Scientific Session 20Venous DiseaselThrombolysis

Research Topics, Study Designs, and Statistical Methodsin VIR.J.M. LaBerge, UCSF, San Francisco, CA, USA· W Huang·Y. Lu • D. V. Glidden

PURPOSE: Statistical analysis is the universal language ofmedical research and is a vital tool for communicating theresults of vascular and interventional radiology (VIR)procedures. Major articles in two radiology journals weresurveyed in order to characterize the research topics, studydesigns, and statistical methods seen in recent VIR researchpublications.

MATERIALS AND METHODS: We retrospectively reviewed130 major clinical VIR articles published from July 2000 ­June 2001: 72 articles (55%) from JVIR and 58 articles (45%)from Radiology. Articles were categorized by research topicand study design. Data were collected on the statisticalmethodology ofeach article.

RESULTS: Research topics included vascular intervention in65 of 130 articles (50%), non-vascular intervention in 26 (20%),vascular imaging in 23 (18%), biopsy in 9 (7%), and other S53

Abstract No. 154

Abstract o. ISS

12:41 PM

topics in 7 (5%). Study design was descriptive in 87 studies(67%), comparative in 39 studies (30%), and involvedsecondary data analysis in 4 studies (3%). Of 126 primaryclinical studies, outcome was cross-sectional (assessed at asingle time point) in 40 studies (32%) and longitudinal(measured over time) in 86 studies (68%). Median samplesize was 61.·Basic tests of association (t-test,' chi square test,etc.) were used in 71 articles (56%) and advanced tests ofassociation (regression analysis) were presented in 25 (20%).Survival analysis was applied in 34 articles (27%). Decisionstatistics such as sensitivity/specificity were useduncommonly (12%). Confidence intervals and powercalculations were reported infrequently (15% and 7%,respectively).

CONCLUSION: VIR publications focus on time-dependentoutcomes following therapeutic interventions. Readers shouldunderstand basic tests of association and survival analysis ­these include only 20 named statistical tests.

Endovenous Laser Treatment: Mid-Term Results on 200Incompetent Greater Saphenous Veins.R.J. Min, Weill Medical College a/Cornell University, NewYork, NY, USA

PURPOSE: The purpose of this study is to assess the mid­term results of endovenous laser treatment of varicose veinsdue to saphenofemoral junction (SFJ) incompetence.

MATERIALSAND METHODS: Following duplex ultrasoundmapping, greater saphenous veins (GSVs) were accessedpercutaneously at the knee level. Tumescent local anesthesia(70 ml to 130 ml of 0.2% lidocaine) was administeredperivenously under sonographic guidance. Laser energy (81 0nm wavelength diode laser; Diomed Inc., Andover, MA) wasdelivered endovenously via a 600 micron laser fiber (LaserPeripherals Inc., Minnetonka, MN). GSVs were treated fromthe SFJ to the point of entry.

RESULTS: 200 incompetent GSVs in 177 patients were treatedwith endovenous laser. Pre-treatment transverse GSVmeasurements ranged from 4.4 mm to 28 mm in diameter. 192/200 (96%) GSV segments were closed following initialendovenous laser treatment with no flow detectable by colorDoppler at one week follow-up. 7/200 (3.5%) GSVs weresuccessfully closed following re-treatment with endovenouslaser. 195/200 (98%) GSVs remain closed at 1-22 monthfollow-up. There have been no skin bums, paresthesias ordeep venous thromboses.

CONCLUSION: Endovenous laser is a new minimally invasivetreatment alternative to surgical ligation and stripping of theincompetent GSv. Mid-term results on 200 GSVs have beenimpressive with low recurrence rates at up to 22 month follow­up. Continued evaluation and longer term follow-up resultsare eagerly anticipated.

12:52 PM

1-8 Year Follow-Up of Infrainguinal Wallstents forTreatment of Venous Obstruction.PE. Thorpe, Lenox Hill Heart and Vascular Institute NewYork, USA

PURPOSE: Success in stenting occluded or stenosed iliacveins is well documented. Untreated femoral stenosis orocclusion may compromise iliac stent patency. We report ourexperience with Wallstent placement across and/or below theinguinal ligament for trcatment of symptomatic post-

Abstract No. 153

Wednesday, April 10, 200212:30 PM· 2:00 PMModerator(s): Kevin W Dickey, MD

Steven F. Millward, MD

12:30 PM

· Arteries with a normal expected diameter of 5-11 mm.

· I cm of normal vessel proximal and distal to aneurysm!traumatic injury to accommodate fixation ofthe endoluminalgraft.

Excluded lesions:

· At essential branch vessel which would be occluded by thestent-graft.

· Lesions which would require that the device cross a joint.Patients are evaluated at follow-up using spiral CT and duplexscans. Patency of the stent-graft is assessed using segmentalpressures, PVR or DWF, and duplex flow velocities.

RESULTS: Immediate post-op exclusion success is achievedin 90/105 (86%) aneurysm and 31/34 (92%) trauma lesions.Among the 18 endoleaks identified post-operatively, 13 (72%)spontaneously resolved at 1-162 days post-op. An additional9 endoleaks have been identified at follow-up (mean length offollow-up= 400 days). 4 (3%) of lesions have requiredadditional endovascular treatment for endoleak, 12 (9%) forthrombosis/stenosis and 5 (4%) have required bypass surgery.One rupture has occurred in this series in a patient for whichan endoleak was not diagnosed post-op. II % of the patientsexperienced a major complication within 30 days of theprocedure. No procedure or device related deaths have beenreported.

CONCLUSION: Interim study results suggest that PETcovered stents are useful in excluding certain types ofaneurysmal and traumatic lesions with less morbidity andmorta'lity than bypass surgery. However, lesions need to be'carefully selected and CT scan surveillance is an importantpart of patient follow-up. Long-term follow-up is necessaryto evaluate durability of this procedure.

thrombotic limbs.

MATERIALS AND METHODS: Consecutive patients referredfor symptomatic PTS were treated with endovascularintervention including thrombolytic therapy, venoplasty andstent placement. Baseline documentation of venousinsufficiency included duplex, airplethysmography andphlebography. Stents were placed in the common andsuperficial femoral segments to treat stenoses causing slowvenous flow. Poor flow was indicated by a pressure gradientofmore than 4 mmHg or contrast stasis. Follow-up includedannual duplex exams and clinical assessment. Reinterventionwith phlebography, IVUS, balloon dilatation and/or additionalstents was performed for recurrent edema/pain. Patients weremaintained on warfarin.

RESULTS: Between April 1993 and March 200 1,84 patients(44M,40F) mean age 45.3 years (range 17-83 years) hadendovenous therapy in 93 lower limbs. A total of 162 Wallstentswere placed in 49 limbs (mean 3.3 stents/limb; range 1-10stents). In 30 limbs (23L, 7R), stents (77) were extended fromthe iliac segment into the femoral vein. Mean follow-up is 48months (range 6-96 months). Primary patency of all stents is77% and assisted patency is 91 %.

, CONCLUSiON: Poor venous flow below iliac stents maylead to thrombosis or persistant post-thrombotic syndrome.Post thrombotic occlusive disease and stenosis in the femoralvein can be successfully treated with stents when there isgood inflow from the tibial and popliteal segments. Placementof femoral stents to establish continuity of flow in the deepsystem may aid in iliac stent patency. Careful patient selectionis important. Although intimal hyperplasia may occur, thiscan be effectively treated with balloon dilatation. Long-termanticoagulation in these patients is recommended.

Lower Extremity Venous Thrombolysis with AdjunctiveMechanical Thrombectomy.S. Vedantham, Mallinckrodt Institute ofRadiology, St. Louis,MO, USA· N. Parti • TM Vesely· MD. Darcy· D.MHovsepian· D. Picus

PURPOSE: To evaluate the usefulness of adding adjunctivemechanical thrombectomy (MT) to our protocol forperfonning pharmacological catheter-directed lower extremityvenous thrombolysis.

MATERIALS AND METHODS: Retrospective review of ourinterventional radiology database revealed that catheter­directed thrombolysis was used to treat 46 symptomatic limbsin 33 patients with lower extremity DVT between January1996 and July 200 I. Adjunctive MT was used in 30 limbs (21patients), of which 25 were treated for native DVT and 5were treated for intraprocedural thrombus formation. Devicesused included the Arnplatz Clotbuster (26 limbs), PossisAngiojet (8 limbs), Arrow PTD (2 limbs), and Oasis (I1irnb).Adjunctive venoplasty and stenting were performed as needed.Thrombolysis time, total pharmacological agent dose, earlyclinical success, and complications were recorded.

RESULTS: For patients undergoing pharmacological lysisalone, mean per-limb lysis time was 26.7 hours comparedwith 16.8 hours for patients with native DVT undergoinglysis with adjunctive MT (p=.038). For patients undergoingpharmacological lysis alone, mean per-limb total dose was 5.1million units urokinase (n=14) and 20.3 units reteplase (n=2),compared with 2.0 million units urokinase (n=8), 18.4 mgTPA (n=7), and 13.8 units reteplase (n= 10) for patients withnative DVT undergoing lysis with adjunctive MT. Overall,early clinical success was achieved in 38/46 (83%) limbs,

Initial Experience and One Year FoUow-Up with the TrapEase Vena Cav-a Filter.D.A. Zuckerman, University ofCincinnati, Cincinnati, OH,USA· 1.L. Foley· K.A. Rahim· A. Agha

PURPOSE: To describe a single-center experience withinsertion and follow-up of the Trap-Ease vena cava filter in aconsecutive series of patients.

Abstract No. 158

Abstract No. 157

1:25PM

The TrapEase Filter: In Vitro and In Vivo Evaluation.K.1. Cho, University ofMichigan, Department ofRadiology,Ann Arbor. Ml, USA· MC. Proctor· C.E. Luke· L.J.Greenfield

PURPOSE: The function ofa vena caval filter is determinedby its design. Prior studies demonstrated that multiple trappingsurfaces and wire type were associated with caval patency.The TrapEase (TE) filter design differs from cone-shapedfilters, so we studied its in vivo and in vitro performancecompared to the Greenfield (GF).

MATERIALS AND METHODS: 48 consecutive patients whohad Trap-Ease filters placed were followed. Clinical andradiographic data were used to evaluate filter position,effectiveness, and complications in the immediate andintermediate-term. Follow-up techniques included chartreview, telephone calls, venacavagrams, CT scans, ultrasound,and conventional abdominal radiographs.

RESULTS: There were 32 male and 16 female patients rangingin age from 22-99 (mean, 57.7). Indications for filter placementincluded contraindication to anticoagulation in the presenceof deep venous thrombosis (33) or pulmonary embolism(PE)(9), prophylaxis in trauma patients (4), and failure ofanticoagulation (3). All filters were successfully deployedfrom either a right femoral (34), left femoral (8) or right jugular(6) access without peri-procedural complication. In threesubjects, the filter was noted to be slightly tilted (less than 10degress) at the time of implantation as documented by inferiorvenacavography. There were no migrations, penetrations ofthe vena cava wall, or other interval change in filter position inthe follow-up period. No vena caval thrombosis was detectedeither by imaging or clinically. 13 of 48 (27%) patients diedduring the I-year follow-up period. Eight ofthese had systemicmalignancy. In no case was PE implicated as the cause ofdeath.

CONCLUSiON: The Trap-Ease filter can be easily deployedwithout significant risk ofcomplication. In the one-year follow­up period, it has proven to be a safe and effective means toprevent PE. More long-term follow-up is necessary to makecomparisons to other currently available fi lters.

1:14 PM

including 13/16 (81%) patients receiving pharmacological lysisalone and 25/30 (83%) patients receiving lysis with adjunctiveMT. Iliac vein and/or caval stents were placed in 20 patients.Bleeding complications occurred during 7136 (19%)procedures, including 4/13 (31 %) receiving pharmacologicallysis alone and 3/23 (13%) receiving lysis with adjunctiveMT (p=.19). No intracranial bleeding occurred.

CONCLUSION: The use of adjunctive MT to augmentpharmacological catheter-directed thrombolysis of lowerextremity DVT provides comparable short-term clinical resultswhile reducing the required thrombolytic dose and infusionduration. . .

Abstract No. 1561:03 PM

854

-- ---- - --- - -- - --- ---- - ~= - ----

MATERIALS AND METHODS: In vitro tests with a simulated circulatory circuit evaluated clot capture, resistance to migration and pressure from volume loading. These tests are predictive of filter performance in animal studies. In vivo studies were done in 6 sheep. I TE and 1 GF were placed in each and 4 were challenged with autologous clot. Plain films, contrast venograms and IVUS were obtained at placement and sacrifice. At necropsy, the cavae were formalin-fixed in situ for histology.

RESULTS: In the in vitro test the TE had better clot capture and fixation (P<.OOI) but generated greater pressure when filled with thrombus (P<.OO 1). The in vivo test showed that the TE filter diameter increased by 8 mm vs 4 mm for the GF (P =.04) and the TE filter ribs penetrated the liver and pericaval tissue (3/6 vs 0/6). The devices were comparable with respect to clot capture and fixation. One animal with the suprarenal TE suffered 30 day renal insufficiency. Fibrous bands webbed the IVC following thrombus resolution in the TE. Penetration was limited with the GF and extruded hook points were covered by a fibrous nodule. Neointimal hyperplasia was heavy with the TE compared to the GF.

CONCLUSION: Filter flow studies demonstrated that stagnant areas develop when thrombus is trapped by multiple tiers. This results in a reduction of flow and increased pressure. Clot lysis is slowed in corresponding areas and heavy fibrous webbing is found. The TE filter effectively captures clot and resists migration but at the expense of caval obstruction and penetration of adjacent structures. The double tiered trapping surface, and the amount of broad wire in contact with the wall results in caval disruption. Outcomes are similar for filters sharing these design characteristics. Conical shaped filters with a single trapping level and limited rvc wall contact capture significant thrombus while maintaining caval patency and limiting penetration.

1:36PM Abstract No. 159

Gunther Tulip Filter Retrievability Study: Initial Single Center Experience. J.A . Kaufman, Dotter Interventional Institute, Portland, OR, USA· R.T. Andrews· R.E. Barton· Pc. Lakin· B.D. Petersen· FS. Keller

PURPOSE: Clinical study to collect data evaluating the safety and outcome of retrieval of the Gunther Tulip Vena Cava Filter as part of a FDA approval process.

MATERIALS AND METHODS: 13 patients (6 female, 7 male, mean age 48 .5 ± 10 years) were enrolled in an IRE approved protocol. Adult patients at temporary high risk (14 days or less) for pulmonary embolism (PE) with patent jugular veins and life expectancy greater than 6 months were eligible. Filters were removed up to 14 days post-implant if patient was no longer at risk ofPE and trapped embolus filled less than 25% of the filter cone. Biplane cavography was obtained pre- and post- implant, and pre-removal. Single plane cavogram was obtained post removal. Follow-up at 3 months after removal will be with abdominal CT scan and access site US. No imaging follow-up was obtained when the filter was intentionally not retrieved.

RESULTS: Indications were trauma (6), high-grade subarachnoid bleed (2), post-op PE (2), pre-op deep vein thrombosis (DVT)(2), and heparin failure (1). Retrieval was not attempted in 7 patients (2 died of pre-existing conditions prior to 14 days, 2 had ongoing contraindications to anticoagulation at 14 days, 2 had large volume of trapped embolus in the filter on pre-removal cavogram, and 1 had recurrent PE with filter occlusion at 9 days). Filter retrieval was attempted and completed in 6/13 (46%). There were no

technical failures. Mean implantation time of retrieved filters was 12.2 ± 2 days. No complications during retrieval occurred. One retrieved filter had migrated to the right atrium (asymptomatic), possibly due to manipulation during surgery. No symptomatic recurrent PE Or DVT occurred after filter removal at mean follow-up of 53.5 ± 24 days. All patients have been on enoxaprin post filter removal. No CT scans or US had been obtained at the time of abstract submission (per protocol).

CONCLUSION: When implanted with the intention to retrieve within 14 days, removal of the Gunther Vena Cava filter was feasible in 46% of patients. All non-retrievals were due to clinical circumstances. No immediate or short-term complications occurred related to filter retrieval.

1:47PM Abstract No. 160

FDA Reported Complications of Inferior Vena Cava Filtration by Device Type. A. V. Rao, Beth Israel Deaconess Medical Center, Boston, MA, USA· R. G. Sheiman • E. V. Lang

PURPOSE: To assess complications of various IVC filter types reported to the FDA.

MATERIALS AND METHODS: The FDA's Manufacturer and User Facility Device Experience (MAUDE) database was searched for filter complications data from the last one hundred reports (from 9/5/00 to 6/25/0 I). The reports were categorized by filter type and complication type. Complication types were categorized as caval thrombosis, deployment (e.g, incomplete opening, entangled legs), migration/malposition, or other (e.g., retroperitoneal hematoma, fractured strut). The companies distributing the respective filters were asked to provide the total number of each filter sold during the study period.

RESULTS: Filter distribution and complications per filter type are shown in the table. The majority of complications for all filters except the Trapease were related to deployment (n=51). Caval thrombosis was reported only with the Trapease filter (n=25). The device manufacturers did not disclose exact sales data. One company supplied the following figures as approximates: 75,000-80,000 filters placed per year with an estimated market share of 50% for the Greenfield filter, 20% for the Trapease filter, 10% for the Simon Nitinol filter, 10% for the Venatech, and 5% for the Bird 's nest filter.

CONCLUSION: The lack of uniform complication reporting and the lack of published market data preclude the calculation of the true incidence of complications. The FDA data show: I) the majority of reported filter complications was related to deployment 2) a large number of thromboses were reported for the Trapease filter as compared to the other filters. While the market share data are estimates, the incidence of thrombosis is out of proportion to this filter's market share.

Comparison

Filter Type ~ ~ ~ Olher Total Trapease 25 8 I 5 39 Slainless Steel Greenfield 0 22 7 33 Titanium Greenfield 0 3 12 Simon Nitinol 0 0 8 Venatech 0 Bird'sNest 0 3 Total 25 51 12 I2 100

S55