There are several ways in which IE can appear by echocardiography. The principal echocardiographic manifestations of IE include the following.

As noted above, echocardiography is considered essential in the diagnosis and management of IE. The major question is whether to perform TTE or use TEE to evaluate for vegetations. In adults, a number of studies have confirmed the superior diagnostic sensitivity of TEE over TTE for IE [37–40]. This is particularly true in those patients with intracardiac abscesses and prosthetic valve endocarditis, in whom TEE appears clearly superior for diagnosis [28, 41]. In children, the advantages of TEE over TTE are less apparent, particularly with the infant and younger child, in whom TTE generally provides excellent imaging. For younger pediatric patients, TTE is usually adequate for the diagnostic evaluation of IE [42]. In most pediatric patients, particularly the younger age groups, the evidence suggests that TEE can be reserved for those patients in whom imaging quality is felt to be suboptimal. A recent study in pediatric patients ages 10 days to 17.5 years (over half of whom had CHD) showed that TTE, when used in conjunction with the Duke criteria, was useful for the diagnosis of IE; the authors recommended TEE only for an “inadequate” TTE study or when there was an organism with a high association of aortic root abscess [43].

The evaluation of IE by TEE can occur in several settings. It can be performed in the ICU or ambulatory setting, serving as either diagnostic evaluation for suspected IE or as a monitoring procedure for a patient receiving treatment for known IE. It also plays a vital role in the operating room. Preoperatively, TEE is used to assess not only the vegetation or abscess, but also valvar function and the surrounding cardiac structures. When applicable, prosthetic valve dehiscence and pseudoaneurysm formation can also be evaluated. Postoperatively, TEE is used to assess the results of operative repair or valve replacement, and to guide perioperative hemodynamic management [19, 36].

If TEE is performed to evaluate for IE, a complete study should be performed using all major TEE views/windows, as outlined in Chap.​ 4. Assessment should focus on a number of details. If a vegetation is present, its appearance and motion should be evaluated in multiple planes, and linear measurements can be obtained. The risk of embolic events appears to be greatest with vegetations >10 mm on the anterior leaflet of mitral valve [44]. Valve leaflet anatomy and motion should also be evaluated both by imaging and color flow Doppler, with attention paid to any valvar perforation or chordal disruption. In the case of aortic valve endocarditis, a careful evaluation should also be made for possible abscess or aneurysm formation. If a prosthetic valve is in place, a thorough assessment should be performed for vegetations, valve leaflet motion, and possible perivalvar abscess/dehiscence. Color flow Doppler is useful to determine valve competency and flow profile. A complete evaluation should be performed of other cardiac structures to rule out structural defects and/or other potential sites of IE, including the aorta and pulmonary artery (when the pulmonary artery can be visualized by TEE). Rarely, cases of endarteritis can also be found by TEE (Fig. 16.5, Video 16.5). Finally, myocardial function should also be assessed.

Three caveats are important to consider. First, even with TEE, not all vegetations will be visible, particularly if the vegetations are smaller than the resolution limits of the TEE probe and/or TEE imaging is suboptimal. This is an important consideration in patients with operated and unoperated CHD, who can have vegetations located in areas not readily visible by TEE (e.g. a Blalock-Taussig shunt). Studies have shown that—irrespective of whether TTE or TEE is used—patients with CHD and IE are less likely to have visible vegetations [5]. Thus, the echocardiographic data should be considered in the context of the entire clinical picture, as noted with the Duke criteria listed above. In some cases, if IE is still suspected, a TTE or TEE can be performed 7–10 days later to determine if a vegetation or abscess has appeared [10]. The second important caveat is that not all echogenic masses represent vegetations. Sterile thrombi, tumors, irregular valve excrescences, and foreign material (such as suture material) can sometimes resemble vegetations. Again, the echocardiogram should be reviewed in conjunction with the entire clinical picture. If previous echocardiograms are available (either transthoracic or transesophageal), these can be very useful to make direct comparisons to determine whether an abnormal finding is new or longstanding. New findings are much more suspicious for IE. The last important caveat is that not all vegetations are infectious. A number of medical conditions can produce sterile vegetations adherent to valvar surfaces. Examples of these include systemic lupus erythematosus (Libman-Sachs endocarditis), and nonbacterial thrombotic endocarditis (NBTE, also known as marantic endocarditis). The latter can occur as complication of malignancy, uremia, burns, hypercoagulable states, or autoimmune diseases, and it has been found in approximately 1.2 % of all autopsy patients, although the reported incidence is between 0.3–9.3 % [1, 45]. In fact, Libman-Sachs endocarditis is felt to be a form of NBTE [46]. These vegetations are usually seen on the valve closure contact line of the atrial surface of the atrioventricular (AV) valves and ventricular surface of the semilunar valves. In many cases, the vegetations are benign and clinically inapparent. However, systemic embolization has been described in up to 30–50 % of patients [45–47], with a tendency towards embolization to the brain, kidney, spleen, mesenteric bed, or extremities [46, 48].

Aside from infective endocarditis, the most common pathologic cardiac masses are thrombus or tumors. The role of TEE in the diagnostic assessment of cardiac masses is generally complementary to TTE, particularly in pediatric patients. In most cases, TTE is more than adequate to evaluate the location and extent of a cardiac mass. However in older patients, and in those with poor echocardiographic windows, TEE might be called upon to assist in evaluation. For instance, TEE has become well established in adults as a requisite tool for evaluation of the etiology of ischemic strokes; approximately 15–20 % are caused by cardiogenic emboli (thrombi, tumors, vegetations, etc.) [49].

Prosthetic valves are integral to the medical and surgical management of pediatric and adult patients with CHD. These valves fall into two major categories—mechanical and biologic—based upon valve composition (Table 16.6).

Mechanical heart valves contain nonbiologic materials (polymers, metal, carbon) in all parts of the prosthesis: the valve ring, sewing cuff, and orifice occluder. A number of mechanical valves have been developed over the past 50 years, essentially of three major types. The first type to be developed was the caged ball valve, consisting of a silastic ball with a circular sewing ring and a cage formed by three metal arches. The most notable of these was the Starr-Edwards valve (Fig. 16.15c), though similar valves have been produced including the Smeloff-Cutter valve. While these valves are no longer implanted, thousands of patients who received these valves are still alive, and continue to be followed regularly. The second type of mechanical valve is the monoleaflet valve, in which a single disk is secured by lateral or central metal struts, and surrounded by a sewing ring. The disk, generally made of extremely hard carbon (pyrolytic carbon), opens by tilting at an angle (about 60°–80°), resulting in two orifices of different sizes. Typical examples of this include the Bjork-Shiley (discontinued), and the Medtronic-Hall valve (Fig. 16.15b). The third major type of mechanical valve is the bileaflet tilting disk valve, made of two semicircular pyrolytic carbon disks attached by hinges to a rigid valve sewing ring. In the open position the valve leaflets tilt to an opening angle of 75°–90°, resulting in three orifices: a small slit-like orifice centrally between the two leaflets, and two semicircular orifices laterally. Of the three types of mechanical valves, the bileaflet tilting disk valve provides the most natural blood flow, greater effective orifice area for a given valve size, and it is also the least thrombogenic. Currently, they are the most commonly implanted mechanical valves, notably the St. Jude Medical (Fig. 16.15a) and Carbomedics bileaflet tilting disk valves. The latter two valves are available in a variety of sizes (from 16–33 mm) suitable for both pediatric and adult patients. The mechanical valves have a proven record of durability, however they require ongoing anticoagulation therapy, and there are ever-present risks of thrombosis and endocarditis of the valve. In many types of mechanical valves, separate aortic and mitral versions are available; nonetheless for a number of the mechanical valves, implantation has been performed in any of the four valve positions.

Biologic heart valves are derived from human or animal tissue, and certain types contain nonbiologic material as well, such as metal and fabric. Human tissue valves fall into two categories: homografts (allografts) and autografts. Homograft valves are cryopreserved human cadaveric aortic and pulmonary valves, generally used as pulmonary or aortic valve replacements (Fig. 16.16). They come in a variety of sizes, depending upon donor availability. In contrast, an autograft represents the patient’s own valve translocated from one site to another. Usually, the autograft is the patient’s pulmonary valve translocated to the aortic position (Ross procedure) or rarely the mitral position (Ross II), with a homograft valve being placed in the original pulmonary position [86–88]. Biologic valves derived from animal tissue are known as xenograft (or heterograft) valves; the most commonly used animal tissues are porcine aortic valve and bovine pericardium, and the tissues are fixed with glutaraldehyde. These valves come in two major forms. Stented biologic valves contain a sewing ring and struts composed of nonbiologic material (metal, cloth), and valve tissue is sewn onto the fabric covering the struts. Both porcine valve (Fig. 16.15d) and bovine pericardium (Fig. 16.15e) are used with these types of valves. Stentless biologic valves contain no struts or sewing ring, which leaves more room for blood flow. Stentless xenograft valves derive primarily from harvested porcine aortic valves (Fig. 16.15f). Of note, human homograft and autograft valves also fall into the category of unstented biologic valves, since they contain no sewing ring or struts. This is because the entire homograft/allograft root (containing the valves) is harvested, thus the intrinsic structural support for the valve leaflets remains intact. Another category of bioprosthesis that has gained popularity is the Contegra pulmonary valve conduit. The Contegra conduit is a bovine jugular vein preserved in glutaraldehyde, and it contains a valve with three leaflets; the leaflets are similar to a human semilunar valve (Fig. 16.17). Since it is derived from a venous vascular structure, it is felt to be best suited for conditions of lower pressure such as the pulmonary circuit, and therefore it is used primarily for congenital heart surgeries in which a right ventricle to pulmonary artery conduit is needed such as the Ross procedure, tetralogy of Fallot, truncus arteriosus, etc. [89]. Thus it serves as an alternative to the homograft, and has achieved comparable short to intermediate term results [90–92]. Strictly speaking, it is a valved conduit (not solely a biologic valve), but it is used in a number of operations in which a valve is necessary. The Contegra conduit is available in both a supported and unsupported model; the supported model contains two external cloth covered polypropylene rings that encircle the valve annulus and commissure to provide additional support (Fig. 16.17a), and is utilized in those cases in which sternal compression could be an issue. As will be discussed below, the bovine jugular valve is also used for transcatheter valve technology.

Over the years, a large number of different valve prostheses (both mechanical and biologic) have been developed, in a variety of valve sizes and profiles. Development continues at a rapid pace, with novel alternatives currently in development or advanced clinical trials. A list of some of the better-known biologic and mechanical valves is given in Table 16.6; this list is by no means exhaustive, and new models and types are periodically being introduced. A full discussion and elaboration of the many individual valves (including details about durability and complications) would require a separate chapter. For further information, the reader is referred to a number of references providing more detailed coverage of the topic [93–97].