Low molecular weight heparins prevent thrombin-induced thrombo-embolism in mice despite low anti-thrombin activity. Evidence that the inhibition of feed-back activation of thrombin generation confers safety advantages over direct thrombin inhibition.

Thrombin-induced thromboembolism in mice is a model in which the feed-back clotting activation produced by the injected enzyme greatly contributes to fibrin accumulation in lungs and to mortality. Using this model we have previously shown that activated human protein C (aPC), by interrupting endogenous clotting activation at a high level (factors Va and VIIIa), prevents mortality inducing only a minor hemostatic impairment. With the same model we have now compared the antithrombotic and prohemorrhagic effects of two low molecular weight heparins (LMWHs), reviparin and tinzaparin, which are expected to inhibit preferentially the positive feed-back triggered by thrombin (anti Xa activity), with those of unfractionated heparin (UFH) and PEG-hirudin, which inhibit mainly or exclusively thrombin activity (anti IIa activity).Pulmonary thromboembolism was induced in mice by i.v. injection of bovine thrombin (1,000U/kg). Drugs (from 0.12 to 1.2 mg/kg) were given as bolus injection 2 min prior to thrombin challenge and mortality was assessed within 15 min. The bleeding time was assessed by a tail tip transection model. Activated partial thromboplastin time (aPTT), thrombin clotting time (TcT), fibrinogen assay and anti Xa activity determination were performed in citrated plasma from saline- or drug-treated animals.All drugs protected mice from thrombin-induced mortality in a dose-dependent way. At comparable antithrombotic dosages, the anti IIa activity generated in plasma (assessed by TcT) was highest with UFH, intermediate with tinzaparin and very low with reviparin. Accordingly, the fibrinogen drop, which is caused mainly by the injected thrombin, was prevented by the heparins to an extent that was fairly well related to their anti IIa activity. aPTT and bleeding time, used as measures of hemorrhagic risk, were markedly more prolonged by UFH than by reviparin. Tinzaparin, instead, had an intermediate effect. Interestingly, PEG-hirudin, at equipotent antithrombotic dosages, caused a prolongation of bleeding time comparable to that observed with UFH.Our data show that, in our model, drugs acting at a high level of the blood clotting cascade, like LMWHs with a high anti Xa/anti IIa ratio, display a better antithrombotic/prohemorrhagic profile than drugs acting prevalently on thrombin.