Chapter 17. Transesophageal Echocardiography for Heart Failure Surgery

The epidemic of heart failure is a worldwide problem that is anticipated to increase with both an aging population and the improved survival from cardiac complications producing left ventricular systolic dysfunction (e.g. myocardial infarction). Increasingly, these patients who survive a serious cardiac injury but have persistent ventricular dysfunction precluding normal end-organ function experience a poor quality of life and high rates of morbidity and mortality. At the age of 40, the lifetime risk of developing heart failure is 20%, and the 1-year heart failure mortality rate is 20%.1 The number of hospitalizations for heart failure has tripled between the 1970s and 2004, and contemporary data indicate that heart failure was the primary or secondary cause of 3.8 million annual admissions in the United States.2 It is estimated that the direct and indirect costs of heart failure in the United States will exceed $37 billion in 2009, highlighting the economic importance of this disease.1

While most heart failure patients are managed medically, surgical options for refractory heart failure include orthotopic heart transplantation and mechanical circulatory support. Advances in donor and recipient selection, organ procurement, and immunosuppressant therapy have led to an increase in the survival of grafted organs. Transplant surgery is currently considered the treatment of choice for end-stage heart, lung, and liver diseases, but the predominant limiting factor is a shortage of donors. Mechanical circulatory support has therefore emerged as a valuable and viable adjunct to transplantation in the management of heart failure patients.

Echocardiography plays an essential role in the donor organ selection process and preoperative screening, perioperative management, and post-transplant follow-up of recipients. Similarly, perioperative transesophageal echocardiography (TEE) provides invaluable anatomic and functional information in patients receiving circulatory support devices, which influence not only anesthetic management but also surgical decision making. The following text will first describe the role of TEE in heart transplantation, followed by a discussion of its value in the implantation of mechanical circulatory support devices.

The application of TEE as a diagnostic and monitoring modality in heart transplant surgery can be divided into five categories:

1. Cardiac donor screening

2. Intraoperative monitoring in the pretransplant period

3. Intraoperative evaluation of cardiac allograft function and surgical anastomoses in the immediate posttransplantation period

4. Management of early postoperative hemodynamic abnormalities in the intensive care unit

5. Postoperative follow-up studies of cardiac allograft function

Role of TEE in Cardiac Donor Screening

As a result of the shortage of available donor hearts, many institutions are now liberalizing their acceptance criteria to include higher-risk (marginal) donor hearts.3 Table 17–1 presents the conventional cardiac contraindications to the use of a donor heart. Despite the potential risk for transmitting atherosclerotic, hypertensive, and valvular heart diseases, organs from older donors are increasingly being used. This aggressive approach has proved particularly successful when matching for higher-risk recipients (alternate recipient list) with a greater short-term mortality risk or with significant comorbid factors.4

Echocardiography plays an important role in the effort to improve the yield of donor evaluation.5 By ruling out donors with structural abnormalities, severe ventricular dysfunction, or significant wall motion abnormalities (WMAs), the need for costly and time-consuming cardiac catheterization can be circumvented. In potential donors on ventilatory support, TEE has been shown to be particularly useful in providing consistent high-quality imaging when transthoracic echocardiography (TTE) has proved inadequate.

An initial echocardiogram should not be obtained before adequate hemodynamic and metabolic resuscitation. In particular, volume status, acidosis, hypoxemia, hypercarbia, and anemia should be corrected, and inotropic support should be weaned to a minimum compatible with adequate blood pressure and cardiac output (CO). The goals of the echocardiogram are to rule out structural abnormalities and assess regional and global functions. It is unclear if donor hearts with left ventricular (LV) hypertrophy, defined as a wall thicker than 11 mm in the absence of underfilling of the ventricle (pseudohypertrophy), can safely be used for transplantation. One study shows that LV hypertrophy (LVH) may increase the incidence of early graft failure,6 but a more recent study demonstrated that hearts with mild (12 to 13 mm) or moderate (13 to 17 mm) LVH do not increase morbidity.3 Most valvular and congenital abnormalities preclude transplantation, with the possible exception of mild lesions such as mitral valve prolapse in the absence of significant regurgitation, a normal functioning bicuspid aortic valve, or an easily repairable secundum-type atrial septal defect.

Segmental WMAs in donor hearts may be the result of coronary artery disease, myocardial contusion, or ventricular dysfunction after brain injury. Contused myocardial tissue resembles infarcted myocardial tissue histologically and functionally.7 The pattern of ventricular dysfunction after spontaneous intracranial hemorrhage is usually segmental and often spares the apex of the left ventricle.8 This pattern correlates with the sympathetic innervation of the ventricle. In contrast, ventricular dysfunction after traumatic brain injury may be global or regional. For both types of brain injury, there is a poor correlation between the distribution of echocardiographic dysfunction and actual histologic evidence of myocardial injury. Some studies have suggested that WMA and global function improve shortly after heart transplantation, but a recent multi-institutional study identified WMA on the donor echocardiogram as a powerful independent predictor of early graft failure.9 WMA on the donor echocardiogram may be particularly important when associated with a donor age older than 40 years and an ischemic time longer than 4 hours.

The lowest fractional area change in a donor heart permitting safe transplantation is unknown, but it has been suggested that a fractional area change greater than 35%, in the absence of other cardiac abnormalities, could be used as a guide.8

Intraoperative Monitoring in the Pretransplant Period

Idiopathic and ischemic cardiomyopathies are the two most common causes of cardiac failure in the transplant recipient. Regardless of the cause of failure, global cardiac dilatation is a common feature and the term dilated cardiomyopathy has been applied to this end-stage condition. These patients have fixed, low stroke volumes and are very dependent on an adequate preload. Further, even mild increases in afterload may result in a marked reduction in stroke volume. Patients in cardiac failure compensate for their low CO by an increase in sympathetic activity, which leads to generalized vasoconstriction and to sodium and water retention. This delicate balance among preload, contractility, and afterload can be dramatically disturbed after the induction of general anesthesia. TEE is therefore ideally suited to rapidly evaluate and guide intraoperative management in these patients. Several factors commonly seen in recipients, including diastolic dysfunction, regurgitant valvular lesions, and positive pressure ventilation, result in a poor correlation between measured filling pressures and LV volumes. Thus, optimization of LV filling and inotropic support can be more readily and rapidly achieved under TEE guidance. Right ventricular (RV) size and function also should be assessed in these patients. The presence of RV hypertrophy is suggestive of long-standing pulmonary hypertension, which may lead to acute RV dysfunction in the transplanted heart.

TEE is similarly sensitive in detecting intracardiac thrombi, with the possible exception of an apical thrombus. Prethrombotic sluggish blood flow is characterized echocardiographically as spontaneous contrast or “smoke.” Patients with dilated cardiomyopathy, especially in the presence of spontaneous echo contrast, have a high incidence of thrombus formation in the apex of the left ventricle. The left atrial (LA) appendage also should be inspected for possible thrombi, particularly in patients with atrial fibrillation. When thrombi are present in the left heart, manipulation of the heart before cardiopulmonary bypass (CPB) should proceed with great caution in an effort to avoid systemic thromboembolism. Other sources of embolism during the pretransplant period include atheromatous plaque from the ascending aorta during aortic cannulation or air entrainment during the explantation of ventricular assist devices. As in all CPB cases, the aorta (ascending aorta, arch, and descending aorta) should be examined for atherosclerotic plaque before aortic cannulation. TEE is extremely sensitive in the detection of intravascular air and early detection and intervention may potentially limit this complication.

It is common practice to place a pulmonary artery (PA) catheter into the PA only after CPB because it is often difficult to pass these catheters through large dilated ventricles, incompetent tricuspid valves, and in low CO states. PA catheter placement is also more prone to induce arrhythmias. TEE therefore can be used to determine CO and PA pressures during the pre-CPB period (see Chapter 4).

Intraoperative Monitoring in the Posttransplantation Period

TEE imaging of the heart during and after weaning from CPB provides invaluable information with important diagnostic and prognostic implications. Before weaning from CPB, TEE is used to detect retained air and to assist venting and de-airing maneuvers. The most common sites of air retention are the right and left upper pulmonary veins, the LV apex, the left atrium, and the coronary sinus. The right coronary artery is commonly affected by air embolism because of its more superior location in the ascending aorta, resulting in a hypocontractile dilated right ventricle and ST-segment changes in the inferior electrocardiographic leads. After separation from CPB, a detailed examination of the transplanted heart should include the elements listed in Table 17–2.

The function of the newly transplanted heart depends on many factors: baseline function before brain death, degree of myocyte damage before and during harvesting, amount of donor inotropic support, ischemic time, myocardial protection during the ischemic interval, reperfusion injury, cardiac denervation, donor-recipient size mismatch, and degree of pulmonary hypertension in the recipient. To accurately assess cardiac allograft anatomy and physiology, the echocardiographer needs to understand the surgical procedure and appreciate the changes that normally occur in the transplanted heart.

The standard or biatrial technique, originally described by Lower and Shumway, was the primary method for nearly 30 years.10 However, more transplantation centers are now using the bicaval anastomotic technique as the method of choice, except in infants and small children. The advantages of the bicaval technique include preserved geometry and function of the atria, improved CO, and less disruption in the geometry of the atrioventricular valves, resulting in reduced valvular regurgitation, fewer conduction abnormalities, less thrombus formation in the left atrium, and decreased perioperative mortality.11 In the standard technique, most of the native atrial walls and the interatrial septum are left in situ, leaving the inferior vena cava (IVC), superior vena cava, and pulmonary venous inflow tracts undisturbed. In the donor heart, an LA cuff is created by incising through the pulmonary vein orifices, whereas the right atrial (RA) cuff is created by incising through the inferior vena caval orifice and extending the incision up toward the base of the RA appendage. When the bicaval technique is performed, most of the native atrial tissue is excised, thereby creating superior vena cava and IVC cuffs for end-to-end anastomoses with the donor vena cavae. Divisions and end-to-end anastomoses of the great vessels are the same for both techniques.

Intraoperative TEE assessment of allograft LV systolic function early after separation from CPB has been shown to better predict early requirements for inotropic and mechanical support than routinely measured hemodynamic variables, particularly when ischemic times are prolonged.12 In general, allograft LV systolic function after CPB is expected to be normal, and impaired LV systolic function at this stage, usually the result of ischemic injury or early acute rejection, is often transient. It is important to document any intraoperative regional WMAs because coronary atherosclerosis and myocardial infarction, often silent, are major causes of morbidity and mortality after heart transplant surgery.

There are several echocardiographic findings that could be considered abnormal in the general population but are characteristic in the allograft left ventricle. These are listed in Table 17–3. Increases in LV wall thickness and LV mass are thought to represent myocardial edema resulting from manipulation and transport of the heart. Because the donor heart is typically smaller than the original dilated failing heart, it tends to be positioned more medially in the mediastinum and tends to be rotated clockwise. This could result in difficulties in obtaining the standard tomographic planes, and nonstandard TEE probe positions and angles may have to be used.

Diastolic compliance is often decreased in the first few days or weeks after cardiac transplant, but typically improves in the first year.4 This is most likely the result of ischemia or reperfusion injury, a smaller donor heart in a larger recipient, or a larger heart implanted into a restricted pericardial space. Unfortunately, Doppler echocardiographic assessment of LV diastolic function is complicated by a variety of factors, outlined in Table 17–4. When remnant atrial tissues retain mechanical activity, atrial contractions become asynchronous, resulting in beat-to-beat variations in transmitral inflow velocities. Atrial dysfunction can also result in abnormal transmitral and pulmonary venous flow patterns.13 LV diastolic dysfunction therefore is not the sole cause of altered transmitral flow patterns, and atrial dysfunction has to be ruled out. The echocardiographic indicators of atrial dysfunction include a decreased ratio of systolic to diastolic maximum pulmonary venous flow velocity in the presence of normal pulmonary capillary wedge pressures, reduced LA area change, and reduced mitral annulus motion.13

The thin-wall right ventricle is particularly susceptible to injury during the period of ischemia and reperfusion and also compensates poorly for any increase in pulmonary vascular resistance, which often is elevated in patients with end-stage heart failure. Therefore, it is not surprising that acute RV failure is more common than LV failure and accounts for 50% of all cardiac complications and 19% of all early deaths after heart transplantation.14 Once the diagnosis of RV dysfunction is established, stenosis at the PA anastomosis or kinking of the PA should first be ruled out. A systolic gradient higher than 10 mm Hg may indicate the need for surgical revision. TEE should then be used to optimize RV filling to avoid overdistention of the ventricle and to assess the response to inotropic support. In the setting of maximum inotropic support and pulmonary vasodilator therapy, the presence of a small hyperdynamic left ventricle with a dilated right ventricle (Figure 17–1), especially when accompanied by marginal urine output, arrhythmias, or coagulopathy, should prompt the consideration of the implantation of an RV assist device.

The size and geometry of the atria and the atrial anastomoses depend entirely on the transplantation technique employed. In the standard biatrial technique, different-sized portions of the native atria are left in situ (Figure 17–2), resulting in biatrial enlargement, asynchronous contraction, and intraluminal protrusion of the atrial anastomoses. This method also often gives the atria a multichamber configuration on the TEE (Figure 17–3). The anastomotic protrusions appear echo-dense and should not be confused with thrombi, although thrombi may form along the suture line. These protrusions may also occasionally contact the posterior mitral leaflet in systole, or even result in a mild constriction with a step-up of intraatrial Doppler flow velocities. Severe cases of supra-mitral valve obstruction, or acquired cor triatriatum, have been described after heart transplantation and should be suspected intraoperatively when the LA remnant is markedly enlarged and LV volume is reduced. Turbulent flow by color-flow Doppler (CFD), fluttering of the mitral valve leaflets, and elevated blood flow velocities by pulsed-wave Doppler also may aid in the confirmation of the diagnosis.

The integrity of the interatrial septum should be assessed intraoperatively by using color-flow Doppler and contrast echocardiography (agitated saline or saline microcavitation). Shunts can occur at the atrial anastomotic site or through a patent foramen ovale (PFO). Although uncommon, shunting through a PFO that is not apparent preoperatively may become hemodynamically significant postoperatively. As the relative pressure difference between the left and right atria changes as a result of pulmonary hypertension, RV dysfunction, or tricuspid regurgitation (TR), right-to-left shunting can occur and present as refractory postoperative hypoxemia.15 Identification of a left-to-right shunt across the interatrial anastomoses also should prompt surgical repair because it can contribute to progressive RV volume overload and TR.

Spontaneous echo contrast can be detected in up to 55% of heart transplant recipients. This is usually confined to the donor atrial component and is associated with thrombi, usually attached to the LA free wall underneath the protruding suture line. The incidence of thrombus formation in the left atrium is reduced with the bicaval anastomosis technique.

The PA anastomosis should be examined for possible stenosis, and, although rare, kinking or torsion of the donor or recipient pulmonary artery should be ruled out, especially in the setting of RV dysfunction.16 Color-flow Doppler may detect turbulent flow, and the pressure gradient should be measured with continuous-flow Doppler. Pulmonary venous inflow also should be assessed with color-flow and pulsed-wave Doppler.

Mild to moderate degrees of TR and mitral regurgitation (MR) are common after heart transplantation. MR is usually mild, produces an eccentric jet toward the LA free wall, and has a reported incidence of 48% to 87%.16,17 TR, the most common valvular abnormality after heart transplantation with a reported incidence of 85%, is usually mild with an eccentric jet directed toward the interatrial septum.18 TR after heart transplantation is best quantified by using the ratio of the maximum area of the regurgitant jet to the RA area.19 The etiology of atrioventricular valve regurgitation in the transplanted heart is thought to be related to distortion of annular geometry. Annular distortion after the standard biatrial anastomotic technique is predominantly the result of disturbed atrial geometry and function, whereas donor heart and recipient pericardial cavity size mismatch is thought to play an important role after the bicaval anastomotic technique. This hypothesis is supported by the fact that the incidence and severity of TR and MR are reduced after the bicaval technique as compared with the standard biatrial technique.

The natural history of these regurgitant lesions varies, but the incidence of severe TR appears to increase with time, and some patients may require tricuspid valve repair or replacement for refractory symptoms. However, in many of these patients, the subvalvular apparatus was damaged during subsequent endomyocardial biopsy.19 When patients were examined 1 year after transplantation, those with significant TR were more symptomatic and had poorer right-side heart function and greater mortality than those with mild or no TR.20,21

Management of Early Postoperative Hemodynamic Abnormalities in the Intensive Care Unit

TEE has become an invaluable tool in the management of seriously ill intensive care patients in whom transthoracic acoustic images may be poor. Particular uses in these circumstances include assessment of biventricular function, anastomotic problems (kinks, torsion, or stenosis), valvular abnormalities, sources of systemic emboli, and the detection of pericardial tamponade.

Postoperative Follow-Up Studies of Cardiac Allograft Function

Echocardiography, a noninvasive means of diagnosing transplant rejection, plays a significant role in the follow-up of recipients after heart transplantation. Proposed echocardiographic indicators of rejection in heart transplant patients are listed in Table 17–5. In addition, two- or three-dimensional echocardiography may be used to guide transvenous endomyocardial biopsies to prevent inadvertent damage to the tricuspid valve and its supporting apparatus. Dobutamine stress echocardiography, used in the detection of allograft vasculopathy, also has been shown to have a high negative predictive value for determining future cardiac events and death in heart transplant recipients.22

Mechanical circulatory support devices include intra-aortic balloon pumps and ventricular assist devices (VADs) that may be inserted for supporting the failing left and/or right ventricle.

Intra-aortic Balloon Pumps

Intra-aortic balloon pumps (IABPs) are placed perioperatively in 2% to 12% of cardiac surgical patients, with the majority being placed intraoperatively.5 When IABPs are placed, TEE can be useful in determining the need for the IABP, assessing for contraindications such as aortic insufficiency or severe aortic atherosclerosis, and guiding its placement into the descending aorta. TEE can also rapidly assess the effects of counterpulsation upon LV function and determine if there were any complications such as aortic dissection or aortic valve perforation. Inappropriate placement is the most common complication, and inadvertent passage of the IABP into the aortic arch, left ventricle, subclavian artery, renal artery, contralateral femoral artery, and right atrium have all been reported.23,24

Assessment of IABP placement begins with visualization of the guidewire within the lumen of the descending aorta. This is particularly important in the setting of aortic dissection, when identification of the true aortic lumen may be challenging. Optimal placement of the IABP tip is 3 to 4 cm distal to the origin of the left subclavian artery, or when the tip is seen at the inferior border of the transverse aortic arch.25 To confirm proper placement, the balloon is first identified in the descending aorta short-axis view. Proper placement has been defined by the disappearance of the tip of the IABP from the aortic arch in the upper esophageal aortic arch long-axis view. Placement below the subclavian artery can also be visualized in a descending aorta long-axis view by slowly withdrawing the probe until the subclavian artery is seen at the level of the aortic arch (which is now seen in cross-section). The common carotid artery is sometimes mistaken for the subclavian artery but can be differentiated by its larger diameter and by turning the probe to the left (to visualize subclavian) and then to the right (to visualize the common carotid). The balloon itself typically appears as an echo-dense image when deflated (Figure 17–4) and a scattered echo image when inflated. A side lobe artifact is commonly seen when the tip of the IABP is visualized in the short-axis view.

Left Ventricular Assist Devices

Transesophageal echocardiography plays a critical role in each step of the management of patients with left ventricular assist devices (LVADs), including the pre-placement evaluation of cardiac structure and function, detection of interatrial shunts, determination of aortic and tricuspid valve pathology, separation from CPB, and assessment of device function in the postoperative period.

Pre-Procedure assessment

A pre-procedure TEE is typically performed in the operating room following induction of general anesthetic and prior to institution of CPB. Determination of the patency of the foramen ovale, aortic valve insufficiency, mitral valve stenosis, tricuspid regurgitation, left heart thrombus, and assessment of right ventricular function are critical to intraoperative planning and management.

Patent Foramen Ovale. While it is important to recognize an atrial or ventricular septal defect, the more common cause of an intracardiac shunt is a patent foramen ovale (PFO). Intracardiac shunts are important to diagnose and repair to reduce the risk of paradoxical embolism or hypoxemia following LVAD placement. An appropriately functioning LVAD will significantly reduce LV diastolic pressures (often to <5 to 10 mm Hg) but right heart filling pressures can remain abnormally elevated, resulting in a right-to-left shunt and hypoxemia. Even small PFOs should be surgically repaired because of the significant incidence of shunting seen in patients with LVADs.

The normal foramen ovale is best seen in a midesophageal (ME) bicaval view; it appears as a thin slice of tissue bound by thicker ridges of tissue, one of which appears as a “flap.” TEE evaluation of the foramen ovale should include two-dimensional (2D) assessment for flap movement and color-flow Doppler assessment, optimized for measurement of lower-velocity flow. Injection of agitated saline (a “bubble study”) along with a Valsalva maneuver is typically used to provoke right-to-left shunting.26 In such a study, the bubbles should be injected after the Valsalva maneuver produces a decrease in RA volume, and the Valsalva should be released (so as to transiently increase RA pressure over LA pressure) when the microbubbles are first seen to enter the RA. Bowing of the septum to the left upon release of Valsalva confirms the transient increase in right atrial pressures. Admixture of agitated saline with small quantities of blood has been reported to improve the acoustic signal of the microbubbles. The bubble study is positive if bubbles appear in the left atrium within five cardiac cycles (Figure 17–5). In patients with severe LV failure, it may be difficult to sufficiently decrease left atrial pressure. In such cases, an alternative method involves partial obstruction of the pulmonary artery by the surgeon after the aortic cannula is placed.27

Aortic Pathology. The LVAD outflow cannula is typically placed in the ascending aorta (except for the Jarvik 2000, which may be attached to the descending aorta). Thus, a thorough examination of the ascending aorta is an essential component of the intraoperative TEE evaluation. The ascending aorta is optimally viewed in the midesophageal ascending aortic short- and long-axis views. An ascending aortic aneurysm may require repair prior to LVAD placement.26 Protruding atheroma or mobile atheroma increase stroke risk, and their presence must be communicated to the surgeon. These plaques can be difficult to palpate; therefore, an epiaortic image at the site of cannulation for CBP and for the outflow cannula may assist the surgeon with precise placement.28

Aortic Valve Insufficiency. Significant aortic valve regurgitation (AR) results in chronic volume overload of the LV with consequent ventricular dilation and dysfunction. Reduction of the transaortic (valve) pressure gradient secondary to elevated LV end-diastolic pressure and reduced aortic diastolic pressure may confound determination of AR severity and lead to underestimation in a heart failure population.26,29 In LVAD patients with AR, LV volume loading may be more pronounced as blood being returned from the LVAD is delivered to the aorta just above the aortic valve and regurgitant volume is increased because the LV end-diastolic pressure is low relative to the aortic pressure. If the resultant regurgitant volume exceeds 1 to 2 L/min, patients may remain in clinical heart failure despite the LVAD, since this volume is not delivered systemically but remains within a circuit formed by the LV, LVAD, and the ascending aorta. Older, volume displacement LVADs typically eject blood each time the device is full; therefore, AR in these patients increases the pump rate.29

Aortic insufficiency is best assessed in the midesophageal aortic valve short- and long-axis views as discussed in Chapter 9. It has been suggested that patients with worse than mild AR should undergo a concomitant aortic valve repair or replacement.30,31 However, the decision to correct AR is complex since the addition of a valve procedure significantly increases procedural mortality.32 Important considerations include the degree of aortic valve calcification and the characteristics of the regurgitant jet. An eccentric regurgitant jet in a heavily calcified valve may be more likely to worsen with VAD support and usually warrants surgical correction. Another consideration relates to the planned duration of LVAD support. If the LVAD is being used to bridge the patient to transplant and a relatively short period of support is anticipated, then moderate AR may be tolerated, anticipating that LVAD speeds/rates may be higher than normal. On the other hand, if the device is being used as a permanent or “destination” treatment, such aortic regurgitation is likely to progress and may impact the durability of the device.

There are several methods of surgically addressing AR. One option is replacement of the aortic valve. Mechanical valves are not typically used because of the potential for thrombus formation on the valve as a consequence of the immobility of the leaflets during most LVAD cycles. Furthermore, intermittent opening of the aortic valve renders the patient at risk for embolization.29,30 Thus, if the valve requires replacement, most surgeons recommended the use of a bioprosthesis.32 Another alternative to managing AR in the LVAD patient is partial or complete surgical ligation of the aortic valve cusps. This should not be performed if there is a chance of ventricular recovery with subsequent removal of the device.29,30 A third option in patients without the possibility for native heart recovery is placement of an occlusive LV outflow tract patch graft. In this situation, all blood must be delivered from the LV to the LVAD, and pump failure may result in severe hemodynamic instability as the native heart would be required to eject through the LVAD.30,32

Mitral Valve Stenosis. A significant mitral gradient will lead to impairment of LVAD filling, persistent elevation of pulmonary venous pressure, and symptoms of heart failure. While rheumatic mitral stenosis (MS) is rare in this group of patients, previous procedures such as mitral valve repair or replacement are common. TEE should evaluate for the presence of severe MS across repaired mitral valves or prosthetic valves. It is recommended that severe MS be surgically repaired, with a commissurotomy or concomitant replacement of the mitral valve.26,29 Mitral valve stenosis is optimally assessed by TEE in the midesophageal four-chamber view using CFD and spectral Doppler to determine peak and mean transvalvular gradients as described in Chapter 7.

Tricuspid Regurgitation. Careful evaluation of tricuspid regurgitation (TR) is also warranted. Severe TR with hepatic vein flow reversal usually warrants concomitant tricuspid repair, as elimination of severe TR may improve right ventricular function and device filling following LVAD placement. The tricuspid valve is optimally viewed in the midesophageal four-chamber and the midesophageal right ventricular inflow-outflow views (see Chapter 10).

Left Heart Thrombus. Abnormal blood flow patterns in the left atrium and ventricle predispose to thrombus formation. Common sites of left heart thrombus include the left atrial appendage (Figure 17–6) and the left ventricular apex.31,33 In an attempt to reduce the risk for embolization, TEE should be utilized to rule out the presence of LV apical thrombus prior to the ventriculotomy for placement of the LVAD inflow cannula. Epicardial scanning may be helpful when the apex cannot be visualized with TEE.

Right Ventricular Dysfunction. An LVAD only supports the left heart and is dependent on a functional right ventricle (RV) to provide adequate preload. While the LVAD may enhance RV performance by decreasing its afterload, it may also worsen RV function by increasing its preload.26 When evaluating the RV, it may be helpful to determine the RV fractional area change (RVFAC), defined as:

A normal RVFAC is greater than 40%, while most patients receiving an LVAD have an RVFAC of 20% to 30%.31 An RVFAC less than 20% predicts a high risk for RV failure following LVAD placement.31 Right ventricular dysfunction remains one of the important clinical challenges in left-side mechanical circulatory support. Combinations of inotropic agents, systemic and inhaled vasodilators, and mechanical RV support may be needed to ensure proper function of the LVAD.

Separation From Cardiopulmonary Bypass

The TEE exam must be repeated during separation from CPB initially to assist in the de-airing process. As the pressure gradients within the heart change dramatically with a functional LVAD, it is also important to repeat the assessment for a PFO, AR, and RV dysfunction. Aortic valve opening, placement and orientation of the inflow cannula, flow in the inflow cannula and outflow graft, and assessment of LV size and ventricular septal position are critical in the intraoperative management of these patients.

De-airing. Following open heart surgery, ambient air (Figure 17–7) can be retained in multiple locations of the heart including the right and left upper pulmonary veins, the LV apex, the left atrial appendage, the right coronary sinus of Valsalva, and the pulmonary artery.26,34 In addition, air can be retained in the VAD cannulas and the pump itself.26 The de-airing process is more complicated for this procedure compared to valvular heart procedures. Most LVAD designs are able to generate negative intraventricular pressure or suction, which can lead to entrainment of extracardiac air. This is most commonly seen when device rate or speed is inappropriately increased during a time when the delivery of blood into the LV is reduced. Thus, the complex de-airing process for LVADs includes removal of intracardiac air as well as vigilance to avoid entrainment and reintroduction. A potential negative impact is when entrained air is delivered into the right coronary artery with subsequent RV dysfunction, reduced LV filling, and further entrainment of air by the pump. In this scenario, the TEE will demonstrate a distended RV, a collapsed LV, and significant air in the aorta. Preserving RV function while weaning from CPB so as to maintain LV preload during the period of reduced LVAD flows and until protamine reversal is therefore highly desirable. TEE examination for air should be conducted continually from before initiation of CPB weaning until after protamine reversal.

Patent Foramen Ovale. Although it is optimal to diagnose a PFO prior to CPB, LVAD-induced reductions in LV end-diastolic pressure and left atrial pressure increase the likelihood of right-to-left shunting. Discovery of a previously unrecognized PFO following CPB has been described and may require reinstitution of CPB for repair if there is significant shunt flow.35,36

Aortic Valve. Ideally the severity of AR should be determined preoperatively to allow for correction during LVAD placement. However, the increased transaortic (valve) gradient associated with reduction of the LV end-diastolic pressure and increased flow into the ascending aorta through the outflow graft after CPB is discontinued may worsen preexistent AR. If worse than mild AR is identified, the aortic valve may need to be surgically corrected.37

In addition to examining the aortic valve for severity of AR, the frequency of aortic valve opening should be determined. A functioning LVAD is capable of reducing LV end-diastolic pressure to a level at which the aortic valve does not open during a normal cardiac cycle. However, if the LVAD is only providing partial or variable support, the aortic valve will open intermittently.26

Right Ventricular Dysfunction. As the LVAD provides a normal cardiac output, a commensurate amount of blood is returning to the right heart as preload. Patients with RV dysfunction may be unable to accommodate for this change, and signs of right heart failure may develop including RV distension, acute severe tricuspid regurgitation, increase in pulmonary pressures, and LV failure secondary to a low preload.31 Another cause of RV dysfunction after LVAD placement is based on the concept of ventricular interdependence. Rapid reductions in LV end-diastolic pressure may result in movement of the ventricular septum toward the LV free wall. Functionally this causes abrupt alternations in RV size and geometry and can influence the severity of tricuspid regurgitation. If identified, the most effective short-term treatment is to reduce the LVAD flow, which subsequently increases the LV end-diastolic pressure and returns the septum to a more normal anatomic position.

Inflow Cannula. The inflow cannula is usually placed in the LV apex and is often directed anteroseptally and toward the mitral valve opening but away from the interventricular septum and lateral wall. It should not abut any of the LV walls in order to avoid obstruction of blood flow into the cannula.31 If the cannula is misdirected, withdrawal and inferior displacement by the surgeon generally rectifies the situation. Proper inflow cannula placement should be evaluated in at least two views: the midesophageal four-chamber view and the midesophageal long-axis view. A CFD sector should be placed across the opening of the inflow cannula and should demonstrate low-velocity, unidirectional, laminar (nonturbulent) flow (Figure 17–8). In addition, unobstructed flow should be demonstrated using continuous-wave Doppler from the inflow cannula with peak velocities less than 2.5 m/s (Figure 17–9).31 The cannula position should be assessed again after chest closure to ensure that it remains correctly positioned.

Outflow Cannula. The outflow cannula of most devices is placed in the ascending aorta. This cannula may be seen in the midesophageal ascending aorta short-axis (Figure 17–10) or long-axis views. In order to assess the blood flow at the cannula anastomotic site, pulsed- or continuous-wave Doppler can be used. The peak velocity should be 1.0 to 2.0 m/s for an axial device and around 2 m/s for a pulsatile device.26

Left Ventricle. After protamine administration, the LVAD speed can be safely increased and TEE should confirm LV unloading. A properly functioning LVAD should reduce the LV diameter, and the interventricular septum should remain in a neutral position. Persistent deviation of the septum to the right suggests inadequate reduction of left ventricular pressure, and an increased pump speed is warranted. Septal displacement towards the LV cavity is indicative of excessive LV unloading and may have adverse implications for RV function as discussed above. Detection of specific wall motion abnormalities and determination of ejection fraction are unreliable with a functional LVAD as preload reduction precludes normal contractility.26

Postoperative Period

TEE can be used to assess patients with reduced LVAD flows or function. The differential diagnosis includes right ventricular failure, pulmonary embolus, cardiac tamponade, hypovolemia, cannula obstruction or malposition, and device failure.31,38

Right Ventricular Failure or Pulmonary Embolus. RV failure and pulmonary embolus present with a similar echocardiographic picture. As mentioned previously, RV failure limits the preload for the LVAD, causing a low output state. The typical findings on TEE include a dilated and dysfunctional RV, severe tricuspid regurgitation, and an underfilled left ventricle. Patients with a pulmonary embolus tend to have elevated pulmonary artery pressures, whereas those with isolated RV failure may actually have low pulmonary artery pressures. If there is concern of a pulmonary embolus, the pulmonary arteries should also be examined.31

Cardiac Tamponade. Tamponade can be difficult to diagnose after LVAD placement, as fluid collections may be loculated, making them difficult to identify. For example, the typical findings of right heart compression may be absent if the fluid collection is located posteriorly and compresses the left atrium. Normal LVAD physiology, which reduces left heart filling pressures, also confounds this assessment.31

Hypovolemia. Clinically, hypovolemia will present with systemic hypotension, reduced jugular venous pressure or central venous pressure, and reduced LVAD output. Although this diagnosis does not typically require imaging, TEE may be a useful adjunct, demonstrating small RV and LV dimensions and ruling out other possible etiologies for the clinical picture, including RV failure and tamponade.

Inflow Cannula. Inflow cannula obstruction has a severely detrimental impact on VAD function and can be caused by a variety of pathological processes (Table 17–6).26,39 The cannula should again be examined in at least two views, typically the midesophageal four-chamber view and the midesophageal long-axis view.31 Color-flow Doppler across the cannula inlet demonstrating turbulent flow during LVAD diastole, and continuous-wave Doppler demonstrating a peak velocity greater than 2.5 m/s are suggestive of obstruction.26,39–41 Three-dimensional (3D) TEE images have also been used to visualize thrombus in the inflow cannula (Figure 17–11).

The pulsatile LVADs contain valved conduits that direct flow through the device and prevent regurgitation in a manner similar to the native heart. Inflow valve regurgitation (IVR) is a common cause of LVAD mechanical dysfunction and is usually due to a torn cusp or commissural dehiscence of the prosthetic valve secondary to high pressures.37,41 The patient with significant IVR presents with clinical evidence of heart failure and elevated pump rates that result from increased device filling. Echocardiography will demonstrate ineffective LV unloading characterized by increased LV dimensions and aortic valve opening in addition to decreased outflow graft velocity and a decreased stroke volume.37,41

Outflow Cannula. Documented complications of the outflow cannulae include perforation and malposition. Air bubbles in the aorta near the outflow cannula anastamosis may suggest cannula perforation.26 An extreme case of cannula misplacement was reported in a hemodynamically unstable patient in whom the LV was dilated and the AV was opening with LVAD systole. The outflow cannula was not visualized on TEE and upon surgical exploration, was found in the right superior pulmonary vein.40

Mechanical assist devices can be used for right ventricular support either in isolation or in combination with a left-sided device. The etiology of right heart failure after cardiac surgery includes prolonged cardiopulmonary bypass time, inadequate myocardial protection, or right coronary occlusion from vasospasm, air embolus, or thrombus.42 Isolated right heart failure is rare, occurring in only about 0.3% of cardiac surgical patients.43 However, it is associated with a very poor prognosis.42,43

Most right ventricular assist devices (RVADs) are placed after a failed attempt to separate from CPB or within the same day as LVAD placement.44 The RVAD inflow cannula is typically placed in the right atrium (but can occasionally be placed in the right ventricle).26 This cannula should be visualized using either the midesophageal four-chamber view or the midesophageal bicaval view. The RVAD outflow cannula is attached to the main pulmonary artery. This cannula can be difficult to image by TEE, but may be seen in the midesophageal right ventricle inflow-outflow view (Figure 17–12). As with the LVAD cannula, flow should be laminar and of low velocity when examined with CFD.

The current surge in the heart failure patient population will produce a concomitant increase in the number of VAD insertions and heart transplants performed worldwide. The option of VAD support is particularly attractive in the face of a shortage of donor organs. As a consequence of the significant technological advances in mechanical cardiac support devices seen in the last decade, patients will benefit from a reduced complication rate and improved survival. Intraoperative TEE imaging is a critical part of successful device placement and cardiac transplantation.