- Email: [email protected]
Reparative operations for interrupted aortic arch with ventricular septal defect
J THoRAc CARDIOVASC SURG 86:832-837, 1983
Reparative operations for interrupted aortic arch with ventricular septal defect From January. 1975. through September. 1982. 24 infants underwent primary or staged repair of interrupted aortic arch (fAA) with ventricular septal defect (VSD). Seven patients had fAA type A and 17 patients had type B. Eleven of the patients. median age 5 days. underwent staged operations and 13 infants. median age 6 days. underwent primary repair. Palliation was by tube graft interposition (six). subclavian-aortic anastomosis (three). left carotid-aortic anastomosis (one). or end-to-side aortic anastomosis (one) combined with pulmonary artery banding (eight) or early VSD closure. With palliation. there were three (27%) early deaths among the eleven patients and one (13%) late death among the eight remaining. Delayed repair at 5 days to 14 months (median 7 months) in seven patients incurred three (43%) early and no late deaths. Primary repair in 13 patients consisted of VSD closure combined with graft interposition (12) or end-to-side aortic anastomosis (one), with three (23 %) early and no late deaths. Nine of 14 survivors had hemodynamic evaluation by catheterization 1 to 3 years following repair. None had a significant residual VSD or pressure gradients between the ascending and thoracic aorta. Six had subaortic stenosis. two mild (gradient < 20 mm Hg) and four severe (gradient> 50 mm Hg). necessitating operation. Results of operations in neonates with fAA continue to improve. Essential in management is an awareness that subaortic stenosis and hypocalcemia may be accompaniments of this anomaly. Based on these data, we prefer primary repair for fAA with VSD.
William I. Norwood, M.D., Peter Lang, M.D. (by invitation), Aldo R. Castaneda, M.D., and Thomas J. Hougen, M.D. (by invitation), Boston, Mass.
Interrupted aortic arch (lAA) is an uncommon congenital cardiovascular malformation which was initially described by Steidele' in 1778. In earlier communications from this institution, this anomaly was reported to have been identified in 1.4% of autopsied infants with congenital heart disease and 1.3% of infants in the New England Regional Infant Cardiac Program.' If IAA is untreated, the median age at death is 4 to 10 days, usually following physiological closure of the ductus arteriosus.' Although IAA can occur in conjunction with many complex congenital cardiac malformations (or rarely in isolation), most often the constellation of IAA, ventricular septal defect (VSD), and patent ductus arteriosus coexist. In the late 1950s, several case reports of reparative operations for IAA and VSD appeared.v' Subsequently, a variety of approaches to surgical palliation or primary repair have been advanced.v'? This From the Departments of Cardiovascular Surgery and Cardiology, Children's Hospital Medical Center, and the Departments of Surgery and Pediatrics, Harvard Medical School, Boston, Mass. Read at the Sixty-third Annual Meeting of The American Association for Thoracic Surgery, Atlanta, Ga., April 25-27, 1983. Address for reprints: Aldo R. Castaneda, M.D., Children's Hospital Medical Center, 300 Longwood Ave., Boston, Mass. 02115.
832
communication reviews a 7 year experience with both primary and staged repair of IAA and VSD with emphasis on late postoperative outcome and hemodynamic evaluation.
Methods and patients From January, 1976, through January, 1983, 24 children have undergone palliative and/or reparative operations for IAA and VSD at the Children's Hospital, Boston, Massachusetts. Patients with severe coarctation or hypoplasia of an aortic arch segment but not interruption are excluded from this review. Also excluded are 10 patients with IAA in whom associated intracardiac malformations other than VSD (single ventricle in three, aortic atresia in one, truncus arteriosus in five, and double-outlet right ventricle with subpulmonary VSD in one) dominated the course of surgical management and outcome. The median age at operation was 5 days; 22 patients were less than 1 month of age, one infant was 10 months old, and one child was 4 years old. All 22 of the neonates had cardiomegaly and increased pulmonary blood flow on chest roentgenogram; at the time of presentation, 16 (73%) had some degree of metabolic acidosis attributed to insufficient systemic perfusion through a restrictive
Volume 86 Number 6 December, 1983
ductus arteriosus. Since 1977, prostaglandin E 1 infusion has been used for preoperative resuscitation of neonates with a closing ductus arteriosus and metabolic acidosis. Seven patients had type A interrupted aortic arch and the remainder (17) had type B." The VSD was muscular in two, malalignment type in eight, and infundibuloventricular in the remainder. The correct diagnosis was established by preoperative cardiac catheterization in 21 patients. IAA was initially diagnosed in two infants undergoing operations for suspected coarctation of the aorta and in one premature neonate (1,200 gm, 27 weeks' gestation) at the time of ligation of the ductus arteriosus. Eleven patients were managed with staged operations and 13 underwent primary reconstruction of all coexisting defects. Palliative operation. Among the patients who underwent initial palliative procedures, six had IAA type A and five had type B. The median age at operation was 5 days; with the exception of the 4-year-old patient, the age range was 2 to 17 days. The ductus arteriosus was ligated in all at the initial operation. The interrupted aortic segment was managed by interposition of a tube graft (8 to 14 mm in diameter) in six patients. An end-to-side anastomosis of the left subclavian artery to the thoracic aorta was used in three infants with IAA type A. In one neonate with IAA type B, an end-to-side anastomosis of the left carotid artery to the aortic isthmus was made. An end-to-end anastomosis was established in one patient with IAA type A. A pulmonary artery band was placed in eight patients to control pulmonary blood flow. In three neonates, no pulmonary artery band was placed and the initial procedure was followed by closure of the VSD 5 days to 6 weeks later. One patient with severe posterior deviation of the infundibular septum causing significant subaortic stenosis had placement of a valved conduit between the apex of the left ventricle and the thoracic aorta at age 5 days combinedwith graft interposition between the ascending and descending thoracic aorta and placement of a pulmonary artery band. Seven survivors of initial palliation have undergone pulmonary artery band removal and/or VSD closure 5 days to 16 months (median 7 months) after initial operation. Primary reconstructive operation. Twelve patients undergoing primary reconstructive procedures had IAA type B and one neonate had type A. These patients ranged in age from 1 to 21 days with the exception of the lo-month-old infant. The interruption was managed by direct anastomosis of the aortic segments in one neonate.One patient had an anomalous right subclavian
fAA with VSD
833
Table I. Mortality from staged and primary repair No. of patients Staged repair Stage I Stage II Primary repair Totals
Deaths Early
Late
II
II 7 13
3 3 3
24
9 (38%)
I (4%)
artery anastomosed end to side to the aortic arch but could not be weaned from cardiopulmonary bypass and a tube graft was subsequently interposed. The remainder had graft interposition between the ascending aorta and the descending thoracic aorta. The distal anastomosis was carried out through a left thoracotomy. Subsequently, the proximal anastomosis was made through a median sternotomy. The ductus arteriosus was then ligated and cardiopulmonary bypass was established for repair of the VSD. The VSD was closed via a right atriotomy with the use of deep hypothermia and circulatory arrest in all but one patient. Perioperative hypocalcemia occurred in 13 patients (55%) and of these, three patients proved to have DiGeorge's syndrome (III-IV pharyngeal pouch deficiency) with an absent thymus gland, no circulating T cells, and a persistent requirement for supplemental calcium. Five patients (20%) had or developed severe subaortic stenosis. In four patients with long-segment obstruction caused by a posteriorly deviated infundibular septum, a conduit was interposed between the apex of the left ventricle and the thoracic aorta at 5 days, 3 weeks, 9 months, and 13 months with no mortality. The other patient had resection of discrete fibrous subaortic stenosis at the age of 3 years. Results Among the 24 patients in this series, 14 (59%) children are alive and clinically well 3 months to 6 years following either primary or staged reconstructive operations (Table I). Among the 11 patients with initial palliative operations, there were three hospital deaths; two infants died at the time of operation and one patient died 4 weeks after IAA repair and pulmonary artery banding. This infant had a coexisting annular pancreas which contributed to the outcome. There was one late death 6 months after operation, caused by an arrhythmia on induction of general anesthesia for cleft lip repair. Of the seven patients who subsequently underwent pulmonary artery band removal and VSD closure, three patients died-one as the result of disruption of the aortic graft on opening the sternum, one of a
The Journal of Thoracic and Cardiovascular
Norwood et al.
834
Surgery
Table II. Hemodynamic data following palliative surgery operation Case No.
Catheterization interval (mo)
1 12 3 16 5
I
2 3 4 5
VSD (QpIQs)
Subaortic PSEG (mm Hg)
Aortic arch PSEG (mm Hg)
2.0
0 0 0 16 0
50 50 40 0 0
NC
3.5 0.9 NC
Comments Subclavian-aorta anastomosis Carotid-aorta anastomosis Subclavian-aorta anastomosis
Legend: VSD. Ventricular septal defect. Qp/Qs, Pulmonary-to-systemic flow. PSEG. Peak systolic ejection gradient. NC, Not calculated.
Table m. Hemodynamic data following reparative operation Case No.
Catheterization interval (mo)
VSD (QpIQs)
6 7 8 9 10 11 12 13 14
14 14 12 27 8 1 36 4 12
l.l l.l
1.0 1.0 1.0 l.l
1.6 1.0 1.0
Subaortic PSEG (mmHg)
0 18 145 72
62 50 14 0
Aortic arch PSEG (mmHg)
0 0 20 16 0 0 0 0 0
For legend see Table II.
technical error in VSD closure, and the third because of late recognition of ventilatory insufficiency. There have been no 'late deaths among the four survivors of staged repair. Among the 13 infants who underwent primary reconstructive procedures, there were three operative deaths and no late deaths. Twelve patients have undergone postoperative cardiac catheterization. In five patients, the study followed palliation and was preliminary to VSD closure (Table 11). Two of these five children had no pressure gradient between the ascending and thoracic aorta 10 and 16 months following tube graft interposition. One child, treated initially by anastomosis of the left carotid artery to the aorta at age 4 days, had a 50 mm Hg peak systolic ejection gradient between the ascending and thoracic aorta. The other two children had a 40 mm Hg systolic gradient across a subclavian-aortic anastomosis. These obstructions were repaired in conjunction with VSD closure and removal of a pulmonary artery band. Nine children were studied 1 to 3 years following primary or staged repair (Table III). One had a residual VSD which was small (pulmonary-to-systemic flow < 1.5). None had a significant systolic pressure gradient between the ascending and thoracic aorta. Six children had evidence of subaortic stenosis. In two
patients, the pressure gradient was less than 20 mm Hg. Four children had severe stenosis (peak systolic ejection gradient> 50 mm- Hg) treated by left ventricular apical-aortic conduit interposition in three infants or resection of a subaortic membrane in one child, with no mortality (Fig. 1).
Discussion The natural history of lAA and VSD dictates early surgical intervention.' Physiological closure of the ductus arteriosus results in diminished systemic perfusion, acidosis, and death. Thus, to alter the outcome of this condition, surgical therapy is most often necessary in the neonatal period. Twenty-two patients in this series were less than 1 month of age at operation. Occasionally, the ductus arteriosus remains patent. In this circumstance, surgical management in infancy is usually necessary because of intractable congestive heart failure, severe failure to thrive, and pulmonary artery hypertension. Rarely, as occurred in one 4-year-old child in this series, the ductus arteriosus remains patent and congestive heart failure can be controlled by anticongestive therapy. In such cases, however, irreversible obstructive pulmonary vascular disease is likely to develop. The goals of surgical therapy for lAA with VSD are to establish continuity between the segments of aorta and to separate the systemic and pulmonary circulations. These goals can be achieved concomitantly or in stages." In this series, primary repair yielded a lower total mortality than the various forms of staged reconstruction. Although staged repair remains a potentially useful technique, our current preference is primary repair. This approach has been facilitated by prostaglandin EI> which allows preoperative resuscitation of the moribund neonate with metabolic acidosis, permitting complete evaluation at cardiac catheterization and contributing to hemodynamic stability during the early phases of operation. Several ingenious techniques have been advanced to establish continuity between the separated ascending and descending thoracic aorta, many of which utilize
Volume 86
lAA with VSD
Number 6 December. 1983
principally autologous vascular tissue to repair the aortic arch.":" Although others advocate the use of the ductus arteriosus in aortic reconstruction, we have previously observed that ductal tissue used to bridge the gap between the ascending and descending aorta in neonates does not provide a reliable conduit since it can constrict and result in significant late stenosis. Utilization of the subclavian artery or carotid artery to bridge the gap, in this experience, resulted in significant restriction of flow early after operation. However, the restrictive region was discrete and easily enlarged with a patch. A direct anastomosis between the ascending and descending thoracic aorta is certainly feasible in some patients and was applied in two patients in this series. The hypothetical advantage of this approach is the potential of providing long-term unobstructed continuity between the ascending and descending aorta. Of note, interposition of a relatively large tube graft (about 10 mID in the neonate) proved a reliable means of establishing unobstructed continuity between the ascending and descending aorta in these patients. Such a reconstruction is intuitively expected to require revision as the child grows. On the other hand, none of the patients who underwent cardiac catheterization 1 to 3 years following graft interposition had a significant pressure gradient between the ascending and descending thoracic aorta . Moreover, neither of the two oldest children in this series, now 5 and 6 years after operation, have upper extremity hypertension. If the tube graft is analogous to coarctation of the aorta, one might expect grafts of the size used in this series not to produce significant hypertension for many years. Seven of the 24 patients in this series (29%) had or developed some degree of subaortic stenosis. Five had surgical management of the stenosis. One patient had a discrete subvalvular fibrous ring, but the remainder had significant long-segment muscular obstruction which did not lend itself to direct reparative techniques. Placement of a valved conduit between the apex of the left ventricle and thoracic aorta provided a .means of decompressing the severely hypertensive left ventricle in these small patients. 16 As has been pointed out by other observers, perioperative hypocalcemia occurs frequently in patients with lAA. 17• 18 In this series, 13 patients had significant hypocalcemia «7 mg/dl) necessitating prolonged infusions of calcium ion. Only three patients proved to have DiGeorge's syndrome by plasma calcium and circulating T cell determinations. The subjective impression was that all patients in this report had distinctly less thymic tissue than is ordinarily seen in neonates and infants.
835
Fig. 1. Frame of cineangiogram in the anteroposterior projection with injection into the left ventricle. An arterial catheter has been placed through the valved conduit interposed between the apex of the left ventricle and thoracic aorta. The stud y visualizes the subaortic stenosis and the conduit.
Although the mortality in this series is appreciable, results of operations in neonates and infants 'with IAA continue to improve. Based on these data, primary repair of IAA with a VSD is our current preference. An awareness that subaortic stenosis and hypocalcemia may be accompaniments of this anomaly is essential in the management of these patients. While direct aorto-aortic anastomosis is appealing and its role in the management of IAA should be further clarified, graft interposition appears to provide unobstructed aortic continuity for years. REFERENCES
2
3
4
5
6
Steidele RJ : Sammig. Verschiedener in der chirug. 'prakt. Lehrschule Gemachten Beobb 2: 114, 1777-1778 Fyler DC, Parisi LF, Berman MA: Regiona1ization of infant cardiac care in New England. Cardiovasc Clin 4:339-340, 1972 Collins-Nakai RL, Dick M, Parisi-Buckley L, Fyler DC, Castaneda AR: Interrupted aortic arch in infancy. J Pediatr 88:959-962, 1976 Merrill DL, Webster CA, Samson PC : Congen ital absence of the aortic isthmu s. J TTIIORAC SURG 33:311320, 1957 Van Praagh R, Bernhard WF, Rosenthal A, Parisi LF, Fyler DC: Interrupted aortic arch. Surgical treatment. Am J Card ioI27:200-211 , 1971 Quie PG, Novick R, Adams P, Anderson RC, Varco RL: Congenital interru ption of the aortic arch. J Pediatr 54:87 -92, 1959
836
The Journal of Thoracic and Cardiovascular Surgery
Norwood et al.
7 Blake HA, Manion WC, Spencer FC: Atresia or absence of the aortic isthmus. 1 TTHORAC CARDIOVASC SURG 43:607-614, 1962 8 Tyson KRT, Harris LC, Ngiem QX: Repair of aortic arch interruption in the neonate. Surgery 67:1006-1010, 1970. 9 Litwin SB, Van Praagh R, Bernhard WF: A palliative operation for certain infants with aortic arch interruption. Ann Thorac Surg 14:369-375, 1972 10 Jones EL, Plauth WH, Hatcher CR Jr: A palliative operation for all types of aortic arch interruption in the neonate. 1 TTHORAC CARDIOVASC SURG 69:579-584, 1975 II Celoria GC, Patton RB: Congenital absence of the aortic arch. Am Heart 1 58:407-413, 1959 12 Barratt-Boyes BG, Nicholls IT, Brandt PWT, Neutze 1M: Aortic arch interruption associated with patent ductus arteriosus, ventricular septal defect, and total anomalous pulmonary venous connection. J TTHORAC CARDIOVASC SURG 63:367-373, 1972 13 Trusler GA, lzukawa T: Interrupted aortic arch and ventricular septal defect. Direct repair through a median sternotomy incision in a 13-day-old infant. J TTHORAC CARDIOVASC SURG 69: 126-131, 1975 14 Monro Jl., Brown W, Conway N: Correction of Type B interrupted aortic arch with ventricular septal defect in infancy. J TTHORAC CARDIOVASC SURG 74:618-623, 1977 15 Moulton AL, Bowman FO lr: Primary definitive repair of type B interrupted aortic arch, ventricular septal defect, and patent ductus arteriosus. Early and late results. 1 TTHORAC CARDIOVASCSURG 82:501-510,1981 16 Norwood WI, Lang P, Castaneda AR, Murphy 10: Management of infants with left ventricular outflow obstruction by conduit interposition between the ventricular apex and thoracic aorta. J THORAC CARDIOVASC SURG 86:771-776,1983 17 Harvey lC, Dungan WT, Elders Ml, Hughes ER: Third and fourth pharyngeal pouch syndrome, associated with vascular anomalies and hypocalcemia seizures. Clin Pediatr 9:496-499, 1970 18 Conley ME, Beckwith lB, Mancer lFK, Tenckhoff L: The spectrum of the DiGeorge syndrome. J Pediatr 94:883-890, 1979
Discussion DR. RONALD C. ELKINS Oklahoma City. Okla.
We have recently reviewed and reported our experience with this entity, comprising 17 patients. Of these 17 patients, 14 have had palliative procedures; of the 14, there were 11 who survived the initial palliation, this being a reconstruction of the aorta with a 6 mm polytetrafluoroethylene (Gore-Tex or Impra) graft and banding of the pulmonary artery in those children with increased pulmonary blood flow. Of the children undergoing palliation, six had primary
VSDs associated with the lAA. Those six patients have been long-term survivors. Five of the patients have had their VSDs corrected. In one of the children, on repeat catheterization, the VSD was of such size that it did not warrant closure. All but one of these patients have had postoperative catheterization, and we have identified one patient who required reoperation due to the size of the child's reconstructed aorta. A significant gradient between the ascending aorta and descending thoracic aorta was present, and the child had placement of an additional 12 mm Dacron graft. This experience had led us to continue to recommend palliative treatment for patients with lAA, particularly those with VSD. We respect the ability of those who do this as a single-stage procedure. However, this is a critically ill group of patients, and if they can be prepared by a procedure limited to a left thoracotomy and then have a VSD repair at a later date with a reasonable success rate, this certainly should be recommended. I would ask two questions of Dr. Norwood. What size of graft material and what type of graft material do you utilize for reconstruction of the aorta? My second question pertains to my suspicion that there is a time frame between your patients who were treated by palliation and those treated by primary repair ofthe VSD as well as correction of the lAA. If you were doing the staged procedure now, it is not likely that your repair of a VSD as a second stage operation would carry a lower mortality than your reported one of >50%? DR. LEONARD L. BAILEY Loma Linda. Calif.
Dr. Norwood has presented an impressive experience in the management of what we have come to call simple lAA complex, that is, those babies who have just a VSD and possibly an atrial septal defect associated with lAA type A or B. He has clarified two issues for me. First, complete repair at the time of initial diagnosis is less risky than a stage procedure, Dr. Elkins' experience notwithstanding. Second, most of these babies should survive the corrective operation. The impressive mortality of 23% in the newborn period in his series is exceptional. I take issue, however, with Dr. Norwood's experience regarding the use of prosthetic tubes in tiny babies to establish aortic continuity. I feel strongly that growing patients should have growth potential of their repairs. Establishing direct aortic continuity is just as easy, if not easier, than inserting a prosthetic tube in these babies, particularly if one uses a bilateral thoracotomy. The experience of Drs. Bowman, Trusler, Moulton, and our own experience would establish ease and efficacy of direct aortic anastomosis. Anastomotic growth has been well documented. Since late 1976 we have repaired lAA complex in II newborn infants. Seven of these had the simple variety of lAA with a VSD. A few of them had atrial septal defects. Five survived total repair, for a mortality of 29%. We have not found subaortic stenosis in the four who have been restudied. I agree with Dr. Norwood that definitive repair is the
Volume 86 Number 6 December, 1983
procedure of choice. I only question the use of prosthetic tubes for the repair. DR. NOR WOOD (Closing) I would like to thank the discussers for raising some interesting aspects of management of these critically ill infants. We have used both Teflon and Dacron grafts. The size almost always was 10 mrn, although a few patients had 8 or 12 mm grafts placed. It is true that our palliative operations tended to be performed earlier in this 7 year period of review. Although it is our current preference to proceed with primary repair, I must say that our experience is not sufficient to clearly choose between palliation and primary repair. I think there is still considerable room for individualization. However, I would add
lAA with VSD
837
that the use of prostaglandin E 1 to establish good metabolic balance prior to the operation has contributed as much as anything to improving the results. Dr. Bailey raised the issue of the aorta-aortic anastomosis. Obviously, it is very appealing to avoid a tube graft that the child will certainly outgrow. However, our experience with aorta-aortic anastomosis is quite limited. One of the patients had compression of the left main-stem bronchus by the aorta following direct anastomosis. This problem was somewhat troublesome in the early postoperative period. There is no question in my mind that these patients will be returning with limited flow through the tube graft. On the other hand, if one thinks in terms of an analogous experience with coarctation of the aorta, upper extremity hypertension may not occur until quite late with a 10 mrn lumen.
Reparative operations for interrupted aortic arch with ventricular septal defect From January. 1975. through September. 1982. 24 infants underwent primary or staged repair of interrupted aortic arch (fAA) with ventricular septal defect (VSD). Seven patients had fAA type A and 17 patients had type B. Eleven of the patients. median age 5 days. underwent staged operations and 13 infants. median age 6 days. underwent primary repair. Palliation was by tube graft interposition (six). subclavian-aortic anastomosis (three). left carotid-aortic anastomosis (one). or end-to-side aortic anastomosis (one) combined with pulmonary artery banding (eight) or early VSD closure. With palliation. there were three (27%) early deaths among the eleven patients and one (13%) late death among the eight remaining. Delayed repair at 5 days to 14 months (median 7 months) in seven patients incurred three (43%) early and no late deaths. Primary repair in 13 patients consisted of VSD closure combined with graft interposition (12) or end-to-side aortic anastomosis (one), with three (23 %) early and no late deaths. Nine of 14 survivors had hemodynamic evaluation by catheterization 1 to 3 years following repair. None had a significant residual VSD or pressure gradients between the ascending and thoracic aorta. Six had subaortic stenosis. two mild (gradient < 20 mm Hg) and four severe (gradient> 50 mm Hg). necessitating operation. Results of operations in neonates with fAA continue to improve. Essential in management is an awareness that subaortic stenosis and hypocalcemia may be accompaniments of this anomaly. Based on these data, we prefer primary repair for fAA with VSD.
William I. Norwood, M.D., Peter Lang, M.D. (by invitation), Aldo R. Castaneda, M.D., and Thomas J. Hougen, M.D. (by invitation), Boston, Mass.
Interrupted aortic arch (lAA) is an uncommon congenital cardiovascular malformation which was initially described by Steidele' in 1778. In earlier communications from this institution, this anomaly was reported to have been identified in 1.4% of autopsied infants with congenital heart disease and 1.3% of infants in the New England Regional Infant Cardiac Program.' If IAA is untreated, the median age at death is 4 to 10 days, usually following physiological closure of the ductus arteriosus.' Although IAA can occur in conjunction with many complex congenital cardiac malformations (or rarely in isolation), most often the constellation of IAA, ventricular septal defect (VSD), and patent ductus arteriosus coexist. In the late 1950s, several case reports of reparative operations for IAA and VSD appeared.v' Subsequently, a variety of approaches to surgical palliation or primary repair have been advanced.v'? This From the Departments of Cardiovascular Surgery and Cardiology, Children's Hospital Medical Center, and the Departments of Surgery and Pediatrics, Harvard Medical School, Boston, Mass. Read at the Sixty-third Annual Meeting of The American Association for Thoracic Surgery, Atlanta, Ga., April 25-27, 1983. Address for reprints: Aldo R. Castaneda, M.D., Children's Hospital Medical Center, 300 Longwood Ave., Boston, Mass. 02115.
832
communication reviews a 7 year experience with both primary and staged repair of IAA and VSD with emphasis on late postoperative outcome and hemodynamic evaluation.
Methods and patients From January, 1976, through January, 1983, 24 children have undergone palliative and/or reparative operations for IAA and VSD at the Children's Hospital, Boston, Massachusetts. Patients with severe coarctation or hypoplasia of an aortic arch segment but not interruption are excluded from this review. Also excluded are 10 patients with IAA in whom associated intracardiac malformations other than VSD (single ventricle in three, aortic atresia in one, truncus arteriosus in five, and double-outlet right ventricle with subpulmonary VSD in one) dominated the course of surgical management and outcome. The median age at operation was 5 days; 22 patients were less than 1 month of age, one infant was 10 months old, and one child was 4 years old. All 22 of the neonates had cardiomegaly and increased pulmonary blood flow on chest roentgenogram; at the time of presentation, 16 (73%) had some degree of metabolic acidosis attributed to insufficient systemic perfusion through a restrictive
Volume 86 Number 6 December, 1983
ductus arteriosus. Since 1977, prostaglandin E 1 infusion has been used for preoperative resuscitation of neonates with a closing ductus arteriosus and metabolic acidosis. Seven patients had type A interrupted aortic arch and the remainder (17) had type B." The VSD was muscular in two, malalignment type in eight, and infundibuloventricular in the remainder. The correct diagnosis was established by preoperative cardiac catheterization in 21 patients. IAA was initially diagnosed in two infants undergoing operations for suspected coarctation of the aorta and in one premature neonate (1,200 gm, 27 weeks' gestation) at the time of ligation of the ductus arteriosus. Eleven patients were managed with staged operations and 13 underwent primary reconstruction of all coexisting defects. Palliative operation. Among the patients who underwent initial palliative procedures, six had IAA type A and five had type B. The median age at operation was 5 days; with the exception of the 4-year-old patient, the age range was 2 to 17 days. The ductus arteriosus was ligated in all at the initial operation. The interrupted aortic segment was managed by interposition of a tube graft (8 to 14 mm in diameter) in six patients. An end-to-side anastomosis of the left subclavian artery to the thoracic aorta was used in three infants with IAA type A. In one neonate with IAA type B, an end-to-side anastomosis of the left carotid artery to the aortic isthmus was made. An end-to-end anastomosis was established in one patient with IAA type A. A pulmonary artery band was placed in eight patients to control pulmonary blood flow. In three neonates, no pulmonary artery band was placed and the initial procedure was followed by closure of the VSD 5 days to 6 weeks later. One patient with severe posterior deviation of the infundibular septum causing significant subaortic stenosis had placement of a valved conduit between the apex of the left ventricle and the thoracic aorta at age 5 days combinedwith graft interposition between the ascending and descending thoracic aorta and placement of a pulmonary artery band. Seven survivors of initial palliation have undergone pulmonary artery band removal and/or VSD closure 5 days to 16 months (median 7 months) after initial operation. Primary reconstructive operation. Twelve patients undergoing primary reconstructive procedures had IAA type B and one neonate had type A. These patients ranged in age from 1 to 21 days with the exception of the lo-month-old infant. The interruption was managed by direct anastomosis of the aortic segments in one neonate.One patient had an anomalous right subclavian
fAA with VSD
833
Table I. Mortality from staged and primary repair No. of patients Staged repair Stage I Stage II Primary repair Totals
Deaths Early
Late
II
II 7 13
3 3 3
24
9 (38%)
I (4%)
artery anastomosed end to side to the aortic arch but could not be weaned from cardiopulmonary bypass and a tube graft was subsequently interposed. The remainder had graft interposition between the ascending aorta and the descending thoracic aorta. The distal anastomosis was carried out through a left thoracotomy. Subsequently, the proximal anastomosis was made through a median sternotomy. The ductus arteriosus was then ligated and cardiopulmonary bypass was established for repair of the VSD. The VSD was closed via a right atriotomy with the use of deep hypothermia and circulatory arrest in all but one patient. Perioperative hypocalcemia occurred in 13 patients (55%) and of these, three patients proved to have DiGeorge's syndrome (III-IV pharyngeal pouch deficiency) with an absent thymus gland, no circulating T cells, and a persistent requirement for supplemental calcium. Five patients (20%) had or developed severe subaortic stenosis. In four patients with long-segment obstruction caused by a posteriorly deviated infundibular septum, a conduit was interposed between the apex of the left ventricle and the thoracic aorta at 5 days, 3 weeks, 9 months, and 13 months with no mortality. The other patient had resection of discrete fibrous subaortic stenosis at the age of 3 years. Results Among the 24 patients in this series, 14 (59%) children are alive and clinically well 3 months to 6 years following either primary or staged reconstructive operations (Table I). Among the 11 patients with initial palliative operations, there were three hospital deaths; two infants died at the time of operation and one patient died 4 weeks after IAA repair and pulmonary artery banding. This infant had a coexisting annular pancreas which contributed to the outcome. There was one late death 6 months after operation, caused by an arrhythmia on induction of general anesthesia for cleft lip repair. Of the seven patients who subsequently underwent pulmonary artery band removal and VSD closure, three patients died-one as the result of disruption of the aortic graft on opening the sternum, one of a
The Journal of Thoracic and Cardiovascular
Norwood et al.
834
Surgery
Table II. Hemodynamic data following palliative surgery operation Case No.
Catheterization interval (mo)
1 12 3 16 5
I
2 3 4 5
VSD (QpIQs)
Subaortic PSEG (mm Hg)
Aortic arch PSEG (mm Hg)
2.0
0 0 0 16 0
50 50 40 0 0
NC
3.5 0.9 NC
Comments Subclavian-aorta anastomosis Carotid-aorta anastomosis Subclavian-aorta anastomosis
Legend: VSD. Ventricular septal defect. Qp/Qs, Pulmonary-to-systemic flow. PSEG. Peak systolic ejection gradient. NC, Not calculated.
Table m. Hemodynamic data following reparative operation Case No.
Catheterization interval (mo)
VSD (QpIQs)
6 7 8 9 10 11 12 13 14
14 14 12 27 8 1 36 4 12
l.l l.l
1.0 1.0 1.0 l.l
1.6 1.0 1.0
Subaortic PSEG (mmHg)
0 18 145 72
62 50 14 0
Aortic arch PSEG (mmHg)
0 0 20 16 0 0 0 0 0
For legend see Table II.
technical error in VSD closure, and the third because of late recognition of ventilatory insufficiency. There have been no 'late deaths among the four survivors of staged repair. Among the 13 infants who underwent primary reconstructive procedures, there were three operative deaths and no late deaths. Twelve patients have undergone postoperative cardiac catheterization. In five patients, the study followed palliation and was preliminary to VSD closure (Table 11). Two of these five children had no pressure gradient between the ascending and thoracic aorta 10 and 16 months following tube graft interposition. One child, treated initially by anastomosis of the left carotid artery to the aorta at age 4 days, had a 50 mm Hg peak systolic ejection gradient between the ascending and thoracic aorta. The other two children had a 40 mm Hg systolic gradient across a subclavian-aortic anastomosis. These obstructions were repaired in conjunction with VSD closure and removal of a pulmonary artery band. Nine children were studied 1 to 3 years following primary or staged repair (Table III). One had a residual VSD which was small (pulmonary-to-systemic flow < 1.5). None had a significant systolic pressure gradient between the ascending and thoracic aorta. Six children had evidence of subaortic stenosis. In two
patients, the pressure gradient was less than 20 mm Hg. Four children had severe stenosis (peak systolic ejection gradient> 50 mm- Hg) treated by left ventricular apical-aortic conduit interposition in three infants or resection of a subaortic membrane in one child, with no mortality (Fig. 1).
Discussion The natural history of lAA and VSD dictates early surgical intervention.' Physiological closure of the ductus arteriosus results in diminished systemic perfusion, acidosis, and death. Thus, to alter the outcome of this condition, surgical therapy is most often necessary in the neonatal period. Twenty-two patients in this series were less than 1 month of age at operation. Occasionally, the ductus arteriosus remains patent. In this circumstance, surgical management in infancy is usually necessary because of intractable congestive heart failure, severe failure to thrive, and pulmonary artery hypertension. Rarely, as occurred in one 4-year-old child in this series, the ductus arteriosus remains patent and congestive heart failure can be controlled by anticongestive therapy. In such cases, however, irreversible obstructive pulmonary vascular disease is likely to develop. The goals of surgical therapy for lAA with VSD are to establish continuity between the segments of aorta and to separate the systemic and pulmonary circulations. These goals can be achieved concomitantly or in stages." In this series, primary repair yielded a lower total mortality than the various forms of staged reconstruction. Although staged repair remains a potentially useful technique, our current preference is primary repair. This approach has been facilitated by prostaglandin EI> which allows preoperative resuscitation of the moribund neonate with metabolic acidosis, permitting complete evaluation at cardiac catheterization and contributing to hemodynamic stability during the early phases of operation. Several ingenious techniques have been advanced to establish continuity between the separated ascending and descending thoracic aorta, many of which utilize
Volume 86
lAA with VSD
Number 6 December. 1983
principally autologous vascular tissue to repair the aortic arch.":" Although others advocate the use of the ductus arteriosus in aortic reconstruction, we have previously observed that ductal tissue used to bridge the gap between the ascending and descending aorta in neonates does not provide a reliable conduit since it can constrict and result in significant late stenosis. Utilization of the subclavian artery or carotid artery to bridge the gap, in this experience, resulted in significant restriction of flow early after operation. However, the restrictive region was discrete and easily enlarged with a patch. A direct anastomosis between the ascending and descending thoracic aorta is certainly feasible in some patients and was applied in two patients in this series. The hypothetical advantage of this approach is the potential of providing long-term unobstructed continuity between the ascending and descending aorta. Of note, interposition of a relatively large tube graft (about 10 mID in the neonate) proved a reliable means of establishing unobstructed continuity between the ascending and descending aorta in these patients. Such a reconstruction is intuitively expected to require revision as the child grows. On the other hand, none of the patients who underwent cardiac catheterization 1 to 3 years following graft interposition had a significant pressure gradient between the ascending and descending thoracic aorta . Moreover, neither of the two oldest children in this series, now 5 and 6 years after operation, have upper extremity hypertension. If the tube graft is analogous to coarctation of the aorta, one might expect grafts of the size used in this series not to produce significant hypertension for many years. Seven of the 24 patients in this series (29%) had or developed some degree of subaortic stenosis. Five had surgical management of the stenosis. One patient had a discrete subvalvular fibrous ring, but the remainder had significant long-segment muscular obstruction which did not lend itself to direct reparative techniques. Placement of a valved conduit between the apex of the left ventricle and thoracic aorta provided a .means of decompressing the severely hypertensive left ventricle in these small patients. 16 As has been pointed out by other observers, perioperative hypocalcemia occurs frequently in patients with lAA. 17• 18 In this series, 13 patients had significant hypocalcemia «7 mg/dl) necessitating prolonged infusions of calcium ion. Only three patients proved to have DiGeorge's syndrome by plasma calcium and circulating T cell determinations. The subjective impression was that all patients in this report had distinctly less thymic tissue than is ordinarily seen in neonates and infants.
835
Fig. 1. Frame of cineangiogram in the anteroposterior projection with injection into the left ventricle. An arterial catheter has been placed through the valved conduit interposed between the apex of the left ventricle and thoracic aorta. The stud y visualizes the subaortic stenosis and the conduit.
Although the mortality in this series is appreciable, results of operations in neonates and infants 'with IAA continue to improve. Based on these data, primary repair of IAA with a VSD is our current preference. An awareness that subaortic stenosis and hypocalcemia may be accompaniments of this anomaly is essential in the management of these patients. While direct aorto-aortic anastomosis is appealing and its role in the management of IAA should be further clarified, graft interposition appears to provide unobstructed aortic continuity for years. REFERENCES
2
3
4
5
6
Steidele RJ : Sammig. Verschiedener in der chirug. 'prakt. Lehrschule Gemachten Beobb 2: 114, 1777-1778 Fyler DC, Parisi LF, Berman MA: Regiona1ization of infant cardiac care in New England. Cardiovasc Clin 4:339-340, 1972 Collins-Nakai RL, Dick M, Parisi-Buckley L, Fyler DC, Castaneda AR: Interrupted aortic arch in infancy. J Pediatr 88:959-962, 1976 Merrill DL, Webster CA, Samson PC : Congen ital absence of the aortic isthmu s. J TTIIORAC SURG 33:311320, 1957 Van Praagh R, Bernhard WF, Rosenthal A, Parisi LF, Fyler DC: Interrupted aortic arch. Surgical treatment. Am J Card ioI27:200-211 , 1971 Quie PG, Novick R, Adams P, Anderson RC, Varco RL: Congenital interru ption of the aortic arch. J Pediatr 54:87 -92, 1959
836
The Journal of Thoracic and Cardiovascular Surgery
Norwood et al.
7 Blake HA, Manion WC, Spencer FC: Atresia or absence of the aortic isthmus. 1 TTHORAC CARDIOVASC SURG 43:607-614, 1962 8 Tyson KRT, Harris LC, Ngiem QX: Repair of aortic arch interruption in the neonate. Surgery 67:1006-1010, 1970. 9 Litwin SB, Van Praagh R, Bernhard WF: A palliative operation for certain infants with aortic arch interruption. Ann Thorac Surg 14:369-375, 1972 10 Jones EL, Plauth WH, Hatcher CR Jr: A palliative operation for all types of aortic arch interruption in the neonate. 1 TTHORAC CARDIOVASC SURG 69:579-584, 1975 II Celoria GC, Patton RB: Congenital absence of the aortic arch. Am Heart 1 58:407-413, 1959 12 Barratt-Boyes BG, Nicholls IT, Brandt PWT, Neutze 1M: Aortic arch interruption associated with patent ductus arteriosus, ventricular septal defect, and total anomalous pulmonary venous connection. J TTHORAC CARDIOVASC SURG 63:367-373, 1972 13 Trusler GA, lzukawa T: Interrupted aortic arch and ventricular septal defect. Direct repair through a median sternotomy incision in a 13-day-old infant. J TTHORAC CARDIOVASC SURG 69: 126-131, 1975 14 Monro Jl., Brown W, Conway N: Correction of Type B interrupted aortic arch with ventricular septal defect in infancy. J TTHORAC CARDIOVASC SURG 74:618-623, 1977 15 Moulton AL, Bowman FO lr: Primary definitive repair of type B interrupted aortic arch, ventricular septal defect, and patent ductus arteriosus. Early and late results. 1 TTHORAC CARDIOVASCSURG 82:501-510,1981 16 Norwood WI, Lang P, Castaneda AR, Murphy 10: Management of infants with left ventricular outflow obstruction by conduit interposition between the ventricular apex and thoracic aorta. J THORAC CARDIOVASC SURG 86:771-776,1983 17 Harvey lC, Dungan WT, Elders Ml, Hughes ER: Third and fourth pharyngeal pouch syndrome, associated with vascular anomalies and hypocalcemia seizures. Clin Pediatr 9:496-499, 1970 18 Conley ME, Beckwith lB, Mancer lFK, Tenckhoff L: The spectrum of the DiGeorge syndrome. J Pediatr 94:883-890, 1979
Discussion DR. RONALD C. ELKINS Oklahoma City. Okla.
We have recently reviewed and reported our experience with this entity, comprising 17 patients. Of these 17 patients, 14 have had palliative procedures; of the 14, there were 11 who survived the initial palliation, this being a reconstruction of the aorta with a 6 mm polytetrafluoroethylene (Gore-Tex or Impra) graft and banding of the pulmonary artery in those children with increased pulmonary blood flow. Of the children undergoing palliation, six had primary
VSDs associated with the lAA. Those six patients have been long-term survivors. Five of the patients have had their VSDs corrected. In one of the children, on repeat catheterization, the VSD was of such size that it did not warrant closure. All but one of these patients have had postoperative catheterization, and we have identified one patient who required reoperation due to the size of the child's reconstructed aorta. A significant gradient between the ascending aorta and descending thoracic aorta was present, and the child had placement of an additional 12 mm Dacron graft. This experience had led us to continue to recommend palliative treatment for patients with lAA, particularly those with VSD. We respect the ability of those who do this as a single-stage procedure. However, this is a critically ill group of patients, and if they can be prepared by a procedure limited to a left thoracotomy and then have a VSD repair at a later date with a reasonable success rate, this certainly should be recommended. I would ask two questions of Dr. Norwood. What size of graft material and what type of graft material do you utilize for reconstruction of the aorta? My second question pertains to my suspicion that there is a time frame between your patients who were treated by palliation and those treated by primary repair ofthe VSD as well as correction of the lAA. If you were doing the staged procedure now, it is not likely that your repair of a VSD as a second stage operation would carry a lower mortality than your reported one of >50%? DR. LEONARD L. BAILEY Loma Linda. Calif.
Dr. Norwood has presented an impressive experience in the management of what we have come to call simple lAA complex, that is, those babies who have just a VSD and possibly an atrial septal defect associated with lAA type A or B. He has clarified two issues for me. First, complete repair at the time of initial diagnosis is less risky than a stage procedure, Dr. Elkins' experience notwithstanding. Second, most of these babies should survive the corrective operation. The impressive mortality of 23% in the newborn period in his series is exceptional. I take issue, however, with Dr. Norwood's experience regarding the use of prosthetic tubes in tiny babies to establish aortic continuity. I feel strongly that growing patients should have growth potential of their repairs. Establishing direct aortic continuity is just as easy, if not easier, than inserting a prosthetic tube in these babies, particularly if one uses a bilateral thoracotomy. The experience of Drs. Bowman, Trusler, Moulton, and our own experience would establish ease and efficacy of direct aortic anastomosis. Anastomotic growth has been well documented. Since late 1976 we have repaired lAA complex in II newborn infants. Seven of these had the simple variety of lAA with a VSD. A few of them had atrial septal defects. Five survived total repair, for a mortality of 29%. We have not found subaortic stenosis in the four who have been restudied. I agree with Dr. Norwood that definitive repair is the
Volume 86 Number 6 December, 1983
procedure of choice. I only question the use of prosthetic tubes for the repair. DR. NOR WOOD (Closing) I would like to thank the discussers for raising some interesting aspects of management of these critically ill infants. We have used both Teflon and Dacron grafts. The size almost always was 10 mrn, although a few patients had 8 or 12 mm grafts placed. It is true that our palliative operations tended to be performed earlier in this 7 year period of review. Although it is our current preference to proceed with primary repair, I must say that our experience is not sufficient to clearly choose between palliation and primary repair. I think there is still considerable room for individualization. However, I would add
lAA with VSD
837
that the use of prostaglandin E 1 to establish good metabolic balance prior to the operation has contributed as much as anything to improving the results. Dr. Bailey raised the issue of the aorta-aortic anastomosis. Obviously, it is very appealing to avoid a tube graft that the child will certainly outgrow. However, our experience with aorta-aortic anastomosis is quite limited. One of the patients had compression of the left main-stem bronchus by the aorta following direct anastomosis. This problem was somewhat troublesome in the early postoperative period. There is no question in my mind that these patients will be returning with limited flow through the tube graft. On the other hand, if one thinks in terms of an analogous experience with coarctation of the aorta, upper extremity hypertension may not occur until quite late with a 10 mrn lumen.