PURPOSE: Three inferior vena cava (IVC) filters of different designs were studied to identify the potential links between published clinical results for thrombosis and recurrent pulmonary embolism (PE) rates and in vitro hemodynamics patterns in the region of the filters. MATERIALS AND METHODS: The filters studied were the Greenfield over-the-wire filter (Medi-tech/Boston Scientific, Watertown, Mass), TrapEase filter (Cordis Europa, Roden, the Netherlands), and Mobin-Uddin umbrella filter (Edwards Laboratories, Santa Ana, Calif). To assess hemodynamics, velocity contour maps were generated for each filter by using the in vitro photochromic flow visualization technique. Results were obtained for both the unoccluded and partially occluded states. Steady flow (R(e) = 600) was used to model physiologic conditions. To estimate the rates of IVC occlusion and recurrent PE, the authors analyzed published clinical studies spanning more than 30 years and a U.S. Food and Drug Administration database. RESULTS: For both the unoccluded and partially occluded Mobin-Uddin and TrapEase filters, regions of flow stagnation and/or recirculation and turbulence developed downstream of the filter. The Greenfield filter did not produce any prothrombotic flow patterns for either the unoccluded or partially occluded states. Results of published clinical studies supported the hemodynamic findings, with the TrapEase and Mobin-Uddin filters having high rates of IVC occlusion and recurrent PE compared with those of the Greenfield filter. CONCLUSIONS: Flow stagnation or recirculation and turbulence have been linked to thrombosis and thrombus and/or PE formation. Thus, the hemodynamic results from this study may help explain the relatively higher rates of filter thrombosis and PE for the Mobin-Uddin and TrapEase filters versus the Greenfield filter.
PURPOSE: Three inferior vena cava (IVC) filters of different designs were studied to identify the potential links between published clinical results for thrombosis and recurrent pulmonary embolism (PE) rates and in vitro hemodynamics patterns in the region of the filters. MATERIALS AND METHODS: The filters studied were the Greenfield over-the-wire filter (Medi-tech/Boston Scientific, Watertown, Mass), TrapEase filter (Cordis Europa, Roden, the Netherlands), and Mobin-Uddin umbrella filter (Edwards Laboratories, Santa Ana, Calif). To assess hemodynamics, velocity contour maps were generated for each filter by using the in vitro photochromic flow visualization technique. Results were obtained for both the unoccluded and partially occluded states. Steady flow (R(e) = 600) was used to model physiologic conditions. To estimate the rates of IVC occlusion and recurrent PE, the authors analyzed published clinical studies spanning more than 30 years and a U.S. Food and Drug Administration database. RESULTS: For both the unoccluded and partially occluded Mobin-Uddin and TrapEase filters, regions of flow stagnation and/or recirculation and turbulence developed downstream of the filter. The Greenfield filter did not produce any prothrombotic flow patterns for either the unoccluded or partially occluded states. Results of published clinical studies supported the hemodynamic findings, with the TrapEase and Mobin-Uddin filters having high rates of IVC occlusion and recurrent PE compared with those of the Greenfield filter. CONCLUSIONS: Flow stagnation or recirculation and turbulence have been linked to thrombosis and thrombus and/or PE formation. Thus, the hemodynamic results from this study may help explain the relatively higher rates of filter thrombosis and PE for the Mobin-Uddin and TrapEase filters versus the Greenfield filter.