- Email: [email protected]
Vascular anomalies causing tracheoesophageal compression
J THoRAc CARDIOVASC SURG 1989;97:725-31
Vascular anomalies causing tracheoesophageal . compression Review of experience in children Two hundred four infants and children (mean age 13 months) have undergone operation for the relief of tracheoesophageal obstruction resulting from vascular anomalies. One hundred thirteen patients had complete vascular rings (group I), 61 with double aortic arch and 52 with right aortic arch with a left ligamentum. Nine patients had a pulmonaryartery sling(group II), 71 had innominateartery compression (group Ill), and 11 had miscellaneous anomalies (group IV). Patients were admitted with respiratory distress, stridor, apnea, dysphagia, or recurrent respiratory infections. Diagnosis was established by barium esophagogram in group I; barium esophagogram, bronchoscopy, and computed tomography or angiography in group II; bronchoscopy in group Ilk and barium esophagogramor angiography in group IV. The operative approach was through a left thoracotomy in groupsI, II, and IV (93% of these patients) and through a right thoracotomy for group m (96% of these patients).The operativemortality rate was 4.9 % and there were seven late deaths (3.4%). There have been no operative deaths in patients with isolated vascular anomalies in the past 28 years. Follow-up data from 1 month to 20 years (mean 8.5 months) were available on 159 patients; 141 (92%) were essentiallyfree of symptoms, and 12 (8%) had residual respiratory problems. Five of six patients in group II having a lung scan postoperatively had a patent left pulmonary artery. A strong index of suspicion is necessary to avoid the complications of vascular rings in children. Barium swallow is the best single diagnostic technique for patients with complete vascular rings. A bronchoscopic study is required to diagnose innominate artery compression. Angiograms or computedtomographic scans are used to confirmthe diagnosisof pulmonaryartery sling. Left thoracotomy provides excellent exposure for all vascular rings except the displaced innominate artery, for which a right thoracotomy is the best approach.
Carl L. Backer, MD, Michel N. Ilbawi, MD, Farouk S. Idriss, MD, and Serafin Y. DeLeon, MD, Chicago, Ill.
h e first successful operation for a vascular ring was reported by Dr. Robert GrOSSI in 1945 when he divided a double aortic arch in a l-year-old boy. He also introduced the term vascular ring to describe mediastinal vascular anomalies causing tracheoesophageal compression. The first successful operation for a vascular From the Division of Cardiovascular-Thoracic Surgery, Children's Memorial Hospital, and the Department of Surgery, Northwestern University Medical School, Chicago, Ill. Supported in part by the A. C. Buehler Foundation, Park Ridge, Ill. Read at the Fourteenth Annual Meeting of The Western Thoracic Surgical Association in Hawaii, June 22-25, 1988. Address for reprints: Carl L. Backer, MD, Divisionof CardiovascularThoracic Surgery, Children's Memorial Hospital, 2300 Children's Plaza, Chicago, IL 60614.
Table I. Clinical groups Group
II III IV Total
Anomaly Complete vascular rings Double aortic arch Right aortic arch Pulmonary artery sling Innominate artery compression Miscellaneous
No. of patients
113 61 52
9 71
_1_1 204
ring at our institution was performed by Dr. Willis J. Potts in 1947.2 He divided a double aortic arch in an 8-year-old boy. In 1972 Nikaidoh, Riker, and Idriss' reported our early experience with 68 children who underwent surgical relief of vascular rings. This current 725
726
The Journal of Thoracic and Cardiovascular Surgery
Backer et a/.
Table II. Age at onset of symptoms and at operation DAA
Symptoms <1 mo 1-6 mo >6 mo Operation <1 mo 1-6 mo 6-12 mo >12 mo
PA sling
RAA
Innominate compression
No.
%
No.
%
No.
%
No.
%
41 5 2
85 10 4
23 12 5
57 30 12
5 1 0
83 16
19 2 2
82 9 9
4 25 15 17
6 41 24 28
0 16 13 23
31 25 44
1 5 3 0
11 55 33
0 34 18 19
48 25 27
DAA, Double aortic arch; PA, pulmonary artery; RAA, right aortic arch with left ligamentum.
Table m. Associated lesions Patients No.
%
Ventricular septal defect Pulmonary atresia Tetralogy of Fallot Ventricular septal defect Ventricular septal defect Complete tracheal rings
1 I 1 4 1 3
2 2 2 7.6 II 33
Ventricular septal defect Tracheoesophageal fistula
2 4
2.8 5.7
Vascular anomaly Double aortic arch (n = 61) Right aortic arch (n = 52) Anomalous left pulmonary artery (n = 9) Displaced innominate artery (n = 71)
Associated anomaly
report reviews our total experience with 204 children who were operated on at Children's Memorial Hospital in Chicago.
Patients and methods Two hundred four patients have undergone surgical treatment for the relief of tracheoesophageal compression by vascular anomalies from 1947 through 1987. Their ages at the time of the operation ranged from 1 week to 17 years (mean 13 months). The majority (68%) were operated on within the first year of life. The sex distribution was unremarkable except in the pulmonary artery sling group, in which there was an 8 to 1 preponderance of boys. Their clinical records were retrospectively reviewed to obtain specificdata. In particular, sex, age at onset of symptoms, age at operation, year of operation, type of vascular ring, symptoms, physical examination, diagnostic tests, type of operation, site of arch division, postoperative complications, long-term results, length of long-term followup, and associated anomalies were recorded for each patient. Four groups of patients were identified (Table I): Group I had complete vascular rings, either a double aortic arch or a right aortic arch with a left ligamentum arising from the descending aorta; group II had an anomalous left pulmonary artery arising from the right (pulmonary artery sling); group
III had a posteriorly displaced innominate artery; and group IV had miscellaneous anomalies of the aortic arch or its branches not forming a complete ring. Clinical manifestations. The most common symptom was stridor or noisy respiration. Respiratory distress was precipitated frequently by crying or feeding or by swallowing a bolus of solid food that pressed on the soft posterior trachea. There were recurrent respiratory tract infections. A brassy cough, similar to a "seal bark," was described in particular with complete rings. Dysphagia for solid foods was noted usually in older children. Two patients with double aortic arch had a penny lodge in the esophagus preoperatively. Choking, severe hypoxia, and even convulsive seizures were noted on occasion. The ages at onset of symptoms and at operation are listed in Table II. In general, among the patients with a complete ring, those with a double aortic arch had earlier onset of symptoms and were operated on within the first year of life (72%). On physical examination, significant findings were present in over 85% of patients. These included rhonchi, stridor, retractions, tachypnea, wheezing, and rales. Hyperextended neck posture that relieved tracheal obstruction by stretching the tracheal wall and splinting the trachea open was seen in some infants. In general, children with vascular rings were well nourished and normal in size for their age, despite the esophageal compression, because they could take liquid feeding well. Associated lesions were uncommon (Table III). Diagnostic procedures. A chest roentgenogram was obtained for all patients and showed abnormalities in over 80% of these children. Hyperaeration of the lung fields was seen and was sometimes unilateral, especially in patients with a pulmonary artery sling. Varying degrees of atelectasis and pneumonic infiltrates were present. The location of the aortic arch in relation to the trachea was helpful in determining the type of vascular ring. When the location was not clear, a double aortic arch was suspected. In patients with significant symptoms and a right aortic arch on the chest x-ray films, a left ligamentum completing the vascular ring was suspected (Fig. I). Lateral chest and neck roentgenograms sometimes showed anterior tracheal compression from a posteriorly displaced innominate artery. The most reliable procedure for the diagnosis of vascular rings (group I) was the barium esophagogram. Different projections and views were necessary to establish the shape,
Volume 97 Number 5 May 1989
Tracheoesophageal compression 7 2 7
Fig. 1. Chest x-ray film of a 5-year-old boy with recurrent upper respiratory tract infections. Arrow points to the arch of the aorta , which is to the right of the tracheal air column.
location, and persistence of the filling defect. In children with a double aort ic arch, barium esophagogram showed two indentations of unequal size on the anteroposterior projection, with the right arch indentation usually superior and larger (Fig. 2). A right aortic arch with a left ligamentum was visualized as a large right impression and a small oblique left impression from the ligamentum. An anterior indentation was suggestive of a pulmonary artery sling. A bronchoscopic study was performed in one half of the patients with a double aortic arch and one quarter of the patients with a right aortic arch . In patients with pulmonary artery sling, bronchoscopy demonstrated complete tracheal rings (n = 3) and other abnormalities (eparterial right bronchus [n = 2]). All patients with innominate artery compression had a bronchoscopic study, which revealed anterior compression of the tracheal lumen. Angiograms were performed on a few patients with complete vascular rings (group I), in six of nine patients with pulmonary artery sling, and in a few patients with innominate artery compression. Recently, computed tomographic (CT) scans have been obtained. They were diagnostic in patients with vascular rings or slings (Fig. 3) and demonstrated, when present, complete tracheal rings. Magnetic resonance imaging was used in an older patients with a right arch, a left ligamentum from the descending aorta, and a Kommerel diverticulum at the origin of the left subclavian artery (Fig. 4). In patients with suspected innominate artery compression, neurologic evaluation, sleep studies, and radionuclide investigation for gastroesophageal reflux were performed to rule out other possible causes of apneic episodes. Operative findings and management Double aortic arch. The approach to this lesion in all instances but two was a left thoracotomy (59 patients) . In two patients the operation was done through a median sternotomy . The right-sided (posterior) arch was dominant in 45 patients (73%), the left-sided (anterior) arch was dominant in 12
Fig. 2. A 4-month-old boy with stridor. Barium esophagogram shows bilateral indentations (arrows) characteristic of a double aortic arch.
(20%), and the arches were equal In four (7%). With a dominant right arch, segments of the left arch were atretic in 40% of the patients. With a dominant left arch , segments of the right arch were atretic in four of 12 (33%) patients . The atretic area occurred most frequently in the posterior or distal end of the lesser arch. The vascular ring eaused by the double aortic arch was released by dividing the lesser of the two arches, usually at the atretic area . When the arches were patent, the lesser was divided between clamps at a site selected to preserve the brachiocephalic blood flow. Both the carotid and radial pulses were monitored by the anesthesiologist after the vascular clamps were applied. The left ligamentum was always divided and careful dissection around the esophagus and the trachea was performed to lyse any residual adhesive bands . The recurrent laryngeal and phrenic nerves were identified and protected. Right aortic arch and left ligamentum. Left thoracotomy was used in 48 patients and a median sternotomy in four. The vascular ring was released by dividing the ductus or ligamentum and dissecting the trachea and esophagus free from adhesive bands. There were two anatomic variations. In 65% a retroesopha-
7 2 8 Backer et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 3. A 7-month-old boy with respiratory distress. CT with contrast shows a pulmonary artery sling. Large arrow points to the left pulmonary artery encircling the trachea (small arrow).
Fig. 4. A 17-year-old patient with dysphagia. Magnetic resonance imaging shows a right aortic arch (small arrow) and a Kommerel diverticulum (large arrow) at the origin of the left subclavian artery.
geal left subclavian artery was present. In 35% there was a mirror-image left innominate artery. In both situations the vascular ring was completed by the ligamentum arteriosum, which stretched from the descending aorta to the pulmonary artery. This distribution is similar to that reported by Felson and Palayew' (56% retroesophageal, 44% mirror image). Theoretically, if there is mirror-image branching with the ligamentum arising from the innominate artery instead of the aorta, a complete ring will not be formed and there will be no symptoms.'
The aorta descended on the right in two thirds of these patients and coursed retroesophageally to descend on the left in the rest. Pulmonary artery sling. The first successful repair of a pulmonary artery sling was reported by Potts, Holinger, and Rosenblum" from our institution in 1954. Since then we have operated on nine patients with an anomalous left pulmonary artery originating from the right. The left pulmonary artery formed a sling as it passed around the origin of the right bronchus between the trachea and esophagus at or above the level of the carina as it coursed toward the left lung. This caused compression of the right bronchus and trachea. In six patients a left thoracotomy was used, in two a median sternotomy, and in one a right thoracotomy. The left pulmonary artery was divided and passed anterior to the trachea and reanastomosed to the right pulmonary artery. In three patients a left ligamentum was also divided. In one patient with complete tracheal rings, two tracheoplasties were performed at later dates; in another, pericardial tracheoplasty with the use of extracorporeal circulation was performed simultaneously. A third patient with complete rings and anomalous tracheal bifurcation has not yet required tracheoplasty. Innominate artery compression. Since the first report by Riker" from our institution in 1954, 71 patients have been operated on for innominate artery compression, the majority between 1977 and 1980. All but three arteriopexies of the innominate artery have been performed through a right anterolateral thoracotomy. The innominate artery was lifted off the trachea by suspending it to the anterior chest wall, usually to the periosteum of the sternum. This was facilitated by leaving a small portion of pericardial reflection attached to the base of the innominate artery for placement of sutures for the suspension. A bronchoscopic examination was sometimes performed during the operation to demonstrate relief of obstruction. Miscellaneous. There were 11 patients in this group. Seven with a left aortic arch and an aberrant right subclavian artery originating from the descending aorta were operated on before
Volume 97 Number 5 May 1989
Tracheoesophageal compression
Table IV. Postoperative mortality in vascular rings-early deaths Year
1946 1949 1951 1952 1953 1954 1957 1959 1973 1974
Age at operation
5 mo 2 mo 5 mo 2 mo 3 mo 2 mo 3 mo 6 yr I mo
I mo
Anatomy
Cause of death
DAA DAA DAA DAA AbRSc RAA RAA RAA RAA
Failure to divide ring Air block syndrome Respiratory failure Hemorrhage intraoperatively Pneumonia, YSD, ASD, PH Respiratory failure, YSD, PH Respiratory failure, YSD Hemorrhage intraoperatively Cardiac arrest in operating room, LTGA, YSD, PS, Waterston shunt Pneumonia, YSD, ASD, PAPYD
INN
AbRSc, Left aortic arch with aberrant right subclavian artery; ASD, atrial septal defect; DAA, double aortic arch; INN, displaced innominate artery; LTGA, L-transposition of the great arteries; PAPVD, partial anomalous pulmonary venous drainage; PH, pulmonary hypertension; PS, pulmonary stenosis; RAA, right aorticarch with left ligamentum; VSD,ventricular septal defect.
1973. Two patients had a left aortic arch with a left ligamentum arteriosum that was divided through a left thoracotomy. There was minimal compression of the trachea and esophagus. Another child had an anomalous right carotid artery compressing the trachea; the artery was suspended to the chest wall through a right thoracotomy approach. A fourth child had aortopexy through a right thoracotomy early in our experience because of tracheal compression by a dilated aortic arch.
Results There were 10 postoperative (up to 30 days) deaths (4.9%) and seven late deaths (3.4%). Most of the deaths occurred in infants under 6 months of age (II / 17) and early in this series (8/10 early deaths before 1960). Over half (9/17) of the patients who died had other complicating cardiac lesions (Tables IV and V). Four early deaths and two late deaths (6.5% and 3.2%, respectively) occurred in the double aortic arch group. Four early deaths (7.6%) and one late death (1.9%) occurred in the right aortic arch group. There were no early deaths but two late deaths (22%) after repair of pulmonary artery sling. Both occurred quite late, at 7 months and 2Y2 years, in patients who required tracheoplasty because of extensive narrowing of the trachea from complete tracheal rings. There was one early (1.5%) and two late deaths (3%) after innominate arteriopexy. Some patients in group I required postoperative treatment with a bronchoscope and four needed a postoperative tracheostomy. In the double aortic arch group, one patient required reoperation for persistent
729
Table V. Postoperative mortality in vascular rings-late deaths Age at Year operation Anatomy
1950 1966 1967 1978
3 wk 6 mo 6 mo 2 rno
DAA RAA DAA INN
1979
6 mo
PAS
1985
7 mo
PAS
1986
6 mo
INN
Cause of death
Postop. survival
Pneumonia, malnutrition 37 days Pneumonia, anoxic eNS 50 days Aspiration, asphyxia 125 days Respiratory failure, aortic 50 days stenosis, coarctation Respiratory failure, 900 days tracheoplasty x2 Respiratory failure, 210 days tracheoplasty Respiratory failure, single 180 days ventricle, pulmonary atresia
CNS, Central nervous system; PAS, pulmonary artery sling. For other abbreviations see Table IV.
stridor, and aortopexy was performed through a right thoracotomy 6 months postoperatively. In the group with right aortic arch, one patient required reoperation at 5 months and had aortopexy through a right thoracotomy. Aortopexy is necessary during the original operation if the two ends of the divided vascular ring do not spring apart naturally to relieve pressure on the trachea and esophagus. The mean hospital stay for this group was 6.5 days. In group II, six of nine patients underwent postoperative bronchoscopic treatment and four of nine required postoperative tracheostomy. One required reoperation after 2 months for postoperative cyanotic spells and had the aorta suspended to the sternum. The mean hospital stay was 22 days. These infants all had a difficult postoperative course. Long-term follow-up with lung scan has been obtained in six patients. Flow was present to the left lung in five of six patients, ranging from 25% to 63% (mean 41%) of the total pulmonary blood flow. One patient had postoperative catheterization and this showed slight stenosis of the anastomosed left pulmonary artery. This patency rate (83% in those studied) is different from the greater than 50% anastomotic occlusion rate reported by Sade and colleagues." In three patients with preoperative tracheostomy, the cannula was successfully removed after innominate arteriopexy. One patient in this group later required a tracheal stent with a rib graft for tracheoma1acia. Two patients were reoperated on, one at 4 months and the other at 5 years for recurrent compression of the trachea. Their mean hospital stay was 7.5 days. The majority of patients had relief of their symptoms. Of 153 patients with a postoperative follow-up that
730
The Journal of Thoracic and Cardiovascular Surgery
Backer et al.
Table VI. Results of surgical treatment Poor
Good
Anatomy Double aortic arch Right aortic arch Pulmonary artery sling Displaced innominate artery Total
No.
%
No.
%
44
90 95 86 93 92
5
10 5
41 6
.2Q
141
2 1
...1
12
14 7 8
varied from 2 weeks to 20 years (mean 8.5 months), 141 (92%) were essentially free of symptoms and the remaining 12 (8%) had residual respiratory problems, often persistent tracheomalacia (Table VI). Generally, the patients were lost to follow-up when their symptoms abated.
Comments and conclusions From this large experience several conclusions can be made. Since compression of the trachea and esophagus by vascular anomalies is uncommon and may be masked by nonspecific respiratory symptoms, a high degree of suspicion needs to be maintained so that the diagnosis will not be missed and increased mortality and morbidity and unnecessary procedures will be avoided. An anteroposterior chest roentgenogram and lateral views of the airway (neck and chest) may be helfpul in initiating the diagnostic workup. A barium esophagogram in the hands of a competent radiologist remains the single most important diagnostic procedure in the vascular ring group. However, CT with contrast and magnetic resonance studies gives an excellent image of the vascular anomaly. This is especially true for pulmonary artery sling, in which case CT may replace invasive angiography and, in addition, allow good visualization of tracheal abnormalities if present. Angiography, however, may be required to investigate associated intracardiac defects. Bronchoscopic examination in many of these patients may not be needed, but it is frequently performed because of the initial obscure nature of the respiratory symptoms. We believe extreme care should be exercised during this procedure to avoid aggravating the condition of the compromised airway. It is needed, however, in older children to investigate the tracheobronchial tree to rule out the presence of a foreign body or any intrinsic tracheobronchial abnormalities. The frequency with which arteriopexy of the innominate artery was performed increased markedly at our institution between 1977 and 1980. This increase may
be related to increased awareness of this entity and change in the bronchoscopic technique to general anesthesia instead of local or no anesthesia in the infant; in the latter case the hurried bronchoscopist may have missed the lesion in the struggling infant," Although we have many examples of patients who have been helped or cured of their respiratory obstructive symptoms and apneic episodes after relief of innominate artery compression, we have adopted more conservative criteria for operating on these infants. Neurologic investigation, sleep studies, and radionuclide gastroesophageal reflux studies are performed to rule out other causes of apneic spells. In addition, the bronchoscopic examination should reveal more than 5.0% obstruction before surgical treatment is indicated. We recommend a left thoracotomy for complete vascular rings or pulmonary artery sling, except in those cases in which repair of intracardiac defects or tracheoplasty has to be performed simultaneously; then median sternotomy should be used. Refinements in operative technique that have subjectively improved results are suturing divided rings and performing pulmonary artery anastomoses with fine Prolene sutures, lysing all surrounding bands, and operating as soon as the diagnosis is made, even in infancy. Diligent postoperative care is required for these infants: intensive respiratory therapy, oxygen when needed, humidity, chest physiotherapy, and repeated suctioning of oral pharyngeal secretions. Attention to these details has also subjectively improved results. Occasionally, intubation and assisted ventilation may be necessary, but prolonged endotracheal intubation should be avoided in order not to aggravate the already severely narrowed airway. Normally, the symptoms and signs of tracheobronchial and esophageal obstruction are relieved immediately, but not completely, after division of the vascular ring. This fact should be kept in mind, since some stridor and noisy respiration may continue, and the parents should be alerted to this to relieve undue anxiety. Less than 10% of children in our series had residual tracheomalacia. On occasion, symptoms may continue for a year or two before they resolve completely. Surgical results have been satisfactory, with low morbidity and almost no mortality. REFERENCES 1. Gross RE. Surgical relieffor tracheal obstruction from a vascular ring. N Engl J Moo 1945;233:586-90. 2. Potts WJ, Gibson S, Rothwell R. Double aortic arch: report of two cases. Arch Surg 1948;57:227-33.
3. Nikaidoh H, Riker WL, Idriss FS. Surgical management of "vascular rings." Arch Surg 1972;105:327-33.
Volume 97 Number 5 May 1989
4. Felson B, Palayew MJ. The two types of right aortic arch. Radiology 1963;81:745-59. 5. D'Cruz lA, Cantez T, Namin EP, Licata R, Hastreiter AR. Right-sided aorta. Br Heart J 1966;28:722-39. 6. Potts WJ, Holinger PH, Rosenblum AH. Anomalous left pulmonary artery causing obstruction to right main bronchus: report of a case. JAMA 1954;155:1409-11. 7. Riker WL. Anomalies of the aortic arch and their treatment. Pediatr Clin North Am 1954;1:181-95. 8. Sade RM, Rosenthal A, Fellows K, Castaneda AR. Pulmonary artery sling. J THORAC CARDIOVASC SURG 1975;69:333-46. 9. Idriss FS, Nikaidoh H, DeLeon SY, Koopot R. Surgery for vascular anomalies causing obstruction of the trachea and esophagus. In: Tucker BL, Lindesmith GC, eds. Congenital heart disease. New York: Grune & Stratton, 1979:125-56.
Discussion Dr. George G. Lindesmith (Los Angeles. Calif). At the Childrens Hospital in Los Angeles, our experience over about the same time interval parallels Dr. Backer's rather closely and our results are certainly not dissimilar. Deaths in our series, like yours, occurred rather early. We have had one death in the total series since 1965, which is not totally different from your experience. In your series, however, I noted that over a third of the cases involved were from innominate artery compression. In our practice that has accounted for less than 5% of the cases. Your table showing the age at onset of symptoms and at operation (see Table II) intrigued me, because about a quarter of your group I patients (double aortic arch and the right aortic arch) were omitted from the "age at onset" data. Interestingly also, in the group with innominate artery compression, over two thirds of the patients were omitted from the "age at onset" data. I have two questions from these data, which I will present as one, since they are closely related. We know that most patients with innominate artery compression of the trachea will develop and grow to adulthood without any symptoms. We also know that an occasional case of double aortic arch will be discovered in an adult who has been free of symptoms throughout his life. My questions are these: What are your current indications for suspending the innominate artery, and what are your current minimal indications for operating on a patient with a double aortic arch? . Dr. Backer. I will answer the innominate artery question first. We have encountered a considerable number of such patients, and most were operated on in the late 1970s. Our bronchoesophagologists at that time became very enthusiastic about this diagnosis when they switched from local to general anesthesia for bronchoscopy. They had a backlog of patients who had been treated for other diagnoses, and in the years between 1977 and 1980, about half of these patients were treated surgically. Since that enthusiastic surge and backlog of
Tracheoesophageal compression 7 3 1
patients, the number of innominate arteriopexies has dropped off to a more realistic figure of three to four per year. Our indications have been refined from the days when these patients had apnea, were found to have innominate compression, and were treated surgically. We now perform full neurologic evaluation, including CT scan, electroencephalogram, sleep studies, and radionuclide studies for gastroesophageal reflux. It is only when all these other studies are negative, and the child has more than 50% compression of the trachea by the innominate artery, that we undertake an arteriopexy. Your second question dealt with the minimal indications for operating on a patient with double aortic arch. Most of these patients don't come to our attention unless they have significant symptoms of stridor, respiratory distress, or dysphagia. I would say that any child with symptoms in whom a double aortic arch is detected by esophagogram, or occasionally by CT or angiography, has indications for an operation. Many of these children can do well, with minimal symptoms, for a period of time, but when an upper respiratory tract infection develops they have tracheal edema and can have severe symptoms, sometimes even requiring intubation and ventilation temporarily. Does that answer your questions? Dr. Lindesmith. Yes. My main concern, as you probably surmised, is that it looked as if a significant number of your patients, particularly in the innominate artery suspension group, didn't have any symptoms. I wondered if you operated only for 50% obstruction of the trachea, or if your patients all had to have symptoms before you subjected them to surgical treatment. I think you answered that question. Dr. J. Nilas Young (Oakland. Calif). I would ask that you describe for us the approach in your institution of intraoperative or postoperative endoscopy in the management of these patients. Dr. Backer. In the patients with innominate artery compression, a bronchoscope always was used to make the diagnosis and then again immediately before innominate arteriopexy. A bronchoscopic exam was also done after we finished the arteriopexy to make sure that the compression had been relieved. For the patients with double aortic arch and right arch with ligamentum, we did not usually use bronchoscopy either preoperatively or postoperatively. The patients with pulmonary artery sling, and particularly those who had complete tracheal rings requiring tracheoplasty, generally did have bronchoscopy during the repair. Therefore, bronchoscopy was chiefly used for the innominate artery group and a select number of the pulmonary artery sling group. In closing, there are two points I would like to emphasize. First, in our experience, the barium esophagogram has been the single most important diagnostic tool, even with the newer advent of CT and magnetic resonance imaging. This examination still seems to be sufficient to make the diagnosis in most of the patients. Second, left thoracotomy provides excellent exposure for all these procedures except for innominate arteriopexy, which is best performed with a right thoracotomy.
Vascular anomalies causing tracheoesophageal . compression Review of experience in children Two hundred four infants and children (mean age 13 months) have undergone operation for the relief of tracheoesophageal obstruction resulting from vascular anomalies. One hundred thirteen patients had complete vascular rings (group I), 61 with double aortic arch and 52 with right aortic arch with a left ligamentum. Nine patients had a pulmonaryartery sling(group II), 71 had innominateartery compression (group Ill), and 11 had miscellaneous anomalies (group IV). Patients were admitted with respiratory distress, stridor, apnea, dysphagia, or recurrent respiratory infections. Diagnosis was established by barium esophagogram in group I; barium esophagogram, bronchoscopy, and computed tomography or angiography in group II; bronchoscopy in group Ilk and barium esophagogramor angiography in group IV. The operative approach was through a left thoracotomy in groupsI, II, and IV (93% of these patients) and through a right thoracotomy for group m (96% of these patients).The operativemortality rate was 4.9 % and there were seven late deaths (3.4%). There have been no operative deaths in patients with isolated vascular anomalies in the past 28 years. Follow-up data from 1 month to 20 years (mean 8.5 months) were available on 159 patients; 141 (92%) were essentiallyfree of symptoms, and 12 (8%) had residual respiratory problems. Five of six patients in group II having a lung scan postoperatively had a patent left pulmonary artery. A strong index of suspicion is necessary to avoid the complications of vascular rings in children. Barium swallow is the best single diagnostic technique for patients with complete vascular rings. A bronchoscopic study is required to diagnose innominate artery compression. Angiograms or computedtomographic scans are used to confirmthe diagnosisof pulmonaryartery sling. Left thoracotomy provides excellent exposure for all vascular rings except the displaced innominate artery, for which a right thoracotomy is the best approach.
Carl L. Backer, MD, Michel N. Ilbawi, MD, Farouk S. Idriss, MD, and Serafin Y. DeLeon, MD, Chicago, Ill.
h e first successful operation for a vascular ring was reported by Dr. Robert GrOSSI in 1945 when he divided a double aortic arch in a l-year-old boy. He also introduced the term vascular ring to describe mediastinal vascular anomalies causing tracheoesophageal compression. The first successful operation for a vascular From the Division of Cardiovascular-Thoracic Surgery, Children's Memorial Hospital, and the Department of Surgery, Northwestern University Medical School, Chicago, Ill. Supported in part by the A. C. Buehler Foundation, Park Ridge, Ill. Read at the Fourteenth Annual Meeting of The Western Thoracic Surgical Association in Hawaii, June 22-25, 1988. Address for reprints: Carl L. Backer, MD, Divisionof CardiovascularThoracic Surgery, Children's Memorial Hospital, 2300 Children's Plaza, Chicago, IL 60614.
Table I. Clinical groups Group
II III IV Total
Anomaly Complete vascular rings Double aortic arch Right aortic arch Pulmonary artery sling Innominate artery compression Miscellaneous
No. of patients
113 61 52
9 71
_1_1 204
ring at our institution was performed by Dr. Willis J. Potts in 1947.2 He divided a double aortic arch in an 8-year-old boy. In 1972 Nikaidoh, Riker, and Idriss' reported our early experience with 68 children who underwent surgical relief of vascular rings. This current 725
726
The Journal of Thoracic and Cardiovascular Surgery
Backer et a/.
Table II. Age at onset of symptoms and at operation DAA
Symptoms <1 mo 1-6 mo >6 mo Operation <1 mo 1-6 mo 6-12 mo >12 mo
PA sling
RAA
Innominate compression
No.
%
No.
%
No.
%
No.
%
41 5 2
85 10 4
23 12 5
57 30 12
5 1 0
83 16
19 2 2
82 9 9
4 25 15 17
6 41 24 28
0 16 13 23
31 25 44
1 5 3 0
11 55 33
0 34 18 19
48 25 27
DAA, Double aortic arch; PA, pulmonary artery; RAA, right aortic arch with left ligamentum.
Table m. Associated lesions Patients No.
%
Ventricular septal defect Pulmonary atresia Tetralogy of Fallot Ventricular septal defect Ventricular septal defect Complete tracheal rings
1 I 1 4 1 3
2 2 2 7.6 II 33
Ventricular septal defect Tracheoesophageal fistula
2 4
2.8 5.7
Vascular anomaly Double aortic arch (n = 61) Right aortic arch (n = 52) Anomalous left pulmonary artery (n = 9) Displaced innominate artery (n = 71)
Associated anomaly
report reviews our total experience with 204 children who were operated on at Children's Memorial Hospital in Chicago.
Patients and methods Two hundred four patients have undergone surgical treatment for the relief of tracheoesophageal compression by vascular anomalies from 1947 through 1987. Their ages at the time of the operation ranged from 1 week to 17 years (mean 13 months). The majority (68%) were operated on within the first year of life. The sex distribution was unremarkable except in the pulmonary artery sling group, in which there was an 8 to 1 preponderance of boys. Their clinical records were retrospectively reviewed to obtain specificdata. In particular, sex, age at onset of symptoms, age at operation, year of operation, type of vascular ring, symptoms, physical examination, diagnostic tests, type of operation, site of arch division, postoperative complications, long-term results, length of long-term followup, and associated anomalies were recorded for each patient. Four groups of patients were identified (Table I): Group I had complete vascular rings, either a double aortic arch or a right aortic arch with a left ligamentum arising from the descending aorta; group II had an anomalous left pulmonary artery arising from the right (pulmonary artery sling); group
III had a posteriorly displaced innominate artery; and group IV had miscellaneous anomalies of the aortic arch or its branches not forming a complete ring. Clinical manifestations. The most common symptom was stridor or noisy respiration. Respiratory distress was precipitated frequently by crying or feeding or by swallowing a bolus of solid food that pressed on the soft posterior trachea. There were recurrent respiratory tract infections. A brassy cough, similar to a "seal bark," was described in particular with complete rings. Dysphagia for solid foods was noted usually in older children. Two patients with double aortic arch had a penny lodge in the esophagus preoperatively. Choking, severe hypoxia, and even convulsive seizures were noted on occasion. The ages at onset of symptoms and at operation are listed in Table II. In general, among the patients with a complete ring, those with a double aortic arch had earlier onset of symptoms and were operated on within the first year of life (72%). On physical examination, significant findings were present in over 85% of patients. These included rhonchi, stridor, retractions, tachypnea, wheezing, and rales. Hyperextended neck posture that relieved tracheal obstruction by stretching the tracheal wall and splinting the trachea open was seen in some infants. In general, children with vascular rings were well nourished and normal in size for their age, despite the esophageal compression, because they could take liquid feeding well. Associated lesions were uncommon (Table III). Diagnostic procedures. A chest roentgenogram was obtained for all patients and showed abnormalities in over 80% of these children. Hyperaeration of the lung fields was seen and was sometimes unilateral, especially in patients with a pulmonary artery sling. Varying degrees of atelectasis and pneumonic infiltrates were present. The location of the aortic arch in relation to the trachea was helpful in determining the type of vascular ring. When the location was not clear, a double aortic arch was suspected. In patients with significant symptoms and a right aortic arch on the chest x-ray films, a left ligamentum completing the vascular ring was suspected (Fig. I). Lateral chest and neck roentgenograms sometimes showed anterior tracheal compression from a posteriorly displaced innominate artery. The most reliable procedure for the diagnosis of vascular rings (group I) was the barium esophagogram. Different projections and views were necessary to establish the shape,
Volume 97 Number 5 May 1989
Tracheoesophageal compression 7 2 7
Fig. 1. Chest x-ray film of a 5-year-old boy with recurrent upper respiratory tract infections. Arrow points to the arch of the aorta , which is to the right of the tracheal air column.
location, and persistence of the filling defect. In children with a double aort ic arch, barium esophagogram showed two indentations of unequal size on the anteroposterior projection, with the right arch indentation usually superior and larger (Fig. 2). A right aortic arch with a left ligamentum was visualized as a large right impression and a small oblique left impression from the ligamentum. An anterior indentation was suggestive of a pulmonary artery sling. A bronchoscopic study was performed in one half of the patients with a double aortic arch and one quarter of the patients with a right aortic arch . In patients with pulmonary artery sling, bronchoscopy demonstrated complete tracheal rings (n = 3) and other abnormalities (eparterial right bronchus [n = 2]). All patients with innominate artery compression had a bronchoscopic study, which revealed anterior compression of the tracheal lumen. Angiograms were performed on a few patients with complete vascular rings (group I), in six of nine patients with pulmonary artery sling, and in a few patients with innominate artery compression. Recently, computed tomographic (CT) scans have been obtained. They were diagnostic in patients with vascular rings or slings (Fig. 3) and demonstrated, when present, complete tracheal rings. Magnetic resonance imaging was used in an older patients with a right arch, a left ligamentum from the descending aorta, and a Kommerel diverticulum at the origin of the left subclavian artery (Fig. 4). In patients with suspected innominate artery compression, neurologic evaluation, sleep studies, and radionuclide investigation for gastroesophageal reflux were performed to rule out other possible causes of apneic episodes. Operative findings and management Double aortic arch. The approach to this lesion in all instances but two was a left thoracotomy (59 patients) . In two patients the operation was done through a median sternotomy . The right-sided (posterior) arch was dominant in 45 patients (73%), the left-sided (anterior) arch was dominant in 12
Fig. 2. A 4-month-old boy with stridor. Barium esophagogram shows bilateral indentations (arrows) characteristic of a double aortic arch.
(20%), and the arches were equal In four (7%). With a dominant right arch, segments of the left arch were atretic in 40% of the patients. With a dominant left arch , segments of the right arch were atretic in four of 12 (33%) patients . The atretic area occurred most frequently in the posterior or distal end of the lesser arch. The vascular ring eaused by the double aortic arch was released by dividing the lesser of the two arches, usually at the atretic area . When the arches were patent, the lesser was divided between clamps at a site selected to preserve the brachiocephalic blood flow. Both the carotid and radial pulses were monitored by the anesthesiologist after the vascular clamps were applied. The left ligamentum was always divided and careful dissection around the esophagus and the trachea was performed to lyse any residual adhesive bands . The recurrent laryngeal and phrenic nerves were identified and protected. Right aortic arch and left ligamentum. Left thoracotomy was used in 48 patients and a median sternotomy in four. The vascular ring was released by dividing the ductus or ligamentum and dissecting the trachea and esophagus free from adhesive bands. There were two anatomic variations. In 65% a retroesopha-
7 2 8 Backer et al.
The Journal of Thoracic and Cardiovascular Surgery
Fig. 3. A 7-month-old boy with respiratory distress. CT with contrast shows a pulmonary artery sling. Large arrow points to the left pulmonary artery encircling the trachea (small arrow).
Fig. 4. A 17-year-old patient with dysphagia. Magnetic resonance imaging shows a right aortic arch (small arrow) and a Kommerel diverticulum (large arrow) at the origin of the left subclavian artery.
geal left subclavian artery was present. In 35% there was a mirror-image left innominate artery. In both situations the vascular ring was completed by the ligamentum arteriosum, which stretched from the descending aorta to the pulmonary artery. This distribution is similar to that reported by Felson and Palayew' (56% retroesophageal, 44% mirror image). Theoretically, if there is mirror-image branching with the ligamentum arising from the innominate artery instead of the aorta, a complete ring will not be formed and there will be no symptoms.'
The aorta descended on the right in two thirds of these patients and coursed retroesophageally to descend on the left in the rest. Pulmonary artery sling. The first successful repair of a pulmonary artery sling was reported by Potts, Holinger, and Rosenblum" from our institution in 1954. Since then we have operated on nine patients with an anomalous left pulmonary artery originating from the right. The left pulmonary artery formed a sling as it passed around the origin of the right bronchus between the trachea and esophagus at or above the level of the carina as it coursed toward the left lung. This caused compression of the right bronchus and trachea. In six patients a left thoracotomy was used, in two a median sternotomy, and in one a right thoracotomy. The left pulmonary artery was divided and passed anterior to the trachea and reanastomosed to the right pulmonary artery. In three patients a left ligamentum was also divided. In one patient with complete tracheal rings, two tracheoplasties were performed at later dates; in another, pericardial tracheoplasty with the use of extracorporeal circulation was performed simultaneously. A third patient with complete rings and anomalous tracheal bifurcation has not yet required tracheoplasty. Innominate artery compression. Since the first report by Riker" from our institution in 1954, 71 patients have been operated on for innominate artery compression, the majority between 1977 and 1980. All but three arteriopexies of the innominate artery have been performed through a right anterolateral thoracotomy. The innominate artery was lifted off the trachea by suspending it to the anterior chest wall, usually to the periosteum of the sternum. This was facilitated by leaving a small portion of pericardial reflection attached to the base of the innominate artery for placement of sutures for the suspension. A bronchoscopic examination was sometimes performed during the operation to demonstrate relief of obstruction. Miscellaneous. There were 11 patients in this group. Seven with a left aortic arch and an aberrant right subclavian artery originating from the descending aorta were operated on before
Volume 97 Number 5 May 1989
Tracheoesophageal compression
Table IV. Postoperative mortality in vascular rings-early deaths Year
1946 1949 1951 1952 1953 1954 1957 1959 1973 1974
Age at operation
5 mo 2 mo 5 mo 2 mo 3 mo 2 mo 3 mo 6 yr I mo
I mo
Anatomy
Cause of death
DAA DAA DAA DAA AbRSc RAA RAA RAA RAA
Failure to divide ring Air block syndrome Respiratory failure Hemorrhage intraoperatively Pneumonia, YSD, ASD, PH Respiratory failure, YSD, PH Respiratory failure, YSD Hemorrhage intraoperatively Cardiac arrest in operating room, LTGA, YSD, PS, Waterston shunt Pneumonia, YSD, ASD, PAPYD
INN
AbRSc, Left aortic arch with aberrant right subclavian artery; ASD, atrial septal defect; DAA, double aortic arch; INN, displaced innominate artery; LTGA, L-transposition of the great arteries; PAPVD, partial anomalous pulmonary venous drainage; PH, pulmonary hypertension; PS, pulmonary stenosis; RAA, right aorticarch with left ligamentum; VSD,ventricular septal defect.
1973. Two patients had a left aortic arch with a left ligamentum arteriosum that was divided through a left thoracotomy. There was minimal compression of the trachea and esophagus. Another child had an anomalous right carotid artery compressing the trachea; the artery was suspended to the chest wall through a right thoracotomy approach. A fourth child had aortopexy through a right thoracotomy early in our experience because of tracheal compression by a dilated aortic arch.
Results There were 10 postoperative (up to 30 days) deaths (4.9%) and seven late deaths (3.4%). Most of the deaths occurred in infants under 6 months of age (II / 17) and early in this series (8/10 early deaths before 1960). Over half (9/17) of the patients who died had other complicating cardiac lesions (Tables IV and V). Four early deaths and two late deaths (6.5% and 3.2%, respectively) occurred in the double aortic arch group. Four early deaths (7.6%) and one late death (1.9%) occurred in the right aortic arch group. There were no early deaths but two late deaths (22%) after repair of pulmonary artery sling. Both occurred quite late, at 7 months and 2Y2 years, in patients who required tracheoplasty because of extensive narrowing of the trachea from complete tracheal rings. There was one early (1.5%) and two late deaths (3%) after innominate arteriopexy. Some patients in group I required postoperative treatment with a bronchoscope and four needed a postoperative tracheostomy. In the double aortic arch group, one patient required reoperation for persistent
729
Table V. Postoperative mortality in vascular rings-late deaths Age at Year operation Anatomy
1950 1966 1967 1978
3 wk 6 mo 6 mo 2 rno
DAA RAA DAA INN
1979
6 mo
PAS
1985
7 mo
PAS
1986
6 mo
INN
Cause of death
Postop. survival
Pneumonia, malnutrition 37 days Pneumonia, anoxic eNS 50 days Aspiration, asphyxia 125 days Respiratory failure, aortic 50 days stenosis, coarctation Respiratory failure, 900 days tracheoplasty x2 Respiratory failure, 210 days tracheoplasty Respiratory failure, single 180 days ventricle, pulmonary atresia
CNS, Central nervous system; PAS, pulmonary artery sling. For other abbreviations see Table IV.
stridor, and aortopexy was performed through a right thoracotomy 6 months postoperatively. In the group with right aortic arch, one patient required reoperation at 5 months and had aortopexy through a right thoracotomy. Aortopexy is necessary during the original operation if the two ends of the divided vascular ring do not spring apart naturally to relieve pressure on the trachea and esophagus. The mean hospital stay for this group was 6.5 days. In group II, six of nine patients underwent postoperative bronchoscopic treatment and four of nine required postoperative tracheostomy. One required reoperation after 2 months for postoperative cyanotic spells and had the aorta suspended to the sternum. The mean hospital stay was 22 days. These infants all had a difficult postoperative course. Long-term follow-up with lung scan has been obtained in six patients. Flow was present to the left lung in five of six patients, ranging from 25% to 63% (mean 41%) of the total pulmonary blood flow. One patient had postoperative catheterization and this showed slight stenosis of the anastomosed left pulmonary artery. This patency rate (83% in those studied) is different from the greater than 50% anastomotic occlusion rate reported by Sade and colleagues." In three patients with preoperative tracheostomy, the cannula was successfully removed after innominate arteriopexy. One patient in this group later required a tracheal stent with a rib graft for tracheoma1acia. Two patients were reoperated on, one at 4 months and the other at 5 years for recurrent compression of the trachea. Their mean hospital stay was 7.5 days. The majority of patients had relief of their symptoms. Of 153 patients with a postoperative follow-up that
730
The Journal of Thoracic and Cardiovascular Surgery
Backer et al.
Table VI. Results of surgical treatment Poor
Good
Anatomy Double aortic arch Right aortic arch Pulmonary artery sling Displaced innominate artery Total
No.
%
No.
%
44
90 95 86 93 92
5
10 5
41 6
.2Q
141
2 1
...1
12
14 7 8
varied from 2 weeks to 20 years (mean 8.5 months), 141 (92%) were essentially free of symptoms and the remaining 12 (8%) had residual respiratory problems, often persistent tracheomalacia (Table VI). Generally, the patients were lost to follow-up when their symptoms abated.
Comments and conclusions From this large experience several conclusions can be made. Since compression of the trachea and esophagus by vascular anomalies is uncommon and may be masked by nonspecific respiratory symptoms, a high degree of suspicion needs to be maintained so that the diagnosis will not be missed and increased mortality and morbidity and unnecessary procedures will be avoided. An anteroposterior chest roentgenogram and lateral views of the airway (neck and chest) may be helfpul in initiating the diagnostic workup. A barium esophagogram in the hands of a competent radiologist remains the single most important diagnostic procedure in the vascular ring group. However, CT with contrast and magnetic resonance studies gives an excellent image of the vascular anomaly. This is especially true for pulmonary artery sling, in which case CT may replace invasive angiography and, in addition, allow good visualization of tracheal abnormalities if present. Angiography, however, may be required to investigate associated intracardiac defects. Bronchoscopic examination in many of these patients may not be needed, but it is frequently performed because of the initial obscure nature of the respiratory symptoms. We believe extreme care should be exercised during this procedure to avoid aggravating the condition of the compromised airway. It is needed, however, in older children to investigate the tracheobronchial tree to rule out the presence of a foreign body or any intrinsic tracheobronchial abnormalities. The frequency with which arteriopexy of the innominate artery was performed increased markedly at our institution between 1977 and 1980. This increase may
be related to increased awareness of this entity and change in the bronchoscopic technique to general anesthesia instead of local or no anesthesia in the infant; in the latter case the hurried bronchoscopist may have missed the lesion in the struggling infant," Although we have many examples of patients who have been helped or cured of their respiratory obstructive symptoms and apneic episodes after relief of innominate artery compression, we have adopted more conservative criteria for operating on these infants. Neurologic investigation, sleep studies, and radionuclide gastroesophageal reflux studies are performed to rule out other causes of apneic spells. In addition, the bronchoscopic examination should reveal more than 5.0% obstruction before surgical treatment is indicated. We recommend a left thoracotomy for complete vascular rings or pulmonary artery sling, except in those cases in which repair of intracardiac defects or tracheoplasty has to be performed simultaneously; then median sternotomy should be used. Refinements in operative technique that have subjectively improved results are suturing divided rings and performing pulmonary artery anastomoses with fine Prolene sutures, lysing all surrounding bands, and operating as soon as the diagnosis is made, even in infancy. Diligent postoperative care is required for these infants: intensive respiratory therapy, oxygen when needed, humidity, chest physiotherapy, and repeated suctioning of oral pharyngeal secretions. Attention to these details has also subjectively improved results. Occasionally, intubation and assisted ventilation may be necessary, but prolonged endotracheal intubation should be avoided in order not to aggravate the already severely narrowed airway. Normally, the symptoms and signs of tracheobronchial and esophageal obstruction are relieved immediately, but not completely, after division of the vascular ring. This fact should be kept in mind, since some stridor and noisy respiration may continue, and the parents should be alerted to this to relieve undue anxiety. Less than 10% of children in our series had residual tracheomalacia. On occasion, symptoms may continue for a year or two before they resolve completely. Surgical results have been satisfactory, with low morbidity and almost no mortality. REFERENCES 1. Gross RE. Surgical relieffor tracheal obstruction from a vascular ring. N Engl J Moo 1945;233:586-90. 2. Potts WJ, Gibson S, Rothwell R. Double aortic arch: report of two cases. Arch Surg 1948;57:227-33.
3. Nikaidoh H, Riker WL, Idriss FS. Surgical management of "vascular rings." Arch Surg 1972;105:327-33.
Volume 97 Number 5 May 1989
4. Felson B, Palayew MJ. The two types of right aortic arch. Radiology 1963;81:745-59. 5. D'Cruz lA, Cantez T, Namin EP, Licata R, Hastreiter AR. Right-sided aorta. Br Heart J 1966;28:722-39. 6. Potts WJ, Holinger PH, Rosenblum AH. Anomalous left pulmonary artery causing obstruction to right main bronchus: report of a case. JAMA 1954;155:1409-11. 7. Riker WL. Anomalies of the aortic arch and their treatment. Pediatr Clin North Am 1954;1:181-95. 8. Sade RM, Rosenthal A, Fellows K, Castaneda AR. Pulmonary artery sling. J THORAC CARDIOVASC SURG 1975;69:333-46. 9. Idriss FS, Nikaidoh H, DeLeon SY, Koopot R. Surgery for vascular anomalies causing obstruction of the trachea and esophagus. In: Tucker BL, Lindesmith GC, eds. Congenital heart disease. New York: Grune & Stratton, 1979:125-56.
Discussion Dr. George G. Lindesmith (Los Angeles. Calif). At the Childrens Hospital in Los Angeles, our experience over about the same time interval parallels Dr. Backer's rather closely and our results are certainly not dissimilar. Deaths in our series, like yours, occurred rather early. We have had one death in the total series since 1965, which is not totally different from your experience. In your series, however, I noted that over a third of the cases involved were from innominate artery compression. In our practice that has accounted for less than 5% of the cases. Your table showing the age at onset of symptoms and at operation (see Table II) intrigued me, because about a quarter of your group I patients (double aortic arch and the right aortic arch) were omitted from the "age at onset" data. Interestingly also, in the group with innominate artery compression, over two thirds of the patients were omitted from the "age at onset" data. I have two questions from these data, which I will present as one, since they are closely related. We know that most patients with innominate artery compression of the trachea will develop and grow to adulthood without any symptoms. We also know that an occasional case of double aortic arch will be discovered in an adult who has been free of symptoms throughout his life. My questions are these: What are your current indications for suspending the innominate artery, and what are your current minimal indications for operating on a patient with a double aortic arch? . Dr. Backer. I will answer the innominate artery question first. We have encountered a considerable number of such patients, and most were operated on in the late 1970s. Our bronchoesophagologists at that time became very enthusiastic about this diagnosis when they switched from local to general anesthesia for bronchoscopy. They had a backlog of patients who had been treated for other diagnoses, and in the years between 1977 and 1980, about half of these patients were treated surgically. Since that enthusiastic surge and backlog of
Tracheoesophageal compression 7 3 1
patients, the number of innominate arteriopexies has dropped off to a more realistic figure of three to four per year. Our indications have been refined from the days when these patients had apnea, were found to have innominate compression, and were treated surgically. We now perform full neurologic evaluation, including CT scan, electroencephalogram, sleep studies, and radionuclide studies for gastroesophageal reflux. It is only when all these other studies are negative, and the child has more than 50% compression of the trachea by the innominate artery, that we undertake an arteriopexy. Your second question dealt with the minimal indications for operating on a patient with double aortic arch. Most of these patients don't come to our attention unless they have significant symptoms of stridor, respiratory distress, or dysphagia. I would say that any child with symptoms in whom a double aortic arch is detected by esophagogram, or occasionally by CT or angiography, has indications for an operation. Many of these children can do well, with minimal symptoms, for a period of time, but when an upper respiratory tract infection develops they have tracheal edema and can have severe symptoms, sometimes even requiring intubation and ventilation temporarily. Does that answer your questions? Dr. Lindesmith. Yes. My main concern, as you probably surmised, is that it looked as if a significant number of your patients, particularly in the innominate artery suspension group, didn't have any symptoms. I wondered if you operated only for 50% obstruction of the trachea, or if your patients all had to have symptoms before you subjected them to surgical treatment. I think you answered that question. Dr. J. Nilas Young (Oakland. Calif). I would ask that you describe for us the approach in your institution of intraoperative or postoperative endoscopy in the management of these patients. Dr. Backer. In the patients with innominate artery compression, a bronchoscope always was used to make the diagnosis and then again immediately before innominate arteriopexy. A bronchoscopic exam was also done after we finished the arteriopexy to make sure that the compression had been relieved. For the patients with double aortic arch and right arch with ligamentum, we did not usually use bronchoscopy either preoperatively or postoperatively. The patients with pulmonary artery sling, and particularly those who had complete tracheal rings requiring tracheoplasty, generally did have bronchoscopy during the repair. Therefore, bronchoscopy was chiefly used for the innominate artery group and a select number of the pulmonary artery sling group. In closing, there are two points I would like to emphasize. First, in our experience, the barium esophagogram has been the single most important diagnostic tool, even with the newer advent of CT and magnetic resonance imaging. This examination still seems to be sufficient to make the diagnosis in most of the patients. Second, left thoracotomy provides excellent exposure for all these procedures except for innominate arteriopexy, which is best performed with a right thoracotomy.