J Gastrointest Surg (2017) 21:927–935

DOI 10.1007/s11605-017-3423-0

ORIGINAL ARTICLE

Postoperative Complications and Functional Outcome

after Esophageal Atresia Repair: Results from Longitudinal
Single-Center Follow-Up
Florian Friedmacher 1 & Birgit Kroneis 1 & Andrea Huber-Zeyringer 1 & Peter Schober 1 &
Holger Till 1 & Hugo Sauer 1 & Michael E. Höllwarth 1

Received: 2 December 2016 / Accepted: 3 April 2017 / Published online: 19 April 2017
# 2017 The Society for Surgery of the Alimentary Tract

Abstract
Background Esophageal atresia (EA) and tracheoesophageal fistula (TEF) represent major therapeutic challenges, frequently
associated with serious morbidities following surgical repair. The aim of this longitudinal study was to assess temporal changes in
morbidity and mortality of patients with EA/TEF treated in a tertiary-level center, focusing on postoperative complications and
their impact on long-term gastroesophageal function.
Methods One hundred nine consecutive patients with EA/TEF born between 1975 and 2011 were followed for a median of 9.6
years (range, 3–27 years). Comparative statistics were used to evaluate temporal changes between an early (1975–1989) and late
(1990–2011) study period.
Results Gross types of EA were A (n = 6), B (n = 5), C (n = 89), D (n = 7), and E (n = 2). Seventy (64.2%) patients had coexisting
anomalies, 13 (11.9%) of whom died before EA correction was completed. In the remaining 96 infants, surgical repair was
primary (n = 66) or delayed (n = 25) anastomosis, closure of TEF in EA type E (n = 2), and esophageal replacement with colon
interposition (n=2) or gastric transposition (n=1). Long-gap EA was diagnosed in 23 (24.0%) cases. Postoperative mortality was
4/96 (4.2%). Overall survival increased significantly between the two study periods (42/55 vs. 50/54; P = 0.03). Sixty-nine
(71.9%) patients presented postoperatively with anastomotic strictures requiring a median of 3 (range, 1–15) dilatations.
Revisional surgery was required for anastomotic leakage (n = 5), recurrent TEF with (n = 1) or without (n=9) anastomotic
stricture, undetected proximal TEF (n = 4), and refractory anastomotic strictures with (n = 1) or without (n = 2) fistula. Normal
dietary intake was achieved in 89 (96.7%) patients, while 3 (3.3%) remained dependent on gastrostomy feedings. Manometry
showed esophageal dysmotility in 78 (84.8%) infants at 1 year of age, increasing to 100% at 10-year follow-up. Fifty-six (60.9%)
patients suffered from dysphagia with need for endoscopic foreign body removal in 12 (13.0%) cases. Anti-reflux medication
was required in 43 (46.7%) children and 30 (32.6%) underwent fundoplication. The rate of gastroesophageal reflux increased
significantly between the two study periods (29/42 vs. 44/50; P = 0.04). Twenty-two (23.9%) cases of endoscopic esophagitis and
one Barrett’s esophagus were identified.
Conclusions Postoperative complications after EA/TEF repair are common and should be expertly managed to reduce the risk of
long-term morbidity. Regular multidisciplinary surveillance with transitional care into adulthood is recommended in all patients
with EA/TEF.

Keywords Esophageal atresia . Tracheoesophageal fistula .

Complications . Follow-up . Outcome . Gastroesophageal
function

* Florian Friedmacher
florian.friedmacher@nhs.net Introduction

1
Department of Pediatric and Adolescent Surgery, Medical University Esophageal atresia (EA) and tracheoesophageal fistula (TEF)
of Graz, Auenbruggerplatz 34, 8036 Graz, Austria represent a spectrum of relatively rare and complex congenital
928 J Gastrointest Surg (2017) 21:927–935

malformations, resulting from disruptions during normal fore- Preoperative assessment for associated anomalies composed
gut separation.1 International prevalence rates of EA/TEF cur- of a full physical examination, plain chest and abdominal x-
rently range between 1.27 and 4.55 per 10,000 births.2–4 Due ray, and in more recent years abdominal ultrasound and echo-
to the complex nature and infrequency of cases, EA/TEF man- cardiography were included into the routine diagnostic work-
agement remains a major therapeutic challenge for most pedi- up. All pediatric surgeons that operated during the study peri-
atric surgeons and other involved specialists.5, 6 In most of the od followed a standard practice protocol. In general, surgery
current reports, prematurity, low birth weight, and additional was scheduled within 48 h after birth except in unstable neo-
congenital anomalies are associated with higher mortality and nates with respiratory distress that required emergency TEF
further complicate the care of these patients,7, 8 whereas sur- ligation. Prior to surgical repair, a tracheobronchoscopy was
vival rates in full-term infants with no other abnormalities can performed in order to ascertain the presence and exact location
approach 100% after EA/TEF surgery.9, 10 Today, improved of a proximal/distal fistula or other anticipated defects. The
survival of newborns with EA/TEF is likely related to multiple standard operative approach was a right posterolateral thora-
factors including advances in neonatal intensive care and an- cotomy via the 4th intercostal space adopting an extrapleural
esthesia, refined surgical techniques, parenteral nutrition, and access, whereas in cases where a right-sided aortic arch was
antibiotics.11, 12 However, early13–15 and late16–18 postopera- found preoperatively, a left thoracotomy was chosen. The type
tive morbidities frequently occur after initial repair of EA/TEF of EA was classified according to the Gross classification de-
despite excellent surgical and neonatal management, and can pending on the pre- and intraoperative anatomy (Fig. 1).
be associated with impaired outcomes. Early recognition and Long-gap EA was defined as a gap between the proximal
treatment of potential complications is therefore essential in and distal segments under tension of greater than four verte-
order to prevent poor long-term results.19–22 At present, there bral bodies. In EA types C and D, following initial TEF liga-
is limited published data from large series of EA/TEF patients tion and mobilization of both esophageal ends, a primary
following surgery with regard to adverse events and their in- single-layer esophagoesophagostomy was attempted.
fluence on functional outcome later in life.24 Thus, the aim of Patients with EA types A and B, and those cases in which a
this longitudinal study was to assess temporal changes in mor- primary anastomosis was not achievable, received a feeding
bidity and mortality of patients with EA/TEF treated in a gastrostomy on the first or second day of life. After a period of
tertiary-level center, focusing on postoperative complications continuous upper pouch suction with or without serial
and their impact on long-term gastroesophageal function. bougienage of the upper and lower pouch, a delayed primary
repair was performed where possible. In those infants, where
an end-to-end anastomosis was still not feasible, esophageal
Patients and Methods replacement was undertaken. In neonates with EA type E, the
isolated TEF was closed through a right cervical incision.
Study Population and Design Early postoperative complications and late sequelae after
EA/TEF repair were reviewed and where possible defined
This was a longitudinal cohort study of all patients with EA/ according to an international consensus paper,25 and conser-
TEF that were born in a tertiary-level children’s hospital be- vative or surgical treatment used was recorded. Standardized
tween 1 January 1975 and 31 December 2011. Each patient follow-up was undertaken in a specialized EA/TEF clinic
was identified by using the hospital inpatient enquiry system consisting of regular physical examination as well as esopha-
and relevant data was extracted from the individual medical geal manometry, 24-h pH monitoring, contrast swallow study,
and operative records. No patient was excluded from this and upper gastrointestinal endoscopy at least once within the
study, which followed institutional ethical committee approval first year of life. All examinations were repeated at 5-year
(EK 27–259 ex 14/15). In 1990, a new head of department intervals. A manometry catheter with 36 solid-state sensors
was employed, who established a specialized pediatric surgi- was used, and after a 5-min resting period, ten wet swallows
cal intensive care unit including new staff members with in- were performed with the patient in supine position. Integrity
creased expertise in the management of newborns with com- of esophageal peristalsis and contraction pattern was catego-
plex congenital malformations. In order to compare potential rized as intact (i.e. propagating), weak, or absent, and lower
changes over time, cases were divided into an early (1975– esophageal sphincter pressure was assessed. A pH probe was
1989) and late (1990–2011) study period. positioned 3 cm above the upper boarder of the lower esoph-
General patients’ characteristics were collected on gender, ageal sphincter and connected to a recorder system.
maternal age, prenatal findings, gestational age, birth weight Gastroesophageal reflux disease (GERD) was defined as re-
(BW), mode of delivery, and coexisting anomalies. Initial flux of gastric contents causing troublesome symptoms such
postnatal management included stabilizing procedures, intra- as recurrent regurgitation with or without vomiting, poor
venous fluid resuscitation, and drainage of the upper esopha- weight gain, irritability, heartburn, or coughing with confir-
geal pouch using an 8F or 10F Replogle suction tube. mation of acidic reflux episodes on 24-h pH monitoring
J Gastrointest Surg (2017) 21:927–935 929

Fig. 1 Types of esophageal atresia (EA) and tracheoesophageal fistula (TEF) according to the Gross classification. Reproduced and modified with
permission of Springer from23

(intraesophageal pH < 4.0) and upper gastrointestinal endos- Forty-two (38.5%) had major cardiac defects that required
copy with biopsy. The 2009 NASPGHAN/ESPGHAN guide- medical, surgical, or palliative treatment. A VACTERL asso-
lines were used as cutoff (i.e. a reflux index >4% was consid- ciation with at least three congenital malformations was diag-
ered as abnormal).26 nosed in a total of 11 (10.1%) newborns. There were no sta-
Those patients with GERD, recurrent respiratory infec- tistical differences in the incidence of BW <1500 g (10/55 vs.
tions, or other conditions such as swallowing disorders, neu- 7/54; P = 0.5989) and major cardiac defects (20/55 vs. 22/54;
rological impairment, or coexisting anomalies were seen as P = 0.6962) between the two study periods.
often as required, and additional surveillance was arranged
by the appropriate specialties. Survival

Statistical Analysis Overall survival was 92/109 (84.4%) with a significant in-
crease between the early and late study period (Table 2). In
All data was extracted into an electronic spreadsheet and sta- total, 13 newborns died before esophageal anastomosis or
tistical analysis was performed using PASW Statistics 18.0 replacement could be performed. One was a preterm neonate
software application (SPSS Inc., Chicago, USA). Results are with EA type C and multiple malformations that died on the
presented as median (range) or mean ± SE. Statistical differ- first day of life and did not receive any surgical treatment.
ences between the two study periods were analyzed using χ2, The remaining 12 newborns had low or very low birth weight
Fisher’s Exact, or Mann-Whitney U test where appropriate. A and received gastrostomy insertion, but died before esopha-
P value of <0.05 was considered statistically significant. geal continuity could be achieved at a median age of 3 days
(range, 1–33 days) due to severe cardiac defects (n = 7), ce-
rebrovascular hemorrhage (n = 3), and respiratory failure
Results (n = 2). Four (4.2%) of the 96 patients that underwent final
surgical repair died of cardiac failure (n = 2), severe aspira-
General Patients’ Characteristics tion pneumonia owing to a recurrent TEF (n = 1), and peri-
tonitis with subsequent multi-organ failure resulting from a
Overall, 109 consecutive patients with EA types A (n = 6; bowel perforation (n = 1) within a median time of 106 days
5.5%), B (n = 5; 4.6%), C (n = 89; 81.7%), D (n = 7; 6.4%), (range, 84–529 days) after successful primary (n = 2) or de-
and E (n = 2; 1.8%) were observed in this study. Sixty-two layed (n = 2) esophageal anastomosis. All deceased infants
(56.9%) were male and 47 (43.1%) female. Median maternal were born prematurely and had additional anomalies. There
age was 27 years (range, 18–47 years) with presence of were no statistical differences in postoperative mortality rates
polyhydramnios in 46 (42.2%) cases. Median gestational between the two study periods (3/45 vs. 1/51; P = 0.3379)
age was 37 weeks (range, 26–43 weeks) with 58 (53.2%) (Table 3).
infants born prematurely. Delivery was vaginal in 67
(61.5%) cases and 42 (38.5%) underwent Cesarean section. Surgical Management
Median BW was 2480 g (range, 700–3860 g) with 17 (15.6%)
neonates having a BW <1500 g. Coexisting anomalies were The final surgical repair of 96 patients operated for EA/TEF is
found in 70 (64.2%) patients and are illustrated in Table 1. displayed in Table 4. Ninety-one newborns underwent
930 J Gastrointest Surg (2017) 21:927–935

Table 1 Coexisting anomalies,

defects, and syndromes in 109 Cardiovascular malformations (n = 74) Gastrointestinal malformations (n = 18)
patients with EA/TEFa Atrial septal defect (ASD) (n = 18) Duodenal atresia (n = 9)
Ventricular septal defect (VSD) (n = 13) Malrotation (n = 8)
Right descending aorta (n = 13) Omphalocele (n = 1)
Pulmonary stenosis (n = 10)
ASD + VSD (n = 4) Anorectal malformations (n = 13)
Tetralogy of Fallot (n = 4) Imperforate anus without fistula (n = 8)
Single atrium (n = 3) Recto-perineal fistula (n = 5)
Coarctation of the aorta (n = 2)
Hypoplastic left heart syndrome (n = 2) Airway or palate malformations (n = 9)
Hypoplastic right heart syndrome (n = 1) Congenital diaphragmatic hernia (n = 3)
Persistent truncus arteriosus (n = 1) Cleft palate (n = 3)
Single ventricle (n = 1) Tracheoesophageal cleft (n = 2)
Tricuspid atresia (n = 1) Mandibulofacial dysostosis (n = 1)
Dextrocardia (n = 1)
Neurological defects (n = 8)
Skeletal abnormalities (n = 41)
Vertebral or rip defects (n = 25) Chromosomal anomalies and syndromes (n = 8)
Limb anomalies: Trisomy 21 (n = 3)
Radial aplasia (n = 6) Trisomy 18 (n = 2)
Supernumerary digit (n = 4) Pierre Robin syndrome (n = 1)
Club foot (n = 4) Potter’s syndrome (n = 1)
Thumb deformity (n = 2) Treacher Collins syndrome (n = 1)

Genitourinary anomalies (n = 41)

a
Patients may have had multiple malformations

primary (n = 66; 68.8%) or delayed (n = 25; 26.0%) end-to- gap EA was diagnosed in a total of 23 (24.0%) patients with
end esophageal anastomosis. Indications for 18 of the 25 de- Gross types A (n = 5), B (n = 5), C (n = 9), and D (n = 4). A
layed primary repairs were long-gap EA types A (n = 4), B primary esophagoesophagostomy was possible in 3 of the 23
(n = 5), C (n = 8), and D (n = 1), which were carried out after a cases with long-gap EA, 18 underwent delayed primary repair,
median period of 53 days (range, 14–175 days) with (n = 6) or and 2 required esophageal replacement. Overall, three infants
without (n = 12) longitudinal bougienage of upper and lower had esophageal replacement with colonic interposition (n = 2)
pouch. The remaining seven patients were preterm neonates or gastric transposition (n = 1) at a median age of 387 days
with EA types C (n = 6) and D (n = 1) that were physiologi- (range, 86–552 days). Indications were long-gap EA types A
cally unstable with severe respiratory distress requiring venti- (n = 1), C (n = 1), and recurrent TEF with mediastinal abscess
lator support. In these cases, an emergency TEF ligation and and anastomotic stricture in an infant with EA type C after
gastrostomy insertion was performed, followed by a delayed primary repair. No chyle leak, conduit necrosis, or vocal cord
primary anastomosis after a median stabilization time of 7 days injury occurred. The median length of initial hospitalization
(range, 3–9 days). Two newborns with EA type E had elective was 59 days (range, 10–351 days), which decreased signifi-
division and suture closure of their isolated TEF through a cantly between the two study periods (108 ± 13 vs. 59 ± 5 days;
right cervical incision on day 13 and 39, respectively. Long- P = 0.0160).

Table 2 Survival rates of all 109

patients with EA/TEF according Group Definition 1975–1989 (n = 55) 1990–2011 (n = 54) P value
to the Spitz classification
I BW > 1500 g, no major cardiac anomaly 33/35 (94.3%) 31/31 (100%) 0.49
II BW < 1500 g or major cardiac anomaly 7/10 (70.0%) 16/17 (94.1%) 0.13
III BW < 1500 g and major cardiac anomaly 2/10 (20.0%) 3/6 (50.0%) 0.30
Overall survival: 42/55 (76.4%) 50/54 (92.5%) 0.03
J Gastrointest Surg (2017) 21:927–935 931

Table 3 Mortality rates in 96

operated patients with EA/TEF Group Definition 1975–1989 (n = 45) 1990–2011 (n = 51) P value
according to the Spitz
classification I BW > 1500 g, no major cardiac anomaly 1/34 (2.9%) 0/31 (0.0%) 1.00
II BW < 1500 g or major cardiac anomaly 1/8 (12.5%) 1/17 (5.9%) 1.00
III BW < 1500 g and major cardiac anomaly 1/3 (33.3%) 0/3 (0.0%) 1.00
Postoperative mortality: 3/45 (6.7%) 1/51 (2.0%) 0.34

Postoperative Complications after initial surgical repair. Two of these suffered from severe
complications. One boy with EA type C, who had previously
Anastomotic Leakage been operated on for dehiscence of the primary esophageal
anastomosis, developed a recurrent distal fistula with medias-
Anastomotic leakage was suspected in 11 (11.5%) patients tinal abscess and anastomotic stricture. He required open re-
following primary (n = 8) or delayed (n = 3) esophageal anas- vision with cervical esophagostomy, feeding jejunostomy, and
tomosis and was eventually confirmed by chest x-ray with later underwent esophageal replacement with colonic interpo-
oral contrast: five in the early and six in the late study period, sition at the age of 17 months. The other one was a premature
respectively. Six of these were localized defects not requiring girl with EA type D and initially missed but then closed prox-
surgical revision and healed spontaneously following conser- imal fistula that died 106 days after primary esophageal repair
vative treatment with chest drainage, broad-spectrum antibi- of aspiration pneumonia before renewed surgical closure of a
otics, and parenteral nutrition support. The remaining five recurrent distal TEF could be performed. The remaining eight
were considered as major leaks with partial disruption of the patients had successful reoperation by thoracotomy and
esophagoesophagostomy and tension pneumothorax, and returned to normal dietary feeding following a failed attempt
were identified after a median time of 2 days (range, 1–5 days) of endoscopic treatment with Histoacryl or fibrin glue in five
postoperatively. In these cases, a rethoracotomy and suture of cases. Seven of these subsequently presented again with anas-
the end-to-end esophageal anastomosis was performed. tomotic strictures.
Overall, eight children with conservatively (n = 5) and surgi-
cally (n = 3) treated leaks later developed further complica- Undiagnosed Proximal TEF
tions: anastomotic strictures (n = 6) and recurrent distal TEF
with (n = 1) or without (n = 1) anastomotic stricture. An initially undiagnosed proximal fistula was found in four
(4.2%) patients after primary anastomosis for supposed EA
type C (n = 3) and following delayed primary repair for sup-
Recurrent Distal TEF
posed EA type A (n = 1) at a median of 74 days (range, 7–
122 days) postoperatively. The proximal TEF was not identi-
A recurrent distal TEF was identified in ten (10.4%) infants
fied during the initial tracheobronchoscopy preceding the
with EA types C (n = 8) and D (n = 2): six in the early and four
esophagoesophagostomy in three newborns, and in one no
in the late study period, respectively. They presented with
tracheobronchoscopy was performed. Each case of missed
feeding difficulties and respiratory symptoms ranging from
upper fistula or recurrent distal TEF was confirmed by repeat
coughing to cyanotic attacks due to recurrent respiratory in-
esophago-/tracheobronchoscopy and if necessary injection of
fections at a median time of 86 days (range, 43–217 days)
diluted methylene blue. The initial diagnosis was reverted
accordingly and all infants underwent open closure of the
Table 4 Final surgical repair in 96 patients with EA/TEF
previously undiagnosed proximal fistula through a cervical
incision. All regained full oral intake afterwards.
Gross type of EA/TEF
Anastomotic Stricture
A B C D E

Primary anastomosis (n = 66) – – 62 4 – Postoperative esophageal strictures at the anastomotic site

Delayed primary anastomosis (n = 25) 4 5 14 2 – were endoscopically confirmed in 69 (71.9%) symptomatic
Closure of TEF (n = 2) – – – – 2 children with dysphagia or respiratory distress during oral
Esophageal replacement (n = 3) feeding and were treated by hydrostatic balloon dilatations
Colonic interposition (n = 2) – – 2 – – or bougienage: 39 in the early and 30 in the late study period,
Gastric transposition (n = 1) 1 – – – – respectively (P = 0.0030). The median number of dilatations
was 3 (range, 1–15) with the last dilatation performed at a
932 J Gastrointest Surg (2017) 21:927–935

median age of 177 days (range, 16 days to 12.8 years). the distal esophageal segment with reduced pressure of the
Between 1975 and 1989, 23 (59.0%) patients had <5 dilata- lower esophageal sphincter, causing increased frequency of
tions and 16 (41.0%) required >5 dilatations. Since 1990, only acidic reflux episodes as a result of a balance created between
8 (26.7%) children underwent >5 dilatations. All cases of the intraesophageal and intragastric pressure. Forty-three
long-gap EA developed an anastomotic stricture with need (58.9%) were on intermittent or permanent anti-reflux medi-
for frequent dilatations. In two infants with EA types A cations, whereas the remaining 30 (41.1%) children did not
(n = 1) and C (n = 1), the stricture was recalcitrant to repeated respond to any conservative therapy and ultimately underwent
dilatations and resection with end-to-end reanastomosis be- fundoplication at a median age of 26 months (range, 31 days
came necessary 10 and 12 months, respectively, following to 10.3 years). Types of fundoplication procedures were
delayed primary repair. One premature boy with EA type A Nissen (n = 17), Thal (n = 5), Toupet (n = 4), and Guarner
developed a fistula at the suture line with severe anastomostic (n = 4).
stricture 2 months after delayed primary anastomosis with Upper gastrointestinal endoscopy revealed moderate or se-
Rehbein’s bougienage. He underwent open reoperation with vere esophagitis in 22 (23.9%) cases at a median time of
resection of the stricture and esophageal reanastomosis. In 15 years (range, 8–22 years) after initial surgical repair, which
total, GERD-associated anastomotic strictures were diagnosed was associated with a significant increase in GERD since the
in 59 (85.5%) of these patients and fundoplication ultimately introduction of routine pH monitoring in 1990 (29/42 vs. 44/
became necessary in 26 cases. 50; P = 0.0377). Barrett’s mucosa without dysplasia was
found in one patient at 20-year follow-up.
Functional Follow-Up Recurrent respiratory infections following EA/TEF surgery
occurred in 59 (64.1%) children, decreasing significantly in
The median age at most recent follow-up was 9.6 years (range, frequency and duration during the first 5 years of life. All of
3–27 years). A total of 71 (74.0%) infants were readmitted these demonstrated esophageal dysmotility and GERD.
more than once in the first year of life and a median of 5 Relevant tracheomalacia was diagnosed in 29 (31.5%) cases
(range, 1–26) hospitalizations with or without surgical inter- based on tracheobronchoscopic findings after a median post-
ventions were required until the age of 3 years. Eighty-nine operative time of 131 days (range, 31 days to 4.1 years). The
(96.7%) of the currently living 92 patients are thriving with majority (n = 21) had mild symptoms and were successfully
normal oral feeds. In 80 (89.9%) of these, full dietary intake treated with non-operative conservative therapy. Eight pa-
without any supplemental feeding gastrostomy or nasogastric tients, which all presented with recurrent dyspnea, dying
tube was achieved within 3 months following surgery, where- spells, and significant tracheal collapse, eventually required
as the remaining 9 (10.1%) had prolonged dependence due to surgical intervention with aortopexy (n = 6) or tracheal
various postoperative complications: anastomotic strictures stenting (n = 2).
(n = 3), recurrent fistula (n = 2), or coexisting anomalies in-
cluding congenital heart defects (n = 2) and syndromic disor-
ders (n = 2). At present, three (3.3%) children remain depen-
dent on feeding gastro- or jejunostomy. All of these have Discussion
neurological impairment and suffer from severe dysphagia
with high aspiration risk. Esophageal manometry showed no Since the first successful primary repair of a newborn with
propagating swallows with reduced pressure of the lower EA/TEF in 1941, survival rates have significantly improved
esophageal sphincter and proven GERD on 24-h pH throughout the world and the majority of today’s patients will
monitoring. reach adulthood.17, 27 Although there was a significant in-
Weak or absent esophageal peristalsis with impaired or crease in overall survival between the early and late study
absent contraction pattern was found postoperatively in a total period in our series, no statistical differences between postop-
of 78 (84.8%) infants at 1 year of age, deteriorating to 100% erative mortality rates were found. These findings suggest
proven dysmotility at 10-year follow-up manometry. Fifty-six considerable improvements in neonatal resuscitation and in-
(60.9%) suffered from varying degrees of swallowing tensive care, thus allowing for more severely ill and complex
difficulties. In 12 (13.0%) cases, there was at least one neonates with EA/TEF (i.e. Spitz group II and III) to survive
episode of esophageal food or foreign body impaction and eventually undergo surgical repair. The overall mortality
requiring endoscopic removal. An underlying anastomotic following surgery for EA/TEF was low at 4.2% and appears
stricture associated with GERD and need for repeat within international results.3, 9–11, 28 In general, the postoper-
dilatation was identified in eight of these. All of them ative mortality seems to be more reflective of a center’s overall
eventually had anti-reflux surgery. neonatal expertise rather than the operative management itself
Overall, GERD was diagnosed in 73 (79.3%) patients after as recently demonstrated by a nationwide study of 3479 chil-
EA/TEF repair and all exhibited weak or absent peristalsis of dren with EA/TEF across the United States.8
J Gastrointest Surg (2017) 21:927–935 933

Because of its rarity and complex nature, EA/TEF repre- highlights the necessity to carefully inspect the posterior tra-
sents one of the major therapeutic challenges not only in mod- cheal wall. Today, we are using a slightly bent tip of a 3F
ern pediatric surgery but also for other specialties involved ureteral catheter to search for potential small opening pits in
due to serious short- and long-term morbidities.13–18 Our re- the posterior tracheal membrane.
sults were consistent with previous studies indicating that de- Anastomotic strictures are one of the most frequent prob-
spite precise esophageal reconstruction and excellent postop- lems after EA/TEF repair and can significantly complicate the
erative care, early and late complications frequently occur and further outcome.34 The majority usually respond well to re-
must be dealt with.19–22 The definition of possible complica- peated dilatations with bougies or hydrostatic balloons with-
tions is however lacking standardization, which makes an ex- out need for further intervention. Only 24 (25.0%) children in
act comparison with the literature difficult and thus results in our series with a stricture had more than five anastomotic
highly variable incidence rates.13, 14 It can be assumed that the dilatations. However, two of the strictures were recalcitrant
development of complications after initial EA/TEF repair is and patients ultimately underwent resection with end-to-end
most likely a multifactorial process that depends on applied reanastomosis. A study from the United Kingdom recently
surgical techniques, peri-/postoperative care, and individual suggested that resection of an anastomotic stricture following
patient factors.29 EA/TEF surgery should be considered if more than ten bal-
Early complications in our cohort included minor (6.3%) or loon dilatations are required.35 Significant tension on the anas-
major (5.2%) anastomotic leaks, recurrent TEF with (1.0%) or tomotic site due to long-gap disease, previous history of anas-
without (9.4%) anastomotic stricture, initially missed proxi- tomotic leakage, and recurrent GERD are known risk factors
mal fistula (4.2%) and refractory anastomotic strictures with for stricture formation.30, 34 Unsurprisingly, all of our long-
(1.0%) or without (2.1%) fistula. gap cases, and seven patients with a preceding leak, later de-
Severe anastomotic leakage due to partial disruption of the veloped an anastomotic stricture. In addition, most of our in-
esophageal anastomosis is one of the most serious and poten- fants with EA/TEF that presented with recurrent strictures also
tially fatal complications following EA/TEF surgery, which had GERD, and in almost half of these a fundoplication be-
should be immediately operated on before inflammation and came necessary around the age of 2 years. GERD is a common
necrosis renders the esophageal wall unsuitable for issue after EA/TEF repair that does not tend to improve over
resuturing.30 Almost 75% of the successfully treated patients time. It most likely originates from the esophageal malforma-
in our series with an anastomotic leak developed afterwards a tion itself, abnormal innervation patterns of the distal esopha-
recurrent distal TEF or stricture, which is in accordance with a geal segment with insufficient propulsive peristalsis and defi-
recent study from Helsinki that identified anastomotic leakage cient function of the lower esophageal sphincter, thus causing
as a significant risk factor for recurrent TEF and stricture a long acid exposure time after each reflux episode.36
formation.30 It is therefore recommended that all infants with The present series concurs with a recent nationwide survey
EA/TEF and a previous history of anastomotic leakage, who from Italy, demonstrating that the treatment of long-gap EA
suffer from nonspecific symptoms like recurrent respiratory can be challenging as nearly all of these patients will experi-
issues or feeding difficulties, should be investigated by con- ence some sort of postoperative difficulties.37 Furthermore,
trast study and esophago-/tracheobronchoscopy with methy- we identified a high rate of gastroesophageal problems in
lene blue test.31 our cohort including dysphagia, impaired esophageal peristal-
Similar to other authors,29 our experience was that despite sis, and GERD, often persisting into adulthood. As patients
multiple applications, endoscopic administration of tissue ad- with EA/TEF have experienced these symptoms since infan-
hesive substances was not very effective in the treatment of cy, most have likely adapted them into their lifestyle and
recurrent TEF, whereas we had good success rates with an would not necessarily report them spontaneously. It has been
open approach, careful surgical separation and interposition pointed out that these cases may have an increased risk of
of vascularized pericardial or pleural tissue even in cases with malignancy of the esophagus later in life, which is why regular
severe adhesion or infection at the primary TEF site. endoscopy and combined impedance/pH monitoring has been
Furthermore, a recurrent distal TEF should be distinguished recommended for EA/TEF.21 On the other hand, a recent
from a previously unidentified proximal fistula, which is very study of 209 patients with a median follow-up of 12 years
rare but can be missed, especially when the gap between the discovered only esophagitis and metaplasia, but no dysplasia
upper and lower esophageal pouch is small without need for or cancer after surgery for EA/TEF.38 Thus, these findings
extensive mobilization.32, 33 In approximately 4% of our cases indicate that routine endoscopic surveillance seems to have
with EA/TEF, an initially undiagnosed fistula above the limited benefit during childhood, but is crucial from adoles-
esophagoesophagostomy was found during reoperation with- cence onwards. The exact cause of the underlying abnormal
out much adhesion and successfully closed through an cervi- esophageal function in EA/TEF is, however, still not fully
cal incision. Nevertheless, the fact that three proximal TEFs understood, and the correlation between symptoms, investiga-
were missed despite preoperative tracheobronchoscopy tional findings, and later outcome often appears to be poor.39
934 J Gastrointest Surg (2017) 21:927–935

In general, patients with EA/TEF needed complex and fre- A multidisciplinary follow-up with proper transitional care
quent hospital-based care as recently also shown by a multi- into adulthood is therefore highly recommended.
disciplinary team from Toronto.40 If a revision of the accom-
plished esophagoesophagostomy unfortunately becomes nec- Author Contributions Study concept and design: F.F., M.E.H.
Data acquisition, analysis, and interpretation: F.F., B.K., A.H.-Z., P.S.,
essary, one has to deal with very difficult and complex cases
H.T., H.S., M.E.H.
that may still require further surgical interventions to ultimate- Drafting of the manuscript: F.F., B.K., H.T., M.E.H.
ly regain esophageal continuity. In comparison with other Final approval: F.F., B.K., A.H.-Z., P.S., H.T., H.S, M.E.H.
large series,10, 41 replacement of native esophagus in our co- Agreement to be accountable for all aspects of the work: F.F., B.K.,
A.H.-Z., P.S., H.T., H.S., M.E.H.
hort was with 3% very low, which is probably associated with
the high rate of delayed primary repair. It has been reported
that with regard to gastroesophageal function, delayed prima- Compliance with Ethical Standards
ry anastomosis generally provides good results, especially in
the management of long-gap EA.42 Permanent dependence on Author Disclosure Statement The authors declare that this study was
gastro-/jejunostomy feedings in this series only occurred in conducted in the absence of any commercial or financial relationships that
three neurologically impaired infants with severe swallowing could be constructed as a potential conflict of interest.
disorders, whereas more than 95% enjoyed full oral intake
despite existing esophageal dysmotility. According to recently
published data from 90 children and adults with EA/TEF reg- References
istered in a German patient support group, the health-related
quality of life after complex and/or complicated surgery ap- 1. Billmyre KK, Hutson M, Klingensmith J. One shall become two:
pears to be excellent.43 separation of the esophagus and trachea from the common foregut
tube. Dev Dyn 2015;244:277–88.
All 96 cases in our cohort operated for EA/TEF were man- 2. Nassar N, Leoncini E, Amar E, et al. Prevalence of esophageal
aged with open surgery. Although thoracoscopic EA/TEF re- atresia among 18 international birth defects surveillance programs.
pair was described more than 10 years ago, this approach is Birth Defects Res A Clin Mol Teratol 2012;94:893–9.
currently performed at only a few centers due to technical 3. Oddsberg J, Lu Y, Lagergren J. Aspects of esophageal atresia in a
population-based setting: incidence, mortality, and cancer risk.
difficulties associated with the minimally invasive access, re-
Pediatr Surg Int 2012;28:249–57.
lated learning curve, and long operative time. Thoracoscopic 4. Pedersen RN, Calzolari E, Husby S, Garne E, EUROCAT Working
correction of EA/TEF trends toward increased anastomotic group. Oesophageal atresia: prevalence, prenatal diagnosis and as-
leakage, and greater need for fundoplication.44 Results pub- sociated anomalies in 23 European regions. Arch Dis Child
lished by surgeons who are pioneers in minimally invasive 2012;97:227–32.
5. Conforti A, Morini F, Bagolan P. Difficult esophageal atresia: trick
EA/TEF repair may not be generalizable and the complication
and treat. Semin Pediatr Surg 2014;23:261–9.
rate from centers with less experience is likely underreported 6. Zani A, Eaton S, Hoellwarth ME, et al. International survey on the
in the literature.45 Furthermore, a recent randomized con- management of esophageal atresia. Eur J Pediatr Surg 2014;24:3–8.
trolled trial reported that thoracoscopic EA/TEF repair can 7. Wang B, Tashiro J, Allan BJ, et al. A nationwide analysis of clinical
be associated with worse intraoperative acidosis and hyper- outcomes among newborns with esophageal atresia and
tracheoesophageal fistulas in the United States. J Surg Res
capnia compared with open management.46 However, in se- 2014;190:604–12.
lected patients and with experienced teams, thoracoscopic cor- 8. Sulkowski JP, Cooper JN, Lopez JJ, et al. Morbidity and mortality
rection of EA/TEF is appropriate. in patients with esophageal atresia. Surgery 2014;156:483–91.
The limitations of this study lie in its follow-up design and 9. Sfeir R, Bonnard A, Khen-Dunlop N, et al. Esophageal atresia: data
thus reliance on individually written chart records. Not report- from a national cohort. J Pediatr Surg 2013;48:1664–9.
10. Koivusalo AI, Pakarinen MP, Rintala RJ. Modern outcomes of
ed confounding factors could not necessarily be ruled out,
oesophageal atresia: single centre experience over the last twenty
which may have caused a potential result bias. years. J Pediatr Surg 2013;48:297–303.
11. Malakounides G, Lyon P, Cross K, et al. Esophageal atresia: im-
proved outcome in high-risk groups revisited. Eur J Pediatr Surg
Conclusion 2016;26:227–31.
12. Orford J, Cass DT, Glasson MJ. Advances in the treatment of oe-
sophageal atresia over three decades: the 1970s and the 1990s.
The overall outcome following EA/TEF surgery was charac- Pediatr Surg Int 2004;20:402–7.
terized by a low mortality rate with the vast majority of cases 13. Dingemann C, Dietrich J, Zeidler J, et al. Early complications after
achieving normal dietary intake. Postoperative complications esophageal atresia repair: analysis of a German health insurance
were successfully treated in nearly all patients with EA/TEF, database covering a population of 8 million. Dis Esophagus.
2016;29:780–6.
but had a significant impact on later esophageal function. 14. Schneider A, Blanc S, Bonnard A, et al. Results from the French
Potential problems should be early recognized and competent- National Esophageal Atresia register: one-year outcome. Orphanet
ly managed in order to reduce the risk of long-term morbidity. J Rare Dis 2014;9:206.
J Gastrointest Surg (2017) 21:927–935 935

15. Alshehri A, Lo A, Baird R. An analysis of early nonmortality out- 31. Coran AG. Diagnosis and surgical management of recurrent
come prediction in esophageal atresia. J Pediatr Surg 2012;47:881– tracheoesophageal fistula. Dis Esophagus 2013;26:380–1.
4. 32. Zhu H, Shen C, Xiao X, Dong K, Zheng S. Reoperation for anas-
16. Rintala RJ, Pakarinen MP. Long-term outcome of esophageal anas- tomotic complications of esophageal atresia and tracheoesophageal
tomosis. Eur J Pediatr Surg 2013;23:219–25. fistula. J Pediatr Surg 2015,50:2012–5.
17. Ijsselstijn H, van Beelen NW, Wijnen RM. Esophageal atresia: 33. Guo W, Li Y, Jiao A, Peng Y, Hou D, Chen Y. Tracheoesophageal
long-term morbidities in adolescence and adulthood. Dis fistula after primary repair of type C esophageal atresia in the neo-
Esophagus 2013;26:417–21. natal period: recurrent or missed second congenital fistula. J Pediatr
18. Legrand C, Michaud L, Salleron J, et al. Long-term outcome of Surg 2010;45:2351–5.
children with oesophageal atresia type III. Arch Dis Child 34. Lévesque D, Baird R, Laberge JM. Refractory strictures post-
2012;97:808–11. esophageal atresia repair: what are the alternatives? Dis
19. Connor MJ, Springford LR, Kapetanakis VV, Giuliani S. Esophagus 2013;26:382–7.
Esophageal atresia and transitional care—step 1: a systematic re- 35. Thyoka M, Barnacle A, Chippington S, et al. Fluoroscopic balloon
view and meta-analysis of the literature to define the prevalence of dilation of esophageal atresia anastomotic strictures in children and
chronic long-term problems. Am J Surg 2015;209:747–59. young adults: single-center study of 103 consecutive patients from
20. Shah R, Varjavandi V, Krishnan U. Predictive factors for complica- 1999 to 2011. Radiology 2014;271:596–601.
tions in children with esophageal atresia and tracheoesophageal 36. Höllwarth ME. Gastroesophageal reflux disease. In: Coran AG,
fistula. Dis Esophagus 2015;28:216–23. Adzick NS, Krummel TM, Laberge JM, Shamberger R,
21. Pedersen RN, Markøw S, Kruse-Andersen S, et al. Esophageal Caldamone AA (eds.). Pediatric surgery. Vol. 2, 7th edn.
atresia: gastroesophageal functional follow-up in 5–15 year old Philadelphia: Elsevier Sauders; 2012:947–58.
children. J Pediatr Surg 2013;48:2487–95. 37. Pini Prato A, Carlucci M, Bagolan P, et al. A cross-sectional nation-
22. Castilloux J, Noble AJ, Faure C. Risk factors for short- and long- wide survey on esophageal atresia and tracheoesophageal fistula. J
term morbidity in children with esophageal atresia. J Pediatr Pediatr Surg 2015;50:1441–56.
2010;156:755–60.
38. Koivusalo AI, Pakarinen MP, Lindahl HG, Rintala RJ. Endoscopic
23. Höllwarth ME, Zaupa P. Oesophageal atresia. In Puri P, Höllwarth
surveillance after repair of oesophageal atresia—longitudinal study
ME (eds.). Pediatric surgery. 1st edn. Heidelberg, Berlin, New
in 209 patients. J Pediatr Gastroenterol Nutr 2016;62:562–6.
York: Springer; 2006:29–48.
39. Sistonen SJ, Koivusalo A, Nieminen U, et al. Esophageal morbidity
24. Somppi E, Tammela O, Ruuska T, et al. Outcome of patients oper-
and function in adults with repaired esophageal atresia with
ated on for esophageal atresia: 30 years’ experience. J Pediatr Surg
tracheoesophageal fistula: a population-based long-term follow-
1998;33:1341–6.
up. Ann Surg 2010;251:1167–73.
25. Low DE, Alderson D, Cecconello I, et al. International Consensus
on Standardization of Data Collection for Complications 40. Zamiara P, Thomas KE, Connolly BL, Lane H, Marcon MA, Chiu
Associated With Esophagectomy: Esophagectomy Complications PP. Long-term burden of care and radiation exposure in survivors of
Consensus Group (ECCG). Ann Surg 2015;262:286–94. esophageal atresia. J Pediatr Surg 2015;50:1686–90.
26. Vandenplas Y, Rudolph CD, Di Lorenzo C, et al. Pediatric gastro- 41. Lacher M, Froehlich S, von Schweinitz D, Dietz HG. Early and
esophageal reflux clinical practice guidelines: joint recommenda- long term outcome in children with esophageal atresia treated over
tions of the North American Society for Pediatric Gastroenterology, the last 22 years. Klin Padiatr 2010;222:296–301.
Hepatology, and Nutrition (NASPGHAN) and the European 42. Friedmacher F, Puri P. Delayed primary anastomosis for manage-
Society for Pediatric Gastroenterology, Hepatology, and Nutrition ment of long-gap esophageal atresia: a meta-analysis of complica-
(ESPGHAN). J Pediatr Gastroenterol Nutr 2009;49:498–547. tions and long-term outcome. Pediatr Surg Int 2012;28:899–906.
27. Maynard S, Bouin M. Follow-up of adult patients with repaired 43. Dingemann C, Meyer A, Kircher G, et al. Long-term health-related
esophageal atresia: how, when, and for how long? Dis Esophagus quality of life after complex and/or complicated esophageal atresia
2013;26:422–4. in adults and children registered in a German patient support group.
28. Burge DM, Shah K, Spark P, et al. Contemporary management and J Pediatr Surg 2014;49:631–8.
outcomes for infants born with oesophageal atresia. Br J Surg 44. Burford JM, Dassinger MS, Copeland DR, Keller JE, Smith SD.
2013;100:515–21. Repair of esophageal atresia with tracheoesophageal fistula via tho-
29. Mortell AE, Azizkhan RG. Esophageal atresia repair with thoracot- racotomy: a contemporary series. Am J Surg 2011;202:203–6.
omy: the Cincinnati contemporary experience. Semin Pediatr Surg 45. Laberge JM, Blair GK. Thoracotomy for repair of esophageal atre-
2009;18:12–9. sia: not as bad as they want you to think! Dis Esophagus 2013;26:
30. Koivusalo AI, Pakarinen MP, Lindahl HG, Rintala RJ. Revisional 365–71.
surgery for recurrent tracheoesophageal fistula and anastomotic 46. Pierro A. Hypercapnia and acidosis during the thoracoscopic repair
complications after repair of esophageal atresia in 258 infants. J of oesophageal atresia and congenital diaphragmatic hernia. J
Pediatr Surg 2015;50:250–4. Pediatr Surg 2015;50:247–9.