Cell activation and thrombin generation in heparin bonded cardiopulmonary bypass circuits using a novel in vitro model.

OBJECTIVE It is generally agreed that when the blood contact surfaces of a cardiopulmonary bypass circuit are treated with a layer of heparin molecules the activation of the humoral pathways is attenuated. However, there is still debate as to whether heparin-bonded circuits reduce thrombin generation. This study aims to examine the effects of immobilized heparin on cell activation and thrombin generation in a novel, well controlled model of cardiopulmonary bypass. METHODS The model used consisted of a heparin-bonded and a non-bonded cardiopulmonary bypass circuit perfused in tandem with the same unit of fresh heparinized (3.3 U/ml) human blood for a period of 6 h. Samples were taken for analysis from the bag just prior to perfusion and at 30, 60, 120 and 360 min of perfusion. Whole blood was used to analyse platelet and white blood cell count, haematocrit and activated coagulation time. Plasma samples were prepared for batch analysis of the cell activation markers p-selectin, elastase and interleukin-8, and the thrombin generation markers thrombin-antithrombin and prothrombin fragment F1 + 2. A sample of tubing was taken from each circuit at the end of the perfusion and prepared for visualization by scanning electron microscopy. RESULTS Platelet counts were significantly reduced in the non-bonded circuits compared with the heparin-bonded circuits at 30 (22 versus 200 x 10(9)/L P < 0.01), 60 (26 versus 193 x 10(9)/L P < 0.01) and 120 min (28 versus 193 x 10(9)/L P < 0.01) as were white blood cell counts at 30(1.5 versus 2.7 x 10(9)/L P < 0.01), 60 (0.9 versus 2.4 x 10(9)/L P < 0.01), 120 (0.9 versus 1.8 x 10(9)/L P < 0.01) and 360 min (0.4 versus 0.9 x 10(9)/L P < 0.05). The concentration of p-selectin was found to be significantly higher in the non-bonded circuits than in the heparin-bonded circuits at 30 (37 versus 29 ng/ml P < 0.01), 60 (37 versus 28 ng/ml P < 0.01). 120 (42 versus 27 ng/ml P < 0.01) and at 360 min (72 versus 46 ng/ml P < 0.01). Elastase was elevated in the non-bonded circuits at 30 (570 versus 145 micrograms/l P < 0.01), 60 (646 versus 278 micrograms/l P < 0.01) and 120 min (613 versus 403 micrograms/l P < 0.05) and interleukin-8 at 120 (705 versus 520 pg/ml P < 0.05) and 360 min (11326 versus 9910 pg/ml P < 0.05). A similar picture was found for the thrombin generation markers. Thrombin-antithrombin complexes were raised in the non-bonded circuits compared with heparin-bonded circuits at 60 (24 versus 13 micrograms/l P < 0.05) and 120 min (46 versus 17 micrograms/l P < 0.05) as was prothrombin fragment F1 + 2 at 30 (1.1 versus 0.7 nmol/l P < 0.01), 60 (1.3 versus 0.7 nmol/l P < 0.01), 120 (1.8 versus 0.9 nmol/l P < 0.01) and 360 min (15.0 versus 13.6 nmol/l P < 0.05). Scanning electron microscopy revealed a greater amount of adherent material on the non-bonded surface relative to the heparin-bonded surface. CONCLUSIONS In a cardiopulmonary bypass circuit perfused with human blood the activation of platelets and white blood cells has been seen to be significantly reduced in the presence of a heparin-bonded surface. Thrombin generation due to contact activation of the intrinsic coagulation pathway is also reduced.