Accepted Manuscript Indications and outcomes of open inferior vena cava filter removal Kristofer M. Charlton-Ouw, MD, Shaikh Afaq, MD, Samuel S. Leake, BS, Harleen K. Sandhu, MD, MPH, Cristina N. Sola, RN, Naveed U. Saqib, MD, Ali Azizzadeh, MD, Hazim J. Safi, MD PII:
To appear in:
Annals of Vascular Surgery
Received Date: 21 March 2017 Revised Date:
15 May 2017
Accepted Date: 30 May 2017
Please cite this article as: Charlton-Ouw KM, Afaq S, Leake SS, Sandhu HK, Sola CN, Saqib NU, Azizzadeh A, Safi HJ, Indications and outcomes of open inferior vena cava filter removal, Annals of Vascular Surgery (2017), doi: 10.1016/j.avsg.2017.05.038. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Indications and outcomes of open inferior vena cava filter removal
Kristofer M Charlton-Ouw, MD1,2, Shaikh Afaq, MD1,2, Samuel S. Leake, BS1, Harleen K. Sandhu, MD, MPH1, Cristina N. Sola, RN1, Naveed U. Saqib, MD1,2, Ali Azizzadeh, MD1,2, Hazim J. Safi, MD1,2 Department of Cardiothoracic and Vascular Surgery, McGovern Medical School, University of Texas Health Science Center at Houston (UTHealth)
M AN U
6 7 8
Memorial Hermann Hospital System, Houston, Texas
Presented at the Vascular and Endovascular Surgery Society Annual Meeting 2017, Steamboat Springs, CO, USA.
16 17 18 19
The authors report no conflicts of interest with regard to this manuscript. Dr. Charlton-Ouw is a consultant with W.L. Gore and Associates. Dr. Azizzadeh is a consultant with W.L. Gore and Associates and Medtronic. No external funding sources were used in connection with this research.
21 22 23 24 25 26 27 28 29
Corresponding Author: Kristofer M. Charlton-Ouw, MD Department of Cardiothoracic and Vascular Surgery University of Texas Health Science Center at Houston 6400 Fannin St, Ste 2850 Houston, TX 77030 Office 713-486-5100 [email protected]
ACCEPTED MANUSCRIPT 2
Introduction: Despite recommendations for retrieval of inferior vena cava (IVC) filters, most are
not removed in a timely manner. Longer IVC filter dwell times are associated with caval wall
perforation and tilting that make percutaneous retrieval more difficult. Open IVC filter removal
is generally reserved for patients with symptoms referable to the filter, such as chronic back and
abdominal pain. We present our management algorithm and review of cases of open IVC filter
Methods: Patients referred for management of implanted IVC filters from May 2010 to May
2016 were included. Demographic and imaging were reviewed for cases requiring open surgical
Results: There were 221 percutaneous retrieval attempts in 218 patients. Successful retrieval
occurred in 196 (89%) attempts. There were 7 patients that had open surgical IVC filter removal
after failure of percutaneous retrieval. One patient had 2 filters and another had 3 filters. Except
for 1 case with complications during the percutaneous retrieval procedure, the remaining patients
all suffered from back or abdominal pain. All had significant filter strut penetration through the
caval wall into adjacent structures. Postoperatively, all patients had relief of pain. There were no
deaths and one patent had a minor ileus that spontaneously resolved.
Conclusions: Patients who fail percutaneous IVC filter retrieval can expect low morbidity and
prompt resolution of symptoms after open surgical removal via mini-laparotomy.
M AN U
ACCEPTED MANUSCRIPT 3
Introduction Retrievable inferior vena cava (IVC) filters were developed to protect patients from fatal pulmonary embolism and to allow for device removal once this risk abated. However, multiple
studies have shown that device retrieval occurs in <50% of cases.1, 2 This is most often due to
lack of referral into a comprehensive follow-up program. The benefit of routine filter retrieval
has recently been called into question and may not be cost-effective.3 The counter argument is
that as many as a quarter of patients do have complications related to long-term indwelling
filters, including IVC thrombosis, filter dislocation, recurrent pulmonary embolism, penetration
of filter struts through pericaval structures, and chronic pain.4, 5 The U.S. Food and Drug
Administration reaffirmed its recommendation in 2014 that retrievable IVC filters should be
removed when reasonable.
M AN U
Most filters can be removed percutaneously using standard techniques but occasionally require creative maneuvers. Assuming the transient risk of pulmonary embolism has passed, the
optimal time of IVC filter retrieval is estimated to be 1-2 months after implantation.6 This is
outside the 12-day retrieval window of the Cordis OPTEASE filter and many filters tend to
become more difficult to remove over time. This is usually due to strut penetrations through the
caval wall leading to tilting and scar formation. Especially with deep filter strut penetration and
severe misalignment, percutaneous retrieval may not be possible without undue risk. We noted
several cases where the filter was nearly sideways in the IVC or entangled with a second filter.
Open IVC filter removal can be done through a retroperitoneal or transperitoneal mini-
laparotomy approach. We describe our indications for open filter removal, technique, and
outcomes in 7 cases.
ACCEPTED MANUSCRIPT 4
Methods Patients referred for management of implanted IVC filters from May 2010 to May 2016
were included. From 2011 to 2015, patients referred for IVC filter management were enrolled in
a prospective registry after signing informed consent. Thereafter, patients were periodically
contacted to assess for ongoing need for the IVC filter. Since 2015, contacting patients regarding
their IVC filter was deemed the practice standard and need for signed informed consent was
waived. Retrospective research access to the database was approved by the Committee for the
Protection of Human Subjects, the local ethics institutional review committee. Demographic and
imaging information was reviewed for cases requiring open surgical removal. Durable resolution
of symptoms was ascertained by accessing outpatient follow-up records.
M AN U
A standard technique is applied with the initial goal of obtaining access to the retroperitoneum and IVC and to allow inspection of the surrounding structures. In each case, a
midline laparotomy incision is made approximately 10 cm in length depending on the patient’s
anatomy. The small bowels are packed to the right side and the retroperitoneum is entered to the
left of the duodenum. The duodenum is elevated off of the IVC, retracted to the patient’s right
side, and inspected for any penetration. If needed, the struts or filter tines are carefully removed
with repair of the bowel wall.
Control of the cava is with proximal and distal clamps after intravenous anticoagulation. However, inflammation and scarring around the cava can be significant making the dissection
ACCEPTED MANUSCRIPT 5
tedious. Direct pressure over the proximal and distal IVC with sponge sticks usually offers
adequate hemostasis. Once control is obtained, a longitudinal venotomy is made (Figure 1). An
endovenectomy is performed as needed to remove scar tissue surrounding the filters within the
lumen. It is important to keep in mind that the filter may embed itself in the renal vein or other
tributaries. Once the luminal surface is free of filter and scar, the venotomy is primarily repaired
with running 5-0 polypropylene suture (Figure 2). We find that a patch angioplasty is not usually
needed except in cases of chronic occlusion and scarring. After the repair, anticoagulation is
reversed and the laparotomy closed.
M AN U
There were 221 percutaneous retrieval attempts in 218 patients during the study period. Only 2% of the filters were placed by our group; the remainder were referred to us for retrieval.
Successful percutaneous retrieval occurred in 196 (89%) attempts. There were 7 patients that had
open surgical IVC filter removal after failure of percutaneous retrieval. One additional patient
required median sternotomy and cardiopulmonary bypass after the filter migrated to the caval-
atrial junction during retrieval attempt at another hospital. Most were referred to us especially for
open removal after unsuccessful percutaneous retrieval by other groups. One patient had 2 filters
and another had 3 filters. Except for 2 cases with complications during the percutaneous retrieval
procedure that were transferred to us, the remaining patients all suffered from back or abdominal
pain. Symptomatic patients all had significant filter strut penetration through the caval wall into
ACCEPTED MANUSCRIPT 6
Case 1 A 73-year-old man with a history of motor vehicle crash and deep vein thrombosis
(DVT) presented with scrotal and back pain. The patient was initially diagnosed and treated for
urinary tract infection. However, a computed tomography (CT) scan of the abdomen revealed
IVC thrombosis at the level of the renal arteries extending from two misaligned IVC filters
(Figure 3). A venogram demonstrated evidence of old organized thrombus in the iliac vein and
two IVC filters. An attempt to pass the wire through the IVC filter for thrombolysis failed. The
patient was started on rivaroxiban. The patient presented a few months later with persistent back
pain and worsening leg swelling despite anticoagulation. He underwent ultrasound-accelerated
thrombolysis with some reconstitution of the iliocaval veins. However, there was still significant
IVC thrombosis within the misaligned filters.
M AN U
After review of CT and venogram, we decided not to attempt another percutaneous
retrieval. A mini-laparotomy was performed and the two IVC filters were removed via a
longitudinal cavotomy. A caval endovenectomy was performed with primary repair of the vein
wall. Several filter barbs penetrating through the caval wall entered the duodenum requiring
suture repair of the small bowel. The patient was slow to recover bowel function but otherwise
recovered uneventfully. He reported resolution of back pain and improved, but not resolved, leg
swelling in office follow-up.
Case 2 A 44-year-old woman presented to interventional radiology for elective removal of an
ACCEPTED MANUSCRIPT 7
ALN IVC filter. The filter was initially placed for recurrent history of pulmonary embolism
while on anticoagulation. Due to tilting of the filter with subsequent penetration into the renal
vein the percutaneous attempts were not successful. The interventional radiologist was unable to
extract the filter using the ALN cone via the right internal jugular vein. An attempt was also
made using a gooseneck snare. The snare became entrapped within the filter barbs and the snare
could not be retrieved (Figure 4) despite multiple attempts. The filter and snare were removed
via mini-midline laparotomy. The snare was entwined within the filter and was cut with a wire
cutter. The remainder of the snare was then removed out of the right internal jugular vein. After
removal of the filter, the inferior vena cava was primarily repaired.
M AN U
A 50-year-old woman with a history of pulmonary embolism presented with severe
abdominal and back pain. Her IVC filter was misaligned on CT scan and she subsequently
underwent three attempts at percutaneous filter removal. Due to significant angulation, the filter
could not be removed. The patient was referred to us for open removal. Several of the struts were
found penetrating through the caval wall and entering the serosa of the duodenum as well as the
anterior spinal ligament. These were carefully removed at mini-laparotomy and a venotomy was
made in the IVC over the filter. The proximal hook of the filter was noted to have lodged into a
posterior lumbar vein explaining the difficulty with percutaneous retrieval. The filter was
dissected off the wall after several of the struts were cut with wire cutters. The venotomy was
primarily repaired. Postoperatively, the patient reported resolution of her back and abdominal
ACCEPTED MANUSCRIPT 8
Case 4 A 25-year old man with a history of motor vehicle crash 6 years ago requiring a C3-C7
spinal fusion presented with right lower quadrant abdominal pain. After undergoing a CT scan,
the patient was found to have an IVC filter with severe angulation and barbs penetrating through
the caval wall. The patient reported that he never knew it was in place so he never sought
removal. An attempt at percutaneous removal was done at an outside hospital that was
He was referred to our center for open removal. We could find no other reason for his
M AN U
abdominal pain. After mini-laparotomy, the filter struts were found to be penetrating through the
caval wall in multiple directions including a barb through the duodenal wall and at least 1 strut
penetrating into the aorta. The inferior vena cava was opened longitudinally. Several of the barbs
were cut with Mayo scissors including the barb penetrating into the aorta. Care was taken to
ensure that the entire filter was removed without damaging the duodenum and the aorta. The
venotomy was primarily repaired. After recovery, the patient reported resolution of his
A 28-year-old woman presented to our facility with persistent abdominal pain. She had a
history of IVC filter placement a year before when she suffered a pelvic fracture. An attempt was
made at that time to remove the IVC filter percutaneously when she regained mobility but it was
noted that the filter was misaligned and imbedded into the caval wall. The filter could not be
removed and the patient later returned with complaints of abdominal pain. We could find no
other cause for her pain and she was scheduled for open filter removal. A midline mini-
ACCEPTED MANUSCRIPT 9
laparotomy incision was made. Several of the barbs were penetrating through the caval wall. One
of the barbs penetrated through the adventitia of the aorta and there was also a filter barb
penetrating into and through the gonadal vein. This was contributing to the severe angulation.
Several of the barbs were transected outside the IVC and a longitudinal venotomy was made to
extract the filter from the IVC as well as the gonadal vein. The venotomy was then primarily
repaired. She reported resolution of her abdominal pain in follow-up.
A 50-year-old woman with chronic kidney disease was transferred from an outside
M AN U
hospital with abdominal pain. She had a history of right leg gunshot wound, below-knee
amputation, deep vein thrombosis, and IVC filter placement 25 years ago. She was diagnosed
with IVC thrombosis 10 years ago and had an additional IVC filter placed superior to the
thrombosis. On arrival, a CT scan showed an occluded IVC with several misaligned filters and
barbs penetrating through the caval wall into adjacent structures. No other cause for her severe
abdominal pain could be found. At laparotomy, 3 tangled filters were found in an occluded IVC.
One of the barbs was penetrating through the wall of the duodenum. The small bowel was
repaired with polyglactin suture. Postoperatively, she had prolonged ileus. She was discharged
home tolerating a regular diet on postoperative day 7.
A 36-year-old woman was admitted via the emergency room complaining of worsening,
severe abdominal and right groin pain. An extensive workup only revealed an angulated IVC
filter placed in the right common iliac vein. The filter was placed several years ago for deep vein
thrombosis and pulmonary embolism. Although she was given a diagnosis of hypercoagulable
ACCEPTED MANUSCRIPT 10
disorder, she was not placed on anticoagulation. An attempt was made at percutaneous retrieval
but this was unsuccessful due to extreme tilting. The filter was removed via transperitoneal mini-
laparotomy. Inspection of the abdomen did not show any other cause for her abdominal pain.
Several of the barbs were penetrating out of the vein wall into the L5 periosteum and
surrounding structures. The longitudinal venotomy in the right common iliac vein was repaired
primarily. She was started on oral anticoagulation postoperatively and was discharged home on
postoperative day 9.
M AN U
Since 2011, we began a comprehensive IVC Filter Retrieval Program. Our goal was to percutaneously remove all filters within 3 months, regardless of insurance status, in patients
whose transient risk of venous thromboembolic events subsided.2 Many centers, including our
own, report retrieval rates of less than 50% and initial technical success of percutaneous retrieval
in 85-99% of cases.7-9 Failure to percutaneously retrieve the filter may be due to thrombus within
the filter or severe angulation; both can worsen over time.
Angulation or tilting of the filter is associated with strut penetration through the caval wall. This appears to occur more frequently depending on IVC diameter, dwell time5 and filter
type. Strut penetration through the caval wall is commonplace and most patients are
asymptomatic.10 All patients referred for open filter removal had significant strut penetration into
adjacent structures, such as small bowel, spine and the aorta. Open surgical IVC filter removal is
generally reserved for symptomatic patients who failed percutaneous retrieval. By itself,
ACCEPTED MANUSCRIPT 11
asymptomatic strut penetration into pericaval structures is not a usual indication for open
removal. Our algorithm for filter retrieval is shown in Figure 5. Several centers published their experience of IVC filter removal via laparotomy and most
show low complication rates using slightly different surgical techniques.11-14 Although there may
be reporting bias, morbidity and mortality appear to be low in experienced centers. We find that
a 10-cm laparotomy is sufficient in most patients. The retroperitoneum can be entered directly
and the duodenum retracted to the right of the IVC without mobilization of the right colon.
Removal of the filter by cavotomy, endovenectomy and primary repair can be safely done
without the need for extensive caval reconstruction as observed in our case series.
M AN U
Regardless of symptoms, most patients are candidates for percutaneous IVC filter retrieval when the risk of pulmonary embolus subsides. We reserve open surgical removal of
misaligned IVC filters to patients suffering from symptoms, such as chronic pain. The symptoms
appear to be caused by filter strut penetration into adjacent structures. Other indications for open
IVC filter removal include complications during percutaneous filter retrieval, such as
dislodgement of the filter into the right atrium and retrieval device entanglement. Although open
surgery is the last resort for retrieval of IVC filters, it can be accomplished with minimal
morbidity and leads to resolution of symptoms.
ACCEPTED MANUSCRIPT 12
Figure Legend Figure 1. Intraoperative image after opening the IVC. Note the filter strut within the iliolumbar tributary.
Figure 2. Primary closure of the venotomy. Figure 3. Abdominal x-ray showing two misaligned IVC filters (arrows).
Figure 4. Fluoroscopy image taken during percutaneous retrieval attempt showing contrast extravasation, severe filter angulation, and goose-neck snare entrapment.
M AN U
Figure 5. Algorithm for retrieval evaluation of implanted IVC filters.
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
ACCEPTED MANUSCRIPT 13
M AN U
1. Angel LF, Tapson V, Galgon RE, Restrepo MI and Kaufman J. Systematic review of the use of retrievable inferior vena cava filters. J Vasc Interv Radiol. 2011;22:1522-1530 e3. 2. Charlton-Ouw KM, Leake SS, Sola CN, Sandhu HK, Albarado R, Holcomb JB, Miller CC, 3rd, Safi HJ and Azizzadeh A. Technical and financial feasibility of an inferior vena cava filter retrieval program at a level one trauma center. Ann Vasc Surg. 2015;29:84-9. 3. Warner CJ, Richey EA, Tower DE, Condino AE, Tapp SJ, Tosteson AN and Walsh DB. Inferior vena cava filter retrieval provides no advantage in the average patient. J Vasc Surg Venous Lymphat Disord. 2015;3:142-6. 4. Stawicki SP, Sims CA, Sharma R, Weger NS, Truitt M, Cipolla J, Schrag SP, Lorenzo M, El Chaar M, Torigian DA, Kim PK and Sarani B. Vena cava filters: a synopsis of complications and related topics. J Vasc Access. 2008;9:102-10. 5. Lee JK, So YH, Choi YH, Park SS, Heo EY, Kim DK and Chung HS. Clinical course and predictive factors for complication of inferior vena cava filters. Thromb Res. 2014;133:53843. 6. Morales JP, Li X, Irony TZ, Ibrahim NG, Moynahan M and Cavanaugh KJ, Jr. Decision analysis of retrievable inferior vena cava filters in patients without pulmonary embolism. J Vasc Surg Venous Lymphat Disord. 2013;1:376-84. 7. Ray CE, Jr., Mitchell E, Zipser S, Kao EY, Brown CF and Moneta GL. Outcomes with retrievable inferior vena cava filters: a multicenter study. J Vasc Interv Radiol. 2006;17:1595604. 8. Tao MJ, Montbriand JM, Eisenberg N, Sniderman KW and Roche-Nagle G. Temporary inferior vena cava filter indications, retrieval rates, and follow-up management at a multicenter tertiary care institution. J Vasc Surg. 2016;64:430-437. 9. Etkin Y, Foley PJ, Wang GJ, Guzzo TJ, Roses RE, Fraker DL, Drebin JA and Jackson BM. Successful venous repair and reconstruction for oncologic resections. J Vasc Surg Venous Lymphat Disord. 2016;4:57-63. 10. Wood EA, Malgor RD, Gasparis AP and Labropoulos N. Reporting the impact of inferior vena cava perforation by filters. Phlebology. 2014;29:471-5. 11. Malgor RD and Labropoulos N. A systematic review of symptomatic duodenal perforation by inferior vena cava filters. J Vasc Surg. 2012;55:856-861 e3. 12. Connolly PH, Balachandran VP, Trost D and Bush HL, Jr. Open surgical inferior vena cava filter retrieval for caval perforation and a novel technique for minimal cavotomy filter extraction. J Vasc Surg. 2012;56:256-9; discussion 259. 13. Malek JY, Kwolek CJ, Conrad MF, Patel VI, Watkins MT, Lancaster RT and LaMuraglia GM. Presentation and treatment outcomes of patients with symptomatic inferior vena cava filters. Ann Vasc Surg. 2013;27:84-8. 14. Rana MA, Gloviczki P, Kalra M, Bjarnason H, Huang Y and Fleming MD. Open surgical removal of retained and dislodged inferior vena cava filters. J Vasc Surg Venous Lymphat Disord. 2015;3:201-6.
262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302
Table. Patients requiring open surgical inferior vena cava removal. Indications for filter placement
Length of implant
Percutaneous Operative attempts findings
2 filters placed, duration unknown
No attempts made after venogram and CT findings
Caval and duodenal penetration
Failed 5 months anticoagulation (recurrent PE)
Recurrent caval thrombosis, leg swelling, back pain None
Multiple percutaneous attempts.
Failed Unknown anticoagulation (History of DVT and PE)
Right groin and mid abdominal pain
At least 3 percutaneous attempts
Trauma Pelvic Fracture (prophylaxis)
Trauma – gunshot wound and DVT
Gooseneck snare entangled with filter. Penetration of caval wall, duodenum, and lumbar vein Caval, duodenal penetration; small aortocaval fistula Penetration of caval wall, aortic adventitia and gonadal vein Penetration of caval wall into duodenum. Filter in right common iliac vein with tilting and penetration through vein wall.
3 filters, placed 10 and 25 years ago Unknown
M AN U
TE D 1 year
Abdominal and right groin pain
Postop length of stay (days) 7
M AN U
M AN U
M AN U
M AN U
M AN U