Reversal Of Dabigatran Anticoagulation By a 4-Factor Prothrombin Complex Concentrate: Correlation Between Effects On Parameters Of Thrombin Generation and Hemostatic Effect In Vivo

Dabigatran etexilate (Pradaxa®) is a prodrug that is metabolized to a small molecule direct thrombin inhibitor. It has the drawback that it is not readily reversible. Several studies have suggested that prothrombin complex concentrates (PCCs) might reverse the effects of dabigatran on laboratory assays. However, the effects of PCCs on laboratory assays do not seem to correlate well with their effects on hemostatic function in several animal models. We have previously reported the effects of a 4-factor PCC (KCentra®, CSL-Behring) on parameters of thrombin generation in a cell-based assay of thrombin generation (cell-based Calibrated Automated Thrombogram Assay, CAT®). This model allowed us to use varying levels of dabigatran and varying levels of PCC in the presence of freshly isolated platelets, a cell-associated source of tissue factor (TF) and plasma levels of coagulation factors. We found that dabigatran lengthens the lag, and reduces the total and peak levels of active thrombin in this model. KCentra can normalize the peak, rate and total amount of thrombin generated in the presence of dabigatran, but does not normalize the lag time. These effects were seen at all levels of dabigatran tested. A level of 1 U/mL KCentra® (equivalent to an dose of 50 U/kg in vivo) was able to restore the rate and peak levels of thrombin generation, as well as the area-under-the-curve, to normal levels. A level of 0.5 U/mL KCentra did not completely normalize the rate, peak and AUC.

We compared results in the cell-based CAT assay to results in a novel mouse saphenous vein bleeding model (Pastoft, et al, A sensitive venous bleeding model in haemophilia A mice: effects of two recombinant FVIII products (N8 and Advate) 2012). We found that an i.v. dose of 15 ug/kg dabigatran doubled the time to hemostasis from 80 sec + 9 in controls to 173 sec + 24. A dose of 25 U/kg KCentra partially corrected hemostasis and a dose of 50 U/kg completely corrected hemostasis in this model. These results are completely consistent with the results of the cell-based CAT assay.

While many workers report the peak or the area under the thrombin generation curve (AUC or ETP) as the key parameter in thrombin generation assays, it is not known which parameters reflect hemostatic function in vivo. Because the parameters of thrombin generation are not independent, it is very difficult to determine which of them are most closely related the hemostasis. However, our current data suggest that a 4-factor PCC is able to completely correct the time to hemostasis is a reproducible mouse model, even though it does not the lag time in an assay of thrombin generation