Overcoming Heparin Limitations in High-Risk Percutaneous Coronary Intervention: The Alternative Strategy — Replacing Heparin wit

Adverse outcomes of PCI include myocardial ischemia and necrosis, the need for endovascular or surgical reintervention, and occasionally death.^1–3 Routine pharmacologic therapy includes antiplatelet and antithrombotic agents. Heparin has been the standard antithrombin, but has a narrow therapeutic window and significant limitations.⁴ In recent years, the pharmacological and clinical limitations of heparin in PCI have been addressed by adding more potent antithrombin agents including GP IIb/IIIa inhibitors and thienopyridines.^5,6 This add-on strategy has reduced ischemic complication rates by 30–50% relative to low-dose, weight-adjusted heparin. However, the incidence of clinically significant bleeding (transfusion, TIMI major bleeding and minor bleeding) has increased almost universally. Furthermore, practice series indicate that both ischemic and bleeding complications remain a substantial clinical and economic problem among the general PCI population.⁷ A potentially attractive alternative strategy is to first overcome the limitations of heparin with more effective and safer antithrombin agents, and then determine the need for potent antiplatelet agents as needed in appropriate patients. Heparin, the historical foundation antithrombin Unfractionated heparin has been the standard intravenous antithrombin therapy in PCI from the outset. Pioneering work used heparin at doses of 3,000 U boluses, but with dextran and prolonged warfarin co-medication.⁸ By the early 1990s, data from case-control studies suggested that in the presence of aspirin, doses of heparin in the range of 10,000–15,000 U targeting activated clotting times (ACT) of 300–350 seconds (Hemochron) were more effective than lower doses,^9–11 but were not associated with increased bleeding risk unless infusions were prolonged after PCI.¹² With the introduction of glycoprotein IIb/IIIa inhibitors, bleeding risk was increased. In response, doses of heparin were reduced, target ACT levels were maintained at 200–250 seconds and vascular access site management was amended.^13,14 How good is heparin — especially in high-risk patients? Formal studies of heparin in PCI have been surprisingly infrequent and heparin treatment remains empiric. But since heparin is ubiquitous in PCI, the reported data prior to the use of GP IIb/IIIa inhibitors provide a basis on which to judge the performance of this indirect antithrombin agent. Risk factors for clinical ischemic complications of PCI are summarized in Table 1. Heparin efficacy: High-risk demographic characteristics. Higher adverse event rates among certain patient groups undergoing angioplasty with heparin alone form the basis for advocating alternative treatment strategies. Specifically, co-investigators of the NHLBI angioplasty registry^2,15 and others^2,16–21 have demonstrated increased risk of death and myocardial ischemia in elderly patients treated with heparin for PCI, with some question surrounding outcomes in octogenarians.^3,17,22 Likewise, female gender has been implicated as an independent risk factor in some series,^23,24 but not others,^2,25–27 while low-normal (2) and low (2) or high (> 35 kg/m²) body mass index appears to increase the risk of death 2–7-fold.²⁸ Interestingly, age, gender and low body weight are also powerful drivers of bleeding risk from heparin in PCI and are often coexistent.^29,30Heparin efficacy: High-risk clinical characteristics. Diabetes mellitus is strongly associated with coronary artery disease^31,32 and reported in approximately 20% of patients undergoing PCI in clinical trials.^14,33–40 Diabetes predicts increased risk of restenosis⁴¹ and may increase the rate of death, MI and urgent revascularization, as observed in the EPISTENT study.³⁴ This increased risk is probably explained by an association of diabetes with increased incidence of plaque ulceration and increased potential for intracoronary thrombus formation. Angioscopic studies in diabetic and non-diabetic patients with unstable angina have revealed ulcerated plaque in 94% of the diabetics versus 60% in non-diabetics (p = 0.01) and intracoronary thrombi was seen in 94% of diabetics versus 55% of non-diabetics (p = 0.004).^31,42 Intracoronary thrombus at the time of PCI therefore seems to be an important factor in the poor outcomes of diabetic patients.³ Renal function has emerged as an important independent predictor of mortality and other complications of coronary artery disease.^43,44 The prevalence of chronic renal disease (CRD) is growing at 7% per year in the United States,⁴⁵ but the incidence is underestimated by serum creatinine so that calculation or measurement of creatinine clearance is a necessary step in patient evaluation. Data from three trials of patients with ACS suggested that the prevalence of renal impairment (defined as creatinine clearance 46 this has been confirmed in the PCI population, where 40% of patients have creatinine clearance 8,47 Chronic renal disease predicts a 2-fold increase in mortality following coronary artery bypass surgery.⁴⁸ PCI results may be even worse;^49–51 even mild renal insufficiency is independently associated with increased mortality.⁷ Mechanisms underlying increased risk in patients with CRD have not been completely defined. Associated risk factors such as diabetes, lipid abnormalities and hypertension may contribute, but this confounding seems to be ruled out by multivariate analyses identifying renal function as an independent risk factor.^7,47 Patients with CRD have more extensive vascular disease and more complex coronary anatomy.⁵¹ Chronic renal disease may create an increased thrombotic tendency independently, mediated by elevated abnormalities in platelet function and activation of the coagulation system,^52–58 which potentiates the risk associated with increased lesion complexity or patient demography. Finally, these patients have an increased risk of bleeding complications as well. ACS patients, i.e., those presenting with unstable angina and non-ST elevated MI, were first classified by Braunwald in 1989 to aid in the decision regarding diagnostic measures and therapy of individual patients, and to provide a more precise basis for including patients in clinical trials, as well as evaluating the outcomes of these trials.⁵⁹ In a subsequent consensus document, patients who failed to respond to heparin within 30 minutes were considered at increased risk of MI or cardiac death.⁶⁰ The classification has been validated prospectively.^61,62 DeFeyter et al.⁶³ proposed adaptations of the Braunwald classification to improve the assessment of patients with unstable angina undergoing PCI. They reviewed 30 published PCI reports of patients treated with heparin, aspirin and PCI (no GP IIb/IIIa inhibitors or thienopyridines were used in these trials), thus representing a reasonable estimate of heparin effectiveness among these patients. Results of the analyses are summarized in Figure 1; they demonstrate the increased risk of ischemic outcomes after PCI in all three “high-risk” clinical presentations. Based on this analysis and observations made in a series of patients treated in Rotterdam, De Feyter proposed the risk stratification of unstable PCI patients (Table 2). The mechanism of increased risk for heparin in UA patients is almost certainly the result of intracoronary thrombus. This link is strongly established by angioscopy, which is a more sensitive method of detection for intracoronary thrombus than angiography. Ramee et al.^31,64 first reported that thrombus was observed in 15/16 UA patients (94%) by angioscopy but only 2/16 patients (13%) using angiography. In a larger, multicenter series reported by the same authors,^31,65 a total of 122 patients undergoing PCI were subjected to angiography and angioscopy. Ninety-five patients (78%) had unstable angina and 27 patients (22%) were stable. Once again, angioscopy (61% incidence) was more sensitive at detecting thrombus than angiography (20%; p 11 reported that patients who experienced adverse clinical outcomes after PCI had significantly lower initial and minimum median ACT times. The odds ratio for adverse outcomes among patients with ACT >= 300 seconds and 9 (0.23; 95% CI, 0.15–0.37) and McGarry et al.¹⁰ (0.22; 95% CI, 0.08–0.65). The combined odds ratio for death or ischemic complications in cases among all three studies was 0.25 (95% CI, 0.17–0.37; p 71 The report by Narins et al. also included a multivariable binary logistic regression model that predicted probability of ischemic complications at any given ACT level. The relation between the initial ACT and adverse outcomes was highly statistically significant (p = 0.015). Moving from the 25th percentile of ACT (324 seconds) to the 75th percentile (413 seconds), the probability of adverse ischemic outcomes declined from 7.9% to 4.5%. These data (published in 1996) are strikingly similar to those reported by Chew et al. in 2001 using aggregate data from 6 randomized trials in 5,216 PCI patients. In this report, an ACT of 350–375 seconds was associated with 6.6% incidence of death, MI or revascularization, compared to 10.1% for an ACT range of 171–295 seconds (p 4 These data support the view that the dose and the pharmacodynamic response to heparin are key determinants in the risks associated with PCI. The best results are seen when the ACT reaches 350 seconds or more. Below this, event rates on heparin are high, as shown in the EPISTENT and ESPRIT trials, in which stents were deployed in elective, low-risk patients. The EPISTENT trial³⁴ reported 10.8% incidence of the so-called triple endpoint, and the ESPRIT trial37 reported 10.5% among elective PCI patients (Table 3). Whether current advances demonstrate a similar magnitude of benefit when compared with higher levels of heparin anticoagulation has not been determined. Unfortunately, this conjecture cannot be adequately put to the test. The only two published randomized trials of heparin dosing have not solved this question due to small samples sizes. Vainer et al.⁷² reported 13.2% ischemic complications with 5,000 U heparin (ACT, 209 seconds) compared to 8.0% with 20,000 U (ACT, 381 ± 117 seconds; odds ratio, 1.74; 95% CI, 0.91–3.35). Conversely, Boccara et al.73 reported freedom from death, MI, unplanned revascularization or bail-out stenting in 91% of patients treated with 15,000 U and 95% treated with 100 U/kg (odds ratio, 1.88; 95% CI, 0.80–4.50). The conflicting and inconclusive nature of these data may be the result of small sample sizes. The smoking gun of PCI complications: Intracoronary thrombus In an era of frequent and extensive stenting, increased ischemic risk can largely be attributed to an elevated thrombotic potential. As reviewed, demographic risk factors are associated with procoagulant states; diabetes and chronic kidney diseases are also associated with arterial pathology leading to surface disruption and increased likelihood of thrombus formation. Unstable coronary syndromes are associated with increased incidence of coronary thrombus and systemic heparin resistance. Finally, even anatomic complexity is perhaps a marker for often undetected red clot in the artery. Why does heparin fail when thrombus is present? Unfractionated heparin is a variable chain-length glycosaminoglycan. It acts indirectly by accelerating conformational changes in antithrombin (AT) that enable inhibition of the clotting factors Xa and thrombin.^74,75 Thrombin is also the key plasma protease responsible for cleavage of fibrinogen to fibrin, and activation of a broad range of critical platelet functions.^76,77 Unfortunately, heparin has pharmacokinetic and biophysical limitations that are dependent on its structure and mechanism of action and these limitations become clear in patients with risk factors for intracoronary thrombosis and ischemic complications in PCI. First, heparin’s non-specific binding to endothelium and to plasma proteins, such as vitronectin, fibronectin, histidine-rich glycoprotein, platelet factor-IV, and high molecular weight multimers of von Willebrand factor result in variable pharmacodynamic effects in patients with various clinical presentations.⁷⁸ Binding to cells and plasma proteins decreases the anticoagulant effect of heparin by limiting its availability to interact with AT.⁷⁹ In addition, platelet factor-IV, a highly cationic protein that binds heparin with high affinity, and high molecular weight moieties of von Willebrand factor are released by activated platelets during clotting.⁸⁰ Furthermore, some clinical studies have demonstrated that heparin increases the likelihood of platelet aggregation in ACS patients.⁸¹ Perhaps the most significant limitation of heparin for management of high-risk PCI patients lies in its inability to inhibit thrombin bound to fibrin.⁸² In addition, factor Xa bound to the surface of activated platelets is also resistant to inactivation by the heparin:AT complex^83,84 and, as part of the prothrombinase complex, converts prothrombin to thrombin, thereby increasing the amount of thrombin available to bind to fibrin. Furthermore, thrombin binds to fibrin via exosite-1, a substrate-binding site that is distinct from thrombin’s active site.^85–87 Current evidence suggests that fibrin-bound thrombin is protected from inactivation by the heparin:AT complex because heparin binds simultaneously to fibrin and to the heparin-binding domain on thrombin, so-called exosite-2, thereby bridging thrombin onto fibrin.⁸⁸ The formation of a ternary heparin:thrombin:fibrin complex increases the affinity of thrombin for fibrin and induces conformational changes in the active site of thrombin. The resistance of thrombin within the ternary complex to inactivation by fluid-phase inhibitors may reflect allosteric modulation of its active site or spacial constraints that impair the reactivity of inhibitors with thrombin.^89,90 Many of the limitations of unfractionated heparin with respect to clot-bound thrombin also apply to low molecular weight heparins — regardless of Xa:IIa ratios.⁹¹ Continued on next page