sign in sign up
<<
Home , Abciximab, Enoxaparin sodium, Eptifibatide

Anticoagulation Strategies for Patients Undergoing Percutaneous Coronary Intervention: Unfractionated , Low-Molecular-Weight , and Direct Inhibitors

Spyros Kokolis, Erdal Cavusoglu, Luther T. Clark, and Jonathan D. Marmur

Low-molecular-weight heparins and direct throm- An alternative strategy to deal with the prob- bin inhibitors are emerging as alternative antico- lems associated with UFH is to substitute this agulants to unfractionated heparin in patients un- agent with such as the low-molec- dergoing percutaneous coronary intervention (PCI). ular-weight heparins (LMWHs) and direct throm- This paper reviews the pharmacologic properties of bin inhibitors (DTIs). The purpose of this review these newer agents and evaluates is to examine the mechanisms of action, phar- the clinical data demonstrating their use in patients macologic properties and clinical uses for UFH, undergoing PCI. LMWHs, and DTIs. © 2004 Elsevier Inc. All rights reserved.

Indirect Thrombin Inhibitors n recent years, an increasing number of antico- Iagulants have become available as alternatives Heparin to unfractionated heparin (UFH) for treatment of Heparin is a large polymer, averaging approxi- patients with acute coronary syndromes (ACS) mately 15,000 daltons in size, which contains neg- and for percutaneous coronary intervention atively charged sulfate and carboxylic acid groups (PCI). Although UFH has served as the primary The effects of heparin can be rapidly reversed with for nearly 50 years, it has several protamine, a basic molecule.2 UFH is a heteroge- well-established limitations1 (Table 1). From the neous preparation of heparin molecules ranging perspective of an interventional cardiologist, per- in size from about 5,000 to 40,000 daltons, and haps the most significant of these limitations is the typically extracted from porcine intestine or bo- ability of UFH to activate . In part to com- vine lung.3 UFH is administered parenterally (in- pensate for these limitations, a number of new travenously [IV] or subcutaneously [SQ]) and ex- therapies and strategies have been implemented in erts its therapeutic effect by indirectly inhibiting the treatment of ACS (e.g., glycoprotein IIb/IIIa factor IIa (thrombin), a critical medi- antagonists, , stents). ator in clot formation. Thrombin is characterized by two binding sites (exosites 1 and 2), in addition From the State University of New York, Health Science to its active catalytic site, as shown in Fig 1. These Center at Brooklyn, Brooklyn, NY. sites are essential for the binding of thrombin sub- Address reprint requests to Jonathan D. Marmur, MD, strates such as fibrinogen.4 Thrombin cleaves fi- FACC, State University of New York, Health Science brinogen to form fibrin, activates other coagula- Center at Brooklyn, 450 Clarkson Ave, Brooklyn, NY tion factors, and carries out other actions 11203; E-mail: [email protected] 0033-0620/$ - see front matter mediated by its binding sites to promote the for- © 2004 Elsevier Inc. All rights reserved. mation, expansion, and organization of blood doi:10.1016/j.pcad.2004.02.002 clots.

506 Progress in Cardiovascular Diseases, Vol. 46, No. 6, (May/June) 2004: pp 506-523 ANTICOAGULATION FOR PCI 507

Table 1. Limitations of Unfractionated Heparin

Limitation Consequence

Nonspecific binding to plasma proteins and Variability in anticoagulant effect, especially in seriously ill patients88 endothelial cells34,69 Release of factor 4 and von Willibrand Results in heparin resistance and a need for higher levels of heparin44 factor from platelets during clotting Inability of heparin to inactivate fibrin-bound Thrombin remains active when bound to fibrin and continues to thrombin6 activate platelets89 Heparin induces platelet activation60 Further activates the clotting cascade and release of heparin-binding proteins Forms heparin Can result in heparin-induced and thrombosis syndrome48 Dose-dependent half-life81 Nonlinear increase in half-life as dose increases Ill-defined dose to achieve target ACT level Supra-anticoagulation resulting in increased risk or in PCI suboptimal anticoagulation resulting in increase risk of ischemic complications90

Abbreviations: ACT, activated clotting time; PCI, percutaneous coronary intervention.

To inactivate thrombin, the heparin molecule absence of heparin, acts over the course of a few must bind to both (AT) and throm- hours to inactivate thrombin. Once heparin binds bin, forming a ternary complex. AT is an ␣-glob- to AT, a molecular change occurs in the hepari- ulin occurring naturally in the body that, in the n:AT complex that accelerates the ability of AT to

Fig 1. Thrombin binding sites and the inhibition of thrombin and factor Xa by UHF and LMWH. Abbreviations: AT, antithrombin; LMWH, low-molecular-weight heparins. (Source: J Invasive Cardiol;14:8B-18B, 2002 [Suppl B]). 508 KOKOLIS ET AL. bind to thrombin. Binding of the heparin mole- thrombin.8-12 Because factor Xa inhibition does cule to thrombin at exosite 2, with bridging be- not require bridging between factor Xa and AT, tween AT and thrombin as shown in Fig 1, is also the smaller (Ͻ18 saccharide units) pentasaccha- necessary for thrombin inhibition. Approximately ride-containing heparin chains in LMWH prepa- one third of the heparin molecules in an UFH rations can inactivate factor Xa, but not thrombin preparation contain the specific pentasaccharide (see Fig 1). LMWHs have anti-Xa to anti-IIa sequence necessary for binding to AT, and most of (thrombin) ratios that vary between 4:1 and 2:1, these have polysaccharide chains long enough (at whereas UFH has nearly equal inhibitory effects least 18 saccharide units) to bind to thrombin.1,5 on factor Xa and thrombin8 for an anti-Xa:anti-IIa Like thrombin, is also inhibited indi- ratio of 1:1. rectly by heparin through the binding of AT. How- One of the major disadvantages of indirect ever, unlike thrombin, factor X does not require thrombin inhibition by UFH and LMWH is their direct heparin binding for inactivation (i.e., ter- inability to inactivate thrombin bound to fibrin, or nary complex formation). Thus, very short hepa- to the soluble fibrin degradation products that rin molecules (Ͻ18 saccharide units in length) increase in concentration after t-PA–induced ly- that contain the pentasaccharide sequence allow- sis.13 At lower concentrations, heparin molecules ing binding to AT can inhibit factor X, but not are unable to inhibit clot- or fibrin-bound throm- thrombin. bin molecules, because the heparin binding site may be less accessible when these thrombin com- Excretion and plexes occur. This clot-bound thrombin is still enzymatically active. The inability to fully inhibit After parenteral administration, heparin mole- clot-bound thrombin may be one of the potential cules are degraded into inactive compounds by the reticuloendothelial system, and then excreted mechanisms for the 3% to 6% reinfarction rate fol- via the kidneys. Liver failure and renal insuffi- lowing initially successful coronary . ciency prolong the half-life of UFH. Heparin is a large molecule that cannot cross the placental bar- Monitoring of Anticoagulation Activity rier, and is therefore considered an acceptable an- Activated clotting time (ACT) and activated par- ticoagulant for use during . tial thromboplastin time (aPTT) are not routinely used to measure the anticoagulant effect of Low-Molecular-Weight Heparins LMWHs. Because only 25% of LMWH molecules LMWHs are derived from UFH through a chemi- are large enough to inactivate thrombin, LMWHs cal and enzymatic depolymerization process re- do not prolong the ACT or the aPTT to the same sulting in preparations of shorter heparin mole- extent as UFH. This is especially true for the cules,6 which like UFH, consist of chains of agents with shorter chain lengths and higher anti- alternating residues of D-glucosamine and uronic Xa:anti-IIa ratios.14 The uncertainty regarding acid.7 LMWHs are characterized by a mean molec- measurements of LMWH anticoagulant activity ular weight of 4000 to 5000 daltons but range in with bedside tests such as the ACT has become a size from 1,000 to 10,000 daltons. Similar to UFH, critical issue in the settings of ACS and PCI . An LMWHs show heterogeneous anticoagulant ac- accepted method to gauge the anticoagulation sta- tivity.1 tus of a patient treated with LMWHs is anti-factor The LMWHs are also indirect thrombin inhibi- Xa testing. However, the test is time consuming, tors that bind to AT via the same pentasaccharide costly, and not readily available as a point-of-care sequence found in UFH.8 The anticoagulant ef- testing. Furthermore, clinical studies have dem- fects of LMWHs result from their inhibition of onstrated weak correlations between anti-Xa lev- factor Xa and thrombin as well as other factors, els and thrombin formation.15-17 including factors XIIa, XIa, and IXa, and tissue Recently, a new test known as ENOX (enoxapa- factor pathway inhibitor (TFPI). LMWHs demon- rin test card) has become available. This test spe- strate less inhibitory activity against thrombin cifically measures the anticoagulant effects of than against factor X, because only 25% to 50% of enoxaparin (Ͼ1.0 IU/mL) using a factor Xa–acti- the LMWH chains are long enough to bridge AT to vated clotting method. However, this test has ANTICOAGULATION FOR PCI 509 never been formally validated in patients under- Table 2. Protamine Neutralization of LMWH going PCI. Furthermore, deficiencies in factors X, V, or prothrombin can result in a prolongation of % Anti-Xa % Heparin Neutralized Sulphate the ENOX clotting test, even in the absence of enoxaparin. UFH has also been found to interfere Fraxiparin 58 35 18 Tinzaparin 86 39 with the test results. In general, assays of antico- Dalteparin 74 37 agulation activity are dependent both on tech- Clivarin 51 35 nique and laboratory standardization, and are dif- Enoxaparin 54 32 ficult to compare and evaluate owing to their Super-sulphonated LMWH 100 42 variability. Source: From Crowther et al.23 In addition, the premise of using anti-factor Xa to gauge the activity of LMWH may in itself be incorrect because it focuses on an isolated step in the coagulation cascade. Anti-factor Xa measure- ies as a consequence of heparin therapy. The dis- ment would not effectively ascertain the extent of order is characterized by a reduction in the num- anticoagulation exerted by LMWHs at multiple ber of platelets approximately 6 to 10 days after levels, through factors XIIa, XIa, and IXa, and TFPI. the initiation of therapy, necessitating the imme- In contrast, assays such as the ACT and the PTT diate discontinuation of heparin. Heparin-in- give a more global assessment of the anticoagulant duced thrombocytopenia and thrombosis syn- effects and more recent data suggest that these drome (HITTS) is characterized by decreased assays may be useful when LMWH are adminis- numbers of circulating platelets and thrombosis tered IV as opposed to SQ.15 Data supporting the within the vasculature. The chronic use of UFH ACT to monitor LMWH on PCI are presented in may lead to a decrease in the production and ac- the “Clinical Data in PCI” section. tivity of AT, which thus increases the risk of thrombosis. Pharmacodynamics LMWH is presumed to have a lower incidence of HIT/HITTS. However, LMWHs do cross-react LMWHs are cleared predominantly by renal with HIT antibodies and have been reported to mechanisms.8 The plasma half-life of LMWH cause thrombocytopenia in a small percentage of ranges from 2 to 4 hours after IV injection and 3 to patients.22 6 hours following a SQ injection.8 The half-life of LMWH increases in patients with renal failure. Reversibility LMWH dose should be reduced by approximately 50% in patients with severe renal impairment Protamine is the reversal agent for UFH. The (creatinine clearance Ͻ30 mL/min), owing to the mechanism through which protamine inhibits risk of accumulation that may lead to major bleed- UFH is dependent on the degree of sulfation on ing, particularly at the sites of instrumentation.19-21 the molecule. The protamine-resistant fraction in LMWHs have a longer half-life and, compared LMWH is due to the lower-molecular-weight frac- to UFH, show less nonspecific binding to endo- tion with a low sulphate charge density. The thelial cells, platelet factor 4 released from acti- amount of anti-Xa neutralized is directly propor- vated platelets, and other plasma proteins. Fol- tional to the number of sulphate groups attached lowing a dose of LMWH, the ability to inhibit to the LMWH. LMWHs are at least partially re- factor Xa persists longer than the ability to sup- versible by protamine and therefore should be press thrombin, owing to more rapid clearance of given if a reversal state is needed rapidly in pa- the longer LMWH chain units. tients who had LMWH administered23 (Table 2).

Heparin-Induced Thrombocytopenia and Heparin-Induced Thrombocytopenia and Direct Thrombin Inhibitors Thrombosis Syndrome In contrast to UFH and LMWH, DTIs (e.g., biva- Heparin-induced thrombocytopenia (HIT) may lirudin, , ) are able to inhibit occur following the formation of platelet antibod- thrombin directly without the need for the cofac- 510 KOKOLIS ET AL.

Table 3. Binding Characteristics of Direct Thrombin Inhibitors

Drug Molecular Weight Mode of Binding Thrombin Binding Site ϳ2.2 Kda (20 amino acids) Reversible bivalent Active site, exosite 1 r-Hirudin (, desirudin) ϳ7.0 Kda (65 amino acids) Irreversible, bivalent Active site, exosite 1 Argatroban ϳ527 Daltons Reversible, univalent Adjacent to active site

tor AT, and are able to inhibit both fibrin-bound Bivalirudin and soluble thrombin. Bivalirudin is a synthetic 20 amino acid polypep- Because of their relatively small size, the inter- tide modeled after hirudin and comprised of an action of the DTI’s with the active site of thrombin active site-directed peptide linked via a tetragly- is not compromised following binding of throm- cine spacer to a dodecapeptide analog of the car- 24 bin to fibrin at exosite 1. Thus, they are able to boxy-terminal of hirudin.27 Bivalirudin binds inactivate both free thrombin and thrombin thrombin with high affinity at both the active site bound to fibrin or fibrin degradation products. and exosite 1. Once bound, thrombin slowly This may be due to a greater affinity of thrombin cleaves bivalirudin at the active site, resulting in for exosite 1 (Table 3). recovery of the function of thrombin’s active Bivalirudin, lepirudin, and hirudin are bivalent site.27 The carboxy-terminal dodecapeptide por- DTIs that bind to thrombin at exosite 1 (the sub- tion of the bivalirudin molecule remains bound to strate recognition site) and at the active site25-27 exosite 1 with low affinity. Therefore, bivalirudin (Figs 2 and 3). The univalent DTI argatroban initially acts as a noncompetitive inhibitor of binds thrombin only at its active site28 (Fig 4). thrombin, but then becomes a competitive inhib- In the direct thrombin inhibitor trialists’ collab- itor, enabling thrombin to subsequently partici- orative group, hirudin and heparin were associ- pate in hemostatic reactions (Fig 2). ated with an increase in major bleeding complica- tions. However, the bivalent inhibitor, bivalirudin Pharmacodynamics was associated with fewer major bleeding compli- Bivalirudin is cleared by a combination of proteo- cations, probably because of its shorter half-life. lytic cleavage and renal mechanisms.30 Bivaliru- Bivalirudin was also associated with a lower death din has a half-life of about 25 minutes31 in patients rate when compared to hirudin , or to univalent with normal renal function, with prolongation inhibitors29 (Fig 5). seen in patients with moderate (34 min) or severe

Fig 2. Molecular interactions between the bivalent direct thrombin inhibitor bivalirudin and thrombin. (a) Bivalirudin binds to two sites on the thrombin molecule: its active site and exosite 1. (b) Bivalirudin is cleaved by thrombin at the active site. (c) The bond between thrombin and the remaining portion of bivalirudin is now weaker, allowing thrombin to resume its hemostatic function. ANTICOAGULATION FOR PCI 511

mechanisms.34 After IV injection, the half-life is approximately 50 to 60 minutes, increasing up to 3 hours depending on comorbid diseases.35 Be- cause the drug is cleared via the kidneys, drug accumulation occurs in patients with renal insuf- ficiency. Consequently, both the bolus and infu- sion dose must be reduced in patients with renal impairment (creatinine clearance Ͻ60 mL/min), and the agent is not recommended for use in pa- tients with creatinine clearance Ͻ15 mL/min.35-37 The plasma concentrations of lepirudin in- crease proportionally to the dose administered. Fig 3. The bivalent binding of hirudin, a direct throm- The standard ACT assay method, however, is un- bin inhibitor. Hirudin binding is highly specific and almost irreversible. Like hirudin, lepirudin is also a suitable for routine monitoring of the anticoagu- bivalent direct thrombin inhibitor, and also demon- lant effects of lepirudin. In general, an aPTT ratio strates nearly irreversible binding to thrombin. (the patient’s aPTT value at any given time divided by an aPTT reference value) is the method recom- mended for monitoring the anticoagulant with (57 min) renal impairment (creatinine clearance lepirudin. The ecarin clotting time assay has also of 30 to 59 mL/min and less than 30 mL/min, been reported as successful in predicting antico- respectively). Dose adjustments of the bivalirudin agulant effect.38 infusion may be required for patients with severe renal impairment and for dialysis-dependent pa- Pharmacodynamics tients.30 Although there is no antidote to counter- act the anticoagulant effects, this does not appear Lepirudin antibodies have been detected in ap- 34,39 to be a major concern with bivalirudin because of proximately 40% of treated patients and ap- its short half-life. The short half-life of bivalirudin pear to have an effect on anticoagulant status. distinguishes it from lepirudin (recombinant Some of these antibodies bind to lepirudin and hirudin) and may contribute to a more favorable potentially prolong its half-life leading to drug safety profile. accumulation and subsequent hemorrhagic com- Bivalirudin exhibits linear dose-proportional plications, and other antibodies reportedly neu- 40 plasma concentration responses with a positive tralize lepirudin’s anticoagulant effect. correlation between dose and anticoagulant effect. Argatroban The anticoagulant effect of bivalirudin is readily measured by both the ACT or the aPTT, making Argatroban is a synthetic derivative of arginine this agent easy to use in the catheterization labo- that binds reversibly to the catalytic site of throm- ratory. Bivalirudin has not demonstrated any cross-re- activity with heparin-induced antibodies in the serum of patients with HIT.

Lepirudin The recombinant agent lepirudin (hirudin), orig- inally isolated from the salivary glands of the me- dicinal leech (Hirudo medicinalis) is a 65-amino acid polypeptide.32 Lepirudin specifically binds to thrombin with such high affinity that the lepiru- din/thrombin complex is considered irreversible, a potential disadvantage for this agent because Fig 4. The univalent binding of argatroban. Argatro- there is no antidote to reverse its anticoagulant ban reversibly binds only near the active site on activity.33 Lepirudin is cleared primarily via renal thrombin, and is characterized by a short half-life. 512 KOKOLIS ET AL.

Fig 5. Direct thrombin inhibitor in ACS meta-analysis. Source: From the Direct Thrombin Inhibitor Trialists’ Collaborative Group.29

bin.28 Argatroban does not inhibit other serine Univalent Versus Bivalent Direct Thrombin proteases, but 54% of the argatroban dose binds to Inhibitors human serum proteins, albumin, and 1-acid gly- coprotein.41 According to a recent meta-analysis of 11 random- ized trials comparing DTIs to UFH in the manage- Pharmacodynamics ment of patients with ACS,42 the authors found a The primary route of clearance for argatroban is 15% reduction in death/ by liver metabolism, with a half-life approxi- (MI) in patients treated with a DTI, compared to mately 54 minutes after IV administration. He- treatment with UFH. An 0.8% absolute risk reduc- patic impairment significantly decreases argatroban tion, maintained at 30 days, was also found. Sim- clearance (approximately 4-fold) and increases its ilar benefits were seen with the DTIs hirudin and half-life to approximately 181 minutes. bivalirudin, but not with the univalent DTI ar- The anticoagulant effect of argatroban can be gatroban. measured using the ACT or the aPTT. Plasma con- In accord with various other trialists, the inves- centrations, dose, and anticoagulant effects are tigators found univalent DTIs (argatroban and well correlated. Argatroban has not demonstrated two other univalent agents, efegatran and inogat- any cross-reactivity to heparin-induced antibodies. ran) to be less effective than the bivalent DTIs ANTICOAGULATION FOR PCI 513

Table 4. LMWH Comparative Studies in PCI

Trial/Study N Study Population Outcomes (%) LMWH vs UFH

REDUCE,54 1996 612 Elective PCI; single lesion Acute coronary events (reviparin) At 24 h 3.9 vs 8.2 Major bleeding 2.3 vs 2.6 CRUISE,52 2003 261 Elective or urgent PCI D/MI/Revas at 30 days 8.5 vs 7.6 (enoxaparin) TIMI major 2.5 vs 1.6 TIMI minor 4.1 vs 10.5 Legalery et al,57 2002 584 enox Unselected PCI D/MI/Revas 1.3 vs 2.0 (enoxaparin) 581 UFH Major bleeding 1.0 vs 0.0 (historical 2.6 vs 1.7 control)

Abbreviations: CRUISE, Coronary Revascularization Using Integrelin and Single Bolus Enoxaparin; D, death; LMWH, low-molecular-weight heparin; MI, myocardial infarction; PCI, percutaneous coronary intervention; REDUCE, Reduction of Restenosis After PTCA, Early Administration of Reviparin in a Double-Blind Unfractionated Heparin and Placebo- Controlled Evaluation; Revasc, revascularization; TIMI, Thrombolysis in Myocardial Infarction; UFH, unfractionated heparin.

(hirudin, bivalirudin) in preventing recurrent The Coronary Revascularization Using Integre- ischemic events. However, the bivalent DTI hiru- lin and Single Bolus Enoxaparin (CRUISE) study din was associated with a higher cost and in- evaluated a total of 261 patients comparing enox- creased risk of bleeding compared with UFH.42 aparin plus eptifibatide with UFH plus eptifi- Hirudin was associated with an increased risk of batide.52,53 The study was too small to reach major bleeding in patients presenting with ST- meaningful conclusions, but there was no evident segment elevation MI. In contrast to these find- benefit for enoxaparin in ischemic events (8.5 ver- ings, bivalirudin was associated with a 50% reduc- sus 7.6, p ϭ .82) or major bleeding (2.5% versus tion in major bleeding in patients with ACS. In a 1.6%, p ϭ .68) compared to UFH.52 This is in subgroup analysis, there was a benefit of DTIs on contrast to Thrombolysis in Myocardial Infarction death or MI in trials of both ACS and PCI. In (TIMI), where major bleeding occurred more fre- contrast, the univalent DTIs showed a less robust quently among patients receiving enoxaparin and decrease in major bleeding.42 TIMI minor bleeding was more frequent with UFH compared with enoxaparin (Table 4). ϭ Clinical Data Dalteparin, a LMWH, in doses of 40 IU/kg (n 27) or 60 IU/kg (n ϭ 76), was evaluated in a small, nonrandomized pilot study of patients undergo- Unfractionated Heparin and Low-Molecular- ing PCI. Early assessment determined that the 40 Weight Heparin IU/kg dose appeared inadequate to provide opti- Large clinical trials evaluating LMWH in the man- mal protection from ischemic events and conse- agement of patients with unstable and quently the majority of patients in this pilot study non–Q-wave MIs have been promising.43-48 Clin- received the 60 IU/kg dose. In this study,55 the ical data from two of these trials comparing enox- composite endpoint of death/Q-wave MI/revascu- aparin to UFH demonstrated reductions in isch- larization occurred at a rate of 3.7%, major bleed- emic events with LMWH versus UFH. However, ing at 2.8%, minor bleeding at 10.3%, and trans- there appeared to be moderate increases in major fusion at 2.8%. and minor bleeding with LMWH compared to Another study with dalteparin evaluated 110 UFH.43,48 patients who were undergoing PCI received doses Evidence for the clinical value of LMWH in PCI of dalteparin 60 or 80 IU/kg alone or followed by is much less clear than that supporting use in ACS. infusion. This study evaluated whether There are a number of publications reporting the the ACT could be used to monitor IV dalteparin use of LMWH in PCI,49-54 but none are large-scale, during PCI. In this trial, dalteparin induced a sig- randomized controlled trials that provide convinc- nificant rise in the ACT with a smaller degree of ing evidence of a meaningful benefit over UFH. variance as compared to UFH. This study had no 514 KOKOLIS ET AL.

Table 5. LMWH Registry Studies in PCI

EPISTENTa Control NICE-1b EPISTENT Treated NICE-4b (UFH) (Enox 1 mg/kg) (UFH ϩ abciximab) (Enox 0.75 mg/kg ϩ abciximab) (N ϭ 809) (N ϭ 828) (N ϭ 794) (N ϭ 813) D/MI/Revasc at 30 days 10.8 7.7 5.3 6.8 TIMI major, all 2.2 1.1 1.5 0.4 TIMI major non-CABG 1.0 0.5 0.8 0.2 TIMI minor all 1.7 6.2 2.9 7.0 Transfusion all 2.2 2.7 2.8 1.8

aThe EPISTENT Investigators, 1998.59 bKereiakes, 2001.51 Abbreviations: CABG, coronary artery bypass graft; D, death; MI, myocardial infarction; Revasc, revascularization; TIMI, Thrombolysis in Myocardial Infarction; MI, myocardial infarction. death or urgent revascularization during the hos- A series of registry studies, National Investiga- pital course of the study population. Two of the 96 tors Collaboration on Enoxaparin (NICE),50,51,58 PCI patients had MI (2.1%), major bleeding with provide the majority of patient exposure to enox- transfusion occurred at a rate of 1.0%, and minor aparin in ACS patients undergoing PCI. The NICE bleeding at a rate of 4.2%. This was the first study registries are open label and noncomparative in to demonstrate that the ACT may be useful in nature and continue to suggest relative equiva- monitoring the anticoagulant effect of IV admin- lence to UFH in ACS patients undergoing elective istered dalteparin during PCI. Similar degrees of or urgent PCI. elevation of the ACT have been seen with enox- The primary endpoint of both NICE-1 (1.0 aparin in 26 PCI patients with the ACT at 150 mg/kg IV bolus of enoxaparin) and NICE-4 (0.75 seconds at baseline to 208 seconds post-10 min- mg/kg enoxaparin plus abciximab) was major utes when enoxaparin was given at a dose of 0.5 bleeding (in-hospital and at 30 days) defined as a mg/kg IV bolus.56 drop in hemoglobin Ն5 g/dL, a drop in hematocrit One randomized study compared reviparin, an- of Ն15%, or intracranial bleeding.51,58 Secondary other LMWH, to UFH in 612 patients.54 Reviparin endpoints included death/MI/revascularization, (n ϭ 306) was administered as a 7000-U bolus need for transfusion, and minor bleeding, which followed by a 10,500-U infusion for 24 hours and was defined as spontaneous and observed hema- a twice-daily 3500-U SQ injection. UFH (n ϭ 306) turia or hematemesis, Ն3 g/dL in hemoglobin or was given as a 10,000-U bolus followed by a hematocrit drop of Ն10%. The investigators 24,000-U infusion for 24 hours. No GP IIb/IIIa found an absolute reduction in the composite inhibitors were used, and stent implantation was endpoint of death/MI/revascularization and in performed only for bailout purposes. The rate of major bleeding compared to the historical UFH rescue stent implantation on the day of PCI was control arm of the EPISTENT trial (a trial of sim- reduced with UFH compared to reviparin, from ilar design). In addition, a corresponding in- 6.9% to 2.0%. However, reviparin did not reduce creased absolute risk in minor bleeding and in the composite of ischemic events at 6 months transfusion rates was also reported for NICE-1 when compared to heparin (1.3% versus 2.0%). (Table 5). Similar observations were made when Therefore, the agent was not promoted further for results of NICE-4 were compared to the abcix- this indication. imab-treated group of EPISTENT (Table 5). Another comparative study employed a histor- The NICE-3 registry assessed the safety of ical control as a basis for assessing the relative enoxaparin in conjunction with one of the three safety and efficacy of enoxaparin in PCI.57 This available GP IIb/IIIa antagonists in ACS pa- study compared 584 unselected ACS patient un- tients.50 This study also addressed the feasibility dergoing PCI with enoxaparin versus a historical of bringing ACS patients treated with enoxaparin UFH control group of 581 patients. There were no to the catheterization laboratory without further significant differences between the two groups anticoagulation using UFH. The NICE-3 study in- (Table 4). cluded 661 ACS patients with documented coro- ANTICOAGULATION FOR PCI 515

(Table 6. Combined Event Rates (UFH ؉ LMWH

ESPRITa ϩ EPISTENTb NICE-3c Combined event rates Enox 1 mg/kg ϩ GPI Control Treated All ACS PCI only (N ϭ 1833) (N ϭ 1834) (616) (292) D/MI/Revasc at 30 days 10.6 6.2 5.7 NR TIMI major, all 1.3 1.25 4.5 1.7 TIMI major non-CABG NR NR 1.9 1.0 TIMI minor all 1.7 2.8 25.0 NR Transfusion all 1.6 1.9 10.5 NR

aThe EPISTENT Investigators, 1998.60 bESPRIT Investigators, 2000.59 cESPRIT Investigators, 2000.50 Abbreviations: ACS, ; CABG, coronary artery bypass graft; D, death; MI, myocardial infarction; NR, not reported; PCI, percutaneous coronary intervention; Revasc, revascularization. nary artery disease who received open-label enox- revascularization. Event rates for transfusion, mi- aparin (1mg/kg SQ twice daily) in addition to a GP nor bleeding and death/MI/revascularization have IIb/IIIa inhibitor. The choice of GP IIb/IIIa inhib- not been reported for the PCI patients. itor was left to the discretion of the participating By combining the event rates from EPISTENT59 institution. For patients requiring subsequent (UFH ϩ abciximab) and ESPRIT60 (UFH ϩ epti- catheterization and intervention, no additional fibatide) an approximate historical control arm anticoagulation was given if enoxaparin had been can be used as a basis of comparison for the results given within the previous 8 hours. If the last dose of NICE-3. This comparison again suggests com- of enoxaparin had been administered more than 8 parability between LMWH and UFH in terms of hours prior to PCI, an additional IV bolus dose of ischemic and major hemorrhagic complications enoxaparin (0.3 mg/kg) was given. For the entire with potential increases in the risk of minor bleed- ACS population (n ϭ 616) the primary endpoint, ing and transfusions (Table 6). non–coronary artery bypass graft (CABG) major Other registries of enoxaparin in PCI, employ- bleeding, occurred in 1.9% of patients and in 1.0% ing a variety of dosing regimens, have reported of patients undergoing PCI (n ϭ 292). The inci- similar outcomes49,56,61 (Table 7). dence of all major bleeding was 4.5% for all ACS A retrospective chart review of 103 patients patients and 1.7% for those undergoing PCI. who received 2 or more doses of enoxaparin for a Transfusions and minor bleeding occurred at a variety of indications suggests a significantly rate of 10.5% and 25.0%, respectively, in the en- higher incidence of bleeding in patients with renal tire cohort with a 5.7% incidence of death/MI/ insufficiency (serum creatinine Ն2.0 mg/dL) (n ϭ

Table 7. LMWH Studies in PCI

Trial/Study N Study Population Outcomes (%)

Kereiakes et al.56 2001 (dalteparin) 40 IU/kg and 60 IU/kg 103 Elective PCI D/MI/Revasc 3.7 Major bleed 2.8 Minor bleed 10.3 Transfusion 2.8 Choussat,61 2002 (enoxaparin) 0.5 mg/kg IV 180 Elective PCI D/MI/Revasc 3.9 Major bleed 0.6 Minor bleed 2.2 Collet et al49 2001 (enoxaparin) 1 mg/kg q 12 h 132/451 Unselected population D/MI 3.0 of ACS patients Major bleed 0.8 undergoing PCI Minor bleed 2.4

Abbreviations: ACS, acute coronary syndrome; D, death; MI, myocardial infarction; PCI, percutaneous coronary intervention; Revasc, revascularization. 516 KOKOLIS ET AL.

Table 8. Bivalirudin Comparative PCI Trials

Trial N Population Outcomes (%) Biv vs UFH

BAT,64 2001 4312 PCI D/MI/R at 7 days 6.2 vs 7.9 Major bleeding 3.5 vs 9.3 CACHET,70 2002 288 PCI D/MI/R at 7 days 2.8 vs 7.8 Major bleeding 1.4 vs 6.3 REPLACE-1,69 2002 1056 PCI D/MI/R at 48 h 5.6 vs 6.9 Major bleeding 2.2 vs 2.7 REPLACE-2,71 2003 6029 PCI D/MI/R/in-hospital major bleed at 30 days 9.2 vs 10.0 D/MI/R at 30 days 7.6 vs 7.1 Major bleeding 2.4 vs 4.1

Abbreviations: D, death; MI, myocardial infarction; PCI, percutaneous coronary intervention; R, revascularization.

53) versus patients with normal renal function (n Ն3 g/dL, the need for transfusion of Ն2U,and ϭ 50) (51% versus 22%, P Ͻ .01).62 Collet et al63 either retroperitoneal or intracranial bleeding. report that patients with severe renal impairment In addition, reduction of both ischemic and undergoing angioplasty should have 64% of the hemorrhagic events was observed in a random- standard dose of enoxaparin administered. ized subgroup of post-MI patients (n ϭ 741). Thus, to date, the results of available LMWH Those treated with bivalirudin had a 51% relative studies have yet to provide convincing clinical risk reduction in death, MI, or revascularization evidence of the therapeutic benefit of LMWH over versus heparin (4.9% versus 9.9%, P ϭ .009) and a and above the use of UHF as a base anticoagulant. 73% relative risk reduction for major bleeding64 (2.4% versus 11.8%, P Ͻ .001). Direct Thrombin Inhibitors The BAT trial was conducted before the wide- Bivalirudin, lepirudin, and argatroban are the only spread use of stents and GP IIb/IIIa inhibitors. The DTIs approved in the United States. Lepirudin and results provided information and suggestions of argatroban are approved for treatment and man- improved outcomes with the use of bivalirudin agement of patients with HIT. Argatroban is also over heparin. To assess bivalirudin in the modern approved for use in patients with HIT or at risk for interventional setting, several pilot trials have HIT undergoing PCI. Bivalirudin is indicated for been conducted. Three randomized studies eval- use in a broader population of patients with un- uating the safety of bivalirudin and GP IIb/IIIa stable angina undergoing PCI. inhibitors versus UFH and GP IIb/IIIa inhibitors have been completed. Although the number of Bivalirudin patients in these studies is small, bivalirudin plus ϭ ϭ Of all the available DTIs, bivalirudin appears to abciximab (n 60), eptifibatide (n 42), and ϭ offer an alternative to UFH in PCI. In the Bivaliru- tirofiban (n 33) all provide positive safety data din Angioplasty Trial (BAT), 4312 patients with consistent with the results seen in BAT for com- 65-67 or post-MI, requiring PCI were bining bivalirudin with GP IIb/IIIa inhibitors. randomized in a double-blind fashion to receive The CACHET B/C and REPLACE-1 trials68,69 bivalirudin or heparin during the procedure.64 provide additional preliminary results with biva- This trial demonstrated a 22% relative risk reduc- lirudin in the setting of stents and GPIIb/IIIa inhi- tion with bivalirudin in the rate of the composite bition. CACHET B/C (n ϭ 208) was an open-label endpoint of death/MI/repeat revascularization at 7 trial of patients undergoing elective PCI, random- days (6.2% versus 7.9%, P ϭ .039) when com- ized to receive either heparin plus abciximab or pared to heparin. In addition, there was signifi- bivalirudin with provisional abciximab.68,70 The cantly less major bleeding (3.5% versus 9.3%, P Ͻ bivalirudin arm (n ϭ 144) had only 24% abcix- .001), with a 62% relative risk reduction in this imab use. The CACHET B/C demonstrated a 64% endpoint (Table 8). Major bleeding was defined as relative risk reduction of death/MI/revasculariza- overt bleeding resulting in a hemoglobin drop of tion for bivalirudin compared with UFH plus ab- ANTICOAGULATION FOR PCI 517 ciximab (n ϭ 64; 2.8% versus 7.8%). Similarly, provisional basis only may be an appropriate an- bivalirudin treatment resulted in a 74% relative ticoagulation strategy in a large portion of PCI risk reduction of major bleeding (1.4% versus patients. 6.3%). This small trial provided preliminary evi- Bivalirudin has also been evaluated in 50 pa- dence of a unique reduction in both ischemic and tients with HIT undergoing PCI, in the (Anticoag- hemorrhagic events associated with the use of ulant Therapy with Bivalirudin to Assist in percu- bivalirudin. taneous coronary intervention in patients with REPLACE-1, a slightly larger trial69 of 1056 pa- heparin-induced Thrombocytopenia [ATBAT]) tients, provided continued evidence of reduced trial. Bivalirudin was well tolerated in these pa- ischemic and hemorrhagic complications in pa- tients. A report on the interim data from the first tients undergoing elective or urgent PCI. Stents 11 patients in this study, in addition to data from were used in 85% of patients, and GPIIb/IIIa inhi- 39 additional patients in previous studies, suggest bition was administered to 72% of patients at the that bivalirudin may provide a superior alterna- discretion of the operator. As found in prior inves- tive to currently available agents.72,73 tigations, this large pilot study demonstrated a Bivalirudin appears to be well-tolerated in high- simultaneous reduction in both ischemic and risk patient populations. Women, patients over bleeding complications. There was a relative risk 65, and patients with serum creatinine Ͼ1.2 reduction of 19% for the composite 48-hour end- mg/dL have been observed to experience fewer point death/MI/revascularization and a 22% re- adverse clinical events, in comparison with UFH duction in major bleeding events observed69 (Ta- treatment.74,75 Although a reduced dose of biva- ble 8). lirudin may be considered for patients with severe The REPLACE-2 trial71 was a randomized, dou- renal insufficiency or for patients on renal dialy- ble-blind trial of bivalirudin conducted in 6010 sis, it is noteworthy that with dose adjustment in patients. The use of bivalirudin plus provisional patients with renal impairment in the BAT trial administration of GP IIb/IIIa inhibitors was com- patients with any degree of renal impairment had pared to UFH with planned GP IIb/IIIa inhibition fewer bleeding complications than similar pa- in patients undergoing PCI. Bivalirudin with pro- tients treated with UFH. visional GP IIb/IIIa inhibitor use demonstrated a numerically reduced incidence of the composite Lepirudin (Hirudin) endpoint (death, MI, revascularization, major Although thrombotic complications are lower bleeding) compared with heparin and GP IIb/IIIa with lepirudin compared to heparin when used to inhibitors. With respect to more traditional end- treat ACS patients, the risk of major bleeding ap- points, the incidences of death and revasculariza- pears greater.76 To date, lepirudin is not indicated tion were lower for bivalirudin-treated patients. for use in PCI patients. A recent analysis suggests However, the incidence of non–Q-wave MIs in 1 reason for the increased risk of bleeding may be this group was higher. No statistically significant related to the development of thrombocytopenia differences were found. The incidences of bleed- with lepirudin (0.9%), which was similar to the ing, transfusions, and thrombocytopenia, how- incidence reported for heparin (1.1%).77 ever, were significantly lower in patients receiving A single large randomized trial of 1141 unstable bivalirudin compared with heparin and provi- angina patients treated with desirudin or UFH was sional GP IIb/IIIa arm (Table 8). The 6-month conducted. Like lepirudin, desirudin is a recom- follow-up data from the REPLACE-2 trial demon- binant DTI based on hirudin. After undergoing strated that patients randomized to heparin with angioplasty, patients in the study demonstrated GPIIb/IIIa inhibitor and bivalirudin experienced significant reduction in death/MI/revascularization very similar rates of MI (1.5%) and revasculariza- at 96 hours. However, there were no significant tion (9.0%). Although it did not reach statistical differences at 7 months for event-free survival.78 significance, the death rate at 6 months and 1 year was numerically lower in the bivalirudin arm compared to UFH and glycoprotein IIb/IIIa in- Argatroban hibitors. This suggests that bivalirudin plus a Argatroban has been evaluated in small clinical glycoprotein IIb/IIIa inhibitor administered on a trials for a number of uses. Two studies have eval- 518 KOKOLIS ET AL. uated the use of argatroban for PCI in HIT pa- ences may be because of statistical power, but no tients. Fifty patients with a current or historical comparative studies between DTIs exist to estab- HIT diagnosis were evaluated by Matthai et al.79 lish clear superiority for 1 agent over the other. Patients who required CABG, who were receiving There are no data suggesting limitations of ar- GP IIb/IIIa inhibitors, or had hepatic dysfunction gatroban use in patients with renal impairment. were excluded. Argatroban was administered as a However, the use of argatroban use in patients bolus of 350 ␮g/kg and followed by an IV infusion with hepatic dysfunction may increase the risk of of 25 to 30 ␮g/kg/min to maintain an ACT of 300 bleeding complications. Even moderate liver dys- to 450 seconds. The result was a 98% success rate, function can prolong the half-life of argatroban by defined as less than 50% stenosis of the vessel more than double compared to its half-life in pa- postprocedure and the absence of bypass surgery, tients with normal hepatic function.82 acute MI, or death. Significant complications in patients given argatroban included 1 retroperito- Strategies for Treatment neal hematoma and 1 abrupt vessel closure that required bypass surgery. Because this was a HIT Heparin has been used as the standard anticoagu- population, inclusion of a control group was con- lant in PCI since the 1970s. Its limitations include sidered unethical. unpredictable anticoagulant effects, inability to Lewis et al80 evaluated 91 HIT patients under- inhibit fibrin-bound thrombin, activation of plate- going PCI in 3 similarly designed studies. They lets, and the lack of a defined optimal dose. This administered a bolus of 350 ␮g/kg was followed has led to the addition of , thienopyridines, by an IV infusion of 25 ␮g/kg/min, with a target GPIIb/IIIa inhibitors, and closure devices with ACT of 300 to 450 seconds. Among patients un- PCI procedures, to reduce ischemic events while dergoing initial PCI (n ϭ 91), 94.5% had a suc- attempting to maintain acceptable rate of hemor- cessful procedure defined as lack of death, emer- rhagic complications. gency coronary bypass or Q-wave MI. Adequate In evaluating optimal strategies for coagulation anticoagulation was demonstrated in 97.8% of the in PCI, it is important to assess ischemic outcomes study population. Death/MI/revascularization oc- and bleeding complications. Assessing bleeding curred within 24 hours in 7.7% of patients. The complications between trials is difficult owing to rate of major bleeding, which was defined as overt various definitions for major and minor bleeding, bleeding resulting in a Ն5 g/dL drop in hemoglo- such as the definitions used by the GUSTO and bin, a transfusion of Ն2 U, intracranial or retro- the TIMI trials.83,84 Many PCI investigations, in- peritoneal bleeding, or bleeding into a major pros- cluding the majority of the LMWH studies, utilize thetic joint, was 1.1%. Minor bleeding was defined the TIMI definitions for the bleeding endpoint.84 as overt bleeding that failed the criteria for major TIMI criteria were initially established for the bleeding, and occurred in 32% of initial patients evaluation of bleeding complications in thrombo- that were treated. It is difficult to assess the overall lytic trials where relatively high rates of such com- effectiveness of this agent given the short 24-hour plications were expected. However, the use of the time point for the primary endpoint, and there TIMI criteria for bleeding in the setting of PCI may was no direct comparison group. not be most appropriate measure for a procedure Because patients with HIT cannot receive hep- where the extent of bleeding complications is not arin, argatroban offers an alternative antico- as high. Furthermore, it has been well established agulant in this special population.81 There are no that bleeding resulting in Ն3 g/dL drop in hemo- published data establishing the com- globin, as well as the need for transfusions, have parative efficacy of argatroban in PCI for patients significant adverse implications for the patient with a history of HIT. morbidity and mortality, as well as for hospital In a meta-analysis of 35,970 patients in compar- costs. Studies have shown an increased length of ative trials of DTIs versus heparin in patients with stay with TIMI minor bleeding and increased risk ACS or undergoing PCI, reductions in death or MI of morbidity and mortality with transfusions.85 with hirudin or bivalirudin were observed. How- There are also some data to suggest that bleeding ever, similar findings were not observed with uni- complications can actually result in ischemic valent agents such as argatroban.42 These differ- complications.86 Consequently, minor bleeding ANTICOAGULATION FOR PCI 519 events should not be dismissed as an insignificant to effectively inhibit fibrin-bound thrombin in the consequence when evaluating outcomes in clini- high-risk ACS population. cal trials. Another small pilot study investigated the pos- LMWHs represent a theoretical advance in an- sibility that bivalirudin may provide an alternative tithrombotic therapy because of their greater to UFH in PCI for ACS patients initially treated pharmacokinetic predictability and reduced pro- with SQ LMWH. In this open-label trial of 40 pensity to stimulate platelet aggregation. This al- patients treated with LMWH for management of lows for a simplified administration scheme (SQ ACS, patients were randomized to either UFH or injections twice daily rather than continuous IV bivalirudin prior to angioplasty. The last dose of infusions) for management of ACS patients. LMWH was administered at least 8 hours prior to Whether these theoretical advantages translate randomization. Coagulation parameters and ad- into clinical benefits in PCI remains to be demon- verse events were assessed. Adverse events were strated. Although there are convincing data that found to be comparable between the two groups. support the use of LMWH in ACS, such data are One patient in the UFH group had a postproce- still lacking in PCI. To date, the most appropriate dure MI; no deaths or major hemorrhages oc- doses of LMWHs and how to combine them opti- curred in either group. Coagulation parameters mally with GPIIB/IIIa inhibitors in the cardiac were more predictable in bivalirudin-treated pa- catheterization laboratory have not been ade- tients compared with the heparin-treated patients, quately established in a large-scale, randomized with 2 to 3 times less variability, and returned to a clinical trial. normal level rapidly in patients receiving bivaliru- LMWH use in open-label pilot and registry PCI din (L. Wallentin, MD, unpublished data, data on file with The Medicines Co., June 2003). These studies, and in ACS patients, have been inter- data suggest that switching from LMWH to biva- preted as indicating that they provide a benefit lirudin for PCI procedures is feasible and may be over UFH. However, critical review of this data advantageous. However, additional investigations suggests that the benefit is limited to being no are needed to confirm these preliminary results. worse than UFH for ischemic and major bleeding complications. In some investigations, LMWH use has been demonstrated to markedly increase Summary the risk of minor bleeding and transfusion.52,53 The evidence supporting the superiority of Among the available DTIs, bivalirudin repre- LMWH over UFH in the non-PCI setting is over- sents the agent with the most robust data in sup- whelming. However, the data favoring the use of port of superior benefits over UFH in PCI. When LMWH over UFH during PCI are more limited. the BAT data or more contemporary trials and Adequate, well-controlled, randomized trials in- registry data are evaluated, bivalirudin remains volving numbers of patients large enough to es- consistent in providing reductions in ischemic tablish the comparative benefit and optimal use of and hemorrhagic events across all patient popula- LMWHs in PCI are lacking. In theory, benefits and tions studied. advantages of LMWH in PCI should be similar to Although the management of anticoagulation that observed in non-PCI. The use of LMWHs via in patients treated with SQ LMWH remains un- IV administration allows for the achievement of a clear, data with bivalirudin suggests that patients rapid and controlled anticoagulation during PCI. receiving either UFH or LMWH prior to PCI may Contrary to traditional belief, recent studies sug- be safely switched to bivalirudin for the PCI pro- gest that the level of anticoagulation induced by cedure. In 1006 patients pretreated with heparin LMWHs can in fact be measured using the stan- and switched to bivalirudin, Bittl87 found that pa- dard ACT test.15 In addition, LMWH does not tients treated with bivalirudin (n ϭ 512) com- appear to activate platelets, thereby potentially re- pared to heparin (n ϭ 494) had fewer ischemic ducing the need for administration of concomi- events (6.3% versus 9.5%; P ϭ .043) and fewer tant glycoprotein IIb/IIIa inhibitors. Such a reduc- bleeding complications (4.1% versus 11.9%, p Ͻ tion could lead to reduced bleeding and costs .001). The significant difference in adverse clini- associated with PCI. LMWHs, unlike DTIs, are at cal outcomes may be because of heparin’s inability least partially reversible with protamine. As a re- 520 KOKOLIS ET AL.

Table 9. Bivalirudin PCI Registry Studies

Trial N Population Outcomes (%) Biv vs UFH

Adamian,91 2002 28 Renal impaired CIN 0 Baseline Cr 1.9 ϩ 0.7 mg/dL Major vascular complications 0 undergoing PCI Bleeding comps 0 Adamian,92 2002 146 Diabetes undergoing PCI Major vascular comps 0 vs 1.4 Transfusion 1.4 vs 0 Hematoma 0 vs 1.4 Cho,93 2003 162 biv Elective or urgent PCI D/MI/R at 120 days 4.2 vs 12.7 172 historical heparin Bleeding 4.3 vs 6.4

Abbreviations: biv, bivalirudin; Cr, creatinine; D, death; MI, myocardial infarction; PCI, percutaneous coronary interven- tion; R, revascularization.

sult, LMWH agents appear to represent an antico- cal Haematology, vol 3. London, Bailliere Tindall, pp agulation option during PCI. This needs further 505-529, 1990 7. Rosenberg RD, Bauer KA: The heparin–Antithrombin investigation to ascertain their relative benefitin system: A natural anticoagulant mechanism, in Col- comparison to UFH, as well as to novel DTIs, such man RW, Hirsch J, Marder VJ, et al (eds): Hemostasis as bivalirudin. and Thrombosis: Basic Principles and Clinical Prac- In contrast to LMWH, the advantages of the tice. 3rd ed. Philadelphia, PA, Lippincott, pp 837-860, DTI, bivalirudin, have been well established. In 1994 particular, the REPLACE-2 trial demonstrated the 8. Weitz JI: Low-molecular-weight heparins. N Engl J Med 337:688-698, 1997 unequivocal superiority of the DTI, bivalirudin, 9. Anderson W, Harthill JE: Molecular weight depen- compared to UFH in PCI. Thus, at present, a dence of heparin anti-factor Xa activity: Influence of stronger case can be made in favor of bivalirudin method. Thromb Res 21:557-564, 1981 than LMWH to replace UFH. Whether or not DTI 10. Danielsson A, Raub E, Lindahl U, et al: Role of ternary is superior to LMWH in patients undergoing PCI complexes, in which heparin binds both antithrombin and proteinase, in the acceleration of the reactions has not been established. A head-to-head random- between antithrombin and thrombin or factor Xa. ized double blind clinical trial of bivalirudin ver- J Biol Chem 261:15467-15473, 1986 sus a high-dose IV bolus of LMWH in PCI (ideally 11. Rosenberg RD, Jordan RE, Favreau LV, et al: Highly with ACT guidance) would be of interest to the active heparin species with multiple binding sites for cardiovascular and interventional communities. antithrombin. Biochem Biophys Res Commun 86: 1319-1324, 1979 Table 9 12. Jordan RE, Favreau LV, Braswell EH, et al: Heparin with two binding sites for antithrombin or platelet References factor 4. J Biol Chem 257:400-406, 1982 1. Hirsh J, Warkentin TE, Raschke R, et al: Heparin and 13. Weitz JI, Leslie B, Hudoba M: Thrombin binds to low molecular weight heparin. Mechanisms of action, soluble fibrin degradation products where it is pro- , dosing considerations, monitor- tected from inhibition by heparin-antithrombin but ing, efficacy and safety. Chest 114:489S-510S, 1998 susceptible to inactivation by antithrombin-indepen- (5 Suppl) dent inhibitors. Circulation 97:544-552, 1998 2. Warkentin TE, Crowther MA: Reversing anticoagu- 14. Duplaga BA, Rivers CW, Nutescu E: Dosing and lants both old and new. Can J Anaesth 49:S11-S25, monitoring of low-molecular-weight heparins in spe- 2002 cial populations. Pharmacotherapy 21:218-234, 2001 3. Linhardt RJ, Gunay NS: Production and chemical 15. Marmur JD, Anand SX, Bagga RS, et al: The activated processing of low molecular weight heparins. Semin clotting time can be used to monitor the low molec- Thromb Hemost 25:5-16, 1999 (Suppl 3) ular weight heparin dalteparin after intravenous ad- 4. Tulinsky A: Molecular interactions of thrombin. Semin ministration. J Am Coll Cardiol 41:394-402, 2003 Thromb Hemost 22:117-124, 1996 16. Leizorovicz A, Bara L, Samama MM, et al: Factor Xa 5. Rosenberg RD, Oosta GM, Jordan RE, et al: The inhibition: Correlation between the plasma levels of interaction of heparin with thrombin and antithrombin. anti-Xa activity and occurrence of thrombosis and Biochem Biophys Res Commun 96:1200-1208, 1980 haemorrhage. Haemostasis 23:89-98, 1993 (Suppl 1) 6. Ofosu FA, Barrowcliffe TW: Mechanisms of action of 17. Bara L, Planes A, Samama MM: Occurrence of low molecular weight heparins and , in thrombosis and haemorrhage, relationship with anti- Hirsh J (ed): Antithrombotic Therapy, Baillieres Clini- Xa, anti-IIa activities, and D-dimer plasma levels in ANTICOAGULATION FOR PCI 521

patients receiving a low molecular weight heparin, 34. Greinacher A, Lubenow N: Recombinant hirudin in enoxaparin or tinzaparin, to prevent deep vein throm- clinical practice: Focus on lepirudin. Circulation 103: bosis after hip surgery. Br J Haematol 104:230-240, 1479-1484, 2001 1999 35. Eikelboom JW, Anand SS, Mehta SR, et al: Prognos- 18. ENOX Test Card [package insert]. Morrisville, NC, tic significance of thrombocytopenia during hirudin Pharmanetics, Inc., 2002 and heparin therapy in acute coronary syndrome 19. Caranobe C, Barret A, Gabaig AM, et al: Disappear- without ST elevation: Organization to Assess Strate- ance of circulating anti-Xa activity after intravenous gies for Ischemic Syndromes (OASIS-2) study. Circu- injection of standard heparin and of a low molecular lation 103:643-650, 2001 weight heparin (CY 216) in normal and nephrecto- 36. Greinacher A: Recombinant hirudin for treatment of mized rabbits. Thromb Res 40:129-133, 1985 heparin-induced thrombocytopenia, in Warkentin TE, 20. Palm M, Mattsson C: Pharmacokinetics of heparin Greinacher A (eds): Heparin-Induced Thrombocyto- and low molecular weight heparin fragment (Fragmin) penia. New York, Marcel Dekker, pp 313-338, 2000 in rabbits with impaired renal or metabolic clearance. 37. Stringer K, Lindenfield J: Hirudins: Antithrombin an- Thromb Haemost 58:932-935, 1987 ticoagulants. Ann Pharmacother 26:1535-1540, 1992 21. Collet JP, Montalescot G, Choussat R, et al: Enox- 38. Potzsch B, Hund S, Madlener K, et al: Monitoring of aparin in unstable angina patients with renal failure. recombinant hirudin: Assessment of a plasma-based Int J Cardiol 80:81-82, 2001 ecarin clotting time assay. Thromb Res 86:373-383, 22. Lovenox ( injection) [package in- 1997 sert]. Bridgewater, NJ, Aventis Pharmaceuticals, Inc., 39. Schiele F, Vuillemenot A, Kramarz P, et al: Use of 2003 recombinant hirudin as antithrombotic treatment in 23. Crowther MA, Berry LR, Monagle PT, et al: Mecha- patients with heparin-induced thrombocytopenia. nisms responsible for the failure of protamine to in- Am J Hematol 50:20-25, 1995 activate low-molecular-weight heparin. Br J Haema- 40. Song X, Huhle G, Wang L, et al: Generation of Anti- tol 116:178-186, 2002 hirudin antibodies in heparin-induced thrombocyto- 24. Bates SM, Weitz JI: Direct thrombin inhibitors for penic patients treated with r-hirudin. Circulation 100: treatment of arterial thrombosis: Potential differences 1528-1532, 1999 between bivalirudin and hirudin. Am J Cardiol 82: 41. Agatroban injection [package insert]. Research Trian- 12P-18P, 1998 gle Park, NC: GlaxoSmithKline, 2003 25. Fenton JW, Villanueva GB, Ofosu FA, et al: Thrombin 42. Direct Thrombin Inhibitor Trialists’ Collaborative inhibition by hirudin: How hirudin inhibits thrombin. Group: Direct thrombin inhibitors in acute coronary Haemostasis 21:27-31, 1991 (Suppl 1) syndromes: Principal results of a meta-analysis 26. Stone SR, Hofsteenge J: Kinetics of the inhibition of based on individual patients’ data. Lancet 359:294- thrombin by hirudin. Biochemistry 25:4622-4628, 302, 2002 1986 43. Antman EM, McCabe CH, Gurfinkel EP, et al: Enox- 27. Maraganore JM, Bourdon P, Jablonski J, et al: De- aparin prevents death and cardiac ischemic events in sign and characterization of hirulogs: A novel class of bivalent peptide inhibitors of thrombin. Biochemistry unstable angina/non-Q-wave myocardial infarction. 29:7095-7101, 1990 Results of the thrombolysis in myocardial infarction 28. Hursting MJ, Alford KL, Becker JC, et al: Novastan (TIMI) 11B trial. Circulation 100:1593-1601, 1999 (brand of argatroban): A small-molecule direct throm- 44. Cohen M, Demers C, Gurfinkel EP, et al: Low-molec- bin inhibitor. Semin Thromb Hemost 23:503-516, ular-weight heparins in non-ST-segment elevation 1997 ischemia: The ESSENCE trial. Efficacy and safety of 29. Direct Thrombin Inhibitor Trialists’ Collaborative subcutaneous enoxaparin versus intravenous unfrac- Group: Direct thrombin inhibitors in acute coronary tionated heparin, in non-Q-wave coronary events. syndromes: Principal results of a meta-analysis Am J Cardiol 82:9L-24L, 1998 based on individual patient’s data. Lancet 359:294- 45. Wallentin L, Bergstrand L, Dellborg M, et al: Low 302, 2002 molecular weight heparin (dalteparin) compared to 30. Robson R, White H, Aylward P, et al: Bivalirudin unfractionated heparin as an adjunct to rt-PA (alte- pharmacokinetics and pharmacodynamics: Effect of plase) for improvement of coronary artery patency in renal function, dose, and gender. Clin Pharmacol acute myocardial infarction—The ASSENT Plus Ther 71:433-439, 2002 study. Eur Heart J 24:897-908, 2003 31. Maraganore JM, Adelman BA, Hirulog: A direct 46. Wallentin L, Husted S, Kontny F, et al: Long-term thrombin inhibitor for management of acute coronary low-molecular-weight heparin (Fragmin) and/or early syndromes. Coron Artery Dis 7:438-448, 1996 revascularization during instability in coronary artery 32. Maraganore JM: Hirudin and Hirulog: Advances in disease (the FRISC II Study). Am J Cardiol 80:61E- antithrombotic therapy. Perspectives in Drug Discov- 63E, 1997 ery and Design 1:461-478, 1993 47. James S, Armstrong P, Califf R, et al: Safety and 33. Stone SR, Hofsteenge J: Kinetics of the inhibition of efficacy of abciximab combined with dalteparin in thrombin by hirudin. Biochemistry 25:4622-4628, treatment of acute coronary syndromes. Eur Heart J 1986 23:1538-1545, 2002 522 KOKOLIS ET AL.

48. Cohen M, Demers C, Gurfinkel EP, et al: A compari- 60. Esprit Investigators: Novel dosing regimen of eptifi- son of low-molecular-weight heparin with unfraction- batide in planned coronary stent implantation (ES- ated heparin for unstable coronary artery disease. PRIT): A randomised, placebo-controlled trial. Lancet Efficacy and safety of subcutaneous enoxaparin in 356:2037-2044, 2000 non-Q-wave coronary events study group. N Engl 61. Choussat R, Montalescot G, Collet JP, et al: A J Med 337:447-452, 1997 unique, low dose of intravenous enoxaparin in elec- 49. Collet JP, Montalescot G, Lison L, et al: Percutane- tive percutaneous coronary intervention. J Am Coll ous coronary intervention after subcutaneous enox- Cardiol 40:1943-1950, 2002 aparin pretreatment in patients with unstable angina 62. Gerlach AT, Pickworth KK, Seth SK, et al: Enoxaparin pectoris. Circulation 103:658-663, 2001 and bleeding complications: A review in patients with 50. Ferguson JJ, Antman EM, Bates ER, et al: The use of and without renal insufficiency. Pharmacotherapy 20: enoxaparin and GPIIb/IIIa antagonists in acute coro- 771-775, 2000 nary syndromes. Final results of the NICE 3 study 63. Collet JP, Montalescot G, Choussat R, et al: Enox- [abstract]. J Am Coll Cardiol 37:1253-1297, 2001 (2 aparin in unstable angina patients with renal failure. Suppl A) Int J Cardiol 80:81-82, 2001 51. Kereiakes DJ, Grines C, Fry E, et al: The NICE 1 and 64. Bittl JA, Chaitman BR, Feit F, et al: Bivalirudin versus NICE 4 Investigators, National Investigators Collabo- heparin during coronary angioplasty for unstable or rating on Enoxaparin. Enoxaparin and abciximab ad- postinfarction angina: Final report reanalysis of the junctive pharmacotherapy during percutaneous cor- Bivalirudin Angioplasty Study. Am Heart J 142:952- onary intervention. J Invasive Cardiol 13:272-278, 959, 2001 2001 65. Reed MD, Bell D: Clinical pharmacology of bivaliru- 52. Bhatt DL, Lee BI, Casterella PJ, et al: Safety of din. Pharmacotherapy 22:105S-111S, 2002 concomitant therapy with eptifibatide and enoxaparin 66. Kleiman NS, Klem J, Fernandes LS, et al: Pharmaco- in patients undergoing percutaneous coronary inter- dynamic profile of the direct thrombin antagonist vention: Results of the Coronary Revascularization bivalirudin given in combination with the glycoprotein Using Integrilin and Single Bolus Enoxaparin Study. IIb/IIIa antagonist eptifibatide. Am Heart J 143:585- J Am Coll Cardiol 41:20-25, 2003 593, 2002 53. Arjomand H, Surabhi SK, Cohen M: Unfractionated 67. Kleiman NS, Lincoff AM, Sapp SK, Maresh KJ, Topol versus fractionated heparin for percutaneous coro- EJ: Pharmacodynamics of a direct thrombin inhibitor nary intervention. Curr Cardiol Rep 4:327-333, 2002 combined with a GP IIb-IIIa antagonist: First experi- 54. Karsch KR, Preisack MB, Baildon R, et al: Low mo- ence in humans [abstract]. Circulation 100:I328, 1999 lecular weight heparin (reviparin) in percutaneous (Suppl 18) transluminal coronary angioplasty. Results of a ran- 68. Kleiman NS, Lincoff AM, Harrington RA, et al: Anti- domized, double-blind, unfractionated heparin and thrombin, antiplatelet therapy or both during PCI: A placebo-controlled, multicenter trial (REDUCE trial). preliminary randomized trial [abstract]. J Am Coll Reduction of Restenosis After PTCA, Early Adminis- Cardiol 37:1293-1242, 2001 (2 Suppl A) tration of Reviparin in a Double-Blind Unfractionated 69. Lincoff AM, Bittl JA, Kleiman NS, et al for the RE- Heparin and Placebo-Controlled Evaluation. J Am PLACE Investigators: The REPLACE 1 Trial: A Pilot Coll Cardiol 28:1437-1443, 1996 Study of Bivalirudin Versus Heparin During Percuta- 55. Cavasoglu E, ElRouby S, Kunamneni PB, et al: Mon- neous Coronary Intervention With Stenting and GP itoring low molecular weight heparin during interven- IIb/IIIa Blockade [abstract]. J Am Coll Cardiol 39:16A, tional coronary procedures using sensitive point of 2002 (5 Suppl A) care assay. Blood 102:1171, 2003 70. Lincoff AM, Kleiman NS, Kottke-Marchant K, et al: 56. Kereiakes DJ, Kleiman NS, Fry E, et al: Dalteparin in Bivalirudin with planned or provisional abciximab ver- combination with abciximab during percutaneous sus low-dose heparin and abciximab during percuta- coronary intervention. Am Heart J 141:348-352, 2001 neous coronary revascularization: Results of the 57. Legalery P, Seronde MF, Schiele F, et al: Enoxaparin Comparison of Abciximab Complications with Hiru- as antithrombotic treatment in patients subject to log for Ischemic Events Trial (CACHET). Am Heart J percutaneous coronary interventions: Comparison 143:847-853, 2002 with unfractionated heparin [abstract]. Am J Cardiol 71. Lincoff AM, Bittl JA, Harrington RA, et al: Bivalirudin 90:163H, 2002 (Suppl 6A) and provisional glycoprotein IIb/IIIa blockade com- 58. Young JJ, Kereiakes DJ, Grines CL: Low-molecular- pared with heparin and planned glycoprotein IIb/IIIa weight heparin therapy in percutaneous coronary in- blockade during percutaneous coronary intervention: tervention: The NICE 1 and NICE 4 trials. National REPLACE-2 randomized trial. JAMA 289:853-863, Investigators Collaborating on Enoxaparin Investiga- 2003 tors. J Invasive Cardiol 12:E14-E18, 2000 (Suppl E) 72. Campbell KR, Mahaffey KW, Lewis BE, et al: Biva- 59. The EPISTENT Investigators: Randomised placebo- lirudin in patients with heparin-induced thrombocyto- controlled and balloon-angioplasty-controlled trial to penia undergoing percutaneous coronary interven- assess safety of coronary stenting with use of platelet tion. J Invasive Cardiol 12:14F-19F, 2000 (Suppl F) glycoprotein-IIb/IIIa blockade. Evaluation of Platelet 73. Campbell KR, Wildermann N, Janning C, et al, for the IIb/IIIa Inhibitor for Stenting. Lancet 352:87-92, 1998 ATBAT Investigators: Bivalirudin during percutaneous ANTICOAGULATION FOR PCI 523

coronary intervention in patients with heparin-in- acute myocardial infarction. N Engl J Med 329:673- duced thrombocytopenia: Interim results of the AT- 682, 1993 BAT Trial [abstract]. Am J Cardiol 86:73i-74i, 2000 84. Rao AK, Pratt C, Berke A, et al: Thrombolysis in (Suppl 8A) Myocardial Infarction (TIMI) Trial—Phase I: Hemor- 74. Bhatt DL, Cho L, Lincoff AM, et al: Reduction in rhagic manifestations and changes in plasma fibrin- percutaneous coronary intervention-related bleeding ogen and the fibrinolytic system in patients treated with bivalirudin is particularly striking in women [ab- with recombinant tissue and stract]. J Am Coll Cardiol 39:1053-1054, 2002 (5 . J Am Coll Cardiol 11:1-11, 1988 Suppl A) 85. Moscucci M, Share D, Kline-Rogers E, et al: The Blue 75. Aguirre FV, Gill JB: Increasing benefit, reducing risk: Cross Blue Shield of Michigan Cardiovascular Con- Focusing on hemorrhagic complications in percuta- sortium (BMC2) collaborative quality improvement neous coronary intervention. J Invasive Cardiol 14: initiative in percutaneous coronary interventions. J In- 48B-54B, 2002 (Suppl B) terv Cardiol 15:381-386, 2002 76. Organisation to Assess Strategies for Ischemic Syn- 86. Ahmed WH, Piana RN, Bittl JA: Does bleeding cause dromes (OASIS-2) Investigators: Effects of recombi- ischemic events after coronary angioplasty [ab- nant hirudin (lepirudin) compared with heparin on stract]? Circulation 94:I-376, 1996 (8 Suppl I) death, myocardial infarction, refractory angina, and 87. Bittl J: Switching from heparin to a thrombin-specific revascularisation procedures in patients with acute anticoagulant (bivalirudin) for PTCA in unstable an- myocardial ischaemia without ST elevation: A ran- gina patients reduces major clinical events [abstract]. domised trial. Lancet 353:429-438, 1999 Circulation 102:813, 2000 (18 Suppl) 77. Eikelboom J, Anand SS, Mehta SR, et al: Prognostic 88. Weitz JI, Hudoba M, Massel D, et al: Clot-bound significance of thrombocytopenia during hirudin and thrombin is protected from inhibition by heparin-an- heparin therapy in acute coronary syndrome without tithrombin but is susceptible to inactivation by anti- ST elevation: Organization to assess strategies for thrombin-independent inhibitors. J Clin Invest 86: ischemic syndromes (OASIS-2) study. Circulation 385-391, 1990 103:643-650, 2001 89. The EPILOG Investigators: Platelet glycoprotein IIb/ 78. Serruys PW, Herrman JP, Simon R, et al: A compar- IIIa receptor blockade and low-dose heparin during ison of hirudin with heparin in the prevention of re- percutaneous coronary revascularization. N Engl stenosis after coronary angioplasty. Helvetica Inves- J Med 336:1689-1696, 1997 tigators. N Engl J Med 333:757-763, 1995 90. Kereiakes DJ, Fry E, Matthai W, et al: Combination 79. Matthai WH: Use of argatroban during percutaneous enoxaparin and abciximab therapy during percutane- coronary interventions in patients with heparin-in- ous coronary intervention: “NICE guys finish first.” duced thrombocytopenia. Semin Thromb Hemost 25: J Invasive Cardiol 12:1A-5A, 2000 (Suppl A) 57-60, 1999 (Suppl 1) 91. Adamian M, Mehran R, Mintz GS, et al: Safety and 80. Lewis BE, Matthai WH, Cohen M, et al: Argatroban feasibility of direct thrombin inhibitor bivalirudin dur- anticoagulation during percutaneous coronary inter- ing percutaneous coronary interventions in patients vention in patients with heparin-induced thrombocy- with chronic renal failure. Results of the pilot study topenia. Catheterization Cardiovasc Interv 57:177- [abstract]. Am J Cardiol 90:162H, 2002 (Suppl 6A) 184, 2002 92. Adamian M, Farkouh M, Mehran R, et al: A matched 81. Lewis BE, Mathai W, Grassman ED: Results of a comparison of safety and feasibility of bivalirudin in phase 2/3 trial of argatroban anticoagulation during patients with diabetes mellitus during percutaneous PTCA of patients with heparin-induced thrombocyto- coronary interventions [abstract]. Am J Cardiol 90: penia [abstract]. Circulation 96:I-217, 1997 (Suppl) 161H, 2002 (Suppl 6A) 82. Moledina M, Chakir M, Gandhi PJ: A synopsis of the 93. Cho L, Chew DP, Moliterno DJ, et al: Safe and effi- clinical uses of argatroban. J Thromb Thrombolysis cacious use of bivalirudin for percutaneous coronary 12:141-149, 2001 intervention with adjunctive platelet glycoprotein IIb/ 83. The GUSTO investigators: An international random- IIIa receptor inhibition. Am J Cardiol 91:742-743, ized trial comparing four thrombolytic strategies for 2003