copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
From the aDepartment of Radiology Section of InterventionalRadiology and the bDepartments of Surgery and RadiologySections of Vascular Surgery and Interventional RadiologyUniversity of Michigan Ann Arbor MichiganSubmitted July 26 2006 accepted for publication August 4 2006The authors have no financial interest arrangement or affiliationwith the manufacturers of any products discussed in the article ortheir competitors Correspondence John E Rectenwald MD 2210 F Taubman Center1500 East Medical Center Drive Ann Arbor MI 48109-0325 E-mail jrectenwumichedu
Inferior Vena Cava Filters in the CancerPatient Current Use and Indications
Todd M Getzen MDa and John E Rectenwald MDb Ann Arbor Michigan
Historically pharmacologic anticoagulation is the pre-ferred treatment for DVT and venous thromboembolismOral warfarin has been the mainstay of treatment sincethe 1950s and is used together with intravenous heparinwhich has been used in humans since the 1930s Recentlythe development of low molecular weight heparins withlonger half-lives and subcutaneous administration has sim-plified treatment of DVT and PE With fractionated he-parins patients no longer must remain in the hospital onintravenous heparin while warfarin is being titrated Lowmolecular weight heparins may also have other advan-tages Experts recently proposed that fractionated heparinsmay improve survival in cancer patients by decreasing tu-mor angiogenesis and interfering with tumor thrombus for-mation that potentates tumor metastasis3 Other neweragents such as the direct thrombin inhibitor ximelaga-tranmelagatran and the specific factor Xa inhibitor pen-tasaccharide are promising for future use
Current Indications for IVC FiltersAlthough pharmacologic treatment of VTE is evolving asubset of patients remains in whom pharmacologic anti-coagulation therapy fails who develop complications re-lated to anticoagulants or who have an unacceptably highrisk for complications from anticoagulants MechanicalIVC filtration is indicated for these patients (Table 1)
Anticoagulation therapy failure is defined as the inabil-ity to titrate to the target international normalized ratioor partial thromboplastin time recurrent or new PE de-spite adequate and therapeutic anticoagulation or thepropagation or new development of DVT despite thera-peutic anticoagulation Complications of anticoagulationare typically related to bleeding including persistent orlarge gastrointestinal hemorrhages intracranial hemor-rhages and hemoptysis Development of significant bleed-ing complications typically precludes further anticoagulantuse unless the underlying cause can be adequately treated
AbstractDeep venous thrombosis and thromboembolism are significanthealth risks with high rates of morbidity and mortality Chronicallyill and hospitalized patients particularly those with cancer have ahigh risk for developing these conditions Mechanical inferior venacava (IVC) filtration has been standard care for patients with theseconditions in whom anticoagulation therapy is contraindicated orhas failed This article reviews caval filters and the current indica-tions for using mechanical IVC filters including retrievable versuspermanent filters focusing on their use in treating venous throm-boembolism in cancer patients (JNCCN 20064881ndash888)
Current Treatment of Venous ThromboembolicDiseaseVenous thromboembolic disease (VTE) which includesdeep venous thrombosis (DVT) and pulmonary embolism(PE) is a common problem in hospitalized patientsReported incidence of VTE is 100 per 100000 peopleper year or approximately 355000 diagnoses per yearTwo thirds of these occurrences are DVT and the re-mainder are PE Overall VTE causes as many as 240000deaths per year12 Some patient populations notably thosewith cancer are at high risk for VTE and are dispropor-tionately affected Treatment of VTE has evolved into 2distinct entities pharmacologic anticoagulation and in-ferior vena cava (IVC) interruption
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 881
Generally patients believed to have unacceptably highrisks for bleeding complications and are therefore poorcandidates for anticoagulant therapy include thosewith 1) recent cerebrovascular accidents 2) significantrisk for falling (with consequent risk for intracranialhemorrhage from minor head trauma) 3) pre-existinggastrointestinal bleeding or lesions that are known tobe at high risk for bleeding 4) brain tumors or otherhigh-risk intracranial lesions and 5) thrombocytope-nia Patients with acute DVT who will undergo or haveundergone major surgery often cannot be treated withanticoagulation Thus they are believed to be at riskfor PE and considered reasonable candidates for IVCfilters Multisystem trauma patients with DVT andhead injury splenic lacerations or liver lacerations arein this category
Free-floating iliocaval thrombi prompt an occa-sional request for filter placement Several authorsevaluated this issue with conflicting results One re-port evaluated 5238 consecutive venous duplex stud-ies to identify 73 patients with 82 free-floating thrombiOf these thrombi 13 were associated with clinicallysignificant pulmonary emboli The authors concludedthat most do not embolize but instead adhere to thevein wall4 Additionally a prospective study evaluat-ing relative risks of embolism (PE) among 57 patientswith free-floating thrombi and 28 with occlusivethrombi found no higher risk for PE in patients witha free-floating thrombus compared with those with a
nonocclusive thrombus5 Conversely Radomski et al6
reported pulmonary emboli in 27 of patients withfree-floating iliocaval thrombi despite adequate an-ticoagulation Although free-floating thrombi are cur-rently an accepted indication for filter placement thisis still debated and a subject for further investigation
A relatively new trend in filter use is the place-ment of prophylactic filters These are typically placedin patients who are at high risk for VTE but have noexisting diagnosis or symptoms of DVT or PE and inthose who are at risk for VTE who cannot safely un-dergo anticoagulation for DVT prophylaxis with low-dose heparin Trauma is a common indication forprophylactic IVC filter placement especially in high-risk patients with pelvic or long bone fractures whowill undergo surgical fixation This indication is aprimary force driving the development of retrievableIVC filters Rutherford7 recently examined this trendconcluding that medical evidence is insufficient tosupport the broad use of prophylactic filters and thatfilter use for this purpose needs further objective study(Table 1)
Currently Available IVC FiltersThe first widely successful percutaneously introducedIVC filter the Kimray-Greenfield filter became avail-able in 1973 This filter used a 30 French introducer largeenough to require surgical exposure of the femoral veinSubsequent refinements have resulted in the 14-Frenchstainless steel and titanium Greenfield filters (BostonScientific Natick MA) The over-the-wire stainlesssteel Greenfield filter is currently the benchmark towhich all filters are compared and has more than 20years of patient follow-up data supporting its use8
Numerous other permanent IVC filters have beendeveloped Most parallel the conical design of theGreenfield filter whereas others more radically departfrom this prototype These include the Gianturco-Roehm Birdrsquos Nest filter (Cook Bloomington IN)9
which is 1 of 2 filters indicated for caval diametersgreater than 28 mm the Bard filters (ie SimonNitinol filter discontinued Recovery Nitinol filterand the G2 filter Bard Tempe AZ) with dual filtrationlevels the B Braun Vena Tech LGM and LP filters (BBraun Bethlehem PA) which have a conical designsimilar to the Greenfield filter with added stabilizingstruts (the Vena Tech LP is the second filter indicatedfor caval diameters greater than 28 mm) and the
882 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 1 Indications for Inferior Vena Cava Filter Placement in Patients with Deep Vein Thrombosis or Pulmonary Embolism
Anticoagulation-related
Failure of anticoagulation (propagation or new PE whiletherapeutic inability to reach therapeutic levels
Significant bleeding complication related to anticoagulation
Contraindication to anticoagulation (eg cerebrovascularaccident central nervous system metastases fall risk)
Recent or planned surgery
Thrombocytopenia
Venous thromboembolism with poor cardiopulmonaryreserve (eg pulmonary hypertension right heart failure)
Free-floating iliocaval thrombus
Trauma (prophylaxis)
Subject to debate see discussion
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 882
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Cordis TRAPEASE filter (Cordis Corporation MiamiLakes FL) which has a modified dual cone design(Figure 1)
Several filters with retrieval indications were re-cently introduced known as either retrievable filters oroptional retrievable filters which allow temporary place-ment and retrieval or may remain permanently
Retrievable FiltersThe Guumlnther Tulip vena cava MREye filter set (Cook)has an apical cone design similar to the Greenfieldfilter with a small hook at the top to allow snaringfor retrieval The filter comes in femoral and jugularinsertion kits with an 8 French introducer sheath Thefilter can be retrieved through the jugular route onlyusing a 12 French retrieval system The OptEase fil-ter (Cordis Corporation) is a dual cone (symmetrical)design nearly identical to the TRAPEASE and canbe inserted from jugular or femoral routes with thesame 6 French introducer sheath (by reorienting thefilter) This filter is retrieved only through the femoralvein by snaring a small hook at the caudal end of thefilter The Guumlnther Tulip and the OptEase filters arethe only filters currently approved by the US Food
and Drug Administration (FDA) as retrievable de-vices A third retrievable filter the Recovery Nitinol(Bard) is no longer marketed The Recovery wasplaced through femoral or jugular routes and was re-trievable through a jugular approach using a uniquecone capturing device Although the Bard G2 filter iscurrently available as a permanent filter it is beingevaluated by the FDA for retrieval use
Retrievable filters have a limited recovery timebefore they become substantially incorporated intothe caval wall and cannot be safely retrieved All re-trievable filters currently available are also approvedby the FDA for permanent placement indications andmay be left in place if the need for caval interruptionbecomes life-long or other factors arise that preventeither retrieval or resumption of anticoagulationUnfortunately studies of retrieval rates have foundthat at best only approximately 50 of retrievablefilters are actually removed when contraindicationsto anticoagulation resolve10
Permanent Versus Retrievable FiltersRetrievable filters seem to be attractive options forpatients with a definable risk period for VTE and a
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 883
limited contraindication to anticoagulation The useof retrievable filters depends on several assumptions(Table 2)
Retrievable filters have other perceived advan-tages including some of the smallest delivery systemsavailable which theoretically may reduce insertion-site thrombosis The PREPIC (Preacutevention du RisquedrsquoEmbolie Pulmonaire par Interruption Cave) study1112
concluded that retrieving filters when the risk for DVTor inability to anticoagulate resolves would provideprotection from PE during the vulnerable time whiletheoretically eliminating the late risk for DVT Severalfactors regarding the use of retrievable filters are un-known For example whether the risk for DVT as de-fined by the PREPIC study justifies the cost ofretrieving these filters and the risk to patients is un-clear Unfortunately all of the available retrievablefilters lack long-term efficacy and safety data compa-rable to those of the Greenfield filter or even newerpermanent filters Retrieval windows are continuallybeing revised upwards and newer technology suchas paclitaxel coating may further lengthen themRetrieval times were initially within days to a fewweeks whereas now they are reported several monthsand in some cases more than 1 year after placement13
IVC Filter Placement ProcedureIVC filters may be placed through several routes Allcommonly used filters come in jugular or femoral kitsor in a single kit that can be used for either routeSome filters allow greater flexibility in placementsuch as the Simon Nitinol which can be placed fromthe antecubital vein approach if necessary with asmall-diameter long delivery system (7 French 103cm) The authors have placed a Simon Nitinol filterthrough an upper-extremity peripherally inserted
central catheter (PICC) line site in a patient with ex-tensive body surface burns (replacing a new PICC lineafter the filter was deployed) and a Guumlnther Tulip fil-ter through the great saphenous vein using intravas-cular ultrasound (IVUS) guidance in a morbidly obesepatient in the intensive care unit (ICU) Probablyonly a few patients truly do not have access suitablefor filter placement
After obtaining venous access a marker flushcatheter is placed in the IVC and a contrast or CO2
(in the event of renal failure) venacavogram is per-formed Positioning the catheter in the left commoniliac vein increases the likelihood of identifying a du-plicated IVC The venacavogram is evaluated for thepresence of intraluminal thrombus vascular variantssuch as IVC duplication or a circumaortic left renalvein and the location of both renal veins (Figure 2)Identifying vascular variants is of vital importancebecause a duplicated IVC or circumaortic left renalvein can provide an alternative route for throm-boemboli to bypass the filter and cause PE If the IVCis free of thrombus the vena cava diameter is deter-mined and the filter is positioned in the infrarenal
884 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 2 Assumptions Guiding Use of Retrievable Filters
bull Equivalent function of retrievable filters and permanentcounterparts
bull Short defined period of susceptibility to thromboembolism
bull Low likelihood of recurrent pulmonary embolism afterretrieval
bull Safety of retrieval procedure
bull The long-term risk of complications from a permanentfilter justify retrieval
Figure 2 The left panel depicts a normal venogram of the inferior venacava (IVC) Note the presence of both the left and right iliac veinsruling out a duplicated vena cava Opacification of the right renal veinreveals its location whereas the level of the left renal vein can be identified by the ldquowash outrdquo of contrast within the IVC by unopacifiedblood from the left renal vein The right panel illustrates proper placement of a Greenfield IVC filter within the IVC
Source Images courtesy of Kyung J Cho MD
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 884
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
IVC and deployed with the tip near the level of therenal veins If any circumstance prevents the use of flu-oroscopy alternative imaging methods for guidingplacement include IVUS and transabdominal ultra-sound in carefully selected patients The use of IVUSor transabdominal ultrasound for IVC filter placementallows filters to be deployed at the bedside which isan advantage in the ICU setting for patients who maynot be stable enough to travel to the fluoroscopy suite
When a vascular variant or IVC thrombus is iden-tified changes in filter position are mandated For aduplicated IVC options include placing a filter ineach IVC or simply placing a suprarenal filter (the leftIVC typically draining to the left renal vein) For acircumaortic left renal vein the IVC filter is posi-tioned below the lowest renal vein inflow or if theinfrarenal cava is not long enough the filter is placedsuprarenally Finally if caval thrombus is identified thefilter is placed above the thrombus suprarenally ifnecessary Care must be taken to ensure that the fil-ter legs do not land in the renal vein ostia becausethis can cause filter instability At the completion offilter placement venous access is removed Typicallylittle bleeding occurs with hemostasis by manual pres-sure for a few minutes even with 15 French deliverysystems
Caval Filtration Efficacy Morbidity and MortalityIn 1992 Becker et al14 reviewed numerous case serieson the 7 filters available or in testing at that time(Original Greenfield Titanium Greenfield SimonNitinol Birdrsquos Nest Venatech LGM Amplatz andGuumlnther) evaluating safety indications and effec-tiveness The combined individual series evaluated2557 patients of whom 64 received Greenfield fil-ters In this group 8 deaths were reported from recur-rent PE after filter placement and 4 deaths were relatedto filter complications The authors concluded thatwhen possible anticoagulation should be continuedor reinstituted after filter placement
The PREPIC study by Decousus et al1112 whichwas the only prospective randomized trial of filter ef-ficacy evaluated the effectiveness of filters forpreventing recurrent PE The study used concomitantanticoagulation with either nonfractionated heparinor enoxaparin no group received filters without un-dergoing anticoagulation and no control was
untreated This study randomized 200 patients pergroup into those who were treated with a filter andthose who were not and then randomized them againinto groups receiving unfractionated heparin or enoxa-parin Four filters were used including the VenaTechLGM Titanium Greenfield Cardial (Bard Saint-Ettienne France) and Birdrsquos Nest with the VenaTechLGM most commonly used (56)
Patients have now been followed-up for 8 yearswith findings first published at 2 years Initially pa-tients treated with a filter experienced fewer recur-rent pulmonary emboli with relatively no instancesof major filter-related complications No difference inmortality was noted between the filter and no filtergroups either immediately or at 2 years The studyalso found that at 2 years filter-related morbidity (re-current DVT filter site thrombosis) increased signif-icantly (208 in the filter group vs 116 in the nofilter group) Findings at 8 years again showed early pro-tection against pulmonary emboli but no change inmortality Recurrent DVT remained higher in patientstreated with a filter probably because of insertion-sitethrombosis however this did not result in a signifi-cantly greater incidence of post-thrombotic syndrome
Previous evaluations of filter-related morbidityand mortality have reported widely varied incidenceSeveral authors analyzed the potential complicationsrelated to IVC filter placement identifying events im-mediately related to placement and later complica-tions Short-term complications include standardpercutaneous risks for infection and bleeding and pro-cedural mishaps such as pneumothorax (for jugularplacement) arterial injury and arteryndashvein fistula for-mation Also included in this category is filter mis-placement or maldeployment (tilting greater than 14degfound to reduce efficacy of filtration) and failure toidentify collateral pathways of thromboembolusmigration (circumaortic renal vein duplicated IVCupper-extremity source) Later complications includefilter migration caval penetration recurrent pul-monary emboli caval occlusion and guidewireentanglement during subsequent venous access place-ment15 Filter migration is a rare event with a reportedincidence of 0216 Caval perforation from retentionhooks is a common event with all filters although itis rarely clinically significant and is the subject of iso-lated case reports
Recurrent pulmonary emboli and caval occlusionare the most significant post-placement complications
Original Article 885
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 885
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
Generally patients believed to have unacceptably highrisks for bleeding complications and are therefore poorcandidates for anticoagulant therapy include thosewith 1) recent cerebrovascular accidents 2) significantrisk for falling (with consequent risk for intracranialhemorrhage from minor head trauma) 3) pre-existinggastrointestinal bleeding or lesions that are known tobe at high risk for bleeding 4) brain tumors or otherhigh-risk intracranial lesions and 5) thrombocytope-nia Patients with acute DVT who will undergo or haveundergone major surgery often cannot be treated withanticoagulation Thus they are believed to be at riskfor PE and considered reasonable candidates for IVCfilters Multisystem trauma patients with DVT andhead injury splenic lacerations or liver lacerations arein this category
Free-floating iliocaval thrombi prompt an occa-sional request for filter placement Several authorsevaluated this issue with conflicting results One re-port evaluated 5238 consecutive venous duplex stud-ies to identify 73 patients with 82 free-floating thrombiOf these thrombi 13 were associated with clinicallysignificant pulmonary emboli The authors concludedthat most do not embolize but instead adhere to thevein wall4 Additionally a prospective study evaluat-ing relative risks of embolism (PE) among 57 patientswith free-floating thrombi and 28 with occlusivethrombi found no higher risk for PE in patients witha free-floating thrombus compared with those with a
nonocclusive thrombus5 Conversely Radomski et al6
reported pulmonary emboli in 27 of patients withfree-floating iliocaval thrombi despite adequate an-ticoagulation Although free-floating thrombi are cur-rently an accepted indication for filter placement thisis still debated and a subject for further investigation
A relatively new trend in filter use is the place-ment of prophylactic filters These are typically placedin patients who are at high risk for VTE but have noexisting diagnosis or symptoms of DVT or PE and inthose who are at risk for VTE who cannot safely un-dergo anticoagulation for DVT prophylaxis with low-dose heparin Trauma is a common indication forprophylactic IVC filter placement especially in high-risk patients with pelvic or long bone fractures whowill undergo surgical fixation This indication is aprimary force driving the development of retrievableIVC filters Rutherford7 recently examined this trendconcluding that medical evidence is insufficient tosupport the broad use of prophylactic filters and thatfilter use for this purpose needs further objective study(Table 1)
Currently Available IVC FiltersThe first widely successful percutaneously introducedIVC filter the Kimray-Greenfield filter became avail-able in 1973 This filter used a 30 French introducer largeenough to require surgical exposure of the femoral veinSubsequent refinements have resulted in the 14-Frenchstainless steel and titanium Greenfield filters (BostonScientific Natick MA) The over-the-wire stainlesssteel Greenfield filter is currently the benchmark towhich all filters are compared and has more than 20years of patient follow-up data supporting its use8
Numerous other permanent IVC filters have beendeveloped Most parallel the conical design of theGreenfield filter whereas others more radically departfrom this prototype These include the Gianturco-Roehm Birdrsquos Nest filter (Cook Bloomington IN)9
which is 1 of 2 filters indicated for caval diametersgreater than 28 mm the Bard filters (ie SimonNitinol filter discontinued Recovery Nitinol filterand the G2 filter Bard Tempe AZ) with dual filtrationlevels the B Braun Vena Tech LGM and LP filters (BBraun Bethlehem PA) which have a conical designsimilar to the Greenfield filter with added stabilizingstruts (the Vena Tech LP is the second filter indicatedfor caval diameters greater than 28 mm) and the
882 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 1 Indications for Inferior Vena Cava Filter Placement in Patients with Deep Vein Thrombosis or Pulmonary Embolism
Anticoagulation-related
Failure of anticoagulation (propagation or new PE whiletherapeutic inability to reach therapeutic levels
Significant bleeding complication related to anticoagulation
Contraindication to anticoagulation (eg cerebrovascularaccident central nervous system metastases fall risk)
Recent or planned surgery
Thrombocytopenia
Venous thromboembolism with poor cardiopulmonaryreserve (eg pulmonary hypertension right heart failure)
Free-floating iliocaval thrombus
Trauma (prophylaxis)
Subject to debate see discussion
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 882
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Cordis TRAPEASE filter (Cordis Corporation MiamiLakes FL) which has a modified dual cone design(Figure 1)
Several filters with retrieval indications were re-cently introduced known as either retrievable filters oroptional retrievable filters which allow temporary place-ment and retrieval or may remain permanently
Retrievable FiltersThe Guumlnther Tulip vena cava MREye filter set (Cook)has an apical cone design similar to the Greenfieldfilter with a small hook at the top to allow snaringfor retrieval The filter comes in femoral and jugularinsertion kits with an 8 French introducer sheath Thefilter can be retrieved through the jugular route onlyusing a 12 French retrieval system The OptEase fil-ter (Cordis Corporation) is a dual cone (symmetrical)design nearly identical to the TRAPEASE and canbe inserted from jugular or femoral routes with thesame 6 French introducer sheath (by reorienting thefilter) This filter is retrieved only through the femoralvein by snaring a small hook at the caudal end of thefilter The Guumlnther Tulip and the OptEase filters arethe only filters currently approved by the US Food
and Drug Administration (FDA) as retrievable de-vices A third retrievable filter the Recovery Nitinol(Bard) is no longer marketed The Recovery wasplaced through femoral or jugular routes and was re-trievable through a jugular approach using a uniquecone capturing device Although the Bard G2 filter iscurrently available as a permanent filter it is beingevaluated by the FDA for retrieval use
Retrievable filters have a limited recovery timebefore they become substantially incorporated intothe caval wall and cannot be safely retrieved All re-trievable filters currently available are also approvedby the FDA for permanent placement indications andmay be left in place if the need for caval interruptionbecomes life-long or other factors arise that preventeither retrieval or resumption of anticoagulationUnfortunately studies of retrieval rates have foundthat at best only approximately 50 of retrievablefilters are actually removed when contraindicationsto anticoagulation resolve10
Permanent Versus Retrievable FiltersRetrievable filters seem to be attractive options forpatients with a definable risk period for VTE and a
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 883
limited contraindication to anticoagulation The useof retrievable filters depends on several assumptions(Table 2)
Retrievable filters have other perceived advan-tages including some of the smallest delivery systemsavailable which theoretically may reduce insertion-site thrombosis The PREPIC (Preacutevention du RisquedrsquoEmbolie Pulmonaire par Interruption Cave) study1112
concluded that retrieving filters when the risk for DVTor inability to anticoagulate resolves would provideprotection from PE during the vulnerable time whiletheoretically eliminating the late risk for DVT Severalfactors regarding the use of retrievable filters are un-known For example whether the risk for DVT as de-fined by the PREPIC study justifies the cost ofretrieving these filters and the risk to patients is un-clear Unfortunately all of the available retrievablefilters lack long-term efficacy and safety data compa-rable to those of the Greenfield filter or even newerpermanent filters Retrieval windows are continuallybeing revised upwards and newer technology suchas paclitaxel coating may further lengthen themRetrieval times were initially within days to a fewweeks whereas now they are reported several monthsand in some cases more than 1 year after placement13
IVC Filter Placement ProcedureIVC filters may be placed through several routes Allcommonly used filters come in jugular or femoral kitsor in a single kit that can be used for either routeSome filters allow greater flexibility in placementsuch as the Simon Nitinol which can be placed fromthe antecubital vein approach if necessary with asmall-diameter long delivery system (7 French 103cm) The authors have placed a Simon Nitinol filterthrough an upper-extremity peripherally inserted
central catheter (PICC) line site in a patient with ex-tensive body surface burns (replacing a new PICC lineafter the filter was deployed) and a Guumlnther Tulip fil-ter through the great saphenous vein using intravas-cular ultrasound (IVUS) guidance in a morbidly obesepatient in the intensive care unit (ICU) Probablyonly a few patients truly do not have access suitablefor filter placement
After obtaining venous access a marker flushcatheter is placed in the IVC and a contrast or CO2
(in the event of renal failure) venacavogram is per-formed Positioning the catheter in the left commoniliac vein increases the likelihood of identifying a du-plicated IVC The venacavogram is evaluated for thepresence of intraluminal thrombus vascular variantssuch as IVC duplication or a circumaortic left renalvein and the location of both renal veins (Figure 2)Identifying vascular variants is of vital importancebecause a duplicated IVC or circumaortic left renalvein can provide an alternative route for throm-boemboli to bypass the filter and cause PE If the IVCis free of thrombus the vena cava diameter is deter-mined and the filter is positioned in the infrarenal
884 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 2 Assumptions Guiding Use of Retrievable Filters
bull Equivalent function of retrievable filters and permanentcounterparts
bull Short defined period of susceptibility to thromboembolism
bull Low likelihood of recurrent pulmonary embolism afterretrieval
bull Safety of retrieval procedure
bull The long-term risk of complications from a permanentfilter justify retrieval
Figure 2 The left panel depicts a normal venogram of the inferior venacava (IVC) Note the presence of both the left and right iliac veinsruling out a duplicated vena cava Opacification of the right renal veinreveals its location whereas the level of the left renal vein can be identified by the ldquowash outrdquo of contrast within the IVC by unopacifiedblood from the left renal vein The right panel illustrates proper placement of a Greenfield IVC filter within the IVC
Source Images courtesy of Kyung J Cho MD
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 884
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
IVC and deployed with the tip near the level of therenal veins If any circumstance prevents the use of flu-oroscopy alternative imaging methods for guidingplacement include IVUS and transabdominal ultra-sound in carefully selected patients The use of IVUSor transabdominal ultrasound for IVC filter placementallows filters to be deployed at the bedside which isan advantage in the ICU setting for patients who maynot be stable enough to travel to the fluoroscopy suite
When a vascular variant or IVC thrombus is iden-tified changes in filter position are mandated For aduplicated IVC options include placing a filter ineach IVC or simply placing a suprarenal filter (the leftIVC typically draining to the left renal vein) For acircumaortic left renal vein the IVC filter is posi-tioned below the lowest renal vein inflow or if theinfrarenal cava is not long enough the filter is placedsuprarenally Finally if caval thrombus is identified thefilter is placed above the thrombus suprarenally ifnecessary Care must be taken to ensure that the fil-ter legs do not land in the renal vein ostia becausethis can cause filter instability At the completion offilter placement venous access is removed Typicallylittle bleeding occurs with hemostasis by manual pres-sure for a few minutes even with 15 French deliverysystems
Caval Filtration Efficacy Morbidity and MortalityIn 1992 Becker et al14 reviewed numerous case serieson the 7 filters available or in testing at that time(Original Greenfield Titanium Greenfield SimonNitinol Birdrsquos Nest Venatech LGM Amplatz andGuumlnther) evaluating safety indications and effec-tiveness The combined individual series evaluated2557 patients of whom 64 received Greenfield fil-ters In this group 8 deaths were reported from recur-rent PE after filter placement and 4 deaths were relatedto filter complications The authors concluded thatwhen possible anticoagulation should be continuedor reinstituted after filter placement
The PREPIC study by Decousus et al1112 whichwas the only prospective randomized trial of filter ef-ficacy evaluated the effectiveness of filters forpreventing recurrent PE The study used concomitantanticoagulation with either nonfractionated heparinor enoxaparin no group received filters without un-dergoing anticoagulation and no control was
untreated This study randomized 200 patients pergroup into those who were treated with a filter andthose who were not and then randomized them againinto groups receiving unfractionated heparin or enoxa-parin Four filters were used including the VenaTechLGM Titanium Greenfield Cardial (Bard Saint-Ettienne France) and Birdrsquos Nest with the VenaTechLGM most commonly used (56)
Patients have now been followed-up for 8 yearswith findings first published at 2 years Initially pa-tients treated with a filter experienced fewer recur-rent pulmonary emboli with relatively no instancesof major filter-related complications No difference inmortality was noted between the filter and no filtergroups either immediately or at 2 years The studyalso found that at 2 years filter-related morbidity (re-current DVT filter site thrombosis) increased signif-icantly (208 in the filter group vs 116 in the nofilter group) Findings at 8 years again showed early pro-tection against pulmonary emboli but no change inmortality Recurrent DVT remained higher in patientstreated with a filter probably because of insertion-sitethrombosis however this did not result in a signifi-cantly greater incidence of post-thrombotic syndrome
Previous evaluations of filter-related morbidityand mortality have reported widely varied incidenceSeveral authors analyzed the potential complicationsrelated to IVC filter placement identifying events im-mediately related to placement and later complica-tions Short-term complications include standardpercutaneous risks for infection and bleeding and pro-cedural mishaps such as pneumothorax (for jugularplacement) arterial injury and arteryndashvein fistula for-mation Also included in this category is filter mis-placement or maldeployment (tilting greater than 14degfound to reduce efficacy of filtration) and failure toidentify collateral pathways of thromboembolusmigration (circumaortic renal vein duplicated IVCupper-extremity source) Later complications includefilter migration caval penetration recurrent pul-monary emboli caval occlusion and guidewireentanglement during subsequent venous access place-ment15 Filter migration is a rare event with a reportedincidence of 0216 Caval perforation from retentionhooks is a common event with all filters although itis rarely clinically significant and is the subject of iso-lated case reports
Recurrent pulmonary emboli and caval occlusionare the most significant post-placement complications
Original Article 885
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 885
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Cordis TRAPEASE filter (Cordis Corporation MiamiLakes FL) which has a modified dual cone design(Figure 1)
Several filters with retrieval indications were re-cently introduced known as either retrievable filters oroptional retrievable filters which allow temporary place-ment and retrieval or may remain permanently
Retrievable FiltersThe Guumlnther Tulip vena cava MREye filter set (Cook)has an apical cone design similar to the Greenfieldfilter with a small hook at the top to allow snaringfor retrieval The filter comes in femoral and jugularinsertion kits with an 8 French introducer sheath Thefilter can be retrieved through the jugular route onlyusing a 12 French retrieval system The OptEase fil-ter (Cordis Corporation) is a dual cone (symmetrical)design nearly identical to the TRAPEASE and canbe inserted from jugular or femoral routes with thesame 6 French introducer sheath (by reorienting thefilter) This filter is retrieved only through the femoralvein by snaring a small hook at the caudal end of thefilter The Guumlnther Tulip and the OptEase filters arethe only filters currently approved by the US Food
and Drug Administration (FDA) as retrievable de-vices A third retrievable filter the Recovery Nitinol(Bard) is no longer marketed The Recovery wasplaced through femoral or jugular routes and was re-trievable through a jugular approach using a uniquecone capturing device Although the Bard G2 filter iscurrently available as a permanent filter it is beingevaluated by the FDA for retrieval use
Retrievable filters have a limited recovery timebefore they become substantially incorporated intothe caval wall and cannot be safely retrieved All re-trievable filters currently available are also approvedby the FDA for permanent placement indications andmay be left in place if the need for caval interruptionbecomes life-long or other factors arise that preventeither retrieval or resumption of anticoagulationUnfortunately studies of retrieval rates have foundthat at best only approximately 50 of retrievablefilters are actually removed when contraindicationsto anticoagulation resolve10
Permanent Versus Retrievable FiltersRetrievable filters seem to be attractive options forpatients with a definable risk period for VTE and a
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 883
limited contraindication to anticoagulation The useof retrievable filters depends on several assumptions(Table 2)
Retrievable filters have other perceived advan-tages including some of the smallest delivery systemsavailable which theoretically may reduce insertion-site thrombosis The PREPIC (Preacutevention du RisquedrsquoEmbolie Pulmonaire par Interruption Cave) study1112
concluded that retrieving filters when the risk for DVTor inability to anticoagulate resolves would provideprotection from PE during the vulnerable time whiletheoretically eliminating the late risk for DVT Severalfactors regarding the use of retrievable filters are un-known For example whether the risk for DVT as de-fined by the PREPIC study justifies the cost ofretrieving these filters and the risk to patients is un-clear Unfortunately all of the available retrievablefilters lack long-term efficacy and safety data compa-rable to those of the Greenfield filter or even newerpermanent filters Retrieval windows are continuallybeing revised upwards and newer technology suchas paclitaxel coating may further lengthen themRetrieval times were initially within days to a fewweeks whereas now they are reported several monthsand in some cases more than 1 year after placement13
IVC Filter Placement ProcedureIVC filters may be placed through several routes Allcommonly used filters come in jugular or femoral kitsor in a single kit that can be used for either routeSome filters allow greater flexibility in placementsuch as the Simon Nitinol which can be placed fromthe antecubital vein approach if necessary with asmall-diameter long delivery system (7 French 103cm) The authors have placed a Simon Nitinol filterthrough an upper-extremity peripherally inserted
central catheter (PICC) line site in a patient with ex-tensive body surface burns (replacing a new PICC lineafter the filter was deployed) and a Guumlnther Tulip fil-ter through the great saphenous vein using intravas-cular ultrasound (IVUS) guidance in a morbidly obesepatient in the intensive care unit (ICU) Probablyonly a few patients truly do not have access suitablefor filter placement
After obtaining venous access a marker flushcatheter is placed in the IVC and a contrast or CO2
(in the event of renal failure) venacavogram is per-formed Positioning the catheter in the left commoniliac vein increases the likelihood of identifying a du-plicated IVC The venacavogram is evaluated for thepresence of intraluminal thrombus vascular variantssuch as IVC duplication or a circumaortic left renalvein and the location of both renal veins (Figure 2)Identifying vascular variants is of vital importancebecause a duplicated IVC or circumaortic left renalvein can provide an alternative route for throm-boemboli to bypass the filter and cause PE If the IVCis free of thrombus the vena cava diameter is deter-mined and the filter is positioned in the infrarenal
884 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 2 Assumptions Guiding Use of Retrievable Filters
bull Equivalent function of retrievable filters and permanentcounterparts
bull Short defined period of susceptibility to thromboembolism
bull Low likelihood of recurrent pulmonary embolism afterretrieval
bull Safety of retrieval procedure
bull The long-term risk of complications from a permanentfilter justify retrieval
Figure 2 The left panel depicts a normal venogram of the inferior venacava (IVC) Note the presence of both the left and right iliac veinsruling out a duplicated vena cava Opacification of the right renal veinreveals its location whereas the level of the left renal vein can be identified by the ldquowash outrdquo of contrast within the IVC by unopacifiedblood from the left renal vein The right panel illustrates proper placement of a Greenfield IVC filter within the IVC
Source Images courtesy of Kyung J Cho MD
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 884
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
IVC and deployed with the tip near the level of therenal veins If any circumstance prevents the use of flu-oroscopy alternative imaging methods for guidingplacement include IVUS and transabdominal ultra-sound in carefully selected patients The use of IVUSor transabdominal ultrasound for IVC filter placementallows filters to be deployed at the bedside which isan advantage in the ICU setting for patients who maynot be stable enough to travel to the fluoroscopy suite
When a vascular variant or IVC thrombus is iden-tified changes in filter position are mandated For aduplicated IVC options include placing a filter ineach IVC or simply placing a suprarenal filter (the leftIVC typically draining to the left renal vein) For acircumaortic left renal vein the IVC filter is posi-tioned below the lowest renal vein inflow or if theinfrarenal cava is not long enough the filter is placedsuprarenally Finally if caval thrombus is identified thefilter is placed above the thrombus suprarenally ifnecessary Care must be taken to ensure that the fil-ter legs do not land in the renal vein ostia becausethis can cause filter instability At the completion offilter placement venous access is removed Typicallylittle bleeding occurs with hemostasis by manual pres-sure for a few minutes even with 15 French deliverysystems
Caval Filtration Efficacy Morbidity and MortalityIn 1992 Becker et al14 reviewed numerous case serieson the 7 filters available or in testing at that time(Original Greenfield Titanium Greenfield SimonNitinol Birdrsquos Nest Venatech LGM Amplatz andGuumlnther) evaluating safety indications and effec-tiveness The combined individual series evaluated2557 patients of whom 64 received Greenfield fil-ters In this group 8 deaths were reported from recur-rent PE after filter placement and 4 deaths were relatedto filter complications The authors concluded thatwhen possible anticoagulation should be continuedor reinstituted after filter placement
The PREPIC study by Decousus et al1112 whichwas the only prospective randomized trial of filter ef-ficacy evaluated the effectiveness of filters forpreventing recurrent PE The study used concomitantanticoagulation with either nonfractionated heparinor enoxaparin no group received filters without un-dergoing anticoagulation and no control was
untreated This study randomized 200 patients pergroup into those who were treated with a filter andthose who were not and then randomized them againinto groups receiving unfractionated heparin or enoxa-parin Four filters were used including the VenaTechLGM Titanium Greenfield Cardial (Bard Saint-Ettienne France) and Birdrsquos Nest with the VenaTechLGM most commonly used (56)
Patients have now been followed-up for 8 yearswith findings first published at 2 years Initially pa-tients treated with a filter experienced fewer recur-rent pulmonary emboli with relatively no instancesof major filter-related complications No difference inmortality was noted between the filter and no filtergroups either immediately or at 2 years The studyalso found that at 2 years filter-related morbidity (re-current DVT filter site thrombosis) increased signif-icantly (208 in the filter group vs 116 in the nofilter group) Findings at 8 years again showed early pro-tection against pulmonary emboli but no change inmortality Recurrent DVT remained higher in patientstreated with a filter probably because of insertion-sitethrombosis however this did not result in a signifi-cantly greater incidence of post-thrombotic syndrome
Previous evaluations of filter-related morbidityand mortality have reported widely varied incidenceSeveral authors analyzed the potential complicationsrelated to IVC filter placement identifying events im-mediately related to placement and later complica-tions Short-term complications include standardpercutaneous risks for infection and bleeding and pro-cedural mishaps such as pneumothorax (for jugularplacement) arterial injury and arteryndashvein fistula for-mation Also included in this category is filter mis-placement or maldeployment (tilting greater than 14degfound to reduce efficacy of filtration) and failure toidentify collateral pathways of thromboembolusmigration (circumaortic renal vein duplicated IVCupper-extremity source) Later complications includefilter migration caval penetration recurrent pul-monary emboli caval occlusion and guidewireentanglement during subsequent venous access place-ment15 Filter migration is a rare event with a reportedincidence of 0216 Caval perforation from retentionhooks is a common event with all filters although itis rarely clinically significant and is the subject of iso-lated case reports
Recurrent pulmonary emboli and caval occlusionare the most significant post-placement complications
Original Article 885
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 885
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
limited contraindication to anticoagulation The useof retrievable filters depends on several assumptions(Table 2)
Retrievable filters have other perceived advan-tages including some of the smallest delivery systemsavailable which theoretically may reduce insertion-site thrombosis The PREPIC (Preacutevention du RisquedrsquoEmbolie Pulmonaire par Interruption Cave) study1112
concluded that retrieving filters when the risk for DVTor inability to anticoagulate resolves would provideprotection from PE during the vulnerable time whiletheoretically eliminating the late risk for DVT Severalfactors regarding the use of retrievable filters are un-known For example whether the risk for DVT as de-fined by the PREPIC study justifies the cost ofretrieving these filters and the risk to patients is un-clear Unfortunately all of the available retrievablefilters lack long-term efficacy and safety data compa-rable to those of the Greenfield filter or even newerpermanent filters Retrieval windows are continuallybeing revised upwards and newer technology suchas paclitaxel coating may further lengthen themRetrieval times were initially within days to a fewweeks whereas now they are reported several monthsand in some cases more than 1 year after placement13
IVC Filter Placement ProcedureIVC filters may be placed through several routes Allcommonly used filters come in jugular or femoral kitsor in a single kit that can be used for either routeSome filters allow greater flexibility in placementsuch as the Simon Nitinol which can be placed fromthe antecubital vein approach if necessary with asmall-diameter long delivery system (7 French 103cm) The authors have placed a Simon Nitinol filterthrough an upper-extremity peripherally inserted
central catheter (PICC) line site in a patient with ex-tensive body surface burns (replacing a new PICC lineafter the filter was deployed) and a Guumlnther Tulip fil-ter through the great saphenous vein using intravas-cular ultrasound (IVUS) guidance in a morbidly obesepatient in the intensive care unit (ICU) Probablyonly a few patients truly do not have access suitablefor filter placement
After obtaining venous access a marker flushcatheter is placed in the IVC and a contrast or CO2
(in the event of renal failure) venacavogram is per-formed Positioning the catheter in the left commoniliac vein increases the likelihood of identifying a du-plicated IVC The venacavogram is evaluated for thepresence of intraluminal thrombus vascular variantssuch as IVC duplication or a circumaortic left renalvein and the location of both renal veins (Figure 2)Identifying vascular variants is of vital importancebecause a duplicated IVC or circumaortic left renalvein can provide an alternative route for throm-boemboli to bypass the filter and cause PE If the IVCis free of thrombus the vena cava diameter is deter-mined and the filter is positioned in the infrarenal
884 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 2 Assumptions Guiding Use of Retrievable Filters
bull Equivalent function of retrievable filters and permanentcounterparts
bull Short defined period of susceptibility to thromboembolism
bull Low likelihood of recurrent pulmonary embolism afterretrieval
bull Safety of retrieval procedure
bull The long-term risk of complications from a permanentfilter justify retrieval
Figure 2 The left panel depicts a normal venogram of the inferior venacava (IVC) Note the presence of both the left and right iliac veinsruling out a duplicated vena cava Opacification of the right renal veinreveals its location whereas the level of the left renal vein can be identified by the ldquowash outrdquo of contrast within the IVC by unopacifiedblood from the left renal vein The right panel illustrates proper placement of a Greenfield IVC filter within the IVC
Source Images courtesy of Kyung J Cho MD
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 884
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
IVC and deployed with the tip near the level of therenal veins If any circumstance prevents the use of flu-oroscopy alternative imaging methods for guidingplacement include IVUS and transabdominal ultra-sound in carefully selected patients The use of IVUSor transabdominal ultrasound for IVC filter placementallows filters to be deployed at the bedside which isan advantage in the ICU setting for patients who maynot be stable enough to travel to the fluoroscopy suite
When a vascular variant or IVC thrombus is iden-tified changes in filter position are mandated For aduplicated IVC options include placing a filter ineach IVC or simply placing a suprarenal filter (the leftIVC typically draining to the left renal vein) For acircumaortic left renal vein the IVC filter is posi-tioned below the lowest renal vein inflow or if theinfrarenal cava is not long enough the filter is placedsuprarenally Finally if caval thrombus is identified thefilter is placed above the thrombus suprarenally ifnecessary Care must be taken to ensure that the fil-ter legs do not land in the renal vein ostia becausethis can cause filter instability At the completion offilter placement venous access is removed Typicallylittle bleeding occurs with hemostasis by manual pres-sure for a few minutes even with 15 French deliverysystems
Caval Filtration Efficacy Morbidity and MortalityIn 1992 Becker et al14 reviewed numerous case serieson the 7 filters available or in testing at that time(Original Greenfield Titanium Greenfield SimonNitinol Birdrsquos Nest Venatech LGM Amplatz andGuumlnther) evaluating safety indications and effec-tiveness The combined individual series evaluated2557 patients of whom 64 received Greenfield fil-ters In this group 8 deaths were reported from recur-rent PE after filter placement and 4 deaths were relatedto filter complications The authors concluded thatwhen possible anticoagulation should be continuedor reinstituted after filter placement
The PREPIC study by Decousus et al1112 whichwas the only prospective randomized trial of filter ef-ficacy evaluated the effectiveness of filters forpreventing recurrent PE The study used concomitantanticoagulation with either nonfractionated heparinor enoxaparin no group received filters without un-dergoing anticoagulation and no control was
untreated This study randomized 200 patients pergroup into those who were treated with a filter andthose who were not and then randomized them againinto groups receiving unfractionated heparin or enoxa-parin Four filters were used including the VenaTechLGM Titanium Greenfield Cardial (Bard Saint-Ettienne France) and Birdrsquos Nest with the VenaTechLGM most commonly used (56)
Patients have now been followed-up for 8 yearswith findings first published at 2 years Initially pa-tients treated with a filter experienced fewer recur-rent pulmonary emboli with relatively no instancesof major filter-related complications No difference inmortality was noted between the filter and no filtergroups either immediately or at 2 years The studyalso found that at 2 years filter-related morbidity (re-current DVT filter site thrombosis) increased signif-icantly (208 in the filter group vs 116 in the nofilter group) Findings at 8 years again showed early pro-tection against pulmonary emboli but no change inmortality Recurrent DVT remained higher in patientstreated with a filter probably because of insertion-sitethrombosis however this did not result in a signifi-cantly greater incidence of post-thrombotic syndrome
Previous evaluations of filter-related morbidityand mortality have reported widely varied incidenceSeveral authors analyzed the potential complicationsrelated to IVC filter placement identifying events im-mediately related to placement and later complica-tions Short-term complications include standardpercutaneous risks for infection and bleeding and pro-cedural mishaps such as pneumothorax (for jugularplacement) arterial injury and arteryndashvein fistula for-mation Also included in this category is filter mis-placement or maldeployment (tilting greater than 14degfound to reduce efficacy of filtration) and failure toidentify collateral pathways of thromboembolusmigration (circumaortic renal vein duplicated IVCupper-extremity source) Later complications includefilter migration caval penetration recurrent pul-monary emboli caval occlusion and guidewireentanglement during subsequent venous access place-ment15 Filter migration is a rare event with a reportedincidence of 0216 Caval perforation from retentionhooks is a common event with all filters although itis rarely clinically significant and is the subject of iso-lated case reports
Recurrent pulmonary emboli and caval occlusionare the most significant post-placement complications
Original Article 885
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 885
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
IVC and deployed with the tip near the level of therenal veins If any circumstance prevents the use of flu-oroscopy alternative imaging methods for guidingplacement include IVUS and transabdominal ultra-sound in carefully selected patients The use of IVUSor transabdominal ultrasound for IVC filter placementallows filters to be deployed at the bedside which isan advantage in the ICU setting for patients who maynot be stable enough to travel to the fluoroscopy suite
When a vascular variant or IVC thrombus is iden-tified changes in filter position are mandated For aduplicated IVC options include placing a filter ineach IVC or simply placing a suprarenal filter (the leftIVC typically draining to the left renal vein) For acircumaortic left renal vein the IVC filter is posi-tioned below the lowest renal vein inflow or if theinfrarenal cava is not long enough the filter is placedsuprarenally Finally if caval thrombus is identified thefilter is placed above the thrombus suprarenally ifnecessary Care must be taken to ensure that the fil-ter legs do not land in the renal vein ostia becausethis can cause filter instability At the completion offilter placement venous access is removed Typicallylittle bleeding occurs with hemostasis by manual pres-sure for a few minutes even with 15 French deliverysystems
Caval Filtration Efficacy Morbidity and MortalityIn 1992 Becker et al14 reviewed numerous case serieson the 7 filters available or in testing at that time(Original Greenfield Titanium Greenfield SimonNitinol Birdrsquos Nest Venatech LGM Amplatz andGuumlnther) evaluating safety indications and effec-tiveness The combined individual series evaluated2557 patients of whom 64 received Greenfield fil-ters In this group 8 deaths were reported from recur-rent PE after filter placement and 4 deaths were relatedto filter complications The authors concluded thatwhen possible anticoagulation should be continuedor reinstituted after filter placement
The PREPIC study by Decousus et al1112 whichwas the only prospective randomized trial of filter ef-ficacy evaluated the effectiveness of filters forpreventing recurrent PE The study used concomitantanticoagulation with either nonfractionated heparinor enoxaparin no group received filters without un-dergoing anticoagulation and no control was
untreated This study randomized 200 patients pergroup into those who were treated with a filter andthose who were not and then randomized them againinto groups receiving unfractionated heparin or enoxa-parin Four filters were used including the VenaTechLGM Titanium Greenfield Cardial (Bard Saint-Ettienne France) and Birdrsquos Nest with the VenaTechLGM most commonly used (56)
Patients have now been followed-up for 8 yearswith findings first published at 2 years Initially pa-tients treated with a filter experienced fewer recur-rent pulmonary emboli with relatively no instancesof major filter-related complications No difference inmortality was noted between the filter and no filtergroups either immediately or at 2 years The studyalso found that at 2 years filter-related morbidity (re-current DVT filter site thrombosis) increased signif-icantly (208 in the filter group vs 116 in the nofilter group) Findings at 8 years again showed early pro-tection against pulmonary emboli but no change inmortality Recurrent DVT remained higher in patientstreated with a filter probably because of insertion-sitethrombosis however this did not result in a signifi-cantly greater incidence of post-thrombotic syndrome
Previous evaluations of filter-related morbidityand mortality have reported widely varied incidenceSeveral authors analyzed the potential complicationsrelated to IVC filter placement identifying events im-mediately related to placement and later complica-tions Short-term complications include standardpercutaneous risks for infection and bleeding and pro-cedural mishaps such as pneumothorax (for jugularplacement) arterial injury and arteryndashvein fistula for-mation Also included in this category is filter mis-placement or maldeployment (tilting greater than 14degfound to reduce efficacy of filtration) and failure toidentify collateral pathways of thromboembolusmigration (circumaortic renal vein duplicated IVCupper-extremity source) Later complications includefilter migration caval penetration recurrent pul-monary emboli caval occlusion and guidewireentanglement during subsequent venous access place-ment15 Filter migration is a rare event with a reportedincidence of 0216 Caval perforation from retentionhooks is a common event with all filters although itis rarely clinically significant and is the subject of iso-lated case reports
Recurrent pulmonary emboli and caval occlusionare the most significant post-placement complications
Original Article 885
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 885
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
of IVC filters However in studies reporting recurrentPE after filter placement some of those events mayhave been caused by alternative embolic pathwaysand not necessarily a failure of the filter One study of318 patients treated with IVC filters reported a re-current PE rate of only 3117 Athanasoulis et al18
reported on a 26-year experience with filters placed in1731 patients observing a 56 rate of recurrent PEThis study also reported a caval thrombosis rate of27
Greenfield and Proctor19 examined prospectivedata on 2109 patients identifying 465 with long-termfollow-up (mean 9 years) from the Michigan filterregistry In this group the rate of recurrent DVT wasfound to be 12 of 241 patients treated with antico-agulation in addition to filter and 15 in 224 pa-tients treated only with filter (Table 3)
IVC Filters in Patients with MalignancyMalignancy is known to carry a significantly increasedrisk for venous thromboembolism Trousseau20 firstnoted this association in the mid-19th centuryAlthough poorly understood for decades several con-tributing factors have been identified recently in-cluding procoagulant factors produced by tumor cellsstimulation of normal coagulant and thrombotic fac-tors in the blood by tumor cells inflammatory medi-ators and even chemotherapy21 The reportedincidence of PE in the literature is somewhere be-tween 7 and 50 in patients with malignancy22 Twostudies estimated that the risks for PE in cancer patientswere approximately 36-fold higher than in those with-
out malignancy2324 The patients at increased risk forVTE also appear to be at increased risk for bleedingwhile undergoing anticoagulation therapy232526
One of the large studies mentioned earlier24 eval-uated the use of anticoagulation therapy in patientswith and without cancer Not surprisingly the recur-rence of thromboembolism in patients with cancerwas significantly greater than in those without cancerAt 12 months 207 of patients with cancer had ex-perienced recurrent thromboembolism compared with68 of those without The same study also found amuch higher rate of major bleeding complications re-lated to anticoagulation in patients with cancer At 12months the incidence of major bleeding in patientswith cancer was 124 compared with 49 in pa-tients without cancer Based on these findings it isreasonable to state that cancer patients have a high riskfor thromboembolic disease and a high risk for signif-icant complications if the standard paradigm of anti-coagulation is used
Debate about the use of IVC filters in malignantdisease has persisted since the 1990s Despite frequentuse in this setting and continued attempts to clarifytheir role their proper use in malignancy remains con-troversial A retrospective study in 2002 questioned theclinical benefit of vena cava filter placement in pa-tients with malignant disease27 In this study 116 pa-tients with typical indications for filter placement anda wide array of primary tumors (25 gastrointestinal24 lung 14 breast 14 gynecologic 12 prostate 8 hema-tologic 4 genitourinary and 15 other) underwentGreenfield filter placement over 8 years (1993ndash2000)Most patients had stage IV disease Proceduralcomplications were negligible and late complicationsincluded 2 patients with clinically apparent recurrentPE and 2 patients with progressive DVT after IVCfilter placement Survival rates were 688 at30 days 494 at 3 months and 268 at 1 year Of 91 patients with stage IV disease 42 had died at 6 weeks and only 13 were alive at 1 year Jarrett et al27 suggested that IVC filter placement may notclinically benefit patients with cancer and may notbe cost-effective given overall survival rates
Another recent retrospective study that exam-ined the benefit of IVC filters for cancer patients22
clearly showed the association between newly diag-nosed PE and the presence of metastasis This studyincluded 99 patients 55 with PE as the presentingsymptom of new metastatic disease In 12 of patients
886 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
Table 3 Summary of Filter Complicationsbull Recurrent pulmonary embolism
bull Progression of or new deep venous thrombosis
bull Iliocaval thrombosis
bull Access-site thrombosis
bull Filter migration
bull Caval perforation
bull Wire entanglement
bull Procedural complications
Arterial injury
Arteryndashvein fistula formation
Infection
Bleeding or hematoma
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 886
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
the diagnosis of PE predated malignancy Acute PEwas present in 52 of patients at cancer diagnosisand 34 of PE events were associated with new metas-tasis These patients had a 40 rate of recurrent PEwith the presence of new metastasis history of PEand multiple neutropenic episodes identified as risk fac-tors for recurrent PE Mean survival was 30 months andwas significantly worse in patients with PE at cancerdiagnosis and those who could not tolerate anticoag-ulant therapy in conjunction with IVC filter place-ment The study suggests that patients with newlydiagnosed metastatic disease a history of PE or mul-tiple episodes of neutropenia may benefit from filterplacement assuming their quality of life and life ex-pectancy are reasonable
Regarding expense and cost-effectiveness of IVCfilters in cancer patients Marcy et al28 retrospectivelyreviewed data from 1994 to 2000 for 30 patients fromone hospital Six of the 30 patients died before dis-charge and 2 died from renal vein thrombosis aftersuprarenal IVC filter placement Even allowing forthese deaths 76 56 and 40 were alive at 1 3and 6 months respectively with an improved qual-ity of life in at least 53 of the patients The authorsconcluded that the low complication rate and lowcost (relative to the mean cost of hospitalization forPE 2) favor IVC filter placement if medically in-dicated
This complex issue has no simple answerUndoubtedly a subset of cancer patients even somewith metastatic disease will benefit from IVC filterplacement However large groups who have advanceddisease and short life expectancy are unlikely to deriveany clinical benefit from filter placement The au-thors contend that patients with cancer who havestandard indications for an IVC filter should not bedenied the procedure on a purely fiscal basis
Recommended Use of IVC Filters in Patients with MalignancyThe best strategy for treating thromboembolic diseasein cancer patients depends on the severity of under-lying disease and life expectancy If possible patientsshould undergo anticoagulation In patients with ad-vanced disease poor life expectancy and significantcontraindications to anticoagulation nontreatmentis a reasonable course and has been proposed by otherauthors27 Patients with VTE and right heart failure
those with pulmonary hypertension or others unlikelyto tolerate PE may undergo prophylactic filter place-ment in addition to anticoagulation
Nonterminal patients with contraindications toanticoagulation would benefit most from filter place-ment VTE may be clinically detectable in as many as15 of cancer patients recurrent thromboembolismis also twice (or more) as likely to occur in cancer pa-tients2124 A subset of patients with a limited period ofsusceptibility to DVT or PE (eg during the periop-erative period) or with a limited period in which theycannot undergo anticoagulation (ie during the pe-rioperative period or while undergoing chemother-apy) may benefit from optional filter placement withsubsequent retrieval assuming that treatment of theirmalignancy has a reasonable likelihood of successThis latter category may become more prevalent inthe future as treatment for malignancies improvesFor patients with suspected or known incurable diseasethe best plan currently seems to be a permanent filterwhen caval interruption is indicated
ConclusionsThe frequency of and indications for caval filter place-ment are increasing with concurrent increases in thenumber and types of IVC filters Sound evidence in-dicates that IVC filters reduce the risk for recurrentPE Published reports regarding morbidity relatedto IVC filters (eg recurrent DVT extension of DVTiliocaval thrombosis) conflict although the overallcomplication rates associated with IVC filters areconsidered acceptable especially considering the mor-bidity and mortality of historical surgical methods forpreventing PE included up to 50 incidence of lowerextremity edema almost 8 incidence of recurrentPE and 15 hospital mortality29 IVC filters are usefulfor treating VTE in patients with contraindicationsto or complications associated with anticoagulationRecent studies suggest that IVC filters may also ben-efit patients with cancer and markedly higher risk for VTE (assuming sufficient life expectancy)Retrievable filters are an excellent option for patientswith a short definable period during which antico-agulation is contraindicated Retrieving the filterwhen the patient can safely undergo anticoagulationwould theoretically reduce the major shortcoming ofpermanent filters
Original Article 887
Inferior Vena Cava Filters
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 887
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
JN049_Jrnl_40909rechtqxd 92706 744 PM Page 888
15 Joels CS Sing RF Heniford BT Complications of inferior vena cavafilters Am Surg 200369654ndash659
16 Athanasoulis CA Complications of vena cava filters Radiology1993188614ndash615
17 David W Gross WS Colaiuta E et al Pulmonary embolus aftervena cava filter placement Am Surg 199965341ndash346
18 Athanasoulis CA Kaufman JA Halpern EF et al Inferior vena cavalfilters review of a 26-year single-center clinical experience Radiology200021654ndash66
19 Greenfield L Proctor M Recurrent thromboembolism in patientswith vena cava filters J Vasc Surg 200133510ndash514
20 Trosseau A Phlegmasia alba dolens In Clinique Medicale de lrsquoHotelDieu de Paris 2nd edition Paris Balliere 1865
21 Caine GJ Stonelake PS Lip GY et al The hypercoagulable stateof malignancy pathogenesis and current debate Neoplasia 20024465ndash473
22 Lin J Proctor MC Varma M et al Factors associated with recurrentvenous thromboembolism in patients with malignant disease J VascSurg 200337976ndash983
23 Gitter MJ Jaeger TM Petterson TM et al Bleeding and throm-boembolism during anticoagulant therapy a population-based studyin Rochester Minnesota Mayo Clin Proc 199570725ndash733
24 Prandoni P Lensing AW Piccioli A et al Recurrent venous throm-boembolism and bleeding complications during anticoagulant treat-ment in patients with cancer and venous thrombosis Blood20021003484ndash3488
25 Ihnat DM Mills JL Hughes JD et al Treatment of patients with ve-nous thromboembolism and malignant disease should vena cavafilter placement be routine J Vasc Surg 199828800ndash807
26 Krauth D Holden A Knapic N et al Safety and efficacy of long-termoral anticoagulation in cancer patients Cancer 198759983ndash985
27 Jarrett B Dougherty M Calligaro K Inferior vena cava filters inmalignant disease J Vasc Surg 200236704ndash707
28 Marcy PY Magne N Gallard JC et al Cost-benefit assessment ofinferior vena cava filter placement in advanced cancer patientsSupport Care Cancer 20021076ndash80
29 Donaldson MC Wirthlin LS Donaldson GA Thirty-year experi-ence with surgical interruption of the inferior vena cava for pre-vention of pulmonary embolism Ann Surg 1980191367ndash372
888 Original Article
Getzen and Rectenwald
copy Journal of the National Comprehensive Cancer Network | Volume 4 Number 9 | October 2006
References1 Streiff MB Vena caval filters a comprehensive review Blood
2000953669ndash3677
2 Hann CL Streiff MB The role of vena caval filters in the manage-ment of venous thromboembolism Blood Rev 200519179ndash202
3 Mousa SA Low-molecular-weight heparin in thrombosis and can-cer Semin Thromb Hemost 200430(suppl 1)25ndash30
4 Baldridge ED Martin MA Welling RE Clinical significance of free-floating venous thrombi J Vasc Surg 19901162ndash67
5 Pacouret G Alison D Pottier JM et al Free-floating thrombus andembolic risk in patients with angiographically confirmed proximaldeep venous thrombosis A prospective study Arch Intern Med1997157305ndash308
6 Radomski JS Jarrell BE Carabasi RA et al Risk of pulmonary em-bolus with inferior vena cava thrombosis Am Surg 19875397ndash101
9 Roehm JO Jr Gianturco C Barth MH et al Percutaneous tran-scatheter filter for the inferior vena cava A new device for treatmentof patients with pulmonary embolism Radiology 1984150255ndash257
10 Hoff WS Hoey BA Wainwright GA et al Early experience withretrievable inferior vena cava filters in high-risk trauma patients J Am Coll Surg 2004199869ndash874
11 Decousus H Leizorovicz A Parent F et al A clinical trial of venacaval filters in the prevention of pulmonary embolism in patients withproximal deep vein thrombosis N Engl J Med 1998338409ndash415
12 PREPIC Study Group Eight-year follow-up of patients withpermanent vena cava filters in the prevention of pulmonary em-bolism the PREPIC (Preacutevention du Risque drsquoEmbolie Pulmonairepar Interruption Cave) randomized study Circulation 2005112416ndash422
13 Binkert CA Sasadeusz K Stavropoulos SW Retrievability of therecovery vena cava filter after dwell times longer than 180 days J VascInterv Radiol 200617(2 Pt 1)299ndash302
14 Becker DM Philbrick JT Selby JB Inferior vena cava filtersIndications safety effectiveness Arch Intern Med 19921521985ndash1994
