- Email: [email protected]
Anatomic reconstruction for recurrent aortic obstruction in infants and children
CARDIOVASCULAR
Anatomic Reconstruction for Recurrent Aortic Obstruction in Infants and Children Daniel J. DiBardino, MD, Jeffrey S. Heinle, MD, Grace C. Kung, MD, Glenn T. Leonard, Jr, MD, Emmett D. McKenzie, MD, Jason T. Su, MD, and Charles D. Fraser, Jr, MD Division of Congenital Heart Surgery, Michael E. DeBakey Department of Surgery, and Division of Pediatric Cardiology, Department of Pediatrics, Baylor College of Medicine, Houston, Texas
Background. Patients undergoing operative repair of aortic obstruction are at a lifelong risk of recurrent obstruction, and there is controversy regarding the optimal surgical technique. We have used an alternative strategy for recurrent aortic obstruction, typically involving anatomic reconstruction by means of a median sternotomy, and describe our techniques and results. Methods. Twenty-one patients presented with recurrent aortic arch obstruction. Mean age and weight were 7.8 ⴞ 5.4 years (range, 0.21 to 15.2 years) and 30.6 ⴞ 21.8 kg (range, 3.6 to 90 kg), respectively. Recurrence involved the aortic arch to some degree in each case, as the mean preoperative transverse aortic arch z score was ⴚ2.9 ⴞ 1.6 (range, ⴚ7.0 to 0.1). Thoracotomy was possible in 2 patients, using re-resection with end-to-end anastomosis (n ⴝ 1) and patch aortoplasty (n ⴝ 1). The remaining 19 patients required median sternotomy, cardiopulmonary bypass, and deep hypothermic circulatory arrest for complete relief of obstruction by aortic arch advancement (n
ⴝ 10), patch aortoplasty (n ⴝ 8), or interposition grafting (n ⴝ 1). Results. There was 1 hospital death. Invasive blood pressure monitoring revealed no residual arm-to-leg gradient in 19 patients and a 20-mm Hg gradient in 2 patients. There have been no late deaths. No patients have undergone subsequent aortic intervention, and all are asymptomatic up to 85 months postoperatively. Two patients are currently followed with a 10-mm Hg arm-toleg blood pressure gradient. Conclusions. Anatomic reconstruction for recurrent aortic obstruction can be safely accomplished in the majority of patients. We favor median sternotomy because of the ability of establishing cardiopulmonary bypass, the facility of anatomic reconstruction techniques, and the ability to repair concomitant cardiovascular lesions. (Ann Thorac Surg 2004;78:926 –32) © 2004 by The Society of Thoracic Surgeons
C
into alternative repair strategies. Extraanatomic aortic bypass techniques, including lateroisthmic bypass grafting (subclavian artery or aortic arch to descending aorta) and ascending-to-descending aortic bypass, were initially met with enthusiasm because they avoided hazardous dissection, eliminated residual gradients, and produced good early results [6, 7]. Vijayanagar and associates [8] first described the technique of incising the posterior pericardium to gain access to the descending aorta for ascending-to-descending aortic bypass through a median sternotomy. This technique continues to be effective in selected adult patients, as recently reported by Connolly and associates [9] and Attenhofer-Jost and colleagues [10], and allows concomitant repair of intracardiac defects. On the basis of our experience with primary anatomic arch repair in neonates and infants, we have used an alternative strategy for recurrent aortic obstruction. Our technique relies on anatomic reconstruction, typically through a median sternotomy and using cardiopulmonary bypass with deep hypothermic circulatory arrest (DHCA). We believe that this strategy allows complete relief of obstruction at the level of the aortic arch and safe, single-stage repair of concomitant intracardiac de-
hildren undergoing operative repair of aortic coarctation and interrupted aortic arch are at a lifelong risk of recurrent aortic obstruction. We have previously reported an aggressive operative strategy for primary neonatal coarctation with associated aortic arch hypoplasia with the goal of reducing the incidence of residual and recurrent obstruction [1]. Since Mustard and associates [2] reported the first surgical experience with neonatal coarctation, these patients are well known to be at increased risk of recurrence. There is currently no consensus on the optimum operation for recurrent aortic obstruction. Initial reports of surgery for recurrent aortic obstruction were discouraging (operative mortality 15% to 33%) [3, 4]. Although Castaneda and Norwood [5] reported mortality less than 10%, a residual gradient of 40 mm Hg or more was noted in 30% of patients immediately after reoperation. These early results prompted investigation Accepted for publication Feb 17, 2004. Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003. Address reprint requests to Dr DiBardino, Congenital Heart Surgery, Texas Children’s Hospital, 6621 Fannin St, MC WT 19345H, Houston, TX 77030; e-mail: [email protected].
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.02.126
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
fects while avoiding or limiting conduit use in growing children. We describe the pattern of recurrence, our operative technique, and results.
Material and Methods Since 1995, 21 patients have presented with recurrent aortic arch obstruction after previous repair, 17 boys and 4 girls. After approval from the local institutional review board, a retrospective chart review was conducted such that all preoperative, operative, and hospital course data were reviewed, and the most recent follow-up data were extracted from the cardiology clinic database. Preoperative echocardiograms and catheterization cineangiograms were reviewed by two cardiologists (G.C.K. and G.L.) blinded to the operative technique and result. Patients in our single-ventricle palliation pathway were excluded from analysis, as obstruction of a Norwood reconstruction is not analogous to obstruction after repair of coarctation or interrupted aortic arch. Aside from patients previously undergoing Norwood reconstruction, all patients referred for surgery for recurrent aortic obstruction since 1995 were included in the analysis. Mean age at reoperation was 7.8 ⫾ 5.4 years (range, 0.21 to 15.2 years), and mean weight was 30.6 ⫾ 21.8 kg (range, 3.6 to 90 kg). Details concerning the initial aortic repair techniques are outlined in Table 1. Median age at primary repair was 9 days (range, 0.1 to 152.2 months), and median interval from primary repair to reoperation for recurrent aortic obstruction was 4.6 years (range, 0.18 to 14.2 years). At the time of referral to surgery, significant arm-to-leg blood pressure gradients were present in all but 1 patient in whom midarch obstruction and an aberrant right subclavian artery eliminated the possibility of a gradient (Table 1). The mean arm-to-leg gradient in the remaining patients was 35.5 ⫾ 15.2 mm Hg (range, 15 to 70 mm Hg).
Preoperative Imaging Preoperative transthoracic two-dimensional echocardiogram with Doppler interrogation was performed on all patients either as part of a routine screening protocol or to confirm the diagnosis in suspected cases. Recurrent aortic obstruction was diagnosed or confirmed by echocardiography in each case, and some degree of transverse arch hypoplasia was present in every patient; blinded echocardiographic review by one cardiologist (G.C.K.) revealed a mean preoperative transverse arch z score of ⫺2.9 ⫾ 1.6 (range, ⫺7.0 to 0.1). The patient with a transverse arch z score of 0.1 had undergone previous aortic arch advancement for coarctation with hypoplastic aortic arch and suffered local recurrence at the anastomotic site. The precise location of maximum obstruction along the aorta was variable, but included the aortic arch or proximal isthmus area in every case. Mean peak velocity across the location of maximum obstruction was 3.3 ⫾ 1.1 m/s (range, 1.3 to 4.8 m/s). Preoperative cardiac catheterization was performed in 13 patients (62%) for hemodynamic assessment of associated lesions or further delineation of arch anatomy
927
(Table 1). The mean gradient from ascending to descending aorta measured at the time of catheterization was 35.0 ⫾ 21.0 mm Hg (range, 10 to 63 mm Hg). One patient was found to have a resting gradient of 10 mm Hg and underwent isoproterenol infusion with a resulting gradient increase to 46 mm Hg. The severity of arch obstruction was underestimated by the gradient in several cases secondary to significant obstruction at the subaortic and aortic valvar level. One patient, for example, was found to have a 54-mm Hg gradient from the left ventricle to the descending aorta, with a 10-mm Hg component in the aortic arch. The appearance of the arch suggested that the severe left ventricular outflow tract obstruction minimized the significance of the arch obstruction and both were, thus, addressed at surgery. Seven patients (33%) underwent balloon angioplasty procedures before surgical referral, and these procedures were unsuccessful in reducing the gradient in each case. Since 1998, we have increasingly used preoperative magnetic resonance angiography (MRA) for further delineation of the aortic arch and descending aorta. Nine patients (43%) underwent preoperative MRA, and Figure 1A demonstrates the typical appearance of recurrence in our series, with significant obstruction visible at the level of the midarch.
Operative Technique At reoperation, thoracotomy was used for relief of aortic obstruction in 2 patients by re-resection with end-to-end anastomosis (n ⫽ 1) and patch aortoplasty (n ⫽ 1) using a clamp-and-sew technique. Thoracotomy was chosen because neither patient had concomitant cardiac defects and it was believed in both cases that the obstruction was distal enough to be safely addressed by thoracotomy and application of a cross-clamp. In the remaining 19 patients, complete relief of obstruction required median sternotomy, cardiopulmonary bypass, and DHCA. Aortic arch advancement was performed in 10 patients, and patch aortoplasty was used in 8 patients including patch extension into the proximal arch in every case. One 18-mm interposition graft was used in a 60-kg boy with recurrent coarctation, hypoplastic aortic arch, and aneurysmal dilation of the proximal descending thoracic aorta. Twenty concomitant cardiovascular procedures were conducted in 8 of the patients undergoing median sternotomy (Table 2). When performing aortic arch advancement, we ligate the aorta at the isthmus, freely mobilize the descending aorta, and perform an end-toside descending-to-ascending aortic anastomosis. Regional low-flow cerebral perfusion (RLFP) was used to minimize DHCA in 2 patients undergoing aortic arch advancement later in the series. A 6-mm Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) conduit was anastomosed to the proximal innominate artery, and this conduit was cannulated for the initiation of cardiopulmonary bypass. After cooling to 18°C, flow was reduced and a vascular clamp was placed on the innominate artery between the Gore-Tex conduit and the aortic arch, allowing regional brain perfusion during aortic arch advancement. Reduction of flow was guided by physiologic cerebral monitoring, as previously described [11].
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
928
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Ann Thorac Surg 2004;78:926 –32
Table 1. Operative Techniques and Blood Pressure Data for 21 Patients Undergoing Operation for Recurrent Aortic Obstruction Primary Repair Type
Patient No. 1
Repair Type for Recurrence
Immediate Postoperative Arm- Cuff BP at Last Leg BP Gradient Follow-up (mm Hg) (mm Hg)
135/44
26
60
Patch aortoplasty
0
135/79
125/64
45
58
0
111/57
108/62
34
15
0
106/50
132/64
31
N/A
0
126/72
150/82
50
N/A
0
136/71
128/65 137/71
70 30
63 N/A
0 0
139/66 121/63
136/71
68
47
0
95/48
107/62
30
30
Aortic arch advancement Aortic arch advancement 18-mm interposition graft Aortic arch advancement Patch aortoplasty Aortic arch advancement Redo aortic arch advancement Patch aortoplasty
0
82/57
116/64
None (aberrant right subclavian)
60
0
131/55
98/73
30
N/A
0
90/60
121/64
50
N/A
0
Death
105/49
15
24
Patch aortoplasty
0
109/71
106/73
37
46
Patch aortoplasty
0
97/50
142/64
40
N/A
Patch aortoplasty
20
135/57
138/76
25
20
0
100/53
Subclavian flap aortoplasty and patch aortoplasty REEA
145/55
40
N/A
0
114/47
128/81
21
N/A
0
96/49
158/66
32
46
20
129/64
20
Subclavian flap aortoplasty Direct anastomosis (IAA)
98/55
16
0
102/59
21
REEA
132/69
20
0
82/52
2 3 4 5 6 7 8 9
10
11 12
13 14 15 16 17
18 19
Ascending to descending aortic bypass Interposition graft (IAA) Subclavian flap aortoplasty Subclavian flap aortoplasty REEA
Preoperative Preoperative Arm- Gradient at Cuff BP Leg BP Gradient Catheterization (mm Hg) (mm Hg) (mm Hg)
REEA Subclavian flap aortoplasty Aortic arch advancement Direct anastomosis (IAA) Carotid artery turndown (IAA) Subclavian flap aortoplasty Subclavian artery advancement Subclavian flap aortoplasty Patch aortoplasty Subclavian flap aortoplasty REEA
BP ⫽ blood pressure; IAA ⫽ interrupted aortic arch; extended end-to-end anastomosis.
Resection with extended end-toend anastomosis Aortic arch advancement Aortic arch advancement
Aortic arch advancement Aortic arch advancement
Aortic arch advancement Patch aortoplasty
10 (severe Patch aortoplasty LVOT obstruction with 54-mm Hg gradient) 10 Patch aortoplasty LVOT ⫽ left ventricular outflow tract;
Results Operative Data There were no intraoperative deaths. The ischemic times for patients undergoing thoracotomy were 15 and 24 minutes. Among patients undergoing median sternotomy, mean cardiopulmonary bypass time was 164.9 ⫾
N/A ⫽ not available;
REEA ⫽ resection with
70.0 minutes (range, 83 to 387 minutes) and mean aortic cross-clamp time was 90.4 ⫾ 53.6 minutes (range, 53 to 291 minutes). Cumulative DHCA time averaged 27.4 ⫾ 8.7 minutes (range, 14 to 50 minutes) in patients not undergoing RLFP and was 6 and 16 minutes in patients undergoing RLFP as an adjunct to circulatory arrest. Among the 9 patients undergoing patch aortoplasty,
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
929
Fig 1. (A) Preoperative magnetic resonance imaging study of a patient with recurrent aortic obstruction after a subclavian advancement operation at 6 days of life. Note the significant anatomic obstruction at the level of the midarch, responsible for a 46-mm Hg gradient with isoproterenol infusion at the time of preoperative catheterization. (B) Postoperative magnetic resonance imaging study of the same patient after complete repair of recurrent aortic obstruction through a median sternotomy and patch aortoplasty under deep hypothermic circulatory arrest. Note the absence of obstruction at the level of the aortic arch.
patch material included autologous pericardium (n ⫽ 2), bovine pericardium (n ⫽ 1), aortic homograft (n ⫽ 2), pulmonary homograft (n ⫽ 1), and synthetic material in 3 patients. Invasive blood pressure monitoring of the upper and lower extremity compartments (proximal and distal to the obstruction) at the conclusion of the case demonstrated a 20-mm Hg gradient in 2 patients and no gradient in the remainder (Table 1). Both patients with residual gradient had undergone patch aortoplasty repair.
Hospital Course There was 1 hospital mortality in a 15-year-old boy who experienced acute suture line dehiscence 2 days after aortic arch advancement. The mean time to extubation was 1.4 ⫾ 1.1 days (range, 0 to 5 days). Eight patients required transient inotropic support, and mean time to weaning of all inotropic agents was 2.4 ⫾ 2.0 days (range, 0 to 6 days) for these patients. Average intensive care unit stay was 3.1 ⫾ 1.4 days (range, 2 to 7 days), and average time to hospital discharge was 6.1 ⫾ 2.3 days (range, 3 to 12 days). Nitroprusside was used for intravenous blood pressure control in the immediate postoperative period with transition to angiotensin-converting enzyme inhibitor therapy once oral intake was established. Complications included the development of a new left vocal cord paralysis in a patient undergoing median sternotomy (n ⫽ 1), reintubation for presumed pneumonia (n ⫽ 1), and narcotic withdrawal requiring readmission to the intensive care unit (n ⫽ 1). Transient arrhythmia occurred in 2 patients, including an episode of nonsustained ventricular tachycardia resolving without specific therapy. No gross neurologic deficits were found at discharge.
Follow-Up There have been no late deaths. All patients remain in a program of close surveillance consisting of clinical and
two-dimensional echocardiographic follow-up. The child undergoing cardiac transplantation experienced a severe episode of rejection requiring 3 days of mechanical circulatory support 10 months after patch aortoplasty. She made a full recovery and is currently without symptoms. The patient requiring pulmonary band replacement underwent successful band takedown 13 months after patch aortoplasty, and no residual ventricular septal defects were found at that time. This was the only patient to undergo follow-up left-heart cardiac catheterization, and direct blood pressure measurement of the ascending aorta and descending aorta revealed no gradient. A 5-year-old patient undergoing patch aortoplasty was diagnosed with epilepsy 1 year after operation and is currently well controlled on carbamazepine. No patients have undergone subsequent aortic intervention, and all are asymptomatic from a cardiovascular standpoint at a mean follow-up of 26.2 ⫾ 24.4 months (range, 0.33 to 85.2 months) postoperatively. Noninvasive arm and leg blood pressure measurements are obtained at clinic visits in all patients. One patient with no gradient immediately after surgery currently has a 10-mm Hg arm-to-leg gradient. Of the 2 patients with a 20-mm Hg gradient immediately after repair, one currently has a 10-mm Hg arm-to-leg gradient and underwent MRA that revealed no significant obstruction (Fig 1B), and the other currently has no arm-to-leg gradient. At least one follow-up two-dimensional echocardiogram with Doppler flow interrogation has been obtained in 17 of 20 surviving patients whereas the last 3 patients still await follow-up imaging. No evidence of significant aortic obstruction is present in any patient. Both patients undergoing thoracotomy at reoperation currently demonstrate a peak velocity of 2.5 m/s across the aortic arch and isthmus. Of those undergoing median sternotomy, 1 patient has a peak velocity of 2.6 m/s and 1 patient a peak velocity of 2.3 m/s across the arch and isthmus after patch
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
930
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Table 2. Twenty Concomitant Procedures Performed in 8 Patients Undergoing Operation for Recurrent Aortic Obstruction Procedure Ventricular septal defect closure Atrial septal defect closure Subaortic membrane resection Branch pulmonary arterioplasty Doty patch Aortic commissurotomy Left ventricular septal myomectomy Pulmonary artery band removal Replacement of pulmonary artery band Arterial switch operation Orthotopic heart transplantation Atrial septectomy Warden procedure
Frequency 3 3 3 2 1 1 1 1 1 1 1 1 1
aortoplasty, whereas 1 patient has a peak velocity of 2.5 m/s after aortic arch advancement. Two patients have evidence of worsening combined subaortic and aortic valvar stenosis with acceleration of flow across the left ventricular outflow tract to 2 to 4 m/s and will likely require operative intervention in the future.
Comment Risk Factors and Pathogenesis for Recurrent Aortic Obstruction There has been much debate concerning the risk factors for recurrent aortic obstruction after repair of coarctation or interrupted aortic arch. Beekman and associates [12] found that the incidence of recurrence was 1.5% of 197 patients undergoing coarctation repair beyond 3 years of age and 38% of 42 patients undergoing repair in the first 3 months of life. In a recent review of 103 patients undergoing primary coarctation repair, younger age and smaller absolute transverse arch diameter were predictors of shorter time to reintervention, whereas initial repair technique was not found to be associated with recurrence [13]. In the most recent published analysis, aortic arch hypoplasia and age younger than 1 month were confirmed as independent predictors of recurrence in 262 children undergoing coarctation repair, whereas initial repair technique was not associated with recurrence [14]. In our series, recurrent aortic obstruction was seen to involve the arch (to some degree) in every case, as evidenced be a mean preoperative aortic arch diameter nearly 3 standard deviations below expected. This was confirmed by catheterization, MRA studies, and intraoperative inspection. Additionally, the median age of initial repair was only 9 days old, and the median interval to recurrence was less than 5 years, when somatic growth was still occurring at a high rate. One possible interpretation of these data is that the recurrent aortic obstruction was related to impaired arch growth that manifested with increasing somatic growth. Another possible hypothesis
Ann Thorac Surg 2004;78:926 –32
is that inadequate relief of arch obstruction at the initial operation was the central factor in recurrence, also manifesting with increasing somatic growth. Although we only performed one of the primary operations in this series and so cannot be certain about the initial arch morphology, we believe that it is essential to completely address significant arch hypoplasia at the primary repair of neonatal coarctation. It is this belief that has led to our current strategy of using arch advancement for repair of neonatal coarctation with significant arch hypoplasia [1], and as experience accumulates, we hope to demonstrate the long-term benefits of this strategy.
Surveillance and Imaging Kappetein and associates [15] warn that the incidence of reoperation for recurrent aortic obstruction is a gross underestimate of the true incidence of recurrence. Despite a 5.8% reoperation rate, they found recurrence in 41% of patients surviving primary coarctation repair with blood pressure gradients of greater than 20 mm Hg in those undergoing catheterization and upper-to-lower limb gradients of more than 50 mm Hg in those undergoing exercise testing. This raises important questions concerning the optimal surveillance modality after repair. Therrien and associates [16] recently assessed a variety of clinical and imaging modalities and concluded that combined clinical assessment and MRA or clinical assessment and transthoracic two-dimensional echocardiogram with MRA reserved for suspicious findings were acceptable screening alternatives. We rely on the latter strategy and stress the need for lifelong cardiovascular surveillance after repair of aortic obstruction.
Options for Recurrent Arch Obstruction Balloon angioplasty has been increasingly used for recurrent aortic coarctation and has more recently been recommended as the procedure of choice by several centers [17]. The safety of this intervention was reinforced by the Valvuloplasty and Angioplasty of Congenital Anomalies Registry Data, which reported 4 deaths in 548 patients (0.7%) undergoing balloon dilation therapy for recurrent aortic obstruction [18]. Although it appears that isolated, short segment recurrent obstruction is amenable to balloon dilation, one third of the patients in our series had failed catheter intervention and were unlikely to have received long-term benefit from balloon dilation alone secondary to aortic arch hypoplasia and long segments of obstruction. Although stent placement can improve longterm patency after balloon dilation, we believe that transverse aortic arch stenting should be avoided in growing children. When faced with recurrence we approach each patient with the intention of providing an anatomic repair. Anatomic reconstruction techniques for recurrent aortic obstruction have traditionally included resection and reanastomosis, patch aortoplasty, and interposition grafting [19 –24]. Extraanatomic bypass techniques were born in an era of disappointing results for the operative treatment of recurrent coarctation. Although Connolly and associates [9] have recently obtained excellent results
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
with ascending-to-descending aortic bypass through a median sternotomy in older patients (15 to 66 years of age), they warn against the use of this technique in children and cite the known risk of anastomotic dehiscence with somatic growth. Although bypass conduits have been used in patients 4 to 11 years of age at other centers [25–27], we agree with Connolly and associates that pediatric patients are not best served by extraanatomic bypass and recommend anatomic repair whenever possible. The concept of aortic coarctation repair through a median sternotomy is not new, particularly in the face of recurrence or concomitant intracardiac defects [21–22, 28]. Because many patients with recurrent aortic obstruction have arch hypoplasia and concomitant intracardiac defects requiring repair, thoracotomy may not be the optimal approach. Two existing reports of patients with recurrent aortic obstruction have specifically described a subset of patients requiring attention to the aortic arch through a median sternotomy [21, 22]. Techniques included patch aortoplasty extended into the proximal arch and interposition grafting performed under DHCA, allowing complete relief of obstruction with excellent results. Our strategy deferrers from these in that we rely heavily on aortic arch advancement with the intention of providing complete relief of obstruction, minimizing potential recurrence at the level of arch, and avoiding conduit use. Although Younoszai and associates [29] and Lacour-Gayet and associates [30] describe similar operations performed by means of a thoracotomy and with a cross-clamp occluding the arch and left brachiocephalic vessels, we emphasize their descriptions that the descending aorta is anastomosed to the underside of the aortic arch and not to the ascending aorta, as in our technique. With each of these procedures, decreased elasticity in the aorta and scar formation may not allow sufficient mobilization of the descending aorta for a tension-free advancement in the older child or adult. We find patch aortoplasty to be an acceptable anatomic technique whenever the possibility of a tension-free anastomosis is in question. Regarding the use of RLFP, this is a relatively new technique for our unit that we have been using regularly since late 2000 for neonatal aortic arch reconstruction. As experience accumulates, we have expanded our indications for this technique and have been encouraged by studies showing its feasibility in the pediatric population [31, 32]. Although the patients undergoing RLFP in this series and in our Norwood population (unpublished data) have enjoyed greatly diminished circulatory arrest times in comparison with those with conventional DHCA, there has not been any published, objective scientific evidence that the use of RLFP improves clinical outcomes. We remain encouraged by its use and report good gross neurologic outcome [11] but cannot speculate on the ability of RLFP to impact on the subtle long-term derangements known to occur with extended DHCA times.
931
References 1. Elgamel M, McKenzie ED, Fraser CD Jr. Aortic arch advancement: the optimal one-stage approach for surgical management of neonatal coarctation with arch hypoplasia. Ann Thorac Surg 2002;73:1267–73. 2. Mustard WT, Rower RD, Keith JD, Sirek A. Coarctation of the aorta with special reference to the first year of life. Am Surg 1955;141:249 –52. 3. Ibarra-Perez C, Casteneda AR, Varco RL, Lillehei CW. Recoarctation of the aorta. Nineteen year clinical experience. Am J Cardiol 1969;23:778 –84. 4. Cerilli J, Lauridsen P. Reoperation for coarctation of the aorta. Acta Chir Scand 1965;129:391–4. 5. Casteneda AR, Norwood WI. Residual coarctation of the aorta: surgical experience. In: Tucker BL, Lindesmith GC, eds. Congenital heart disease. New York: Grune and Stratton, 1979:167–78. 6. Weldon CS, Hartmann AF, Steinhoff NG, Morrissey JD. A simple, safe and rapid technique for the management of recurrent coarctation of the aorta. Ann Thorac Surg 1973;15: 510 –9. 7. Edie RN, Janani J, Attai LA, Malm JR, Robinson G. Bypass grafts for recurrent or complex coarctation. Ann Thorac Surg 1975;20:558 –64. 8. Vijayanagar R, Natarajan P, Eckstein PF, Bognolo DA, Toole JC. Aortic valvular insufficiency and postductal aortic coarctation in the adult. J Thorac Cardiovasc Surg 1980;79:266 –8. 9. Connolly HM, Schaff HV, Izher U, et al. Posterior pericardial ascending-to-descending aortic bypass. Circulation 2001; 104(Suppl 1):I-133–7. 10. Attenhofer-Jost CH, Schaff HV, Connolly HM, et al. Spectrum of reoperations after repair of aortic coarctation: importance of an individualized approach because of coexistent cardiovascular disease. Mayo Clin Proc 2002;17:646 –53. 11. Andropoulos DB, Stayer SA, McKenzie ED, Fraser CD Jr. Novel cerebral physiologic monitoring to guide low-flow cerebral perfusion during neonatal aortic arch reconstruction. J Thorac Cardiovasc Surg 2003;125:491–9. 12. Beekman RH, Rocchini AP, Behrendt DM, Rsonethal A. Reoperation for coarctation of the aorta. Am J Cardiol 1981;48:1108 –14. 13. McElhinney OB, Yang S, Hogarty AN, et al. Recurrent arch obstruction after repair of isolated coarctation of the aorta in neonates and young infants: is low weight a risk factor? J Thorac Cardiovasc Surg 2001;122:883–90. 14. Walhout RJ, Lekkerkerker JC, Oron GH, et al. Comparison of polytetrafluoroethylene patch aortoplasty and end-to-end anastomosis for coarctation of the aorta. J Thorac Cardiovas Surg 2003;126:521–8. 15. Kappetein AP, Zwinderman AH, Rogers AJJC, Rohmer J, Huysmans HA. More than 35 years of coarctation repair: an unexpected high relapse rate. J Thorac Cardiovasc Surg 1994;107:87–95. 16. Therrien J, Thorne SA, Wright A, Kilner PJ, Somerville J. Repaired coarctation: a “cost effective” approach to identifying complications in adults. J Am Coll Cardiol 2000;35:997– 1002. 17. Siblini G, Rao PS, Nouri S, et al. Long-term follow-up results of balloon angioplasty of postoperative aortic recoarctation. Am J Cardiol 1998;81:61–7. 18. McGrindle BW, Jones TK, Morrow R, et al. Acute results of percutaneous balloon angioplasty of native aortic coarctation versus recurrent aortic obstruction are equivalent. J Am Coll Cardiol 1996;28:1810 –7. 19. Sweeney MS, Walker WE, Duncan JM, et al. Reoperation for aortic coarctation: technique, results and indications for various approaches. Ann Thorac Surg 1985;40:46 –9. 20. Kron IL, Flanagan TL, Rheuban KS, et al. Incidence and risk of reintervention after coarctation repair. Ann Thorac Surg 1990;49:920 –6. 21. Ralph-Edwards AC, Williams WG, Coles JC, Rebeyka IM,
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
932
22.
23.
24.
25.
26.
27.
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Trusler GA, Freedom RM. Reoperation for recurrent aortic coarctation. Ann Thorac Surg 1995;60:1303–7. Lange R, Thielmann M, Schmidt KG. Spinal cord protection using hypothermic cardiocirculatory arrest in extended repair of recoarcation and persistent hypoplastic aortic arch. Eur J Thorac Cardiovasc Surg 1997;11:697–702. Manganas C, Iliopoulos J, Chard RB, Nunn GR. Reoperation and coarctation of the aorta: the need for lifelong surveillance. Ann Thorac Surg 2001;72:1222–4. Wong CH, Watson B, Smith L, Hamilton JRL, Hasan A. The use of left heart bypass in adult and recurrent coarctation repair. Eur J Thorac Cardiovasc Surg 2001;20:1199 –201. Jacob T, Cobanoglu A, Starr A. Late results of ascending aorta-descending aorta bypass grafts for recurrent coarctation of aorta. J Thorac Cardiovasc Surg 1988;95:782–7. Grinda JM, Mace L, Dervanian P, Folliguet TA, Neveux JY. Bypass graft for complex forms of isthmic aortic coarctation. Ann Thorac Surg 1995;60:1299 –302. Barron DJ, Lamb RK, Ogilvie BC, Monro JL. Technique for
28. 29. 30.
31.
32.
Ann Thorac Surg 2004;78:926 –32
extraanatomic bypass in complex aortic coarctation. Ann Thorac Surg 1996;61:241–4. Ungerleider RM, Ebert PA. Indications and techniques for midline approach to aortic coarctation in infants and children. Ann Thorac Surg 1987;44:517–22. Younoszai AK, Reddy VM, Hanley FL, Brook MM. Intermediate term follow-up of the end-to-side aortic anastomosis for coarctation of the aorta. Ann Thorac Surg 2002;74:1631–4. 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. Pigula FA, Nemoto EM, Griffith BP, Siewers RD. Regional low-flow perfusion provides cerebral circulatory support during neonatal aortic arch reconstruction. J Thorac Cardiovasc Surg 2000;119:331–9. Pigula FA, Gandhi SK, Siewers RD, Davis PJ, Webber SA, Nemoto EM. Regional low-flow perfusion provides somatic circulatory support during neonatal aortic arch surgery. Ann Thorac Surg 2001;72:401–7.
DISCUSSION DR FRANCIS ROBICSEK (Charlotte, NC): Certainly a very interesting presentation. I have one question to the authors. Why do you stick to the “anatomic” approach? Our experience is that in larger children, and especially in adults, you can do an ascending aorta with an infrarenal graft, which takes about an hour, no cardiopulmonary bypass, minimum morbidity and mortality. We did it in about 16 such cases and to us it worked very well in obstructive lesions of the distal aortic arch. Naturally you can do it with recurrent aneurysms as a two-staged procedure as well. DR DIBARDINO: Thank you, Doctor Robicsek, for your comments. That is an excellent question, and it really drives home
our central point. The Mayo Clinic has recently described a similar technique, an ascending to descending bypass conduit, which is performed through the posterior pericardium via median sternotomy. They can do it off bypass or use bypass for any concomitant intracardiac repair. It is a great technique with excellent results in their hands. Being that we are a congenital heart surgery service, most of the patients that we have seen are younger children. The average age was only 7.8 years and the youngest were only infants. So we believe in approaching younger patients with the intention of performing an anatomic or a viable repair that can provide for viable growth. We feel that this is important in the long-term management of these patients, many of whom may require several reoperations in their lifetime.
Anatomic Reconstruction for Recurrent Aortic Obstruction in Infants and Children Daniel J. DiBardino, MD, Jeffrey S. Heinle, MD, Grace C. Kung, MD, Glenn T. Leonard, Jr, MD, Emmett D. McKenzie, MD, Jason T. Su, MD, and Charles D. Fraser, Jr, MD Division of Congenital Heart Surgery, Michael E. DeBakey Department of Surgery, and Division of Pediatric Cardiology, Department of Pediatrics, Baylor College of Medicine, Houston, Texas
Background. Patients undergoing operative repair of aortic obstruction are at a lifelong risk of recurrent obstruction, and there is controversy regarding the optimal surgical technique. We have used an alternative strategy for recurrent aortic obstruction, typically involving anatomic reconstruction by means of a median sternotomy, and describe our techniques and results. Methods. Twenty-one patients presented with recurrent aortic arch obstruction. Mean age and weight were 7.8 ⴞ 5.4 years (range, 0.21 to 15.2 years) and 30.6 ⴞ 21.8 kg (range, 3.6 to 90 kg), respectively. Recurrence involved the aortic arch to some degree in each case, as the mean preoperative transverse aortic arch z score was ⴚ2.9 ⴞ 1.6 (range, ⴚ7.0 to 0.1). Thoracotomy was possible in 2 patients, using re-resection with end-to-end anastomosis (n ⴝ 1) and patch aortoplasty (n ⴝ 1). The remaining 19 patients required median sternotomy, cardiopulmonary bypass, and deep hypothermic circulatory arrest for complete relief of obstruction by aortic arch advancement (n
ⴝ 10), patch aortoplasty (n ⴝ 8), or interposition grafting (n ⴝ 1). Results. There was 1 hospital death. Invasive blood pressure monitoring revealed no residual arm-to-leg gradient in 19 patients and a 20-mm Hg gradient in 2 patients. There have been no late deaths. No patients have undergone subsequent aortic intervention, and all are asymptomatic up to 85 months postoperatively. Two patients are currently followed with a 10-mm Hg arm-toleg blood pressure gradient. Conclusions. Anatomic reconstruction for recurrent aortic obstruction can be safely accomplished in the majority of patients. We favor median sternotomy because of the ability of establishing cardiopulmonary bypass, the facility of anatomic reconstruction techniques, and the ability to repair concomitant cardiovascular lesions. (Ann Thorac Surg 2004;78:926 –32) © 2004 by The Society of Thoracic Surgeons
C
into alternative repair strategies. Extraanatomic aortic bypass techniques, including lateroisthmic bypass grafting (subclavian artery or aortic arch to descending aorta) and ascending-to-descending aortic bypass, were initially met with enthusiasm because they avoided hazardous dissection, eliminated residual gradients, and produced good early results [6, 7]. Vijayanagar and associates [8] first described the technique of incising the posterior pericardium to gain access to the descending aorta for ascending-to-descending aortic bypass through a median sternotomy. This technique continues to be effective in selected adult patients, as recently reported by Connolly and associates [9] and Attenhofer-Jost and colleagues [10], and allows concomitant repair of intracardiac defects. On the basis of our experience with primary anatomic arch repair in neonates and infants, we have used an alternative strategy for recurrent aortic obstruction. Our technique relies on anatomic reconstruction, typically through a median sternotomy and using cardiopulmonary bypass with deep hypothermic circulatory arrest (DHCA). We believe that this strategy allows complete relief of obstruction at the level of the aortic arch and safe, single-stage repair of concomitant intracardiac de-
hildren undergoing operative repair of aortic coarctation and interrupted aortic arch are at a lifelong risk of recurrent aortic obstruction. We have previously reported an aggressive operative strategy for primary neonatal coarctation with associated aortic arch hypoplasia with the goal of reducing the incidence of residual and recurrent obstruction [1]. Since Mustard and associates [2] reported the first surgical experience with neonatal coarctation, these patients are well known to be at increased risk of recurrence. There is currently no consensus on the optimum operation for recurrent aortic obstruction. Initial reports of surgery for recurrent aortic obstruction were discouraging (operative mortality 15% to 33%) [3, 4]. Although Castaneda and Norwood [5] reported mortality less than 10%, a residual gradient of 40 mm Hg or more was noted in 30% of patients immediately after reoperation. These early results prompted investigation Accepted for publication Feb 17, 2004. Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003. Address reprint requests to Dr DiBardino, Congenital Heart Surgery, Texas Children’s Hospital, 6621 Fannin St, MC WT 19345H, Houston, TX 77030; e-mail: [email protected].
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
0003-4975/04/$30.00 doi:10.1016/j.athoracsur.2004.02.126
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
fects while avoiding or limiting conduit use in growing children. We describe the pattern of recurrence, our operative technique, and results.
Material and Methods Since 1995, 21 patients have presented with recurrent aortic arch obstruction after previous repair, 17 boys and 4 girls. After approval from the local institutional review board, a retrospective chart review was conducted such that all preoperative, operative, and hospital course data were reviewed, and the most recent follow-up data were extracted from the cardiology clinic database. Preoperative echocardiograms and catheterization cineangiograms were reviewed by two cardiologists (G.C.K. and G.L.) blinded to the operative technique and result. Patients in our single-ventricle palliation pathway were excluded from analysis, as obstruction of a Norwood reconstruction is not analogous to obstruction after repair of coarctation or interrupted aortic arch. Aside from patients previously undergoing Norwood reconstruction, all patients referred for surgery for recurrent aortic obstruction since 1995 were included in the analysis. Mean age at reoperation was 7.8 ⫾ 5.4 years (range, 0.21 to 15.2 years), and mean weight was 30.6 ⫾ 21.8 kg (range, 3.6 to 90 kg). Details concerning the initial aortic repair techniques are outlined in Table 1. Median age at primary repair was 9 days (range, 0.1 to 152.2 months), and median interval from primary repair to reoperation for recurrent aortic obstruction was 4.6 years (range, 0.18 to 14.2 years). At the time of referral to surgery, significant arm-to-leg blood pressure gradients were present in all but 1 patient in whom midarch obstruction and an aberrant right subclavian artery eliminated the possibility of a gradient (Table 1). The mean arm-to-leg gradient in the remaining patients was 35.5 ⫾ 15.2 mm Hg (range, 15 to 70 mm Hg).
Preoperative Imaging Preoperative transthoracic two-dimensional echocardiogram with Doppler interrogation was performed on all patients either as part of a routine screening protocol or to confirm the diagnosis in suspected cases. Recurrent aortic obstruction was diagnosed or confirmed by echocardiography in each case, and some degree of transverse arch hypoplasia was present in every patient; blinded echocardiographic review by one cardiologist (G.C.K.) revealed a mean preoperative transverse arch z score of ⫺2.9 ⫾ 1.6 (range, ⫺7.0 to 0.1). The patient with a transverse arch z score of 0.1 had undergone previous aortic arch advancement for coarctation with hypoplastic aortic arch and suffered local recurrence at the anastomotic site. The precise location of maximum obstruction along the aorta was variable, but included the aortic arch or proximal isthmus area in every case. Mean peak velocity across the location of maximum obstruction was 3.3 ⫾ 1.1 m/s (range, 1.3 to 4.8 m/s). Preoperative cardiac catheterization was performed in 13 patients (62%) for hemodynamic assessment of associated lesions or further delineation of arch anatomy
927
(Table 1). The mean gradient from ascending to descending aorta measured at the time of catheterization was 35.0 ⫾ 21.0 mm Hg (range, 10 to 63 mm Hg). One patient was found to have a resting gradient of 10 mm Hg and underwent isoproterenol infusion with a resulting gradient increase to 46 mm Hg. The severity of arch obstruction was underestimated by the gradient in several cases secondary to significant obstruction at the subaortic and aortic valvar level. One patient, for example, was found to have a 54-mm Hg gradient from the left ventricle to the descending aorta, with a 10-mm Hg component in the aortic arch. The appearance of the arch suggested that the severe left ventricular outflow tract obstruction minimized the significance of the arch obstruction and both were, thus, addressed at surgery. Seven patients (33%) underwent balloon angioplasty procedures before surgical referral, and these procedures were unsuccessful in reducing the gradient in each case. Since 1998, we have increasingly used preoperative magnetic resonance angiography (MRA) for further delineation of the aortic arch and descending aorta. Nine patients (43%) underwent preoperative MRA, and Figure 1A demonstrates the typical appearance of recurrence in our series, with significant obstruction visible at the level of the midarch.
Operative Technique At reoperation, thoracotomy was used for relief of aortic obstruction in 2 patients by re-resection with end-to-end anastomosis (n ⫽ 1) and patch aortoplasty (n ⫽ 1) using a clamp-and-sew technique. Thoracotomy was chosen because neither patient had concomitant cardiac defects and it was believed in both cases that the obstruction was distal enough to be safely addressed by thoracotomy and application of a cross-clamp. In the remaining 19 patients, complete relief of obstruction required median sternotomy, cardiopulmonary bypass, and DHCA. Aortic arch advancement was performed in 10 patients, and patch aortoplasty was used in 8 patients including patch extension into the proximal arch in every case. One 18-mm interposition graft was used in a 60-kg boy with recurrent coarctation, hypoplastic aortic arch, and aneurysmal dilation of the proximal descending thoracic aorta. Twenty concomitant cardiovascular procedures were conducted in 8 of the patients undergoing median sternotomy (Table 2). When performing aortic arch advancement, we ligate the aorta at the isthmus, freely mobilize the descending aorta, and perform an end-toside descending-to-ascending aortic anastomosis. Regional low-flow cerebral perfusion (RLFP) was used to minimize DHCA in 2 patients undergoing aortic arch advancement later in the series. A 6-mm Gore-Tex (W.L. Gore & Assoc, Flagstaff, AZ) conduit was anastomosed to the proximal innominate artery, and this conduit was cannulated for the initiation of cardiopulmonary bypass. After cooling to 18°C, flow was reduced and a vascular clamp was placed on the innominate artery between the Gore-Tex conduit and the aortic arch, allowing regional brain perfusion during aortic arch advancement. Reduction of flow was guided by physiologic cerebral monitoring, as previously described [11].
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
928
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Ann Thorac Surg 2004;78:926 –32
Table 1. Operative Techniques and Blood Pressure Data for 21 Patients Undergoing Operation for Recurrent Aortic Obstruction Primary Repair Type
Patient No. 1
Repair Type for Recurrence
Immediate Postoperative Arm- Cuff BP at Last Leg BP Gradient Follow-up (mm Hg) (mm Hg)
135/44
26
60
Patch aortoplasty
0
135/79
125/64
45
58
0
111/57
108/62
34
15
0
106/50
132/64
31
N/A
0
126/72
150/82
50
N/A
0
136/71
128/65 137/71
70 30
63 N/A
0 0
139/66 121/63
136/71
68
47
0
95/48
107/62
30
30
Aortic arch advancement Aortic arch advancement 18-mm interposition graft Aortic arch advancement Patch aortoplasty Aortic arch advancement Redo aortic arch advancement Patch aortoplasty
0
82/57
116/64
None (aberrant right subclavian)
60
0
131/55
98/73
30
N/A
0
90/60
121/64
50
N/A
0
Death
105/49
15
24
Patch aortoplasty
0
109/71
106/73
37
46
Patch aortoplasty
0
97/50
142/64
40
N/A
Patch aortoplasty
20
135/57
138/76
25
20
0
100/53
Subclavian flap aortoplasty and patch aortoplasty REEA
145/55
40
N/A
0
114/47
128/81
21
N/A
0
96/49
158/66
32
46
20
129/64
20
Subclavian flap aortoplasty Direct anastomosis (IAA)
98/55
16
0
102/59
21
REEA
132/69
20
0
82/52
2 3 4 5 6 7 8 9
10
11 12
13 14 15 16 17
18 19
Ascending to descending aortic bypass Interposition graft (IAA) Subclavian flap aortoplasty Subclavian flap aortoplasty REEA
Preoperative Preoperative Arm- Gradient at Cuff BP Leg BP Gradient Catheterization (mm Hg) (mm Hg) (mm Hg)
REEA Subclavian flap aortoplasty Aortic arch advancement Direct anastomosis (IAA) Carotid artery turndown (IAA) Subclavian flap aortoplasty Subclavian artery advancement Subclavian flap aortoplasty Patch aortoplasty Subclavian flap aortoplasty REEA
BP ⫽ blood pressure; IAA ⫽ interrupted aortic arch; extended end-to-end anastomosis.
Resection with extended end-toend anastomosis Aortic arch advancement Aortic arch advancement
Aortic arch advancement Aortic arch advancement
Aortic arch advancement Patch aortoplasty
10 (severe Patch aortoplasty LVOT obstruction with 54-mm Hg gradient) 10 Patch aortoplasty LVOT ⫽ left ventricular outflow tract;
Results Operative Data There were no intraoperative deaths. The ischemic times for patients undergoing thoracotomy were 15 and 24 minutes. Among patients undergoing median sternotomy, mean cardiopulmonary bypass time was 164.9 ⫾
N/A ⫽ not available;
REEA ⫽ resection with
70.0 minutes (range, 83 to 387 minutes) and mean aortic cross-clamp time was 90.4 ⫾ 53.6 minutes (range, 53 to 291 minutes). Cumulative DHCA time averaged 27.4 ⫾ 8.7 minutes (range, 14 to 50 minutes) in patients not undergoing RLFP and was 6 and 16 minutes in patients undergoing RLFP as an adjunct to circulatory arrest. Among the 9 patients undergoing patch aortoplasty,
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
929
Fig 1. (A) Preoperative magnetic resonance imaging study of a patient with recurrent aortic obstruction after a subclavian advancement operation at 6 days of life. Note the significant anatomic obstruction at the level of the midarch, responsible for a 46-mm Hg gradient with isoproterenol infusion at the time of preoperative catheterization. (B) Postoperative magnetic resonance imaging study of the same patient after complete repair of recurrent aortic obstruction through a median sternotomy and patch aortoplasty under deep hypothermic circulatory arrest. Note the absence of obstruction at the level of the aortic arch.
patch material included autologous pericardium (n ⫽ 2), bovine pericardium (n ⫽ 1), aortic homograft (n ⫽ 2), pulmonary homograft (n ⫽ 1), and synthetic material in 3 patients. Invasive blood pressure monitoring of the upper and lower extremity compartments (proximal and distal to the obstruction) at the conclusion of the case demonstrated a 20-mm Hg gradient in 2 patients and no gradient in the remainder (Table 1). Both patients with residual gradient had undergone patch aortoplasty repair.
Hospital Course There was 1 hospital mortality in a 15-year-old boy who experienced acute suture line dehiscence 2 days after aortic arch advancement. The mean time to extubation was 1.4 ⫾ 1.1 days (range, 0 to 5 days). Eight patients required transient inotropic support, and mean time to weaning of all inotropic agents was 2.4 ⫾ 2.0 days (range, 0 to 6 days) for these patients. Average intensive care unit stay was 3.1 ⫾ 1.4 days (range, 2 to 7 days), and average time to hospital discharge was 6.1 ⫾ 2.3 days (range, 3 to 12 days). Nitroprusside was used for intravenous blood pressure control in the immediate postoperative period with transition to angiotensin-converting enzyme inhibitor therapy once oral intake was established. Complications included the development of a new left vocal cord paralysis in a patient undergoing median sternotomy (n ⫽ 1), reintubation for presumed pneumonia (n ⫽ 1), and narcotic withdrawal requiring readmission to the intensive care unit (n ⫽ 1). Transient arrhythmia occurred in 2 patients, including an episode of nonsustained ventricular tachycardia resolving without specific therapy. No gross neurologic deficits were found at discharge.
Follow-Up There have been no late deaths. All patients remain in a program of close surveillance consisting of clinical and
two-dimensional echocardiographic follow-up. The child undergoing cardiac transplantation experienced a severe episode of rejection requiring 3 days of mechanical circulatory support 10 months after patch aortoplasty. She made a full recovery and is currently without symptoms. The patient requiring pulmonary band replacement underwent successful band takedown 13 months after patch aortoplasty, and no residual ventricular septal defects were found at that time. This was the only patient to undergo follow-up left-heart cardiac catheterization, and direct blood pressure measurement of the ascending aorta and descending aorta revealed no gradient. A 5-year-old patient undergoing patch aortoplasty was diagnosed with epilepsy 1 year after operation and is currently well controlled on carbamazepine. No patients have undergone subsequent aortic intervention, and all are asymptomatic from a cardiovascular standpoint at a mean follow-up of 26.2 ⫾ 24.4 months (range, 0.33 to 85.2 months) postoperatively. Noninvasive arm and leg blood pressure measurements are obtained at clinic visits in all patients. One patient with no gradient immediately after surgery currently has a 10-mm Hg arm-to-leg gradient. Of the 2 patients with a 20-mm Hg gradient immediately after repair, one currently has a 10-mm Hg arm-to-leg gradient and underwent MRA that revealed no significant obstruction (Fig 1B), and the other currently has no arm-to-leg gradient. At least one follow-up two-dimensional echocardiogram with Doppler flow interrogation has been obtained in 17 of 20 surviving patients whereas the last 3 patients still await follow-up imaging. No evidence of significant aortic obstruction is present in any patient. Both patients undergoing thoracotomy at reoperation currently demonstrate a peak velocity of 2.5 m/s across the aortic arch and isthmus. Of those undergoing median sternotomy, 1 patient has a peak velocity of 2.6 m/s and 1 patient a peak velocity of 2.3 m/s across the arch and isthmus after patch
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
930
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Table 2. Twenty Concomitant Procedures Performed in 8 Patients Undergoing Operation for Recurrent Aortic Obstruction Procedure Ventricular septal defect closure Atrial septal defect closure Subaortic membrane resection Branch pulmonary arterioplasty Doty patch Aortic commissurotomy Left ventricular septal myomectomy Pulmonary artery band removal Replacement of pulmonary artery band Arterial switch operation Orthotopic heart transplantation Atrial septectomy Warden procedure
Frequency 3 3 3 2 1 1 1 1 1 1 1 1 1
aortoplasty, whereas 1 patient has a peak velocity of 2.5 m/s after aortic arch advancement. Two patients have evidence of worsening combined subaortic and aortic valvar stenosis with acceleration of flow across the left ventricular outflow tract to 2 to 4 m/s and will likely require operative intervention in the future.
Comment Risk Factors and Pathogenesis for Recurrent Aortic Obstruction There has been much debate concerning the risk factors for recurrent aortic obstruction after repair of coarctation or interrupted aortic arch. Beekman and associates [12] found that the incidence of recurrence was 1.5% of 197 patients undergoing coarctation repair beyond 3 years of age and 38% of 42 patients undergoing repair in the first 3 months of life. In a recent review of 103 patients undergoing primary coarctation repair, younger age and smaller absolute transverse arch diameter were predictors of shorter time to reintervention, whereas initial repair technique was not found to be associated with recurrence [13]. In the most recent published analysis, aortic arch hypoplasia and age younger than 1 month were confirmed as independent predictors of recurrence in 262 children undergoing coarctation repair, whereas initial repair technique was not associated with recurrence [14]. In our series, recurrent aortic obstruction was seen to involve the arch (to some degree) in every case, as evidenced be a mean preoperative aortic arch diameter nearly 3 standard deviations below expected. This was confirmed by catheterization, MRA studies, and intraoperative inspection. Additionally, the median age of initial repair was only 9 days old, and the median interval to recurrence was less than 5 years, when somatic growth was still occurring at a high rate. One possible interpretation of these data is that the recurrent aortic obstruction was related to impaired arch growth that manifested with increasing somatic growth. Another possible hypothesis
Ann Thorac Surg 2004;78:926 –32
is that inadequate relief of arch obstruction at the initial operation was the central factor in recurrence, also manifesting with increasing somatic growth. Although we only performed one of the primary operations in this series and so cannot be certain about the initial arch morphology, we believe that it is essential to completely address significant arch hypoplasia at the primary repair of neonatal coarctation. It is this belief that has led to our current strategy of using arch advancement for repair of neonatal coarctation with significant arch hypoplasia [1], and as experience accumulates, we hope to demonstrate the long-term benefits of this strategy.
Surveillance and Imaging Kappetein and associates [15] warn that the incidence of reoperation for recurrent aortic obstruction is a gross underestimate of the true incidence of recurrence. Despite a 5.8% reoperation rate, they found recurrence in 41% of patients surviving primary coarctation repair with blood pressure gradients of greater than 20 mm Hg in those undergoing catheterization and upper-to-lower limb gradients of more than 50 mm Hg in those undergoing exercise testing. This raises important questions concerning the optimal surveillance modality after repair. Therrien and associates [16] recently assessed a variety of clinical and imaging modalities and concluded that combined clinical assessment and MRA or clinical assessment and transthoracic two-dimensional echocardiogram with MRA reserved for suspicious findings were acceptable screening alternatives. We rely on the latter strategy and stress the need for lifelong cardiovascular surveillance after repair of aortic obstruction.
Options for Recurrent Arch Obstruction Balloon angioplasty has been increasingly used for recurrent aortic coarctation and has more recently been recommended as the procedure of choice by several centers [17]. The safety of this intervention was reinforced by the Valvuloplasty and Angioplasty of Congenital Anomalies Registry Data, which reported 4 deaths in 548 patients (0.7%) undergoing balloon dilation therapy for recurrent aortic obstruction [18]. Although it appears that isolated, short segment recurrent obstruction is amenable to balloon dilation, one third of the patients in our series had failed catheter intervention and were unlikely to have received long-term benefit from balloon dilation alone secondary to aortic arch hypoplasia and long segments of obstruction. Although stent placement can improve longterm patency after balloon dilation, we believe that transverse aortic arch stenting should be avoided in growing children. When faced with recurrence we approach each patient with the intention of providing an anatomic repair. Anatomic reconstruction techniques for recurrent aortic obstruction have traditionally included resection and reanastomosis, patch aortoplasty, and interposition grafting [19 –24]. Extraanatomic bypass techniques were born in an era of disappointing results for the operative treatment of recurrent coarctation. Although Connolly and associates [9] have recently obtained excellent results
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
with ascending-to-descending aortic bypass through a median sternotomy in older patients (15 to 66 years of age), they warn against the use of this technique in children and cite the known risk of anastomotic dehiscence with somatic growth. Although bypass conduits have been used in patients 4 to 11 years of age at other centers [25–27], we agree with Connolly and associates that pediatric patients are not best served by extraanatomic bypass and recommend anatomic repair whenever possible. The concept of aortic coarctation repair through a median sternotomy is not new, particularly in the face of recurrence or concomitant intracardiac defects [21–22, 28]. Because many patients with recurrent aortic obstruction have arch hypoplasia and concomitant intracardiac defects requiring repair, thoracotomy may not be the optimal approach. Two existing reports of patients with recurrent aortic obstruction have specifically described a subset of patients requiring attention to the aortic arch through a median sternotomy [21, 22]. Techniques included patch aortoplasty extended into the proximal arch and interposition grafting performed under DHCA, allowing complete relief of obstruction with excellent results. Our strategy deferrers from these in that we rely heavily on aortic arch advancement with the intention of providing complete relief of obstruction, minimizing potential recurrence at the level of arch, and avoiding conduit use. Although Younoszai and associates [29] and Lacour-Gayet and associates [30] describe similar operations performed by means of a thoracotomy and with a cross-clamp occluding the arch and left brachiocephalic vessels, we emphasize their descriptions that the descending aorta is anastomosed to the underside of the aortic arch and not to the ascending aorta, as in our technique. With each of these procedures, decreased elasticity in the aorta and scar formation may not allow sufficient mobilization of the descending aorta for a tension-free advancement in the older child or adult. We find patch aortoplasty to be an acceptable anatomic technique whenever the possibility of a tension-free anastomosis is in question. Regarding the use of RLFP, this is a relatively new technique for our unit that we have been using regularly since late 2000 for neonatal aortic arch reconstruction. As experience accumulates, we have expanded our indications for this technique and have been encouraged by studies showing its feasibility in the pediatric population [31, 32]. Although the patients undergoing RLFP in this series and in our Norwood population (unpublished data) have enjoyed greatly diminished circulatory arrest times in comparison with those with conventional DHCA, there has not been any published, objective scientific evidence that the use of RLFP improves clinical outcomes. We remain encouraged by its use and report good gross neurologic outcome [11] but cannot speculate on the ability of RLFP to impact on the subtle long-term derangements known to occur with extended DHCA times.
931
References 1. Elgamel M, McKenzie ED, Fraser CD Jr. Aortic arch advancement: the optimal one-stage approach for surgical management of neonatal coarctation with arch hypoplasia. Ann Thorac Surg 2002;73:1267–73. 2. Mustard WT, Rower RD, Keith JD, Sirek A. Coarctation of the aorta with special reference to the first year of life. Am Surg 1955;141:249 –52. 3. Ibarra-Perez C, Casteneda AR, Varco RL, Lillehei CW. Recoarctation of the aorta. Nineteen year clinical experience. Am J Cardiol 1969;23:778 –84. 4. Cerilli J, Lauridsen P. Reoperation for coarctation of the aorta. Acta Chir Scand 1965;129:391–4. 5. Casteneda AR, Norwood WI. Residual coarctation of the aorta: surgical experience. In: Tucker BL, Lindesmith GC, eds. Congenital heart disease. New York: Grune and Stratton, 1979:167–78. 6. Weldon CS, Hartmann AF, Steinhoff NG, Morrissey JD. A simple, safe and rapid technique for the management of recurrent coarctation of the aorta. Ann Thorac Surg 1973;15: 510 –9. 7. Edie RN, Janani J, Attai LA, Malm JR, Robinson G. Bypass grafts for recurrent or complex coarctation. Ann Thorac Surg 1975;20:558 –64. 8. Vijayanagar R, Natarajan P, Eckstein PF, Bognolo DA, Toole JC. Aortic valvular insufficiency and postductal aortic coarctation in the adult. J Thorac Cardiovasc Surg 1980;79:266 –8. 9. Connolly HM, Schaff HV, Izher U, et al. Posterior pericardial ascending-to-descending aortic bypass. Circulation 2001; 104(Suppl 1):I-133–7. 10. Attenhofer-Jost CH, Schaff HV, Connolly HM, et al. Spectrum of reoperations after repair of aortic coarctation: importance of an individualized approach because of coexistent cardiovascular disease. Mayo Clin Proc 2002;17:646 –53. 11. Andropoulos DB, Stayer SA, McKenzie ED, Fraser CD Jr. Novel cerebral physiologic monitoring to guide low-flow cerebral perfusion during neonatal aortic arch reconstruction. J Thorac Cardiovasc Surg 2003;125:491–9. 12. Beekman RH, Rocchini AP, Behrendt DM, Rsonethal A. Reoperation for coarctation of the aorta. Am J Cardiol 1981;48:1108 –14. 13. McElhinney OB, Yang S, Hogarty AN, et al. Recurrent arch obstruction after repair of isolated coarctation of the aorta in neonates and young infants: is low weight a risk factor? J Thorac Cardiovasc Surg 2001;122:883–90. 14. Walhout RJ, Lekkerkerker JC, Oron GH, et al. Comparison of polytetrafluoroethylene patch aortoplasty and end-to-end anastomosis for coarctation of the aorta. J Thorac Cardiovas Surg 2003;126:521–8. 15. Kappetein AP, Zwinderman AH, Rogers AJJC, Rohmer J, Huysmans HA. More than 35 years of coarctation repair: an unexpected high relapse rate. J Thorac Cardiovasc Surg 1994;107:87–95. 16. Therrien J, Thorne SA, Wright A, Kilner PJ, Somerville J. Repaired coarctation: a “cost effective” approach to identifying complications in adults. J Am Coll Cardiol 2000;35:997– 1002. 17. Siblini G, Rao PS, Nouri S, et al. Long-term follow-up results of balloon angioplasty of postoperative aortic recoarctation. Am J Cardiol 1998;81:61–7. 18. McGrindle BW, Jones TK, Morrow R, et al. Acute results of percutaneous balloon angioplasty of native aortic coarctation versus recurrent aortic obstruction are equivalent. J Am Coll Cardiol 1996;28:1810 –7. 19. Sweeney MS, Walker WE, Duncan JM, et al. Reoperation for aortic coarctation: technique, results and indications for various approaches. Ann Thorac Surg 1985;40:46 –9. 20. Kron IL, Flanagan TL, Rheuban KS, et al. Incidence and risk of reintervention after coarctation repair. Ann Thorac Surg 1990;49:920 –6. 21. Ralph-Edwards AC, Williams WG, Coles JC, Rebeyka IM,
CARDIOVASCULAR
Ann Thorac Surg 2004;78:926 –32
CARDIOVASCULAR
932
22.
23.
24.
25.
26.
27.
DIBARDINO ET AL RECURRENT AORTIC OBSTRUCTION IN INFANTS AND CHILDREN
Trusler GA, Freedom RM. Reoperation for recurrent aortic coarctation. Ann Thorac Surg 1995;60:1303–7. Lange R, Thielmann M, Schmidt KG. Spinal cord protection using hypothermic cardiocirculatory arrest in extended repair of recoarcation and persistent hypoplastic aortic arch. Eur J Thorac Cardiovasc Surg 1997;11:697–702. Manganas C, Iliopoulos J, Chard RB, Nunn GR. Reoperation and coarctation of the aorta: the need for lifelong surveillance. Ann Thorac Surg 2001;72:1222–4. Wong CH, Watson B, Smith L, Hamilton JRL, Hasan A. The use of left heart bypass in adult and recurrent coarctation repair. Eur J Thorac Cardiovasc Surg 2001;20:1199 –201. Jacob T, Cobanoglu A, Starr A. Late results of ascending aorta-descending aorta bypass grafts for recurrent coarctation of aorta. J Thorac Cardiovasc Surg 1988;95:782–7. Grinda JM, Mace L, Dervanian P, Folliguet TA, Neveux JY. Bypass graft for complex forms of isthmic aortic coarctation. Ann Thorac Surg 1995;60:1299 –302. Barron DJ, Lamb RK, Ogilvie BC, Monro JL. Technique for
28. 29. 30.
31.
32.
Ann Thorac Surg 2004;78:926 –32
extraanatomic bypass in complex aortic coarctation. Ann Thorac Surg 1996;61:241–4. Ungerleider RM, Ebert PA. Indications and techniques for midline approach to aortic coarctation in infants and children. Ann Thorac Surg 1987;44:517–22. Younoszai AK, Reddy VM, Hanley FL, Brook MM. Intermediate term follow-up of the end-to-side aortic anastomosis for coarctation of the aorta. Ann Thorac Surg 2002;74:1631–4. 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. Pigula FA, Nemoto EM, Griffith BP, Siewers RD. Regional low-flow perfusion provides cerebral circulatory support during neonatal aortic arch reconstruction. J Thorac Cardiovasc Surg 2000;119:331–9. Pigula FA, Gandhi SK, Siewers RD, Davis PJ, Webber SA, Nemoto EM. Regional low-flow perfusion provides somatic circulatory support during neonatal aortic arch surgery. Ann Thorac Surg 2001;72:401–7.
DISCUSSION DR FRANCIS ROBICSEK (Charlotte, NC): Certainly a very interesting presentation. I have one question to the authors. Why do you stick to the “anatomic” approach? Our experience is that in larger children, and especially in adults, you can do an ascending aorta with an infrarenal graft, which takes about an hour, no cardiopulmonary bypass, minimum morbidity and mortality. We did it in about 16 such cases and to us it worked very well in obstructive lesions of the distal aortic arch. Naturally you can do it with recurrent aneurysms as a two-staged procedure as well. DR DIBARDINO: Thank you, Doctor Robicsek, for your comments. That is an excellent question, and it really drives home
our central point. The Mayo Clinic has recently described a similar technique, an ascending to descending bypass conduit, which is performed through the posterior pericardium via median sternotomy. They can do it off bypass or use bypass for any concomitant intracardiac repair. It is a great technique with excellent results in their hands. Being that we are a congenital heart surgery service, most of the patients that we have seen are younger children. The average age was only 7.8 years and the youngest were only infants. So we believe in approaching younger patients with the intention of performing an anatomic or a viable repair that can provide for viable growth. We feel that this is important in the long-term management of these patients, many of whom may require several reoperations in their lifetime.