Endothelial seeding of venous prostheses

To evaluate the potential benefit of endothelial seeding of venous prostheses, 20 dogs were subjected to iliocaval reconstruction with expanded polytetrafluoroethylene grafts protected by an arteriovenous fistula. Grafts seeded with enzymatically derived endothelial cells were compared with control grafts that were sham seeded with culture medium and blood. Five seeded and seven sham-seeded grafts remained patient and were perfusion fixed in situ 4 to 6 weeks after operation. Specimens were examined by light, scanning, and transmission electron microscopy. No statistical difference in early patency was noted. The mean thrombus-free surface area was 80% in the seeded and 71% in the sham-seeded group. Light microscopy of these areas revealed a monocellular layer lining the lumen in all grafts. Scanning electron microscopy demonstrated a thin cellular lining covering 50% to 100% of the specimens' surface area in four of the five seeded and five of seven sham-seeded grafts. Transmission electron microscopy revealed these cells to exhibit characteristics typical of endothelial cells. The subcellular layer was equally thin in both groups. Early patency rates were not benefited by endothelial seeding of grafts placed in the venous system. Seeding of grafts with enzymatically derived endothelial cells provides a good endothelial cover with a thin subendothelial layer but not to a greater extent than does sham seeding of the venous prostheses.