Risk of atherothrombotic events in patients after proximal deep-vein thrombosis

Several studies have shown high incidence of atherothrombotic events (ATEs) in patients with unprovoked pulmonary embolism. This association remains understudied in patients presenting with deep-vein thrombosis (DVT). We evaluated the risk factors for and the incidence of ATEs in patients with unprovoked proximal DVT, and compared it to patients with provoked DVT and reference patients without DVT. Patients with compression ultrasonography (CUS)-proven unprovoked proximal DVT, provoked DVT, and symptomatic patients in whom DVT was excluded by CUS were followed and scored for the occurrence of ATEs. A total of 1235 patients were enrolled: 170 patients with provoked, 74 patients with unprovoked DVT, and 991 patients without DVT. During follow-up, 128 ATEs occurred (incidence 6.5/100 patient-years). Adjusted hazard ratio was not different between patients with DVT and without DVT [hazard ratio 1.4; 95% confidence interval (CI) 0.76-2.4]. In contrast, patients with unprovoked DVT suffered from ATEs more frequently than provoked DVT patients (hazard ratio 3.16; 95% CI 1.1-9.1) and reference patients (hazard ratio 2.8; 95% CI 1.3-5.7). Notably, when fully adjusted for the known ATE risk factors, the risk differences between references, provoked, and unprovoked DVT patients diminished: hazard ratio 1.1 (95% CI 0.47-2.5) and 1.7 (95% CI 0.80-3.6), respectively. Our study confirms that the risk of ATEs in patients with unprovoked DVT was higher than in those with provoked DVT or reference patients. Interestingly, our results raise the question whether the known risk factors for ATE and venous thromboembolism attribute equally in pulmonary embolism and DVT patients, and contradict a causal relation between ATE and proximal DVT.