Article Text

Download PDFPDF

Inferior vena cava filters for the cerebrovascular patient
  1. Sanjeeva P Kalva1,2,
  2. Bhanusupriya Somarouthu1,2,
  3. Joshua Hirsch1,2
  1. 1Department of Imaging, Massachusetts General Hospital, Boston, Massachusetts, USA
  2. 2Department of Radiology, Harvard Medical School, Boston, Massachusetts, USA
  1. Correspondence to Dr S P Kalva, GRB-297, 55 Fruit Street, Boston, MA 02114, USA; skalva{at}


Acute cerebrovascular disease is often complicated by deep venous thrombosis and pulmonary embolism. Many of these patients are at high risk of intracranial hemorrhage with therapeutic anticoagulation. These patients may benefit from insertion of inferior vena cava filters. Studies specifically dealing with stroke patients are lacking, but it is the authors' opinion that filters reduce the incidence of pulmonary embolism. There is little evidence to support the use of these devices prophylactically in patients who do not have venous thromboembolism. Retrievable filters are an attractive option but there are concerns about their safety; and if regularly used, a system for successful filter retrieval in all patients should be instituted. The role of concurrent anticoagulation with filters is not clear. However, we believe anticoagulation, in the absence of a contraindication, is beneficial in patients with active venous thromboembolism.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Stroke remains the third leading cause of death in the USA.1 The stroke per se and subsequent medical complications affect the survival of stroke patients. Complications such as pulmonary embolism (PE), deep venous thrombosis (DVT), urinary tract infections and pneumonia following stroke indicate a poor quality of stroke care.2 Half of all patients admitted to hospital for an acute stroke develop DVT.3 4 The risk of DVT is even higher in patients with hemorrhagic stroke5 6 and in patients with brain malignancy presenting as stroke.7 The incidence of DVT during the first 2 weeks after stroke has been reported as 27–75% among untreated control patients in various randomized prophylaxis trials.8 The risk of recurrent DVT is also high in patients with limb paresis following stroke.7 More than a quarter of stroke patients who develop DVT subsequently develop PE.3 4 In addition, nearly half of clinically evident PE after stroke present as sudden death and many of these patients do not have clinical evidence of DVT.9 In a study that encompassed 19 years, the corrected rate of fatal PE among patients with ischemic stroke ranged from 1.5% to 2.1%.10

Prophylactic measures for prevention of venous thromboembolism (VTE) are now routinely employed in stroke patients. Physical methods such as thigh length graduated compression stockings and intermittent pneumatic compression devices are ineffective in reducing the risk of DVT after stroke.11 Unfractionated heparin is effective in decreasing the incidence of DVT but was associated with death and recurrent stroke in 0.5% of participants in the International Stroke Trial.4 The same study reported a 1.5% incidence of non-fatal extracranial bleeding with unfractionated heparin. A subsequent randomized study demonstrated 43% reduction in the risk of VTE with Enoxaparin compared with unfractionated heparin with no significant change in the occurrence of any bleeding (8% in either group) or symptomatic intracranial and major extracranial hemorrhage (1% in either group).12 A meta-analysis further suggested the higher benefit/risk ratio of low dose low molecular weight heparins (LMWH) for prevention of DVT and PE in patients with acute ischemic stroke13; however, the odds ratio of developing an intracranial hemorrhage was 1.39 with low dose LMWH, 1.67 with low dose unfractionated heparin, 2.1 with high dose LMWH and 3.86 with high dose unfractionated heparin. Antiplatelet agents have little role in the prevention of VTE.14 15 Despite the risks, DVT prophylaxis was associated with improvement of acute stroke outcomes.16

Despite effective prophylaxis with LMWH and unfractionated heparin, certain patients (0–36%) develop DVT following stroke.17 These patients require therapeutic doses of heparin or warfarin for treatment of DVT. Such therapeutic doses are often associated with an unacceptably high incidence of intracranial hemorrhage, especially in patients with hemorrhagic stroke and intracranial malignancy presenting as stroke. These patients may benefit from inferior vena cava interruption.

What are inferior vena cava filters?

Inferior vena cava (IVC) filters are mechanical devices that are designed to be placed in the IVC.18 They are made from various materials—stainless steel, titanium, Nitinol and Elgiloy. Their shapes vary; most of them are umbrella or cone shaped and a few are basket shaped or mesh-type. All of them have some sort of hooks to attach to the wall of the cava. Although the initial iteration of the filters were permanent devices, the new designs incorporate a hook or hub that allows an operator to retrieve the filter after its placement. The IVC filters protect against PE by trapping large emboli (but not all emboli) that originate in the deep veins of the pelvis and lower extremities.

Are inferior vena cava filters effective?

A randomized study comparing the effectiveness of a combination of anticoagulation and IVC filters versus anticoagulation alone in patients with proximal DVT demonstrated a significant decrease in the incidence of PE in the filter group (1.1% vs 4.8%).19 This benefit was observed even after 8 years.20 The reduction in the incidence of PE was not associated with any significant improvement in overall mortality; however, the population studied was an average geriatric and the majority of deaths were secondary to cancer and cardiovascular diseases.21 No other randomized study exists on the effectiveness of the IVC filters. Multiple retrospective studies demonstrated their effectiveness through reduction of mortality from PE. One large study that encompassed over 26 years and involved 1731 patients and nine different designs of permanent IVC filters demonstrated a 5.6% incidence of post-filter PE and 3.7% of post-filter fatal PE.22 The reported incidence of post-filter PE ranged from 2% to 5% in various retrospective studies.18 23 Current designs of retrievable filters (also called ‘optional’ filters, as these are permanent devices with an option to retrieve) demonstrate similar efficacy.24

What are the risks associated with inferior vena cava filters?

The Prevention du Risque d'Embolie Pulmonaire par Interruption Cave (PREPIC) study demonstrated a significant increase in the incidence of DVT in patients who received anticoagulation and an IVC filter for proximal DVT compared with those who received anticoagulation alone (20.8% vs 11.6% at 2 years and 35.7% vs 27.5% at 8 years)19 20; however, the incidence of post-phlebitic syndrome was similar in both groups at 8 years.20 Other retrospective studies also confirmed the increased incidence of DVT and vena caval occlusion with the use of IVC filters.18 22 23 The post-filter DVT occurred at a frequency of 6–32% and caval occlusion 3.6–11.2% in various studies.23 25 Other complications specifically associated with IVC filters include migration of the filter or of fractured filter components, penetration of the filter legs into other organs, venous access site thrombosis and operator errors during placement.18 22 23

What are the accepted indications for inferior vena cava filters?

IVC filters are indicated in patients who have active VTE and cannot receive anticoagulation for any reason18 25—for example, in the presence of a contraindication to anticoagulant therapy such as active gastrointestinal bleeding, intracranial hemorrhage, following craniospinal and orthopedic surgeries, and in the occurrence of a complication following standard dose anticoagulant therapy such as development of retroperitoneal or rectus sheath hematoma following heparin or warfarin therapy. Filters are also indicated in patients who develop recurrent thromboembolism despite therapeutically adequate anticoagulation.

Additional extended or soft indications for IVC filters, that are not well supported by evidence, exist. Filters are often suggested for patients who have a high risk of cardiopulmonary collapse (such as in patients with pre-existing pulmonary hypertension or severe chronic obstructive pulmonary disease) in the presence of a massive PE, despite being therapeutically anticoagulated. Filters are also suggested in patients who have a free floating thrombus in the iliocaval veins. Filters are now increasingly used prophylactically in patients who are at a high risk of VTE (patients with intracranial trauma, hemorrhagic stroke, post-hip arthroplasty, femoral neck fracture) and who cannot receive regular pharmacological thromboprophylaxis. In addition, filters are increasingly used in older patients who are at risk of fall and may not receive adequate anticoagulation and in patients who are potentially unreliable to comply with medication regimens.

What is the role of filters in the cerebrovascular patient?

Cerebrovascular patients (patients with acute ischemic or hemorrhagic stroke and patients with intracranial tumors presenting as stroke) in the presence of acute DVT or PE who cannot receive therapeutic anticoagulation benefit from IVC filters. Given the high incidence of PE in patients with DVT following stroke, filters reduce the incidence of PE. However, there are no data to indicate improved survival with the use of filters in this population. The use of IVC filters as a prophylactic measure in the absence of VTE is unjustified.

Similarly, stroke patients during rehabilitation benefit from filters if they cannot receive adequate anticoagulation. There is little role for prophylactic filter placement in this group.

How effective are inferior vena cava filters in stroke patients?

No long term studies, either prospective or retrospective, exist on the efficacy of IVC filters in stroke patients. Multiple retrospective studies on the efficacy of filters included stroke patients but a separate analysis on the efficacy of filters is lacking. Athanasoulis et al included 22 patients with neurological disease in their study of 1731 patients22 and reported 5.6% of post-filter PE in the entire group. Kalva et al included 112 stroke patients in their study of 751 patients and reported 7.5% incidence of symptomatic post-filter PE in the entire group with the use of the TrapEase (Cordis Endovascular, New Jersey, USA) filter.26 In a recent review of our hospital data (Somarouthu et al, unpublished; data reported at the Massachusetts General Hospital clinical research day, May 2010), there was a 15% incidence of symptomatic post-filter PE in 371 stroke patients although it was fatal in only 0.8%. Of the15% who developed symptoms of PE, only 4% had PE on imaging. The same authors reported the use of nine designs of IVC filters in their study. This suggests a comparable efficacy of IVC filters in this group compared with the available large retrospective studies that included various disease groups and various designs of IVC filters.

How safe are inferior vena cava filters in stroke patients?

Caval occlusion and DVT remain the most common complications associated with IVC filters in various studies.18 25 No study elucidated the specific risks of IVC filters in stroke patients but these two remain the most concerning, given the prothrombogenic state of stroke patients. Review of our hospital data (Somarouthu et al, unpublished) revealed a 16% incidence of symptomatic post-filter DVT and 5.1% incidence of symptomatic and asymptomatic caval occlusion. These findings are similar to previously reported data on complications of the IVC filters, suggesting similar safety of IVC filters in stroke patients. In addition, filter fracture is a real concern in patients with quadreparesis as the filter may be compressed during physical therapy maneuvers.

Are all filters similar in terms of safety and efficacy?

Prospective comparison studies on various designs of IVC filters are lacking. One prospective randomized study compared the Greenfield filter (Boston Scientific, Natick, Massachusetts, USA) and TrapEase filter (Cordis Endovascular). This study demonstrated similar efficacy of these filters but a higher rate of symptomatic iliac or iliocaval thrombosis with the TrapEase filter (6.9% vs 0%).27 However, another larger, retrospective study on the TrapEase filter reported a 0.7% incidence of caval occlusion.26 A high rate of caval occlusion was observed with Mobin-Uddin filter and Simon Nitinol filter (Cook Medical, Bloomington, Illinois, USA) in one large retrospective study.22 The variations in the rate of DVT and caval occlusion appear to be related to the inherent differences in the patient population studied. Various factors, such as the presence and type of the underlying hypercoagulable state, pre-existing DVT and anticoagulation after filter placement affect the rate of post-filter caval thrombosis.

What role do retrievable filters have in stroke patients?

Retrievable (or option) filters are approved for permanent use and subsequent retrieval if desired. The retrieval window—the time between the filter placement and retrieval—is variable among the filters: 23 days for OptEase filter (Cordis Endovascular); undefined for Gunther Tulip (Cook Medical), Celect (Cook Medical), G2/G2 Express/Eclipse (Bard Peripheral, Murray Hill, New Jersey, USA) and Option (Rex Medical, Gainesville, Florida, USA) filters. The package inserts in these filters do not define the retrieval window but it is generally accepted that the success of retrieval of these filters is limited beyond 1 year of their placement. The retrieval option makes these devices highly attractive as the long term filter complications can be avoided while providing adequate protection against PE in patients who cannot receive anticoagulation during the acute phase of stroke, provided these patients are rigorously followed and filters successfully retrieved. This remained the basis for increasing use of these devices, especially for prophylaxis against PE in high risk patients. However, in many studies, only 15–30% of the retrievable filters are actually retrieved.24 In addition, the long term safety of the retrievable devices is still not known given the seemingly high incidence of filter fractures and penetration leading to organ injuries.28 Other factors that need consideration include the limited survival of the stroke patients. As high of 50% of patients who require an IVC filter for VTE following stroke die within 2 years (Somarouthu et al, unpublished data). Given the risks associated with retrievable filters and limited survival of stroke patients, a cautious approach to integration of retrievable filters to clinical practice is warranted. A dedicated follow-up and serious attempts to remove these filters make them highly desirable in stroke patients.

How should we treat filter thrombosis?

Symptomatic filter thrombosis requires therapy. Acute symptomatic filter thrombosis is treated with anticoagulation with or without local pharmacomechanical thrombolysis. However, these options may be limited in stroke patients given the risk of intracranial hemorrhage. Alternative approaches include pure mechanical thrombolysis with heparin infusion. Compression stockings reduce limb edema. Treatment of asymptomatic filter thrombosis is unclear. A recent retrospective study demonstrated no significant difference in the progression or regression rates of filter thrombus whether or not anticoagulation is instituted.29 In patients who are at risk of complications, a therapeutic anticoagulation may be avoided in the presence of asymptomatic filter thrombosis. Alternatively, if no contraindication exists, anticoagulation may be instituted in patients with asymptomatic filter thrombosis.

Do all patients with an inferior vena cava filter need anticoagulation?

No data exist to support the use of anticoagulants in all patients with an IVC filter in the absence of VTE. However, given the high incidence of DVT in patients with an IVC filter, many physicians elect to ‘treat an IVC filter’ with anticoagulants. Our recommendation is to treat patients who have active VTE with anticoagulation for at least 3 months. A recent review suggested a higher incidence of post-thrombotic syndrome and venous ulcers with IVC filters although this was not observed in the PREPIC study.20 30 However, the role of anticoagulation in the prevention of post-thrombotic syndrome in patients with an IVC filter is not clear.


IVC filters provide clinical benefit by reducing the incidence of pulmonary embolism in cerebrovascular patients who otherwise cannot receive standard therapeutic anticoagulation. The effect of IVC filters on improving survival is not known. Prophylactic use of these devices should be discouraged. Retrievable filters appear to be an attractive option but limited data exist on their safety. A vigorous program to retrieve these filters must be instituted when the devices are used regularly. No device is superior to others in terms of efficacy and all filters are associated with an increased incidence of DVT and caval occlusion.



  • Competing interests None.

  • Provenance and peer review Not commissioned; not externally peer reviewed.