Abstract
Inferior vena cava (IVC) filters are believed by many to be an effective method of preventing clot from a lower extremity deep vein thrombosis (DVT) travelling centrally and causing a pulmonary emobolism (PE). 1 Prior to the development of percutaneous devices, open ligation/clip placement was utilized for PE prevention. 2 The potential role of the IVC filter has been diminished by the effectiveness and low morbidity of systemic anticoagulation.3–5
The complications of vena cava filter insertion can be significant with a mortality rate of 0.12% reported. The common complications include DVT, vena cava perforation, filter migration, total occlusion due to clot despite anticoagulation and PE despite filter placement.5,6 In the USA, IVC filters are placed in up to 15% of all DVTs, and indeed in 2008 over 65,000 filters were placed (20 per 100,000 population).1,4,7 In the UK, Hospital Episode Statistics reveal that 1173 IVC filters were placed in 2011 (2 per 100,000 population), an increase from 532 in 2006 (1 per 100,000 population).8,9
It is estimated that there are 466,000 cases of DVT and 296,000 cases of PE per annum in the European Union, with 370,000 fatal cases (150, 95 and 120 per 100,000, respectively). 10 In the USA, there are an estimated 900,000 venous thromboembolism cases, with 300,000 fatalities from PE alone (290 and 97 cases per 100,000, respectively). 11
Little evidence governs their insertion and even less for their effectiveness.2,12 This has led to extreme discordance between guidelines and practice.1,3,13 A disconcerting feature of the literature is the poor rate of randomized controlled trials – only two studies were methodologically sound enough to be included in Young et al.'s 2 Cochrane review.
The only indications for insertion with guideline recommendation are for cases of DVT where anticoagulation is contra-indicated or where there is recurrent PE despite adequate anticoagulation.3,13
Most concerning of all is the poor rate of retrieval of so-called ‘temporary’ filters (also known as optional filters).This is despite the evidence that anticoagulation need not be stopped for safe retrieval, 14 and the development of new filter designs which allow removal safely after some months. Importantly, technique of insertion to achieve accurate centre-lining is vital for removal and haemodynamic flow.15,16
The USA data show that only 1.5–2.1% of temporary filters were removed in 2008 1 with even prospective studies displaying disappointing removal figures of 10–70%.17–19
In the UK, the retrieval rate has improved from 13% in 2006 to 26% in 2011, though this remains poor. 9 The rate of IVC thrombosis ranges from 6–30%. The rate of IVC thrombosis without anticoagulation has been found to be as low as 0% in one study, with the large eight-year PREPIC study finding a rate of 13%, with only 35% of the IVC filter undergoing long-term anticoagulation. Those advocating placement must accept responsibility for removal.
Trauma is a field where there is great interest in the placement of IVC filters. Trauma patients are often immobile and often have contraindications to anticoagulation due to ongoing bleeding. This has led to the development of insertion under ultrasound guidance at the bedside. 20 However, even here the evidence is conflicting at best: initial trials highlight the problems with retrieval. 21 and further studies have shown the inconsistencies with US treatment. 22 Indeed, a recent study of cost-effectiveness questioned the use of IVC filters even in this highly selected group. 23
IVC filter placement in the context of cancer patients has been advocated in some circles, especially when IVC manipulation is required; however, a recent review of patients with renal cell carcinoma undergoing resection recommended that filters should not be placed unless the conditions of contra-indicated anticoagulation or continued PE despite anticoagulation are met. 24 IVC filters are also a hazard in this and similar groups of patients due to possibility of incorporation in tumour thrombus and technical difficulty in surgery secondary to the filter.
Prophylactic use of IVC filters during bariatric surgery has also gathered interest despite studies showing no reduction in rates of PE. 25 Caution should be used in the interpretation of small scale studies in favour of the procedure.26,27
Thrombophilic patients remain a difficult group, with little evidence to guide decision-making, due to the complex nature of coagulation. There is a significantly increased risk of IVC filter thrombosis. 28
The recent implementation of the UK Inferior Cava Filter Registry in conjunction with the British Society of Interventional Radiology 29 has led to improved treatment and guideline adherence – in the latest report spanning 2008–2010, 1255 IVC filters were placed, with 25% overall retrieval rate (41% of temporary filters). 30 This is mirrored by work on trauma patients in the USA.
Overall it is clear that IVC filters remain a favoured procedure despite limited evidence of its benefits and poor retrieval of the implanted filters. Further high-quality randomized studies are required to inform us of the benefits or detriments to placing filters. Currently, the practice is unguided despite decades of filter insertion, which it is vital to improve.