Tissue Factor Mediated Activation of Prothrombin Complex Concentrates (PCCs) Is Differently Inhibited by Dabigatran, Rivaroxaban and Apixaban. Potential Clinical Implications

Dabigatran, rivaroxaban, and apixaban were synthesized and/or commercially obtained.

Profilnine (Grifols Biologicals Inc.) was also commercially obtained. One commercial lot of a recombinant thrombin preparation Recothrom® was obtained from ZymoGenetics Inc for the development of polyclonal antibodies. To generate specific antisera, individual groups of rabbits (n = 3–6) were challenged repeatedly with human recombinant thrombin, over a 9-month period. At the end of this time the antisera from each rabbit was collected and pooled. Immunglobulin (IgGs) were isolated using a protein G column (HiTrap Protein G HP – GE Helathcare Bio-Science Crop). Buffered profilnine (2.5 u/ml) was activated with routinely used tissue factor reagents by adding commercially available PT reagents such as thromboplastin C, neoplastinPlus, and simplastin at a 1:4 ratio and incubated for 30 minutes. The activation of profilnine was measured by using thrombin generation utilizing a fluorogenic substrate method (Technoclone) and the protease generation profile was evaluated using mass spectrometry method (SELDI), SDS-PAGE analysis and immunoblotting using a specific antithrombin (Recothrombin) antibody to profile the activation products. Similar studies were carried out in profilnine supplemented with graded amount of various oral anticoagulants in the concentration range of 0–2.5ug/ml.

All tissue factors produce varying degrees of time dependent activation of profilnine as measured by consumption of prothrombin peak at 71 KDa and generation of thrombin peaks at 3l–37 KDa as observed in the SELDI. Varying amounts of prothrombin generation at 52 KDa was also evident. Distinct immunoblot for thrombin in western blotting analysis was consistent with SDS-PAGE and SELDI analysis showing the generation of thrombin. The anti-Xa agents blocked the generation of thrombin whereas dabigatran failed to produce this effect. This phenomenon was also observed in all three methods used to study generation of the thrombin when using other PCCs such as octaplex and thromboplex activated by various tissue factors. In the fluorometric thrombin generation assays both apixaban and rivaroxaban produced a relatively stronger inhibition of thrombin generation (IC₅₀= 20–200ng/ml) wheras > 500ng/ml for dabigatran in various PCCs.

These results suggest that in contrast to dabigatran both rivaroxaban and apixaban produce a much stronger inhibition of tissue factor mediated generation of the thrombin in PCCs. Inhibition of the functional generation of thrombin was weaker with dabigatran in contrast to apixiban and rivaroxiban. The observed ex-vivo neutralization profile of these agents by PCCs may be due to the differential interactions with the protease generated during their activation. These differences along with the compositional variations in the PCCs should be taken into account while considering prothrombin complexes for the neutralization of new oral anticoagulants.

No relevant conflicts of interest to declare.