The surgically created vascular anastomosis is a thrombogenic zone of uncertain etiology. This study was designed to investigate the importance of tissue factor as a cause of human microvascular thrombogenicity. The ability of tissue factor pathway inhibitor (TFPI) to block the effect of tissue factor was also tested in this whole-vessel model system. Tissue factor activity in the presence of absence of TFPI was assayed on the luminal surface of dissected human placental arteries, on the advential surface, and also at the site of a microvascular anastomosis. Vessel wall thrombin activity was measured in the presence and absence of TFPI. Platelet deposition onto a vessel surface using a perfusion system was measured with and without TFPI. Tissue factor activity was greater on the adventitia (4.6 +/- 2.8 x 10(-4) units factor Xa generated/min) than on the endothelium (1.8 +/- 1.6 x 10(-4), P < 0.03) or at a surgically created anastomosis (2.1 +/- 1.2 x 10(-4), P < 0.04). TFPI reduced Xa generation to undetectable levels in 21 of 23 endothelial, adventitial, and anastomotic segments (P < 0.002). TFPI significantly reduced vessel wall thrombin activity in comparison to control anastomoses (control, 3.2 +/- 1.7 ng fibrinopeptide A (FPA)/(ml x min); TFPI, 1.4 +/- 1.2 ng FPA/(ml x min); P < 0.0001). TFPI reduced the platelet deposition on vessel segments with intact endothelium (no TFPI, 0.88 +/- 0.69 x 10(6) platelets/cm2; TFPI, 0.49 +/- 0.29 x 10(6) platelets/cm2; P < 0.06) and on vessel segments with anastomoses (no TFPI, 1.3 +/- 0.70 x 10(6) platelets/cm2; TFPI, 0.76 +/- 0.35 x 10(6) platelets/cm2; P < 0.02). This study demonstrates the importance of tissue factor as a thrombogenic element in a human whole-vessel model system. TFPI is effective in reducing this thrombogenicity.