Long-term Retrieval Success Rate Profile for the Günther Tulip Vena Cava Filter

Article Outline

• Abstract
• Materials and Methods
  • Patient Population
  • Device Description
  • Filter Placement
  • Filter Retrieval
  • Patient Follow-up
  • Data Analysis
• Results
  • Filter Placement
  • Filter Retrieval
  • Patient Follow-up
  • Mortality
  • Adverse Events
• Discussion
• Acknowledgments
• References
• Copyright

Purpose

To evaluate the likelihood of successful retrieval of the Günther Tulip vena cava filter after various implant durations (up to 494 days).

Materials and Methods

Retrievable Günther Tulip filters were placed in 554 patients. All patients satisfied requirements for filter placement; the primary indication for placement was specified for 394 patients (71%), as follows: unspecified trauma (n = 164), bariatric procedures (n = 128), orthopedic procedures (n = 36), and other (n = 66). Filter tilt and vena cava injury were assessed at implantation. Filters were not repositioned after placement. At retrieval, filter orientation, vena cava injury, other device-related incidents, and the degree of difficulty associated with retrieval were reported.

Results

Filter retrieval was attempted in 275 patients and successful in 248 (90.2%). The mean filter indwell time was 58.9 days (range, 3–494 days). Unsuccessful retrievals (n = 27) were attributed primarily to improper hook orientation (n = 10) or excessive tissue in-growth at the filter legs (n = 16). Of the remaining 279 patients, 223 withdrew from the study, 41 were associated with a decision to keep the filter as a permanent device, 13 died for reasons unrelated to the study, and two had no reported endpoint data. A Kaplan-Meier product-limit survival estimate revealed that the probability of successful device retrieval remained greater than 94% at 12 weeks and greater than 67% at 26 weeks.

Conclusions

This study contributes to the body of clinical data related to retrievable filters, demonstrating reliable retrieval rates at 12 weeks, with successful retrievals up to 17 months after implantation.

Abbreviations: DVT, deep venous thrombosis, IVC, inferior vena cava, PE, pulmonary embolism

PERMANENT inferior vena cava (IVC) filters have been used in clinical practice since the 1970s (1, 2); these devices are indicated for use in patients with contraindications to or failure of anticoagulant therapy. Optional IVC filters are desirable in circumstances where patients need only temporary treatment against pulmonary embolism (PE), such as trauma, pregnancy, or surgeries that would be complicated by anticoagulant therapy. These retrievable devices have been investigated since the mid-1980s (3), as they are assumed to avoid the potential long-term risks associated with permanent filters, such as filter migration, strut fracture, IVC perforation, IVC occlusion, or recurrent deep vein thrombosis (DVT) (4, 5).

Vena cava filters have proved effective for the prevention of recurrent PE when anticoagulant therapy is contraindicated (2). The Günther Tulip vena cava filter (manufactured by William Cook Europe [ApS Bjaeverskov, Denmark] and sold in the United States by Cook Medical [Bloomington, Indiana]) has proved safe and effective in this regard (6). The U.S. Food and Drug Administration first cleared the Günther Tulip filter for permanent placement in 2000 and cleared the device for use as a retrievable filter in 2003. Early animal studies showed successful retrieval at up to 14 days after implantation (6). Similarly, results of the first registry study were favorable; Millward et al (7) reported a 98% retrieval rate for filters with implantation times of 2–25 days (mean, 9 days). Despite this early success and continued successful studies on retrievability, the true time frame for successful clinical retrievals remains undefined. This information is important, as the possibility of longer indwell times offers protection for patients in whom PE remains a risk beyond 2 weeks after implantation.

In addition to the Günther Tulip vena cava filter, there are three other IVC filters cleared by the Food and Drug Administration with permanent and retrieval indications: the Recovery G2 filter (Bard Peripheral Vascular, Tempe, Arizona), the OptEase filter (Cordis Endovascular, Miami Lakes, Florida), and the Celect filter (William Cook Europe, Bjaeverskov, Denmark). Long-term clinical data establishing a time frame for successful retrieval of these devices would assist physicians weighing the risks and benefits associated with attempted device retrieval. Specifically, the probability of safe retrieval as related to indwell time is important when choosing a retrievable filter for individual patients. With regard to the Günther Tulip vena cava filter, data are available regarding successful long-term retrieval in smaller clinical studies (8, 9, 10, 11), and several case reports (12, 13) describe successful retrievals after hundreds of days. However, no large-scale study has provided long-term data on successful retrieval versus indwell time for the Günther Tulip vena cava filter. Therefore, the objective of this prospective study was to determine the probability of retrieval success with respect to filter indwell time.

Back to Article Outline

Materials and Methods 

Patient Population 

Between March 2005 and July 2006, 554 patients at 25 U.S. institutions underwent placement of a Günther Tulip vena cava filter. The study group included 262 men and 292 women with a mean age (±standard deviation) of 50 years ± 17 (See Table 1 for patient demographics). Patients were prospectively enrolled in the study after satisfying the study inclusion criteria and none of the exclusion criteria (Table 2, Table 3, respectively). Before study enrollment, each patient signed an informed consent form and satisfied at least one of the criteria for filter placement (Table 4, Table 5). The patient's involvement in the study ended 14 days after retrieval or once the patient had reached another study endpoint (ie, withdrawal, death, successful filter retrieval, unsuccessful filter retrieval, decision not to retrieve the filter). Study sites were queried until each enrolled patient had reached a study endpoint. The study was conducted under approval of the investigational research board at each institution and advised by an independent data safety monitoring board.

Table 1. Summary of Patient Demographics

Parameter	No. of Patients (n = 554)
Sex
M	262(47.3)
F	292(52.7)
Comorbidity
Hypertension	214(38.6)
DVT	135(24.4)
Gastrointestinal disease	108(19.5)
Diabetes type 2	105(19.0)
PE	84(15.2)
Pulmonary disease	79(14.3)
Cancer	69(12.5)
Coronary artery disease	48(8.7)
Stroke	33(6.0)
Coagulation disorder	30(5.4)
Congestive heart failure	24(4.3)
Renal disease	23(4.2)
Peripheral disease	21(3.8)
Hepatic disease	15(2.7)
Diabetes type I	14(2.5)

Note.—Patients ranged in age from 18 to 92 years (mean age, 50 years ± 17). Numbers in parentheses are percentages.

Table 2. Inclusion Criteria

1. Patient is at high risk for PE, but placement of a permanent IVC filter is not likely to be required and falls into one of the following categories.

A. Surgical patients undergoing procedures associated with high risk for venous thromboembolism;

B. Patients with a documented DVT or PE with temporary contraindications to anticoagulation therapy;

C. Patients with documented DVT or PE who must temporarily interrupt anticoagulation therapy;

D. Patients with documented DVT or PE who are at high risk for PE during initiation of anticoagulation; or

E. Trauma patients.

2. Patient must have a patent internal jugular vein.

3. Patient or legal guardian must have signed informed consent document.

4. Patient agrees to return for recommended retrieval visit and comply with 14-day post-retrieval phone follow-up.

Table 3. Exclusion Criteria

Table 4. Satisfaction of Requirements for Vena Cava Filter Placement

Table 5. Primary Indications for Filter Placement

Indication for Placement	No. of Patients (n = 554)	No. of Retrieval Attempts	No. of Successful Retrievals	Mean Indwell Time (d)
Unspecified trauma	164(29.6)	70	63(90)	93.0
Bariatric procedure	128(23.1)	99	94(95)	53.6
Orthopedic procedure	36(6.5)	17	13(76)	60.8
Other⁎	66(12)	22	20(91)	55.2
Reason not provided	160(28.8)	67	58(86)	53.1

Note.—Numbers in parentheses are percentages.

Device Description 

The Günther Tulip vena cava filter has been described previously (8, 14). Briefly, the filter deployment set consists of a preloaded filter introducer, a coaxial introducer sheath system (8.5 F), a hydrophilic-coated dilator, and a three-way stopcock. The Günther Tulip vena cava filter retrieval set (described previously by Millward et al [14]) consists of a retrieval loop system with a braided platinum wire loop, a coaxial retrieval sheath system (11 F), an entry needle, a wire guide, and a dilator. The filter can be placed with either a femoral or jugular approach, but retrieval must occur by means of a jugular approach.

Filter Placement 

Patients were prepared for the procedure according to the standard of care at each institution, and all implant procedures were performed according to the instructions for use. Briefly, an angiographic catheter or the 8 F introducer sheath was placed at the confluence of the common iliac veins and a venogram obtained in the anteroposterior view. Filters were placed by using either a jugular or femoral approach (15) with the coaxial introducer sheath system. The details associated with filter placement are summarized in Table 6. A postdeployment venogram was obtained to verify proper placement and assess for vena cava injury and filter tilt (the degree of filter tilt was categorized as follows: 0°, 0°–5°, 6°–10°, 11°–15°, and 16°+). The delivery sheath was removed and hemostasis achieved according to the standard of care. After verification of proper placement, the filter was left undisturbed for the duration of the implantation.

Table 6. Filter Implantation Details

Parameter	No. of Implantations (n = 554)
Route of filter introduction
Right internal jugular vein	257(46.4)
Right common femoral vein	249(44.9)
Left common femoral vein	47(8.5)
Left internal jugular vein	1(0.2)
Filter placement site
Infrarenal IVC	547(98.7)
Suprarenal IVC	6(1.1)
Not specified	1(0.2)
Caval injury	1(0.2)
Hematoma at vascular access site	0(0.0)
Filters not deployed due to a defect	3(0.5)

Note.—Numbers in parentheses are percentages.

Filter Retrieval 

Filter retrieval was attempted when deemed clinically appropriate by the investigating physician. A pre-retrieval vena cavogram was obtained to assess for possible filter migration, tilt, thrombus, injury to the IVC, or other events potentially related to the filter. Filter retrieval was not attempted in patients with trapped thrombi occupying more than 25% of the filter's cone volume, in patients at continued high risk for PE, or in patients with a poor overall prognosis. The Günther Tulip vena cava filter retrieval set was recommended for filter retrieval, although retrieval with other snares was not prohibited. With each retrieval attempt, the technical difficulty associated with retrieval was recorded by the physician as none, minimal, moderate, major, not possible, or not assessed. A successful retrieval was defined as one in which the filter was retrieved with none of the following events occurring within 14 days of the retrieval: acute IVC injury requiring intervention, hemorrhage, fracture of the retrieved filter or the retrieval set, or injury at the vascular access site. In the event that a filter could not be retrieved, the cause of failure was provided, vena cavograms were obtained in two orthogonal views, and the filter was left as a permanent device. All adverse events that occurred before and after retrieval were recorded and the device's role in each was categorized as not related (due to an underlying or concurrent illness or effect of another device, not related to study device or procedure), probably not related (little or no temporal relationship to the study device or procedure), possibly related (strong temporal relationship to the study device or procedure and alternative cause is equally or less likely compared to the potential relationship to study device), or probably related (strong temporal relationship to the study device or procedure and another cause is unlikely or significantly less likely).

Patient Follow-up 

Postprocedure care was provided according to the standard of care at each institution. Adverse events that occurred while the filter was in place and during at least a 2-week post-retrieval period were documented and assessed for their relationship to the device or the retrieval procedure. This information was collected from either the filter retrieval data form or a follow-up phone conversation with the patient approximately 14 days after retrieval.

Data Analysis 

Five hundred fifty-four patients were enrolled in this study; 275 of these patients underwent retrieval attempts. The Kaplan-Meier product limit method was used to estimate the probability of a successful retrieval across the time frame encompassed by the study, that is, retrieval from 0 to 494 days.

Back to Article Outline

Results 

Filter Placement 

All 554 filters were placed. Five hundred twenty-eight filters were placed by using fluoroscopic guidance and venography. Twenty-six filters were placed by using ultrasonographic guidance; 22 of these procedures were associated with trauma patients. Additional details regarding filter placement are summarized in Table 6. At placement, one patient had evidence of IVC injury; however, there were no reported complications associated with this patient during follow-up. No patients had evidence of hematoma at the vascular access site.

No evidence of tilt (ie, 0°) was detected in 345 of the 554 filters (62.3%) at placement. Filter tilt was detected at implantation in the remaining 209 filters (37.7%). Data pertaining to the frequency and extent of filter tilt immediately after implantation and at retrieval are presented in Table 7.

Table 7. Filter Tilt at Implantation and Retrieval

Filter Tilt	At Implantation		At Retrieval (n = 275)
	All Patients (n = 554)	Patients with Subsequent Retrieval Attempts (n = 275)
0°	345(62.3)	182(66.2)	191(69.5)
0°–5°	76(13.7)	35(12.7)	28(10.2)
6°–10°	69(12.5)	31(11.3)	23(8.4)
11°–15°	39(7.0)	16(5.82)	17(6.2)
>16°	25(4.5)	11(4.0)	13(4.7)
Not reported	0(0)	0(0)	3(1.1)

Note.—Numbers in parentheses are percentages.

Filter Retrieval 

Filter retrieval was attempted in 275 of the 554 patients (49.6%); the details associated with retrieval attempts are provided in Table 8. Retrieval attempts occurred between 0 and 494 days after placement. Overall, 248 of the 275 retrieval attempts (90.2%) were successful, with a mean indwell time of 58.9 days. Eight patients later underwent a second retrieval attempt due to an unsuccessful first attempt. Of these, five of the second attempts were successful and three were unsuccessful. In all, retrieval attempts were unsuccessful in 27 patients due primarily to an inability to remove the filter without excessive force, associated with excessive tissue in-growth at the filter legs (n = 16 [59.3%]) or the inability to grasp the hook due to its orientation (ie, tilt) toward the vessel wall (n = 10 [37%]). The other reason cited for nonretrieval was a small clot in the filter (n = 1 [3.7%]). Filters were left as permanent devices in these patients. The mean filter indwell time at the retrieval attempt for the group of patients with unsuccessful retrieval attempts was 114 days (range, 21–206 days). Sixty-five percent (17/26) of filters associated with unsuccessful retrievals were in place for more than 3 months, and 23% (6/26) were in place for more than 6 months; the indwell time was not provided in one case. One hundred seventy-two of the 275 retrieval attempts were made with the Günther Tulip retrieval set. Failed retrievals were never attributed to the filter retrieval set.

Table 8. Filter Retrieval Details

Parameter	No. of Retrieval Attempts (n = 275)
Route of filter retrieval
Right jugular vein	271(98.5)
Left jugular vein	4(1.5)
Technical difficulty with retrieval
None	165(60.0)
Minimal	44(16.0)
Moderate	30(10.9)
Major	6(2.2)
Not possible	26(9.4)
Not assessed	2(0.7)
Not reported	2(0.7)

Note.—Numbers in parentheses are percentages.

A decision not to retrieve the filter was made for the following reasons: 31 patients had an ongoing high risk for PE, five had an overall poor prognosis, two had a substantial amount of trapped thrombus in the filter (>25% of the cone volume), and three did not have reasons provided. For these 41 patients, this decision was a study endpoint.

A Kaplan-Meier product limit estimate for the success of filter retrieval is presented in the Figure. This plot predicts the probability of successful retrieval on the basis of indwell time. From this analysis, the estimated probability of successful retrieval for the Günther Tulip filter is greater than 99% at 4 weeks, 94% at 12 weeks, 67% at 26 weeks, and 37% at 52 weeks after implantation.

• 
  • View Large Image
  • Download to PowerPoint

• Figure. 

  Kaplan-Meier analysis for retrieval success. The overall filter retrieval success rate was 91%. The probability of successful retrieval remained greater than 99% 4 weeks after implantation, greater than 94% 12 weeks after implantation, greater than 67% 26 weeks after implantation, and greater than 37% 52 weeks after implantation.

Patient Follow-up 

A total of 223 patients withdrew from the study for the following reasons: 172 chose to withdraw from the study after 6 months, assuming their filter would not be retrieved, and 51 left the study on the recommendation of their doctor or on their own choosing. Of these 51 patients, eight patients decided to end their participation in the study (seven patients did not want to undergo another procedure, ie, the retrieval procedure; one patient did not want to continue to contribute to study data), one patient had the filter retrieved at a hospital outside of the participating institutions and was withdrawn, one patient did not show for a scheduled retrieval, one patient was unable to return for a retrieval attempt, eight patients had not yet fully recovered from their procedures or were having additional operations in the future and retained their filters, and 31 patients retained their filters as permanent devices after an assessment of their overall medical condition (a ‘decision not to retrieve’ data form was not completed for these patients). Data on two patients removed from the study were not provided.

Mortality 

Thirteen patients died while enrolled in the study; no death was device-related.

Adverse Events 

All medical events for each patient were documented throughout this study, regardless of whether the event was thought to be device-related. Pre-retrieval events reported in this study included, but were not limited to, surgical procedures, pneumonia, arrhythmia, and transfusions. A total of 385 pre-retrieval events were reported; of these, none were reported to be related to the filter. These events included three incidents of PE and one vascular access site complication; following the initial observation of these events and appropriate treatment, there were no further complications with these patients. There were 22 post-retrieval events reported during the study; of these, only two (9%) were categorized as ‘probably’ related to the filter. One patient whose post-retrieval event was classified as ‘probably’ related to the filter had 27% stenosis in the vena cava at retrieval, but was asymptomatic, and a repeat venogram obtained approximately 2 months later showed resolution of the stenosis. A second patient had a suspected small PE immediately after retrieval evidenced by complaints of chest pain, which resolved spontaneously shortly after. After retrieval, this patient also had a vena cava injury associated with thrombus and was hospitalized overnight to initiate anticoagulation therapy and for observation.

Back to Article Outline

Discussion 

This large (n = 554) prospective study provides a detailed profile of retrieval success over time for a retrievable vena cava filter. In 275 patients, attempts were made to retrieve a Günther Tulip filter between 0 and 494 days (71 weeks) after the initial implantation. A Kaplan-Meier plot (Figure) developed from these results provides physicians with a valuable clinical tool to estimate the likelihood of successful filter retrieval at various times after implantation.

Limited previous clinical evidence has suggested that the filter indwell time for the Günther Tulip filter could be safely extended beyond 14 days (8, 9, 10, 11, 16). Terhaar et al (8) reported on filter retrieval attempts in 19 patients with a mean filter indwell time of 43.6 days (range, 7–126 days). Retrieval was successful in 16 patients (84.2%). Unsuccessful retrievals in their study were attributed to filter adherence to the IVC wall (n = 1) and thrombus within the filter (n = 2). Hoppe et al (16) reported a technical and clinical success rate of 100% for retrieval efforts in 23 patients with Günther Tulip filters. The mean indwell time, however, was only 11.1 days (range, 2–14 days). Difficulty with retrieval occurred in 29.1% of the patients and was associated with filter migration and tilting. De Gregorio et al (9) reported on retrieval rates approximately 30 days after implantation (range, 29–34 days). Of 32 filter retrieval attempts, 31 (97%) were successful; most (79%) were performed without difficulty. In a similarly sized study, Looby et al (10) reported on 45 patients who had Günther Tulip filters implanted with the intent of removal. The mean indwell time for filters in that study was 33.6 days (range, 14–112 days), and successful retrievals were performed in 36 of the 45 patients (80%). The nine unsuccessful retrieval attempts were classified as follows: two patients with clinical indications for a permanent filter, five patients in whom the hooks of the struts were embedded in the wall of the IVC and the force required for retrieval was considered too high, and two patients in whom the hook on the proximal end of the filter was tilted against the vena cava wall and could not be grasped.

A larger study by Rosenthal et al (11) of 177 patients with multiple trauma investigated the feasibility of Günther Tulip filter retrieval after indwell times of more than 180 days. Forty-one patients retained their filters for more than 180 days; the mean indwell time for this group was 261.5 days (range, 182–403 days). Within this group, 31 filters (75.6%) were retrieved uneventfully, whereas 10 were irretrievable (all were associated with some degree of tilt) and remained in place as permanent devices. Another subset of this study was a group of 64 patients who retained their filters for less than 180 days; the mean indwell time for this group of patients was 51 days (range, 42–180 days), and the retrieval success rate was 93.7%.

In addition to these clinical studies, individual case reports have documented successful retrieval of the Günther Tulip filter at times up to 475 days after implantation. Binkert et al (12) reported on a 22-year-old patient who received a Günther Tulip filter following trauma injuries. Successful filter retrieval occurred 317 days after implantation with some increased level of force. In a separate case, Kachura (13) reported on a 33-year-old patient who received a Günther Tulip filter for ileofemoral DVT extending into the lower IVC. At 475 days after implantation, moderate filter incorporation into the caval wall was evident, but the device was successfully retrieved.

Together, these studies suggest that safe retrieval of the Günther Tulip filter can occur well beyond 14 days. The present study found that the probability of successful filter retrieval remained greater than 99% at 4 weeks, greater than 94% at 12 weeks, greater than 67% at 26 weeks, and greater than 37% at 52 weeks after implantation (Figure). Overall, the degree of technical difficulty (Table 8) associated with retrieval in the present study was similar to or slightly improved compared to that reported by De Gregorio et al (9) and Hoppe et al (16), although each reported on a smaller patient population (n = 35 and n = 41, respectively) with shorter indwell times (range, 29–34 days and 2–14 days, respectively). The results of this study indicate that the likelihood of successful filter retrieval is quite stable for the first 12 weeks; retrieval became progressively more challenging after 12 weeks and seemed to stabilize at a success rate of approximately 40% after 26 weeks. This finding is supported by the retrieval results presented by Rosenthal et al (11) in their study of patients with multiple trauma. Successful retrievals after this time are possible; however, as indwell time increases, the option of the filter remaining as a permanent implant must be weighed against the likelihood of successful retrieval. This point was emphasized by Kachura (13) in his recent case study report. He suggested that retrieval decisions should be patient-specific and must be based on the risk of filter permanence versus the risks associated with delayed retrieval.

In the present study, filter retrieval attempts were unsuccessful in 27 patients (9.8%), due primarily to excessive tissue in-growth or filter tilt. The explanations for failed retrievals in the present study are consistent with those from other clinical studies of the Günther Tulip filter (8, 9, 10). This small group of patients also had a longer mean indwell time than the overall study population; the mean indwell time for this group of patients was 114 days (range, 21–206 days), and most (65%) of these retrieval attempts occurred after a 3-month indwell period.

Filter tilt is assessed in most clinical vena cava filter studies because it may affect both function and retrieval success. Specifically, a tilted filter may have a reduced ability to trap clots (17, 18), although more recent studies with the Günther Tulip filter have found that this difference is not statistically significant (19). In addition, tilt may inhibit retrieval by preventing snaring of the hook. In the present study, filter tilt was assessed at both implantation and retrieval (Table 7). Interestingly, by the time of filter retrieval, more filters had shifted into categories of lesser tilt (n = 9) than into categories of higher tilt (n = 3). At retrieval, only 81 of 275 filters had some degree of tilt, which is a low occurrence rate compared to a recent report by Sag et al (20) in which 159 of 175 filters had some degree of tilt at retrieval. In the 27 unsuccessful retrieval attempts in the present study, only eight (29.6%) were attributed to hook orientation toward the vena cava wall (ie, filter tilt). In these cases, filter tilt was reported to be 6°–10° (n = 3), 11°–15° (n = 1), or 16°+ (n = 3); data were not provided in one case. Recent studies by Ray et al (21) and Looby et al (10) have also documented failed Günther Tulip filter retrieval attempts due to tilt. Ray et al (21) attributed 50% of their failed retrieval attempts to technical difficulties (which included both tilt and excessive tissue growth), and Looby et al (10) attributed 22% of their failed attempts to tilt.

Tissue in-growth at the site of a filter may also contribute to failed retrievals (22, 23); this growth is likely the result of a biologic process termed transmural incorporation, which occurs slowly (weeks to months) (23). Excessive tissue growth was reported as contributing to failed retrieval efforts in more than half (59.3%) of the failed retrieval attempts in the present study, as these filters could not be retrieved without excessive force. The mean indwell time for these filters was 122.5 days (range, 60–198 days), which is consistent with the timeframe for transmural incorporation. This growth does not appear to be related to filter tilt; only four of 17 failed retrievals attributed to excessive tissue growth were also associated with some degree of filter tilt. These data are similar to those reported by Looby et al (10), who attributed 55% of their failed retrievals to excessive tissue growth at the site of the filter.

Importantly, the manufacturer cautions against the use of excessive retrieval force. In a 32-patient study, de Gregorio et al (24) measured the force required to retrieve filters following an average indwell period of 30 days. The authors suggested that the force required to retrieve the filters ranged from 1.96 N to 9.8 N and indicated that if the required forces are greater than 9.8 N it may not be possible to retrieve a filter. However, physicians continue to develop and report on their own techniques to retrieve filters embedded in the vena cava wall as a result of filter tilt and neointimal growth. For instance, Burke et al (25) described successful retrieval by using rigid bronchoscopy foreps to free filters from the vena cava wall, and Van Ha et al (26) recently reported on four different nonroutine retrieval techniques to assist with difficult retrievals. In deciding to attempt retrieval with these techniques, physicians must weigh the potential for vascular injury against the benefits of filter retrieval, especially considering that the filter may be used as a permanent device.

It has long been suspected that the implant time of the Günther Tulip filter could be safely extended without additional risk to the patient. Retrieval data presented herein are both informative and encouraging in this regard, as the success rate for retrieval remained greater than 94% at 12 weeks (Figure). These data continue to establish the Günther Tulip filter as a safe, reliable, and versatile device that may be implanted temporarily, or permanently, for the prevention of PE. These results may also be useful for establishing a basis for the expectations associated with other retrievable/optional filters and may be used as a benchmark against which to compare the results of other trials with similar parameters. Furthermore, these data are valuable clinically for weighing the probability of successful filter retrieval (estimated from implant duration) versus the overall medical condition of each patient.

Back to Article Outline

Acknowledgments 

The authors thank Scott Snyder, PhD, and Shraddha Mehta for their assistance with statistical analyses, Leslie Rooze for assistance in data collection and monitoring, and Elizabeth Basquin Krause, MD, and Lorraine Van Meter for their assistance in preparing the manuscript.

Back to Article Outline

References 

1. Mobin-Uddin K, McLean R, Bolooki H, Jude JR. Caval interruption for prevention of pulmonary embolism: long-term results of a new method. Arch Surg. 1969;99:711–715
   • View In Article
   • MEDLINE
2. Greenfield LF, McCurdy JR, Brown PP, Elkins RC. A new intracaval filter permitting continued flow and resolution of emboli. Surgery. 1973;73:599–606
   • View In Article
   • MEDLINE
3. Hunter DW, Lund G, Rysavy JA, et al. Retrieving the Amplatz retrievable vena cava filter. Cardiovasc Intervent Radiol. 1987;10:32–36
   • View In Article
   • MEDLINE
   • CrossRef
4. Kaufman JA. Filter placement in deep venous thrombosis. Am J Roentgenol. 1996;166:457–458
   • View In Article
5. Kinney TB. Update on inferior vena cava filters. J Vasc Interv Radiol. 2003;14:425–440
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (2280 KB)
   • CrossRef
6. Neuerburg JM, Handt S, Beckert K, et al. Percutaneous retrieval of the Tulip vena cava filter: feasibility, short- and long-term changes—an experimental study in dogs. Cardiovasc Intervent Radiol. 2001;24:418–423
   • View In Article
   • MEDLINE
   • CrossRef
7. Millward SF, Oliva VL, Bell SD, et al. Günther Tulip retrievable vena cava filter: results from the registry of the Canadian interventional radiology association. J Vasc Interv Radiol. 2001;12:1053–1058
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (457 KB)
   • CrossRef
8. Terhaar OA, Lyon SM, Given MF, Foster AE, McGrath F, Lee MJ. Extended interval for retrieval of Günther Tulip filters. J Vasc Interv Radiol. 2004;15:1257–1262
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (1263 KB)
   • CrossRef
9. De Gregorio MA, Gamboa P, Bonilla DL, et al. Retrieval of Günther Tulip optional vena cava filters 30 days after implantation: a prospective clinical study. J Vasc Interv Radiol. 2006;17:1781–1789
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (1857 KB)
   • CrossRef
10. Looby S, Given MF, Geoghegan T, McErlean A, Lee MJ. Günther Tulip retrievable inferior vena caval filters: indications, efficacy, retrieval and complications. Cardiovasc Intervent Radiol. 2007;30:59–65
    • View In Article
    • MEDLINE
    • CrossRef
11. Rosenthal D, Wellons ED, Hancock SM, Burkett AB. Retrievability of the Günther Tulip vena cava filter after dwell times longer than 180 days in patients with multiple trauma. J Endovasc Ther. 2007;14:406–410
    • View In Article
    • CrossRef
12. Binkert CA, Bansal A, Gates JD. Inferior vena cava filter removal after 317-day implantation. J Vasc Interv Radiol. 2005;16:395–398
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (2370 KB)
    • CrossRef
13. Kachura JR. Inferior vena cava filter removal after 475-day implantation. J Vasc Interv Radiol. 2005;16:1156–1158
    • View In Article
    • Full Text
    • Full-Text PDF (857 KB)
    • CrossRef
14. Millward SF, Bhargava A, Aquino J, et al. Günther Tulip filter: preliminary clinical experience with retrieval. J Vasc Interv Radiol. 2000;11:75–82
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1792 KB)
    • CrossRef
15. Neuerburg JM, Günther RW, Vorwerk D, et al. Results of a multicenter study of the retrievable tulip vena cava filter: early clinical experience. Cardiovasc Intervent Radiol. 1997;20:10–16
    • View In Article
    • MEDLINE
    • CrossRef
16. Hoppe H, Nutting CW, Smouse HR, et al. Günther Tulip filter retrievability multicenter study including CT follow-up: final report. J Vasc Interv Radiol. 2006;17:1017–1023
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (871 KB)
    • CrossRef
17. Katsamouris AA, Waltman AC, Delichatsios MA, Athanasoulis CA. Inferior vena cava filters: in vitro comparison of clot trapping and flow dynamics. Radiology. 1988;166:361–366
    • View In Article
    • MEDLINE
18. Günther RW, Schild H, Fries A, Störkel S. Vena caval filter to prevent pulmonary embolism: experimental study (Work in progress). Radiology. 1985;156:315–320
    • View In Article
    • MEDLINE
19. Hammer FD, Rousseau HP, Joffre FG, Sentenac BP, Tran-Van T, Barthelemy RP. In vitro evaluation of vena cava filters. J Vasc Interv Radiol. 1994;5:869–876
    • View In Article
    • Abstract
    • Full-Text PDF (3264 KB)
    • CrossRef
20. Sag AA, Stavas JM, Burke CT, Dixon RG, Marquess JS, Mauro MA. Analysis of tilt of the Günther Tulip filter. J Vasc Interv Radiol. 2008;19:669–676
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (238 KB)
    • CrossRef
21. Ray CE, Mitchell E, Zipser S, Kao EY, Brown CF, Moneta GL. Outcomes with retrievable inferior vena cava filters: a multicenter study. J Vasc Interv Radiol. 2006;17:1595–1604
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1682 KB)
    • CrossRef
22. Simon M. Vena cava filters: prevalent misconceptions. J Vasc Interv Radiol. 1999;10:1021–1024
    • View In Article
    • Full-Text PDF (416 KB)
    • CrossRef
23. Proctor MC, Greenfield LJ, Cho KJ, Moursi MM, James EA. Assessment of apparent vena caval penetration by the Greenfield filter. J Endovasc Surg. 1998;5:251–258
    • View In Article
    • MEDLINE
    • CrossRef
24. De Gregorio MA, Gamboa P, Bonilla DL, et al. Retrieval of Günther Tulip optional vena cava filters 30 days after implantation: a prospective clinical study. J Vasc Interv Radiol. 2006;17:1781–1789
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1857 KB)
    • CrossRef
25. Burke CT, Dixon RG, Stavas JM. Use of rigid bronchoscopic forceps in the difficult retrieval of the Günther Tulip inferior vena cava filter. J Vasc Interv Radiol. 2007;18:1319–1323
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (805 KB)
    • CrossRef
26. Van Ha TG, Vinokur O, Lorenz J, et al. Techniques used for difficult retrievals of the Günther Tulip inferior vena cava filter: experience in 32 patients. J Vasc Interv Radiol. 2009;20:92–99
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (2101 KB)
    • CrossRef