V alies may be a major cause of tracheoesophageal

Transcription

1 Diagnosis and Management of Congenital Vascular Rings: A 22-Year Experience Karen Chun, MD, Paul M. Colombani, MD, David L. Dudgeon, MD, and J. Alex Haller, Jr, MD Division of Pediatric Surgery, The Johns Hopkins University School of Medicine, Baltimore, Maryland Between 1968 and 1990, we operatively treated 39 patients (19 boys, 20 girls) with congenital aortic arch anomalies. Median age was 7 months (range, 1.5 months to 23 years). Thirty-seven patients (95%) had respiratory symptoms. Barium swallow was diagnostic in 95%. Right arch with aberrant left subclavian artery and double aortic arch were the most common types (11 each). Treatment of an aortic diverticulum was documented in 19 patients; the aortic diverticulum was excised (91, managed by aortopexy (7), or left in situ (3). Postoperative recovery was rapid, with a median intensive care unit stay of 2 days, time to oral feeding of 1 day, and postoperative time to discharge of 7 days. Two deaths occurred: 1 infant had undergone emergent operation for control of hemorrhage from an aortotracheal fistula due to tracheostomy tube erosion, and the other had multiple associated congenital heart defects. Postoperative complications included bleeding (l), pneumonia (5), and chylothorax (4). One boy had persistent severe symptoms due to an untreated aortic diverticulum and underwent subsequent excision of the aortic diverticulum with complete relief of symptoms. Median length of follow-up was 12.5 months, with at least 97% of survivors completely or nearly completely free of symptoms from the vascular ring. These results suggest that early repair of congenital aortic vascular rings, including fixating or excising an associated serious aortic diverticulum, is safe and effective and allows for normal tracheal growth. (Ann Thorac Surg 1992;53: ) ascular rings and other congenital aortic arch anom- V alies may be a major cause of tracheoesophageal obstruction in neonates. Division of the ring has become the standard of surgical treatment; operative management of an associated aortic (Kommerell's) diverticulum, however, is much less clearly defined. Since 1968 we have operatively managed 39 cases of symptomatic aortic arch anomalies, including 19 aortic diverticula. We undertook this study to assess general outcome, including improvement in airway function, with a particular focus on treatment of an associated aortic diverticulum. Clinical Materials and Results Between January 1968 and December 1990, 39 cases of congenital vascular rings were surgically treated at the Johns Hopkins Children's Medical and Surgical Center on the Pediatric Surgery service. Patients with associated congenital heart anomalies, for which open heart surgical correction was simultaneously undertaken, were excluded from the retrospective study except for 1 boy whose vascular ring was repaired before definitive treatment of concomitant congenital heart defects. Aortic arch anomalies were categorized according to the Mayo Clinic classification [l]. In this series, the most common types were double aortic arch (Mayo type IA, 11 Presented at the Thirty-eighth Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 7-9, Address reprint requests to Dr Haller, Division of Pediatric Surgery, The Johns Hopkins University School of Medicine, 601 N Broadway, Baltimore, MD or 28%) (Fig 1) and right aortic arch with aberrant left subclavian artery (Mayo type IIIB1, 11 or 28%) (Fig 2). Of the double arch rings, the right posterior arch was dominant in 7 (64%). Seven patients (18%) had double arch with atretic left arch (IB); 4 (10%) had right arch with mirror-image branching (IIIA); 3 (8%) had isolated anomalous innominate arteries; and 2 (5%) had isolated aberrant right subclavian arteries that arose from a left descending thoracic aorta and coursed behind the esophagus. The presence or absence of an associated aortic diverticulum was documented by angiography or at operation in 29 patients (74%). Median age at presentation was 7 months (range, 6 weeks to 23 years). Eighty percent of the patients were less than 2 years of age when first seen, and 64% were less than 1 year of age. There were 19 boys (49%) and 20 girls (51%). Thirty-seven patients (95%) had respiratory symptoms, which consisted of stridor, cough, wheezing, or a combination of these symptoms. Thirteen (35%) of these patients also had dysphagia or evidence of poor feeding. Only 2 patients had dysphagia alone. One of these was a 23-year-old woman who had had an incomplete repair of her vascular ring at an outside facility during childhood. The other was a 4-month-old infant with gastroesophageal reflux and failure to thrive. The median duration of symptoms before presentation was 4.5 months; the average duration was 16.3? 4.8 months (range, 1 month to 11 years). Seventeen patients had had from 1 to 4 episodes of by The Society of Thoracic Surgeons /92/$5.00

2 598 CHUNETAL Ann Thorac Surg 1992;53: t Fig 1. Double aortic arch, Mayo type IA vascular ring. Trachea and esophagus are encircled by the vascular ring. (LCC = left common carotid artery; LD = left ductus; LPA = left pulmonary artery; LS = left subclavian artery; RCC = right common carotid artery; RD = right ductus; RPA = right pulmonary artery; RS = right subclavian artery.) pneumonia for which they were treated on an outpatient basis or were hospitalized. Often symptoms were noted at birth but diagnosis was delayed because medical attention was not sought or because the underlying condition was not suspected by the primary physician. Most commonly, in these latter cases, symptoms were ascribed to bronchitis, tracheitis, or asthma. Associated congenital anomalies were present in 6 patients. A vascular ring was not initially suspected in 1 patient with a laryngeal web that was treated by laser multiple times before eventual division of his vascular ring. Another had Rubenstein-Tabyi syndrome with mental retardation. Imperforate anus with perineal fistula were associated anomalies in 1 other patient. The latter also had horseshoe kidney and vesicoureteral reflux. Two additional patients also had vesicoureteral reflux, and 1 of these had pectus carinatum. The only patient with associated congenital heart disease was found, after division of his vascular ring, to have a double-outlet right ventricle and pulmonary stenosis which required an open heart operation. One patient presented with a major complication of her anomaly. She was initially seen in extremis with exsanguinating hemorrhage from a tracheoaortic fistula resulting from erosion of a previously placed tracheostomy tube into a vascular ring. A familial incidence occurred only once: 2 sisters were affected with type IIIBl rings. At the time of repair of the older sister s vascular ring at 8 years of age, her 1-year-old sister was being evaluated for the same condition. Diagnostic Eva1 uat ion Barium esophagograms were obtained in all patients and were diagnostic in all but 2, both of whom were found to have anomalous innominate arteries. Aortography was performed in 87% with 100% accuracy. Computed tomographic scanning was not performed on any patient; more recently, however, magnetic resonance imaging (MRI) has been used instead of angiography to avoid intravenous contrast medium injection. In this study, MRI was diagnostic for a vascular ring per se in all 3 of the patients for whom it was obtained but was incorrect as to the type of ring in 1 patient. In this patient a double arch was diagnosed by MRI but was found to be a right arch with aberrant left subclavian artery at operation. Operative Management Exposure was through a standard left posterolateral thoracotomy incision in 34 cases (87%). Median sternotomy was performed for the 3 cases of anomalous innominate artery and for the emergent case of tracheoaortic fistula. One patient underwent bilateral thoracotomy for repair of a type IA ring with a dominant left arch. In this unusual patient, a left patent ductus arteriosus was first ligated and divided through a left thoracotomy; a right thoracotomy was then required for division of the nondominant right arch and for reimplantation of an aberrant right subclavian artery. In the 7 patients with double arch and dominant right arch, the nondominant left arch was divided either between the left common carotid and left subclavian arteries or just distal to the left subclavian artery. In the case of a nondominant right arch, the arch R R CG Fig 2. Right aortic arch with aberrent left subclavian artery, Mayo type IllBl vascular ring. Left ductus completes the vascular ring formed by the anomalous right aortic arch and aberrent left subclavian artery. (ALS = aberrent left subclavian artery; LCC = left common carotid artery; LD = left ductus; PT = pulmonary trunk; RCC = right common carotid artery; RS = right subclavian artery.)

3 Ann Thorac Surg 1992;53: CHUNETAL 599 Fig 3. (Upper) Exposure of Mayo type IA vascular ring through a standard left posterolateral thoracotomy incision. The left anterior aortic arch gives rise to the left subclavian artery just proximal to Kommerell's diverticulum. Trachea and esophagus are compressed within the vascular ring. (Lower) Repair of vascular ring-division of ligamenturn arteriosum. Ligamentum arteriosum has been doubly ligated and divided. Kommerell's diverticulum intact. was divided between the right common carotid and right subclavian arteries. The ligamentum arteriosum or patent ductus arteriosus was suture ligated and divided in all cases (Figs 3, 4). Patients with type IIIBl vascular rings generally underwent division of the left ligamentum arteriosum (Fig 5). In all patients with complete or partial vascular rings, after division of the ring, the esophagus and trachea were extensively freed from the surrounding adventitial tissues. Treatment of an aortic diverticulum was documented in 19 of 29 cases. The diverticulum was excised flush with the aorta in 9 patients (see Figs 4, 5), managed by aortopexy in 7 (Fig 6), and left in situ in 3. The mode of treatment was determined by the operating surgeon's preference. There was no significant difference in mode of treatment of a diverticulum among the 4 types of vascular rings (IA, IB, IIIA, IIIB). Anomalous innominate arteries were treated by suspension to the sternum. One isolated aberrant retroesophageal right subclavian artery was managed by division without reimplantation in an 8-month-old boy. No evidence of subclavian steal has been noted. In the patient for whom emergent operation was required, exsanguinating hemorrhage from tracheostomy tube erosion was found arising from a tracheoaortic fistula just distal to the right subclavian artery in a type IA vascular ring. Division of the right posterior arch at the bleeding site controlled the hemorrhage. The tracheal defect was then repaired with a pericardial patch. Outcome Among the 37 survivors, postoperative recovery was rapid with a median intensive care unit stay of 2 days (range, 1 to 46 days), median time to oral feedings of 1 day (range, 1 to 14 days), and median postoperative time to discharge of 7 days (range, 4 to 135 days). None of these patients required tracheostomy or prolonged ventilatory support. No instances of recurrent laryngeal nerve injury were documented. Complications occurred in 13 patients Fig 4. Excision of Kommerell's diverticulum. Completed repair of Mayo type IA vascular ring. The left anterior arch has been divided just distal to the origin of the left subclavian artery. Left ductus has been doubly ligated and divided. Kommerell's diverticulum has been excised Push with the descending aorta.

4 600 CHUNETAL Ann Thorac Surg 1992; Fig 5. (Upper) Repair of type liibl vascular ring. Left ligarnentum has been doubly ligated and divided. Kommerell s diverticulum remains intact. (Lower) Excision of Kommerell s diverticulum. Vascular ring has been divided. Aortic diverticulum has been excised flush with the descending aorta. (33%) and included postoperative bleeding requiring reoperation for control (3), respiratory arrest with reintubation (2), pneumonia (5), pneumothorax (l), and sepsis (1). Four patients had development of chylothorax, 1 bilaterally. Two of these required operative ligation of the thoracic duct; the others were successfully managed by chest tube drainage and dietary restriction. One boy had persistent severe symptoms due to partial obstruction of the trachea by an untreated aortic diverticulum. He underwent subsequent excision of the diverticulum with Fig 6. Aortopexy of Kommerell s diverticulum. (Upper) Vascular ring has been divided. (Lower) Kommerell s diverticulum is secured to endothoracic fascia (aortopexy), leaving the adjacent trachea and esophagus completely free of extrinsic compression.

5 Ann Thorac Surg 1992; CHUNETAL 601 complete relief of symptoms. This is the patient who prompted our belief that a large retained diverticulum may cause residual obstruction. Another infant's hospital stay was prolonged by a recurrent congenital laryngeal web that required repetitive laser treatments. Two deaths occurred (5%); one was in the infant with tracheoaortic fistula. Twenty-four hours after emergent repair of the fistula and division of her vascular ring, the patient was returned to the operating room for control of bleeding from the proximal right arch stump. The patient's condition then deteriorated over the ensuing 20 days. The other death occurred in a 2-month-old boy with double-outlet ventricle and pulmonary stenosis who underwent a routine type IA vascular ring repair. His postoperative course was complicated by apneictcyanotic spells that forced a decision to proceed with definitive correction of the congenital heart defects. The open heart repair was unsuccessful and the patient eventually succumbed to multiple organ failure. Median length of follow-up of 32 survivors was 12.5 months (range, 10 weeks to 20 years). Five patients (13%) were lost to follow-up. Of the patients with documented clinic or emergency room visits after discharge, 12 (38%) were asymptomatic and 15 (47%) had mild residual symptoms (cough, brassy voice) at 6 months to 1 year after repair. Respiratory symptoms required no specific treatment and improved gradually over approximately 1 year concomitant with tracheal growth. Less than optimal outcome was related to persistent clinical conditions that were unrelated to vascular ring repair, ie, reactive airway disease, esophageal stricture, recurrent laryngeal web, and gastroesophageal reflux with failure to thrive in 1 patient each. The single patient with a poor operative result was ultimately asymptomatic after reoperation for his aortic diverticulum. Comment The safety and effectiveness of immediate operative repair of symptomatic congenital vascular rings has been confirmed by multiple authors [2-61. Correction in early infancy relieves dyspnea and dysphagia, prevents aspiration-induced pneumonia and other respiratory infections, and most importantly, allows for normal tracheal growth. Since the mid-1970s we have been performing selective excision of an associated Kommerell's aortic diverticulum in addition to division of the ring. Aortopexy of the aortic diverticulum is an alternative to excision in some cases. Clinical evidence suggests that the diverticulum may be a major cause of residual tracheal or esophageal compression and if left in situ after division of the vascular ring, may result in continuing obstructive symptoms [7]. Resection of the diverticulum has also been advised by Campbell [8] because aneurysmal dilation of the diverticulum is known to occur. In our series 2 patients required reoperation for persistent symptoms after incomplete ring division. One of these patients was the 23-year-old woman who had had a vascular ring operation in childhood and presented with a 4-month history of dysphagia. Despite reoperation with excision of an aortic diverticulum and lysis of adhesions around the esophagus, the patient continued to complain of dysphagia. She was subsequently diagnosed as having an esophageal stricture and spastic esophageal dysmotility. Her case was complicated by concomitant chronic psychiatric difficulties including drug dependency, and thus a functional rather than organic cause may, at least partly, account for the less-than-optimal result. In the case of the other patient who underwent reoperation for an incompletely divided vascular ring, the results were remarkable. The patient was seen at 4 months of age with respiratory symptoms since birth. He underwent division of his type IB ring, but postoperatively continued to have serious respiratory symptoms including apnea requiring reintubation. After reoperation and excision of an associated aortic diverticulum his symptoms disappeared. Thus, although we could not make any valid statistical comparisons relative to diverticulum treatment because of the small number of cases in each group, we believe that aggressive management of the Kommerell diverticulum by aortopexy or excision of the diverticulum is indicated in the surgical management of congenital aortic vascular rings. Exposure through a left thoracotomy incision is generally accepted as the operation of choice for division of a vascular ring. Jung and associates [7] once recommended a right thoracotomy, but we found that a standard left posterolateral thoracotomy provided excellent exposure in all but 1 case. A right thoracotomy or median sternotomy is, however, the approach of choice for division and anastomosis of an isolated anomalous right subclavian artery [5]. An anomalous innominate artery may be approached through a median sternotomy or a right anterolateral thoracotomy to suture the artery to the undersurface of the sternum. A barium esophagogram is extremely reliable in diagnosing a vascular ring but does not delineate the precise vascular anatomy. Aortography has been the gold standard in providing a preoperative road map of the anomaly [9]. Angiograms may be confusing, however, because of overlapping of various vessels [lo] and may not distinguish a IB (one arch atresia) from a IIIBl anomaly (right arch with aberrant left subclavian). Recently we have been using MRI as a noninvasive alternative to aortography. Diagnostic accuracy was 100% for a vascular ring in 3 patients, but accurate delineation of the vascular anomaly appeared to be less clear than with aortography because vascular ring type was incorrectly interpreted in 1 of the 3 patients. With the rapid evolution of this technology, however, and with more experience interpreting the scans and because it is not invasive, MRI may soon supersede aortography, especially in the pediatric population, as the diagnostic modality of choice in the definitive work-up of patients with congenital vascular rings [ll, 121. Our review reiterates the effectiveness of surgical intervention in infancy or early childhood for congenital vascular rings. For the routine patient with an isolated aortic arch anomaly that is diagnosed in infancy, early surgical repair results in complete or nearly complete cure. Most

6 602 CHUNETAL Ann Thorac Surg 1992;53: patients can be discharged between 4 and 7 days postoperatively. A slower regression of respiratory symptoms is expected in some patients because of preexisting pulmonary disease and mild to moderate tracheomalacia. Concomitant major congenital anomalies were the causes of prolonged hospital stay, rehospitalization, poorer outcome, and death. Clinical suspicion of a vascular ring as the cause of serious tracheoesophageal obstruction in a young child should lead to early diagnosis and referral for operative treatment. Serious airway compromise as well as recurrent respiratory infections and destructive pneumonitis can produce chronic illness and debilitation in older children and adults. Such a course can be prevented by early diagnosis and definitive operative treatment, including aggressive management of an associated Kommerell's diverticulum. References 1. Stewart JR, Kincaid OW, Edwards JE. Atlas of vascular rings and related malformations of the aortic arch system. Springfield, IL: Charles C. Thomas, Binet JP, Langlois J. Aortic arch anomalies in children and infants. J Thorac Cardiovasc Surg 1977;73: Arciniegas E, Hakimi M, Hertzler JH, Farooki ZQ, Green EW. Surgical management of congenital vascular rings. J Thorac Cardiovasc Surg 1979;77: Richardson JV, Doty DB, Rossi NP, Ehrenhaft JL. Operation for aortic arch anomalies. Ann Thorac Surg 1981;31:42&32. Wychulis AR, Kincaid OW, Weidman WH, Danielson GK. Congenital vascular ring: surgical considerations and results of operation. Mayo Clin Proc 1971;46: Backer CL, Ilbawi MN, ldriss FS, DeLeon SY. Vascular anomalies causing tracheoesophageal compression: review of experience in children. J Thorac Cardiovasc Surg 1989;97: Jung JY, Almond CH, Saab SB, Lababidi Z. Surgical repair of right aortic arch with aberrant left subclavian artery and left ligamentum arteriosum. J Thorac Cardiovasc Surg 1978;75: Campbell CF. Repair of an aneurysm of an aberrant retroesophageal right subclavian artery arising from Kommerell's diverticulum. J Thorac Cardiovasc Surg 1971;62: Vallette RC, Arensman RM, Falterman KW, Oschner JL. Tracheoesophageal compression syndromes related to vascular ring. South Med J 1989;82: Tucker BL, Meyer BW, Lindesmith GG, Stiles QR, Jones JC. Congenital aortic vascular ring. Arch Surg 1969;99: Gomes AS, Lois JF, George 8, Alpan G, Williams RG. Congenital abnormalities of the aortic arch: MRI imaging. Radiology 1987;165: Gomes AS. MR imaging of congenital anomalies of the thoracic aorta and pulmonary arteries. Radio1 Clin North Am 1989;27: DISCUSSION DR CONSTANTINE MAVROUDIS (Chicago, IL): I would like to congratulate Dr Haller on a beautifully presented landmark report and also to share a piece of nostalgic history. J. Willis Potts, whom he knew well, was the first chief of Cardiac Surgery at Children's Memorial Hospital in Chicago. His clinical contribution to the surgery of vascular rings is well documented. This resulted in a large experience which included the first description of successful treatment of a pulmonary artery sling in Since 1947, surgeons at Children's Memorial Hospital have performed more than 225 operations for vascular rings with a 5% early and 5% late mortality. The mortality was highest in the early years. Over the last 15 years, the mortality has been quite low and relates to preexisting associated cardiac anomalies. The excellent results presented by Dr Haller and his associates mirrors the great traditions of his own and sister institutions. I noticed that the authors did not present their experience with pulmonary artery sling. We have employed pericardial patch tracheoplasty with cardiopulmonary bypass with excellent results. I would like to ask the authors how they manage this difficult group of patients. I enjoyed this report very much. DR HALLER I am very pleased that you brought up Dr Potts' name in the discussion, because he was one of the early surgeons managing children with congenital heart abnormalities as well as the vascular rings. The reason we did not include any pulmonary artery slings in this group of patients is because we excluded all patients who required a concomitant open heart operation. Those who required open heart operations and happened to have associated vascular rings, such as an aberrant left pulmonary artery, are not included in this series. In answer to your question, we would manage those pulmonary slings with open heart surgical techniques because of the necessity of dividing the ring within the mediastinum. DR JOSEPH T. WALLS (Columbia, MO): I enjoyed your report very much. We have just reviewed a 34-year experience with vascular ring; however, there were fewer patients in our series. We also noticed that only a few patients had intracardiac anomalies. One of the important aspects of treatment reinforced in the review was that in addition to any other maneuvers required to divide the vascular ring, one should explore the ductus or ligamentum area to make sure it is also released. One of the patients had a pexing procedure, and I was somewhat skeptical of the technique, but a 17-year follow-up showed that it held up very nicely. Of other importance in our experience was that 2 people, 1 with an aberrant right subclavian artery arising as the fourth branch off of the arch, had development of an aneurysm later in life-this was not one that was initially treated but was picked up de novo as an aneurysm. Also, one who had a double aortic arch surgically interrupted had an ascending aortic aneurysm and aortic valve insufficiency develop in adulthood which required aortic valve and root replacement. In follow-up, did any of your patients have degenerative disease of the remaining arch vessels? DR HALLER Thank you very much for your comments. I am familiar with 1 patient who had an aneurysm develop in the aortic diverticulum, which is another reason for excising it, but only 1 or 2 such patients have been reported. None of our patients has had development of degenerative diseases up to this point in a 22-year follow-up of them. You notice that there were some patients who were lost to follow-up, and that always makes

7 Ann Thorac Surg 1992;53: CHUNETAL 603 you nervous, but of those we were able to follow, there were no long-term vascular problems associated with dividing the vascular ring. DR ED L. HOOVER (Buffalo, NY): I too enjoyed the report very much, and I wonder if you ever see problems akin to tracheomalacia or persistent tracheal obstruction, particularly in the children who are operated on at a somewhat older age. If you have, can you predict those children who will need a tracheal resection at the time, or have you had to go back at a later date to repair them secondarily? DR HALLER We did not have a chance to emphasize some of the associated tracheomalacia-type symptoms, but they are in our group called good. All of these children ultimately outgrew airway symptoms, and it was not necessary to carry out any tracheal reconstruction in this group of patients. I believe most of them are on the basis of compression, and the earlier the ring is divided the better likelihood that the trachea will grow and develop and not have long-term sequelae. If you do not operate on them until they are older children or young adults, then you may have persistent problems with tracheomalacia. DR MICHAEL S. SWEENEY (Houston, TX): The last questioner asked the question that I was going to ask, but I would like to add my congratulations to Dr Haller. I can remember slogging through a few of these with you. As a matter of fact I think I can most easily identify with your talk when you show the complications slide. But at any rate, congratulations. It is a truly noteworthy experience from a very creative and very talented surgeon. DR HALLER Thank you very much, Dr Sweeney. It is certainly true that the residents take care of most of the complications, am I right, President Sade? Thank you very much.