Prevalence and Clinical Outcome of Free-Floating Thrombus Formation in Lower Extremity Deep Veins

Most published guidelines for treatment of acute and subacute deep venous thrombosis (DVT) recommend early ambulation over initial bed rest. However, formation of free-floating thrombus in the lower extremity deep veins carries a high risk of fatal pulmonary embolism (PE) during ambulation. Furthermore, the natural history of free-floating thrombus in this patient population is not known. The purpose of this study was to identify the prevalence and outcome of free-floating thrombus among patients with acute DVT. Between January 2013 and December 2013, 427 patients were diagnosed as having DVT by compression ultrasound. The anatomic distributions of the thrombi were divided into iliofemoral, femoropopliteal, and calf DVT. The ultrasound features of the thrombi were classified as firmly or loosely attached and free-floating. PE was confirmed with computed tomography angiography. Among 427 patients with DVT, seven (1.7%) were found to have free-floating thrombus. The risk factors for DVT included recent knee surgery in two (28.5%), recent hip surgery in one (14.3%), immobilization in one (14.3%), malignant neoplasm in one (14.3%), hormone replacement therapy in one (14.3%), and stroke in one (14.3%). Initial treatment of these patients included bed rest and anticoagulation therapy. Two of these patients underwent temporary insertion of an inferior vena cava filter. The thrombus was located predominantly in the femoropopliteal segment (five patients, 71.4%), and the prevalence of femoropopliteal DVT was significantly higher in patients with free-floating thrombus than in those without (P = .0002). PE was found in three patients (42.9%), and the incidence of PE was significantly higher in patients who had free-floating thrombus than in those who did not (P = .0001). The free-floating thrombus resolved spontaneously in four patients (57.1%) within a mean period of 19 days. On the other hand, the thrombus became firmly attached to the vein wall in the remaining three patients (42.9%) within a mean period of 9 days, leading to secondary thrombus recanalization. No further thromboembolic complications were encountered in these patients. Although the prevalence of free-floating thrombus formation is low among patients with DVT, it can be detected by compression ultrasound in daily practice. Our data show that the natural history of free-floating thrombus treated by anticoagulation can be largely classified into two patterns: early spontaneous thrombus resolution and initial adhesion to the vein wall with secondary thrombus recanalization. Insertion of an inferior vena cava filter can be reserved for patients in whom anticoagulation is contraindicated.