Caution: Inferior vena cava filters in distal deep vein thrombosis

This is an invited commentary on Huang Y et al. Independent association between IVC filter placement and VTE risk in patients with upper gastrointestinal bleeding and isolated distal DVT: A retrospective cohort study. Vasc Med 2024; 29: 424–432.

Very little research has examined the safety and efficacy of placing inferior vena cava (IVC) filters for the management of isolated distal (calf) deep vein thrombosis (DVT). Current guidelines endorse IVC filter placement in patients with DVT and a contraindication to anticoagulation; however, the guidelines also recognize some circumstances, including isolated distal DVT and a short-term contraindication to anticoagulation, when monitoring for recurrence or extension is recommended and likely preferred. ¹ In a retrospective study by Huang and colleagues, ² the authors sought to better understand the outcomes of IVC filter use in a population of hospitalized patients with upper gastrointestinal (GI) bleeding diagnosed with isolated distal DVT. The decision to place an IVC filter was made clinically with multidisciplinary input. Patients with a follow-up duplex ultrasound within 3 months were included in the study. Per local clinical practices, duplex ultrasound was typically performed one to two times weekly on inpatients and weekly or biweekly on outpatients to follow distal DVT for propagation. Prophylactic anticoagulation was used when the hemoglobin was stable for over 1 week, and mechanical prophylaxis with graduated compression stockings or intermittent compression devices were the standard of care (including on the leg with distal DVT). Multivariable logistic regression and propensity score matching were used to address differences between the patients who received an IVC filter and those who did not. The primary outcome of interest was venous thromboembolism (VTE) recurrence or propagation of the distal DVT to the proximal veins.

Among the 450 patients included in the analysis, 102 received an IVC filter and 348 were followed with surveillance alone (106 without follow-up Duplex ultrasound were excluded). ² In the unmatched population, those who received IVC filters were on average 3 years younger but otherwise did not have significant differences in comorbidities or laboratory values (there was limited statistical power due to sample size). A similar percentage of patients in the IVC filter and no IVC filter groups had axial distal DVTs (posterior tibial/peroneal); however, there was a higher percentage of patients with muscular vein distal (gastrocnemius/soleal) DVT in the no IVC filter group (45.4% vs 26.5%) and a higher percentage of patients with both axial and muscular distal DVT in the IVC filter group (42.2% vs 22.7%). A measure of the burden of distal DVT, a potential confounder to the outcome of VTE recurrence/propagation, was included in the statistical adjustment and propensity score matching, accounting for both locations (axial vs muscular vs both), as well as unilateral or bilateral involvement. The primary finding from this study was that IVC filter placement resulted in a higher crude odds ratio of VTE recurrence or propagation (OR 2.37, 95% CI 1.5–3.75) that remained significantly elevated after adjustment for potential confounding factors (OR range 2.2–3.3), suggesting not only a lack of efficacy but a signal concerning for harm. Symptomatic pulmonary embolism (PE) occurred in two patients in the no IVC filter group. Propagation to the proximal veins occurred in 36% of patients with an IVC filter compared to 17.8% of patients without a filter (p < 0.001), whereas recurrent VTE was similar between the groups (7.8% vs 7.2%, p = 0.8).

In general, the overuse of IVC filters, the potential harms of placement, and lack of efficacy in some settings are now well recognized, with more recent studies demonstrating a reduction in the number of IVC filters being placed over time.^3,4 Recent data also demonstrate that even when IVC filters are thought to be indicated by some physicians, vascular medicine specialists often find appropriate alternative VTE management strategies.^4,5 The efficacy or lack of efficacy for filter placements in the setting of distal DVT is important to better understand, and this study adds value to the literature in this area. ² The finding of higher VTE recurrence in the IVC filter group in this study is reminiscent of data from the 8-year follow-up of the PREPIC study. ⁶ Patients with initial proximal DVT (with or without PE) were randomized to IVC filter placement or no filter, and all patients were treated for at least 3 months with therapeutic anticoagulation. PREPIC demonstrated a higher risk for recurrent DVT (35.7% vs 27.5%) but a lower risk for PE (6.2% vs 15.1%) in patients with an IVC filter compared to no filter. ⁶

Owing to the nonrandomized and retrospective nature of the study by Huang et al., the results must be interpreted cautiously. It is possible that details of the distal DVT, such as proximity to the popliteal vein, bilateral involvement, or overall burden, influenced the decision to place an IVC filter. These details were not accounted for in the statistical adjustments. Although mechanical ‘prophylactic’ strategies and anticoagulant prophylactic doses were described as ‘routine’, the actual data suggest a significant minority of patients received this, and the authors do not provide data on the true adherence to these strategies within each of the groups. It is plausible that the placement of an IVC filter influenced patient or physician behaviors as it relates to additional prophylaxis.

From a purely mechanical standpoint, one must question the practice of IVC filter placement for distal DVT, which by nature are smaller and known to have less risk for embolization (at least symptomatic embolization). IVC filters are less of a filter and more of a sieve, largely designed to capture larger emboli while otherwise not obstructing blood flow. Any effectiveness of IVC filters for distal DVT might be derived from those patients with proximal DVT propagation. Though the risk of propagation tends to be fairly low for ambulatory patients, the risk in hospitalized patients is not as well defined. The risk for DVT propagation or VTE recurrence in this study is substantial at 25–44% within 3 months, which is about eight times the risk seen in the CACTUS study of low-risk ambulatory patients without cancer who were diagnosed with distal DVT,^6,7 and is much higher than what has been seen in other observational studies.^8,9

There are several take-home messages from Huang et al.’s report. First, the practice of serial duplex ultrasound follow-up demonstrates a critically important component of monitoring high-risk hospitalized patients with distal DVT. This practice is consistent with the American College of Chest Physicians’ guideline that recommends serial ultrasound monitoring in patients with distal DVT at high risk for bleeding who do not receive therapeutic anticoagulation. ¹⁰ The study also conveys another important detail – although an active bleeding event has occurred simultaneously with a DVT, there are additional tools to reduce VTE recurrence, and we should employ mechanical and anticoagulant prophylactic measures as appropriate for the clinical circumstances – and this alone may be the only therapy needed for many patients.

Although the study by Huang and colleagues is imperfect, as is often the case for outcomes studies of ‘real world’ experience, these data are an important addition to the literature. The clear lack of any benefit and the possibility of harm indicate that IVC filter placement for isolated distal DVT in patients with a bleeding contraindication for anticoagulation should not be routine practice.