Long-term risk of venous thrombosis after stopping anticoagulants for a first unprovoked event: A multi-national cohort

Background: Choosing short-term (3-6 months) or indefinite anticoagulation after a first unprovoked venous thromboembolic event (VTE) is a common and difficult clinical decision. The long-term absolute risk of recurrent VTE after a first unprovoked VTE, in all patients and sub-groups, is not well established, hindering decision making. Methods: We conducted a multi-center multi-national prospective cohort study in first unprovoked VTE patients to establish the long-term risk of recurrent VTE after short-term anticoagulation in first unprovoked VTE patients (and sub-groups). We followed patients for symptomatic suspected VTE off of OAT. Suspected recurrent VTE was investigated with reference to baseline imaging and then independently and blindly adjudicated. Findings: We recruited 663 participants between October, 2001 and March 2006 with the last follow-up in April 2014. During a mean 5.0 years of follow-up, 165/663 suspected VTE (in 408 patients) were adjudicated as recurrent VTE resulting in an annualized risk of recurrent VTE of 5.0% (95% CI: 4.2-5.8%) with a cumulative risk of 29.6% at 8 years. Men had a 7.6% (95% CI: 6.3-9.2%) annual risk of recurrent VTE. High risk women (2 or more HERDOO2 points; see text) had an annual risk of recurrent VTE of 5.9% (95% CI: 4.2-8.1%). Low riskwomen (1 or 0 HERDOO2 points) had 1.1% (95% CI: 0.6-2.0%) annual risk of recurrent VTE with a cumulative risk of 8.7% at 8 years. Interpretation: Men and high risk women with unprovoked VTE should be considered for long-term anticoagulant therapy given a high risk of recurrent VTE after long-term follow-up. Women with a low HERDOO2 score may be able to safely discontinue anticoagulants. Funding: This study was funded by the Canadian Institutes of Health Research (Grant # MOP 64319) and Heart and Stroke Foundation of Ontario (Grant # NA 6771). Registered at www.clinicaltrials.gov identifier: NCT00261014 (C) 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).