Increasing Benefit, Reducing Risk: Focusing on Hemorrhagic Complications in Percutaneous Coronary Intervention

Despite technologic and therapeutic advances, both ischemic and hemorrhagic complications remain the most common risks associated with percutaneous coronary interventions (PCI). Contemporary PCI trials typically focus on reducing ischemic events with secondary considerations given to the associated hemorrhagic complications. This standard paradigm equates the decreased ischemic event rates with a potentially increased risk of hemorrhagic complications. Re-evaluation of this paradigm may be warranted, however, as evidence suggests that reductions in both ischemic and hemorrhagic events are possible. Hemorrhagic complications remain an often under-appreciated complication of PCI. In randomized clinical trials, the incidence of minor post-PCI ischemic complications are generally within the range of 5–7% of patients, while major hemorrhage (a drop in hemoglobin >= 5 g/dl) varies widely from 0.4–16.6%. Actual bleeding event rates, however, are inherently related to the underlying risk profile of patient populations, accounting for the higher rates of TIMI major and minor bleeding observed in both RAPPORT and ADMIRAL, trials of antithrombotic therapy in primary PCI. Related to this issue is the applicability of these data within daily practice. Beyond the clinical trial population, evidence suggests that the incidence of hemorrhagic complications is likely to be even higher than that seen in clinical trials.¹ Despite the broad inclusion criteria of these trials, patients enrolled within studies remain a selected population. Specific clinical factors that may contribute to this bleeding risk differential between clinical trials and clinical practice include a greater prevalence of advanced age, renal impairment and extensive vascular disease. Continued efforts to reduce ischemic event rates coupled with reduced procedure-related hemorrhagic complications are vital to improving morbidity and mortality in patients undergoing PCI. This benefit should be demonstrated within clinical trials and daily clinical practice. Heparin: A drug whose time has come...and gone Unfractionated heparin (UFH) has served as the standard anticoagulant used during PCI for nearly 20 years. Unfortunately, the pharmacokinetics and pharmacodynamics of UFH, including the inability to inactivate fibrin-bound thrombin, heparin-mediated platelet activation and distribution characteristics, limit the clinical utility of the drug. Additionally, the lack of a clearly specified therapeutic PCI heparin dose contributes to inconsistencies in its catheterization lab administration, potentially predisposing patients to excess hemorrhagic complications. While performance of PCI without heparin anticoagulation substantially increases the risk of post-procedural ischemic complications,^2,3 heparin use does not appear to eliminate this risk (Table 1) and substantially contributes to the risk of procedure-related hemorrhagic complications. The activated clotting time (ACT) is clinically used to define the degree of heparin-induced anticoagulation achieved after heparin administration. Limitations with this technique exist, including variations in ACT values between differing ACT monitors, and inconsistent correlation between ACT values and clinical outcomes. For example, retrospective analyses of varying clinical databases have shown an inverse correlation between ACT levels and post-PCI ischemic complications, such as abrupt vessel closure⁴ and need for emergent coronary artery bypass graft surgery;⁵ however these same studies failed to demonstrate a correlation between elevated ACT levels and post-PCI hemorrhagic events. These limitations with UFH provide an opportunity to explore the clinical utility of novel anticoagulants, which may enhance both the clinical efficacy and safety profile of adjunct pharmacotherapy among a broad spectrum of patients undergoing PCI. Procedure-related hemorrhagic rates declining with better techniques, therapy and patient management Over the past 15 years, there has been a decline in the incidence of PCI-related bleeding complications (Table 1), partially attributed to improved procedural techniques and patient management practices in recent years. Some degree of caution is needed when comparing event rates between trials, since different inclusion/exclusion criteria and variations in bleeding definitions make absolute comparisons difficult. However, certain trends are evident. Prior to the introduction of glycoprotein (GP) IIb/IIIa receptor inhibitors, heparin doses prescribed to achieve an ACT of 300–350 seconds were recommended. The EPIC trial, which was the first large-scale randomized PCI trial to add abciximab to standard dose (10,000–12,000 U bolus) heparin, resulted in a reduced incidence of ischemic complications among patients undergoing high-risk coronary intervention but at the cost of a two-fold excess risk of post-procedural major bleeding and need for blood product transfusion.⁶ Increased post-PCI hemorrhagic complications were found to be associated with higher procedural heparin doses, longer duration of post-PCI heparin administration and low patient body weight.⁷ The PROLOG study suggested for the first time that a dissociation of clinical efficacy and hemorrhagic complications could be achieved by modifying procedural heparin dosing and altering post-PCI vascular access-site care. In PROLOG, the use of abciximab with low- or standard- (high) dose heparin together with a post-PCI strategy of early versus late vascular access-site sheath removal was examined.⁸ In multivariable modeling, sheath removal time and heparin dose were independent predictors of hemorrhagic complications. The combination of early sheath removal and use of low-dose heparin was of additive benefit in reducing post-PCI bleeding complications (Table 2). Multiple logistic regression analysis identified the variables in this study associated with a higher incidence of the composite bleeding endpoint as older age (p = 0.025), late sheath removal (p = 0.041), standard-dose heparin (p = 0.057) and female gender (p = 0.059). Bleeding endpoints were uniformly least favorable among patients randomized to standard-dose heparin (100 U/kg plus adjustments to achieve ACT ~ 300 seconds and late sheath removal at approximately hour 16) and most favorable for patients randomized to low-dose heparin (70 U/kg and not adjusted to achieve ACT) and early sheath removal (6 hours after procedure). This study also provided evidence that femoral vessel hemostasis in this low-shear setting may be less dependent upon platelet aggregation than on the formation of a fibrin-rich thrombus. The larger EPILOG study examined the efficacy and safety among 2,792 patients randomly assigned to receive either: 1) abciximab + low-dose (70 U/kg) heparin; 2) abciximab + standard dose (100 U/kg) heparin; or 3) standard dose (100 U/kg) heparin. Interestingly, while minor post-PCI hemorrhagic complications were significantly higher among patients receiving abciximab + standard-dose heparin as compared to the other two treatment groups (p 7 Knowledge of the incremental risk of hemorrhage associated with a specific patient and associated procedural characteristics may suggest strategies for reducing bleeding complications in PCI. Within the context of GP IIb/IIIa receptor inhibitors, results of a multivariable linear regression model using data from the EPIC trial defined several independent clinical and procedural variables that were associated with an increase risk of blood loss.⁷ Table 3 presents those variables with a p = 120 U/kg) (11.8% versus 3.3%) compared to placebo-treated (heparin) patients.⁷ A similar multivariate analysis of the Bivalirudin Angioplasty Trial (BAT) revealed similar clinical predictors for bleeding events (Table 4).¹⁰ Randomization to treatment with heparin, age and female gender all increased risk of bleeding by 2–3 fold. Interestingly, in this trial, treatment with the direct antithrombin bivalirudin was observed to be associated with a reduction in major hemorrhagic events compared to heparin- treated patients by approximately 60–65% in women (5.3% versus 15.3%; p = 65 years (5.2 versus 12.9%; p 11–13 In each of these groups, patients assigned to receive bivalirudin also demonstrated reductions in post-PCI ischemic events compared to heparin-treated patients. The predictors of procedure-related bleeding complications from these multivariable analyses closely match the findings of a recent analysis of predictors of post-PCI blood product transfusion (Table 5).¹⁴ Moscucci et al.¹⁴ demonstrated increased utilization of blood product transfusions among elderly individuals, female patients, those with elevated creatinine levels, acute myocardial infarction, post-procedural heparin use, and concomitant use of platelet GP IIb/IIIa receptor antagonists. Patient selection influences ischemic and hemorrhagic event rates Ischemic and hemorrhagic risks associated with PCI are tightly intertwined. Most PCI patients identified to be at increased risk for bleeding are also at increased risk for ischemic complications. Specific high-risk PCI patient subsets deserve special attention, given their high prevalence in clinical practice. Elderly patients. Elderly patients are at increased risk for complications due to the greater prevalence of poor left ventricular function, impaired renal function, increased lesion complexity and multivessel disease. Procedure-related bleeding is also more common in these patients, which is at least partially due to the presence of co-existing comorbidity such as peripheral vascular disease, cerebrovascular disease and systolic hypertension. A review of several clinical trials shows higher post-PCI ischemic event rates in elderly patient subgroups as compared with younger patients (Table 6). As previously noted, increasing patient age is also an independent predictor of excess post-PCI hemorrhagic complications.^15,20 Female gender and low body weight. The risk of death across all PCI trials is approximately double in women.^15,16 In EPIC, the rates of blood product transfusion were approximately double in women compared to men and in low body weight (6 The increased risk of post-PCI ischemic and hemorrhagic complications among women presenting with coronary artery disease has been repeatedly demonstrated and repeatedly challenged. The increased risk is partially explained by the increased prevalence of other risk factors such as age, diabetes, hypertension, poor left ventricular function and smaller body surface area. Nevertheless, as demonstrated in several multivariate analyses, female sex is an independent risk factor for increased bleeding risk.^10,14,17 Hemorrhagic event rates for this population have not been reported for the majority of PCI clinical trials. In the limited number of trials (EPIC and BAT) that do present data for these populations, event rates are elevated in women and low body weight patients. Acute coronary syndromes. Over half of the patients currently undergoing PCI present with an acute coronary syndrome.¹⁸ Patients presenting with acute myocardial infarction and cardiogenic shock constitute a special high-risk group, since these patients are at greater risk of ischemic and bleeding events.^7,19,20 The increased bleeding risk in this patient subset appears to be related to the greater use of fibrinolytics and antiplatelet agents, concomitant use of venous and arterial sheaths, and use of intra-aortic balloon pumps (IABP). Acute coronary syndromes are associated with platelet activation and aggregation.²¹ Acute ischemic events occur with increased frequency in patients with elevated states of platelet activation. In vivo studies report that therapeutic concentrations of unfractionated heparin are associated with a 2-fold increase in the number of activated platelets as reflected by a significant increase in the number of circulating p-selectin-positive (degranulated) platelets as well as monocyte- and neutrophil-mediated platelet aggregation.²² In vitro studies of platelet activation and aggregation with the thrombin-specific anticoagulant bivalirudin demonstrated complete inhibition of thrombin-mediated platelet activation and caused no independent stimulation of platelet aggregation.²³ It can be hypothesized that heparin activation of platelets leads to a paradoxical stimulation of thrombosis in some cases and contributes to therapeutic failure under certain conditions. Evidence lending some credence to this hypothesis is seen in the Bivalirudin Angioplasty Trial, where heparin-treated patients with more advanced degrees of coronary artery disease had higher rates of ischemic and hemorrhagic complications compared to patients without additional risk factors, while bivalirudin-treated patients provided more consistent protection against both ischemic and hemorrhagic complications in these patient populations (Figures 1A and 1B).^20,24 Renal failure. There has been interest in the impact of renal impairment on morbidity and mortality associated with PCI. Even subtle elevations in serum creatinine levels may be associated with marked reductions in creatinine clearance and substantial risk among PCI patients.²⁵ The metabolic derangements associated with renal disease are complex and produce inflammatory activation, increased thrombin generation and defects in platelet function.^26–28 Both ischemic and bleeding risks have been shown to substantially increase among these patients. Compounding these issues is the fact that the extent, severity and complexity of coronary artery disease are often greater in these patients. These factors all contribute to relatively poor outcomes following PCI among patients with renal disease. Evidence suggests that abnormal accumulations of heparin, low molecular weight heparin, and lepirudin over time may contribute to increased bleeding risks, especially in patients with renal insufficiency.^29–31 In addition, there is sparse evidence supporting the efficacy and safety of GP IIb/IIIa inhibitors use among patients with serum creatinine levels > 2.5 mg/dl, since the small molecule agents (i.e., tirofiban and eptifibatide) undergo substantial renal excretion and randomized trials have largely excluded these patients. In the BAT study, however, nearly 75% of the enrolled patients had some degree of renal impairment. No matter the degree of renal impairment, bivalirudin-treated patients experienced both fewer bleeding complications and fewer ischemic complications than patients treated with heparin (Figure 2).³² Need for safer anticoagulation Bleeding complications are associated with significant morbidity and mortality, as well as increased costs of care. In a review of hospital records for patients undergoing PCI from 1989–1993, deaths occurred in 121 patients (1%) and 2.5% of the deaths were attributable to bleeding complications.¹⁶ In another study, bleeding complications increased the patient length of stay to an average of 7 days, compared to an average of 3 days in an uncomplicated procedure.⁷ It has even been suggested that bleeding complications may indirectly lead to increased post-PCI ischemic complications. In a study of 4,098 unstable angina patients undergoing angioplasty, the incidence of death and myocardial infarction among patients with hemorrhage was 2.4% and 13%, respectively, versus 0.2% and 3% in patients with no major hemorrhage (p 33 Given the significant contribution of bleeding complications to increased morbidity and mortality, attention to factors that can reduce these risks is warranted. There is evidence to suggest that use of bivalirudin during PCI can reduce both post-PCI ischemic and hemorrhagic complications. In the BAT study, bivalirudin was found to reduce the incidence of hemorrhagic complications by 62% compared to heparin during hospitalization following PCI. This effect was consistent across all patient subgroups evaluated (Figure 3). There was also a statistically significant 22% reduction in the combined clinical endpoint of death, myocardial infarction or revascularization within 7 days in the overall patient population. Even more promising was a dramatic 51% reduction in ischemic events and a 73% reduction in major bleeding complications in the pre-specified, separately randomized subset of patients with post-infarction angina.²⁰ Thus, bivalirudin appears to be the first antithrombotic agent to decrease both ischemic and hemorrhagic complications during and after PCI. Unlike heparin, bivalirudin does not bind to plasma proteins other than thrombin and does not initiate platelet activation or aggregation. Bivalirudin does, however, inhibit both fibrin-bound and free circulating thrombin. The agent’s predictable and short plasma half-life of approximately 25 minutes may also contribute to its improved safety profile over unfractionated heparin,32,34 especially among patient subgroups clinically predisposed to experience higher rates of post-PCI hemorrhagic complications. Conclusion Over the past decade, the advancement of device technology and implementation of potent adjunct pharmacotherapies have resulted in a gradual reduction in post-procedural ischemic complications. However, there remains a continued need to focus on reducing the rates of post-PCI hemorrhagic complications, especially among high-risk PCI patient subgroups. Based on the observed reductions in both ischemic and hemorrhagic post-PCI event rates, bivalirudin appears to be an attractive alternative to UFH during PCI, especially among high-risk patient populations. Outcome results from ongoing PCI clinical trials will provide additional support for the clinical utility of bivalirudin use among a broad spectrum of patients undergoing PCI.