T gaining increasing popularity as a means of continuous

Value of Transesophageal Echocardiography During Repair of Congenital Heart Defects Maurizio Dan, MD, Raffaele Bonato, MD, Alessandro Mazzucco, MD, Uberto Bortolotti, MD, Giuseppe Faggian, MD, Gianpiero Giron, MD, and Vincenzo Gallucci, MD Departments of Anesthesiology and Cardiovascular Surgery, University of Padova Medical School, Padova, Italy Two-dimensional transesophageal color Doppler echocardiography was employed intraoperatively in 30 children undergoing repair of a variety of simple and complex cardiac malformations. There were 6 female and 4 male patients, with a mean age of 9 f 3 years (range, 4 to 3 years) and a mean weight of 3 f 9 kg (range, 6 to 50 kg), 6 children weighing less than 30 kg. A standard, commercially available transesophageal echocardiography probe (5 MHz, 64 elements) was used in all patients without complications. Transesophageal echocardiography proved helpful in selecting the surgical approach, in assessing the adequacy of surgical repair, in detecting residual intracardiac shunts, and in allowing uninterrupted monitoring of ventricular performance throughout the procedure. Our initial experience suggests that transesophageal echocardiography is a valuable tool to be used in children with congenital cardiac malformations, particularly in those requiring complex intracardiac procedures. The amount of information obtained by the surgeon should favor the routine use of transesophageal echocardiography during open heart procedures and stimulate the development of probes to be safely used even in infants and newborns. ( 990;50:63743) ransesophageal echocardiography (TEE) is currently T gaining increasing popularity as a means of continuous cardiac imaging during open heart operations [l, 2; its ability to monitor left ventricular function and wall motion abnormalities and to assess the adequacy of sura, gical repair is well defined in adult patients [3, 4. With the recent introduction of color-flow Doppler technology in transesophageal imaging, TEE has become even more suitable for intraoperative application and its indications are extended to all situations in which detection of abnormal intracardiac blood flow is needed [5]. For this reason patients with congenital heart defects should benefit most from the intraoperative use of TEE. Experience in this field, however, is at present very limited [6, 7, mainly because pediatric-size probes were not commercially available until very recently, while epicardial color flow imaging has received increasing acceptance for intraoperative application in children [8]. The latter, however, presents many limitations, such as the need for temporary discontinuation of the operative procedure, impossibility of continuous imaging, poor surface contact, heart compression with potential rhythm disturbances, and, above all, limited time allowed for evaluation [9]. We present here our initial experience with twodimensional (2-D) color Doppler TEE used intraoperatively in a series of children undergoing repair of various congenital heart malformations. This report focuses on Accepted for publication June, 990 Address reprint requests to Dr Dan, Istituto di Anestesiologia e Rianimazione, Universita di Padova, Via C. Battisti, 267, 352 Padova, Italy. the technical feasibility of 2-D TEE and its potential advantages for the cardiac surgeon. Material and Methods Patient Population During the interval October 988 to September 989, 2-D TEE was used intraoperatively in 30 children undergoing surgical repair of various congenital heart defects. The criteria to be enrolled in the study were a body weight preferably above 20 kg and absence of any documented pathology of the upper gastrointestinal tract. There were 6 female and 4 male patients, with a mean age at operation of 9 & 3 years (range, 4 to 3 years) and a mean weight of 3? 9 kg (range, 6 to 50 kg), 6 weighing less than 30 kg. Twenty-seven patients had primary repair of various intracardiac malformations, patient had reoperation for residual pulmonary stenosis and ventricular septa defect (VSD) due to patch detachment after correction of tetralogy of Fallot, had reoperation because of subaortic stenosis after repair of interrupted aortic arch and VSD, and had an open chest biopsy of a left ventricular hemangioma (Table ). The preoperative diagnosis was made with standard 2-D echocardiographic and hemodynamic investigations in 3 and by 2-D echocardiography alone in 7. With the exception of the patient undergoing left ventricular biopsy all children were operated on with standard cardiopulmonary bypass and moderate systemic hypothermia. Myocardial protection was achieved in all cases with the use of cold potassium cardioplegia combined with topical cool- 0 990 by The Society of Thoracic Surgeons 0003-4975/90/$3.50

638 DANETAL 990;50:63743 Table. Preoperative Diagnosis and Surgical Procedures No. of Disease Patients Operation Ostium I ASD + PAPVC +PS Ostium I ASD Fibrous SAS VSD + PS + A Aortic stenosis TOF +CAVC TGA, VSD, PS Single ventricle, CAVC Ebstein s anomaly TV VSD patch detachment, PSb Fibrous SAS, AIc LV hemangioma 5 5 4 2 2 ASD patch closure ASD patch closure, intraatrial tunnel ASD patch closure, pulmonary valvotomy ASD patch closure, mitral valvoplasty (3) Blunt resection, septal myotomy (2) VSD patch closure VSD patch closure, pulmonary valvotomy (l), right ventricular muscle band resection () VSD patch closure, aortic valvoplasty Commissurotomy and valvoplasty TOF repair with transannular patch TOF repair with transannular patch, CAVC repair with double patch Rastelli procedure with aortic homograft Modified Fontan, atrial partitioning Tricuspid valve replacement VSD closure, transannular patch Blunt resection Left ventricular biopsy a Underwent successful orthotopic heart transplantation because of biventricular failure after 20 hours of mechanical support. Reoperation after previous repair of TOF. Reoperation after previous repair of interrupted aortic arch and VSD closure. A = aortic incompetence; ASD = atrial septal defect; CAVC = complete atrioventricular canal; LV = left ventricular; PAPVC = partial anomalous pulmonary venous connection; PS = pulmonary stenosis; SAS = subaortic stenosis; TGA = transposition of great arteries; TOF = tetralogy of Fallot; TV = tricuspid valve; VSD = ventricular septal defect. ing. Details of the surgical techniques employed are summarized in Table. Technique of Two- Dimensional Transesophageal Echocard iography A 64-element, 5-MHz, commercially available phased array transducer (HP 2362 A, Hewlett-Packard Co, Andover, MA) interfaced with a Hewlett-Packard ultrasonograph (HP 77020 AC) was used in all cases. This system allows color-flow Doppler examination as well as highresolution 2-D imaging. Insertion of the probe was performed in the anesthetized patient after endotracheal intubation, either blindly or under direct visualization of the pharynx with the neck moderately flexed. Advancement of the generously lubricated probe was accomplished very gently in the flexible mode with the control unlocked. Tilting of the transducer tip in both the anteroposterior and lateral planes was intentionally avoided in patients weighing less than 30 kg. Transesophageal echocardiographic monitoring was started before skin incision and continued throughout the entire procedure, being temporarily suspended during cardiopulmonary bypass to minimize the risk of esophageal damage due to probe heating and temperature gradient during this period. At the end of each procedure, just before transfer of the patient from the operating room to the intensive care unit, the probe was removed. When indicated, 2-D TEE studies were also performed in the postoperative period while the patient was still intubated. Cardiac views were obtained from both the esophagus and stomach as described by Seward and associates (0. Longitudinal four-chamber or four-chamber and left ventricular outflow tract views were mainly used; additional views such as basal short-axis, transgastric short-axis, and transgastric four-chamber were also obtained when considered of particular interest for the study. For color-flow mapping long-axis views were almost exclusively used, as blood flow must be parallel to the Doppler source for optimal signal processing. Results Surgical Results There were two hospital deaths (less than 30 days). One 2-year-old boy died of septic shock 4 days after total repair of tetralogy of Fallot and complete atrioventricular canal, and a 2-year-old girl died of heart failure on postoperative day after a modified Fontan procedure with atrial partitioning for single ventricle and complete atrioventricular canal. Finally, a 2-year-old boy showed severe biventricular failure after tricuspid valve replacement for Ebstein s anomaly; he was supported with extracorporeal circulation for 20 hours before undergoing successful orthotopic heart transplantation. No major postoperative complications were observed in the remaining patients. Technical Aspects of Transesophageal Echocardiography Two-dimensional transesophageal echocardiography monitoring was performed 32 times, once in 28 patients and twice in 2. In the latter it was repeated in the intensive care unit within 24 hours after operation: in the patient with low output state after a modified Fontan procedure, to assess left ventricular performance, and in the patient with tetralogy of Fallot and complete atrioventricular canal repair to verify the effect of vasodilator administration on the degree of residual mitral incompetence. The probe was successfully inserted without problems in all cases and was left in place for the entire surgical procedure in all but. In this patient, the smallest of our series (a 6-year-old, 6-kg boy), the probe was withdrawn during the period of cardiopulmonary bypass. The quality of 2-D TEE was considered satisfactory in all cases,

990;50:637-43 DANETAL 639 Fig 2. (a) Transesophageal two-dimensional modified long-axis view in a 5-year-old, 22-kg boy undergoing reoperation for ventricular patch detachment farrow) after correction of tetralogy of Fallot. (b) Two-dimensional color-flow Doppler echocardiography after slight rotation of the probe detecting a mosaic jet from the angle between the left ventricular outflow tract (LVOT) and the tricuspid plane (arrow). (LA = left atrium; RA = right atrium; RV = right ventricle.) including smaller patients (less than 30 kg in weight) in whom probe manipulation was intentionally avoided. The addition of color-flow mapping, performed at a mean pulse repetition frequency of 3 to 4 KHz, was considered to represent a substantial improvement to the examination in all cases. Nevertheless, in some instances "ghost" color signals were derived from fast-moving cardiac structures instead of only from the bloodstream. This phenomenon was observed particularly when the heart rate exceeded 40 beats/min, as occurred in some of our children during weaning from cardiopulmonary bypass; in such cases time was allowed for rhythm stabilization while evaluation of color data was in the meantime performed through frame by frame analysis. Atrial Septum Both atria and the interatrial septum were always well visualized by a modified long-axis view, and adequacy of surgical repair after simple atrial septal defect patch closure was confirmed in all cases. Useful information also was provided by TEE, with the adjunct of color-flow Doppler, when an atrial septal defect was part of a more complex intracardiac malformation. In fact, in patient, who showed signs of marked arterial desaturation immediately after a modified Fontan procedure for single ventricle and complete atrioventricular canal, a partial detachment of the pericardial baffle used for atrial septation was detected, thus allowing immediate repair of this complication; in this particular patient, however, the atriopul-

640 DANETAL 990;50:637-43 Fig 2. (a) Preoperative transesophageal two-dimensional long-axis view of the left ventricular outflow tract (LVOT) in a 0-year-old, 37-kg boy showing a subaortic ridge (arrow). (b) Postoperative control in the same view demonstrating complete resection of the ridge. (LA = left atrium; LV = left ventricle.) monary anastomosis could not be effectively visualized. In another patient, in whom preoperative catheterization had shown a VSD and raised the suspicion of an atrial septa defect on the basis of oxymetric data, intraoperative TEE showed an intact interatrial septum; this finding was confirmed at direct inspection and the high oxygen saturation was shown to be caused by the combination of ventricular shunting and tricuspid regurgitation secondary to valve dysplasia. Ventricular Septum The ventricular septum was best visualized with a standard long-axis view. In all patients with perimembranous and malalignment VSD TEE was helpful to assess the adequacy of operation and to exclude any residual interventricular shunt. Particularly, in child undergoing reoperation after tetralogy of Fallot repair it allowed exact identification of the site of patch detachment, not clearly shown at angiography (Fig ). Left Ventricular Outflow Tract and Aortic Valve Both left ventricular outflow tract and aortic valve were always well identified through a long-axis view; in addition a basal short-axis view was also helpful in imaging the aortic valve. In patients with subaortic stenosis TEE allowed us to immediately verify the complete resection of the subaortic membrane and to exclude any iatrogenic complication such as a VSD or aortic regurgitation (Fig 2). Both aortic valve and left ventricular outflow tract were easily examined also in patients with aortic incompetence or stenosis to assess the early result of valve repair. Pulmonary Artery When needed, the main pulmonary trunk and right pulmonary artery were well visualized through a basal

990;5063743 DANETAL 64 Fig 3, Transesophageal two-dimensional basal short-axis view after closure of a ventricular septal defect in a 7-year-old, 24-kg boy showing stenosis of the origin of the right pulmonary artery (arrowheads). (A0 = aorta; PA = main pulmonary artery; RVOT = right ventricular outflow tract.) short-axis view. In patient stenosis of the right pulmonary artery could be demonstrated after VSD patch closure even though intraoperative measurements failed to reveal any clinically significant pressure gradient (Fig 3). However, in the child who had a Rastelli procedure for transposition of the great arteries the right ventricular to pulmonary artery conduit could not be visualized, nor could the right ventricular outflow tract and pulmonary valve in many patients. Atrioventricular Valves Both atrioventricular valves were always well identified in a long-axis, four-chamber view. Transesophageal echocardiography was of particular value in the evaluation of mitral anatomy and dynamics because of the favorable position of this valve with regard to the esophageal window; consequently, immediate evaluation of the adequacy of mitral valvoplasty was always possible. Furthermore, in the patient with tetralogy of Fallot and complete atrioventricular canal, TEE helped to assess the degree of residual mitral incompetence after repair of a severely dysplastic mitral component and to monitor the effect of vasodilator therapy in weaning the patient from cardiopulmonary bypass (Fig 4). Cardiac Performance Variations of cardiac function due to hypovolemia or hypocontractility were instantaneously recognized by TEE in all patients, allowing us to select the most proper treatment in each situation. However, quantitative analysis was often impractical mainly because of the hemodynamic instability of most patients with complex intracardiac repair early after cardiopulmonary bypass. Comment The clinical experience with the intraoperative use of TEE in children so far has been very limited. Recently, Cyran and co-workers (6 demonstrated the efficacy of this technique in children undergoing operation for congenital heart disease. They employed a 32-element, 3.5-MHz transducer mounted on a gastroscope with an -mm external diameter shaft and a 5-mm external diameter tip in all patients except, in whom they used a smaller and more technologically advanced second-generation probe. They demonstrated the feasibility of TEE in patients older than 7 years and the great potential of this technique in assessing the adequacy of surgical repair of complex cardiac lesions. In all patients in our series we used a second-generation TEE imaging system that has a higher number of elements, 64, with an emission frequency of 5 MHz and smaller dimensions (0-mm external diameter shaft and 2 x 4-mm external dimension tip) and incorporates color-flow Doppler echocardiography. These features have further enhanced the utility of intraoperative TEE in pediatric patients undergoing open heart operations by allowing a higher resolution in the near field and, what is more important, a better understanding of normal and abnormal intracardiac blood flow before and after correction. Transesophageal echocardiography was employed in 30 children between 4 and 3 years of age who had a mean weight of 3 * 9 kg and a variety of simple and complex cardiac malformations. We were able to study children with a body weight as low as 6 kg by employing probes of smaller dimensions when compared with those used by others [6]. Although subsequent endoscopic examination was not performed in any patient in this series, no apparent complications related to probe insertion were observed, most likely because some technical guidelines were routinely followed. These included generous lubrication of the probe, moderate nuchal flexion, direct laryngoscopic visualization in the presence of even minimal resistance to probe advancement into the pharynx, and

642 DANETAL 990;50:637-43 Fig 4. (a) Postoperative transesophageal two-dimensional four-chamber view in a 2-year-old, 50-kg boy with tetralogy of Fallot and complete atrioventricular canal. The two patches used to repair the septal defects are evident (arrows).(b) The same image with color Doppler shows a regurgitant jet (mosaic systolic pattern) into the left atrium due to residual mitral incompetence. (LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle.) temporary suspension of monitoring during cardiopulmonary bypass. Whether the probe should be withdrawn in children weighing less than 20 kg or during deep hypothermic circulatory arrest is still not clear; however, we believe the latter may be a useful precaution despite the safety of prolonged use of TEE as demonstrated experimentally by the absence of injury to the esophageal mucosa even in small animals [ll]. The results of this study, although performed in a group of patients with a wide spectrum of cardiac anomalies, show the potential advantages of the intraoperative use of TEE. Transesophageal echocardiography is probably not needed when dealing with simple congenital heart defects such as an atrial septal defect; nevertheless, application of this technique in these patients has helped to achieve enough expertise, especially from the technical point of view, enabling us to confidently employ it in more complex cases and in progressively smaller children. The benefits of TEE, greatly enhanced by the addition of color-flow mapping, are particularly evident for patients with complex intracardiac anatomy who require elaborate surgical procedures. Intracardiac flow abnormalities, for example, can be easily detected, thus rendering TEE equivalent to an intraoperative angiography. This appears particularly helpful in detecting the presence and exact site of any residual intracardiac shunt and the amount of any transvalvular reflux, thus allowing the surgeon to make the most appropriate decision for each individual patient before discontinuation of cardiopulmonary bypass. This occurred in of our patients in whom partial detachment of the atrial partitioning patch was detected after a modified Fontan procedure. Transesophageal echocardiography permits uninterrupted monitoring of cardiac performance throughout the

990;5063743 DANETAL 643 surgical procedure and therefore evaluation of the effectiveness of any inotropic or vasoactive drug administered [2]. This clearly represents an advantage over intraoperative epicardial 2-D echocardiography as poor surface contact and heart compression with consequent arrhythmias are avoided and temporary discontinuation of the operation is not required. Nevertheless, TEE has some deficiencies that at the present time may be represented mainly by the limited number of scanning planes, which does not allow in most cases a clear visualization of structures such as the right ventricular outflow tract and the pulmonary valve. Furthermore, despite the fact that in all our patients TEE monitoring was judged to be satisfactory, in some cases, particularly when the heart rate rose above 40 beats/min, color signals derived from fast-moving cardiac structures temporarily impaired the quality of images. The low signal to noise ratio observed in such situations may probably be solved with more sophisticated probes especially designed for small children, as recently reported by Kyo and colleagues [7]. Our preliminary data suggest that TEE is a valuable tool to be used intraoperatively in children during open heart operations, particularly during correction of complex congenital heart malformations. From this experience it appears that a standard, commercially available, adult probe can be safely used in children weighing more than 20 kg. However, the amount of information obtained by the surgeon with this technique should favor its routine use and stimulate development of probes to be employed even in small infants and newborns. References. Kremer P, Cahalan M, Beaupre P, et al. Intraoperative monitoring using transesophageal 2-dimensional echocardiography. Anaesthesist 985;34:-7. 2. Goldman ME, Mindich BP. Intraoperative 2-dimensional echocardiography. New application of an old technique. J Am Coll Cardiol 986;7:37&82. 3. Kyo S, Takamoto S, Matsumura M, et al. Immediate and early postoperative evaluation of results of cardiac surgery by transesophageal two-dimensional Doppler echocardiography. Circulation 987;76(Suppl 5):3-2. 4. Shiller NB. Evaluation of cardiac function during surgery by transesophageal 2-dimensional echocardiography. In: Hanrath P, Bleifeld W, Souquet J, eds. Cardiovascular diagnosis by ultrasound: transesophageal, computerized, contrast, Doppler echocardiography. The Hague: Martinus Nijhoff, 982:289-93. 5. De Bruijn NP, Clements FM, Kisslo JA. Intraoperative transesophageal color flow mapping: initial experience. Anesth Analg 987;66:386-90. 6. Cyran SE, Kimball TR, Meyer RA, et al. Efficacy of intraoperative transesophageal echocardiography in children with congenital heart disease. Am J Cardiol 989;63:594-8. 7. Kyo S, Koike K, Takanawa E, et al. Impact of transesophageal Doppler echocardiography on pediatric cardiac surgery. Int J Card Imag 989;4:4-2. 8. Ungerleider R. Decision making in pediatric cardiac surgery using intraoperative echo. Int J Card Imag 989;4:33-5. 9. Sutherland GR, van Daele MERM, Stumper OFW, et al. Epicardial and transesophageal echocardiography during surgery for congenital heart disease. Int J Card Imag 989;4: 3740. 0. Seward JB, Khandheria BK, Oh JK, et al. Transesophageal echocardiography: technique, anatomic correlations, implementation and clinical applications. Mayo Clin Proc 988;63: 649-80.. OShea JP, DAmbra MN, Magro C, et al. Transesophageal echocardiography: is it safe to the esophagus? An in vivo study [Abstract]. Circulation 988;78(Suppl 2):440. 2. Switzer DF, Nanda NC. Doppler color flow mapping. Ultrasound Med Biol 985;:403-6.