Aspirin and recurrent venous thromboembolism in patients with symptomatic atherosclerosis: retrospective cohort study


Paolo Prandoni, Department of Cardiothoracic and Vascular Sciences, Clinica Medica 2, University of Padua, Via Giustiniani 2, 35128 Padua, Italy.
Tel.: +39 49 8212656. fax: +39 49 8218731;

In a recent placebo-controlled, randomized clinical trial, aspirin (100 mg once daily) was found to reduce by approximately 40% the risk of recurrent venous thromboembolism (VTE) in patients after a first episode of acute unprovoked deep vein thrombosis (DVT) and/or pulmonary embolism (PE) who had completed a 6- to 18-month period of conventional anticoagulation [1]. As patients with symptomatic atherosclerosis were excluded from the study, whether these findings also apply to this category of patients is uncertain. We had the opportunity to retrieve information from a large cohort of patients with VTE, with and without symptomatic atherosclerosis, who had a long-term follow-up after their thrombotic episode [2]. Below is a report of our findings.

Between May 1991 and April 2003, 1919 consecutive patients presented at three study centers with a first episode of VTE, which was either unprovoked or triggered by transient risk factors, such as recent surgery or trauma, immobilization, hormonal treatment, pregnancy or puerperium. They were followed-up for an average period of 4 years after discontinuing anticoagulation. Of these patients, 256 (13.3%) had a history of symptomatic atherosclerosis, defined as a history of ischemic stroke, transient ischemic attack, acute myocardial infarction, angina or intermittent claudication. In all 256 patients aspirin in doses ranging from 80 to 160 mg once daily was prescribed for an indefinite length of time soon after discontinuing vitamin K antagonists (VKA), whereas the remaining 1663 patients were left without any antithrombotic treatments. The aim of the present study was to calculate on an intention-to-treat basis the adjusted hazard ratio (HR) and its 95% confidence intervals (CI) of recurrent VTE, as confirmed by objective tests, in patients with and without symptomatic atherosclerosis. Otherwise unexplained sudden deaths were excluded from the evaluation. The calculation was repeated after analyzing only the patients (1063) with unprovoked VTE.

Table 1 shows the main characteristics of the study patients, separately for those with and without previous symptomatic atherosclerosis. As shown in Table 1, patients with atherosclerosis were on average older and had a higher prevalence of unprovoked VTE. Recurrent VTE during follow-up was diagnosed in 44 of the 256 patients (17.2%) with symptomatic atherosclerosis who were given aspirin, and in 330 of the 1663 patients (19.8%) without atherosclerosis who were left without any antithrombotic treatments. The cumulative incidence of recurrent VTE, calculated with the log-rank test in Kaplan–Meier survival analysis, was 8.7% (95% CI, 5.2–12.2) after 1 year, 16.6% (11.5–21.7) after 3 years, 23.3% (16.2–30.3) after 5 years and 27.1% (18.7–35.5) after 10 years in patients with symptomatic atherosclerosis (patient-years recurrence risk, 0.054). The corresponding figure in patients without symptomatic atherosclerosis was 6.1% (95% CI, 4.9–7.3), 15.7% (13.7–17.7), 23.1% (20.6–25.6) and 33.4% (29.3–37.5), respectively (patient-years recurrence risk, 0.047). The difference was not statistically significant (P = 0.74). After adjustment for age, gender and duration of VKA treatment, the HR of recurrent VTE in patients with symptomatic atherosclerosis in comparison to the remaining patients was 0.92 (95% CI, 0.66–1.27). These findings did not change when the analysis was confined to the patients with unprovoked VTE (HR 0.82; 95% CI, 0.56–1.21).

Table 1.   Characteristics of patients with and without symptomatic atherosclerosis
FeaturesPatients with atherosclerosis (n = 256)Patients without atherosclerosis (n = 1663) P
  1. *Medical diseases in 41, recent surgery in 33, recent trauma in 11, hormonal treatment in 2.

  2. Recent trauma in 265, recent trauma in 244, medical diseases in 84, hormonal treatment in 65, pregnancy in 29, puerperium in 6. Unless otherwise indicated, numbers in parenthesis indicate percentages.

  3. SD, standard deviation; DVT, deep vein thrombosis; PE, pulmonary embolism.

Age (mean ± SD)73.0 ± 11.661.3 ± 17.5<0.001
Male gender138 (54)725 (44)0.002
Type of atherosclerosis
 Coronary heart disease103 (74.6) 
 Peripheral arterial disease64 (46.4) 
 Cerebrovascular disease59 (42.7) 
 Others13 (9.4) 
Thrombophilia (in tested)22/84 (26.2)270/994 (27.2)0.84
Index event
 DVT alone150 (58.6)1104 (66.4)0.073
 PE alone54 (21.1)255 (15.3)
 DVT and PE52 (20.3)304 (18.3)
Modality of presentation
 Unprovoked163 (63.7)900 (54.1)0.003
 Provoked93 (36.3)*763 (45.9)
Antithrombotic treatment, months (median, interquartile range)6 (8)6 (4)0.46

As all patients with previous symptomatic atherosclerosis were prescribed aspirin after discontinuation of VKA, we cannot conclude that aspirin was ineffective for the prevention of recurrent VTE in this category of patients. However, the absolute rate of recurrent VTE events (more than 17%) in these patients did not differ from that (< 20%) observed in those without atherosclerosis who were left without any antithrombotic treatments and, therefore, is not reassuring [3]. We acknowledge that information about the duration of aspirin treatment and about compliance to treatment was not available. However, the lack of this information is unlikely to impact on the interpretation of the study results, which was essentially based on an intention-to-treat analysis.

The reason why aspirin does not seem to confer any appreciable protection against recurrent VTE in patients with symptomatic atherosclerosis is virtually unknown. The practical implication of our observation is that, contrary to the general expectation [1,4], whenever patients with symptomatic atherosclerosis are deemed to require long-term protection against recurrent VTE, they are unlikely to benefit from (resuming) aspirin. Aspirin should be given on top of VKA only to those patients in whom the combination with conventional anticoagulation is expected to increase the benefit-to-risk in comparison to anticoagulation alone [5].

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.