Predicting excess bleeding due to haemostatic failure following cardiac surgery requiring cardiopulmonary bypass

• Main Author: Percy, Charles L.
• Published: Cardiff University 2015
• Subjects: 617.4; R Medicine (General)
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.665881

Bleeding following cardiac surgery requiring cardiopulmonary bypass (CPB) is associated with increased morbidity. Identification of patients at increased risk of bleeding might allow intervention to prevent bleeding developing. In this thesis, clotting factors, anticoagulants and calibrated automated thrombin generation were investigated as potential methods for identifying such patients. Post-CPB FXIII, fibrinogen and platelet count were significantly lower in those who bleed more than 2 mL/kg/hr for two consecutive hours and in those who bleed in excess of 1 litre at 24 hours. ROC analysis demonstrated these had modest predictive value. Calibrated automated thrombography was unable to identify patients at risk of bleeding. Calibrated automated thrombography was also used to investigate the effects of haemostatic treatment (FFP, rFVIIa, PCC and TFPI inhibition) on thrombin generation in vitro. Blocking the effect of TFPI produced the greatest improvement in thrombin generation. The effect of CPB on platelet phospholipids was investigated using mass spectrometry. Post-CPB the ability to externalise phosphatidylethanolamine and phosphatidylserine was impaired. The ability to externalise and synthesise 12-HETE-PC and 12-HETE-PE in response to both thrombin and collagen post-CPB was also reduced. The effect of these phospholipids on thrombin generation and the ability to identify patients at risk of bleeding was then investigated. Thrombin generation using liposomes containing 12-HETE-PC or 12-HETE was lower in patients who required haemostatic treatment for post-CPB bleeding compared to those who did not. This suggests there are variations between individuals in the way their coagulation factors interact with oxidised phospholipids and that this may influence bleeding. Finally a cell based model of thrombin generation was developed using monocytes as a source of tissue factor and incorporating the observed changes in phospholipids, clotting factors and anticoagulants. This model provides a basis to further investigate the influence of different TF expressing cells on thrombin generation which may affect bleeding.