Dr. Macedo will be a Senior Associate Consultant with the Department of Radiology, Mayo Clinic and Foundation, Rochester, MN. Dr. Macedo received her MD from Universidade Federal de Goiás Medical School, Brazil, in 1995. She completed an internship at Albert Einstein Medical Center, Philadelphia, PA in 1996. She is completing her fellowship at the Mayo Clinic, Rochester, MN. Dr. Bjarnason is a Senior Associate Consultant with the Department of Radiology, Mayo Clinic and Foundation, Rochester, MN.

Thrombolysis is a safe and effective treatment alternative for acute peripheral thromboembolic disease. Good results have been described for both acute and subacute arterial and venous thromboembolic events. While the best pharmacologic regimen for catheter-directed infusion remains to be determined, a variety of effective regimens have been reported. Emerging alternatives such as the use of combined platelet inhibitors and thrombolytic agents as well as mechanical thrombolysis are promising. This article will discuss some of the concepts, techniques, and results of thrombolysis as well as present a brief report on the data presented at the 2002 Society of Cardiovascular and Interventional Radiology (SCVIR) annual meeting.

In 1933, Tillett and Garner discovered the lytic properties of streptokinase (SK) produced by group C beta-hemolytic streptococci. ¹ More than a decade later, clinical usefulness of SK in liquefaction and drainage of loculated hemothorax was described, ² and it was not until 1963 that the first intra-arterial use of SK was described by McNicol. ³ In 1972, Dotter et al ⁴ reported results that warranted clinical application of SK for the treatment of thromboembolic disease. Soon it became noted that for reasons such as drug resistance, high complication rates, and erratic clot lysis, SK was not an optimal agent for frequent lengthy fibrinolysis.

That prompted the development of several other drugs, mostly based on tissue plasminogen activators (t-PAs) and urokinase (UK) (Abbokinase, Abbott Laboratories, Abbott Park, IL). Urokinase was recognized in the early 1950s, and it became available for clinical use in the early 1960s, but was used mostly on an experimental basis. It was the most widely used drug for catheter-directed thrombolysis until 1999 when it was removed from the market for reasons related to the manufacturing process. This caused significant changes in the interventional radiology community, which had to adapt to the use of drugs such as alteplase (rt-PA) (Activase, Genentech, Inc., San Francisco, CA), reteplase (r-PA) (Retavase, Centocor, Inc., Malvern, PA), and tenecteplase (TNK) (TNKase, Genentech, Inc., San Francisco, CA). We are still learning about doses and other management-related issues such as concomitant anticoagulation. Urokinase is now expected to be re-released on the market later this year in its original form but with all manufacturing issues addressed.

Indications and contraindications

Thrombolysis is a routine procedure in interventional radiology practice. Most thrombolytic procedures are performed for peripheral native arterial and graft occlusions and deep venous thrombosis (DVT) of the iliac and femoral veins. Best results are seen in the acute presentation (<14 days); however, thrombolysis can be attempted in older occlusions. The indications and major contraindications for thrombolysis are summarized in Table 1. Other relative contraindications include pregnancy, bacterial endocarditis, hepatic failure, and diabetic hemorrhagic retinopathy. It is important to individualize each case, analyze the risks versus the benefits, and make a wise decision in the best interest of the patient.

Drugs

Thrombolytic drugs all work through a common path, activation of plasminogen to plasmin. There are other important drugs commonly used concomitantly with thrombolysis that have very different final effects and paths. For clarification and simplification, a summary of the relevant points in the coagulation path, as well as when the concomitant drugs act, is illustrated in Figure 1.

Plasminogen activators

The plasminogen activators or fibrinolytic drugs work by catalyzing the conversion of plasminogen to plasmin, which then causes cleavage of fibrin strands and dissolution of fibrin, resulting in thrombolysis. The plasminogen activator inhibitor (PAI-1) and alpha-2-antiplasmin counteract the action of these drugs. Important concepts in differentiating these drugs are fibrin affinity and specificity. The most commonly used drugs in this class are rt-PA and r-PA. Despite multiple reports of experience with these drugs, no randomized prospective study is available to confirm the superiority of one drug over the other. These drugs are manufactured by recombinant technology. The use of thrombolytics in the peripheral vascular system discussed in this article is an off-label indication. An overview for these drugs is found in Table 2.

Alteplase­­ Alteplase has the shortest half-life (5 min) and higher fibrin specificity when compared with UK and r-PA. It is supplied in 50- and 100-mg vials as a lyophilized powder or as 2-mg vials for catheter clearance. It is reconstituted in sterile water for injection (SWFI), USP, to give a concentration of 1 mg/mL and diluted in nonheparinized saline to a final concentration of 0.2 mg/mL. The recommended dosage by the SCVIR advisory panel is between 0.12 to 2.0 mg/h, median 0.5 to 1.0 mg/h, for a total dose of <40 mg, not exceeding 2 mg/h. ⁵ This drug has Food and Drug Administration (FDA) approval for use in myocardial infarction, pulmonary embolism, stroke, and catheter clearance.

Reteplase­­ Reteplase has high fibrin specificity and, compared with rt-PA, has longer plasma half-life (15 min), and lower fibrin affinity. Theoretically, although it has not been proven by clinical trials, the low fibrin affinity offers superior clot penetration. It is supplied in 10 U vials as a lyophilized powder and reconstituted in SWFI, USP, to a final concentration of 1 U/mL. It should be reconstituted just before use since it does not have an antibiotic preservative. Dilution with nonheparinized normal saline may yield a concentration of no more than 0.2 U/mL. The dosage varies from 0.25 to 1 U/h, to avoid exceeding 20 U or infusion times of more than 24 hours in acute thrombosis. Dosages lower than 0.25 U/h are associated with long infusion times. ⁶ This drug has FDA approval for use only in myocardial infarction.

Tenecteplase­­ The newest drug in this class has the highest fibrin specificity, longest plasma half-life (22 min), and increased resistance to PAI-1. It is supplied in 50-mg lyophilized powder vials and reconstituted in SWFI, USP, to a concentration of 5 mg/mL. For the same reason as r-PA, it should be reconstituted just before use. Concentration for a peripheral intervention (diluted in normal saline) is between 0.01 to 0.05 mg/mL. Initial studies have used a dosage of 0.25 to 0.5 mg/h. ⁷ This drug has FDA approval for use only in myocardial infarction.

Urokinase­­ This drug was used widely in the past, had a good reputation for safety, and its dosages were understood clearly for use in catheter-directed thrombolysis. However, UK is not currently available in the United States. Its fibrin specificity is relatively low compared with the recombinant plasminogen activators. The half-life is 16 min and typical dosages were 100,000 to 200,000 U/h.

Streptokinase­­ Streptokinase is an indirect plasminogen activator. Since SK is a foreign protein, infusion will initiate immune response with production of antibodies, which limits its use, especially repetitive use. The drug has also been shown to have a higher complication rate and lower performance than UK, and its use in the peripheral vascular system has been abandoned. This drug is approved by the FDA for myocardial infarction, pulmonary embolism, DVT, arterial thrombosis or embolism, and occlusion of arteriovenous shunts.

Anticoagulants

Heparin is the best known anticoagulant even though multiple new drugs with the same indications have been introduced recently. It prevents formation of thrombus but will not cause clot lysis directly. Its mechanism of action is through antithrombin III (Figure 1). The use of anticoagulation with thrombolytic treatment is somewhat controversial. The SCVIR advisory panel agreed on safe use of heparin at an "ultra-low" dosage of 30 to 40 U/kg bolus and 7 U/kg/h infusions. Low-dose sheath infusion was recommended to prevent pericatheter thrombosis. ⁶ Some practices will avoid the bolus and some will not use heparin at all. It is typically administered intravenously through a peripheral venous access or intra-arterially through the side arm of the sheath. It should be emphasized that heparin and rt-PA or r-PA are incompatible when combined in solution. Low-molecular weight heparin has been used concomitantly with thrombolytics, but is impractical because of difficulties in reversing its effect quickly.

Anti-platelet drugs

Glycoprotein (GP) IIb/IIIa receptor inhibitors -- The role of GP IIb/IIIa receptor inhibitors is a topic of debate in peripheral thrombolysis. The three drugs available in the United States are abciximab (ReoPro, Eli Lilly & Co., Indianapolis, IN), eptifibatide (Integrilin, COR Therapeutics, San Francisco, CA) and tirofiban (Aggrastat, Merck & Co., Inc., West Point, PA). These drugs have different properties and Table 3 presents a summary comparison. Typically, these drugs are administered through peripheral venous access. The lower specificity to the GP IIb/IIIa receptor implies additional binding to the vitronectin receptor and leukocyte integrin Mac-1. This may have clinical implications in inflammatory response and intimal hyperplasia, although this has not been proven in laboratory studies or clinical data. ^8,9

Abciximab is the prototype of a low-specificity drug, but a recent study has suggested eptifibatide also binds to the vitronectin receptor under certain conditions. ¹⁰ These drugs have been studied in the coronary arteries and have shown accelerated thrombolysis and the need for smaller doses of thrombolytic drugs. ^11-16 Only one study, the TARGET trial (Do Tirofiban and Reo-Pro Give Similar Efficacy Outcomes?) has compared abciximab and tirofiban directly, ¹⁷ and the latter had lower efficacy in reducing adverse cardiac ischemic events in the first 30 days after coronary stenting. The differences might have been due to dosage or drug properties. Early experience in the peripheral arterial system has been encouraging and national multicenter trials are in progress.

Thrombolysis technique

Thrombolysis, as discussed here, is given locally with infusion of the agent directly into the thrombus. Mechanical thrombectomy, which will not be discussed here, is not yet established as an alternative to pharmacologic thrombolysis.

Typically, a multiple side-hole catheter is placed across the thrombosed segment, taking care that there are side holes proximal to the thrombus in order to secure lysis proximally. In most facilities, the patient is in the intensive care unit for early detection of possible complications during the infusion. If heparin is being used, the activated partial thromboplastin time level should be monitored frequently. If the international normalized ratio is elevated prior to initiation of therapy, it should be corrected before lysis is started. The significance of following the fibrinogen level is still controversial. No correlation has been found between the rate of bleeding complications and decrease in fibrinogen level. ¹⁸ A conservative approach would be to check the baseline fibrinogen and again at 4 hours; if it has fallen below 50% at that time, the measurement should be repeated at 12 hours. If the fibrinogen level falls below 100 mg/dL, one would consider cryoglobulin infusion, or stopping or reducing the dose of the thrombolytic agent. Usually, repeat angiogram is performed within 8 to 12 hours. At this point, treatment may be terminated if all of the thrombus has been lysed. Often there is residual thrombus present and further infusion is deemed necessary with or without manipulation of the infusion catheter. Often, a "culprit" lesion may be identified and treated with angioplasty or stent placement. If treatment is performed in the same setting, outcome is usually improved. Anticoagulation can be started 2 to 4 hours after hemostasis is secured.

Pulse-spray pharmacomechanical thrombolysis is an alternative used for dialysis graft thrombosis but generally has not been found to be practical for peripheral thrombolysis.

Arterial thrombolysis

Thrombolytic therapy for arterial limb ischemia is a generally accepted treatment for acute thromboembolic occlusions of the peripheral arteries. Landmark papers such as the Rochester, ¹⁹ Thrombolysis or Peripheral Arterial Surgery (TOPAS), ²⁰ and Surgery versus Thrombolysis for Ischemia of the Lower Extremity (STILE) trials ²¹ compared surgery and thrombolysis. The results of these trials indicated that patients initially treated with thrombolytic treatment had decreased mortality rates, increased amputation-free survival at 1 year, and lowered amputation rates in patients with <14 days of symptoms. An example of this procedure is illustrated in Figures 2 and 3.

Different drugs and regimens are being used currently, but to date, there are no solid blinded studies that prove one regimen is superior. Many reports on thrombolytic treatment were presented at the recent 2002 SCVIR meeting in Baltimore, MD. A prospective study by Castaneda et al ²² revealed an optimal r-PA dosage of 0.25 U/h. The higher dosage of 0.5 U/h was associated with higher complication rates, and the lower dosage of 0.125 U/h resulted in longer infusion times. Other studies have confirmed good efficacy and reasonable safety of r-PA using dosages ranging from as low as 0.125 U/h to 1 U/h. ^23,24 The clinical success ranged from 74% to 79%, and severe complications were seen in about 4.2%. ^23,24 Alzate et al ²⁵ reported rt-PA similarly safe and effective in both native and graft occlusions (complete lysis in 91% to 93% of cases and significant complication in about 3.5%). In study of 17 matched pairs, Tepe et al ²⁶ compared r-PA and UK and found a tendency for higher bleeding complications but better efficacy and improved long-term results in the r-PA group. Initial clinical experience with TNK at 0.25 and 0.5 mg/h were presented and revealed safety and efficacy results similar to those reported in both arterial and venous occlusions with no major bleeding complications. ⁷

Combination therapy with the GP IIb/IIIa may change the future of peripheral thrombolytic therapy; however, its definite role and regimen has not yet been defined. Encouraging early results indicate faster clot dissolution and improved amputation-free survival. ²⁷ A recent pilot study also reported shorter lysis time (with a mean of 17.5 h), 93% clinical success, and no major hemorrhagic complications. ²⁸ A study presented at the 2002 SCVIR meeting compared two combination therapies, r-PA plus abciximab versus rt-PA plus eptifibatide and reported superior clinical success, complete thrombolysis, lower complication rates, and shorter times with r-PA plus abciximab. ²⁹

Venous thrombolysis

Deep venous thrombosis is associated with significant long-term sequelae due to incomplete thrombus resolution, persistent venous obstruction, and valvular insufficiency. The resultant postthrombotic syndrome includes leg-swelling, pain, hyperpigmentation, and ulceration that can be seen in up to two-thirds of patients. Medical therapy using systemic anticoagulation is the standard treatment used to prevent pulmonary embolism and recurrent thrombosis; however, it does not address the already established thrombus directly. Catheter-based thrombolysis proposes rapid relief of acute symptoms and prevention of long-term complications. Although no randomized prospective study has established the long-term effects of thrombolysis for DVT, the report of a national multicenter registry has suggested good technical success rates and intermediate patency rates. ³⁰

Romano and colleagues ³¹ presented a poster at the 2002 SCVIR meeting reporting on 11 treatments of subclavian vein thrombosis. All treatments were successful, but angioplasty was performed in 7 of the 11 patients. They used rt-PA for all treatments and reported 1 minor complication. ³¹ Hofkin et al ³² reported on the treatment of 14 patients with iliofemoral DVT using r-PA. Of the 14 patients, 11 patients had partial or complete thrombolysis with improvement in symptoms and complications prevented completion of treatment in 2 patients. ³² Razavi et al ⁷ used 0.25 mg/h or 0.5 mg/h TNK for the treatment of iliofemoral DVT in 16 patients. Both regimens were similarly effective and bleeding complications were few. Of the 16, 8 (50%) patients had complete thrombolysis, 3 patients had <50% lysis, and 5 patients had between 50% and 90% lysis. ⁷

The role of combined pharmacologic and mechanical thrombolysis is not clear. Vedantham et al ³³ presented a poster paper at SCVIR that demonstrated shorter infusion times and lower drug dose with comparable short-term result. Successful thrombolysis with mechanical device only has been reported in central venous obstruction. ^34,35

Cost analysis

Thrombolytic treatment is rather expensive, comparable to that of surgical thrombectomy as shown by Ouriel and colleagues. ¹⁹ A large proportion of the total was attributed to the cost of UK, the thrombolytic drug. Table 4 lists the average published wholesale prices ³⁶ for the 4 agents: rt-PA, r-PA, TNK, and UK. Although UK is not available in the United States, it can still be found on the wholesale list. It is not known what the new price for UK will be when it is back on the market. Table 4 lists the hourly rates for each of the drugs, using arbitrary infusion rates and the cost of the smallest usable unit of agent. When comparing their hourly prices, the similarity of rt-PA, r-PA, and TNK is remarkable. Many institutions can get considerably lower prices for some of these agents, and these variations are not accounted for in this table.

Conclusion

Thrombolysis is a well-established procedure in an interventional radiology practice. Although it is safe and effective for the treatment of peripheral arterial and venous occlusions, some issues related to drug choices and dosage are yet to be determined. Currently, rt-PA and r-PA are the most commonly used drugs. Combination therapy with GP IIa/IIIb drugs and mechanical devices are promising adjunctive treatment options still undergoing evaluation.