Intraoperative Echocardiography

	INTRODUCTION

Top Next References

View larger version (46K): [in this window] [in a new window]   FIGURE 10-1 The 20 multiplane views of a comprehensive intraoperative multiplane transesophageal examination. (Reproduced with permission from Shanewise JS, Cheung AT, Aronson S, et al: ASE/SCA Guidelines for Performing a Comprehensive Intraoperative Multiplane Transesophageal Echocardiography Examination: Recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999; 89:870.)

	VALVULAR DISEASE

Top Previous Next References

Over the last several years, the impact of TEE's influence on perioperative cardiac surgical decision making has become increasingly more appreciated. TEE, however, has been particularly useful in the perioperative evaluation of mitral valve (MV) disease. Early studies commented on the use of MV scanning on the field by the surgeon⁸; however, because of the MV's orientation in relation to the esophagus, excellent intraoperative echocardiographic windows for two-dimensional (2-D) and both qualitative and quantitative Doppler evaluation can easily be obtained. Accordingly, TEE has supplanted surgeon-guided epicardial scanning for evaluation of the MV.⁹

A thorough TEE examination of the MV requires the utilization of all echocardiographic modalities. In the initial pre–cardiopulmonary bypass (CPB) period, 2-D echocardiography defines the extent and etiology of MV pathology. The severity and location of MV disease, including subvalvular involvement, annular calcification and dilation, bileaflet prolapse, and anterior mitral leaflet (AML) disease, along with the degree of left ventricular (LV) function, should be assessed. This information, when integrated and evaluated, can be helpful in determining whether MV replacement (MVR), repair, or neither is indicated. It is also important to diagnose if existing severe mitral regurgitation (MR) is primarily associated with ischemia as opposed to rheumatic or myxomatous disease, since the extent of surgical treatment for ischemic MR remains controversial. Although the 2-D anatomic images and the color flow Doppler (CFD) modalities are used most often during interrogation of the MV in the operating suite, a complete echocardiographic evaluation using pulse wave (PWD) and continuous wave Doppler (CWD) can assist in determining the severity of blood flow disturbances associated not only with MR, but also with mitral stenosis (MS).¹

Clinical studies have suggested that the pre–CPB TEE exam prompts changes in surgery in 9% to 14%^8–11 of patients undergoing MV surgery. The population of patients influenced by the pre–CPB TEE exam includes those initially scheduled only for coronary bypass grafting (CABG) who also require an MV procedure. In addition, the pre–CPB exam also influences surgical decision-making in those patients scheduled for an MV procedure based upon the preoperative transthoracic echocardiographic (TTE) or TEE exam, who actually only require a CABG due to "improvement" in the severity of MV disease as determined by the pre–CPB TEE examination in the operating room. The influence of changes in hemodynamic loading conditions during general anesthesia on transmitral flow and pressure, the potential changes in the patient's overall medical condition since the time of the preoperative echocardiographic exam, the technical limitations of TTE in comparison to TEE, and the general subjectivity of the assessment among different echocardiographers must all be taken into consideration when making the final surgical decision regarding the most appropriate MV procedure.

The initial post–CPB TEE exam is essential in helping to determine the competency of the replaced or repaired MV, and to evaluate persistent MR and/or systolic anterior motion (SAM). It is especially important for the echocardiographer to fully understand the normal flow patterns and acceptable transvalvular pressure gradients following various types of MV repairs and replacements. Almost all MV replacements will exhibit physiologic, regurgitant transvalvular flow patterns that are acceptable and anticipated, although they vary significantly depending upon the specific prosthetic (Fig. 10-2). Most small perivalvular jets detected by CFD are insignificant, although larger "leaks" may become associated with hemolysis, hemodynamic instability, or valvular dehiscence. Hemodynamic loading conditions and LV function must also be taken into consideration in the assessment of residual MR. Following MV procedures, several studies have suggested that, in approximately 5% to 11% of cases,^3,8–11 the post–CPB TEE exam may identify persistent lesions that require additional, immediate surgical intervention.

View larger version (53K): [in this window] [in a new window]   FIGURE 10-2 Normal prosthetic mitral valve regurgitation. All prosthetic valves demonstrate some degree of transvalvular regurgitation. In this color flow Doppler examination of a St. Jude mitral valve prosthesis (St. Jude Medical Inc., St. Paul, MN), two jets of mitral regurgitation are demonstrated (arrows). These regurgitant jets are normal for this valve and originate from the hinge points of the dual tilting disks. The valve sewing ring prevents ultrasound propagation, thus producing "drop out" beyond the ring (asterisks). Left ventricular cavity visualization is also obscured by the valve disks. LA = left atrium; RA = right atrium; RV = right ventricle.

In an early study of the utility of epicardial scanning in 100 patients undergoing MV repair, Stewart et al⁸ noted that 6 patients required an immediate return to CPB for further MV evaluation and repair. Three of these patients had SAM, 2 patients had residual 2+ or higher MR, and 1 patient had a flail leaflet. In this study, 2 patients with 3+ MR were not reoperated on immediately, but required a second procedure within 5 days due to hemodynamic compromise.

Another series employing TEE reported outcomes in 586 cardiac valve procedures⁹ included 92 patients undergoing MV procedures. In these 92 MV patients, the pre–CPB TEE impacted surgical decision-making in a "minor" way in 14% of the cases. Minor changes included discovering an unknown LV thrombus, repairing an unknown ruptured chord, and identification of a valvular vegetation. Pre–CPB TEE had a "major" impact on an additional 11% of the mitral patients in this series including discovery of unknown lesions on the aortic (AV), tricuspid (TV), or MV (n = 7), existence of an ASD (n = 2), or unknown asymmetric septal hypertrophy (n = 1). The post–CPB TEE in 10 mitral patients (11%) determined the need for further MV surgery because the repairs were not judged to be adequate.

In a large series of 3245 adult patients undergoing a wide variety of cardiac surgical procedures, new information resulting in a modification of the surgical procedure was identified during the pre–CPB intraoperative TEE examination in 12% of 1265 patients undergoing MV surgery. Of note, 96 patients scheduled for MV surgery and a concomitant procedure were found to have a "normal" valve, and consequently did not require an MV procedure. Post–CPB, 6% of the MV surgical patients were noted to have a "new echocardiographic finding."¹⁰ A smaller study, which prospectively followed 203 cardiac surgical patients who underwent intraoperative TEE, found a similar incidence of 17% unknown pre–CPB findings in 8 of 46 patients requiring MV procedures, and a return to CPB in 4.7% (2 of 46) mitral surgical patients.¹¹ Another large series, reporting on 6340 patients who had intraoperative echocardiographic examinations over a 10-year period, noted that return to CPB was required in 7% of 2226 MV repair procedures.³ Despite the experience in MV repairs in this surgical group, the incidence of returning to bypass because of an unsatisfactory initial repair did not decrease over the 10 years, suggesting that more difficult repairs were being attempted.

To date, there are no prospective, large-scale, randomized trials that have specifically identified a consistent, independent advantage for intraoperative TEE. The wide-scale acceptance of intraoperative TEE during MV procedures has made it difficult if not impossible to conduct such a study. Data from several clinical investigations, however, have implicated an important, clinically significant and cost-effective role for TEE as a safe and valuable hemodynamic monitor in identifying high-risk patients, in assisting in the determination of the definitive surgical approach, and in providing a timely post–CPB evaluation of the procedure, thereby allowing for immediate intervention or at least the opportunity to triage patients appropriately. In the near future, the role of intraoperative TEE in patients undergoing MV procedures will most certainly become even more important with the development of three-dimensional (3-D) TEE and the further introduction of newer, minimally invasive approaches in which direct surgical visualization and inspection may be more limited.

PERSISTENT MITRAL REGURGITATION FOLLOWING MITRAL VALVE REPAIR

It is clear from the studies cited that intraoperative echocardiography can provide the surgeon with an immediate evaluation of MV function after repair. Moderate to severe residual MR (Fig. 10-3) will usually necessitate a return to CPB for further evaluation and definitive surgery. One main question arises regarding MV repairs and the utility of intraoperative echo: What is the outcome of patients who have trace to mild residual MR after repair?

View larger version (72K): [in this window] [in a new window]   FIGURE 10-3 Moderate mitral regurgitation is demonstrated in this color flow Doppler image following mitral valve repair. The regurgitation jet extends to the back of the left atrium (LA) and is broad-based. The arrows indicate the borders of the ring placed during the repair. LV = left ventricle.

SYSTOLIC ANTERIOR MOTION FOLLOWING MITRAL VALVE REPAIR

Following MV repair, the presence of left ventricular outflow tract (LVOT)obstruction has typically been associated with SAM. Risk factors for SAM include preoperative evidence of excessive or redundant posterior (PML) and/or AML tissue (Fig. 10-4), nondilated LV, and a narrow mitroaortic angle. Postoperative risk factors for SAM include persistent AML redundancy, excessive PML resection, the selection of an annuloplasty ring that is either too small or incorrectly oriented, and an hypertrophied/hyperdynamic LV. All of those factors contribute to a Venturi effect that "pulls" the AML and coaptation point toward the ventricular septum during systole.

View larger version (68K): [in this window] [in a new window]   FIGURE 10-4 Elongated, redundant mitral valve leaflets are potential risk factors for systolic anterior motion following mitral valve repair. In the example shown, a redundant posterior mitral leaflet (PML) has prolapsed into the left atrium (LA), and the elongated anterior mitral leaflet (AML) extends deep into the left ventricle (LV) during systole. RA = right atrium; RV = right ventricle.

Another echocardiographic study comparing 8 patients who developed SAM¹⁴ with a control population who did not develop SAM following repair found that the AML was significantly longer than the PML (3.5 ± 0.40 cm vs. 2.8 ± 0.32 cm) and the ratio of the AML/PML to be almost identical. In this investigation, however, measurements of the mitral leaflet lengths were performed in mid-diastole, whereas in the previous study the mitral leaflet lengths were measured at the start of ventricular systole, making a comparison between the studies somewhat difficult.

Several studies have identified SAM as a contribution to the requirement for early and late MV reoperation.^15,16 Freeman et al¹⁶ reported on 143 MV procedures for MR associated with myxomatous degeneration. Over 9% of these patients demonstrated echocardiographic evidence of SAM following MV repair. Although all of these patients had PML prolapse, 7 also had AML prolapse. The majority underwent PML resection (n = 10), and 12 underwent ring annuloplasty (Duran = 10, Carpentier = 2). In this investigation, reinstitution of CPB was required in 5 patients due to persistent 3+ to 4+ MR, which resolved following the second procedure. In the remaining 8 patients, SAM and concurrent MR were associated with hypovolemia and a relative hyperdynamic heart. Resolution of hemodynamic instability and pathology was achieved with volume resuscitation and by reducing the administration of inotropic therapy. Marwick et al¹⁵ reported that 10 of 309 mitral patients were observed to have LVOT obstruction and all 10 required reinstitution of CPB for either removal of the MV ring (n = 5) or MVR (n = 5).

In the absence of an echocardiographic evaluation, the diagnosis of and treatment of SAM following MV repair can be challenging, since commonly employed post–CPB therapeutic interventions (such as the use of inotropic agents) may actually be relatively contraindicated and lead to further hemodynamic compromise. Thus, it is essential to include a comprehensive evaluation for the presence of SAM in the post–CPB TEE examination following MV repair. The echocardiographic evaluation of SAM following MV repair includes a 2-D assessment of the mitral apparatus (Fig. 10-5A), a qualitative and quantitative CFD analysis of the MR jet (Fig. 10-5B), and a quantification of the pressure gradient across the LVOT obstruction. The CFD jet is typically directed toward the left pulmonary vein and away from the atrial septum as the AML is pulled toward the LVOT. Obstruction in the LVOT can produce pressure gradients as high as 100 mm Hg. Typically the Doppler flow velocity profile associated with dynamic LVOT obstruction can be differentiated from the profile associated with aortic stenosis (AS) by its "dagger" shape and mid-systolic acceleration. It would seem reasonable to assume that in the presence of echocardiographic confirmation of SAM and a hypovolemic and/or hyperdynamic LV, a trial of volume resuscitation and reduction of intropic therapy as suggested by Freeman et al¹⁶ is initially warranted, followed by a reevaluation of any persistent MR and LVOT obstruction before returning to CPB. Maslow et al¹³ reported that SAM resolved completely in 4 patients with this treatment and, in another 4 patients, SAM was still present but without CFD evidence of LVOT obstruction or MR. The authors do not comment on whether the 3 remaining patients required further surgery.

View larger version (70K): [in this window] [in a new window]   FIGURE 10-5 Systolic anterior motion following mitral valve repair. (A) The anterior mitral leaflet (arrows) is hinged toward the left ventricular outflow tract (LVOT) and intraventricular septum (IVS), thereby obstructing flow during systole. The hinge point of the leaflet is indicated by the arrowhead. (B) Color flow Doppler demonstrates high-velocity flow in the left ventricular outflow tract (black arrow) and severe mitral regurgitation (white arrow) associated with systolic anterior motion. RA = right atrium; RV = right ventricle; LA = left atrium; LV = left ventricle.

By definition, ischemic MR presents in patients who have significant coronary artery disease with relatively normal mitral leaflet morphology. Mitral regurgitation associated with ischemia may be caused by annular dilatation in the presence of normal mitral leaflet motion or by restricted systolic leaflet motion with or without annular dilatation (Fig. 10-6).¹⁷ Echocardiography is an essential diagnostic technique for evaluating the mechanism of ischemic MR and quantifying valvular dysfunction.

View larger version (70K): [in this window] [in a new window]   FIGURE 10-6 Ischemic mitral regurgitation. (A) Mid-esophageal, four-chamber transesophageal echocardiographic view demonstrating impaired anterior and posterior mitral leaflet coaptation (arrowhead) and tethering of the posterior mitral leaflet in a patient with ischemic mitral regurgitation. (B) Color flow Doppler demonstrating central and slight posterolateral eccentricity of a severe ischemic mitral regurgitation jet. LA = left atrium; LVOT = left ventricular outflow tract; RA = right atrium; RV = right ventricle.

A retrospective review of 269 patients referred for surgery with the diagnosis of moderate ischemic MR (determined by catheterization, preoperative echocardiogram, or both) yielded 136 patients who underwent CABG only.¹⁷ The remaining 133 underwent an MV repair along with the CABG and were consequently excluded from the analysis. In the group of 136 patients who underwent CABG alone, the authors compared the pre- and postoperative TTE evaluation with the intraoperative TEE assessment. Because these patients were scheduled only for CABG and intraoperative echocardiography was left to the discretion of the surgical and anesthesia team, intraoperative TEE was performed in only 38 of these patients. Similarly, only 68 patients underwent a postoperative TTE within 6 weeks of the hospitalization. By definition, all patients arrived in the operating room with 3+ MR, but the 38 patients who had intraoperative TEE had a decrease in average grade of MR to 1.4+. In addition, the postoperative TTE assessment of persistent MR revealed an increase in grade to an average of 2.2+. These results suggested that, under general anesthesia, intraoperative TEE may underestimate the degree of ischemic MR.

Bach et al¹⁹ studied MR jet area and diameter in 46 patients who underwent MR surgery and noted that in the 12 patients who had ischemic heart disease, both these measures decreased under general anesthesia. Jet area decreased from 10.0 ± 3.8 cm² to 5.7 ± 3.5 cm² and jet diameter decreased from 1.10 ± 0.29 cm to 0.79 ± 0.33 cm, lending further credence to a cautionary note on interpretation of intraoperative TEE results in patients with ischemic MR.

In patients with ischemic MR, intraoperative echo results should be interpreted with the following considerations in mind. First, changes in loading conditions have been shown to dramatically affect the degree of MR evaluated intraoperatively in the pre–CPB period.²⁰ Consequently, it is important to evaluate the degree of MR intraoperatively (while a patient is under general anesthesia and receiving positive pressure ventilation) only after hemodynamics have been optimized with volume infusions and/or vasconstriction to reflect preoperative values. Second, the assessment of MR jet size by CFD is a relatively subjective assessment. Consequently, interobserver variability and differences between TTE and TEE techniques must also be considered when there is a discrepancy between grades of MR severity. Thus, other factors including clinical signs and symptoms of congestive heart failure as well as further supportive evidence of a significant regurgitant lesion such as elevated pulmonary artery pressures and Doppler echocardiographic evidence of jet eccentricity, increased LA size, and abnormal pulmonary venous flow patterns should also be considered.²¹

PROSTHETIC MITRAL VALVE REPLACEMENT

All prosthetic valves have characteristic regurgitant CFD jets that must be recognized and distinguished from pathologic transvalvular and paravalvular lesions.²² For example, the St. Jude prosthetic valve (St. Jude Medical, Inc., St. Paul, MN) is manufactured to permit 3 to 5 small, transvalvular, wash-out jets along the leaflet hinge points, which serve to prevent thrombus formation. Pathologic transvalvular regurgitant jets are often associated with leaflets that are torn or stented open with clot, abscesses, or excessive subvalvular tissue.

Small paravalvular jets are not uncommon and frequently resolve spontaneously (Fig. 10-7A). Meloni et al studied 27 cardiac surgical patients undergoing MVR to assess regurgitant characteristics of MV biologic and mechanical prostheses immediately after implantation.²² Physiologic transvalvular regurgitant jets were observed in both biologic and mechanical prostheses. Trivial periprosthetic jets were observed in 14 patients (maximum jet area 0.46 cm²; range 0.1–1.5 cm²). Cardiopulmonary bypass was reinitiated in 2 patients with MV regurgitant jet areas of 3.6 cm² and 5.5 cm²; however, prosthetic dehiscence was discovered by direct inspection in only the latter case. Morehead et al also studied 27 cardiac surgical patients undergoing aortic valve replacement (AVR) (n = 11), MVR (n = 15), or AVR/MVR (n = 1), and hypothesized that although paravalvular jets visualized by intraoperative TEE immediately post–CPB are common, they tend to improve with restoration of normal coagulation parameters.²³ In this patient population, only 29 of the original 55 jets were still identified following protamine administration, and jet areas decreased from 2.0 cm² ± 2.2 cm² to 0.86 ± 1.7cm². None of the patients required immediate reexploration. Thus the results from these trials would suggest that most residual regurgitant jets visualized by intraoperative echocardiography following MVR are trivial. Alternatively, large paravalvular jets usually have to be repaired immediately to prevent excessive hemolysis, hemodynamic instability, and possible valve dehiscence (Fig. 10-7B).

View larger version (87K): [in this window] [in a new window]   FIGURE 10-7 Perivalvular leaks following mitral valve replacement. (A) Small perivalvular leaks (asterisk) visualized around the mitral valve sewing ring often close following heparin reversal. (B) More widely based perivalvular leaks usually require surgical intervention. LA = left atrium; arrows = sewing rings.

For most cases in which the AV is replaced, the pathology has been well delineated by preoperative echocardiography and angiography (Fig. 10-8). Consequently, the likelihood of the intraoperative TEE exam impacting directly on the surgical management of the AV per se is low. In a retrospective study by Nowrangi et al, none of the 383 patients in their series who underwent aortic valve replacement (AVR) for AS required aortic prosthesis modification post–CPB, based on the intraoperative TEE examination.²⁴ However, in that same study, intraoperative TEE altered the planned operation in 13% of the patients. The changes in surgical plan based solely on new data from the intraoperative pre–CPB TEE exam included performance or cancellation of an MV procedure, PFO closure, and removal of LAA thrombi and other intracardiac masses. In addition, intraoperative TEE measurement of aortic annular size, which the investigators were able to do reliably, allowed homografts to be thawed and prepared prior to aortotomy, potentially decreasing the duration of CPB.

View larger version (83K): [in this window] [in a new window]   FIGURE 10-8 Aortic stenosis. (A) This two-dimensional transesophageal echocardiographic image demonstrates a calcified and stenotic aortic valve. (B) A continuous wave Doppler flow velocity profile is used to measure the velocity (arrow) across the stenotic aortic valve. The velocity across the aortic valve is nearly 4 m/s, which is consistent with a transvalvular pressure gradient (P) of 64 mm Hg according to the modified Bernoulli equation: P = 4(v)2. LA = left atrium; RA = right atrium; AV = aortic valve; PA = pulmonary artery; RV = right ventricle.

Because of its ability to provide detailed information about the structure and function of the entire valvular apparatus, TEE assessment of the AV has a growing impact upon surgical decision-making in cases of aortic insufficiency (AI), especially those due to aortic dissection. Moskowitz et al studied 50 consecutive patients with acute type A aortic dissection, and defined the severity and mechanism of AI in each case.²⁹ They identified three mechanisms of AI which were amenable to surgical repair: (1) incomplete closure of intrinsically normal leaflets due to leaflet tethering by a dilated sinotubular junction; (2) leaflet prolapse due to disruption of leaflet attachments by a dissection flap that extends into the aortic root; and (3) prolapse of the dissection flap through intrinsically normal leaflets, which disrupts leaflet coaptation. Generally, accepted echocardiographic criteria for a normal aortic root include a valvular annular diameter of 19 to 23 mm and symmetrical sinuses, with a sinotubular junction diameter at most 3 mm larger than the valve annulus. In the presence of an anatomically abnormal AV (i.e., bicuspid AV, Marfan's syndrome, leaflet thickening, or aortitis), AVR was determined to be necessary. Of the 16 patients with AI and intrinsically normal leaflets, 15 underwent successful AVR, and did not require subsequent surgery at a median follow-up of 23 months. The lone patient who did not undergo a valve-sparing procedure was excluded at the surgeon's discretion after a complicated intraoperative course unrelated to the AV. The authors concluded that TEE can successfully define the severity and mechanisms of AI in cases of type A aortic dissection, and assist the surgeon in distinguishing repairable valves from those requiring replacement.

Pulmonic Valve

Pulmonic stenosis is rare in the adult patient and is usually associated with a congenital syndrome such as tetraology of Fallot, or a primary etiology including unicuspid, bicuspid, or quadricuspid forms of the valve. Severe pulmonary regurgitation is most usually seen in patients with signs of pulmonary hypertension due to congenital defects (i.e., intracardiac shunts) or due to MV disease. However, mild degrees of pulmonary regurgitation occur in 40% to 70% of the normal adult patient population.³⁰

Pulmonic valve (PV) replacement associated with primary pathology is usually limited to procedures involving congenital heart disease. However, the introduction of the Ross procedure, in which a diseased AV is replaced with the native PV, has promoted interest in the intraoperative echocardiographic evaluation of this structure. Elkins et al reported on 206 patients who underwent Ross procedures.³¹ Eleven patients required autograft reoperations for valve insufficiency (Fig. 10-9). Three patients required a procedure within 6 months and the remaining 8 required reoperation from 1 to 6.2 years later. Intraoperative echocardiography demonstrated 1+ to 2+ AI in 2 of the 3 early reoperation patients. Autograft mismatch was most likely responsible for the AI in 1 patient with a dilated aorta. The second patient had "malalignment of the autograft valve commissures in a scalloped subcoronary implant." In 5 of the 8 patients in whom reoperation was required over a postoperative period of 1 to 6.2 years, the intraoperative TEE evaluation demonstrated abnormal findings. Three of the patients who had significant aortic dilation and left the operating room with 1+ to 2+ AI required reoperation at 2.1, 4.4, and 5.4 years. Two of these 3 patients underwent successful reduction annuloplasty while the third patient required valve replacement after an initial failed attempt at repair. The 2 remaining patients, who had bicuspid AV disease and mild AI preoperatively and only trace or mild AI after autograft replacement, developed progressive AI and aortic annular dilatation by the time reoperation was required.

View larger version (57K): [in this window] [in a new window]   FIGURE 10-9 Moderate to severe aortic insufficiency (arrows) following a Ross procedure. Patients with significant aortic valve dilatation are at risk for postoperative aortic insufficiency unless the surgical procedure is altered to downsize the aortic annulus. LA = left atrium; LV = left ventricle; Asc Ao = ascending aorta.

View larger version (58K): [in this window] [in a new window]   FIGURE 10-10 Pulmonary artery and aortic measurements prior to Ross procedure. (A) Measurement of the pulmonic valve (PV) annular diameter (straight line). (B) Aortic valve and ascending aortic diameters measured in a mid-esophageal, aortic valve long axis transesophageal echocardiographic view. Line a = aortic valve annulus; line b = sinus of Valsalva; line c = sinotubular junction; line d = ascending aorta; RVOT = right ventricular outflow tract; PA = pulmonary artery; LV = left ventricle; LA = left atrium; LVOT = left ventricular outflow tract.

Tricuspid Valve

A comprehensive echocardiographic examination of the tricuspid valve (TV) includes 2-D evaluation for leaflet pathology (i.e., prolapse, calcification, chordal rupture) and annular dimensions. Utilization of Doppler echocardiography is essential for the evaluation of severity of tricuspid regurgitation (TR) and tricuspid stenosis (TS). In one series of patients undergoing TV replacement (TVR), TR was the primary pathology in 75% compared to only 2% of the patients who had pure TS.³⁵ Although a variety of congenital and acquired disorders of the TV are responsible for valve pathology,³⁶ TR is most commonly associated with normal leaflet morphology in the presence of RV and/or LV failure with concurrent pulmonary hypertension.

Early studies using epicardial echocardiography intraoperatively were able to establish the utility of CFD in the evaluation of TV disease.^37,38 In 18 patients undergoing TVR, the immediate post–CPB measurement of TR correlated with echocardiographic evaluation several weeks postoperatively.³⁸ Interestingly, there was poor correlation in this study with RA V wave and degree of TR. Grading of TR, however, is not as standardized at grading of MR. Early studies measured the degree of TR according to CFD jet extension into the RA. A jet that extended just beyond the TV was graded as 1+. A step-up in grade was assigned for each one-third extension into the RA up to 4+ for a jet that extended to the back of the RA.

A study of 66 patients in a database with "moderate or severe" TR by TTE compared 19 echocardiographic findings in an attempt to develop a TR severity score.³⁹ Severe TR was defined by the presence of at least two of the following three clinical findings: (1) V waves in the jugular venous pulsations, (2) pulsatile liver, and (3) seesaw movement in the parasternal area. In this case series, 38 patients had clinical evidence of TR. Four of the 19 echocardiographic signs had a greater than 80% positive predictive value of detecting clinical TR. Hepatic vein systolic flow reversal, paradoxical atrial septal movement, TR jet area greater than 9 cm², and RA area greater than 30 cm² had predictive values of severe TR of 91.2%, 82.8%, 81.4%, and 80.6%, respectively (Fig. 10-11).

View larger version (86K): [in this window] [in a new window]   FIGURE 10-11 Severe tricuspid regurgitation. (A) Enlarged right atrium and color flow Doppler evidence of severe tricuspid regurgitation. (B) Hepatic vein Doppler flow velocity profile demonstrating systolic flow reversal (arrow) consistent with severe tricuspic regurgitation. RV = right ventricle; TV = tricuspid valve.

	VENTRICULAR FUNCTION: REGIONAL AND GLOBAL

Top Previous Next References

 
The primary purposes for any cardiac surgical preoperative risk stratification scheme are to identify patients for whom the risk/benefit ratio of coronary revascularization is most advantageous, to facilitate consultations with other medical specialists, to delineate appropriate postoperative triaging, and to assist in containing hospital expenses. There are several variables describing cardiac surgical patient demographics and comorbidity that have been studied extensively as potential predictors of postoperative adverse events.^43–48 The preoperative estimate of ventricular systolic function is considered one of the more important risk factors and predictors of postoperative outcome in patients undergoing surgical coronary revascularization.^49,50 The beneficial effect of revascularization of dysfunctional myocardium in chronic ischemic heart disease has traditionally been measured by its effect on improvement of resting regional and global LV function.^51–53 Although several studies have described the prognostic value of preoperative echocardiographic evaluation of ventricular function for predicting survival after revascularization, many of these studies have been limited by relatively small sample sizes, lack of serial echocardiographic measurements, and only short-term follow-up.^54–58 In addition, there is still some controversy regarding the independent value of preoperative ejection fraction (EF),⁵⁹ the relative value of global versus regional measures of LV function,⁶⁰ and the role of RV dysfunction⁶¹ as predictors of morbidity and mortality following CABG. Thus, further investigation of the relationship between perioperative echocardiographic measures of global and regional ventricular systolic function and postoperative adverse events is warranted.

Regional Left Ventricular Function

Segmental wall motion abnormalities (SWMA) of the LV develop within seconds of regional myocardial oxygen deprivation, often preceding or even occurring in the absence of ECG changes.^62–64 Although segmental shortening may decline with as little as a 10% to 20% reduction in coronary blood flow, akinesis and dyskinesis require a greater than 90% and greater than 95% reduction, respectively. The echocardiographic evaluation of SWMA includes an assessment of both inward endocardial border excursion and wall thickening during systole, which can be determined by both qualitative visual assessment or quantitative linear measurements.^65,66 Abnormalities in wall thickening tend to occur before changes in endocardial excursion and better define the boundaries of an ischemic area, but both techniques are often used. Normally, endocardial excursion is greater than 30% and myocardial wall thickness increases by 40% to 50% during systole.⁶⁷ Ideally a significant difference in visual SWMA requires a change of at least two grades (e.g., normal to severe hypokinesis or severe hypokinesis to dyskinesis) since intra- and interobserver variability has been demonstrated with this technique.⁶⁸ Acute changes in loading conditions, afterload, and rhythm can also interfere with the interpretation of SWMA.⁶⁹ Furthermore, the appearance of SWMA in the absence of ischemia can be seen with tangential views or with translational and rotational movement of the heart especially in the septal region after pericardiotomy.⁷⁰ Finally, tethering of infarcted or stunned segments may present as SWMA, although a sudden, severe decrease or cessation of LV segmental contraction is almost certainly due to myocardial ischemia. Despite these limitations, the diagnosis of new SWMA by TEE remains an important influence in guiding anti-ischemia therapy in CABG patients.⁷¹

The TEE visual evaluation of regional LV function is often confined to the transgastric, short axis view at the mid-papillary level (TG mid SAX; see Fig. 10-1D) since all four LV walls corresponding to the territorial distribution of the three respective epicardial coronary arteries may be visualized simultaneously (Fig. 10-12). Split screens or multiple saved images can be helpful for making comparisons and determining changes in SWMA over time (i.e., pre– versus post–CPB). The use of the TG mid SAX view as a single plane is practical for continuous long-term evaluation; however, this view does not include apical and basal levels. In addition, evaluation of the lateral and septal walls, which are parallel to the ultrasound beam, may be limited by echocardiographic "drop-out" in this short axis view. A more complete assessment can be performed using the 16-segment model and scoring system established by the American Society of Echocardiography (Fig. 10-13).¹ The use of additional TEE viewing planes has been shown to improve SWMA detection. In one recent study, only 17% of SWMA were detected in the TG-mid SAX view.⁷² Although the introduction of other short axis views increased the detection of SWMA to 65%, the introduction of multiplane views was required to capture the remaining 35%.

View larger version (29K): [in this window] [in a new window]   FIGURE 10-12 Regional coronary artery distribution relative to standard transesophageal echocardiographic views. LAD = left anterior descending coronary artery; Cx = left circumflex coronary artery; RCA = right coronary artery. (Reproduced with permission from Shanewise JS, Cheung AT, Aronson S, et al: ASE/SCA Guidelines for Performing a Comprehensive Intraoperative Multiplane Transesophageal Echocardiography Examination: Recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999; 89:870.)

View larger version (41K): [in this window] [in a new window]   FIGURE 10-13 Regional left ventricular wall segments according to standard transesophageal echocardiographic views. (Reproduced with permission from Shanewise JS, Cheung AT, Aronson S, et al: ASE/SCA Guidelines for Performing a Comprehensive Intraoperative Multiplane Transesophageal Echocardiography Examination: Recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999; 89:870.)

Intraoperative administration of low-dose dobutamine combined with TEE has also been proposed to be a valuable technique for predicting functional recovery in regional LV function following CABG surgery. It is extremely important to distinguish dysfunctional myocardial regions that are viable from those that are irreversibly injured in patients undergoing coronary revascularization procedures. Experimental and clinical evidence have confirmed that low-dose dobutamine increases coronary blood flow and contractility without an associated increase in regional myocardial oxygen demand. Aronson et al analyzed 560 LV segments in 40 patients scheduled for elective CABG surgery for SWMA at four stages in the procedure: baseline (after general anesthesia induction and endotracheal intubation), with administration of low-dose dobutamine (5 µg/kg/min) before CPB, after separation from CPB (early), and after protamine administration (late).⁷⁵ Changes in regional myocardial function after low-dose dobutamine were highly predictive of early and late changes in regional myocardial function compared to baseline regional scores. Thus, intraoperative low-dose dobutamine stress testing may be considered a reliable method for predicting regional myocardial reserve and determining anticipated functional recovery after coronary revascularization.

Endocardial motion can also be displayed in color, using Color Kinesis technology (CK; Philips, Andover, MA) or color Doppler tissue imaging (DTI). CK automatically detects endocardial motion in real time using integrated back-scatter data to identify pixel transitions between blood-tissue interfaces throughout the cardiac cycle. The speed and direction of endocardial motion are displayed as a composite of different, user-selected, frame-specific colors superimposed upon the 2-D gray scale at end systole and at end diastole. Although CK facilitates the detection of endocardial border excursion, wall thickening is not further optimized. In comparison to conventional 2-D analysis of endocardial motion, CK is a sensitive but not particularly specific method for evaluating SWMA.^76–78 In addition, CK technology may be limited by interference from ultrasound "drop-out," image clutter, and translational motion of the heart. Regional LV function may also be evaluated using DTI, which uses Doppler frequency shift signals to continuously display the velocity and direction of myocardial wall motion either in color or as a spectral waveform analysis of velocities over time.

The use of TEE for evaluating perioperative ischemic myocardial injury is considered a class II indication according to the guidelines published by the SCA/ASE.² In several studies, the echocardiographic demonstration of new SWMA in the perioperative period has been shown to be a more reliable sign of myocardial ischemia and infarction compared to ST-segment changes.^79,80 For example, in a study of 50 patients undergoing cardiovascular surgery, new significant SWMA occurred in 24 patients whereas ischemic ST-segment changes only occurred in 6 patients.⁸¹ Three patients who sustained intraoperative myocardial infarction (MI) developed significant SWMA in the corresponding area of myocardium, which persisted until the end of surgery, but only 1 of these 3 patients had ischemic ST-segment changes intraoperatively.

In another study, investigators diagnosed new SWMA in 19 of 179 (10.6%) patients undergoing CABG surgery, and 14 of 71 patients (19.7%) undergoing other cardiac surgical procedures (combined CABG/valve, valve, and aortic surgery).⁸² Intraoperative TEE directed the management of 18 of 33 (54%) of these patients who developed new SWMA. Finally, in another study, TEE was compared to Holter ECG in a series of 351 patients undergoing CABG surgery, to determine if either technique had clinical value in identifying those patients in whom MI (creatine kinase-MB 100 ng/mL) was likely to develop.⁸³ Electrocardiographic or TEE evidence of intraoperative ischemia was present in 126 (36%) of the patients; however, the concordance between the two techniques was poor (positive concordance = 17%; Kappa statistic = 0.13). Myocardial infarction occurred in 62 (17%) of the patients, and 32 (52%) of them had previous intraoperative ischemia of which 28 (88%) were identified by TEE, yet only 13 (41%) were identified by ECG. In this study, SWMA were more commonly observed compared to ECG changes and concordance between the two modalities was low. Logistic regression analysis revealed that TEE was twice as predictive as ECG in identifying patients who develop a perioperative MI. Thus perioperative identification of SWMA remains an important technique for diagnosing myocardial ischemic events and identifying patients at risk for MI following cardiac surgery.

Global Left Ventricular Function

Global LV function is an important predictor of outcome in patients undergoing major surgery. In patients undergoing CABG surgery, global ventricular function, as assessed by EF, is more important in determining survival than the number of diseased vessels.^84,85 The ASE Guidelines for Quantification of the Left Ventricle recommend that, following optimization of image quality (assuring the best resolution of endocardial border definition by using the proper transducer frequency, dynamic range, depth, and transmit zone), LV dimensions, areas, and volume can be acquired to facilitate quantification of contractile function and cardiac performance.⁶⁹ Therefore, a comprehensive evaluation of LV function involves an integration of different ventricular performance constituents (i.e., preload, afterload, and contractility) in addition to an evaluation of global measures. Furthermore, echocardiographic evaluation of trans-MV (LV filling) and pulmonary venous (LA filling) Doppler flow velocity profiles can be useful in the assessment of LV diastolic function.

FRACTIONAL SHORTENING AND FRACTIONAL AREA CHANGE

The simplest echocardiographic method for evaluating global LV function involves indirect measures of EF. Fractional shortening (FS) is derived from the difference between end-diastolic internal diameter (EDID) and end-systolic internal diameter (ESID) divided by the EDID as imaged in the TGSA-MP view (mean: 33%; range: 28%–41%):

FS is usually measured in the TG-mid SAX view and evaluated using M-mode echocardiography with the ultrasound beam oriented through the center of the LV (Fig. 10-14).

View larger version (73K): [in this window] [in a new window]   FIGURE 10-14 Fractional shortening of the left ventricle is derived from an M-mode view of a transgastric mid short axis echocardiogram. The internal diameters of the LV are measured during diastole (D) and systole (S).

Circumferential areas are obtained by tracing around the LV endocardial borders (planimetry) during the corresponding phase of the cardiac cycle in the TG-mid SAX view since 85% of cardiac contractility occurs along this axis (Fig. 10-15). FAC is limited by the ability to obtain optimal cardiac images, errors in tracing or dimensional measurements, and the inherent limitations of equating 2-D measurements with 3-D physiological data. Although FAC correlates reasonably well with other approximations of the global EF, significant limitations apply. For example, FAC like other measures of EF is load-dependent and thus markedly affected by acute changes in preload and afterload without changing the intrinsic systolic function of the LV. Despite these limitations, FAC has been shown to significantly correlate with radionucleotide EF (r = 0.85; SEE = 9.6%).⁸⁶

View larger version (81K): [in this window] [in a new window]   FIGURE 10-15 Fractional area of change of the left ventricle is derived by planimetering the area of the left ventricular cavity in the transgastric mid short axis echocardiogram during (A) diastole and (B) systole.

Since TEE is primarily a 2-D technology and volume is a 3-D property, one must accept certain assumptions about ventricular geometry and use mathematical formulae to calculate volume from internal dimensions by "summing areas" or disks along one or more ventricular axes. Measurement of EF requires that LV volumes be measured in the TG-mid SAX view and one or more long axis views (i.e., mid-esophageal 4-chamber and mid-esophageal 2-chamber views) (Fig. 10-16). Acoustic quantification and automated border detection (AQ and ABD: Philips Medical Systems; Andover, MA) utilize unprocessed radio frequency data to discriminate blood from myocardial tissue to provide continuous, real-time, on-line estimates of EF and stroke volume. These estimates correlate favorably with radionucleotide-derived and thermodilution-acquired estimates.⁸⁷

View larger version (70K): [in this window] [in a new window]   FIGURE 10-16 According to Simpson's rule, the LV end-diastolic volume and LV end-systolic volume can be determined by assuming that the LV is composed of a series of disk-shaped slices, which can each be measured and summed. (A) LV end-diastolic volume. (B) LV end-systolic volume.

Cardiac output and its individual components can also be evaluated echocardiographically. Blood velocity can be measured using PWD or CWD. Stroke volume is calculated from the product of velocity-time integral (VTI: integrated area within the Doppler flow velocity profile) and the measured area through which blood is flowing. Furthermore, cardiac output (CO) can be obtained according to the following formula:

Cardiac output measurement using Doppler echocardiography assumes laminar blood flow, a uniform Doppler flow velocity profile, parallel alignment of the Doppler beam with blood flow, and visualization of a circular and unchanging measured area. Despite these limiting factors, CO measurement using echocardiography has been performed in several different intracardiac locations (MV, AV, pulmonary artery/ RVOT) and has been validated against classical techniques such as thermodilution.^88–90

Left Ventricular Diastolic Function

Diastolic dysfunction describes a limitation of the ventricle to fill to an adequate end-diastolic volume without an abnormal increase in end-diastolic pressure. The main determinants of diastolic filling include myocardial relaxation, passive LV filling characteristics (LV compliance), LA function, heart rate, and MV integrity. Doppler echocardiographic evaluation of the trans-MV (LV-filling) and pulmonary venous (LA-filling) flow velocity profiles can provide information pertaining to LV diastolic function (Fig. 10-17).⁹¹ The normal trans-MV Doppler flow velocity profile is characterized by an early E-wave velocity, which reflects LV relaxation and compliance as well as a later diastolic A-wave, which is more dependent upon the contractile state of the LA and the balance between LA and LV compliance. The normal pulmonary venous Doppler flow velocity profile is characterized by a systolic forward velocity reflecting LA relaxation and atrioventricular interactions, an early diastolic velocity reflecting LA filling while the MV is open, and a small late diastolic flow reversal during atrial contraction.

View larger version (14K): [in this window] [in a new window]   FIGURE 10-17 Transmitral and pulmonary venous Doppler flow velocity profiles. The impact of progressive left ventricular diastolic dysfunction on transmitral (TMDF; top) and pulmonary venous Doppler flow (PVDF; bottom) velocity profiles. E = E-wave; A = A-wave; IVRT = LV isovolumic relaxation time; PVAR = late diastolic retrograde velocity; PVS1 = first systolic component; PVS2 = second systolic component; PVD = diastolic component.

Preoperative diastolic dysfunction has been reported in 30% to 70% of cardiac surgical patients and has been associated with difficult weaning from CPB, the need for more frequent inotropic and vasoactive pharmacological support, and increased morbidity.^93,94 Following CPB, acute or worsening diastolic dysfunction of varying degrees associated with ischemia-reperfusion injury, hypothermia, metabolic disturbances, or myocardial edema occurs in many patients and may persist for several minutes to days.^95,96 Therefore the availability and feasibility of echocardiography for diagnosing the presence and progression of perioperative diastolic function should assist in the identification of high-risk cardiac surgical patients who may benefit from appropriate triage and therapeutic intervention.

Right Ventricular Function

The ASE/SCA guidelines divide the RV into the free wall (with basal, apical, anterior, and inferior segments), the septal wall, and RV outflow.¹ The four conventional TEE views for evaluating RV function include the ME 4-chamber (see Fig. 10-1A), TG mid SAX (see Fig. 10-1D), ME RV inflow-outflow (see Fig. 10-1M), and TG RV inflow (see Fig. 10-1N). Compared to the evaluation of LV function, objective quantification of RV function is more difficult by both TTE and TEE.^97,98 The RV free wall is extremely thin, with an end-diastolic thickness of less than 5 mm. Consequently, evaluation of wall thickening and evaluation of RWMA may be difficult except in the presence frank akinesis or dyskinesis. In addition, the asymmetrical and crescent-shaped RV does not lend itself to volumetric measurement by ultrasound. Furthermore, the RV contains multiple trabeculations making endocardial border detection and accurate volume tracing difficult. Standard geometric formulas for volume calculations have only limited applicability given the shape of the RV. Attempts at modeling the RV in the TTE literature according to complex geometric shapes (i.e., ellipse, half cylinder, half conical structure, prism, a pyramid with a triangular base, or a three-dimensional parallelogram) have met with varying degrees of success when compared to volumetric methods (angiography, nuclear magnetic resonance, or formalin casting).³ Few studies have attempted to evaluate geometric models of the RV in a varying number of patients with a wide spectrum of RV pathology, and they have no immediate intraoperative application.^102,103

Despite these limitations, RV dysfunction is conventionally diagnosed echocardiographically by evaluating measures of both RV dimensions and RWMA. Because of its thin wall, the RV will dilate in response to volume overload. A comparison of RV to LV area is often used as a subjective measure of RV dilation, being described as mild (RV area smaller than LV area), moderate, (RV equals LV area), or severe (RV larger than LV area).² Because of its highly unusual geometry, it is essential to scan the RV in multiple planes in order to diagnose pathological RV dilation.

Regional dysfunction of the RV can also be evaluated by diagnosing the presence of hypokinetic, akinetic, or dyskinetic systolic RWMA. Additional echocardiographic features including RV free wall hypokinesis, systolic and diastolic septal flattening,⁹⁹ and dilation of the RA, IVC, and pulmonary artery in the presence of TR may support a diagnosis of RV failure.^100–102 However, in an intraoperative clinical setting, semiquantification of the RV area compared to the LV area is usually sufficient. Although objective echocardiographic signs of RV systolic dysfunction may be difficult to determine, subjective signs of RV dysfunction have been used to demonstrate its incidence in the perioperative period.¹⁰³ In one retrospective study involving 66 emergent intraoperative TEE exams in 46 cardiac surgical patients, 2 of the 66 had a main TEE finding of "severe right heart dysfunction."¹⁰³ In this study the main diagnoses were aortic dissection (n = 14) and severe LV dysfunction (n = 17). Another similar case series looked at urgently requested TEE exams in the cardiac surgical intensive care unit, and found, out of 130 emergent TEEs, 18 resulted in a diagnosis of RV dysfunction.¹⁰⁴ The incidence of RV dysfunction was surpassed only by LV dysfunction (n = 39).

Nonetheless, there are few studies that systematically investigate the utility of intraoperative echocardiography in evaluating RV function. In a study involving 16 patients undergoing CABG, Rafferty et al attempted to correlate RV EF (as measured by a rapid-response thermodilution pulmonary artery catheter) with an echocardiographic calculation using an off-line analysis of planimetered RV systolic and diastolic volumes in the ME 4-chamber and the TG mid SAX views, as well as systolic and diastolic length changes along a major axis and minor axis in the ME 4-chamber view.¹⁰⁵ Measurements were taken after general anesthesia induction, following sternotomy/pericardotomy, and after CPB. Although thermodilution RVEF did not change after CPB (EF PRE = 41%; EF POST = 37%), planimetered EF changed from 50% before CPB to 37% after CPB, indicating how difficult it is to correlate echocardiographic images of the RV.

A more recent study¹⁰⁶ demonstrated the utility of echocardiography for evaluating RV function postoperatively in cardiac surgical patients by administering incremental doses of dobutamine and obtaining RV pressure-area measurements using automated endocardial border detection. In addition, the echocardiographic measurement of RV cross-sectional area was validated intraoperatively as a surrogate for RV volume. Echocardiographic measurements were obtained from the difference between RV systolic and diastolic cross-sectional areas. Stroke volume was also measured directly by the investigators by placing an electromagnetic flow probe around the main pulmonary artery. Snares were placed around the IVC and then slowly tightened until a systolic radial artery pressure reached 70 mm Hg before being released to vary loading conditions. RV cross-sectional area as determined by automated border detection was shown to correlate with changing flow rate. The authors emphasized, however, that the utility of echocardiography for measuring RV stroke volume should be interpreted with caution since maintaining the image of the RV in a ME 4-chamber view required strict vigilance on behalf of the ultrasonographer, and had not been verified in previous studies.

Intraoperative RV systolic and diastolic function have also been evaluated echocardiographically by assessing Doppler hepatic venous flow velocity (HVF) patterns before and after CPB. The normal HVF pattern is characterized by forward flow in systole that exceeds the forward flow velocity in early diastole. The systolic component is produced by RV contraction with a combination of atrial relaxation and descent of the TV annulus. Alternatively, the diastolic component is a result of TV opening and RV relaxation and compliance. In addition, reversed flow associated with atrial contraction and late systole may also be observed. The results from two studies that evaluated changes in HVF patterns in cardiac surgical patients implicated that the development of post–CPB diastolic predominance suggested the presence of a reduction in RA compliance or relaxation associated with impaired RV relaxation and contraction. In addition, the demonstration of increased end-systolic reversal supports a diagnosis of RV stunning and dysfunction.^107,108

Ventricular Dysfunction Associated with Pulmonary Embolism

Pulmonary embolism (PE) is a relatively rare but important cause of acute RV heart failure. The sensitivity (58%–97%) and specificity (88%–100%) of TEE for directly visualizing a pulmonary embolus in situ is somewhat variable especially for peripheral pulmonary emboli.¹⁰⁹ Consequently, indirect echocardiographic signs including the presence of mobile thrombus in the RA, RV, vena cavae, and/or main pulmonary artery are often considered highly suggestive of an acute PE, especially when visualized concurrently with evidence of acute RV failure in a high-risk patient. Initial RV dilation accompanied by tricuspid regurgitation is associated with an inability to compensate for sudden increases in resistance to flow in the pulmonary artery that occur following a clinically significant PE. An end-diastolic ratio of RV/LV diameter of 0.6 correlates with a significant RV pressure overload, whereas a ratio of 0.8 indicates a PE.¹¹⁰ RV systolic dysfunction follows acute dilation, often manifesting with a paradoxical shift and flattening of the interventricular septum and echocardiographic evidence of an underfilled LV.¹¹¹

A characteristic abnormal motion of the RV free wall relative to the apex has been demonstrated in patients with documented PE.¹¹² Normally, the RV contracts with a spiral motion that pulls the TV toward the apex. The RV free wall subsequently contracts inward toward the septum, creating the bellows-like effect. Finally, the RV is pulled by the contracting LV near the site of mutual attachment. In contrast, patients who have a PE demonstrate a bulging or akinesis of the middle portion of the RV free wall, while the apex continues to contract normally. In the study by McConnell et al, this unique pattern of RV motion was initially described in 14 patients with documented PE, and was validated retrospectively by viewing the TTE of 85 patients who had RV dysfunction from a variety of etiologies and sources.¹¹² The echocardiographic finding of RV free wall akinesis with normal apical contraction had a 77% sensitivity and a 94% specificity for the diagnosis of PE. The authors speculated that the mechanism for this abnormal RV motion could be due to localized RV ischemia or RV apical tethering to a normal or hyperdynamic contracting left ventricle. Despite the apparent importance of this "pathognomonic" finding, the use of TEE may still be limited by difficulty in routinely visualizing the RV apex.

	PERICARDIAL EFFUSIONS AND CARDIAC TAMPONADE

Top Previous Next References

 
Pericardial Effusions

Two-dimensional echocardiographic visualization of the pericardial cavity as an "echo-free space" depends upon the volume and pattern of fluid distribution around the heart. Small pericardial effusions (<100 mL; <10 mm in width) tend to be localized behind the posterior LV wall.¹¹³ Moderate sized effusions (100–500 mL; 10–15 mm in width) expand laterally, apically, and anteriorly and can be visualized throughout the entire cardiac cycle. Large pericardial effusions (>500 mL; >15 mm in width) tend to be more evenly distributed around the heart, although preferential posterior accumulation is still common.¹¹⁴ Large, chronically accumulating pericardial fluid effusions are often associated with excessive anteroposterior heart motion, and counterclockwise rotation in the horizontal plane.

Pericardial effusions can be identified postoperatively in approximately 85% of cardiac surgical patients although their size may vary, usually peaking by the 10th postoperative day.¹¹⁵ The clinical presentation of postoperative effusions may also vary from a spurious finding in an asymptomatic patient to hemodynamic instability associated with local or circumferential cardiac chamber compression. The absolute size of the effusion, however, does not necessarily correlate with clinical signs since even relatively small, rapidly accumulating effusions may produce symptoms. Although postoperative effusions may present as an isolated anterior effusion, they are often located posteriorly and loculated by restraint from adhesions within the pericardial space.¹¹⁶ Postoperative hemopericardium can present with thrombus usually along the anterolateral RA free wall, and may also cause compression and hemodynamic compromise.¹¹⁷ Moderate or large pericardial effusions requiring surgical intervention also develop in patients with renal failure and following pericardial infiltration from lung and breast carcinoma.¹¹⁸

Pericardial Tamponade

Cardiac tamponade is a clinical syndrome defined as the decompensated phase of cardiac compression resulting from increased intrapericardial compression. Severe cardiac tamponade causing significant impairment of ventricular filling often requires surgical intervention to prevent circulatory collapse. Patients with severe cardiac tamponade usually become symptomatic when an excessively large or rapidly accumulating pericardial effusion develops (i.e., perioperative coagulopathy, vascular incisional or anastomotic dehiscence, ascending aortic dissection, coronary artery dissection following angioplasty, cardiac trauma, etc.). Postoperative pericardial tamponade has been reported in up to 2% of cardiac surgical patients.¹¹⁹

TWO-DIMENSIONAL ECHOCARDIOGRAPHY

Although a variety of 2-D echocardiographic signs have been reported in patients with symptomatic pericardial tamponade, the most important include evidence of RA and RV diastolic collapse (Fig. 10-18).¹²⁰ Many of the described classic echocardiographic features of cardiac tamponade may not be present when small increases in intrapericardial volume are associated with rapid rises in pericardial pressure (i.e., cardiac trauma) or when effusions are loculated.¹²¹ Following pericardiocentesis, cardiac pressure and volume corrections are more likely to reflect the presence and degree of cardiac compression rather than the amount of fluid removed. After drainage, the more compliant RV may dilate significantly and develop abnormal septal motion consistent with volume overload and congestive heart failure.¹²² LV dilation ("pericardial shock") and pulmonary edema may also develop following pericardiocentesis in patients with underlying myocardial disease, due to the acute increase in venous return in the presence of persistent compensatory increases in peripheral resistance.¹²²

View larger version (59K): [in this window] [in a new window]   FIGURE 10-18 Two-dimensional transesophageal echocardiogram reveals the classic echocardiographic finding of diastolic right atrial (RA) collapse (arrows) and a large pericardial effusion (PE). LA = left atrium; RV = right ventricle; LVOT = left ventricular outflow tract.

In comparison to 2-D echocardiography, Doppler echocardiographic features of pericardial tamponade are more sensitive measures of clinical severity.¹²³ Doppler findings of pericardial tamponade are based upon alterations of intrathoracic and intracardiac pressures that occur during respiration and an exaggeration of ventricular interdependence. Normally during spontaneous respiration, changes in intrathoracic pressures are almost equally transmitted to the pericardial space and intracardiac chambers. Significant pericardial effusions often blunt the transmission of intrathoracic pressure. Consequently, LA and LV filling pressure gradients are decreased during spontaneous inspiration, resulting in diminished pulmonary venous forward diastolic velocities, delayed MV opening, prolonged isovolumic relaxation time, and decreased mitral E-wave velocity.¹²⁴ Similarly, relative increases in LA and LV filling pressure gradients during spontaneous expiration are responsible for corresponding increases in LA and LV Doppler inflow velocities. Exaggeration of ventricular interdependence with pericardial tamponade is responsible for reciprocal changes in right-sided intracardiac flows, resulting in increased tricuspid early E-wave filling velocities during spontaneous inspiration. These transvalvular Doppler flow velocities tend to normalize following therapeutic pericardiocentesis.¹²⁵ Alterations in the respiratory variation of RA inflow velocities (i.e., vena cavae and hepatic veins) are also altered by cardiac tamponade pathophysiology. For example, hepatic vein forward velocities decrease and reverse flows increase during spontaneous expiration.¹²³

The identification of a significant pericardial effusion in a hemodynamically unstable, symptomatic patient can provide sufficient evidence to warrant prompt surgical intervention, thus avoiding a delay in diagnosis and further morbidity or mortality. In patients who require surgical intervention, perioperative echocardiography including 2-D and Doppler can provide important information related to the determination of postoperative functional status. Understanding the utility of perioperative echocardiography in evaluating patients with significant pericardial effusions is therefore an important component of the ultrasonographer's knowledge base.

	AORTIC DISSECTION AND ATHEROMATOUS DISEASE

Top Previous Next References

 
Ascending Aortic Dissection

The sensitivity of TEE in the diagnosis of thoracic aortic dissections has been well documented in the literature. For example, in the previously cited study by Moskowitz et al,²⁹ an intimal dissection flap was identified in all 50 patients studied, highlighting the high sensitivity of TEE for diagnosis of aortic dissection (Fig. 10-19). This has been a consistent finding in other studies of multiplane TEE in the setting of aortic dissection.¹²⁶ With reported sensitivities of 98% to 100%, TEE is comparable to MRI and superior to CT or aortography as a screening test. Characteristic echocardiographic findings in cases of aortic dissection include the intimal flap, which appears as an undulating, intralumenal density; the presence of two separate echoes in the aortic wall, one representing the intima and inner media, the other the outer media and adventia; and the presence of two lumens with different CFD patterns.¹²⁷ The limitation of TEE as a diagnostic modality for aortic dissection is its tendency to produce false-positive results, as a consequence of reverberation artifacts, in dilated ascending aortas. The specificity of multiplane TEE in these cases is 94% to 95%, which is inferior to the specificity of 98% to 100% quoted for MRI and the basis of the recommendation by one group that MRI should be the initial diagnostic test on hemodynamically stable patients.¹²⁸ This limitation can be minimized with utilization of M-mode technology, which in one study improved the specificity of TEE to 100%.¹²⁹

View larger version (55K): [in this window] [in a new window]   FIGURE 10-19 Mid-esophageal long axis ascending aortic transesophageal echocardiographic view of an ascending aortic aneurysm (AAA) and dissection. Arrows point toward the intimal flap. LA = left atrium; LVOT = left ventricular outflow tract.

Echocardiographic Evaluation of Aortic Atheromatous Disease

Postoperative neurologic dysfunction in the cardiac surgical population continues to be a significant cause of mortality, morbidity, and increased cost. In 1996, a prospective multicenter study of 2108 patients undergoing CABG showed an incidence of adverse cerebral outcome of 6.1%.¹³² This figure is of particular concern in view of the fact that the patients who suffered postoperative neurologic dysfunction in this study had a significantly higher mortality rate (16% vs. 2%, p < .001), had a longer length of hospital stay (22 days vs. 9.5 days, p < .001), and were less likely to be discharged home (46% vs. 90%, p < .001) than patients who did not sustain neurologic injury. The risk factor most predictive of poor neurologic outcome in this study was the presence of ascending aortic atheroma, which increased the incidence of adverse neurologic outcome more than 4-fold over that seen in patients without aortic disease. These authors concluded that this finding supported "the theory that most strokes are caused by large atherosclerotic emboli liberated by surgical manipulation of the aorta."¹³²

The association between ascending aortic atheroma and postoperative neurologic dysfunction was originally hypothesized in case reports from the surgical literature in the 1970s,^133,134 and was further advanced by autopsy studies¹³⁵ and several retrospective series that examined the link between aortic atheromas, as diagnosed by intraoperative TEE, and postoperative embolic stroke. Most notable among these investigations was a study by Davila-Roman,¹³⁶ which observed a significantly higher occurrence of intraoperative stroke (15%) in patients with aortic arch atheromas than in those without aortic arch disease (2%). In addition, a study by Stern¹³⁷ similarly reported an intraoperative stroke incidence of 11.6%, versus a general institutional intraoperative stroke incidence of 2.2%, in patients with aortic arch atheromas larger than 5 mm in diameter. Aortic atheromatous disease is generally classified according to plaque thickness and mobility (Fig. 10-20). ¹³⁸ The causal relationship of aortic atheromas to embolic stroke was ultimately demonstrated in prospective studies^139–141 of nonsurgical patients during the 1990s. These studies showed an approximately 4-fold increase in the risk of future embolic stroke for patients with significant aortic atheromatous disease, as diagnosed by echocardiographic imaging.

View larger version (60K): [in this window] [in a new window]   FIGURE 10-20 Grading of descending aortic atheromatous disease by transesophageal echocardiography.

Historically, the preoperative chest x-ray, cardiac catheterization, and manual palpation of the aorta constituted the predominant diagnostic regimen for identification of aortic atheroma. Unfortunately, although each of these modalities has reasonable specificity for the presence of atheromatous disease, none is sensitive enough to be particularly useful as a screening tool. Chest radiography, which has a predictive value of 84% for identifying atheroma if aortic knob calcification is seen, has a sensitivity of only 52%; cardiac catheterization provides no additional information beyond that obtained from the chest x-ray alone.¹⁴⁵ The diagnostic utility of manual palpation of the aorta has been studied more extensively, but with equally unimpressive results. For example, Ohteki et al¹⁴⁶ found that palpation detected only 25% of atherosclerotic plaques seen echocardiographically, and Sylviris found that palpation undergraded the severity of ascending aortic atheromas, compared with ultrasound imaging, in up to 69% of patients.¹⁴⁷

Newer imaging modalities such as CT, magnetic resonance imaging (MRI), and ultrasound have improved the clinician's ability to diagnose thoracic aortic atheromatous disease. Several studies have investigated the sensitivity and specificity of these tests for the detection of this entity. CT scanning, when compared with TEE, has been shown to have a sensitivity of 87% and a specificity of 82% for the detection of protruding aortic atheromas.¹⁴⁸ Better results have been observed with MRI, which has a sensitivity and specificity equivalent to TEE for detection of the presence of atheroma, but its tendency to undergrade the thickness of the plaques represents a diagnostic limitation.¹⁴⁹

As a result of these and similar studies, echocardiography has emerged as the gold standard for detection of atheromatous disease of the thoracic aorta. In the cardiac surgical patient, however, the opportunity to place an ultrasound transducer directly upon the ascending aorta has generated an additional debate as to the optimal mode of echocardiographic scanning. This debate has arisen because of the fact that ultrasound is unable to adequately penetrate air-filled structures, such as the trachea and proximal bronchi. This gives epiaortic scanning a physical advantage over TEE for the assessment of the distal ascending aorta and proximal arch, which lie above the trachea and bronchi; these aortic segments represent a "blind spot" in the TEE field of vision. Several recent studies investigated whether the theoretical advantages of epiaortic scanning resulted in improved sensitivity for the diagnosis of aortic atheromas in cardiac surgical patients. Konstadt et al¹⁵⁰ showed that TEE failed to visualize as much as 42% of the ascending aorta, and as a result failed to diagnose severe plaques, as seen by epiaortic scanning, in 36% of patients. Davila-Roman et al¹⁵¹ compared biplane TEE with epiaortic ultrasound and found that TEE underdiagnosed atheromatous disease in 39% of proximal ascending aortic segments and 66% of distal ascending aortic segments. Similarly, Sylviris et al¹⁴⁷ reported that TEE undergraded atheromas in 50% of patients in both mid and distal ascending aortic segments, but only in 9% of patients in the proximal segments. Thus, for the detection of atheroma in the areas of the aorta most frequently manipulated during cardiac surgery, it can be concluded that epiaortic scanning is superior to TEE.

However, not every patient undergoing cardiac surgery requires epiaortic scanning prior to aortic manipulation.¹⁵² In the absence of significant risk factors for aortic atheromatous disease, such as age older than 65 years, diabetes mellitus, smoking, and peripheral vascular disease, the incidence of ascending aortic disease is very low, and epiaortic scanning is unlikely to provide useful information. Furthermore, the negative predictive value of TEE examination of the thoracic aorta for atheromatous disease of the ascending aorta has been shown to be 100%¹⁵³ so that an intraoperative TEE exam that is negative for atheroma effectively obviates the need for epiaortic scanning. The optimal approach for detection of ascending aortic disease in cardiac surgical patients should incorporate consideration of risk factors and results of the TEE exam before time is spent on the epiaortic examination.

	INTRACARDIAC MASSES

Top Previous Next References

 
One of the major indications for the use of perioperative TEE in cardiac surgical patients is to assess and evaluate intracardiac masses prior to resection. Since the ultrasound beam is not traveling through the chest cavity as in transthoracic echocardiography (TTE), the images of cardiac masses obtained by TEE are usually clearer, more defined, and often quite dramatic.¹⁵⁴ In a series of 93 nonsurgical patients, TEE identified intracardiac and paracardiac masses in all but 89 (96%) of the cases, whereas TTE identified only 54 (58%).¹⁵⁵ In addition, TTE was not able to visualize thrombi in the left atrial appendage (LAA) (n =12), LA (n = 6), superior vena cava (SVC) (n = 3), or right pericardial space (n = 5). LV thrombi detected in 5 patients by TTE, however, were not visualized by TEE.

If TEE is not performed prior to surgery, the intraoperative echocardiographic examination must reconfirm the diagnosis from other preoperative imaging modalities and assure the surgical team that the intracardiac mass does not have any extracardiac extension. Proper determination of the size, shape, and mobility of the intracardiac mass can assist the surgical team in determining the CPB cannulation technique and sites if it appears that the attachment of the structure is tenuous, suggesting the possibility of embolism from surgical manipulation. In addition, knowledge of potential obstructive effects of the mass on transvalvular flow patterns may be helpful in diagnosing the etiology of pre–CPB hemodynamic instability and supports the potential benefits of changing patient positioning, administering fluid resuscitation or vasoactive medications, and/or determining the urgency of initiating CPB.

Intracardiac Thrombus

Transesophageal echocardiography is a particularly useful technique for diagnosing the presence and location of thrombus in all cardiac chambers with the exception of the LV. In particular, thrombus in the LV apex may be difficult to visualize via a TEE approach compared to TTE.¹⁵⁵ TEE is particularly useful for identifying thrombus in the LAA. In a study by Click et al, unexpected thrombi were identified by TEE in 37 out of 3245 patients (1.1%), thus altering the surgical procedure in 36 patients.¹⁰

Patients at risk for intracardiac thrombus formation include those with stasis of blood flow associated with atrial fibrillation, LA enlargement, isolated MS, or decreased CO. Echocardiographic spontaneous contrast indicating relative stasis often appears as a smoky texture within the LA and alerts the ultrasonographer to interrogate the LAA for thrombus.¹⁵⁶ The echocardiographic characteristics of an atrial thrombus include layering, an area of central echolucency, and broad-based attachment to the atrial wall or septum. In addition, the absence of contraction can be used to distinguish thrombus from endocardium especially in the presence of protruding trabeculations and atrial ridges, which can often be confused with isolated clots (Fig. 10-21).

View larger version (64K): [in this window] [in a new window]   FIGURE 10-21 Transesophageal echocardiographic image of the left atrial appendage (LAA). (A) Normal LAA. (B) LAA with clot. PV = pulmonary vein; AV = aortic valve.

Primary malignant cardiac tumors are fortunately rare in the adult. However, malignant tumors that originate in the thorax may extend directly to the heart, compressing cardiac chambers or compromising the integrity of adjacent great vessels.¹⁵⁸ Echocardiographic delineation of the extension and origination of intra- and pericardial tumors can have major impact in determining the extent of the surgery.

The majority of benign primary cardiac tumors are atrial myxomas. Almost 80% of these solitary tumors originate in the LA, but they have been known to occur in the RA and RV. Rarely, myxomas have been reported simultaneously in multiple locations. Typical echocardiographic characteristics of myxomas include a rounded shape, smooth margins, lack of laminated appearance, and a definitive site of attachment (stalk) most frequently from the fossa ovalis of the interatrial septum (Fig. 10-22). The tumor's site of origin may have particularly important implications for involvement of adjacent, indirectly involved structures including the MV. In addition, identifying the location and extent of RA myxomas may be important for guiding the appropriate positioning of inferior vena cava (IVC) and SVC cannulae.¹⁵⁸ Finally, following resection of a myxoma, intraoperative post–CPB TEE is necessary to determine the completion of resection and to assure that the atrial septum is intact.

View larger version (58K): [in this window] [in a new window]   FIGURE 10-22 Left atrial myxoma attached by stalk (arrow) to atrial septum. LV = left ventricle; MV = mitral valve; LA = left atrium; RA = right atrium.

The accurate diagnosis of abnormal cardiac masses requires a fundamental appreciation for variants of normal anatomy. Intracardiac structures that may be confused with pathologic masses include the eustachian valve (EV),¹⁵⁹ Chiari network,¹⁶⁰ RV trabeculations,¹⁶¹ moderator band, the LAA, and lipomatous infiltrations of the atrial septum.^162,163

The EV and Chiari network are remnants of an embryologic membrane in the RA that directs flow from the IVC to the foramen ovale. The EV, located at the junction of the IVC and RA, is a mobile, thin structure that can often be confused with thrombus (Fig. 10-23). ^159,160 The Chiari network, found in 1.0% to 1.5% of patients studied by TTE, is located near the coronary sinus and also appears as a highly mobile, thin filamentous structure.¹⁶¹ The Chiari network is usually much longer than the EV and may be distinguished from thrombus by the presence of a broader base of attachment.

View larger version (53K): [in this window] [in a new window]   FIGURE 10-23 A mid-esophageal bicaval transesophageal echocardiographic view of a prominent eustachian valve (arrow). LA = left atrium; RA = right atrium; SVC = superior vena cava; IVC = inferior vena cava; CSS = coronary sinus.

View larger version (54K): [in this window] [in a new window]   FIGURE 10-24 A transgastric mid short axis transesophageal echocardiographic view of a large moderator band in the right ventricle (arrows). RV = right ventricle; LV = left ventricle.

	CONGENITAL CARDIAC DISEASE

Top Previous Next References

Muhiudeen et al compared the pre- and postoperative diagnostic utility of TEE with epicardial scanning in 90 pediatric patients (mean age: 4.1 years) with congenital heart disease.¹⁶⁶ In the pre–CPB evaluation of 130 different congenital lesions (shunts and regurgitant and obstructive lesions), TEE failed to diagnosis correctly only 7 anomalies. Epicardial scanning missed 2 anomalies and was superior only in the pre–CPB diagnosis of obstructive lesions of the RVOT. Given that epicardial scanning requires interrupting the surgical procedure and raises the potential for breaking sterility, the conclusion from this study was that TEE was comparable to epicardial scanning for the perioperative evaluation of most congenital lesions.

Post–CPB findings in another study of children undergoing open heart surgery for congenital defects demonstrated that TEE had a major impact in reinitiating CPB.¹⁶⁷ Successful TEE exams performed in 227 out of 230 pediatric patients (mean age: 4.4 years) demonstrated an incomplete repair in 17 (7.5%) of the cases. The initial diagnoses in these patients were VSD (n = 5), subaortic stenosis (n = 9), and 1 each of transposition, aortic valve (AV) disease, and double outlet RV. Although these data represent a pediatric population, inferences can be made pertaining to the overall utility of perioperative echocardiography in the evaluation of congenital heart lesions.

The importance of performing a comprehensive intraoperative TEE exam and understanding anomalies associated with the primary diagnosis are essential in the evaluation of the adult patient undergoing surgery for congenital heart disease. Both are needed to avoid overlooking unsuspected lesions that would warrant surgical repair. In adult patients with extremely complex congenital heart disease, close collaboration with an expert ultrasonographer who specializes in complex congenital adult heart disease may be necessary to assure the highest standard of care. Discussion of the ultrasound characteristics of complex congenital lesions is beyond the scope of this chapter; however, the impact of intraoperative ultrasound on common lesions of the atrial septum provides an example of the utility of intraoperative echocardiography in adult congenital heart surgery.

Atrial Septal Defects

Atrial septal defects (ASD) including ostium secundum, ostium primum, sinus venosus, and coronary sinus defects are among the most common congenital heart defects found in adults. Although many of these lesions may be well defined preoperatively, a comprehensive TEE examination must confirm not only the presence of the known congenital lesion but also the direction and quantification of shunting. In addition, it is imperative for the ultrasonographer to rule out the presence of collateral pathology (e.g., cleft anterior MV leaflet associated with ostium primum defects) and any additional concurrent congenital lesions that may have a direct impact on the surgical procedure. Immediately following CPB, a thorough intraoperative TEE exam including 2-D and CFD must also be performed to ensure anatomic integrity of the surgical repair. The intravenous injection of saline contrast may be particularly useful for detecting trans-septal flow.¹⁶⁶

A sinus venosus ASD may be difficult to thoroughly evaluate preoperatively by TTE. Consequently, a comprehensive intraoperative pre–CPB TEE examination focusing on the location and size of the lesion as well as the presence of any concurrent congenital pathology (e.g., anomalous pulmonary venous return) is of critical importance (Fig. 10-25). The post–CPB TEE examination may also be particularly important since repair of these defects, using a patch or a baffle, may involve the pulmonary veins, may obstruct the IVC, or may not completely close the defect. In each of these instances, the diagnosis may be made immediately by the intraoperative TEE.

View larger version (46K): [in this window] [in a new window]   FIGURE 10-25 The close proximity of the transesophageal echo probe to the left atrium (LA) permits accurate visualization of atrial septal defects. A sinus venosus defect (white dots) is clearly evident in this mid-esophageal bicaval view. SVC = superior vena cava; RA = right atrium; LA = left atrium.

A patent foramen ovale (PFO) at least 0.2 to 0.6 cm in diameter can be found in more than 25% of surgical patients (Fig. 10-26).¹⁶⁸ In comparison to TTE, TEE with its clear views of the of the atrial septum has been shown in numerous studies to permit a more thorough evaluation of PFO anatomy and the presence of trans-septal shunting.^169,170

View larger version (53K): [in this window] [in a new window]   FIGURE 10-26 Color flow Doppler demonstrating patent foramen ovale (arrow). LA = left atrium; RA = right atrium.

	HEART TRANSPLANTS AND VENTRICULAR ASSIST DEVICES

Top Previous Next References

 
Heart Transplantation

In the heart transplant recipient, the value of the pre–CPB echocardiographic examination is usually limited to its utility as a monitor of ventricular performance. In addition, since most transplant recipients present with a history of severe, chronic cardiomyopathy and cardiomegaly, the potential for the stasis of blood flow is significant. Consequently, a comprehensive echocardiographic investigation for the presence of intracardiac thrombus is warranted prior to excessive manipulation and cannulation.

Once the donor heart is placed, a thorough post–CPB echocardiographic examination of RV and LV regional and global function is essential to guide appropriate pharmacologic or surgical intervention. Depending on the duration of donor ischemia time, the adequacy of myocardial protection during CPB, the duration of CPB, and the presence of recipient pulmonary hypertension, the incidence of RV and/or LV dysfunction is variable. Interestingly, interventricular septal hypokinesis persisting through 12 postoperative months has been demonstrated in the majority of cardiac transplant recipients in at least two investigations.^175,176 Explanations for this echocardiographic finding included perioperative septal ischemia, development of RV dysfunction in the presence of pulmonary hypertension, or the development of posttransplantation subendocardial coronary artery disease. Consequently, the presence of isolated interventricular septal hypokinesis in an otherwise hemodynamically stable cardiac transplant recipient should not necessarily cause concern.

The atrial suture lines between the donor and recipient cuffs may have a distinctive "hour-glass" echocardiographic appearance.^177–180 Stenosis of the suture line that impedes intracardiac flow has been reported. In one study of 64 transplant recipients who underwent echocardiographic examination at an average of 31 months postoperatively, spontaneous echo contrast was seen in 35 of the patients by TEE, and 18 patients had evidence of LA thrombus (38%).¹⁷⁹ Another study demonstrated only a 25% incidence of spontaneous LA echo contrast by TEE, but a 50% incidence of suspicious LA or RA masses suggestive of intra-atrial thrombus.¹⁷⁷ Therefore, a comprehensive intraoperative echocardiographic examination should include confirmation of atrial size and geometry, which may correlate with the potential for intra-atrial thrombus formation.

Postoperative MR and TR following cardiac transplantation have been well documented in the literature.^{177–179,181} In these TEE studies acquired from 5 to 31 months postoperatively, an incidence of 73% to 100% of mild to moderate TR and 53% to 77% of mild to moderate MR has been reported. The etiology of posttransplant MR includes the development of an enlarged LA with mitral annular dilatation and resultant annuloventricular disproportion,^182,183 cardiac edema,¹⁸⁴ SAM,¹⁸⁵ and subacute bacterial endocarditis.¹⁸⁶ Posttransplant TR has been attributed to annular dilatation secondary to alterations in RV geometry associated with subclinical rejection or adaptation to higher pulmonary pressures.¹⁸¹ In addition, direct and indirect injury to the TV during endomyocardial biopsy has also been reported.^187–189 Consequently, the documentation of valve function by intraoperative echocardiography in the immediate post–CPB period should be an essential component of a routine examination in a heart transplant recipient.

Ventricular Assist Devices

Several important principles should be incorporated into a comprehensive intraoperative echocardiographic examination of patients receiving ventricular assist devices (VAD). Before implantation, it is important to document the extent of ventricular regional and global dysfunction. Patients undergoing VAD placement are also susceptible to preoperative intracardiac thrombus formation, which has been demonstrated in 12% of this high-risk patient population.¹⁷⁷ Interrogating the extent and location of significant concurrent valvular disease is also important. The demonstration of significant AI in 4 patients undergoing the placement of a HeartMate (Thermocardiosystems Inc, Woburn, MA) resulted in the requirement for an additional AV procedure to assure appropriate VAD function via the ascending aortic outflow cannula.¹⁷⁷ TR, diagnosed by intraoperative echocardiography in 23 of the patients, did not require a concurrent repair. Finally, the documentation of a PFO may have important consequences in patients scheduled for a VAD. Following left VAD (LVAD) placement, the subsequent decrease in LA and LV pressures could promote predominant right to left shunting via a PFO.^175,177 In the 91 HeartMate patients discussed above, 8 were found to have a PFO that was closed during surgery and another patient was subsequently discovered to have a PFO after implantation, which was subsequently closed after reinitiating CPB.

Immediately following the placement of a single VAD, it is imperative to monitor function of both the recipient and nonassisted ventricle. In the HeartMate series, 11 patients undergoing LVAD placement required a subsequent right VAD upon the confirmation of RV failure by intraoperative TEE.¹⁷⁷ It is also important to document that the assisted chamber is being effectively unloaded and that the inflow cannula is properly positioned (Fig. 10-27). Of the 91 patients in this series, 3.3% of the cases required cannula repositioning based on TEE evidence. Finally, it is important to document the presence of intracavitary air following any open cardiac procedure because of the inherent risk of embolization. Significant "microair" was found in 91% of patients, which required halting of VAD flow until the air could be vented for several minutes.¹⁷⁷

View larger version (48K): [in this window] [in a new window]   FIGURE 10-27 A mid-esophageal two-chamber transesophageal echocardiographic view demonstrating appropriate positioning of the left ventricular (LV) inflow cannula (arrows) of a HeartMate ventricular assist device. The cannula inlet (top arrow) is aligned with transmitral inflow and not obstructed by the left ventricular walls. LA = left atrium.

	SAFETY AND COMPLICATIONS OF INTRAOPERATIVE TRANSESOPHAGEAL ECHOCARDIOGRAPHY

Top References

 
A discussion on the impact of a new technology on outcome is not complete without consideration for safety and potential complications. The evidence that TEE is a relatively safe procedure comes mainly from the cardiology literature, but there are several case series focusing on intraoperative complications. Initial safety data were reported in 1991 from a survey of 15 European centers, which included 10,419 TEE examinations performed from 1982–89.¹⁹⁰ In 201 patients (1.9%), the probe could not be placed in the esophagus due to operator inexperience (n = 198) and/or the presence of a tracheostomy (n = 1) or esophageal diverticulum (n = 2). In 88 awake patients (0.6%), the exam could not be completed due to either probe intolerance (n = 65) or non–life-threatening pulmonary (n = 8), cardiac (n = 8), or bleeding (n = 1) complications. The echocardiography exam had to be terminated prematurely due to vomiting in 5 patients. Finally, 1 patient who had an undiagnosed esophageal tumor died from a bleeding complication. Comparable complication rates have been reported in a large series of patients undergoing TEE examinations from the Mayo Clinic.^191,192 Among 15,381 patients in this series, 2 deaths, 3 esophageal tears, and 18 cases of hematemesis were reported. In addition, the incidence of hypoxia/laryngospasm (0.7%) and minor cardiovascular problems including superventricular tachycardia, hypotension, and hypertension (1%) was also noted.

Several case series have reported specifically on the low rate of significant complications associated with intraoperative TEE.^193–195 A 1993 study involving 846 intraoperative TEE examinations reported no deaths or esophageal perforations but did include 1 incidence of a chipped tooth, 3 pharyngeal abrasions, and 1 unilateral vocal cord paralysis.¹⁹⁵ A large case series retrospectively studied 7200 patients who underwent TEE in the operating room and reported no deaths that were directly attributed to the TEE and an overall 0.2% incidence of morbidity.¹⁹³ In 2 patients, the endotracheal tube was advanced into the left mainstem bronchus during the intraoperative study, resulting in transient hypoxia until the endotracheal tube was repositioned. No hemodynamic perturbations were reported in these anesthetized patients associated specifically with the TEE placement or examination. Swallowing abnormalities were reported in 7 patients postoperatively. Finally, 1 patient suffered an esophageal perforation that required surgical repair on the second postoperative day. Although esophageal perforation remains a relatively infrequent complication associated with TEE, at least 12 cases of perioperative TEE-associated pharyngeal or esophageal perforations have been reported in neonates and adult surgical patients with the vast majority of these cases involving patients undergoing cardiac surgical procedures.^{193,196–204} The high morbidity and mortality associated with this potential complication suggests that the placement of a TEE probe should be avoided in patients with esophageal or gastric pathology and in those patients in whom probe insertion or advancement within the esophagus is difficult.

The most frequent controversy in the cardiac surgical literature regarding intraoperative TEE complications focuses on a perceived association with postoperative dysphagia. An initial study in 1991, retrospectively reviewed 1245 cardiac surgical patients and reported a 1.8% incidence of dysphagia in the 217 patients who had a TEE placed compared to a 1.6% incidence in the 1028 patients in whom TEE was not performed.²⁰⁵ Although there was no statistical difference between the two populations, the authors concluded that the number of patients was insufficient to completely exclude the possibility of a very weak association between TEE and dysphagia. Hogue et al also investigated the association between intraoperative TEE and dysphagia in 879 cardiac surgical patients.²⁰⁶ Barium swallows demonstrating esophageal dysfunction were obtained in 34 patients who experienced dysphagia or coughing while drinking water following extubation. In comparison to the rest of the population, those patients with dysphagia were older, had lower CO, were more likely to require an intra-aortic balloon counterpulsation, had longer intubation times, and had a higher incidence of TEE probe placement (68% to 48%). Even though age, duration of tracheal intubation, and intraoperative TEE were independent predictors of swallowing dysfunction, the authors noted that confounding variables may have accounted for this relationship.

A retrospective chart review of 838 consecutive cardiac surgical patients also found 10 of 126 TEE-studied patients had postoperative dysphagia compared to 13 of 712 with dysphagia who did not have an intraoperative TEE.¹⁹⁵ Again, the patients who underwent TEE examination had longer intubation times (34.1 hours vs. 17.5 hours) and more frequently underwent valve or combined valve/coronary artery bypass graft procedures (60.3% vs. 20.7%) than the non-TEE cohort. In addition, the need for TEE was not assigned randomly but was determined at the discretion of the surgeon and/or anesthesiologist. Other smaller studies of 57 and 41 patients have not demonstrated this correlation between TEE and dysphagia or any esophageal injury.^194,207

Many studies may underestimate the true incidence of TEE-associated complications due to conservative definitions of clinically significant adverse events, limitations in standardizing the evaluation of patient's subjective complaints, and ascribing truly related complications to other etiologies. In addition, it is difficult to assess consequences associated with the "distraction factor" of intraoperative multitasking and patient care while performing intraoperative TEE. Morbidity associated with intraoperative TEE may be minimized by avoiding its use in patients with obvious contraindications to gastroesophageal manipulation, adhering to conservative judgment in abandoning probe insertion when significant resistance is encountered, and maintaining strict vigilance while continuing to observe and care for patients during the examination. The relatively low reported incidence of morbidity suggests that intraoperative TEE is a relatively safe diagnostic monitor that continues to serve as an invaluable tool in the perioperative management of cardiac surgical patients.