Article Text
Abstract
Right ventricular (RV) size and function are important predictors of cardiovascular morbidity and mortality in patients with various conditions. However, non-invasive assessment of the RV is a challenging task due to its complex anatomy and location in the chest. Although conventional echocardiography is widely used, its limitations in RV assessment are well recognised. New techniques such as three-dimensional and speckle tracking echocardiography have overcome the limitations of conventional echocardiography allowing a comprehensive, quantitative assessment of RV geometry and function without geometric assumptions. Cardiac magnetic resonance (CMR) and CT provide accurate assessment of RV geometry and function, too. In addition, tissue characterisation imaging for myocardial scar and fat using CMR and CT provides important information regarding the RV that has clinical applications for diagnosis and prognosis in a broad range of cardiac conditions. Limitations also exist for these two advanced modalities including availability and patient suitability for CMR and need for contrast and radiation exposure for CT. Hybrid imaging, which is able to integrate anatomical information (usually obtained by CT or CMR) with physiological and molecular data (usually obtained with positron emission tomography), can provide optimal in vivo evaluation of Rv functional impairment. This review summarises the clinically useful applications of advanced echocardiography techniques, CMR and CT for comprehensive assessment of RV size, function and mechanics.
- advanced cardiac imaging
- echocardiography
Statistics from Altmetric.com
Introduction
For a long time, the right ventricle (RV) has been considered as a dispensable cardiac chamber, with limited contribution to overall cardiac performance. The introduction of the Fontan procedure, that excludes the RV from circulation in patients with congenital heart diseases, and the lack of non-invasive accurate and reproducible parameters of RV function had further contributed to this misconception.
The advent of new echocardiography techniques, such as deformation imaging and three-dimensional echocardiography (3DE), and technical developments of cardiac magnetic resonance (CMR) and cardiac CT (CCT) have allowed the accurate assessment of RV morphology and function and changed this erroneous perception of the contribution of the RV to overall cardiac performance.1 It is now evident that the function of the RV is directly related to outcomes in many cardiac conditions including pulmonary arterial hypertension (PAH),2 3 ischaemic heart disease,4 5 heart failure with reduced6 7 and preserved8 ejection fraction (EF), heart failure after implantation of left ventricular assist devices (LVADs)9 and also in the general population of patients undergoing clinically indicated echocardiography.10–12
However, the non-invasive assessment of RV geometry and function remains challenging because of its complex anatomy and mechanics. The ideal imaging technique should allow comprehensive, accurate and reproducible assessment of the RV morphology, contraction mechanics and haemodynamic performance, and effectively coping with its complex 3D geometry, unique myocardial fibre architecture, unfavourable location within the chest, limited number of well-defined anatomical landmarks and complex mechanism of contraction. In addition, the ideal imaging technique should be widely available, safe and cheap, allowing the assessment of the RV in a timely manner and in different clinical settings, including acute conditions and intraoperative, perioperative care.
Currently, there is no ideal non-invasive imaging modality (online supplementary table 1) that may allow comprehensive, accurate and reproducible assessment of the RV geometry, function and contraction mechanics in different clinical settings, including acute conditions and intraoperative, perioperative care.13 Accordingly, the use of multimodality cardiovascular imaging is recommended by current guidelines.14
Supplemental material
The purpose of this review is to summarise available data about the use of state-of-the-art non-invasive imaging modalities for advanced imaging of RV anatomy and function.
RV anatomy
RV dilatation is a powerful predictor of morbidity and mortality in the general population,15 and in patients with acute pulmonary embolism,16 PAH17 and chronic obstructive pulmonary disease.18 Moreover, RV enlargement is the most frequent finding in probands of patients with arrhythmogenic RV cardiomyopathy.19 Therefore, quantitation of RV size is pivotal to effectively manage patients with cardiovascular diseases.
Echocardiography
Conventional linear and area measurements by two-dimensional echocardiography (2DE) are challenging because of the complex shape of the RV and the lack of specific right-sided anatomic landmarks to be used as reference points. The conventional apical four-chamber view (ie, focused on the left ventricle (LV)) results in too much variability on how the right side of the heart is projected, and consequently, RV linear dimensions and areas may vary widely in the same patient with relatively minor rotations in transducer position (figure 1). Accordingly, RV linear dimensions are best estimated from a RV-focused apical four-chamber view with an upper reference limit of 42 mm.20
However, conventional 2DE views encompass only a limited amount of the endocardial surface and RV myocardium, and parts of the RV (ie, the RV outflow tract which contributes to 25%–30% of the RV volume) are usually overlooked (figure 2).
Conversely, 3DE enabled the acquisition of pyramidal data sets which include all the three structural components of the RV (inflow, outflow and apex). RV data sets can be displayed in volume-rendering mode to visualise the functional anatomy of the RV, sliced in multiple tomographic views to analyse regional wall motion, distribution of hypertrophy or extension of masses, or postprocessed by dedicated software packages, to map the RV endocardial surface and measure the RV volumes without using geometrical assumptions about the shape of the RV (figure 3).21 3DE measurements of RV volumes are comparable with those obtained from CMR,22 23 and 3DE has been reported to be more accurate and reproducible than 2DE in assessing RV size.24 Sugeng et al 25 compared RV volumes obtained using 3DE, CMR and CCT, using a vendor-independent software package to eliminate analysis-related intermodality differences. They showed good agreement between both 3DE and CCT measurements and CMR data that were used as reference (correlation coefficients 0.89 to 0.79, respectively) with low interobserver and intraobserver variability (4%–13%). Finally, normative data for 3DE RV volumes and EF are available, providing age-specific, body size-specific and sex-specific reference values derived from large cohorts of healthy adults and children (table 1).26 27
Accordingly, 3DE is the recommended technique to assess RV size and function by echocardiography.28 An RV end-diastolic volume of 87 mL/m2 in adult males and 74 mL/m2 in adult females, and an RV end-systolic volume of 44 mL/m2 for adult males and 36 mL/m2 for adult females should be used as the upper threshold values to diagnose dilated RVs.28
Limitations of 3DE include dependence on patient’s acoustic window, frequent dropout of the RV anterior wall, difficulty to include the whole RV in the 3DE pyramidal data set in case of severe dilation, and when a multibeat acquisition is needed, patients should be able to cooperate with breatholding and have a regular heart rhythm. A large meta-analysis that included 807 patients reported a significant underestimation of larger RV volumes and overestimation of RV volumes in older patients.22
Nevertheless, 3DE yields significant benefits in comparison with the other imaging modalities in terms of portability, no use of ionising radiation and absence of contraindications, making it a versatile and cost-effective technique to assess the RV. However, not all patients can have their RV assessed by 3DE. Certain categories of patients (with highly irregular arrhythmias, marked obesity, dyspnoea, chronic obstructive pulmonary disease, very enlarged and deformed RVs, mechanically ventilated patients or with mechanical support devices, etc) are particularly challenging for transthoracic 3DE acquisition. Overall, the feasibility of transthoracic 3DE for assessing the RV is 85%–90%26 27 and may decrease to 56% in LVAD recipients.29 When transthoracic 2DE image quality is poor, contrast enhancement is only rarely used for improving the quality of 3DE data sets. These patients are generally addressed either to transoesophageal echocardiography or other imaging modalities (CMR or CCT).
Cardiac magnetic resonance
CMR allows an accurate evaluation of RV anatomy, function and tissue characterisation. Thanks to the possibility to arbitrarily set slice orientation, its excellent spatial resolution and 3D volume rendering, CMR can overcome the challenges related to the complex RV geometry and motion (figure 4). This technique has been proved accurate and reproducible, thus is currently regarded as the gold standard for non-invasive assessment of cardiac volumes, mass, flows and systolic function, including on repeated examinations. The main limiting factor is represented by RV-reduced wall thickness (3–5 mm), which makes it unsuitable for certain CMR features.
RV anatomical evaluation is generally performed using steady-state free precession ‘cine’ images, useful to assess the presence of morphological abnormalities such as aneurysms and outpouchings but also regional wall motion abnormalities and T1-weighted black-blood turbo spin-echo (TSE) sequences.
Standard cine imaging30 include: (1) a short-axis stack, covering RV from the atrioventricular valvular plane to the apex, which is also used to calculate mass, volumes and systolic function by a disc summation method with delineation of epicardial and endocardial borders; (2) a transaxial stack, covering from diaphragm to pulmonary bifurcation; (3) an RV inflow/outflow image, which allows the simultaneous visualisation of the inlet portion, trabecular apex and infundibular portion; (4) RV two-chambers and (5) sagittal RV outflow tract. Further pictures may explore pulmonary artery and its branches, pulmonary veins, systemic veins and aorta.
Various CMR-based approaches have been applied to the morphological and functional study of the RV, and their association with patients’ outcome has been reported.15 31 In a recent paper, Mauger et al 32 reported that atlas shape features (in particular RV size, tricuspid excursion, longitudinal shortening and sphericity index variation) had stronger associations than traditional mass and volume measures for all tested risk factors (dyslipidaemia, hypertension, obesity and smoking; p<0.005 for each) on a wide cohort. Further research is needed to understand the prognostic value of these new morphometric scores.
Furthermore, multiparametric properties of CMR allow for an accurate, non-invasive tissue characterisation, performed with or without contrast administration, which aids in the differential diagnosis or RV pathologies (figure 5). Precontrast T1-weighted TSE images, with and without fat saturation, are used to identify fat infiltration/fatty metaplasia, on the same planes used for cine images. Finally, late-enhancement images, acquired after the administration of a gadolinium-based contrast mean, can be used to identify pathological processes including oedema, focal fibrosis and fatty infiltration, although the interpretation may be challenging because of the reduced thickness of RV wall. Unfortunately, at the moment, other advanced tissue characterisation techniques such as, for example, T1 mapping and extracellular volume cannot be reliably applied to RV.
Normal values for RV volumes and EF measured with CMR are available1 2 33 (table 1).
Cardiac computed tomography
Although CCT is mainly used for the assessment of coronary artery disease, it can also be used to evaluate LV and RV function and morphology thanks to its isotropic submillimetre spatial resolution, acceptable high temporal resolution and good contrast resolution between ventricular blood pool and myocardium.34 Moreover, such as for 3DE or CMR it allows full coverage of the ventricular chamber without the use of any geometric assumptions.
Assessment of RV function by CCT has been validated in multiple studies using CMR as the reference standard showing that volume are slightly overestimated and as consequence LVEF and stroke volume are underestimated (figure 6). In a recent meta-analysis, Pickett et al found that CCT overestimating RVEF by 4.67% with a an excellent correlation (r=0.79) as compared with CMR.35 Several reasons can explain this minimal discrepancy. First, the limited temporal resolution could be responsible for a not optimal selection of right end-diastolic and end-systolic phases. Second, the motion artefacts that usually occur in case of high heart rate during the acquisition could impair the delineation of endocardial contours. Third, the administration of beta-blockade premedication could affect the RV performance.36 Finally, the acquisition in a shorter breath hold in inspiration phase could play a role as compared with CMR acquisition that is usually performed in a longer breath hold in expiration phase.
CCT is not a routinely used technique for RV assessment due to the significant radiation exposure and the use of iodinated contrast medium. Therefore, it is usually used for the concomitant RV function evaluation in clinical contexts where CCT is indicated (figure 7).
The appropriate use criteria for the quantitative assessment of RV function by CCT37 include patients with heart failure in which CCT is required to rule out coronary artery disease, pulmonary embolism to rule out perfusion defect in pulmonary artery, pulmonary disorder to evaluate lung disease or congenital heart disease to better visualise anatomical relationship. Finally, it is a valuable alternative to CMR in patients with pacemaker, CMR incompatible prosthetic material and claustrophobia (figure 5).
Normal values for RV volumes and EF measured with CCT are available (table 1).
RV function
RVEF is an independent predictor of cardiovascular morbidity and mortality in various diseases and also in the general population of patients undergoing clinically indicated echocardiography.3 4 38
RV systolic function
In the absence of a single reliable 2DE measure of the RV systolic function, a number of surrogate echo parameters (RV fractional area change (FAC), tricuspid annular plane systolic excursion (TAPSE), peak S wave velocity of the lateral tricuspid annulus by tissue Doppler imaging and RV myocardial performance index (RIMP)) have been proposed for clinical use.39 However, these 2DE measures do not capture the contraction of the entire RV. FAC neglects the contribution of the RV outflow tract, whereas TAPSE and S wave velocity also neglect the contribution of RV free-wall and septum. Conversely, 3DE is the only echocardiographic technique able to provide accurate and reproducible measurement of RVEF.28
3DE measurements of the RVEF have been reported to be accurate and reproducible in comparison with CMR both in adults and children.22 23 40 In a recent meta-analysis, 3DE was the most accurate among six evaluated imaging modalities, by overestimating the RVEF only 1.16% and with the lowest limits of agreement (from −0.59% to 2.92%) when compared with CMR.35 Moreover, 3DE RVEF has shown to be predictive of prognosis in a variety of cardiac conditions.3 10 11 The lower limits of normality for RVEF are lower than those used for LVEF and have been set at 45%.26 28 41 Recently, partition values to define mild (45%<RVEF≤40%), moderate (40%<RVEF≤30%) and severe (RVEF<30%) degrees of RV dysfunction have been defined and prognostically validated.38
However, RVEF is heavily dependent on the loading conditions of the RV. Accordingly, myocardial deformation or strain is increasingly being used to assess RV function because RV myocardial strain is less sensitive to loading conditions than EF, S’, TAPSE or RIMP.42 Strain is defined as the percentage change in dimension (lengthening, shortening or thickening) of the myocardium during the cardiac cycle. Currently, 2DE speckle tracking is the most frequently used echocardiography technique to measure RV strain. Since the predominant RV myocardial fibres are arranged in a longitudinal fashion from the tricuspid annulus to the apex, the most relevant strain component to assess the RV function is the longitudinal strain (LS).43 2DE RV LS is obtained from an RV-focused apical four-chamber view and two strain components can be measured (figure 6): RV free-wall LS, which is the average value of the LS measured in the three myocardial segments of the RV free-wall; and RV four-chamber LS, which is the average between the strain values of the RV free-wall and interventricular septum.43 RV free-wall LS is generally higher than RV four-chamber LS.44 Currently, reference values of RV free-wall strain were given as −29.0%±4.5% in the 2015 American Society of Echocardiography/European Association of Cardiovascular Imaging (ASE/EACVI) guidelines.28
An increasing body of literature demonstrates the utility of RV strain for outcome prediction in multiple acute or chronic cardiac conditions (online supplementary tables 2-4) and even in the general population of patients undergoing clinically indicated echocardiography.12
Supplemental material
Supplemental material
Supplemental material
RV diastolic function
RV diastolic function has been reported to be an important parameter in the assessment of cardiac dysfunction in patients with heart failure.45 Measurement of RV diastolic function is almost exclusively performed by echocardiography. Transtricuspid E/A ratio, E/e’ ratio, right atrial volume and hepatic vein flow have been the most validated parameters to assess RV diastolic function. Mild RV diastolic dysfunction is defined by a transtricuspid E/A ratio <0.8 (suggesting impaired relaxation) or transtricuspid E/A ratio between 1 and 2 with S/D >1 in hepatic vein flow and early component of the tricuspid annular tissue Doppler velocity (e’) less than the atrial component of the tricuspid annular tissue Doppler velocity (a’). Moderate or severe RV diastolic dysfunction can be assumed to be present either when the transtricuspid E/A ratio is between 0.8 and 2.1 with an E/e’ ratio >6 or a diastolic flow predominance in the hepatic veins suggests pseudonormal filling, or the transtricuspid E/A ratio >2.1 with a deceleration time <120 ms (suggestive of restrictive filling) and an inverted systolic wave form is present on the Doppler hepatic vein flow signal.
However, the assessment of the RV diastolic function is particularly challenging because of the thin wall of the RV that makes it very sensitive to loading conditions, in particular in presence of RV myocardial diseases such as ischaemia or infarction, and the peculiar haemodynamics of the RV. During inspiration, there is an increase of the preload of the RV that causes an increase in E and therefore an increase in the E/A ratio. The presence of tachycardia causes an increase in E but a relatively greater increase in A and therefore a decrease in the E/A ratio. A reduction in preload (eg, diuresis) causes a decrease in E wave velocity but a relatively smaller decrease in A wave velocity and therefore a decrease in E/A. Tissue Doppler is less load dependent than spectral Doppler of transtricuspid blood flow, since a reduction in preload causes an equal decrease in e’ and a’, but the e’/a’ ration will not change. Accordingly, tissue Doppler should be used to differentiate normal from pseudonormal filling patterns with elevated filling pressures.39 Of note, the presence of moderate to severe TR or atrial fibrillation could confound diastolic parameters, making the assessment of RV diastolic dysfunction feasible and reliable in a minority of patients.
Evaluation of RV function by CMR and CT
Calculation of RV volumes and EF is routinely achieved by CMR using the disk summation method which avoids geometrical assumptions about the shape of the RV. Age and gender-specific reference values are available for both the adult and the paediatric patients.46 47 This is crucial in ensuring an accurate quantification of RV, characterised by a complex geometry. Reproducibility is high, with a reported interobserver variabilities with intraclass correlation coefficient 0.92 for RV end-diastolic volume and 0.77 for end-systolic volume.47
Furthermore, interest is growing in the application of feature-tracking CMR, already established in the evaluation of LV to RV. This technology allows quantification of myocardial global and segmental deformation, including longitudinal (performed on long axis), radial and circumferential (performed on short axis) strain measurements and is less dependent on loading conditions than EF. It overcomes the limitation of traditional myocardial tagging, for which accuracy was reduced because of the thin RV wall and required time-consuming postprocessing. Nevertheless, further studies are needed to understand the clinical and prognostic relevance of this information.
CCT is also considered a reliable technique to measure global RV function. Pickett et al reported that CCT was the second most accurate imaging modality, after 3DE, to assess RVEF with only a 4.67% overestimation of CMR measurements, and narrow limits of agreement (from −3.71% to 5.62%).35
Role of hybrid imaging
Hybrid imaging, which is able to integrate anatomical information (usually obtained by CT or CMR) with physiological and molecular data (usually obtained with positron emission tomography), can provideoptimal in vivo evaluation of Rv functional impairment. PET/CT has become commonplace in clinical practice and research settings and has significantly contributed to our insights in pathophysiological processes affecting the RV. Hybrid PET/CMR is a relatively new, fast-upcoming hybrid imaging modality that not only allows simultaneous combination of PET with anatomic imaging but also provides functional imaging, perfusion, tissue characterisation and flow imaging, thereby improving the assessment of patients with RV failure. However, one limitation of hybrid systems with CMR is that it does not provide information required for attenuation correction of nuclear images.
Conclusion
RV dysfunction is increasingly being recognised as a marker of poor prognosis in a variety of cardiovascular diseases. Therefore, the accurate quantification of the degree of RV impairment is crucial to address management of patients. Although 2DE and Doppler echocardiography remains the mainstay for RV evaluation, 3DE and speckle tracking echocardiography provide a more comprehensive and accurate assessment of RV geometry and function. Other non-invasive cardiac imaging modalities provide additional useful parameters. CMR is particularly useful for accurate and reproducible quantification of RV volumes and EF, as well as myocardial tissue characterisation. Cardiac CT can reliably supplement echocardiography and CMR in patients with unfavourable acoustic windows and contraindications to CMR.
Most of the patients with suspected or confirmed RV dysfunction will need a multimodality approach for the assessment of their RV geometry and function. Combined results and collective evidence obtained from the different imaging modalities will provide deeper insight into the pathophysiology of this intriguing cardiac chamber, translating into more accurate diagnosis and improved clinical management of patients.
Footnotes
Twitter @lpbadano
Contributors GP is an internationally renown expert in CCT. He provided the paragraphs and images related to CCT. CT is an internationally renown expert in cardiac magnetic resonance. She provided the paragraphs and images related to CMR. DM, KA, and RML wrote most of the echo-related paragraphs, particularly those related to 3D and speckle tracking echocardiography. LPB and GP collected all contributions, merged them and drafted the final version of the paper. DM provided most of the echo images. All the authors have read and approved the final version of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, conduct, reporting or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Commissioned; externally peer reviewed.