Keywords
Broncho-esophageal fistula - esophageal fistula - esophageal leak - esophageal stent
- esophageal stricture - tracheo-esophageal fistula
Introduction
Although the use of esophageal stents has been relatively common in adults,[1],[2] for both benign and malignant conditions, their use in children is still evolving.[3],[4],[5],[6],[7],[8],[9],[10],[11] There are a few reports on the use of covered stents in children with caustic strictures
or strictures after esophageal atresia repair in which dilatations were unsuccessful.[3],[6],[10] In regards to leak or perforation after repair of esophageal atresia or esophageal
dilatations, nonoperative management is usually effective. However, there are some
cases in which the esophagus fails to close with nonoperative management in which
covered stents could be used to manage such perforations.[5],[7] Small size of the pediatric esophagus and the lack of availability of dedicated
pediatric esophageal stents [Table 1] makes this process difficult. This report describes the application of esophageal
stents in children with esophageal strictures, leaks, or airway-esophageal fistulae
refractory to conventional treatment at a tertiary care children’s hospital.
Table 1
Characteristics of self-expanding metal esophageal stents available in USA
|
Esophageal Stent
|
Material
|
Diameter (mm)
|
Length (cm)
|
Advantages
|
Disadvantages
|
|
Merit Medical Endotek®
|
|
|
|
|
|
|
Alimaxx-ES™ Esophageal stent (fully covered)
|
Laser cut nitinol, Polyurethane, silicone
|
12-22
|
7-12
|
No stent foreshortening or elongation, polyurethane coating decreases tissue ingrowth,
silicone lining provides smooth inner lumen, anti-migration struts reduce stent migration,
purse-string design of the proximal suture knot helps with repositioning and removal
|
Stent cannot be reconstrained once deployment is initiated.
|
|
EndoMAXX™ Esophageal stent (fully covered)
|
Laser cut nitinol, silicone
|
19-23
|
7-15
|
In addition to advantages of Alimaxx-ES™, flared ends reduce migration, proximal and
distal metal suture allow stent repositioning
|
Stent cannot be reconstrained once deployment is initiated
|
|
Boston Scientific
|
|
|
|
|
|
|
WallFlex™ Esophageal stent (partially and fully covered)
|
Braided nitinol, silicone
|
18,23
|
10,12,15
|
Silicone coating decreases tissue ingrowth, flared ends reduce migration, proximal
polyester suture helps stent removal, reconstrainable for up to 75% deployment
|
Stent foreshortening
|
|
Ultraflex™ Esophageal NG stent (covered and non-covered)
|
Knitted nitinol, polyurethane
|
18,23
|
7-15
|
Proximal flare to reduce migration, sutured loop for stent repositioning, available
as proximal and distal release systems
|
|
|
Cook Medical
|
|
|
|
|
|
|
Evolution® Esophageal Controlled-Release stent (partially and fully covered)
|
Woven nitinol, silicone
|
18,20
|
8-15
|
Silicone coating decreases tissue ingrowth, flared ends reduce migration, lasso for
stent repositioning, reconstrainable for up to 50% deployment
|
Stent foreshortening
|
Case Series
Three children aged 7 months, 1 year, and 4 years underwent esophageal stenting at
a tertiary care pediatric hospital between December 2012 and November 2016. Institutional
review board approval was obtained for this retrospective review.
The stents used in this series were covered nitinol esophageal (Alimaxx-ES™) or tracheo-bronchial
(AERO ®) commercially available stents (Merit Medical Systems, Inc., South Jordan,
UT, USA). The features of the stent include polyurethane coating to decrease granulation
tissue formation, anti-migration struts to decrease migration, and a proximal suture
knot to aid removal. All stents were placed in the interventional radiology suite
under general anesthesia with flexible endoscopy and fluoroscopy guidance using a
flexible over-the-wire delivery system. The esophageal diameter was estimated using
previous contrast study and endoscopic assessment, and the stent diameter was oversized
by approximately 2 mm. The length of the stent was selected to provide at least 2
cm overlap on either side of the leak but without extension into the upper or lower
gastrointestinal sphincter. After stent deployment, an esophagogram was performed
to document stent position and absence of extravasation. Follow-up chest radiographs
were performed as needed during follow-up.
Case 1
A 1-year-old boy developed a large acquired tracheo-esophageal fistula [Figure 1]A after endoscopic removal of a swallowed foreign body. This was surgically repaired
but complicated postoperatively with leak from the esophageal anastomosis and tracheo-esophago-pleural
fistula. This led to tension pneumothorax for which multiple chest tubes were placed.
The patient however continued to have difficulty with ventilation and could not be
extubated. A 12 mm × 7 cm Alimaxx-ES™ stent was placed in the esophagus [Figure 1]B to exclude the large fistula. This allowed the patient to be weaned off the ventilator
and extubated on postoperative day 8. On postoperative day 15, however, the stent
migrated proximally [Figure 1]C and was removed. Thereafter, the patient had successful surgical repair of the
partially healed fistula.
Figure 1 (A-C): Case 1: M/1 year with large acquired tracheoesophageal fistula (TEF) after foreign
body removal. (A) Esophagogram shows large TEF (black arrow) in the upper third of
the esophagus. (B) 12 mm χ 7 cm Alimaxx-ES™ esophageal stent placed across TEF. (C)
Proximal migration of stent on day 15
Case 2
A 4-year-old boy had congenital esophageal atresia and distal tracheo-esophageal fistula
repaired at an outside hospital. Investigation of recurrent pneumonia led to the finding
of an acquired esophago-bronchial fistula [Figure 2]A. An Amplatzer™ Vascular Plug II (St. Jude Medical, St. Paul, MN, USA) was placed
within the fistula along with a 16 mm × 7 cm Alimaxx-ES™ esophageal stent [Figure 2]B to completely exclude the fistula. The stent was removed electively at 8 weeks
and esophagogram immediately post removal was normal. However, a repeat study two
weeks later prior to initiation of oral feeding showed persistent fistula. During
this period, the patient had improved symptomatically and all signs of infection had
cleared. The patient underwent successful surgical repair with no recurrence of the
fistula at the 3-year follow-up.
Figure 2 (A and B): Case 2: M/4 year with esophageal atresia and distal TEF post surgical repair. (A)
Esophagogram shows acquired esophago-bronchial fistula (black arrow). (B) Placement
of 4 mm Amplatzer™ Vascular plug II (black arrow) and 16 mm χ 7 cm Alimaxx-ES™ esophageal
stent to occlude fistula
Case 3
A 7-month-old girl with Trisomy 21, long-gap esophageal atresia, duodenal atresia,
and Hirschprung’s disease underwent delayed primary esophageal anastomosis at 4 months
of age during which the esophageal ends were tacked together without fashioning an
anastomosis because of excessive tension. One month later, the ends of the esophagus
were reapproximated over a trans-anastomotic tube. This was complicated by contained
leak and anastomotic stricture that did not respond to repeated dilatations [Figure 3]A. A 12 mm × 4 cm AERO ® tracheo-bronchial stent was placed across the stricture
and leak [Figure 3]B. The AERO ® stent was selected because the shortest available Alimaxx-ES™ esophageal
stent was 7 cm that would have extended into the stomach. Distal migration of the
AERO ® stent after one week was managed with telescoping a 14 mm × 4 cm AERO ® stent
within the previous stent. Over the ensuing 4 weeks, the coaxial stents not only migrated
distally but also eroded into the right lateral esophageal wall creating a large diverticulum
[Figure 3]C. This made removal of the stent challenging by obscuring a large part of the upper
border of the stent and its suture knot. Over the following 2 weeks, the diverticulum
progressed to a esophago-bronchial fistula and there was recurrence of the stricture.
The patient was not a candidate for colon interposition given Hirschprung’s disease
and had frozen chest from repeated thoracotomies and soiling, therefore, a proximal
esophageal diversion was performed. The patient, however, succumbed to progressive
respiratory deterioration and aspiration pneumonia from the persistent esophago-bronchial
fistula.
Figure 3 (A-C): Case 3: F/7 month with long gap esophageal atresia post repair (A). Esophagogram
shows high grade anastomotic stricture in the mid esophagus (black arrow) associated
with a contained leak (white arrow). (B) Placement of a 12 mm χ 4 cm AERO® tracheo-bronchial
stent across stricture. (C) Distal migration and erosion of stent into right lateral
esophageal wall
Discussion
Esophageal leaks and strictures in children can occur after a variety of procedures
including repair of esophageal atresia, dilatation of stricture, or ingestion of caustic
agents. Esophageal anastomotic leaks and iatrogenic esophageal perforations often
heal over time with nonoperative management. However, the management of complex esophageal
problems that do not heal spontaneously or respond to dilatations is challenging.
The use of esophageal stents in children is evolving and has mostly been described
for treatment of strictures refractory to dilatation.[3],[4],[6],[10],[12] The advent of covered retrievable stents led to their utilization particularly for
caustic strictures as well as for strictures following esophageal atresia repair with
high success rates. Zhang et al. described eight patients with corrosive esophageal stenosis who underwent placement
of covered retrievable nitinol stents.[3] All patients were able to eat solid food without dysphagia after stent removal.
Rollins et al. reported a series of three patients with persistent esophageal leaks in which covered
nitinol stents were successfully used,[5] and patients were able to receive nutrition orally with the stents in place. They
concluded that the use of esophageal stents should be considered as an effective option
for these cases. In a large pediatric experience of esophageal stenting, Manfredi
et al. concluded that esophageal stents proved more useful in the management of esophageal
perforations than recalcitrant anastomotic strictures.[8] In the current report, esophageal stents were useful mainly as a bridge to definitive
surgical repair. In case 1, the stent played a critical role in controlling massive
leak of tidal volume through the tracheo-esophago-pleural fistula. This made it possible
to extubate the child but the stent migrated prematurely and required surgical management
of the residual fistula. In case 2, the stent temporarily sealed the esophago-bronchial
fistula allowing the child to recover from severe pneumonia and systemic illness.
The child was then a better candidate for surgical repair. In both these cases, however,
the stent could only provide partial or temporary healing of the perforation and surgical
repair was subsequently required.
One of the main complications of esophageal stents is stent migration. Migration has
been reported in 0–29% of pediatric patients [12] and 30–60% in the adult literature.[4] The Alimaxx-ES™ esophageal and AERO ® tracheo-bronchial stents have anti-migration
struts that are designed to decrease the risk of migration. The struts are oriented
such that they counter distal migration of the esophageal stent and proximal migration
of the tracheo-bronchial stent. It has also been reported that stents placed in the
upper and lower esophagus are more likely to migrate and more difficult to manage.[13] Stent migration was a significant problem in case 1 and case 3 in the current series.
In case 1, although an esophageal stent with distal anti-migration struts was used,
placement in the upper esophagus in proximity to the cricopharyngeal sphincter was
most likely responsible for its early proximal migration. In case 3, distal migration
occurred despite stent revision. In this patient, a 4-cm long AERO ® tracheo-bronchial
stent was placed initially because the commercially available esophageal stents were
too long for this 7-month-old child. The tracheo-bronchial stents have struts oriented
to prevent proximal migration, but do not prevent distal migration. In addition, a
high-grade anastomotic stricture with a tortuous lumen in this child also contributed
to stent migration.
Esophageal perforation secondary to esophageal stents has been reported as one of
the delayed complications.[14],[15],[16] Manfredi et al. reported esophageal perforation and migration of an esophageal stent into the pleural
space within two weeks of placement and recommended close monitoring of all children
undergoing esophageal stenting.[8] In case 3, angulation of the rigid AERO ® stent into the right lateral wall of the
esophagus resulted in erosion of the esophagus and formation of a large diverticulum
which subsequently progressed to esophago-bronchial fistula. The AERO ® stent is made
from a laser cut solid nitinol tube which makes its structure rigid, straightens the
natural esophageal curves, and causes pressure necrosis of the esophageal wall.
Esophageal stenting in children presents several technical challenges. In infants
and smaller sized children, the commercially available esophageal stents may be too
long. In these patients, AERO ® tracheo-bronchial stents available in shorter lengths
may be utilized cautiously. The size mismatch and rigidity of the stents also makes
the esophagus vulnerable to pressure necrosis, for which close monitoring and timely
removal is suggested. There is certainly a need for development of smaller sizes (both
length and diameter) of stents as well as more flexible stents which conform to the
natural curvatures of the pediatric esophagus.
In conclusion, esophageal stents may serve as a useful tool in the management of children
with esophageal leaks and acquired airway-esophageal fistulae refractory to conventional
treatment. In this case series, the stent allowed for time to control ventilation
and resolve infection, acting as a bridge to definitive surgical repair. The small
size and tortuosity of the pediatric esophagus and the limited availability of commercial
esophageal stents designed for the pediatric esophagus are challenges in this field.
Development of esophageal stents with improved flexibility, anti-migration features
and smaller diameters and lengths could further increase their successful application
in children.
Declaration of patient consent
The authors certify that they have obtained all appropriate patient consent forms.
In the form the patient has given her consent for her images and other clinical information
to be reported in the journal. The patient understands that name and initial will
not be published and due efforts will be made to conceal identity, but anonymity cannot
be guaranteed.
