Enhancement of tissue factor expression by vein segments exposed to coronary arterial hemodynamics.

PURPOSE: Although saphenous vein is the most reliable conduit for arterial interposition procedures in the coronary circulation, graft thrombosis remains a clinical problem. We hypothesized that an important factor in early graft thrombosis is sudden change in the hemodynamic environment of the vein as it is placed in the coronary circulation. METHODS: We used an ex vivo perfusion system to study freshly excised segments of human saphenous vein (HSV) and pig internal jugular vein. For coronary graft (CAVG) simulation, sections of HSV were subjected to arterial pulsatile pressure and flow and twisting and stretching to mimic deformations caused by the beating heart. Using functional and immunohistochemical assays, we investigated the effect of these conditions on expression of tissue factor (TF), an important prothrombotic surface molecule. RESULTS: In each of 11 experiments (6 human, 5 porcine), vein segments from a single donor were subjected to venous conditions (VEN), CAVG perfusion, or no perfusion. Expression of TF was measured as the amount of factor Xa generated per unit area of luminal vein surface. VEN perfusion did not cause a significant change in mean TF expression over nonperfused control values (human: 14.3 +/- 1.5 versus 11.4 +/- 2.3 U/cm2, p = 0.31; pig: 11.6 +/- 1.5 versus 12.5 +/- 1.4 U/cm2, p = 0.70). CAVG perfusion led to significant enhancement of TF expression over VEN perfusion (human: 36.8 +/- 6.2 versus 14.3 +/- 1.5 U/cm2, p < 0.05; pig: 40.0 +/- 9.9 versus 11.6 +/- 1.5 U/cm2, p < 0.05). Immunohistochemical analysis showed positive TF staining on the luminal side of a CAVG-stimulated HSV segment, but not on a VEN-stimulated segment. In four additional studies, HSV segments were subjected to arterial perfusion without twist and stretch to mimic lower extremity arterial interposition grafts. TF expression for lower extremity venous graft perfusion was significantly higher than for VEN perfusion (25.3 +/- 2.5 versus 14.3 +/- 1.5, p < 0.01) but not significantly different from CAVG perfusion. CONCLUSIONS: Our studies in a unique perfusion system suggest that exposure of vein to coronary arterial hemodynamic conditions results in elevated expression of the important prothrombotic molecule TF. This phenomenon may contribute to early graft thrombosis.

Effect of retroviral transduction on human endothelial cell phenotype and adhesion to Dacron vascular grafts.

PURPOSE: Retroviral transduction for genetic enhancement of endothelial cell (EC) anti-thrombotic phenotype offers potential for improving the clinical success of vascular graft seeding; however, application of this technique may bring concomitant alteration in cell functionality. METHODS: Human microvascular ECs were transduced with a retroviral vector encoding for the marker gene beta-galactosidase. Transduced endothelial cells (rtECs) and nontransduced endothelial cells (ntECs) were evaluated by flow cytometry for expression of intercellular adhesion molecule (ICAM)-1 and tissue factor (TF) on both smooth (coverslips) and graft (Dacron, 6 mm inside diameter) surfaces under static and shear exposed conditions. Graft EC retention was measured after 6-hour pulsatile perfusions. Platelet and neutrophil adherence was measured on perfused coverslips. RESULTS: Lower levels of ICAM-1 were expressed by rtECs on coverslips under both static (p < 0.01 vs static ntECs) and shear exposed conditions (p < 0.01 vs static and shear ntECs). Accordingly, fewer polymorphonuclear leukocytes adhered to rtEC monolayers (p < 0.01 vs ntECs). No difference in ICAM-1 and TF expression by static graft seeded rtECs and ntECs was observed. However, graft-seeded rtECs that were exposed to wall shear stress displayed less TF than sheared ntECs (p < 0.05). Transduction did not affect EC retention to the sheared graft surface. CONCLUSIONS: These data suggest that retroviral transduction does not elicit a prothrombotic/proinflammatory phenotype, rather indices of these states appear in some conditions to be reduced. Further, transduction does not adversely affect EC adherence to Dacron graft surfaces under arterial hemodynamics.

Platelet deposition on ePTFE grafts coated with fibrin glue with or without FGF-1 and heparin.

INTRODUCTION: The disappointing long-term patency of small-caliber prosthetic grafts may be due in part to early thrombogenicity of the prosthetic surface. We previously reported that the coating of expanded polytetrafluoroethylene (ePTFE) with fibrin glue (FG) containing fibroblast growth factor type 1 (FGF-1) and heparin accelerated spontaneous endothelial coverage of ePTFE grafts in an animal model; however, FG's effect on platelets remains unclear. This study was done to evaluate platelet deposition onto GF/FGF-1/ heparin-coated vs FG-coated vs whole-blood-preclotted ePTFE surfaces. METHODS: Twelve 5-cm ePTFE grafts were treated either with FG (thrombin, 0.32 U/ ml, and fibrinogen, 32.1 mg/ml, n = 8) or with FG containing FGF-1 (11 ng/ml) plus heparin (250U/ml, n = 4). Twelve control ePTFE grafts were preclotted with canine (n = 8) or human (n = 4) whole blood. These treated grafts were placed onto a loop pulsatile perfusion system in pairs (preclotted with either FG or FG/ FGF-1/heparin) and perfused with a M-199/10% FBS/ 111indium-labeled platelet suspension. After 60 min the grafts were gamma counted and CPM/mm2 were determined. RESULTS: In both trials, the preclotted ePTFE grafts demonstrated similarly increased platelet deposition when compared to grafts treated with FG/FGF-1/heparin or FG alone (P < 0.001 for each). CONCLUSION: The decrease in platelet deposition on the FG/FGF-1/ heparin-coated grafts vs preclotted grafts is not due to heparin and is not specific to canine or human platelets. FG-coated grafts may induce a decrease in early graft thrombogenicity when compared to whole blood preclotting.

ePTFE coating with fibrin glue, FGF-1, and heparin: effect on retention of seeded endothelial cells.

In an attempt to improve the resistance of seeded endothelial cell (EC) to desquamation due to shear stress, we evaluated the effect of coating expanded polytetrafluoroethylene (ePTFE) grafts with fibrin glue (FG) containing fibroblast growth factor 1 (FGF1) and heparin on the retention of EC exposed to pulsatile flow ex vivo. Five pairs of ePTFE grafts (30 microm internodal distance, 4 mm internal diameter, 7 cm long) were coated with either FG/FGF-1/heparin (fibrinogen 32.1 mg/ml, thrombin 0.32 U/ml, FGF-1 11 ng/ml, heparin 250 U/ml) or fibronectin (FN) (20 microgram/ml). Canine jugular vein endothelial cells (Factor VIII, passages 5-7), were radiolabeled with indium-111 (100 microCi/1 million cells). Cell seeding (3 x 10(5) cells/cm2) was achieved by four successive inoculations of cells separated by 90 degree graft rotations. After overnight incubation (37 degrees C), pairs of FG and FN grafts (5 cm long) were simultaneously perfused ex vivo with culture media containing 10% fetal bovine serum (120/80 mm Hg, 90 cc/min, 60 pulsations/min). During the 1-hr perfusion, perfusate samples were taken at 0, 5, 15, 30, and 60 min to determine radioactivity loss. Pre- and postperfusion whole graft radioactivity data were compared to estimate cell retention and confirmed by histologic evaluation. Mean adherent radioactivity on FG-coated grafts (96 +/- 5%) was significantly higher (P = 0.0029, Student's t test) than on FN-coated grafts (85 +/- 3%). Maximum radioactivity loss in perfusate was seen after 5 min, with lower sustained loss thereafter. The improved retention of seeded EC on ePTFE grafts coated with FG containing FGF-1 and heparin compared to FN will need to be confirmed for longer durations of perfusion and using in vivo models.

A device for the application of cyclic twist and extension on perfused vascular segments.

Cyclic strain is known to influence many aspects of vascular cell biology, including macromolecular biosynthesis, cell proliferation, and cell morphology. The simulation of the cyclic strain associated with the vasculature in situ has been carried out mostly using devices that manipulated flat membranes on which the cells of choice are cultured. The purpose of this work was to create an apparatus wherein cyclic strains consistent with those found in native blood vessels could be applied to intact three-dimensional (tubular) specimens perfused ex vivo. The cyclic strain protocols using our apparatus may be any combination of axial stretch and twist. Concurrently, the perfusion flow and shear stress are controlled as desired. Vessel diameter, intraluminal pressure, volume flow, and regional strain are measured and stored on line. To illustrate the application of the device, we present hemodynamic and kinematic data collected from a human saphenous vein segment perfused under steady-flow conditions while subjected to cyclic stretch and twist.