European Journal of Cardio-thoracic Surgery 28 (2005) 11–15 www.elsevier.com/locate/ejcts
Tracheomalacia in oesophageal atresia: morphological considerations by endoscopic and CT study* Vito Brigantia,*, Lucia Orioloa, Vitaliano Buffab, Salvatore Garofaloc,d, Sebastiano Cavallaroc, Alessandro Calistia a
U.O.C. di Chirurgia Pediatrica, Azienda Ospedaliera “S. Camillo-Forlanini”, Via Cicerone 60, 00193 Roma, Italy b Dipartimento di Radiologia Generale, Azienda Ospedaliera “S. Camillo-Forlanini”, Roma, Italy c U.O.C. di Chirurgia Pediatrica A, Azienda Ospedaliera Materno Infantile “O.I.R.M-S.Anna”, Torino, Italy d Dipartimento di Radiologia Generale, Azienda Ospedaliera Materno Infantile “O.I.R.M-S.Anna”, Torino, Italy Received 7 February 2005; received in revised form 31 March 2005; accepted 4 April 2005
Abstract Objective: A Tracheomalacia complicates 11–33% of cases of Oesophageal Atresia with distal Tracheo-Oesophageal Fistula. The lesion generally involves only the thoracic segment of the trachea, and it has close anatomical relationships with the mediastinal structures, specially with the aortic arch. We therefore tried to define the most important morphotypes of tracheobronchial malacia by using dynamic fiberoptic bronchoscopy (DFB) and spiral multilayer computed tomography (CT). Methods: Between 1999 and 2003 we studied 40 children from two different institutions who had been operated on at birth for oesophageal atresia. All patients were been submitted to DFB, and the positive cases underwent examination by CT with an iodinated contrast medium. CT angiographic images of great vessels and multiplanar and threedimensional images of the airways (virtual broncoscopy and broncography) were obtained for morphological evaluation. Results: Twenty-five patients (62%) tested positive for malacia using DBF and all were also confirmed by CT study. In 11 cases (46%), the malacia was located at the thoracic section of the trachea, which was occluded by compression of the aorto-innominate complex. A simple intrinsic tracheomalacia without any vascular compression was present in eight cases (33%), while in five cases (21%), the malacia was complex. Conclusions: A correct morphological analysis of the malformed segment permitted ‘tailored surgery’ for each individual patient, allowing us to take account of the type of malacia, its length, and the compressive action exercised by the mediastinal great vessels. Q 2005 Elsevier B.V. All rights reserved. Keywords: Tracheal disease; Trachea; Oesophageal atresia; Child
1. Introduction A severe tracheomalacia (TM) is found in 11–33% of cases of oesophageal atresia (OA) with distal tracheo-oesophageal fistula (TOF) present at birth, while it is very rare in other forms of OA [1,2]. TM is generally segmentary, located in the thoracic segment above the carina; it does not exceed 1 cm in length, and the aortic arch often significantly influences the dynamics of the obstruction [3]. Subsequent studies have, nevertheless, underlined that the obstructive dynamics of a TM can also be influenced by the epiaortic vessels [4,5], and/or the pulmonary artery [6]. The situation can be further complicated if it is present in a right aortic arch, or the malacia involve one or both main-stem bronchi,
* Classification of different forms of Tracheomalacia associated to Oesophageal Atresia. It is a morphological analysis by endoscopic and Spiral Multilayer Computed Tomography study. * Corresponding author. Tel.: C39 06 3212234; fax: C39 06 58204592. E-mail address:
[email protected] (V. Briganti).
1010-7940/$ - see front matter Q 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.ejcts.2005.04.003
or the TM is intrinsic and is not correlated to compression by the great mediastinal vessels. We have therefore tried to define the most important tracheo-bronchial malacia morphotypes by means of a dynamic fiberoptic bronchoscopic (DFB) study and the use of axial helical computed tomography (CT).
2. Methods In the period between January 1999 and December 2003 we studied 40 children (28 males and 12 females), from two different institutions, with an average age of 31.8 months (ranging from 21 days to 120 months), who had been operated on at birth for OA. Thirty-two children were born with an OA with distal TOF, while in eight cases an OA without fistula was present and a gastrostomy operation was performed at birth, while ‘delayed’ reconstructive oesophageal surgery was carried out using a different procedure [7]. All patients were given DFB during spontaneous breathing. The positive cases underwent examination by helical computed tomography with a non-ionic iodinated contrast medium. CT was performed under sedation supervised by an
12
V. Briganti et al. / European Journal of Cardio-thoracic Surgery 28 (2005) 11–15
anaesthesiologist. Initially, the children were observed for respiratory problems; in the last few years instead, we have routinely used tracheoscopical checks on all children submitted for neonatal surgery for OA. All children also underwent a radiological examination of the upper intestinal tract, using barium meal, and an Oesophagogastroscopy (EGDscopy) for a morphological evaluation and for a study of the Gastroesophageal Reflux Disease (GERD).
2.3. Oesophagogastroscopy Standard paediatric videoendoscopy equipment was used to study the oesophagus and Gastroesophageal junction. Mucosa lesions were scored according to the Savary–Miller classification [12,13], while scoring of cardias competence was achieved through the J-manoeuvre following the Hill classification [14]. 2.4. Radiology
2.1. Dynamic fiberoptic bronchoscopic For this examination we used standard fiberoptic bronchoscopy equipment, with a video monitor and a VHS or, more recently, a DVD recording system. The endoscopes used were 2.8 and 3.5 mm diameter, and examination was performed during spontaneous breathing prior to topical anaesthesia of the vocal cords and the trachea. The endoscopic examinations were always carried out by the same surgeon and subsequently analysed in detail with the help of the video recordings. In some cases, when a recurrent tracheo-oesophageal fistula was strongly suspected, an endoscopic examination by rigid instrumentation was also carried out, to perform a correct manipulation of the posterior tracheal wall in the search for fistulas. During endoscopic examination an evaluation was carried out of the morphology of the airways and the carina, of the mucous layer aspect, of the presence of masses (tumours, CE) and fistulas, of the presence of mediastinal compression and/or vascular pulsations, and the degree of dynamic obstruction of the airway during the breathing cycle. Using the Benjamin score [8,9], we considered all obstructions that exceeded 50% of the antero-posterior diameter of the trachea as pathological. Consequently, we defined obstructions below 80% occlusion as Partial TM, and obstructions over 90% as Severe TM. Particular attention was paid to the morphology and dynamics of the sopraglottic and glottic area, while for the evaluation of subglottic stenosis we used the Cotton score [10]. Intraoperative tracheobroncoscopy was used in all cases submitted for surgery so as to obtain a correct resolution of the obstructive problems [11]. 2.2. CT scan A CT scan examination was carried out using spiral CT in seven patients, and spiral multislice CT in 18 patients. All the examinations were performed with non-ionic iodinated contrast medium to the dose of 2 cm3/kg, in sedation. Technical parameters were optimised to obtain mutiplanar and three-dimensional images of good quality (thickness of 3 mm for spiral CT, 1.25 mm for multislice CT). Radiation dose was reduced on basis of ALARA criterium (0.5 mSv). Axial images were processed with reconstruction of multiplanar images (coronal and sagittal), three-dimensional angiography of aorta and its branches, virtual broncoscopy and virtual broncography, using Shaded Surface Display or Volume Rendering technique.
Standard barium esophagograms were always obtained under fluoroscopic control. The presence of hiatal hernias, spontaneous GER, strictures, mucosa changes, and oesophageal peristalsis were recorded.
3. Results Of the patients examined, 26 (60%) showed a malacia of the tracheobronchial tract, while 16 (40%) did not. Indeed, in four patients, one with TM and three without, a persistent tracheoesophageal fistula was found. An ‘active’ subglottic stenosis, 18 degree Cotton, secondary to GER and regressed after antireflux surgical treatment was present in two patients. DFB and CT scans allowed us to describe three morphological variations of thoracic TM: (a) Anterior TM (ATM) caused by compression of the Aorto-Innominate Complex (Mustard Complex); (b) Intrinsic TM (ITM), in which signs of vascular imprinting were not evident; (c) Complex Malacia (CM). Within this term we included both the malacias that involved the main-stem bronchus systems, with or without associated tracheomalacia, as well as those forms in which vascular compression was not due to the aorto-innominate complex. In these cases, a vascular map was produced by CT angiographic images (multiplanar and three-dimensional) for every patient using axial helical CT and a three-dimensional reconstruction so as to permit a comparison between the DBF and the virtual tracheobroncoscopy. We defined ATM as a malacia that primarily involved the anterior arch of the tracheal wall. This malformation is always above the carina and presents at DBF a typical endoscopic aspect with an eccentric triangular shape (Fig. 1) caused by the pulsating displacement–compression on the anterior wall of the trachea by the aortic arch and especially by the innominate artery. In the forms in which compression was applied primarily by the innominate vessel, mild compression of the anterior tracheal wall by rigid tracheoscopy enabled us to observe a reduction of the right radial pulse. A subsequent CT examination allowed us to analyse the true compressive role of the aortic arch and innominate artery as well as the length of the segment compressed by this artery (Fig. 2). ATM was present in 11 patients (46%). We defined ITM as an intrinsic weakness of the tracheal rings, with reduction of the antero-posterior diameter. Endoscopically, the thoracic trachea appeared flattened, with evident antero-posterior collapse on exhalation and characteristic prominence of the pars membranacea
V. Briganti et al. / European Journal of Cardio-thoracic Surgery 28 (2005) 11–15
13
Fig. 3. Intrinsic TM. A conventional tracheoscopic view of ITM with reduction of the antero-posterior diameter and dynamic bulging of the posterior tracheal wall.
Fig. 1. Anterior TM. (A) Virtual endoscopic image showing the pathognonomic eccentric triangular shape of the trachea. (B) Schematic outline showing localised bulging caused by aorto-innominate artery compression.
Fig. 2. Anterior TM. (A) 3D reconstruction images of the airway in a patient with ATM: 3D external rendering of the trachea shows a malacic segment caused by compression of the innominate artery (black arrow). The white arrow indicates a residual tracheoesophageal pouch. (B) Schematic outline showing antero-posterior relationships by compressed trachea and innominate artery.
(back wall bulging) during breathing (Fig. 3). In these cases the lesion was always above the level of the aortic arch [15] and could extend in length up to the narrow thoracic inlet. This form was present in eight (33%) patients. We defined CM as present in cases in which the malacic segment extended to one or both main bronchi, either alone or in association with TM, and cases in which different types of vascular airway compression were present. Each case was therefore evaluated individually. CM was present in five patients (21%). In one patient the malacia involved the left main bronchus which was compressed by the left pulmonary artery. The patient, who had shown a poor respiratory symptoms, improved spontaneously by the age of two. In another patient, instead, a long segmental malacia of the thoracic trachea and of the first tract of the right main bronchus was present. The antero-posterior pulsating obstruction was almost complete and due to an ectasia of the ascending aorta in this patient with Fallott’s tetralogy. Correction of the cardiac and airway defects allowed resolution of the clinical symptoms. Another two patients, instead, had a right aortic arch whose position created differing relationships with the short length airway segments. The first case, in which it had not been possible to perform respiratory weaning after corrective operation for OA, presented an anterior segmentary TM, caused by compression of the left carotid vessel. It came from the ascending aorta in right mediastinum and crossed the chest transversally before arriving at the superior thoracic inlet (Fig. 4). The left main bronchus was also obstructed by the fibrous residue of the Botallo ligament, while the left pulmonary artery completed the obstruction of the left bronchus. In this patient an intrapericardic– extrapericardic correction of the defect was carried out by sectioning the Botallo duct, suspending the pulmonary trunk by the intrapericardic route and suspending the left carotid artery, thoracic trachea by the extrapericardic route. All procedures were performed under intraoperative endoscopic control.
14
V. Briganti et al. / European Journal of Cardio-thoracic Surgery 28 (2005) 11–15
Fig. 4. A Complex Malacia by the right aortic arch. (A) A contrast-enhanced axial CT image showing the right aortic arch (white arrow). (B) A 2D sagittal reformation image showing reduction of the tracheal lumen by the left carotid artery (black arrow). (C) A 3D CT scan reconstruction of the aorta and epiaortic vessels. The right origin of left carotid artery (white arrow) is seen completely crossing the mediastinum from right to left before entering the neck.
The second case, with poor clinical symptoms, was treated in a conservative manner. This child was affected by two mild symmetrical compressions of the thoracic trachea on its lateral side, that CT examination revealed as two common carotids of orthotopic origin. There were no evident marks of occlusion of the principal bronchus in this patient (Fig. 5). One patient had an intrinsic TM associated with a severe left bronchial hypoplasia in association with a left pulmonary hypoplasia that had, nevertheless, allowed respiratory weaning after a neonatal intervention for OA. The patient, however, died after 27 days of life of severe respiratory complications probably caused by ab ingestis.
4. Discussion A TM is present in 11–33% of cases of OA. Bargy supports the hypothesis that there is a high risk of tracheomalacia in cases of OA where a high distal fistula is present [16], while a TM is very rare in OA without Fig. 6. Obstructive Upper Esophageal Pouch Syndrome. (A) A contrastenhanced axial CT image showing posterior bulging of the trachea (black arrow) caused by an hypotonic dilated upper oesophagus (white arrow). (B) In the same case: external 3D rendering of the oesophagus with anastomotic stenosis (white arrow). (C) An oesophagogram using standard barium meal: upper oesohageal stenosis (white arrow).
Fig. 5. An Asymptomatic Malacia caused by the right aortic arch. (A) A 3D CT scan reconstruction of the tracheal image. The white arrows show tracheal compression areas caused by the right and left common carotids. (B) A 3D reconstruction of the aorta and epiaortic vessels. The orthotopic origin may be seen of the right and left (white arrow) carotid arteries, which gently ‘thrash’ the thoracic trachea side by side before entering the neck. (C) A contrast-enhanced axial CT image showing the right aortic arch (black arrow).
TOF. In our experience, only one case in eight of OA sine fistula showed an intrinsic TM with a prevalence of posterior collapse of the pars membranacea. This was caused by an ectasia of the atonic oesophageal upper pouch which pulled back the posterior tracheal wall during feeding. We defined the condition as Obstructive Upper Oesophageal Pouch Syndrome (Fig. 6). In this particular case, the problem of the obstruction was resolved by several cycles of esophageal dilation and subsequent antireflux surgery [17]. In the last few years, several different hypotheses have been formulated which explain the pathogenesis of a TM in OA as either an anomaly in the separation groove between the oesophagus and the trachea, a tracheal compression resulting from a hypertrophied and dilated upper esophageal
V. Briganti et al. / European Journal of Cardio-thoracic Surgery 28 (2005) 11–15
pouch [18], and/or a lack of amniotic liquid inside the trachea causing a reduction in intrafetal tracheal stenting [19]. Gross invented the traditional form of treatment for a TM in the 1950s and this involves suspension of the aortic arch from the inside surface of the sternum (Gross Aortopexy) so that it maintains tracheal luminal patency [3]. Subsequent morphological studies, supported by endoscopical assessment and CT scans, revealed different morphotypes of OA associated malacia. In fact, in our experience, we think that it is necessary to bear in mind that a complex rearrangement of all the mediastinal structures always occurs in OA. It is not the case, in fact, that a high incidence of cardiac and vascular malformations are to be found. Our experience has allowed us to identify two principal TM morphotypes associated with OA: an anterior TM and an intrinsic TM, which represent almost 80% of cases. All the other cases that did not match these two principal forms (20% of cases), were grouped together and labelled as complex forms. These cases differed from each other from time to time, as our two cases of TM with right aortic arch displacement. Our analysis of the results, therefore, permits us to draw some conclusions: (a) OA exhibits a high number of morphological variations of the mediastinal structure, some of which are not clinically influential; (b) the use of routine airway assessment by endoscopic procedure is mandatory in all cases of OA before the fourth month of life, and in TM positive cases CT examination was carried out using axial helical computed tomography with a contrast medium; (c) correction of the TM cannot be always effected using a standard operation that does not include consideration of any multiform structural anomalies of the mediastinum, and for this reason we always recommend a surgically ‘tailored’ approach based on the endoscopic and radiology findings; (d) corrective surgery must be performed using one ‘window of access’ to the airway, differently from Gross’ traditional approach, and we think that the use of a low cervical incision with a split of the sternal manubrium, as recommended by some surgeons [20,21], meets this need better; (e) the importance of the use of intraoperative broncoscopy in the correction of a TM.
15
References [1] Spitz L, Kiely E, Brereton RJ. Esophageal atresia: five years experience with 148 cases. J Pediatr Surg 1987;22:103–8. [2] Slany E, Holzki J, Holschneider AM, Gharib M, Hugel W, Mennicken U. Flaccid trachea in tracheoesophageal malformations. Z Kinderchir 1990; 45:78–85. [3] Gross RE, Neuhauser EBD. Compression of the trachea by an anomalous innominate artery. Am J Dis Child 1948;75:570–4. [4] Triglia JM, Guys JM, Borrione CL. Tracheomalacia caused by arterial compression in esophageal atresia. Ann Otol Rhinol Laryngol 1994;103: 516–21. [5] Fletcher BD, Cohn RC. Tracheal compression and the innominate artery: ME evaluation in infants. Radiology 1989;170:103–7. [6] Kamata S, Usui N, Sawai T, Nose K, Kitayama Y, Okuyama H, Okada A. Pexis of great vessels for patients with tracheobronchomalacia in infancy. J Pediatr Surg 2000;35:454–7. [7] Puri P, Ninan GK, Blake NS, Fitzgerald RJ, Guiney EJ, O’Donnell B. Delayed primary anastomosis for esophageal atresia: 18 months’ to 11 years’ follow-up. J Pediatr Surg 1992;27:1127–30. [8] Benjamin B. Endoscopy in congenital tracheal anomalies. J Pediatr Surg 1980;15:164–71. [9] Campbell AH, Young IF. Tracheobronchial collapse: a variant of obstructive respiratory disease. Br J Dis Chest 1963;57:174–81. [10] Myer 3rd CM, O’Connor DM, Cotton RT. Proposed grading system for subglottic stenosis based on endotracheal tube sizes. Ann Otol Rhinol Laryngol 1994;103:319–23. [11] Abdel-Rahman U, Ahrens P, Fieguth HG, Kitz R, Heller K, Moritz A. Surgical treatment of tracheomalacia by bronchoscopic monitored aortopexy in infant and children. Ann Thorac Surg 2002;74:315–9. [12] Armstrong D, Blum AL, Savary M. Reflux disease and Barrett’s oesophagus. Endoscopy 1990;24:9–17. [13] Savary M, Miller G. L’oesophage. Manuel et atlas d’endoscopie. Solothurn: Gassmann; 1977. p. 80–92. [14] Hill LD, Snopkowski P. The Hill repair. In: Sabiston DC, Spencer FC, editors. Surgery of the chest, vol. 1. Philadelphia, PA: W.B. Saunders; 1992. p. 1039–52. [15] Kimura K, Soper RT, Kao SCS, Smith WL, Franken EA. Aortosternopexy for tracheomalacia following repair of esophageal atresia: evaluation by cine-CT and technical refinement. J Pediatr Surg 1990;25:769–72. [16] Bargy F, Manach Y, Wakim A. La tracheomalacie dans l’atresie de l’oesophage. Chir Pediatr 1984;25:261. [17] Briganti V, Oriolo L, Calisti A. Reflux strictures of the oesophagus in children: personal experience with preoperative dilatation followed by anterior fundoplication. Pediatr Surg Int 2003;19:544–7. [18] Davis MRQ, Cywes S. The flaccid trachea and tracheoesophageal congenital anomalies. J Pediatr Surg 1978;13:363. [19] Schwartz MZ, Filler RM. Tracheal compression as a cause of apnea following repair of tracheoesophageal fistula: treatment by aortopexy. J Pediatr Surg 1980;15:842. [20] Morabito A, MacKinnon E, Alizai N, Asero L, Bianchi A. The anterior mediastinal approach for management of tracheomalacia. J Pediatr Surg 2000;35:1456–8. [21] Vaishnav A, MacKinnon AE. New cervical approach for tracheopexy. Br J Surg 1986;73:441–2.
