Thrombocytopenia and Outcome in Invasive Cardiology (Part I)

Since Andreas Gruentzig first introduced coronary angioplasty in 1977, clinicians have been driven to lower the risks associated with the procedure and improve patient outcomes. Improvements in catheter technology as well as the clinician’s pharmaceutical armamentarium have resulted in significant reductions in the risks associated with percutaneous coronary interventions (PCI), so that current event rates (including bleeding) for elective PCI are in the 10% range, down several-fold from the initial prognosis for morbidity with the procedure. Risk stratification strategies using appropriate pharmaceutical and device choices for patients presenting with higher-risk profiles have improved outcomes. A decrease in relative or absolute platelet count is a risk factor more common than previously recognized, contributing to increased bleeding and thrombotic events. Awareness of thrombocytopenia is growing, with the incidence likely to rise with increased use of drugs like heparin and the glycoprotein (GP) IIb/IIIa receptor antagonists that affect platelet receptors and can elicit an immune response that ultimately removes platelets from functional participation in coagulation. Many drugs used within the interventional cardiology suite can initiate and/or exacerbate thrombocytopenia and affect patient outcomes. This review centers on the drug-induced platelet abnormalities that occur primarily through formation of antibody complexes to induce platelet destruction and will examine the incidence, risks and contributors to thrombocytopenia and discuss some strategies that may be employed to lessen this risk. Types, origins, and incidence of thrombocytopenia. Thrombocytopenia is characterized by decreases in platelet counts occurring on both an absolute and a relative basis. Thrombocytopenia is a result of disturbances in platelet production or distribution, or their destruction caused by hereditary, acquired, iatrogenic factors, and pharmacological therapies that alter platelet ability to activate and participate in clot formation. Medical conditions that can produce absolute or functional thrombocytopenia, such as sepsis, disseminated intravascular coagulation (DIC), thrombotic thrombocytopenic purpura (TTP) and hemolytic-uremic syndrome (HUS), also occur in patients presenting for PCI. The literature separates thrombocytopenia into mild (1 When platelet counts are between 50,000 and 100,000/µl, patients are at increased risks for bleeding associated with surgical procedures; at counts between 20,000 and 50,000/µl, bleeding may occur with minor trauma; and with counts less than 10,000/µl, patients are at high risk for spontaneous bleeding.² Relative thrombocytopenia occurs when platelet counts drop precipitously. Severe thrombocytopenia occurring in conjunction with qualitative platelet dysfunction, as in the case of thienopyridine or GP IIb/IIIa use, presents a higher risk for bleeding than severe thrombocytopenia without dysfunction.³ The non-linearity of the bleeding time associated with thrombocytopenia of defined value and different etiology was demonstrated by Harker approximately 20 years ago.³ This has resulted in differences in the absolute platelet count at which clinicians are guided to initiate platelet transfusion (the so-called “trigger”) for thrombocytopenia of different etiologies. Pseudothrombocytopenia, an artificial laboratory-indicated low platelet count, is relatively common. Platelet clumping in blood samples collected into EDTA can affect the size of platelets “seen” by the instrumentation and cause functional platelets to be classified as leukocytes. Platelets can also adhere to leukocytes, resulting in an apparent thrombocytopenic condition. While conferring little actual morbidity as a result of this condition, there is definite harm associated with unnecessary platelet transfusion and therapies, as well as the cancellation or postponement of scheduled interventional procedures. The frequency of pseudothrombocytopenia increases with the use of GP IIb/IIIa inhibitors. Review of the CAPTURE, EPIC, EPILOG and EPISTENT trials showed pseudothrombocytopenia to occur in 2.1% of abciximab-treated patients [95% confidence intervals (CI) 1.7%, 2.5%] and in only 0.6% of placebo-treated patients (p 4 Identification of thrombocytopenia as spurious in origin may require collecting blood into ammonium oxalate diluting fluid and counting platelets manually. Incidence: In the general medical population, thrombocytopenia with a platelet count of less than 100,000/µl occurs more commonly than expected and is generally associated with drug therapy. The morbidity of thrombocytopenia as an isolated phenomenon has been studied and is significant. Cumulative death rates within a 5-year period after a hospital admission carrying a diagnosis of thrombocytopenia are 4 times greater than that of the general age-matched, medical population,⁵ and individual event rates are similarly increased. Thrombocytopenia following thrombolytic therapy is also relatively common. The incidence of thrombocytopenia in 1,001 patients enrolled in Phases 2, 3 and 5 of the Thrombolysis and Angioplasty in Myocardial Infarction (TAMI) trial and the urokinase trial was 16.4%, with no difference among the various thrombolytic regimens (urokinase, tPA, combination therapy).¹ Apart from the higher likelihood of bleeding in this group, patients with thrombocytopenia also had a lower median ejection fraction and a higher likelihood of 3-vessel disease than a non-thrombocytopenic comparable group. The study was further able to demonstrate a higher in-hospital mortality rate and a more complicated hospital course, which was statistically significant versus to the comparable group, even after consideration of other variables. Thrombocytopenia at presentation for PCI. While there has been a substantive amount written on the occurrence of thrombocytopenia following PCI, in relation to drugs administered, there has been little systematic review of the morbidities posed by patients presenting with low platelet counts of unclear etiology. A European study of patients admitted to the intensive care unit (ICU) revealed a significant incidence and morbidity of thrombocytopenia in the critically ill. Of 329 patients studied, the incidence of at least mild thrombocytopenia was 41.3%, with a 5-fold increase in bleeding rate with mild thrombocytopenia (21.4% versus 4.1%) and a 12-fold increase in bleeding for moderate thrombocytopenia (52.6%).⁶ In addition, patients with mild thrombocytopenia demonstrated increased length of stay in the ICU [8 days (range, 4–16 days) vs. 5 days (range, 2–9 days)], higher ICU mortality (odds ratio, 5; 95% CI, 2.7–9.1), and hospital mortality (p 6 In a prospective study of 1,000 patients who underwent femoral arterial puncture for a diagnostic or therapeutic vascular procedure, platelet counts below 100,000/µl (of any etiology) were statistically correlated with a higher incidence of hematoma (p = 0.002).⁷ No clear evaluation of the progressive risk of complications with stratified classifications of increased severity of thrombocytopenia has been published. To some extent, such a classification would be meaningless, since thrombocytopenia of different etiologies carries different risk profiles. Accordingly, it is much more meaningful to review the preventable and drug-induced causes of thrombocytopenia in PCI and review risk profile and outcome data in these groups. Heparin-induced thrombocytopenia and thrombosis syndrome (HITTS) Heparin interacts with platelet factor-4 (PF4) to form antibodies that can cause platelets to agglutinate and increases their destruction. Antibody formation is relatively common, occurring in as many as 61% of patients who are exposed to heparin boluses for cardiac surgery.⁸ Beef heparin has been shown to be more likely to stimulate antibody generation than porcine heparin in 4 randomized studies in medical patients, although a recent randomized study in cardiac surgical patients did not demonstrate this conclusively.^9,10 It was previously thought that antibody formation to heparin conferred no particular morbidity or risk in the absence of a discernible decrease in platelet count, but Mattioli recently reported a 50% increase in the incidence of thrombotic events in a prospective, year-long study of the risks associated with antibody formation in the cardiac surgical population.¹¹ This is currently being confirmed at other institutions in the United States. The incidence of heparin-induced thrombocytopenia (HIT) is estimated at between 2% and 7% of patients exposed to heparin.^12,13 Given the 12 million annual exposures to heparin, the at-risk population is estimated at 360,000 in the United States.¹⁴ Certainly repeated exposure to heparin over a short time period, as may occur in patients undergoing multiple diagnostic and interventional procedures, increases the risk of developing HIT. The terminology should be defined, since both heparin-associated thrombocytopenia (HAT) and HIT appear in the literature and there is some mild confusion over the terms. HIT denotes pathologic antibody formation where heparin exposure is demonstrated to have produced the platelet decline, while HAT describes a condition in which a pathogenic role for heparin antibodies cannot be convincingly demonstrated and the thrombocytopenia present may be non-immune in origin. The biologic basis for the variability in heparin antibody incidence among patient populations, ranging from 1% in cardiac medical patients to 5% in orthopedic patients and 50% in the cardiac surgical patients, is understood.¹² The HIT antigen is a neoantigen on PF4 that is formed when PF4 binds to heparin. Only stoichiometric concentrations of heparin and PF4 will form the antigen. Thus, frequency of antibody formation will depend not only upon the heparin dose, but also upon circulating levels of PF4. Cardiac surgery increases circulating PF4 levels and this most likely contributes to the very high incidences of antibody formation. Mattioli et al. demonstrated a 66% incidence of out-of-hospital thrombotic events occurring in the year following discharge after treatment for unstable angina in patients with HIT antibodies.¹¹ Six retrospective cohort studies spanning a 16-year period have examined the incidence of thrombosis complicating HIT in-hospital, and have pointed to an incidence of thrombosis between 7–76%.¹² The increased incidences of bleeding complications and thrombotic events are not only costly, but in some circumstances life-threatening;¹³ therefore, a thorough understanding of the available alternatives to heparin and the evidence supporting their use is warranted. Therapeutic adjuncts and alternatives to heparin Given that thrombocytopenia occurs in roughly 5% of patients currently treated in invasive angioplasty and stenting, and that this confers an increased morbidity profile for patients, the clinician may elect to consider alternative management strategies designed to reduce the risk of thrombocytopenia and improve outcomes. It is further important to examine whether or not morbidity associated with the thrombocytopenia that occurs with alternative anticoagulants carries a difference in outcome prognosis. Each adjunctive platelet receptor inhibitor as well as alternative anticoagulant that may be used in conjunction with or instead of heparin will thus be reviewed in some detail, specifically addressing the issue of thrombocytopenia and available outcome data. GP IIb/IIIa inhibition. The incidence of thrombocytopenia in the 6 large ACS trials ranges from 1.0–7.3%.15–20 Different trials utilized somewhat different definitions for thrombocytopenia, which may explain some of the differences in the incidences reported. In PURSUIT (Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy), thrombocytopenia occurred in 7.0% of the non-ST segment elevation patients and was found to predict significantly more bleeding events (2-fold increase in moderate/severe bleeding) and increased risk of ischemic events (p 15 This analysis of PURSUIT used a definition of thrombocytopenia as a platelet nadir of 16 In GUSTO-IIb,¹⁷ ESSENCE,¹⁸ TAMI¹⁹ and the conservative arm of TIMI-II,20 the rates of thrombocytopenia by this same definition were 1.0%, 3.0%, 7.3% and 1.5%, respectively. The differences in the rates observed can be primarily accounted for by differences in the baseline characteristics of the study groups, duration and frequency of platelet monitoring, and the differing amount of exposure to GP IIb/IIIa antagonists among the studies. Differences among the GP IIb/IIIa inhibitors with respect to incidence and severity of thrombocytopenia. Among the available GP IIb/IIIa inhibitors, a higher incidence of thrombocytopenia is associated with abciximab use, although more than one third of the thrombocytopenia occurring with abciximab is classified as pseudothrombocytopenia.⁴ In 9 multicenter studies comprising 24,946 patients, the incidence of thrombocytopenia of 21 Using the static definition, the incidence of thrombocytopenia was 4.9% with eptifibatide in PURSUIT compared to 3.9% with abciximab in EPIC, although this incidence was greater (5.2%) in the bolus plus infusion arm,^15,16 which was statistically validated through a logistic regression model (p = 0.016). PURSUIT failed to show any difference between placebo and active arms in the incidence of thrombocytopenia, while a meta-analysis of 4 large placebo-controlled abciximab trials (CAPTURE, EPIC, EPILOG and EPISTENT) demonstrated a significant difference between the two groups, with an occurrence of 3.7% in abciximab-treated patients and 1.8% in placebo-treated patients (p 4 In this meta-analysis, pseudothrombocytopenia also occurred significantly more frequently in abciximab-treated patients than placebo (2.1% vs. 0.6%; p 15 Severe profound thrombocytopenia with abciximab (platelet count 22 This level of thrombocytopenia occurs less frequently with the small molecule GP IIb/IIIa antagonists, eptifibatide and tirofiban. However, Hongo reported 4 cases of acute profound thrombocytopenia with eptifibatide in their institution’s experience.²³ In addition, the literature contains several reports of acute anaphylactic reactions with markedly dangerous levels of thrombocytopenia in relation to abciximab therapy, whereas they are less so for other GP IIb/IIIa inhibitors. Iakovou reported two recent cases with platelet nadirs of 24 Makoni reported a similar case of a patient who did not manifest any symptoms of anaphylaxis, but whose platelet count decreased to 5,000/µl from a baseline of 174,000/µl 7 hours after administration of abciximab.²⁵ Tirofiban produces a smaller incidence of thrombocytopenia than abciximab, with varying reports on the incidence of the condition. In PRISM (Platelet Receptor Inhibition in Ischemic Syndrome Management), a study of 3,232 unstable angina patients, thrombocytopenia occurred more frequently with tirofiban than with heparin (1.1% versus 0.4%; p = 0.04),²⁶ but the RESTORE (Randomized Efficacy Study of Tirofiban for Outcomes and Efficacy) trial of 2,139 patients undergoing balloon angioplasty or directional atherectomy failed to show a significant difference in the incidence of thrombocytopenia between tirofiban and placebo (1.1% versus 0.9%; p = 0.709).27 This may have more to do with the higher incidence of thrombocytopenia accompanying PCI alone than in the former study. Figure 1 illustrates the rates of severe thrombocytopenia of platelet counts 28 In addition, this study observed the antibody response specifically and demonstrated that although 4.8% of patients demonstrated the antibody after initial administration, an additional 19% of patients demonstrated the antibody after the first re-administration of the drug. The authors concluded that platelet counts should be followed closely in patients receiving abciximab readministration, and perhaps beyond the first 24 hours due to the timing of thrombocytopenia witnessed. The authors made no conclusion as to whether alternative GP IIb/IIIa receptor antagonists should be considered when readministration is necessary, but tirofiban has been used safely in a patient who developed profound thrombocytopenia in response to prior abciximab therapy.²⁹ Continued on next page