GENERAL THORACIC
Surgical Management of Benign Acquired
Tracheoesophageal Fistulas: A Ten-Year
Experience
Benoit Jacques Bibas, MD, Paulo Francisco Guerreiro Cardoso, MD, PhD,
Helio Minamoto, MD, PhD, Leandro Picheth Eloy-Pereira, MD,
Mauro Federico L. Tamagno, MD, Ricardo Mingarini Terra, MD, PhD, and
ˇ
Paulo Manoel Pego-Fernandes, MD, PhD
Division of Thoracic Surgery, Heart Institute (InCor), Hospital das Clínicas da Faculdade de Medicina da Universidade de S~
S~
ao Paulo, Brazil
Background. Benign tracheoesophageal fistulas (TEFs)
are rare, and surgical correction is the ideal method of
treatment. The objective of this study was to evaluate the
results of operative treatment of benign TEFs in patients
from a tertiary referral center.
Methods. Retrospective study of patients with benign
TEFs who were treated between January 2005 and
December 2014. Preoperative evaluation included
computed tomography of the chest, bronchoscopy, and
upper endoscopy. Preoperative treatment included
nutritional support by gastrostomy and treatment of lung
infections. Surgical repair was done with tracheal resec-
tion and reconstruction, laryngotracheal resection, or
membranous tracheal repair without resection. Esopha-
geal management consisted of two-layer closure.
Results. Twenty patients (11 men) with mean age 48 ±
17 years were included. The most frequent cause was
postintubation injury (n [ 16; 80%). The median TEF
length was 9 mm (interquartile range [IQR], 2 to 25 mm).
cquired benign tracheoesophageal fistula (TEF)
A usually derives from tracheal postintubation injury
despite the improvements in endotracheal tube man-
agement and the use of low-pressure and high-volume
cuffs [1]. Benign TEFs are almost exclusively located in
the upper and middle thirds of the trachea [2].
The onset of the TEF often imposes a dramatic impact
on the patient’s health status and quality of life because
of swallowing difficulties, recurrent aspiration pneu-
monia, and severe weight loss. Likewise, the association
with tracheal stenosis adds an ominous variable that
frequently demands different management strategies
such as the use of tracheal appliances to maintain
airway patency while the patient is prepared for
Accepted for publication April 11, 2016.
Presented at the Fifty-second Annual Meeting of The Society of Thoracic
Surgeons, Phoenix, AZ, Jan 23–27, 2016.
Address correspondence to Dr Bibas, Rua Dr. Eneas de Carvalho Aguiar
44, bloco 2, 2o andar, sala 9, Cerqueira Cezar, S~
ao Paulo, SP, 05403-000
Brazil; email: benoitbibas@hotmail.com.
Ó 2016 by The Society of Thoracic Surgeons
Published by Elsevier
ao Paulo,
The most commonly used surgical approaches were cer-
vicotomy (n [ 15; 75%) and cervicosternotomy (n [ 3;
15%). Eleven patients required tracheal resection;
median length was 3 cm (IQR, 3 to 5.5 cm). Seven
patients (35%) required intraoperative tracheostomy.
Complications occurred in 55% of patients. There was
one dehiscence of the tracheal anastomosis, and one
procedure-related death. Ninety-five percent of patients
had complete closure of the TEF occurred in 95% of cases.
Two patients had tracheal stenosis recurrence, and one
patient had both TEF and tracheal stenosis recurrence.
Two patients have indwelling silicone tracheal stents.
Conclusions. Surgical treatment of TEF is effective.
Nonetheless, morbidity and mortality are not negligible,
even when performed at a referral center and after
appropriate preoperative evaluation.
(Ann Thorac Surg 2016;102:1081–7)
Ó 2016 by The Society of Thoracic Surgeons
definitive surgical treatment [1–3]. Preparation for the
surgical procedure may take weeks or even months and
includes definitive weaning from mechanical ventila-
tion, treatment of infection, respiratory physiotherapy,
and correction of malnutrition by means of enteral
feeding [3].
Surgical correction is a major surgical undertaking
that typically requires division and closure of the TEF in
conjunction with airway reconstructive techniques to
restore airway continuity with a muscle flap interposition
[4]. Nevertheless, recurrence of fistula will add to the
complexity of the problem, requiring laborious reopera-
tions and even unorthodox techniques [5]. In high-
volume centers the operative mortality and recurrence
of the TEF can be as high as 11% [2].
As a reference center for airway surgical procedures,
we have performed an increasing number of operations
for correction of benign TEFs over the years. The aim of
this study is to report our experience with the surgical
correction of acquired benign TEF, focusing on early and
late results.
0003-4975/$36.00
http://dx.doi.org/10.1016/j.athoracsur.2016.04.029
 1082 BIBAS ET AL
GENERAL THORACIC
BENIGN TRACHEOESOPHAGEAL FISTULAS
Material and Methods
This retrospective study included patients who under-
went surgical treatment of benign TEF at the Division of
Thoracic Surgery of the University of S~ao Paulo, Brazil,
between January 2005 and December 2014. Our institu-
tion is a tertiary academic hospital and a national
referral center for tracheal diseases. Patients were
identified from a prospectively filled electronic data-
base. All data were collected until March 2015. Confi-
dentiality of patient information was ascertained by
identifying patients through a code known by one
investigator (R.M.T.). Our hospital’s ethics committee
approved the project, and informed consent was waived
because of the retrospective nature of the data. Patients
with missing or incomplete data were excluded from the
cohort.
Preoperative Evaluation
All patients were initially evaluated at the thoracic sur-
gical outpatient clinic. Assessment included computed
tomography scan of the larynx, trachea, and lungs; flex-
ible or rigid bronchoscopy or both; and upper gastroin-
testinal endoscopy in all patients. Recurrent pulmonary
infections and aspiration control were dealt with phys-
iotherapy, cessation of all oral intake, and antibiotics.
Enteral nutritional support was ascertained preferably
through a percutaneous endoscopic gastrostomy or jeju-
nostomy. Routine consultations with the nutrition
department were scheduled. Patients with TEFs larger
than 10 mm and recurrent pulmonary infections under-
went a preoperative tracheostomy placed at the fistula
site, or as close to it as possible, to minimize tracheal
damage. The procedures were guided by bronchoscopy,
and the cuff of the cannula was placed distal to the
esophageal defect. Four patients had tracheal stents
(three T-Tubes and one Dumon Stent) placed before
the operation. They remained in place for a median of
77 days (interquartile range [IQR], 60 to 112 days); mean
80
25 days.
Operative Management
All surgical procedures were elective, and all patients
were breathing spontaneously before the operation. TEFs
located in the upper and mid esophagus were operated
on through a low cervical collar incision, and, if neces-
sary, a partial sternotomy was performed. Distal fistulas
were operated on through a right thoracotomy in the
fourth intercostal space.
The TEF diameter and the presence of tracheal stenosis
dictated the approach and repair as follows: TEF less than
5 mm in diameter without tracheal stenosis were treated
through a lateral approach with membranous tracheal
and esophageal closure. Larger fistulas (10 mm) usually
required concomitant tracheal resection and reconstruc-
tion. Tracheal anastomosis was performed with a
continuous running suture of polydioxanone 4-0 (PDS II;
Ethicon, Bridgewater, NJ) in the membranous wall and
separated sutures of polyglactin 3-0 (Vicryl; Ethicon) in
Ann Thorac Surg
2016;102:1081–7
the cartilaginous wall [6]. Laryngotracheal reconstruction
was performed according to the technique described by
Pearson and colleagues [7] and Grillo and colleagues [8].
Intraoperative tracheostomy was performed according to
the surgeon’s judgment. Fistulas associated with long-
segment stenosis (>5 cm) were approached by incision
of the anterior wall of the trachea at the site of the fistula,
with exposure of the esophagus. The esophageal defect
was closed, and tracheal reconstruction was accom-
plished. A tracheostomy or a silicone T-Tube was used.
Large TEFs (>10mm) without tracheal stenosis were
treated similarly, but without any tracheal appliances, as
described by Macchiarini and colleagues [9].
In all instances, the repair of the esophageal defect
included a two-layer closure technique. Vascularized
muscle flaps were interposed between the airway and
esophageal suture lines in all cases. Strap muscles were
used in proximal fistulae and intercostal muscle for distal
fistulae. Patients without gastrostomy or jejunostomy
received a nasogastric feeding tube in the operating room
that remained in place postoperatively until oral intake
was fully resumed.
Postoperative Management
Patients were initially sent to the surgical intensive care
unit for the first 24 to 48 hours and then discharged to
the ward. Between the 5th and 7th postoperative day a
barium swallow was performed to check for leaks or
aspiration before oral intake was resumed. Esophago-
scopy was not performed in the early postoperative
period. Patients were discharged when adequate
feeding was resumed either by oral intake or by feeding
tube in case of dysphagia or aspiration or both. Patients
were seen at the outpatient thoracic surgical clinic at 2
weeks; 1, 3, and 6 months after hospital discharge; and
twice yearly thereafter. Flexible bronchoscopy and up-
per endoscopy were routinely done 3 months after the
surgical procedure. In the event of any complaints
related to the operation or clinical signs or symptoms of
complications, imaging studies, flexible bronchoscopy,
and upper gastrointestinal endoscopy were also
performed.
Definition of Outcomes
We defined complications as binary (yes/no) variables.
Only events that required an intervention or led to pro-
longed hospital stay counted as a complication for the
purpose of the analysis. Operative mortality was defined
as deaths occurring within 30 days of the surgical pro-
cedure and deaths during the same hospitalization. Pa-
tients were considered lost to follow-up if no personal or
phone contact was available for 12 months.
Statistical Analysis
Parametric continuous variables are presented as means
and standard deviations. Nonparametric data are pre-
sented as medians and IQRs. The Shapiro-Wilk test was
used for the assessment of normality. Categorical vari-
ables are presented as absolute numbers and percentage.
Ann Thorac Surg
2016;102:1081–7
Statistics analysis was performed with Stata 13 (Stata-
Corp, College Station, TX).
Results
Twenty patients (11 men; 9 women) underwent surgical
procedures during the study period. They were all extu-
bated in the operating room and sent to the intensive care
unit. Mechanical ventilation was not necessary in the
early postoperative period. The flowchart of the study is
depicted in Figure 1. Demographic characteristics are
shown in Table 1. The most common TEF cause was
postintubation injury (80%). Other causes included direct
tracheoesophageal trauma and infections. One patient
experienced a TEF 6 months after a tracheal resection and
primary anastomosis. He had a postoperative indwelling
tracheostomy and was subsequently treated with a T-
Tube. After the diagnosis, the TEF was surgically
corrected.
Previous treatments before definitive closure of the
TEF can be seen in Table 2. The majority of patients had
a tracheostomy before the operation (75%). Ten patients
already had a tracheostomy on the first consultation at
our institution (4 had fistulas wider than 10 mm). Four
patients had tracheal stents while awaiting for definitive
treatment of the TEF; 3 patients had T-Tubes, and 1 pa-
tient had a silicone Dumon stent.
Endoscopic closure of the TEF was attempted in two
cases. In 1 patient, a self-expanding esophageal metallic
stent was placed at another institution. Another patient
had a 5-mm TEF and was treated with an atrial septal
defect occluder [10]. The patient remained asymptomatic
for 7 months, when the occluder dislodged, and respi-
ratory symptoms recurred that required surgical
correction of the TEF. Nineteen operations (95%) were
performed in one stage. One patient with HIV and
cytomegalovirus infection had a multistaged repair. A
cervical esophagostomy, closure of the esophageal
defect, and diversion were performed. Four months later
Fig 1. Study design. (Post-op. ¼ postoperative; success ¼ patient breathing without tracheal appliance and with adequate oral intake;
TEF ¼ tracheoesophageal fistula.)
BIBAS ET AL 1083
GENERAL THORACIC
BENIGN TRACHEOESOPHAGEAL FISTULAS
a colon interposition was performed to restore gastro-
intestinal continuity.
TEFs were located primarily in the upper and mid
esophagus and measured 9.4 mm on average (range, 2 to
25 mm). A low cervical collar incision was the operative
choice for most cases (75%). Three of 10 TEFs located
in the mid esophagus required a partial sternotomy.
Two lesions close to the carina were approached
through a right thoracotomy in the fourth intercostal
space (Table 3). Eleven patients (55%) required tracheal
reconstruction. The median length of trachea resected
was 3 cm (IQR, 3 to 5.5 cm). Suprahyoid release maneu-
vers were not used.
Complications occurred in 11 patients (55%) and are
depicted in Table 4. The most common were subcu-
taneous emphysema and pneumonia. Three patients
had previously failed TEF surgical repair in other in-
stitutions. One had undergone surgical procedures
twice. At our institution, he underwent surgical proce-
dure to a tracheal resection with a protective tracheos-
tomy and 2-layer closure of the esophagus. On the 5th
postoperative day there was progressive subcutaneous
emphysema, and bronchoscopy showed a partial
dehiscence of the anterior portion of the tracheal anas-
tomosis. Conservative treatment was performed by
means of a Montgomery T-Tube placed on the 14th
postoperative day. He had recurrence of the TEF (2 to 3
mm in diameter) and still has an indwelling T-Tube.
Nonetheless, he is capable of oral intake and has no
signs or symptoms of pulmonary infections. The other 2
patients with previous TEF repairs recovered unevent-
fully from the operation. One of them had a late recur-
rence of the tracheal stenosis, was managed initially with
a Montgomery T-Tube, and now has a silicone Dumon
stent. One patient with a 6-cm tracheal stenosis and a 2-
cm TEF was treated with an anterior incision of the
trachea and 2-layer closure of the esophagus. He had a
prior tracheostomy that was maintained in the post-
operative period. Three weeks after the initial operation
 1084 BIBAS ET AL Ann Thorac Surg
GENERAL THORACIC

BENIGN TRACHEOESOPHAGEAL FISTULAS 2016;102:1081–7

Table 1. Demographic Characteristics (n ¼ 20) Table 2. TEF Characteristics

Characteristic Value Characteristic Value

Sex, n (%) Previous TEF failed surgical repair, n (%) 4 (20)

Male 11 (55) Previous tracheal treatments, n (%)
Female 9 (45) T-Tube 3 (15)
Age, years 48
17 (13–84) Silicone stent 1 (5)
TEF cause, n (%) Tracheal resection 1 (5)
Postintubation injury 16 (80) Balloon dilation of the stenosis 1 (5)
Esophageal trauma/operation 1 (5) Tracheostomy 15 (75)
Postoperative tracheal resection 1 (5) Previous esophageal treatments, n (%)
Cytomegalovirus 1 (5) Gastrostomy 7 (35)
Tuberculosis 1 (5) Jejunostomy 2 (10)
Comorbidities, n (%) Self-expanding metallic stent 1 (5)
Hypertension 5 (25) Endoscopic closure of the TEF 1 (5)
Diabetes 4 (20) TEF size, mm 9.4
7.5 (2–25)
Chronic kidney disease 2 (10) TEF size, n (%)
Asthma 1 (5) 5 mm 8 (40)
Heart failure 1 (5) 5–10 mm 4 (20)
Bipolar disorder 1 (5) >10 mm 8 (40)
COPD 1 (5) TEF location, n (%)
HIV 1 (5) Upper esophagus 8 (40)
Intubation time, days 23
15 (2–60) Mid esophagus 10 (50)
Distal esophagus 2 (10)
Values are mean
standard deviation (range) unless otherwise specified.
Concomitant tracheal stenosis, n (%) 11 (55)
COPD ¼ chronic obstructive pulmonary disease; HIV ¼ human im- Upper third 7 (64)
munodeficiency virus; TEF ¼ tracheoesophageal fistula.
Subglottic 2 (18)
Distal third 2 (18)
a silicone T-Tube was placed. It was successfully
removed after 18 months. Four other patients required Values are mean
standard deviation (range) unless otherwise specified.
an intraoperative tracheostomy. One had a previous TEF ¼ tracheoesophageal fistula.
tracheal resection. In the remaining three cases, the
tracheal anastomosis was judged tense, and a protective postintubation injury, and the average intubation time
tracheostomy was performed. No esophageal leaks were was approximately 3 weeks. Successful closure of the
diagnosed in the postoperative period. TEF was accomplished in 95% of the cases. Complications
One postoperative death occurred in this series. It occurred in 55% of patients, and there was a 5% mortality
consisted of a 4-cm tracheal stenosis that required rate. Three patients had recurrent tracheal stenosis; all
resection of the anterior portion of the cricoid cartilage. were managed with silicone stents.
A T-Tube was placed at the time of the operation. Early
postoperative period was complicated by deep venous Table 3. Operative Management
thrombosis and pulmonary embolism. Pneumonia and
sepsis ensued; the patient was intubated 45 days after the Variable n (%)
operation and died at the 60th postoperative day. Post- Approach
mortem examination did not show TEF recurrence. Cervicotomy 15 (75)
At the end of data collection all the surviving patients
Cervicotomy þ partial sternotomy 3 (15)
were breathing normally and with normal oral intake.
Right thoracotomy 2 (10)
Two patients have tracheal appliances (one Montgomery
Tracheal management
T-Tube and one silicone Dumon Stent). Patients were
Primary tracheal resection and anastomosis 9 (45)
followed for 27.9
22 months (range, 3 to 80 months).
Anterior tracheal incision without tracheal 6 (30)
Six patients were considered lost to follow-up although
resection
they were followed for at least 24 months.
Membranous wall suture 3 (15)
Primary laryngotracheal reconstruction 2 (10)
Comment Postoperative tracheostomy 7 (35)
Esophageal management
Acquired benign TEF usually derives from a complication
2-Layer closure 19 (95)
of intubation with a cuff-related tracheal injury, has
Esophagostomy, closure of the defect and 1 (5)
serious consequences to the patient’s health status, and diversion þ jejunostomy
its surgical management can be challenging [4]. In the Muscle flap interposition 20 (100)
current series, 80% of TEFs occurred as a consequence of
Ann Thorac Surg BIBAS ET AL 1085

GENERAL THORACIC
2016;102:1081–7 BENIGN TRACHEOESOPHAGEAL FISTULAS

Table 4. Complications outcomes. Similar data have been reported by Altorjay

Complication
Subcutaneous emphysema
Pneumonia
Wound infection
Deep vein thrombosis
Acute renal failure (non-dialysis)
Pulmonary embolism
Pulmonary sepsis
Bacteremia
Tracheal dehiscence
Respiratory failure þ mechanical ventilation
Death
TEF recurrence
Hoarseness (laryngeal nerve palsy)
Recurrent tracheal stenosis
TEF ¼ tracheoesophageal fistula.
The success rates after surgical treatment of benign
acquired TEFs vary according to the literature (Table 5),
and fistula recurrence rates of up to 11% have been
reported [1, 2, 5, 11, 9, 12–14]. Although there are no
published predictors for TEF recurrence, previous
attempted repairs, prior esophagectomy, and laryngec-
tomy seem to be associated with poor outcomes [2, 12]. In
the report by Muniappan and colleagues [2] 13 patients
had failed initial surgical or endoscopic repairs; one fis-
tula recurred and could not be closed. Furthermore, all
four recurrent fistulas occurred after one prior surgical
failure or presented after esophagectomy or laryngec-
tomy. Other series have reported complications and
fistula recurrence after prior failed operations, often
needing a permanent tracheal appliance [1, 12]. In our
cohort, 3 patients had failed prior TEF repairs. Only one
individual had an uneventful recovery. The other two
have indwelling tracheal stents. However, both are able
to eat and breathe normally, without disturbance in
their daily activities.
Therefore, individuals with initial failed repairs of TEF
should not be denied surgical procedures [2]. Camargo
and colleagues [11] operated on 3 cases of recurrent TEF,
and only 1 patient experienced a small tracheal dehis-
cence. All 3 reported cases had good long-term functional
and colleagues [5]. In a 10-year series, the group treated
n (%)
8 patients with recurrent TEF; 1 patient had been oper-
4 (20) ated on four times. The investigators used combined
3 (15) approaches in 7 cases, and 6 patients required a thora-
2 (10) cotomy. All TEFs were successfully repaired.
2 (10) The surgical management of the tracheal and esopha-
2 (10) geal defects is still controversial. A myriad of tech-
1 (5) niques have been described to treat a TEF [5, 9, 12, 13].
1 (5) Nonetheless, the procedure described by Grillo and col-
1 (5) leagues [15], with resection of the stenotic segment (when
1 (5) present), primary tracheal anastomosis, and two-layer
1 (5) esophageal repair is the most straightforward [9, 12, 14].
1 (5)
It does not require extensive tissue devascularization,
allows complete exposure of the entire tracheoesophageal
1 (5)
damage, and the recurrent laryngeal nerves are less likely
1 (5)
to be injured [9]. The decision to resect the trachea is
3 (15)
based on two factors: concomitant tracheal stenosis, or
extensive damage to the posterior wall of the trachea
because of the TEF. In our series, 11 patients needed
tracheal resection (55%), and this subgroup of patients
had the greatest number of severe complications, such
as TEF and tracheal stenosis recurrence. One patient
that underwent a laryngotracheal resection and closure
of the TEF died 60 days after the operation. The high
number of tracheal complications (15%) in our series
could be explained because most of our patients had
postintubation injury, and 75% of the cohort had a tra-
cheostomy before the definitive repair of the TEF.
The need for tracheostomy in TEF patients is yet to be
defined. Reed and Mathisen [4] stated that to minimize
contamination through the TEF, a new tracheostomy
tube should be placed, with the cuff located distal to the
fistula. We believe that tracheostomy is not mandatory
in all cases and should be done only if patients have
large fistulas, uncontrollable salivary aspiration, and
continuous contamination of the airway (Fig 2A). Half of
the patients referred to our center already had a trache-
ostomy. Because we are a reference center for airway
diseases, we receive patients from remote sites that
often do not have an attending general thoracic surgeon.
Thus, we believe that in some cases the tracheostomy
could have been avoided, because patients had fistulas
smaller than 5 mm and had mild pulmonary symptoms
(Figs 2B, 2C). In this subgroup of patients, aspiration
might be controlled with cessation of oral intake and

Table 5. Summary of Published Results in TEF Repair

Study Year n PII, % TS, % PFR, % Morbidity, % Mortality, % TEF Recurrence, %

Bibas and colleagues 2016 20 80 55 15 55 5 5

Muniappan and colleagues [2] 2013 36 47 61 36 56 2.8 11
Shen and colleagues [12] 2010 21 9.5 9.5 22.8 54.3 5.7 8.6
Camargo and colleagues [11] 2010 16 93.8 100 18.8 25 0 0
Macchiarini and colleagues [9] 2000 32 NI 43.8 63 22 3.1 3.2
Baisi and colleagues [13] 1999 29 74.2 3.4 0 NI 3.4 0
Mathisen and colleagues [1] 1991 38 71.1 81.6 21.1 NI 10.5 7.9

NI ¼ not informed; PFR ¼ previous failed TEF repair; PII ¼ postintubation injury; TEF ¼ tracheoesophageal fistula; TS ¼ tracheal
stenosis.
 1086 BIBAS ET AL Ann Thorac Surg
GENERAL THORACIC

BENIGN TRACHEOESOPHAGEAL FISTULAS 2016;102:1081–7

decision to place preoperative tracheal stents was indi-

vidualized. One patient had a 5-mm tracheal stenosis
and no tracheostomy. It was decided to dilate and put a
Dumon stent rather than do a tracheostomy. One patient
had a previous tracheal resection and already used a
T-Tube. The other 2 patients already had tracheos-
tomies, but they were unable to speak and were not fully
adapted to the tracheostomy. In our institution, we do not
advocate the use of esophageal stents to decrease aspi-
ration. Full closure of the defect is not warranted, even if
the stent is covered [18]. Furthermore, they are technically
difficult to remove, may enlarge the fistula, or even
migrate into the trachea [18, 19].
In this series, tracheal complications were dealt with
silicone stenting [1, 2, 9, 11, 12, 13]. It is our routine to
initially use silicone T-Tubes because they are easy to
place, are safe, and have a low migration rate. Moreover,
the lateral limb of the tube can be opened if needed
[16, 17]. After adaptation, the T-Tube can be switched to
a Dumon stent. After long-term airway stenting, removal
of the tracheal appliance is possible if there is no residual
stenosis or malacia and if the mucosa has no signs of
inflammatory activity [17].
The main limitation of this study is its retrospective
nature. However, data were collected from a prospec-
tively filled electronical database, which ensured the
quality and the reliability of the information. Further-
more, it is a single-center experience, so our results
might not be fully generalizable to other institutions,
because TEF causes differ according to the reported
studies [1, 2, 5, 12].
In conclusion, this single-institution experience with
the surgical treatment of TEF illustrates that despite good
long-term results, even in high-volume academic centers,
operation for TEF carries important morbidity and mor-
tality. Because TEF is a rare disease, it is unlikely that
prospective trials will be performed. Therefore, retro-
spective multicenter studies should be performed to
identify predictors for TEF and tracheal stenosis recur-
rence, mortality, and the impact of the different tech-
niques in the surgical outcomes.

References
Fig 2. (A) Large tracheoesophageal fistula (TEF) with nasoenteric
tube. Patient was referred without tracheostomy and had recurrent
pulmonary infection. Tracheostomy and gastrostomy were done.
(B) Small TEF (2 to 3mm); patient had a gastrostomy and no signs of
pulmonary aspiration or infections. Tracheostomy was not performed.
(C) Small TEF with tracheostomy previously done in another
institution.
nutrition through a gastrostomy tube. If necessary, a
jejunostomy tube can be placed through the gastrostomy.
Another issue that is still subject to debate is the use
of tracheal and esophageal stents before TEF closure. In
our series, 4 patients had tracheal stents before the
operation. Endoscopic placement of tracheal stents adds
little damage to the tracheal wall and allows phonation
and humidification of the upper airways [16, 17]. The
1. Mathisen DJ, Grillo HC, Wain JC, Hilgenberg AD. Man-
agement of acquired nonmalignant tracheoesophageal fis-
tula. Ann Thorac Surg 1991;52:759–65.
2. Muniappan A, Wain JC, Wright CD, et al. Surgical treatment
of nonmalignant tracheoesophageal fistula: a thirty-five year
experience. Ann Thorac Surg 2013;95:1141–6.
3. Couraud L, Ballester MJ, Delaisement C. Acquired trache-
oesophageal fistula and its management. Semin Thorac
Cardiovasc Surg 1996;8:392–9.
4. Reed MF, Mathisen DJ. Tracheoesophageal fistula. Chest
Surg Clin N Am 2003;13(2):271–89.
5. Altorjay A, Mucs M, R€ ull M, et al. Recurrent, nonmalignant
tracheoesophageal fistulas and the need for surgical impro-
visation. Ann Thorac Surg 2010;89:1789–96.
6. Bibas BJ, Terra RM, Oliveira Junior AL, et al. Predictors for
postoperative complications after tracheal resection. Ann
Thorac Surg 2014;98:277–82.
7. Pearson FG, Brito-Filomeno L, Cooper JD. Experience with
partial cricoid resection and thyrotracheal anastomosis. Ann
Otol Rhinol Laryngol 1986;95(6 Pt 1):582–5.
Ann Thorac Surg
2016;102:1081–7
8. Grillo HC, Mathisen DJ, Wain JC. Laryngotracheal resection
and reconstruction for subglottic stenosis. Ann Thorac Surg
1992;53:54–63.
9. Macchiarini P, Verhoye JP, Chapelier A, Fadel E,
Dartevelle P. Evaluation and outcome of different surgical
techniques for postintubation tracheoesophageal fistulas.
J Thorac Cardiovasc Surg 2000;119:268–76.
10. Scordamaglio PR, Tedde ML, Minamoto H, Pedra CA,
Jatene FB. Endoscopic treatment of tracheobronchial tree
fistulas using atrial septal defect occluders: preliminary re-
sults. J Bras Pneumol 2009;35:1156–60.
11. Camargo JJ, Machuca TN, Camargo SM, Lobato VF,
Medina CR. Surgical treatment of benign tracheo-
oesophageal fistulas with tracheal resection and oesopha-
geal primary closure: is the muscle flap really necessary? Eur
J Cardiothorac Surg 2010;37:576–80.
12. Shen KR, Allen MS, Cassivi SD, et al. Surgical management
of acquired nonmalignant tracheoesophageal and bron-
choesophageal fistulae. Ann Thorac Surg 2010;90:914–8; dis-
cussion 919.
DISCUSSION
DR CHADRICK E. DENLINGER (Charleston, SC): Thank you.
These are often challenging problems. One question I have for
you is a technical question.
When you incise the trachea anteriorly, do you resect a
segment of the trachea or just simply open and then close it?
Could this affect your postoperative stenosis rate?
DR BIBAS: Yes. Initially if a patient had a small fistula, we would
do a lateral approach. But then as described by Macchiarini
and colleges in 2000, it is easier to just open the trachea, close
the defect, and then suture the trachea back again.
We do not consider that a formal resection, but you might
just resect one or two tracheal rings, but it is often not necessary.
So you can just open the trachea, dissect the posterior wall of the
trachea and the esophagus, close the defect, put a muscle flap,
and then suture the trachea back on.
DR DENLINGER: Great. Thank you.
DR JESSICA S. DONINGTON (New York, NY): Nutrition is
incredibly important in these patients, do you support them with
percutaneous endoscopic gastrostomy (PEG) or feeding tubes, or
do most of these patients eat immediately after the surgical
procedure?
BIBAS ET AL 1087
GENERAL THORACIC
BENIGN TRACHEOESOPHAGEAL FISTULAS
13. Baisi A, Bonavina L, Narne S, Peracchia A. Benign trache-
oesophageal fistula: results of surgical therapy. Dis Esoph-
agus 1999;12:209–11.
14. Fiala P, Cernohorský S, Cerm ak J, Patek J, Krepela E,
Mouckov a M. Tracheal stenosis complicated with trache-
oesophageal fistula. Eur J Cardiothorac Surg 2004;25:127–30.
15. Grillo HC, Moncure AC, McEnany MT. Repair of inflamma-
tory tracheoesophageal fistula. Ann Thorac Surg 1976;22:112–9.
16. Bibas BJ, Bibas RA. A new technique for T-tube insertion in
tracheal stenosis located above the tracheal stoma. Ann
Thorac Surg 2005;80:2387–9.
17. Terra RM, Bibas BJ, Minamoto H, et al. Decannulation in
tracheal stenosis deemed inoperable is possible after long-
term airway stenting. Ann Thorac Surg 2013;95:440–4.
18. Eleftheriadis E, Kotzampassi K. Temporary stenting of ac-
quired benign tracheoesophageal fistulas in critically ill
ventilated patients. Surg Endosc 2005;19:811–5.
19. Han Y, Liu K, Li X, et al. Repair of massive stent-induced
tracheoesophageal fistula. J Thorac Cardiovasc Surg 2009;
137:813–7.
DR BIBAS: Half of our patients had some sort of nutritional
support before the operation, either with a PEG or jejunostomy.
In the postoperative period, if the patient did not have a PEG
or a jejunostomy, we would put in a nasoenteral tube. We did
not do a gastrostomy in all patients. We do not think it is
necessary.
DR DONINGTON: What is the time frame between seeing them
in the clinic and taking them for the surgical procedure? You say
most of them come to see you as outpatients. Some of them
appear to have had very large fistulas.
Is there a standard time frame before you are able to operate
on them, and does it take a while to get them ready for this type
of large operation?
DR BIBAS: That is a good question, and it is challenging because
they get to us most of the time in our outpatient clinic, so they
have lost a lot of weight.
We have scheduled routine consultations in the nutrition
department. We only operate when they are cleared to go. It is
challenging sometimes to keep them at the ideal weight. It is not
easy.
In the meantime, we might have to do a tracheostomy to
control pulmonary infections.