Abstract
IntroductionExtracorporeal membrane oxygenation (ECMO) is a supportive therapy for severe cardiac or respiratory failure associated with high rates of bleeding and thrombosis. Their development is associated with increased rates of mortality. Haemostatic and inflammatory changes are recognised to occur in patients with critical illness and because of the extracorporeal circuit. This thesis aims to assess bleeding and thrombosis during ECMO in more detail and assess specific temporal changes in haemostatic parameters.
Methods and materials
Three components were looked at in this thesis: 1) a single-centre registry review of patients receiving veno-venous ECMO for the development of bleeding and thrombosis and outcomes associated with the use of blood products transfusions, 2) haemostatic changes in an ex vivo circuit of ECMO, and 3) haemostatic changes in patients during veno-venous ECMO, focusing on thrombin generation, fibrinolysis and circulating histones.
Consent was obtained from patients receiving ECMO or nominated consultees with local ethics approval. For the ex vivo circuit, blood samples (whole blood with 3.2% trisodium citrate) were collected over a 24-hour period and compared to time-matched control samples. For patient samples, blood samples were taken at multiple time points prior to, during and after ECMO and correlated with bleeding and thrombotic events. ELISA-based techniques were used for thrombin generation markers, fibrinolytic assays and histone levels.
Results
A cohort of 365 patients with influenza, COVID-19 and bacterial pneumonia requiring ECMO were reviewed. Key findings demonstrated the underlying disease aetiology affected rates of thrombosis, particularly COVID-19 infection. Small volume intracranial haemorrhage (ICH) and small volume pulmonary artery filling defects were seen frequently at initiation of ECMO. Increased mortality was seen in those who had progression of ICH and non-intracranial major haemorrhage as opposed to thrombotic events.
In reviewing blood transfusion-associated outcomes, patients receiving a ‘restrictive’ red cell transfusion approach (haemoglobin target 80-90g/L) had similar survival outcomes to those with a ‘liberal approach (100-120g/L). Other blood product transfusions had high usage with major bleeding and an increased mortality rate was seen with platelet transfusions.
In the ex vivo circuit model, the key changes were an increase in plasminogen activator inhibitor-1 (PAI-1) and decrease in tissue plasminogen activator (tPA) after 24-hours. There was no increase in thrombin generation markers over 24 hours. However, circulating H3 histone levels were significantly increased because of the circuit.
In patients requiring ECMO, there were haemostatic changes in keeping with a reactive profile were seen prior to initiation. An early increase in thrombin generation was seen during ECMO followed by a subsequent increase in fibrinolysis. Both decreased following decannulation from ECMO. Low thrombin-activatable fibrinolysis inhibitor (TAFI) and elevated PAI-1 levels were seen in bleeding events.
Conclusions
The development of thrombosis and bleeding have temporal changes in patients during ECMO and this is reflected by changes in thrombin generation and fibrinolytic activation markers. Additionally, retrospective registry data suggests that more ‘liberal’ blood transfusion approaches may not necessarily improve survival outcomes and transfusion triggers during ECMO need to be investigated further.
| Date of Award | 1 Aug 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Beverley Hunt (Supervisor) & Karen Breen (Supervisor) |