Original Article
Tertiary Surgery for Complicated Repair of Esophageal Atresia Ruben Ortiz1 Alba Sánchez Galán1 Leopoldo Martinez1 Manuel Lopez Santamaria1 Juan Antonio Tovar1 1 Department of Pediatric Surgery, Hospital Universitario La Paz,
Madrid, Spain
Eva Dominguez1
Francisco Hernández1
Address for correspondence Ruben Ortiz, MD, Department of Pediatric Surgery, La Paz Universitary Hospital, Paseo de la Castellana 261, Madrid 28046, Spain (e-mail: [email protected]).
Eur J Pediatr Surg 2015;25:20–26.
Abstract
Keywords
► esophageal atresia ► esophageal replacement ► long-gap ► elongation ► complications
received May 15, 2014 accepted June 23, 2014 published online August 21, 2014
Aim The ideal repair of esophageal atresia (EA) is primary anastomosis with closure of the fistula if present. Long gap or local circumstances prompt other procedures that occasionally lead to disastrous complications. The aim of this study was to analyze the management of these complications in a tertiary referral center. Patients and Methods A retrospective review of patients treated for EA between 1993 and 2013 was conducted. Both the patients were primarily treated by us, and referrals from elsewhere after two or more failed operations were included. Results In total, 23 patients were included (3/176 cases of EA treated primarily by us and 20 referrals). Of the 23 patients, 6 had type I EA, 15 type III (four long gaps), 1 type IV, and 1 type V. Cardiac anomalies were associated in seven cases, duodenal atresia in three, and Down syndrome in two patients. Primary anastomosis was initially achieved in 12 patients. Primary or secondary Foker lengthening was used in seven cases. The causes of the failure were anastomotic leaks in nine, unmanageable strictures in seven, and refistulization in five patients. These patients required 66 reoperations (median of 3 [2–7]) before inclusion in the study. Radical tertiary treatment consisted of 15 esophageal replacements (11 colonic grafts and 4 gastric pull-ups), and 1 esophageal-gastric disconnection. Five patients previously treated with esophageal replacement and referred for graft problems required 13 interventions. Two families did not give consent for one replacement and one disconnection. Complications appeared in 12 patients, and 9 additional operations were required in 7 patients. With a follow-up of 31 months (range, 4–139 months) 15 patients take all their meals per os, 5 occasionally use the gastrostomy, and 2 and 1 are fed exclusively via gastrostomy or jejunostomy. All tracheoesophageal fistulas were closed, but 15 cases are below p3 for weight and 12 for height. Three patients (13%) ultimately died 32 months (range, 9–56 months) after the first operation (due to aspiration in one, and for causes unrelated to it in the other two [tracheostomy obstruction and Guillain–Barré syndrome]). Conclusions When repeated complications appear after EA repair, radical surgical attitudes may be justified. If esophageal continuity cannot be reestablished, the native esophagus may have to be discarded and replaced. Many complications should be expected, but the end result can be good. These patients should be referred to centers with large experience in the management of this complex condition.
© 2015 Georg Thieme Verlag KG Stuttgart · New York
DOI http://dx.doi.org/ 10.1055/s-0034-1386645. ISSN 0939-7248.
Downloaded by: University of Georgia Libraries. Copyrighted material.
20
Tertiary Surgery for Complicated Repair of EA
Esophageal atresia (EA) occurs in 1 of every 2,500 to 3,000 live births,1 and represents a paradigm of pediatric surgery because the results of its treatment are an index of the quality of the care provided by individuals and instituions.2 The aim of EA repair is to reestablish esophageal patency by performing a tension-free, primary anastomosis of both ends of the native organ, with closure of the tracheoesophageal fistula (TEF) if present. This is possible in most cases because the atretic proximal esophageal pouch is more or less close to the fistula that communicates the trachea with the stomach. Pediatric surgeons learned along the years that, even under some tension, primary anastomosis could be done in most cases and that only in a few of them, both ends are too far apart to achieve primary repair. These “long-gap” cases and the 10% that have “pure” EA (proximal atresia of the esophagus without fistula) are not amenable to neonatal primary anastomosis and require other procedures. Delayed primary anastomosis,3,4 esophageal lengthening procedures,5,6 and esophageal replacement7,8 are therefore mandatory in a limited proportion of cases in which the risks involved are high and the complications frequent and severe. These, and the few cases in which primary repair fails because of massive leaks or recurrent TEF leading to repeated unsuccessful operations, represent major reconstructive challenges. Every pediatric surgeon considers himself/herself entitled to perform all the surgical procedures related to EA and TEF. Our large experience in the field has shown us along the years that much of the morbidity and mortality of this condition are hidden in unpublished small series, and we felt that it might be useful to show that only after risky operations in tertiary
21
centers and at the price of many complications, most of these unfortunate patients can be put back into track. The aim of this study was to analyze the management of a large group of these patients at a tertiary referral center with particular interest on esophageal conditions.
Patients and Methods A retrospective review of patients treated at our Institution for EA between 1993 and 2013 was conducted. Only children having two or more failed major operations were included in this study and this cohort involved both children primarily treated by us and referred from elsewhere. Clinical data, associated comorbidities, initial procedure, complications, reoperations, and long-term outcome were recorded and analyzed. Stricture dilations were not taken into account as initial reoperations, but they are discussed in the Results section. The current status of the patient was assessed in every case by imaging and endoscopic procedures. The viability of direct repair or the need for discarding and replacing the esophagus was individually judged on the basis of careful examination. The aims of the radical procedure were to close fistulas, to reestablish patent anastomoses, or to remove and replace the esophagus if it was deemed unsalvageable. In these cases, it was resected except when this involved too much risk. Procedures previously described were used for replacement with either the colon or the stomach. Transhiatal route was preferred, but it was not always usable. Operations on previously performed replacements were done transsternally. Previously placed gastrostomies or new ones were liberally used in the postoperative period and closed as soon as possible.
Table 1 First repair interventions performed in the 23 patients of the study with complicated EA First repair procedure
n
Type of EA (number)
Thoracotomy, fistula ligation, and primary anastomosis
12
type III (11); type IV (1)
Gastrostomy and fistula ligation
1
type III long-gap (1)
Gastrostomy an delayed primary anastomosis
1
type I (1)
Gastrostomy
5
type I (4); type V (1)
Foker procedure
2
type III long-gap (2)
Gastrostomy þ esophagostomy
1
type I (1)
Gastrostomy þ dynamic stump-lengthening technique
1
type I (1)
Complications after first repair Major leaks
9
Stricture
7
Refistulization
5
Aspiration and pneumonia
7
Gastrostomy–esophagostomy problems
1
Chylothorax and chylous ascites
1
Abbreviation: EA, esophageal atresia. Note: Total times of each procedure, type of EA and posterior complications after first attempt of repair are shown. European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Downloaded by: University of Georgia Libraries. Copyrighted material.
Introduction
Ortiz et al.
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
Patients were seen at the outpatient clinic every 6 months. Height and weight were recorded at that time and nutritional assessment was done if required. Data are reported as median (range). Imaging and endoscopic assessment of the results were performed as appropriate.
Results In total, 23 patients (15 males/8 females) were included in this review. Of the 23 patients, 3 were treated primarily by ourselves (from 176 cases of EA treated at Hospital Universitario La Paz over the last 20 years), and the remaining 20 (86%) patients were referred from other institutions after they had been unsuccessfully treated. Median age at birth was 35.5 weeks (range, 31–40 weeks), with a mean weight at birth of 2, 085 g (range, 1,200–2,900 g). Six patients had pure EA (type I), 15 failed repair of type III (in four patients considered “long-gap”), and 1 case of type IV and type V each. Associated anomalies were present in 17 patients: 7 major cardiac defects, 3 VACTERL associations, 3 duodenal atresias, 2 Down syndromes, 1 Goldenhar syndrome and other malformations (Morgagni hernia, Meckel diverticulum, congenital subglottic stenosis, and ear agenesis) in 1 patient each. The first attempt at repair had been performed by thoracotomy in 21 cases and by thoracoscopy in 2 cases. Fistula closure and primary anastomosis were initially achieved in 12 patients, Foker lengthening procedure in 2, delayed primary anastomosis in 1, dynamic stump-lengthening techniques in 1, and gastrostomy þ esophagostomy or fistula ligation in the remaining 7 patients. The causes of the failure were major anastomotic leaks in nine, unmanageable strictures in seven, and repeated refistulization in five patients. Massive reflux with aspiration and pneumonia was also present in seven cases after the first procedure, and one patient had chylothorax and chylous ascites(►Table 1). These patients had a median of 3 (range, 2–7) reoperations (total of 66) before inclusion in the study, on top of esophageal dilations that were necessary in 9 (up to 57 procedures). Overall, 14 Nissen fundoplications, 12 thoracotomies for recurrent fistula closure or redoanastomosis, 5 secondary Foker lengthening techniques, and 3 repeated endoscopic attempts at recurrent TEF closure had been performed. Five esophageal replacements (four colonic [three retroesternal, one mediastinal] and one Schärli procedure) had been performed elsewhere and ultimately referred to us for complications with the graft (►Table 2). All Foker lengthenings and attempts at endoscopic fistula closure had been performed elsewhere. Radical tertiary treatment at an age of 20.5 months (range, 5–70 months) consisted of 15 of 23 (65%) esophageal resections and replacements (11 colonic grafts [8 intramediastinal and 3 retroesternal] and 4 gastric pull-ups) and one esophageal-gastric disconnection in a patient with intractable reflux. The five patients previously treated with esophageal replacement and referred for graft problems were evaluated and required 13 interventions (three redo esophageal-graft anastomosis, four pyloroplasties, two gastro-duodenostoEuropean Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Table 2 Reoperations performed after initial repair before inclusion in the study Procedure
n
Thoracotomy and esophageal redo-anastomosis
4
Thoracotomy and recurrent TEF ligation
8
Gastrostomy esophagostomy
16
Foker lengthening procedure
5
Nissen fundoplication
14
Endoscopic techniques for recurrent TEF closure
3
Esophagostomy dilation
1
Cervicotomy and proximal fistula ligation
2
Thoracotomy and lobectomy
1
Laparotomy
2
Tracheostomy
1
Feeding jejunostomy
1
Esophagectomy
1
Endoscopic dilation of esophageal anastomotic stricturea
57 in 9 patients
Mediastinal colonic replacement
1
Retroesternal colonic replacement
3
Schärli procedure
1
Redo esophageal-graft anastomosis
1
Endoprosthesis stent for esophageal-graft stricture
1
Abbreviation: TEF, tracheoesophageal fistula. Note: Total of 66 in 23 patients, with a median of 3 (2–7). a Total of reoperations (n ¼ 66), excluding esophageal stricture dilations (57 procedures in 9 patients).
mies, one esophagostomy, and one mucocele esophagectomy). One presternal coloplasty after another stenoticischemic colonic graft performed elsewhere was achieved, and one feeding jejunostomy in a patient who did not consent to esophageal-gastric disconnection. Finally, of the remaining two patients, one family did not give consent for replacement, and the other patient with a Down syndrome and had three Nissen procedures (►Fig. 1). Complications appeared in 12 of 15 patients (80%) (anastomotic stricture in four, salivary fistula in five, delayed gastric emptying in two, severe reflux in three, and intestinal obstruction in two patients), and 9 additional operations were required in 7 excluding stricture dilations (27 procedures in 8 patients). These reoperations after radical treatment consisted in four redo esophageal-graft anastomosis, one gastro-duodenostomy, two laparotomy for intestinal obstruction, one esophagostomy, and one incisional hernia repair) (►Table 3). We have seen that patients who had the association of EA and duodenal atresia (n ¼ 3) needed a high number of
Downloaded by: University of Georgia Libraries. Copyrighted material.
22
Ortiz et al.
23
Downloaded by: University of Georgia Libraries. Copyrighted material.
Tertiary Surgery for Complicated Repair of EA
Fig. 1 Tertiary treatment for 23 patients with complicated esophageal atresia treated at Hospital La Paz in the period of 1993 to 2013. Radical treatments, complications, and outcomes are summarized.
reoperations due to secondary complications. One of them received three thoracotomies for esophageal reanastomosis elsewhere and finally we had to replace it. In another case, we opted for an esophageal-gastric disconnection after two failed Nissen fundoplications with persistent intractable reflux; and the third patient who has Down syndrome, had three Nissen procedures and is fed by gastrostomy. With a follow-up of 31 months (range, 4–139 months), most patients (15/23) take all their meals per os, five use the gastrostomy only occasionally, and two and one are fed exclusively via gastrostomy or jejunostomy, respectively. All TEFs were closed, but 15 of 23 cases are below p3 for weight and 12 of 23 below p3 for height. Three patients (13%) ultimately died 32 months (range, 9–56 months) after opera-
tion, because of aspiration in one, and unrelated causes in the other two patients (tracheostomy obstruction and Guillain– Barré syndrome).
Discussion Repair of EA still remains a challenge for the pediatric surgeon because not all patients have the same spectrum of disease. Nowadays, mortality in EA is rare and survival is mainly determined by the presence of associated anomalies; however, the outcome will also depend on the nature of the surgical procedures performed during the first attempt at repair and this is particularly true for the “difficult cases.” It is widely accepted that the own child’s esophagus is the best material for European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
Table 3 Radical-tertiary treatment performed at our institution for 23 patients with complicated EA Radical tertiary treatment (total 21/23 patients)
Procedures (n)
1. Esophageal resection and replacement (15 patients)
15
• Mediastinal colonic graft
8
• Retroesternal colonic graft
3
• Gastric pull-up
4
2. Patients referred with graft problems after esophageal replacement elsewhere (5 patients
13
• Redo esophageal-graft anastomosis,
3
• Pyloroplasty
4
• Gastro-duodenostomy
2
• Esophagsotomy
1
• Esophagectomy (mucocele)
1
• Pre-esternal coloplasty
1
• Feeding jejunostomy
1
3. Esophageal-gastric disconnection (1 patient)
1
Complications after radical-tertiary treatment (12 patients)
n
Anastomotic stricture
4
Salivary fistula
5
Delayed gastric emptying
2
Severe reflux
3
Aspiration and pneumonia
3
Intestinal obstruction
2
Dumping syndrome
1
Anastomotic dehiscence
1
Additional reoperations (7 patients)
Number of procedures (n)
Redo esophageal-colonic anastomosis
3
Redo esophageal-gastric anastomosis
1
Gastro-duodenostomy
1
Esophagostomy
1
Laparotomy for intestinal obstruction
2
Incisional hernia repair
1
Dilation of anastomotic esophageal-graft stricture
a
27 in 8 patients
Note: Type and number of procedures performed, complications appeared and additional reoperations needed are detailed. a Total additional reoperations after tertiary treatment (n ¼ 9), excluding anastomotic stricture dilations.
repair. However in “long-gap” or no-esophagus situations, to maintain the aim of obtaining a primary anastomosis at any price may lead to repeated operations and complications that can have serious long-term consequences for the child. Different procedures have been proposed for owner-establishing digestive patency with esophageal material. Kimura multistage extra thoracic esophageal elongation6 and particularly the Foker traction lengthening procedure5,9 had been used for achieving primary anastomoses even in extreme cases in which esophageal ends are too separated or in pure EA with a minimum distal esophageal nubbin. For achieving this goal, these techniques need repeated traction procedures (extrathoracic in Kimura and intrathoracic in Foker) to get esophageal lengthening ultimately allowing European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
anastomosis at any price. This can be indeed achieved in many cases, but the price to pay, in terms of functional quality of the reconstructed organ, may be too high.2 In a series of extrathoracic esophageal elongation,6 11 of 12 (91.6%) patients required a Nissen fundoplication and all received several dilations for stricture and even stricture resection in three (25%) patients. In the largest series of intrathoracic esophageal lengthening,9 traction sutures had to be relocated in several patients during the postoperative period involving several thoracotomies. All patients required Nissen fundoplication (42/42) due to the constant presence of reflux and 12 of 42 (28.5%) had a redofundoplication. Dilations for stricture and resection (with more thoracotomies) were necessary in many cases. In addition, 15 patients had patches of gastric
Downloaded by: University of Georgia Libraries. Copyrighted material.
24
metaplasia above the gastroesophageal junction. Both lengthening techniques at two different European centers were recently compared10 in a limited group of patients. The results were only fair, with long periods of traction to achieve the anastomosis, the need for repeated thoracotomies and severe complications in three patients (two finally need an esophageal replacement). The authors concluded that more data are needed (they suggested a multicenter registry) to elucidate the efficacy and safety of each elongation procedures. Other authors prefer a delayed primary anastomosis in pure EA3,4 based on the spontaneous growth of the upper pouch of the esophagus without any mechanical traction, because of swallowing and reflux into the lower pouch. They observed that this maximal growth occurred during the first 8 to 12 weeks of life, suggesting that the ideal time for delayed primary anastomosis was approximately 12 weeks of age. In the study by Sri Paran et al,4 a total of 21 patients with pure EA with a median initial gap of 3.7 cm received a delayed primary anastomosis. However, this procedure implied a late repair (median, 80 days) with a prolonged hospitalization period (median of 5.5 months) and continuous aspiration of the upper pouch. Moreover, symptomatic gastroesophageal reflux (GER) occurred in 66% of cases, Nissen fundoplication was required in 43% of cases, repeated stricture dilations were necessary in 47%, and stricture resection with redoanastomosis was eventually performed in 28.5% of these children. In a recent meta-analysis of complications and long-term outcome of this procedure comparing 44 articles with 451 newborns treated with this procedure,11 the incidence of postoperative interventions were as follows: stricture dilation in 60% of cases, stricture resection and reanastomosis in 21%, fundoplication for symptomatic GER in 36%, and esophageal replacement for unsatisfactory results in 14%. Authors highlighted that the incidence of GER and strictures requires early intervention to prevent feeding problems. The available alternative to preserving the native esophagus at any price is esophageal replacement. It can be considered primarily in log-gap situations and whenever survival is at risk after several attempts to preserve the patency of native esophagus.1 The ideal esophageal substitute should function as closely as possible to the original one, and it should work well during the many years without early deterioration.7,12,13 Gastric transposition “gastric pull-up”7,14 and colonic interposition8,15–18 are the two most extended procedures. Satisfactory results in terms of quality of life and normal feeding pattern without digestive or respiratory symptoms are possible,7,8,18 but the price to pay is an additional major operative risk with frequent and potentially serious complications.12,13 Another important scenario met after repair of EA is recurrent TEF that occurs in approximately 5 to 10% of cases, and involves difficult reoperations. Repeated thoracotomies and/or endoscopic closure techniques generate scarring and inflammation that add further difficulties to repair,19,20 cause prolonged hospital stays, and threaten quality of life. When facing a difficult repair, pediatric surgeons have to decide whether they can deal with all the possible scenarios
Ortiz et al.
involved or, alternatively, whether they will benefit more the patient by referring him to a tertiary institution. The experience summarized in this study suggests that most of them face one of these problems rarely and that, like for any other rare condition, centralization gives the patient more chances of a successful outcome. It can be argued that there are no evidence-based data supporting this proposal, but decisions based on subjective experience are still legitimate in rare conditions.2 These repairs in multioperated patients involve complex and heavy surgery, require the best imaging, anesthetic and intensive care support, and are better dealt with in large referral centers with an interest on esophageal diseases. Our series reveal that esophageal removal and replacement was selected as a suitable option after failure/s in previous repairs in several cases, but it was only one of many other operations necessary. Our results are more than acceptable, but the risks involved, the amount of suffering, the cost, and some permanent sequelae, such as chronic respiratory disease and growth stunting invite to strongly support that “preserving the own esophagus at all costs” and trying individually all new procedures in small individual or institutional series do not serve the interest of the patient better than early referral.
Conclusion Unsatisfactory repair of EA or repeated complications after it, prompt radical surgical attitudes directed to reestablish the patency of the digestive tract, close any fistula, and provide the patient with the best possible outcome. Several operations are generally necessary to achieve this goal and sometimes, but not always, the native esophagus is better discarded and replaced. Many complications should be expected, but the end result can be good. These patients should probably benefit of being referred earlier to centers with large experience in the management of complex esophageal conditions.
Conflict of Interest None.
References 1 Spitz L. Oesophageal atresia. Orphanet J Rare Dis 2007;2:24 2 Tovar JA, Fragoso AC. Current controversies in the surgical treat-
ment of esophageal atresia. Scand J Surg 2011;100(4):273–278 3 Puri P, Khurana S. Delayed primary esophageal anastomosis
for pure esophageal atresia. Semin Pediatr Surg 1998;7(2): 126–129 4 Sri Paran T, Decaluwe D, Corbally M, Puri P. Long-term results of delayed primary anastomosis for pure oesophageal atresia: a 27year follow up. Pediatr Surg Int 2007;23(7):647–651 5 Foker JE, Linden BC, Boyle EM Jr, Marquardt C. Development of a true primary repair for the full spectrum of esophageal atresia. Ann Surg 1997;226(4):533–541, discussion 541–543 6 Kimura K, Nishijima E, Tsugawa C, et al. Multistaged extrathoracic esophageal elongation procedure for long gap esophageal atresia: Experience with 12 patients. J Pediatr Surg 2001;36(11): 1725–1727 European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
25
Downloaded by: University of Georgia Libraries. Copyrighted material.
Tertiary Surgery for Complicated Repair of EA
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
7 Spitz L. Gastric transposition in children. Semin Pediatr Surg 2009; 8 9
10
11
12 13
18(1):30–33 Hamza AF. Colonic replacement in cases of esophageal atresia. Semin Pediatr Surg 2009;18(1):40–43 Foker JE, Kendall Krosch TC, Catton K, Munro F, Khan KM. Long-gap esophageal atresia treated by growth induction: the biological potential and early follow-up results. Semin Pediatr Surg 2009; 18(1):23–29 Sroka M, Wachowiak R, Losin M, et al. The Foker technique (FT) and Kimura advancement (KA) for the treatment of children with longgap esophageal atresia (LGEA): lessons learned at two European centers. Eur J Pediatr Surg 2013;23(1):3–7 Friedmacher F, Puri P. Delayed primary anastomosis for management of long-gap esophageal atresia: a meta-analysis of complications and long-term outcome. Pediatr Surg Int 2012;28(9): 899–906 Loukogeorgakis SP, Pierro A. Replacement surgery for esophageal atresia. Eur J Pediatr Surg 2013;23(3):182–190 Gallo G, Zwaveling S, Groen H, Van der Zee D, Hulscher J. Long-gap esophageal atresia: a meta-analysis of jejunal interposition, colon
14 15 16
17 18
19
20
interposition, and gastric pull-up. Eur J Pediatr Surg 2012;22(6): 420–425 Spitz L, Kiely E, Sparnon T. Gastric transposition for esophageal replacement in children. Ann Surg 1987;206(1):69–73 Hendren WH, Hendren WG. Colon interposition for esophagus in children. J Pediatr Surg 1985;20(6):829–839 Stone MM, Fonkalsrud EW, Mahour GH, Weitzman JJ, Takiff H. Esophageal replacement with colon interposition in children. Ann Surg 1986;203(4):346–351 Mitchell IM, Goh DW, Roberts KD, Abrams LD. Colon interposition in children. Br J Surg 1989;76(7):681–686 Burgos L, Barrena S, Andrés AM, et al. Colonic interposition for esophageal replacement in children remains a good choice: 33-year median follow-up of 65 patients. J Pediatr Surg 2010;45(2):341–345 Bruch SW, Hirschl RB, Coran AG. The diagnosis and management of recurrent tracheoesophageal fistulas. J Pediatr Surg 2010;45(2): 337–340 Coran AG. Redo esophageal surgery: the diagnosis and management of recurrent tracheoesophageal fistula. Pediatr Surg Int 2013;29(10):995–999
Downloaded by: University of Georgia Libraries. Copyrighted material.
26
European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Tertiary Surgery for Complicated Repair of Esophageal Atresia Ruben Ortiz1 Alba Sánchez Galán1 Leopoldo Martinez1 Manuel Lopez Santamaria1 Juan Antonio Tovar1 1 Department of Pediatric Surgery, Hospital Universitario La Paz,
Madrid, Spain
Eva Dominguez1
Francisco Hernández1
Address for correspondence Ruben Ortiz, MD, Department of Pediatric Surgery, La Paz Universitary Hospital, Paseo de la Castellana 261, Madrid 28046, Spain (e-mail: [email protected]).
Eur J Pediatr Surg 2015;25:20–26.
Abstract
Keywords
► esophageal atresia ► esophageal replacement ► long-gap ► elongation ► complications
received May 15, 2014 accepted June 23, 2014 published online August 21, 2014
Aim The ideal repair of esophageal atresia (EA) is primary anastomosis with closure of the fistula if present. Long gap or local circumstances prompt other procedures that occasionally lead to disastrous complications. The aim of this study was to analyze the management of these complications in a tertiary referral center. Patients and Methods A retrospective review of patients treated for EA between 1993 and 2013 was conducted. Both the patients were primarily treated by us, and referrals from elsewhere after two or more failed operations were included. Results In total, 23 patients were included (3/176 cases of EA treated primarily by us and 20 referrals). Of the 23 patients, 6 had type I EA, 15 type III (four long gaps), 1 type IV, and 1 type V. Cardiac anomalies were associated in seven cases, duodenal atresia in three, and Down syndrome in two patients. Primary anastomosis was initially achieved in 12 patients. Primary or secondary Foker lengthening was used in seven cases. The causes of the failure were anastomotic leaks in nine, unmanageable strictures in seven, and refistulization in five patients. These patients required 66 reoperations (median of 3 [2–7]) before inclusion in the study. Radical tertiary treatment consisted of 15 esophageal replacements (11 colonic grafts and 4 gastric pull-ups), and 1 esophageal-gastric disconnection. Five patients previously treated with esophageal replacement and referred for graft problems required 13 interventions. Two families did not give consent for one replacement and one disconnection. Complications appeared in 12 patients, and 9 additional operations were required in 7 patients. With a follow-up of 31 months (range, 4–139 months) 15 patients take all their meals per os, 5 occasionally use the gastrostomy, and 2 and 1 are fed exclusively via gastrostomy or jejunostomy. All tracheoesophageal fistulas were closed, but 15 cases are below p3 for weight and 12 for height. Three patients (13%) ultimately died 32 months (range, 9–56 months) after the first operation (due to aspiration in one, and for causes unrelated to it in the other two [tracheostomy obstruction and Guillain–Barré syndrome]). Conclusions When repeated complications appear after EA repair, radical surgical attitudes may be justified. If esophageal continuity cannot be reestablished, the native esophagus may have to be discarded and replaced. Many complications should be expected, but the end result can be good. These patients should be referred to centers with large experience in the management of this complex condition.
© 2015 Georg Thieme Verlag KG Stuttgart · New York
DOI http://dx.doi.org/ 10.1055/s-0034-1386645. ISSN 0939-7248.
Downloaded by: University of Georgia Libraries. Copyrighted material.
20
Tertiary Surgery for Complicated Repair of EA
Esophageal atresia (EA) occurs in 1 of every 2,500 to 3,000 live births,1 and represents a paradigm of pediatric surgery because the results of its treatment are an index of the quality of the care provided by individuals and instituions.2 The aim of EA repair is to reestablish esophageal patency by performing a tension-free, primary anastomosis of both ends of the native organ, with closure of the tracheoesophageal fistula (TEF) if present. This is possible in most cases because the atretic proximal esophageal pouch is more or less close to the fistula that communicates the trachea with the stomach. Pediatric surgeons learned along the years that, even under some tension, primary anastomosis could be done in most cases and that only in a few of them, both ends are too far apart to achieve primary repair. These “long-gap” cases and the 10% that have “pure” EA (proximal atresia of the esophagus without fistula) are not amenable to neonatal primary anastomosis and require other procedures. Delayed primary anastomosis,3,4 esophageal lengthening procedures,5,6 and esophageal replacement7,8 are therefore mandatory in a limited proportion of cases in which the risks involved are high and the complications frequent and severe. These, and the few cases in which primary repair fails because of massive leaks or recurrent TEF leading to repeated unsuccessful operations, represent major reconstructive challenges. Every pediatric surgeon considers himself/herself entitled to perform all the surgical procedures related to EA and TEF. Our large experience in the field has shown us along the years that much of the morbidity and mortality of this condition are hidden in unpublished small series, and we felt that it might be useful to show that only after risky operations in tertiary
21
centers and at the price of many complications, most of these unfortunate patients can be put back into track. The aim of this study was to analyze the management of a large group of these patients at a tertiary referral center with particular interest on esophageal conditions.
Patients and Methods A retrospective review of patients treated at our Institution for EA between 1993 and 2013 was conducted. Only children having two or more failed major operations were included in this study and this cohort involved both children primarily treated by us and referred from elsewhere. Clinical data, associated comorbidities, initial procedure, complications, reoperations, and long-term outcome were recorded and analyzed. Stricture dilations were not taken into account as initial reoperations, but they are discussed in the Results section. The current status of the patient was assessed in every case by imaging and endoscopic procedures. The viability of direct repair or the need for discarding and replacing the esophagus was individually judged on the basis of careful examination. The aims of the radical procedure were to close fistulas, to reestablish patent anastomoses, or to remove and replace the esophagus if it was deemed unsalvageable. In these cases, it was resected except when this involved too much risk. Procedures previously described were used for replacement with either the colon or the stomach. Transhiatal route was preferred, but it was not always usable. Operations on previously performed replacements were done transsternally. Previously placed gastrostomies or new ones were liberally used in the postoperative period and closed as soon as possible.
Table 1 First repair interventions performed in the 23 patients of the study with complicated EA First repair procedure
n
Type of EA (number)
Thoracotomy, fistula ligation, and primary anastomosis
12
type III (11); type IV (1)
Gastrostomy and fistula ligation
1
type III long-gap (1)
Gastrostomy an delayed primary anastomosis
1
type I (1)
Gastrostomy
5
type I (4); type V (1)
Foker procedure
2
type III long-gap (2)
Gastrostomy þ esophagostomy
1
type I (1)
Gastrostomy þ dynamic stump-lengthening technique
1
type I (1)
Complications after first repair Major leaks
9
Stricture
7
Refistulization
5
Aspiration and pneumonia
7
Gastrostomy–esophagostomy problems
1
Chylothorax and chylous ascites
1
Abbreviation: EA, esophageal atresia. Note: Total times of each procedure, type of EA and posterior complications after first attempt of repair are shown. European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Downloaded by: University of Georgia Libraries. Copyrighted material.
Introduction
Ortiz et al.
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
Patients were seen at the outpatient clinic every 6 months. Height and weight were recorded at that time and nutritional assessment was done if required. Data are reported as median (range). Imaging and endoscopic assessment of the results were performed as appropriate.
Results In total, 23 patients (15 males/8 females) were included in this review. Of the 23 patients, 3 were treated primarily by ourselves (from 176 cases of EA treated at Hospital Universitario La Paz over the last 20 years), and the remaining 20 (86%) patients were referred from other institutions after they had been unsuccessfully treated. Median age at birth was 35.5 weeks (range, 31–40 weeks), with a mean weight at birth of 2, 085 g (range, 1,200–2,900 g). Six patients had pure EA (type I), 15 failed repair of type III (in four patients considered “long-gap”), and 1 case of type IV and type V each. Associated anomalies were present in 17 patients: 7 major cardiac defects, 3 VACTERL associations, 3 duodenal atresias, 2 Down syndromes, 1 Goldenhar syndrome and other malformations (Morgagni hernia, Meckel diverticulum, congenital subglottic stenosis, and ear agenesis) in 1 patient each. The first attempt at repair had been performed by thoracotomy in 21 cases and by thoracoscopy in 2 cases. Fistula closure and primary anastomosis were initially achieved in 12 patients, Foker lengthening procedure in 2, delayed primary anastomosis in 1, dynamic stump-lengthening techniques in 1, and gastrostomy þ esophagostomy or fistula ligation in the remaining 7 patients. The causes of the failure were major anastomotic leaks in nine, unmanageable strictures in seven, and repeated refistulization in five patients. Massive reflux with aspiration and pneumonia was also present in seven cases after the first procedure, and one patient had chylothorax and chylous ascites(►Table 1). These patients had a median of 3 (range, 2–7) reoperations (total of 66) before inclusion in the study, on top of esophageal dilations that were necessary in 9 (up to 57 procedures). Overall, 14 Nissen fundoplications, 12 thoracotomies for recurrent fistula closure or redoanastomosis, 5 secondary Foker lengthening techniques, and 3 repeated endoscopic attempts at recurrent TEF closure had been performed. Five esophageal replacements (four colonic [three retroesternal, one mediastinal] and one Schärli procedure) had been performed elsewhere and ultimately referred to us for complications with the graft (►Table 2). All Foker lengthenings and attempts at endoscopic fistula closure had been performed elsewhere. Radical tertiary treatment at an age of 20.5 months (range, 5–70 months) consisted of 15 of 23 (65%) esophageal resections and replacements (11 colonic grafts [8 intramediastinal and 3 retroesternal] and 4 gastric pull-ups) and one esophageal-gastric disconnection in a patient with intractable reflux. The five patients previously treated with esophageal replacement and referred for graft problems were evaluated and required 13 interventions (three redo esophageal-graft anastomosis, four pyloroplasties, two gastro-duodenostoEuropean Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Table 2 Reoperations performed after initial repair before inclusion in the study Procedure
n
Thoracotomy and esophageal redo-anastomosis
4
Thoracotomy and recurrent TEF ligation
8
Gastrostomy esophagostomy
16
Foker lengthening procedure
5
Nissen fundoplication
14
Endoscopic techniques for recurrent TEF closure
3
Esophagostomy dilation
1
Cervicotomy and proximal fistula ligation
2
Thoracotomy and lobectomy
1
Laparotomy
2
Tracheostomy
1
Feeding jejunostomy
1
Esophagectomy
1
Endoscopic dilation of esophageal anastomotic stricturea
57 in 9 patients
Mediastinal colonic replacement
1
Retroesternal colonic replacement
3
Schärli procedure
1
Redo esophageal-graft anastomosis
1
Endoprosthesis stent for esophageal-graft stricture
1
Abbreviation: TEF, tracheoesophageal fistula. Note: Total of 66 in 23 patients, with a median of 3 (2–7). a Total of reoperations (n ¼ 66), excluding esophageal stricture dilations (57 procedures in 9 patients).
mies, one esophagostomy, and one mucocele esophagectomy). One presternal coloplasty after another stenoticischemic colonic graft performed elsewhere was achieved, and one feeding jejunostomy in a patient who did not consent to esophageal-gastric disconnection. Finally, of the remaining two patients, one family did not give consent for replacement, and the other patient with a Down syndrome and had three Nissen procedures (►Fig. 1). Complications appeared in 12 of 15 patients (80%) (anastomotic stricture in four, salivary fistula in five, delayed gastric emptying in two, severe reflux in three, and intestinal obstruction in two patients), and 9 additional operations were required in 7 excluding stricture dilations (27 procedures in 8 patients). These reoperations after radical treatment consisted in four redo esophageal-graft anastomosis, one gastro-duodenostomy, two laparotomy for intestinal obstruction, one esophagostomy, and one incisional hernia repair) (►Table 3). We have seen that patients who had the association of EA and duodenal atresia (n ¼ 3) needed a high number of
Downloaded by: University of Georgia Libraries. Copyrighted material.
22
Ortiz et al.
23
Downloaded by: University of Georgia Libraries. Copyrighted material.
Tertiary Surgery for Complicated Repair of EA
Fig. 1 Tertiary treatment for 23 patients with complicated esophageal atresia treated at Hospital La Paz in the period of 1993 to 2013. Radical treatments, complications, and outcomes are summarized.
reoperations due to secondary complications. One of them received three thoracotomies for esophageal reanastomosis elsewhere and finally we had to replace it. In another case, we opted for an esophageal-gastric disconnection after two failed Nissen fundoplications with persistent intractable reflux; and the third patient who has Down syndrome, had three Nissen procedures and is fed by gastrostomy. With a follow-up of 31 months (range, 4–139 months), most patients (15/23) take all their meals per os, five use the gastrostomy only occasionally, and two and one are fed exclusively via gastrostomy or jejunostomy, respectively. All TEFs were closed, but 15 of 23 cases are below p3 for weight and 12 of 23 below p3 for height. Three patients (13%) ultimately died 32 months (range, 9–56 months) after opera-
tion, because of aspiration in one, and unrelated causes in the other two patients (tracheostomy obstruction and Guillain– Barré syndrome).
Discussion Repair of EA still remains a challenge for the pediatric surgeon because not all patients have the same spectrum of disease. Nowadays, mortality in EA is rare and survival is mainly determined by the presence of associated anomalies; however, the outcome will also depend on the nature of the surgical procedures performed during the first attempt at repair and this is particularly true for the “difficult cases.” It is widely accepted that the own child’s esophagus is the best material for European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
Table 3 Radical-tertiary treatment performed at our institution for 23 patients with complicated EA Radical tertiary treatment (total 21/23 patients)
Procedures (n)
1. Esophageal resection and replacement (15 patients)
15
• Mediastinal colonic graft
8
• Retroesternal colonic graft
3
• Gastric pull-up
4
2. Patients referred with graft problems after esophageal replacement elsewhere (5 patients
13
• Redo esophageal-graft anastomosis,
3
• Pyloroplasty
4
• Gastro-duodenostomy
2
• Esophagsotomy
1
• Esophagectomy (mucocele)
1
• Pre-esternal coloplasty
1
• Feeding jejunostomy
1
3. Esophageal-gastric disconnection (1 patient)
1
Complications after radical-tertiary treatment (12 patients)
n
Anastomotic stricture
4
Salivary fistula
5
Delayed gastric emptying
2
Severe reflux
3
Aspiration and pneumonia
3
Intestinal obstruction
2
Dumping syndrome
1
Anastomotic dehiscence
1
Additional reoperations (7 patients)
Number of procedures (n)
Redo esophageal-colonic anastomosis
3
Redo esophageal-gastric anastomosis
1
Gastro-duodenostomy
1
Esophagostomy
1
Laparotomy for intestinal obstruction
2
Incisional hernia repair
1
Dilation of anastomotic esophageal-graft stricture
a
27 in 8 patients
Note: Type and number of procedures performed, complications appeared and additional reoperations needed are detailed. a Total additional reoperations after tertiary treatment (n ¼ 9), excluding anastomotic stricture dilations.
repair. However in “long-gap” or no-esophagus situations, to maintain the aim of obtaining a primary anastomosis at any price may lead to repeated operations and complications that can have serious long-term consequences for the child. Different procedures have been proposed for owner-establishing digestive patency with esophageal material. Kimura multistage extra thoracic esophageal elongation6 and particularly the Foker traction lengthening procedure5,9 had been used for achieving primary anastomoses even in extreme cases in which esophageal ends are too separated or in pure EA with a minimum distal esophageal nubbin. For achieving this goal, these techniques need repeated traction procedures (extrathoracic in Kimura and intrathoracic in Foker) to get esophageal lengthening ultimately allowing European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
anastomosis at any price. This can be indeed achieved in many cases, but the price to pay, in terms of functional quality of the reconstructed organ, may be too high.2 In a series of extrathoracic esophageal elongation,6 11 of 12 (91.6%) patients required a Nissen fundoplication and all received several dilations for stricture and even stricture resection in three (25%) patients. In the largest series of intrathoracic esophageal lengthening,9 traction sutures had to be relocated in several patients during the postoperative period involving several thoracotomies. All patients required Nissen fundoplication (42/42) due to the constant presence of reflux and 12 of 42 (28.5%) had a redofundoplication. Dilations for stricture and resection (with more thoracotomies) were necessary in many cases. In addition, 15 patients had patches of gastric
Downloaded by: University of Georgia Libraries. Copyrighted material.
24
metaplasia above the gastroesophageal junction. Both lengthening techniques at two different European centers were recently compared10 in a limited group of patients. The results were only fair, with long periods of traction to achieve the anastomosis, the need for repeated thoracotomies and severe complications in three patients (two finally need an esophageal replacement). The authors concluded that more data are needed (they suggested a multicenter registry) to elucidate the efficacy and safety of each elongation procedures. Other authors prefer a delayed primary anastomosis in pure EA3,4 based on the spontaneous growth of the upper pouch of the esophagus without any mechanical traction, because of swallowing and reflux into the lower pouch. They observed that this maximal growth occurred during the first 8 to 12 weeks of life, suggesting that the ideal time for delayed primary anastomosis was approximately 12 weeks of age. In the study by Sri Paran et al,4 a total of 21 patients with pure EA with a median initial gap of 3.7 cm received a delayed primary anastomosis. However, this procedure implied a late repair (median, 80 days) with a prolonged hospitalization period (median of 5.5 months) and continuous aspiration of the upper pouch. Moreover, symptomatic gastroesophageal reflux (GER) occurred in 66% of cases, Nissen fundoplication was required in 43% of cases, repeated stricture dilations were necessary in 47%, and stricture resection with redoanastomosis was eventually performed in 28.5% of these children. In a recent meta-analysis of complications and long-term outcome of this procedure comparing 44 articles with 451 newborns treated with this procedure,11 the incidence of postoperative interventions were as follows: stricture dilation in 60% of cases, stricture resection and reanastomosis in 21%, fundoplication for symptomatic GER in 36%, and esophageal replacement for unsatisfactory results in 14%. Authors highlighted that the incidence of GER and strictures requires early intervention to prevent feeding problems. The available alternative to preserving the native esophagus at any price is esophageal replacement. It can be considered primarily in log-gap situations and whenever survival is at risk after several attempts to preserve the patency of native esophagus.1 The ideal esophageal substitute should function as closely as possible to the original one, and it should work well during the many years without early deterioration.7,12,13 Gastric transposition “gastric pull-up”7,14 and colonic interposition8,15–18 are the two most extended procedures. Satisfactory results in terms of quality of life and normal feeding pattern without digestive or respiratory symptoms are possible,7,8,18 but the price to pay is an additional major operative risk with frequent and potentially serious complications.12,13 Another important scenario met after repair of EA is recurrent TEF that occurs in approximately 5 to 10% of cases, and involves difficult reoperations. Repeated thoracotomies and/or endoscopic closure techniques generate scarring and inflammation that add further difficulties to repair,19,20 cause prolonged hospital stays, and threaten quality of life. When facing a difficult repair, pediatric surgeons have to decide whether they can deal with all the possible scenarios
Ortiz et al.
involved or, alternatively, whether they will benefit more the patient by referring him to a tertiary institution. The experience summarized in this study suggests that most of them face one of these problems rarely and that, like for any other rare condition, centralization gives the patient more chances of a successful outcome. It can be argued that there are no evidence-based data supporting this proposal, but decisions based on subjective experience are still legitimate in rare conditions.2 These repairs in multioperated patients involve complex and heavy surgery, require the best imaging, anesthetic and intensive care support, and are better dealt with in large referral centers with an interest on esophageal diseases. Our series reveal that esophageal removal and replacement was selected as a suitable option after failure/s in previous repairs in several cases, but it was only one of many other operations necessary. Our results are more than acceptable, but the risks involved, the amount of suffering, the cost, and some permanent sequelae, such as chronic respiratory disease and growth stunting invite to strongly support that “preserving the own esophagus at all costs” and trying individually all new procedures in small individual or institutional series do not serve the interest of the patient better than early referral.
Conclusion Unsatisfactory repair of EA or repeated complications after it, prompt radical surgical attitudes directed to reestablish the patency of the digestive tract, close any fistula, and provide the patient with the best possible outcome. Several operations are generally necessary to achieve this goal and sometimes, but not always, the native esophagus is better discarded and replaced. Many complications should be expected, but the end result can be good. These patients should probably benefit of being referred earlier to centers with large experience in the management of complex esophageal conditions.
Conflict of Interest None.
References 1 Spitz L. Oesophageal atresia. Orphanet J Rare Dis 2007;2:24 2 Tovar JA, Fragoso AC. Current controversies in the surgical treat-
ment of esophageal atresia. Scand J Surg 2011;100(4):273–278 3 Puri P, Khurana S. Delayed primary esophageal anastomosis
for pure esophageal atresia. Semin Pediatr Surg 1998;7(2): 126–129 4 Sri Paran T, Decaluwe D, Corbally M, Puri P. Long-term results of delayed primary anastomosis for pure oesophageal atresia: a 27year follow up. Pediatr Surg Int 2007;23(7):647–651 5 Foker JE, Linden BC, Boyle EM Jr, Marquardt C. Development of a true primary repair for the full spectrum of esophageal atresia. Ann Surg 1997;226(4):533–541, discussion 541–543 6 Kimura K, Nishijima E, Tsugawa C, et al. Multistaged extrathoracic esophageal elongation procedure for long gap esophageal atresia: Experience with 12 patients. J Pediatr Surg 2001;36(11): 1725–1727 European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
25
Downloaded by: University of Georgia Libraries. Copyrighted material.
Tertiary Surgery for Complicated Repair of EA
Tertiary Surgery for Complicated Repair of EA
Ortiz et al.
7 Spitz L. Gastric transposition in children. Semin Pediatr Surg 2009; 8 9
10
11
12 13
18(1):30–33 Hamza AF. Colonic replacement in cases of esophageal atresia. Semin Pediatr Surg 2009;18(1):40–43 Foker JE, Kendall Krosch TC, Catton K, Munro F, Khan KM. Long-gap esophageal atresia treated by growth induction: the biological potential and early follow-up results. Semin Pediatr Surg 2009; 18(1):23–29 Sroka M, Wachowiak R, Losin M, et al. The Foker technique (FT) and Kimura advancement (KA) for the treatment of children with longgap esophageal atresia (LGEA): lessons learned at two European centers. Eur J Pediatr Surg 2013;23(1):3–7 Friedmacher F, Puri P. Delayed primary anastomosis for management of long-gap esophageal atresia: a meta-analysis of complications and long-term outcome. Pediatr Surg Int 2012;28(9): 899–906 Loukogeorgakis SP, Pierro A. Replacement surgery for esophageal atresia. Eur J Pediatr Surg 2013;23(3):182–190 Gallo G, Zwaveling S, Groen H, Van der Zee D, Hulscher J. Long-gap esophageal atresia: a meta-analysis of jejunal interposition, colon
14 15 16
17 18
19
20
interposition, and gastric pull-up. Eur J Pediatr Surg 2012;22(6): 420–425 Spitz L, Kiely E, Sparnon T. Gastric transposition for esophageal replacement in children. Ann Surg 1987;206(1):69–73 Hendren WH, Hendren WG. Colon interposition for esophagus in children. J Pediatr Surg 1985;20(6):829–839 Stone MM, Fonkalsrud EW, Mahour GH, Weitzman JJ, Takiff H. Esophageal replacement with colon interposition in children. Ann Surg 1986;203(4):346–351 Mitchell IM, Goh DW, Roberts KD, Abrams LD. Colon interposition in children. Br J Surg 1989;76(7):681–686 Burgos L, Barrena S, Andrés AM, et al. Colonic interposition for esophageal replacement in children remains a good choice: 33-year median follow-up of 65 patients. J Pediatr Surg 2010;45(2):341–345 Bruch SW, Hirschl RB, Coran AG. The diagnosis and management of recurrent tracheoesophageal fistulas. J Pediatr Surg 2010;45(2): 337–340 Coran AG. Redo esophageal surgery: the diagnosis and management of recurrent tracheoesophageal fistula. Pediatr Surg Int 2013;29(10):995–999
Downloaded by: University of Georgia Libraries. Copyrighted material.
26
European Journal of Pediatric Surgery
Vol. 25
No. 1/2015
