The most common indication for referral of patients with severe AS to CMR after echocardiography is to define aortic morphology and to exclude aortic aneurysms in patients with bicuspid aortic valves. This is generally done by CMR angiography, but can be done with non-contrast cine CMR if renal dysfunction is significant. A related indication is the exclusion of additional pathology such as coarctation or aneurysm of the aorta or other vascular pathology, which is easily accomplished using CMR as it provides detailed imaging of all intra-thoracic contents. Body habitus and the presence of intercurrent pathology such as chronic obstructive pulmonary disease do not degrade imaging as long as patients can cooperate with breath-holding instructions. A further requirement for cine MRI, is that patients have a stable cardiac rhythm with no frequent ectopy or rapid atrial fibrillation as this degrades image quality; however, this is not essential for aortic evaluation by CMR angiography.

The next most common indication is the broad category of discrepant diagnostic information from echocardiography, physical examination and/or cardiac catheterization. In particular, direct planimetry of aortic valve area (AVA), and complete morphologic visualization of aortic valve geometry in multiple oblique planes can resolve many clinical issues. In this context, CMR has become the gold standard for quantitative evaluation of chamber morphology and myocardial function, allowing very accurate assessment of stroke volume as long as mitral regurgitation is not a significant coincident problem [5–10]. Even in such cases, velocity encoded CMR (VENC) imaging, which provides analogous information to Doppler echocardiography and cardiac catheterization [11], can provide estimates of aortic or mitral regurgitant volume when needed to help determine the net forward cardiac output. Evaluation of the relatively laminar flow in the aorta and left ventricular outflow tract is very reliable in quantifying stroke volume and regurgitant volume [7–9]. However, determination of valve gradients is generally extremely technically demanding in comparison to echocardiography, for two reasons. First, very small changes in the direction of evaluation are easily accomplished by an experienced echocardiographer with a hand held probe, while each change in VENC imaging requires a separate breath-hold imaging sequence which can be very time consuming, on the order of 1–2 min of additional time per sequence acquisition [2, 3]. In addition, turbulent flow induces magnetic currents that change the velocity encoded flow rate. Thus, CMR gradient measurements are generally not robust enough to be decisive. Additional valuable information about the direction and pattern of high velocity outflow into the aorta downstream from the valve can be defined from cine CMR and velocity encoded imaging. Recent publications have confirmed the influence of angulated jets in affecting the degree of pressure recovery [12–14]. Based on in-vitro experiments and a limited number of patient studies, recent publications explain the discrepancy between relatively large geometric valve area and high gradients frequently seen in bicuspid aortic valves (reverse area gradient mismatch), and have provided quantitative formulas to calculate pressure recovery from aortic size and flow angulation in the proximal aorta [12–16].

Further relevant information from chamber morphology and function is often useful in patients with area/gradient mismatch as low-flow low-gradient physiology with normal or reduced ejection fraction (EF). The volume of the left ventricle and stroke volume is vital in assessing the hemodynamics of these patients and recent studies affirm the valuable accuracy CMR can provide in such cases. Further evidence that may be helpful in decision-making includes the degree of hypertrophy, fibrosis and elevated left ventricle mass that may confirm evidence of elevated left ventricular hemodynamic load, and chronic severe resistance to left ventricular outflow which portends a worse prognosis longer term for such patients [17–19]. Specific techniques such as delayed gadolinium contrast enhancement may reveal prior infarction or focal myocardial fibrosis that may mediate previously confusing symptomatology. In some cases the new technique of T1 or T2 quantitative mapping provide evidence of diffuse fibrosis that is not clearly revealed by traditional delayed enhancement. At this point a limited number of specialized centers have access to such technology but this is rapidly becoming a routine method. In LF/LG and NF/LG patients, MRI demonstrates larger AVA, less LVH, and similar focal fibrosis to LF/HG AS. This challenges the notion of the more advanced disease of the LF/LG AS patients.

Secondary causes of elevated left ventricular pressure such as intra-ventricular gradients from hypertrophic cardiomyopathy or subvalvular membranes that were missed on echocardiography may be revealed by CMR. In certain cases associated congenital anomalies that are occult or difficult to diagnose without access to thoracic angiography also may be revealed. MRI has become an integral part of evaluation of patients with HOCM for diagnostic, prognostic, and therapeutic considerations. It also can provide quantitative methods for assessment of concomitant aortic valve regurgitation as regurgitant volumes and ratios.

In cases where the severity of aortic stenosis is not in question and patients are symptomatic, echocardiography may be the only diagnostic tool needed to proceed to the operating room directly [27].

However, other important questions must be asked prior to surgery. Is coronary artery disease present and if so, what is the severity? Is there a concomitant aortopathy in need of repair at the time of aortic valve replacement? In the past, such questions would inevitably lead to a cardiac catheterization for coronary evaluation prior to surgery, and then lead to a static CT angiogram (CTA) of the chest or possibly a chest magnetic resonance angiogram to evaluate aortic size. Today, surgeons are more likely to accept a CTA for coronary evaluation (Fig. 7.1), and at the same time use that data to gather information about the ascending and descending thoracic aorta, as well as view the entire chest for potential ‘pitfalls’ when proceeding with possible sternotomy. It is now possible to get such accurate imaging of the chest to know if prior chest surgery would even allow for a repeat sternotomy, or when a minimally invasive mini-thoracotomy would be the preferred option. In cases of previous bypass, a CTA can clearly show the left internal mammary artery graft adhesion to the chest wall, or saphenous vein grafts which have become adherent to the chest or sternum, typically the graft to the right coronary artery (RCA) as it runs anterior to the heart and sits behind the sternum (Fig. 7.1). In this case, an MRI/MRA of the heart and chest would be less helpful as coronary and graft anatomy would not be well visualized. However, for patients who are unable to tolerate iodinated contrast, MRA can provide excellent imaging of aortic dimensions and pathology. With the chest and coronary anatomy in hand, the patient can be taken to the OR with a detailed surgical plan in place.

However, other valuable information can be gleaned from the CT or CMR. Information such as left ventricular outflow tract and aortic annular sizing as is done now for trans-catheter aortic valve replacement (TAVR) cases (see Chap. 9), as well as accurate valve planimetry for CT [21–24] and CMR [28–34] when compared to both TTE, and CMR. Lastly, CMR and cine CT will allow for a very accurate and reproducible left ventricular (LV) volume and EF [6–9, 26, 35]. Such information may be available from echocardiography, however in cases where echocardiograms may prove difficult to interpret due to patient body habitus or valvular calcification, CMR or cine CT can be helpful (Fig. 7.2). An overview of these and more applications and limitations of both CMR and CT can be found in Tables 7.1a, 7.1b, 7.2a, and 7.2b.

A 71-year-old male with H/O CAD and prior myocardial infarctions presents with shortness of breath and fatigue. On auscultation, a single second heart sound, and a loud, harsh, III/VI mid to late peaking systolic ejection murmur which is mid to late peaking at the right upper sternal border.

His echocardiogram reveals a calcified AV with a peak velocity of 3.1 m/s, a mean gradient of 32 mmHg, an AVA of 0.6 cm², a calculated stroke volume index of 37 ml/m², and an EF of 38 %.

A dobutamine stress echo shows that the patient’s EF rises to 50 %, the stroke volume rises by 25 %–51 ml/m², the mean gradient increases to 44 mmHg, and the AVA remains at 0.6 cm². The patient has a projected AVA of 1.0 cm². All signs in this case point to a low-flow, low gradient severe aortic stenosis with contractile reserve, which would portend a favorable prognosis [36–40] (see Chap. 5). The patient undergoes a cardiac MRI for additional risk stratification prior to aortic valve surgery. This confirms his low ejection fraction of 34.75 %, however notable was a previous infarct in the lateral wall seen on delayed enhancement images (Fig. 7.3). His AVA planimetry reveals confirming the presence of severe AS as well as a low SVI confirming the low flow state. Though patients with poor contractile reserve have a higher mortality than those with reserve, their comparative mortality is much lower with surgery than with medical management [40–43]. Independent variables which predict poor surgical outcomes also include patients with previous myocardial infarction, any concomitant CAD (>50 % lesion stenosis) and those with immediate need for circulatory support following surgery (i.e. IABP or inotropes) [44].

In this setting, CMR is helpful to define the etiology of myocardial dysfunction. In addition to its ability to give accurate and reproducible ejection fraction and LV volumes pre and post operatively in this low flow group [45] the use of delayed hyper-enhancement imaging can describe in detail if a patient has suffered a previous myocardial infarction, another independent preoperative risk factor [17–19] or additional etiologies such as inter-current amyloid heart disease. Other prospective studies in patients with LFLG aortic stenosis have shown that cardiac fibrosis in general is an independent risk factor for decreased functional class post aortic valve replacement [45]. Thus, an assessment of myocardial fibrosis with CMR pre-operatively among patients with depressed ejection fraction and confirmed severe aortic stenosis may prove beneficial to understand the level to which a patient’s symptoms may improve.

Another means by which CMR can be helpful in this case would be calculation of valvulo-arterial impedance (Z_va), which is calculated as follows:
SAP=systolic arterial pressure in mmHg, MPG= mean pressure gradient in mmHg, and SVI is the stroke volume index in ml/m² .

Z_va has also been validated among symptomatic patients with paradoxical LFLG AS to determine prognosis, with values >4.5 mmHg/ml/m² portending a much poorer prognosis and higher mortality [46–48]. Though TTE is fully capable of calculating this value, CMR has robust capabilities in measuring left ventricular volumes and stroke volume, and via flow mapping, calculation of the MPG can be made [2], or alternatively can be taken from echocardiographic findings. This positions CMR as an alternate method to accurately calculate the Z_va. Taking the SAP from the patient’s physical exam, and MPG obtained from the time velocity integral of the trans-aortic flow by echocardiography, the patient’s Z_va can be calculated readily. In this patient, we calculate . This would place the patient in a severe range of >4.5 mmHg/ml/m². This would portend a poor prognosis for this patient, and valve surgery would be recommended [47–49].

Several times practitioners will face patients who clearly suffer from aortic stenosis with conflicting evidence about disease severity. The difficulties generally arise when there are conflicting test findings or when the patient’s symptoms may be due to intercurrent non-cardiac disease as demonstrated below.