Clin Pharmacol Ther. 2026 Sep 21. doi: 10.1002/cpt.70484. Online ahead of print.
ABSTRACT
After short dual antiplatelet therapy, ticagrelor monotherapy may reduce bleeding while maintaining ischemic protection after percutaneous coronary intervention, but frequentist analyses do not directly estimate benefit-harm probabilities, limiting the quantitative interpretation of the ischemic-bleeding trade-off. In this post hoc Bayesian analysis of ULTIMATE-DAPT, we assessed the probabilistic risk-benefit profile of ticagrelor monotherapy using trial data with non-informative priors and four hierarchical informative priors derived from external randomized evidence. The likelihood and informative priors used published aggregate summary statistics (not individual participant data). Posterior probabilities for absolute risk reduction and relative risk were estimated across clinically relevant thresholds. Primary endpoints were the ischemic composite of all-cause death, myocardial infarction, or stroke and major bleeding; secondary endpoints included individual ischemic components, stent thrombosis, and any bleeding. Under non-informative priors, the probability of increased ischemic risk was 29.0%, whereas that of reduced bleeding exceeded 99.5%. Increased stent thrombosis risk was near neutral, while increased myocardial infarction risk was notable (87.0%); other ischemic endpoints were < 50.0%. Informative priors confirmed substantial bleeding benefit (probability of RR < 0.80, > 97.5%). In mixed cohorts including chronic coronary syndrome patients, probabilities of increased stent thrombosis rose to 75.8%-87.7% and myocardial infarction exceeded 91.0%, particularly with aspirin withdrawal at ≤ 1 month. These findings provide complementary quantitative, population-level evidence indicating that ticagrelor monotherapy after short DAPT was associated with reduced bleeding and a low posterior probability of increased overall ischemic risk, although potential ischemic risks remained with very early aspirin withdrawal and in mixed cohorts including chronic coronary syndrome patients.
PMID:42768782 | DOI:10.1002/cpt.70484