Retrievable inferior vena caval filter for thromboembolic disease in pregnancy

Abstract

An 18 year old nonsmoking woman presented at 37 weeks of gestation in her first pregnancy with a painful swollen left leg. The pregnancy had been uneventful until her admission to hospital. The swelling and tenderness in the left leg extended from the ankle to the groin. There were no other significant clinical findings. An ultrasound scan showed thrombus in the left iliac and left femoral vein. There were limited views of the inferior vena cava but it appeared patent from the bifurcation, with normal flow changes in response to the Valsalva manoeuvre. Intravenous heparin (5000 IU) was started as a loading dose followed by 40 000 IU daily by continuous intravenous infusion. At this dose the activated partial thromboplastin time was maintained between 1.5 to 2.5 times the control. Forty-eight hours later she developed a sinus tachycardia. She had no cough, haemoptysis, dyspnoea or cyanosis. She had a low grade temperature (37.5°C) and was normotensive. An ECG confirmed the sinus tachycardia and the classical changes associated with a pulmonary embolus, namely deep S waves in lead (I), with Q waves and inverted T waves in lead (III). Arterial blood gases showed that she was hypoxic (partial pressure of oxygen 9 kPa) with normocarbia and decreased oxygen saturation (92%). The activated partial thromboplastin time level was rechecked and found to be twice the control. A cardiotocograph showed a baseline of 140 bpm with good variability. In view of the extensive deep vein thrombosis, electrocardiogram findings and hypoxia a diagnosis of pulmonary embolus was made. Oxygen was administered by a mask. To prevent further pulmonary emboli occurring a retrievable inferior vena cava filter was inserted the following day in the angiography suite. The heparin infusion was stopped four hours before the procedure. Under local anaesthetic, the right common femoral vein was percutaneously catheterised. An inferior vena cavogram was performed which confirmed that the left common iliac vein was blocked with thrombus extending just into the cava. The inferior vena cava was otherwise patent. A Gunther Tulip inferior vena cava filter (William Cook Europe, Bjaeversikov, Denmark) was inserted through a 3.3 mm (10 French) sheath and positioned in the cava below the entry of the renal veins. At 38 weeks of gestation a 3.4 kg male infant was born in good condition by elective caesarean section. The delivery was planned to provide good control over the timing of the woman's anticoagulation. Heparin was stopped four hours before delivery and restarted six hours later. On the fifth post-operative day (eight days after insertion) the inferior vena cava filter was removed. A 4 mm (12 French) sheath was inserted percutaneously into the right jugular vein under local anaesthetic. A cavogram was performed, from the right femoral vein, to check that the filter was free of thrombus. The Tulip filter has a tiny hook on its top and this must be snared using a retrieval loop designed to exit from its sheath at right angles. Once the loop was tightened around the top hook (Fig. la) a catheter was slid over the filter, collapsing it and disengaging the four strut hooks from the wall of the inferior vena cava (Fig. 1b). The collapsed filter, including its hooks, is completely covered by a further coaxial catheter and withdrawn through the sheath. Warfarin was started on the seventh post- operative day and heparin was discontinued when the international normalised ratio was 2. Warfarin was continued for three months. She was screened for protein C, protein S and antithrombin III deficiency, lupus anticoagulant and anticardiolipin antibodies and Leiden V mutation (activated protein C resistance). No abnormalities were detected. In future pregnancies she will be advised to take aspirin 75 mg from 12 weeks, subcutaneous heparin will be used in labour and either warfarin or heparin prescribed for six weeks following delivery. (a) Snare tightened around hook on Tulip retrievable filter, (b) Collapsed filter ready for withdrawal. This is the first case report of a retrievable inferior vena cava filter being inserted on a temporary basis in pregnancy to prevent recurrent pulmonary emboli. Venous thromboembolic disease is a serious complication of pregnancy. Pregnancy increases the risk of thromboembolism sixfold. The overall incidence varies between 0.3 and 1.2% of all pregnancies1. In the latest Confidential Enquiries into Maternal Deaths in the United Kingdom (1991–1993) thromboembolic disease was the most common cause of direct maternal deaths, an increase compared with the previous triennium2. To reduce morbidity and death in pregnancy and the puerperium due to thromboembolism, treatment should be started as soon as diagnosis is made. Anticoagulation with heparin is the standard management for deep venous thrombosis. If thrombus is demonstrated extending from the pelvic veins into the cava, especially if the thrombus is free-floating or recurrent emboli have occurred despite anticoagulation then the patient is at increased risk of fatal pulmonary embolus and further intervention should be considered3. Treatment options include surgical removal of thrombus, thrombolytic agents, such as streptokinase and urokinase and insertion of an inferior vena cava filter. Thrombectomy has the advantage of removing the source of emboli but can result in considerable blood loss and is reserved for patients where limb swelling is so great as to cause venous gangrene4. Thrombolytic agents have been successfully used in pregnancy but they may precipitate labour and produce an atonic uterus because of the interference of fibrin degradation products with uterine contraction5. There use should be confined to treating established life threatening pulmonary embolus. Inferior vena cava filter insertion was first introduced in 1967 by Greenfield6. Since then the procedure has been shown to reduce the risk of significant recurrent pulmonary emboli to 1% to 3% in the nonpregnant population7. Use of a filter in pregnancy was first described in 1973 and there have been 20 case reports of inferior vena cava filter insertion in pregnant women. The gravid uterus does not hinder accurate positioning of the filter. Filter insertion via the femoral vein is technically easier than via the jugular vein and the jugular route has been associated with air embolism, cardiac arrythmia, haemopericardium and inadvertent release within the heart8. Some authors recommend that the Greenfield filter is placed above the level of the renal veins in pregnant women to avoid contact between the gravid uterus and the filter. In this site the filter will not be displaced by the contracting uterus and will provide additional protection against embolism from thrombus in pelvic or ovarian veins. The suprarenal location also has the advantage of accelerated venous flow which promotes lysis of entrapped thrombi. The favourable results with suprarenal filter placement have made this an appropriate choice in pregnant women and in young women anticipating pregnancy9. The disadvantage is that if the filter clots completely the renal veins are at risk of thrombosis. Vena caval filters are effective and follow up studies have demonstrated that some filters have a patency rate of up to 95% at 12 year follow up10. However, concerns about possible long term complications arising in young patients with a long life expectancy have led to the development of both permanent filters that can be retrieved (retrievable filters) and temporary filters which must be removed. Retrievable filters become endothelialised and should be removed within ten days of insertion to pre-empt this. However, they can be left in situ if they are found to contain thrombus. Temporary filters differ in that they are attached to a catheter or wire which remains in place through the insertion vein. The disadvantages of temporary filters are that they restrict patient's mobility and that they cannot be left permanently should this become necessary. A retrievable filter offers a short period of protection from embolism without the long term complications of a permanent filter. The maximum risk of embolism is early in the thrombotic process and in the case we have presented the imminent delivery of the fetus, with the associated trauma to and changes in compression of the pelvic veins, posed an increased risk. In this short term high risk situation we suggest the use of a retrievable filter and we considered this safer than the use of streptokinase which may have caused severe obstetric haemorrhage. Permanent filters may be more useful as adjunctive therapy for patients in early pregnancy with extensive thrombus or recurrent emboli despite anti-coagulation11. The safety and efficacy of thrombolytic agents and both retrievable and permanent filters in pregnancy needs to be assessed in long term follow up studies. The authors would like to thank Mrs S. E. Gull for allowing us to report her patient.