Risk of Major Bleeding With Potent Antiplatelet Agents After an Acute Coronary Event: A Comparison of Ticagrelor and Clopidogrel in 5116 Consecutive Patients in Clinical Practice

Abstract

Abstract
Background
Major bleeding after acute coronary syndrome predicts a poor outcome but is challenging to define. The choice of antiplatelet influences bleeding risk.

Methods and Results
Major bleeding, subsequent myocardial infarction (MI), and all-cause mortality to 1 year were compared in consecutive patients with acute coronary syndrome treated with clopidogrel (n=2491 between 2011 and 2013) and ticagrelor (n=2625 between 2012 and 2015) in 5 English hospitals. Clinical outcomes were identified from national hospital episode statistics. Bleeding and MI events were independently adjudicated by 2 experienced clinicians, blinded to drug, sequence, and year. Bleeding events were categorized using Bleeding Academic Research Consortium 3 to 5 and PLATO (Platelet Inhibition and Patient Outcomes) criteria and MI by the Third Universal Definition. Multivariable regression analysis was used to adjust outcomes for case mix. The median age was 68 years and 34% were women. 39% underwent percutaneous coronary intervention and 13% coronary artery bypass graft surgery. Clinical outcome data were 100% complete for bleeding and 99.7% for MI. No statistically significant difference was seen in crude or adjusted major bleeding for ticagrelor compared with clopidogrel (Bleeding Academic Research Consortium 3–5, hazard ratio [HR], 1.23; 95% CI, 0.90–1.68; P=0.2, PLATO major adjusted HR, 1.30; 95% CI, 0.98–1.74; P=0.07) except in the non-coronary artery bypass graft cohort (n=4464), where bleeding was more frequent with ticagrelor (Bleeding Academic Research Consortium 3–5, adjusted HR, 1.58; 95% CI, 1.09–2.31; P=0.017; and PLATO major HR, 1.67; 95% CI, 1.18–2.37; P=0.004). There was no difference in crude or adjusted subsequent MI (adjusted HR, 1.20; 95% CI, 0.87–1.64; P=0.27). Crude mortality was higher in the clopidogrel group but not after adjustment, using either Cox proportional hazards or propensity matched population (HR, 0.90; 95% CI, 0.76–1.10; P=0.21) as was the case for stroke (HR, 0.82; 95% CI, 0.52–1.32; P=0.42).

Conclusions
This observational study indicates that the apparent benefit of ticagrelor demonstrated in a clinical trial population may not be observed in the broader population encountered in clinical practice.