Bleeding Is Bad…. Isn’t It?
Advances in antithrombin and antiplatelet therapy have traditionally been associated with reductions in myocardial infarction and other thrombotic events in patients who have experienced an acute coronary syndrome (ACS), are undergoing a percutaneous coronary intervention (PCI), or are receiving long-term therapy for secondary prevention of stable vascular disease. Although individual trials of antithrombotic therapy have rarely been able to demonstrate that the reduction in ischemic events leads to a reduction in mortality (the most important end point), meta-analyses have often suggested that this is the case.1 And because myocardial infarction is an independent correlate of mortality, a link between a reduction in ischemic events and reduced mortality makes intuitive sense.
Article p 2793
Unfortunately, more potent anticoagulants and antiplatelet agents are also associated with an increased risk of bleeding, especially when used in combination with one another, as is usually the case. Furthermore, there is a well-established body of evidence that indicates an association between bleeding and ischemic events.2,3 It has been unclear, however, whether the link between bleeding and thrombosis is the result of bleeding after the initial development of a thrombotic complication (and its treatment) or if bleeding precedes the development of an ischemic complication and actually leads to (actually causes) thrombosis. In support of the former possibility is that patients who, because of the presence of thrombus, undergo a longer or more complicated procedure, receive a higher dose or longer duration of antithrombotic medication, or require an intra-aortic balloon pump are surely more prone to bleed. However, the possibility that hemorrhage actually leads to thrombosis rather than results from it is supported by the observation that hemorrhage is a potent stimulus for thrombosis; all patients would surely die if bleeding did not trigger at least a local thrombotic reaction. Furthermore, as the Global Registry of Acute Coronary Events (GRACE) investigators report in an analysis of 40 087 patients in this issue of Circulation, patients who experience a hemorrhagic complication often have their antithrombotic medications discontinued, which further increases the risk of thrombosis.4 Although it might seem easy to distinguish between these 2 very different explanations of the relationship between bleeding and death, most studies have not been able to do so. Also, historically, in patients with coronary disease and particularly those undergoing PCI, death usually resulted from thrombosis. Now, more frequently than ever before, it is recognized that death may result from hemorrhagic complications. In fact, fatal hemorrhage is probably a more common cause of death among patients undergoing PCI than is refractory coronary thrombosis that leads to fatal myocardial infarction.
Even though the relationship between bleeding and mortality has often been described as an independent one, many of the correlates of bleeding, as reported in the analysis of GRACE, are themselves associated with mortality. Older age, female sex, impaired renal function, and others are all associated with mortality, and no amount of statistical adjustment can provide certainty of a cause-and-effect relationship between correlates, particularly when a large number of covariates differ between the 2 groups of interest that are themselves associated with the outcome of interest (ie, death in patients who did and did not bleed).4
Perhaps as a result of the recognition of how difficult it is to separate whether bleeding leads to or results from thrombotic complications and of the clinical importance of bleeding, recent trials have begun to include bleeding as a component of their primary end point. Terminology such as “net clinical outcome” and “net adverse cardiac events (NACE)” has begun to enter the lexicon. However, composite end points are most appropriate when the individual components are of roughly equal importance, occur with approximately equal frequency, and correlate with one another. The “quadruple end point” of death, myocardial infarction, repeat revascularization, and bleeding satisfies few of these requirements—fewer, even, than the “triple end point” of death, myocardial infarction, and repeat revascularization. Because more potent anticoagulant and antiplatelet therapies typically reduce the frequency of thrombotic events but increase the frequency of hemorrhagic events, composite end points that include both thrombotic and hemorrhagic end points are problematic and have not been universally accepted.
The Organization for the Assessment of Strategies for Ischemic Syndromes (OASIS)-5 Trial, in which the Xa inhibitor fondaparinux was compared with enoxaparin in patients with non–ST-elevation ACS, is being touted as the strongest evidence that a reduction in bleeding may lead to a reduction in subsequent mortality.5 In OASIS-5, the frequencies of death, myocardial infarction, and refractory ischemia at 9 days (the prespecified primary end point) were virtually identical, but fondaparinux caused significantly less bleeding (an absolute 1.9% reduction) at 9 days. Remarkably, in the next 6 months, a difference in death (an absolute 0.7%) was detected between the enoxaparin and fondaparinux groups, in favor of fondaparinux. However, one is reminded of a somewhat analogous situation in which the addition of a glycoprotein IIb/IIIa inhibitor to heparin reduced procedural infarctions by ≈40% and was associated with a reduction in mortality not when the drugs were administered, but in late mortality, long after their discontinuation.1,6 Additional analyses, however, revealed that >80% of the patients who died in the placebo arms of these trials had not experienced a procedural infarction.6 This finding was a blow to the theory linking procedural infarction to death and its corollary, that a reduction in early procedural infarction leads to greater long-term survival rates; many people now believe that the difference in late death was most likely the result of chance. The causes of death among patients assigned to enoxaparin in OASIS-5 have not yet been reported, and it has not yet been reported whether late deaths were related in any way, directly or indirectly, to the bleeding that occurred earlier. Preliminary analysis of the Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction (HORIZONS) trial, in which bivalirudin reduced major bleeding in patients with ST-elevation myocardial infarction undergoing primary PCI and, to the surprise of many, was also associated with a significantly lower mortality at 30 days, may also support a link between bleeding and a reduction in mortality; further analysis and longer-term follow-up are needed.6a
If procedural infarction is associated with mortality, however, and bleeding is associated with mortality, why would it not be easy to demonstrate that drugs that reduce these events also reduce mortality? There are many possible explanations. One is that drugs that are potent inhibitors of platelet aggregation, such as glycoprotein IIb/IIIa inhibitors, reduce procedural infarction but also increase bleeding, so the beneficial effects from a reduction in infarction may be partially offset by the increase in bleeding. This was also just seen in the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombosis in Myocardial Infarction 38 (TRITON-TIMI 38), in which the novel, more potent P2Y12 inhibitor prasugrel reduced infarction significantly more than clopidogrel but, as a result of its greater potency, was also associated with a significant increase in major and life-threatening bleeding.6b Similarly, drugs that reduce bleeding, such as the direct thrombin inhibitor bivalirudin, may be associated with a trend toward more frequent procedural infarction, particularly in higher-risk, troponin-positive patients; this, if true, might in part offset the expected improvement in clinical outcome that results from a reduction in bleeding.7
One of the principle findings in the analysis of GRACE highlights an important concept to remember whenever registry outcomes are analyzed. Spencer and colleagues4 report that many of the correlates of bleeding in GRACE are themselves independently associated with mortality. No antithrombotic therapy can be expected to reduce the mortality that results from nonmodifiable risk factors associated with bleeding. Again, a parallel exists between the many PCI studies that have shown that older patients and those with diabetes, vein grafts, a greater number of lesions, and longer lesions, for example, have an increased frequency of procedural infarction. No antithrombotic agent can reduce the mortality that results from these nonmodifiable characteristics associated with procedural infarction.
There are many remaining issues, however, about which we have much to learn in order to better understand the relationship between bleeding and outcome. Many different definitions of bleeding have been used; not all are appropriate for both antiplatelet and anticoagulant agents, or for both chronic and acute clinical situations, or for both ACS and PCI.8,9 The definition of bleeding used in GRACE is yet another relatively unique definition; however, reported rates of bleeding with this definition are similar to those reported in other recent studies.10 Bleeding in GRACE was associated with in-hospital, although not longer-term, mortality.4 Other studies have suggested a relationship between bleeding and longer term (1-year) mortality.3,11 The GRACE analysis avoids a pitfall that many other analyses have not: A transfusion without a documented bleed was not sufficient for moderate or severe bleeding to be considered to have occurred.4 Many other analyses consider a transfusion to represent a moderately severe or major bleed, even without documented bleeding. However, we know that many ACS patients (several absolute percent) receive a transfusion because of anemia or procedural blood loss without a documented hemorrhagic complication. Indeed, baseline anemia is a powerful correlate, among the most powerful correlates, of transfusion.12 When long-term analyses indicate a great mortality in such patients, it may not necessarily be because they bled, even though they are “defined” as having bled.8 Does the mortality result from more complex coronary disease that leads to a more difficult and complicated procedure, with greater blood loss? Or underlying comorbid illnesses that led to the anemia? Or is transfusion itself deleterious? The answer to all of these questions is unknown but under active investigation. Regardless, the GRACE analysis provides further insight into the frequency, correlates, and clinical impact of bleeding on adverse outcomes in >40 000 patients with ST-elevation and non–ST-elevation myocardial infarction.4 It is now clearer than ever that mortality may result from either thrombosis or hemorrhage. Although antithrombotic medications that cause less bleeding may not reduce mortality as much as we would like or expect, bleeding is dangerous, uncomfortable to the patient, associated with much nonfatal morbidity, and expensive, and it is to be avoided whenever possible.13 The GRACE analysis reinforces just how much we have to learn about the most appropriate ways to characterize, prevent, and treat bleeding during the administration of antithrombotic therapy to optimize outcomes.
Dr Berger has spoken at CME-approved scientific symposia supported by BMS, Sanofi-Aventis, the Medicines Company, AstraZeneca, Medtronic, Schering-Plough, and Lilly/Daiichi Sankyo. He has served as a consultant to PlaCor, Lilly/Daiichi Sankyo, Molecular Insight Pharmaceuticals, and CV Therapeutics. He owns equity in Lumen, Inc. Dr Manoukian is on the speakers’ bureau for the Medicines Company, has served as a consultant for Guerbet, Sanofi-Aventis, Schering-Plough, and the Medicines Company, and has received research support from Guerbet.
The opinions expressed in this editorial are not necessarily those of the editors or of the American Heart Association.
Manoukian SV, Feit F, Mehran R, Voeltz MD, Ebrahimi R, Hamon M, Dangas GD, Lincoff AM, White HD, Moses JW, King SB III, Ohman EM, Stone GW. Impact of major bleeding on 30-day mortality and clinical outcomes in patients with acute coronary syndromes: an analysis from the ACUITY trial. J Am Coll Cardiol. 2007; 49: 1362–1368.
Spencer FA, Moscucci M, Granger CB, Gore JM, Goldberg RJ, Steg PG, Goodman SG, Budaj A, FitzGerald G, Fox KAA. Does comorbidity account for the excess mortality in patients with major bleeding in acute myocardial infarction? Circulation. 2007; 116: 2793–2801.
Fifth Organization to Assess Strategies in Acute Ischemic Syndromes Investigators; Yusuf S, Mehta SR, Chrolavicius S, Afzal R, Pogue J, Granger CB, Budaj A, Peters RJ, Bassand JP, Wallentin L, Joyner C, Fox KA. Comparison of fondaparinux and enoxaparin in acute coronary syndromes. N Engl J Med. 2006; 354: 1464–1476.
Stone GW, for the HORIZONS Investigators. HORIZONS AMI: a prospective, randomized comparison of bivalirudin vs heparin plus glycoprotein IIb/IIIa inhibitors during primary angioplasty in acute myocardial infarction—30 day results. Presented at: TCT 2007; October 24, 2007; Washington, DC.
Wiviott SD, Braunwald E, McCabe CH, Montalescot G, Ruzyllo W, Gottlieb S, Neumann FJ, Ardissino D, De Servi S, Murphy SA, Riesmeyer J, Weerakkody G, Gibson CM, Antman EM, for the TRITON-TIMI 38 Investigators. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2007; 357: 2001–2015.
Stone GW, McLaurin BT, Cox DA, Bertrand ME, Lincoff AM, Moses JW, White HD, Pocock SJ, Ware JH, Feit F, Colombo A, Aylward PE, Cequier AR, Darius H, Desmet W, Ebrahimi R, Hamon M, Rasmussen LH, Rupprecht HJ, Hoekstra J, Mehran R, Ohman EM; ACUITY Investigators. Bivalirudin for patients with acute coronary syndromes. N Engl J Med. 2006; 355: 2203–2216.
Steinhubl SR, Kastrati A, Berger PB. Variation in the definitions of bleeding in clinical trials of patients with acute coronary syndromes and undergoing percutaneous coronary intervention and its impact on the apparent safety of antithrombotic drugs. Am Heart J. 2007; 154: 3–11.
Rao SV, O’Grady K, Pieper KS, Granger CB, Newby LK, Mahaffey KW, Moliterno DJ, Lincoff AM, Armstrong PW, Van de Werf F, Califf RM, Harrington RA. A comparison of the clinical impact of bleeding measured by two different classifications among patients with acute coronary syndromes. J Am Coll Cardiol. 2006; 47: 809–816.
Moscucci M, Fox KA, Cannon CP, Klein W, Lopez-Sendon J, Montalescot G, White K, Goldberg RJ. Predictors of major bleeding in acute coronary syndromes: the Global Registry of Acute Coronary Events (GRACE). Eur Heart J. 2003; 24: 1815–1823.
Ndrepepa G, Berger PB, Mehilli J, Seyfarth M, Neumann F-J, Schömig A. Periprocedural bleeding and 1-year outcome after percutaneous coronary interventions: appropriateness of including bleeding as a component of a quadruple end point. J Am Coll Cardiol. In press.