Safety and Efficacy of an Absorbable Filter in the Inferior Vena Cava to Prevent Pulmonary Embolism in Swine

Abstract

Purpose To evaluate the immediate and long-term safety as well as thrombus-capturing efficacy for 5 weeks after implantation of an absorbable inferior vena cava (IVC) filter in a swine model. Materials and Methods This study was approved by the institutional animal care and use committee. Eleven absorbable IVC filters made from polydioxanone suture were deployed via a catheter in the IVC of 11 swine. Filters remained in situ for 2 weeks (n = 2), 5 weeks (n = 2), 12 weeks (n = 2), 24 weeks (n = 2), and 32 weeks (n = 3). Autologous thrombus was administered from below the filter in seven swine from 0 to 35 days after filter placement. Fluoroscopy and computed tomography follow-up was performed after filter deployment from weeks 1-6 (weekly), weeks 7-20 (biweekly), and weeks 21-32 (monthly). The infrarenal IVC, lungs, heart, liver, kidneys, and spleen were harvested at necropsy. Continuous variables were evaluated with a Student t test. Results There was no evidence of IVC thrombosis, device migration, caval penetration, or pulmonary embolism. Gross pathologic analysis showed gradual device resorption until 32 weeks after deployment. Histologic assessment demonstrated neointimal hyperplasia around the IVC filter within 2 weeks after IVC filter deployment with residual microscopic fragments of polydioxanone suture within the caval wall at 32 weeks. Each iatrogenic-administered thrombus was successfully captured by the filter until resorbed (range, 1-4 weeks). Conclusion An absorbable IVC filter can be safely deployed in swine and resorbs gradually over the 32-week testing period. The device is effective for the prevention of pulmonary embolism for at least 5 weeks after placement in swine. © RSNA, 2017.

Figures

Figure 1:

Photograph of the absorbable IVC filter made from polydioxanone suture. The filter is composed of a stent portion that adheres to the caval wall and a cone portion.

Figure 2:

Changes involving the IVC at baseline and after deployment of IVC filter (cone and stent delineated by double-headed arrows in top row) on fluoroscopic images (top row), axial CT images (middle row, white arrows), and hematoxylin-eosin–stained slides (bottom row) for five different swine. On the baseline fluoroscopic image, note right (arrowhead) and left (double arrowhead) renal vein inflow. Filters were deployed in an infrarenal position. Baseline histologic slides were not obtained because animals were not immediately killed after filter deployment. Axial CT and hematoxylin-eosin–stained images were obtained through the stent component of the filter and demonstrate hyperplasia of the inferior caval wall, which is most pronounced at weeks 2 and 5 after filter deployment. The embedded fragments of polydioxanone (black arrows) are located within the caval wall and gradually degrade with evidence of granulomatous inflammation (ie, multinucleated giant cells, histiocytes, and lymphocytes; arrowheads in bottom row), which are difficult to identify by week 32.

Figure 3:

Graph of the average luminal area of the IVC within the stent portion of the filter plotted against time. Area calculations were on the basis of CT imaging and the area of the lumen was modeled as an ellipse. Vertical bars represent maximum and minimum values at each time point.

Figure 4:

Venacavogram immediately after iatrogenic thrombus deployment from the IVC below the level of the filter. The thrombus (*) was captured completely with no evidence of PE at immediate follow-up CT examination of the chest during the pulmonary arterial phase of contrast-agent enhancement (not shown).

Figure 5a:

Slides show filter fragments in the lung tissue of a swine. (a) Hematoxylin-eosin–stained lung tissue filter fragments (arrow and arrowhead) within terminal pulmonary arterioles in a swine that was euthanized at 24 weeks after filter deployment. (b) Confirmation of the polydioxanone filter fragments (arrow and arrowhead) was performed by passing the slide under polarized light. Fragments were surrounded by granulomatous inflammation, which is similar to findings identified in the cava wall. Fragments measured 11 × 31 μm (arrow) and 13 × 13 μm (arrowhead).

Figure 5b:

Slides show filter fragments in the lung tissue of a swine. (a) Hematoxylin-eosin–stained lung tissue filter fragments (arrow and arrowhead) within terminal pulmonary arterioles in a swine that was euthanized at 24 weeks after filter deployment. (b) Confirmation of the polydioxanone filter fragments (arrow and arrowhead) was performed by passing the slide under polarized light. Fragments were surrounded by granulomatous inflammation, which is similar to findings identified in the cava wall. Fragments measured 11 × 31 μm (arrow) and 13 × 13 μm (arrowhead).