HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Correction (v73,p2038) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Henning A. Gaissert Cameron D. Wright James S. Allan John C. Wain Hermes C. Grillo Douglas J. Mathisen Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mangi, A. A. Articles by Mathisen, D. J. Search for Related Content PubMed PubMed Citation Articles by Mangi, A. A. Articles by Mathisen, D. J.

Ann Thorac Surg 2002;73:911-915
© 2002 The Society of Thoracic Surgeons

---

Original article: general thoracic

Benign broncho-esophageal fistula in the adult

^a Division of Thoracic Surgery, Department of Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA

Accepted for publication November 20, 2001.

* Address reprint requests to Dr Mathisen, Division of Thoracic Surgery, Massachusetts General Hospital, Blake 1570, 55 Fruit St, Boston, MA 02114 USA

	Abstract

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Background. Benign broncho-esophageal fistula (BEF) in the adult is rare, and occurs as a complication of inflammatory disorders, foreign body ingestion, or congenital anomalies. Nonspecific symptoms lead to a delay in diagnosis.

Methods. The charts of 13 patients from 1960 to 2001 at the Massachusetts General Hospital were retrospectively reviewed.

Results. Nine patients had chronic cough, which worsened upon ingestion. Four patients developed BEF after prior thoracic surgery, and 3 after histoplasmosis. Silicosis, foreign body ingestion, lye ingestion, bronchogenic cyst, esophageal diverticulum, and a congenital anomaly caused BEF in 1 patient each. Barium swallow was the most useful diagnostic test. Fistulas most often arose from the right bronchial tree and communicated with the distal esophagus. Management involved excision of the tract, primary closure of the bronchus and esophagus, and interposition of vascularized tissue. There was one perioperative failure, but no long-term recurrences after successful surgical closure.

Conclusions. The majority of benign BEF in adults are acquired, and result from mediastinal inflammation. Accurate recognition and surgical closure prevents persistent uncontrolled aspiration and pulmonary sepsis.

	Introduction

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Most broncho-esophageal fistulas (BEF) in the adult population develop as a result of locally advanced bronchogenic or esophageal malignancy. These patients often succumb to sudden aspiration. In contrast, patients with benign BEF often have a relatively insidious clinical course. The combination of nonspecific symptoms and the rarity of benign BEF often result in long intervals without diagnosis and proper treatment. Significant morbidity and death can occur. We treated 10 patients with benign BEF over the last decade. Of these patients, 5 were treated within the last 3 years. A total of 75% experienced symptoms that went undiagnosed for variable periods.

We have reviewed our experience with benign BEF treated at our hospital over 40 years. This series, the largest from a single hospital, delineates the mode of presentation of BEF in adults and describes appropriate surgical management.

	Material and methods

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Patient selection and chart review
Case records of the Massachusetts General Hospital were reviewed from 1960 to 2001. A total of 228 cases of broncho-esophageal fistulas were identified. Of these, 215 were due to bronchogenic or esophageal malignancy; these patients were excluded from the study. The remaining 13 patients had benign etiologies. With approval by the institutional review board, the charts of these 13 patients were retrospectively reviewed. The world literature was also reviewed using a MEDLINE search (1966 to 2001), and a secondary review of referenced works was performed.

Surgical technique
Among the 13 patients, the median age was 53 years (range 17 to 80 years, mean 47 years); 7 patients were male and 6 female. All 13 patients underwent bronchoscopy and esophagoscopy in the operating room before thoracotomy to verify presence of the fistula. The standard operative procedure consisted of right posterolateral thoracotomy, exposure and division of the fistula, primary repair of bronchus and esophagus, and pedicled tissue interposition. The bronchus was closed primarily in all cases with interrupted 4-0 Vicryl (Ethicon, Somerville, NJ) sutures, the esophagus was repaired with two layers of 4-0 silk, and viable tissue was interposed between the suture lines. A posteriorly pedicled intercostal muscle flap was used in 7 cases, omentum in 2, and pericardial fat or thickened pleura in 1 case each. Three patients did not undergo the standard procedure, and technical variations are presented in this report.

	Results

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
Etiology
Three of 13 patients lived in areas endemic for histoplasmosis (patients 7, 9 and 10; Table 1). Characteristic radiologic, clinical, and pathologic features (such as sinus histiocytosis in large calcified lymph nodes), along with positive serum titers for histoplasmosis, supported this as the etiology. An ingested wire (patient 8), and lye ingestion (patient 1) were responsible in 1 case each. Patient 4 developed a BEF related to an infected bronchogenic cyst, which had not been completely removed on three prior attempts at resection. The last attempt was 2 years before the patient was referred to our hospital. Patient 5, a worker in a talc mine, developed extensive mediastinal silicosis. The erosion of hyalinized and silicotic mediastinal lymph nodes resulted in a fistula. In 4 patients, BEF developed after esophageal surgery. After esophagectomy, 1 patient had an anastomotic leak (patient 11) and another (patient 12) developed dehiscence of the gastric staple line, resulting in communication to the right lower lobe and in right main bronchus, respectively. The third patient (patient 3) had an antireflux procedure 8 years before presentation of a fistula between the left lower lobe bronchus and hernia repair. The fourth patient (patient 13) was noted to have a fistula after attempted primary esophageal repair of spontaneous postemetic esophageal rupture (Boerhaave’s syndrome) at another hospital. Patient 6 developed a right lower lobe fistula at the tip of a long esophageal diverticulum. Only patient 2 had a congenital broncho-esophageal fistula. This 25-year-old woman had had symptoms her entire life with multiple admissions for pulmonary suppuration, and had undergone extensive study before the diagnosis was established by barium swallow. The fistula was repaired as described and was lined with stratified squamous epithelium.

Physical findings were nonspecific except in 2 cases. These included drainage of ingested methylene-blue from an empyema drain in patient 4 after three abortive attempts at excising an infected bronchogenic cyst; and an air leak through a gastrostomy tube in patient 8, an autistic child who was transferred to our hospital intubated with mediastinal sepsis after ingesting a wire.

The clinical event that precipitated the appropriate diagnostic study included coughing on ingesting food (2 patients), hemoptysis (2 patients), severe chest pain, severe exacerbation of a chronic cough, worsening gastroesophageal reflux, new pleural effusion, and fever/obtundation in a patient with autism (1 patient each). The diagnosis was established by barium swallow in 8 patients (Fig 1), bronchoscopy in 3 patients, esophagoscopy in 1, and computed tomography in 1. The diagnosis of BEF was made at outside institutions in 12 patients, after which they were transferred to the Massachusetts General Hospital for further management. The median delay in diagnosis was 18 months (mean 46 months, range 1 week to >20 years). Fistulas in 2 patients (patient 12 and 13) were diagnosed in the immediate perioperative period by bronchoscopy and barium swallow, respectively.

View larger version (63K): [in this window] [in a new window]   Fig 1. Demonstration of broncho-esophageal fistula on barium swallow. (CA = cancer; LLL = left lower lobe bronchus; L. MAIN = left main stem bronchus; RLL = right lower lobe bronchus; R. MAIN = right main stem bronchus; s/p = status post.)

Bronchoscopic and endoscopic findings
Bronchoscopic abnormalities included inflammatory changes in all 13 patients, a discrete focus of heaped up granulations in 4, and visualization of the fistula orifice in 5. An endobronchial mass was seen in two patients. In one patient, the immediate appearance in the bronchus of dyed saline instilled in the esophagus confirmed the fistula. The most common finding on esophagoscopy was that of a papilla or a dimple in 6 patients. Insufflation of the esophagus resulted in an escape of air through the endotracheal tube in 1 patient. One had esophageal and gastric necrosis after lye ingestion.

Anatomy of fistula
The course of the fistula tract varied (Fig 2). The most common course (in 7 of 13 patients) was from the right bronchial tree to the distal esophagus, 27 to 30 cm from the incisors. The right mainstem bronchus was the site of airway entry in 5 patients, followed by the right lower lobe bronchus (3 patients) and the bronchus intermedius (2 patients). The left mainstem bronchus was involved twice and the left lower lobe bronchus once. The fistula drained into the distal esophagus in 9 patients and the midesophagus in 4. The diameter and length were recorded in 3 patients, and ranged from 2 mm to 2 cm in diameter, and 4 mm to 4 cm in length. Because fistula diameter was recorded in only 3 patients, we were unable to satisfactorily demonstrate a correlation between fistula size and severity of symptoms. The largest fistula (2 x 4 cm) occurred in patient 1, who succumbed to complications of lye ingestion. A total of 12 patients had one fistula orifice, and 1 had two pinpoint adjacent orifices. The fistula was lined with epithelium in only 2 of the 10 specimens submitted for pathologic examination. The fistula tract from patient 4 was lined with metaplastic respiratory epithelium; that from patient 6 was lined stratified squamous epithelium. Only one fistula swab grew any organisms (enterococcus); this was in patient 4.

View larger version (107K): [in this window] [in a new window]   Fig 2. Schematic representation of course of broncho-esophageal fistulas in adults, showing correlation with etiology.

The standard procedure could not be performed in 3 patients. The first was a critically ill patient (patient 1) who developed a large 2 x 1 cm fistula between a necrotic esophagus and the left mainstem bronchus after lye ingestion. This patient underwent cervical division and diversion of the esophagus with placement of gastrostomy and jejunostomy tubes, but succumbed to pulmonary sepsis. The second was an 80-year-old woman (patient 3) who presented with intermittent copious hemoptysis, respiratory failure secondary to continuous aspiration, and progressive abdominal distension. This patient went into pulmonary edema during bronchoscopy, and later expired in the intensive care unit. The third patient (patient 12) was malnourished, with an albumin of 1.0; she developed BEF secondary to dehiscence of the gastric staple line after Ivor-Lewis esophago-gastrectomy. The bronchial defect was repaired primarily and buttressed with a posteriorly pedicled intercostal flap; the stomach was debrided and returned to the abdomen; and cervical esophagostomy was performed. The patient was discharged home on postoperative day 17; she had intestinal continuity restored by colonic interposition 4 months later, from which she recovered uneventfully.

Five patients developed postoperative complications. Patient 13 had a persistent pinhole fistula on barium swallow, probably related to stasis in the esophageal remnant. Patient 7, with mediastinal histoplasmosis, required thoracotomy and ligation of the thoracic duct for postoperative chyle leak. She was discharged home on postoperative day 16. Patient 8 needed prolonged postoperative intubation for pulmonary toilet and airway protection. He was eventually discharged to rehabilitation after 43 days. Transient subglottic edema in patient 6, necessitated reintubation on postoperative day 4, with extubation 2 days later, and discharge on postoperative day 13. Patient 12 needed thoracentesis for a persistent pleural effusion.

The other 6 patients were discharged on postoperative day 10. Median length of stay for all patients was 13 days. Nine of 10 surviving patients remain free of recurrence 2 to 27 years after surgery. Patient 13 refused further attempts at closure. The persistence of symptoms despite several treatment modalities preoperatively, and the freedom from symptoms postoperatively, suggest that symptoms were indeed due to BEF.

	Comment

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
In 1916, Heiderich provided the first demonstration of a fistula that coursed from the esophagus to the right main stem bronchus [1]. Since then, only about 120 cases of benign BEF have been reported in the world literature. The largest multi-institution series, published in 1966, discussed 260 cases of benign and malignant fistulas in infants, children, and adults [2]. The largest single institutional series discussing benign BEF in adults was published in 1965 by Braimbridge and Keith [3], who reviewed 3 cases from their own experience and 20 cases that had been reported in the literature until that time. They suggested a classification system based on this data that still has utility today. Type I fistula originates at the tip of a wide-necked esophageal diverticulum. Type II extends from the esophagus to a lobar or segmental bronchus. These are most common, comprising approximately 60% of all broncho-esophageal fistulas. Type III fistulas extend from a bronchogenic cyst to the esophagus, and type IV fistulas from the esophagus to a pulmonary sequestration. In our series, type II fistulas were seen in 85% (11 of 13) of patients, types I and III once each, and we encountered no patients with type IV fistulas.

Given the intimate anatomic relationship between the left main bronchus and the esophagus, the preponderance of right-sided fistulas appears counterintuitive. One hypothesis that explains this observation differentiates BEF that occur as a result of intrinsic esophageal disease (such as ingestions or Boerhaave’s syndrome) from those that are caused by conditions resulting from primary mediastinal inflammation (such as histoplasmosis and silicosis). Broncho-esophageal fistulas resulting from intrinsic esophageal disease are caused by direct erosion of the esophagus into the membranous wall of the bronchus, and therefore tend to be left-sided, as in patients 1, 3, and 13. Right-sided fistulas are more common in primary mediastinal inflammatory states, presumably because of the involvement of lymph nodes that lie between the esophagus and the less-angulated right-sided bronchial tree, as in patients 4, 5, 7, 9, and 10.

There are important differences between benign tracheoesophageal fistulas (TEF) and BEF. In adults, TEF most often occurs as a complication of endotracheal intubation, particularly in association with a naso- or orogastric tube. Symptoms related to aspiration and inadequate ventilation in TEF are often pronounced; and diagnostic delays are unusual as opposed to BEF. Surgical treatment is preferably deferred for TEF until weaning from mechanical ventilation is accomplished [5]. In most cases of TEF, circumferential resection of the airway is required, unlike BEF, which does not typically require a resection.

Risher and associates [4] termed 109 of their collected BEF congenital. The liberal use of this term may mask a lack of comprehensive efforts to definitively establish the diagnosis of a BEF—a condition that, in our experience, is largely an acquired one (Table 1). When congenital fistulas do occur in adults, they tend to be small and are not associated with esophageal atresia. It has been suggested that, even in adults, the most common etiology of BEF is the congenital failure of separation of the tracheal band from the esophagus during embryogenesis. Explanations of why symptoms manifest themselves only in adulthood are less satisfying [3, 4, 6–10].

A diagnosis of broncho-esophageal fistula should be suspected in any patient who presents with persistent unexplained cough. Most of our patients were given a diagnosis of chronic bronchitis or adult onset asthma. Such a diagnosis should be questioned and further evaluation pursued when symptoms fail to respond to established medical therapy. The rarity of this entity as well as the relatively nonspecific symptoms and signs makes the diagnosis challenging. Hemoptysis and coughing on ingestion are late signs, but this is often when the diagnosis is established for the first time. Often bronchoscopy is needed to discover the fistula. The potential risks of an untreated small BEF are chronic sepsis and pulmonary failure. Although we have not seen severe progressive deterioration due to a fistula of limited size, closure is nonetheless important to avoid chronic pneumonia and disability.

Early and definitive surgical intervention with meticulous dissection of the fistula, division, and primary closure of bronchus and esophagus, along with interposition of viable tissue (eg, pedicled intercostal muscle flap, omentum, pericardial fat, or thickened pleura) is the procedure of choice [11]. Alternatives to this strategy mentioned in the literature include division of the fistula using a surgical stapling device [12], and endoscopic application of sodium hydroxide and acetic acid to both the bronchial and esophageal sides of the tract [13]. These approaches seem to have little relevance to most patients.

A high degree of suspicion is required to make a prompt diagnosis. A single-stage repair of the fistula with a viable tissue buttress interposed between the esophageal and bronchial repair will achieve success in the vast majority of patients.

	Acknowledgments

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 
The authors thank Edith Tagrin for the excellent artwork that accompanies this article.

	References

Top Abstract Introduction Material and methods Results Comment Acknowledgments References

 

1. Heiderich H. Pathologich-anatomische demonstrationen. Dtsch Med Wochenschr 1916;42:340.
2. Blackburn W.R., Amoury R.A. Congenital esophago-pulmonary fistulas without esophageal atresia: an analysis of 260 fistulas in infants, children and adults. Rev Surg 1966;23:153-175.[Medline]
3. Braimbridge M.V., Keith H.I. Oesophago-bronchial fistula in the adult. Thorax 1965;20:226-233.
4. Risher W.H., Arensman R.M., Ochsner J.L. Congenital broncho-oesophageal fistula. Ann Thorac Surg 1990;49:500-505.[Abstract]
5. Mathisen D.J., Grillo H.C., Wain J.C., Hilgenberg A.D. Management of acquired nonmalignant tracheoesophageal fistula. Ann Thorac Surg 1991;52:759-765.[Abstract]
6. In: Jackson C., Coates G.M., eds. The nose, throat, and ear, and their diseases. Philadelphia: WB Saunders, 1929:1124.
7. Negus V.E. Oesophagus from a middle aged man, showing a congenital opening into the trachea. Proc R Soc Med 1929;22:527.
8. Demong C.V., Grown J.B., Heitzman G.C. Congenital tracheo-esophageal fistula without atresia of the oesophagus. Am Surg 1959;25:156-162.[Medline]
9. Weissburg D., Kaufman M. Bronchoesophageal fistula in adults: congenital or acquired?. J Thorac Cardiovasc Surg 1990;99:756-757.[Medline]
10. Parry W., Juma A. Congenital bronchoesophageal fistula. Ann Thorac Surg 1991;51:346-347.[Medline]
11. Grillo H.C., Wilkins E.W., Jr Esophageal repair following late diagnosis of intrathoracic perforation. Ann Thorac Surg 1975;20:387-399.[Abstract]
12. Smith B.D., Jr, Mikaelian D.O., Cohn H.E. Congenital broncho-esophageal fistula in the adult. Ann Otol Rhinol Laryngol 1987;96:65-67.[Medline]
13. Smith D.C. A congenital broncho-esophageal fistula presenting in adult life without pulmonary infection. Br J Surg 1970;57:398-400.[Medline]

This article has been cited by other articles:

				H. K. Kim, Y. S. Choi, K. Kim, J. Kim, and Y. M. Shim Long-term results of surgical treatment in benign bronchoesophageal fistula J. Thorac. Cardiovasc. Surg., August 1, 2007; 134(2): 411 - 414. [Abstract] [Full Text] [PDF]

				S. Griffo, P. Stassano, G. Iannelli, L. Di Tommaso, M. Cicalese, M. Monaco, and G. Ferrante Benign bronchoesophageal fistula: Report of four cases J. Thorac. Cardiovasc. Surg., May 1, 2007; 133(5): 1378 - 1379. [Full Text] [PDF]

				O. N. Pages, S. Rubin, and B. Baehrel Intra-esophageal rupture of a bronchogenic cyst Interactive CardioVascular and Thoracic Surgery, August 1, 2005; 4(4): 287 - 288. [Abstract] [Full Text] [PDF]

				M. P. Devbhandari, L. Raco, M. T. Hendrickse, and A. J. Duncan Congenital Bronchoesophageal Fistula in a Patient With Crohn's Disease: A Cautionary Tale Ann. Thorac. Surg., May 1, 2005; 79(5): 1776 - 1777. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Correction (v73,p2038) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Henning A. Gaissert Cameron D. Wright James S. Allan John C. Wain Hermes C. Grillo Douglas J. Mathisen Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mangi, A. A. Articles by Mathisen, D. J. Search for Related Content PubMed PubMed Citation Articles by Mangi, A. A. Articles by Mathisen, D. J.