Recurrent Tracheoesophageal Fistula William Stanford, Col, Raymond G. Armstrong, L t Col, Robert E. Cline, Maj, and M. J. Williams, Col, all MC, USAF ABSTRACT Three cases of recurrent tracheoesophageal fistula have been added to the 64 reported in the literature. Recurrence is suspected in patients with a previous repair who develop recurrent pneumonia, asthmatic attacks, or choking spells. The diagnosis can be established by barium swallow roentgenography or, in difficult cases, by staining of the esophagus by contrast medium introduced into the trachea. Prevention is best accomplished by an accurate repair without tension, by strict adherence to aseptic techniques, and by using fine, nonabsorbable suture material. The interposition of a pleural flap between the trachea and esophagus diminishes the chance of an initial recurrence and is very useful at the time of reoperation for a recurrent fistula.
he occurrence of repeated episodes of choking and pneumonia months to years following repair of a tracheoesophageal fistula should suggest that there may have been a recurrence of the fistula. Admittedly this diagnosis is difficult to make 'and not infrequently is established only by persistence in efforts at its demonstration. Such was the case in each of the 3 instances of recurrent tracheoesophageal fistula seen at Wilford Hall USAF Medical Center in a ten-year period ending January, 1971. These patients form the basis of this report.
Case Reports Patient 1. A 17-month-old baby boy had had primary repair of a type IIIb (Vogt) tracheoesophageal fistula at 48 hours of age. Postoperatively he had repeated episodes of pneumonia. A barium swallow roentgenographic study one month after repair demonstrated what was initially thought to be barium aspiration into the tracheobronchial tree. A repeat barium swallow and endoscopic examination at 17 months of age showed a definite recurrence of the fistula (Figure). At reoperation a 3 mm. fistula was found at the site of the previous repair. This was divided, and pleura was interposed. T h e postoperative course was uneventful. Patient 2 . A 4-year-old boy had had complete repair of a tracheoesophageal fistula at another hospital at 3 days of age. The type of From the Thoracic Surgery Service, Department of Surgery, Wilford Hall USAF Medical Center (AFSC), Lackland AFB, Tex. Presented a t the Nineteenth Annual Meeting of the Southern Thoracic Surgical Association, Port of Spain, Trinidad, Nov. 2 4 . 1972. Address reprint requests to Dr. Stanford, Wilford Hall USAF Medical Center (SGHST), Lackland AFB, Tex. 78236.
THE ANNALS OF THORACIC SURGERY
Recurrent Tracheoesophageal Fistula
A repeat barium swallow roentgenographic study at 17 months of age shows definite recurrence of the fistula in Patient 1 . (Oblique views.)
fistula was not known. At the age of 6 months he required gastrostomy and frequent dilations for stricture. Recurrent episodes of pneumonia led to his admission to Wilford Hall USAF Medical Center, where repeat barium swallow studies disclosed a small recurrence at the site of the previous repair. T h e fistula and paraesophageal scar were excised, and pleura was interposed. T h e esophageal stricture was opened longitudinally and closed transversely. T h e postoperative course was uneventful. Patient 3 . This 3-year-old girl was born with an H type fistula 4 cm. above the carina. This was divided soon after birth. Because of repeated choking and episodes of pneumonia, she was evaluated by esophagoscopy. A previous barium swallow study had failed to reveal the fistula. Introduction of methylene blue dye into the endotracheal tube produced staining of the esophagus and verified the fistula. At reoperation a 3 mm. fistula 2 to 3 cm. above the carina was divided and pleura interposed. T h e postoperative course was uneventful.
Comment T h e diagnosis of recurrent tracheoesophageal fistula should be suspected in infants who have undergone a previous repair and who subsequently develop recurrent pneumonia, bronchitis, asthmatic attacks, or choking spells. T h e incidence of recurrence would appear to be relatively rare, with only 64 cases reported in the literature . However, we suspect the true
STANFORD E T AL.
incidence to be considerably higher, especially in patients with leaking anastomosis or stricture or both following primary repair. Demonstration of the recurrent fistula is difficult and may require persistence on the part of the radiologist to obtain a variety of oblique views of the barium-filled esophagus. Even with multiple roentgenograms the fistula may not be demonstrable, as evidenced in our third patient. However, a recurrence might be suspected if one were to detect the odor of the anesthetic in the esophagoscope at the time of endoscopy. A useful maneuver that aided in the demonstration of the fistula in 1 of our patients was the introduction of methylene blue dye into the endotracheal tube, as suggested by Kafrouni and associates . This produced staining of the esophageal mucosa and confirmed the presence of the recurrent fistula. Other techniques consist of the instillation of methylene blue dye into the esophagus during bronchoscopy [Z, 71 and the instillation of saline into the esophagus while putting positive pressure on the airway. T h e presence of bubbles in the saline sometimes signifies a fistula . The cause of recurrent tracheoesophageal fistula is unknown. It may be due to small abscesses developing in response to tissue reaction around suture material [l, 41. We currently use small Tycron suture" to minimize tissue reaction: however, some surgeons believe this to be an important enough consideration that they advocate the use of fine wire [l] to further diminish inflammatory response. Mucosal damage from stretching of an anastomotic area by repeated dilations may also lead to recurrence, as was suspected in our second patient. Prevention of recurrence follows the established surgical principles of accurate repair, which include absence of tension , careful aseptic technique, and the use of nonabsorbable sutures. These factors are especially important in esophageal operations, in which there is a greater incidence of anastomotic leak because of lack of serosal support. Finally, as with fistulas elsewhere in the body, the interposition of tissue between the esophageal and tracheal suture lines is of great importance. It not only buttresses the closure, but also prevents the possibility of leak developing from suture-to-suture approximation C1-41. We routinely interpose tissue in all primary repairs and have found a pleural flap dissected from the chest wall to be an excellent source of tissue for this purpose.
References 1. Cowley, L. L. Congenital tracheo-esophageal fistula: Recurrence after repair. A m . Surg. 33:409, 1967. 2. Falletta, G. P. Recommunication on repair of congenital tracheo-esophageal fistula. Arch. Surg. 88:779, 1964. 3. Haight, C. Congenital tracheo-esophageal fistula without esophageal atresia. J . Thorac. Surg. 17:600, 1948. *Tycron suture, Davis & Geck, American Cyanamid Co., Pearl River, N.Y.
THE ANNALS OF THORACIC SURGERY
Recurrent Tracheoesophageal Fistula 4. Hewlett, T. H. Recurrent tracheo-esophageal fistula (case report). J . Thorac. Cardiovasc. Surg. 29:658, 1955. 5. Kafrouni, G., Baick, C. H., and Wooley, M. M. Recurrent tracheoesophageal fistula: A diagnostic problem. Surgery 68:889, 1970. 6. Shaw, R. R., Paulson, D. L., and Siebel, E. K. Congenital atresia of the esophagus with tracheo-esophageal fistula: Treatment of surgical complications. A n n . Surg. 142:204, 1955. 7. Wychulis, A. R., Ellis, F. H., and Anderson, H. A. Acquired non-malignant esophago-tracheo-bronchial fistula: Report of 36 cases. J.A.M.A. 196: 117, 1966.
NOTICE FROM THE SOUTHERN THORACIC SURGICAL ASSOCIATION T h e Twentieth Annual Meeting of the Southern Thoracic Surgical Association will be held at the Galt House, Louisville, Ky., November 1-3, 1973. Reservations may be made by writing to the Reservations Manager, T h e Galt House, Louisville, Ky. 40202. Members wishing to participate in the scientific program should submit abstracts-typed double-spaced and in triplicate-to Arthur C. Beall, Jr., M.D. (Chairman of the Program Committee), 1200 Moursund Ave., Houston, Tex. 77025. T h e deadline for submission of abstracts is June 1, 1973. T h e Council of the Southern Thoracic Surgical Association has ruled that, beginning with the 1973 meeting, all slides used during presentation of papers and discussions must be of 35-mm. size. If additional information is required, inquiries should be directed to Dr. Beall. Application for membership in the Southern Thoracic Surgical Association, on forms provided by the Association, should be sent directly to Richard H. Blank, M.D. (Chairman of the Membership Committee), 1 Davis Blvd., Tampa, Fla. 33606. T h e deadline for application to membership is September 1, 1973. Papers that are accepted for the program will be considered for publication in T h e Annals and must be submitted to the Editor by October 15, 1973. W. BROOKS, M.D. Secretary-Treasurer