‘Aspirin versus ticagrelor for the management after coronary revascularization - a systematic review and meta‑analysis of randomised trials’

Abstract

Objective

To compare the efficacy and safety of ticagrelor monotherapy versus aspirin monotherapy in patients with coronary artery disease undergoing coronary revascularization.

Methods

We conducted a systematic review and meta-analysis of RCTs comparing ticagrelor monotherapy with aspirin monotherapy after PCI or CABG. MEDLINE, Embase, Cochrane Central, Scopus, ClinicalTrials.gov and Google Scholar were searched up to 19th January 2026. The primary outcome was major adverse cardiovascular events (MACE), a composite of all-cause mortality, myocardial infarction and stroke. Secondary outcomes included all-cause mortality, myocardial infarction, stroke, major bleeding, repeat revascularization and stent thrombosis. Risk ratios (RRs) with 95% CIs were pooled using a random-effects model with restricted maximum likelihood estimation and Knapp-Hartung adjustment.

Results

Five randomised trials involving 25,994 participants were included, of whom 12,998 received ticagrelor monotherapy and 12,996 received aspirin monotherapy. Ticagrelor monotherapy was associated with a significantly lower risk of MACE than aspirin monotherapy (RR 0.86, 95% CI 0.78 to 0.95; p = 0.012; I² = 0%). All-cause mortality was also reduced with ticagrelor (RR 0.86, 95% CI 0.77 to 0.97; p = 0.023; I² = 0%). No significant differences were observed for myocardial infarction (RR 0.87, 95% CI 0.70 to 1.07; p = 0.138; I² = 0%), stroke (RR 1.01, 95% CI 0.85 to 1.19; p = 0.913; I²=0%), major bleeding (RR 1.00, 95% CI 0.84 to 1.20; p = 0.976; I² = 0%), repeat revascularization (RR 0.89, 95% CI 0.58 to 1.37; p = 0.452; I² = 45.8%) or stent thrombosis (RR 0.88, 95% CI 0.20 to 3.90; p = 0.481; I² = 0%).

Conclusion

Ticagrelor monotherapy was associated with a potential reduction in MACE and all-cause mortality compared with aspirin monotherapy after coronary revascularisation, without increasing major bleeding. However, these findings were driven primarily by PCI trials, particularly the GLOBAL LEADERS and GLASSY trial program, and should be interpreted cautiously because PCI and CABG populations were analyzed together.

Trial registratrion

CRD420261286239

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12872-026-06186-w.

Introduction

Coronary artery disease is the leading cause of morbidity and mortality worldwide [1]. For the management of chronic coronary artery disease, optimal medical therapy (OMT) is used consisting of antithrombotics, statins, angiotensin-converting enzyme inhibitors, beta-blockers, short-acting nitrates, and lifestyle modification alongside revascularization when indicated through percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) [2, 3]. PCI is performed when the patient has lifestyle-limiting angina despite OMT or significant stenosis (≥ 70%) in a major epicardial vessel. The choice between PCI and CABG depends mainly on the coronary anatomy, surgical risk, comorbidities of the patient and the cardiac function [3].

Antiplatelet therapy is required for secondary prevention after coronary revascularization but the regimen depends upon the revascularization type. After PCI, dual antiplatelet therapy (DAPT) is recommended which consists of aspirin with P2Y12 inhibitor followed by monotherapy later. In contrast, after CABG, aspirin monotherapy is recommended early after surgery and it is continued indefinitely [3]. Aspirin and P2Y12 inhibitors act through complementary pathways: aspirin irreversibly inhibits cyclooxygenase-1, reducing thromboxane A2-mediated platelet activation, while P2Y12 inhibitors (clopidogrel and ticagrelor) block ADP-mediated activation [4]. Ticagrelor is a direct, reversible P2Y12 inhibitor that provides faster and more consistent platelet inhibition compared to clopidogrel, an irreversible thienopyridine prodrug requiring variable CYP-mediated activation [5, 6].

DAPT reduces the risk of early and late thrombotic events among patients with acute coronary syndrome (ACS). Prolonged DAPT may increase bleeding risk through simultaneous inhibition of complementary platelet pathways, while premature discontinuation may increase ischemic events in selected patients [7]. After an appropriate DAPT period, transition to single antiplatelet therapy is a common long-term secondary prevention strategy, although the optimal monotherapy agent and timing of aspirin withdrawal remain areas of active investigation [3]. Current evidence suggests that aspirin discontinuation and ticagrelor monotherapy reduce the risk of bleeding [8]. Clopidogrel can also be used instead of ticagrelor and has comparable protective effect with similar incidence of MACE [9].

Recent evidence further underscores why this question remains clinically relevant in contemporary practice. A 2026 review of high-bleeding-risk PCI emphasized that abbreviated DAPT followed by single antiplatelet therapy can be used with bleeding-sparing procedural strategies, but that the optimal timing of PCI, DAPT duration, and monotherapy selection remains uncertain in high-risk settings [10]. In parallel, newer secondary-prevention meta-analyses have challenged uniform long-term use of several non-antiplatelet therapies in special populations, including statins in dialysis-dependent patients with established ASCVD and beta-blockers after MI without reduced ejection fraction [11, 12]. Together, these data illustrate a broader shift toward individualized, risk-adapted post-revascularization care rather than fixed lifelong treatment pathways, making the comparative role of aspirin versus potent P2Y12 inhibitor monotherapy especially important.

Previous systematic reviews and meta-analyses have evaluated aspirin monotherapy, clopidogrel monotherapy, and broader P2Y12 inhibitor monotherapy strategies after coronary revascularization. However, direct randomized evidence comparing ticagrelor monotherapy with aspirin monotherapy remains limited across both PCI and CABG populations. Although patients undergoing PCI or CABG require antiplatelet therapy after revascularization, these populations are not biologically interchangeable. Post-PCI ischemic risk is largely related to stent thrombosis and recurrent coronary events, whereas post-CABG risk is influenced by graft failure, graft patency, and progression of native coronary disease. Because the number of direct ticagrelor versus aspirin trials is small, we pooled PCI and CABG trials to summarize the totality of randomized evidence; however, the pooled findings should be interpreted as hypothesis-generating and requiring cautious interpretation. This systematic review and meta-analysis therefore evaluate the efficacy and safety of ticagrelor monotherapy versus aspirin monotherapy after coronary revascularization while explicitly considering differences in revascularization type, monotherapy timing, and trial dominance.

Materials and methods

Inclusion criteria and exclusion criteria

The inclusion criteria were as follows:

• Population: Adults (≥18 years) with coronary artery disease who underwent PCI or CABG.

• Intervention: Ticagrelor (any formulation/dose; oral) as a primary agent for prevention of MACE

• Comparator: Aspirin (any formulation/dose; oral) as a primary agent for prevention of MACE

• Outcomes:

• Primary outcome

• MACE (All-cause mortality, MI, Stroke)

• Secondary outcomes

• All-cause mortality

• Myocardial Infarction

• Stroke

• Major bleeding (BARC ≥3 or TIMI criteria)

• Stent thrombosis

• Repeat revascularization

• Study design: Randomised controlled trials (RCTs)

The exclusion criteria were as follows:

• Duplicated publication

• Case reports, case series without comparator, editorials, and reviews

• Incomplete or wrong data

• Studies which compared aspirin monotherapy with DAPT in patients with CAD

• Studies with insufficient data or unavailable full texts despite reasonable effort

• Studies did not report the relevant adverse clinical outcomes

Design

This systematic review and meta-analysis followed the principles of the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) [11]. The PRISMA diagram detailing the selection process is shown in Fig. 1.

Fig. 1.

PRISMA flow diagram illustrating the study identification, screening, eligibility, and inclusion process

Protocol registration and deviations

The protocol was prospectively registered on PROSPERO (CRD420261286239) on 24th of January 2026. The full protocol is provided in Supplementary Appendix S2. The registered protocol title was “Aspirin vs Ticagrelor for the Management of Coronary Heart Disease - A Systematic Review and Meta-analysis”. The manuscript title was revised to “Aspirin versus Ticagrelor for the Management After Coronary Revascularization - A Systematic Review and Meta-analysis of Randomised Trials” for greater precision and clarity, as the review specifically focuses on patients who underwent percutaneous coronary intervention or coronary artery bypass grafting. The following additional deviations from the registered protocol occurred:

• Only randomised controlled trials were included (protocol planned RCTs and non-randomised studies with ROBINS-I). No eligible non-randomised studies were identified.

• Primary outcome defined as MACE (all-cause mortality, MI, stroke) instead of MACCE (protocol used cardiovascular/cerebrovascular death).

• Major bleeding defined as BARC ≥ 3 or TIMI (protocol planned BARC ≥ 2 or TIMI).

• Planned subgroups (PCI vs. CABG, ACS vs. stable, timing, dosing, stent type) and fixed-effect sensitivity analyses were not performed due to only five trials and incomplete reporting; leave-one-out sensitivity analysis was conducted instead.

• Search was expanded to include Embase, Scopus, and Google Scholar (protocol listed only CENTRAL, MEDLINE, ClinicalTrials.gov).

• An additional author (NG) contributed to manuscript drafting after registration.

These changes were made for feasibility and did not affect the primary research question or eligibility criteria. The PROSPERO record will be updated with this amendment upon manuscript acceptance.

Search strategy

MEDLINE (via PubMed), Embase, Cochrane Central, Scopus, ClinicalTrials.gov and Google Scholar databases were searched up to 19th January, 2026. Additional studies were identified through manual searching of reference lists and other supplementary.

Cross-references from the published articles were manually searched to retrieve additional literature.

To create an extensive search strategy that encompassed all fields in the records as well as Medical Subject Headings (MeSH words) for broadening the search in an advanced PubMed search, the predefined phrases were identified. For PubMed, MeSH terms and keywords combined with Boolean operators (AND, OR) were used for the systematic identification of records. For Google Scholar, the keywords “Aspirin”, “Ticagrelor” and “coronary artery disease” were used. The preliminary search strategy is described in the Supplementary Appendix S1.

Selection of studies

Two reviewers, SR and SU, rigorously and independently screened titles and abstracts for eligibility. The search result was uploaded to the reference management software Zotero. Deduplication of the records was done automatically and then manually. Full texts of potentially eligible studies were retrieved and assessed in duplicate. Any disagreement was resolved by consulting with a third author DRR whenever necessary. A PRISMA flow diagram detailing the study identification and selection process is shown in Fig. 1.

Data collection process

The data extraction was done independently by the two authors (SR, SU) using a standardized data extraction form developed in Google Sheets. Any disagreements were settled by communication with a third author (DRR).

For this meta-analysis, data were retrieved from the selected articles using a standardized data extraction form (see Supplementary Table S1 for the complete list of extracted variables).

All reported results compatible with each outcome domain, including different measurement scales, time points, and analyses, were extracted from the included studies. No data conversions or imputations were required, as the necessary statistics were directly available from the included studies. Studies that did not report the number of events for a particular outcome were excluded from the meta-analysis for that outcome.

For multi-arm trials, only study arms relevant to the predefined comparison (ticagrelor monotherapy versus aspirin monotherapy) were included in the analysis. In the DACAB trial, which included ticagrelor monotherapy, aspirin monotherapy, and ticagrelor–aspirin dual therapy arms, only the ticagrelor monotherapy and aspirin monotherapy groups were included, while the dual-therapy arm was excluded.

Primary Outcome:

• MACE (Major Adverse Cardiovascular Events)

• MACE was considered a composite of all-cause mortality, myocardial infarction, and stroke.

Secondary Outcomes

• Major bleeding (BARC ≥3 or TIMI criteria)

• Major bleeding was defined as bleeding events meeting Bleeding Academic Research Consortium (BARC) grade ≥3 or Thrombolysis in Myocardial Infarction (TIMI) major bleeding criteria, as reported in the included studies.

• All-cause mortality

• Myocardial infarction

• Stroke

• Stent thrombosis

• Repeat revascularization

Risk of bias assessment

RoB 2.0 [12] is a revised tool for assessing the risk of bias in randomised trials that addresses five specific domains: (1) bias arising from the randomization process; (2) bias due to deviations from intended interventions; (3) bias due to missing outcome data; (4) bias in outcome measurement; and (5) bias in reported result selection. Two authors (SR and SU) independently assessed the quality of each included study using the RoB 2.0 and documented the rationale for their judgments in each domain as low risk, high risk, or some concerns. Any differences in risk of bias assessments or explanations for assessments were handled by communication with a third author, DRR. Following guidance given for RoB 2.0, we derived an overall summary of Risk of Bias judgment (low; some concerns; high) for each specific outcome, whereby the overall RoB for each study was determined by the highest RoB level in any of the domains that were assessed.

Data synthesis and statistical analysis

Data on the number of events and sample sizes in the intervention and comparator groups were extracted. Risk ratios (RRs) with 95% confidence intervals (CIs) were calculated as the effect measure for each outcome using the raw event counts and group sizes. A random-effects model was used to pool the effect sizes because significant between-study heterogeneity was anticipated. The restricted maximum likelihood (REML) estimator was used to estimate the between-study variance (𝜏2). Knapp–Hartung adjustments were applied to obtain the confidence intervals around the pooled estimates [13]. Because fewer than 10 studies were included, no subgroup analyses or meta-regression were performed to explore sources of heterogeneity. Study characteristics (intervention and comparator details, population characteristics, and outcomes) were tabulated and compared with the pre-specified synthesis groups in the protocol to determine eligibility for quantitative synthesis. Pooled effect estimates, measures of between-study heterogeneity, and p-values were presented in forest plots. A two-sided p-value < 0.05 was considered statistically significant. All analyses were performed using R software (version 4.5.2), primarily with the meta and metafor packages and metabin function. Continuity correction was not required for the primary analyses because no included comparison had zero events in both treatment arms for the analyzed outcomes. For sparse outcomes, particularly stent thrombosis, pooled estimates were interpreted cautiously because of the limited number of contributing studies and events.

Reporting bias assessment

Risk of bias due to missing results (publication bias) was assessed by visual inspection of funnel plots for each outcome. Formal statistical testing of funnel plot asymmetry was not performed because fewer than 10 studies were included, in accordance with Cochrane and PRISMA recommendations [11, 14].

Certainty of evidence

The certainty of evidence for each outcome was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach [15].

Randomised controlled trials were initially considered high-certainty evidence and were downgraded, where appropriate, for risk of bias, inconsistency, indirectness, imprecision, and publication bias. Risk-of-bias judgments were informed by the RoB 2.0 assessments performed for the included randomised trials. The absolute risk for each outcome was estimated using the event rate in the aspirin group as the baseline risk, and the corresponding risk in the ticagrelor group was calculated by applying the pooled risk ratio. A Summary of Findings table was prepared using GRADEpro GDT to present the pooled relative effects, absolute risks, and overall certainty of evidence for the primary and secondary outcomes.

Results

Identification and selection of the study

Electronic search of literature resulted in a total number of 3098 articles. After careful assessment of the abstract and application of the pre-specified eligibility criteria, 3022 articles were excluded. A total of 76 articles were read in full, and finally, five trials were included for meta-analysis. Figure 1 reveals the detailed search process.

Basic characteristics of the included studies

Five trials involved a total of 25,994 participants that compared the benefits and risks of aspirin versus ticagrelor in patients with coronary artery disease who underwent percutaneous coronary intervention or coronary artery bypass grafting. Out of 25,994 participants, 12,996 participants received aspirin monotherapy and 12,998 participants received ticagrelor monotherapy. Table 1 lists the basic features of the studies which were included in this meta-analysis.

Table 1.

Characteristics of the included studies

Study	Year	Revascularization	Design	Total (N)	Aspirin	Ticagrelor
GLOBAL LEADERS [16]	2018	PCI	RCT	15,968	7,988	7,980
TiCAB [17]	2019	CABG	RCT	1,859	928	931
DACAB [18]	2018	CABG	RCT	332*	166	166
GLASSY [8]	2019	PCI	RCT	7,585	3,791	3,794
TARGET [19]	2021	CABG	RCT	250	123	127

*Only the ticagrelor monotherapy and aspirin monotherapy arms were included from the DACAB trial; the dual-therapy arm was excluded as it was not relevant to the comparison (total trial n = 500)

RCT Randomised Controlled Trial, PCI Percutaneous Coronary Intervention, CABG Coronary Artery Bypass Graft

Baseline characteristics of the participants

Table 2 lists the baseline characteristics of the participants in each study. The weighted mean age of the participants was 64.8 ± 10.3 years. The overall male-to-female ratio was 3.4:1, and comorbidities such as diabetes mellitus, hypertension, dyslipidemia, and smoking were prevalent in the population.

Table 2.

Baseline characteristics of participants

Study	Age (years)*	Male (%)	Diabetes (%)	Hypertension (%)	Dyslipidemia (%)	Smoking (%)
GLOBAL LEADERS [16]	64.6 / 64.5	76.9 / 76.6	24.9 / 25.7	73.3 / 74.0	70.0 / 69.3	26.3 / 25.9
TiCAB [17]	67.0 / 66.4	84.6 / 85.3	35.6 / 36.3	90.1 / 89.8	81.3 / 82.2	20.2 / 21.5
DACAB [18]	64.0 / 63.3	84.9 / 80.7	40.4 / 45.2	72.3 / 73.5	71.7 / 74.7	52.4 / 44.6
GLASSY [8]	64.8 / 64.9	76.5 / 76.0	23.7 / 24.3	72.3 / 72.5	65.3 / 63.3	29.1 / 28.6
TARGET [19]	68.0 / 67.5	87.0 / 81.9	43.9 / 41.7	79.7 / 80.3	89.4 / 91.3	21.1 / 18.9

*Values shown as Aspirin / Ticagrelor

Risk of bias assessment

Figure 2 shows the risk of bias assessment. Three of the five studies had a low risk, while two of them showed some concerns. GLOBAL LEADERS was judged to have some concerns for both MACE and bleeding because the trial was open-label and secondary outcomes in the studies were investigator reported without blinded central adjudication.

Fig. 2.

Risk of bias assessment

The DACAB trial was also judged to have some concerns as it was an open-label trial, which introduces potential bias due to deviations from intended interventions despite high participant retention. The GLASSY trial even though open-label was considered to have low risk as it had incorporated blinded central adjudication of clinical outcomes, reducing the risk of measurement bias. The TiCAB and TARGET trials demonstrated low risk of bias across domains, with adequate randomization, minimal missing outcome data, and objective outcome assessment. A detailed risk-of-bias assessment is presented in the Supplementary Appendix S3.

Main results of the analysis

The duration of follow-up across the included trials ranged between one and two years. The pooled effect estimates for the primary endpoint (MACE) and all secondary outcomes are depicted in the forest plots (Figs. 3, 4, 5, 6, 7, 8 and 9), while a comprehensive summary of the overall results is presented in Table 3.

Fig. 3.

Pooled analysis of MACE

Fig. 4.

Pooled analysis of mortality

Fig. 5.

Pooled analysis of myocardial infarction

Fig. 6.

Pooled analysis of stroke

Fig. 7.

Pooled analysis of major bleeding

Fig. 8.

Pooled analysis of repeat revascularization

Fig. 9.

Pooled analysis of stent thrombosis

Table 3.

Summary of pooled results

Outcomes	RR with 95% CI	p-value	I2(%)
MACE	0.86[0.78;0.95]	0.012	0%
All-cause mortality	0.86[0.77;0.97]	0.023	0%
Myocardial Infarction	0.87[0.70;1.07]	0.137	0%
Stroke	1.01[0.85;1.19]	0.912	0%
Major Bleeding	1.00[0.84;1.20]	0.975	0%
Repeat Revascularization	0.89[0.58;1.37]	0.452	45.8%
Stent Thrombosis	0.88[0.20;3.90]	0.480	0%

The forest plot presenting risk ratios with 95% CI for the primary outcome (MACE) and secondary outcomes—all-cause mortality, myocardial infarction, stroke, major bleeding, repeat revascularization, and stent thrombosis is depicted in Figs. 3, 4, 5, 6, 7, 8 and 9. TiCAB, DACAB, and TARGET did not report data on stent thrombosis, and DACAB also lacked information on repeat revascularization. Except for stent thrombosis (analyzed in two studies) and repeat revascularization (analyzed in four studies), all five studies were included in the evaluation of the remaining outcomes. Although statistical heterogeneity was low for most outcomes, with I² = 0%, this should not be interpreted as evidence of clinical homogeneity because only a small number of trials were included and the pooled estimates were heavily influenced by large PCI-based trials. The pooled effect estimates were predominantly influenced by two large randomised controlled trials, GLOBAL LEADERS and GLASSY, which together contributed to more than 80% of the total statistical weight due to the large number of participants and correspondingly small standard errors. Accordingly, the pooled estimates were largely reflective of this PCI-based trial program.

Pooled analysis of MACE

The primary outcome analysis included five studies involving 12,998 participants allocated to ticagrelor and 12,996 participants allocated to aspirin. The pooled effect estimate demonstrated a 14% relative reduction in MACE with ticagrelor compared with aspirin (risk ratio [RR] = 0.86, 95% CI 0.78–0.95; p = 0.012). Heterogeneity across studies was negligible with I² = 0% and a prediction interval 0.75–0.99. The absence of statistical heterogeneity (I² = 0%) across most outcomes strengthens the consistency of the observed treatment effects.

Pooled analysis of mortality

The pooled analysis of all-cause mortality included five randomised trials comprising 25,994 participants. Ticagrelor monotherapy was associated with a statistically significant reduction in all-cause mortality compared with aspirin monotherapy (RR 0.86, 95% CI 0.77–0.97; p = 0.023). Prediction interval was 0.71–1.05. These results suggest that ticagrelor monotherapy may provide a modest survival benefit compared with aspirin monotherapy in patients undergoing coronary revascularization. The observed reduction in all-cause mortality should be interpreted cautiously and considered hypothesis-generating rather than definitive, as the pooled estimate was largely influenced by PCI-dominant trials and was not consistently supported by the CABG-specific evidence.

Pooled analysis of myocardial infarction

The pooled analysis for myocardial infarction included five randomised trials involving 25,994 participants. Ticagrelor monotherapy was associated with a numerically lower incidence of myocardial infarction compared with aspirin monotherapy; however, the difference was not statistically significant (RR 0.87, 95% CI 0.70–1.07; p = 0.137; prediction interval 0.55–1.36). These results suggest that ticagrelor monotherapy does not significantly reduce the risk of myocardial infarction compared with aspirin monotherapy in patients undergoing coronary revascularization.

Pooled analysis of stroke

Five studies reporting stroke outcomes were included in the pooled analysis. The pooled effect estimate demonstrated no significant difference between ticagrelor and aspirin monotherapy in the risk of stroke (RR 1.01, 95% CI 0.85–1.19; p = 0.913; prediction interval 0.74–1.38). Overall, ticagrelor monotherapy did not confer a significant advantage or disadvantage compared with aspirin for stroke prevention in this population.

Pooled analysis of major bleeding

Major bleeding events were reported across five included trials. Heterogeneity across studies was negligible with I² = 0% and a prediction interval 0.80–1.25. Between-study heterogeneity was negligible (I² = 0%), indicating consistent safety outcomes across studies. These findings suggest that ticagrelor monotherapy does not significantly increase the risk of major bleeding compared with aspirin monotherapy following coronary revascularization.

Pooled analysis of repeat revascularization

Only four of the studies included revascularization as one of their outcomes. There was no significant difference for repeat revascularization (RR = 0.89, 95% CI = 0.58–1.37, p = 0.452; prediction interval 0.42–1.90). Overall, ticagrelor monotherapy did not significantly reduce the repeat revascularization compared with aspirin monotherapy following coronary revascularization.

Pooled analysis of stent thrombosis

Stent thrombosis was reported in only two of the included trials. The pooled effect estimate demonstrated no statistically significant difference between ticagrelor and aspirin monotherapy (RR 0.88, 95% CI 0.20–3.90; p = 0.481; prediction interval 0.19 – 4.07). Only two studies, GLOBAL LEADERS and GLASSY, were included in this analysis because these trials involved patients undergoing percutaneous coronary intervention (PCI), whereas the remaining studies evaluated coronary artery bypass grafting (CABG), which involves a different pathophysiological mechanism of thrombotic events. Consequently, the limited number of studies and events resulted in a wide confidence interval, resulting in substantial imprecision of the pooled estimate.

Descriptive comparision by revascularization type

Given the small number of included trials, formal subgroup meta-analysis by revascularization type was not performed. However, because PCI and CABG represent clinically distinct revascularization settings, we descriptively summarized event counts and crude risk ratios according to revascularization type. PCI data were derived from GLOBAL LEADERS and GLASSY, while CABG data were derived from TiCAB, DACAB, and TARGET. The direction of effect for MACE and all-cause mortality was similar in PCI and CABG strata, although the absolute number of events and statistical weight were substantially larger in PCI trials. Repeat revascularization showed a direction favoring ticagrelor in PCI trials but not in CABG trials. Major bleeding was similar in PCI trials, while CABG trials showed numerically more bleeding events with ticagrelor, although event counts were small. These descriptive findings should not be interpreted as evidence of treatment-effect modification because no formal subgroup interaction test was performed (Table 4).

Table 4.

Comparision by revascularization type

Outcome	Revascularization type	Ticagrelor events/total	Aspirin events/total	Crude Risk Ratio	Direction of effect
MACE	PCI	625/11,774	731/11,779	0.86	Favours ticagrelor
MACE	CABG	72/1,224	79/1,217	0.91	Slightly favours ticagrelor
All-cause mortality	PCI	335/11,774	389/11,779	0.86	Favours ticagrelor
All-cause mortality	CABG	24/1,224	27/1,217	0.88	Slightly favours ticagrelor
Myocardial infarction	PCI	356/11,774	385/11,779	0.93	Slightly favours ticagrelor
Myocardial infarction	CABG	22/1,224	34/1,217	0.64	Favours ticagrelor numerically
Stroke	PCI	124/11,774	126/11,779	0.98	No clear difference
Stroke	CABG	33/1,224	30/1,217	1.09	Slightly favours aspirin
Major bleeding	PCI	263/11,774	270/11,779	0.97	No clear difference
Major bleeding	CABG	38/1,224	29/1,217	1.30	Numerically favours aspirin
Repeat revascularization	PCI	460/11,774	545/11,779	0.84	Favours ticagrelor
Repeat revascularization	CABG	45/1,058	35/1,051	1.28	Numerically favours aspirin
Stent thrombosis	PCI	129/11,774	146/11,779	0.88	Slightly favours ticagrelor
Stent thrombosis	CABG	Not reported	Not reported	Not applicable	Not applicable

Crude RRs were calculated from summed event counts within PCI and CABG strata and are presented for descriptive purposes only. They do not represent formal subgroup meta-analysis. Formal subgroup testing was not performed because only five trials were included and the analysis was underpowered. Stent thrombosis was reported only in PCI trials.

Sensitivity analysis

A leave-one-out sensitivity analysis was performed for all major outcomes. Individual trials for MACE, major bleeding, myocardial infarction, stroke, thrombosis, repeat revascularization, and mortality were sequentially omitted, but the aggregated impact estimates remained steady. Most iterations showed little or no heterogeneity. Although the direction of effect was generally stable across leave-one-out analyses, the statistical significance of some outcomes changed after exclusion of large trials such as GLOBAL LEADERS and GLASSY. Therefore, the sensitivity analyses support cautious interpretation rather than definitive robustness of the pooled estimates. Detailed sensitivity analyses are presented in the Supplementary Appendix S4.

Publication bias

Publication Bias was assessed using a funnel plot for each outcome. Formal statistical testing of the asymmetry was not done as recommended by methodological guidelines.

Visual inspection of the funnel plot did not show obvious asymmetry but interpretation is limited as fewer than ten studies were included. Funnel plot for each outcome is presented in the Supplementary Appendix S5.

Certainty of evidence

Using the GRADE approach, the certainty of evidence was rated low for major adverse cardiovascular events, all-cause mortality, stroke, and major bleeding, very low for myocardial infarction, repeat revascularization, and stent thrombosis. The certainty ratings were mainly lowered because two of the included randomised trials had some concerns in the RoB 2.0 assessment, particularly related to deviations from intended interventions in open-label settings. Additional downgrading was done for indirectness because PCI and CABG populations were pooled despite distinct mechanisms of thrombotic risk, and for dominance bias because estimates were heavily influenced by GLOBAL LEADERS and GLASSY. Further downgrading for imprecision was applied to myocardial infarction and repeat revascularization because the confidence intervals crossed the line of no effect and remained compatible with both clinically important benefit and no clear effect. Stent thrombosis was rated to have very low-certainty evidence because only two studies contributed data and the confidence interval was very wide. No serious statistical inconsistency was identified for most outcomes, however, clinical heterogeneity remained important because PCI and CABG populations, treatment pathways, and trial weights differed across studies. The summary of findings table is presented in Supplementary Table S2.

Discussion

Managing the transition from dual antiplatelet therapy (DAPT) to monotherapy involves balancing ischemic protection against bleeding risks, particularly after coronary revascularization. Our meta-analysis of five randomised controlled trials, including 25,994 participants, addressed this clinical challenge by comparing ticagrelor monotherapy directly against aspirin monotherapy. We found that ticagrelor monotherapy was associated with a 14% relative reduction in major adverse cardiovascular events (MACE) (RR 0.86, 95% CI 0.78–0.95) and a significant reduction in all-cause mortality (RR 0.86, 95% CI 0.77–0.97).

Conversely, we observed no statistically significant differences between the two strategies regarding myocardial infarction, stroke, major bleeding, repeat revascularization, or stent thrombosis.

Recent literature provides an important context for interpreting these results. In high-bleeding-risk PCI, current evidence supports abbreviated DAPT followed by single antiplatelet therapy in selected patients, particularly when radial access, newer-generation devices, and bleeding risk stratification are used, but it does not resolve which monotherapy agent should be preferred after aspirin withdrawal [10]. Our findings therefore add to this evolving field by focusing on the direct randomized comparison of ticagrelor monotherapy versus aspirin monotherapy after revascularization. However, our pooled estimate should not be read as a simple mandate to replace aspirin with ticagrelor, because treatment choice depends on ischemic risk, bleeding risk, revascularization type, and the clinical pathway leading to monotherapy.

The observed reduction in MACE suggests that ticagrelor monotherapy may provide modest ischemic benefit compared with aspirin monotherapy in this setting. This finding aligns with the PANTHER individual patient data meta-analysis, which demonstrated that P2Y12 inhibitor monotherapy reduced composite coronary outcomes versus aspirin in patients with established coronary artery disease [20]. Mechanistically, this benefit may stem from the more potent and consistent platelet inhibition provided by ticagrelor. Pharmacodynamic data from the TEMPLATE trial indicate that ticagrelor monotherapy achieves profound suppression of ADP-induced aggregation while functionally overlapping with aspirin’s effects on thromboxane pathways [21]. Furthermore, studies in healthy volunteers confirm that ticagrelor alone provides broad platelet inhibition across multiple agonists [22]. These pharmacologic findings provide biological plausibility for the observed association between ticagrelor monotherapy and lower composite ischemic events, but they should not be interpreted as definitive evidence of clinical superiority across all post-revascularization populations.

A notable finding of our analysis is the reduction in all-cause mortality with ticagrelor monotherapy. This signal is consistent with a recent network meta-analysis of antithrombotic strategies in coronary disease, which reported a hazard ratio of 0.68 for mortality with ticagrelor monotherapy versus aspirin [23]. However, this result warrants cautious interpretation. Other broad secondary prevention meta-analyses have found virtually identical mortality rates between P2Y12 inhibitors and aspirin [24]. In our pooled analysis, the mortality estimate is heavily influenced by the GLOBAL LEADERS trial and its GLASSY substudy, which together contributed more than 80% of the statistical weight [8, 16]. Conversely, the TiCAB trial in surgical patients showed no difference in mortality (RR 0.96), suggesting that the mortality signal may not be consistent across all revascularization settings [17].

Regarding safety and specific thrombotic outcomes, our results indicate a neutral bleeding profile and divergent findings based on revascularization type. The pooled relative risk for major bleeding was 1.00, suggesting that ticagrelor does not inherently increase major bleeding compared to aspirin in this monotherapy context. This contrasts with the landmark analysis of GLOBAL LEADERS, where ticagrelor monotherapy was associated with higher BARC type 3 or 5 bleeding (adjusted HR 1.89) [25]. In the surgical domain, both the DACAB and TARGET trials found no significant difference in vein graft patency or occlusion rates between ticagrelor and aspirin monotherapy [19, 26]. This distinction is critical because graft occlusion differs mechanistically from stent thrombosis. While we found no significant difference in stent thrombosis, this outcome was reported in only two trials with a wide confidence interval (RR 0.88, 95% CI 0.20–3.90), limiting our ability to draw definitive conclusions for PCI-specific thrombotic risks.

In the descriptive comparison by revascularization type, the crude direction of effect for MACE and all-cause mortality was broadly similar in PCI and CABG strata. However, the PCI trials contributed most participants and events, whereas CABG estimates were based on substantially smaller trials with fewer outcome events. Outcome-specific differences were also observed: repeat revascularization appeared more favorable with ticagrelor in PCI trials, while CABG trials showed numerically more repeat revascularization and major bleeding events with ticagrelor. These findings reinforce that PCI and CABG populations should not be considered biologically interchangeable and that the pooled estimates should be interpreted as hypothesis-generating.

These findings may have clinical relevance for selected patients but should be interpreted within the limitations of the available evidence. For selected patients transitioning off DAPT, particularly those at high ischemic risk who can tolerate potent P2Y12 inhibition, ticagrelor monotherapy may be considered a potential alternative, although the apparent reductions in MACE and mortality should be viewed as hypothesis-generating. This supports the strategy of early aspirin withdrawal, a concept reinforced by the TWILIGHT trial, which showed that ticagrelor monotherapy reduced bleeding without sacrificing ischemic protection compared to ticagrelor-based DAPT [27]. However, decision-making must remain individualized. The lack of benefit seen in the TiCAB trial suggests that routine substitution of aspirin with ticagrelor after CABG may not be warranted based on current evidence [28].

Heterogeneity in monotherapy initiation and treatment pathways

The included trials also differed in how and when monotherapy was initiated. GLOBAL LEADERS used 1 month of DAPT followed by prolonged ticagrelor monotherapy, whereas GLASSY evaluated a prespecified population within the GLOBAL LEADERS strategy.

CABG trials evaluated ticagrelor or aspirin in surgical patients with different treatment objectives, including graft patency and postoperative clinical outcomes. These differences mean that the pooled estimate does not represent a single uniform monotherapy pathway.

Rather, it summarizes several related but distinct strategies involving aspirin withdrawal, direct ticagrelor monotherapy, and post-surgical antiplatelet management. This treatment-pathway heterogeneity should temper clinical interpretation.

Recent evidence on other components of optimal medical therapy also reinforces the need for individualized interpretation of antiplatelet findings. The dialysis ASCVD statin meta-analysis suggested lower all-cause mortality and MACE with statin therapy but was based on observational evidence and did not show significant effects on cardiovascular mortality, myocardial infarction, or stroke [29]. Similarly, the reconstructed individual-patient-data meta-analysis of beta-blockers after MI without reduced ejection fraction found no significant reduction in mortality, recurrent MI, or heart failure, supporting selective use rather than universal long-term therapy [30]. Although these studies do not directly evaluate antiplatelet monotherapy, they place the present findings within a contemporary secondary-prevention landscape in which therapy is increasingly tailored to patient phenotype, procedural context, and competing ischemic and bleeding risks.

Our study has several strengths, including the use of exclusively randomised data and a large pooled sample size that allowed for precise estimation of common outcomes. Nevertheless, limitations exist. The inclusion of only five trials limits the power for subgroup analyses. The findings are also limited by dominance of the GLOBAL LEADERS and GLASSY trial program. These studies contributed more than 80% of the statistical weight for the primary analysis, making the pooled estimate partly reflective of a single large PCI-based treatment strategy. In addition, GLOBAL LEADERS was open-label, and its design evaluated a complex antiplatelet pathway rather than a direct head-to-head comparison of ticagrelor monotherapy versus aspirin monotherapy initiated at the same time point. These factors limit the generalizability of the pooled estimate and support cautious, hypothesis-generating interpretation. Furthermore, we pooled PCI and CABG populations, yet post-PCI stent thrombosis and post-CABG graft occlusion reflect different pathophysiological mechanisms.

Finally, differences in the timing of monotherapy initiation and outcome definitions across trials required harmonization, which may introduce heterogeneity.

In conclusion, ticagrelor monotherapy after coronary revascularization was associated with a potential reduction in MACE and all-cause mortality compared with aspirin monotherapy, with no observed difference in major bleeding. These findings are hypothesis-generating and primarily reflect PCI-based evidence, especially the GLOBAL LEADERS trial program. They should not be interpreted as definitive evidence of a universal survival benefit or as support for routine replacement of aspirin with ticagrelor after revascularization. Future studies should prioritize adequately powered randomised comparisons stratified by revascularization type, clinical presentation, and timing of monotherapy initiation to better define the optimal target population for this strategy.

Authors' contributions

DRR was responsible for the conception and design, SR and SU searched literature, extracted data and analysed the data, SR, SU and NG wrote the manuscript, DRR revised the manuscript. All the authors agreed to the final version of the manuscript.

Funding

No funding or sponsorship was received for this study.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. All data analyzed in this study are derived from previously published studies, which have been cited appropriately in the manuscript.

Declarations

Ethical approval and consent to participate

Ethical approval and informed consent were not required as this study is a systematic review and meta-analysis of previously published data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.