Assessment of anatomy and function

Magnetic resonance imaging has been shown to deliver improved reproducibility for both volumetric and functional parameters when compared with echocardiography, and is currently considered to be the gold standard for evaluating many forms of congenital cardiac disease in general, and those involving the right ventricle in particular, especially in the adult population (Table 1). Magnetic resonance imaging allows for measurements of right ventricular volumes, systolic function, and severity of pulmonary valvar regurgitation with good reproducibility.Reference Davlouros, Kilner and Hornung^1, Reference Tulevski, Dodge-Khatami, Groenink, van der Wall, Romkes and Mulder²

Table 1

Indications for cardiac magnetic resonance in congenital cardiac disease (Footnote *).

(*) from reference 8.

Using magnetic resonance imaging in 64 patients with idiopathic pulmonary arterial hypertension, van Wolferen et al. demonstrated a stroke volume ⩽25 ml/m², a right ventricular end diastolic volume ⩾84 ml/m², and a left ventricular end diastolic volume ⩽40 ml/m² to be strong independent predictors of mortality (Fig. 1).Reference van Wolferen, Marcus and Boonstra³ Interval changes in these parameters at one year of follow-up were strong predictors of mortality, while right ventricular mass and wall thickness were not. Furthermore, van Wolferen et al. also demonstrated, in a separate study (including idiopathic and congenital cardiac disease-associated pulmonary arterial hypertension), significant improvement in right ventricular performance and cardiac index after three months of sildenafil added to baseline bosentan therapy.Reference van Wolferen, Boonstra and Marcus⁴ This study, however, is small and data should be interpreted with caution.

Figure 1

Survival in patients with idiopathic pulmonary arterial hypertension as a function of cardiac magnetic resonance parameters. LVEDVI, left ventricular end-diastolic volume index; RV, right ventricle; RVEDVI, right ventricular end-diastolic volume index; SVI, systolic volume index. Reproduced with permission from van Wolferen SA et al., Reference Reference van Wolferen, Marcus and Boonstra3.

Gan et al. examined right ventricular diastolic function and the acute effects of sildenafil in 25 pulmonary hypertensive patients (mainly idiopathic pulmonary arterial hypertension) using magnetic resonance imaging, brain natriuretic peptide and invasive measurements.Reference Gan, Holverda and Marcus⁵ Interventricular relaxation time correlated positively with brain natriuretic peptide, (r = 0.70), and inversely with cardiac index (r = −0.70) and right ventricular ejection fraction (r = −0.69). Both interventricular relaxation time and right ventricular ejection fraction improved after sildenafil and this may be a result of afterload reduction or direct effects of sildenafil on the right ventricle.

Flattening of the interventricular septum as an indicator of right ventricular hypertension is routinely applied in the echocardiographic evaluation of congenital cardiac disease and pulmonary arterial hypertension. Roeleveld et al. reported on 37 patients (including idiopathic and congenital cardiac disease-related pulmonary arterial hypertension) and demonstrated a linear relationship (r = 0.68) between the shape of the septum determined by magnetic resonance imaging and catheterization derived systolic pulmonary arterial pressure.Reference Roeleveld, Marcus and Faes⁶ This correlation is not as good as that reported in the original series involving echocardiography (r = 0.86), and is likely due, in part, to the fact that septal curvature is more closely related to the ratio of right-to-left ventricular pressures than an absolute number.

Numerous other studies have attempted to find echocardiographic correlates with magnetic resonance measurements with limited success. Parameters such as right ventricular fractional area change, tricuspid annular velocity, transverse fractional shortening and others have all been proposed. Due to either poor correlation with magnetic resonance imaging, poor reproducibility or the inherent time involved for acquisition and data analysis, none have gained wide acceptance. Recent studies using three-dimensional echocardiography have demonstrated excellent correlation with magnetic resonance and radionuclide estimates of right ventricular function.Reference Nesser, Tkalec and Patel⁷ This and some of the newer echocardiographic indices such as tissue Doppler imaging and myocardial performance index (the so-called Tei index), may challenge some of the current advantages of cardiac magnetic resonance in the future. The current availability and use of three-dimensional echocardiography is much less than that of magnetic resonance however, and indices require further evaluation, particularly in adult patients, where transthoracic views of the right ventricle can be problematic.

In some forms of left heart disease, magnetic resonance imaging exceeds echocardiography in the quality and type of information it can provide.Reference Pennell, Sechtem and Higgins⁸ Magnetic resonance imaging can define the pulmonary venous anatomy with improved accuracy as compared to echocardiography or cardiac catheterization.Reference Masui, Seelos, Kersting-Sommerhoff and Higgins⁹ In Shone’s complex, delayed enhancement has been used to document endocardial fibroelastosisReference Tworetzky, del Nido, Powell, Marshall, Lock and Geva¹⁰ and may offer a surgical option to reverse pulmonary hypertension due to left atrial hypertension.

The application of delayed enhancement to examine the right ventricle has been far more limited than for the left ventricle. In patients with corrected tetralogy of Fallot, delayed enhancement due to presumed fibrosis has been demonstrated in the right ventricular outflow tract and in the area of the ventricular septal defect patch.Reference Babu-Narayan, Kilner and Li¹¹ Multiple studies in patients with systemic right ventricles found areas of abnormal myocardium as demonstrated by delayed enhancement.Reference Babu-Narayan, Goktekin and Moon¹² The presence of abnormal myocardial regions was associated with known markers of adverse outcome, such as right ventricular dysfunction, poor exercise tolerance, arrhythmia, and clinical deterioration over time. In adult patients with pulmonary hypertension, a small number of studies have suggested that the presence of delayed enhancement at the right ventricular insertion points is common, regardless of etiology, and is inversely related to right ventricular function and pulmonary hemodynamics, with systolic pulmonary arterial pressure correlating most closely (Fig. 2).Reference Blyth, Groenning and Martin^13, Reference Sanz, Dellegrottaglie and Kariisa¹⁴ In patients with pulmonary arterial hypertension associated with congenital cardiac disease, the relative contribution of pulmonary hypertension versus the cardiac defect itself to delayed enhancement maybe difficult to differentiate.

Figure 2

Delayed enhancement is a common finding in patients with pulmonary hypertension of all etiologies. Most commonly seen in the interventricular septum at the “hinge points,” enhancement has also been demonstrated at the site of surgical fibrosis, papillary muscles and right ventricular outflow tract.

Computed tomography scanning is a reasonable alternative to magnetic resonance imaging for the acquisition of anatomic and functional data when magnetic resonance is not feasible, unavailable or contraindicated.Reference Nicol, Gatzoulis, Padley and Rubens¹⁵ Current resolution of computed tomography presents a good alternative to pulmonary angiography in the evaluation of clot or pulmonary arterial stenoses in all but the smallest vessels. Furthermore, computed tomography acquisition times are very short (a few minutes) which is an advantage for patients with claustrophobia or those with other disabilities such as Down’s syndrome.

The Eisenmenger population

Because of the commonplace inability adequately to image adolescents and adults with congenital cardiac disease, magnetic resonance imaging is increasingly used in this population.Reference Broberg, Ujita and Prasad¹⁷ In addition to assisting in providing complete anatomic definition, cardiac magnetic resonance can replace echocardiography and catheterization partially (e.g. yearly) in the follow up of patients with the Eisenmenger’s syndrome. The known sequelae such as pulmonary arterial thrombosis and pulmonary arterial dissection have all been recently documented by magnetic resonance screening in the literature. However, it should be emphasized that neither magnetic resonance imaging nor computed tomography can replace confirmatory right heart catheterization for diagnosis of pulmonary arterial hypertension. Although these studies are excellent additional tools, right (and eventually left) heart catheterization remains the gold standard for confirmation of diagnosis.

Assessment of hemodynamic and mechanical properties

One of the most important hemodynamic parameters obtained during cardiac catheterization is the determination of cardiac output (cardiac index). As a strong predictor of mortality, cardiac index is often used in clinical decision making and therapeutic choices. Excellent correlation between catheterization and cardiac magnetic resonance velocity mapping methods have been demonstrated on many occasions in terms of both absolute and relative systemic and pulmonary flows in the setting of shunts.Reference Brenner, Caputo and Mostbeck^18, Reference Hundley, Li and Hillis¹⁹ A distinct advantage of magnetic resonance measurements over catheterization based techniques is the indifference of the former to pulmonary and tricuspid valvar regurgitation in measuring cardiac output.

One further advantage of magnetic resonance over cardiac catheterization is its ability to measure flow in essentially any reasonably sized structure whether arterial or venous. This is particularly important in cases of multiple sources of pulmonary blood flow or anatomic abnormalities which prevent accurate thermodilution or Fick determination of flow; in some of these instances, measurement of total pulmonary vein flow may be used to provide the most accurate measurement of total pulmonary flow. Additionally, the ability of magnetic resonance to measure relative right and left lung perfusion as part of a comprehensive evaluation may obviate the need for radionuclide studies. In contradistinction to measurement of cardiac output, magnetic resonance estimation of pulmonary arterial pressures has not been successful to date.

Biomechanical properties of the pulmonary arteries have also been examined as markers of the disease process. Specifically, pulmonary arterial stiffness (capacitance, compliance, distensibility) which is felt normally to comprise approximately 30% of the total workload of the right ventricle, has been investigated as a measure of right ventricular work in pulmonary hypertension. The ratio of stroke volume to pulse pressure (expressed in ml per mmHg) as an indirect measure of arterial compliance, and ventricular work, was originally investigated in the systemic circulation in patients with hypertension and was found to be predictive of cardiovascular death.Reference de Simone, Roman, Koren, Mensah, Ganau and Devereux²⁰ This concept was then evaluated in the setting of pulmonary hypertension by Muthurangu et al. who described a method of measuring pulmonary arterial capacitance utilizing magnetic resonance flow data and invasive pressure measurements.Reference Muthurangu, Atkinson and Sermesant²¹ An inverse relationship between capacitance and both resistance (r = 0.89) and mean pulmonary artery pressure (r = 0.72) was found.

Capacitance was subsequently found to be a predictor of mortality as well. Mahapatra et al. studied 104 patients with idiopathic pulmonary arterial hypertension and calculated the capacitance based on invasive measurements of pulse pressure and echocardiographic calculations of stroke volume.Reference Mahapatra, Nishimura, Sorajja, Cha and McGoon²² During the 4-year follow-up period, the capacitance index was the strongest univariate predictor of mortality (hazard ratio 17.0 per ml·mmHg⁻¹ decrease) when compared to conventional markers. Patients in the lowest capacitance index quartile had a mortality of 61% while those in the highest had no deaths. Other investigators observed good correlations between echocardiographic and catheterization derived measures of capacitance and that impedance was better at predicting patient outcomes than pulmonary vascular resistance alone.

Adapting these methods, Gan et al., in 70 patients consisting primarily of idiopathic pulmonary arterial hypertension or systemic sclerosis used magnetic resonance imaging to examine proximal pulmonary artery stiffness.Reference Gan, Lankhaar and Westerhof²³ The relative cross-sectional area change in pulmonary hypertensive patients was one-third that of normal controls and showed an inverse relationship with mean pulmonary arterial pressure. Patients with a relative cross-sectional area change ⩽16% had worse prognosis than those with values >16% (Fig. 3).

Figure 3

Kaplan Meier survival curves demonstrating improved survival with relative area change (RAC, i.e. pulmonary artery compliance measured by magnetic resonance) > 16%. Reproduced with permission from Gan et al., Reference Reference Gan, Lankhaar and Westerhof23.

Heart-lung transplant assignment

Right ventricular changes in response to chronic pressure overload regress after lung transplantation. Neither magnetic resonance imaging nor computed tomography have been used routinely to decide between lung or heart-lung transplant in patients with pulmonary arterial hypertension with or without associated congenital cardiac defects, and no studies have been published examining this approach. The choice of heart-lung transplant is usually based on the severity of tricuspid regurgitation (quantitative measurements are not felt to be clinically important), and in the case of unrepaired congenital cardiac disease, the complexity of the underlying cardiac anatomy and the extent of intracardiac repair required (personal communication, Dr. Bruce Reitz).