Abstract
Background
Retrievable inferior vena cava (IVC) filters are used in patients with acute venous thromboembolism (VTE) who have temporary contraindications to anticoagulation. Although guidelines recommend removing filters once anticoagulation is feasible, many filters remain in place. Long-term consequences of retention remain unclear.
Objectives
This study compared clinical outcomes in patients with VTE and retrievable IVC filters, according to filter status and anticoagulation use.
Methods
We analyzed patients from the international RIETE (Registro Informatizado de Enfermedad TromboEmbólica) registry who received retrievable IVC filters. Outcomes from day 90 onward were stratified by filter status (removed vs retained) and anticoagulation (on vs off). Primary endpoints included VTE recurrence, major bleeding, and all-cause mortality. Filter-related complications were assessed.
Results
Among 1186 patients (mean age, 62 years; 51% male), 51.9% had filters removed by day 90. Patients with retained filters more often had active cancer and received longer anticoagulation, primarily low-molecular-weight heparin. VTE recurrence was higher with retained filters (4.5-6.5 vs 1.2-1.3 events per 100 patient-years), mainly as deep vein thrombosis; removal was associated with lower risk (adjusted incidence rate ratio, 0.29; 95% CI, 0.15-0.54). All-cause mortality was higher in patients with retained filters (7.3-17.7 vs 3.8-4.5; removal: adjusted odds ratio, 0.53; 95% CI, 0.34-0.81). Major bleeding was infrequent, with higher rates in patients with retained filters off anticoagulation. Filter-related complications were rare and did not differ significantly between groups after adjustment.
Conclusion
In this large registry, retained IVC filters were associated with higher risk of VTE recurrence, even with extended anticoagulation. Findings support structured follow-up and timely filter retrieval.
Keywords: anticoagulants, vena cava filters, venous thromboembolism
Visual Abstract
Essentials
-
•
IVC filters are used when anticoagulation for VTE is unsafe.
-
•
We analyzed outcomes in patients with VTE receiving retrievable IVC filters within the RIETE registry.
-
•
Retained IVC filters were associated with more VTE recurrences and higher mortality.
-
•
Findings highlight the importance of follow-up and timely filter removal.
1. Introduction
Venous thromboembolism (VTE) is primarily treated with anticoagulation, which effectively reduces the risk of recurrence and mortality. However, in patients with acute VTE who have a temporary or permanent contraindication to anticoagulant therapy, insertion of an inferior vena cava (IVC) filter is a widely used alternative aimed at preventing pulmonary embolism (PE) [[1], [2], [3]].
Over the past few decades, IVC filter use has increased markedly, extending beyond well-established indications such as active bleeding or urgent surgery. Following 2 US Food and Drug Administration safety communications warning of device-related complications, filter use has declined, particularly in the United States; yet, retrieval rates remain low [4,5]. Many filters intended to be temporary are never removed, often due to persistent contraindications, limited life expectancy, or failures in follow-up systems [4,5].
Although IVC filters may reduce fatal PE in selected high-risk patients, they do not treat established thrombus and do not prevent deep vein thrombosis (DVT). Extended filter dwell time has been associated with complications including filter thrombosis, IVC occlusion, vessel wall perforation, migration, and fracture [6,7]. Accordingly, multiple guidelines—including those from the American Society of Hematology and the American College of Chest Physicians—recommend prompt removal once anticoagulation becomes feasible [1,2].
Despite these recommendations, filters are frequently left in place, particularly in patients with ongoing comorbidities or poor follow-up [[8], [9], [10], [11]]. Extended anticoagulation has been proposed to reduce risk in these patients, but supporting evidence is limited and inconsistent [[12], [13], [14], [15]]. To address this knowledge gap, we analyzed data from the international RIETE (Registro Informatizado de Enfermedad TromboEmbólica) registry to compare long-term clinical outcomes in patients with retrievable IVC filters, stratified by filter status (removed vs retained) and anticoagulation use.
2. Methods
2.1. Study design and participants
RIETE is an ongoing, prospective, international registry of patients with acute VTE (ClinicalTrials.gov identifier: NCT02832245). It collects standardized data on demographics, risk factors, treatments, and clinical outcomes across multiple participating centers. The registry’s design and methodology have been described previously [16].
For this analysis, we included patients enrolled between 2001 and 2024 with acute VTE who received a retrievable IVC filter. Patients in whom IVC filter retrieval was attempted but technically unsuccessful were excluded from the analytic cohort. Patients were categorized according to filter status (removed or retained) and anticoagulant status at day 90 after the index VTE diagnosis. We excluded patients with missing follow-up data, atypical VTE locations (eg, upper extremity thrombosis), or undocumented filter status.
2.2. Study groups and follow-up
Patients were stratified into 4 mutually exclusive groups based on filter and anticoagulation status at day 90: group A—retained filter, on anticoagulation; group B—retained filter, off anticoagulation; group C—filter removed, on anticoagulation; and group D—filter removed, off anticoagulation. Patients whose filter was removed after day 90 were classified as retained. Similarly, those who discontinued anticoagulation before day 90 were considered off anticoagulation. Anticoagulation status at day 90 was determined based on prospectively recorded treatment data in the registry. Patients were considered off anticoagulation only if anticoagulant therapy had been discontinued and a stop date was recorded. Day 90 was selected as the landmark time point because 3 months represents the guideline-recommended minimum duration of anticoagulant therapy for VTE [1,2] and a clinically relevant window for reassessing the indication for IVC filter retrieval. Patients with unclear or missing status were flagged for sensitivity analyses. Follow-up began at day 90 [17]. All events and person-time were counted from this time point onward.
2.3. Data acquisition
Clinical data were collected prospectively at each site using a standardized electronic case report form. Investigators were responsible for data entry and verification. Centralized quality checks ensured consistency across sites. Data collected included demographics, VTE characteristics, treatment details (including anticoagulation type and duration), and clinical outcomes (eg, VTE recurrence, bleeding, mortality, and filter-related complications).
2.4. Variables and outcomes definitions
The primary clinical outcomes assessed were as follows: (1) VTE recurrence, defined as a new, objectively confirmed DVT or PE; (2) major bleeding, based on International Society on Thrombosis and Haemostasis (ISTH) criteria (eg, critical site bleeding, fatal bleeding, or transfusion of ≥2 units); and (3) all-cause mortality, with subcategories for PE-related, bleeding-related, and cancer-related deaths.
Filter-related complications included filter thrombosis or IVC thrombosis, tilt (>15°), migration (>2 cm), perforation, fracture, and other reported mechanical events. For the purposes of this analysis, filter-related complications were assessed in aggregate.
2.5. Statistical analyses
Continuous variables are reported as means (±SD) or medians (IQR), depending on distribution. Categorical variables are expressed as counts and percentages. Comparisons between groups used Student’s t-test or Mann–Whitney U-test for continuous variables and chi-squared or Fisher exact test for categorical variables.
Event rates were calculated per 100 patient-years. Multivariable Poisson regression (or negative binomial regression if overdispersion was detected) was used to estimate adjusted incidence rate ratios (aIRRs) for VTE recurrence, major bleeding, and mortality. Adjusted variables included age, sex, active cancer, VTE type, anticoagulant regimen and duration, and filter indication. Logistic regression was used for binary outcomes (eg, mortality and filter-related complications), with both crude and adjusted odds ratios reported. Comparisons included the following: (1) group A vs group B; (2) group C vs group D; (3) secondary—group A vs group C and group B vs group D; and (4) pooled—retained (A+B) vs removed (C+D). Calendar year of VTE diagnosis was additionally evaluated in sensitivity analyses to assess potential temporal confounding. Cumulative incidence curves were generated to illustrate time-to-event data for VTE recurrence, with unadjusted comparisons using the log-rank test.
2.6. Role of the funding source
Funding bodies had no role in study design, data collection, analysis, interpretation, or manuscript preparation. They do not review or influence any part of this work.
3. Results
3.1. Patient characteristics
After applying predefined inclusion and exclusion criteria, a total of 1186 patients with retrievable IVC filters were included in the final analysis. Details of patient selection and exclusions are shown in Supplementary Figure 1. These patients were classified at day 90 into 4 mutually exclusive groups: group A (retained IVC filter/on anticoagulant treatment, n = 342); group B (retained/off anticoagulation, n = 229); group C (IVC filter removed/on anticoagulation, n = 415); and group D (removed/off anticoagulation, n = 200).
Mean age was 62 ± 17 years and 50.8% were male. Overall, the filter was removed in 615 (51.9%) patients. Patients with retained filters were more likely to have active cancer at VTE diagnosis (32.2% vs 21.0%; P < .001). Baseline characteristics of these patients are presented in Table 1.
Table 1.
Baseline characteristics of patients with VTE in RIETE, according to IVC filter placement.
| Variable | Group A, retained on AC | Group B, retained off AC | Group C, removed on AC | Group D, removed off AC | Overall |
|---|---|---|---|---|---|
| No. patients | 342 | 229 | 415 | 200 | 1186 |
| Age (y), mean ± SD | 63 ± 16 | 68 ± 15∗ | 57 ± 18‡ | 65 ± 17 | 62 ± 17 |
| Male sex | 180 (52.6) | 119 (52.0) | 205 (49.4) | 99 (49.5) | 603 (50.8) |
| Initial VTE presentation | |||||
| DVT only | 106 (31.0) | 76 (33.2) | 119 (28.7) | 68 (34.0) | 369 (31.1) |
| PE only | 103 (30.1) | 77 (33.6) | 138 (33.3) | 60 (30.0) | 378 (31.9) |
| DVT + PE | 133 (38.9) | 76 (33.2) | 158 (38.1) | 72 (36.0) | 439 (37.0) |
| Initial VTE risk factor | |||||
| Active cancer | 108 (31.6) | 76 (33.2) | 80 (19.3) | 49 (24.5) | 313 (26.4) |
| Surgery | 79 (23.1) | 40 (17.5) | 101 (24.3) | 37 (18.5) | 257 (21.7) |
| Hormonal | 17 (5.0) | 6 (2.6) | 38 (9.2)† | 5 (2.5) | 66 (5.6) |
| Prolonged immobility | 95 (27.8) | 67 (29.3) | 104 (25.1) | 59 (29.5) | 325 (27.4) |
| Unprovoked | 83 (24.3)∗ | 71 (31.0) | 114 (27.5) | 67 (33.5) | 335 (28.3) |
| Indication for IVC filter | |||||
| Bleeding during treatment | 69 (20.2)‡ | 115 (50.2) | 65 (15.7)‡ | 118 (59.0) | 367 (30.9) |
| Risk of bleeding | 103 (30.1)‡ | 56 (24.5)† | 113 (27.2)‡ | 24 (12.0) | 296 (25.0) |
| Recurrence despite treatment | 23 (6.7) | 4 (1.7) | 22 (5.3)† | 2 (1.0) | 51 (4.3) |
| Need for surgical intervention | 90 (26.3)† | 39 (17.0) | 111 (26.7) | 44 (22.0) | 284 (23.9) |
| Other | 57 (16.7)‡ | 15 (6.6) | 104 (25.1)‡ | 12 (6.0) | 188 (15.9) |
Values are n (%) unless specified. Comparisons between group D and the other subgroups of patients: ∗P < .05; †P < .01; ‡P < .001.
AC, anticoagulation; DVT, deep venous thrombosis; IVC, inferior vena cava; PE, pulmonary embolism; VTE, venous thromboembolism.
Initial VTE presentation was PE (with or without concomitant DVT) in 817 patients (68.9%), with similar distributions across groups. Isolated DVT accounted for 31%. Recent surgery was a risk factor in 22% of patients. Hormonal risk factors were more frequent in group C (9.2%) compared with the other groups, where frequencies were ≤5%. Unprovoked VTE accounted for 28% of cases. The main indications for filter placement were active bleeding (30.9%), need for urgent surgery (23.9%), and perceived high bleeding risk (24.9%).
3.2. Treatment strategies
Low-molecular-weight heparin was the most common initial therapy (73%), followed by unfractionated heparin (15%) and direct oral anticoagulants (DOACs; 1.3%). In 4.1% of patients, no initial anticoagulation was administered (Table 2).
Table 2.
VTE treatment strategies according to IVC filter status.
| Variable | Group A, retained on AC | Group B, retained off AC | Group C, removed on AC | Group D, removed off AC | Overall |
|---|---|---|---|---|---|
| Initial AC | |||||
| UFH | 52 (15.2) | 38 (16.6) | 61 (14.7) | 27 (13.5) | 178 (15.0) |
| LMWH | 248 (72.5) | 161 (70.3) | 308 (74.2) | 154 (77.0) | 871 (73.4) |
| DOAC | 7 (2.0) | 2 (0.9) | 7 (1.7) | 0 (0.0) | 16 (1.3) |
| VKA | — | — | — | — | — |
| No initial AC | 15 (4.4) | 8 (3.5) | 21 (5.1) | 5 (2.5) | 49 (4.1) |
| Anticoagulant at day 90 | |||||
| UFH | — | — | — | — | — |
| LMWH | 195 (57.0) | 0 (0.0) | 160 (38.6) | 0 (0.0) | 355 (29.9) |
| DOAC | 88 (25.7) | 0 (0.0) | 153 (36.9) | 0 (0.0) | 241 (20.3) |
| VKA | 46 (13.5) | 0 (0.0) | 101 (24.3) | 0 (0.0) | 147 (12.4) |
| Off anticoagulation at day 90 | 0 (0.0) | 229 (100.0) | 0 (0.0) | 200 (100.0) | 429 (36.2) |
| Duration of anticoagulation (d), median (IQR) | 165 (72-402) | — | 113 (40-281) | — | 133 (59-315) |
| Switched anticoagulant class during follow-upa | 63 (18.4) | — | 37 (8.9) | — | 100 (13.2) |
| Duration of filter dwell, since VTE (d), median (IQR) | 210 (113-483)‡ | 193 (116-440)‡ | 29 (15-49) | 34 (21-55) | 138 (16-302) |
Values are n (%) unless specified. Comparisons between group D and the other subgroups of patients: ∗P < .05; †P < .01; ‡P < .001.
AC, anticoagulation; DOAC, direct oral anticoagulant; IVC, inferior vena cava; LMWH, low-molecular-weight heparin; UFH, unfractionated heparin; VKA, vitamin K antagonist.
Switched anticoagulant class during follow-up (eg, VKA→DOAC or DOAC→LMWH).
At day 90, low-molecular-weight heparin remained the most frequently used agent (30%), especially in group A (57%), while DOACs were more often used in group C (37%). Median anticoagulation duration was longer for retained-filter patients: group A vs group C, 165 days (IQR, 72-402) vs 113 days (IQR, 40-281). Overall, 13.2% of patients switched anticoagulant class during follow-up.
3.3. Clinical outcomes
VTE recurrence was higher in patients with retained filters: group A, 7.3% (4.53 per 100 patient-years; 95% CI, 2.96-6.94); group B, 7.9% (6.54; 95% CI, 4.04-10.59); group C, 1.7% (1.21; 95% CI, 0.62-2.35); and group D, 2.0% (1.31; 95% CI, 0.49-3.50). Most recurrent events in patients with retained filters were DVT.
Major bleeding was infrequent across all groups. The highest rate was observed in group B (3.14; 95% CI, 1.64-6.01) and the lowest in group D (1.31; 95% CI, 0.49-3.50).
All-cause mortality was significantly higher in patients with retained filters: 22.3% (17.7; 95% CI, 13.9-22.6) in group B; 12.6% (7.34; 95% CI, 5.71-9.44) in group A; 6.5% (4.53; 95% CI, 3.44-5.97) in group C; and 6.0% (3.79; 95% CI, 2.44-5.90) in group D (Table 3; Supplementary Figure 2). Cancer-related mortality contributed substantially to the overall mortality in groups A and B.
Table 3.
Clinical outcomes during follow-up.
| Variable | Group A retained on AC | Group B retained off AC | Group C removed on AC | Group D removed off AC |
|---|---|---|---|---|
| No. patients | 342 | 229 | 415 | 200 |
| VTE recurrence | 25 (7.3) | 18 (7.9) | 7 (1.7) | 4 (2.0) |
| PE recurrence | 4 (1.2) | 4 (1.7) | 5 (1.2) | 1 (0.5) |
| DVT recurrence | 21 (6.1) | 14 (6.1) | 2 (0.5) | 3 (1.5) |
| Major bleeding | 8 (2.3) | 9 (3.9) | 10 (2.4) | 4 (2.0) |
| All-cause mortality | 43 (12.6) | 51 (22.3) | 27 (6.5) | 12 (6.0) |
| PE-related mortality | 1 (0.3) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Cancer-related mortality | 24 (7.0) | 22 (9.6) | 13 (3.1) | 5 (2.5) |
| Filter-related complications | 3 (0.9) | 4 (1.7) | 2 (0.5) | 1 (0.5) |
| Days from insertion to complication, median (IQR) | 234 (97-328) | 95 (88-169) | 296 (195-396) | 90 (90-90) |
| Days from removal to complication, median (IQR) | — | — | 260 (182-337) | 20 (20-20) |
| Person-years of follow-up, median (IQR) | 0.9 (0.3-2.2) | 0.6 (0.2-1.7) | 0.8 (0.3-1.7) | 0.8 (0.3-2.1) |
| Event rates per 100 patient-years (95% CI) | ||||
| VTE recurrence | 4.53 (3.00-6.59)∗ | 6.54 (4.00-10.1)† | 1.21 (0.53-2.39) | 1.31 (0.42-3.16) |
| PE recurrence | 0.69 (0.22-1.66) | 1.39 (0.44-3.36) | 0.86 (0.32-1.91) | 0.32 (0.02-1.58) |
| DVT recurrence | 3.77 (2.40-5.66)∗ | 5.08 (2.89-8.31)† | 0.34 (0.06-1.11) | 0.97 (0.25-2.64) |
| Major bleeding | 1.40 (0.65-2.65) | 3.14 (1.53-5.77) | 1.71 (0.87-3.05) | 1.31 (0.42-3.15) |
| All-cause mortality | 7.34 (5.38-9.80)∗ | 17.7 (13.3-23.1)‡ | 4.53 (3.04-6.50) | 3.79 (2.06-6.45) |
| Filter complications | 0.52 (0.13-1.41) | 1.44 (0.46-3.48) | 0.34 (0.06-1.11) | 0.32 (0.02-1.57) |
Values are n (%) unless specified. Comparisons between group D vs other subgroups of patients: ∗P < .05; †P < .01; ‡P < .001.
AC, anticoagulation; DVT, deep venous thrombosis; IVC, inferior vena cava; PE, pulmonary embolism; VTE, venous thromboembolism.
In pooled comparisons, retained filters (groups A+B) were associated with significantly higher risks of VTE recurrence and all-cause mortality. Filter removal was associated with a 71% lower adjusted risk of VTE recurrence (aIRR, 0.29; 95% CI, 0.15-0.54) and a 47% lower adjusted risk of all-cause mortality (aIRR, 0.53; 95% CI, 0.34-0.81) compared with filter retention. These associations remained significant after adjustment for baseline covariates, including active cancer, and were unchanged in sensitivity analyses after additional adjustment for calendar year of diagnosis. The pooled difference was concentrated in DVT recurrence favoring filter removal (aIRR, ≈ 0.12; 95% CI, 0.05-0.32), whereas PE recurrence differences were smaller and imprecisely estimated because of few events (Table 4). Time-to-event analyses showed higher recurrence and mortality in patients with retained filters, particularly those without anticoagulation, while bleeding rates were similar across groups (Figure A–C).
Table 4.
Crude and adjusted estimates for clinical outcomes in retained vs removed IVC filters.
| Outcome | Retained (A+B) | Removed (C+D) | Crude IRR | Adjusted IRR (95% CI) |
|---|---|---|---|---|
| VTE recurrence | 43 (7.5) | 11 (1.8) | 0.25 (0.14-0.46) | 0.29 (0.15-0.54) |
| 1 VTE recurrence | 38 (6.7) | 9 (1.5) | ||
| 2 VTE recurrences | 5 (0.9) | 2 (0.3) | ||
| PE recurrence | 8 (1.4) | 7 (1.1) | 0.93 (0.35-2.47) | 1.09 (0.39-3.08) |
| 1 PE recurrence | 8 (1.4) | 6 (1.0) | ||
| 2 PE recurrences | 0 | 1 (0.2) | ||
| DVT recurrence | 35 (6.1) | 5 (0.8) | 0.12 (0.05-0.29) | 0.12 (0.05-0.32) |
| 1 DVT recurrence | 30 (5.3) | 5 (0.8) | ||
| 2 DVT recurrences | 5 (0.9) | 0 | ||
| Major bleeding | 17 (3.0) | 14 (2.3) | 0.78 (0.40-1.52) | 1.07 (0.53-2.15) |
| 1 Major bleeding | 16 (2.8) | 12 (2.0) | ||
| 2 Major bleeding | 0 (0.0) | 2 (0.3) | ||
| 3 Major bleeding | 1 (0.2) | 0 (0.0) | ||
| Outcome | Retained (A+B) | Removed (C+D) | Crude OR | Adjusted OR (95 CI) |
| All-cause mortality | 94 (17.0) | 39 (6.3) | 0.34 (0.23-0.51) | 0.53 (0.34-0.81) |
| PE-related mortality | 1 (0.2) | 0 (0.0) | — | — |
| Cancer-related mortality | 46 (8.1) | 18 (2.9) | 0.34 (0.20-0.60) | 0.53 (0.28-0.98) |
| Filter-related complications | 7 (1.2) | 3 (0.5) | 0.40 (0.10-1.54) | 0.26 (0.06-1.15) |
Poisson regression models were used for recurrent outcomes (VTE recurrence and major bleeding), as patients could experience >1 event during follow-up. Logistic regression was used for binary outcomes (all-cause mortality, cancer-related mortality, and filter-related complications), where only 1 event per patient could occur. No overdispersion was detected in recurrent-event analyses, so negative binomial models were not required.
AC, anticoagulation; DVT, deep vein thrombosis; IRR, incidence rate ratio; OR, odds ratio; PE, pulmonary embolism; VTE, venous thromboembolism.
Figure.
Cumulative incidence curves for clinical outcomes by inferior vena cava filter status and anticoagulation (AC) use (from day 90 onward). (A) Recurrent venous thromboembolism. Retained-filter groups showed a higher cumulative incidence of recurrence, especially in patients not receiving AC (group B). (B) Major bleeding. Event rates were low across groups, with no major divergence by filter status or AC. (C) All-cause mortality. Mortality was highest among patients with retained filters not receiving AC (group B), consistent with adjusted analyses.
3.4. Filter-related complications
Filter-related complications were infrequent overall, but occurred more often in patients with retained filters. Incidence rates were 0.52 per 100 patient-years (95% CI, 0.13-1.41) in group A and 1.44 (95% CI, 0.46-3.48) in group B, compared with 0.34 (95% CI, 0.06-1.11) in group C and 0.32 (95% CI, 0.02-1.57) in group D (Table 3). Details on filter-related complication subtypes (migration, occluding thrombus, vena cava thrombosis, infection, and other) are provided in Supplementary Table 1.
In pooled analyses, the crude complication rate was 1.2% in patients with retained filters (Groups A+B) vs 0.5% in those with removed filters (Groups C+D). After multivariable adjustment, this difference did not reach statistical significance (adjusted odds ratio, 0.26; 95% CI, 0.06-1.15) (Table 4).
4. Discussion
In this large, international registry of patients with retrievable IVC filters, filter retention was associated with significantly higher rates of VTE recurrence, even among patients receiving prolonged anticoagulation. In contrast, major bleeding and filter-related complications were infrequent, with no statistically significant difference between retained and removed filters.
Our findings emphasize that IVC filter retention should not be viewed as a benign or default strategy. Retained filters appear to confer a persistent thrombotic risk, as evidenced by higher rates of recurrent DVT and overall VTE. Extended anticoagulation in these patients may reduce but does not fully offset this excess risk. These results underscore the importance of timely filter retrieval once anticoagulation is feasible, as recommended in multiple clinical guidelines [[1], [2], [3]].
The higher VTE recurrence rates observed in patients with retained filters may reflect both mechanical and biological mechanisms. Filters can disrupt caval flow and promote stasis, increasing thrombotic risk. They may also trigger endothelial dysfunction or serve as a nidus for clot propagation [18]. Although our data cannot establish causality, the observed association with recurrent VTE is clinically relevant and aligns with concerns regarding prolonged filter exposure. Consistent with prior reports, most recurrent events were DVT rather than PE, suggesting that the excess risk in patients with retained filters primarily reflects local thrombosis around or below the filter rather than embolic recurrence.
Evidence regarding optimal anticoagulation in patients with retained filters remains mixed. A randomized trial (Prevention du Risque d'Embolie Pulmonaire par Interruption Cave 2) in patients with acute PE showed that adding a retrievable IVC filter to anticoagulation did not reduce the risk of recurrent PE compared with anticoagulation alone. However, it was associated with an increased risk of DVT, indicating limited benefit in the short term [12]. Observational studies on anticoagulation in patients with retained IVC filters have yielded conflicting results. One retrospective study reported fewer filter-related complications with rivaroxaban [13], while another study suggested filters alone may offer PE protection [14]. US national data show filter placement in patients unable to receive anticoagulation is associated with higher short-term mortality [15]. Our results extend these findings to a large, unselected registry population, showing that retention is associated with more recurrences despite prolonged anticoagulation. Hence, anticoagulation appears necessary but insufficient to normalize outcomes in patients with retained filters. The relatively low use of DOACs reflects the long inclusion period of the registry. However, adjustment for calendar year did not alter the results, suggesting that temporal changes in treatment practice did not explain the observed associations.
All-cause mortality was higher in patients with retained filters, even after adjustment for cancer and other risk factors. However, this finding should be interpreted cautiously as an association rather than evidence of a causal mortality benefit from filter retrieval, since prognostic differences between patients with retained vs removed filters may have contributed to this association. Although active cancer likely contributed to differences in outcomes, adjustment for malignancy did not eliminate the observed association. Residual confounding and baseline prognostic differences may still have contributed to the mortality gradient. Only 1 PE-related death was reported, underscoring that most excess mortality was not directly attributable to embolism. IVC filters are frequently used in oncology patients, who also have lower retrieval rates and a higher risk of complications [11,19].
Filter-related complications were uncommon overall, although numerically higher in the retained-filter groups. Consistent with the short median dwell time (≤ 1 year), individual complication subtypes were rare. This likely explains the lower complication rates observed in this study compared with prior studies that reported more frequent mechanical events with long-term retention [20,21].
Importantly, our study found relatively high retrieval rates (51.9%) compared with other real-world reports, which typically show retrieval in fewer than 40% of cases [8,22,23]. This may reflect more systematic follow-up and VTE-focused care in participating RIETE centers, although the registry itself is noninterventional and cannot dictate clinical practice. Notably, a study from the United States has shown that active surveillance and system-level interventions can substantially improve retrieval success [9], further supporting guideline-based follow-up protocols. Since RIETE is noninterventional, no causal inference about follow-up pathways can be made.
This study has limitations. As a nonrandomized observational registry, residual confounding is possible despite multivariable adjustment. We lacked detailed data on filter types, placement techniques, and imaging confirmation of complications. Cause-specific mortality data were limited, and we could not account for all potential confounders in anticoagulant selection or dosing. Group B (patients with retained filters who were off anticoagulation at day 90) may represent a clinically heterogeneous population, including patients in whom anticoagulation was discontinued for different reasons, which may introduce confounding by indication. Finally, information on repeat or reinsertion of IVC filters after retrieval is not systematically captured in the RIETE registry and, therefore, could not be analyzed. Such events, if present, may influence the interpretation of outcomes following filter retrieval. Strengths of this study include its large international scope, prospective design, and systematic outcome capture in RIETE.
Our findings have important clinical implications. Patients with retained filters remain at elevated risk of recurrent thrombotic events despite extended anticoagulation. These results reinforce guideline recommendations for structured follow-up and timely retrieval whenever feasible, particularly to reduce prolonged filter exposure and thrombotic complications. Future studies should explore optimal anticoagulant type and duration in patients with retained filters and assess interventions to increase safe filter removal.
In conclusion, among patients with VTE receiving retrievable IVC filters, retained filters were associated with higher rates of recurrent VTE and higher all-cause mortality than removed filters. However, the mortality difference should be interpreted as associative rather than causal. The main practical implication of these findings is the need for active reassessment of filter indication and retrieval status during follow-up, with timely filter removal pursued once clinically feasible.
Appendix
Coordinator of the Registro Informatizado de Enfermedad TromboEmbólica (RIETE) Registry: Manuel Monreal.
RIETE Steering Committee Members: Paolo Prandoni, Benjamin Brenner, and Dominique Farge-Bancel.
RIETE National Coordinators: Raquel Barba (Spain), Peter Verhamme (Belgium), Hugo Hyung Bok Yoo (Brazil), Radovan Malý (Czech Republic) Laurent Bertoletti (France), Sebastian Schellong (Germany), Inna Tzoran (Israel), Pierpaolo Di Micco (Italy), Abilio Reis (Portugal), Marijan Bosevski (R. North Macedonia), Lucia Mazzolai (Switzerland), Joseph A. Caprini (USA), and My Hanh Bui (Vietnam).
RIETE Registry Coordinating Center: S&H Medical Science Service.
Members of the RIETE Group—Spain: A. Abad-Fernández, M. D. Adarraga, J. Aibar, A. Alda, J. Alfonso, C. Amado, M. Angelina-García, J. I. Arcelus, A. Ballaz, R. Barba, C. Barbagelata, M. Barrón, B. Barrón-Andrés, F. Beddar-Chaib, E. I. Bencosme de Méndez, Á. Blanco-Molina, I. Cañas, N. Chamorro, L. Chasco, G. Claver, P. Crecente, J. Creu-Paris, J. Criado, C. De Ancos-Aracil, C. De Juana-Izquierdo, J. Del Toro, N. Delgado-Casado, P. Demelo-Rodríguez, M. C. Díaz-Pedroche, J. A. Díaz-Peromingo, Á. Dubois-Silva, D. Durán, J. C. Escribano, A. I. Farfán-Sedado, L. A. Fernández-Bermejo, C. Fernández-Martínez de Septién, J. L. Fernández-Reyes, M. Á. Fidalgo, M. Fuentes-Spínola, C. Gabara, C. García-González, A. García-Ortega, O. Gavín-Sebastián, A. Gil-Díaz, C. Gómez-Cuervo, J. Gorostidi-Pérez, L. Guirado, J. Gutiérrez-Guisado, L. Hernández-Blasco, J. Hernández-Borge, D. Jiménez, A. I. Jiménez-Morales, M. D. Joya, R. Lecumberri, P. Llamas, H. López-Brull, L. López-Jiménez, P. López-Miguel, J. J. López-Núñez, M. López-Rubio, A. López-Ruiz, J. B. López-Sáez, E. Madroñal-Cerezo, A. Maestre, J. Mancebo, P. J. Marchena, M. Marcos, M. Martín-del Pozo, L. Mas-Maresma, M. I. Mercado, J. Moisés-Lafuente, M. Monreal, L. Monzón, M.V. Morales, S. Moreno-Casas, N. S. Navas, J. A. Nieto, M. J. Núñez-Fernández, M. Olid, J. L. Orcastegui, L. Ordieres-Ortega, M. Ortiz, J. Osorio, S. Otálora, E. M. Padín-Paz, J. Pagán, P. Parra-Caballero, P. Parra-Rosado, J. M. Pedrajas, C. Pérez-Ductor, M. Pérez-Pinar, M. L. Peris, M, L. Pesce, J. A. Porras, L. M. Prieto-Gañán, G. Puche, A. Rivas, F. Rivera-Cívico, A. Rodríguez-Cobo, P. Ruiz-Artacho, N. Ruiz-Giménez, P. Ruiz-Torregrosa, V. Sendín, G. Sierra-Palomares, P. Sigüenza, A. Solé, S. Suárez-Fernández, A. Suárez-Simón, R. Tascón-Rodríguez, L. Terés, R. Tirado, M. R. Toda, C. Tolosa, J. Trujillo-Santos, L. Tudela, F. Uresandi, R. Valle, J. F. Varona, E. Vega-Romero, D. Vicente-Navarro, G. Vidal, P. Villares, and J. S. You; Austria: C. Ay, S. Nopp, and I. Pabinger; Belgium: M. Erard, P. Verhamme, and A. Verstraete; Brazil: H. H. B. Yoo; Colombia: S. E. Carbonell, A. C. Montenegro, and J. Roa; Czech Republic: P. Grenar, J. Hirmerova, R. Malý; France: S. Accassat, L. Bertoletti, I. Boulmerka, A. Bura-Riviere, J. Catella, R. Chopard, F. Couturaud, B. Crichi, O. Espitia, R. Le Mao, I. Mahé, P. Morange, F. Moustafa, A. Ngaide, L. Plaisance, G. Poenou, I. Quéré, G. Sarlon-Bartoli, P. Suchon, and E. Versini; Germany: S. Schellong; Iran: Y. Jenab, A. Khodayari, P. Sadeghipour, and S. Yadangi; Israel: B. Brenner, O. Cohen, N. Dally, G. Kenet, and I. Tzoran; Italy: G. Barillari, M. Basaglia, F. Bilora, D. Bissacco, C. Bortoluzzi, M. Bortoluzzi, B. Brandolin, R. Casana, M. M. Ciammaichella, P. Di Micco, A. Guida, E. Imbalzano, D. Lambertenghi-Deliliers, C. Marcon, D. Mastroiacovo, U. Mazzarelli, V. Pizzuti, A. Poz, P. Prandoni, T. Scandiuzzi-Piovesan, P. Scarinzi, C. Siniscalchi, N. Vo Hong, and B. Zalunardo; Latvia: H. Cesnieks, A. Skride, and M. Zicans; Mexico: C. Flores-Ramírez, M. Á. Mendoza Romo-Ramírez, and D. Tarín-Recéndez; Portugal: S. Fonseca, C. Fontes, and J. Meireles; Republic of North Macedonia: M. Bosevski, M. Geceska, and M. Zdraveska; Switzerland: S. Barco and L. Mazzolai; USA: D. J. Angiolillo, J. A. Caprini, L. Ortega-Paz, and A. J. Tafur; and Vietnam: M. H. Bui, S. T. Nguyen, K.Q. Pham, and G. B. Tran.
Acknowledgments
We express our gratitude to ROVI for supporting this registry with an unrestricted educational grant. We also thank the Registro Informatizado de Enfermedad TromboEmbólica registry Coordinating Center, S&H Medical Science Service, for their quality control, data management, logistic, and administrative support and Prof. Salvador Ortiz, Universidad Autònoma de Madrid, Statistical Advisor in S&H Medical Science Service, and Irene González Recio, Statistician in S&H Medical Science Service, for the statistical analysis of the data presented in this paper.
Funding
Registro Informatizado de Enfermedad TromboEmbólica is an investigator-initiated registry, supported by Red Respira from the Instituto Carlos III, Spain (Red Respira-ISCiii-RTIC-03/11), by Sanofi and ROVI. None of these sponsors have had any role in the design of the registry nor do they have rights to access the database, review, or comment on prepublished studies from Registro Informatizado de Enfermedad TromboEmbólica.
Author contributions
O.C. and D.S. performed study conception and design and wrote the manuscript. All authors interpreted the data and critically revised the manuscript and approved the final version of the article.
Relationship Disclosure
There are no competing interests to disclose.
Declaration of generative AI and AI-assisted technologies in the writing process
During the preparation of this manuscript, the authors used ChatGPT 5 (OpenAI) to assist with language editing and phrasing. After using this tool, the authors reviewed and revised the text as needed and take full responsibility for the content of the final version.
Footnotes
Handling editor: Professor Michael Makris
The online version contains supplementary material available at https://doi.org/10.1016/j.rpth.2026.106797.
Contributor Information
Omri Cohen, Email: omricmd@gmail.com.
Members of the RIETE Group:
Manuel Monreal, Paolo Prandoni, Benjamin Brenner, Dominique Farge-Bancel, Raquel Barba, Peter Verhamme, Hugo Hyung Bok Yoo, Radovan Malý, Laurent Bertoletti, Sebastian Schellong, Inna Tzoran, Pierpaolo Di Micco, Abilio Reis, Marijan Bosevski, Lucia Mazzolai, Joseph A. Caprini, My Hanh Bui, A. Abad-Fernández, M.D. Adarraga, J. Aibar, A. Alda, J. Alfonso, C. Amado, M. Angelina-García, J.I. Arcelus, A. Ballaz, R. Barba, C. Barbagelata, M. Barrón, B. Barrón-Andrés, F. Beddar-Chaib, E.I. Bencosme de Méndez, Á. Blanco-Molina, I. Cañas, N. Chamorro, L. Chasco, G. Claver, P. Crecente, J. Creu-Paris, J. Criado, C. De Ancos-Aracil, C. De Juana-Izquierdo, J. Del Toro, N. Delgado-Casado, P. Demelo-Rodríguez, M.C. Díaz-Pedroche, J.A. Díaz-Peromingo, Á. Dubois-Silva, D. Durán, J.C. Escribano, A.I. Farfán-Sedado, L.A. Fernández-Bermejo, C. Fernández-Martínez de Septién, J.L. Fernández-Reyes, M.Á. Fidalgo, M. Fuentes-Spínola, C. Gabara, C. García-González, A. García-Ortega, O. Gavín-Sebastián, A. Gil-Díaz, C. Gómez-Cuervo, J. Gorostidi-Pérez, L. Guirado, J. Gutiérrez-Guisado, L. Hernández-Blasco, J. Hernández-Borge, D. Jiménez, A.I. Jiménez-Morales, M.D. Joya, R. Lecumberri, P. Llamas, H. López-Brull, L. López-Jiménez, P. López-Miguel, J.J. López-Núñez, M. López-Rubio, A. López-Ruiz, J.B. López-Sáez, E. Madroñal-Cerezo, A. Maestre, J. Mancebo, P.J. Marchena, M. Marcos, M. Martín-del Pozo, L. Mas-Maresma, M.I. Mercado, J. Moisés-Lafuente, M. Monreal, L. Monzón, M.V. Morales, S. Moreno-Casas, N.S. Navas, J.A. Nieto, M.J. Núñez-Fernández, M. Olid, J.L. Orcastegui, L. Ordieres-Ortega, M. Ortiz, J. Osorio, S. Otálora, E.M. Padín-Paz, J. Pagán, P. Parra-Caballero, P. Parra-Rosado, J.M. Pedrajas, C. Pérez-Ductor, M. Pérez-Pinar, M.L. Peris, M.L. Pesce, J.A. Porras, L.M. Prieto-Gañán, G. Puche, A. Rivas, F. Rivera-Cívico, A. Rodríguez-Cobo, P. Ruiz-Artacho, N. Ruiz-Giménez, P. Ruiz-Torregrosa, V. Sendín, G. Sierra-Palomares, P. Sigüenza, A. Solé, S. Suárez-Fernández, A. Suárez-Simón, R. Tascón-Rodríguez, L. Terés, R. Tirado, M.R. Toda, C. Tolosa, J. Trujillo-Santos, L. Tudela, F. Uresandi, R. Valle, J.F. Varona, E. Vega-Romero, D. Vicente-Navarro, G. Vidal, P. Villares, J.S. You, C. Ay, S. Nopp, I. Pabinger, M. Erard, P. Verhamme, A. Verstraete, H.H.B. Yoo, S.E. Carbonell, A.C. Montenegro, J. Roa, P. Grenar, J. Hirmerova, R. Malý, S. Accassat, L. Bertoletti, I. Boulmerka, A. Bura-Riviere, J. Catella, R. Chopard, F. Couturaud, B. Crichi, O. Espitia, R. Le Mao, I. Mahé, P. Morange, F. Moustafa, A. Ngaide, L. Plaisance, G. Poenou, I. Quéré, G. Sarlon-Bartoli, P. Suchon, E. Versini, S. Schellong, Y. Jenab, A. Khodayari, P. Sadeghipour, S. Yadangi, B. Brenner, O. Cohen, N. Dally, G. Kenet, I. Tzoran, G. Barillari, M. Basaglia, F. Bilora, D. Bissacco, C. Bortoluzzi, M. Bortoluzzi, B. Brandolin, R. Casana, M.M. Ciammaichella, P. Di Micco, A. Guida, E. Imbalzano, D. Lambertenghi-Deliliers, C. Marcon, D. Mastroiacovo, U. Mazzarelli, V. Pizzuti, A. Poz, P. Prandoni, T. Scandiuzzi-Piovesan, P. Scarinzi, C. Siniscalchi, N. Vo Hong, B. Zalunardo, H. Cesnieks, A. Skride, M. Zicans, C. Flores-Ramírez, M.Á. Mendoza Romo-Ramírez, D. Tarín-Recéndez, S. Fonseca, C. Fontes, J. Meireles, M. Bosevski, M. Geceska, M. Zdraveska, S. Barco, L. Mazzolai, D.J. Angiolillo, J.A. Caprini, L. Ortega-Paz, A.J. Tafur, M.H. Bui, S.T. Nguyen, K.Q. Pham, and G.B. Tran
Supplementary material
References
- 1.Stevens S.M., Woller S.C., Baumann Kreuziger L., Doerschug K., Geersing G.J., Klok F.A., et al. Antithrombotic therapy for VTE disease: compendium and review of CHEST guidelines 2012-2021. Chest. 2024;166:388–404. doi: 10.1016/j.chest.2024.03.003. [DOI] [PubMed] [Google Scholar]
- 2.Ortel T.L., Neumann I., Ageno W., Beyth R., Clark N.P., Cuker A., et al. American Society of Hematology 2020 guidelines for management of venous thromboembolism: treatment of deep vein thrombosis and pulmonary embolism. Blood Adv. 2020;4:4693–4738. doi: 10.1182/bloodadvances.2020001830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kaufman J.A., Barnes G.D., Chaer R.A., Cuschieri J., Eberhardt R.T., Johnson M.S., et al. Society of Interventional Radiology clinical practice guideline for inferior vena cava filters in the treatment of patients with venous thromboembolic disease: developed in collaboration with the American College of Cardiology, American College of Chest Physicians, American College of Surgeons Committee on Trauma, American Heart Association, Society for Vascular Surgery, and Society for Vascular Medicine. J Vasc Interv Radiol. 2020;31:1529–1544. doi: 10.1016/j.jvir.2020.06.014. [DOI] [PubMed] [Google Scholar]
- 4.Axley J.C., May M.M., Novak Z., Aucoin V.J., Spangler E.L., McFarland G.E., et al. Clinical practice and volume trends of inferior vena cava filter usage at a single tertiary care center during a 19-year period. J Vasc Surg Venous Lymphat Disord. 2022;10:887–893. doi: 10.1016/j.jvsv.2022.01.005. [DOI] [PubMed] [Google Scholar]
- 5.Ferro E.G., Mackel J.B., Kramer R.D., Torguson R., Whatley E.M., O’Connell G., et al. Postmarketing surveillance of inferior vena cava filters among US medicare beneficiaries: the SAFE-IVC Study. JAMA. 2024;332:2091–2100. doi: 10.1001/jama.2024.19553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kelkar A.H., Rajasekhar A. Inferior vena cava filters: a framework for evidence-based use. Hematology Am Soc Hematol Educ Program. 2020;2020:619–628. doi: 10.1182/hematology.2020000149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kesselman A., Oo T.H., Johnson M., Stecker M.S., Kaufman J., Trost D. Current controversies in inferior vena cava filter placement: AJR expert panel narrative review. AJR Am J Roentgenol. 2021;216:563–569. doi: 10.2214/AJR.20.24817. [DOI] [PubMed] [Google Scholar]
- 8.Ip S., Salata K., Luo J., de Mestral C., Roche-Nagle G. Predictors of inferior vena cava filter retrieval in a population-based Canadian cohort. J Vasc Surg Venous Lymphat Disord. 2025;13 doi: 10.1016/j.jvsv.2025.102283. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sterbis E., Lindquist J., Jensen A., Hong M., Jr., Gupta S., Ryu R., et al. Inferior vena cava filter retrieval rates associated with passive and active surveillance strategies adopted by implanting physicians. JAMA Netw Open. 2023;6 doi: 10.1001/jamanetworkopen.2023.3211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jia Z., Fuller T.A., McKinney J.M., Paz-Fumagalli R., Frey G.T., Sella D.M., et al. Utility of retrievable inferior vena cava filters: a systematic literature review and analysis of the reasons for nonretrieval of filters with temporary indications. Cardiovasc Intervent Radiol. 2018;41:675–682. doi: 10.1007/s00270-018-1880-9. [DOI] [PubMed] [Google Scholar]
- 11.Fontyn S., Bai Y., Bolger S., Greco K., Wang T.F., Hamm C., et al. Inferior vena cava filter use at a large community hospital: a retrospective cohort study. Sci Rep. 2024;14 doi: 10.1038/s41598-024-60868-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mismetti P., Laporte S., Pellerin O., Ennezat P.V., Couturaud F., Elias A., et al. Effect of a retrievable inferior vena cava filter plus anticoagulation vs anticoagulation alone on risk of recurrent pulmonary embolism: a randomized clinical trial. JAMA. 2015;313:1627–1635. doi: 10.1001/jama.2015.3780. [DOI] [PubMed] [Google Scholar]
- 13.Wang B., Jiang C., Zhang Y., Li X., Xu H. Outcome of anticoagulation with rivaroxaban in patients with non-retrieved inferior vena cava filters for the prevention of filter thrombosis: a retrospective cohort study. BMC Cardiovasc Disord. 2022;22:406. doi: 10.1186/s12872-022-02849-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zektser M., Bartal C., Zeller L., Nevzorov R., Jotkowitz A., Stavi V., et al. Effectiveness of inferior vena cava filters without anticoagulation therapy for prophylaxis of recurrent pulmonary embolism. Rambam Maimonides Med J. 2016;7 doi: 10.5041/RMMJ.10246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Turner T.E., Saeed M.J., Novak E., Brown D.L. Association of inferior vena cava filter placement for venous thromboembolic disease and a contraindication to anticoagulation with 30-day mortality. JAMA Netw Open. 2018;1 doi: 10.1001/jamanetworkopen.2018.0452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bikdeli B., Jimenez D., Hawkins M., Ortíz S., Prandoni P., Brenner B., et al. Rationale, design and methodology of the computerized registry of patients with venous thromboembolism (RIETE) Thromb Haemost. 2018;118:214–224. doi: 10.1160/TH17-07-0511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Joyce E., Haymart B., Kong X., Ali M.A., Carrigan M., Kaatz S., et al. Length of anticoagulation in provoked venous thromboembolism: a multicenter study of how real-world practice mirrors guideline recommendations. J Am Heart Assoc. 2022;11 doi: 10.1161/JAHA.122.025471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Grewal S., Chamarthy M.R., Kalva S.P. Complications of inferior vena cava filters. Cardiovasc Diagn Ther. 2016;6:632–641. doi: 10.21037/cdt.2016.09.08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Brunson A., Ho G., White R., Wun T. Inferior vena cava filters in patients with cancer and venous thromboembolism (VTE): patterns of use and outcomes. Thromb Res. 2016;140(Suppl 1):S132–S141. doi: 10.1016/S0049-3848(16)30112-8. [DOI] [PubMed] [Google Scholar]
- 20.Wang S.L., Siddiqui A., Rosenthal E. Long-term complications of inferior vena cava filters. J Vasc Surg Venous Lymphat Disord. 2017;5:33–41. doi: 10.1016/j.jvsv.2016.07.002. [DOI] [PubMed] [Google Scholar]
- 21.Tsui B., Moon E., Zheng J., Kapoor B., Wang W. Outcomes of the TrapEase inferior vena cava filter over 10 years at a single health care system. J Vasc Surg Venous Lymphat Disord. 2018;6:599–605. doi: 10.1016/j.jvsv.2018.03.012. [DOI] [PubMed] [Google Scholar]
- 22.Sarosiek S., Crowther M., Sloan J.M. Indications, complications, and management of inferior vena cava filters: the experience in 952 patients at an academic hospital with a level I trauma center. JAMA Intern Med. 2013;173:513–517. doi: 10.1001/jamainternmed.2013.343. [DOI] [PubMed] [Google Scholar]
- 23.Brown J.D., Raissi D., Han Q., Adams V.R., Talbert J.C. Vena cava filter retrieval rates and factors associated with retrieval in a large US cohort. J Am Heart Assoc. 2017;6 doi: 10.1161/JAHA.117.006708. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.