Aortic Root Replacement With the Pulmonary Autograft in Children With Complex Left Heart Obstruction

Aortic Root Replacement With the Pulmonary Autograft in Children With Complex Left Heart Obstruction Vaughn A. Starnes, MD, Giovanni Battista Luciani, MD, Winfield J. Wells, MD, Rosemary B. Allen, BSN, and Alan B. Lewis, MD Divisions of Cardiothoracic Surgery and Cardiology, Children's Hospital Los Angeles, and Departments of Surgery and Pediatrics, USC School of Medicine, Los Angeles, California Background. The optimal surgical treatment of complex (multiple level or recurrent) left ventricular outflow tract obstruction (LVOTO) in infancy is controversial. Staged procedures expose the children to the need for reoperation, and currently available techniques of aortoventriculoplasty are associated with the morbidities of biological and mechanical prostheses. Methods. Between July 1992 and January 1996, we have performed 24 aortic root replacements with the pulmonary autograft in pediatric patients (<18 years). Of this group, 8 were infants and children with complex LVOTO aged 9 days to 22 months (mean, 8.6 ± 8 months) and weighing 3.3 to 10.2 kg (mean, 6.3-2.6 kg). The diagnoses were interrupted aortic arch/ventricular septal defect/subaortic stenosis in 3, recurrent aortic stenosis in 2, aortic stenosis and subaortic stenosis in 1, and aortic stenosis/subaortic stenosis/mitral stenosis/regurgitation in 2. All patients had undergone one to three previous operative procedures (mean, 1.5 ± 0.8 procedures/ patient). Preoperative echocardiographic peak LVOT gradient was 71.7 + 25 mm Hg (range, 40 to 110 mm Hg) and aortic annulus size was 7.2 ± 2.3 mm (range, 4 to 10.6 mm). The surgical technique included replacement of the aortic root with the pulmonary autograft combined with incision of the conal septum to relieve subaortic stenosis or accommodate for size discrepancy between the aortic and pulmonary autograft root and a pulmonary homograft placed in the right ventricular outflow tract. Results. There were no perioperative or late deaths at follow-up (range, 2 to 25 months; mean, 13.5 ± 8 months). Mean hospital stay was 15 -+ 17 days (range, 4 to 53 days). Three children had the following complications: diaphragmatic paresis (1), delayed pericardial effusion (1), and atrioventricular block requiring a pacemaker (1). In follow-up, echocardiographic findings showed absent aortic regurgitation in 3 and trivial aortic regurgitation in 5, and no significant LVOTO (mean peak gradient, 6.2 ± 7.6 mm Hg; range, 0 to 16 mm Hg). Pulmonary homograft regurgitation was absent in 5, trivial in 2, and moderate in 1. Peak right ventricular outflow tract gradient by echocardiogram was trivial in 7, and a significant gradient of 55 mm Hg has developed in 1 infant. There were no infective or embolic complications during follow-up. Conclusions. Our experience shows that aortic root replacement with the pulmonary autograft can be performed in children with excellent clinical results. The technique of root replacement combined with ventriculoplasty allows definitive and simultaneous relief of complex and multiple-level obstructive lesions. Considering the growth potential of the pulmonary autograft, this should be regarded as the optimal treatment modality in infants with complex LVOTO. (Ann Thorac Surg 1996;62:442-9) hildren presenting with complex left ventricular C outflow tract obstruction (LVOTO) represent a challenging medical and surgical problem. The treatment modalities are often palliative because the initial therapy frequently leaves residual lesions. More definitive approaches have involved valve replacement with mechanical prostheses, which require anticoagulation, or bioprostheses (aortic homograft), which have limited durability and no growth potential. Based on our positive experience with the Ross procedure in older patients, we Presented at the Thirty-second Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Jan 29-31, 1996. Address reprint requests to Dr Starnes, Division of Cardiothoracic Surgery, USC School of Medicine, Childrens Hospital Los Angeles, 4650 Sunset Blvd MS 66, Los Angeles, CA 90027. have extended the technique of aortic root replacement with the pulmonary autograft to the population of children less than 24 months of age with complex left heart obstruction. The early results with this procedure are presented. Material and Methods Patients Between July 1992 and January 1996, 24 pediatric patients (<18 years of age) underwent replacement of the aortic root with the pulmonary autograft at Children's Hospital Los Angeles (Fig 1). Eight patients (33%) were children aged less than 24 months with complex left heart obstruction, defined as the coexistence of multiple-level or recurrent obstructive lesions of the left heart. 1996 by The Society. of Thoracic Surgeons 0003-49751961515.00 Published by Elsevier Science Inc PII S0003-4975(96)00384-0

Ann Thorac Surg STARNES ET AL 443 1996;62:442-9 PULMONARY AUTOGRAFT IN INFANCY J lll.iij... I Fig 1. Histogram showing the age distribution of! children (n = 24) undergoing aortic root replacement with the pulmonary autograft between July... 1992 and January 1996. <I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Aga (yours) Demographics There were 5 male and 3 female patients, with ages ranging from 9 days to 22 months and weights ranging from 3.3 to 10.2 kg (Table 1). Previous Procedures Patients had presented in the neonatal period with a complex obstructive lesion of the left heart and underv~ent an invasive procedure at a mean of 5.1 days of life (range, 2 to 16 days). There were 1.5 invasive or surgical procedures per patient (range, 1 to 3) before aortic root replacement with the pulmonary autograft (see Table 1). Clinical Status Patients were in New York Heart Association class III or IV before their aortic root replacement. Anatomic Diagnosis The primary disease condition was interrupted aortic arch/ventricular septal defect/subaortic stenosis (IAA/ VSD/SAS) in 3, recurrent aortic stenosis on bicuspid valve in 3, and critical aortic stenosis with mitral regurgitation in 2 (Table 2). Two patients with IAA/VSD/SAS presented with aortic valve regurgitation, persistence of interventricular communication, and arch obstruction. The third infant with IAA/VSD/SAS had undergone one-stage neonatal repair at our institution [1]. This patient also had a mildly stenotic bicuspid aortic valve with hypoplastic annulus, which progressed toward a moderate-to-severe lesion within 12 months. The 3 infants with critical neonatal aortic stenosis had received balloon valvuloplasty without lasting relief of the obstruction. Two patients with complex LVOTO and mitral regurgitation had a previous balloon valvuloplasty before the aortic root replacement (see Table 2). Preoperative Assessment Transthoracic echocardiography was used for initial evaluation in all patients, and subsequent catheterization and angiographic examination was done. Diameter of the aortic annulus ranged from 4 to 10.6 mm (mean, 7.2-2.3 mm) as measured by echocardiography, whereas cumulative LVOTO peak pressure gradient by echocardiography ranged from 40 to 110 mm Hg (mean, 71.7-25 mm Hg). No gradient across the right ventricular outflow tract (RVOT) was demonstrated in 7 patients, whereas I with palliated IAA/VSD/SAS presented a peak gradient of 40 mm Hg due to the presence of a pulmonary band (Table 3). All patients had intraoperative transesophageal echocardiographic examination. Operative Technique Cardiopulmonary bypass was established using one aortic and two caval cannulas in 6 patients. Two cases in which a period of deep hypotherrnic (18 C rectal temperature) circulatory arrest was used for repair of the aortic arch had a single right atrial cannula. The remaining patients (n = 6) were cooled to a rectal temperature of 22 to 24 C before aortic cross-clamping. An initial dose of cold (0 to 4 C) blood cardioplegia was administered in the aortic root, before transection of the aorta well above the valve commissures. Continuous cold (0 to 4 C) saline perfusion of the pericardial sac was performed during the operation to maintain myocardial temperature between 4 and 8 C. In the children having reconstruction of the aortic arch (Table 4), the continuity of the arch was reestablished with an end-to-end posterior anastomosis of the native aorta and anterior augmentation with a pulmonary homograft after resection of the stenotic segment (recurrent coarctation in 2 patients, stenotic polytetrafluoroethylene graft in 1). The aortic valve was then excised and the two coronary buttons harvested. The main pulmonary artery was transected at the bifurcation and completely separated posteriorly from the left main coronary artery (Fig 2). The anterior surface of the right ventricular infundibulum was then incised transversely 4 Table 1. Patient Demographics" Variable Range Mean -- SD Median Age (mo) 0.3-22 8.6 + 8 6 Weight (kg) 3.3-10.2 6.3 + 2.6 5.9 BSA (m 2) 0.22-0.495 0.346 -- 0.11 0.341 Previous procedures 1-3 1.5 z 0.8 1 (no./patient) Age at first procedure 2-16 5.1 -- 5 3 (days) Interval from last procedure 0.25-18 7.4 7.9 4.6 to Ross (months) " n - 8.5 male and 3 female. BSA = body surface area; standard deviation. F = female; M male; SD -

444 STARNES ET AL Ann Thorac Surg PULMONARY AUTOGRAFT IN INFANCY 1996;62:442-9 Table 2. Preoperative Diagnosis and Previous Procedures Diagnosis Patients Previous Procedures Residual Lesion IAA/VSD/SAS 3 IAA/VSD repair 1, SAS resection 1 AR, VSD, arch obstruction IAA/prosthetic graft palliation/pa band 1 AR, SAS, VSD, arch obstruction IAA/VSD/SAS repair 1 AS, mild SAS Recurrent AS 2 Balloon valvuloplasty 2, valvotomy 1 AS, AR Balloon valvuloplasty 1, valvotomy 1 AS AS/SAS/AoCo 1 AoCo repair 2, balloon valvuloplasty 2 AS, SAS ASISAS/MS/MR 2 Balloon valvuloplasty 1 AS, SAS, MR Balloon valvuloplasty 1 AS, SAS, MR AoCo = aortic coarctation; AR - aortic regurgitation; AS - aortic stenosis; IAA = interrupted aortic arch; MR = mitral regurgitation; MS - mitral stenosis; PA pulmona D' artery; SAS subaortic stenosis; VSD = ventricular septal defect. to 5 mm below the valve level, and dissection of the posterior aspect of the pulmonary root was completed with care to avoid large septal arteries. No excess portion of the RVOT muscle was harvested to accommodate the completion of the septal ventriculoplasty. The subaortic region was then assessed and any fibrous membrane resected. The incision of the conal septum, starting at the level of the intercoronary commissure, was carried leftward to avoid the conduction system. The size of the ventriculoplasty was adjusted according to the severity of the LVOTO and the degree of size mismatch between the aortic and pulmonary autograft root. The continuity of the interventricular septum was reconstructed by one of two techniques. If a relatively shallow septal incision was sufficient to open the subaortic area, reconstruction was completed with the subpulmonary conal muscle of the autograft (n = 4). When a deeper incision into the septum was required, a Dacron patch was used (n ~ 4). In both instances repair was accomplished using Teflonpledgeted mattress sutures along the muscular septum (see Fig 2). Repair of coexistent outlet VSDs (n = 2) was carried out at this time. A second dose of cold blood cardioplegia was infused directly into the coronary ostia to disclose any significant bleeding source at the level of the incised pulmonary valve. The proximal anastomosis of the pulmonary autograft was then completed with three 5-0 Prolene (Ethicon, Somerville, NJ) monofilament sutures, starting at each commissure and anchoring the autograft root onto the crest of the ventricular septum and aortic annulus. Reimplantation of the coronary arteries was done as buttons onto the right and left sinuses of the pulmonary valve using a running 7-0 Prolene monofilament suture. The RVOT was reconstructed using pulmonary homografts (range, 10 to 21 ram; mean, 15 _+ 3 mm) and running 6-0 and 5-0 Prolene sutures for the distal and proximal anastomoses, respectively (Fig 3). Ascending aortic continuity was then reestablished with an end-to-end anastomosis between the distal pulmonary autograft and the ascending aorta with a running 6-0 Prolene suture. Weaning from cardiopulmonary bypass was aided by infusion of dopamine and dobutamine starting at 5 /zg kg 1. min 1 each. Intraoperative and perioperative aprotinin (Bayer AG, Leverkusen, Germany) was routinely used. Associated procedures included repair of arch obstruction in 3 infants, modification of the septal ventriculoplasty to accommodate for concomitant VSD repair in 2, removal of a pulmonary artery band at the time of autograft harvesting in 1, and mitral valvuloplasty in 2. Table 3. Clinical and Echocardiographic Follow-up Data Variable Preoperative Postoperative Follow-up p Value Mean weight (kg) Mean NYHA class Mean aortic valve annulus diameter (mm) Mean PV annulus diameter (ram) Mean peak LVOT gradient (mm Hg) Mean peak RVOT gradient (mm Hg) Absent/mild AR (no.) Moderate/severe AR (no.) Absent/mild PR (no.) Moderate/severe PR (no.) 6.3 + 2.6... 10.8 ± 2.0 <0.001 "~ 3.4 + 0.5... 1.3 ± 0.5 <0.001 ~ 7.2 _+ 2.3 12.2 ± 2.6 15.7 ± 3.6 <0.002 b 12.4 + 2.6 13.4 ± 3.9 14.0-4.0 NS b 71.7 + 25 6.2 ± 7.6 8.2 -~ 6.4 <0.001 a 5.7 + 4.3 4.3 ± 4.2 6.4 + 3.5 NS a 6/8 8/8 8/8 NS a 2/8...... <0.01 a 8/8 8/8 7/8 NS ~...... 1/8.., Comparison of preoperative versus follow-up data. b Comparison of postoperative versus follow-up data. AR - aortic regurgitation; LVOT - left ventricular outflow tract; NS not significant; NYHA = New York Heart Association; pulmona D" regurgitation; PV pulmona D' valve; RVOT right ventricular outflow tract. PR

Ann Thorac Surg STARNES ET AL 445 1996;62:442-9 PULMONARY AUTOGRAFT IN INFANCY Table 4. Operative Technique (n = 8) Technique Autograft root replacement RVOT homograft reconstruction Conal septum incision Aortic arch reconstruction VSD patch repair PA band removal Mitral valvuloplasty No. of Patients,\\ PA - pulmonary artery; RVOT - right ventricular outflow tract; VSD = ventricular septal defect. Postoperative Follow-up All patients underwent predischarge and follow-up transthoracic echocardiographic assessment. Echocardiographic examinations included quantitative and qualitative estimates of ventricular function, semilunar valve diameters and function, anatomy of the ventricular septurn, and left heart-aorta complex. Statistical Analysis Data were expressed as means _+ standard deviation. A two-tailed, paired Student's t test was used to compare continuous variables. Discrete variables were compared using Pearson's )(2 test. A p value less than 0.05 was considered significant. Results Operative Results Eight children underwent extended aortic root replacement with the pulmonary autograft with the above- Fig 2. Operative view of the excised aortic and pulmonary roots. The coronary buttons have already been harvested. The left ventricular outflow tract has been widened by incision or resection of the conal septum, when appropriate. The continuity of the ventricular septum has been reestablished using a wedge-shaped Dacron patch. Fig 3. Operative view of the completed repair with pulmonary autografl in the aortic position, implanted coronary arteries, and the pulmonary homograft in the orthotopic position. described technique, and all survived the operation. Mean aortic cross-clamp time was 93.5 -+ 25 minutes (range, 77 to 152 minutes; median, 84.5 minutes) and mean cardiopulmonary bypass time was 131.8 -- 22 minutes (range, 103 to 165 minutes; median, 128.5 minutes). All patients were easily weaned from bypass and had good valvular function as determined by intraoperative transesophageal echocardiography. Postoperative Course The postoperative recovery in the intensive care unit included an average of 2 days (range, 1 to 14 days) of inotropic support with dopamine and dobutamine infusion and an average of 2 days of mechanical ventilation (range, 1 to 13 days). Prolonged (>2 days) inotropic and ventilatory support were necessary in 2 infants, 1 with previously palliated IAA/VSD/SAS who had been intubated and receiving an intravenous catecholamine infusion since the last surgical procedure. In addition, 1 patient also with previously palliated IAA/VSD/SAS experienced left hemidiaphragm paresis, confirmed by fluoroscopic examination. Function of the hemidiaphragm recovered after 13 days of mechanical ventilation, without the need for surgical plication. All children recovered regular sinus rhythm except 1 infant, in whom complete heart block developed after arch reconstruction, autograft root replacement, and mitral valvuloplasty. In this case, a permanent pacemaker was required followed by a mitral valve replacement. Three patients showed transient ST-T wave abnormalities, which were not detected at subsequent follow-up electrocardiograms. Cumulative hospital stay averaged 15 days (range, 4 to 53 days). Length of stay was a mean of 8.8 days for 7 of the 8

446 STARNES ET AL Ann Thorac Surg PULMONARY AUTOGRAFT IN INFANCY 1996;62:442-9 inmhg n=8 p<0.001 120 100 80 60 40 20 0 Preoperative 6 mos Follow-up Fig 4. Diagram showing the decrease in cumulative left ventricular outflow tract gradient at follow-up in all infants undergoing aortic root replacement with the pulmonary autograft. patients. One patient required a prolonged hospitalization (53 days) because of the need for subsequent mitral valve replacement (17-mm St. Jude Medical [St. Paul MN]). Predischarge echocardiographic examination demonstrated absence of semilunar valve regurgitation or stenosis with normal biventricular kinetics in all patients (see Table 3). Six patients were discharged on oral furosemide and digoxin therapy and 2 infants required no discharge medications. One patient with previously palliated critical aortic stenosis presented with a low output syndrome 10 days after extended aortic root replacement with the pulmonary autograft. Emergency echocardiographic examination showed a large pericardial effusion that caused cardiac tamponade. The infant underwent surgical drainage through a subxiphoid window with immediate improvement in the hemodynamic status and recovered uneventfully. Follow-up There were no deaths or reinterventions during an average follow-up of 13.5 _+ 8 months (range, 2 to 25 months). Clinical assessment of patients demonstrated appropriate weight gain in all infants and significant improvement in functional status (Table 3). No infective or embolic events were reported by the families or referring physicians following up the infants. The echocardiographic data demonstrated lasting resolution of the LVOTO with peak pressure gradients ranging from 0 to 16 mm Hg (mean, 6.2 + 7.6 mm Hg) (Fig 4; see Table 3). Aortic regurgitation was absent in 3 patients and trivial or mild in 5. Comparing the diameter of the aortic annulus at the early postoperative and follow-up examination, there was aortic root enlargement in proportion to somatic growth suggesting growth of the autograft (Table 5). Five infants had absent pulmonary regurgitation, 2 had trivial regurgitation, and 1 had moderate regurgitation. In I infant significant pulmonary homograft obstruction has developed during follow-up, with an estimated gradient of 55 rnm Hg by echocardiogram. The remaining 7 patients have no demonstrable RVOT gradients. Left and right ventricular function were normal in all patients. Comment Current Approach to Left Ventricular Outflow Tract Obstruction The surgical management of complex LVOTO, defined as multiple-level or recurrent obstruction, in childhood remains a true challenge. Because of the high propensity of these lesions to recur and of the necessity to accommodate for growth, even radical operations often result in transient palliation. Critical congenital aortic valve stenosis presenting in the neonate is far from a benign lesion. A recent review of multiinstitutional experience with the invasive treatment of neonatal aortic stenosis reported an early mortality exceeding 50% and late mortality approaching 20% of survivors [2]. Thus, only 2 of 5 neonates initially treated for this lesion will survive up to 15 years. Furthermore, nearly 100% of long-term survivors will require reintervention, in half of the patients consisting of aortic valve replacement with or without aortic root enlargement procedures [2]. Even more complex are the infants presenting with multiple levels of LVOTO, which may include the mitral valve, subaortic region, aortic valve, and aortic arch as represented by 3 infants in this series. These infants may have even higher mortality than infants with critical aortic stenosis. Traditional operative strategies applied to recurrent aortic stenosis in babies have included multiple balloon valvuloplasty and open valvotomy procedures to postpone the problem of valve replacement to childhood or adolescence. Often this leaves the child with residual lesions resulting in a compromised left ventricle when definitive repair is undertaken. Repairs have included valve replacement techniques, which often require annulus enlargement [3-6]. However, analysis of the clinical results with such techniques, including the Nicks posterior patch annuloplasty, the Manougian mitral-aortic annuloplasty, and the Konno aortoventriculoplasty, when used in the child, has demonstrated disappointingly high mortality and morbidity [7]. Alternative operations have also been proposed for the older child but not Table 5. Serial Echocardiographic Measurements of Pulmonary Autograft Diameter in the Left Ventricular OuO:low Tract (n = 4) Age at Operation Follow-up Time Autograft Diameter (too) (too) (mm) 3.75 0.04 8.6 8 9 20 9.2 4 6 13.3 12 16.9 18 0.1 11.3 6 12.3 22 4 10.6 11 16.3

Ann Thorac Surg STARNES ET AL 447 1996;62:442-9 PULMONARY AUTOGRAFT IN INFANCY yet successfully reported for the infant [8, 9]. In addition, all of the above-mentioned options maintain the disadvantage of tissue degeneration (bioprostheses), the devastating consequences of thromboembolic and hemorrhagic complications (mechanical prostheses), and the lack of growth potential requiring reoperation [7, 10]. Similar conclusions must be drawn when performing these operations for associated aortic valvular and subvalvular stenosis. Alternative surgical strategies advocating extraanatomic bypass of the obstruction with left ventricle-aorta conduits have assumed primarily historical importance, given the precarious palliation offered and the unsolved problem of reoperation [11]. Bypassing the obstructed left ventricular outflow tract using the Norwood approach has gained some recent favor; however, this operation predisposes these infants to univentricular physiology [121. In spite of the satisfactory early survival reported with this approach, justification of its application becomes difficult in the presence of two welldeveloped ventricular chambers, as in IAA/VSD. In fact, as the morbidity and intraprocedural mortality of the Norwood staged palliation is still very high, various techniques of biventricular one-stage repair seem justified [1, 13]. A recent advancement in the surgical management of complex LVOTO in the child has come from the idea of combining the extended aortic root replacement offered by the Konno aortoventriculoplasty with the implantation of aortic homografts [14]. Although this is effective in relieving the obstructive lesions, the vulnerability of the aortic homograft to early degeneration predisposes the child to the need for repeat operations [15]. With the expected longer duration of the pulmonary autograft in the aortic position, we, like others, have combined the aortoventriculoplasty technique with pulmonary autograft insertion [16-20]. Encouraged by our preliminary results and faced with a growing population of infants with complex LVOTO for whom alterative options have shown dismal outcome, we have applied the extended aortic root replacement with the pulmonary autograft to this unusual patient population. Pulmonary Autografls in Infants Clinical experience with the use of pulmonary autografts in children is currently limited. Preliminary results have been encouraging except in patients with important endocardial fibroelastosis [19-24]. Our data suggest that aortic root replacement with the pulmonary autograft is a safe operation even in young infants with body weights between 3 and 10 kg. Although our current approach to critical aortic stenosis in the neonate includes attempts at percutaneous or surgical valvotomy, and the approach to IAA/VSD/SAS complex involves one-stage primary repair [1], we have been largely satisfied with the present technique of aortic root replacement when valvular and subvalvular lesions recur. Several technical aspects are important to the intraoperative and perioperative clinical course, including rou- tine administration of aprotinin and myocardial protection using cold induction and maintenance followed by warm reperfusion blood cardioplegia. Despite relatively long duration of aortic cross-clamp and cardiopulmonary bypass, especially in patients undergoing associated procedures, recovery of mechanical and electrical cardiac activity was prompt in all cases. Ventricular septoplasty by means of incision and resection of the conal septum and any other obstruction at the aortic subvalvular level may be completed by reconstruction of the septum either with Dacron or by implanting the autograft lower down in the aortic root. Unlike other authors [17, 24], we have avoided harvesting additional muscle from the RVOT free wall to accommodate the septoplasty. Although the use of prosthetic material potentially exposes the patient to infectious complications, we believe the untoward effects of extended right ventricular resection in a small child in terms of contractility far outweigh its expected benefits. Our data also suggest that aortoventriculoplasty using pulmonary autografts can be performed in children with a very low rate of complications, including postoperative conduction disturbances (12% in our series) previously reported as 7% to 8% [17, 24]. Functional Results Analysis of midterm results in this series of infants and children undergoing aortic root replacement with pulmonary autografts parallels previous reports on older children and adolescents [16, 17, 21]. Both the autograft and the homograft valves are functioning well with the exception of one pulmonary homograft with a demonstrable stenosis by echocardiography. Contrary to other authors [17], we have not observed any tendency toward progressive autograft root dilatation in our patients, although we recognize the short-term follow-up in this group of infants. Furthermore, the aortoseptoplasty operation has resulted in lasting normalization of left ventricular outflow tract anatomy and flow. In spite of the complexity of the procedure, postoperative and follow-up ventricular function have shown prompt recovery, as previously reported by others [20]. Finally, the present surgical approach would represent the most physiologic procedure if appropriate autograft valve growth were clearly demonstrated. The major criticism to the pioneering experience of Ross and associates in children is that autograft growth has never been proven [19]. More recent experience by Elkins and coworkers [21] has shown enlargement of the autograft root with time, which is proportional to somatic growth. Once again, the infant and young child represent an unusual subset of patients in that the somatic growth is rapid in the baby and proportional autograft enlargement should be demonstrable. The preliminary data of our experience would tend to confirm autograft enlargement that parallels somatic growth (see Table 5). We are therefore looking with great expectations at this population as the most dependable clinical model to prove or disprove the growth of autograft valves.

448 STARNES ET AL Ann Thorac Surg PULMONARY AUTOGRAFT IN infancy 1996;62:442-9 We recognize there are limitations of this study. The follow-up is short, and meaningful conclusions are yet to be determined. The use of a pulmonary homograft for RVOT reconstruction is likely to expose the patient to at least one reoperation to accommodate for lack of growth of the cadaveric implant. However, as the choice of slightly oversized pulmonary homografts (average diameter, 15 ram) may delay the timing of reoperation, the durability of pulmonary homografts has yet to be determined. Finally, as with all clinical applications of the autograft valve, one should remember that our operation exposes two valves to risk during follow-up. Conclusions The population of young children with complex LVOTO have a uniform lack of valid therapeutic alternatives. The present experience with aortoventriculoplasty using pulmonary autografts offers excellent early survival and functional outcome. Considering the freedom from prosthetic valve-related complications, wider use of this technique is recommended. If growth of the autograft valve were to be confirmed by longer follow-up, this operation could become the ultimate surgical therapy for left-heart obstruction in infants and young children. References 1. Luciani GB, Ackerman RJ, Chang AC, Wells WJ, Starnes VA. One-stage repair of interrupted aortic arch, ventricular septal defect, and subaortic obstruction in the neonate: a novel approach. J Thorac Cardiovasc Surg 1996;111:348-58. 2. Gaynor WJ, Bull C, Sullivan ID, et al. Late outcome of survivors of intervention for neonatal aortic valve stenosis. Ann Thorac Surg 1995;60:122-6. 3. Konno S, Imai Y, lida Y, Nakajima M, Tastsuno K. A new method of prosthetic valve replacement in congenital aortic stenosis associated with hypoplasia of the aortic valve ring. J Thorac Cardiovasc Surg 1975;70:909-17. 4. Schaffer MS, Campbell DN, Clarke DR, Wiggins JW Jr, Wolfe RR. Aortoventriculoplasty in children. J Thorac Cardiovasc Surg 1986;92:391-5. 5. Nicks R, Cartmill T, Bernstein L. Hypoplasia of the aortic root. Thorax 1970;25:339-46. 6. Manougian S, Seybold-Epting W. Patch enlargement of the aortic valve ring by extending the aortic incision into the anterior mitral leaflet. J Thorac Cardiovasc Surg 1979;78: 402-12. 7. Ross DB, Trusler GA, Coles JC, et al. Small aortic root in childhood: surgical options. Ann Thorac Surg 1994;58: 1617-25. 8. Vouhe PR, Poulain H, Bloch G, et al. Aortoseptal approach for optimal resection of diffuse subvalvular aortic stenosis. J Thorac Cardiovasc Surg 1984;87:887-93. 9. Caspi J, Ilbawi MN, Roberson DA, Piccione W, Monson DO, Najafi H. Extended aortic valvuloplasty for recurrent valvular stenosis and regurgitation in children. J Thorac Cardiovasc Surg 1994;107:1114-20. 10. McIntyre B, Guyton RA, Jones EL, Graver JM, Williams WH, Hatcher CR. Reoperation for prosthetic valve degeneration after Konno aortoventriculoplasty. J Thorac Cardiovasc Surg 1986;91:934-6. 11. Norwood WI, Lang P, Castafieda AR, Murphy JD. Management of infants with left ventricular outflow obstruction by conduit interposition between the ventricular apex and thoracic aorta. J Thorac Cardiovasc Surg 1983;86:771-6. 12. Jacobs ML, Rychik J, Murphy JD, Nicolson SC, Steven JM, Norwood WI. Results of Norwood's operation for lesions other than hypoplasfic left heart syndrome. J Thorac Cardiovasc Surg 1995;110:1555-62. 13. Bove EL, Minich L, Pridjian A, et al. The management of severe subaortic stenosis, ventricular septal defect, and aortic arch obstruction in the neonate. J Thorac Cardiovasc Surg 1993;105:289-96. 14. McKowen RL, Campbell DN, Woelfel GF, Wiggins JW Jr, Clarke DR. Extended aortic root replacement with aortic allografts. J Thorac Cardiovasc Surg 1987;93:366-74. 15. Clarke DR. Extended aortic root replacement with cryopreserved allografts: do they hold up? Ann Thorac Surg 1991;52: 669-75. 16. Gerosa G, McKay R, Ross DN. Replacement of the aortic valve or root with a pulmonary autograft in children. Ann Thorac Surg 1991;51:424-9. 17. Daenen WJ, Vanhove MA, Gewillig MH. Extended aortic root replacement with pulmonary autografts: experience in 14 cases. Ann Thorac Surg 1995;60:$180-4, 18. Hvass U, Chatel D, Calliani J, Pansard Y. Relief of complex left ventricular outflow tract obstruction with pulmonary autografts. J Thorac Cardiovasc Surg 1995;109:1019. 19. Reddy VM, Rajasinghe HA, McElhinnay DB, et al. Extending the limits of the Ross procedure. Ann Thorac Surg 1995;60: $600-3. 20. Calhoon JH, Bolton JWR. Ross/Konno procedure for critical aortic stenosis in infancy. Ann Thorac Surg 1995;60:$597-9. 21. Elkins RE, Knott-Craig CJ, Ward KE, McCue C, Lane MM. Pulmonary autograft in children: realized growth potential. Ann Thorac Surg 1994;57:1387-94. 22. School PH, Cromme-Dijkhuis AH, Bogers JJ, et al. Aortic root replacement with pulmonary autograft in children. J Thorac Cardiovasc Surg 1994;57:367-73. 23. Quaegebeur JM, Solowiejczyk D, Hsu DT, Bourlon F, Hess J, Gersony W. Does the Ross operation provide a definitive solution for children with complex left ventricular outflow tract obstruction? [Abstract]. Presented at the 74th Annual Meeting of The American Association for Thoracic Surgery, New York, NY, April 23-26, 1994. 24. Reddy VM, Rajasinghe HA, Teitel DF, Haas GS, Hanley FL. Aorticoventriculoplasty with the pulmonary autograft: the "Ross-Konno" procedure. J Thorac Cardiovasc Surg 1996; 111:158-67. DISCUSSION DR JOHN H. CALHOON (San Antonio, TX): I thank The Society for the opportunity of discussing this fine report and commend Dr Starnes on taking a very difficult set of children and doing a wonderful job with them. I also thank him for giving me this text; it is very well done and it will be of benefit to all of us. In San Antonio we, too, have performed three Ross proce- dures in children 1 would call neonates with no early deaths, one with interrupted aortic arch and two with isolated aortic stenosis. Our first patient is now about 4 years old, has not required reoperation, and has shown growth in the pulmonary autograft. l agree with Dr Starnes and others now that the Ross procedure is probably the procedure of choice for left ventricular outflow tract obstruction in infants.

Ann Thorac Surg STARNES ET AL 449 1996;62:442-9 PULMONARY AUTOGRAFT IN INFANCY l have two questions. In our patient with interrupted aortic arch, left main bronchial compression developed. Do you have any tips that you would like to share with us on avoiding this complication? Also, now that you have experience in this age group, all of whom had prior procedures, do you think there is a place for no prior intervention with a balloon dilation or other palliation in isolated neonatal aortic valve stenosis and proceeding straight to a Ross operation? DR STARNES" Thank you, Dr Calhoon, for those nice comments. I think in relation to reconstructing the aortic arch and trying to avoid the complication of left bronchial stenosis, we have found that actually bringing the posterior native aortic arch together with the proximal descending aorta can be done without tension if you really immobolize the head vessels completely and also add a homograft augmentation anteriorly. Whether or not this should be offered as a primary procedure I think is obviously being debated more and more. I think neonates presenting with neonatal critical aortic stenosis probably still need the benefit of a balloon valvuloplasty. And if it fails, I think we should go directly then to the neonatal Ross procedure. I think some of the palliative operations that we have encountered from other institutions should not be performed. I think the complex lesions of ventricular septal defect and critical aortic stenosis should be approached with the methodology that we propose today. REVIEW OF RECENT BOOKS Coronary Bypass Surgery in the Elderly. Ethical, Economical and Quality of Life Aspects Edited by Paul J. Walter Dordrecht-Boston-London, Kluwer Academic, 1996 257 pages, illustrated, $127.00 ISBN 0-7923-3188-5 Reviewed by Agustin Arbulu, MD This book is a compilation of the papers presented at an international symposium where world experts in health economics, ethics, clinical cardiology, cardiac surgery, epidemiology, and health care policy discussed the need and outcome of the operation in old and very old patients. The meeting took place from March 9 to 11, 1994, in Antwerp, Belgium. The book is divided into six parts. In part I, the demographics and health policy are presented in relation to the economic trends in Europe. The increase of the elderly within a range of 24% to 64% by the year 2020 will raise the prevalence of coronary. disease. This elderly segment will need intervention techniques such as coronary artery bypass grafting and percutaneous transluminal coronary angioplasty more than ever. In part II, the results of coronary bypass grafting are presented in a series of four papers. These presentations confirm the fact that the elderly tolerate coronary bypass operations almost as well as patients who are less than 65 years of age. For the great majority of elderly patients (octagenarians) the quality of life is vastly improved by a coronary operation. In part III, the cost of coronary bypass grafting is carefully analyzed in a paper from Germany that concludes that the average costs per heart operation of younger patients are slightly lower than the corresponding costs for older patients. In part IV, the controversies in relation to economic and ethical questions related to the surgical management of the elderly are discussed. Most of the viewpoints presented are based on the philosophical considerations of the authors of the five papers presented. In part V, which I consider perhaps the most revealing of the whole book, the influence of coronary bypass grafting on the quality of life for the elderly patient is carefully studied in six presentations. Three of these papers were from the United States. It appears that the improvement in the quality of life is a result of great importance among the elderly, far more important than survival rates. Finally, in part VI, the characteristics of the rehabilitation of the elderly are discussed. This area is often neglected and in need of improvement. The book closes with discussions that developed during the proceedings of the symposium. This portion of the book is quite interesting and very informative. I consider the book to be valuable to primary doctors, internists, cardiologists, cardiac surgeons, health care officers, administrators, and economists of the health care industry. The book is timely as we enter the era of various managed care systems in health care. Detroit, Michigan 1996 by The Society. of Thoracic Surgeons 0003-4975196/$15.00 Published by Elsevier Science lnc PII S0003-4975(96)00408-0