Coarctation of the aorta is a congenital narrowing of the

Operative Risk Factors and Durability of Repair of Coarctation of the Aorta in the Neonate Walter H. Merrill, MD, Steven J. Hoff, MD, James R. Stewart, MD, Charles C. Elkins, MD, Thomas P. Graham, [r, MD, and Harvey W. Bender, [r, MD Department of Cardiac and Thoracic Surgery and Division of Pediatric Cardiology, Vanderbilt University School of Medicine, Nashville, Tennessee The risk factors for the operative mortality and long-term durability of repair after surgical correction of coarctation of the aorta in neonates remain controversial. Between January 1970 and January 1993, 139 patients under 1 month of age underwent repair of coarctation of the aorta. Complex intracardiac defects were present in 59 patients. Another 44 patients had an associated ventricular septal defect. Subclavian artery flap repair was performed in 92 patients; end-to-end anastomosis (38 patients) and patch angioplasty (9 patients) were performed less commonly. The hospital mortality was significantly higher in patients with complex intracardiac defects (9 of 59 patients; 15.2%) than in those with a ventricular septal defect (1 of 44 patients; 2.3%) or with isolated coarctation (none of 36 patients; p = 0.007). Elevated pulmonary artery diastolic pressure ip = 0.041) and complex intracardiac anomalies (p = 0.048) were found to be indepen- dent predictors of hospital mortality. The presence of a complex cardiac defect ip < 0.001) was an independent predictor of poor long-term survival. Recurrent stenosis requiring reoperation had occurred or balloon dilation had been necessary in 27.9% of the children at 5 years postoperatively. In patients followed up for at least 5 years, the recurrence-free survival was better in those who had undergone subclavian artery flap repair than in those who had undergone end-to-end repair (p = 0.017). When coarctation of the aorta must be repaired in the neonate, operative mortality and long-term survival are affected by the complexity of associated intracardiac anomalies. The subclavian artery flap repair of aortic coarctation in neonates was found to result in a lower rate of recurrent stenosis in patients followed up for at least 5 years postoperatively. (Ann Thorae Surg ) Coarctation of the aorta is a congenital narrowing of the lumen of the thoracic aorta. Its manifestations may cover a broad spectrum of abnormalities, from minimal narrowing to luminal atresia. As the age of the child at presentation becomes younger, the incidence of isolated coarctation diminishes. Coarctation may be associated with a range of complex obstructive lesions, including tubular hypoplasia of the aortic arch, left ventricular outflow tract obstruction, and left ventricular hypoplasia (hypoplastic left heart syndrome). In addition, coarctation of the aorta is often associated with other congenital intracardiac anomalies. Patients who have coarctation with or without associated congenital intracardiac defects may present very early in life with severe congestive heart failure, and this may require urgent operation. The specific factors that might affect the risk of operative intervention in neonates with coarctation of the aorta remain controversial. Many neonates have undergone successful repair of coarctation of the aorta, only to experience restenosis subsequently. Therefore, the durability of repair is a major issue, and has also become a controversial one because of the inevitable comparisons of the results of various operative techniques. Presented at the Fortieth Annual Meeting of the Southern Thoracic Surgical Association, Panama City Beach, FL, Nov 4-6, 1993. Address reprint requests to Dr Merrill, Department of Cardiac and Thoracic Surgery, Room 2986, The Vanderbilt Clinic, Nashville, TN 37232. 1994by The Society of Thoracic Surgeons To address these issues, we report on our 23-year experience with the repair of aortic coarctation in neonates. Material and Methods The medical records of all infants who underwent operative repair of aortic coarctation in the first month of life at the Vanderbilt University Medical Center between January 1970 and January 1993 were retrospectively reviewed. We recorded details of the patients' admission clinical status, including the presenting symptoms, the use of mechanical ventilation, the use of inotropic support, and the use of prostaglandins. All patients who were evaluated had undergone cardiac catheterization and angiography before repair. Details of the surgical procedure as well as of the immediate and long-term follow-up were noted. Special attention was paid to the recurrence of coarctation of the aorta during the long-term follow-up by analyzing the clinical, echocardiographic, and catheterization data. All survivors were evaluated, and their current status was assessed between January and July 1993. The overall and recurrence-free survival were estimated using the Kaplan-Meier method. Statistical evaluation included univariate and multivariate analyses using the Cox proportionate-hazards regression method. Statistical significance was defined as a p value of less than 0.05. 0003-4975/94/$7.00

400 MERRILL ET AL Ann Thorae Surg Table 1. Associated Complex Intracardiac Anomalies Lesion Double-outlet right ventricle Aortic stenosis Hypoplastic left heart Transposition Subaortic stenosis Mitral stenosis Atrioventricular septal defect Pulmonic stenosis Tricuspid atresia Partial anomalous pulmonary venous connection Truncus arteriosus No. 18 11 11 10 8 6 6 4 3 2 14 Percent 12 III Isolated DVSD 13 10.Complex (/) 8 e ::l "'0 8 8 Q) o e 6 7 Co 6 ci 4 z 4 4 2 3 2 0 1 1970 1975 1980 1985 1990 Fig 1. Types of defect (isolated coarctation, coarctation plus ventricular septaldefect [VSD], and coarctation plus associated complex intracardiac defects) seen each year. Results During the 23-year study period, 139 infants younger than 1 month underwent repair of aortic coarctation. There were 80 male and 59 female infants. The mean age at operation was 10.8 days (range, 1 to 28 days) and the mean weight at operation was 3,330 g (range, 1,650 to 5,650 g). In 92% (n = 128) of these infants, there were symptoms of congestive heart failure that served as the primary indication for operation. Prostaglandin infusion was used in 55% of the patients (n = 76). Twenty-four percent (n = 34) were intubated and placed on mechanical ventilation preoperatively, and 23% (n = 32) required inotropic support. At cardiac catheterization, the average pressure gradient across the coarctation was 31 mm Hg (range, 5 to 84 mm Hg), the mean left ventricular end-diastolic pressure was 15 mm Hg (range, 2 to 50 mm Hg), and the mean pulmonary artery diastolic pressure was 29 mm Hg (range, 13 to 60 mm Hg), Isolated coarctation (with or without patent ductus arteriosus, foramen ovale, or a bicuspid aortic valve) was present in 26% (n = 36) of the patients. Coarctation plus an associated ventricular septal defect (VSD) was present in 32% (n = 44) of the patients. Coarctation plus an associated complex intracardiac anomaly existed in 42% (n = 59) of the patients. The coarctation was complicated by the presence of a hypoplastic transverse aortic arch in 37% (n = 52) of the patients, and the incidence of this was relatively constant during the entire period of this study. The distribution of the associated lesions is shown in Table 1. The distribution of the types of defect (isolated, associated VSD, and associated complex defect) treated during the time frame of this study is illustrated in Figure 1. A subclavian artery flap repair was performed in 92 infants (67%), an end-to-end anastomosis was performed in 38 (27%), and a patch angioplasty was carried out in 9 (6%). The distribution of the types of repair performed over the course of this experience is illustrated in Figure 2. Concomitant pulmonary artery banding was performed in 43% (n = 60) of the patients, and the indication for doing this was the presence of a large (>2:1) left-to-right shunt resulting from an intracardiac defect. The band was placed in 24 of the 44 patients with an associated VSD and in 36 of the 59 patients with associated complex defects. The overall hospital mortality was 7.2% (n = 10). No deaths occurred in the 36 patients with an isolated coarctation. In the patients with an associated VSD, the mortality was 2.3% (n = 1), and it was 15.2% (n = 9) in the patients with associated complex cardiac anomalies. All hospital mortalities were related to the initial operation performed to repair the coarctation, with or without placement of a pulmonary artery band. In a multivariate analysis using a Cox proportionate-hazards regression model of multiple preoperative and operative factors, only elevated pulmonary artery diastolic pressure (p = 0.041) and associated complex intracardiac anomalies (p = 0.048) proved to be independent predictors of operative mortality. Age, weight, whether a pulmonary artery band was placed, the presence or absence of a hypoplastic transverse aortic arch, and the type of repair were not statistically significant factors. The causes of death were congestive heart failure (n = 5), pulmonary infection and sepsis (n = 3), cerebral hemorrhage and sepsis (n = 1), and arrhythmias and bleeding (n = 1). There have been a total of 20 late deaths, but none have occurred in the patients with an isolated coarctation. There have been three late deaths in the patients with an associated VSD, and 17 late deaths have occurred in the group I 1970 1975 1980 1985 1990 Fig 2. Types of operative repair performed each year.

Ann Thorae Surg MERRILL ET AL 401 100 -- - - - - - - - - - - - - - - -... e 'Ẹ ė &. 25 p<.ool 100 > 'Ẹ.~.. 0-25 - - Subclavian Flap -End-to-End 0 0 60 o + - - - - - - + - - - - - - + - - - - - - + - - - - - ~ o 60 Fig 3. Actuarial survival for all patients according to the type of defect. (VSD = ventricular septal deieci.) Fig 4. Recurrence-free survival for all patients who initially underwent repair consisting of either the subclavian artery flap or end-toend technique. (NS = not signiiicant.) with an associated complex intracardiac defect. Late deaths in 16 patients were due to a subsequent operative repair or palliation of a coexisting lesion. The interval between the coarctation repair and subsequent operative intervention related to the VSD or complex cardiac defect was highly variable, and ranged from several months to several years. In the remaining 4 patients, late deaths were due to congestive heart failure or to pulmonary complications. The long-term survival was significantly lower in patients with complex associated cardiac defects (p < 0.001; Fig 3). This was the only independent predictor of poor long-term survival. All late survivors have been reassessed recently. Follow-up for the entire survivor group extends to a mean of 44.2 months (range, 1 to 272 months). A subgroup of 40 patients has been followed up for a minimum of 5 years, and consists of 15 patients who underwent end-to-end repair, 24 patients who underwent subclavian artery flap repair, and 1 patient who had a patch aortoplasty. All survivors have been assessed to determine if coarctation has recurred at the site of repair. Recurrent stenosis was considered present if the resting systolic blood pressure gradient between the arm and the legs was 20 mm Hg or greater. All such patients then underwent repeat angiography, which in all documented at least 50% luminal narrowing at the site of repair. For the survivor group as a whole, the incidence of recurrent stenosis requiring balloon dilation or reoperation was 13.9% at 1 year and 27.9% at 5 years. The presence or absence of arch hypoplasia, the presence or absence of associated cardiac anomalies, and the type of repair did not affect the incidence of recurrence (p = not significant) (Fig 4). However, in patients followed up for at least 5 years, the recurrence-free survival was found to be better for subclavian artery flap repair than that for end-to-end repair tp = 0.017) (Fig 5). Comment Neonates with symptoms of congestive heart failure stemming from coarctation of the aorta require operative intervention to improve their otherwise poor prognosis [1]. Over the years, improved preoperative preparation [2] and refined surgical techniques have resulted in better survival rates. The operative mortality rates for coarctation repair vary widely in reported series, ranging from 0% to 45% [3-5]. Several factors have been associated with higher mortality rates, and these include younger age at operation, tubular hypoplasia of the transverse aortic arch, and severe associated intracardiac anomalies. In the current study, the only factors that were shown to have a statistically significant influence on operative outcome were an elevated pulmonary artery diastolic pressure and the presence of an associated complex cardiac defect. Age, weight, the presence or absence of a pulmonary artery band, the presence or absence of a hypoplastic aortic arch, and the type of repair carried out were factors that did not achieve statistical significance. Shortly after operation, a persistent pressure gradient (residual coarctation) may be noted. In most instances, a true recurrent coarctation is detected only months or years after an initially successful repair [6, 7]. Recurrence after repair consisting of resection and endto-end anastomosis has been reported to affect 36% to 75% of those patients undergoing the operation when neonates [8, 9]. Restenosis may occur, either from lack of growth of an anastomotic suture line or from active narrowing resulting from fibrosis. Waldhausen and Nahrwold [l01 introduced the subcla- 100 > 'Ẹ.. l:!. p=.017 0-25 0 0 60 - - Subclavian Flap - End-to-End Fig 5. Recurrence-free survival for patients followed up for at least 5 years postoperatively.

402 MERRILL ET AL Ann ThoraeSurg 1994;58:399-403 vian artery flap repair, which became very popular and widely used because it was associated with an apparent decrease in the prevalence of recurrent coarctation. This technique is simple, it preserves the growth potential of the flap, and it avoids a circumferential suture line. However, this technique allows for only partial excision of residual ductal tissue from within the aorta, and the residual ductal tissue that is left behind may foster the development of an obstructing shelf [6]. Several concerns have been raised regarding the longterm results from the subclavian flap aortoplasty, particularly with regard to the rate of recurrence in neonates. Despite the absence of convincing data from comparisons of one type of repair to another, may surgeons currently prefer resection and end-to-end anastomosis. This decision is based primarily on the belief that the coarctation segment containing abnormal ductal tissue should be resected to help prevent recurrence [11]. In patients with coarctation and an associated VSD or complex intracardiac defect, we prefer to repair the coarctation as an initial operative procedure, and have reserved correction of the associated lesion for a subsequent intervention. This approach has simplified the initial procedure, and allowed a precise, unhurried repair to be carried out. The results seem to have been satisfactory in terms of hemodynamic improvement and survival. Our study findings indicate the importance of long-term follow-up for assessing the risk of recurrent narrowing after coarctation repair in neonates. For the group as a whole, there was no statistically significant difference in the rate of recurrent coarctation after repair consisting of either resection and end-to-end anastomosis or the subclavian artery flap repair. However, the evaluation of those patients with a minimum of 5 years of follow-up revealed a statistically significant higher rate of recurrence after resection and end-to-end repair than after subclavian artery flap repair. In summary, in our experience with coarctation repair in neonates, the operative mortality is increased only in the presence of an elevated pulmonary artery diastolic pressure or an associated complex intracardiac defect. In patients followed up for more than 5 years postoperatively, the rate of recurrent stenosis seems to be lower in those who have undergone the subclavian artery flap repair than in those who have undergone either end-to-end anastomosis or patch angioplasty. References 1. Kirklin JW, Barratt-Boyes BG. Coarctation of the aorta and aortic arch interruptions. In: Kirklin JW, Barratt-Boyes BG,eds. Cardiac surgery. 1st ed. New York: Churchill-Livingstone, 1988: 1035-80. 2. Graham TP [r, Bender HW Jr. Preoperative diagnosis and management of infants with critical congenital heart disease. Ann Thorac Surg 1980;29:272-88. 3. Lacour-Gayet F, Bruniaux J, Serraf A, et al. Hypoplastic transverse arch and coarctation in neonates. Surgical reconstruction of the aortic arch. A study of sixty-six patients. J Thorac Cardiovasc Surg 1990;100:808-16. 4. Behl R, Sante P, Blesovsky A. Surgical treatment of isolated coarctation of the aorta: 18 years' experience. Thorax 1987;42: 309-14. 5. Bobby JJ, Emami JM, Farmer RDT, Newman CGH. Operative survival and 40 year follow up of surgical repair of aortic coarctation. Br Heart J 1991;65:271-6. 6. Beekman RH, Rocchini AP, Behrendt DM, et al. Long-term outcome after repair of coarctation in infancy: subclavian angioplasty does not reduce the need for reoperation. J Am Coli Cardiol 1986;8:1406-11. 7. Ziemer G, Jonas RA, Perry SB, Freed MD, Castaneda AR. Surgery for coarctation of the aorta in the neonate. Circulation 1986;74(pt 2):1-25-31. 8. Williams WG, Shindo G, Trusler GA, Dische MR, Olley PM. Results of repair of coarctation of the aorta during infancy. J Thorac Cardiovasc Surg 1980;79:603-8. 9. Van Son JAM, Daniels 0, Vincent JG, van Lier HJJ, Lacquet LK. Appraisal of resection and end-to-end anastomosis for repair of coarctation of the aorta in infancy: preference for resection. Ann Thorac Surg 1989;48:496-502. 10. Waldhausen JA, Nahrwold DL. Repair of coarctation of the aorta with a subclavian flap. J Thorac Cardiovasc Surg 1966; 51:532-3. 11. Jonas RA. Coarctation. Do we need to resect ductal tissue? Ann Thorac Surg 1991;52:604-7. DISCUSSION DR CHRISTOPHER J. KNOTT-CRAIG (Oklahoma City, OK): I enjoyed this presentation and the data immensely, and congratulate Dr Merrill and associates on their outstanding results. We have recently reviewed our own experience and published our findings in a similar group of patients, consisting of over a hundred neonates undergoing coarctation repair. We essentially found many of the same results as those reported by Dr Merrill, with a few exceptions, which I would like to describe and ask Dr Merrill to comment on. In the first place, we found that, although our recurrence rate was similar, the majority of the recurrences occurred within the first year of repair during that period when the infants are growing the fastest. My impression from your abstract is that, in fact, the majority of your recurrences occurred after 1 year of repair. Could you explain this, at least in terms of the analysis of your data. Second, although in our experience we found no complications related to arm length or the vascularity of the left arm, did you encounter any complications in this regard in the patients who underwent subclavian artery flap repair? Finally, from the data that you presented, it would seem that the patients undergoing resection and end-to-end anastomoses who had been followed up for more than 5 years were those who had been operated on very early in your experience, and the suture material and techniques of operation may have had an impact on at least the incidence of late recurrence in this group. Could you comment on this? Do you really believe that the subclavian artery flap repair is in fact a superior operation to resection in the treatment of coarctation in the neonate? DR IRVING 1. KRON (Charlottesville, VA): Doctor Merrill was kind enough to send me this paper ahead of time; it is a very well-written paper and I think you will enjoy reading it. We agree with his results. We presented similar material at this meeting 3 years ago, and we noted that 13 of 44 infants who underwent end-to-end anastomoses required repeat intervention, versus only

Ann Thorac Surg MERRILL ET AL 403 5 of 41 who underwent subclavian artery flap repair. So, he is certainly preaching to the choir as far as we are concerned. One of the questions we have is: when do you do a subclavian artery flap repair and when do you do an end-to-end anastomosis? I notice by your slide that you became reinterested in the latter technique in the late 1980s, but it looks like this past year you did hardly any end-to-end anastomoses. Could you explain the reason for this? How often do you do concomitant repair of these complex defects together with coarctation repair? Finally, have you witnessed any subclavian flap aneurysms? We had a lot of experience with Dacron patch aneurysms, but have not thus far intervened in a subclavian flap aneurysm. DR MERRILL: Thank you all for your comments. I really cannot explain why the recurrences occurred primarily in the first year in Oklahoma and later than this in Nashville. One of the possible explanations for any recurrence is that, in the subclavian artery flap procedure, the coarctation shelf is resected partially, if at all, and some of the abnormal ductal tissue remains behind in the aorta. Many persons have argued that this is one of the major flaws of the subclavian artery flap procedure, and all this abnormal ductal tissue ought to be removed rather than left behind. We have not encountered any evidence of arm ischemia after the left subclavian artery flap procedure. There may well have been some slight modifications in the operative technique over time, but we think that the main aspects of the repair have been done in a fairly consistent fashion over the entire course of the study period. The final question from Dr Knott-Craig is whether the subclavian artery flap repair is really better. I will admit that I am not sure, but I would say that, based on our data, it appears that the long-term results seem to be better for the subclavian artery flap procedure. With regard to the comments of Dr Kron, it certainly is true that, beginning in the mid to late 1970s and on into the mid 1980s, the left subclavian artery flap repair was the preferred procedure. There has been some change in our thinking over the past 5 or 6 years or so, in that we, like most everyone else, had hoped 10 years ago that the subclavian artery flap repair was the definitive answer; it was going to solve all the problems and there would not be any recurrences. We were wrong, as most everybody else was also wrong. It is a good operation, but it is not perfect, and it does not completely solve the problem of recurrence. Therefore, some years ago there was increased interest in going back to the end-to-end repair. We thought that perhaps the coarctation shelf should be completely resected, and perhaps all the abnormal ductal tissue should be removed. I think we will just have to wait for some more experience to accumulate until we know which repair technique gives the best long-term results. We have employed a treatment protocol that is fairly simple. When there is hemodynamically important coarctation of the aorta, that is dealt with in the neonatal period. If there is a significant left-to-right shunt, then the patient also undergoes pulmonary artery banding. We have not performed concomitant repair of complex lesions. Those have generally been left for at least several weeks, if not several months, before being dealt with. Finally, we have not seen any evidence of aneurysm formation after the subclavian artery flap procedure.