Venous thromboembolism (VTE) is a common multicausal disease in the general population. von Willebrand factor (VWF) has been associated with VTE in epidemiological studies and recently VWF-mediated platelet adhesion has been shown to be critical for deep vein thrombosis (DVT) in mouse models. The large VWF multimers are cleaved by ADAMTS13 into smaller, less active multimers. Because of the key role of ADAMTS13, VWF and factor VIII (FVIII) in hemostasis and their close biological relationship, we aimed at investigating in the same population the effect of these proteins on VTE risk.

Materials and Methods

We included consecutive patients admitted to our anticoagulation clinic with a first objectively confirmed VTE event between January 2007 and July 2011. The study included only patients with DVT of the lower limbs and/or pulmonary embolism aged between 18-70 years, and without medical history of overt malignancy, arterial thrombosis, liver cirrhosis, renal failure on hemodialysis, connective tissue disease and other conditions characterized by chronic inflammation. Blood was collected at least 1 month after stopping anticoagulation and ≥ 6 months after VTE. Controls were recruited from friends or partners of patients, had no personal history of VTE and were matched to patients by gender and age. The other exclusion criteria for controls were the same as for patients, and in both groups pregnant or postpartum women at blood collection were excluded. ADAMTS13 and VWF antigens were determined by commercial ELISA and FVIII activity was measured using FVIII deficient plasma. High VWF (>150%) and FVIII (>150%) were defined by plasma levels of these proteins exceeding the 88th and 94th percentiles of the control group, respectively. To define low ADAMTS13 levels we used the 10th percentile of the controls (≤0.64 μg/ml). Odds ratios (OR) and 95% confidence intervals (CI) were presented adjusted for gender and age and medians with interquartile variation (25th-75th percentiles).


358 patients with a first objectively confirmed VTE event were admitted to our clinic, of whom 282 did not participate in the study due to the exclusion criteria (n= 249) and loss of follow-up or refusal (n= 33). Therefore, we included 76 patients (53 women, 70%) with a median age of 43 years (33-55 years) and 96 controls (66 women, 69%), with a median age of 42 years (31-52 years). In controls, there was a negative albeit weak correlation between ADAMTS13 levels and VWF (rs= -0.213 by Spearman coefficient, p= 0.037) and FVIII (rs= -0.251, p= 0.014). As expected, VTE was associated with high levels of VWF (OR 2.80, 95% CI 1.20-6.54) and FVIII (OR 3.02, 95% CI 1.08-8.43). Low ADAMTS13 levels were detected in 12 patients and 9 controls (OR 1.76, 95% CI 0.70-4.46). The population was dichotomized according to the 88th percentile of VWF of the controls. Median ADAMTS13 was lower in the group with VWF >p88 compared to the group with VWF ≤p88 in controls (0.80μg/ml, 0.63-0.94μg/ml vs. 0.98μg/ml, 0.81-1.23μg/ml, p= 0.051) and in patients (0.75μg/ml, 0.53-0.95μg/ml vs. 0.95μg/ml, 0.81-1.14μg/ml, p= 0.001). Next the population was categorized into subjects with VWF ≤p88 and ADAMTS13 >p10 (reference category: 52 patients, 79 controls), VWF ≤p88 and ADAMTS13 ≤p10 (4 patients, 6 controls), VWF >p88 and ADAMTS13 >p10 (12 patients, 8 controls) and VWF >p88 and ADAMTS13 ≤p10 (8 patients, 3 controls). The combination of high VWF and low ADAMTS13 increased VTE risk when compared to the reference category (OR 4.14; 95% CI 1.03-16.71). The risk did not substantially change with adjustment for acute phase proteins and ABO group, and it was attenuated when further adjusted for high FVIII (OR 3.41, 95% CI 0.80-14.70). No correlation was found between the time since the VTE event and levels of VWF, FVIII and ADAMTS13 (p>0.05).


our data suggest a potential effect of high VWF and low ADAMTS13 on VTE risk. An unbalance between these two interconnected proteins might have a role in the pathophysiology of VTE. Results should be interpreted with caution due to the wide 95% CI and to the attenuation of VTE risk with further adjustment for FVIII. In order to better explain our results studies with larger sample size are warranted.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.