Over-the-Wire Inferior Vena Cava Filter Placement: How We Do It

Inferior vena cava (IVC) filter placement is routinely performed in selected clinical settings to protect patients from thromboembolism to the lungs. The absolute indications for IVC filter placement include patients who have failed or have absolute contraindications to anticoagulation in the setting of acute venous thromboembolic disease (VTE), such as pulmonary embolism (PE) and deep vein thrombosis (DVT). The relative indications are variable based on society guidelines and include patients who are candidates for anticoagulation but have large free-floating proximal DVT or poor cardiopulmonary reserve. Along with the expansion of relative indications, the advent of retrievable IVC filter has also led to a significant growth in the number of filter placements in previous years. Overall from 1993 to 2010, the number of annual filter placement increased from 28,000 to 130,000, representing a 358% increase. ¹ In the subsequent years, the filter utilization has decreased due to a combination of Food and Drug Administration (FDA) advisory and legal effort, although the number of IVC filter placements remains high. ¹ Along with the high filter placement, there has been a growing number of filter complications, including filter fracture, migration, tilting, and long-term complications such as postthrombotic syndrome. ^{2 3} Among those, filter tilt often occurs during the immediate periprocedure timeframe and is associated with a decrease in filtering efficiency and an increase in retrieval difficulties. Many methods have been utilized to minimize filter tilt and one possible option is the use of the over-the-wire (OTW) technique, during which the OptionElite filter (Argon Medical Devices, Plano, TX) is deployed over a 0.035-inch guidewire. Similar to an OTW stent placement, the guidewire facilitates centering the filter within the IVC, thereby preventing filter tilt. Based on our experience, placing IVC filter in OTW fashion has certain unique benefits over the standard technique, specifically in cases of a tortuous IVC course. In this article, the technique of OTW IVC filter placement is described and depicted in detail along with a review of the technique in the context of the literature.

Technique

The IVC, bilateral renal veins, and access site anatomy should be evaluated on cross-sectional imaging. If there is no available cross-sectional imaging, IVC venogram during the procedure is usually sufficient to review the relevant anatomy. Under ultrasound guidance, venous access is obtained. A short, 5-Fr vascular sheath is then introduced into the venous access site. Either the internal jugular or the common femoral vein can be used as access sites. However, in our practice, the jugular access is preferred as it offers a straighter route to the IVC. Furthermore, the internal jugular approach avoids the potential, more proximal DVTs in the iliac veins that may be missed with ultrasound. After venous access, a 0.035-inch guidewire is negotiated into the IVC followed by placement of a 5-Fr pigtail catheter and IVC venogram ( Fig. 1 ). The venogram is obtained to assess the venous anatomy (anatomical variants, such as left retroaortic renal vein, and duplicated IVC) and to take note of any intraluminal stenosis or IVC thrombus present. It is also important to assess the infrarenal IVC size, as the OptionElite filter (Argon Medical Devices, Plano) has a maximal deployable caval diameter of 30 mm. It is important to be aware that the IVC is a dynamic vessel and its shape is subjected to respiratory motion. If the diameter is above the 30-mm threshold in the anteroposterior projection, a lateral view can be helpful to assess whether the IVC is collapsed at the time of anteroposterior measurement or if the diameter is truly above 30 mm. Once the IVC venogram demonstrates a suitable IVC for filter placement, a 260-cm 0.035-inch Bentson wire (Cook Medical, Bloomington, IN) is advanced through the pigtail catheter into one of the iliac venous limbs. It is helpful to maintain the table position once the IVC venogram has been performed to ensure precise infrarenal IVC filter deployment, although some practitioners prefer using the anatomic landmarks. It is important to note the tip of the guidewire for filter placement will be outside of the field of view. Therefore, for safety purposes, it is preferred to use a Bentson wire which has a long floppy tip as opposed to a stiff wire such as an Amplatz guidewire (Boston Scientific, Marlborough, MA). Then, the vascular sheath is exchanged for the 6-Fr delivery system consisted of a 6-Fr sheath and an inner dilator. The delivery system is advanced until the marker tip of the delivery sheath is at the desired location: in the jugular approach, the location of the filter legs is at the more distal infrarenal IVC, and in case of the femoral venous approach the location of the filter hook is just below the renal veins ( Fig. 2 ). Once the delivery system is in place, the dilator is removed and will be used in later steps. Then, the filter cartridge is loaded over the Bentson wire (thread the wire carefully through the center hole of the filter) and care should be taken at this point to not move the Bentson wire or change the location of the outer delivery sheath position. It is also important to check the orientation of the filter cartridge to make sure the filter orientation is correct as the OptionElite comes in one model for both femoral and jugular placements. At this point, the filter cartridge is snapped into the outer delivery sheath and typically a click is heard. To deploy filter, back tension is needed on the anchoring Bentson wire. Care should be taken to not move the wire, otherwise a loop may form and the filter may get stuck. Then, the inner dilator from the delivery system is advanced over the wire and is used to push the filter while maintaining good back tension on the Bentson wire. Once the delivery marker of the inner dilator aligns with the end of the filter cartridge, the filter has been pushed to the tip of the delivery sheath. In one smooth motion, while maintaining the position of the inner dilator, the outer delivery sheath is pulled back over the 0.035-inch Bentson guidewire, thus unsheathing the filter ( Fig. 3 ). Under fluoroscopy guidance, the Bentson guidewire is retracted carefully in its entirety. Very rarely, the guidewire may be stuck within the filter, and in these cases a gentle spinning motion of the wire under fluoroscopy should be pursued to dislodge the wire from areas of resistance ( Figs. 4 and 5 ). Postdeployment, a fluoroscopic image is typically obtained to assess the filter position. In our practice, we do not routinely perform a postdeployment venogram and in our experience, the described OTW technique often results in minimal filter tilt. In selected circumstances, such as very tortuous IVC, a postdeployment venogram is performed through the 6-Fr delivery sheath with the tip of the sheath being positioned just above the filter hook. In these cases, we do a rapid injection with a 10-mL syringe filled with 5 mL of contrast and 5 mL of saline. It is important to remember to avoid pushing the delivery sheath beyond the newly placed IVC filter (even with a guidewire) as such maneuver can dislodge and possibly tilt the freshly placed filter. Finally, the delivery sheath is removed and hemostasis is obtained with manual compression at the venous access site.

Fig. 1.

A 73-year-old woman with past medical history of adenocarcinoma of the esophagus who presented with right middle cerebral artery stroke and left lower extremity proximal deep vein thrombosis. Because of the recent acute ischemic stroke, anticoagulation was contraindicated and an inferior vena cava (IVC) filter placement was requested. ( a ) Venogram obtained with a power injector showed the infrarenal IVC measuring 19.5 mm and did not demonstrate any intraluminal thrombosis or stenosis. ( b ) The 0.035-inch 260-cm Bentson wire looped within the left common iliac vein (white open arrow). The filter was pushed into “tip-to-tip” position with the delivery sheath using the inner dilator (note that the legs of the filters were at the level of the inferior-most radiopaque tip marker of the sheath—solid white arrow). The top two radiopaque markers (open black arrows) indicated the location of the pusher. ( c ) Completely deployed filter showed good symmetrical expansion of the filter legs and no tilting. The filter hook was in infrarenal position at the lower L2 level.

Fig. 2.

A 62-year-old man with a history of esophageal adenocarcinoma, and proximal lower extremity deep vein thrombosis (DVT) complicated by pulmonary embolism (PE). Patient was scheduled for surgery and therefore anticoagulation had to be discontinued. Inferior vena cava (IVC) filter placement was requested given the increased risk of pulmonary thromboembolism in the perioperative period in an immobilized patient with a history of DVT and PE. ( a ) Pigtail catheter cavogram demonstrated patent bilateral renal veins. The infrarenal IVC measured 24 mm and was free of any intraluminal thrombus, stenosis, or web formation. ( b ) Advancement of the 6-Fr delivery sheath over the 0.035-inch 260-cm Bentson wire, along with the inner dilator (note all three markers are present—arrows). The Bentson wire was anchored in the left external iliac vein. The most superior radiopaque marker (white arrow) indicated the location of the legs of the filter upon deployment. ( c ) The filter was in “tip-to-tip” position with the lowest radiopaque marker of the delivery sheath, immediately prior to unsheathing. ( d ) The IVC filter was in satisfactory position, free of tilting and incomplete opening. ( e, f ) Care should be taken to not advance the sheath tip beyond the filter (even if a guidewire is still in place), as this could potentially dislodge the recently placed filter. To confirm the intraluminal position, a 10-mL bolus of contrast mixed with saline was given through the side arm of the sheath. The image showed the contrast column surrounding the filter, thus confirming that the filter was centered within the IVC without tilting. Solid redlines outline the contour of the IVC and dashed redlines mark the center of the IVC filter.

Fig. 3.

A 57-year-old man with proximal lower extremity deep vein thrombosis (DVT) complicated by pulmonary embolism. While on anticoagulation, the patient developed recurrent DVT. Due to failure of anticoagulation, he was referred for filter placement. ( a ) Cavogram showed the infrarenal inferior vena cava (IVC) measuring 23 mm without evidence of intraluminal thrombus, stenosis, or malformation. ( b ) Image taken immediately after unsheathing the filter over the guidewire. The radiopaque marker (arrow) was located immediately above the filter hook indicating that the unsheathing process has been completed. Note that the Bentson wire helped straighten and center the filter within the IVC. The Bentson wire was then carefully removed without disturbing the filter position. ( c ) Postdeployment IVC filter demonstrated satisfactory intraluminal position without significant tilting and complete opening.

Fig. 4.

A 62-year-old man with a history of proximal deep vein thrombosis and pulmonary embolism. Anticoagulation was contraindicated due to the risk of intracranial bleeding in the setting of brain metastasis from thyroid cancer. ( a ) Cavogram showed the infrarenal inferior vena cava (IVC) measured 23 mm and was free of intraluminal thrombus or stenosis. ( b ) The IVC filter was in “tip-to-tip” position with the delivery sheath. The inner dilator was seen immediately superior to the filter. ( c ) Immediately post unsheathing, the filter was maintained in appropriate intraluminal position by the long 0.035-inch Bentson guidewire. The wire was then carefully removed under fluoroscopy. ( d ) Postdeployment, the filter hook was in infrarenal position at the lower L2 level.

Fig. 5.

A 52-year-old woman with a relevant history of proximal acute deep vein thrombosis and pulmonary embolism. The patient was initially placed on a heparin drip which had to be stopped due to a significant hemoglobin drop. An inferior vena cava (IVC) filter placement was requested. ( a ) Cavogram with infrarenal IVC measuring 23 mm and an IVC that was free of intraluminal thrombus, stenosis, or malformation. ( b ) The sheath was advanced along with the inner stiffener over the 260-cm 0.035-inch Bentson wire. The thick indicator band (arrow) represented the eventual location of the filter legs (when placing the filter via the internal jugular approach) during final deployment. ( c ) The sheath was maintained with tip in the infrarenal position after the inner stiffener has been removed. ( d ) The filter (white arrows) was advanced over the wire using the inner stiffener until the legs of the filer reached the thick indicator (“tip-to-tip” position). The upper two thinner radiopaque markers (black arrows) were from the inner stiffener. ( e ) In one smooth motion, the sheath was pulled back to deploy the filter over the Bentson wire. ( f ) The filter was shown to be in appropriate position centered within the IVC. Note the sheath (arrow) was situated within the IVC just above the filter.

Discussion

Typically, IVC filter is placed through a sheath without a guidewire. The IVC filter is advanced through the sheath with a pusher and is deployed with a variety of different mechanisms depending on the type of filter placed. In comparison, with OTW deployment, the sheath is stabilized longitudinally via the guidewire that is anchored in the proximal IVC when using a femoral access approach and in the iliac venous system when using a jugular venous access approach. Theoretically, the anchoring wire provides the longitudinal stability that minimizes the filter tilt.

Filter tilt is defined as an angulation more than 15 degrees from the filter's long axis relative to the IVC, usually measured in the lateral axis on the anteroposterior projection. In the literature, tilt more than 15 degrees is sometimes referred to as a severe tilt. ⁴ On the other hand, other authors contended that filter tilt should be defined by the ratio of the distance between filter tip and the vessel wall to the vessel radius. ⁵ A review of literature has shown that the filter tilt rates for various retrievable filters were between 10 and 20%. ⁶ Prior generations of IVC filters had a higher tilt rate. For example, one study has shown that 41% of the titanium Greenfield filters (Boston Scientific) had a tilt angle greater than 15 degrees, although the finding was limited by the small sample size and the long indwelling time. ⁷

The clinical significance of a tilted filter is multifold. For one, it decreases the efficiency of thrombus filtration. In vitro studies have shown that severe tilt (>15 degrees) can decrease trapping efficiency of small thrombus (3 × 5 mm and 3 × 10 mm), while maintaining efficient trapping of large thrombus (7 × 10 mm and 10 × 24 mm). Furthermore, there is a significant decrease in the overall thrombus capture rate in tilted filters versus concentrically located filters. ⁸ Singer and Wang explored the hemodynamic changes in tilted filters. Using a computational model, the authors found the tilted filters led to a stagnation zone near the vessel wall, which could become a nidus for in situ thrombosis. ⁹ The decrease in filter efficacy and altered hemodynamics may promote an increase in recurrent PE. Indeed, a review of literature has shown some evidence of this theoretical association. Greenfield et al have shown that 8.5% of the patients with tilted filters suffered recurrent PE compared with 3.3% of the patients with well-positioned filters. ¹⁰ Rogers et al have found three cases of recurrent PE in the tilted filter group and none in the properly positioned filter group. ¹¹ It should be noted that in these studies, sample sizes were too small to power any meaningful statistical analysis and proper studies are required to draw any definitive conclusion. Finally, tilted filters are predictive of challenging retrievals. Dinglasan et al have shown that 90% of the patients with filter-retrieval failure had titled filters with a mean mediolateral tilt angle of 19 degrees. ¹² Similarly, Avgerinos et al have found that severe tilt was predictive of challenging filter retrieval (odds ratio: 2.607, 95% confidence interval: 1.045–6.508). ¹³

Given the significant clinical sequelae of a tilted filter, efforts have been made to minimize the risk of filter tilting. These include design modifications and post placement adjustment techniques. When first introduced in 1972, the Greenfield filter (Boston Scientific) was initially placed OTW. There was a hollow opening at the apex of the Greenfield filter that permitted guidewire passage. Over the guidewire, the filter is manipulated into the ideal location and deployed, which in theory properly centers the filter. However, there were multiple other placement issues plaguing the device, mostly due to its 24-Fr introducer sheath, which increased the risk of access-site thrombosis. ¹⁴

The percutaneous model of the Greenfield filter was modified from the original design and was made out of titanium instead. Due to the more flexible and stronger material, the introducer sheath was downsized to 12 Fr. However, as previously mentioned, the titanium Greenfield filter had a high tilt rate as well as a significant rate of asymmetrical opening, numbering as high as 71%. ¹⁵ Then, the OTW stainless steel Greenfield filter was introduced to facilitate the proper alignment of the filter in the IVC. Kinney et al compared the OTW Greenfield to the titanium Greenfield filters. The authors retrospectively reviewed 141 titanium Greenfield and 104 stainless steel OTW Greenfield placements. Right common femoral, left common femoral, and right internal jugular approaches were compared as well. In this study, the authors found that the OTW configuration did not improve the tilting angle in all placement routes (4.1 vs. 5.5, −2 vs. −5.1, 0.4 vs. 1.1, p  = 0.06, 0.12, and 0.5 for right common femoral, left common femoral, and right internal jugular vein approaches, respectively). ¹⁶ This finding seems to be congruent with other studies. Johnson et al found 55% of the OTW Greenfield filters demonstrated filter tilt at the time of placement. Severe tilt in the study was defined as more than 15-degree angulation from the vertical center of the IVC. ¹⁷

One possible reason that OTW technique failed to improve Greenfield filter tilt may have been that the guidewire was not advanced more distally into the iliac venous system, thereby forcing filter deployment over the floppy tip. The original guidewire had a floppy, 10-cm tip that might be advantageous in traversing challenging anatomy but does not offer any structural support in anchoring the filter. ¹⁸ Furthermore, the floppy guidewire tip has been reported to tangle between the filter struts and lead to guidewire entrapment that necessitated open surgical retrieval. ¹⁹ Although the guidewire floppy tip has since been shortened to 1.5 cm, published reports described continued guidewire entrapment in the filter. ¹⁸ Additionally, the Greenfield filter may have been too bulky to enable the guidewire to provide sufficient longitudinal stability. Therefore, given that OTW technique did not seem to confer any additional benefits in centering the Greenfield IVC filter, and the associated complications, OTW techniques for filter placement lost popularity in subsequent years.

In the meantime, other design features have been introduced into the filter design. For example, the Celect filter (Cook Medical) is designed with four long struts and eight short anchoring struts in the hope to improve the self-centering capability of the filter. On the other hand, the G2 filter (Bard Peripheral Vascular, Tempe, AZ) also utilized shorter anchoring hooks and long centering struts to reduce the risk of filter tilting. However, in clinical practice, the theoretical benefit did not lead to significantly decreased filter tilt rates. Shelgikar et al studied 302 filter placements and compared Celect filters with a centering design to Günther Tulip filters (Cook Medical) that did not include the self-centering design. The authors have found no significant difference in filter tilt between the two groups ( p  = 0.34). However, the filter tilt (defined as tilt > 15 degrees) rates of Celect and Günther Tulip filters were 4.6 and 2.6%, respectively, which were significantly lower than that of the Greenfield filters. Furthermore, on secondary analysis, at the time of retrieval, Celect filters demonstrated better self-centering capability compared with the Günther Tulip filter (62 vs. 42%, p  = 0.03), which indicated that the design modification may have reduced filter tilting in the long term. ²⁰ Similarly, the Option filter (Argon Medical Devices, Frisco, TX) has demonstrated improved filter tilt in comparison to the previous generation of IVC filters. Tsui et al studied 516 Option filter placements in a single institution. The authors have found the severe tilt (tile angle > 15 degree) occurred in 0.6% of the cases, while 5% of the cases demonstrated 10 to 15 degrees of filter tilt. ²¹ In a different study, Johnson et al prospectively evaluated the safety and effectiveness of Option filter placements in 100 patients. No immediate postplacement filter tilt was reported. At the time of retrieval (mean retrieval time: 67.1 days with a range of 1–175 days), 7.7% of the patients had significant filter tilt (> 15 degrees). ²²

The OptionElite filter is a modification of the Option filter. It incorporated a modified anchoring hook for better stability and a hollow tip where a guidewire can traverse through. The filter is deployed through a 6-Fr sheath which is relatively small compared with delivery systems of other filter designs. Because of the hollow tip, OTW placement technique became again feasible. In clinical practice, the OptionElite filter is associated with a low tilt rate. Hightower et al studied 72 OptionElite filter placements and followed up the patient postprocedurally for 2 years. No significant filter tilt was encountered (tilt angle > 15 degrees). ²⁴ In the meantime, Kim et al conducted a retrospective study comparing OTW OptionElite to non-OTW filter placement. A total of 78 filters were placed with 39 being OTW OptionElite and the rest being Option, Celect, Tulip, or Denali (Bard) filters. An exchange-length Bentson wire was used as deployment guidewire. The authors used the ratio of filter tip to midline distance over the caval radius as the measurement of filter tilt. There was a decrease in the filter tilt ratio (9.8 vs. 19.9%) in the OTW OptionElite group in comparison to the non-OTW group. Additionally, there was no significant periplacement complications associated with the OTW technique. ⁵ On the other hand, Park et al performed an interesting in vitro study comparing the efficacy of different OTW techniques in a phantom model. The authors compared bent stiff wire, original guide wire, and hydrophilic stiff wire in centering the filter. The authors have found the bent stiff wire had the lowest filter tilt ratio (0.49 vs. 0.78 and 0.67 for original push wire and stiff hydrophilic wire, respectively, p  = 0.019) in the femoral access approach. In addition, the original guide wire had the lowest filter tilt ratio (0.34 vs. 0.57 and 0.58 for stiff hydrophilic wire and bent stiff wire, respectively, p  = 0.045) in the jugular access approach. ²⁵ Note the filter tilt ratio was more significant in this study in comparison to the study of Kim et al, ⁵ which could be secondary to the elasticity of the phantom model and the lack of physiologic variables such as respiratory motion and cardiac cycles. In our practice, we believe that the OTW technique can improve the filter stability and reduce the filter tilt, particularly in a tortuous IVC. More importantly, since filter tilt is predictive of difficult filter retrieval, the improvement in filter tilt may facilitate retrieval. Furthermore, a properly positioned filter can spare patients the need for advanced retrieval technique, which may necessitate dual, large bore access and increased fluoroscopy time.

Other than device modification, other authors have advocated for device adjustment at the time of or shortly after the deployment. Hastings et al described a case where a permanent OTW Greenfield filter was resheathed with the use of a balloon and a sheath. The filter was partially collapsed and partially collected into the sheath. It was then relocated and recentered to the desirable location and redeployed. ²⁶ On the other hand, Laroia et al described a method to correct filter tilt at the time of placement. If the postplacement filter had a tilt angle more than 15 degrees, the authors advocated for using a SOS catheter (AngioDynamics, Latham, NY) to push the newly deployed filter into a more upright position. In the small case series, 96.4% of the patients with initial severe filter tilt had successful correction. ²⁷ However, it needs to be considered that SOS catheter manipulation in a freshly placed filter can markedly change the final placement level, thereby moving the filter away from the ideal immediate infrarenal IVC location.

Conclusion

IVC filters are placed for a variety of indications and the major clinical utility is to prevent fatal (recurrent) PE. With the widespread use and placement of IVC filters, there has also been an increase in filter complications. Filter tilt is a common complication that is associated with a decrease in filter efficacy and an increase in difficulty during retrieval. Multiple design improvements and technical modifications have been incorporated into practice to decrease the tilting risk. These include self-centering struts, anchoring struts, filter manipulation at the time of placement, and OTW technique. A review of literature has shown some evidence of the efficacy of the OTW technique, particularly with newer filter designs delivered through small sheaths. We routinely place IVC filter in the OTW fashion particularly in a tortuous IVC. The technique has the potential to decrease filter tilt and thereby facilitate future retrieval. However, further research is warranted to validate the efficacy of the OTW method.