Introduction
The merits of antiplatelet therapy for patients undergoing percutaneous coronary intervention (PCI) are well established. The outcome benefits outweigh the inherent, increased bleeding risk, which is the most common, noncardiac adverse event of PCI. Although this increased risk usually manifests as easily bruising, when severe, it is associated with an increased propensity for morbidity and mortality. This chapter will focus on the bleeding risk that accompanies the antiplatelet therapy associated with PCI.
Scope of problem
The exact rate of bleeding associated with each antiplatelet agent is not completely clear. Reported bleeding rates vary greatly mainly due to heterogeneity in bleeding definitions and patient populations included. When different definitions were applied to evaluate the rates of “major bleeding” in a population of patients undergoing PCI in the context of non-ST-elevation myocardial infarction, the rate ranged from 1.2% to 8.2% [1]. Further, the ability to properly assess the bleeding rate of a specific agent, even when uniform definition is being used, is hampered by adjunctive therapies, such as other antithrombotic agents and concomitant anticoagulant usage.
A thorough discussion regarding the different bleeding definitions will follow this chapter. In brief, bleeding definitions commonly utilize both clinical events and laboratory parameters to define “major” and “minor” bleeding. Thrombolysis in Myocardial Infarction (TIMI) [2] and Global Utilization of Strategies To Open Occluded Arteries (GUSTO) [3] are two of the most commonly used definitions, although originated in the thrombolysis era. More recently, a professional consortium has suggested a standardized set of definitions to be used in clinical trials [4]. When these and other definitions were used in pivotal clinical trials [5, 6, 7, 8], the major bleeding rate ranged from 1.5% to 3.7%. Data from “real-life” registries have reported higher incidences of major bleeding (9.2–12.0%), which demonstrate the aforementioned heterogeneity [9, 10]. Increased bleeding risk translates into increased mortality risk [11, 12].
Bleeding predictors
Percutaneous coronary intervention-related major bleeding stems are the result of both patient and procedural parameters. Old age, female gender, impaired renal function, anemia, and cardiogenic shock at presentation are patient characteristics associated with major bleeding, while the use of intra-aortic balloon pump counterpulsation and concomitant use of antithrombotic and glycoprotein IIb/IIIa inhibitors (GPI) are two important procedural aspects associated with major bleeding [13, 14, 15]. Contemporary data from Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38), which compared the risk and benefits of two dual antiplatelet therapies in acute coronary syndrome (ACS) patients, demonstrated that within the entire patient cohort, history of stroke or transient ischemic event, age greater than 75 years, and body weight less than 60 kg were all associated with a higher risk for serious bleeding irrespective of the antiplatelet regimen used (HR 1.58, 95% CI 1.10–2.29, p = 0.01; HR 2.58, 95% CI 2.13–3.13, p < 0.001; and HR 2.30, 95% CI 1.74–3.05, p < 0.001, respectively). More specifically, within the subgroups of elderly patients and those with low body weight, prasugrel was associated with a significantly higher risk for bleeding compared with clopidogrel [16].
Femoral access bleeding is reported to be the most common bleeding site associated with PCI, accounting for greater than 50% of all major bleeding events with an incidence of approximately 2–8% [13, 14, 16, 17, 18, 19], although contemporary evidence has demonstrated a decline in the rates of femoral artery-related major bleeding in the last decade [19, 20]. Femoral access bleeding is followed by gastrointestinal, retroperitoneal, and other major organ bleeding. Femoral sheath size is one of major determinates of access site complications [19, 21]. For example, the odds ratio for major vascular bleeding when using a 7–8Fr sheath size compared with 6Fr is 1.53 (1.28–1.83, 95% CI) and is 1.91 (1.29–2.82, 95% CI) when a 9Fr sheath is compared with 6Fr [19]. Other factors associated with femoral access-related bleeding are procedure length and the intensity and duration of anticoagulation used. Gastrointestinal bleeding (GIB) complicates approximately 1.0% of PCI and is significantly associated with mortality at 30 days and at 1 year (HR 4.87 [2.61–9.08], p < 0.001, and HR 3.97 [2.64–5.99], p < 0.001, respectively) [22, 23]. Although the incidence of GIB has decreased throughout the years, GIB-related mortality rates have remained constant and are mostly associated with gastrointestinal malignancy [22].
Bleeding and outcome
The association between bleeding events and both short- and long-term mortality was established in numerous clinical trials and registries [11, 14, 16, 19, 24]. These and other studies have demonstrated a 3- to 10-fold increase in mortality risk in patients who bled versus those who did not. Bleeding severity was correlated with this risk [12, 24]. The ability to properly assess the effect of PCI-related bleeding on outcome is limited by the population selected to participate in the clinical trial and the overlap between bleeding risk and mortality risk.
Chhatriwalla et al. have collected data from greater than 3.3 million patients who underwent PCI between 2004 to 2011 and sought to evaluate the population’s adjusted mortality risk associated with bleeding after PCI. The population-adjusted risk for inhospital mortality was 12% (11.4–12.7%, 95% CI). By utilizing a propensity-matched cohort of patients, it was evident that the number needed to harm was 29 (28–31, 95% CI), with this being consistent across the spectrum of preprocedural patient risk profiles [11]. Beyond the intuitive association between bleeding and short-term mortality, the association of bleeding with long-term mortality was also broadly reported.
