The standard treatment for patients with acute deep venous thromboembolism (DVT) is five to seven days of unfractionated or low-molecular-weight heparin, followed by three months of oral anticoagulant therapy. Three months of anticoagulation is believed to be sufficient to prevent the recurrence of DVT in patients with no definable risk factors. Kearon and associates hypothesized that patients may actually benefit from more than three months of warfarin therapy to further reduce the risk of thromboembolic disease. They performed a double-blind, randomized trial comparing an additional 24 months of warfarin therapy with placebo therapy in patients who had already received the standard three months of anticoagulation for an idiopathic DVT. In addition, the authors attempted to determine if prothrombotic abnormalities, including factor V Leiden, G20210A prothrombin gene mutation or antiphospholipid antibodies, might identify patients who are at high risk for recurrent DVT.

Patients enrolled in the study had completed three months of uninterrupted warfarin therapy after their first incident of idiopathic venous thromboembolism. The incident could be a confirmed DVT or a pulmonary embolism that occurred in the absence of any major thrombotic risk factors. Such risk factors might include a fracture or plaster casting of a lower limb; three consecutive days of either hospitalization with bed rest or use of general anesthesia, each occuring within the previous three months; a known deficiency of protein C, protein S or antithrombin; or a diagnosis of cancer in the previous five years. Patients were also excluded if they had contraindications to long-term anticoagulation therapy, including the need for long-term treatment with aspirin or nonsteroidal medication, a familial bleeding diathesis, pregnancy or a major psychiatric disorder.

At baseline, all patients had a ventilation-perfusion lung scan, ultrasonography of proximal leg veins and, if possible, bilateral impedance plethysmography to increase the accuracy of later diagnosis of recurrent venous thromboembolism. They also had assays for factor V Leiden, G20210A prothrombin gene mutation, anticardiolipin antibodies and lupus anticoagulant. Patients were then randomly assigned to continue treatment with warfarin or placebo tablets. The initial dose of warfarin was based on the International Normalized Ratio (INR), which was obtained on the day of randomization. For the patients in the placebo arm, sham INR results were given. The doses of warfarin and placebo were each adjusted so that an INR (or sham INR) of 2.0 to 3.0 was maintained.

Of the 327 patients who initially met the inclusion criteria, 123 were dropped for meeting the exclusion criteria. Of the 204 remaining patients, 162 patients gave informed consent and were randomized to receive treatment with warfarin or placebo. The mean age of the patients was 59 years, and 60 percent of the patients were men.

Only one of 79 patients who received warfarin had a recurrent venous thromboembolism. Of the 83 patients who were given placebo, 17 had recurrent venous thromboembolism, including 11 DVTs, five nonfatal pulmonary embolisms and one fatal pulmonary embolism. The calculated recurrence rate of venous thromboembolism was 1.3 percent per patient per year in the warfarin group and 27.4 percent in the placebo group. Because of this significant difference, the trial was terminated early. Follow-up was originally intended to be 24 months but, because of the observed outcomes, lasted only an average of 10 months. Three patients in the warfarin group had nonfatal major bleeding episodes. When the biochemical markers for hypercoagulability were assessed, 26 percent of patients had factor V Leiden and 5 percent had antiphospholipid antibodies and the G20210A prothrombin gene mutation. However, the presence of lupus anticoagulant was the only variable significantly associated with recurrent venous thromboembolism.

The authors conclude that patients who have had a first episode of an idiopathic DVT are at high risk for a recurrent episode and should receive anticoagulation therapy for longer than three months. However, the optimal duration of therapy beyond this point is still unknown.

editor's note: This study has significant clinical ramifications for practicing physicians who treat patients with DVT. Unfortunately, biochemical markers, with the exception of lupus anticoagulant, did not help to predict who would have recurrent disease. In addition, it seems obvious that patients should receive more than three months of warfarin, but the critical question will now be when to stop treatment. Perhaps treatment should never stop, if indeed the propensity to develop DVT is a marker for chronic disease.—j.t.k.