Evaluation of Diastolic Function by Cardiac Magnetic Resonance Imaging

Introduction

In the past few decades, cardiac MRI has become an integral tool for the diagnosis and management of cardiovascular disease. Cardiac MRI offers excellent spatial and temporal resolution that allows for anatomic and functional assessment in a noninvasive manner. It is widely viewed as the gold standard for measurement of cardiac chamber volume in terms of precision and reproducibility. Cardiac MRI provides the most comprehensive phenotypic assessment of cardiac disease by providing quantitative assessment of ventricular size/function, myocardial perfusion, myocardial fibrosis, tissue characterization, and flow quantification. While TTE remains the first-line diagnostic tool for cardiac imaging and diastolic assessment, imaging quality can be suboptimal if there are limited acoustic windows, which is common in patients with increased body habitus. Because of the ability for cardiac MRI to provide comprehensive quantitative assessment, it has emerged as an important tool for the evaluation of cardiac disease resulting in abnormal diastolic function.

Physics of MRI

Cardiac MRI relies on the inherent feature of hydrogen nuclei known as nuclear spin, in which the hydrogen proton produces a small positive electric charge that allows it to spin at a rapid rate and to ultimately produce a small but measurable magnetic field. Cardiac MRI capitalizes on the fact that hydrogen atoms are extremely abundant in the human body and produce these small magnetic fields randomly oriented throughout the human body. Therefore, when the protons are placed in an MRI scanner, they will accordingly align or repel against the MRI scanner’s strong magnetic field. This alignment can be disturbed by applying additional small radiofrequency pulses that stimulate the protons to enter into a more excited state. When the hydrogen protons subsequently relax and return to their original state, a radiofrequency signal will be emitted, which can be measured and used to produce a clinical image.

Basic Imaging Sequences

Various MRI images can be obtained by using different gradient coils and radiofrequency pulses in specified sequences. The characteristics of these imaging sequences result in different signal characteristics of the hydrogen protons. The imaging sequences can be used in combination to specifically evaluate various properties of the cardiac structures, including tissue characterization, flow quantification, contractility, and diastolic function. The most common imaging sequences and those that are helpful for evaluating diastolic function are outlined in the following sections.

Spin Echo Technique—Black Blood Imaging

Black blood imaging sequences provide morphologic images of the heart in still frames. In these sequences, flowing blood moves quickly and thereby will not be consistently exposed to the radiofrequency pulses. This results in a signal void that appears black on final imaging ( Fig. 15.3 ). Conversely, because the myocardium and great vessels that remain within the imaging plane are exposed to the sequence of radiofrequency pulses, they yield high contrast signals that allow for accurate morphologic assessment. Spin echo sequences can be processed to emphasize different relaxation properties of the protons, which are known as T1-weighted or T2-weighted images. Black blood imaging is not used specifically in the assessment of diastolic function. However, their excellent tissue contrast and anatomic detail allow for careful assessment of morphology, which provide essential context for assessing the underlying etiology of diastolic dysfunction.

Phase Velocity or Phase Contrast Imaging

The term phase in cardiac MRI refers to the manner in which protons rotate about a particular axis. As hydrogen nuclei move through a magnetic field, they will shift their phase in proportion to their velocity. This allows for the direct measurement of blood flow velocity by phase contrast cardiac MRI, whereby an imaging plane perpendicular to the path of interest is selected to ultimately create a velocity encoded cine image. This technique will then provide data for the measurement of transvalvular velocities and gradients, regurgitant volumes, and shunt flow calculation (Qp/Qs).

Transmitral flow (TMF) serves to directly quantify the filling gradient between the left atrium and the left ventricle. Numerous studies have shown sufficient correlation between cardiac MRI and TTE for measurement of flow velocities ( Fig. 15.4 ). As with TTE, there are limitations to the use of TMF such as in patients with tachycardia or irregular rhythms, where interpretation will be difficult due to fusion or absence of the conventional filling pattern waves. It may also be more difficult to interpret TMF filling patterns in the more elderly population, in which there is an expected, age-related reduction in E wave velocity that is thought to be most often related to systemic hypertension resulting in slow LV relaxation rate. This may be deceiving in that a normal filling pattern in patients over 60 years of age may indeed represent a pseudonormal filling pattern.

Pulmonary venous flow can be measured using phase contrast cardiac MRI ( Fig. 15.5 ) and can help differentiate between normal TMF and pseudonormalization. In such cases, a normal TMF pattern (where E > A) may actually be pseudonormalization. In patients with pseudonormal TMF, the duration of the pulmonary A wave will be longer than that of the mitral A wave. The main advantage of cardiac MRI is that pulmonary venous flow (PVF) can be measured in virtually all cardiac MRI cases as opposed to only 68% of TTE cases due to improved image quality with cardiac MRI.

Phase velocity cardiac MRI can be used to assess longitudinal myocardial velocities; these are comparable to those obtained by tissue Doppler imaging on TTE, although they are only validated in small sample sizes as of yet. Midwall longitudinal fractional shortening has been shown to correlate with LV diastolic function and can also be reliably and easily measured by cardiac MRI.

Myocardial Strain

Myocardial strain can be assessed by myocardial tagging, which is unique to cardiac MRI and essentially allows for visualization of myocardial deformation during contraction by use of magnetic labels with a rectangular or radial grid ( Fig. 15.6 ). Tagging of these grids can be obtained every 20 msec, allowing for high temporal resolution. Myocardial tagging may allow for earlier recognition of subclinical diastolic dysfunction, as demonstrated by Edvarsen et al. in a study of 218 asymptomatic patients with LV hypertrophy (LVH) who were found to have regional diastolic dysfunction despite no clinical cardiovascular disease or LV dysfunction. Several other methods for measuring myocardial strain have emerged recently, including strain-encoded imaging (SENC), phase velocity mapping, and deformation encoding with stimulated echoes (DENSE), and feature tracking with cine steady-state free precession (SSFP) imaging. Recently, left atrial (LA) function and strain have been recognized as important predictors of cardiovascular morbidity and mortality in patients with diastolic dysfunction. Myocardial tracking can also be applied to the left atrium as well for quantification of LA function. Early recognition may allow for aggressive risk factor modification and management to prevent progression of disease, although this has not been well established in clinical practice as of yet.

Magnetic Resonance Spectroscopy

Magnetic resonance spectroscopy utilizes P signals as opposed to H . This thereby allows for definition of regional adenosine triphosphate (ATP) and phosphocreatine (PCr) contents, which can be indirectly suggestive of energy status and viability of the myocardium. Diastolic function is an active, energy-dependent process and therefore changes in PCr and ATP levels can be quantified by P–cardiac MRI; altered high-energy phosphate metabolism has been associated with diastolic dysfunction. Currently the clinical utility of P–cardiac MRI is limited by the low intrinsic signal-to-noise ratio of P in humans, thereby making its acquisition technically demanding. Recent data using 7T cardiac MRI allowed for more precise quantification of the spectra when compared to conventional 3T, which may be useful in future practice.

Late Gadolinium Enhancement

Intravenous gadolinium-chelated contrast agents can be used to detect areas of fibrosis, often defined as late gadolinium enhancement (LGE), as the prolonged washout of the contrast correlates with a reduction in functional capillary density in the irreversibly injured myocardium. Myocardial fibrosis has a well-established role in the development and progression of both systolic and diastolic heart failure. LGE typically represents areas with replacement fibrosis or myocardial infarction, and the pattern of LGE can be useful in regard to identifying the underlying myocardial pathology ( Fig. 15.7 ). Recent studies have indeed shown a correlation between LGE suggestive of fibrosis and diastolic dysfunction in various cardiovascular disease states. The clinical significance of LGE in patients with diastolic dysfunction is yet to be determined, although hypertrophic cardiomyopathy (HCM) patients with LGE have worse outcomes than counterparts without LGE.

Quantitative T1 Mapping and Extracellular Volume Measurements

Deposition of myocardial collagen results in diffuse myocardial fibrosis or myocyte replacement/replacement fibrosis. Diffuse myocardial fibrosis can negatively impact diastolic filling and systolic function of the left ventricle. Recently there have been numerous studies demonstrating the utility and accuracy of quantifying myocardial fibrosis with T1 mapping and extracellular measurements in patients with heart failure with preserved ejection fraction (HFpEF), hypertensive heart disease, valvular heart disease, HCM, amyloidosis, and Fabry disease ( Fig. 15.8 ).

T1 times and extracellular volume measurements have been shown to correlate with diastolic dysfunction.