Comparison of the Antithrombotic Effectiveness of the Direct Thrombin Inhibitor, Dabigatran Etexilate, and Two Factor Xa Inhibitors in an A-V Shunt Model of Thrombosis in Rabbits.

Inhibition of two key serine proteases in the coagulation cascade, thrombin (IIa) and factor Xa, are currently being exploited for direct, oral antithrombotic activity in the clinic. However, it is still unclear if one form of coagulation factor inhibition is more effective than the other. Thus, the objective of this study was to test the antithrombotic efficacy of the clinically advanced compounds, the potent direct thrombin inhibitor, dabigatran etexilate and rivaroxaban, a potent direct factor Xa inhibitor in the rabbit A-V shunt model of thrombosis. In addition, another internally developed factor Xa inhibitor, BI42551, with properties similar to those in clinical development was tested. All three compounds have affinities (Ki) for their respective coagulation factor in the low nM range, i.e. human thrombin with dabigatran or human factor Xa with rivaroxaban or BI42551. In addition, each is at least >700-fold selective for its human coagulation factor, dabigatran etexilate for IIa vs Xa and the factor Xa inhibitors for Xa vs IIa. These compounds are highly selective inhibitors not only of the human enzyme, but also have similar values for rabbit thrombin and Xa, respectively. All experiments were performed according to German animal ethics guidelines. The femoral artery and vein of anesthetised rabbits were connected with polyethylene tubing containing a fixed length of suture, pre-soaked in tissue factor. Blood flow through the shunt was maintained over 40 min, after which the suture with any thrombus was removed from the shunt and weighed. The prodrug dabigatran etexilate and the factor Xa inhibitors were given in doses of 3 and 10 mg/kg orally and the rabbits were anesthetised either 90 min or at the highest dose, also 6.5 hrs after drug administration. There was a dose-dependent reduction of thrombus formation with all three compounds as compared to control. Antithrombotic efficacy at 3 and 10 mg/kg is shown as % inhibition of control measured 2 hrs after drug administration (table, columns 2&3). These effects were long-lasting, as significant antithrombotic activity was also measured 7 hrs post administration (last column). Plasma levels of all compounds were dose-dependent and clotting tests correlated well with dose.

These results show that both thrombin and factor Xa inhibition are effective methods of inhibiting thrombosis in a rabbit AV shunt model. All drugs had potent and long-lasting effects after a single oral administration in this model, though dabigatran showed a trend to elevated antithrombotic efficacy at both 2 and 7 hrs. However, in the clinical setting differences in antithrombotic treatment may also be related to differences in pharmacokinetic profiles, drug interactions or metabolism, or the individual side effect profiles of each compound.