- Email: [email protected]
A standardised investigative strategy prior to revisional oesophageal surgery in children: High incidence of unexpected findings
Journal of Pediatric Surgery (2013) 48, 2241–2246
www.elsevier.com/locate/jpedsurg
A standardised investigative strategy prior to revisional oesophageal surgery in children: High incidence of unexpected findings Nigel J. Hall a,b,⁎, Michelle Wyatt c , Joe I. Curry b , Edward M. Kiely b a
Surgery Unit, UCL Institute of Child Health, London, UK Department of General Surgery, Great Ormond Street Hospital NHS Foundation Trust, London, UK c Department of Otorhinolaryngology, Great Ormond Street Hospital NHS Foundation Trust, London, UK b
Received 17 January 2013; revised 24 February 2013; accepted 4 March 2013
Key words: Oesophageal atresia; Revisional surgery; Tracheomalacia; Airway reconstruction; Oesophageal replacement
Abstract Background/Purpose: Revisional oesophageal reconstructive surgery carries uncommon and unusual risks related to previous surgery. To provide maximum anatomical detail and facilitate successful outcome, we report a standardised pre-operative investigative strategy for all such patients. Methods: Prospective 8-month cohort study following the introduction of this strategy. All patients underwent high resolution thoracic contrast CT scan and micro-laryngo-bronchoscopy by a paediatric ENT surgeon in addition to upper gastrointestinal contrast study, oesophagoscopy, and echocardiogram. Results: Seven children (median age 5.6 months [range 2.2–60]) completed the pathway. Four were referred with recurrence of a previously divided tracheo-oesophageal fistula (3 congenital, 1 acquired) and 3 (all with oesophagostomy) for oesophageal replacement for congenital isolated oesophageal atresia (OA, n = 1) and failed repair of OA with distal TOF with wide gap (n = 2). Overall, unanticipated findings were demonstrated in 6/7 children and comprised severe tracheomalacia and right main bronchus stenosis requiring aortopexy (n = 1), vocal cord palsy (n = 2), extensive mediastinal rotation (n = 1), proximal tracheal diverticulum (n = 1), severe subglottic stenosis requiring airway reconstruction (n = 1), proximal tracheal diverticulum (n = 1), right sided aortic arch (n = 1) and left sided aortic arch (previously reported to be right sided, n = 1). Conclusions: This standardised approach for this complex group of patients reveals a high incidence of unexpected anatomical and functional anomalies with significant surgical and possible medico-legal implications. We recommend these investigations during the pre-operative work-up prior to all revisional oesophageal surgery. © 2013 Elsevier Inc. All rights reserved.
⁎ Corresponding author. Surgery Unit UCL Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. Tel.: +44 020 7905 2733. E-mail address: [email protected] (N.J. Hall). 0022-3468/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.03.051
When compared with primary oesophageal surgery, revision surgery is technically more difficult and carries higher operative risks. These are most commonly the result of anatomical distortion from operative scarring and obliteration of normal tissue planes and appearance. Children requiring
2242 revisional surgery following failed or complicated repair of congenital or acquired oesophageal structural abnormalities are often referred to tertiary referral centres. As a result, uncertainty is introduced which may increase the risk of subsequent surgical complications. The surgeon is reliant on second-hand information regarding the original anatomy, and the precise nature of any previous surgery may be unclear. We have recently encountered a number of patients referred for revisional oesophageal surgery for a variety of indications. As a result of adverse outcome in 2 of these patients we introduced a standardised investigative pathway within our institution. The aim was to improve our knowledge of the preoperative anatomy and function in these patients in order to reduce operative misadventure and post-operative complications. Herein we describe our investigative pathway and our initial findings following its implementation.
1. Methods Our current investigative pathway has 3 stages. Firstly we perform whatever investigations are clinically indicated to confirm the nature of the abnormality for which the patient has been referred. On occasion, clinical examination and review of the radiological investigations that arrive with the child are all that is necessary. However most will need further targeted radiological imaging such as a prone tube oesophagogram in the case of suspected recurrent tracheooesophageal fistula. All children undergo rigid or flexible oesophagoscopy and echocardiogram. We then perform a high resolution computed tomography (CT) scan of the chest with intravenous contrast under general anaesthesia if necessary. The aim of this CT scan is to provide anatomical detail of the thoracic vascular anatomy, detail of the oesophageal and airway anatomy and to identify the presence of pulmonary disease. Finally an endoscopic examination of the airway is performed by an experienced paediatric otorhinolaryngologist with the general paediatric surgical team in attendance. The specific purpose of this examination is to identify structural or dynamic airway abnormalities including significant tracheomalacia, the presence of a tracheo-oesophageal fistula and its anatomical location and to assess vocal cord movement. We collected data prospectively over the 8 month period up to December 2011 and report all patients referred for revisional oesophageal surgery during this time period. We firstly report brief details of 2 patients on whom we operated prior to this and who had adverse outcomes. Our current investigative pathway was devised following the treatment of these initial 2 patients.
2. Results The first child was referred at the age of 6 months. At the age of 1½ months he had suffered an oesophageal injury due
N.J. Hall et al. to impaction of a battery in the upper oesophagus. The battery was retrieved from the oesophagus at an emergency open operation and a cervical oesophagostomy fashioned because of damage to the oesophageal wall. A gastrostomy was created to allow feeding. The child was referred to our centre for oesophageal reconstruction. Following assessment of oesophageal length, oesophageal reconstruction was performed. Post-operatively the child was noted to have a hoarse voice and subsequent laryngoscopy revealed evidence of left recurrent laryngeal nerve (RLN) palsy. It was unclear if this had been present prior to surgery. The hoarseness settled. A second child was referred for oesophageal replacement surgery at the age of 18 months having been born preterm at 34 weeks gestation with oesophageal atresia and distal tracheo-oesophageal fistula, an anorectal malformation and VACTERL association. Following a complicated operative course, a gastrostomy and cervical oesophagostomy were in situ at the time of referral. In addition there had been a previous episode of cardiac arrest requiring intensive care treatment and an episode of sepsis-related acute renal failure requiring haemofiltration. A gastric transposition was successfully performed. Twenty four hours following surgery signs of superior vena cava (SVC) obstruction developed with acute cardiorespiratory compromise. Echocardiography revealed a markedly narrowed SVC with thrombus and no flow. In view of further deterioration an attempt was made to commence extra-corporeal life support but it was not possible to achieve venous cannulation due to stenosis in multiple vessels. The child died. Following the treatment of these two children we derived a standardised investigative pathway for all children referred for revisional oesophageal surgery. We have evaluated 7 children in this manner. Demographic and clinical details of these children are shown in Table 1. The combination of contrast CT scan and endoscopic evaluation or the airway has revealed a high incidence of abnormalities which were not suspected clinically but which have had clinical, surgical and in some cases potentially medico-legal implications. The abnormalities detected and subsequent clinical course of each child are detailed in Table 1. All children were orally fed following surgery. Overall unexpected findings were detected in 6 of the 7 patients. The most significant airway anomaly detected was critical subglottic stenosis in child C. The airway would accept only a size 2.5 endotracheal tube at the age of 6 months. He required larnygotracheal reconstruction prior to oesophageal replacement surgery as it was deemed unsafe to proceed with major oesophageal surgery requiring a period of postoperative ventilation with a critically narrow airway. Examples of the other abnormalities detected are shown in the Figures. Fig. 1 shows pulmonary disease secondary to chronic aspiration alongside significant rotation of the mediastinal anatomy such that the trachea lies lateral and to the right side of the oesophagus in child A who had previously undergone division of a congenital trachea-oesophageal fistula
Strategy prior to revisional oesophageal surgery Table 1
2243
Patients, investigation findings and outcome.
Patient Age at Initial referral pathology (m)
Previous oesophageal surgery & investigation findings
Reason for referral
Bronchoscopy CT chest findings findings
A
2
OA with distal Primary repair OA/TOF TOF, prematurity (28 weeks)
B
10
C
3
OA with distal Primary repair OA/TOF; TOF; tracheomalacia recurrent fistula repair × 2; aortopexy OA with distal Thoracotomy, Oesophageal Severe subTOF cervical reconstruction glottic stenosis oesophagostomy & gastrostomy
D
21
E
5
F
3
G
60
OA with distal TOF, tracheomalacia Isolated OA, tetralogy of Fallot-repaired
Isolated OA
Primary repair OA/TOF; aortopexy Cervical oesophagostomy & gastrostomy; left sided aortic arch
Thoracotomy, cervical oesophagostomy & gastrostomy; right sided arch Acquired TOF Cervical repair (battery of acquired TOF ingestion)
Repair of Left vocal cord Rotation of recurrent TOF paralysis mediastinum to the right; trachea to right of oesophagus Right lower lobe consolidation Inadequate Repair of Significant recurrent TOF tracheomalacia aortopexy; right main bronchus (ventilator stenosis dependent)
Surgery performed
Outcome
Repair Good recurrent TOF
Trans-sternal Good aortopexy; repair recurrent TOF Good Laryngotracheal reconstruction; Laparoscopic gastric transposition Repair Good recurrent TOF
Repair of recurrent TOF Oesophageal Blind ending reconstruction pouch 3 cm below vocal cords
Laparoscopic Right aortic arch; double superior vena gastric cava; tracheal transposition diverticulum
Oesophageal reconstruction
Left sided aortic arch Laparoscopic gastric transposition
Repair of Left vocal cord recurrent TOF paralysis
Temporary right Horner's syndrome (resolved after 2 months) Good
Repair Good recurrent TOF
OA, oesophageal atresia; TOF, trachea-oesophageal fistula.
and repair of oesophageal atresia. Fig. 2 shows a wide distance between the aortic arch and the sternum in child B despite a previous aortopexy for critical tracheomalacia in addition to pulmonary disease due to chronic aspiration. The child remained ventilator dependent. A repeat aortopexy was performed via a trans-sternal route to allow the child to breathe spontaneously prior to surgical correction of the rerecurrent trachea-oesophageal fistula. Fig. 3A shows the endoscopic appearance of a blind ending upper tracheal diverticulum in child E who had been born with pure oesophageal atresia and managed with a cervical oesophagostomy and gastrostomy. This is depicted on 3 dimensional airway reconstruction images from the CT scan in Fig. 3B. It was not possible to achieve delayed anastomosis of the
oesophagus so a laparoscopic gastric transposition was performed. Prior knowledge of the presence of the upper tracheal diverticulum was of surgical relevance when performing the dissection of the cervical oesophagostomy and cervical oesophagogastric anastomosis.
3. Discussion In this short series of children referred for revisional oesophageal surgery, we have implemented a standardised investigative pathway and been impressed by the high incidence of clinically unexpected findings. We arrived at this investigative regimen following adverse outcomes in 2
2244
N.J. Hall et al.
Fig. 1 Thoracic CT scan of patient A. T, trachea; O, oesophagus containing nasogastric tube. There is obvious mediastinal rotation such that the trachea lies to the right of the oesophagus, and pulmonary consolidation consistent with recurrent aspiration.
earlier and similar patients; our practice was influenced as a direct consequence of these outcomes. While we cannot say with certainty that the ‘incidental findings’ we have unearthed have impacted directly on the clinical outcome of the subsequent 7 children, we believe that there are a number of benefits to the knowledge they have provided. We do believe that had we performed these investigations in the initial 2 patients we would have been in a better position to address the clinical scenarios and significant uncertainties that arose during their course of treatment. Firstly, in some cases, we have been able to explain previous treatment failure and guide subsequent treatment (for example in child B who had undergone an unsuccessful aortopexy prior to referral to us). Secondly, in child C we believe it would have been unsafe to proceed with major oesophageal surgery requiring post-operative paralysis and ventilation (as per our current protocol following gastric transposition) in the presence of the critically narrow subglottic stenosis. Third, we believe that knowledge of
Fig. 2 CT scan of patient B. The line indicates the distance between the aorta and posterior aspect of the sternum in a child who has previously had an aortopexy. Pulmonary consolidation is also present.
Fig. 3 Bronchoscopy findings (A) and rendered CT scan of the airway (B) from patient E. A catheter has been passed into a blindending upper tracheal diverticulum. The trachea is seen to the left (A). The diverticulum can be seen arising from the right side of the upper trachea (B).
the anatomy that has been achieved due to the CT scan has allowed us to plan surgery more effectively, interpret our surgical findings with a greater degree of certainty and alter our surgical approach where indicated. The mediastinal rotation detected preoperatively in child A meant that intraoperative findings could be predicted more accurately and surgical approach adjusted accordingly with the potential to reduce the risk of intra-operative complications. Although it is difficult to prove, the ‘route map’ provided may have helped to prevent inadvertent intraoperative perforation of either trachea or oesophagus. Knowledge of the presence of the upper tracheal diverticulum in child E was certainly extremely useful when it was encountered in the neck during dissection of the cervical oesophagostomy prior to cervical oesophagogastric anastomosis. As an aside we believe that this upper tracheal diverticulum most likely results from either a ‘missed’ upper pouch tracheooesophageal fistula or a fistula that was occluded at one
Strategy prior to revisional oesophageal surgery end in this child who did not have endoscopic evaluation of his airway prior to formation of the cervical oesophagostomy in the referring centre. Finally, some of our findings may have medico-legal implications. Two children in this series were found to have a unilateral vocal cord palsy despite no clinical suspicion. Had this pathology come to light later in life in these children, it would not be possible to say with certainty whether the abnormality was present before revisional surgery or not. As in the case of one of the 2 children we treated prior to implementing our pathway we shall never know whether the vocal cord palsy was a result of our revisional surgery or not. Due to the inherently higher risks involved in revisional surgery for the reasons we have indicated we believe that thorough investigation is desirable in this regard. The available literature on revisional oesophageal surgery is limited. Although there are a number of impressive series reports of recurrent tracheo-oesophageal fistula repair [1–3], mainly following repair of congenital oesophageal atresia with tracheo-oesophageal fistula, none have subjected patients to extensive pre-operative investigations except for the purpose of confirming the diagnosis. In regard to pre-operative investigations Babu et al. have previously documented that echocardiogram may be unreliable in determining the side of the aortic arch in infants with OA/TOF, a feature in 2 of our patients [4]. A few reports have suggested that CT may be useful to gain additional anatomical information prior to primary surgical repair of OA/TOF in selected cases [5–7] and specifically to identify anatomical detail that may cause operative difficulty. We believe its utility in this regard is even greater prior to a revisional procedure. Whether further investigations such as magnetic resonance imaging studies including angiography and cardiac catheter studies would yield even greater information prior to revisional surgery is unclear. Magnetic resonance angiography is certainly an excellent modality for imaging the mediastinal vascular anatomy in children [8] and cardiac catheter studies will likely yield higher resolution information than can be obtained from echocardiogram. However cardiac catheter studies in particular are not without risk. We would suggest that these latter two investigations be considered and used selectively in this population. There are, however, a number of reports of vocal cord paralysis in children who have undergone repair of OA/TOF or had other oesophageal surgery [9–13]. In fact the number of reports of this phenomenon have called some to suggest that it should be considered a recognised complication of surgical repair of OA/TOF with an incidence as high as 12% [9]. Many reports of OA/TOF repair (both open and thoracoscopic) [14,15] and of recurrent TOF [1–3] fail to mention it. Grundfast and Harley have proposed that any child presenting with a hoarse voice or stridor following surgery for OA/TOF should be assumed to have a vocal cord paralysis until proven otherwise [16]. The proposed mechanism for vocal cord paralysis is recurrent laryngeal nerve injury or traction injury to the vagus during
2245 oesophageal manipulation [17]. Aberrant neural anatomy may contribute to the risk of intra-operative nerve damage; Qi et al. have demonstrated aberrant RLN branches and deviation of the vagi from their usual course in the Adriamycin rat model of OA [18] and in one series of endoscopic repair of patent ductus arteriosus ligation an unusual course of the vagus was identified in 13% of patients [19]. In addition to intra-operative neural damage it is possible that some children with OA/TOF are born with vocal cord paralysis. In a series of 6 children with OA/TOF and vocal cord paralysis Oestreicher-Kedem et al. reported a child with isolated OA who presented with bilateral vocal cord paralysis who had never undergone thoracotomy [20]. They suggest that vocal cord paralysis may be a congenital anomaly in some children with OA/TOF and cite the aberrant neural anatomy demonstrated in the experimental animal model as being compatible with this theory [18]. They recommend endoscopic evaluation of the vocal cords in all children prior to OA/TOF repair. In relation to the child they report, it is not clear if a cervical oesophagostomy was fashioned; this would potentially account for bilateral RLN damage. Indeed there is a higher than expected incidence of children who have undergone H-type TOF via a cervical approach in most series reporting vocal cord paralysis suggesting that cervical oesophageal surgery may be a particular risk factor for this complication [13]. We are not aware of other reports of congenital vocal cord paralysis in a child born with OA/TOF although vocal cord paralysis can certainly be a congenital abnormality [21]. Furthermore vocal cord paralysis may be completely asymptomatic as in 2 cases in our series [9]. We believe that the additional knowledge gained as a result of this standardised investigative pathway has impacted positively on our ability to treat this group of complex patients. We acknowledge the lack of a comparative group in order to confirm our beliefs but given the frequency and breadth of ‘incidental’ findings to date we recommend these investigations prior to revisional oesophageal surgery in all children.
References [1] Bruch SW, Hirschl RB, Coran AG. The diagnosis and management of recurrent tracheoesophageal fistulas. J Pediatr Surg 2010;45:337-40. [2] Ghandour KE, Spitz L, Brereton RJ, et al. Recurrent tracheooesophageal fistula: experience with 24 patients. J Paediatr Child Health 1990;26:89-91. [3] Ein SH, Stringer DA, Stephens CA, et al. Recurrent tracheoesophageal fistulas seventeen-year review. J Pediatr Surg 1983;18:436-41. [4] Babu R, Pierro A, Spitz L, et al. The management of oesophageal atresia in neonates with right-sided aortic arch. J Pediatr Surg 2000;35: 56-8. [5] Ratan SK, Varshney A, Mullick S, et al. Evaluation of neonates with esophageal atresia using chest CT scan. Pediatr Surg Int 2004;20: 757-61.
2246 [6] Su P, Huang Y, Wang W, et al. The value of preoperative CT scan in newborns with type C esophageal atresia. Pediatr Surg Int 2012;28: 677-80. [7] Chang EY, Chang HK, Han SJ, et al. Clinical characteristics and treatment of esophageal atresia: a single institutional experience. J Korean Surg Soc 2012;83:43-9. [8] Lee EY, Browne LP, Lam W. Noninvasive magnetic resonance imaging of thoracic large vessels in children. Semin Roentgenol 2012;47:45-55. [9] Robertson JR, Birck HG. Laryngeal problems following infant esophageal surgery. Laryngoscope 1976;86:965-70. [10] Hartnick CJ, Brigger MT, Willging JP, et al. Surgery for pediatric vocal cord paralysis: a retrospective review. Ann Otol Rhinol Laryngol 2003;112:1-6. [11] Daya H, Hosni A, Bejar-Solar I, et al. Pediatric vocal fold paralysis: a long-term retrospective study. Arch Otolaryngol Head Neck Surg 2000;126:21-5. [12] Miyamoto RC, Parikh SR, Gellad W, et al. Bilateral congenital vocal cord paralysis: a 16-year institutional review. Otolaryngol Head Neck Surg 2005;133:241-5. [13] Olsen L, Meurling S, Grotte G. H-type tracheo-oesophageal fistula in children with special reference to surgical management and to repair of recurrent nerve injury. Z Kinderchir 1982;36:27-9.
N.J. Hall et al. [14] Burford JM, Dassinger MS, Copeland DR, et al. Repair of esophageal atresia with tracheoesophageal fistula via thoracotomy: a contemporary series. Am J Surg 2011;202:203-6. [15] Castilloux J, Noble AJ, Faure C. Risk factors for short- and long-term morbidity in children with esophageal atresia. J Pediatr 2010;156: 755-60. [16] Grundfast KM, Harley E. Vocal cord paralysis. Otolaryngol Clin North Am 1989;22:569-97. [17] Hulscher JB, van Sandick JW, Devriese PP, et al. Vocal cord paralysis after subtotal oesophagectomy. Br J Surg 1999;86:1583-7. [18] Qi BQ, Merei J, Farmer P, et al. The vagus and recurrent laryngeal nerves in the rodent experimental model of esophageal atresia. J Pediatr Surg 1997;32:1580-6. [19] Odegard KC, Kirse DJ, del Nido PJ, et al. Intraoperative recurrent laryngeal nerve monitoring during video-assisted throracoscopic surgery for patent ductus arteriosus. J Cardiothorac Vasc Anesth 2000;14:562-4. [20] Oestreicher-Kedem Y, DeRowe A, Nagar H, et al. Vocal fold paralysis in infants with tracheoesophageal fistula. Ann Otol Rhinol Laryngol 2008;117:896-901. [21] de Jong AL, Kuppersmith RB, Sulek M, et al. Vocal cord paralysis in infants and children. Otolaryngol Clin North Am 2000;33:131-49.
www.elsevier.com/locate/jpedsurg
A standardised investigative strategy prior to revisional oesophageal surgery in children: High incidence of unexpected findings Nigel J. Hall a,b,⁎, Michelle Wyatt c , Joe I. Curry b , Edward M. Kiely b a
Surgery Unit, UCL Institute of Child Health, London, UK Department of General Surgery, Great Ormond Street Hospital NHS Foundation Trust, London, UK c Department of Otorhinolaryngology, Great Ormond Street Hospital NHS Foundation Trust, London, UK b
Received 17 January 2013; revised 24 February 2013; accepted 4 March 2013
Key words: Oesophageal atresia; Revisional surgery; Tracheomalacia; Airway reconstruction; Oesophageal replacement
Abstract Background/Purpose: Revisional oesophageal reconstructive surgery carries uncommon and unusual risks related to previous surgery. To provide maximum anatomical detail and facilitate successful outcome, we report a standardised pre-operative investigative strategy for all such patients. Methods: Prospective 8-month cohort study following the introduction of this strategy. All patients underwent high resolution thoracic contrast CT scan and micro-laryngo-bronchoscopy by a paediatric ENT surgeon in addition to upper gastrointestinal contrast study, oesophagoscopy, and echocardiogram. Results: Seven children (median age 5.6 months [range 2.2–60]) completed the pathway. Four were referred with recurrence of a previously divided tracheo-oesophageal fistula (3 congenital, 1 acquired) and 3 (all with oesophagostomy) for oesophageal replacement for congenital isolated oesophageal atresia (OA, n = 1) and failed repair of OA with distal TOF with wide gap (n = 2). Overall, unanticipated findings were demonstrated in 6/7 children and comprised severe tracheomalacia and right main bronchus stenosis requiring aortopexy (n = 1), vocal cord palsy (n = 2), extensive mediastinal rotation (n = 1), proximal tracheal diverticulum (n = 1), severe subglottic stenosis requiring airway reconstruction (n = 1), proximal tracheal diverticulum (n = 1), right sided aortic arch (n = 1) and left sided aortic arch (previously reported to be right sided, n = 1). Conclusions: This standardised approach for this complex group of patients reveals a high incidence of unexpected anatomical and functional anomalies with significant surgical and possible medico-legal implications. We recommend these investigations during the pre-operative work-up prior to all revisional oesophageal surgery. © 2013 Elsevier Inc. All rights reserved.
⁎ Corresponding author. Surgery Unit UCL Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK. Tel.: +44 020 7905 2733. E-mail address: [email protected] (N.J. Hall). 0022-3468/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.03.051
When compared with primary oesophageal surgery, revision surgery is technically more difficult and carries higher operative risks. These are most commonly the result of anatomical distortion from operative scarring and obliteration of normal tissue planes and appearance. Children requiring
2242 revisional surgery following failed or complicated repair of congenital or acquired oesophageal structural abnormalities are often referred to tertiary referral centres. As a result, uncertainty is introduced which may increase the risk of subsequent surgical complications. The surgeon is reliant on second-hand information regarding the original anatomy, and the precise nature of any previous surgery may be unclear. We have recently encountered a number of patients referred for revisional oesophageal surgery for a variety of indications. As a result of adverse outcome in 2 of these patients we introduced a standardised investigative pathway within our institution. The aim was to improve our knowledge of the preoperative anatomy and function in these patients in order to reduce operative misadventure and post-operative complications. Herein we describe our investigative pathway and our initial findings following its implementation.
1. Methods Our current investigative pathway has 3 stages. Firstly we perform whatever investigations are clinically indicated to confirm the nature of the abnormality for which the patient has been referred. On occasion, clinical examination and review of the radiological investigations that arrive with the child are all that is necessary. However most will need further targeted radiological imaging such as a prone tube oesophagogram in the case of suspected recurrent tracheooesophageal fistula. All children undergo rigid or flexible oesophagoscopy and echocardiogram. We then perform a high resolution computed tomography (CT) scan of the chest with intravenous contrast under general anaesthesia if necessary. The aim of this CT scan is to provide anatomical detail of the thoracic vascular anatomy, detail of the oesophageal and airway anatomy and to identify the presence of pulmonary disease. Finally an endoscopic examination of the airway is performed by an experienced paediatric otorhinolaryngologist with the general paediatric surgical team in attendance. The specific purpose of this examination is to identify structural or dynamic airway abnormalities including significant tracheomalacia, the presence of a tracheo-oesophageal fistula and its anatomical location and to assess vocal cord movement. We collected data prospectively over the 8 month period up to December 2011 and report all patients referred for revisional oesophageal surgery during this time period. We firstly report brief details of 2 patients on whom we operated prior to this and who had adverse outcomes. Our current investigative pathway was devised following the treatment of these initial 2 patients.
2. Results The first child was referred at the age of 6 months. At the age of 1½ months he had suffered an oesophageal injury due
N.J. Hall et al. to impaction of a battery in the upper oesophagus. The battery was retrieved from the oesophagus at an emergency open operation and a cervical oesophagostomy fashioned because of damage to the oesophageal wall. A gastrostomy was created to allow feeding. The child was referred to our centre for oesophageal reconstruction. Following assessment of oesophageal length, oesophageal reconstruction was performed. Post-operatively the child was noted to have a hoarse voice and subsequent laryngoscopy revealed evidence of left recurrent laryngeal nerve (RLN) palsy. It was unclear if this had been present prior to surgery. The hoarseness settled. A second child was referred for oesophageal replacement surgery at the age of 18 months having been born preterm at 34 weeks gestation with oesophageal atresia and distal tracheo-oesophageal fistula, an anorectal malformation and VACTERL association. Following a complicated operative course, a gastrostomy and cervical oesophagostomy were in situ at the time of referral. In addition there had been a previous episode of cardiac arrest requiring intensive care treatment and an episode of sepsis-related acute renal failure requiring haemofiltration. A gastric transposition was successfully performed. Twenty four hours following surgery signs of superior vena cava (SVC) obstruction developed with acute cardiorespiratory compromise. Echocardiography revealed a markedly narrowed SVC with thrombus and no flow. In view of further deterioration an attempt was made to commence extra-corporeal life support but it was not possible to achieve venous cannulation due to stenosis in multiple vessels. The child died. Following the treatment of these two children we derived a standardised investigative pathway for all children referred for revisional oesophageal surgery. We have evaluated 7 children in this manner. Demographic and clinical details of these children are shown in Table 1. The combination of contrast CT scan and endoscopic evaluation or the airway has revealed a high incidence of abnormalities which were not suspected clinically but which have had clinical, surgical and in some cases potentially medico-legal implications. The abnormalities detected and subsequent clinical course of each child are detailed in Table 1. All children were orally fed following surgery. Overall unexpected findings were detected in 6 of the 7 patients. The most significant airway anomaly detected was critical subglottic stenosis in child C. The airway would accept only a size 2.5 endotracheal tube at the age of 6 months. He required larnygotracheal reconstruction prior to oesophageal replacement surgery as it was deemed unsafe to proceed with major oesophageal surgery requiring a period of postoperative ventilation with a critically narrow airway. Examples of the other abnormalities detected are shown in the Figures. Fig. 1 shows pulmonary disease secondary to chronic aspiration alongside significant rotation of the mediastinal anatomy such that the trachea lies lateral and to the right side of the oesophagus in child A who had previously undergone division of a congenital trachea-oesophageal fistula
Strategy prior to revisional oesophageal surgery Table 1
2243
Patients, investigation findings and outcome.
Patient Age at Initial referral pathology (m)
Previous oesophageal surgery & investigation findings
Reason for referral
Bronchoscopy CT chest findings findings
A
2
OA with distal Primary repair OA/TOF TOF, prematurity (28 weeks)
B
10
C
3
OA with distal Primary repair OA/TOF; TOF; tracheomalacia recurrent fistula repair × 2; aortopexy OA with distal Thoracotomy, Oesophageal Severe subTOF cervical reconstruction glottic stenosis oesophagostomy & gastrostomy
D
21
E
5
F
3
G
60
OA with distal TOF, tracheomalacia Isolated OA, tetralogy of Fallot-repaired
Isolated OA
Primary repair OA/TOF; aortopexy Cervical oesophagostomy & gastrostomy; left sided aortic arch
Thoracotomy, cervical oesophagostomy & gastrostomy; right sided arch Acquired TOF Cervical repair (battery of acquired TOF ingestion)
Repair of Left vocal cord Rotation of recurrent TOF paralysis mediastinum to the right; trachea to right of oesophagus Right lower lobe consolidation Inadequate Repair of Significant recurrent TOF tracheomalacia aortopexy; right main bronchus (ventilator stenosis dependent)
Surgery performed
Outcome
Repair Good recurrent TOF
Trans-sternal Good aortopexy; repair recurrent TOF Good Laryngotracheal reconstruction; Laparoscopic gastric transposition Repair Good recurrent TOF
Repair of recurrent TOF Oesophageal Blind ending reconstruction pouch 3 cm below vocal cords
Laparoscopic Right aortic arch; double superior vena gastric cava; tracheal transposition diverticulum
Oesophageal reconstruction
Left sided aortic arch Laparoscopic gastric transposition
Repair of Left vocal cord recurrent TOF paralysis
Temporary right Horner's syndrome (resolved after 2 months) Good
Repair Good recurrent TOF
OA, oesophageal atresia; TOF, trachea-oesophageal fistula.
and repair of oesophageal atresia. Fig. 2 shows a wide distance between the aortic arch and the sternum in child B despite a previous aortopexy for critical tracheomalacia in addition to pulmonary disease due to chronic aspiration. The child remained ventilator dependent. A repeat aortopexy was performed via a trans-sternal route to allow the child to breathe spontaneously prior to surgical correction of the rerecurrent trachea-oesophageal fistula. Fig. 3A shows the endoscopic appearance of a blind ending upper tracheal diverticulum in child E who had been born with pure oesophageal atresia and managed with a cervical oesophagostomy and gastrostomy. This is depicted on 3 dimensional airway reconstruction images from the CT scan in Fig. 3B. It was not possible to achieve delayed anastomosis of the
oesophagus so a laparoscopic gastric transposition was performed. Prior knowledge of the presence of the upper tracheal diverticulum was of surgical relevance when performing the dissection of the cervical oesophagostomy and cervical oesophagogastric anastomosis.
3. Discussion In this short series of children referred for revisional oesophageal surgery, we have implemented a standardised investigative pathway and been impressed by the high incidence of clinically unexpected findings. We arrived at this investigative regimen following adverse outcomes in 2
2244
N.J. Hall et al.
Fig. 1 Thoracic CT scan of patient A. T, trachea; O, oesophagus containing nasogastric tube. There is obvious mediastinal rotation such that the trachea lies to the right of the oesophagus, and pulmonary consolidation consistent with recurrent aspiration.
earlier and similar patients; our practice was influenced as a direct consequence of these outcomes. While we cannot say with certainty that the ‘incidental findings’ we have unearthed have impacted directly on the clinical outcome of the subsequent 7 children, we believe that there are a number of benefits to the knowledge they have provided. We do believe that had we performed these investigations in the initial 2 patients we would have been in a better position to address the clinical scenarios and significant uncertainties that arose during their course of treatment. Firstly, in some cases, we have been able to explain previous treatment failure and guide subsequent treatment (for example in child B who had undergone an unsuccessful aortopexy prior to referral to us). Secondly, in child C we believe it would have been unsafe to proceed with major oesophageal surgery requiring post-operative paralysis and ventilation (as per our current protocol following gastric transposition) in the presence of the critically narrow subglottic stenosis. Third, we believe that knowledge of
Fig. 2 CT scan of patient B. The line indicates the distance between the aorta and posterior aspect of the sternum in a child who has previously had an aortopexy. Pulmonary consolidation is also present.
Fig. 3 Bronchoscopy findings (A) and rendered CT scan of the airway (B) from patient E. A catheter has been passed into a blindending upper tracheal diverticulum. The trachea is seen to the left (A). The diverticulum can be seen arising from the right side of the upper trachea (B).
the anatomy that has been achieved due to the CT scan has allowed us to plan surgery more effectively, interpret our surgical findings with a greater degree of certainty and alter our surgical approach where indicated. The mediastinal rotation detected preoperatively in child A meant that intraoperative findings could be predicted more accurately and surgical approach adjusted accordingly with the potential to reduce the risk of intra-operative complications. Although it is difficult to prove, the ‘route map’ provided may have helped to prevent inadvertent intraoperative perforation of either trachea or oesophagus. Knowledge of the presence of the upper tracheal diverticulum in child E was certainly extremely useful when it was encountered in the neck during dissection of the cervical oesophagostomy prior to cervical oesophagogastric anastomosis. As an aside we believe that this upper tracheal diverticulum most likely results from either a ‘missed’ upper pouch tracheooesophageal fistula or a fistula that was occluded at one
Strategy prior to revisional oesophageal surgery end in this child who did not have endoscopic evaluation of his airway prior to formation of the cervical oesophagostomy in the referring centre. Finally, some of our findings may have medico-legal implications. Two children in this series were found to have a unilateral vocal cord palsy despite no clinical suspicion. Had this pathology come to light later in life in these children, it would not be possible to say with certainty whether the abnormality was present before revisional surgery or not. As in the case of one of the 2 children we treated prior to implementing our pathway we shall never know whether the vocal cord palsy was a result of our revisional surgery or not. Due to the inherently higher risks involved in revisional surgery for the reasons we have indicated we believe that thorough investigation is desirable in this regard. The available literature on revisional oesophageal surgery is limited. Although there are a number of impressive series reports of recurrent tracheo-oesophageal fistula repair [1–3], mainly following repair of congenital oesophageal atresia with tracheo-oesophageal fistula, none have subjected patients to extensive pre-operative investigations except for the purpose of confirming the diagnosis. In regard to pre-operative investigations Babu et al. have previously documented that echocardiogram may be unreliable in determining the side of the aortic arch in infants with OA/TOF, a feature in 2 of our patients [4]. A few reports have suggested that CT may be useful to gain additional anatomical information prior to primary surgical repair of OA/TOF in selected cases [5–7] and specifically to identify anatomical detail that may cause operative difficulty. We believe its utility in this regard is even greater prior to a revisional procedure. Whether further investigations such as magnetic resonance imaging studies including angiography and cardiac catheter studies would yield even greater information prior to revisional surgery is unclear. Magnetic resonance angiography is certainly an excellent modality for imaging the mediastinal vascular anatomy in children [8] and cardiac catheter studies will likely yield higher resolution information than can be obtained from echocardiogram. However cardiac catheter studies in particular are not without risk. We would suggest that these latter two investigations be considered and used selectively in this population. There are, however, a number of reports of vocal cord paralysis in children who have undergone repair of OA/TOF or had other oesophageal surgery [9–13]. In fact the number of reports of this phenomenon have called some to suggest that it should be considered a recognised complication of surgical repair of OA/TOF with an incidence as high as 12% [9]. Many reports of OA/TOF repair (both open and thoracoscopic) [14,15] and of recurrent TOF [1–3] fail to mention it. Grundfast and Harley have proposed that any child presenting with a hoarse voice or stridor following surgery for OA/TOF should be assumed to have a vocal cord paralysis until proven otherwise [16]. The proposed mechanism for vocal cord paralysis is recurrent laryngeal nerve injury or traction injury to the vagus during
2245 oesophageal manipulation [17]. Aberrant neural anatomy may contribute to the risk of intra-operative nerve damage; Qi et al. have demonstrated aberrant RLN branches and deviation of the vagi from their usual course in the Adriamycin rat model of OA [18] and in one series of endoscopic repair of patent ductus arteriosus ligation an unusual course of the vagus was identified in 13% of patients [19]. In addition to intra-operative neural damage it is possible that some children with OA/TOF are born with vocal cord paralysis. In a series of 6 children with OA/TOF and vocal cord paralysis Oestreicher-Kedem et al. reported a child with isolated OA who presented with bilateral vocal cord paralysis who had never undergone thoracotomy [20]. They suggest that vocal cord paralysis may be a congenital anomaly in some children with OA/TOF and cite the aberrant neural anatomy demonstrated in the experimental animal model as being compatible with this theory [18]. They recommend endoscopic evaluation of the vocal cords in all children prior to OA/TOF repair. In relation to the child they report, it is not clear if a cervical oesophagostomy was fashioned; this would potentially account for bilateral RLN damage. Indeed there is a higher than expected incidence of children who have undergone H-type TOF via a cervical approach in most series reporting vocal cord paralysis suggesting that cervical oesophageal surgery may be a particular risk factor for this complication [13]. We are not aware of other reports of congenital vocal cord paralysis in a child born with OA/TOF although vocal cord paralysis can certainly be a congenital abnormality [21]. Furthermore vocal cord paralysis may be completely asymptomatic as in 2 cases in our series [9]. We believe that the additional knowledge gained as a result of this standardised investigative pathway has impacted positively on our ability to treat this group of complex patients. We acknowledge the lack of a comparative group in order to confirm our beliefs but given the frequency and breadth of ‘incidental’ findings to date we recommend these investigations prior to revisional oesophageal surgery in all children.
References [1] Bruch SW, Hirschl RB, Coran AG. The diagnosis and management of recurrent tracheoesophageal fistulas. J Pediatr Surg 2010;45:337-40. [2] Ghandour KE, Spitz L, Brereton RJ, et al. Recurrent tracheooesophageal fistula: experience with 24 patients. J Paediatr Child Health 1990;26:89-91. [3] Ein SH, Stringer DA, Stephens CA, et al. Recurrent tracheoesophageal fistulas seventeen-year review. J Pediatr Surg 1983;18:436-41. [4] Babu R, Pierro A, Spitz L, et al. The management of oesophageal atresia in neonates with right-sided aortic arch. J Pediatr Surg 2000;35: 56-8. [5] Ratan SK, Varshney A, Mullick S, et al. Evaluation of neonates with esophageal atresia using chest CT scan. Pediatr Surg Int 2004;20: 757-61.
2246 [6] Su P, Huang Y, Wang W, et al. The value of preoperative CT scan in newborns with type C esophageal atresia. Pediatr Surg Int 2012;28: 677-80. [7] Chang EY, Chang HK, Han SJ, et al. Clinical characteristics and treatment of esophageal atresia: a single institutional experience. J Korean Surg Soc 2012;83:43-9. [8] Lee EY, Browne LP, Lam W. Noninvasive magnetic resonance imaging of thoracic large vessels in children. Semin Roentgenol 2012;47:45-55. [9] Robertson JR, Birck HG. Laryngeal problems following infant esophageal surgery. Laryngoscope 1976;86:965-70. [10] Hartnick CJ, Brigger MT, Willging JP, et al. Surgery for pediatric vocal cord paralysis: a retrospective review. Ann Otol Rhinol Laryngol 2003;112:1-6. [11] Daya H, Hosni A, Bejar-Solar I, et al. Pediatric vocal fold paralysis: a long-term retrospective study. Arch Otolaryngol Head Neck Surg 2000;126:21-5. [12] Miyamoto RC, Parikh SR, Gellad W, et al. Bilateral congenital vocal cord paralysis: a 16-year institutional review. Otolaryngol Head Neck Surg 2005;133:241-5. [13] Olsen L, Meurling S, Grotte G. H-type tracheo-oesophageal fistula in children with special reference to surgical management and to repair of recurrent nerve injury. Z Kinderchir 1982;36:27-9.
N.J. Hall et al. [14] Burford JM, Dassinger MS, Copeland DR, et al. Repair of esophageal atresia with tracheoesophageal fistula via thoracotomy: a contemporary series. Am J Surg 2011;202:203-6. [15] Castilloux J, Noble AJ, Faure C. Risk factors for short- and long-term morbidity in children with esophageal atresia. J Pediatr 2010;156: 755-60. [16] Grundfast KM, Harley E. Vocal cord paralysis. Otolaryngol Clin North Am 1989;22:569-97. [17] Hulscher JB, van Sandick JW, Devriese PP, et al. Vocal cord paralysis after subtotal oesophagectomy. Br J Surg 1999;86:1583-7. [18] Qi BQ, Merei J, Farmer P, et al. The vagus and recurrent laryngeal nerves in the rodent experimental model of esophageal atresia. J Pediatr Surg 1997;32:1580-6. [19] Odegard KC, Kirse DJ, del Nido PJ, et al. Intraoperative recurrent laryngeal nerve monitoring during video-assisted throracoscopic surgery for patent ductus arteriosus. J Cardiothorac Vasc Anesth 2000;14:562-4. [20] Oestreicher-Kedem Y, DeRowe A, Nagar H, et al. Vocal fold paralysis in infants with tracheoesophageal fistula. Ann Otol Rhinol Laryngol 2008;117:896-901. [21] de Jong AL, Kuppersmith RB, Sulek M, et al. Vocal cord paralysis in infants and children. Otolaryngol Clin North Am 2000;33:131-49.