Assessment of Risk of Recurrent Venous Thromboembolism

Editor's Note: Commentary based on Eichinger S1, Heinze G, Kyrle PA.D-dimer levels over time and the risk of recurrent venous thromboembolism: an update of the Vienna prediction model. J Am Heart Assoc 2014;3(1):e000467. doi: 10.1161/JAHA.113.000467.


Previous thrombotic episodes have been known to be a major risk factor for recurrent venous thromboembolism (VTE). The magnitude of the risk is highly dependent on patient-specific factors, including permanent risk factors such as inherited thrombophilia and reversible risk factors such as recent surgery and immobilization. Risk stratification is important in determining a patient's anticoagulation regimen and its duration, especially in patients with a high risk of bleeding. There have been a number of methods proposed to assess the risk of recurrent VTE by integrating multiple factors, including D-dimer level, gender, age, and thrombus location.1,2,3 In the current study, the authors included more patients and expanded their original model, the Vienna Prediction Model,4 to assess the recurrent risk at multiple time points later than three weeks after completion of anticoagulation.


A prospective cohort study of 553 patients from four thrombosis centers in Vienna, Austria was conducted. Patients with an initial unprovoked VTE seen between January 2000 and August 2008 who were on oral anticoagulants for more than three months were included. Exclusion criteria included patients with VTE provoked by surgery, trauma, pregnancy, female hormone intake, thrombophilia, or cancer. The primary efficacy endpoint of the study was recurrent symptomatic deep vein thrombosis (DVT) or symptomatic pulmonary embolism (PE) confirmed by imaging studies. The D-dimer levels were measured three weeks and three, nine, 15, and 24 months after discontinuation of vitamin K antagonists. Using a dynamic prediction model, the authors generated a web-based calculator and normograms to calculate risk scores and recurrence rates of VTE based on gender, location of VTE, and D-dimer level.

Table 1: Hazard Ratios for Recurrent Venous Thromboembolism from the Multivariable Dynamic Prediction Model


Time after discontinuation of anticoagulation

Hazard Ratio (95% CI)

Male vs female

3 Weeks

(1.57 to 377)

3 Months 2.27 (1.48 to 3.48)
9 Months 1.98 (1.30 to 3.02)
15 Months 1.73 (1.11 to 2.69)

Proximal DVT or PE
vs distal DVT

3 Weeks

1.84 (1.00 to 3.43)

3 Months 1.68 (0.91 to 3.10)
9 Months 1.49 (0.79 to 2.81)
15 Months 1.44 (0.76 to 2.72)

D-dimer (per doubling)

3 Weeks

1.30 (1.07 to 1.58)

3 Months 1.27 (1.06 to 1.51)
9 Months 1.20 (1.02 to 1.41)
15 Months 1.13 (0.95 to 1.36)


Patients were followed in median for almost six years. Among the 553 patients, the cumulative probability of recurrence was 7.2% (95% CI, 5.0% to 9.3%) after one year and 23.0% (95% CI, 19.2% to 26.6%) after five years. A multivariable dynamic prediction model demonstrated that male versus female sex, proximal versus distal DVT, and high D-dimer level were associated with high hazard ratios for recurrent VTE at different time points after the initial event, although their effect decreased over time (Table 1). This study established the utility of a web-based calculator that can predict recurrence risk at any point between three weeks and 60 months after discontinuation of anticoagulation. The investigators also established normograms to predict recurrence risk at three weeks, three, nine, and 15 months based on these data.


The risk of recurrent VTE can be estimated by combining a patient's gender, location of previous VTE, and D-dimer measurements at multiple time points after discontinuation of anticoagulation.


The annual incidence of VTE is one to two cases per 1000 people, with a recurrence rate of five to 10 percent per year. The pathogenesis of recurrent VTE is multifactorial and predisposing factors include location of previous VTE, male gender, age, and D-dimer level. Previous studies have suggested various ways to estimate rate of recurrent VTE to guide the duration of anticoagulation therapy given the risk of bleeding with prolonged anticoagulation. For example, Rodger et al. showed in the Canadian prediction rule that female sex, leg without hyperpigmentation, edema or redness, D-dimer <250 µg/L, body mass index (BMI) <30, and age <65 confer low risk of VTE recurrence.1Tosetto el al. demonstrated that a prognostic recurrence score called "DASH (D-dimer, Age, Sex, and Hormonal therapy)" correlated with rate of recurrent VTE.2

The authors of the current study updated their previously published Vienna Prediction Model,4 which combined a patient's gender, location of initial VTE, and serial D-dimer measurements to predict the rate of recurrent VTE using a web-based calculator. However, this study has limitations, as it does not include patients with VTE provoked by surgery, trauma, pregnancy, female hormone intake, thrombophilia, or cancer. Moreover, external validation would be necessary to assess whether this study can be generalized to populations other than the study population from Austria. It will be interesting to analyze subgroups of the study population to evaluate if the risk calculator works well across major demographic groups including women, elderly, racial minorities, diabetics, and patients with impaired renal function. In addition, it would be of interest to evaluate whether risk factors identified in previous studies, such as the presence of clinical symptoms and BMI, should be incorporated in the web-based VTE risk calculator.

Estimating the rate of recurrent VTE to help identify the optimal duration of anticoagulation therapy remains a clinical challenge. Current guidelines recommend at least three months of anticoagulation therapy in patients with unprovoked DVT, and recommend longer therapy duration in patients with a favorable risk/benefit assessment.5 Despite its limitations, this study provides a useful clinical practice tool to assess recurrent VTE risk up to 60 months, which can be used to guide anticoagulation therapy for patients with unprovoked VTE.


  1. Rodger MA, Kahn SR, Wells PS, et al. Identifying unprovoked thromboembolism patients at low risk for recurrence who can discontinue anticoagulant therapy.CMAJ 2008;179:417-26.
  2. Tosetto A, Iorio A, Marcucci M, et al. Predicting disease recurrence in patients with previous unprovoked venous thromboembolism: a proposed prediction score (DASH). J Thromb Haemost 2012;10:1019-25.
  3. Palareti G, Cosmi B, Legnani C, et al. D-dimer testing to determine the duration of anticoagulation therapy. N Engl J Med 2006;355:1780-9.
  4. Eichinger S, Heinze G, Jandeck LM, Kyrle PA. Risk assessment of recurrence in patients with unprovoked deep vein thrombosis or pulmonary embolism: the Vienna prediction model. Circulation 2010;121:1630-6.
  5. Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:e419S-94S

Keywords: Anticoagulants, Cohort Studies, Austria, Fibrin Fibrinogen Degradation Products, Neoplasm Recurrence, Local, Pulmonary Embolism, Risk Factors, Thrombophilia, Thrombosis, Venous Thromboembolism, Venous Thrombosis

< Back to Listings