Dual antiplatelet therapy (DAPT) with a potent P2Y12 inhibitor and aspirin remains the standard post-PCI (percutaneous coronary intervention) antithrombotic regimen in patients presenting with acute coronary syndrome (ACS). However, safety of withdrawing aspirin immediately after the procedure remains uncertain.
The NEO-MINDSET trial failed to demonstrate noninferiority for the composite ischemic endpoint at 30 days, and a prespecified subgroup analysis of STOPDAPT-3 exhibited a numerical excess risk for cardiovascular events. Thus, we performed a reconstructed patient-level data analysis of NEO-MINDSET and STOPDAPT-3 trials to assess the efficacy and safety of very early withdrawal of aspirin following PCI for ACS patients.
This study followed the Cochrane Handbook for Systematic Reviews of Interventions, with a protocol prospectively registered in PROSPERO ( CRD420251243246 ). We searched PubMed, Embase, and Cochrane up to December 2025 using the terms: “early,” “aspirin,” “percutaneous coronary intervention,” and “acute coronary syndrome.” We included studies if they were randomised clinical trials (RCTs) or subgroup analyses of RCTs; directly compared immediate aspirin withdrawal followed by potent P2Y12i monotherapy versus 1-month DAPT; reported outcomes in Kaplan–Meier (KM) curves.
Our search resulted in 1,358 results, of which one RCT and one prespecified RCT subgroup analysis were included (NEO-MINDSET and STOPDAPT-3), comprising 7,886 patients. Mean age was 65 years, and 3,954 patients were randomised to immediate potent P2Y12i monotherapy. Timing of randomization differed between trials: before PCI in STOPDAPT-3 and within 4 days after PCI in NEO-MINDSET. Notably, STOPDAPT-3, conducted exclusively in Japan, utilized prasugrel 3.75 mg once daily, the standard approved dose in that population. In contrast, NEO-MINDSET employed the conventional 5 or 10 mg prasugrel once daily and 90mg ticagrelor twice daily, doses used in Western populations.
Our composite coprimary endpoints were ischemic events at 30 days; defined as a composite of cardiovascular death, myocardial infarction, definite stent thrombosis, or ischemic stroke at 30 days in STOPDAPT-3 and as death from any cause, myocardial infarction, stroke, or urgent-vessel revascularization in NEO-MINDSET. The other coprimary endpoint was bleeding events at 30 days; defined as major bleeding or clinically relevant nonmajor bleeding (BARC 2,3,5) in NEO-MINDSET and as major bleeding (BARC 3 to 5) in STOPDAPT-3.
Individual patient data (IPD) were reconstructed from KM curves reported in the included trials. We extracted time and survival probability coordinates from the published KM curves using the “IPDfromKM” package in R. Hazard ratios were calculated using stratified Cox regression. The proportional hazards assumption was verified through the Schoenfeld residuals test. p Values <0.05 were considered statistically significant. All analyses were performed using R version 4.5.0. We assessed bias using the RoB-2 tool.
Immediate aspirin discontinuation followed by potent P2Y12i monotherapy was associated with a higher risk of ischemic events at 30 days (HR 1.40; 95% CI 1.11 to 1.78; p = 0.005; Figure 1 ), with an absolute risk increase of 1.12% and a number-needed-to-harm of 89. Moreover, potent P2Y12i monotherapy did not significantly reduce the risk of bleeding (HR 0.85; 95% CI 0.67 to 1.08; p = 0.19; Figure 1 ). The Schoenfeld test confirmed the proportional hazards assumption for both endpoints (p = 0.94 for the ischemic endpoint and p = 0.064 for bleeding). The risk of bias demonstrated both studies as having some concerns.
Kaplan–Meier curves for the co-primary endpoints of (A) ischemic events, and (B) bleeding events. ARI = absolute risk increase; NNH = number needed to harm.
In this reconstructed individual patient-level analysis of NEO-MINDSET and STOPDAPT-3, immediate aspirin discontinuation after PCI for ACS was associated with a significant increase in early ischemic events without a corresponding reduction in bleeding.
The early excess of ischemic events observed with P2Y12 inhibitor monotherapy is consistent with the mechanistic vulnerability of the first post-PCI weeks. In this context, the synergistic inhibition provided by DAPT may be necessary to mitigate the risk of acute thrombotic complications. Both trials individually reported numerically higher early stent thrombosis with monotherapy, supporting the biological plausibility of our findings.
The absence of a significant difference in bleeding at 30 days is consistent with the very low early bleeding rates observed in contemporary PCI trials. In this setting, the potential incremental benefit of removing aspirin immediately after PCI may be intrinsically limited, as the absolute bleeding risk in the first month is already low. Therefore, in contrast to studies evaluating aspirin withdrawal at later timepoints, the early peri‑PCI period appears less sensitive to modulation of bleeding risk by de-escalating antiplatelet therapy. Importantly, these findings should be interpreted in the context of the bleeding endpoint heterogeneity between the included trials. Whereas NEO-MINDSET captured BARC type 2, 3, and 5 events, STOPDAPT-3 was restricted to BARC type 3 to 5, thus excluding clinically relevant nonmajor bleeding. BARC type 2 events, although classified as less severe, can meaningfully affect patient quality of life and adherence to therapy. Therefore, the present analysis may not fully capture the real-world bleeding burden associated with continued DAPT, and the true benefit of aspirin withdrawal on bleeding may be underestimated by our pooled results.
Moreover, the absolute magnitude of the ischemic–bleeding tradeoff becomes crucial when interpreting these data. The 1.1% absolute increase in ischemic events with P2Y12 inhibitor monotherapy corresponds to a number-needed-to-harm of 89 at 30 days. Importantly, even small absolute effects acquire greater weight when extrapolated to real-world practice.
These findings, however, apply specifically to the immediate post-procedural period and should not be extrapolated to later de-escalation strategies. The recently published TARGET-FIRST trial demonstrated that among low-risk patients with acute myocardial infarction who underwent early complete revascularization and completed 1 month of uneventful DAPT, P2Y12 inhibitor monotherapy was noninferior to continued DAPT for ischemic events and significantly reduced clinically relevant bleeding. Taken together, these data suggest that while immediate post-PCI aspirin withdrawal increases ischemic risk without reducing bleeding, aspirin discontinuation at the 1-month timepoint in carefully selected, low-risk patients with complete revascularization may represent a safer window for antiplatelet de-escalation.
This analysis has some limitations, including variability in the ischemic and bleeding endpoints used for reconstructing pooled data, differences in P2Y12 inhibitor regimens between studies, and the use of reconstructed rather than original IPD.
Overall, these findings suggest that in ACS patients undergoing PCI, very early aspirin withdrawal does not reduce early bleeding and significantly increases ischemic events at 30 days. In the contemporary PCI era, aspirin appears to remain essential in the first weeks after ACS, where the ischemic hazard remains clinically meaningful.
CRediT authorship contribution statement
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
