The optimal therapy for patients who present with markedly prolonged INR values during warfarin therapy is undefined. Evidence suggests that the risk of bleeding increases directly with the degree of prolongation of the INR. Traditional therapy for patients with excessive warfarin associated anticoagulation has included warfarin withdrawal with or without vitamin K, transfusion therapy and admission to the hospital. As part of an ongoing international study, we prospectively enrolled 32 consecutive warfarin-treated patients presenting with an INR of more than 10.0. Eligible patients had no evidence of active bleeding or need for immediate correction of their INR. All patients received 2.5 mg of oral vitamin K, were not treated with coagulation factor replacement, and were followed over 90 days for clinical events and their INR response. Seventeen of the 32 patients were women, with an average age of 65.8 (range 31 to 89). Treatment of venous thrombosis and atrial fibrillation (12 patients each) were the most common indications for warfarin. Twenty-five patients had a target INR of 2.0 to 3.0; the remainder had a target of 2.5 to 3.5. The mean INR at presentation was 12.9 (range 10.0 to 21.2). Of the 25 patients with a recorded INR value on the day following vitamin K, 19 (76%) had an INR of 6.0 or less (range 1.6 to 17.5, mean 5.0, 2 less than 2.0). On day 7 after study drug, the mean INR was 3.5 (range 1.6 to 17.5). Six (19%) patients reported bleeding over the 90 days after study drug: 1 (3%) bleed was major (retroperitoneal hemorrhage diagnosed the day after study drug) and 3 patients reported epistaxis or bruising within three days of study drug. Two patients reported late bleeding; one had a fall on day 8 at which time the INR was 10.5, and the other had bleeding associated with a surgical procedure 25 days after study drug. No patients had suspected or confirmed thrombosis, and no patients died during follow-up. This the first prospective study of vitamin K monotherapy for patients with INR values above 10, and confirms prior retrospective analyses which suggest that low dose oral vitamin K effectively lowers the INR in such patients. Our preliminary results also suggest that coagulation factor replacement may be unneeded in such patients. The true risk of bleeding, and the impact of degree of prolongation of the INR on the vitamin K response, will require additional patient recruitment.

Author notes

Corresponding author

Sign in via your Institution