Abstract
Echocardiography is an essential tool in the evaluation and management of patients with prosthetic valves. Its use requires an understanding of valve design, normal prosthetic appearance and function, imaging artifacts introduced by valve elements, and the spectrum of valve dysfunction. This chapter will cover these topics with a focus on aortic and mitral prostheses, since these are the most commonly implanted valves, and those for which the largest evidence base exists.
Keywords
echocardiography, patient prosthesis mismatch, pressure recovery, prosthetic valve, 3D echocardiography
Introduction
Echocardiography is an essential tool in the evaluation and management of patients with prosthetic valves. Its use requires an understanding of valve design, normal prosthetic appearance and function, imaging artifacts introduced by valve elements, and the spectrum of valve dysfunction. This chapter will cover these topics with a focus on aortic and mitral prostheses since these are the most commonly implanted valves and those for which the largest evidence base exists. A valuable reference is the current joint document of the American Society of Echocardiography (ASE), European Association of Cardiovascular Imaging (EACVI; formerly European Association of Echocardiography), and other professional societies. The reader is directed to Chapter 40 (Echocardiography in Infective Endocarditis) for a discussion of prosthetic valve endocarditis and to sections in the chapters on native valves ( Chapter 28 , Chapter 29 , Chapter 30 ), which deal with the echocardiographic quantitation of valve stenosis and regurgitation, since similar approaches are used, with exceptions that are noted in this chapter. The unique considerations for echocardiography in transcatheter valves are covered in greater detail in Chapter 32 and the appropriate use of echocardiography in prosthetic valves is covered in Chapter 47 .
While all echo modalities are used in the evaluation of prosthetic valves, there should be a low threshold for transesophageal echocardiography (TEE) in virtually all cases of known or suspected prosthetic valve abnormality. Also helpful are three-dimensional (3D) techniques, particularly in the setting of valve stenosis or dehiscence with paravalvular regurgitation.
Normal Appearance and Function
The most commonly encountered mechanical prostheses are bileaflet or single tilting-disc valves, although ball and cage valves, which are no longer implanted, may be seen as well ( Fig. 31.1 ). The majority of bioprosthetic valves are stented porcine or bovine pericardial valves, although freestyle (stentless) xenografts, cadaveric homograft, autograft (Ross procedure), transcatheter and sutureless surgical valves, which represent a hybrid between conventional stented and transcatheter bioprostheses, are also available ( Fig. 31.2 ). Prosthetic annular rings are also commonly used in mitral and tricuspid repair. The sewing rings of all valves, as well as the leaflets/discs of mechanical valves, may cause acoustic shadowing that limits imaging and Doppler assessment. Additionally, the material of the ball-in-cage valves transmits sound more slowly than human tissue with the result that the ball appears much larger than its actual size when imaged echocardiographically.
In general, the range of problems that can involve prosthetic valves includes pathologic stenosis and/or regurgitation and other instances where the appearance is abnormal, although function remains within normal limits. It is very helpful to know the valve type and size. At the time of implantation, patients are given wallet cards that include this information and, at the time of scheduling for echocardiographic evaluation, patients should be told to bring these cards. An alternative source of information is the operative note, which may be available in the patient’s medical record. The operative note will also provide details about any deviations from standard surgical techniques, such as atypical positioning of the valve and the use of surgical glue that may translate into an atypical baseline echocardiographic appearance. It is also very helpful to have access to the intraprocedural transesophageal echocardiogram.
A core concept for prosthetic valves is that valve replacement is not curative. Thus, even normal prostheses are variably stenotic with the degree of stenosis inversely related to valve size. Additionally, trivial degrees of valvular regurgitation are normal findings and, while not normal, trivial paravalvular regurgitation is not uncommon. Intraventricular micro-cavitations are often seen in the presence of mechanical valves. These findings underscore the importance of a baseline echocardiographic evaluation, which is critically important to establish gradients and the degree of regurgitation, if any, at a time when the valve is presumably normal. Typically, the baseline transthoracic echo is performed before patients are discharged or within the first 4–8 weeks post-discharge. The latter is important if the post-discharge echo is technically difficult due to residual intrathoracic air and/or the inability of the patient to move without discomfort.
Table A.15 in Appendix A provides normal echocardiographic values for the most commonly implanted valves. A rule of thumb, which is helpful when the valve size is unknown, is that for commonly sized prostheses with physiologic heart rates (HRs) and stroke volumes, the peak transaortic velocity should be less than 3 mps and the mean transmitral gradient should be ≤5 mm Hg.
Fig. 31.1 and corresponding 
show the most common mechanical prostheses and their transesophageal echocardiographic counterparts. All echocardiograms show valves in the mitral position. Note should be made of the reverberation artifacts caused by the mobile disc elements for the bileaflet and tilting-disc valves, as well as the acoustic shadowing caused by the sewing rings of all valves. In the case of tilting-disc valves, such an artifact also emanates from the disc pivot point. A different type of artifact is associated with ball-in-cage valves. This arises because echocardiography assumes that ultrasound waves will encounter only biologic tissue and move at a constant speed. However, sound moves through the ball of a ball-in-cage valve more slowly than it does through tissue and, thus, the trailing edge of the ball is represented as if the ball were much larger than it actually is. While ball-in-cage valves are no longer implanted, given their durability it is not unusual to encounter such valves in the echocardiography lab. 
demonstrate physiologic degrees of regurgitation for these valves. Note that for the bileaflet valve there are multiple jets around the perimeter as well as centrally. With a single tilting-disc valve there are also perimeter jets but the central jet is larger. Very little if any regurgitation is seen with a normally functioning ball-in-cage valve.
Fig. 31.2 shows the most commonly encountered bioprosthetic valves and their echocardiographic appearance. In these examples, all prostheses have been implanted into the aortic position. Stented bioprostheses, of which one example is provided, are the most common bioprosthetic valves. However, transcatheter aortic valves, either balloon-expandable or self-expanding, as shown here, are increasingly encountered, and are discussed in detail in Chapter 32 . Note that with the balloon expandable aortic prosthesis, the support is provided by a metal frame rather than by three stents. For the balloon-expandable valve, the frame is shorter than that for the self-expanding valve, but for both there is a variable offset between the lower end of the frame and the position of the cusps. This is best appreciated in the 
. Since valvular or paravalvular regurgitation is not invariable in bioprostheses, although still relatively common in transcatheter valves, no representative color images are provided. Similarly, since the echocardiographic appearance of stentless heterograft prostheses as well as homografts (cadaveric aortic valves) or autografts (pulmonic valve transplanted to the aortic position during the Ross procedure) is typically no different from that of the native valve, no figures are provided.
Prosthetic Valve Stenosis
The echocardiographic tools used to identify and quantitate prosthetic valve stenosis are similar to those used for native valves. They include mean and peak gradients, and valve area calculated by the continuity equation, which is most widely used for aortic prostheses and termed the effective orifice area (EOA). While the pressure half-time can be calculated for mitral prostheses, it cannot be extrapolated to an absolute valve area, but rather can be used per se for longitudinal monitoring. The Doppler velocity index (DVI) is the prosthetic valve equivalent of the dimensionless index.
Note should be made of the impact of cardiac output and HR on prosthetic gradients, particularly for valves in the mitral position, with gradients being higher when stroke volume and/or HR is/are increased. The corollary of this observation is that echocardiographic reports should always make note of the HR at the time of hemodynamic evaluation.
Gradients
For native valves, it is very important to note the angle dependence of Doppler as it is used to capture velocities from which gradients are derived. Fig. 31.3 shows aortic prosthetic gradients recorded from the apical, suprasternal, and right parasternal windows demonstrating that, as is common for native valves, the gradients are typically higher when recorded from the right parasternal window.
To interpret gradients, it is important to have normal reference values, which are valve type and size specific. Table A.15 in Appendix A reproduces the values provided in the ASE–EACVI recommendations. Note that these are derived from echocardiographic rather than in vitro flow tank values, thus adjusting for pressure recovery—a concept that is discussed below. Another useful reference will be the implantation values recorded either intraoperatively or on the first postoperative transthoracic study.
Pressure Recovery
The concept of pressure recovery is not unique to prosthetic valves, although, most commonly, it is a clinical consideration in the setting of small mechanical aortic prostheses. When blood encounters an area of narrowing, pressure energy is converted to kinetic energy with the result that the pressure at the vena contracta or effective orifice will be at its lowest ( Fig. 31.4 ), and the pressure gradient recorded at this site will be highest. Distal to the obstruction, kinetic energy can be either dispersed as thermal energy or recovered as pressure energy so that the pressure recorded distal to the obstruction will be higher, and the pressure gradient lower than that recorded at the effective orifice. Pressure recovery is more prominent when the aortic root and ascending aorta are small. Based, in part, on in vitro studies, a small aortic root is one with a diameter of ≤3 cm at the sinotubular junction, and a small prosthesis is ≤19 mm. The pressure recovery explains the discrepancy between peak instantaneous gradients measured by echocardiography and catheterization, as echocardiography measures gradients at the vena contracta, while catheterization measures the gradient distally after pressure has recovered. It is also a consideration in the differential diagnosis of high Doppler-measured prosthetic gradients in small mechanical prostheses. Of course, there is also a fundamental difference between the peak instantaneous gradient measured by echo and the peak-to-peak gradient, which is more commonly measured invasively.
A related concept is that of the relative gradients measured across the central and lateral orifices in bileaflet mechanical prosthesis. It is recognized that the gradients across the central orifice will be higher than those across the lateral orifices, and that a continuous-wave (CW) spectrum optimized to capture the highest velocity will indeed capture that across the central orifice. For aortic prostheses, even with TEE, it is rarely possible to selectively interrogate central and lateral orifices, although this may be possible with TEE evaluation of mitral valves. Pressures measured distal to the valve will reflect a mixing of the central and lateral jets with the result that the pressure gradient will be lower than that calculated at the vena contracta. The differences are most dramatic in small valves.
Valve Area and Other Measures of Valve Stenosis
Applying the continuity equation in a manner analogous to that used in native aortic valve stenosis can provide the EOA for aortic prostheses. As with native aortic stenosis, it is critical that left ventricular outflow tract sampling be proximal to the site of flow acceleration, and in the case of transcatheter or sutureless valves, that it be proximal to the inlet of the metal frame, since, in these valves, there is flow acceleration at the inlet to the metal frame as well as at the level of the cusps with both elements contributing to the total obstruction provided by the valve. While the continuity equation approach to calculating EOA has been proposed for mitral prostheses, it has not been extensively validated. It has been suggested that it is best suited for mitral bioprostheses and tilting-disc mechanical prostheses, as it is limited in bileaflet valves by the tendency of Doppler to record the higher central orifice rather than lateral orifice flows. Normal valve- and size-specific values for EOA are provided in Table A.15 in Appendix A . The pressure half-time should not be used to calculate EOA in the setting of mitral prostheses, although the absolute value of the pressure half-time in milliseconds can be used for longitudinal follow-up in individual patients with progressive lengthening suggesting the development of prosthetic stenosis.
Given the challenges of measuring the left ventricular outflow tract diameter and calculating the cross-sectional area, the DVI has been proposed as an alternative measure of the degree of valve obstruction. Conceptually, this is the same as the dimensionless index for native aortic stenosis and is calculated as the ratio of the peak velocity in the left ventricular outflow tract to the peak velocity of the aortic valvular jet. A DVI less than 0.25 is considered to be highly suggestive of significant aortic valve obstruction.
The DVI calculated for the mitral valve is done somewhat differently and reflects the ratio of the velocity time integral (VTI) proximal to the mitral prosthesis to the VTI in the left ventricular outflow tract. For this index, a ratio greater than 2.5 is highly suggestive of significant mitral valve prosthetic obstruction.
For aortic prostheses, an additional semiquantitative measure of stenosis is the acceleration time of transvalvular flow, as measured by CW Doppler, with a rounded jet envelope being more suggestive of significant intrinsic dysfunction than one with a sharp upstroke. The acceleration time is calculated from the time of onset of flow to the peak velocity, and a value greater than 100 ms is suggestive of intrinsic prosthetic dysfunction.
Patient-Prosthesis Mismatch
With prosthetic valves there are scenarios where the gradients are high but the valve has a normal appearance and a normal-for-valve type and size calculated EOA. In this case, patient prosthesis mismatch (PPM) must be considered. Simply put, this is a scenario where the valve is too small for the patient, which is a situation that can arise when native valve geometry, typically excessive calcium, limits the size of the valve that can be implanted. In the setting of PPM, typically there will be persistence of abnormally high postoperative gradients, although PPM is defined on the basis of body surface area–indexed EOA in units of cm 2 /m 2 .
For valves in the aortic position, an acceptable indexed EOA is greater than 0.85 cm 2 /m 2 . Values of 0.66–0.85 cm 2 /m 2 are considered to reflect moderate PPM, and values of ≤0.65 cm 2 /m 2 are considered to reflect severe PPM. This physiology is less commonly encountered with mitral prostheses, but definitions have been proposed as follows: mild or no PPM-indexed EOA greater than 1.2 cm 2 /m 2 ; moderate PPM-indexed EOA 0.9–1.2 cm 2 /m 2 ; and severe PPM-indexed EOA less than 0.9 cm 2 /m 2 .
The prevalence and consequences of PPM have been debated. For aortic PPM, it has been reported that there is less improvement in postoperative functional class, an increased incidence of late cardiac events, and less regression of left ventricular hypertrophy. Studies have suggested a major impact on perioperative mortality, particularly if left ventricular dysfunction is present and there is a moderate impact on late mortality; that is, after 7 years. The impact of mitral PPM has been less clearly established.
An algorithm has been proposed for the interpretation of high gradients in valve prostheses as shown in Box 31.1 and an application of the algorithm is shown in Fig. 31.5 . In summary, calculated EOA is compared with reference values for the same type and size of prosthesis, and if the EOA is within the normal range for the valve as published, an indexed EOA should be calculated. Depending on the value, a diagnosis of PPM can be confirmed or excluded. However, it should be emphasized that because of the potential for error in the calculations, and uncertainty regarding the role of indexing, particularly in obese patients, PPM should be a diagnosis of exclusion, one made after TEE and/or fluoroscopy have been performed to definitively exclude intrinsic valve dysfunction.
