Vitamin K decreased INR but did not reduce bleeding in overanticoagulated patients on warfarin

Low-dose oral vitamin K reduced the international normalized ratio among overanticoagulated patients who received warfarin, but the intervention was not superior to placebo in reducing the frequency of bleeding events or thromboembolism, according to data from a randomized trial.

Non-bleeding patients with INR values between 4.5 and 10.0 receiving warfarin were randomly assigned to 1.25 mg oral vitamin K (n=355; 347 analyzed) or placebo (n=369; 365 analyzed). The primary outcome was bleeding events; secondary outcomes were thromboembolism and death.

Bleeding events occurred among 56 patients in the vitamin K group and 60 patients in the placebo group. Major bleeding events occurred among nine patients in the vitamin K group and four patients in the placebo group. Four patients in the vitamin K group and three in the placebo group had thromboembolsim. Seven patients in the vitamin K group and seven in the placebo group died throughout the 90-day period; no deaths were considered the result of thromboembolism.

According to the researchers, patients assigned to vitamin K experienced a more rapid decline in the INR. On the day after treatment, the INR had decreased by a mean of 1.4 among the placebo group and 2.8 among the vitamin K group (P<.001). Three patients assigned to placebo and 25 assigned to vitamin K had an INR <2.0, and 34 assigned to placebo and 136 assigned to vitamin K had an INR between 2.0 and 3.0 (P<.001 for both).

“Although a small dose of oral vitamin K helped to correct supratherapeutic INR values, it did not substantially reduce the frequency of bleeding events,” the researchers wrote. “Furthermore, major bleeding was uncommon in such patients, regardless of whether vitamin K was administered.” – by Stacey L. Adams

Crowther MA. Ann Intern Med. 2009;150:293-300.

As with all 'randomized trials' the devil is in the details. I hope this study is not used as a simplistic crutch not to use low-dose vitamin K in patients with excessive INRs. The risk of 2 mg oral vitamin K is negligible (except perhaps in patients with mitral valve prostheses), while the benefit may continue to be substantial in subgroups of patients at more-than-average risk of severe hemorrhage. Subgroups get buried in “evidence-based medicine” driven trials, which should not replace careful assessment of the individual patient. My recent patient with atrial fibrillation on chronic coumadin and an INR of 6.0 — but with past gastric bleeds — needs rapid improvement in INR, as a not-unusual example. How about the many patients on aspirin and/or clopidogel plus coumadin with super-therapeutic INRs?

– Harry S. Jacob, MD

HemOnc Today Chief Medical Editor