Simultaneous Right and Left Heart Real-Time, Free-Breathing CMR Flow Quantification Identifies Constrictive Physiology

Study Questions:

What is the diagnostic accuracy of free-breathing real-time phase contrast cardiac magnetic resonance imaging (RT-CMR) for measuring respiratory-dependent changes in mitral valve (MV) and tricuspid valve (TV) inflow velocities as a marker of constrictive pericarditis (CP)?


CMR with multiple analyses for morphology, including pericardial thickness, assessment of respiratory-dependent septal shift, and early diastolic septal motion abnormalities as well as respiratory-dependent phasic variation in MV and TV flow velocities, were undertaken in 16 patients with known or suspected CP and 10 normal controls. Patients were excluded if they had mechanical valves, greater than moderate valvular disease, right ventricular infarction, restrictive cardiomyopathy, significant pulmonary disease, or atrial fibrillation. All patients underwent combinations of standard detailed transthoracic echocardiography, cardiac CT, and hemodynamic catheterization to establish the diagnosis of CP. CMR inflow velocities were obtained in a horizontal plane with regions of interest placed in the MV and TV inflow region, and flow simultaneously assessed. Respiratory variation was timed by assessment of diaphragmatic motion during 10 seconds of relaxed free-breathing. The percent respiratory variation for each valve was calculated as maximum (expiratory E – inspiratory E)/inspiratory E for both the MV and TV. For both MV and TV, maximum respiratory-dependent velocity change >25% was assessed for diagnostic accuracy.


CP was present by standard criteria in 10 of 16 patients, all of whom had classic symptoms of right heart failure. Etiologies included prior pericarditis in one, previous cardiac surgery in two, mediastinal radiation in one, and idiopathic in six. Surgical confirmation of CP was available in seven of ten patients, and CP was confirmed based on noninvasive imaging and cardiac catheterization in two patients. Respiratory variation of inflow averaged 46 ± 20% for the MV and 60 ± 15% for the TV in patients with CP, compared to 16 ± 8% and 24 ± 11% in patients without CP, and 17 ± 5% and 31 ± 13% in controls (comparisons to CP p < 0.004). There was no difference in MV and TV velocity variation in patients without CP and controls. Respiratory variation exceeded 25% across the MV in all patients with CP and in no patient without CP or controls, resulting in sensitivity and specificity of 100%. All CP patients had TV respiratory variation exceeding 25%, whereas 17% of non-CP patients and 50% of controls had similar variation. A TV respiratory variation of 45% identified nine of ten patients with CP and was present in only one of six non-CP patients and one of ten controls, resulting in a sensitivity of 90% and specificity of 88%. All patients with CP had pericardial thickening of >4.0 mm (6.2 ± 1.0 mm), and cine CMR demonstrated diastolic septal bounce in 50% of CP patients and respiratory-dependent septal shift in 80%.


Assessment of respiratory variation in MV and TV inflow can be assessed with RT-CMR during a free-breathing technique, and identifies diagnostic respiratory variation in patients with CP. These data complement the morphologic data also available by CMR.


Patients with CP often represent very difficult diagnostic and management problems. Currently, multiple modalities including transthoracic echocardiography, cardiac CT, and MR, as well as detailed hemodynamic catheterization studies, are required to establish the diagnosis. CMR and CT have previously both been utilized to identify pericardial thickness, and using cine techniques, the abnormalities of septal motion in early diastole as well as respiratory interdependence of left and right ventricular filling can also be identified. Doppler echocardiography has previously been used extensively to study the hemodynamics of MV and TV inflow, with previous studies demonstrating thresholds of respiratory variation highly diagnostic of CP. This study nicely used 10 seconds of free-breathing to assess the mitral and tricuspid velocities and create similar curves of respiratory-dependent flow velocity change to that previously described with Doppler echocardiography. In a single CMR session, both morphology and the physiology of ventricular inflows could be assessed, resulting in a high degree of diagnostic accuracy. It should be noted, however,, that this was a highly selected patient population which excluded significant valvular heart disease, concurrent pulmonary hypertension, and patients with atrial fibrillation, all of which may coexist with and confound the diagnosis of pericardial constriction.

Clinical Topics: Cardiac Surgery, Heart Failure and Cardiomyopathies, Invasive Cardiovascular Angiography and Intervention, Noninvasive Imaging, Pericardial Disease, Pulmonary Hypertension and Venous Thromboembolism, Valvular Heart Disease, Cardiac Surgery and Heart Failure, Cardiac Surgery and VHD, Acute Heart Failure, Pulmonary Hypertension, Interventions and Imaging, Interventions and Structural Heart Disease, Echocardiography/Ultrasound, Magnetic Resonance Imaging

Keywords: Constriction, Tricuspid Valve, Infarction, Pericarditis, Constrictive, Cardiac Catheterization, Magnetic Resonance Imaging, Hemodynamics, Complement System Proteins, Echocardiography, Doppler, Cardiomyopathy, Restrictive, Hypertension, Pulmonary, Heart Failure, Catheterization, Heart Valve Diseases, Cardiac Surgical Procedures, Respiration, Diastole, Mitral Valve, Lung Diseases, Echocardiography

< Back to Listings