Influence of Definition and Treatment Strategies

The reported incidence of major bleeding events in the ACS period varies significantly across clinical trials with best estimates between less than 1% and 14%. Numerous factors account for the discrepancy among trials and registries. Currently, no standard definition of bleeding severity exists. Second, bleeding rates have been directly correlated with the increased utilization of invasive revascularization procedures. Therefore, trials that employ a more conservative approach may have correspondingly lower rates of observed bleeding. Moreover, clinical trials of anticoagulant and antiplatelet therapies use various combinations of agents, which may also affect the measured rate of bleeding. ^, Recent trials, representing a more aggressive anti-ischemic treatment approach (reflected as concomitant use of multiple antiplatelet and antithrombin agents) may contribute to the higher reported incidence of bleeding during an ACS event. Finally, differences in an individual patient’s propensity to experience a bleeding event may also contribute to the observed variance in bleeding incidence during an ACS. ^,

The existence of multiple bleeding definitions accounts in part for the disparity of the reported incidence of bleeding complications among ACS patients. Historically, the two most commonly employed bleeding severity classification schemes are the TIMI and GUSTO scales. ^, The GUSTO scale categorizes bleeding as severe or life-threatening, moderate, mild, or none, and defines bleeding based on clinical outcomes such as hemodynamic compromise or intracranial hemorrhage. In contrast, the TIMI definition categorizes bleeding as major, minor, minimal, or none and is dependent on changes in laboratory parameters (hemoglobin or hematocrit) and not necessarily on clinically defined events (although intracranial hemorrhage is considered a TIMI major bleed). The use of different bleeding definitions across studies can influence the reported incidence of bleeding ( Table 29-1 ). Additional definitions of bleeding events adopt variations and combinations of the GUSTO and TIMI classification schemes, such as those developed in the SYNERGY, PURSUIT, and OASIS-5 trials. These trials uniquely defined bleeding events with combinations of both clinical and laboratory findings. ACS trial-specific definitions for bleeding complications have been developed, such as those in FRISC, ESSENCE, CURE, and ACUITY trials. Variations in transfusion thresholds, reductions in hemoglobin, and interruption of treatment regimens uniquely define bleeding events among these trials. For instance in the CURE trial, Yusuf and colleagues reported major bleeding as any life-threatening event resulting in hypotension or requiring more than 4 units of red blood cells (RBCs) and minor bleeding as any blood loss that led to the discontinuation of the study drug. The ACUITY trial investigators included access site bleeding requiring intervention, hematoma greater than 5 cm in diameter, and reoperation for treatment of bleeding in their trial-specific criteria for defining a major bleeding event. Recent trials assessing the bleeding risks associated with novel, antiplatelet agents, prasugrel and ticagrelor have been completed. In assessment of prasugrel, The TRITON TIMI 38 trial utilized the previous TIMI bleeding risk scheme with additional life-threatening bleeding-event classification that was defined as a TIMI major bleeding event that is fatal, leads to hypotension requiring treatment with intravenous inotropic agents, requires surgical intervention for ongoing bleeding, necessitates the transfusion of four or more units of blood (whole blood or packed RBCs) over a 48-hour period, or is a symptomatic intracranial hemorrhage (ICH) (see Table 29-1 ). Ticagrelor, a reversible inhibitor of the purinergic G protein–coupled P2Y 12 (P2Y12) receptor, was studied in the PLATO trial, which employed yet another ACS trial-specific bleeding classification scheme (see Table 29-1 ). The definition used in this trial was much more broad than in the TRITON TIMI 38 trial, and so led to higher reported rates of major bleeding. The incorporation of TIMI with GUSTO bleeding scales and existence of other trial-specific definitions of bleeding events precludes the ability to compare bleeding complication rates across ACS trials and to determine a single estimate of the incidence of bleeding complications among ACS patients.

Multiple treatment regimens have been assessed and approved for the management of ACS. Estimating the bleeding incidence is therefore confounded by the variability of pharmacologic therapies employed in the ACS literature. The addition of a novel, dual antiplatelet regimen (clopidogrel and aspirin) for the treatment of ACS in the CURE trial was associated with reductions in ischemic events (relative risk [RR] with clopidogrel as compared with placebo, 0.80; 95% confidence interval [CI], 0.72-0.90; P < .001), but at the expense of an increased risk for bleeding (3.7% in clopidogrel group vs. 2.7% in the placebo group; RR, 1.38; P = .001). The ACUITY trial highlighted the importance of adding antithrombotic therapy to antiplatelet therapy as a component of the ACS management algorithm. The administration of bivalirudin alone compared with heparin plus a glycoprotein (GP) IIb/IIIa inhibitor for the ACS patient in the ACUITY trial was associated with reductions in major bleeding events at 30 days (3.0% vs. 5.7%; RR, 0.53; 95% CI, 0.43-0.65; P < .001) and demonstrated the non-inferiority of bivalirudin to heparin plus a GP IIb/IIIa inhibitor in relation to ischemic outcomes.

Another major issue is the variation in the use of invasive risk stratification across ACS trials. For example, the CURE trial included patients who were managed medically, as well as those who underwent percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG). In contrast, the TRITON TIMI 38 trial was designed specifically as a trial that included patients undergoing PCI; very few patients underwent CABG or received medical therapy only. Similarly, the ACUITY trial included patients who underwent very rapid invasive risk stratification. The differences in the use of invasive procedures can have a large impact on bleeding risk and the subsequent reporting of bleeding events. Therefore, the applicability of the ACUITY, CURE and similar trials that have assessed the addition of novel, pharmacologic agents into the ACS management algorithm must be carefully considered.

Prognosis

Several studies describe an association between bleeding (regardless of definition) in the ACS population and adverse outcomes including death, stroke, myocardial infarction (MI) and unplanned revascularization (see Table 29-1 ). For example, Moscucci and colleagues examined the GRACE registry of 24,045 patients with ACS (including UA, NSTEMI, and ST-segment elevation myocardial infarction [STEMI]) and found an association between GRACE major bleeding and increased in-hospital mortality. Similarly, an analysis by Rao and colleagues, examined 26,452 ACS patients enrolled in the PURSUIT, PARAGON B, and GUSTO IIb trials and demonstrated a stepwise increase in risk between bleeding severity and 30-day and 6-month death. Congruent with those studies, recent studies by Eikelboom and colleagues, Manoukian and colleagues, and Segev and colleagues, describe a significant association between different definitions of major bleeding in ACS patients and adverse short-term and long-term outcomes including stroke and stent thrombosis.

Only one published study has directly compared the prognostic significance of different bleeding definitions. Rao and colleagues examined 15,898 patients from two clinical trials of ACS patients that used both the GUSTO and TIMI definitions of bleeding to classify the severity of bleeding complications. When applied separately, each scale identified patients as having had a bleeding event that was missed by the other scale. Additionally, both GUSTO and TIMI bleeding were associated with an increased risk for 30-day death or MI when examined separately. When both definitions were included in the same model, increasing GUSTO bleeding severity was associated with a stepwise increase in the adjusted hazard of death or MI, whereas TIMI bleeding did not correlate with prognosis. Based on these data, bleeding defined by clinical events is likely more important in terms of prognosis rather than bleeding defined solely on the basis of reductions in hemoglobin concentration.

One issue worth noting is the inclusion of blood transfusion as part of the bleeding definition in many studies. The administration of a transfusion is a discrete, documented clinical event that is easily captured. Data suggest that approximately 5% to 10% of ACS patients receive blood transfusions, and use of transfusions is higher in the United States than in other countries. ^, The “appropriate” use of blood transfusion administration is highly subjective due to lack of definitive data on transfusion triggers in patients with coronary artery disease ; therefore, transfusion may confound the bleeding–adverse outcomes relationship. Three large observational studies have found an association between transfusion and short-term mortality in patients with ACS. Yang and colleagues analyzed data from 74,271 patients with ACS and found a significant association between blood transfusion and in-hospital mortality. Wu and colleagues analyzed 78,974 elderly patients with acute MI and discovered that blood transfusion was associated with a significant increased risk of 30-day death among patients whose baseline hematocrit was greater than 33%. Rao and colleagues examined 24,111 NSTE (non-ST-elevation) ACS patients and found that blood transfusion was associated with a significantly higher risk of 30-day mortality if the nadir hematocrit was greater than 24%. The latter two studies highlight the controversy of defining a hematocrit threshold for administration of blood products in patients with ischemic heart disease. In practical terms, both studies underscore the concept that aggressive blood transfusion is associated with worse clinical outcomes. At issue is the inflection point at which the harm becomes manifest. Until further randomized data are available, it seems reasonable to transfuse ACS patients only if their hemoglobin is below 8 g/dL or their hematocrit is below 24% (in the absence of ongoing chest pain or active bleeding), per current American College of Cardiology/American Heart Association (ACC/AHA) and European Society of Cardiology (ESC) guideline recommendations. As antiquated practice guidelines supporting the liberal administration of blood products to ACS patients with a hematocrit of less than 24% may not be as benign as previously thought, it seems reasonable to conclude in addition, that routine use of transfusion to maintain arbitrary hemoglobin levels in asymptomatic patients should probably be avoided.