Human saphenous veins were cryopreserved in 4% human albumin and 10% dimethyl sulfoxide. The effect of cryopreservation on endothelial cells was studied in terms of the anticoagulant activity of thrombomodulin and in terms of cell proliferation. After storage for 2 weeks at -150 degrees C, 0.45 +/- 0.07 x 10(5) endothelial cells/cm2 were detected in cryopreserved veins and 1.03 +/- 0.04 x 10(5) endothelial cells/cm2 in fresh veins (p < 0.01). The thrombin-catalyzed activation of protein C decreased after cryopreservation, indicating altered thrombomodulin activity in the endothelial cells. On a cell number basis, the release of soluble thrombomodulin was three times higher from the cryopreserved endothelium than from the fresh endothelium (p< 0.05). The amount of spontaneous release of von Willebrand factor from the endothelial surface was not significantly different between fresh and cryopreserved veins. Endothelial cells were cultured from fresh veins and from their cryopreserved counterparts. On plating of endothelial cells in primary culture, the number of adhered cells was 0.9 +/- 0.09 x 10(3) cells/cm2 from fresh veins and 0.25 +/- 0.03 x 10(3) cells/cm2 from cryopreserved veins (p < 0.01). The positive immunohistochemical stain for von Willebrand factor indicated that the endothelial cell character was maintained after cryopreservation. The endothelial desquamation with loss of anticoagulant function and the slow proliferation of surviving cells in vitro suggest an impaired endothelial healing in vivo. The loss of anticoagulant activity complicates the problems of the exposure of thrombogenic subendothelial matrix to blood in implanted cryopreserved veins.

The prevention of circumferential distension could reduce structural damage in arteriovenous grafts. We studied the effect of an external biodegradable supporting conduit on the endothelium and extracellular matrix in vein graft in a pig model. Cephalic vein control grafts (Group I) and jugular veins wrapped in a vicryl mesh tube (I.D. 4 mm) (Group II) were implanted into autologous carotid arteries (n = 14). The grafts were explanted after 1 and 24 hours and at 1 and 3 weeks and evaluated by ELISA for endothelial DNA synthesis and by immunohistoenzymic assays for cells and extracellular matrix. In group I an initial loss of endothelial and smooth muscle cells along with elastin breakdown was followed by an impaired endothelial regeneration and significant graft wall thickening. The elastic tissue was replaced by collagen type I and chondroitin sulfate accumulations, which included a disarray of alpha-smooth muscle actin positive cells. The endothelium was preserved in group II. After 3 weeks the circumferential elastin layers were densified, distended and separated from the endothelium by a neointimal growth of irregular thickness. Biodegradable perivenous conduit minimized endothelial injury and allowed the partial preservation of elastin fibers and smooth muscle cells in the arteriovenous graft. It did not however, prevent myofibroblastic cell proliferation and triggered a macrophagic reaction.

To elucidate the role of the molecular structure of adhesive proteins in an endothelialization of synthetic vascular prosthesis in vitro, a recombinant fibronectin-like engineered adhesion factor (FP) constructed from the specific Arg-Gly-Asp cell adhesion repeats was assayed as adhesive substrate to culture human saphenous vein endothelial cells on ePTFE. ePTFE samples (1 cm2) inserted into cell culture chambers were coated by incubation with increasing amounts of FP (up to 100 micrograms/cm2) prior to cell seeding. At 24 hours after low density cell seeding and 20 micrograms/ml/cm2 FP concentration, the number of adhered cells reached a plateau and the adhered cells did not proliferate up to 6 days of culture. At 24 hours after high density seeding (10(5) cells/cm2), the number of adhered cells was significantly higher on ePTFE with preadsorbed FP than on fibronectin coated PTFE. About 55% of the initially adhered cells survived up to 7 days on FP, whereas cell debris and free nuclei were predominant on fibronectin coated PTFE. In the investigated model the engineered RGD polymer potentialized a short-term adhesion of vascular endothelial cells to PTFE, nevertheless it did not ensure proliferation and long-term survival of these cells.