Who should receive inferior vena cava filters?

Consults, new data can help hospitalists decide.

To make recommendations about inferior vena cava (IVC) filter placement, hospitalists must navigate mixed evidence, conflicting recommendations, and varying practice norms. Some cases are more straightforward than others. Consults with specialists and newly published data can help hospitalists decide what's best for the patients they treat.

Filters are generally appropriate for patients with acute venous thromboembolism who have absolute contraindications for anticoagulation, according to guidelines from both the American College of Chest Physicians (CHEST) and the Society of Interventional Radiology. But for other patients, the guidelines “vary all over the place,” said Matthew Johnson, MD, professor of radiology and surgery at Indiana University School of Medicine in Indianapolis. “There is not agreement around the world as to who should have a filter.”

The confusion stems from a shortage of informative randomized trials, as well as reports of filters that fractured or perforated, migrated to the heart, or completely thrombosed at the level of the cava. These risks are hard to quantify, but more than 1,600 filter-related complications were reported to the FDA between 2009 and 2012, according to a study in the August 2014 Journal of Vascular and Interventional Radiology.

Moreover, studies indicate that IVC filters contribute to lower-extremity deep venous thrombosis (DVT). Among patients with acute symptomatic venous thromboembolism who had substantial bleeding risk, filters significantly increased the risk of DVT but also were associated with significantly lower 30-day mortality from pulmonary embolism, as compared with anticoagulants, researchers reported in the April 29, 2014, Journal of the American College of Cardiology.

Interventional radiologists often cite a 1% increase in thrombosis each year a filter is in place, said Wael Saad, MB, BCH, professor of radiology at the University of Michigan in Ann Arbor. “That is a lay number, but it is often quoted. If you stack up 10 years, you've got a 10% risk,” he said.

Hospitalists should also understand that there is no perfect filter, Dr. Saad added. The more clots a filter catches, the better it is at preventing pulmonary embolism, but the more likely it is to thrombose the entire cava. Likewise, the easier a filter is to retrieve, the more likely it is to migrate.

Such complexities show why clinicians need to tread cautiously when recommending for or against filters, experts said. The decision should “be made judiciously, after careful consideration of the patient's circumstances and the available literature,” said Timothy Morris, MD, ACP Member, professor of medicine and chief of the pulmonary and critical care medicine service at University of California, San Diego. “A group discussion involving the hospitalist, local experts in this particular disease, and those who are trained in IVC placement will likely yield the best decision for any particular patient.”

Shoulds, should-nots, and gray areas

Saying that any patient “definitely” needs a filter “may be too strong a word to use, because much of the IVC issue is controversial,” said Dr. Morris. But filters are certainly a good option for acute pulmonary embolism patients who cannot receive anticoagulants, he added. “That truism is widely accepted and will likely never be tested in a randomized trial.”

On the other end of the spectrum, well-designed trials convincingly show that most pulmonary embolism patients who can take anticoagulants do not need filters, Dr. Morris said. For example, the randomized, multicenter, single-blinded PREPIC2 trial showed that in patients with severe acute pulmonary embolism, placing retrievable filters did not improve 3-month outcomes compared with anticoagulation alone. Researchers reported the results in the April 28, 2015, JAMA.

“That conclusion was reaffirmed in the current antithrombotic guidelines,” said Dr. Morris. “However, the trials and the CHEST recommendations concerned stable pulmonary embolism patients. The area of uncertainty concerns patients with severe unstable acute pulmonary emboli.”

Very few patients with acute pulmonary embolism are unstable, but their mortality can exceed 30%, Dr. Morris noted. The infrequency of this condition makes it hard to study in prospective randomized trials, so clinicians must base their decisions about filters on physiological arguments, case series, and retrospective reviews that are subject to varying interpretations, he added.

“Some experts would treat unstable pulmonary embolism patients exclusively with fibrinolytic drugs, hoping to rapidly reduce the load of clot in the pulmonary arteries. Some would use IVC filters, especially if the unstable patients had coexisting deep vein thrombi, hoping to prevent acute increases in the pulmonary artery load from recurrence. Others would do both, or use other extraordinary means to relieve the pulmonary artery load,” he said.

Some observational data do support using filters in unstable patients. In a 10-year analysis of the Nationwide Inpatient Sample, patients with pulmonary embolism who were on ventilators or in shock and received filters were significantly less likely to die in the hospital compared with those who did not receive filters—regardless of whether patients also received thrombolytics.

Notably, the protective effect spanned all age groups and was greatest in patients older than 80 years. In-hospital mortality was 28% lower with vena cava filters in elderly patients not on thrombolytic therapy and 19% lower in those with a vena cava filter and thrombolytic therapy, researchers reported in the March 2014 American Journal of Medicine.

“I personally would recommend filters in unstable patients,” said lead author Paul Stein, MD, FACP, professor of osteopathic medical specialties at Michigan State University in East Lansing. “Any small recurrent pulmonary embolism might be fatal in an unstable patient. Filters might prevent such a recurrence of even small emboli.”


Another controversial population is patients who need surgery and are considered at high risk for clot, noted Margaret Fang, MD, medical director of the anticoagulation clinic at the University of California, San Francisco. A classic example is obese patients scheduled for bariatric surgery. Such patients often receive filters, although evidence supporting the practice is weak, she said. “Previously, we also sometimes considered placing a filter in patients with poor pulmonary reserve and large lower-extremity clot burden, but practice is moving away from that.”

Obtaining truly informed consent for such a complex treatment choice can be very difficult. “Certainly, patients should be involved in the decision as much as possible, just as they should be involved with every decision,” said Dr. Morris. “However, the physician must not abdicate the responsibility of presenting the potential benefits and risks in as clear a way as possible and making a recommendation based on [his or] her understanding of the situation.”

Clinicians often overstate the long-term risks of filters and may not fully understand their potential benefits in unstable patients, according to Dr. Morris. “Adding to the problem is the fact that the population that stands to benefit most may be hemodynamically unstable, which could interfere with their comprehension and decision-making. For these reasons, I think that patients are entitled to a definite recommendation from their physicians before they make the final decisions for themselves,” he said.

Hospitalists also should notify a patient's primary care clinician when recommending a filter, and they should discuss the fact that the filter might be retrievable in the future, the experts said. On the other hand, “everyone should know that all filters are potentially permanent. You might never be able to get them out,” Dr. Johnson said. Both he and Dr. Saad schedule filter retrieval at the time of placement and then follow up to determine whether removing the filter is possible and indicated, they said. Hospitalists can help facilitate retrieval when they encounter a patient who received a filter for an acute indication that is no longer present, added Dr. Fang.

Rising use with unclear rationale

Rises in IVC filter use have outpaced the estimated incidence of unstable PE, according to Dr. Morris. For example, in a study of Medicare fee-for-service beneficiaries, the annual number of filters placed in hospitalized patients with pulmonary embolism rose by 78% between 1999 and 2010. Mortality from pulmonary embolism fell independently of filter use during that period, researchers reported in the March 8, 2016, Journal of the American College of Cardiology.

“I would urge practitioners against [such] widespread and routine use of IVC filters,” said lead author Behnood Bikdeli, MD, ACP Resident/Fellow Member, a resident at Yale University in New Haven, Conn. “Old age, per se, is likely not enough to warrant the use of these devices.” While filters might save lives, “there is little proof of such concept in the few high-quality studies performed to date.” Therefore, clinicians should minimize the use of IVC filters to indications recommended in the guidelines, he said.

But current filter use appears to be far more liberal. At 1 academic medical center, 23% of patients who received filters had no documented indication based on the Society of Interventional Radiology guidelines, and 40% had no indication based on the CHEST guidelines, investigators reported in the March 2015 American Journal of Medical Sciences.

“This result could reflect genuine differences in opinion regarding patient management. It may be that the guidelines are not applicable in every patient,” said lead author Kenneth Nugent, MD, FACP, professor of medicine at Texas Tech University Health Sciences Center in Lubbock. But the findings could also reflect overuse of filters in lieu of reasonable alternatives, he acknowledged.

Controversies will likely persist until better data are available, but randomized trials of IVC filters face a fundamental problem: It is unethical to assign patients who cannot be anticoagulated to a “no filter” group, Dr. Johnson noted. Therefore, existing trials provide little or no information about the safety and efficacy of filters in patients who need them most, he said.

To help bridge that gap, Dr. Johnson and colleagues are enrolling adults with guideline-based indications for IVC filters into a multicenter, prospective, nonrandomized, 5-year study. For up to 2 years after filter placement, they will use CT scans and other objective measures to track adverse events such as filter perforation, migration, caval thrombotic occlusion, DVT, and perioperative complications. The PRESERVE study aims to enroll 2,100 patients from 60 sites in the United States and currently has enrolled about 120 patients, Dr. Johnson said.

“This trial is examining 7 different filters, almost all the types of filters sold in America,” Dr. Johnson said. “We will be able to use this enormous amount of data to look at which filters were placed and why, if and when they were retrieved and why or why not, and complications overall and according to individual devices over the next 5 years.”