Thrombocytopenia and Outcome in Invasive Cardiology (Part II)

Continued from previous page Mechanism for thrombocytopenia with IIb/IIIa antagonists. A conclusive explanation for the occurrence of thrombocytopenia in conjunction with GP IIb/IIIa receptor antagonism has not been provided; however, immune mediation is certainly causative. Binding of the antagonist to the GP IIb/IIIa receptor may lead to the exposure of a ligand-induced binding site, which is then recognized by either pre-existing or induced antibodies. Alternatively, the metabolite complex of the receptor drug may induce an immune response.³⁰ Yet antibodies are also detected in a high percentage of normal controls and in patients not developing thrombocytopenia, so the pathogenesis remains a bit of a conundrum.³¹ It seems that both the generation of drug-induced antibodies and the formation of neoepitopes induced by drug changes of a restricted set of glycoproteins on the platelet surface need to both occur in order to set in motion the cascade of platelet destruction.³² Mapping of the antigenic sites may in the future help to characterize the genetic polymorphisms, predisposing to the generation of antibody binding sites and helping identify a patient population at risk for the development of thrombocytopenia in addition to the clinical criteria described above. Morbidity associated with thrombocytopenia secondary to GP IIb/IIIa receptor blockade. Of the 2,099 patients who received abciximab bolus, bolus + infusion or placebo in the EPIC trial, patients who developed thrombocytopenia had a higher incidence of unfavorable clinical outcomes (death, MI, bypass surgery, additional percutaneous revascularization or balloon pump insertion) and longer median lengths of stay in the coronary care unit, than did patients who did not develop thrombocytopenia.¹⁶ However, among patients who developed thrombocytopenia, better outcomes were seen if they received abciximab than if they received placebo, particularly a lower incidence of MI, CABG, additional PCI procedure, and the composite endpoint. However, the causality between thrombocytopenia and poorer outcomes could not be established since several of the outcome measures occurred prior to the demonstration of thrombocytopenia by recorded assay. Similarly, while hemorrhage occurred significantly more frequently in patients with thrombocytopenia, it occurred less frequently in patients who had been treated with abciximab than in placebo-treated thrombocytopenic patients (p = 0.046) (Table 1). Abciximab, therefore, increased the rate of major bleeding nearly 3-fold in non-thrombocytopenic patients, but abciximab was associated with a decrease in the incidence of major bleeding in patients with thrombocytopenia of approximately 37%. This relationship does not appear to hold out with the small molecule GP IIb/IIIa inhibitors, with which thrombocytopenia was generally not associated with a higher incidence of major bleeds compared with controls.³³ While platelet transfusions are recommended as therapy for profound (34 Thienopyridines and thrombotic thrombocytopenic purpura (TTP). The thienopyridines are somewhat unique in that the variant of thrombocytopenia rarely induced is TTP, which is characterized not only by severe thrombocytopenia but by microangiopathic hemolytic anemia and neurologic abnormalities in most patients affected with the malady. Fever and renal involvement are also present in the majority of afflicted patients. The mechanism for TTP with the thienopyridines is not well understood, though it appears to be immune-mediated. The thienopyridines clopidogrel and ticlopidine both act to inhibit the platelet ADP receptor by conversion to metabolites that non-competitively antagonize the receptor. This in turn inhibits expression of the GP IIb/IIIa receptor in the high-affinity configuration that binds fibrinogen and large multimers of von Willebrand factor. Although there were no reports of TTP in pivotal clinical trials, post-approval studies revealed an incidence between 1 in 1,600 to 1 in 5,000 patients.^35,36 Eleven cases of severe thrombotic thrombocytopenic purpura have been reported in the literature for clopidogrel, after more than 3 million people had received the drug.²⁵ Clopidogrel-associated TTP occurs early, within 2 weeks of treatment, while ticlopidine-associated TTP occurs after 2 weeks in 95% of the cases reported. Low molecular weight heparin. Low molecular weight heparins (enoxaparin, dalteparin) are derived from UFH by chemical or enzymatic depolymerization to yield fragments that are approximately one third the size of heparin. Like UFH, they are heterogeneous with respect to molecular size and anticoagulant activity. LMWHs have a mean molecular weight of 4,000–5,000 daltons, with a molecular weight distribution of 1,000–10,000 daltons.³⁷ There is a lower incidence of HIT associated with the use of low molecular weight heparins than with unfractionated heparin. Nevertheless, moderate thrombocytopenia occurs at a rate of 1.3% in patients given enoxaparin, 1.2% of patients receiving heparin and 0.6% of patients on placebo in clinical trials, while severe thrombocytopenia occurred at a rate of 0.1% in enoxaparin-treated patients, 0.2% for patients receiving heparin and 0% in patients receiving placebo.³⁸ In 1995, Warkentin looked at the incidence of HIT in a patient population receiving LMWH or unfractionated heparin for DVT prophylaxis following hip surgery and was unable to demonstrate HIT in patients receiving LMWH, while 2.7% of the patients receiving unfractionated heparin developed HIT (p = 0.002). However, heparin-dependent IgG antibodies did occur in the subgroup that received LMWH at just below a third of that seen with unfractionated heparin (2.2% versus 7.8%).³⁹ Despite a lowered immunogenicity, LMWH displays nearly 100% in vitro cross-reactivity to HIT antibodies using sensitive assays, making its use in therapy of patients with HITS contraindicated. Studies have reported a 0.9% incidence of thrombocytopenia with low molecular weight heparin, and severe, profound immune-mediated thrombocytopenia has been reported as well.⁴⁰Heparin partial constructs. Danaparoid, a mixture of glycosaminoglycans with anti-Xa activity, has been used in treatment and therapy of patients with HIT. The drug demonstrated a higher incidence of thrombosis resolution (92% versus 71%), but does cross-react with heparin antibodies and so is not a widely used alternative.⁴¹ Fondaparinux, a new pentasaccharide sequence that was recently shown to lower the incidence of DVT in the orthopedic patient population compared with LMWH,42 should theoretically not react with HIT antibodies or cause HIT, but there is no reported experience with this agent for patients with HIT. Direct thrombin inhibitors. The direct thrombin inhibitors offer an alternative to heparin in that they do not interact with heparin antibodies. They inhibit coagulation by directly interfering with thrombin function without the requirement of antithrombin. These agents can also inhibit clot-bound thrombin, thereby increasing the efficacy of anticoagulation. The anticoagulant effect of these agents is more predictable than with heparin because the direct thrombin inhibitors have less non-specific binding and are not neutralized by PF4. Rebound coagulation after discontinuation of direct thrombin inhibitors is also less likely than with other anticoagulants. The FDA approved direct thrombin inhibitors, including argatroban, lepirudin (Refludan®), and bivalirudin (Angiomax®). Argatroban is indicated as an anticoagulant for prophylaxis or treatment of thrombosis in patients with heparin-induced thrombocytopenia. Lepirudin is indicated for anticoagulation in patients with heparin-induced thrombocytopenia (HIT) and associated thromboembolic disease in order to prevent further thromboembolic complications. Bivalirudin is the only one of the three specifically approved as an anticoagulant for use in patients with unstable angina undergoing PCI in combination with aspirin. Bivalirudin further appears to provide adequate anticoagulation for HITS patients undergoing PCI.⁴³ Each of these drugs will be reviewed separately. Argatroban is a small (molecular weight, 527 daltons) synthetic direct thrombin inhibitor that binds reversibly in a univalent fashion to the catalytic domain of thrombin.^44,45 Argatroban is metabolized by the liver and has not been associated with antibody formation.⁴⁶ There is no reversal agent for argatroban and it has a terminal elimination half-life of 39–51 minutes. Argatroban is highly protein-bound and has an active metabolite. Argatroban was well tolerated in clinical trials of patients with HIT and had a similar bleeding risk compared to historical controls.⁴⁷ Argatroban was associated with lower composite event rates (new thrombosis, limb amputation, death) in patients with clinically diagnosed HIT compared to historic control subjects [26% versus 39% for patients with HIT alone (p = 0.014) and 44% versus 57% for patients with thrombosis (p = 0.13)]. Argatroban use does not increase the incidence of bleeding compared with placebo.⁴⁸ Argatroban has not been extensively studied in PCI. A series of 91 patients with HIT undergoing PCI with argatroban was reported in 2001 with no complications related to argatroban use.^49,50 An application for use of argatroban in PCI has been submitted to the FDA based on a study of 91 patients. The investigators of the study concluded that procedural outcomes and bleeding rates were similar to those reported historically for heparin use during PCI.⁵⁰ Use of argatroban is not recommended for patients with hepatic impairment due to its metabolic pathway. In addition, argatroban prolongs the INR and therefore requires a treatment strategy for overlapping treatment with coumadin. Lepirudin is a recombinant molecule identical to natural hirudin except for the substitution of leucine for isoleucine at the N-terminal end and the absence of sulfate on tyrosine at position ⁶³. Lepirudin binds tightly and irreversibly to thrombin and thus has a relatively long duration of action. Lepirudin is excreted primarily via renal mechanisms; therefore, patients with renal impairment are at increased risk for severe bleeding.⁵¹ There is no known reversal agent for lepirudin. More than 40% of patients treated with lepirudin develop anti-hirudin antibodies that sometimes act to increase the compound’s anticoagulant effects and require a dose reduction.⁵² The incidence of thrombocytopenia occurring with lepirudin use is similar to that of unfractionated heparin in patients with acute coronary syndromes.⁵³ The primary trials that evaluated lepirudin use in thrombocytopenic (heparin-induced) patients were HAT-1 and HAT-2. Lepirudin was associated with a lower composite event rate of death, amputation or new thromboembolic complication than historical controls in patients with HIT (25 versus 52 at 35-day follow-up; p = 0.014), and a trend toward a lower endpoint in patients with thrombosis (p = 0.12). Meta-analysis showed significant benefit in HIT patients with thrombosis (p = 0.004). Bleeding, however, occurs more frequently with lepirudin therapy. In HAT-2, 44.6% of patients experienced at least one bleeding event compared to 27. 2% for the heparin historical control group (p = 0.0001).⁵⁴ While lepirudin seems to offer a similar prognosis as heparin for patients with thrombocytopenia who have acute coronary syndromes with regard to the incidence of death and MI, it is associated with a 170% higher risk of major bleeding than unfractionated heparin (OR, 5.4 and 95% CI, 2.6–11.3 versus OR, 2.0 and 95% CI 0.3–14.4).⁵³ In the OASIS-2 trial, a total of 10,141 patients with acute coronary syndromes without ST elevation were randomized to receive 72 hours of intravenous hirudin or standard heparin.55 Compared to heparin, significantly fewer patients treated with lepirudin (6.7% versus 5.6%; p = 0.0125) experienced cardiovascular death, new myocardial infarction or refractory angina (the secondary endpoint). There is minimal evidence for lepirudin use in PCI. There have been case reports,⁵⁶ but no randomized prospective trial comparing lepirudin to heparin in non-HITs patients demonstrating any improvement in safety or efficacy. Lepirudin has been used as the primary anticoagulant in PCI for at least 4 patients with HITS without reported complications.⁵⁶ The major problems with lepirudin include the significantly increased need for transfusions (p = 0.02), as documented in a meta-analysis of the two large prospective trials⁵⁷ and the need for caution in using the drug in renally impaired patients. The Organization to Assess Strategies for Ischemic Syndromes (OASIS)-2 study was analyzed to specifically examine whether thrombocytopenia occurring with lepirudin therapy carried a morbidity risk, and concluded that thrombocytopenia, which occurred in a similar number of unfractionated heparin and lepirudin-treated patients (1.1% versus 0.9%), carried an increased risk of major bleeding.⁵³ Desirudin (Revasc®) is a recombinant molecule that is identical to natural hirudin except for the absence of sulfate on tyrosine at position 63, and was studied in HELVETICA for its effectiveness in preventing restenosis after PCI.⁵⁸ The administration of desirudin was associated with a significant reduction in early cardiac events, which occurred in 11.0%, 7.9%, and 5% of patients in the respective groups (combined relative risk with hirudin, 0.61; 95% CI, 0.41–0.90; p = 0.023).⁵⁸ Group 1 received 10,000 IU heparin bolus + 24-hour heparin infusion + subcutaneous placebo twice daily for three days; Group 2 received 40 mg hirudin bolus + 24-hour hirudin infusion + subcutaneous placebo twice daily for three days; Group 3 received 40 mg hirudin bolus + 24-hour hirudin infusion + 40 mg subcutaneous hirudin twice daily for three days.⁵⁸ A significant reduction with desirudin compared to heparin in the incidence of death or non-fatal (re)infarction at 24 hours in patients with acute myocardial infarction was reported in the Global Use of Strategies to Open Occluded Coronary Arteries (GUSTO)-IIb trial, but not in the Thrombolysis and Thrombin Inhibition in Myocardial Infarction (TIMI)-9B trial.⁵⁹ Despite the early reduction shown in GUSTO-IIb, desirudin was not associated with an improved long-term clinical benefit at 30 days compared with heparin. Thrombocytopenia has not been shown to occur with desirudin use. Desirudin is not an FDA-approved agent in the United States. Bivalirudin (Angiomax®) is approved for use in PCI in conjunction with aspirin. The agent appears unique among the direct thrombin inhibitors in that it has the shortest half-life (approximately 25 minutes) and was associated with a significantly lower incidence of ischemia (22% reduction; p = 0.039) as well as bleeding (62% reduction; p = 0.001) in a population of 4,312 patients with unstable angina undergoing angioplasty.⁶⁰ Bivalirudin is a 20-amino acid polypeptide with a molecular weight of 2,180 daltons that binds to thrombin at both its active and catalytic sites.⁶¹ Although bivalirudin binds to thrombin with high affinity, the portion of the molecule that binds to the active site is cleaved, allowing recovery of thrombin functionality.⁶¹ Additionally, only 20% of the unaltered active drug substance is excreted via renal mechanisms, the rest being metabolized by proteolytic cleavage, making bivalirudin a good choice in patients with mild to moderate renal and hepatic impairment.⁶² In clinical studies, bivalirudin has been successfully used without significant attributable adverse events in the treatment of 62 patients with HIT or HITTS.^63,64 Of those undergoing PCI, the procedure was successful in 94% and failed in 1 patient due to an unapproachable lesion. In an interim analysis of the ongoing ATBAT trial, bivalirudin was reported to be successfully used without complications in 11 patients with active HITS.⁶⁴ Bivalirudin provided adequate anticoagulation in these patients. There is more reported patient experience with bivalirudin (13,223 patients in completed Phase 1–3 trials) than any of the other direct thrombin inhibitors. In REPLACE-1, however, the incidence of confirmed thrombocytopenia with bivalirudin in combination with GP IIb/IIIa inhibitors was 3 of 365 patients (0.8%),⁶⁵ which is the lowest reported figure for the incidence of thrombocytopenia in association with abciximab and may lend support to the contributory role of heparin in producing thrombocytopenia with GP IIb/IIIa inhibitor use, particularly abciximab. Overall, given the preponderance of experience with bivalirudin, as well as its favorable risk profile, it warrants consideration as an alternative anticoagulant in patients at increased risk for thrombocytopenia. Practical recommendations. While cardiologists cannot completely eliminate the risk of thrombocytopenia, it seems clear that appropriate monitoring of patients with frequent platelet counts in the post-procedure period and appropriate selection of anticoagulants and/or platelet receptor antagonists can improve outcomes. The take-away points from this review relate to understanding the risks related to heparin, LMWH, GP IIb/IIIa inhibition and repeat dosing, and the evolving role for the direct thrombin inhibitors. A good deal of the problem relates to the relative inability to diagnose quickly and in a practical fashion the presence of antibodies formed from exposure to these drugs. At present, the enzyme-linked immunoabsorbent assays commonly require up to five days to obtain results and are expensive tests with limited availability. Should such a rapid diagnostic tool become practically available, then some but not all of the potential risks for the development of thrombocytopenia might be avoided by improved selection of anticoagulants and platelet inhibitors. Repeat dosing with unfractionated heparin and LMWH carries a significantly higher risk for the development of thrombocytopenia related to the unexpectedly large portion of patients who develop heparin PF4 antibodies. With as many as 5–10% of patients carrying these antibodies and a potential interaction in as many as 50% of these, it seems prudent to consider an alternative anticoagulant that does not carry the risk of thrombocytopenia, particularly if this re-exposure is within an 8-week period of the initial exposure. The GP IIb/IIIa antagonists, and abciximab in particular, should also be used with great caution on repeat procedures. While the incidence of thrombocytopenia in primary cases with abciximab was 5.4%, it is reasonable to suspect that the incidence of antibody formation is significantly larger than this figure, as in the case of heparin antibody incidence. Thus, repeat dosing with abciximab within the 8-week to 100-day window in which circulating antibodies may be encountered, coupled with the reports mentioned earlier of “anaphylactic” responses to repeat dosing with GP IIb/IIIa antagonists, should lead one to consider whether or not addition of a GP IIb/IIIa receptor antagonist is prudent on repeat procedures. Herein, the clinician is faced with the choice of selecting a different GP IIb/IIIa receptor antagonist, with a lesser but still defined risk for thrombocytopenia occurrence, or potentially considering the benefit obtained with the use of GP IIb/IIIa receptor blockade and seeking alternative anticoagulants that provide a similar magnitude of benefit. Here, it is again interesting to consider bivalirudin. Thus, we return to the emerging role and place of direct thrombin inhibitors in PCI. Their use is associated with a reduced incidence of thrombocytopenia, although lepirudin is associated with increased risk of bleeding. Argatroban and bivalirudin appear to offer attractive profiles for use in PCI. Significantly, only bivalirudin is approved for use in this indication. Clinical experience with bivalirudin in patients with HIT suggests bivalirudin is safe and provides adequate anticoagulation in the thrombocytopenic population or patients at risk for the development of the condition. Further, bivalirudin has demonstrated reductions in both bleeding and ischemic events in unstable patients undergoing PCI. Thus, under conditions that may result or contribute to thrombocytopenia, consideration of bivalirudin for anticoagulation may be warranted. Proposed clinical trials with bivalirudin should further clarify the comparative endpoint occurrence rate in patients with HIT/HITTS. Ongoing investigations will measure the relative benefit of bivalirudin-based anticoagulation to heparin-GP IIb/IIIa inhibitor use. If these studies prove bivalirudin to provide improved safety and efficacy in a broad population undergoing contemporary PCI, use of bivalirudin could potentially lower the risks associated with thrombocytopenia development and ultimately improve patient outcomes.