Acquired Adult Aerodigestive Fistula: Classification and Management

• Published in: Journal of gastrointestinal surgery Vol. 22; no. 10; pp. 1785 - 1794
• Main Authors: Qureshi, Yassar A., Muntzer Mughal, M., Fragkos, Konstantinos C., Lawrence, David, George, Jeremy, Mohammadi, Borzoueh, Dawas, Khaled, Booth, Helen
• Format: Journal Article
• Language: English
• Published: New York Springer US 01-10-2018; Springer Nature B.V
• Subjects: Adult; Aged; Anastomosis, Surgical - adverse effects; Classification; Esophageal cancer; Female; Gastroenterology; Gastrointestinal surgery; Humans; Male; Medicine; Medicine & Public Health; Middle Aged; Neoplasms - complications; Neoplasms - radiotherapy; Original; Original Article; Postoperative Complications; Radiation therapy; Radiotherapy - adverse effects; Retrospective Studies; Stents; Surgery; Survival Rate; Tracheoesophageal Fistula - classification; Tracheoesophageal Fistula - etiology; Tracheoesophageal Fistula - surgery; Young Adult; Oesophageal cancer; Aerodigestive fistula; Tracheo-oesophageal fistula
• ISSN: 1091-255X; 1873-4626
• DOI: 10.1007/s11605-018-3811-0

Background
Acquired aerodigestive fistulae (ADF) are rare, but associated with a high mortality rate. We present our experience of the diagnosis, management and outcomes of patients with ADFs treated at a tertiary centre. Utilising our findings, we propose an anatomical classification system, demonstrating how specific features of an ADF may determine management.
Methods
A clinical database was searched and 48 patients with an ADF were identified. A classification system was developed based on anatomical location of the ADF and differences in clinico-pathological features based on this categorisation were performed, with the chi-squared test used for inferential analyses and Kaplan-Meier curves with log-rank test to assess survival.
Results
Twenty (41.6%) patients developed an ADF secondary to malignancy, with previous radiotherapy (18.7%), post-operative anastomotic dehiscence and endotherapy (14.6% each) representing other causes. Thirty-one patients were managed with tracheal and/or oesophageal stents and eight underwent surgical repair. The classification system demonstrated benign causes of ADF were either proximally or distally sited, whilst a malignant cause resulted in mid-tracheal fistulae (
p
 = 0.001), with the latter associated with poorer survival. ADFs over 20 mm in size were associated with poor survival (
p
 = 0.011), as was the use of previous radiotherapy. Proximal and distal ADFs were associated with improved survival (
p
 = 0.006), as were those patients managed surgically (
p
 = 0.001).
Conclusion
By classifying ADFs, we have demonstrated that anatomical location correlates with the size, history of malignancy, previous radiotherapy and aetiology of ADF, which can affect management. The proposed classification system will aid in formulating multi-modality individualised treatment plans.