Numerous thrombin/Xa-inhibitors have been used in prophylaxis/treatment of thromboembolic diseases. We were interested to compare argatroban (direct thrombin inhibitor, DTI) with bivalirudin (DTI), lepirudin (DTI) and danaparoid (indirect Xa inhibitor) concerning the effects on fibrin gel permeability (Ks), APTT and INR. We were also interested to assess whether the drug-induced impairment of haemostasis reacts differently to the addition of two coagulants-rVIII or rVIIa. Ks was measured in a normal plasma pool (NPP) mixed with one of the inhibitors in increasing concentrations. Coagulation was initiated by adding recombinant tissue factor (rTF, 5pM), purified phospholipids (PPL, 4mM) and CaCl2. Since the activity of rVIIa is dependent on the presence of activated platelets, rTF and PPL were replaced by washed-frozen platelets when the influence of rVIIa was evaluated. APTT and INR were tested in the same inhibitor-spiked NPP using reagents from Stago, France. We found that all inhibitors investigated made the fibrin network more porous in a concentration-dependent way. However, danaparoid was less potent to change Ks than argatroban or bivalirudin. When a comparison was performed between samples containing argatroban or bivaluridin, the former drug brought increases in Ks which appeared gentler than those by the latter. Regarding the effect by lepirudin, a dose-response curve in Ks assay was very steep. APTT and INR were prolonged by all the inhibitors examined, but only argatroban and bivaluridin led to prolongations which were significantly correlated to Ks. rFVIII overcame the anticoagulant effect on Ks to some degree. Similar influence by rVIIa was seen in samples containing argatroban or bivalirudin but not in those containing lepirudin. The effect of danaparoid on Ks was almost neutralised by addition of the two recombinant coagulants.

Conclusions: In the in vitro study, argatroban, bivalirudin, lepirudin and danaparoid can inhibit thrombin generation and thus depress coagulation, shown as prolonged time-to-start of detectable fibrin formation and increased porosity of fibrin network. The latter effect may favour fibrinolysis partly by facilitating the transportation of fibrinolytic components. The steep dose-response curve for Ks shown in the samples containing lepirudin may implicate a narrow therapeutic window. Further work are needed to clarify 1) whether the dissimilarities between argatroban and other inhibitors concerning the effects on Ks, APTT or INR implicate different therapeutic effects or safety in patients; 2) whether the dose-dependent influence of argatroban/bivalirudin on APTT and INR can be shown in vivo, thereby determining if these two simple methods are sensitive enough for therapeutic monitoring; 3) whether rVIII or rVIIa can be used to treat patients with haemorrhagic complications to the therapies. Additionally, the fibrin gel assay which employs a tiny dose of TF (together with PPL) as the thrombin generation trigger to test the end stage of coagulation, i.e. fibrin network formation, may reflect the effects of multiple protease activation and inhibition in a physiologically relevant way. We consider that in contrast to the conventional methods such as APTT and INR, the fibrin gel assay can give additional and important information on drug effects, and thus contribute with important information in the development of new anticoagulant compounds.

Author notes

Disclosure: No relevant conflicts of interest to declare.