Evidence of ECMO induced changes to haemostasis in an ovine model

Abstract

Patients requiring ECMO have acquired haemostatic disorders due to a vast interplay of factors including drugs, disease and device. In this heterogenous population it is impossible to define the isolated and combined effects of critical illness and ECMO on haemostasis. This study aimed to characterise changes in haemostasis attributable to ECMO in an in vivo ovine model using ROTEM® and Multiplate® analysis. In our validated model, 10 healthy sheep and 10 sheep with smoke-induced acute lung injury (S-ALI) were supported on venovenousECMO. Heparin infusioncommencedat4 U/kg/h to achieve activated clotting time of 200–300 s. Samples were collected at baseline, pre-ECMO and 0.25, 1, 1.5 and 2 h post-ECMO initiation. Analysis included platelet counts, EXTEM and FIBTEM on ROTEM®, and Multiplate® analysis with ADP and collagen agonists. For both groups, platelet count remained within normal range. ECMO induced no significant changes to ADP or collagen-induced platelet aggregation or EXTEM clotting time. EXTEM clot formation time (CFT) was increased (p less than 0.01) and EXTEM maximum clot firmness (MCF) (p less than 0.001) was decreased compared to baseline for both groups throughout ECMO. A smaller decrease in FIBTEM-MCF was documented during ECMO in healthy sheep (p less than 0.05) compared to sheep with S-ALI prior (p less than 0.01). ECMOinduces an increase in clot formation time and decrease in fibrinogen function and overall clot quality in healthy sheep. S-ALIprior to ECMO exacerbates these haemostatic alterations. Platelet number and function was maintained during ECMO, which, when assessed with the decrease in EXTEM-MCF and FIBTEM-MCF, suggests fibrinogen may be central to these haemostatic alterations.