Abstract

Background: Venous thromboembolism (VTE) has a tendency to recur after anticoagulation is stopped. The optimal duration of anticoagulation is influenced by the risk of recurrence: indefinite anticoagulation is recommended after a first VTE associated with a persistent and strong risk factor (e.g. cancer), whereas three months of anticoagulation is adequate if VTE was provoked by a transient risk factor (e.g. surgery). The optimal duration of anticoagulation after a first unprovoked VTE is, however, still a matter of debate. An attractive approach is to identify subgroups of patients with unprovoked VTE who have a high risk of recurrence and treat them with prolonged anticoagulation, and subgroups with a lower risk of recurrence and treat them for only three months. A critical issue, therefore, is to determine a cut-off value for risk of recurrence that is low enough to justify stopping anticoagulant therapy at three months (i.e., a rate of recurrence that is acceptable to patients and physicians). We propose that the rate of recurrence after VTE that was provoked by a transient risk factor represents such a value.

Aim of the study: To accurately estimate the risk of recurrence in patients with VTE provoked by a transient risk factor who have completed at least three months of anticoagulant therapy.

Materials and Methods: Medline, Embase and Cochrane Collaboration Registry of Randomized Trials were searched for any studies reporting the recurrence rate of VTE after a first episode of VTE associated with a transient risk factor (surgery, trauma, plaster, bed rest, pregnancy, puerperium, hormone treatment). The references of retrieved articles were scanned for any additional relevant studies. Studies were included if enrolling patients met the following criteria:

  • a first episode of VTE provoked by a transient risk factor;

  • a course of at least three months of oral anticoagulant therapy;

  • a follow up of 12 or 24 months after treatment discontinuation with assessment of VTE recurrence rate.

Recurrence rate, or data that allowed its calculation, needed to be reported. An overall estimate of the recurrence rate was calculated following Laird (Stats Med 1990), having predefined to use a fixed-effect model if no heterogeneity was found among the studies or a random-effect model otherwise. The inverse variance method was used to calculate weights for the studies.

Results: The literature search yield 1089 references. After careful scanning of the potentially relevant papers, 15 papers were included in the final analysis. All studies except one were prospective. 12/15 and 11/15 studies reported data about the recurrence rate at 12 and 24 months, respectively. Overall there were 106 events among of 2217 patients at 12 months and 160 events among 2321 patients at 24 months. The pooled recurrence rate was 4.0% (95% C.I. 3.0%–5.2%) at 12 months and 6.7% (95% C.I. 5.2%–8.6%) at 24 months.

Conclusion: The cumulative risk of recurrence in patients with VTE provoked by a reversible risk factor is 6.7% (95% C.I. 5.2%–8.6%) two years after anticoagulant withdrawal. We suggest that it acceptable to stop anticoagulant therapy after three months in subgroups of patients with unprovoked VTE who have been shown to have a risk of recurrence that is similar to, or lower than, this rate.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author