Objectives To assess the incidence, severity and risk factors of
bronchial fistula following
pneumonectomy for
cancer. Patients and methods From 1989 to 2003, 690 consecutive patients underwent a
pneumonectomy for thoracic
cancer in Sercive of Thoracic Surgery of the Teaching Hospital of Sainte Marguerite in Marseilles (France). The M/F sex ratio was 5,44 . Mean age was 59+/-9,9 years [16 - 81]. Clinical and surgical variables were studied retrospectively, and their possible association with the occurrence of a
bronchial fistula was assessed by univariate and multivariate analysis. Results Fifty one patients (7,7%) experienced a
bronchial fistula. This complication accounted for 56% (45/80) of the cases of reoperation and 25,5% (13/51) of early deaths. At univariate analysis, the following factors were identified as statistically significant: tobacco consumption (p<0,003), presence of
COPD (p =0,02), preoperative
radiotherapy (p=0,03), previous thoracic surgery (p=0,03), right side of the resection (p<0,001), hand-fashioned bronchial
suture (p=0,05) and squamous cell histology (p= 0,04). Multivariate logistic regression analysis disclosed tobacco consumption (p=0,002), presence of
COPD (p=0,01), previous thoracic surgery (p=0,03), extended procedures (p=0,05), right
pneumonectomy (p<0,001) and squamous cell histology (p=0,02) as independent predictors of
bronchial fistula. Conclusion The occurrence of a
bronchial fistula following
pneumonectomy is a frequent life threatening event, especially in cases of right sided resections and extended procedures. Tobacco cessation,
preoperative rehabilitation, and reinforcement of the bronchial
suture are possible means of prevention.