Purpose: The purpose of this study was to develop a cost-effective canine graft healing model that gives information on various implant sites and controls for variable factors between graft locations and between animals and to compare the influence of implant site (retropleural, retroperitoneal, and subcutaneous areas) on arterial graft healing in the same subject under such controlled study conditions.
Methods: Five mongrel dogs were studied for 8 weeks, and one was studied for 3 years. Each received three porous Dacron grafts during the same surgery: a carotid-femoral bypass (C-FB) and interposition grafts in the descending thoracic aorta and abdominal aorta. To produce comparable shear stress calibers of the C-FB and abdominal aorta grafts were 2 mm less than those of the descending thoracic aorta, and a distal arterio-venous fistula was created to further increase the C-FB flow. For comparable blood aggregation status platelet aggregation was preevaluated and adjusted with antiplatelet agents. Graft flow surfaces were assessed for thrombus-free surface and endothelial-like cell coverage scores. Tissue samples were studied with hematoxylin-eosin, factor VIII/ von Willebrand factor, smooth muscle alpha-actin staining, and scanning electron microscopy and transmission electron microscopy.
Results: All grafts were patent. Shear stress for the three grafts and platelet aggregation among the study subjects were comparable. Healing of descending thoracic aorta and abdominal aorta grafts was similar, but C-FB healing was slow, incomplete, and uneven, with a high incidence of seroma. Eight-week and 3-year results were comparable.
Conclusions: This model gives broad healing information about the areas where grafts are often implanted in humans. Eight weeks appears to be a sufficient period to reflect basic and general healing characteristics. Grafts heal better in the retropleural and retroperitoneal areas than in the subcutaneous tissues.