Recent technical advances and increasing experience

Transcription

1 Pediatric Open Heart Operations Without Diagnostic Cardiac Catheterization Jean-Pierre Pfammatter, MD, Pascal A. Berdat, MD, Thierry P. Carrel, MD, and Franco P. Stocker, MD Division of Pediatric Cardiology, and Department of Cardiovascular Surgery, University Hospital, Berne, Switzerland Background. Echocardiography allows for an adequate noninvasive preoperative evaluation in an increasing proportion of pediatric patients before open heart operations. The present study assessed the diagnostic accuracy of both invasive and noninvasive preoperative evaluation in children with congenital heart disease. Methods. We prospectively evaluated the accuracy of preoperative noninvasive or invasive diagnostic methods. Preoperatively established diagnosis was compared with the intraoperative diagnosis made by surgical inspection and routine perioperative transesophageal echocardiography. Results. During the study period of 30 months (ending in December 1997) 209 open-heart procedures were performed. Eighty-one patients (39%) were in the first year of life at the time of surgery, and 43% of all patients had symptoms. Noninvasive preoperative diagnosis using echocardiography was done exclusively in 142 patients (68%). Of the 67 children who had preoperative catheterization, 4 (6%) showed an additional intraoperative finding that modified the surgical approach in 2 of them. In the 142 patients who had echocardiographic preoperative assessment, the surgeons were confronted with a previously undiagnosed finding in 12 patients (8.5%). The finding was considered significant (prolongation of cardiopulmonary bypass time) in 2 patients and might have affected the outcome in 1 of them, a neonate with transposition of the great arteries and a preoperatively undiagnosed intramural coronary artery, who died postoperatively despite a technically adequate repair. Conclusions. In many infants and children, diagnostic work-up before open heart operations could be adequately based on an exclusively noninvasive basis by relying on echocardiography alone. (Ann Thorac Surg 1999;68:532 6) 1999 by The Society of Thoracic Surgeons Accepted for publication Feb 9, Address reprint requests to Dr Pfammatter, Division of Pediatric Cardiology, Children s Hospital, Freiburgstrasse, CH-3010 Berne, Switzerland. Recent technical advances and increasing experience in echocardiography allow for a detailed description of intracardiac anatomy. With Doppler analysis of blood flows, accurate estimation of hemodynamic characteristics has become possible in most patients, and especially in children. Early in the use of echocardiography, it was studied whether in certain circumstances echocardiography alone would enable adequate preoperative planning of cardiac corrective or palliative operations in children with congenital heart defects. Initially, echocardiography as the only preoperative diagnostic imaging modality had been proposed for more simple procedures, such as closure of atrial septal defects (ASD) [1]. More recently, exclusively noninvasive preoperative diagnostic work-up has been extended to more complex cardiac lesions, such as complete atrioventricular canal or selected patients with tetralogy of Fallot [2, 3]. The extent of preoperative diagnostic procedures must be based on a consensus between the pediatric cardiologist and the operating surgeon. At our institution, this consensus has led to a predominantly noninvasive preoperative evaluation for even complex cardiac defects in infants and children. In the present study, we prospectively assessed the accuracy of the preoperative diagnostic evaluation by echocardiography alone or by combined echocardiography and catheterization, in a population of unselected, consecutive pediatric patients who had open heart operations. Patients and Methods Between July 1995 and December 1997, all infants and children admitted at our institution for surgical palliation or correction of congenital heart disease using extracorporeal bypass were evaluated prospectively for the present study. Preoperative Diagnostic Procedures All patients had the standard preoperative evaluation according to the requirements of the surgical procedure and based on the institutional policy regarding preoperative diagnostic work-up requiring consensus between cardiovascular surgeons and pediatric cardiologists. The extent of preoperative evaluation was unaffected by the ongoing study and the diagnostic work-up was usually completed at the time of hospital admittance for the operation. Diagnostic evaluation consisted of complete two-dimensional and color Doppler echocardiography, This article has been selected for the open discussion forum on the STS Web site: by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)

2 Ann Thorac Surg PFAMMATTER ET AL 1999;68:532 6 PEDIATRIC OPERATIONS WITHOUT CATHETERIZATION 533 Table 1. Patient Characteristics and Distribution of Cardiac Defects Diagnosis n Catheterization n (%) Mean Extracorporeal Bypass Time (min) Mean Aortic Cross-Clamp Time (min) Deaths Atrial septal defect 49 0 (0) Atrial septal defect and partial 8 5 (62) anomalous pulmonary vein connection Ventricular septal defect 33 4 (12) Complex ventricular septal 9 4 (44) defect Complete atrioventricular 26 3 (11) canal Partial atrioventricular canal 6 1 (17) Univentricular heart 10 9 (90) Tetralogy of Fallot (100) Double-outlet right ventricle 4 4 (100) Left outflow tract or aorta 13 2 (15) Right outflow tract or 6 5 (83) pulmonary artery Transposition of great arteries 14 3 (21) Complex transposition 3 3 (100) Total anomalous pulmonary 3 3 (100) vein connection Conduit replacement 4 4 (100) Other 5 1 (20) Total (32) 8 chest roentgenogram, and a 12-lead electrocardiogram in all cases. Based on these data, the team of surgeons and cardiologists decided individually for each case on the need for preoperative cardiac catheterization. The institutional consensus between cardiovascular surgeons and pediatric cardiologists during the study period defined was that cardiac catheterization was required before palliative procedures in children with univentricular hearts (except for newborns with hypoplastic left heart syndrome, where institutional policy discouraged operation), before repair of complex transposition, tetralogy of Fallot, and before replacement of conduits or operations on the pulmonary arteries. In all other cases, preoperative echocardiographic evaluation alone was considered adequate, but the decision was always made on an individual basis. All preoperative echocardiographic examinations were done by one of two pediatric cardiologists using an Acuson 128 XP/10 machine (Mountain View, CA) with a 7.5- or 5-MHz transducer according to age and weight of the patient. The detailed preoperative diagnosis was defined at the surgical briefing the day before the surgical intervention. The intraoperative diagnosis was obtained during surgical inspection and by routine perioperative transesophageal echocardiography (except in neonates because of the lack of an adequately sized echo-probe at the beginning of the study) using an Acuson 128 XP/10 machine with a 5-MHz biplane transesophageal transducer. Surgical Technique Throughout the study period, the same two surgeons did all cardiac procedures in the patients. All operations were performed using extracorporeal bypass, and myocardial protection was done using intermittent anterograde cold blood cardioplegia. Except for closure of ASD (normothermia) all other procedures were done with the child in hypothermia between 24 C and 28 C. Perioperative death was defined as death within 30 days postoperatively. Results Patient Characteristics During the study period, a total of 209 open heart procedures were done at our institution. The age of the patients was between 0.1 and 18 years (mean, 4.4 years); 23 patients were newborns and 81 were in their first year of life. Weight at operation was between 2 and 66 kg (mean, 17.6 kg). Of all the children operated on 91 (43%) were symptomatic at the time of surgical correction, including 80% (65 of 81) of the infants and 19% (26 of 128) of the older children. Preoperative Diagnostic Procedures Table 1 shows the distribution of cardiac defects in the study population, with the respective proportions of children who had preoperative invasive diagnosis. Of all

3 534 PFAMMATTER ET AL Ann Thorac Surg PEDIATRIC OPERATIONS WITHOUT CATHETERIZATION 1999;68:532 6 Table 2. Age Dependency of Preoperative Diagnostic Procedures Echocardiography Alone Preoperative Total Number Catheterization of Patients Infants 62 (77%) 19 (23%) 81 (39%) ( 1 year) Children 80 (63%) 48 (37%) 128 (61%) ( 1 year) Total 142 (68%) 67 (32%) 209 (100%) 209 patients, 67 (32%) had preoperative cardiac catheterization, whereas most patients (142 of 209 children, 68%) had echocardiography alone. In children with ASD, none of the 49 patients with isolated secundum ASD had catheterization, whereas 1 of 6 children with primum type ASD had catheterization (for suspected pulmonary hypertension), and 5 of 8 children with anomalous pulmonary veins were also studied invasively. Ventricular septal defect (VSD) was treated predominantly noninvasively (in 88% of the 33 children with isolated VSD and in 56% of the 9 children with complex forms of VSD). Eleven of 13 (85%) had operations in the left outflow tract or in the aorta also were diagnosed solely on a noninvasive basis. Of the children with more complex cardiac defects, newborns with transposition of the great arteries were operated on with echocardiographic diagnosis alone (11 of 14, 79%). In case of the need to proceed to a Rashkind atrioseptostomy, we preferred to do that procedure in the intensive care unit under echocardiographic control (n 6), and the children were taken to the catheterization laboratory only in case of failure to proceed by the umbilical route. Whereas none of the 22 infants with complete atrioventricular canal who were younger than 1 year old had catheterization before surgical correction, those older than 1 year at the time of diagnosis (n 3) had invasive hemodynamic testing for assessment of pulmonary vascular reactivity. Of the rare cardiac malformations, one newborn each with truncus arteriosus and aortopulmonary window were successfully operated on after only echocardiographic preoperative evaluation. In one infant with a univentricular heart, a Damus-Kaye- Stansel procedure was done without preoperative catheterization. For infants and children with double-outlet right ventricle or tetralogy of Fallot, catheterization before surgical correction was the rule (100% of patients). The proportion of children who had catheterization was dependent not only on the type of cardiac anomaly but also on the patient s age as shown in Table 2. Only 23% (19 of 81) of newborns and infants less than 1 year old had invasive diagnostic procedures, whereas catheterization was done in 37% (48 of 128) of older children ( p 0.01). Perioperative mortality for the whole series of pediatric patients was 3.8% (Table 1) or 8 of 209, of which 4 were infants and 4 older children. Table 1 also shows a comparison of procedural times. Incomplete Diagnosis After Invasive Preoperative Diagnostics Of 67 children who had preoperative catheterization, an additional undiagnosed finding was discovered intraoperatively in 4 (6%). In two cases the surgical plan had to be modified. In one child with a univentricular heart and heterotaxy syndrome, suspected but undiagnosed total anomalous pulmonary vein connection had to be repaired before the planned bidirectional shunt. In the other child who was preoperatively diagnosed as having tetralogy of Fallot, the VSD had to be enlarged for the presence of double-outlet right ventricle. Both patients had a good outcome. In the other 2 children, the additional finding was thought to be of minor importance (presence of a large ductus in a patient with VSD and pulmonary hypertension, and presence of double-outlet right ventricle instead of simple tetralogy but without the need for enlarging the VSD). Incomplete Diagnosis in Children Diagnosed by Echocardiography Alone Of 142 patients in whom echocardiography was the only preoperative imaging modality, 12 (8.5%) had an additional intraoperative finding. In children with secundum ASD, on two occasions, an additional small defect was seen, 2 more children had one pulmonary vein draining into the superior vena cava, and one patient had a persistent left superior vena cava. All these additional findings did not affect the surgical approach or the outcome. In 3 infants with complete atrioventricular canal, an additional intraoperative observation was made. One child had a large ductus, in 1 child only four instead of five leaflets of the common atrioventricular valve were found, and in 1 patient there was no ventricular component of the defect and the diagnosis in fact was partial atrioventricular canal. These additional findings did not affect the outcome and were not considered unexpected by the surgeons. The case of the child with absence of the VSD in presumed atrioventricular canal was the only instance in the series where the surgeon did not find a malformation diagnosed previously in the echocardiogram. In 3 newborns with transposition of the great arteries, the diagnosis had to be completed intraoperatively or postoperatively. In one patient an additional small muscular VSD, which closed spontaneously, was observed postoperatively. In 2 patients, a coronary anomaly was not observed on echocardiography. In the first patient the left anterior descending coronary artery originated from the right coronary artery and was treated surgically without technical difficulties. In the other newborn, the left anterior descending coronary artery was diagnosed intraoperatively as taking an intramural course; it was treated surgically with only short prolongation of the procedural time, but the child died postoperatively. At autopsy there was a large area of ischemia although the coronary artery was patent. For comparison, a coronary

4 Ann Thorac Surg PFAMMATTER ET AL 1999;68:532 6 PEDIATRIC OPERATIONS WITHOUT CATHETERIZATION 535 Fig 1. Evolution of the ratio between diagnostic catheterizations and open heart procedures during the past 18 years at our institution. anomaly was diagnosed correctly by echocardiography in 4 other newborns with transposition of the great arteries. In one older patient with valvular aortic stenosis, it was noticed only during an operation for valve replacement that the ascending aorta was dilated to such an extent that it had to be replaced at the same time. Figure 1 shows the evolution of noninvasive diagnostic procedures in our unit. In the era before echocardiography, the ratio between diagnostic catheterizations and open heart procedures was After introduction of echocardiography, that ratio decreased to 1. With the introduction of color Doppler technology that ratio further decreased to its recent level of 0.34 in Comment Open heart procedures in infants and children carry inherent risks, even in the best surgical hands. Because the length of open heart operations is an important determinant of successful outcome, unexpected intraoperative findings should be carefully avoided by a proper preoperative diagnostic evaluation. However, cardiac catheterization, especially in neonates and small infants, also carries a risk. Therefore many children with congenital heart disease would benefit if a preoperative evaluation by purely noninvasive methods could produce a diagnostic yield comparable to that of invasive diagnostic assessment and if the surgical outcomes of these patients with only echocardiography-based diagnostic work-up would not differ substantially from the outcomes of children who had preoperative catheterization. The main finding of this prospective evaluation of preoperative diagnostic procedures was that safety of pediatric open heart operations steadily improved despite the increasing use of only echocardiography as the primary preoperative imaging technique and the decreasing proportion of children who had preoperative catheterization. These results are highlighted by the fact that perioperative mortality rate decreased from 8% (in 1988) to its recent level of 4% during the study period, although the operative complexity increased. Soon after echocardiography had been established as a reliable diagnostic tool in pediatric cardiology, it was shown that in the simpler congenital cardiac malformations such as ASD it was possible to successfully correct them by relying mainly on echocardiography for the diagnostic evaluation rather than preoperative cardiac catheterization [1]. Concurrently it was shown that for ASD the diagnostic accuracy of echocardiography did not differ significantly from that of invasive diagnostic evaluation. Other authors showed that abandoning preoperative cardiac catheterization in children with a wider spectrum of congenital cardiac defects did not influence the surgical outcome negatively; however, these results were obtained in relatively small series of patients and included many children who had closed heart procedures [4, 5]. In one large study it was shown that, especially in infants, developments in echocardiography allowed for an increasing number of patients to be treated adequately without preoperative catheterization [6]. That study was undertaken shortly before the introduction of color Doppler technology. The proportion of children who had exclusively noninvasive diagnostic examination for an open heart procedure was 23% in that population. The present study assessed the contemporary diagnostic accuracy of color Doppler echocardiography in a population of consecutive pediatric patients who had open-heart procedures. During the study period, the proportion of patients who did not have catheterization before surgery was 68%, which again showed a significant evolution compared to the 23% of patients not catheterized in the late 1980s [6]. However, a direct comparison of these figures might not be accurate because institutional policies could differ significantly. Simple cardiac defects such as ASD or VSD, where pathologic-echocardiographic correlations have led to a thorough understanding of surgical anatomic significance of echocardiographic images [7], and more complex cardiac lesions, such as transposition of the great arteries or complete atrioventricular canal, were treated adequately during the study period by using predominantly echocardiography alone as the preoperative imaging mode. In complete atrioventricular canal, there is no advantage of angiography over echocardiography with regard to intracardiac anatomy. In infants younger than 1 year of age it was shown that successful surgical correction was possible in patients with severe elevation of pulmonary vascular resistance as calculated from catheterization data [2]. On the basis of our own results in children with that cardiac lesion we currently consider catheterization an unnecessary risk in infants. In newborns with transposition of the great arteries, the intracardiac anatomy was adequately shown by echocardiography. The anatomy of the coronary arteries could be shown with a very high accuracy by echocardiography, even in the presence of intramural coronary arteries [8, 9]. In the present study an anomaly of the coronary arteries was correctly predicted in most patients, although in one newborn, an intramural coronary artery

5 536 PFAMMATTER ET AL Ann Thorac Surg PEDIATRIC OPERATIONS WITHOUT CATHETERIZATION 1999;68:532 6 was not diagnosed preoperatively, which might have affected the patient s outcome. Now that the Rashkind procedure can be done safely in the pediatric intensive care unit by echocardiographic guidance [10], coronary angiography should not be considered a prerequisite to a successful arterial switch operation, except when institutional policy considers an intramural coronary artery a contraindication to the arterial switch operation. The present study was not undertaken to compare the diagnostic accuracy of either invasive or exclusively noninvasive preoperative diagnostic assessment, because of the spectrum of cardiac defects in the two groups. References 1. Freed MD, Nadas AS, Norwood WI, Castaneda AR. Is routine preoperative cardiac catheterization necessary before repair of secundum and sinus venosus atrial septal defects? J Am Coll Cardiol 1984;4: Zellers TM, Zehr R, Weinstein E, Leonard S, Ring WS, Nikaidoh H. Two-dimensional and doppler echocardiography alone can adequately define preoperative anatomy and hemodynamic status before repair of complete atrioventricular septal defect in infants 1 year old. J Am Coll Cardiol 1994;24: Santoro G, Marino B, Di Carlo D, et al. Echocardiographically guided repair of tetralogy of Fallot. Am J Cardiol 1994; 73: Huhta JC, Glasow P, Murphy DJ, et al. Surgery without catheterization for congenital heart defects: management of 100 patients. J Am Coll Cardiol 1987;9: Krabill KA, Ring S, Foker JE, et al. Echocardiographic versus cardiac catheterization diagnosis of infants with congenital heart disease requiring cardiac surgery. Am J Cardiol 1987; 60: Sreeram N, Colli AM, Monro JL, et al. Changing role of non-invasive investigation in the preoperative assessment of congenital heart disease: a nine year experience. Br Heart J 1990;63: Gatzoulis MA, Li J, Ho SY. The echocardiographic anatomy of ventricular septal defects. Cardiol Young 1997;7: Pasquini L, Sanders SP, Parness IA, et al. Coronary echocardiography in 406 patients with transposition of the great arteries. J Am Coll Cardiol 1994;24: Pasquini L, Parness IA, Colan SD, Wernovsky G, Mayer JE Jr, Sanders SP. Diagnosis of intramural coronary artery in transposition of the great arteries using two-dimensional echocardiography. Circulation 1993;88: Ward CJB, Hawker RE, Cooper SG, et al. Minimally invasive management of transposition of the great arteries in the newborn period. Am J Cardiol 1992;69: INVITED COMMENTARY Cardiovascular operations without preoperative cardiac catheterization began two generations ago, when physical findings, electrocardiogram, and chest roentgenogram were the sole preoperative evaluation methods in children with an atrial septal defect, patent ductus arteriosus, or coarctation of the aorta. In the next generation of pediatric cardiologists, this practice spread to more complicated lesions by combining anatomic data from two-dimensional echocardiographic imaging with flow and pressure dynamics from Doppler echocardiography. This technique required careful echocardiographic studies tracing every detail of cardiovascular anatomy. Cardiac catheterization was done in fewer patients until interventional procedures were developed. The present generation of pediatric cardiologists is trained in both technologies, and the two methods are used in combination. Noninvasive evaluation now determines when cardiac catheterization must be used. This reduces risk by shortening procedures and decreasing contrast load. A marriage of the two technologies, with cardiac catheterization for selected indications, provides the safest preoperative evaluation. The issue, then, is defining when cardiac catheterization is needed. Echocardiographic evaluation is only as good as the information available (garbage in, garbage out). Accuracy depends on high-quality images (top-ofthe-line equipment with high-frequency probes, good penetration which makes the newborn infant an ideal candidate, and a cooperative patient) and the experience of the pediatric sonographer and echocardiographer. When necessary information cannot be determined noninvasively, cardiac catheterization should be used to provide it. Each institution must have its own approach, as determined by physician experience and resources. It is important that the approach to each patient is controlled by medical indications determined by a team of cardiologists and surgeons. There is no place for pressure from competition or third-party payers in decisions on whether to perform an invasive evaluation. Studies defining the accuracy of noninvasive evaluation are important. Pfammatter and associates reported the accuracy of preoperative diagnosis in 142 patients who had clinical and echocardiographic assessment and 67 who also had diagnostic cardiac catheterization. An undiagnosed finding was discovered perioperatively in 8.5% of patients who had noninvasive procedures only and in 6% of patients who also had catheterization. An exact comparison between groups cannot be made, because patients were not randomly assigned to treatment, but accurate noninvasive preoperative evaluation is clearly possible in many patients. The surgeon and cardiologist must recognize the limitations of preoperative evaluation both of the techniques and quality of information and with consideration of the type of congenital heart disease. Some conditions might be difficult to detect using either approach, and an unrecognized patent ductus arteriosus with high pulmonary artery pressure or an intramural coronary artery might have to be dealt with perioperatively. John L. Bass, MD Division of Pediatric Cardiology University of Minnesota Box 94 Mayo, 420 Delaware St SE Minneapolis, MN by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)