Endoscopic Closure of Esophagobronchial Fistula With Fibrin Glue

Ogunmola, Nicholas*; Wyllie, Robert*; McDowell, Karen†; Kay, Marsha*; Mahajan, Lori*

Journal of Pediatric Gastroenterology & Nutrition: May 2004 - Volume 38 - Issue 5 - pp 539-541
Case Reports

Departments of *Pediatric Gastroenterology and Nutrition and †Pediatric Pulmonology, The Cleveland Clinic Foundation, Cleveland, Ohio, U.S.A.

Address correspondence and reprint requests to Dr. Robert Wyllie, Chairman, Department of Pediatric Gastroenterology and Nutrition, The Cleveland Clinic Foundation, Desk A111, 9500 Euclid Avenue, Cleveland, OH, U.S.A. 44195 (e-mail: wyllier@ccf.org).

Article Outline

Operative repair of primary and recurrent tracheoesophageal fistula (TEF) is associated with a high rate of failure and postoperative complications (1–10). Surgical division and closure of recurrent fistulas may be difficult or unsuccessful because of dense adhesions or poor blood supply to the ends of the fistula (1). The first successful bronchoscopic fistula closure using a tissue adhesive was reported in 1975. Subsequently, a variety of obliterating agents have been used (3,11). Histoacryl cement (n-butyl-2-cyanoacrylate), aethoxysclerol, and 30% NaCl have been used in addition to fibrin glue alone or in combination, with mixed results (1–4). Endoscopic diathermy has also been used but typically requires multiple sessions (4,5). We report the successful closure of a recurrent esophagobronchial fistula with fibrin glue in a pediatric patient during simultaneous endoscopy and bronchoscopy. To our knowledge, this technique has not been reported previously in pediatric patients.

Back to Top | Article Outline


The patient was a 5-year-old girl with VACTERL syndrome, bronchopulmonary dysplasia and asthma who had a 2-month history of increasing respiratory symptoms, midsternal chest pain, recurrent emesis and choking. Her symptoms were worse with oral feeding. She was born prematurely with an omphalocele, pulmonary atresia, ventricular septal defect and interrupted inferior vena cava, all of which were treated surgically during the newborn period. She had a type C TEF characterized by esophageal atresia with a distal fistula. She had a gastrostomy placed on day one of life and a subsequent Nissen fundoplication for chronic gastroesophageal reflux, which required multiple revisions. The repair of her TEF required a left main-stem bronchus transection and repair because of a common wall between the bronchus and fistula. She subsequently experienced homograft stenosis that required surgical revision. Distal tracheal and left main-stem bronchial stents were inserted for severe tracheobronchial malacia, and they also required multiple surgical revisions. The tracheal stent was removed because of disintegration and migration of the stent. She had an iatrogenic tracheoesophageal cleft that required surgical repair. She had recurrent pneumonias and reactive airway disease.

At the time of evaluation her weight and height were at the fiftieth percentile for age. She had bilateral expiratory wheezing, basilar rales, and a Grade II/VI systolic murmur. Examination results were otherwise unremarkable. A barium swallow demonstrated a dilated proximal esophagus with poor motility, a bronchoesophageal fistula at the origin of the left main bronchial stent, a hiatal hernia and a partial breakdown of the fundoplication. Reflux into the hiatal hernia was noted without gross reflux into the esophagus.

Her cardiac disease, previous thoracic surgeries, and chronic respiratory disease placed her at a high risk for an open surgical procedure. Therefore, endoscopic therapy was attempted. Under general anesthesia, the patient was ventilated with a size 2 laryngeal mask airway (LMA). A flexible pediatric fiberoptic bronchoscope (Olympus 3C30, 3.5 mm OD, Olympus America, Melville, NY, U.S.A.) was inserted via the LMA. Her larynx was anesthetized with 2% lidocaine to prevent laryngospasm. She had edematous vocal cords, a patent subglottic space, a tortuous trachea with moderate tracheomalacia, and a 50% narrowing of the airway lumen. Her left main-stem bronchial stent was epithelialized. The orifice of the bronchial fistula seen immediately proximal to the bronchial stent was too small to cannulate. The bronchoscope was then removed to minimize laryngospasm, and an Olympus N230 (5.3 mm OD) pediatric endoscope was inserted around the LMA into the esophagus. The fistula orifice was seen 15 cm from the central incisors (Fig. 1). The proximal esophagus was dilated, but the remainder of the endoscopic examination was unremarkable. The bronchoscope was reinserted through the LMA into the trachea and positioned once again at the orifice of the bronchial fistula. A tapered tip 5.5 Fr endoscopic retrograde cholangiopancreatography (ERCP) catheter prefilled with saline was inserted into the fistula from the esophageal side (Fig. 2). For the remainder of the procedure, the fistula was under direct endoscopic and bronchoscopic view. Saline was injected through the ERCP catheter until it appeared on the bronchial side. The catheter was withdrawn, fibrin glue (Tisseel VH, Baxter Healthcare Corporation, Glendale, CA, U.S.A.) was mixed, and the catheter was filled with fibrin glue. Approximately 0.5 mL of fibrin glue was injected into the orifice of the esophageal fistula until a white patch was seen on the bronchial side. The catheter was withdrawn, and both luminal sides of the fistula were revisualized. There was no spillage of fibrin glue into the bronchus. The patient tolerated the procedure well and was discharged home within 4 hours.

Respiratory symptoms recurred 6 months after the procedure. A repeat combined procedure was performed. Esophagoscopy revealed a small red papular swelling (reepithelialization) at the site of the fistula. An intraoperative esophagram with Omnipaque showed no fistulous connection. To confirm the absence of a fistulous connection, 5 mL of 1:1000 methylene blue was injected at the old fistula site. No dye appeared on the bronchial side. A barium swallow after this endoscopy failed to demonstrate a fistula. Twelve months after endoscopic closure of the fistula, the patient had gained more than 2 kg as a result of dramatic improvement in her oral intake of liquids and solids. Her respiratory status had significantly improved, with a reduction in her aerosol and oxygen requirements.

Back to Top | Article Outline


Combined morbidity and mortality rates as high as 50% have been reported in association with the surgical repair of TEFs, especially in patients with medically complex conditions. In recent series, the mortality rates reported have been lower (1,5,10,12,13). The most serious complication is recurrent fistula, reported in 3% to 15% of patients undergoing TEF repair (1–5,9,10,12). The success rate for endoscopic TEF closure, primarily via bronchoscopy, using a variety of agents is 55% to 80% (2,4).

The fibrin sealant used in our patient is composed of fibrinogen, thrombin, the antifibrinolytic agent aprotinin (bovine) and calcium chloride. It was introduced for endoscopic applications in Europe in the mid-1980s and approved for use in the United States by the Food and Drug Administration in 1998 (6). When mixed, the separate components of the fibrin sealant give rise to a stable, cross-linked fibrin clot, which stimulates local tissue fibroplasia. Cell rich granulation tissue is seen after three days, with proliferation of collagen fibers after 4 to 7 days. After 14 days, collagen-rich granulation tissue with markedly decreased numbers of infiltrating cells is observed (14). Clotting and adhesion time of fibrin sealant is dependent on the concentration of thrombin in the sealant after mixing.

Fibrin sealant has been used in the treatment of a variety of gastrointestinal disorders, including fistulasinano, enterocutaneous fistulas, and upper gastrointestinal bleeding (6,8,14,15). Multiple endoscopic applications may be required to complete fistula closure. The risk of viral transmission is thought to be minimal because of a vapor heating process. Five-year postmarketing surveillance data from more than one million applications contain only two reports of adverse events related to the fibrin sealant, both linked to the bovine aprotinin component (14). Fibrin, unlike other tissue adhesives, does not produce a foreign-body reaction (7). In our patient, the only technical problem encountered during the procedure was clotting of the agent in the catheter. The use of a dual lumen catheter system can prevent this problem.

To our knowledge, the use of combined endoscopic injection with simultaneous bronchoscopic visualization has not been reported in a pediatric patient. The combined approach was chosen to monitor for spillage of fibrin sealant into the airway. In one previous report, spillage of fibrin sealant was thought to have produced an acute postoperative pneumonia (7). Endoscopic closure of recurrent TEF is an attractive option in pediatric patients because it is less invasive, can be done on an outpatient basis, is less costly, and can be repeated if necessary (3). Small caliber, relatively straight fistulas of recent onset may be the most amenable to this technique (7). The technique could also be considered before surgery in patients with recurrent fistulas to improve the preoperative nutritional status and prevent pulmonary infections (1). We believe this approach is a reasonable alternative to surgery, especially in high-risk patients with complex needs.

Back to Top | Article Outline


1. Vandenplas Y, Helven R, Derop H, et al. Endoscopic obliteration of recurrent tracheoesophageal fistula. Dig Dis Sci 1993;38(2): 374–7.
2. Willetts IE, Dudley NE, Tam PK. Endoscopic treatment of recurrent tracheooesophageal fistulae: long-term results. Pediatr Surg Int 1998;13(4):256–8.
3. Hoelzer DJ, Luft JD. Successful long-term endoscopic closure of a recurrent tracheoesophageal fistula with fibrin glue in a child. Int J Pediatr Otorhinolaryngol 1999;48(3):259–63.
4. Wiseman NE. Endoscopic closure of recurrent tracheoesophageal fistula using Tisseel. J Pediatr Surg 1995;30(8):1236–7.
5. Rangecroft L, Bush GH, Lister J, et al. Endoscopic diathermy obliteration of recurrent tracheoesophageal fistulae. J Pediatr Surg 1984;19(1):41–3.
6. Groitl H, Scheele J. Initial experience with the endoscopic application of fibrin tissue adhesive in the upper gastrointestinal tract. Surg Endosc 1987;1(2):93–7.
7. Cellier C, Landi B, Faye A, et al. Upper gastrointestinal tract fistulae: endoscopic obliteration with fibrin sealant. Gastrointest Endosc 1996;44(6):731–3.
8. Rabago LR, Ventosa N, Castro JL, et al. Endoscopic treatment of postoperative fistulas resistant to conservative management using biological fibrin glue. Endoscopy 2002;34(8):632–8.
9. Gutierrez C, Barrios JE, Lluna J, et al. Recurrent tracheoesophageal fistula treated with fibrin glue. J Pediatr Surg 1994;29(12): 1567–9.
10. Engum SA, Grosfeld JL, West KW, et al. Analysis of morbidity and mortality in 227 cases of esophageal atresia and/or tracheoesophageal fistula over two decades. Arch Surg 1995;130(5): 502–8.
11. Gdanietz K, Wiesner B, Krause I, et al. [Tissue-adhesive for sealing of oesophagotracheal fistulae in children (author's translation)]. [German]. Zeitschrift fur Erkrankungen der Atmungsorgane 1974;141(1):46–50.
12. Tsai JY, Berkery L, Wesson DE, et al. Esophageal atresia and tracheoesophageal fistula: surgical experience over two decades. Ann Thorac Surg 1997;64(3):778–83.
13. Alexander F, Johanningman J, Martin LW. Staged repair improves outcome of high-risk premature infants with esophageal atresia and tracheoesophageal fistula. J Pediatr Surg 1993;28(2):151–4.
14. Dunn CJ, Goa KL. Fibrin sealant: a review of its use in surgery and endoscopy. Drugs 1999;58(5):863–86.
15. Sentovich SM. Fibrin glue for anal fistulas: long-term results. Dis Colon Rectum 2003;46(4):498–502.
© 2004 Lippincott Williams & Wilkins, Inc.