Novel endovascular retrieval of duodenal-penetrating inferior vena cava filters: a report of two cases and technical considerations

Introduction

Background

Inferior vena cava (IVC) filters are suitable for patients at high risk of venous thromboembolism. Despite significant improvements in filter designs and significant advances in filter removal technology, the actual removal rate is still <50%. Common complications associated with prolonged filter placement (e.g., filter adhesion and tilt, rupture, or displacement; venous wall penetration; thrombotic obstruction) and rare complications (e.g., pseudoaneurysm of the abdominal aorta, tear of a lumbar artery, perforation of the duodenum and small intestine, ureter injury) are increasing annually (1-4).

Rationale and knowledge gap

Most of these rare complications, such as duodenal penetration by the filter foot, are treated surgically. Current limitations of advanced retrieval techniques: laser-assisted removal achieves 99% success for embedded filters but requires the hook or tip to remain within the IVC, is not Food and Drug Administration (FDA)-approved for this indication, and is unavailable in China. Robotic-assisted retrieval remains experimental with unresolved technical issues and high costs. Rigid guidewire and balloon-assisted techniques have increased retrieval rates to >90% but are contraindicated for permanent filters or when the recovery hook has penetrated the IVC wall. No established interventional methods exist for these specific scenarios.

Objective

For permanent filters or cases where the recovery hook has penetrated the IVC wall, current international guidelines recommend open surgery. We herein describe a novel endovascular intervention method that was successfully used to remove the filter foot piercing the duodenum in two patients. This study aims to explore the feasibility of endovascular techniques in these extreme scenarios. We present this article in accordance with the CARE reporting checklist (available at https://acr.amegroups.com/article/view/10.21037/acr-2026-0053/rc).

Case presentation

All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. The patients provided written informed consent for the publication of this report and the accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Case 1

A 66-year-old male admitted with ‘epigastric pain for 3 months, worsening for 1 week’. Pain was located in the subxiphoid region, persistent, unrelated to meals, without hematemesis or melena. The patient’s vital signs are stable, with a temperature of 36.5 ℃, pulse 78 bpm, respiration 18/min, and blood pressure 132/84 mmHg. The abdomen is flat and soft, with mild tenderness in the subxiphoid region; there is no rebound tenderness or muscle rigidity, and bowel sounds are normal. Gastroscopy showed a metal rod within the descending duodenum (Figure 1), but removal under gastroscopy failed. A permanent filter and left iliac vein stent were implanted because of lower extremity venous thrombosis and remained in place for 9 years. The patient refused surgery and was transferred to our hospital. Preoperative assessment revealed the following: (I) the filter was a permanent filter (Simon Nitinol Filter; Bard, New Providence, NJ, USA), and the current domestic (Chinese) and foreign guidelines all indicate contraindications for interventional removal; (II) one foot in the lower part of the filter had pierced the duodenum; (III) the foot was located far from the abdominal aorta, and the risk of damage to the abdominal aorta was small; (IV) the filter had no recovery hook, and the upper and middle connections of the filter were subject to weak force; these connections were designated LOOP points (based on the guidewire loop technique). We concluded that the filter could not be removed with the conventional endovascular intervention method.

Figure 1 Gastroscopy revealed a metal rod in the duodenum.

The surgical procedure was performed as follows: (I) preoperative preparation: the patient fasted (both food and water) for 2 days before surgery, with administration of preoperative intravenous fluids and prophylactic antibiotics. (II) Anesthesia and positioning: the surgery was performed under local anesthesia. The right jugular vein area and bilateral inguinal areas were routinely disinfected and draped. (III) Vascular access and sheath placement: an 18F vascular sheath (Ultimum; Abbott, Abbott Park, IL, USA) was implanted through the right femoral vein. (IV) Protective filter implantation: a protective recoverable filter (Denali; Bard) was implanted above the original filter through the right internal jugular vein to prevent filter fragmentation and distal embolization to the heart or pulmonary artery. (V) Arterial roadmap guidance: through the left femoral artery approach, a pigtail catheter was placed in the subrenal aorta for real-time marking, enabling safe visualization of the filter LOOP direction during the procedure. (VI) Balloon-assisted intimal peeling: a 12- to 14-mm balloon was introduced via guidewire through the right femoral vein, advanced through the gap between the left filter foot and the IVC wall, and inflated to 6–8 atm to gradually peel the filter from the intima. The leftward displacement of the filter foot was closely monitored under fluoroscopy. (VII) LOOP technology application: a LOOP point was formed at the middle connection of the filter. The modified sheath core (proximal end truncated to accommodate four 0.035-inch guidewires) was advanced to the middle section of the filter. Under balanced mechanics, force was directed to the filter itself rather than the vascular wall, enabling safe inversion and retrieval into the sheath. (VIII) Post-procedural confirmation: the IVC was re-examined by venography, and abdominal aortography confirmed no contrast extravasation, dissection, or stenosis (Figure 2A-2F). The removed IVC filter is shown in Figure 3.

Figure 2 Endovascular retrieval of the duodenal-penetrating permanent Simon Nitinol filter (Case 1). (A) Implantation of the protective filter and positioning of the pigtail catheter in the abdominal aorta. (B) Inferior vena cava angiography suggested severe intimal hyperplasia. (C) Use of balloon to peel filter off. (D) LOOP technology. (E) Inferior vena cava angiography. (F) Abdominal aortography.

Figure 3 Removed inferior vena cava filter.

Follow-up:

• At 3 months: CT venography demonstrated a patent IVC without stenosis or thrombosis; gastroscopy showed well-healed duodenal mucosa without stricture.
• At 6 months: repeat CT showed normal IVC diameter without thrombosis; patient reported no abdominal pain or distension.
• At 12 months: telephone follow-up confirmed good general condition, no gastrointestinal symptoms, no lower extremity edema.

Case 2

A 58-year-old woman was admitted to another hospital because of a 1-week history of epigastric discomfort. The patient presents with a temperature of 36.8 ℃, pulse of 82 bpm, respiration of 20/min, and blood pressure of 128/80 mmHg. Abdominal examination reveals a flat and soft abdomen with mild tenderness in the upper mid-abdomen, no palpable mass, and negative shifting dullness. Laboratory findings show hemoglobin 132 g/L, white blood cell count 7.2×10⁹/L, platelet count 195×10⁹/L; coagulation profile: prothrombin time (PT) 12.0 s, international normalized ratio (INR) 1.1, activated partial thromboplastin time (APTT) 30.2 s; renal function: creatinine 65 µmol/L, blood urea nitrogen 4.8 mmol/L. Gastroscopy suggested foreign matter in the descending duodenum (Figure 4). The patient had undergone implantation of the filter because of femoral neck fracture and lower extremity deep vein thrombosis 8 years previously. Two attempted interventions to remove the filter failed. The patient continued therapy with anticoagulant drugs. The hospital recommended open surgery to remove the filter with vascular and bowel repair if necessary; however, the patient refused and was transferred to our hospital. Preoperative assessment revealed the following: (I) the filter was a first-generation product (Günther Tulip; Cook Medical, Bloomington, IN, USA), and the retaining time after implantation had exceeded 8 years, far beyond the recovery time window. Additionally, the sub-filter wire was prone to adhere to the IVC wall, and the risk of IVC tearing was high with the use of a conventional intervention method. (II) The foot of the filter was hard, and it was prone to penetrating the IVC. One foot of the filter had pierced the duodenum, and the risk of intestinal perforation was high. (III) The foot was close to the back of the abdominal aorta, and the risk of damage to the abdominal aorta was also high. (IV) The filter recovery hook had pierced the IVC (Figure 5A,5B). We concluded that conventional transjugular retrieval was not technically feasible given the IVC-penetrating recovery hook; therefore, we performed a novel endovascular intervention as follows.

Figure 4 One foot of the filter was visible in the duodenum under gastroscopy.

Figure 5 Abdominal computed tomography showed that the filter recovery hook and one foot had pierced the inferior vena cava near the posterior wall of the abdominal aorta. (A) Coronal view showing the filter strut penetrating the inferior vena cava wall. (B) Axial view showing the filter foot in close proximity to the posterior wall of the abdominal aorta.

The surgical procedure was performed as follows: (I) preoperative preparation: the patient fasted (both food and water) for 2 days before surgery, with administration of preoperative intravenous fluids and prophylactic antibiotics. (II) Anesthesia and positioning: the surgery was performed under local anesthesia. The right jugular vein area and bilateral inguinal areas were routinely disinfected and draped. (III) Vascular access and sheath placement: an 18F vascular sheath (Ultimum; Abbott) was placed through the right femoral vein after predilation of an 8F catheter sheath. Anteroposterior and lateral IVC imaging confirmed that the filter recovery hook had pierced the IVC wall (Figure 6A). (IV) Protective filter implantation: a vena cava filter recovery system (Cook Medical) was implanted through the right internal jugular vein above the renal vein of the original filter. (V) Arterial roadmap guidance: through the left femoral artery, a pigtail catheter was placed at the upper end of the renal artery. Anteroposterior and lateral imaging demonstrated that the left filter foot was close to the posterior wall of the abdominal aorta. The catheter tip was adjusted to the posterior aortic wall and used as a real-time marker. (VI) Balloon-assisted intimal peeling: utilizing guidewire and catheter technology via the right femoral vein approach, a 12- to 14-mm balloon was introduced into the gap between the left filter foot and the IVC wall. The adherent filter feet were peeled off laterally in sequence, with close monitoring of the distance between the left foot and the pigtail catheter. If this distance shortened, the balloon position was immediately adjusted to prevent aortic protrusion. (VII) LOOP technology application: at the proximal end of the left foot, the guidewire exited the modified sheath core (front end removed). The sheath core was advanced to the LOOP point under balanced mechanics using guidewire capture technology. The filter foot was flipped, and the recovery hook was slowly maneuvered toward and into the IVC lumen. Continuous monitoring of the foot-to-catheter distance was maintained throughout this step. (VIII) Safety monitoring: no abdominal pain should occur during any step; if pain occurred, the surgical plan was adjusted immediately. (IX) Filter retrieval completion: the recovery hook was grasped from the neck approach and collected into the sheath. The retrieved filter showed one foot bent upward but was otherwise intact. (X) Post-procedural confirmation: imaging revealed no filling defect at the original filter site and no contrast extravasation, vascular dissection, or stenosis (Figure 6B-6F).

Figure 6 Endovascular retrieval of the duodenal-penetrating Günther Tulip filter with IVC-penetrating recovery hook (Case 2). (A) Filter recovery hook piercing the inferior vena cava. (B) Positioning of the pigtail catheter. (C) Inferior vena cava balloon dilation, LOOP surgery. (D) Abdominal aortography. (E) Inferior vena cava angiography. (F) Removed inferior vena cava filter. IVC, inferior vena cava.

Follow-up:

• At 3 months: CT venography demonstrated a patent IVC without thrombosis; the patient reported no abdominal pain.
• At 12 months: repeat CT showed no IVC stenosis or thrombosis; patient resumed normal diet without gastrointestinal discomfort.

Discussion

Key findings

This report describes successful endovascular removal of two IVC filters with duodenal penetration using balloon-assisted intimal peeling, LOOP technology, and intra-arterial roadmap guidance. Both filters—one permanent Simon Nitinol (9 years) and one Günther Tulip (8 years) with a recovery hook penetrating the IVC—were safely retrieved into an 18F sheath without vascular or intestinal injury.

Following long-term filter implantation, persistent mechanical compression by filter feet against the IVC wall leads to local ischemia, inflammatory response, and intimal hyperplasia, ultimately penetrating the IVC wall. Given the close anatomical relationship between the descending duodenum and IVC (particularly at the L3–4 level), the filter foot further erodes the duodenal wall after penetrating the IVC, forming a chronic rather than acute perforation. This explains the patients’ presentation with chronic abdominal pain rather than an acute abdomen. This mechanistic understanding supports the feasibility of endovascular retrieval without immediate bowel repair, as the perforation is sealed by the filter foot itself, minimizing peritoneal contamination.

Strengths and limitations

Strengths

This is the first report demonstrating endovascular removal of filters with duodenal penetration, including permanent filters and those with IVC-penetrating recovery hooks. The technique offers a minimally invasive alternative to open surgery, avoiding laparotomy, bowel repair, and general anesthesia risks.

Limitations and future directions

This report includes only two cases, and the technique requires advanced endovascular skills available only in specialized centers. While our 12-month follow-up data are favorable, long-term outcomes and potential late complications (e.g., duodenal strictures, IVC thrombosis) remain unknown and require continued surveillance. The method may not be applicable to all filter types or anatomical variations. Additionally, the procedures were performed by a single experienced operator, and reproducibility in other centers requires validation. Collaboration between interventional radiologists, vascular surgeons, and gastroenterologists is crucial for optimal patient selection and management. Multi-center prospective studies with standardized protocols are needed to confirm safety, efficacy, and generalizability before widespread adoption.

Comparison with similar research

Laser-assisted removal using excimer laser sheaths achieves 99% success for embedded filters but requires the hook or tip to remain within the IVC lumen. This fundamental limitation excludes permanent filters without hooks and cases where the hook has penetrated the IVC wall—precisely the scenarios addressed by our technique (5). It is also not FDA-approved for this use and is unavailable in China. Robotic-assisted retrieval offers precision but remains experimental, expensive, and unavailable in most centers (6). Consequently, surgical removal remains a last resort due to its invasiveness, anesthesia risks, and other drawbacks. Our technique, while having a slightly lower success rate (98.75% based on our center’s experience with >300 difficult filters), offers broader indications and requires no expensive equipment.

Due to varying technical expertise across centers, current guidelines recommend avoiding: (I) interventional removal of permanent filters; (II) retrieval when the hook has pierced the IVC; and (III) intimal peeling, which may cause IVC contracture and thrombosis—more serious than obstruction alone (7). Although permanent filter implantation has declined, complications from prior implants are emerging. Based on our 13-year clinical experience and analysis of new technologies, we propose expanded indications for interventional filter removal (including permanent filters): (I) indwelling filters causing psychological disorders; (II) IVC occlusion leading to lower limb ulcers or pelvic congestion requiring reconstruction; (III) hook penetration of the IVC or other rare complications; (IV) long-term anticoagulation causing bleeding complications (e.g., gastrointestinal, intracranial hemorrhage).

The core principle of interventional filter removal is to avoid forceful extraction, which risks IVC laceration, hemorrhage, and thrombosis. Advanced techniques (hard guidewires, balloon assistance, LOOP technology, bronchial forceps) have increased retrieval rates to >90% but carry higher complication risks (8,9). Preoperative CT venography is essential to assess IVC patency, thrombus, and the spatial relationship of filter components with the IVC and abdominal aorta (10). Procedures should be performed under local anesthesia, using the patient’s pain perception to guide mechanical force.

Explanations of findings

Technical points specific to these two cases: (I) sheath modification: an 18F vascular sheath core was truncated proximally to accommodate four 0.035-inch guidewires, forming double or single LOOP configurations. For Case 1 (Simon Nitinol filter), the LOOP point was selected at the junction of the upper and lower filter segments, leveraging the filter’s structural weakness at this connection to enable inversion. For Case 2 (Günther Tulip filter), the LOOP point was at the proximal left foot, allowing controlled flipping while monitoring the foot-to-aorta distance via pigtail catheter marking. (II) Force application: the sheath core was advanced to the LOOP point to direct force onto the filter itself rather than the IVC wall, preventing vessel laceration. (III) Balloon dissection: a 12–14 mm balloon was introduced through the gap between the filter foot and IVC intima, inflated to 6–8 atm to achieve gradual peeling, with real-time adjustment based on arterial roadmap guidance to prevent aortic injury in Case 2.

In Case 1, the structural characteristics of this filter were cleverly used to flip and remove the filter using balloon peeling and LOOP technology. In Case 2, one foot of the filter was close to the abdominal aorta, increasing the risk. In such cases, the pigtail catheter should be located properly, and balloon closure and covered stent implantation should be performed if necessary. When using the intimal peeling and LOOP methods, the clinician must pay attention to the direction of force to avoid damage to the abdominal aorta. After the filter foot or the recovery hook is pulled into the IVC, it can be removed with the conventional methods.

With the continuous advancement of endovascular intervention technology, the indications for filter removal have also been updated. Through long clinical practice, we have mastered a set of effective and safe interventional methods that have broken through the technical barriers of intervention for rare filter complications and solved the problem of surgical filter removal. Of course, the implementation of these techniques requires a skilled endovascular intervention technology center to avoid serious complications.

Nevertheless, the primary risks of this technique include: (I) IVC laceration: LOOP technology concentrates force on the filter itself rather than the vascular wall, significantly reducing this risk; (II) aortic injury: in Case 2, the filter foot was adjacent to the abdominal aorta; we utilized intra-arterial pigtail catheter real-time marking to ensure balloon dissection direction away from the aorta; (III) exacerbation of duodenal perforation: preoperative fasting, gentle intraoperative manipulation, and postoperative antibiotic therapy reduce this risk; (IV) filter fracture: preoperative implantation of a protective filter (Denali) prevents fragment embolization.

Conclusions

This case report demonstrates the technical feasibility and safety of this novel endovascular approach in two high-risk patients with duodenal-penetrating IVC filters, including a permanent filter and one with an IVC-penetrating recovery hook. While promising, these findings are based on a limited case series and require validation in larger cohorts with longer follow-up before widespread adoption. The technique may serve as a minimally invasive alternative to open surgery in specialized interventional centers with appropriate expertise.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the CARE reporting checklist. Available at https://acr.amegroups.com/article/view/10.21037/acr-2026-0053/rc

Peer Review File: Available at https://acr.amegroups.com/article/view/10.21037/acr-2026-0053/prf

Funding: This study was supported by Chongqing Jiangbei District Science & Health Joint Medical Research Project (No. 2025JBKWLH006).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://acr.amegroups.com/article/view/10.21037/acr-2026-0053/coif). All authors report that this study was supported by Chongqing Jiangbei District Science & Health Joint Medical Research Project (No. 2025JBKWLH006). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All procedures performed in this study were in accordance with the ethical standards of the institutional and/or national research committee(s) and with the Declaration of Helsinki and its subsequent amendments. Written informed consent was obtained from the patients for the publication of this case report and accompanying images. A copy of the written consent is available for review by the editorial office of this journal.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Rodriguez AK, Goel A, Gorantla VR. Complications Associated With Inferior Vena Cava Filter Retrieval: A Systematic Review. Cureus 2024;16:e55052. [Crossref] [PubMed]
2. Lee JS, Hwang JK, Park SC, et al. Surgical removal of an inferior vena cava filter with duodenal penetration. Interact Cardiovasc Thorac Surg 2019;28:487-8. [Crossref] [PubMed]
3. Capasso K, Awad NA, Alvarez N, et al. Urinary excretion after transcaval renal penetration of a fragmented Bird’s Nest filter. J Vasc Surg Venous Lymphat Disord 2021;9:254-7. [Crossref] [PubMed]
4. Slyko R, Vulev I, Smoter I. Acute Iatrogenic Distal Aortic Pseudoaneurysm Caused by Metallic Arm of Displaced Vena Cava Filter—Case Report and Review of the Literature. Bratislava Medical Journal 2025;126:317-321.
5. Kuo WT, Doshi AA, Ponting JM, et al. Laser-Assisted Removal of Embedded Vena Cava Filters: A First-In-Human Escalation Trial in 500 Patients Refractory to High-Force Retrieval. J Am Heart Assoc 2020;9:e017916. [Crossref] [PubMed]
6. Owji S, Lu T, Loh TM, et al. Robotic-Assisted Inferior Vena Cava Filter Retrieval. Methodist Debakey Cardiovasc J 2017;13:34-6. [Crossref] [PubMed]
7. Saeed H, Buxey K, Fitzgerald M. Caval wall injury secondary to inferior vena cava filter removal: a rare complication. ANZ J Surg 2018;88:240-2. [Crossref] [PubMed]
8. Al-Hakim R, Kee ST, Olinger K, et al. Inferior vena cava filter retrieval: effectiveness and complications of routine and advanced techniques. J Vasc Interv Radiol 2014;25:933-9; quiz 940. [Crossref] [PubMed]
9. Stavropoulos SW, Chen JX, Sing RF, et al. Analysis of the Final DENALI Trial Data: A Prospective, Multicenter Study of the Denali Inferior Vena Cava Filter. J Vasc Interv Radiol 2016;27:1531-1538.e1. [Crossref] [PubMed]
10. Gong M, He X, Gu J. A more appropriate modality may be desired for the measurement of inferior vena cava filter position. J Vasc Surg Venous Lymphat Disord 2024;12:101735. [Crossref] [PubMed]