Inferior vena cava filter-related thrombosis: Mechanisms, clinical features, prevention, diagnosis, and risk-stratified management.

Inferior vena cava (IVC) filters can provide mechanical protection from pulmonary embolism in selected patients who cannot receive anticoagulation, but filter-related thrombosis is an important complication. A uniform definition is lacking. Thrombus in the filter region may represent emboli captured by the filter, extension of pre-existing deep vein thrombosis, or local thrombus formation around the device. Filter-induced flow disturbance and vessel-wall responses may contribute to local thrombosis, although direct mechanistic evidence in humans remains limited. Reported occurrence varies substantially because study populations, imaging schedules, and outcome definitions differ. Clinically detected caval thrombosis or thrombotic occlusion is uncommon in large databases and prospective cohorts. Planned imaging before retrieval, however, often detects small intrafilter thrombi, most of which are asymptomatic. Accordingly, the clinical importance of an imaging-detected thrombus depends on whether it is accompanied by meaningful caval obstruction or symptoms. Prevention should continue throughout the filter life cycle. Filters should be used only for a time-limited, evidence-based indication. Anticoagulation should be resumed when a separate indication persists and bleeding risk permits, and retrieval should be assessed once filter-mediated pulmonary embolism protection is no longer required. Structured follow-up can help prevent loss to care. Management of established thrombosis should be guided by symptoms, impaired caval outflow, thrombus progression and chronicity, bleeding risk, and the need for filter retrieval. Anticoagulation remains foundational when indicated for active venous thromboembolism. Catheter-directed thrombolysis, mechanical thrombectomy, or combined treatment may be considered for selected symptomatic, progressive, or obstructive cases, but most supporting studies are nonrandomized. Future studies require standardized minimum reporting and patient-centered outcomes.