Predictors of bleeding and time dependence of association of bleeding with mortality: insights from the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel--Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38).

The balance between benefit (ischemia protection) and risk (bleeding) is a key consideration in choosing the intensity of antiplatelet therapy for patients with acute coronary syndromes. The goals of this analysis were to identify baseline characteristics that independently predict bleeding and to determine how bleeding events impact the subsequent mortality in the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel--Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38). Multivariable Cox regression analyses adjusted for treatment, baseline, and procedural variables were used to determine the predictors for serious (TIMI major or minor) bleeding. To analyze the hazard ratio and time dependency of bleeding on mortality, we used iterative day-to-day landmark analyses after the bleed. From the 13 420 patients with acute coronary syndromes included in this analysis, 534 (4.0%) experienced a serious bleeding event. Variables with the highest strength of association with risk of serious bleeding were female sex, use of a glycoprotein IIb/IIIa inhibitor, duration of intervention, age, assignment to prasugrel, regional characteristics, admission diagnosis of ST-elevation myocardial infarction, femoral access for angiography, creatinine clearance, hypercholesterolemia, and arterial hypertension. Serious bleeding was associated with a significantly increased adjusted hazard ratio of 5.84 (95% confidence interval 4.11 to 8.29) for mortality. However, the hazard ratio did not differ statistically from baseline risk by 40 days after the bleeding event. The major predictors of serious bleeding were a combination of patient and procedural characteristics and antiplatelet therapies. Although serious bleeding was strongly associated with mortality within the first month of the bleeding event, this association was not significant beyond 40 days. http://www.clinicaltrial.gov. Unique identifier NCT00097591.