EHD-04041; No of Pages 4 Early Human Development xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Early Human Development journal homepage: www.elsevier.com/locate/earlhumdev
Best practice guidelines
Oesophageal atresia and tracheo-oesophageal fistula Nicola Smith ⁎ Department of Paediatric Surgery, Cambridge University Hospitals NHS Foundation Trust, Addenbrooke's Hospital, Hills Road, Cambridge CB2 0QQ, United Kingdom
a r t i c l e
i n f o
Available online xxxx Keywords: Tracheo-oesophageal fistula Oesophageal atresia Oesophageal stricture Neonatal Paediatric
a b s t r a c t Oesophageal atresia with tracheo-oesophageal fistula is a relatively common congenital anomaly occurring in around 1:2500 births. The aetiology and embryology of the condition remain unclear, whilst associations with other significant anomalies are common. Studies in rodent models are contributing to our understanding of the condition. Advances in surgical care and neonatal management have improved survival considerably to around 90%. Long-gap and isolated oesophageal atresia present significant management challenges. Post-operative and long-term complications including oesophageal stricture, gastro-oesophageal reflux and respiratory compromise however remain relatively common and continue to pose a challenge for the ongoing management of patients. © 2014 Elsevier Ltd. All rights reserved.
Contents 1. 2. 3. 4.
Definition/incidence . . . . . . . . . . . Embryology . . . . . . . . . . . . . . . Diagnosis . . . . . . . . . . . . . . . . Associated anomalies & genetic associations 4.1. Prognosis . . . . . . . . . . . . . 4.2. Surgery . . . . . . . . . . . . . 4.3. Post-operative management . . . . 4.4. Complications . . . . . . . . . . 5. Growth and development . . . . . . . . 6. Variations/challenges . . . . . . . . . . . 7. Long-gap oesophageal atresia . . . . . . . 8. Thoracoscopic repair . . . . . . . . . . . 8.1. Areas for research . . . . . . . . . Conflict of interest statement . . . . . . . . . References . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
1. Definition/incidence Oesophageal atresia and tracheo-oesophageal fistula have an incidence of around 1:2500 live-births [1]. It may be divided anatomically into 5 types (see Fig. 1) with the most common being oesophageal atresia with a distal tracheo-oesophageal fistula — found in around 85% cases. The condition consists of a discontinuity or atresia of the oesophagus; with the majority of infants exhibiting a connection or
⁎ Tel.: +44 1223 256276. E-mail address: [email protected].
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
fistula between the oesophagus and trachea. The exceptions to this are children born with an isolated oesophageal atresia and those with an H-type tracheo-oesophageal fistula. The latter of these tends to present after the neonatal period, is challenging to diagnose and is treated by a different operation. Thus, its management will not be discussed further here. 2. Embryology During early development the airway develops as a diverticulum from the primitive foregut. Development starts as early as 4 weeks of gestation. Once the diverticulum is formed, it extends caudally into
http://dx.doi.org/10.1016/j.earlhumdev.2014.09.012 0378-3782/© 2014 Elsevier Ltd. All rights reserved.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
2
N. Smith / Early Human Development xxx (2014) xxx–xxx
Fig. 2. Chest radiograph of an infant with OA-TOF. The nasogastric tube can be seen terminating in the upper pouch. Sub-diaphragmatic air indicates the presence of a distal tracheo-oesophageal fistula.
Fig. 1. Line drawing showing the five anatomical variants of congenital oesophageal atresia/ tracheo-oesophageal fistula. 1a: oesophageal atresia with distal fistula (85% cases); 1b: oesophageal atresia without fistula (75); 1c: H-type TOF (4%); 1d: oesophageal atresia with proximal and distal fistula (3%); 1e: oesophageal atresia with proximal fistula (1%).
possible to allow the diagnosis to be made — without of course damaging the upper pouch. Surgical teaching is for a 10Fr nasogastric tube to be passed — which typically cannot be advanced beyond 10 cm. The initial film will allow assessment of the presence or absence of distal gas, which indicates whether a distal fistula is present. The absence of gas in sub-diaphragmatic bowel loops indicates that a distal fistula is not present — this raises concerns of a long oesophageal gap; and the possibility of technical difficulties at surgery. If there is ongoing doubt regarding the diagnosis, on rare occasions a contrast study may be helpful. However, extreme care should be taken as there is a high risk of aspiration of contrast during the study.
4. Associated anomalies & genetic associations the splanchnic mesoderm and divides into two tubules which will form the right and left main bronchi. Simultaneously the formation of the trachea-oesophageal groove begins to separate the two tubes in a rostro-caudal direction. This common origin provides one explanation for the embryology of tracheo-oesophageal fistulae — presumed to result from a failure of invagination of the lateral tracheo-oesophageal grooves. The teratogenic Adriamycin rodent model of OA-TOF has been very useful in helping to uncover some of the developmental features of the condition. However the embryology remains unclear.
3. Diagnosis In the vast majority of infants the diagnosis is made in the early postnatal period. Antenatally, polyhydramnios along with a small stomach has around a 55% predictive value for oesophageal atresia. In addition the presence of recognised associated anomalies (see below) may raise the possibility of the diagnosis, however it is not possible to definitively diagnose OA-TOF on an antenatal ultrasound scan. Diagnosis is therefore made when attempts are made to feed the infant. The baby may then be described as ‘bubbly’, be unable to tolerate feeds, may experience choking episodes or become cyanotic when attempting to feed. If the diagnosis is suspected, an NG tube should be passed. Failure to pass a tube into the stomach, along with a characteristic x-ray (Fig. 2) is diagnostic of OA-TOF. Caution should however be exercised to ensure the NG tube is of sufficient calibre to not simply coil in the back of the pharynx, and that it is pushed down as far as
OA-TOF may be associated with other anomalies, and these should be sought during the early assessment. Most common is the VACTERL association. This mnemonic indicates defects in Vertebrae, Anorectal malformations, Cardiac anomalies; Tracheo-Esophageal fistula; Renal abnormalities and Limb defects (most commonly radial hypoplasia or agenesis). Many other associations have been noted with varying degrees of frequency, including duodenal atresia, intestinal malrotation, and other skeletal anomalies. Other associations include the CHARGE syndrome and trisomy 18. To date no definitive genetic basis for OA-TOF in humans has been identified. Around 1% of cases show familial inheritance; the remainder are sporadic diagnoses [2]. Murine models including the Shh−/− knockout demonstrate a phenotype including OA-TOF and have been used to study some aspects of the condition. However, patients with abnormalities in this pathway, including mutations in the GLI family of transcription factors (downstream effectors of the SHH pathway) do not seem to exhibit OA-TOF as part of their phenotype. More recently microdeletions of the FOX gene cluster at 16q24.1 have been found in patients with wide-ranging abnormalities including respiratory and foregut malformations [2]. Further detailed investigations are needed to elucidate the exact nature of any genetic abnormality underlying OA-TOF in isolation. A full clinical examination including inspection of the perineum, spine and assessment of the limbs is essential. Babies diagnosed with OA-TOF should also undergo echocardiography prior to anaesthesia, with ultrasound of the renal system and spine being performed in the early neonatal period.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
N. Smith / Early Human Development xxx (2014) xxx–xxx Table 1 Table showing modified survivals based on the Spitz classification.
Birthweight N 1500 g without major cardiac anomaly Birthweight b 1500 g OR major cardiac anomaly Birthweight b 1500 g AND major cardiac anomaly
1994 survival
2006 survival
97% 59% 22%
98% 82% 50%
4.1. Prognosis Prior to the onset of surgical repair of OA-TOF in the 1940s and 50s the condition was universally fatal. Development and refinement of surgical technique along with advances in neonatal care have radically changed the outlook for affected babies and their families. Survival is now largely dependent on the presence or absence of other anomalies rather than the OA-TOF itself. In 1994 the revised Spitz classification for the outcome of babies with OA-TOF was published. The classification involved two determinants, birthweight greater or less than 1500 g and the presence or absence of significant cardiac anomalies (Table 1). The author has more recently revisited this data and found improved survival rates for all groups [3]. 4.2. Surgery Before starting the operation some surgeons will undertake a bronchoscopy, to visualise the fistula and allow the rare upper pouch fistula to be seen. The procedure is most commonly performed through a rightsided thoracotomy. The operation involves dividing the fistula connection from the bronchial tree and anastomosing this to the lower oesophageal pouch. Post-operatively most surgeons prefer to keep a trans-anastomotic nasogastric tube to allow for early feeding and leave a chest drain to manage potential oesophageal anastomotic leaks. 4.3. Post-operative management There is some variation regarding the early post-operative management of babies with OA-TOF. Approaches range from the truly ‘tubeless TOF’ in which the baby is extubated at the end of the procedure, to surgeons who request the baby remains paralysed and intubated for at least 48 h post-operatively. Much depends on the preference of the operating surgeon, although the perceived tension of the anastomosis at the time of surgery is also important. It is assumed that an anastomosis under a lot of tension will be less likely to disrupt if movements of the baby's neck are minimised and the baby is nursed with its neck flexed. Most babies will remain intubated and ventilated for the night following surgery, and then be allowed to gently wake up the following day. Many surgeons will allow NG feeds to commence around 48 h postoperatively. If these are tolerated and the baby is otherwise well oral feeds are usually started at around 5 days. If there is any suspicion of an anastomotic leak, if there were technical operative difficulties or sometimes by surgeon preference, a contrast swallow may be requested before oral feeds are started. Babies are often started on anti-reflux medication immediately after the operation. Gastro-oesophageal reflux is common following TOF-OA repair and may contribute to the formation of an oesophageal stricture.
3
baby; and the majority may be managed conservatively. However in a minority of cases, the baby is symptomatic, and some of these will require a second operation to allow surgical closure of the leak [4]. Tracheomalacia is seen in a high proportion of post-operative OA-TOF patients. In its mildest form this may result in the classic persistent harsh ‘TOF cough’ which is largely asymptomatic. However, more severe cases can lead to respiratory distress and at worst the need for prolonged ventilation. In the neonatal period symptoms are often most marked during feeding, making the differentiation from gastroesophageal reflux challenging. Symptoms tend to improve with age, however some infants may benefit from aortopexy to mitigate their symptoms. Later on, anastomotic stricture is relatively common, with around 50% of patients requiring oesophageal dilatation [5]. Prophylactic dilatation has previously been suggested but has not been found to be useful [6]. Most surgeons will undertake a watch and wait approach being guided by the onset of dysphagia symptoms in the child and findings from a contrast swallow. If a stricture is seen this is treated by balloon dilatation with radiographic guidance. Many strictures will need multiple dilatations before satisfactory resolution is seen. The presence of gastro-oesophageal reflux increases the risk of stricture formation; satisfactory resolution of a stricture will also be difficult to achieve in the presence of ongoing reflux. Oesophageal dysmotility is commonly seen in children OA-TOF, and seems to be multi-factoral in origin. Abnormal development of the vagus nerve and Auerbach's plexus; surgical mobilisation and trauma; gastro-oesophageal reflux and possible stricture formation may all contribute. Oesophageal manometry studies on OA-OF patients — even those with minimal symptoms show very disordered peristalsis so it certainly exists [7]. Parents and children often describe the child taking a mouthful of drink following each mouthful of food to help with swallowing. There is little that can be done to improve these symptoms but most children learn to live perfectly happily with them. Reported symptoms tend to decrease with age, although closer questioning of adults reveals a higher prevalence of symptoms including dysphagia and heartburn in OA-TOF patients than in the general population [7]. Recurrent chest infections are common, particularly in the first year of life [8]. The mechanism underlying this remains unclear although it is thought that micro-aspiration may contribute [10]. As well as recurrent infections many children exhibit symptoms of airway reactivity and require bronchodilator inhalers. Larger units often offer multi-disciplinary follow-up clinics involving respiratory paediatricians to optimise management of the children. The worst affected cases may benefit from prophylactic antibiotics particularly over the winter months. In the past skeletal deformities including winging of the scapula (caused by inadvertent division of the long thoracic nerve) have been described. Adoption of the serratus sparing approach to the thoracic cavity should largely prevent this complication. In children who require more than one thoracotomy the risk of skeletal deformity is clearly increased. As mentioned above patients often show symptoms of gastrooesophageal reflux. ‘Prophylactic’ medication with anti-acid therapy (ranitidine or proton pump inhibitors) is undertaken by many surgeons. If the child is asymptomatic these may be weaned off later in the first year of life. However, many children – particularly those with a longgap repair – will continue to show symptoms of gastro-oesophageal reflux and around 15–25% will go on to need a fundoplication to control those symptoms [5]. 5. Growth and development
4.4. Complications The most common early operative complication is a leak at the site of the anastomosis. This may be detected by swinging pyrexia, any frothy mucous in the chest drain if one is present — and definitively identified by contrast swallow. A leak may occur in around 15% of cases [1]. The majority of these are detected on contrast swallow in an asymptomatic
Some studies have shown that in early childhood some degree of failure to thrive is common [9]. OA-TOF babies tend to be smaller at the time of delivery, and this finding continues during early childhood. The majority of parents of affected babies report some difficulty with feeding, vomiting or at least slow feeding [9]. This along with later than average introduction of solids (due to a combination of dysphagia
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
4
N. Smith / Early Human Development xxx (2014) xxx–xxx
and some babies undergoing stricture dilatations) results in poor weight gain. However, the few long-term outcome studies which have looked at adults have shown that height and weight measurements have recovered.
gastric tube or pull-up are the most commonly undertaken procedures, whilst ileal interposition has also been described [12].
6. Variations/challenges
Another recent advancement has been in the thoracoscopic repair of OA-TOF. Although the technique presents challenges both technically for the surgeon and certainly from the anaesthetists' point of view, successful repairs have been described by several experienced minimally invasive surgeons. Recent outcome reports suggest that early repairs may have had an increased stricture rate, although with increasing experience outcomes and complications are comparable to those from open series [13]. It is however a considerable technical challenge and should only be attempted by surgeons with considerable expertise and experience.
The absence of a stomach bubble or distal bowel gas on initial x-ray suggests an isolated oesophageal atresia. This in turn should raise concerns about a long-gap between upper and lower pouches. A long-gap is variably defined but may be considered as a gap greater than two vertebral bodies or 4 cm in distance. This may mean it is impossible to bring the two ends together at operation — or at least the anastomosis is under significant tension. A high tension anastomosis is more likely to develop a leak in the short term and to develop a stricture in the longer term. 7. Long-gap oesophageal atresia Management of a long-gap oesophageal atresia is aimed at attempting to preserve the native oesophagus. Despite the problems mentioned above, the outcome from oesophageal replacement is almost invariably worse than when the patients' own oesophagus is saved. Where the problem is suspected, it may be prudent to avoid even attempting a primary anastomosis in the early days. In such patients a gastrostomy should be formed to allow enteral feeding. This in itself can be challenging as the lack of foetal swallowing results in the development of microgastria. Satisfactory positioning of a gastrostomy to allow subsequent formation of a gastric tube if required for oesophageal replacement can be extremely difficult. With a gastrostomy in situ a ‘gap assessment’ can be undertaken; by passing a radio-opaque catheter via the gastric opening into the lower oesophageal stump whilst simultaneously passing a similar catheter into the upper pouch. Radiographs can then allow an assessment of the distance between the two ends. If this is greater than two vertebral bodies or around 4 cm many surgeons elect to simply watch and wait. Over a period of around 6 weeks the upper pouch can grow considerably. Serial gap assessments with x-rays guide the timing of delayed primary anastomosis. In this way recurrent thoracotomies may be avoided, making the definitive surgery less technically demanding and minimising the risk of skeletal deformities. Of course not all long-gaps are predictable prior to surgery. Unexpected cases can be found at the time of operation — in which case any tracheo-oesophageal fistula is divided (to improve the baby's ventilation whist definitive surgery is awaited) and a decision made as to how to deal with the oesophageal ends. The upper pouch may either be left — and drained by a replogle tube until anastomosis is achieved or brought out as a cervical oesophagostomy. Techniques to encourage growth of the oesophagus have been described. One technique involves attaching multiple sutures to either end and passing them out through the thoracotomy wound. These ends are gradually pulled over a period of several weeks to bring the oesophageal ends closer together. Proponents of the technique say it improves the likelihood of achieving an anastomosis and hastens this possibility [11]. However, it is not universally supported. If it is not possible to join the ends of the native oesophagus, or if a significant anastomotic leak means the oesophagus has to be abandoned then oesophageal replacement will have to be undertaken. The outcomes from oesophageal replacement are rarely as good as those from patients where the native oesophagus has been retained. Several techniques to replace the oesophagus have been described, colonic interposition and
8. Thoracoscopic repair
8.1. Areas for research As outlined above the pathophysiology and embryology of OA-TOF remain poorly understood. Genetic associations warrant further investigation, as do the role of teratogens in the development of this condition. Almost universally, patients who have undergone OA-TOF repair will demonstrate significant oesophageal dysmotility, with a huge variety in the severity of symptoms. This is highly problematic to treat and causes some patients considerable morbidity well into adult life. Currently available pharmacological agents to improve oesophagogastric dysmotility are of very limited benefit and have worrying sideeffect profiles. Advances in this area would be very welcome. Conflict of interest statement I confirm that I have no conflict of interests relating to this article. References [1] Beasley S. In: Stringer Mark D, Oldham Keith T, Mouriquand Pierre DE, editors. Pediatric surgery and urology long-term outcomes. Cambridge University Press; 2006. [2] Shaw-Smith Charles. Genetic factors in esophageal atresia, tracheo-esophageal fistula and the VACTERL association: roles for FOXF1 and the 16q24.1 FOX transcription factor gene cluster, and review of the literature. Eur J Med Genet Jan 2010; 53(1–3):6–13. [3] Lopes PJ, Keys C, Pierro A, Drake DP, Kiely EN, Curry JI, et al. Oesophageal atresia: improved outcome in high risk groups? J Pediatr Surg 2006;41:331–4. [4] Spitz L. Lessons I, have learned in a 40-year experience. J Pediatr Surg 2006;41: 1635–40. [5] Tovar JA. Fragoso AC gastroesophageal reflux after repair of esophageal atresia. Eur J Pediatr Surg Jun 2013;23(3):175–81. [6] Koivusalo A, Pakarinen MP, Rintala RJ. Anastomotic dilatation after repair of esophageal atresia with distal fistula. Comparison of results after routine versus selective dilatation. Dis Esophagus 2009;22(2):190–4. [7] Sistonen SJ, Koivusalo A, Nieminen U, Lindahl H, Lohi J, Kero M, et al. Esophageal morbidity and function in adults with repaired esophageal atresia with tracheoesophageal fistula: a population-based long-term follow-up. Ann Surg Jun 2010;251(6):1167–73. [8] Fragoso AC, Tovar JA. The multifactorial origin of respiratory morbidity in patients surviving neonatal repair of esophageal atresia. Front Pediatr 2014;2(39):1–5. [9] Chetcuti PAJ, Puntis JWL. In: Stringer Mark D, Oldham Keith T, Mouriquand Pierre DE, editors. Pediatric surgery and urology long-term outcomes. Cambridge University Press; 2006. [10] Kovesi T, Rubin S. Long-term complications of congenital esophageal atresia and/or traceoesophageal fistula. Chest 2004;126:915–25. [11] Nasr A, Langer JC. Mechanical traction techniques for long-gap esophageal atresia: a critical appraisal. Eur J Pediatr Surg Jun 2013;23(3):191–7. [12] Gallo G, Zwaveling S, Groen H, et al. Long-gap esophageal atresia: a meta-analysis of jejunal interposition, colon interposition, and gastric pull-up. Eur J Pediatr Surg Dec 2012;22(6):420–5. [13] Huang J, Tao J, Chen K, et al. Thoracoscopic repair of oesophageal atresia: experience of 33 patients from two tertiary referral centres. J Pediatr Surg Dec 2012;47(12):2224–7.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
Contents lists available at ScienceDirect
Early Human Development journal homepage: www.elsevier.com/locate/earlhumdev
Best practice guidelines
Oesophageal atresia and tracheo-oesophageal fistula Nicola Smith ⁎ Department of Paediatric Surgery, Cambridge University Hospitals NHS Foundation Trust, Addenbrooke's Hospital, Hills Road, Cambridge CB2 0QQ, United Kingdom
a r t i c l e
i n f o
Available online xxxx Keywords: Tracheo-oesophageal fistula Oesophageal atresia Oesophageal stricture Neonatal Paediatric
a b s t r a c t Oesophageal atresia with tracheo-oesophageal fistula is a relatively common congenital anomaly occurring in around 1:2500 births. The aetiology and embryology of the condition remain unclear, whilst associations with other significant anomalies are common. Studies in rodent models are contributing to our understanding of the condition. Advances in surgical care and neonatal management have improved survival considerably to around 90%. Long-gap and isolated oesophageal atresia present significant management challenges. Post-operative and long-term complications including oesophageal stricture, gastro-oesophageal reflux and respiratory compromise however remain relatively common and continue to pose a challenge for the ongoing management of patients. © 2014 Elsevier Ltd. All rights reserved.
Contents 1. 2. 3. 4.
Definition/incidence . . . . . . . . . . . Embryology . . . . . . . . . . . . . . . Diagnosis . . . . . . . . . . . . . . . . Associated anomalies & genetic associations 4.1. Prognosis . . . . . . . . . . . . . 4.2. Surgery . . . . . . . . . . . . . 4.3. Post-operative management . . . . 4.4. Complications . . . . . . . . . . 5. Growth and development . . . . . . . . 6. Variations/challenges . . . . . . . . . . . 7. Long-gap oesophageal atresia . . . . . . . 8. Thoracoscopic repair . . . . . . . . . . . 8.1. Areas for research . . . . . . . . . Conflict of interest statement . . . . . . . . . References . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
1. Definition/incidence Oesophageal atresia and tracheo-oesophageal fistula have an incidence of around 1:2500 live-births [1]. It may be divided anatomically into 5 types (see Fig. 1) with the most common being oesophageal atresia with a distal tracheo-oesophageal fistula — found in around 85% cases. The condition consists of a discontinuity or atresia of the oesophagus; with the majority of infants exhibiting a connection or
⁎ Tel.: +44 1223 256276. E-mail address: [email protected].
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
. . . . . . . . . . . . . . .
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
fistula between the oesophagus and trachea. The exceptions to this are children born with an isolated oesophageal atresia and those with an H-type tracheo-oesophageal fistula. The latter of these tends to present after the neonatal period, is challenging to diagnose and is treated by a different operation. Thus, its management will not be discussed further here. 2. Embryology During early development the airway develops as a diverticulum from the primitive foregut. Development starts as early as 4 weeks of gestation. Once the diverticulum is formed, it extends caudally into
http://dx.doi.org/10.1016/j.earlhumdev.2014.09.012 0378-3782/© 2014 Elsevier Ltd. All rights reserved.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
2
N. Smith / Early Human Development xxx (2014) xxx–xxx
Fig. 2. Chest radiograph of an infant with OA-TOF. The nasogastric tube can be seen terminating in the upper pouch. Sub-diaphragmatic air indicates the presence of a distal tracheo-oesophageal fistula.
Fig. 1. Line drawing showing the five anatomical variants of congenital oesophageal atresia/ tracheo-oesophageal fistula. 1a: oesophageal atresia with distal fistula (85% cases); 1b: oesophageal atresia without fistula (75); 1c: H-type TOF (4%); 1d: oesophageal atresia with proximal and distal fistula (3%); 1e: oesophageal atresia with proximal fistula (1%).
possible to allow the diagnosis to be made — without of course damaging the upper pouch. Surgical teaching is for a 10Fr nasogastric tube to be passed — which typically cannot be advanced beyond 10 cm. The initial film will allow assessment of the presence or absence of distal gas, which indicates whether a distal fistula is present. The absence of gas in sub-diaphragmatic bowel loops indicates that a distal fistula is not present — this raises concerns of a long oesophageal gap; and the possibility of technical difficulties at surgery. If there is ongoing doubt regarding the diagnosis, on rare occasions a contrast study may be helpful. However, extreme care should be taken as there is a high risk of aspiration of contrast during the study.
4. Associated anomalies & genetic associations the splanchnic mesoderm and divides into two tubules which will form the right and left main bronchi. Simultaneously the formation of the trachea-oesophageal groove begins to separate the two tubes in a rostro-caudal direction. This common origin provides one explanation for the embryology of tracheo-oesophageal fistulae — presumed to result from a failure of invagination of the lateral tracheo-oesophageal grooves. The teratogenic Adriamycin rodent model of OA-TOF has been very useful in helping to uncover some of the developmental features of the condition. However the embryology remains unclear.
3. Diagnosis In the vast majority of infants the diagnosis is made in the early postnatal period. Antenatally, polyhydramnios along with a small stomach has around a 55% predictive value for oesophageal atresia. In addition the presence of recognised associated anomalies (see below) may raise the possibility of the diagnosis, however it is not possible to definitively diagnose OA-TOF on an antenatal ultrasound scan. Diagnosis is therefore made when attempts are made to feed the infant. The baby may then be described as ‘bubbly’, be unable to tolerate feeds, may experience choking episodes or become cyanotic when attempting to feed. If the diagnosis is suspected, an NG tube should be passed. Failure to pass a tube into the stomach, along with a characteristic x-ray (Fig. 2) is diagnostic of OA-TOF. Caution should however be exercised to ensure the NG tube is of sufficient calibre to not simply coil in the back of the pharynx, and that it is pushed down as far as
OA-TOF may be associated with other anomalies, and these should be sought during the early assessment. Most common is the VACTERL association. This mnemonic indicates defects in Vertebrae, Anorectal malformations, Cardiac anomalies; Tracheo-Esophageal fistula; Renal abnormalities and Limb defects (most commonly radial hypoplasia or agenesis). Many other associations have been noted with varying degrees of frequency, including duodenal atresia, intestinal malrotation, and other skeletal anomalies. Other associations include the CHARGE syndrome and trisomy 18. To date no definitive genetic basis for OA-TOF in humans has been identified. Around 1% of cases show familial inheritance; the remainder are sporadic diagnoses [2]. Murine models including the Shh−/− knockout demonstrate a phenotype including OA-TOF and have been used to study some aspects of the condition. However, patients with abnormalities in this pathway, including mutations in the GLI family of transcription factors (downstream effectors of the SHH pathway) do not seem to exhibit OA-TOF as part of their phenotype. More recently microdeletions of the FOX gene cluster at 16q24.1 have been found in patients with wide-ranging abnormalities including respiratory and foregut malformations [2]. Further detailed investigations are needed to elucidate the exact nature of any genetic abnormality underlying OA-TOF in isolation. A full clinical examination including inspection of the perineum, spine and assessment of the limbs is essential. Babies diagnosed with OA-TOF should also undergo echocardiography prior to anaesthesia, with ultrasound of the renal system and spine being performed in the early neonatal period.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
N. Smith / Early Human Development xxx (2014) xxx–xxx Table 1 Table showing modified survivals based on the Spitz classification.
Birthweight N 1500 g without major cardiac anomaly Birthweight b 1500 g OR major cardiac anomaly Birthweight b 1500 g AND major cardiac anomaly
1994 survival
2006 survival
97% 59% 22%
98% 82% 50%
4.1. Prognosis Prior to the onset of surgical repair of OA-TOF in the 1940s and 50s the condition was universally fatal. Development and refinement of surgical technique along with advances in neonatal care have radically changed the outlook for affected babies and their families. Survival is now largely dependent on the presence or absence of other anomalies rather than the OA-TOF itself. In 1994 the revised Spitz classification for the outcome of babies with OA-TOF was published. The classification involved two determinants, birthweight greater or less than 1500 g and the presence or absence of significant cardiac anomalies (Table 1). The author has more recently revisited this data and found improved survival rates for all groups [3]. 4.2. Surgery Before starting the operation some surgeons will undertake a bronchoscopy, to visualise the fistula and allow the rare upper pouch fistula to be seen. The procedure is most commonly performed through a rightsided thoracotomy. The operation involves dividing the fistula connection from the bronchial tree and anastomosing this to the lower oesophageal pouch. Post-operatively most surgeons prefer to keep a trans-anastomotic nasogastric tube to allow for early feeding and leave a chest drain to manage potential oesophageal anastomotic leaks. 4.3. Post-operative management There is some variation regarding the early post-operative management of babies with OA-TOF. Approaches range from the truly ‘tubeless TOF’ in which the baby is extubated at the end of the procedure, to surgeons who request the baby remains paralysed and intubated for at least 48 h post-operatively. Much depends on the preference of the operating surgeon, although the perceived tension of the anastomosis at the time of surgery is also important. It is assumed that an anastomosis under a lot of tension will be less likely to disrupt if movements of the baby's neck are minimised and the baby is nursed with its neck flexed. Most babies will remain intubated and ventilated for the night following surgery, and then be allowed to gently wake up the following day. Many surgeons will allow NG feeds to commence around 48 h postoperatively. If these are tolerated and the baby is otherwise well oral feeds are usually started at around 5 days. If there is any suspicion of an anastomotic leak, if there were technical operative difficulties or sometimes by surgeon preference, a contrast swallow may be requested before oral feeds are started. Babies are often started on anti-reflux medication immediately after the operation. Gastro-oesophageal reflux is common following TOF-OA repair and may contribute to the formation of an oesophageal stricture.
3
baby; and the majority may be managed conservatively. However in a minority of cases, the baby is symptomatic, and some of these will require a second operation to allow surgical closure of the leak [4]. Tracheomalacia is seen in a high proportion of post-operative OA-TOF patients. In its mildest form this may result in the classic persistent harsh ‘TOF cough’ which is largely asymptomatic. However, more severe cases can lead to respiratory distress and at worst the need for prolonged ventilation. In the neonatal period symptoms are often most marked during feeding, making the differentiation from gastroesophageal reflux challenging. Symptoms tend to improve with age, however some infants may benefit from aortopexy to mitigate their symptoms. Later on, anastomotic stricture is relatively common, with around 50% of patients requiring oesophageal dilatation [5]. Prophylactic dilatation has previously been suggested but has not been found to be useful [6]. Most surgeons will undertake a watch and wait approach being guided by the onset of dysphagia symptoms in the child and findings from a contrast swallow. If a stricture is seen this is treated by balloon dilatation with radiographic guidance. Many strictures will need multiple dilatations before satisfactory resolution is seen. The presence of gastro-oesophageal reflux increases the risk of stricture formation; satisfactory resolution of a stricture will also be difficult to achieve in the presence of ongoing reflux. Oesophageal dysmotility is commonly seen in children OA-TOF, and seems to be multi-factoral in origin. Abnormal development of the vagus nerve and Auerbach's plexus; surgical mobilisation and trauma; gastro-oesophageal reflux and possible stricture formation may all contribute. Oesophageal manometry studies on OA-OF patients — even those with minimal symptoms show very disordered peristalsis so it certainly exists [7]. Parents and children often describe the child taking a mouthful of drink following each mouthful of food to help with swallowing. There is little that can be done to improve these symptoms but most children learn to live perfectly happily with them. Reported symptoms tend to decrease with age, although closer questioning of adults reveals a higher prevalence of symptoms including dysphagia and heartburn in OA-TOF patients than in the general population [7]. Recurrent chest infections are common, particularly in the first year of life [8]. The mechanism underlying this remains unclear although it is thought that micro-aspiration may contribute [10]. As well as recurrent infections many children exhibit symptoms of airway reactivity and require bronchodilator inhalers. Larger units often offer multi-disciplinary follow-up clinics involving respiratory paediatricians to optimise management of the children. The worst affected cases may benefit from prophylactic antibiotics particularly over the winter months. In the past skeletal deformities including winging of the scapula (caused by inadvertent division of the long thoracic nerve) have been described. Adoption of the serratus sparing approach to the thoracic cavity should largely prevent this complication. In children who require more than one thoracotomy the risk of skeletal deformity is clearly increased. As mentioned above patients often show symptoms of gastrooesophageal reflux. ‘Prophylactic’ medication with anti-acid therapy (ranitidine or proton pump inhibitors) is undertaken by many surgeons. If the child is asymptomatic these may be weaned off later in the first year of life. However, many children – particularly those with a longgap repair – will continue to show symptoms of gastro-oesophageal reflux and around 15–25% will go on to need a fundoplication to control those symptoms [5]. 5. Growth and development
4.4. Complications The most common early operative complication is a leak at the site of the anastomosis. This may be detected by swinging pyrexia, any frothy mucous in the chest drain if one is present — and definitively identified by contrast swallow. A leak may occur in around 15% of cases [1]. The majority of these are detected on contrast swallow in an asymptomatic
Some studies have shown that in early childhood some degree of failure to thrive is common [9]. OA-TOF babies tend to be smaller at the time of delivery, and this finding continues during early childhood. The majority of parents of affected babies report some difficulty with feeding, vomiting or at least slow feeding [9]. This along with later than average introduction of solids (due to a combination of dysphagia
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
4
N. Smith / Early Human Development xxx (2014) xxx–xxx
and some babies undergoing stricture dilatations) results in poor weight gain. However, the few long-term outcome studies which have looked at adults have shown that height and weight measurements have recovered.
gastric tube or pull-up are the most commonly undertaken procedures, whilst ileal interposition has also been described [12].
6. Variations/challenges
Another recent advancement has been in the thoracoscopic repair of OA-TOF. Although the technique presents challenges both technically for the surgeon and certainly from the anaesthetists' point of view, successful repairs have been described by several experienced minimally invasive surgeons. Recent outcome reports suggest that early repairs may have had an increased stricture rate, although with increasing experience outcomes and complications are comparable to those from open series [13]. It is however a considerable technical challenge and should only be attempted by surgeons with considerable expertise and experience.
The absence of a stomach bubble or distal bowel gas on initial x-ray suggests an isolated oesophageal atresia. This in turn should raise concerns about a long-gap between upper and lower pouches. A long-gap is variably defined but may be considered as a gap greater than two vertebral bodies or 4 cm in distance. This may mean it is impossible to bring the two ends together at operation — or at least the anastomosis is under significant tension. A high tension anastomosis is more likely to develop a leak in the short term and to develop a stricture in the longer term. 7. Long-gap oesophageal atresia Management of a long-gap oesophageal atresia is aimed at attempting to preserve the native oesophagus. Despite the problems mentioned above, the outcome from oesophageal replacement is almost invariably worse than when the patients' own oesophagus is saved. Where the problem is suspected, it may be prudent to avoid even attempting a primary anastomosis in the early days. In such patients a gastrostomy should be formed to allow enteral feeding. This in itself can be challenging as the lack of foetal swallowing results in the development of microgastria. Satisfactory positioning of a gastrostomy to allow subsequent formation of a gastric tube if required for oesophageal replacement can be extremely difficult. With a gastrostomy in situ a ‘gap assessment’ can be undertaken; by passing a radio-opaque catheter via the gastric opening into the lower oesophageal stump whilst simultaneously passing a similar catheter into the upper pouch. Radiographs can then allow an assessment of the distance between the two ends. If this is greater than two vertebral bodies or around 4 cm many surgeons elect to simply watch and wait. Over a period of around 6 weeks the upper pouch can grow considerably. Serial gap assessments with x-rays guide the timing of delayed primary anastomosis. In this way recurrent thoracotomies may be avoided, making the definitive surgery less technically demanding and minimising the risk of skeletal deformities. Of course not all long-gaps are predictable prior to surgery. Unexpected cases can be found at the time of operation — in which case any tracheo-oesophageal fistula is divided (to improve the baby's ventilation whist definitive surgery is awaited) and a decision made as to how to deal with the oesophageal ends. The upper pouch may either be left — and drained by a replogle tube until anastomosis is achieved or brought out as a cervical oesophagostomy. Techniques to encourage growth of the oesophagus have been described. One technique involves attaching multiple sutures to either end and passing them out through the thoracotomy wound. These ends are gradually pulled over a period of several weeks to bring the oesophageal ends closer together. Proponents of the technique say it improves the likelihood of achieving an anastomosis and hastens this possibility [11]. However, it is not universally supported. If it is not possible to join the ends of the native oesophagus, or if a significant anastomotic leak means the oesophagus has to be abandoned then oesophageal replacement will have to be undertaken. The outcomes from oesophageal replacement are rarely as good as those from patients where the native oesophagus has been retained. Several techniques to replace the oesophagus have been described, colonic interposition and
8. Thoracoscopic repair
8.1. Areas for research As outlined above the pathophysiology and embryology of OA-TOF remain poorly understood. Genetic associations warrant further investigation, as do the role of teratogens in the development of this condition. Almost universally, patients who have undergone OA-TOF repair will demonstrate significant oesophageal dysmotility, with a huge variety in the severity of symptoms. This is highly problematic to treat and causes some patients considerable morbidity well into adult life. Currently available pharmacological agents to improve oesophagogastric dysmotility are of very limited benefit and have worrying sideeffect profiles. Advances in this area would be very welcome. Conflict of interest statement I confirm that I have no conflict of interests relating to this article. References [1] Beasley S. In: Stringer Mark D, Oldham Keith T, Mouriquand Pierre DE, editors. Pediatric surgery and urology long-term outcomes. Cambridge University Press; 2006. [2] Shaw-Smith Charles. Genetic factors in esophageal atresia, tracheo-esophageal fistula and the VACTERL association: roles for FOXF1 and the 16q24.1 FOX transcription factor gene cluster, and review of the literature. Eur J Med Genet Jan 2010; 53(1–3):6–13. [3] Lopes PJ, Keys C, Pierro A, Drake DP, Kiely EN, Curry JI, et al. Oesophageal atresia: improved outcome in high risk groups? J Pediatr Surg 2006;41:331–4. [4] Spitz L. Lessons I, have learned in a 40-year experience. J Pediatr Surg 2006;41: 1635–40. [5] Tovar JA. Fragoso AC gastroesophageal reflux after repair of esophageal atresia. Eur J Pediatr Surg Jun 2013;23(3):175–81. [6] Koivusalo A, Pakarinen MP, Rintala RJ. Anastomotic dilatation after repair of esophageal atresia with distal fistula. Comparison of results after routine versus selective dilatation. Dis Esophagus 2009;22(2):190–4. [7] Sistonen SJ, Koivusalo A, Nieminen U, Lindahl H, Lohi J, Kero M, et al. Esophageal morbidity and function in adults with repaired esophageal atresia with tracheoesophageal fistula: a population-based long-term follow-up. Ann Surg Jun 2010;251(6):1167–73. [8] Fragoso AC, Tovar JA. The multifactorial origin of respiratory morbidity in patients surviving neonatal repair of esophageal atresia. Front Pediatr 2014;2(39):1–5. [9] Chetcuti PAJ, Puntis JWL. In: Stringer Mark D, Oldham Keith T, Mouriquand Pierre DE, editors. Pediatric surgery and urology long-term outcomes. Cambridge University Press; 2006. [10] Kovesi T, Rubin S. Long-term complications of congenital esophageal atresia and/or traceoesophageal fistula. Chest 2004;126:915–25. [11] Nasr A, Langer JC. Mechanical traction techniques for long-gap esophageal atresia: a critical appraisal. Eur J Pediatr Surg Jun 2013;23(3):191–7. [12] Gallo G, Zwaveling S, Groen H, et al. Long-gap esophageal atresia: a meta-analysis of jejunal interposition, colon interposition, and gastric pull-up. Eur J Pediatr Surg Dec 2012;22(6):420–5. [13] Huang J, Tao J, Chen K, et al. Thoracoscopic repair of oesophageal atresia: experience of 33 patients from two tertiary referral centres. J Pediatr Surg Dec 2012;47(12):2224–7.
Please cite this article as: Smith N, Oesophageal atresia and tracheo-oesophageal fistula, Early Hum Dev (2014), http://dx.doi.org/10.1016/ j.earlhumdev.2014.09.012
