Epithelial grafting of a decellularized whole-tracheal segment: an in vivo experimental model.

OBJECTIVES: Prerequisites for successful trachea transplantation include the use of a biocompatible construct, submucosal vascularization and an epithelial covering. Implantation of non-epithelialized tracheal scaffolds may lead to stenosis. However, epithelial grafting or seeding can only be attempted onto a well-vascularized submucosal bed. Our aim was to investigate a method to prevent stenosis during prelamination of non-epithelialized, gently decellularized rabbit tracheae and to evaluate whether grafting of revascularized constructs with buccal mucosa is feasible.

METHODS: Allotracheae underwent two 48-h cycles of detergent-enzymatic decellularization using sodium deoxycholate and DNAse. In the first series, 12 circular scaffolds were implanted bilaterally in lateral thoracic artery flaps (n = 6 rabbits). Right-sided transplants were covered internally with Integra™. In the second series, 10 decellularized tracheae covered with Integra were prelaminated in flaps (n = 10 rabbits). Twenty-one days after implantation, revascularized tracheae were grafted with buccal mucosa. A macroscopic, histological analysis and immunohistochemistry were performed on explants.

RESULTS: In the first series, tracheae without Integra covering developed significantly greater intraluminal (P = 0.032) and subepithelial narrowing (P = 0.0345) compared with tracheae with Integra covering. All tracheae exhibited insufficient submucosal revascularization. In the second series, submucosal revascularization was incomplete in the first 2 constructs, which were implanted circularly. These tracheae only showed marginal buccal graft ingrowth. To accelerate revascularization, the subsequent 8 transplants were opened longitudinally before implantation. Compared to circularly implanted tracheae, submucosal revascularization of these transplants was superior (P = 0.0008). Graft adherence was complete in 6 opened constructs. Mild lymphocytic infiltration within the buccal graft was detected in 5 specimens.

CONCLUSIONS: We observed satisfactory host integration of opened tracheae that were temporarily covered with Integra during revascularization and subsequently grafted with buccal mucosa. Integra successfully prevented stenosis during revascularization. This model may provide an example of an immunosuppressive-free approach in the treatment of long-segment tracheal lesions. With the aid of further refinements such as a respiratory epithelial lining, an orthotopically transplantable construct could be created.