Rethinking warfarin reversal

Even under optimal circumstances a small proportion of patients treated with warfarin will develop serious complications such as intracranial or gastrointestinal hemorrhage. In such situations, the clinician needs to reverse the effect of warfarin as quickly as possible. Administering phytonadione (vitamin K) is a critical element of any warfarin reversal strategy; an intravenous dose of 5 to 10 mg of vitamin K will correct the international normalized ratio (INR; a measure of the anticoagulant effect of warfarin) within 24 hours in most patients.¹  However, with life- or limb-threatening hemorrhage the coagulant potential of the blood must be restored more quickly. Fresh frozen plasma (FFP) has been used for warfarin reversal for decades and is still recommended in consensus guidelines.^2,3  FFP's drawbacks include the inability of quick administration, a significant intravascular volume challenge, and the possibility of rare complications such as transfusion-associated lung injury or blood-borne infection. Recombinant activated factor VII (rVIIa) is a compound approved for the treatment of hemophilia patients with a specific inhibitor. Authors of several anecdotal reports and small case series have suggested that rVIIa may be an effective treatment for warfarin-associated bleeding.

In this issue of Blood, Skolnick and colleagues present a series of experiments designed to evaluate the in vivo and ex vivo effects of rVIIa on healthy subjects treated to a therapeutic INR with warfarin.⁴  The study found that a punch-biopsy–induced bleeding model did reflect the anticoagulant effect of warfarin: both bleeding duration and bleeding volume were increased with warfarin treatment. rVIIa did, at least partially, correct ex vivo coagulation measurements such as the prothrombin time and the activated partial thromboplastin time. The effect of rVIIa was also examined on whole-blood assays that include platelets. When 80 mcg/kg rVIIa was given to 24 warfarin-treated subjects with a mean INR of 2.8, both thrombin generation and thromboelastography improved significantly compared with the same measurements in 24 warfarin-treated subjects who received placebo. Despite corrections in these ex vivo measurements, rVIIa failed to reduce warfarin-induced increases in both the bleeding time and total blood loss in the punch-biopsy model. There are 2 possible interpretations of this last finding: either the model did not fully reflect the complex biology of in vivo clot formation (and thus this model does not accurately portray the efficacy of rVIIa in patients with warfarin-associated bleeding), or rVIIa, despite its favorable effect on laboratory measurements, is not an effective antidote for warfarin. Interpreted in the light of a recent review that highlighted the lack of high-quality evidence supporting the efficacy of rVIIa for warfarin reversal,⁵  I suggest the current paper should lead clinicians to consider other warfarin reversal strategies (which may include FFP or prothrombin complex concentrates) before giving rVIIa to a warfarin-treated patient who has life-threatening bleeding.