Abstract
Objective
This study aimed to characterize radiographic characteristics on computed tomography venography and risk factors of inferior vena cava thrombosis (IVCT) in situ after retrievable vena cava filter (VCF) placement.
Methods
Between September 2018 and June 2023, a single-center retrospective cohort study was conducted in patients with or without IVCT in situ following VCF placement. Patient baseline demographics, presentation of lower extremity deep vein thrombosis (LEDVT), thrombus characteristics, concurrent pulmonary embolism, comorbidities and risk factors for LEDVT, and IVCT and VCF-related information were collected and analysed. Univariable analysis followed by multivariable analysis was performed to evaluate the odds ratio (OR) with a 95% confidence interval (CI).
Results
One hundred and seventeen eligible patients were included, regionally isolated filling-defect surrounding the support pillars of VCF and contacting inferior vena cava (IVC) wall on computed tomography venography images were identified, clots were more frequently found on the minor axis or anterior wall of IVC. Univariable analyses suggested that the incidence of IVCT in situ (31.6%, 37/117) was closely associated with age (P = .001), thrombus limb (left (P = .001) and bilateral side (P = .001)), hypertension (P = .008), filter shapes (P < .001), short IVC diameter (P = .009) or magnification percentage (P = .004), and long IVC diameter (P = .006). Multivariable analyses suggested that bilateral side LEDVT (OR, 4.92; 95% CI, 1.56-15.51; P = .007) and increased short IVC magnification percentage (OR, 1.01; 95% CI, 1.00-1.03; P = .013) statistically significant increase the IVCT in situ risk, whereas increased age (OR, 0.96; 95% CI, 0.94-0.99; P = .013) and short IVC diameter (OR, 0.87; 95% CI, 0.77-0.98; P = .026) were associated with decreased odds against IVCT in situ.
Conclusions
IVCT in situ represents regionally isolated filling-defect at points of filter contact with IVC wall. Bilateral side LEDVT and increased short IVC magnification percentage may be potential risk factors impacting the occurrence of IVCT in situ, while increased age and short IVC diameter may decrease the incidence of IVCT in situ and seem to be protective factor against IVCT in situ emergence.
Keywords: Venous thromboembolism, Vena cava, inferior, Filter, Complications, Risk factors
Article Highlights.
-
•
Type of Research: This is a single-center retrospective cohort study.
-
•
Key Findings: This study may have extended the existing literature concerning inferior vena cava (IVC) filter in several important ways. This single-center retrospective cohort study, which aimed to characterize the radiographic characteristics and risk factors of retrievable IVC thrombosis (IVCT) in situ after vena cava filter placement using computed tomography venography in 117 patients, found that IVCT in situ represents an isolated filling defect at points of filter contact with IVC wall, clots were more frequently found on the minor axis or anterior wall of IVC. Bilateral side lower extremity deep vein thrombosis and increased minimum IVC magnification percentage were risk factors for the occurrence of IVCT in situ, while older age and minimum IVC diameter were protective factors against IVCT in situ.
-
•
Take Home Message: IVCT in situ represents isolated filling defects at points of filter contact with IVC wall; clots were found more frequently on the minor axis or anterior wall of the IVC. Bilateral side lower extremity deep vein thrombosis and increased the minimum IVC magnification percentage were risk factors for the occurrence of IVCT in situ, while increased age and minimum IVC diameter were protective factors against IVCT in situ.
Vena cava filter (VCF) is an availably intravascular device that is widely introduced in the infrarenal inferior vena cava (IVC) to prevent lower extremity deep vein thrombosis (LEDVT) from fatal pulmonary embolism (PE).1,2 According to reports, >250,000 VCFs were used in the U.S. Medicare population alone from 2012 to 2016, and more than 40,000 were used in 53 medical centers in China between 2009 and 2019.3,4 However, the use of VCFs has been a subject of debate because it services as both a help and a hinderance, offering prophylaxis for potentially fatal PE in patients with LEDVT while inevitably leading to VCF-related complications.5, 6, 7 The most frequently reported complications in previous studies included VCF fractures (2%-10%), IVC wall penetration (0%-41%), and VCF-related thrombosis (2.0%-32.7%).5, 6, 7 Among the last one, IVC thrombosis (IVCT) in situ after VCF placement is a component scenario and has been under-reported.
IVCT in situ refers to isolated thrombi present on IVC originating from the reaction of IVC to VCF placement and has the potential to trigger secondary IVCT either below, above, within, or cephalad to the VCF, or lead to the progressive LEDVT. Despite clinical significance, there remains a lack of exclusive studies focused on IVCT in situ after VCF placement. Its clinical manifestations are often insidious, nonspecific, and can be easily concealed or confused in patients with LEDVT. IVCT in situ is typically discovered incidentally during cross-sectional computed tomography venography (CTV) imaging or intravascular venography at the time of attempted retrieval for retrievable VCF. However, routinely screening all suspected or asymptomatic patients with a VCF with these methods prove to be tedious, low yield, and cost ineffective, especially considering that the majority of patients with LEDVT with VCFs do not develop IVCT in situ. Hence, the detection of IVCT in situ based on clinical features or radiologic imaging alone remains a challenge. Risk factors, available screening methods, and alternative predictive programs remain unclear.
In this study, we therefore adopted baseline demographics, clinical characteristics, and radiographic findings on cross-sectional CTV imaging for analyses, which is hoped to characterize radiographic characteristics and risk factors of IVCT in situ after retrievable VCF placement, develop an individualized plan, to be aware of higher risk patients, in whom VCF have been implanted, as well to provide preventive intervention for high-risk patients who may be prone to thrombosis after VCF placement.
Methods
Study population and design
This retrospective cohort study was conducted using data from an academic hospital. A comprehensive search of the Medical Database System and the Picture Archiving and Communication Systems was performed. A total of 588 consecutive inpatients with LEDVT aged >18 years received VCFs and CTV, between September 13, 2018, and June 19, 2023, were retrospectively reviewed. The eligible patient, met following inclusion criteria: (i) received infrarenal VCF, (ii) had a CTV before VCF placement, and (iii) underwent CTV within 1 month after VCF placement, were included in the present study. There were 471 patients who were subsequently excluded owing right heart failure, hypovolemic disease, massive ascites, duplication of IVC, permanent filter use, complete VCF thrombosis, or nonfilter-related IVCT (Fig 1). Institutional review board review of this study hospital approved the protocol, and requirement for written informed consent was waived owing to the retrospective nature. All study protocols and procedures were conducted in accordance with the Declaration of Helsinki.
Fig 1.
Study flowchart of inclusion and exclusion criteria for patients with and without IVCT in situ following VCF placement. CTV, computed tomography venography; IVC, inferior vena cava; IVCT, inferior vena cava thrombosis; LEDVT, lower extremity deep vein thrombosis.
Patient baseline demographics, clinical characteristics, and radiographic variables
Of the 117 eligible inpatients, 37 patients with IVCT in situ were classified into the IVCT in situ group and 80 without were divided into non-IVCT in situ group. Patient clinical characteristics including demographics (age, gender, and body mass index), presentation of LEDVT (symptoms and signs, and onset time), thrombus characteristics (thrombus segments [proximal and distal LEDVT], thrombus limbs [left, right, and bilateral side LEDVT]), concurrent PE, comorbidities (hypertension, diabetes, cardiologic artery disease, and history of cerebral vascular disease), risk factors for LEDVT (trauma, recent major surgery history, immobilization, rheumatic diseases of immune system [antiphospholipid antibody syndrome, systemic lupus erythematosus and rheumatoid arthritis], previous venous thromboembolism, cancer, and inflammation), morphological variables of IVC (IVC morphology [oval and circular shaped], IVC diameter [minimum and maximum IVC diameter], and IVC length), VCF-related features (filter location, shapes [spindly and conically shaped filters], and filter length on CTV), measured location of IVC, follow-up time (ie, time duration between prefilter and postfilter placement), anticoagulation treatment, and the conjunctive endovascular treatments (catheter-directed thrombolysis [CDT], percutaneous mechanical thrombectomy [PMT], percutaneous angioplasty, and percutaneous stents) were collected and analyzed.
Indications, diagnostics, measurement modalities, and definitions
The VCFs were placed under venographic guidance, and the indications for VCF placement in this study included anticoagulation contraindications, planned PMT or CDT in patients with acute proximal LEDVT, and prophylactic use in patients with severe trauma or major surgery. The management strategies were selected based on operator preference, economic considerations, and device availability. The indications for PMT and CDT therapy included acute proximal DVT, a life expectancy of >1 year, and a low bleeding risk. For patients without anticoagulation contraindications, we initiated anticoagulant treatment using low-molecular-weight heparin at a bolus dose of 100 U/kg twice daily.
For inpatients who received VCF, CTVs pre- and post-VCF placement were performed using a single CT (a 128-slice dual-source CT; SOMATOM Definition Flash, Siemens, Germany). These images were reconstructed with 1.0-mm-thick slices to allow for the diagnosis and subsequent measurements. CTV images were acquired during a single inspiratory breath hold and under a nonshock status. A contrast agent (Iopromide, Bayer HealthCare, Berlin, Germany) was administered via an 18G cubital intravenous access.
IVCT in situ was defined as isolated filling-defect presented on the IVC wall around VCF in the venous phase of CTV (typical and clot aggravated images were illustrated in Fig 2, A-D). Initial interpretation of the diagnosis of IVCT in situ was based on the radiologist's reading and subsequently verified by interventional radiologists (G.M.F. and H.X.). IVC diameters were measured using quantitative measures based on cross-sectional CTV images (Fig 2, E,F). The quantitative evaluation included IVC minimum and maximum diameter at the identical CT level before and after VCF placement at the position that 2 to 5 cm below the left renal vein. The magnification percentage of IVC diameter was calculated using the following equation: (D1 − D2)/D2 × 100%. Where D1 pertained to the IVC diameter measured after VCF placement, and D2 referred to the IVC diameter measured before VCF placement at the corresponding position. IVC morphology was assessed according to the criteria by Xiao et al.8 An oval-shaped IVC indicated the patient had an inequality between the minimum and maximum diameters, and a circular-shaped IVC represented an approximately equal diameter (Fig 2, G). The proximal LEDVT and distal LEDVT were defined according to the clinical practice guidelines.9
Fig 2.
Typical images of IVCT in situ, IVC diameter measurements, and IVC shapes. The radiographic characteristics of IVCT in situ represented regionally isolated filling-defect (white arrows) surrounding the support pillars of VCF and contacting IVC wall on CTV images (A, B). The clots volume on sagittal CTV (C) tended to increase 3 days later on venography (D). IVC minimum and maximum diameter were measured and magnification percentage of the minimum IVC were calculated using the following formula: (D1 − D2)/D2 × 100%, where D2 is the minimum diameter before IVC placement (E), and D1 is the minimum diameter after VCF placement (F). Oval-shaped IVC (E) and circular-shaped IVC (G) on CTV were presented. CTV, computed tomography venography; IVC, inferior vena cava; IVCT, inferior vena cava thrombosis.
Statistical analysis
Statistical analyses were performed by SPSS statistical software package (version 23.0; SPSS statistical software, Chicago, IL) and R statistical language software (version 3.6.3; The R Foundation for Statistical Computing, Vienna, Austria). The distribution of continuous variables was tested using the Kolmogorov-Smirnov test. Continuous data were presented as mean ± standard deviation, and categorical data were given as counts (percentage). When assessing the correlation between two groups and comparing continuous data, an independent samples t test was used. The significance of categorical data was tested with a χ2 test or Fisher's exact test. The risk factors for IVCT in situ were estimated with the logistic regression models; univariable analysis was followed by multivariable analysis. Findings with a P value of <.050 (two-tailed) were deemed statistically significant.
Results
Baseline demographics and characteristics of included patients with LEDVT
During the study period, a total of 117 patients (mean age, 58.6 ± 17.5 years; 58.1% male) were finally included. Overall, 31.6% of cases (37/117) had IVCT in situ and were allocated to the IVCT in situ group; the remaining 68.4% (80/117) were in non-IVCT in situ group. More than one-half of the included patients (73.5%) experienced limb swelling, and 77.8% (91/117) of the time from symptom onset to admission were ≤7 days. Of these patients, 93.2% (109/117) suffered from proximal LEDVT, with the distribution of thrombus focused mainly on left side limbs (41.9%). Notably, 49.6% of these patients (58/117) were found to have concurrent PEs identified by CT angiography images. The most prevalent comorbidities and risk factors among the included patients with LEDVT were hypertension (39.3%) and recent major surgery history (29.9%).
In terms of IVC- and VCF-related information, IVC measurements were taken at the position that 32.0 ± 9.8 mm under the left renal vein. The mean length of the IVC, measured from the lowest renal vein to the initial of bilateral common iliac veins, was 103.6 ± 13.9 mm, with the most common morphology observed being an oval-shaped IVC. The mean minimum diameter of the IVCs measured was 14.9 ± 4.1 mm, which increased to 19.9 ± 3.3 mm at the identical position after VCF placement (P < .050), and the magnification percentage of minimum IVC diameter was 44.6 ± 46.0%. However, the mean maximum diameter of the IVC ranged from 22.4 ± 4.1 mm to 22.3 ± 3.2 mm before and after filter placement (P > .050). All included patients had VCFs (100%) placed under the renal veins. According to the filter morphology, spindly shaped filters used included Aegisy and OptEase filters; conically shaped filter included Günther Tulip, Celect, and Denali filters. Conically shaped VCFs, accounting for 65.0% of cases (76/117), were the filters used most frequently in the present study. The mean length of filters on CTV was 60.0 ± 4.0 mm. After VCF placement, various conjunctive endovascular treatments, including PMT (30/117), percutaneous angioplasty (33/117), percutaneous stents (12/117), and CDT (58/117), were performed. Anticoagulation treatment between patients with and without IVCT in situ was not significantly different (P = .644). Patients with IVCT in situ had a younger age; a lower proportions of left side LEDVT, hypertension, minimum and maximum IVC diameter, and conically shaped VCF use; but a higher proportion of bilateral side LEDVT and minimum IVC diameter magnification percentage than those without (P < .05). Detailed information is summarized in Table I.
Table I.
Baseline demographics, presentation of lower extremity deep vein thrombosis (LEDVT), thrombus characteristics, concurrent pulmonary embolism (PE), comorbidities, risk factors, and inferior vena cava (IVC) and vena cava filter (VCF)-related information for patients with LEDVT who received VCFs
| Characteristic | All patients (n = 117) | IVCT in situ (n = 37) | Non-IVCT in situ (n = 80) | P value |
|---|---|---|---|---|
| Demographics | ||||
| Age, years | 58.6 ± 17.5 | 50.6 ±17.1 | 62.3 ± 16.5 | .001 |
| Gender, male | 68 (58.1) | 19 (51.4) | 49 (61.3) | .313 |
| BMI ≥30 kg/m2 | 13 (11.1) | 4 (10.8) | 9 (11.3) | .999 |
| Symptoms and signs | ||||
| Limb swelling | 86 (73.5) | 27 (73.0) | 59 (73.8) | .929 |
| Limb pain | 48 (41.0) | 18 (48.6) | 30 (37.5) | .254 |
| Symptoms or signs onset time ≤7 days | 91 (77.8) | 28 (75.7) | 63 (78.8) | .710 |
| Thrombus segmentsa | ||||
| Proximal LEDVT | 109 (93.2) | 34 (91.9) | 75 (93.8) | .999 |
| Distal LEDVT | 8 (6.8) | 3 (8.1) | 5 (6.3) | |
| Thrombus limbs | ||||
| Left side | 49 (41.9) | 7 (18.9) | 42 (52.5) | .001 |
| Right side | 30 (25.6) | 10 (27.0) | 20 (25.0) | .815 |
| Bilateral side | 38 (32.5) | 20 (54.1) | 18 (22.5) | .001 |
| Concurrent PE | 58 (49.6) | 18 (48.6) | 40 (50.0) | .892 |
| Comorbidities | ||||
| Hypertension | 46 (39.3) | 8 (21.6) | 38 (47.5) | .008 |
| Diabetes mellitus | 20 (17.1) | 7 (18.9) | 13 (16.3) | .721 |
| CAD | 14 (12.0) | 2 (5.4) | 12 (15.0) | .238 |
| CVD | 9 (7.7) | 1 (2.7) | 8 (10.0) | .315 |
| Risk factors for LEDVT | ||||
| Trauma | 18 (15.4) | 4 (10.8) | 14 (17.5) | .351 |
| Recent major surgery history | 35 (29.9) | 14 (37.8) | 21 (26.3) | .203 |
| Immobilization | 18 (15.4) | 8 (21.6) | 10 (12.5) | .204 |
| Rheumatic diseases of immune systemb | 12 (10.3) | 4 (10.8) | 8 (10.0) | .999 |
| Previous VTE | 1 (0.9) | 1 (2.7) | 0 (0) | .316c |
| Cancer | 13 (11.1) | 7 (18.9) | 6 (7.5) | .131 |
| Inflammation | 4 (3.4) | 3 (8.1) | 1 (1.3) | .177 |
| Morphologic variables of IVC | ||||
| IVC morphology | ||||
| Oval shaped | 98 (83.8) | 28 (75.7) | 70 (87.5) | .107 |
| Circular shaped | 19 (16.2) | 9 (24.3) | 10 (12.5) | |
| IVC diameter before filter placement, mm | ||||
| Minimum IVC diameters | 14.9 ± 4.1 | 13.4 ± 3.7 | 15.6 ± 4.2 | .009 |
| Maximum IVC diameters | 22.4 ± 4.1 | 20.8 ± 3.4 | 23.1 ± 4.3 | .006 |
| IVC diameter after filter placement, mm | ||||
| Minimum IVC diameters | 19.9 ± 3.3 | 20.4 ± 2.6 | 19.7 ± 3.6 | .313 |
| Maximum IVC diameters | 22.3 ± 3.2 | 22.5 ± 2.3 | 22.1 ± 3.5 | .498 |
| Magnification percentage of minimum IVC diameter, % | 44.6 ± 46.0 | 62.2 ± 49.5 | 36.4 ± 42.1 | .004 |
| VCF-related characteristics | ||||
| Filter locations | ||||
| Infrarenal vein | 117 (100.0) | 37 (100.0) | 80 (100.0) | N/A |
| Filter shapesd | ||||
| Spindly shaped | 41 (35.0) | 27 (73.0) | 14 (17.5) | <.001 |
| Conically shaped | 76 (65.0) | 10 (27.0) | 66 (82.5) | |
| Filter length, mm | 60.0 ± 4.00 | 60.7 ± 3.8 | 59.7 ± 4.0 | .184 |
| Measured location of IVC, mm | 32.0 ± 9.8 | 33.7 ± 10.9 | 31.2 ± 9.3 | .209 |
| Follow-up time, days | 13.2 ± 11.3 | 13.4 ± 9.8 | 13.1 ± 12.0 | .907 |
| Anticoagulation treatment | 105 (89.7) | 32 (86.5) | 73 (91.3) | .644 |
| Conjunctive endovascular treatments | ||||
| PMT | 30 (25.6) | 12 (32.4) | 18 (22.5) | .253 |
| PTA | 33 (28.2) | 12 (32.4) | 21 (26.3) | .490 |
| PTS | 12 (10.3) | 2 (5.4) | 10 (12.5) | .396 |
| CDT | 58 (49.6) | 17 (45.9) | 41 (51.3) | .594 |
BMI, Body mass index; CAD, cardiologic artery disease; CDT, catheter-directed thrombolysis; CVD, cerebral venous disease; IVCT, inferior vena cava thrombosis; PTA, percutaneous angioplasty; PMT, percutaneous mechanical thrombectomy; PTS, percutaneous stent; VTE, venous thromboembolism.
Continuous data are presented as the means ± standard deviations; categorical data are given as the counts (percentage).
Proximal LEDVT included thrombus in the common iliac vein, external iliac vein, common femoral vein, proximal and distal segments of the femoral vein, and/or popliteal vein. Distal LEDVT included thrombus in distal veins, including the anterior tibial vein, posterior tibial vein, peroneal vein, gastrocnemius muscle vein, and soleus muscle vein.
Including antiphospholipid antibody syndrome, systemic lupus erythematosus and rheumatoid arthritis.
Fisher exact.
According to the filter morphology, spindly shaped filter includes Aegisy and OptEase filters; conically shaped filter includes Günther Tulip, Celect and Denali filters.
Radiographic characterization of IVCT in situ after filter placement detected by CTV
The mean time from VCF placement to CTV examination performed was 13.2±11.3 days (range, 1-30 days). Among 37 patients (31.6%) with IVCT in situ, the regionally isolated filling defects at points of filter contact with the lateral IVC wall were identified. Twenty-eight patients (75.7%) had an oval-shaped IVC, 85.7% of the clots (24/28) were detected in the minor axis of caval wall, and 14.3% (4/28) were in the major axis. Furthermore, 53.6% of the clots (15/28) were in the anterior wall of the caval minor axis, 32.1% (9/28) in the posterior wall, and 14.3% (4/28) in the lateral wall. When it comes to patients with circular-shaped IVCs, 55.5% of the clots (5/9) were detected in the anterior wall of the IVC, 11.1% (1/9) in the posterior wall, and 33.3% (3/9) in the lateral walls.
Relationship between associated factors and IVCT in situ after VCF placement
Univariable analysis showed, compared to the patients without IVCT in situ, the factors of age [P = .001], thrombus limb (left side [P = .001] and bilateral [P = .001]), hypertension (P = .008), filter shapes (P < .001), minimum IVC diameter (P = .009) or magnification percentage of minimum IVC diameter (P = .004), and maximum IVC diameter (P = .006) were statistically significantly associated with IVCT in situ. The remaining baseline demographics, presentation of LEDVT, thrombus characteristics, concurrent PE, comorbidities and risk factors, and IVC and VCF-related information were not statistically significant (P > .050). Patients with an IVCT in situ had a statistically significantly lower mean minimum diameter (13.4 ± 3.7 mm vs 15.6 ± 4.2 mm; P = .004) and higher minimum IVC magnification percentage after VCF placement (62.2 ± 49.5% vs 36.4 ± 42.1%; P = .004) than those without IVCT in situ. The IVC diameter was expanded and not statistically significant difference (P = .313). Multivariable logistic regression analysis showed that age (odds ratio [OR], 0.96; 95% confidence interval [CI], 0.94-0.99; P = .013), minimum IVC diameter (OR, 0.87; 95% CI, 0.77-0.98; P = .026) or increased minimum IVC magnification percentage (OR, 1.01; 95% CI, 1.00-1.03; P = .013), and bilateral side thrombus (OR, 4.92; 95% CI, 1.56-15.51; P = .007) were associated with the occurrence of IVCT in situ following filter placement (Table II). The results revealed that older age and minimum IVC diameter may be protective factors for IVCT in situ after VCF placement, which was associated with a decreased odds as age advances and minimum IVC diameter grows. Meanwhile, patients with bilateral LEDVT and increased minimum IVC magnification percentage were more likely to develop IVCT in situ than those with left-sided LEDVT.
Table II.
Multivariate regression analysis of risk factors for inferior vena cava filter (IVCF) thrombosis in patients with lower extremity deep vein thrombosis (LEDVT)
| Risk factors | OR | 95% CI | P value |
|---|---|---|---|
| Age, years | 0.96 | 0.94-0.99 | .013 |
| Hypertension | 0.54 | 0.18-1.62 | .273 |
| Filter shapes | 1.90 | 0.67-5.41 | .227 |
| Minimum IVC diameter, mm | 0.87 | 0.77-0.98 | .026 |
| Maximum IVC diameter, mm | 0.90 | 0.79-1.02 | .087 |
| Minimum IVC diameter magnificent percentage | 1.01 | 1.00-1.03 | .013 |
| Thrombus segment | |||
| Left | 1 | - | - |
| Right | 2.35 | 0.68-8.18 | .179 |
| Bilateral | 4.92 | 1.56-15.51 | .007 |
CI, Confidence interval; OR, odds ratio; IVC, inferior vena cava.
Discussion
As is well-known, there have been few studies concerning IVCT in situ after the placement of VCF,7 which may stem in part from the fact that digital subtraction venography, a two-dimensional image, remains the most commonly used approach for detection of filter-related thrombosis.10 Although filling defects would be identified on it, it remains difficult to determine whether they represent trapped emboli or in situ clots.10,11 Three-dimensional venography of IVC, such as CTV examination, helps overcome some limitations of the two-dimensional images and provides more accurate sectional morphology information.9,10 Hence, in this retrospective cohort study, we used the radiographic findings on CTV images coupled with baseline demographics and clinical characteristics to investigate the characteristics and predisposing risk factors associated with IVCT in situ. IVCT in situ was characterized by regionally isolated filling defects surrounding the support pillars of VCF and contacting IVC wall on CTV images. Based on the baseline demographics, clinical characteristics, and radiographic findings of IVC morphology, diameter, VCF-related features, we found that hypertension (OR, 0.54, 95% CI, 0.18-1.62; P = .273), filter shapes (OR, 1.90, 95% CI, 0.67-5.41; P = .227), and maximum IVC diameter (OR, 0.90, 95% CI, 0.79-1.02; P = .087) did not demonstrate significant relationship between patients with and without IVCT in situ. Bilateral LEDVT (OR, 4.92; 95% CI, 1.56-15.51, P =.007) and increased minimum IVC magnification percentage (OR, 1.01; 95% CI, 1.00-1.03; P = .013) were risk factors for the occurrence of IVCT in situ, whereas older age (OR, 0.96; 95% CI, 0.94-0.99; P = .013) and minimum IVC diameter (OR, 0.87; 95% CI, 0.77-0.98; P = .026) were associated with decreased odds against IVCT in situ. These factors are likely to have complex interactions in fostering its occurrence.
The reported incidence of filter-related IVCT varied widely owing to time differences in available screening.1,10, 11, 12, 13 Teo et al10 conducted a retrospective study involving 440 patients who experienced VCFs placement, revealing an incidence of 8.0% within the first 30 days, which decreased to 3.1% at a dwell time of 151 to 180 days. As follow-up time went by, the incidence decreased to 2% at 17 months12 and 1.3% at 2 years.13 The findings suggested a decreasing trend in the incidence of filter related IVCT with prolonged dwell time, which were also in line with our previous study.11 Hence, CTV was performed within 1 month after placement of VCF chosen as a data point to capture a critical period for potential thrombosis while minimizing the impact of potential other confounders, such as intimal hyperplasia. However, filter-related IVCT is speculated to originate from a in situ event in the IVC wall or progressive clots owing to the interception of LEDVT debris, the incidence of IVCT in situ has not been extensively investigated. In the present study, with a mean follow-up 13.2 days, the overall rate of IVCT in situ was seen in 31.6% of patients, which was in accordance with our data of filter-related IVCT from a multicenter RCT,7,11 but higher than that in the study by Teo et al.10 The difference may be partly attributable to the high percentage of intravascular venography or CTV used. Hence, the true incidence may be underestimated because thrombosis sometimes is not detected consistently or reported. It is important to note that our study had a relatively small sample size and may not be representative across different institutions.
The pathogenesis of IVCT in situ is still not well-understood. Therefore, carrying out this study in patients with or without IVCT in situ to investigate related risk factors may yield some clinical implications, which have captured the attention of clinicians.7,11 As classical hypothesis,14,15 Virchow's triad, which describes blood hypercoagulability, hemodynamic alterations, and endothelial/vessel wall injury as factors for thrombosis. Interestingly, in the present study, patients who suffered from IVCT in situ tended to have a higher likelihood of clots on the minor axis of caval wall compared to the major axis. IVCT in situ is assumed to share a common mechanism with Virchow's triad. First, filter placement potentially triggers off unnoticed or severe IVC vessel/endothelial wall injury as the radial forces transmitted by the distal part of the support pillars of spindly or conically shaped filter legs, especially when a filter is inserted into a relatively smaller IVC diameter and higher IVC magnification percentage. Subsequently, these events may invite a secondary inflammatory reaction, leukocyte, and platelet aggregation, and so on, further stimulating thrombosis in situ. Second, the filter blockage and LEDVT may collectively or separately alter IVC hemodynamics, because bilateral LEDVT may decrease the blood flow into the IVC, which had a 4.92-fold increased risk of IVCT in situ compared with left-sided LEDVT. Third, blood hypercoagulability may aggravate thrombus, which should be validated in the future. Moreover, other factors are likely to have an additional complex interaction in promoting the occurrence of IVCT in situ. However, this hypothesis is speculative and has not been explored.
A typical manifestation of IVCT in situ remains unclear; to our knowledge, no studies have investigated the radiographic characteristic of it after filter placement. In this study, based on the CTV images observations, a possible hypothesis concerning IVCT in situ images may be that IVCT in situ presents as isolated thrombosis at points of filter contact with the IVC wall, and is unlikely to vanish with the use of anticoagulation, which mainly depends on our experience. IVCT in situ commonly presents in the minor axis of caval wall in patients with an oval-shaped IVC, but on the anterior wall of the IVC in patients with a circular-shaped IVC. In contrast, filter-related thrombosis, which was intercepted by filter migrating from LEDVT, presents as free-floating clots in the centre of VCF. However, these hypotheses are merely speculative and based on limited structural data and do not seem to be evidence based. With these limitations in mind, further investigation and analysis in patients with filter prophylaxis is needed to enrich the characteristics and provide a better understanding.
Interestingly, the mean age of patients with IVCT in situ was significantly younger than those without. The older age was found to be protective against IVCT in situ after filter placement, which is counterintuitive, given the predisposition of elderly populations that may be implicated in venous thrombosis.11 This scenario is in line with a report from Liu et al14 on IVCT. Regardless of other risk factors, age is independently associated with IVCT in situ. Although prior studies regarding filter-related thrombosis illustrated that the rheumatic diseases of immune system were associated with a 14-fold increased risk of IVCF thrombosis,11 it did not reach statistical significance after analysis, implying that it may not be an independent risk factor for IVCT in situ. Liu et al14 hypothesized that inheritable thrombophilia is one of the main risk factors for IVCT; the onset age of IVCT in patients with an inheritable thrombophilia is relatively very young. However, whether the negative association of age was attributable to inheritable thrombophilia has not been identified; screenings for factor V Leiden, prothrombin 20,210G>A, and deficiencies of antithrombin, protein C, and protein S was not performed in this study owing to the constraints of our detection conditions. Therefore, a complete screen may be warranted in patients with IVCT in situ. Further studies are needed to confirm these assumptions and provide a better understanding of the factors.
The present study has several important limitations that merit discussion. First, the definition of IVCT in situ was not strict and empirically determined, depending mainly on CTV images, which might have affected the findings. Further study, including filter use as prophylaxis, would be insightful to confirm the radiologic characteristics. Second, confounding variables involving filter tilt angle and hood wall apposition,16, 17, 18 as well variety of IVC filter, may be related to the IVCT in situ; however, the indicators were lacking. Third, the drawbacks of selective basis may exist, because eligible patients were recruited from an Asian population; further studies in diverse populations are needed to confirm these findings. Fourth, given that our study was limited by the small numbers of cases, a relatively low follow-up rate, and other undefined risk factors, our conclusions should be interpreted with caution. In addition, a well-designed prospective trial with a longer follow-up period is warranted. In the future, studies with larger sample sizes that include more factors and exclude confounding factors are needed to overcome these limitations and provide a more comprehensive understanding of IVC filter changes.
Conclusions
This study showed that IVCT in situ presents regionally isolated filling defects at points of filter contact with the lateral IVC wall. Bilateral LEDVT and increased minimum IVC magnification percentage were found to be risk factors for the occurrence of IVCT in situ after VCF placement, which suggested that intense monitoring of patients with this sign is essential because of the relatively high incidence of IVCT in situ in this population. Older age and a larger minimum IVC diameter decrease the incidence of IVCT in situ and seem to be a protective factor against IVCT in situ emergence. These findings may be helpful in better understanding the radiographic characteristics on CTV and risk factors of IVCT in situ, which may help to create awareness of higher risk patients, in whom some preventive intervention needs to be taken. Future research is crucial to validate these findings and guide evidence-based adjustments.
Author Contributions
Conception and design: MG, RJ, ZL, BZ, JK, XH, JG
Analysis and interpretation: MG, RJ
Data collection: MG, RJ
Writing the article: MG, RJ
Critical revision of the article: MG, RJ, ZL, BZ, JK, XH, JG
Final approval of the article: GM, RJ, ZL, BZ, JK, XH, JG
Statistical analysis: MG, RJ, JK
Obtained funding: MG, JG
Overall responsibility: JG
Disclosures
None.
Footnotes
Supported by the National Natural Science Foundation of China (81871463), Jiangsu Medical Association Special Fund Project [SYH-3201140-0088(2023035)], Nanjing Medical Science and Technology Development Project (YKK23116), and Nanjing Medical University Science and Technology Development Fund Project (NMUB20230163).
The datasets generated and analyzed during the current study are not publicly available, as the study data are related to other studies that are progressing but are available from the corresponding author upon reasonable request.
The editors and reviewers of this article have no relevant financial relationships to disclose per the Journal policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest.
References
- 1.Warren R.E., Dhruva S.S., Kinard M., Neuhaus J.M., Redberg R.F. Trends in FDA adverse events reporting for inferior vena cava filters and estimated insertions in the US, 2016 to 2020. JAMA Intern Med. 2023;183:271–272. doi: 10.1001/jamainternmed.2022.6161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ramakrishnan G., Willie-Permor D., Yei K., et al. Immediate and delayed complications of inferior vena cava filters. J Vasc Surg Venous Lymphat Disord. 2023;11:587–594.e3. doi: 10.1016/j.jvsv.2022.08.011. [DOI] [PubMed] [Google Scholar]
- 3.Ahmed O., Wadhwa V., Patel K., Patel M.V., Turba U.C., Arslan B. Rising retrieval rates of inferior vena cava filters in the United States: insights from the 2012 to 2016 summary Medicare Claims data. J Am Coll Radiol. 2018;15:1553–1557. doi: 10.1016/j.jacr.2018.01.037. [DOI] [PubMed] [Google Scholar]
- 4.Zhang F., Gu J., Li H.L. China venous thromboembolism and inferior vena cava filter clinical application cooperation group. Diagnosis and treatment of venous thromboembolism and clinical application of inferior vena cava filter in China. J Vasc Surg Venous Lymphat Disord. 2023;11:1149–1156. doi: 10.1016/j.jvsv.2023.05.003. [DOI] [PubMed] [Google Scholar]
- 5.Jia Z., Wu A., Tam M., Spain J., McKinney J.M., Wang W. Caval penetration by inferior vena cava filters: a systematic literature review of clinical significance and management. Circulation. 2015;132:944–952. doi: 10.1161/CIRCULATIONAHA.115.016468. [DOI] [PubMed] [Google Scholar]
- 6.Andreoli J.M., Thornburg B.G., Hickey R.M. Inferior vena cava filter-related thrombus/deep vein thrombosis: data and management. Semin Intervent Radiol. 2016;33:101–104. doi: 10.1055/s-0036-1581087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gong M., Jiang R., Zhao B., et al. Relationship between vascular access and angulation of vena cava filter at placement and retrieval: a multicenter retrospective cohort study. Ther Adv Chronic Dis. 2023;14 doi: 10.1177/20406223231200254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Xiao Y.D., Zhang Z.S., Ma C. Cavographic vs. cross-sectional measurement of the inferior vena cava diameter before filter placement: are we routinely oversizing? Eur Radiol. 2019;29:3281–3286. doi: 10.1007/s00330-018-5820-5. [DOI] [PubMed] [Google Scholar]
- 9.Vedantham S., Desai K.R., Weinberg I., et al. Society of Interventional Radiology position statement on the endovascular management of acute iliofemoral deep vein thrombosis. J Vasc Interv Radiol. 2023;34:284–299.e7. doi: 10.1016/j.jvir.2022.10.038. [DOI] [PubMed] [Google Scholar]
- 10.Teo T.K., Angle J.F., Shipp J.I., et al. Incidence and management of inferior vena cava filter thrombus detected at time of filter retrieval. J Vasc Interv Radiol. 2011;22:1514–1520. doi: 10.1016/j.jvir.2011.08.006. [DOI] [PubMed] [Google Scholar]
- 11.Gong M., Liu Z., Kong J., et al. Incidence and risk factors for inferior vena cava filter thrombosis detected at time of filter retrieval in patients with lower extremity deep vein thrombosis: a multicenter retrospective cohort study. Quant Imaging Med Surg. 2023;13:8313–8325. doi: 10.21037/qims-23-724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xiao N., Karp J., Lewandowski R., et al. Inferior vena cava thrombosis risk in 1582 patients with inferior vena cava filters. Radiology. 2022;303:300–302. doi: 10.1148/radiol.211169. [DOI] [PubMed] [Google Scholar]
- 13.King R.W., Wooster M.D., Veeraswamy R.K., Genovese E.A. Contemporary rates of inferior vena cava filter thrombosis and risk factors. J Vasc Surg Venous Lymphat Disord. 2022;10:313–324. doi: 10.1016/j.jvsv.2021.07.016. [DOI] [PubMed] [Google Scholar]
- 14.Liu X.R., Zhou W., Chen F. Severe compression of left iliac vein is a protective factor for the risk of inferior vena cava thrombosis. J Vasc Surg Venous Lymphat Disord. 2022;10:1107–1112. doi: 10.1016/j.jvsv.2022.04.018. [DOI] [PubMed] [Google Scholar]
- 15.Hume M. Venous thrombosis: mechanisms and treatment. Adv Exp Med Biol. 1978;102:215–224. doi: 10.1007/978-1-4757-1217-9_14. [DOI] [PubMed] [Google Scholar]
- 16.Grubman S., Kostiuk V., Brahmandam A., et al. Effect of inferior vena cava filter placement position on device complications. J Vasc Surg Venous Lymphat Disord. 2023;11:1165–1174.e2. doi: 10.1016/j.jvsv.2023.05.020. [DOI] [PubMed] [Google Scholar]
- 17.Hajduk B., Tomkowski W.Z., Malek G., Davidson B.L. Vena cava filter occlusion and venous thromboembolism risk in persistently anticoagulated patients: a prospective, observational cohort study. Chest. 2010;137:877–882. doi: 10.1378/chest.09-1533. [DOI] [PubMed] [Google Scholar]
- 18.Gillespie D.L., Spies J.B., Siami F.S., Rectenwald J.E., White R.A., Johnson M.S. Predicting the safety and effectiveness of inferior vena cava filters study: design of a unique safety and effectiveness study of inferior vena cava filters in clinical practice. J Vasc Surg Venous Lymphat Disord. 2020;8:187–194.e1. doi: 10.1016/j.jvsv.2019.07.009. [DOI] [PubMed] [Google Scholar]