Vena cava filters

Question 25 from the second paper of 2021 asked some detailed questions about the insertion of an IVC filter into a trauma patient. Prior to that, the other main SAQ dealing with this matter was Question 2e from the first paper of 2000, which presented the candidate with a pulmonary embolism scenario. In each case the college wanted a list of pros and cons. Since the early 2000 paper, the cons seem to have multiplied. The rushed exam candidate should probably limit their reading to the content of this  highly critical 2013 article from PulmCCM, as well as the excellent LITFL page on IVC filter placement. For the time-rich exam candidate, this issue of CHEST published all of the most recent (9th edition) ACCP guidelines. Whereas previously they had a specific set of guidelines for IVC filters, they have now split them all up between different indications, and one needs to hunt though the articles for the recommendations. PulmCCM link to a summary statement which describes them.  The British also have a set of discrete recommendations available (from 2006).

In the interest of future-proofing this topic against as-yet unwritten CICM SAQs, this question will be answered according to a "critically evaluate" template.

Rationale of IVC filter insertion

• Pulmonary embolis is common, and carries with it a substantial mortality and morbidity.
• Preventative measures such as mechanical thromboprophylaxis and prophylactic anticoagulation are not uniformly effective, and not always appropriate.
• Certain at-risk groups will therefore be at increased risk of fatal pulmonary embolism.
• A mechanical filter may prevent such fatal embolism at the cost of some filter-related complications
• Therefore, there are certain groups in whom the risk of filter-related complications weighs favourably against the risk of massive PE.

Indications for insertion

• Absolute contraindication to anticoagulation in a patient with high risk of DVT/PE
  • These might include unsecured aneurysm after SAH,  or recent major gastrointestinal haemorrhage eg. from varices.
• Complication of anticoagulation which requires it to be reversed
  • eg. significant bleeding while anticoagulated, or HITTS
• Inability to achieve full anticoagulation for whatever reason
• Pulmonary emboli while fully anticoagulated
• Previously, the indications also included pulmonary embolectomy and large free-floating iliofemoral venous thrombus.
• The ACCP also lament the "extended indications" for placement;  people seem to have started inserting these things prophylactically into all sorts of people who have no evidence of DVT or PE, but who are at risk of it. These groups include trauma, burns, cancer patients, those with clot-induced pulmonary hypertension and those with predictably poor compliance with anticoagulation (eg. the demented elderly).

Advantages of the IVC filter

• Can be inserted in patients with a contraindication to anticoagulation
• May decrease the risk of fatal PE
• Is retrievable, supposedly

Disadvantages of the IVC filter

• These things do not prevent or treat DVT; they are not a replacement for anticoagulation
• Venous stasis of lower limbs will occur
• Though retrievable, in practice fewer than 60% are ever retrieved.
• Filter related complications, eg malposition,  IVC damage, perforation, IVC thrombosis, and embolism of filter fragments. The filter may even migrate into the pulmonary artery.
• It may offer no mortality benefit whatsoever.

Insertion of the IVC filter:

• Interventional radiology procedure
• Access via the femoral or IJ
• venogram is performed to define the anatomy
• Seldinger technique of insertion, with fluoroscopic guidance
• Desired position is infrarenal

Evidence to support or refute the use of these devices

• PREPIC trial (2005): 400 patients with proximal DVT, all anticoagulated for a minimum of 3 months, were randomised to either receive a filter or not. The IVC filter reduced the rate of PE from 15% to 6%. However, at 8-year follow-up there was no change in mortality, and the IVC filter group actually had more DVTs.
• Lots of problems with this trial. For instance, all the patients were anticoagulated, and so did not meet the major indication for IVC filter insertion (i.e. contraindication to anticoagulation).
• FILTER-PEVI trial (2012): a very specific group; 141 patients undergoing PEVI (percutaneous endovenous intervention) for DVT. The argument was that this clot retrieval procedure can dislodge emboli, and IVC filters should be deployed before you fiddle with the clot material. The trial confirmed that the IVC filters decrease this risk of iatrogenic PE eightfold.  However, the numbers were: 1 in 14 patients vs. 8 in 22 patients.
• A more pragmatic study (White et al, 2000) looked prospectively at a large group of patients (n = 75,000) of whom 3632 had IVC filters inserted, presumably with conventional indications (i.e. many would have had contraindications for anticoagulation). Depressingly, there did not seem to be any difference in hospitalisation for PE recurrence.
• A recent retrospective study of major trauma patients (Hemmila et al, 2015) also did not demonstrate any mortality benefit.
• Closer to home, a 2019 RCT from Ho et al did not find much of a difference among a group of 240 Australian trauma patients. They looked at a composite endpoint of death and symptomatic PE;  with or without the filter their risk of this composite outcome remained around 14%. However, in a small and powerless subgroup of patients who could not be anticoagulated for whatever reason, none of the filtered patients developed a PE, whereas five of the unfiltered patients did. 
• In 2020, Young & Sriram performed a meta-analysis of the available data, scraping together data from six studies and 1388 participants. Unfortunately, only two studies were worth meta-analysing, from the standpoint of methodological quality. They were PREPIC from 2005, and the abovementioned work by Ho et al (2019). Predictably, not a lot of additional information could be wrung from further analysis of these two. "We can draw no firm conclusions", the authors sulked.