Venous Thrombolysis

Thursday, March 1, 1996 diameter. Delayed penetration of the caval wall and surrounding structures (reported rate, 10/02%) has been documented for most filters (apparently extremely rare for the VenaTech filter), but infrequently has clinical consequence. Cbaracteristics of an Ideal Filter No perfect filter exists, and complications have been documented for all filters. Due to design considerations, each of the available filters may have advantages or disadvantages in a particular case. The ideal filter should be biocompatible over a long period, be effective in preventing PE, not migrate or penetrate, and not cause or contribute to thrombosis. Highly desirable characteristics would also include ease of insertion, lack or interference with other imaging studies, low cost, and potential for retrieval.

The safety and effectiveness of vena caval filters are seriously questioned by some investigators due to the lack of any large-scale, well-controlled clinical trials or comparisons of the different filters. Further evaluation of these devices is indicated. All radiologists inserting filters should be aware of the indications and potential complications of vena caval filters and use appropriate judgement in deciding the need for a vena caval filter in each patient and in each clinical situation.

Sellected Bibliography Athanasoulis CA. Complications of vena cava filters. Radiology 1993; 188:614-615. Becker DM, Philbrick JT, Selby JB. Inferior vena cava filters: indications, safety, effectiveness. Arch Intern Med 1992; 152: 1985-1994. Dorfman GS. Evaluating the roles and function of vena caval filters: will data be available before or after these devices are removed from the market? Radiology 1992; 185:1517. Dorfman GS. Percutaneous inferior vena cava filters. Radiology 1990; 174:987-992. Ferris EJ, McCowan TC, Carver DK, McFarland DR. Percutaneous inferior vena caval filters: follow-up of seven designs in 320 patients. Radiology 1993; 181:851-856. Grassi CJ. Inferior vena caval filters: analysis of five currently available devices. AJR 1991; 156:813-821. Greenfield LJ, Cho KJ, Proctor M, et al. Results of a multicenter study of the modified

hook-titanium Greenfield filter. J Vasc Surg 1991; 14:253-257. Greenfield LJ, Proctor MC, Cho KJ, et al. Extended evaluation of the titanium Greenfield vena caval filter. J Vasc Surg 1994; 20:458-465. McCowan TC, Ferris EJ, Carver DK, Molpus WM. Complications of the Nitinol vena cava filter. ]VIR 1992; 3:401-408. Molgaard CP, Yucel EK, Geller SC, Knox TA, Waltman AC. Access-site thrombosis after placement of inferior vena cava filters with 12-14F delivery sheaths. Radiology 1992; 185:257. Murphy TP, Dorfman GS, Yedlicka JW, et al. LGM vena caval filter: objective evaluation of early results. ]VIR 1991; 2:107-115. Roehm JOF, Johnsrude IS, Barth KH, Gianturco C. Bird's nest inferior vena cava filter: progress report. Radiology 1988; 168:745749. Rose BS, SImon DC, Hess ML, Van Arnan ME. Percutaneous transfemoral placement of the Ki=ay-Greenfield vena cava filter. Radiology 1987; 165:373-376. Simon M, Athanasoulis CA, Kim D, et al. Simon Nitinol inferior vena cava filter: initial clinical experience. Radiology 1989;172:99103. Vesely TM. Technical problems and complications associated with inferior vena cava filters. Semin Intervent Radiol 1994; 11:121-133. 11:10 am

Venous Tbrombolysis Charles P. Semba, MD

Learning objective: To discuss the rationale, techniques, and results of catheter-directed thrombolysis for treating iliofemoral deep venous thrombosis at Stanford University.

Catheter-directed thrombolysiS is a new experimental technique used for treating acute proximal deep venous thrombosis (DVT) of the lower extremity 0-4). The most common form of treatment for DVT is systemic anticoagulation. It has been the cornerstone of therapy since the 1960s. In our experience at Stanford University, anticoagulation alone in patients with iliofemoral DVT rarely leads to recanalization of the obstructed venous segment. While the fibrinolytic system can lyse small venous thrombi, such as isolated popliteal or calf vein

337

Thursday, March 7, 1996

II

DVT, complete endogenous thrombolysis of the larger femoral and iliac is a rare occurrence. Consequently, the thrombus organizes. This leads to permanent occlusion of the vein frequently with valvular damage. Anticoagulation regimens with unfractionated heparin, lowmolecular weight heparin, or sodium warfarin do not actively promote thrombolysis but merely provide prophylaxis from further thrombus formation. Anticoagulation alone does not actively promote fibrinolysis. While technologic advances in vascular imaging and surgery have greatly improved our understanding of vascular disease, therapeutic achievements for DVT lag behind clinical breakthroughs in arterial disease for three major reasons:

Perception that "thinning the blood" with heparin promotes clot lysis. Despite the widespread use of heparin and warfarin in daily clinical practice, a continued perception among many primary physicians is that "blood thinning" helps speed the enzymatic clearing of occlusive thrombi. The clinical data are to the contrary. Krupski et al (5) demonstrated with Doppler studies that up to 40% of patients will continue to propagate venous thrombus even with adequate doses of intravenously administered heparin. A comprehensive literature review by Sherry (6) estimated that DVT resolved with heparin therapy alone within 10 days in only 6% of patients, Unfractionated heparin is a mucopolysaccharide isolated from bovine or porcine intestine or lung and has no anticoagulation properties alone. Heparin binds to antithrombin III to form an active complex that neutralizes thrombin and other coagulation proteins (XII, XI, IX, X). Even the newer low-molecular weight heparin (LMWH) compounds do nothing to activate thrombolysis directly and merely prevent more thrombus from forming; they may offer better prophylaxis than unfractionated heparin (7). The strategic advantage of LMWH is to reduce overall costs by reducing or eliminating hospital stay. Administration of the compound is by subcutaneous injection and based on total body weight. No interim laboratory values (such as partial thromboplastin time) are measured while the patient is given orally administered warfarin.

Lack of understanding of the venous pathophysiology and the long-term sequelae of DIT. The

338

major focus of managing the patient with DVT is on preventing the complication of pulmonary emboli. In our experience, the long-term sequelae of DVT in the lower extremities is typically underappreciated. Patients with chronic venous disease are as disabled as patients with acute arterial ischemia, yet aggressive treatment of lower extremity DVT is not often pursued. Often, anticoagulation is followed by a hope for symptomatic improvement by development of venous collaterals. Over the long term, patients with postphlebitic syndrome become a significant socioeconomic burden due to chronic disability and frequent hospitalizations (8,9), After 5 years, 95% of patients with iliofemoral DVT who were treated with anticoagulation alone never have normal flow dynamics of the limb due to development of severe valvular insufficiency and outflow obstruction (10).

Lack of a simple surgical solution. Surgery for venous thrombosis consists either of a balloon embolectomy for acute DVT or a bypass for venous outflow obstruction (11). Despite early favorable reports about venous thrombectomy, most vascular surgeons have abandoned the procedure due to high re-thrombosis rates (12). A cross-over femoral bypass, typically reserved for patients with severe venous outflow obstruction, requires reinforced polytetrafluoroethylene combined with a small arteriovenous fistula to maintain brisk flow (13). Surgical interventions in venous disease have long met with frustration because of low velocity flow, tendency for early thrombosis, and the difficult nature in handling thin-walled, compliant veins. Overall, no surgical technique for treating acute or chronic DVT is generally accepted in the United States. Understandably, most vascular surgeons are frustrated in their attempts to help these patients with chronic venous disease. Rationale for Thrombolytic Therapy Because of our experience with venography for patients with chronic lower extremity veno-occlusive disease and because we are aware of the frustrations of patients, internists, and vascular surgeons, we began to ask whether a better therapeutic alternative to anticoagulation for the patient with an iliofemoral DVT was possible. Over the past decade, catheter-directed thrombolysis for treating occluded dialysis access fistulas, arterial thromboses, and upper extremity DVT (eg, effort vein thrombosis) has developed (14-16). The rationale for catheterdirected thrombolytic therapy is to remove the

Thursday, March 7, 1996 thrombus quickly, alleviate painful limb edema, preserve valve function, prevent pulmonary emboli, and prevent postphlebitic syndrome. Are thrombolytic agents superior to anticoagulation alone? In a comprehensive review of 13 clinical studies by Comerota and Aldridge (17), involving nearly 600 patients with acute DVT, 63% of the patients treated with systemic thrombolytic agents had complete or partiallysis of the DVT compared with 18% of patients treated with anticoagulation alone. Emerging evidence shows that early thrombolysis preserves valve function based on serial duplex examinations (18). Catheter-directed Endovascular Therapy for Iliofemoral DVT The main advantages of a catheter-based approach are that it provides superior efficiency of thrombolysis compared with systemic infusions, reduces the risk of systemic fibrinolysis and bleeding, and allows endovascular access for adjunctive therapeutic techniques such as balloon angioplasty and stent placement.

Patient Selection Patients undergoing consideration for catheterdirected thrombolysis have US-documented lower-extremity DVT or ascending venography involving the iliac veins with or without involvement of the inferior vena cava or the femoral, popliteal, or calf veins. Main contraindications are patients who cannot undergo anticoagulation, those with bleeding disorders, those who are pregnant or have recently given birth, those with metastatic disease involving the central nervous system, and those who have had a hemorrhagic stroke within the previous year. Recent surgery is a relative contraindication; we generally do not treat patients within 7 days of a major operation. Technique The endovascular procedure is performed entirely in the angiography suite with a high-resolution, 40 cm diameter (1024 X 1024 matrix) digital subtraction image intensifier and a tilting examination table. Besides the interventional radiologist, two licensed vascular technologists and one registered critical care nurse staff the angiography suite. The patients are fully monitored (EKG, pulse oximeter, blood pressure) and given intravenously administered sedation. Vascular access is either by way of the right internal jugular vein (with the patient supine) or

the ipsilateral popliteal vein (with the patient prone). Following sheath placement, a baseline inferior vena cava venogram is obtained to document any caval extension of the thrombus. The thrombosed iliac vein is probed with a steerable hydrophilic guide wire and catheter under fluoroscopic control. After complete passage of the wire across the thrombosed segment of vein, the catheter is advanced and venography is repeated to document the inferior extent of thrombus and to visualize the major collateral pathways.

II

Occasionally, the right internal jugular vein approach is difficult due to the buckling of catheters into the right atrium or inability to engage an iliac thrombus that is flush with the lateral wall of the inferior vena cava. The patient is then placed prone, and a popliteal access is used. The popliteal vein is visualized under US guidance with a 7- or 9-MHz phased-array transducer. A single wall puncture is made with a 5-F rnicroaccess set. The popliteal approach is simpler in several ways. It provides better mechanical advantage, making it easier to pass the catheter into the occluded iliofemoral segment. It involves less valve damage during catheter placement into the thrombosed femoral segment. It also allows use of shorter catheters compared with the jugular approach. Thrombolysis is delivered locally by using catheter-directed techniques. The agent of choice is urokinase due to its reliability, consistency, wide margin of safety, and predictability. The use of streptokinase is hampered by immunogenic complications and high bleeding complications. The tissue plasminogen activator (t-PA) is used for bolus infusions for acute myocardial infarction and results in more bleeding complications when used for prolonged infusions. Thrombolysis systems are either co-axial (5-F end-hole catheter and 0.035-in thrombolytic guide wire) or catheter (4.7-F dual-lumen multi-side-hole catheter) systems. Generally, both work equally well with acute thrombus. We have not found a particular advantage of one over the other. When the jugular approach is used, the catheters are positioned so that the lysis wire or multi-side-hole catheter is placed with its tip at the leading edge of the clot. The end-hole catheter or side holes are placed approximately one-third of the way.into the thrombus. When the popliteal approach is used, the tip of the guide wire is placed ap-

339

Thursday, March 7, 1996 proximately two-thirds into the thrombus, and the side holes or end-hole catheter is placed at the lead edge of the thrombus closest to the access site. The quantity of urokinase infusions is typically 100,000 to 200,000 U/hr overnight (maximum, 4,400 U/hr/kg). We mix 500,000 U of urokinase in 250-mL bags of 0.9% sodium chloride to give a concentration of 2,000 U/mL. Using the coaxial or multi-side-hole catheter system requires use of three intravenous pumps: two pumps for urokinase and one pump for systemic heparin. Urokinase is given in split doses (eg, 60,000 U/hr through the thrombolysis wire and 60,000 U/hr through the end-hole catheter) for a total infusion of 120,000 U/hr (60 mL/hr). The heparin is given through the side arm of the venous sheath. The thrombus is not pre-laced. Neither do we perform pulse-spray pharmacomechanical thrombolysis. Patients are sent to a step-down unit (not an intensive care unit) where the nurses are trained in managing vascular access sheaths and are familiar with following patients being given heparin -and urokinase. The next day, the patients are re-evaluated in the angiography suite. If the thrombus is entirely resolved with excellent venographic flow, the procedure is terminated, and the patient undergoes anticoagulation. More commonly, an underlying anatomic defect is present with the iliac vein, which requires adjunctive treatment. The use of urokinase is also beneficial in selected cases of chronic iliac vein occlusion. If passing a guide wire across the occluded segment during the initial session is impossible, urokinase is infused at the leading aspect of the thrombus. In about 95% of cases, we can pass a hydrophilic wire across the occluded segment the following day, although no significant improvement is noted venographically. For long-standing iliac or caval occlusions (months to years), urokinase can be given to assist in softening the thrombus, but direct angioplasty and stenting can also be done without antecedent lysis. Angioplasty in the large veins is often not adequate to support patency of the treated segment. We use angioplasty principally to dilate the vein before stent placement and to create more room for the catheters by breaking up any organized, fibrotic synechiae or webs. Due to compliance and elastic recoil, angioplasty is

340

generally not useful as a stand-alone procedure and is typically followed with stent placement. Choice of balloon diameter depends on the size of the native iliac vein. We attempt to create the vein as large as allowable (10-14 mm). The use of high-pressure balloons capable of 12 to 20 atm of pressure are usually required with chronic occlusions. The three choices of stent are the balloon-expandable, rigid, stainless-steel Palmaz stent; the self-expanding, flexible stainless-steel Wallstent; and the rigid, self-expanding Rosch-Gianturco Z stent. We generally prefer the Wallstent for long-segment iliac vein reconstruction due to its length and ability to conform to the natural curves of the vessel. The Palmaz stent is ideal for shorter, focal lesions (:::;3 cm in length and :::;12 mm in diameter). The basic principle of stent placement is to provide as large a lumen as possible and to provide continuous, laminar, in-line flow. Any venographic imperfections, such as a local stasis, may doom the reconstructive procedure due to early re-thrombosis. Our experience in stent placement in the iliac veins and inferior vena cava is extremely positive. The large diameters of these vessels can provide for excellent durability with I-year patency [,,
Thursday, March 7, 1996 Table 2 Treatment Outcomes in 41 Limbs Treated for Iliofemoral DVI

TaMe 1 Prc~enting

Symptoms and Location and Cause of DVI in 41 Limbs

No. of Limbs (%)

No of Limbs (%) Initial symptoms Lower-extremity edema Lower-extremity pain Phlegmasia

40 41 4

(98) 000)

Type of symptoms Acute' Chronict

25 16

(61) (39)

Location of thrombus Inferior vena cava, iliac vein Inferior vena cava, iliac and femoral veins mac vein Iliac and femoral veins Left lower extremity Right lower extremity Cause Femoral vein catheter May-Thurner syndrome Oral contraceptives Postoperative DVT Radiation injury Recurrent DVT Retroperitoneal fibrosis Unknown

5 11 7 18 28 13

3 2 3 8 6 3 2 14

(10)

(2) (27) (17)

(44) (68) (32)

(7)

(5) (7) (20) (5) (7)

(5) (34)

'Average duration of acute symptoms was 10.5 days (range, 7-28 days). t Median duration of chronic symptoms was 270 days (range, 35-5,475 days). women; 24-74 years old; mean, 53 years). Analysis was based on the number of treated limbs since 3 patients with bilateral disease had improvement in one leg but not the other. Table 1 summarizes the initial clinical presentation and location of thrombi.

II

No thrombolysis Angioplasty, stent placement Thrombolysis Complete No further intervention Angioplasty Angioplasty, stent placement Partial No further intervention Angioplasty, stent placement None No further intervention Angioplasty, stent placement

5

(2)

21 8 2 11

(51)

9

(22)

1 8 2

(5)

1 1

Technical failure Occluded vein unable to be crossed with guide wire No thrombolysis achieved, no further intervention Partial thrombolysis, no further intervention Resolution of leg edema and pain Complete Partial None

4

(0)

33 2 6

(80) (5) (5)

Technical success Clinical success

35 35

(85) (85)

further intervention (25%). Adjunctive endovascular techniques were used in 22 limbs following thrombolysis because of significant residual stenoses (>50%) following urokinase treatment.

Twenty-four of the 32 patients (75%) were treated with anticoagulation with no symptomatic relief for an average of 10 days before endovascular intervention. Table 2 summarizes treatment outcomes for catheter-directed therapy.

In our limited follow-up period (range, 0-45 mo), no successfully treated patients have experienced worsening or progression of symptoms following intervention. Six patients have died during this period due to underlying malignancies. Four patients have been lost to follow-up. Clinical and Doppler US evaluation of 19 of 24 limbs treated using endovascular stents show a primary patency rate of 95% at 6 months.

Overall, the technical and clinical success rate was 85% (35 of 41). Thirty-two limbs were treated with catheter-directed thrombolysis by using an average urokinase dose of 3.5 million IU (range, 1.4 million to 16.0 million IU) infused over an average of 30 hours (range, 1574 hr). Thrombolysis was complete in 21 limbs (66%), partial in nine limbs (28%), and no lysis was achieved in two limbs (6%). Only eight limbs with complete thrombolysis required no

Complications No major complications, including pulmonary emboli or death, occurred. Surprisingly, the risk of clinically detectable pulmonary emboli is extremely low. We have abandoned the practice of obtaining baseline perfusion lung scans and using inferior vena cava filters. From a transjugular approach, working through the deployed inferior vena cava filter can be difficult. Minor risks include venous access hematoma and an-

341

Thursday, March 7, 1996 aphylactoid reactions to the urokinase infusion. Other rare complications include embolization of the stent or balloon rupture during stent placement. The interventionalist must be prepared to capture the errant stent and redeploy it in the target region or place it in a "safe-harbor" (eg, the iliac vein or inferior vena cava).

Summary Catheter-directed thrombolytic therapy and endovascular stenting is a new and promising approach for treating acute and chronic thrombotic iliofemoral venous occlusions based on our initial study of a small group of patients. With acute DVT, catheter-directed techniques provide more complete lysis than systemic infusions. Early, aggressive therapy may spare the patient from the lifelong ravages of postphlebitic syndrome by preserving valve function and eliminating the venous outflow obstruction. Immediate post-thrombolysis venography can evaluate the underlying vein and assess the need for adjunctive treatment with angioplasty and stents. Urokinase has a high degree of safety with few complications when a catheter-directed approach rather than systemic infusion is used. Patients with chronic DVT can benefit since the obstruction to venous outflow is reduced if the occlusion is limited to the iliac vein and the inferior vena cava. Long-term follow-up studies are necessary to evaluate patency rates of the treated veins, to determine whether successfully treated limbs have a lower frequency of recurrent DVT, and to determine the frequency of chronic venous insufficiency compared with that in patients treated with anticoagulation alone. Based on our initial experience, the National Venous Thrombosis Registry was established in October 1994. The purpose of this multidisciplinary registry is to prospectively document the long-term results of catheter-directed thrombolytic therapy for patients with iliofemoral DVT from 40 leading medical centers around the United States. We hope that endovascular techniques for iliofemoral DVT will significantly reduce the immediate and long-term complications commonly associated with this difficult and often misunderstood clinical problem.

References 1. Okrent D, Messersmith R, Buckman J. Transcatheter fibrinolytic therapy and angioplasty for left iliofemoral venous thrombosis.]VIR 1991; 2:195-197.

342

2. Molina JE, Hunter DW, Yedlicka]W. Thrombolytic therapy for iliofemoral venous thrombosis. Vasc Surg 1992; 26: 630-637. 3. Comerota AJ, Aldridge SC, Cohen G, et al. Strategy of aggressive regional therapy for acute iliofemoral venous thrombosis with contemporary venous thrombectomy or catheter-directed thrombolysis. J Vasc Surg 1994; 20:244-254. 4. Semba CP, Dake MD. Iliofemoral deep venous thrombosis: aggressive therapy using catheter-directed thrombolysis. Radiology 1994; 191:487-494. 5. Krupski WC, Bass A, Dilley RB, et al. Propagation of deep venous thrombosis identified by duplex ultrasonography. J Vasc Surg 1990; 12:467-475. 6. Sherry S. Thrombolytic therapy for deep venous thrombosis. Semin Intervent Radiol 1985; 4:331-337. 7. Turpie AGG, Levine MN, Hirsh J, et al. Randomized controlled trial of lowmolecular weight heparin (enoxaprin) to prevent deep vein thrombosis in patients undergoing hip surgery. New Engl J Med 1986; 315:925-929. 8. O'Donnell1F, Browse WL, Burnand KE, Thomas ML. Socioeconomic effects of an iliofemoral deep venous thrombosis. J Surg Res 1977; 22:483-488. 9. Strandness DE, Langlois YE, Cramer M, et aL Long-term sequelae of acute venous thrombosis. JAMA 1983; 250:1289-1292. 10. Akesson H, Brudin L, Dahlstrom JD, et al. Venous function assessed during a five-year period after acute iliofemoral venous thrombosis treated with anticoagulation. Eur J Vasc Surg 1990; 4:43-48. 11. Haller )A, Abrams BL. Use of thrombectomy in the treatment of acute iliofemoral venous thrombosis in forty-five patients. Ann Surg 1963; 158:561-569. 12. Edwards WH, Sawyer JL, Foster JH. Fiveyear follow-up study of iliofemoral venous thrombosis: reappraisal of thrombectomy. Ann Surg 1970; 171:961-970. 13. Plate G, Akesson H, Einarsson E, et al. Long-term results of venous thrombectomy combined with a temporary arteriovenous fistula. Eur J Vasc Surg 1990; 4:483-489. 14. Becker GJ, Rabe FE, Richmond BD, et aL Low-dose fibrinolytic therapy: results and

Thursday, March 7, 1996

15.

16.

17.

18.

new concepts. Radiology 1983; 148:663667. Bookstein ]], Valji K. Pulse-spray phannacomechanical thrombolysis. Cardiovasc Intervent Radiol 1992; 15:228233. Machleder HI. Evaluation of a new treatment strategy for Paget-Schroetter syndrome: spontaneous thrombosis of the axillary-subclavian vein. ] Vasc Surg 1993; 17:305-315. Comerota A], Aldridge sc. Thrombolytic therapy for acute deep vein thrombosis. Sem Vasc Surg 1992; 5:76-81. Meissner MH, Manzo RA, Bergelin RO, et al. Deep venous insufficency: relationship between lysis and subsequent reflux. J Vasc Surg 1993; 18:596--608.

11:30 am V~,nous

Angioplasty and Stents

Anthony C. Venbrux, MD

uarning objectives: (1) To discuss the etiology of central venous stenosis in the nondialysis patient. (2) To describe current techniques used to treat the nondialysis patient with central venous stenoses. (3) To review the medical literature (clinical and technical successes, patency and restenosis rates, and postprocedural patient management leg, the role of anticoagulant therapyD. (4) To note indications, contraindications, and complications associated with percutaneous transluminal angioplasty and stcnting of patients with central venous stenoses. (5) To discuss patient management with specific clinical examples. Central venous stenoses are the result of intrinsic or extrinsic conditions. The etiology of intrinsic venous disease includes venous thrombosis (either acute or chronic), benign conditions such as venous stenosis due to vein wall thickening and fibrosis (intimal hyperplasia), traumatic injury to the vein wall, and malignant tumors extending into the central veins or rarely tumors arising from the wall of the vein. The etiology of extrinsic venous disease includes benign and malignant processes that compress the central veins and compromise the lumen of the vessel. If extrinsic compression is severe, venous thrombosis may result. Clinical examples include perivascular fibrosis from benign cases such as inflammation or mechanical

compression. Malignant causes of extrinsic venous compression are numerous and may result in superior vena cava syndrome or inferior vena cava occlusion.

II

Overview of Currendy Available Techniques Venous obstruction, which occurs gradually, is often clinically asymptomatic due to the development of numerous venous collaterals. Yet, when central venous stenosis or occlusion occurs rapidly or sufficient collaterals fail to develop (eg, in a scarred surgical field), the patient may become symptomatic due to venous congestion. Venous obstruction may be treated by using percutaneous transluminal angioplasty (PTA), thrombolytic therapy, intravascular stenting, or a combination of interventional techniques. Thrombolytic therapy applied locally through catheter-directed infusion techniques or administered systemically may change a lengthy complete venous occlusion to a shorter, focal stenosis. PTA, though frequently applied in the central veins, is only partially effective and requires frequent repeat procedures (1). Too often, intrinsic or extrinsic conditions result in elastic recoil immediately or soon after vein PTA. Venous PTA has been applied extensively in dialysis patients with central venous and graft anastomotic stenoses. These lesions are notoriously difficult to treat because of their fibrotic nature. Percutaneous therapy of patients with venous occlusions and Budd-Chiari syndrome has traditionally consisted of PTA of the inferior vena cava or the hepatic veins. Although treatment has met with an initial high success rate with 1year patency rates of 80% 100% (including some patients requiring repeat PTA) (1), restenosis is a frequent occurrence requiring repeated angioplasty (2-5). Given the known limitations of PTA for central venous stenoses and occlusions, the percutaneous application of venous stents has become an important therapeutic option (6-11). The Wallstent, Palmaz stent, and Z stent are most commonly used for treating patients with central venous lesions. Though technical and clinical successes are impressive, long-term outcomes are not known. Prospective, randomized, multicenter trials are needed. Before percutaneous intervention, conventional supportive therapy for venous disease is implemented. This generally consists of anticoagulation, compression stockings, and elevation of

343