Intraoperative Transesophageal Echocardiography for Pediatric Patients with Congenital Heart Disease

Intraoperative transesophageal echocardiography (TEE) has been used in adult cardiac patients since the mid-1980s for the evaluation of valvular repair ^[1,2] and prosthetic valve function ^[3,4], and for monitoring of myocardial ischemia ^[5,6] and left ventricular preload ^[7-9]. Immediate detection of inadequate surgical repairs by TEE improves surgical corrections, thereby avoiding subsequent reoperations and reducing morbidity, mortality, and cost ^[10]. Until 1990, intraoperative evaluation of infants and children undergoing congenital heart surgery was not feasible with TEE because probe sizes were too large for most children. The subsequent development of miniaturized probes has generated a number of studies, which have demonstrated that TEE can be performed safely in the pediatric population and provide substantial benefit as well ^[11-15]. This review focuses on the evolution of intraoperative echocardiography for the evaluation of congenital heart surgery and the current practice of pediatric TEE in the operating room (OR). Our goal is to provide a reference source to those involved in the perioperative management of infants and children with congenital heart disease.

History of Pediatric Intraoperative Epicardial Echocardiography

One of the first demonstrations of the use of intraoperative echocardiography to direct specific details of corrective congenital heart surgery was provided by Ungerleider et al. ^[16]. They applied transducers covered in sterile latex directly to the anterior surface of the heart to assess ventricular function before and after cardiopulmonary bypass (CPB) and to document the adequacy of the surgical repair before leaving the OR ^[16]. Eighteen consecutive patients undergoing surgical repair of atrioventricular septal defects were evaluated by epicardial imaging pre- and postbypass. Imaging before CPB demonstrated morphologic features not previously appreciated in 7 of 18 patients (39%) and, after bypass, revealed significant residual shunts in 2 of 18 patients (11%). These findings "directed specific and efficient repair immediately so that all patients left the operating room with documented surgically acceptable results. Comparison of ventricular function between pre- and post-bypass studies enabled appropriate application of pharmacologic agents in the operating room if necessary." This landmark article by a leading congenital heart center team established that, in complex congenital heart surgery, intraoperative echocardiography could guide specific surgical or anesthetic adjustments to optimize outcome. These results substantiated those of Gussenhoven et al. ^[17] and were also documented in a larger series of 328 pediatric patients undergoing congenital heart surgery ^[18]. More recently, a study from Duke University described the usefulness of epicardial imaging for evaluation of patients undergoing surgical repair of tetralogy of Fallot ^[19]. These investigators concluded that postrepair identification of sizable residual abnormalities by epicardial echocardiography indicated that a reestablishment of CPB was required. This report and others ^[20] underscore the useful role of epicardial echocardiography in the repair of congenital heart disease. Some medical centers still prefer this imaging approach to TEE, but it requires participation and some experience in epicardial echocardiography by the surgeon. It also excludes others who might be independent advocates for patients from judgments about the adequacy of repair.

History of Pediatric Intraoperative TEE

The first experience with pediatric intraoperative TEE was described by Cyran et al. ^[21] in 1989 in children as young as 7.5 yr of age using an adult-sized probe. This report documented that this technique was useful for evaluating surgical repair of congenital heart lesions. The first probe specially designed for infants weighing as little as 3 kg was subsequently designed by Kyo et al. ^[22] and Omoto et al. ^[23] with the Aloka Company (UST 5234-5; Aloka/Corometrics, Tokyo, Japan). This was a single-plane, 5-MHz, 26-element, phased array probe with a maximal diameter of 6.9 mm. Reports regarding this probe appeared from around the world in the early 1990s, documenting the accuracy, immediacy, and feasibility of TEE in the assessment of pediatric surgical repairs ^[11-14,24]. Initial experience and comparative studies showed that TEE might improve the diagnostic accuracy in some patients relative to precordial echocardiography and that it had the advantage of accuracy equal to that obtained using epicardial imaging ^[24-30]. However, the described use of TEE had some limitations. For example, it failed to detect the exact size and location of lesions involving the right ventricular outflow tract.

Despite the limitations of TEE, these studies ^[25-30] and others suggested that, compared with epicardial echocardiography, TEE did not interrupt surgery, cause potential dysrhythmias or hypotension, or increase the infection risk ^[14,24]. Research therefore continued into improving the technology of pediatric TEE probes. The principal limitation of the technique was the inability to adequately visualize the ventricular outflow tracts and the poor quality of the resulting images ^[31]. Therefore, clinical investigations began in 1992 using a new high-resolution, single-plane probe with continuous wave Doppler capabilities (48-element, 5.0-MHz, phased array TEE probe) ^[32,33]. The images provided with this new probe were superior to those obtained with the previous technology, with resolution sensitive enough to define the coronary arteries and permit visualization of blood flow within these arteries, even in infants weighing 2.7 kg ^[34].

Concurrent refinement of the Doppler modes (pulsed, continuous, and color) allowed for optimal analysis of jet direction and of degrees of regurgitation, both of which are especially important in the surgical setting, in which assessment of residual lesions is necessary.

Imaging the outflow tracts and, in particular, obtaining pressure gradients across them, was still unresolved, even after the development of pediatric biplane probes with both transverse and longitudinal planes ^[35-39]. The addition of the longitudinal plane allowed for a more complete examination of the ventricular outflow tracts, but the Doppler angle of incidence still did not allow for assessment of gradients in this plane. Several investigators then demonstrated that advancing the single-plane probe into the fundus of the stomach could provide a view that favorably imaged the outflow tracts, providing for intraoperative assessment of potential areas of residual obstruction ^[40,41]. Accordingly, this new imaging plane, the transgastric view, overcame the limitations of transverse single-plane and biplane imaging.

To date, most institutions use a combination of all approaches (single-plane, biplane, and transgastric views) for complete evaluation of congenital heart lesions. A new multiplane probe for children is currently in development ^[42] in addition to one developed by Vingmed Sound (Vingmed, Horten, Norway), which will obtain images in several planes, an obvious advantage ^[43]. Because of these advancements, pediatric TEE has been a rapidly evolving field and has now become a standard of care in many cardiac surgical centers for intraoperative and early postoperative evaluation of patients undergoing surgery for congenital heart disease.

Indications for Intraoperative TEE

Indications for TEE in children have been proposed by the Committee on Standards for Pediatric Transesophageal Echocardiography, Society of Pediatric Echocardiography in 1992 ^[44]. In 1996, task forces of the American Society of Anesthesiologists and the Society of Cardiovascular Anesthesiologists ^[45] published practice guidelines for perioperative transesophageal echocardiography. At the same time, the American College of Cardiology/American Heart Association Task Forces, in collaboration with the American Society of Echocardiography ^[46], published guidelines for the clinical application of echocardiography. The indications for pediatric intraoperative TEE from the three reports are summarized as follows:

1. Intraoperative and postrepair examinations are indicated when operations are performed on cardiac defects in which there are significant residual abnormalities such as outflow tract obstruction, valve regurgitation or stenosis, or intracardiac communications are anticipated or suspected ^[44].

2. Most cardiac defects requiring repair under cardiopulmonary bypass are a category 1 indication for intraoperative TEE, including pre- and postcardiopulmonary imaging (defined as that being supported by the strongest evidence or expert opinion substantiating that TEE is useful in improving clinical outcomes) ^[45].

3. Monitoring and guidance during cardiothoracic procedures associated with the potential for residual shunts, valvular regurgitation, obstruction, or myocardial dysfunction is a class 1 indication (defined as conditions for which there is evidence and/or general agreement that a given procedure or treatment is useful and effective) ^[46].

Thus, three separate reports by different bodies of experts categorically state that TEE is indicated in infants and children undergoing repair of congenital heart disease.

Intraoperative TEE Technique

Equipment

Both single-plane and biplane pediatric probes that provide excellent resolution are commercially available. We use two imaging systems (Hewlett Packard, Anderson, MA and Acuson, Mountain View, CA) with equal frequency. Hewlett Packard has a 48-element, single-plane HP[registered sign] TEE probe with dimensions of 10 x 8 mm and a dual-frequency biplane probe with 64 elements per transducer and dimensions of 9.1 x 8.8 mm. These probes interface with several of their imaging systems. Acuson has a dual-frequency biplane probe with 32 elements per transducer and dimensions of 9.5 x 8.7 mm, which currently interfaces with the Aspen system (Acuson). Company specifications state that the probes are suitable for patients weighing as little as 3 kg. Other TEE probes suitable for pediatric use are marketed by various companies.

Probe Selection

The usual weight range for infants and children who can be safely imaged with the current commercially available pediatric single-plane probe varies from 3 to 20 kg ^[26,47,48], although there are reports documenting the use of TEE in patients weighing as little as 2.4-2.7 kg ^[34,49]. Vigilance should be directed toward the hemodynamics and peak airway pressures during probe insertion, particularly in the extremely small neonate. The probe should be repositioned or removed immediately if any airway or hemodynamic compromise is suspected. A relatively recent report has described the use of a pediatric biplane transducer with 64 imaging elements in each pallet in 46 infants and children (lowest weight 2.9 kg) ^[35] undergoing surgery for complex congenital heart disease. Probe insertion was successful in all patients, but imaging was discontinued in one neonate because of possible airway compression by the probe. There is little difference between the actual transducer dimensions of the single- and biplane probes, although, clinically, the biplane probes are sometimes easier to insert. The multiplane probe developed by Vingmed has an external diameter of 10.5 mm; the smallest patient in whom it has been used weighed 4.9 kg ^[43].

In patients >20 kg, we prefer to use adult-sized probes because of their superior resolution capabilities. Our experience with the multiplane probes strongly favors this over the biplane transducers, as the long axis of the heart deviates approximately 30[degree sign] from the long axis of the esophagus, and ideal positioning of the interrogating plane is not dictated by the fixed plane of the transducer.

TEE Probe Insertion

The patient should be supine with the head in the midline or lateral position. After the induction of general anesthesia and endotracheal intubation, gastric contents may be suctioned to optimize image quality. The lubricated probe, which should be in the unlocked position, should be passed gently into the esophagus. A jaw thrust of the mandible frequently assists in the passage of the probe. If difficulty is encountered, a laryngoscope should be used to advance the transducer into the esophagus under direct visualization. Once the transducer is positioned behind the heart, the patient's head can be positioned to one side to avoid interference with the surgical procedure during manipulation of the probe.

Imaging Technique

The pediatric TEE probe can be manipulated in three directions: advanced or withdrawn, anteflexed or retroflexed, and rotated clockwise or counter-clockwise relative to the sagittal plane. As general rules of transducer manipulation, anteflexion of the transducer brings structures anterior and toward the base of the heart into view, clockwise rotation allows imaging of rightward structures, and counter-clockwise rotation permits viewing of left-sided structures.

We recommend the image orientation and nomenclature guidelines proposed by the American Society of Echocardiography for the standard basal short-axis, four-chamber, long-axis, and transgastric short-axis TEE views ^[50]. In the smallest patients (3-4 kg), very small adjustments in probe position, i.e., movements of only a few millimeters, are required to change from view to view. The distance between the different views increases as the patient's weight increases (Figure 1).

Figure 1: Distances for obtaining the standard views from the central incisors or gums according to weight. PA valve = pulmonic valve, Ao SA = aortic short-axis view, LVOT = left ventricular outflow tract, 4C = four-chamber view, SA = left ventricular short-axis view. Reproduced with permission from Echocardiography 1993;10:599-608.

Transverse Plane Examination. Short-axis views consist of sectional planes oriented in an axial plane and are therefore slightly oblique relative to a true short-axis of the heart, ranging from the basal to the transgastric level. The most cranial short-axis view is obtained at the base of the heart, from which the probe is anteflexed slightly to display the aorta, the proximal pulmonary artery, and its bifurcation (Figure 2A). Advancement of the probe displays the aortic valve in short axis and the proximal ascending aorta and the origins of the coronary arteries (Figure 2B). Counterclockwise rotation of the probe in this position shows the left pulmonary veins; clockwise rotation shows the right pulmonary veins, superior vena cava, and right atrial appendage. Advancing the probe further, the five- and four-chamber views are obtained, demonstrating the atrial, atrioventricular, and ventricular septae and atrioventricular valves (Figure 2C, and Figure 2D). By advancing the probe into the stomach, the left ventricular short-axis images are obtained with multiple cross-sectional views of the left ventricle, mitral valve, papillary muscles, and oblique sections of the right ventricle demonstrated by slight probe flexion and/or rotation at this level (Figure 2E).

Figure 2: Transverse plane examination. The diagrams demonstrate the different views that can be obtained during the transverse plane examination. Ao = aorta, MPA = main pulmonary artery, RPA = right pulmonary artery, LPA = left pulmonary artery, AoV = aortic valve, RA = right atrium, LA = left atrium, RVOT = right ventricular outflow tract, RV = right ventricle, LV = left ventricle, LVOT = left ventricular outflow tract.

Longitudinal Plane Examination. Once the standard transverse TEE is completed, a longitudinal examination can be performed. The TEE long-axis view is obtained by anteflexion and clockwise rotation of the probe to display the interatrial septum and entrance of the superior vena cava (Figure 3A). Counter-clockwise rotation of the probe provides visualization of the left ventricular outflow tract and the ascending aorta (Figure 3B), and further counter-clockwise rotation provides visualization of the right ventricular outflow tract and main pulmonary artery (Figure 3C) ^[38,51-53].

Figure 3: Longitudinal plane examination. The diagrams demonstrate the views that can be obtained with this plane. SVC = superior vena cava, IVC = inferior vena cava, RA = right atrium, LA = left atrium, RPA = right pulmonary artery, Ao = aorta, LVOT = left ventricular outflow tract, RVOT = right ventricular outflow tract, PA = pulmonary artery.

Transgastric Plane Examination. When only a single-plane probe is available, a complete examination can still be achieved by using transgastric imaging planes: the probe is passed into the stomach, anteflexed maximally, and advanced anteriorly to the fundus. When the patient's abdomen is exposed, it is often possible to see the tip of the probe slightly distending the abdominal wall during this maneuver. If there is difficulty in achieving the views, the probe is then withdrawn, readvanced, and withdrawn with maximal anteflexion to ensure good probe contact. This anteflexion should not be performed if resistance is encountered. From this position, rotation of the probe to the left with moderate deflexion provides images of the anterior right ventricular outflow tract and the proximal pulmonary trunk as it courses anteriorly across the surface of the heart (Figure 4A); clockwise rotation and slight flexion from this position permits similar evaluation of the left ventricular outflow tract (Figure 4B). The flexion of the probe is then increased slightly to define the inlet and outlet components of the ventricular septum, as well as the atria and atrioventricular valves (Figure 4C). The entrance of the pulmonary veins into the left atrium is demonstrated from this plane (Figure 4D), and with rotation of the probe to the right, the venoatrial connections of the right atrium can also be seen. Because the probe is some distance from the heart with a portion of the liver interposed, small movements of the transducer subtend large imaging arcs, permitting examination of the heart from the posterior atrial wall to near the anterior surface of the right ventricle. After imaging of the outflow tracts, pulsed, continuous-wave and Doppler color-flow mapping are performed from this transgastric location and from all other transducer locations as required by the suspected pathology. This transgastric approach is also useful with a biplane probe in view of the alignment of the outflow tracts to the Doppler angle of interrogation. Pathological slices obtained from normal anatomical specimens of hearts substantiate the transgastric imaging planes (Figure 5 and Figure 6) ^[41].

Figure 4: Transgastric examination. The diagrams demonstrate the four views commonly obtained from the transgastric window. RV = right ventricle, LV = left ventricle, PA = pulmonary artery, SVC = superior vena cava, RA = right atrium, Ao = aorta, MPA = main pulmonary artery, LAA = left atrial appendage, LA = left atrium.

Figure 5: The morphology of the right and left ventricular outflow tracts from their respective ventricles with the corresponding cut specimens. Top left, anatomical specimen of the right ventricular outflow tract as viewed from the transgastric approach cut to display the morphology of an anteriorly directed plane through the right ventricular outflow tract and the pulmonary artery. Top right, transgastric echocardiogram displaying a normal right ventricular outflow tract with the anterior papillary muscles and septomarginal trabeculation. The left ventricle is shown in its short axis. PA = pulmonary artery, LV = left ventricle, RV = right ventricle. Bottom left, anatomical specimen of the left ventricular outflow tract to simulate the transgastric view. This slice is posterior to the previous cut, showing the aorta arising from the left ventricle. The tricuspid valve is seen in the posterior inlet portion of the right ventricle. The mitral valve and its tendinous attachments are seen. Bottom right, transgastric echocardiogram of a left ventricular outflow tract showing the left ventricle and the ventriculoarterial connection to the aorta. The supravalvar aortic area is well displayed. AO = aorta, LV = left ventricle, RA = right atrium, RV = right ventricle. Reproduced by permission of the American Society of Echocardiography.

Figure 6: The morphology of the membranous and muscular components of the atrioventricular septum. Top left, anatomical specimen cut to display the atrioventricular component of the membranous septum (interposed between right atrium and subaortic outlet). Top right, corresponding transgastric echocardiogram displaying the membranous septum. AO = aorta, LV = left ventricle, RA = right atrium. Bottom left, the most posterior cut in anatomical specimen revealing the musculoatrioventricular septum, inlet valves, and entrance of the pulmonary veins. Bottom right, transgastric echocardiogram demonstrating the inlet valves, left atrium, pulmonary veins, and a portion of the left atrial appendage. LA = left atrium, LAA = left atrial appendage, PV = pulmonary vein, RA = right atrium. Reproduced by permission of the American Society of Echocardiography.

Morphologic Analysis

Assessment of the connections of the various cardiac segments, atrial arrangement or situs, venoatrial, atrioventricular, and ventriculoarterial connections should be performed. Septal and valvar structures should then be evaluated, including assessment of flow velocities with Doppler echocardiography. Both transgastric imaging and midesophageal longitudinal plane imaging accurately display the ventriculoarterial connections in detail: we prefer the transgastric approach because it has certain diagnostic advantages ^[40,41]. This approach permits intraoperative imaging of the right and left ventricular outflow tracts from a vantage point similar to that achieved with the subcostal (subxyphoid) location.

Recognition of Common Lesions

The transesophageal planes suitable for each of the most commonly found lesions are defined in Table 1 and Table 3, and representative echocardiograms for some of these are provided in Figure 7, Figure 8, Figure 9, Figure 10 and Figure 11. (For comprehensive reference atlases, the reader is referred to Pediatric Echocardiography edited by N. Silverman, Echocardiography in Pediatric Heart Disease by R. Snider, or Transesophageal Echocardiography in Congenital Heart Disease edited by O. Stumper and G. Sutherland.)

Table 1: Congenital Heart Lesions: Which TEE Plane to Use

Table 3: Table 1. Continued

Figure 7: Atrial septal defects. Top, transverse plane view of secundum atrial septal defect with and without color-flow Doppler to demonstrate the left-to-right atrial shunt. Middle, longitudinal plane view of a sinus venosus atrial septal defect. Bottom, transverse plane view of a primum atrial septal defect before and after surgical repair. Arrows indicate sutures in the associated cleft mitral valve. ASD = atrial septal defect, RA = right atrium, LA = left atrium, RV = right ventricle, LV = left ventricle, IAS = interatrial septum, SVASD = sinus venosus atrial septal defect, SVC = superior vena cava.

Figure 8: Ventricular septal defects. Top left, transverse plane view of a doubly committed subarterial ventricular septal defect (VSD) with color flow demonstrating the left-to-right ventricular shunt. Top right, longitudinal plane view of the same defect showing the size of the VSD adjacent to the pulmonary valve and locating the defect as a doubly committed subarterial type. Middle left, transverse plane view of a perimembranous VSD with a ventricular septal aneurysm (arrow). Middle right, transverse plane view of a perimembranous VSD with color-flow Doppler to demonstrate the left-to-right shunt. Bottom, transverse plane view of an atrioventricular septal defect (AVSD). This diastolic frame shows the single valvar orifice overlying both ventricles. RV = right ventricle, LV = left ventricle, RVOT = right ventricular outflow tract, AO = aorta, LA = left atrium, RA = right atrium.

Figure 9: The morphology of a patient with tetralogy of Fallot and the corresponding transesophageal and transgastric echocardiograms. Left, transesophageal five-chamber view showing the significant override of the aorta. LA = left atrium, LV = left ventricle, RA = right atrium, RV = right ventricle. Middle, transgastric echocardiogram showing the aortic override and ventricular septal defect (labeled subcostal equivalent). AO = aorta. Right, anatomical specimen displaying the right ventricular hypertrophy and the aorta overriding the ventricular septal defect. Reproduced by permission of the American Society of Echocardiography.

Figure 10: The use of transgastric imaging for evaluating the ventriculoarterial connections in truncus arteriosus (top) and transposition of the great arteries (bottom). Top left, transgastric echocardiogram of truncus arteriosus showing the ventriculoarterial connections with a large central vessel arising above the ventricular septal defect. Anterior angulation shows the aortic arch arising from the overriding truncal vessel. AO = aorta, RV = right ventricle, LV = left ventricle. Top right, transgastric echocardiogram of a second patient with truncus arteriosus (type II) with origins of the right and left pulmonary artery (RPA and LPA, respectively) displayed clearly from a large common trunk. TR = truncal root. Bottom left, transgastric echocardiogram showing the pulmonary artery arising from the posterior left-sided chamber of left ventricular morphology in a patient with transposition of the great arteries. The pulmonic valve is shown (arrow). RA = right atrium, PA = pulmonary artery. Bottom right, transgastric echocardiogram in a patient with transposition of the great arteries. The transposed aorta (AO) arises from the more anterior right-sided chamber with right ventricular morphology. Reproduced by permission of the American Society of Echocardiography.

Figure 11: Two congenital heart lesions: a hypoplastic left ventricle and an atrioventricular septal defect. Top left, transgastric view of a patient with hypoplastic left ventricle, with a small aorta (Ao) and right and left coronary arteries (RCA and LCA, respectively) arising from this small vessel. The descending aorta (DAo) is also shown, along with the coronary sinus (CS). RA = right atrium, RV = right ventricle. Top right, transgastric imaging of the right ventricular outflow tract and normal-sized pulmonary artery (PA) in the same patient. Bottom left, transgastric echocardiogram in a patient with an atrioventricular septal defect, showing the interatrial septal defect (lower arrow) and the entrance of the pulmonary veins in addition to a secundum atrial septal defect (upper arrow). LA = left atrium, LV = left ventricle. Bottom right, transgastric echocardiogram in a patient with an atrioventricular septal defect, showing the free-floating bridging leaflets (arrow) and the muscular atrioventricular septum. Reproduced by permission of the American Society of Echocardiography

Assessment of Residual Pathology

A complete postbypass echocardiographic examination requires attention to the details of specific lesions. Guidelines for the evaluation of each lesion after repair are suggested in Table 2 and Table 4. Some commonly encountered residual defects that may require return to bypass include a residual ventricular septal defect, left or right outflow tract obstruction, and atrioventricular valvular regurgitation (Figure 12, Figure 13, Figure 14 and Figure 15).

Table 2: Postbypass TEE Checklist for Various Diagnoses

Table 4: Table 2. Continued

Figure 12: Subaortic stenosis. Top, transgastric echocardiogram showing membranous ridge of subaortic stenosis underneath the aortic valve (arrows). Middle, color-flow Doppler demonstrating aliasing across the obstruction. Bottom, continuous-wave Doppler across the left ventricular outflow tract demonstrating the gradient.

Figure 13: A residual mid-muscular ventricular septal defect. Top left, two-dimensional four-chamber view with contrast showing a large ventricular mid-muscular septal defect prebypass. Top right, color Doppler echocardiogram displaying left-to-right shunt (blue flow) through the ventricular septal defect. Bottom left, two-dimensional four-chamber view postbypass showing apparent intact patch repair. Bottom right, color Doppler showing a residual defect with aliased flow at the inferior portion of the ventricular septal defect (VSD) patch. LA = left atrium, LV = left ventricle.

Figure 14: Residual subpulmonic stenosis. Top, longitudinal view of the right ventricular outflow tract with the arrows indicating residual subpulmonic stenosis. Bottom, Doppler color flow through the residual pulmonic stenosis showing disturbed flow. Ao = aorta, PA = pulmonary artery, RV = right ventricle.

Figure 15: Residual mitral and tricuspid regurgitation. Top, two-dimensional four-chamber view postbypass showing residual mitral regurgitation. Bottom, two-dimensional four-chamber view postbypass showing residual tricuspid regurgitation. LA = left atrium, LV = left ventricle, MR = mitral regurgitation, RA = right atrium, TR = tricuspid regurgitation.

In all examinations, careful two-dimensional colorpulsed and continuous-wave Doppler should be used. To evaluate for residual shunts, we use contrast agents because the microbubbles are readily apparent even when a very small number cross a defect ^[54]. Contrast echo can also be used for identification of systemic venous connections, particularly in patients with left superior vena cava to coronary sinus connections. Saline may be used, but, in our experience, 0.5-1.0 mL of the patient's blood agitated vigorously between two syringes with 3-5 mL of saline produces a very satisfactory effect ^[55].

Assessment of Pressure Gradients

The transgastric approach allows axial alignment for Doppler estimates of gradients across the outflow tracts and their quantitative assessment, thereby permitting intraoperative detection of residual obstructions ^[41,47,69]. This approach overcomes the limitations of transverse single-plane and biplane imaging of these structures; for example, biplane imaging with a longitudinal plane can visualize the right ventricular outflow tracts but provides only anatomic and color Dopper information; gradients cannot be assessed because of poor alignment of the Doppler beam with the direction of blood flow.

Doppler measurements of the right and left ventricular outflow gradients should be performed preferentially from the transgastric position (using the modified Bernoulli equation [peak pressure difference = 4 (V)²]) (Figure 16). In our series and others ^[41,47,69], reasonable correlation has been confirmed between Doppler estimates and direct pressure measurements (Figure 17). Thus, this technique can provide the operative team with an online assessment of pressure gradients before and after the surgical repair. It is important to note that preoperative estimates of gradients obtained when patients are awake or lightly sedated may differ from those obtained by TEE at the time of operation when the patient is under general anesthesia.

Figure 16: The calculation of residual right ventricular outflow tract obstruction using the Bernoulli principle. The flow disturbance caused by this condition is shown by using color Doppler flow imaging. Top, continuous-wave Doppler interrogation across the right ventricular outflow tract predicts a pressure gradient of 54 mm Hg postbypass. Bottom, after further surgery, the residual gradient is predicted to be 28 mm Hg.

Figure 17: The comparison of the difference in pressure gradients across the right and left ventricular outflow tracts derived by Doppler and cardiac catheterization or direct pressure gradient measurements. The vertical axis shows the difference between the measurements, and the horizontal axis shows the mean measurement obtained by each method. Reproduced by permission of the American Society of Echocardiography.

Assessment of Ventricular Preload and Function

TEE can measure those factors that directly affect ventricular function, such as preload, contractility, and afterload. TEE has been shown to be a reliable monitor of cardiac filling changes in pediatric patients. In a study of patients undergoing surgical repair of congenital heart lesions, blood was withdrawn until the systolic blood pressure decreased by 5 and 10 mm Hg ^[56]. Experienced anesthesiologist-echocardiographers blinded to study events were able to identify with high sensitivity (80%-95%) and specificity (80%) these mild reductions in blood volume by TEE changes in left ventricular end-diastolic area.

TEE assesses global left ventricular function indirectly by measurement of cardiac output, stroke volume, or ejection fraction ^[9,57]. Compared with radionuclide ejection fraction, TEE has a correlation coefficient of r = 0.80 ^[58]. TEE estimation reached a good correlation (r = 0.98) compared with measures of ejection fraction by transthoracic echocardiography ^[59]. Subsequent studies have also proved that TEE provides a direct method by which to assess left ventricular preload and ejection. It should be emphasized that ejection fraction is preload- and afterload-dependent, and, as such, should be viewed as an index of overall ventricular function. A comparison was made between transthoracic echocardiography and TEE for assessment of ventricular function in pediatric patients, but the results were poor because of technical limitations ^[60].

Postbypass ventricular dysfunction has been correlated to poorer outcome ^[70]. The presence of new abnormality in wall motion after a repair carries a significantly higher mortality rate; left ventricular 25%, right ventricular 33%, and biventricular 69%. When there were no changes in ventricular function, the mortality rate was 4%. Thus, intraoperative assessment of ventricular function can be of benefit not only as a marker of contractility, but also as a predictor of outcome.

Influences of TEE in Intraoperative Surgical Management

The overall incidence of change in surgical management with the aid of intraoperative TEE has been shown to be 7% in our series and others ^{[30,34,41,47]}, and it does not differ significantly from that reported with epicardial echocardiography ^[16,18,19]. Stevenson et al. ^[47] evaluated surgical repairs in 230 patients undergoing congenital heart surgery in which cardiopulmonary bypass was used. TEE prompted a return to bypass in 7.4% of their patients: 9 of 28 for left ventricular outflow tract obstruction, 5 of 78 for residual or additional ventricular septal defects, and 3 of 51 because of other residual problems needing further surgery, such as revision of a coronary reimplantation site or residual gradients. All their findings were confirmed at additional surgery. They concluded that TEE offers substantial utility in detection of residual problems requiring reoperation. In addition, data suggest that preoperative diagnoses can be changed 3%-5% ^[26,29] and that transgastric imaging of residual gradients reveals an additional 2.2% of patients in need of a return to cardiopulmonary bypass ^[41].

Outcome and Effectiveness of Pediatric Intraoperative Echocardiography

There is little information available regarding improvement in outcome based on intraoperative transesophageal or epicardial echocardiographic findings. One prospective study was designed to analyze the efficacy of intraoperative epicardial color-flow mapping to predict outcome after surgical repair ^[70]. If patients left the operating room with residual pathology detected by echocardiography, there was a higher rate of reoperation (42% vs 3%; P < 0.006) and of early death (29% vs 10%; P < 0.006). If patients had an alteration of their ventricular function after bypass, they had a greater incidence of early death (35% vs 4%; P < 0.004). Patients with acceptable surgical results as substantiated by echocardiography had a >90% likelihood of a long-term acceptable outcome.

Information regarding the efficacy of pediatric TEE to predict outcome after surgical repair is lacking. In one series of 230 patients in which the issue of outcome was addressed, if a residual abnormality was detected by TEE and revised, the outcome was excellent in 76% (13 of 17) of patients ^[49]. Twelve patients who had suboptimal outcomes were identified: 2 of 12 (16%) had residual defects that may have been amenable to further surgery, and 10 (84%) had final outcomes that were believed to be unrelated to a problem identifiable by TEE. Thus, if immediate revision of a repair was performed, the outcome was improved; if the patient was left with a residual defect, the outcome was suboptimal. However, there are no large, controlled studies that determine the impact of intraoperative TEE in anesthetic and surgical management and overall influence in patient outcome in congenital heart surgery.

Training/Certification in Pediatric TEE Training

The subject of training and certification in pediatric TEE has received very little attention. The American Society of Echocardiography published guidelines for physician training in TEE in 1992 ^[61]. Recommendations regarding pediatric TEE were as follows:

"One should have a thorough knowledge of cardiac disease, and the hemodynamic alterations associated with acquired and congenital disorders, understanding of ultrasonic image formation and Doppler assessment of intracardiac blood flow, the range of normal structural findings and echocardiographic manifestations of a large number of cardiac disorders. Physicians performing and interpreting diagnostic TEE studies should first develop experience with general echocardiographic techniques consistent with at least level II training in echocardiography. This training involves the performance and primary interpretation (with supervision) of approximately 300 general echocardiographic studies over a period of 6 months or achievement of the equivalent level of experience. Training in the TEE examination needs to include development of cognitive and technical skills. Recommended practical experience includes introduction of the TEE probe and manipulation, optimization of instrument controls, and interpretation of the findings."

The issue of education for intraoperative TEE was also addressed by Savage et al. ^[62], who state, "Regardless of any suggested numerical guidelines, the end product of individual competency in intraoperative echocardiography is more important than the numerical milestones along the way." The basic skills recommended by the American Society of Echocardiography should probably be mastered before one considers advancement to the arena of congenital heart disease.

Specific recommendations have been made regarding training and performing pediatric TEE as defined by the Committee on Standards for Pediatric Transesophageal Echocardiography ^[44]. They suggested that the guidelines for training of nonpediatric cardiologists in pediatric TEE be consistent with those required for pediatric cardiology trainees and should be individualized to include the existing level of familiarity a physician may have with children and their heart disease. They strongly recommended that a pediatric cardiologist knowledgeable in TEE participate in the performance and interpretation of studies in infants and young children and those who have complex heart disease.

Given these recommendations and training requirements, how should one proceed as an anesthesiologist interested in pediatric TEE? There has been increasing experience by those involved in pediatric cardiac surgery and intraoperative TEE, including members of our group, and the issue of who is responsible for the intraoperative interpretation of TEE has been raised. We believe that one should follow the general guidelines established by the American Society of Anesthesiologists and the Society of Cardiovascular Anesthesiologists ^[45]:

"Anesthesiologists with advanced training in perioperative TEE should be able to exploit the full diagnostic potential of TEE in the perioperative period. Specific objectives for advanced training in TEE for congenital heart disease include cognitive skills-knowledge of echocardiographic manifestations of congenital heart disease and technical skills-and the ability to utilize TEE to evaluate congenital heart lesions and to assess surgical intervention in congenital heart disease."

Because it is essential for many intraoperative applications to obtain a definitive interpretation of the TEE examination at the time of surgery, the task force strongly recommends that anesthesiologists actively pursue collaboration with surgeons, cardiologists, or other physicians involved in a patient's care. Our practice has evolved such that we have a core group of individuals (anesthesiologists, pediatric cardiologists, and cardiothoracic surgeons) who collaborate as the guidelines suggest.

Certification

The Society of Cardiovascular Anesthesiologists has appointed the Task Force on Certification for Perioperative TEE to develop a certification process for knowledge and proficiency in interpretation of perioperative TEE. The inaugural official examination was given in 1998. However, the examination includes only adult congenital heart disease; pediatric TEE has not been addressed.

Quality Assurance

All quality assurance programs for echocardiography should include the following key processes: the indications for performing TEE, technical aspects of performing and recording the examination, application of examination findings to physiologic conditions, documentation, equipment, professional communication, education, and billing ^[45].

At the University of California-San Francisco, the pediatric TEE examinations are viewed and interpreted immediately, either in the OR or by satellite transmission to the echocardiography laboratory, and reports for the pre- and postbypass examinations are then generated.

Limitations

Experience has shown that there are limitations in TEE that must be considered during the evaluation of a patient. For example, during assessment of mitral regurgitation after repair of atrioventricular septal defects, a number of factors must be included, such as loading conditions, systolic pressure, contractility, and instrument settings ^[27,63]. Preliminary data in a small series of patients suggest that intraoperative TEE may not predict the severity of postoperative mitral regurgitation (MR); patients who had no or mild MR before leaving the OR progressed to moderate (25%) or severe (16%) MR in follow-up studies ^[64]. However, in our experience, only 2 of 34 patients with no or mild MR by intraoperative TEE developed moderate regurgitation requiring surgical intervention on short-term follow-up. Estimates of residual regurgitation and obstruction can vary dramatically depending on the loading conditions; therefore, every attempt must be made to perform the evaluation during conditions that reflect the patients' baseline hemodynamics.

Contraindications

Conditions generally associated with increased risk of complications such as esophageal pathology ^[45], severe respiratory decompensation, or inadequate control of the airway ^[44] are contraindications for TEE.

Complications

Serious complications during pediatric TEE are very rare ^[48]. However, hemodynamics and respiration must be closely monitored. Blood pressure generally remains stable but can decrease precipitously if anteflexion or retroflexion compresses the aorta (Figure 18). Therefore, we recommend placement of the TEE probe after arterial line placement ^[34,65]. Blood pressure changes induced by probe manipulation resulting in aortic compression may or may not be evident, depending on the location of the arterial catheter. Respiratory compromise can occur, peak inspiratory pressures can increase, and desaturation can be observed. In addition, there may be movement of the endotracheal tube during the TEE examination, resulting in right mainstem intubation or extubation of the trachea ^[48]. If desaturation occurs suddenly ^[34,41,66], correct positioning of the endotracheal tube must be confirmed, and the TEE probe must occasionally be with-drawn immediately. For these reasons, we recommend frequent monitoring of the peak inspiratory pressures throughout TEE examinations.

Figure 18: Hypotension during transesophageal echocardiographic examination. The electrocardiogram and arterial blood pressure tracings are displayed. Reproduced with permission from Echocardiography 1993;10:599-608.

Endocarditis after TEE examinations has been reported in the literature but is considered an unlikely occurrence ^[67]. The most recent guidelines by the American Heart Association state that endocarditis prophylaxis is not recommended for TEE but should be optional for high-risk patients ^[68].

Conclusion

In many medical centers, TEE is now the standard of care for the intraoperative assessment of most congenital heart repairs before removal of bypass cannulas and closure of the sternotomy. TEE has become a valuable adjunct to intraoperative anesthetic and surgical management. TEE clearly has a significant perioperative impact in the care of infants and children undergoing surgical repair of congenital heart disease. Diagnoses can be confirmed preoperatively or altered ^{[16,18,24,26]}, and the surgical plan can be revised ^[26,41,47]. Multicenter studies are needed to address the efficacy of TEE on the perioperative outcome of patients undergoing surgical repair of congenital heart disease.

The authors are grateful to Michel Balea, MS, for his assistance with the technical drawings.