DOI: https://doi.org/10.21203/rs.3.rs-1681929/v1
Background: Inferior vena cava (IVC) filter retrieval is typically accomplished with a standard loop snare. When this fails, more advanced techniques are necessary, especially when removal falls outside a 12-month window. Complications during filter retrieval depend heavily on technique, type of filter, and filter position. In this study we examined safety and efficacy of 536 filter retrievals at a tertiary care center and compared complication rates between standard loop snare and endobronchial forcep retrieval.
Method: We reviewed 536 cases between August 2015 and August 2020, recording retrieval success rates, patient comorbidities, and complication rates at the time of removal.
Results: Total overall retrieval success was 97.9% (525/536), and complications occurred in approximately 6.0% (32/536) of all cases. Success and complications with standard snare technique alone were 99.4% (345/347) and 1.7% (5 Grade I/II, 1 Grade III) and advanced forcep technique 98.8% (171/173) and 14.5% (22 Grade I/II, 2 Grade III, and 1 Grade IV), respectively. There was no significant difference between the technical success rates of the isolated loop snare and forceps techniques (p=0.60) despite a significantly longer dwell time in patients undergoing forceps retrieval ( p<0.001).
Conclusion: To our knowledge, this is the largest cohort of forceps directed IVC filter retrieval present in the literature. Successful endobronchial forceps and loop-snare retrieval rates in this study are similar to previous reports. Although use of endobronchial forceps may be associated with higher complication rates, this is likely due to prolonged dwell times, filter tilt, and attempted removal of non retrievable filters. Overall, forceps-directed retrieval offers a safe, effective means of removal in difficult cases.
Level of Evidence: Level 3, Large Retrospective Study
Inferior vena cava (IVC) filters, although excellent in the short term to prevent clinically significant pulmonary embolism (PE) in appropriate patients, are typically associated with a wide range of adverse events in the chronic setting, including increased risk of distal deep venous thrombosis (DVT) formation, filter migration, caval penetration, and filter fracture. Therefore, both the Food and Drug Administration (FDA) and Society of Interventional Radiology (SIR) recommend they be removed as soon as clinically indicated [1, 2]. Despite these recommendations, retrieval rates have been historically low. Mohapatra et al published data of a large cohort of IVC filters and reported only 6.6% of 131,791 filters were successfully retrieved, while Everhart et al in another study described retrieval rates of approximately 16% of prophylactic filters and 5.69% of therapeutic filters [3, 4]. Recent recognition of poor retrieval rates and the development of online filter registries and other standardized methods of patient follow up at select institutions have shown significantly improved filter removal rates, up to 66% [5–11].
Retrieval of IVC filters can be complex, and incorporation of advanced techniques is often necessary to maintain adequate removal rates. Filter retrieval has been shown to be successful in approximately 80–90% of cases using standard loop snare technique [12]. However, in cases when the filter hook cannot be engaged directly by a snare, alternative methods of retrieval are necessary [12]. Use of adjunctive and/or advanced removal techniques increase when the indwelling filter demonstrates high filter tilt, embedded struts, filter migration, fracture, endothelial overgrowth, and thrombus [12, 13]. Recently, the utilization of endobronchial forceps has been widely accepted as a promising alternative technique for complex retrievals. Although a handful of published studies have demonstrated success using this technique ranging from 85–100% [13–16], the majority of the literature is limited to case reports and case series [17–20].
In this paper, we further investigate the use of endobronchial forceps as an alternative to the traditional loop snare technique by evaluating overall filter removal success rates, patient outcomes, and complication profiles in 536 patients over a 5 year period. To our knowledge, our study is the largest cohort of forceps directed IVC filter retrieval present in the literature, providing additional evidence for the efficacy of this technique.
A single-center retrospective cohort study analysis of all consecutive filter retrievals between August 2015 and August 2020, was performed after institutional review board approval. The requirement for informed consent was waived due to the retrospective nature of the review. Overall, 536 IVC filter retrievals were attempted during this period. Medical records were reviewed to obtain patient age, sex, indication for filter placement, indication for retrieval, retrieval success rate, complications, the brand of filter, type of filter (permanent/temporary), dwell time, removal technique, access site, procedure time, fluoroscopic time, and volume of iodinated contrast administered. Filters were classified as temporary if they were designed for later retrieval, and they were classified as permanent if they were not designed specifically for later retrieval. This designation was made based on the brand of filter. All complications were classified using the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03. Specifically, the category “vascular disorders- others, specify” was utilized for the categorization of complications.
The data was stratified into isolated loop snare and forcep retrieval subgroups based on removal technique. Any attempted retrievals that utilized forceps were characterized as a forceps directed retrieval, including those in which snare was used initially, concurrently, or after attempted forceps retrieval. The association of categorical variables was assessed by the chi-square or Fischer’s exact test to report the Mantel-Haenszel common odds ratios (OR) and 95% confidence intervals (95%CIs). The differences in the distribution of continuous variables (for example, dwell times) across the categorical variables (for example, Retrieval Technique) was investigated by the Mann-Whitney U test. P values less than 0.05 denoted statistical significance. The statistical analyses were conducted by SPSS software version 26 (IBM, Chicago, Illinois, USA).
Indications for insertion of IVC filters were separated into 7 categories and are listed in Table 1. Of the 536 filters, 31 filters were permanent, nonretrievable, while 505 were classified as retrievable. Ten complications occurred during the removal of permanent filters, while 22 occurred during the removal of retrievable filters, yielding complication rates of 32.3% and 4.4%, respectively. The complication rates associated with the removal of specific types of permanent filters and retrievable filters are summarized in Tables 2 and 3, respectively. Overall, 97.9% (525/536) of the filters were successfully removed, of which a 6.0% complication rate (32/536) was observed. The isolated snare technique was attempted in 347 retrievals, isolated forceps were utilized in 105 cases, and a combination of both loop snare and forceps was used in 68 cases. Removal was not attempted in seven cases due to chronic thrombus where risks of removal outweighed benefits and/or recurrent pulmonary embolism. Success rates and graded complication rates for these different retrieval techniques are summarized in Tables 4 and 5.
| Indication for IVC filter placement | Number of filters |
|---|---|
| Contraindication to anticoagulation in the setting of VTE | 187/536 (34.9%) |
| Placement in the pre/post operative period in setting of major surgery | 134/536 (25.0%) |
| Pharmaco-mechanical thrombectomy for iliofemoral DVT | 22/536 (4.1%) |
| Venous thromboembolism on therapeutic anticoagulation | 20/536 (3.7%) |
| Extensive VTE | 39/536 (7.3%) |
| Prophylactic placement (for example in trauma) | 20/536 (3.7%) |
| Placement at outside hospital or with an unknown indication | 114/536 (21.3%) |
| Permanent Filter Complications | |||
|---|---|---|---|
| Filter Type | Grade 1–2 Complications (Minor) | Grade 3–5 Complications (Major) | Complication Rate |
| Trapease | 2 | 1 | 17.6% (3/17) |
| Venatech | 1 | 0 | 50% (1/2) |
| Simon Nitinol | 3 | 0 | 60% (3/5) |
| Greenfield | 1 | 1 | 33.3% (2/6) |
| Birdsnest | 1 | 0 | 100% (1/1) |
| Total | 8 | 2 | 32.3% (10/31) |
| Retrievable Filter Complications | |||
|---|---|---|---|
| Filter Type | Grade 1–2 Complications (Minor) | Grade 3–5 Complications (Major) | Complication Rate |
| Denali | 2 | 0 | 1.3% (2/157) |
| Gunther Tulip | 4 | 1 | 3.4% (5/147) |
| Optease | 0 | 0 | 0% (0.8) |
| Bard G2 | 1 | 0 | 3.8% (1/26) |
| Bard Recovery | 2 | 0 | 9.5% (2/21) |
| Eclipse | 0 | 0 | 0% (0/5) |
| Celect | 5 | 1 | 10.7% (6/56) |
| Option | 3 | 0 | 4.1% (3/74) |
| Bard | 3 | 0 | 42.9% (3/7) |
| Total | 20 | 2 | 4.4% (22/505) |
| TECHNIQUE | # PATIENTS | DWELL TIME AVG (DAYS) | % SUCCESS | % COMPLICATIONS |
|---|---|---|---|---|
| Solo SNARE | 347 | 658 | (345/347)99.4% | 1.7% (6/347) |
| Solo FORCEPS | 105 | 2778 | (104/105) 98.8% | 13.3% (14/105) |
| Loop Snare + Forceps | 68 | 1364 | (67/68) 98.0% | 16.2% (11/68) |
| OVERALL | 536 | 1203 | (525/536) 97.9% | 6.0% (32/536) |
| Complications of IVC Filter Retrieval Based on Techniques of Removal | ||||||
|---|---|---|---|---|---|---|
| Grade 1 | Grade 2 | Grade 3 | Grade 4 | Grade 5 | Total | |
| Loop Snare | 0.58% (2/347) | 0.86% (3/347) | 0.29% (1/347) | 0% (0/347) | 0% (0/347) | 1.7% (6/347) |
| Forceps | 7.6% (8/105) | 2.9% (3/105) | 1.9% (2/105) | 0.95% (1/105) | 0% (0/105) | 13.3% (14/105) |
| Forceps + Loop Snare | 7.4% (5/68) | 8.8% (6/68) | 0% (0/68) | 0% (0/68) | 0% (0/68) | 16.2% (11/68) |
| Other Methods and Failed Filter Access | 6.3% (1/16) | 0% (0/16) | 0% (0/16) | 0% (0/16) | 0% (0/16) | 6.3% (1/16) |
| Total | 3.0% (16/536) | 2.2% (12/536) | 0.56% (3/536) | 0.19% (1/536) | 0% (0/537) | 6.0% (32/536) |
The isolated loop snare technique exhibited a success rate of 99.4% (345/347), with 1.7% (6/347) of patients experiencing complications. Among these filters, 9 were permanent, and 338 were retrievable. The majority of the complications in this cohort demonstrated low grade complications (Two Grade 1; 2/347, 0.58% and three Grade 2; 3/347, 0.86%). One patient experienced retroperitoneal bleeding after standard IVCF removal (Grade 3; 1/347, 0.29%). A higher complication profile was seen in removal of permenant filters (2/5 complications, 40%). Filters removed with the snare technique had a mean dwell time of 658 days, ranging from 1 day to 5713 days. The median dwell time was 221 days.
Among all cases in which forceps retrieval was attempted, there was an overall 98.8% success rate (171/173). Solo forceps filter retrieval had a 99.05% (104/105) success rate, though it also reported a higher complication profile compared to isolated loop snare: 13.3% (14/105). Furthermore, higher grade complications were also more prevalent with forceps retrieval (one Grade 4 ; 1/105, 0.95%; IVC rupture requiring intra-procedural placement of thoracic aortic stent graft, and two Grade 3; 2/105, 1.9%; retroperitoneal bleeding delaying discharge or requiring later hospitalization). Higher grade complications (grade 3 and 4) soley occurred during the removal of permanent filters. Eight low grade complications were observed (one Grade 1: 8/105, 7.6%; and three Grade 2 complications 3/105, 2.9%). The majority of low grade complications occurred with the removal of retrievabale filters (7/11, 63.6%). The mean dwell time was 2778 days, with a minimum dwell time of one day and a maximum of 10,075 days. The median dwell time was 2917 days.
Combined snare and forceps retrieval (failed snare removal converted to forceps or planned concurrent use of forceps and snare) was successful in 98.5% (67/68) of attempted removals, and complications occurred in 16.2% (11/68) of these procedures. Primarily, only low grade complications were observed with combined technique (five Grade 1 5/68, 7.4% and six Grade 2 6/68, 8.8%). Two of these minor complications (2/11, 18.2%) occurred during the attempted retrieval of a permanent filter. The mean dwell time was 1364 days (range: 4–7470 days). The median dwell time was 568 days.
Any attempted filter removal that utilized forceps, regardless of concurrent technique, was classified as a forceps-directed retrieval. Filters retrieved with forceps had a statistically significant longer dwell time (median: 1734 days; range(1-10075 days) compared to filters removed with the solo snare technique (median: 221 days; range(1-5713 days), (p < 0.001). Despite this difference in dwell time, there was no statistically significant difference in retrieval rate between these two techniques (p = 0.60). However, complications were less likely in the solo snare group compared to the forceps retrieval group (OR:0.10; 95%CI:0.04–0.25), (p < 0.001).
VTE is often managed with anticoagulation therapy, but IVC filters provide necessary protection in the setting of DVT/PE in patients with contraindications to anticoagulation [1, 2]. While the use of IVC filters reduces the risk of PE, prolonged filter dwell time is associated with a range of potentially severe complications [1] and more difficult retrieval.
This center’s data was consistent with prior studies, demonstrating high retrieval success rates and low overall complication rates using both standard and advanced techniques [12, 13, 21]. Procedures that utilized the endobronchial forceps method of retrieval had an overall success rate of 99.05% (104/105). Similarly, Stavropoulos et al. demonstrated a success rate of 96% with endobronchial forceps filter retrieval [13]. Furthermore, this institution’s results demonstrated a complication rate of 13.3% (14/105) with 3 major complications (classified as Grade 3 or higher), which was higher than the rate of 3.5% with one major complication reported by Stavropoulos et al. [13]. Though Stavropoulos et al. had lower complication rates, the mean filter dwell time among their patients was 465 days (range 31-2976), while this institution demonstrated a mean dwell time of 2,778 days (range 1–10,075 and median 2917). Prolonged dwell time is associated with an increased risk of filter fracture, migration, thrombosis formation in and around the filter, and strut penetration of surrounding structures [22]. This type of embedded filter makes retrieval much more difficult and may result in higher complication rates.
Based on previous studies and this institution’s data, endobronchial forceps appear to provide a viable if not primary alternative to the loop snare technique in advanced filter retrievals. The success rates were similar- 99.4% (345/347) with loop snare, compared to 98.8% (171/173) with forceps (p = 0.60), despite the significantly longer dwell times of filters removed with forceps (p < 0.001).
This center’s complication rate was similar to published data with solo loop snare retrieval but higher for forceps retrieval. Longer dwell times and a higher proportion of permanent filters may explain this discrepancy. Standard loop snare technique had a much lower complication rate than forceps-directed filter retrieval. Additionally, major complications (defined as grade three or higher) were also more common with forceps removal, similar to published literature. Forceps retrieval of filters with longer dwell time, larger angles of tilt and concurrent caval thrombosis, if present, likely resulted in longer intraprocedural times, higher radiation doses, increased contrast doses and ultimately higher rates of complications. The higher complication profile of forceps mediated filter retrieval was therefore attributed to a combination of filter characteristics and concurrent medical comorbidities within this cohort, rather than the technique itself.
Overall, this data reaffirms that the use of advanced techniques, including endobronchial forceps filter removal, is equivocal in terms of success rates to standard loop snare retrieval. While this data provides good evidence for the safety and efficacy of IVC filter retrieval with endobronchial forceps, further data is necessary to assess the effect of confounding variables on complication rates. Additionally further studies are required to assess when forceps should be used as the primary method of retrieval as this is not yet known or studied.
Endobronchial forceps and the loop snare technique had similar rates of successful IVC filter retrieval, despite the significantly higher dwell time seen in filters retrieved with forceps. Complication rates were higher in the forceps subgroup, though further studies will be needed to determine whether the complications are due to the increased dwell time or the removal technique itself.
IVC (inferior vena cava), PE (pulmonary embolism), DVT (deep vein thrombosis), VTE (venous thromboembolism), FDA (Food and Drug Administration), SIR (Society of Interventional Radiology), CTCAE (common terminology criteria for adverse events), OR (odds ratio), CI (confidence interval).
Author contributions
LK, AG, and PS performed the majority of the raw data collection. PS wrote the initial draft of the manuscript and played a role in each round of revision. LK, AG, EK, RM, GH, CD, KM, and JS helped with revisions and directed raw data collection. RM performed all statistical analysis. All authors approved the final manuscript.
Acknowledgments
No acknowledgements.
Ethical Approval and Consent to Participate
IRB approval was obtained from the USF IRB and the need for participant informed consent was waived due to retrospective review.
Consent for Publication
For this type of study, consent for publication from the patients is not required.
Availability of Data and Materials
Reasonable requests for de-identified data may be accomadated by contacting corresponding author.
Competing Interests
The authors declare that they have no conflict of interest
Funding
This research was not supported by any funding.