Improved airway healing after lung transplantation. An analysis of 348 bronchial anastomoses

We evaluated various clinical factors to identify predictors of airway complication after lung transplantation. Two hundred twenty-nine consecutive single (n = 110) and bilateral (n = 119) lung transplants were done between September 1988 and August 1994. These 348 bronchial anastomoses were retrospectively analyzed. Airway complication that necessitated clinical intervention affected 33 anastomoses (9.5%) in 29 patients (12.8%). Satisfactory healing was achieved in 22 of these patients by conservative therapy such as one or a combination of dilation, stent, and laser. There were five deaths (2.2%) attributable to airway complications. One patient had an early postoperative death unrelated to airway complication and one patient has a recalcitrant bronchus intermedius stricture. Complication occurred more often in single-lung than in bilateral lung transplants (16/110, 14.4%, versus 17/238, 7.1%; p < 0.05). The use of a mattress suture (21/153, 13.7%) was associated with more frequent complications than was simple interrupted suture (8/122, 6.6%) or figure-of-eight suture (4/73, 5.5%) (p < 0.05). For patients in whom airway complications subsequently developed, the duration of postoperative mechanical ventilation was greater than that for those in whom an airway complication did not develop. The prevalence of airway complications as our program evolved was evaluated by separating the 229 transplants into three groups: phase I, the first 77 transplants; phase II, the next 76 transplants; and phase III, the most recent 76 transplants. The airway complication rate per anastomosis was significantly lower in phase III (5/126, 4.0%) than in phase I (12/110, 10.9%; p < 0.05) and phase II (16/112, 14.3%; p < 0.01). The majority of airway complications are successfully treated and rarely fatal. The recent reduction in prevalence of airway complications is likely a result of better maintenance immunosuppression and rejection surveillance.