Commentary

5 previous, well-designed, randomised trials have compared different durations of anticoagulation for venous thromboembolism (VTE).^1–5 3 of these studies tested whether the duration of anticoagulation could be reduced from 3^{1, 3} or 6 months² to only 4 or 6 weeks in patients who, for the most part, had a first episode of VTE and did not have cancer. All 3 studies found that shortening the duration of treatment resulted in about a 2-fold increase in the rate of recurrent VTE after anticoagulants were stopped, without being associated with a convincing reduction in the rate of bleeding. These studies also found that the risk for recurrent VTE was about 4 times higher for “unprovoked” or “idiopathic” VTE than for thrombosis that was provoked by a transient risk factor, such as recent surgery.

The other 2 trials evaluated an extended duration of anticoagulant treatment in 2 groups of patients with VTE who were thought to have a high risk for recurrence: those with a second episode of VTE⁴ and those with a first unprovoked episode of VTE.⁵ Both studies found that extending oral anticoagulant treatment (international normalised ratio of 2.0 to 3.0) was effective for preventing recurrent VTE; however, long term use was associated with a substantial risk for bleeding. Schulman et al⁴ concluded that after a second episode of thrombosis the risk for recurrent VTE after completing 6 months of treatment (about 6% per year) was not high enough to routinely recommend extended treatment. Our group found that after a first episode of unprovoked thrombosis, the risk for recurrent VTE after completing 3 months of treatment was higher than expected (about 25% during the first year), which justified treatment for longer than 3 months; however, we could not say how much longer than 3 months was required.⁵

In a mix of patients with provoked and unprovoked VTE, Pinede et al found the risk for recurrent VTE to be similar after completing 3 and 6 months of treatment. In patients with a first episode of unprovoked deep venous thrombosis, Agnelli et al found a similar risk for recurrent VTE after completing 3 months and 12 months of treatment. In the latter study, this risk was only about 5% per year, much lower than the 25% observed after 3 months of treatment in a similarly defined patient population.⁵

A consistent finding across studies is that 3 months of anticoagulant treatment is adequate for patients with VTE that has been provoked by a major transient risk factor. The subsequent risk for recurrence is low (< 5% per year); however, uncertainty exists when the risk for recurrence after an unprovoked VTE is as low after 3 months of treatment as it is after 6 months. Although the 2 new studies strongly suggest that 3 months of treatment achieves as low a subsequent risk for recurrent VTE as 6 or 12 months of treatment, neither study had the power to establish this definitively. On the basis (mostly) of the results of our own study,⁵ which is representative of our patient population, I will continue to recommend a minimum of 6 months of anticoagulant treatment for a first unprovoked episode of VTE, provided patients do not have a high risk for bleeding. Current evidence suggests that patients with unprovoked VTE who have been treated for 6 months or longer have a risk for recurrent VTE of about 10% during the year after anticoagulant treatment is stopped. Therefore, once patients have successfully completed 6 months of treatment, I encourage them to remain on long term treatment provided they have a low risk for bleeding and do not find treatment a burden.