Abstract 20496: Time Dependence of Association of Bleeding Events With Mortality in Patients Treated With Prasugrel or Clopidogrel: Insights From the TRITON-TIMI 38 Trial
Introduction: Recent data have suggested an impact of bleeding events with antithrombotic treatment on short- and long-term mortality. Therefore, we analyzed the impact and time-dependency of TIMI major or minor bleeding events on mortality in patients with acute coronary syndrome and planned early percutaneous coronary intervention enrolled in the TRITON-TIMI 38 trial that compared clopidogrel versus prasugrel.
Methods: We performed multivariable Cox models including all available baseline variables and a propensity score for bleeding in all patients treated with either clopidogrel or prasugrel. The effect of time since the bleeding event on mortality was assessed using multiple iterative landmark analyses starting from day 1 up to day 180 after the bleeding event and analyzing events until the end of follow up (e.g., day 1 until end of follow-up, day 2 until end of follow-up, etc.). The hazard ratio (bleeding vs. no bleeding) for death of each analysis was plotted against the landmark day. Bleeding was stratified by instrumented and spontaneous events.
Results: From the 13,421 patients included in the analysis, 209 incurred an instrumented and 325 a spontaneous TIMI major or minor bleeding event. The time-dependency of instrumented and spontaneous TIMI major or minor bleeding events is shown in the figure. Hazard ratios for instrumented and spontaneous bleeding did not statistically differ from 1.0 after 30 days.
Conclusions: Spontaneous bleeding events were associated with a higher hazard for death compared to instrumented bleeding events. TIMI major or minor bleeding events show an early high association with mortality, which levels out at a lower value after ∼ 30 days.
- Acute coronary syndromes
- Antiplatelet drugs
- Clinical trials
- Percutaneous coronary intervention
- © 2010 by American Heart Association, Inc.