• pregnancy;
  • temporary IVC filters;
  • venous thromboembolism

Anticoagulation remains the standard therapy in cases of major venous thromboembolic disease [1,2]. When adequate anticoagulation fails or is contraindicated, interruption of the inferior vena cava (IVC) is indicated to prevent life-threatening pulmonary embolism (PE). Although surgical plication is an option, percutaneous placement of an IVC filter is now the most common practice [1,3]. These filters have generally been found to be safe; however, delayed complications such as caval thrombosis, caval penetration, filter fracture and filter migration are recognized causes of morbidity and mortality [3].

In the context of peri-partum thromboembolic disease, the ability to remove an implanted metallic device from a young woman in order to avoid the potential for complications many years later is highly desirable. Various designs of temporary IVC filters have been studied. We report the use of two specific filters – one currently approved device (Gunther Tulip; Cook, Canada, Stouffville, ON, Canada) and another novel filter available on a compassionate use basis (Recovery IVC Filter; Bard Canada, Mississauga, ON, Canada) – for the peri-partum prevention of PE. A healthy 30-year-old primigravida presented at 35 weeks gestational age to a community hospital with left leg swelling. Antenatal care and previous medical history had been unremarkable. Color Doppler ultrasound confirmed a 4 cm length occlusive thrombus in the left femoral vein. Dalteparin [100 international units (IU) kg−1 s.c. b.i.d.] was administered. Repeat Doppler studies at 38 and 40 weeks gestational age showed minimal recannulization of the thrombosis. The dual concern of labor and delivery causing clot detachment and PE, and the risk of continued anticoagulation prior to delivery prompted consultation with Interventional Radiology.

A Gunther Tulip temporary IVC filter was inserted to an infra-renal location without complication via the right internal jugular vein at 40 + 2 weeks gestation. Anticoagulation with unfractionated heparin was continued until the patient entered spontaneous labor. She delivered a healthy baby girl the following day. Coumadin therapy was instituted on postpartum day 2 with continued heparin until the INR was >2.0.

The filter was removed on postpartum day 9 (10 days following insertion) without complication. Pre- and post-filter venous cavograms did not show any evidence of intraluminal thrombus. Anticoagulation was discontinued after 3 months. A follow-up investigation for inherited or acquired thrombophilic states did not reveal any abnormalities (including tests for abnormalities in protein S, protein C, anti-thrombin III levels; factor V Leiden mutation; the G20210A mutation in the prothrombin gene; and the methylene tetrahydrofolate reductase mutation; lupus anti-coagulant screening, and anticardiolipin IgG antibody levels).

A healthy 29-year-old, gravida 2, para 1 presented with a painful swollen left leg at 32 weeks gestational age to a community hospital. Color Doppler ultrasound revealed an occlusive iliac venous thrombosis. Low molecular weight (LMW) heparin (Dalteparin 200 IU kg−1 s.c. o.d.) was commenced. At 36 weeks, a repeat color Doppler study revealed distal extension of the thrombus, with an extensive common femoral occlusive thrombus present despite ongoing anticoagulation.

The patient was transferred to our institution at 37 weeks. Ultrasound revealed a breech presentation and a Caesarean section was planned. In view of the high risk of PE at the time of Caesarean section, a Recovery IVC Filter was inserted into the IVC to an infrarenal location using a right femoral vein approach. One week later the fetus had spontaneously converted to a vertex presentation. The patient was therefore induced at 38 weeks with temporary cessation of LMW heparin treatment. She had a normal vaginal birth of a healthy male infant.

Abdominal radiography postpartum showed a 1.5 cm caudal migration of the filter. Four weeks following delivery a routine follow-up abdominal radiograph at the community hospital demonstrated one of the filter arms to be projecting above the filter, suggesting a fracture. When the patient was seen at our institution for the pre-removal CT scan, the filter fracture was confirmed. No trapped thrombus was seen. Elective filter removal was subsequently performed 75 days post-insertion. At the time of filter removal, the broken arm was retrieved; however, one of the leg hooks had broken off. CT imaging on several occasions has failed to identify this tiny fragment, and the patient has remained asymptomatic. A postpartum complete thrombophilia investigation revealed no abnormalities. The woman continued postpartum on Coumadin and has remained asymptomatic on subsequent follow-up visits.

IVC filters are considered standard management in patients when anticoagulation is contraindicated or unsuccessful in the setting of venous thromboembolic disease. However, unique issues arise in pregnant women. First, these patients are typically young and healthy with long life expectancies. Thus, the need for a temporary device is essential because permanent filters have important long-term complications. In a review by Streiff [3] insertion of a Greenfield (permanent) filter in non-pregnant patients was associated with insertion site thrombosis (23% of patients), IVC thrombosis (3.6%), penetration into the IVC wall (4.4%), migration (5%) and development of post-phlebitic syndrome (19%). Furthermore, permanent metallic IVC filters preclude the future use of magnetic resonance (MR) imaging.

Studies have suggested that there is no survival advantage provided by continued IVC filter placement once the acute risk for venous thromboembolic disease has passed and the clot has dissolved. This observation, together with the known risks of long-term use of permanent IVC filters, has stimulated research into the development of temporary IVC filters. Because of the transient contraindication to anticoagulation at the time of delivery, temporary devices are of particular interest in obstetrics.

The Gunther Tulip filter was the first retrievable filter to be approved in Canada. This filter was found to be an effective device in the Canadian Registry but a 10-day limit on its insertion precludes its use in many clinical situations, other than peri-partum protection against PE [4]. The current manufacturer's instructions to remove the filter within 10–14 days of placement is based on studies that demonstrate endothelial hyperplasia, with subsequent failure to remove the device at 16 days post-insertion [5].

The use of a temporary IVC filter specifically during late pregnancy has been previously documented in the European literature. Owen and Krarup [6] described the insertion of a temporary Gunther Tulip IVC filter in late pregnancy followed by a Caesarean delivery. The filter was removed without complication. Although the Gunther Tulip filter was used without complication in our first case, its use would cause concern in situations where a complicated and/or prolonged peri-partum course might extend the timeframe during which vena caval interruption is required. This is particularly true following Caesarean section. There is therefore a need for a long-term temporary filter that permits flexibility in the timing of removal. Preliminary experience with the Recovery Filter suggests that this filter would be optimal for the peri-partum patient [1].

The Recovery Nitinol Filter (RNF) is a new retrievable IVC filter composed of 0.013 in. nitinol wires that extend from a central nitinol sleeve. It has six arms and six legs, resulting in dual level protection. This device has been widely used at our institution over the past three years in non-pregnant patients. The preliminary clinical use of the RNF filter in humans is described in a published article by Asch [1]. Since that time, the Recovery Filter has been retrieved successfully as long as 158 days post-insertion.

Removal of the RNF in our second case occurred 75 days after insertion. This allowed time for the patient to be discharged from hospital, return to normal activities and to ensure that it was safe to restart anticoagulation therapy. Thus, the RNF obviated the need for immediate removal of the device in the recovering post-partum patient. The timing of the removal can be flexibly planned according to individual circumstances.

There were no significant complications in the use of the IVC filter in our patients. The filter fracture that occurred in the RNF patient was the only fracture that has occurred in our institution's series of 58 filters. This is a commonly accepted complication of IVC filters, and has been previously reported with the Simon Nitinol filter in up to 14% of patients [3], although the clinical consequences remain unknown. Fortunately, our second patient has remained asymptomatic for 18 months following filter removal. There was a minimal degree of filter migration in the second case, which had no clinical consequences. Routine abdominal radiography is, thus, recommended for all filter patients in order to identify filter fracture/migration.

The option of a retrievable filter is particularly attractive in the setting of pregnancy, due to the transient nature of thromboembolic risk factors, including physiological changes of pregnancy and immobility in the third trimester. Insertion of any IVC filter remains controversial. We recommend consideration of this approach for women in late pregnancy with large or unstable deep iliac vein venous thrombi prior to delivery. The novel RNF device is an attractive option when peri-partum circumstances might extend the duration during which caval interruption should be maintained, for example in the case of an underlying thrombophilia disorder. Further research is required to evaluate the benefits and risks of this intervention.


  1. Top of page
  2. References
  • 1
    Asch MR. Initial experience in humans with a new retrievable inferior vena cava filter. Radiology 2002; 10: 1148.
  • 2
    Hyers TM, Agnelli G, Hull RD, Morris TA, Samama M, Tapson V, Weg JG. Antithrombotic therapy for venous thromboembolic disease. Chest 2001; 119(Suppl. 1): 176S93S.
  • 3
    Streiff MB. Vena caval filters: a comprehensive review. Blood 2000; 95: 366977.
  • 4
    Millward SF. Gunther Tulip retrievable filter: why, when and how? Can Assoc Radiol J 2001; 52: 18892.
  • 5
    Burbridge BE, Walker DR, Millward S. Incorporation of the Gunther temporary inferior vena cava filter into the caval wall. J Vasc Interv Radiol 1996; 7: 28990.
  • 6
    Owen RJT, Krarup KC. Case report: the successful use and removal of the Gunther Tulip inferior vena caval filter in pregnancy. Clin Radiol 1997; 52: 2413.