RSS Feeds
Ultrasonography to Guide Duration of Anticoagulation in DVT
- Paolo Prandoni, MD, PhD;
- Anthonie W.A. Lensing, MD, PhD; and
- Martin H. Prins, MD, PhD
IN RESPONSE:
In response to Dr. Kim, in patients with acute symptomatic proximal DVT who have been treated for 3 to 6 months with conventional anticoagulation, residual thrombosis on ultrasonography was consistently shown to occur in 60%, 40%, and 30% of patients at 6, 12, and 24 months, respectively (1, 2). In our study, 29.5% of patients randomly assigned to the recommended fixed duration of anticoagulation still had residual thrombosis at their last ultrasonography assessment, which was performed after 1 year in patients with secondary DVT and after 2 years in patients with unprovoked DVT. This rate was statistically significantly higher than that (20.0%) found in patients randomly assigned to the flexible duration. As the 2 study groups had fully comparable baseline characteristics, the prolonged use of anticoagulants in the flexible-duration group probably contributed to the lower observed rate of residual venous thrombosis. Because residual thrombus after DVT treatment is a well-known risk factor for recurrent DVT (2–4), we agree with Dr. Kim that the reduction in recurrent venous thromboembolism (VTE) observed in the flexible-duration group may have resulted from the lower rate of residual thrombosis.
We agree with Dr. Imfeld and colleagues that the better outcome in patients randomly assigned to flexible-duration anticoagulation versus fixed-duration anticoagulation reflects the longer duration of anticoagulation. The advantage of our approach lies in the identification of a subgroup of patients who can benefit from prolonging anticoagulation without exposing many people who are less likely to develop recurrent VTE to the risk of anticoagulants. Our decision model, which stipulates continuation or cessation of anticoagulation in response to ultrasonography results over time, reflects increasing understanding that postbaseline variables may be as important as (or more important than) baseline characteristics, such as sex, thrombophilia, and type of DVT, for predicting risk for VTE recurrence. As far as the statistical approach is concerned, we performed the primary analysis with a Cox proportional hazards model, which allows adjustment for confounders. The P value in Kaplan–Meier log-rank analysis showed similar results (P = 0.047).
Anthonie W.A. Lensing, MD, PhD
University of Amsterdam
Amsterdam, the Netherlands
Martin H. Prins, MD, PhD
University of Maastricht
NL-6200 MD Maastricht, the Netherlands
Article and Author Information
-
Potential Conflicts of Interest: None disclosed.
