Optimal graft diameter: effect of wall shear stress on vascular healing.

Arterial walls tend to adapt to maintain a specific wall shear stress. The formation of neointimal hyperplasia and endothelial cell healing of polytetrafluoroethylene grafts may also be governed by wall shear stress, which suggests that an optimal graft diameter may exist. To test this, 40 polytetrafluoroethylene grafts with internal diameters of 3, 6, and 8 mm were inserted end to end in the femoral and carotid arteries of 10 mongrel dogs. Total flow and diameter were measured, and grafts were stained with Evans blue dye, fixed by pressure perfusion, and analyzed by computer for anastomotic neointimal thickening, graft pseudointimal thickening, and degree of endothelial coverage. Mean calculated shear stress was 41 dyne/cm2 for the 3 mm grafts, 7 dyne/cm2 for the 6 mm grafts, and 3 dyne/cm2 for the 8 mm grafts. Fifteen weeks later the patency rate was 0 of 10 for the 3 mm grafts, 16 of 20 for the 6 mm grafts, and 7 of 10 for the 8 mm grafts. The mean graft shear stress was calculated to be 10 dyne/cm2 for the 6 mm grafts and 4 dyne/cm2 for the 8 mm grafts. Pseudointima lining the graft was composed of disorganized protein and cell remnants. The rough surface contained no overlying endothelium. Anastomotic neointima contained a layer of well-organized smooth muscle cells covered by a single layer of polygonal-shaped endothelial cells. A transition zone of thrombus, which is sandwiched by a wedge of smooth muscle cells near the graft surface and covered by endothelial cells, is described. Mean thickness of pseudointima of the patent 8 mm grafts was 150 microns thicker than that of the 6 mm grafts. Anastomotic neointimal thickness was 110 microns thicker in the 8 mm grafts compared with the 6 mm grafts. Among the 6 mm grafts, the carotid grafts had an average initial shear stress of 10 dyne/cm2, whereas the femoral grafts averaged a lower 5 dyne/cm2 and yielded pseudointima and neointima that were 40 microns thicker. The percent graft surface area covered with neointima did not differ among the grafts of differing diameter either proximally or distally. Lower shear stresses produced greater amounts of pseudointimal thickening within polytetrafluoroethylene grafts and neointimal thickening at their anastomoses. Conversely, the high shear stress from small-diameter grafts was associated with poor graft patency. These results suggest that an optimal graft diameter may help to prevent neointimal hyperplasia and graft thrombosis.