| Objective:Consensus have been reached on the employment of vena cava filter to reduced lethal pulmonary embolism, however, there are still some of studies and our clinical data proved that implantation of filter might lead to several types of complications, including filter related thrombosis, shifting, tilt and embolism of the filter. After establishment of the in vitro and in vivo filter implantation model, we aimed to assess the impact of the filter on vena cave flow-field by particle image velocimetry(PIV) test,and the pathological changes and surface coverage of the filter.Methods:1.A retrospective study of 164 consecutive patients with retrievable inferior vena cava filter placement during June 2008 to June 2014 was performed. We recorded the morphology, size and location of the retrievable inferior vena cava filter associated thrombus, and analyzed the relationship between the thrombus and filter tilt, filter hook wall apposition, dwell time, vein expansion or ancoagulation factors.2.We collected the imaging data from 100 patients who underwent digital substraction angiography or color doppler examination during January 2010 to January 2013 and measured the mean diameter of the vena cava dilation after filter implantation and normal vena cava adjacent to the site of filter implantation. An in vitro physical tube model was then established by using 1:1 ratio according to the above data.3.Thirty-five percentage glycerine solution, which has similar viscosity to human blood, was used as the circulation fluid and the flow-field test was determined by using PIV test in the in vitro model.4.Twenty beagle dogs were employed for in vivo evaluation. Four dogs in contrasr-team.Sixteendogs in filter team. After anesthesia, a retrievalable filter wasintroduced through femoral vein and implanted into 1 cm position at the inferior segment of vena cava. Color doppler examination, dual source CT scan and multiple dimensional imaging reconstruction was performed at 2, 4, 8 and 12 week time point. We also sacrified these dogs at 2, 4, 8 and 12 weeks to obtain the vena cava located at site of filter implantation and 1 cm adjacent. The surface coverage inside of the filter was observed and analyzed by direct eye observation, light and electronic microscopy. Moreover, we also performed the comparison of the neointimal formation at 2, 4, 8 and 12 weeks.Results1. The total risk offilter associated thrombosis was 31.1%. In the setting of filter tilt ≥ 15°, hook wall apposition, vein expansion rate ≥ 50% and dwell time > 2 weeks,the risk of filter associated thrombosis was 70.3% 、 79.6% 、 64.7% and 86.2%,respectively, and the presence of filter related thrombosis were significantly higher than the control group in these settings. The mean length of throbus was 8.6 ± 1.7mm, 5.3 ±1.4 mm, 10.1 ± 2.2 mm and 3.6 ± 0.9 mm in the setting of filter tilt ≥ 15°, hook wall apposition, vein expansion rate ≥ 50% and dwell time > 2 weeks and a significant difference was found between dwell time > 2 weeks and filter tilt ≥ 15° or vein expansion rate ≥ 50%2. The in vitro physical model of filter implantation was successfully established.The laminar flow was main form of blood flow before filter implantation, and a significant vena cava dilation with reduced blood flow was found after filter implantation. When the filter tilt was 0°, 5° and 10°, the laminar form of blood flow without axial flow deviation was observed, suggesting a lower shear force of blood flow to the vessel wall. When the filter tilt was 15° and presence of hook wall apposition, we did not observed a laminar flow and the axial flow deviated from filter was found, indicating a significantly increased shear force of blood flow to the vessel wall, which could result in generation of turbulence and vortex.3. In filter team,all the 16 dogs were implanted with vena cava filter successfully.The color doppler examination revealed that preoperative mean diameter of vena cava at lower segement of kidney was 13.47 ±0.38 mm, and the postoperative mean diameter of vena cava 17.81±0.47 mm, 17.70±0.46 mm, 17.65±0.40 mm and17.39±0.56 mm at 2, 4, 8 and 12 weeks. A significant difference was found on the diameter of vena cava before and after filter implantation(p<0.001). The flow velocityof vena cava before operation was 45.6±1.69cm/s, whereas the mean axial flow23.3±2.01cm/s, 23.4±1.8cm/s, 25.5±2.3cm/s and 25.4±3.4cm/s at 2, 4, 8 and 12 week after operation, which result in a significant difference on the flow velocity of vena cava before and after filter implantation(p<0.001). The pressure gradient maintaining the blood velocity was 0.844±0.078 mm Hg at preoperation and was 0.186±0.036 mm Hg, 0.190±0.035 mm Hg, 0.197±0.034 mm Hg and 0.212±0.058 mm Hg at postoperation, which show a significant difference between pre- and postoperation(p<0.001).No change in contrast-team.4. In filter team,the postoperative dual source CT scan and multidimensional imaging reconstruction revealed that filter dilation at segement of filter implantation and open of the collateral vessels of lumbar vertebral venous plexus. At Week 2, the struts of all the 4 filter were restricted inside the lumen of IVC. At Week 4, a large part of one strut was piercing through IVC wall. At Week 8, 2 cases of struts were fully piercing through IVC wall while other 2 were largely piercing through; thickened vein wall could be also found at the site of filter piercing and a blood clot strip was found in one of cases. At Week 12, all the 4 struts of filter were located outside IVC lumen,eccentric thickening of vein wall and formation of thrombi stripe inside IVC. CTVE could found the sunken of lumen at the site of strut, significant thickening of endothelial vessel at adjacent struts, which result in crest bumps. Moreover, uneven and stripped floated thrombi could be observed at the filter implanted IVC, whereas patency lumen could be found in the normal vessel.No change in contrast-team.5. Intact adventitia could be found by eye and light microscopy at 2, 4, 8 and 12 week postoperation. In filter team,no filter was found to pierce through the IVC and into adjacent tissue and organs. The large part of filter strut was embeded into the vessel wall, covered with semi-transparent thin tissue, and could be easily separated between filter and vessel wall. The struts of the filter was fully covered into the proliferated endothelial at postoperative Week 4, 8 and 12, and thickened and uneven tissues were covered on the surface of the filter. We also observation the adhesion between filter hook and vessel wall, which could result in difficulties on filter peeling off at postoperative Week 4. The small amount of thrombi formed at Week 8 could lead to the severe difficulties on filter peeling, whereas struts and hooks of filters were all embeded into the vessel wall with the increasing of the attached thrombi, thickenedintimal coverage, uneven of IVC lumen at Week 12, which could result in inability on filter peeling. Significantly increased intimal was found on the surface of filter at Week8 and 12 than at Week 2 and 4. The light microscopy at Week 2 revealed that flatted endothelial cells at the site of filter implantation, thinned internal elastic lamina,fractures inside vessel wall and injury of endothelial cells at strut of filter with incosistent intimal and exposed collagen fibers. The intimal on the filter surface with significantly increased collagen tissue and focal collagen could be found, and the neointimal was composed of smooth muscle cells, collagen and a small amount of macrophages. A large amount of collage fiber proliferation and fractures between vessels could result in structure injury of vessel wall at the site of filter retrieval.No change in contrast-team.6. Scan eletronic microscopy revealed that stripped, uneven thickeness collagen fibers with a small amount of cells were cross linked on the surface of filter at Week 2.Enlarged and aggregated collagen fibers could be found on the surface of filter at Week4, and most of which were covered by collage fibers. Enlarged and thickened collage fibers with platelets adhesion were found on the surface of the filter at Week 8, which was composed of collagen fiber. Fully covered neointimal with increased and aggerated collagen fibers, thrombi formed by platelets and erythrocyte sedimentation was found on the surface of filter at Week 12. The coverage percentage of neointimal at Week 2, 4, 8 and 12 was(14.34±8.85)%、(29.52±12.38)%、(62.63±12.94)%、and(68.88±10.74)%, respectively. Significant differences could be found between postoperative Week 2, Week 4 and Week 8 and 12(P<0.001), while no significant difference was found between week 8 and week 12(P>0.05). The endothelial coverage percentage at postoperative 2, 4, 8 and 12 weeks was(42.57±8.20)%、(93.58±5.46)%,100% and 100%. We found a significant difference between Week 2 and Week 4, 8 and12(P<0.001), while no significant difference was found among Week 4, 8 and 12(P>0.05).Conclusions:1. Partial dilation of vessel with reduced blood flow could be found after filter implantation. The laminar flow could be disappeared and deviated axial flow when the filter tilt was larger than 15°, which could result in the formation of turbulence and vortex flow and increased shear force from the blood flow to vessel wall. The filtercould exert a blocking function on the vena cava flow and the flow velocity, flow state,deviation of axial flow and shear force of blood flow to vessel wall could be changed with increasing of filter tilt, which could lead to the secondary vein thrombosis and should be considered as risk factor for vein thrombosis.2. Imageological examination of the filter implanted animals suggested that filter implantation could result in dilation of vena cava, reduction of blood flow, and formation of lumbar collateral vessels, suggesting a blocking function to the lumen blood flow generated by the filter. The eye and light microscopy results proved that increased thickness of collagen fiber inside the vessel wall, generation of multiple fractures change, which could lead to the vessel injury and remodeling. With the increasing of filter implantation, the cross-linked and thickened collagen could be presented on the surface of filter, which could result in the platelet activation, adhesion of red cells and subsuquently thrombosis. A significant increased neointimal formation,fully coverage of strut and hook of filter by neointimal and secondary thrombosis was found at Week 8 after filter implantation, which could result in difficulties on filter retrieval and tearing the endothelial tissue. |
PDF Full Text Request