In contrast-enhanced (CE) MRA, gadolinium (Gd) compounds are routinely used although, at times, other contrast agents like super paramagnetic iron oxide and chromium-labeled red blood cells can also be used. Prince et al. [20] first introduced first-pass CE-MRA in 1993. Gadolinium contrast agents, based on their pharmacological properties, result in significant T1 relaxation (shortened 10–25-fold) which results in significant signal production. Technological advancements have resulted in reliable, faster, and dynamic acquisition protocols that have significantly advanced the capabilities of MR angiography. Serial scan technique, time-resolved MRA, can be implemented that can provide a dynamic assessment of the vasculature as in DSA. This technique lessens the need for precise bolus timing and at the same time provides physiological information. In addition, time-resolved sequences allow for reduced volumes of Gd contrast for imaging. Gd compounds that are used at present are also less nephrotoxic and result in a reduced burden of hypersensitivity reactions than computed tomographic iodinated contrast media and thus are well tolerated by most patients [21–23].

Gadolinium administration has been associated with low incidence of nephrogenic systemic fibrosis (NSF). This condition following CE MRI has been observed exclusively in patients whose renal function is compromised. Grobner et al. [24] first identified this association in 2006 after which several studies confirmed this association and virtually with all types of Gd contrast agents [25–31]. NSF is a debilitating disorder resulting in fibrosis of the skin, connective tissue, and occasionally of internal organs associated with a high morbidity and mortality. As Gd compounds have been implicated in NSF, the disease entity has been thoroughly investigated and guidelines to help reduce the likelihood and impact of NSF have been proposed. Adherence to these guidelines has drastically reduced the incidence of NSF [32–34]. Current recommendations are to avoid injection of Gd contrast agents in patients with reduced glomerular filtration rate (GFR) <30 ml/min/1.73 m². If MRA is necessary, contrast agents with predominant hepatic clearance or non-enhanced MRA should be considered.

Non-Gd enhanced MRA is a viable alternative in patients whose renal functional parameters are compromised [35–37]. Blood flow can cause signal loss or gain on different types of MRA sequences. This property can be utilized to produce inherent endogenous contrast between stationary tissue and flow of blood in vessels to produce black-blood or bright-blood imaging. Black-blood imaging relies on flow-related signal loss, whereas bright-blood imaging relies on flow-related enhancement. Various MRA techniques and their major advantages and disadvantages are mentioned in Table 22.2.

The utility of MRA as applied to TAVR patients is manifold. The indications for use include the following: to delineate relevant vascular anatomy; to identify the presence and extent of atherosclerotic disease, stenosis, thromboembolic occlusion, or any extrinsic compression over the major vessels/vascular pathways; and to identify the presence, location, and extent of vascular abnormalities (e.g., aneurysms, dissections, arteriovenous malformations, vasculitis) (Figs. 22.1, 22.2, and 22.3). For all TAVR patients, care should be taken to avoid performing MRI/MRA when contraindicated (Box 22.1).

An example of a MRA protocol for TAVR and the steps involved in acquiring images are listed in Table 22.3
. Baseline cardiovascular functional parameters like LVEF, end-diastolic volume, end-systolic volume, and stroke volume are calculated by acquiring SSPE cine images from base to apex either in short axis (SA) or in horizontal long axis (HLA) (4 chamber) views. These views also easily identify the presence of any thrombus or mass within the left ventricle.

An optimal arterial MRA acquisition requires patient preparation, screening the patients for any contraindications to MRA, injection of adequate amount of contrast, and more importantly to time the study while the contrast is in transit in the arterial vessels. In doing so the protocol allows for isolation of the contrast within the arterial system. At our institution, during performance of the scan, we define the true orthogonal anatomical plane of the aortic valve from contrast-enhanced MRA images. Balanced SSFP (True FISP) cine images are acquired: few slices above, through and below the aortic valve plane. These views help in determining the number of valve leaflets, assessing the leaflets mobility, any commissural fusion or bulky mass, or vegetations on the aortic valve. This view also helps in calculating the area of stenosis. Using the same MRA source images, we also obtain a cross-sectional images of aorta and pulmonary artery (just above their respective valves). The images obtained are used to derive flow-related measurements through phase contrast (PC) angiography.

Herein listed are the steps to acquire MRA images that we employ at our institution. First, we acquire pre-contrast MRA images or mask images. This is obtained to know the proper anatomic coverage, to plan the field of view, to identify unwanted artifacts occurring over the region of interest (ROI), and to familiarize the pattern of breath hold technique to be followed and the expected length of breath hold. Second, we acquire MRA source images and time-resolved MRA. This is followed by administration of contrast agent @ 0.1 mmol/kg body weight and saline flush of 20–30 ml. Source images are obtained in identical fashion to mask image. Usually a series of three to four MRAs can be obtained with the same or on different imaging planes over a period of time, that is, time-resolved MRA, thus providing 4D spatial and temporal vascular imaging [38, 39]. Third, we perform subtracted MRA. Subtraction of pre- and post-contrast 3D MRA data can be done to provide images without any noisy background tissue and thus improving the resolution of MRA. This method is successful only if the two data sets have identical imaging parameters and spatial orientation. Misregistration of two data sets might lead to poor image quality and artifacts. Next, we perform image post-processing. For assessing the aortic root and the aortoiliofemoral vasculature, an MRA series which shows good luminal enhancement of the vessels is selected. At our institution, we routinely use a dedicated offline 3D workstation. Workstations with similar post-processing tools are available from different vendors. 3D volumetric source or subtracted images are evaluated in multiple planes and are post processed further to highlight the ROI. Importantly, high quality source MRA images are essential to ensure accurate diagnoses as all post-processing tools rely on adequate image quality. If doubt exists or if subtracted images are not of diagnostic quality, MRA interpreting physicians should rely on the source images rather than on subtracted images.

Various post-processing tools are used to analyze the 3D volumetric data [40–42]. Some of the most commonly used tools include multiplanar reformation (MPR), maximum intensity projection (MIP), and volume-rendered techniques (VRT).

MPR is a post-processing tool where the volumetric data is automatically arranged in three orthogonal planes, namely, axial (or transverse), coronal, and sagittal planes. MPR tool is unique as it allows image processing in any desired plane (Fig. 22.4). By toggling the imaging planes to the axis of a vessel, a profile and enface viewing plane can be obtained for assessment of vascular structures. This is especially helpful in evaluating structures that are oriented obliquely to the standard orthogonal planes. Slice thickness can be manipulated so as to visualize a large field of view albeit with slight decrease in spatial resolution. Advanced software algorithms provide curved MPR option. Curved MPR helps in laying out any curved vessel flat in a 2D plane. It can also be rotated 360° for easy assessment of vessel analysis. This is especially helpful in assessing a real versus pseudostenosis within a tortuous vessel.

MIP is the simplest and most widely used post-processing technique. In this algorithm, a set of parallel rays is projected on to the chosen viewing plane. The rays in their pathway take into consideration only the maximum signal intensity present in the voxel and the low intensity signals are discarded. The end result is a 3D volumetric data with brightest signal projected as a 2D image. This is useful for visualizing the entire vascular tree but with the expense of loss of depth projection (Fig. 22.5). To overcome this, MIP images are rotated in multiple imaging planes to gain an appreciation of the spatial relationship of the various structures being analyzed. The interpreting physician must be aware that any other entities with high signal intensity may also project within the plane of the vasculature and act to obscure it. This is of particular concern when images are significantly degraded by background noise. As a result, MIP imaging is more helpful when background noise has been adequately subtracted from the mask images.

VRT images allow for analyses of the voxels in the entire volume of data and provide depth information. It assigns a color based on each voxel’s signal intensity and opacity on a scale of 0–100 % (Fig. 22.6).