The orally available activated factor X (FXa) and thrombin inhibitors have often been referred to as the ‘new anticoagulants’. They have the potential to replace warfarin and other vitamin K antagonists in the treatment and secondary prevention of venous thromboembolism and in the prevention of stroke in patients with atrial fibrillation. They have been reviewed in this journal (Bates & Weitz, 2006) and more recently elsewhere (Garcia et al, 2010; Weitz, 2011). As their use becomes more widespread, an issue of concern is how to manage bleeding in patients taking these drugs. When a patient on warfarin presents with major bleeding the clinician is reassured that anticoagulation can be completely reversed with a readily available antidote, namely prothrombin complex concentrate (PCC) (Keeling et al, 2011). The new agents have no specific antidote despite reviews and guidance on their ‘reversal’ (Levi et al, 2011; Kaatz et al, 2012). In this issue of the journal Makris et al present a British Committee for Standards in Haematology (BCSH) guideline on the management of bleeding in patients on antithrombotic agents including the new orally available FXa and thrombin inhibitors (Makris et al, 2012). In the absence of a specific antidote the authors, as others have (Kaatz et al, 2012), emphasize the importance of general measures, such as discontinuation of the drug, mechanical compression, surgical or radiological intervention, maintaining blood pressure and urine output, and replacement of blood components in massive haemorrhage. The use of activated charcoal may be an option if the drug was taken in the previous two hours. Dabigatran (but not rivaroxaban and apixaban) has relatively low plasma protein binding and haemodialysis is an option (Warkentin et al, 2012) although, as Makris et al (2012) (under)state, ‘deployment in settings outside intensive care units is likely to be challenging’.

What of non-specific pro-haemostatic agents? This is usually taken to mean PCC, activated PCC [Factor Eight Inhibitor Bypassing Activity (FEIBA)] or recombinant activated factor VII (rFVIIa; Novoseven), but we should not forget the potential of the antifibrinolytic, tranexamic acid, which has been shown to reduce bleeding following trauma (CRASH-2, 2010). The use of PCC, FEIBA or rFVIIa in patients bleeding on new oral anticoagulants has little theoretical basis. PCCs contain Fll, FVll, FlX and FX, but it is difficult to see how simple factor replacement might offset the effects of an inhibitor. The key may simply be providing large amounts of the substrate prothrombin to increase thrombin generation. rFVIIa is used for the management of haemophilia patients with antibodies to FVIII. It is able to directly activate FIX and FX by-passing FVlll (Hedner & Lee, 2011), but inhibition of FXa or thrombin interferes with coagulation below this step and will not be by-passed. FEIBA is also used as a haemostatic agent for patients with antibodies to FVlll. FEIBA contains FVIIa (Luu & Ewenstein, 2004) and it may simply combine the effects of FVIIa and PCC (Marlu et al, 2012), alternatively a FXa:prothrombin complex may be the main active component (Turecek et al, 2004). Again, inhibition of FXa or thrombin interferes with coagulation at or below these points in the coagulation cascade.

We are therefore dependent on studies that have examined the effect of these pro-haemostatic agents on coagulation paramenters in humans or on their effect in animals. These studies are well reviewed by Makris et al but unfortunately only provide poor evidence for weak recommendations. The human studies have examined the ability to reversal abnormalities of standard coagulation tests, such as prothrombin time and thrombin time (Eerenberg et al, 2011) or to normalize impaired thrombin generation (Eerenberg et al, 2011; Marlu et al, 2012). This is very different to stopping bleeding. Why should we be impressed that administration of PCC does not restore a prolonged thrombin time in patients on dabigatran (Eerenberg et al, 2011) when we know that FII, FVII, FIX and FIX cannot influence a thrombin time in which exogenous thrombin is added to clot fibrinogen directly to fibrin? Thrombin generation assays may have more meaning and, in some experiments, FEIBA has looked the most promising agent but the authors acknowledge that using platelet-poor plasma may disadvantage rFVIIa (Marlu et al, 2012). The animal experiments have mostly been presented only in abstract form and have not given clear or consistent answers and we need to be mindful that human proteins will behave differently in animals compared with humans.

In reality, very few patients using these drugs will require a pro-haemostatic agent. If they do, the BCSH give a weak recommendation that ‘In situations with ongoing life-threatening bleeding PCC, APCC and rFVIIa should be considered’ (Makris et al, 2012). We rather unhelpfully recommend that all haematologists should pick the one they will use in extremis and ensure it is available in their hospital.


  1. Top of page
  2. References