The Anticoagulant Therapy with Bivalirudin to Assist in the <br />
Performance of Percutaneous Coronary Intervention in Patients with

Heparin-induced thrombocytopenia (HIT) is infrequent and often goes unrecognized. It occurs in ~1% to 5% of the patients given unfractionated heparin,¹ about 25–50% of whom will develop HIT with thrombotic syndrome (HITTS).² Morbidity and mortality are high, and more than 50% of patients suffering thrombotic complications will die.³ Patients with recognized HIT may require anticoagulation for acute coronary syndromes (ACS) or percutaneous coronary intervention (PCI), thus presenting a clinical challenge given the limited alternatives for anticoagulation treatment. The direct thrombin inhibitors provide an attractive alternative for anticoagulation in the settings of ACS and PCI. These agents inhibit thrombin directly, acting to prevent thrombin activation and generation. Bivalirudin, a direct thrombin inhibitor, is a synthetic, 20-amino-acid polypeptide that interacts with both the active site and exosite 1 on thrombin. Once bound to thrombin, the bivalirudin molecule is slowly cleaved by thrombin with consequent recovery of function of the thrombin active site, but competitive inhibition with low-affinity binding to the exosite 1 remains.⁴ Therefore, bivalirudin has several advantages over heparin, including high specificity and potency for thrombin inhibition, independence from antithrombin for activity, inhibition of both free and clot-bound thrombin, lack of platelet activation, and no association with immune-mediated thrombocytopenia and thrombosis or cross-reactivity with HIT antibodies.⁵ Direct thrombin inhibitors have been studied extensively in non-ST segment elevation ACS, acute ST segment elevation myocardial infarction (MI), and during PCI. While no study has shown superiority of direct thrombin inhibitors in ACS, a systematic overview of trials comparing direct thrombin inhibitors with heparin in ACS showed the superiority of direct thrombin inhibitors in reducing the composite of death or MI, particularly with the bivalent inhibitors (hirudin and bivalirudin) versus univalent inhibitors (argatroban, efegatran, and inogatran).⁶ A second systematic overview and a recent trial in contemporary PCI support these finding.^7,8 The Anticoagulant Therapy with Bivalirudin to Assist in the performance of PCI in patients with heparin-induced Thrombocytopenia (ATBAT) trial was designed to evaluate the safety and efficacy of direct thrombin inhibition with bivalirudin during PCI in patients with HIT or HITTS. Methods Study design. ATBAT was a prospective, multicenter, open-label, single-arm study to evaluate the safety and efficacy of bivalirudin in patients undergoing PCI with newly diagnosed or previous HIT or HITTS. Patients were enrolled in 24 centers in the United States and Germany (Appendix A). Data on the first 11 patients have been reported.^9,10Patient population. Patients aged 18 years or older with a new diagnosis of or previous clinically suspected or objectively recorded HIT/HITTS were eligible for inclusion if the decision to perform PCI had been made. The protocol defined HIT as: 1) a positive heparin-induced platelet aggregation or other functional assay for HIT or immunoassay for HIT antibodies, or 2) as thrombocytopenia associated with heparin therapy in which the platelet count had decreased to = 30% from the pretreatment platelet count) or to = 40% from the pretreatment platelet count). The protocol-defined HITTS by thrombocytopenia and arterial or venous thrombosis (deep-vein thrombosis, pulmonary embolism, mesenteric venous or arterial thrombosis, acute MI, left ventricular thrombus, ischemic stroke, or occlusion of limb arteries) was diagnosed by physical examination/laboratory evidence or appropriate imaging studies (duplex ultrasound, venography, ventilation-perfusion scan, venous or arterial angiography, magnetic resonance imaging or angiography, or catheterization). Exclusion criteria included overt or known active, significant, bleeding; definitive evidence of an alternative explanation for thrombocytopenia (such as disseminated intravascular coagulation, sepsis, other consumptive coagulopathy, immune thrombocytopenic purpura, thrombotic thrombocytopenia purpura, hemolytic uremia syndrome, or bone-marrow failure); coagulation disorder or bleeding diathesis unrelated to HIT/HITTS; renal impairment, defined as calculated creatinine clearance = 180/100 mmHg at study entry); ongoing anticoagulant therapy at study entry unless anticoagulant therapy could be safely discontinued, or, in the case of oral anticoagulation with warfarin, if the international normalized ratio could be safely reduced to Study treatment and follow-up. Bivalirudin was to be given >= 5 minutes before the PCI as a 1 mg/kg intravenous bolus and continued as a 2.5 mg/kg/hour infusion for 4 hours (high-dose group). If the attending physician determined that the patient required anticoagulant therapy for > 4 hours, then the bivalirudin dose was to be reduced to 0.2 mg/kg/hour for up to 20 hours after PCI. In January 2002, the bivalirudin dosing was changed to a 0.75 mg/kg bolus and 1.75 mg/kg/hour infusion during PCI and for up to 4 hours after PCI, if the physician considered additional anticoagulation necessary (low-dose group). These changes were based on data from the Comparison of Abciximab Complications with Hirulog ischemia Events Trial (CACHET) and Randomized Evaluation in PCI Linking Angiomax to reduced Clinical Events (REPLACE)-1 trial showing efficacy and safety with the lower dose.^11,12 Functional and immune assays for HIT were not required, but if done, were performed by local laboratories. Baseline coagulation (activated clotting time [ACT], activated partial thromboplastin time [aPTT/PT, optional]), hematology (complete blood count), and routine chemistries [blood urea nitrogen (BUN), creatinine, bilirubin, liver-function tests] samples were to be drawn within 24 hours of starting bivalirudin. Hemoglobin and platelet counts were measured each morning during hospitalization. Samples for ACT, aPTT, and platelet counts also were drawn just after the study-drug bolus, but before PCI began (pre-PCI), just before sheath removal (post-PCI), and 48 hours after drug discontinuation or at hospital discharge, whichever occurred first (predischarge). Serial creatine kinase (CK) or CK-MB assessments were recommended but not mandated, and were not systematically collected. The use of concomitant medications was at the discretion of the physician, including clopidogrel, aspirin, ticlopidine, and glycoprotein IIb/IIIa inhibitors. PCI was to be performed according to institutional standards. Vascular access, use of devices, sheath removal, and use of vascular closure devices was at the discretion of the attending physician. Endpoints and definitions. The primary endpoint was the incidence of major bleeding within 48 hours of study-drug administration or until hospital discharge, whichever occurred first. Key secondary endpoints included event rates for components of the primary endpoint and the ACT, aPTT, and platelet counts at baseline, pre-PCI, post-PCI, and pre-discharge. We also assessed the percent change in hematological variables from baseline to each sampling time. Major bleeding was defined as intracranial bleeding, retroperitoneal bleeding, bleeding resulting in hemodynamic compromise (reduction in systolic blood pressure from baseline to = 3 units of whole blood or packed red cells; or a decrease in hemoglobin of >= 3 g/dL or in hematocrit of >= 9%. Procedural success was defined as Thrombolysis In Myocardial Infarction (TIMI) grade 3 flow after PCI and a final lesion stenosis Statistical analysis. Categorical variables were summarized as percentages, and continuous variables as medians with 25th and 75th percentiles. The point estimate and 95% confidence interval (CI) were calculated for the incidence of the primary outcome. Two patients were enrolled twice and were included in all counts and percentages except for the primary outcome of bleeding, for which they were counted once for their first enrollment. Results Patient recruitment and baseline characteristics. From July 1999 to February 2003, a total of 52 patients were recruited. Baseline characteristics suggest a typical high-risk HIT/HITTS patient population (Table 1). About 50% of the patients were female, 70% of patients had a prior MI, 40% had diabetes, and 30% had prior congestive heart failure. Overall, 11 patients (21%) had a history of HITTS, 22 (43%) had a history of HIT, and 19 (37%) had a new diagnosis of HIT. All patients received bolus drug therapy and some study-drug infusion. Procedural details. Lesion information and procedural results are detailed in Table 2. Two patients in the high-dose group did not undergo attempted PCI because of unsuitable anatomy determined by angiography after enrollment. One patient was treated medically and the other underwent elective bypass surgery. Of those undergoing PCI, thirty-one percent had more than one lesion treated. Procedural success was achieved in 98% of patients, and clinical success was achieved in 96%. Ten patients received concomitant glycoprotein IIb/IIIa inhibitor therapy. Aspirin and clopidogrel were given to 98% and 86% of patients, respectively. Of the 41 patients treated with a stent, forty received aspirin, thirty-six were treated with clopidogrel and one patient received ticlopidine. Primary outcomes. The primary endpoint of major bleeding occurred in 1 patient (1.9%; 95% CI, 0.04–10.65%). The one major bleeding event occurred in the high-dose group, in the patient who underwent elective bypass surgery. No bleeding source was identified, and the bleeding was defined by a drop in hemoglobin/hematocrit without hemodynamic compromise. Only 1 unit of packed red blood cells was given (Table 3). Hematological details. No patient had a platelet count Clinical outcomes. Major cardiac events were rare. One patient in the low-dose bivalirudin group died ~46 hours after a clinically uncomplicated PCI from asystolic cardiac arrest without evidence of recurrent ischemia, infarction or congestive heart failure. No patient suffered acute MI, transient ischemic attack, or stroke. In all, 30 adverse events were reported by investigators in 16 patients (31%) (Table 6). The two patients with reported thrombocytopenia had platelet decreases of 40–50% following study-drug treatment but absolute platelet counts >75X109/L with no bleeding or thrombotic complications. Discussion In the ATBAT trial, bivalirudin treatment appeared safe, and 98% of patients undergoing PCI had a successful procedure. The patient population had a number of risk factors for increased risk of ischemic and bleeding complications, including a higher proportion of women than expected, majority of patients with prior MI, and 21% reported a history of HITTS. The primary end point of major bleeding occurred in only one patient. No patient had significant thrombocytopenia (platelet count Experience with bivalirudin in PCI. Bivalirudin has been studied in a series of clinical investigations in patients undergoing elective and urgent PCI for various indications, including stable coronary artery disease, unstable angina, and postinfarction angina. Following a dose-ranging study in 291 patients,¹³ a large trial of 4,312 patients undergoing PCI showed no significant reduction in the primary composite endpoint of death, MI, abrupt vessel closure, or recurrent ischemic events with bivalirudin compared with heparin (11.4% versus 12.2%, respectively) but significant reductions in the composite endpoint of death, MI, or revascularization (6.2% versus 7.9%; p = 0.039) and bleeding (3.5% versus 9.3%; p 14,15 After the encouraging results of the CATCHET11 and REPLACE-1 trials,¹² the 6,002-patient REPLACE-2 study better defined the role of bivalirudin and provisional glycoprotein IIb/IIIa inhibition during PCI.⁸ The bivalirudin dose was a 0.75 mg/kg bolus and 1.75 mg/kg/hour infusion during PCI and was compared with heparin (64 U/kg bolus) and planned glycoprotein IIb/IIIa inhibition (abciximab or eptifibatide, at the physician’s discretion). The primary composite outcome of death, MI, urgent repeat revascularization, or in-hospital major bleeding occurred in 9.2% of patients in the bivalirudin group compared with 10.0% of patients in the heparin group, which met the specified statistical criteria for noninferiority. In-hospital major bleeding was significantly reduced in the bivalirudin group (2.4% versus 4.1%; p Experience with bivalirudin in HIT. Bivalirudin has been evaluated in two registries including 39 patients with previous HIT requiring anticoagulation. The clinical indication was PCI for 17 of 39 of these patients, of which 16 were successful. Overall clinical outcomes were comparable to historical datasets.⁹Clinical implications. These data, in concert with those from REPLACE-2, support the use of bivalirudin as an alternative choice as an anticoagulant in patients with HIT undergoing PCI. Two other direct thrombin inhibitors also are commercially available. Argatroban has been evaluated in patients with HIT and HITTS requiring anticoagulation for a various clinical indications in two multicenter registries with 304 patients.¹⁶ Two reports have described the experience with argatroban in a total of 50 patients with HIT undergoing PCI. Procedural success was achieved in 49 of these patients.¹⁷ Argatroban has also been evaluated in three prospective, multicenter, open-label registries of 91 patients with HIT undergoing 112 coronary interventions in total.¹⁸ Glycoprotein IIb/IIIa inhibition was not used. Satisfactory outcome of procedure (freedom from death, emergency bypass surgery, repeat PCI, MI, or bleeding within 24 hours of infusion cessation or hospital discharge, whichever came first) was achieved in 94.5% of patients. ACT values between 300 and 450 seconds were reported in 97.8% of patients. Unlike bivalirudin, argatroban has not been evaluated specifically for use during PCI in patients without HIT. Argatroban is metabolized by the liver and should not be used in patients with liver dysfunction, but there do not appear to be any safety concerns about the use of argatroban in patients with renal impairment. Lepirudin has been evaluated in two studies of patients with HIT requiring anticoagulation.^19,20 The rate of thrombotic complications at 35 days was lower in patients treated with lepirudin, but there were excess bleeding events compared with historical controls with HIT treated with alternative anticoagulants or no anticoagulation. No studies have specifically evaluated lepirudin therapy in patients with HIT undergoing PCI. Lepirudin was been evaluated extensively in the OASIS investigations^21,22 that included more than 10,000 patients with non-ST segment elevation ACS, of which ~5% underwent PCI. Lepirudin is excreted through the kidneys and may accumulate in patients with renal dysfunction. The recommended dose of bivalirudin during PCI in patients with HIT is a 0.75 mg/kg bolus and a 1.75 mg/kg/hour infusion during the PCI and optionally for up to 4 hours after the procedure as warranted. Although the low-dose and high-dose groups in ATBAT showed no apparent differences in measured safety and efficacy outcomes, the lower-dose recommendation is supported by these data and the results of the REPLACE-2 trial, which used the same dosing regimen. The median ACT in patients assigned to receive bivalirudin in REPLACE-2 was 358 (320, 400) seconds, comparable with that observed in this study [335 (274, 394) seconds]. The routine use of glycoprotein IIb/IIIa inhibitors with bivalirudin is not supported by ATBAT or REPLACE-2, but a provisional or “bail-out” approach has been shown to be non-inferior to heparin in elective PCI and the limited experience in ATBAT is supportive Limitations. The major limitation of this study is the relatively small sample size. More than 5,000 patients undergoing PCI have been treated with bivalirudin, however, in trials evaluating its safety and efficacy compared with unfractionated heparin. The lack of a comparator in the ATBAT registry also is a limitation of the study, but there is no current standard for treatment of patients with HIT undergoing PCI, and a placebo-controlled trial would have been unethical. Patients with severe renal impairment (creatinine clearance Conclusion. The results of the ATBAT study, in conjunction with the extensive experience with bivalirudin in patients undergoing PCI, support the use of bivalirudin as an anticoagulant in patients with HIT who require coronary intervention. Acknowledgments. We acknowledge Pat French for editorial assistance and Dr. Peter Teuber and Linda Rootkin for their important roles in trial design, project leadership, data interpretation, and manuscript review.