakers/ coils, etc.).In the absence of contraindications, CMR could be performed supplementary to echocardiography for the comprehensive serial evaluation of patients with TGA after atrial repair and patients with ccTGA 1 (Figure 5).However, there is a lack of evidence about the definition of normal sRV function via CMR imaging.More research is required to establish standard cutoff values for sRV volumes and sRVEF that could be pivotal for the routine imaging evaluation of these patients.An essential practical aspect when assessing sRV by CMR is whether the sRV trabeculations and papillary muscles should be included in the volume estimations or manually outlined and excluded.In patients with systemic LV, both analysis methods conclude with similar measurements.Yet, in patients with sRV, incorporating trabeculations and papillary muscles into CMR measurements leads to considerably increased systolic and diastolic volumes and lower EF.Subsequently, values acquired by the sRV delineation outside the trabeculations and papillary muscles are more accurate, and therefore this method is suggested in patients with sRV. 35Regarding the reference value for the indexed right ventricular end-diastolic volume (RVEDVI), a cut-point of 130 mL/m 2 was associated with prognosis in a population of 158 patients with sRV. 36pecially in patients with TGA and atrial repair, CMR constitutes an efficient imaging tool not only for their regular evaluation during the postsurgery follow-up period but also for their essential guidance to further therapeutic options.Delineation of the SAVV morphology and estimation of the flow rates through it, are of high importance. 37terestingly, in a retrospective cohort study of 33 patients with ccTGA, no association was reported between the degree of SAVVR and sRV function or sRV end-diastolic volume.Therefore, the aetiology of sRV dysfunction seems multifactorial and substantial clinical decisions should not rely solely on the degree of the SAVVR measured by CMR. 38R techniques also allow the detection of fibrotic areas in the myocardial tissue using late gadolinium enhancement (LGE).The presence of myocardial fibrosis in patients with sRV, although not sufficiently studied yet, may be attributed to ischemia caused by the increased oxygen demand of the hypertrophied sRV due to chronic systemic loading. 39In patients with TGA after Mustard procedure, the visualization of myocardial fibrosis by LGE has been associated with dyssynchrony, reduced long-axis function and SAVVR. 40Additionally, LGE has been related to poor exercise tolerance, accelerating clinical deterioration, 39 arrhythmias and sudden cardiac death. 41However, additional larger, more robust and, if possible multicenter studies are needed to investigate the predictive value of LGE in patients with sRV.

Cardiovascular computed tomography
Radiation exposure has limited the use of CCT in clinical practice; it is mainly reserved for patients with contraindications for CMR regarding the assessment of sRV (e.g., patients with pacemaker or implantable cardioverter-defibrillator leads).CCT is not subjected to geometric assumptions, has high spatial resolution and provides pretty good visualization of the great vessels, coronary arteries, and venous baffle anatomy. 1,37However, CCT measurements are less accurate than those acquired by CMR, and therefore it is not the preferred modality in determining sRV size and volumes. 1

CONCLUSIONS
The assessment of sRV in patients with TGA and ccTGA is complex and requires a comprehensive approach by ACHD experts.CMR is the gold standard, whereas echocardiographic parameters and especially novel techniques, including STE and 3DE, are valuable in everyday clinical practice.The high rates of variability among observers when assessing the sRV qualitatively underscore the need to establish diagnostic algorithms that involve the evaluation of quantitative imaging parameters.Efforts should be made to identify standardized thresholds for the echocardiographic and CMR indices in patients with sRV.

F I G U R E 1
Most commonly used conventional variables for the sRV echocardiographic evaluation of a patient with TGA after atrial switch.Longitudinal contraction of the sRV evaluated in the apical four-chamber window [(A) TAPSE by M-mode, (B) RV S' by tissue Doppler imaging].FAC (C) is obtained from the apical four-chamber view, by tracking the systemic right ventricular endocardial border beneath the trabeculations and is calculated as the difference in end-diastolic area and end-systolic area divided by the end-diastolic area.FAC is calculated in this patient 18.2 %, very low, which is in agreement with the other systolic parameters of this patient.FAC, fractional area change; RV S' , right ventricular systolic excursion velocity; sRV, systemic right ventricle; TAPSE, tricuspid annular plane systolic excursion; TGA, transposition of the great arteries.TA B L E 1 Overview of echocardiographic parameters for the evaluation of the systemic right ventricle and correlation with CMR. , Affected by SAVVR, Subject to tethering phenomenon, Affected by pericardial adhesions due to prior cardiac surgeries, Questionable prognostic value due to the predominant circumferential contraction pattern of the sRV <14 mm 9,10

F I G U R E 2
Additional conventional echocardiographic parameters calculated in the same patient.(A) sRV dp/dt is the instantaneous rate of sRV pressure rise during early systole.It is obtained by continuous-wave Doppler of the systemic atrioventricular valve (tricuspid) regurgitation jet and measuring the time interval necessary for the sRV to build up the predefined pressure difference (32 mmHg).(B) Measurement of IVA during isovolumic contraction at the level of the tricuspid annulus using tissue Doppler imaging (IVA = IVV/t, m/s2).dp/dt, rate of systolic RV pressure increase from the SAVVR continuous wave signal; IVA, isovolumic acceleration; IVV, peak isovolumic velocity; t, time from zero crossing to peak isovolumic velocity; sRV, systemic right ventricle.

F I G U R E 3
Speckle tracking two-dimensional longitudinal strain imaging in the same patient.Using the apical four-chamber RV-focused view, the sRV borders were manually traced.Speckle tracking was semi-automated.Global longitudinal strain analyses were performed using either six segments [(A) sRVGLS, three free wall segments and three septal segments] or only the free wall segments [(B) sRVFWS].The peak systolic strain values of the six segments were averaged to derive global sRV longitudinal strain (−12.8%,A). sRV, systemic right ventricle; sRVGLS, systemic right ventricular global longitudinal strain; sRVFWLS, systemic right ventricular free wall longitudinal strain; TGA, transposition of the great arteries; 2D, two-dimensional.F I G U R E 4 Three-dimensional full-volume acquisition to calculate systemic right ventricular EDV, ESV and EF in the same patient.The upper images (green boxes) are four-chamber views at end-diastole (ED) and end-systole (ES).The middle (white boxes) and lower images (purple boxes) are short-axis views at different levels.The main issue is encompassing the entire enlarged sRV in the pyramidal sector size.Right ventricular ejection fraction was calculated with three-dimensional 18.8%, which is significantly reduced.EDV, end-diastolic volume; EF, ejection fraction; ESV, end-systolic volume; sRV, systemic right ventricle.F I G U R E 5 CMR assessment of a 56-year-old male patient post Mustard operation for TGA.Panel A: Horizontal long axis view in end-diastole, end-systole, and late post gadolinium injection.Panel B: Short axis view at mid-ventricular level in end-diastole, end-systole and late post gadolinium injection.The sRV is moderately dilated with no evident fibrosis on late gadolinium images apart from the RV/LV insertion areas (arrows).CMR, cardiac magnetic resonance; LGE, late gadolinium enhancement; LV, left ventricle; sRV, systemic right ventricle; TGA, transposition of the great arteries.