How Short Is Safe? Duration of Dual Antiplatelet Therapy Before Aspirin Withdrawal After Percutaneous Coronary Intervention: A Dose-Response Network Meta-Analysis

Background: Trials of P2Y12 inhibitor monotherapy after percutaneous coronary intervention (PCI) withdrew aspirin at fixed times, mostly against 12-month dual antiplatelet therapy (DAPT), and categorical syntheses disagree. Methods: Bayesian model-based network meta-analysis of randomized trials of aspirin withdrawal with P2Y12 inhibitor continuation, with 12-month myocardial infarction (MI) and major bleeding modeled as functions of DAPT duration (Emax and spline). The prespecified estimand was the shortest duration at which the upper 95% credible limit of the odds ratio (OR) for MI versus 12-month DAPT fell below 1.15. Results: Sixteen trials (17 randomized comparisons, 58,068 patients) contributed 929 MIs and 1,102 major bleeding events. The OR for MI was 1.31 (95% credible interval 0.95–2.09) at 0 days, 1.17 (0.96–1.43) at 30 days, and 1.09 (0.98–1.25) at 90 days. Under every model the point estimate exceeded 1.15 below about 3 weeks and fell below it from about 90 days; a 30-day change-point did not fit better. The credible bound was met at 158 days (Emax) and 282 days (spline). Bleeding decreased with shorter duration (OR 0.40 at 0 days, 0.62 at 90 days). Certainty was low for MI below 90 days and the estimand, moderate for bleeding. Conclusions: Shorter DAPT reduced major bleeding, and the ischemic cost of aspirin withdrawal declined with duration; a step at 1 month was not supported. A 15% excess of MI could be excluded from about 5 months (Emax) or 9 months (spline). The curves match duration to each patient's ischemic margin.