Enhanced effect of inhibition of thrombin on endothelium in murine endotoxaemia: specific inhibition of thrombocytopenia

Co-administration of iloprost and eptifibatide in septic shock (CO-ILEPSS)-a randomised, controlled, double-blind investigator-initiated trial investigating safety and efficacy

BACKGROUND

Part of the pathophysiology in septic shock is a progressive activation of the endothelium and platelets leading to widespread microvascular injury with capillary leakage, microthrombi and consumption coagulopathy. Modulating the inflammatory response of endothelium and thrombocytes might attenuate this vicious cycle and improve outcome.

METHOD

The CO-ILEPSS trial was a randomised, placebo-controlled, double-blind, pilot trial. Patients admitted to the intensive care unit with septic shock were randomised and allocated in a 2:1 ratio to active treatment with dual therapy of iloprost 1 ng/kg/min and eptifibatide 0.5 μg/kg/min for 48 h or placebo. The primary outcomes were changes in biomarkers reflecting endothelial activation and disruption, platelet consumption and fibrinolysis. We compared groups with mixed models, post hoc Wilcoxon signed-rank test and Mann-Whitney U test.

RESULTS

We included 24 patients of which 18 (12 active, 6 placebo) completed the full 7-day trial period and were included in the per-protocol analyses of the primary outcomes. Direct comparison between groups showed no differences in the primary outcomes. Analyses of within-group delta values revealed that biomarkers of endothelial activation and disruption changed differently between groups with increasing levels of thrombomodulin (p = 0.03) and nucleosomes (p = 0.02) in the placebo group and decreasing levels of sE-Selectin (p = 0.007) and sVEGFR1 (p = 0.005) in the active treatment group. Platelet count decreased the first 48 h in the placebo group (p = 0.049) and increased from baseline to day 7 in the active treatment group (p = 0.023). Levels of fibrin monomers declined in the active treatment group within the first 48 h (p = 0.048) and onwards (p = 0.03). Furthermore, there was a significant reduction in SOFA score from 48 h (p = 0.024) and onwards in the active treatment group. Intention-to-treat analyses of all included patients showed no differences in serious adverse events including bleeding, use of blood products or mortality.

CONCLUSION

Our results could indicate benefit from the experimental treatment with reduced endothelial injury, reduced platelet consumption and ensuing reduction in fibrinolytic biomarkers along with improved SOFA score. The results of the CO-ILEPSS trial are exploratory and hypothesis generating and warrant further investigation in a large-scale trial.

TRIAL REGISTRATION

Clinicaltrials.com, NCT02204852 (July 30, 2014); EudraCT no. 2014-002440-41.

Endothelial cell-specific anticoagulation reduces inflammation in a mouse model of acute lung injury

Tissue factor (TF)-dependent coagulation contributes to lung inflammation and the pathogenesis of acute lung injury (ALI). In this study, we explored the roles of targeted endothelial anticoagulation in ALI using two strains of transgenic mice expressing either a membrane-tethered human tissue factor pathway inhibitor (hTFPI) or hirudin fusion protein on CD31⁺ cells, including vascular endothelial cells (ECs). ALI was induced by intratracheal injection of LPS, and after 24 h the expression of TF and protease-activated receptors (PARs) on EC in lungs were assessed, alongside the extent of inflammation and injury. The expression of TF and PARs on the EC in lungs was upregulated after ALI. In the two strains of transgenic mice, expression of either of hTFPI or hirudin by EC was associated with significant reduction of inflammation, as assessed by the extent of leukocyte infiltration or the levels of proinflammatory cytokines, and promoted survival after LPS-induced ALI. The beneficial outcomes were associated with inhibition of the expression of chemokine CCL2 in lung tissues. The protection observed in the CD31-TFPI-transgenic strain was abolished by injection of an anti-hTFPI antibody, but not by prior engraftment of the transgenic strains with WT bone marrow, confirming that the changes observed were a specific transgenic expression of anticoagulants by EC. These results demonstrate that the inflammation in ALI is TF and thrombin dependent, and that expression of anticoagulants by EC significantly inhibits the development of ALI via repression of leukocyte infiltration, most likely via inhibition of chemokine gradients. These data enhance our understanding of the pathology of ALI and suggest a novel therapeutic strategy for treatment.

Characteristics of thrombin generation in a fulminant porcine sepsis model

INTRODUCTION

The activation of blood coagulation has been demonstrated in most cases of sepsis, however previous studies in humans could not detect hypercoagulability with global hemostasis assays. In a fulminant porcine sepsis model we analysed coagulation screening tests and thrombin generation to evaluate hemostatic alterations.

MATERIALS AND METHODS

Live Escherichia coli bacteria were inoculated to female pigs and prothrombin time, activated partial thromboplastin time, thrombin time and fibrinogen were measured by coagulometry. Platelet counts, platelet aggregates and platelet phosphatidyl serine (PS) expression were studied, furthermore in in vitro experiments the PS-inducing ability of septic and control plasmas was investigated by flow cytometry. Thrombin generation was carried out by the Ascent Fluoroscan reader and results were evaluated by the Thrombinoscope software.

RESULTS

Clotting assays showed a large variability, but no systematic changes during the 4-hour observation period. Platelet count significantly decreased and the number of platelet aggregates increased already by 2h compared to baseline values and to control animals. Although the increase in platelet PS expression was non-significant in the septic group, the septic plasma elicited PS expression on normal human red blood cells. Thrombin generation became significantly faster, but the quantity of formed thrombin demonstrated both hypo- and hypercoagulability depending on the setting of the assay.

CONCLUSIONS

Enhanced thrombin generation without activators and the PS-inducing capacity of septic plasma are signs of hemostatic activation during fulminant sepsis while the decreased amount of generated thrombin upon tissue factor and phospholipid induced activation demonstrates attenuated thrombin forming ability.