This study is registered with PROSPERO (CRD42024588250) and was conducted in accordance with the 2020 version of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement (the PRISMA checklist is provided in Checklist 1) [17]. Four English databases (Cochrane Library, PubMed, Embase, and Web of Science) and 1 Chinese database (Wanfang) were searched, and the publication deadline for the literature was October 1, 2024. We searched each database to identify MeSH (Medical Subject Headings) (eg, “aspirin,” “indobufen,” and “clinical trial”) and related free terms (eg, Acetylsalicylic Acid*, Acetysal*, Acylpyrin*, Aloxiprimum*, Colfarit*, Ibustrin*, controlled trial*, and Trial*). The search strategy was developed via a combination of MeSH terms and related free terms . Moreover, the references of all included studies were browsed and screened to complement the eligible literature not retrieved. Owing to the limited number of randomized controlled trials (RCTs) in the relevant field, we also included eligible well-conducted cohort studies in our analysis to enhance statistical power and generalizability. This approach of combining RCTs and nonrandomized studies is acceptable when the objective is to evaluate both efficacy and effectiveness in broader clinical settings [18].

For inclusion in the systematic review and meta-analysis, we selected clinical trials that investigated the antiplatelet impact of indobufen and aspirin and their safety in human participants. The following inclusion criteria were applied: (1) the study population comprised adults (aged 18 years and older) with CHD caused by coronary artery atherosclerosis or stroke caused by intracranial atherosclerosis; (2) the intervention was represented by indobufen in the study groups versus aspirin in the control groups; (3) the primary outcome was the incidence of MACCEs or any bleeding or BARC 2/3/5 bleeding; (4) the secondary outcomes were cardiovascular death, MI, and ischemic stroke; adverse cardiovascular events such as coronary thrombus reformation, heart failure, MI, stroke, angina pectoris, and cardiovascular death; stroke; MI; and cardiovascular death; and (5) the studies were randomized clinical trials with either crossover or parallel designs or prospective observational trials. We have supplemented the secondary outcome measures from the registered protocol in order to enhance the guiding value of clinical practice. Our exclusion criteria were as follows: (1) lack of control group studies, (2) no studies reporting cardiovascular or cerebrovascular events or bleeding during the follow-up or research group period, (3) follow-up time of less than 6 months or no mention of follow-up time, and (4) case control studies, cross-sectional studies, ecological studies, case reports, and other descriptive studies. Reviews, conference abstracts, and commentaries were also excluded. Two authors (WHP and LEG) independently searched the appropriate studies from all the databases. Two authors (WHP and LEG) independently screened the studies and removed the papers that were duplicates or not relevant after carefully screening the titles and abstracts. In the next step, we further independently examined the remaining trials and excluded the ineligible trials according to the full text. We resolved disagreements between the authors by discussion. And a third researcher (HCZ) further resolved any conflicts.

Two researchers (WHP and LEG) independently extracted the following characteristics from the trials via a standardized data extraction form: year of publication, lead author, country of publication, total sample size of the trial in the intervention and control groups, duration of follow-up (months), age at entry, participants’ sex percentage, drug dosage (mg per day), and sample size for the occurrence of outcome indicators.

Our data were managed via Excel, and we used Review Manager 5.3 (The Cochrane Collaboration) and Stata SE 18.0 (Stata Corp LLC) to combine and analyze outcomes from the selected trials. All the data were pooled to calculate the risk ratio (RR) and 95% CIs via a fixed-effects model. Between-trial heterogeneity was assessed by using an I² test, and a value of >50% was regarded as considerable heterogeneity. If heterogeneity was high, we used a random-effects model. Statistical significance was defined as a P value <.05. If the number of pooled studies was greater than 10, publication bias was surveyed by visual inspection of the funnel plot and Egger test.

We divided all the included studies into RCTs and non-RCT trials. Two researchers (WHP and LEG) independently assessed the quality of RCTs on the basis of the Cochrane risk-of-bias criteria according to the following criteria [18]: (1) allocation concealment, (2) random sequence generation, (3) blinding of participants and personnel, (4) blinding of outcome assessment, (5) incomplete outcome data, (6) selective reporting, and (7) other probable sources of bias. The Newcastle‒Ottawa Scale was used to evaluate risk of bias for non-RCTs across domains such as selection, intervention classification and comparability, missing data, outcome measurement, and result reporting [19]. Studies with a score of ≥7 were considered high-quality studies. We performed leave-one-out sensitivity analyses to test the robustness of the results.

As reported in Figure 1, 307 papers from the English database were initially retrieved. A total of 86 duplicate studies were removed. Among the remaining 221 papers, 173 were excluded after the titles and abstracts were carefully screened. In the next step, 48 trials were further examined, and 41 additional trials were excluded for the following reasons: the population studied in 3 papers did not meet the inclusion criteria, the intervention methods used in 5 papers did not meet the requirements, 4 studies did not report the required outcome measures, 18 studies did not provide a proper comparison with aspirin, 1 study’s design method did not meet the criteria, 4 papers had a follow-up time of less than 6 months or did not mention the follow-up time, and 6 papers did not retrieve the results. Finally, 7 studies (Dai et al [14], Wu et al [15], Bai et al [20], Dai et al [21], Pan et al [22], Shi et al [23], and Ren et al [24]) were included in our meta-analysis. An additional 11 studies (Chen et al [25], Chen [26], Lin et al [27], Liu [28], Ma [29], Tu et al [30], Wen and Cui [31], Wu [32], Yang [33], Rui et al [34], and Zhou et al [35]) from the Chinese database Wanfang were included in the study. A total of 12,981 patients were included in this analysis.

The mean age of the participants varied from 47.20 to 72 years. The sample sizes of the eligible studies ranged from 19 to 2723 participants. All the studies were performed in China. The literature we have included has focused primarily on populations with ischemic stroke caused by cerebral vascular atherosclerosis and CHD, which includes stable angina pectoris, acute coronary syndrome, and coronary artery stent implantation or coronary artery bypass grafting due to CHD. The dose of indobufen given to these participants ranged from 100 mg daily to 100 mg twice a day, and the daily dose of aspirin was 100 mg. The characteristics of the included studies are summarized in Table 1. Figure 2 and Table 2 provide information regarding the quality assessment for the risk of bias for each study included in the meta-analysis.

In this meta-analysis, the results of any bleeding and BARC studies revealed a high level of heterogeneity, with I² values of 51% and 84%, indicating substantial heterogeneity, possibly due to the small sample size. To further explore the source of this heterogeneity, we conducted a sensitivity analysis. This involved sequentially excluding individual studies to assess their influence on the overall pooled estimate. Although the limited number of studies restricts the depth of the analysis, the results showed that the pooled effect estimate remained stable, indicating relative robustness (Figures 3 and 4). However, variations in heterogeneity were observed when individual studies were excluded, suggesting that 1 or more studies might have contributed to the observed heterogeneity. We attribute the source of the heterogeneity primarily to the use of the second antiplatelet agent in the dual antiplatelet therapy (DAPT) regimen, as well as to variations in sample size and concomitant use of proton pump inhibitors.

For outcomes with more than 10 papers included in the literature, we plotted funnel plots and performed Egger tests. There are 16, 13, 12, and 13 studies on outcomes of any bleeding events, adverse cardiovascular events, stroke events, and MI, respectively. The P values of the Egger tests for any bleeding or adverse cardiovascular events were .249 and .235, respectively, both of which are greater than .05. We believe that the publication bias is small, but the P values of the Egger tests for stroke and MI were .000 and .001, indicating a certain degree of publication bias. All funnel plots are included in Multimedia Appendix 1.

In this systematic review and meta-analysis involving 18 clinical trials and 12,981 patients, we compared antiplatelet therapy involving indobufen with that involving aspirin. Although we did not impose any language restrictions for the included studies, all the clinical trials ultimately retrieved were conducted in China. This may be because indobufen is not widely approved as a routine antiplatelet agent in Europe and the Americas and is used only as a reference drug in specific studies [36]. These results suggest that, compared with aspirin, indobufen can effectively reduce the incidence of mucosal bleeding (including both major and minor bleeding) in patients with cardiovascular and cerebrovascular atherosclerosis. Additionally, on the basis of the studies included in our analysis, indobufen was shown to reduce the incidence of adverse cardiovascular events (a composite outcome of coronary thrombus reformation, heart failure, MI, stroke, angina pectoris, and cardiovascular death) as well as MI. However, MACCEs, stroke, and cardiovascular death rates were similar between the 2 groups, which may be related to the limited follow-up time. The heterogeneity for any bleeding and BARC 2/3/5 bleeding was high, but the sensitivity analysis revealed no difference in outcomes. There was significant publication bias for stroke, MI, and cardiovascular death. The effects of indobufen on MACCEs, stroke, and cardiovascular death were similar to those of aspirin, demonstrating no inferiority in efficacy.

Antiplatelet therapy in patients with atherosclerosis remains a topic of considerable debate. Current guidelines recommend aspirin for secondary prevention in patients with coronary atherosclerotic heart disease, and DAPT is advised for patients with a high risk of thrombosis and a low risk of bleeding or those undergoing interventional treatment [37]. For patients who require DAPT but are intolerant to aspirin, indobufen is being investigated as a potential alternative to aspirin [38,39].

The antiplatelet effect of aspirin is achieved by irreversibly inhibiting and permanently blocking COX-1 in platelets, thereby preventing the formation of thromboxane A2 (TXA2) [40]. This effect lasts for the lifespan of the platelet (7‐10 days) [41]. In contrast, indobufen is a reversible COX-1 inhibitor, meaning that it temporarily inhibits the enzyme but does not cause persistent inhibition [42]. This reversible action allows for shorter-term effects on platelet function than does aspirin [43]. The inhibitory effects of indobufen on platelet function returned to baseline within 24 hours after discontinuation. Additionally, indobufen does not substantially impact prostacyclin, potentially minimizing the risk of aspirin-gastric side effects [25]. While both drugs seek to reduce platelet aggregation and associated thrombosis, the irreversible inhibition of aspirin leads to protracted platelet inhibition, whereas indobufen offers a reversible and potentially safer alternative, particularly in patients at greater risk for bleeding.

Aspirin’s first-line status in the prevention and treatment of cardiovascular and cerebrovascular diseases is supported by a large body of evidence-based research [10,44,45]. In comparison, although the clinical application of indobufen has evolved, it remains very limited at present. Initially, owing to its reversible inhibition of COX, it was not recommended, as it was considered insufficient to effectively inhibit platelet aggregation in vivo [46]. However, subsequent clinical trials have led some meta-analyses to consider indobufen similarly to aspirin. Cataldo G et al [47], in their combined analysis of 2 multicenter coronary artery bypass grafting studies, concluded that indobufen is as effective as aspirin and dipyridamole in preventing graft occlusion. Bhana and McClellan [48], in their review, suggested that indobufen may be an effective alternative to aspirin for patients with nonrheumatic atrial fibrillation who have contraindications to anticoagulation or a greater risk of bleeding. In a meta-analysis of 9 clinical studies, Zhang et al [49] reported that indobufen had fewer gastrointestinal reactions, a lower risk of bleeding, and could be a viable option for patients at high risk of bleeding or with gastrointestinal ulcers.

The strengths of our study include the following: compared with previous studies, our meta-analysis incorporated a larger number of clinical trials and patients with homogeneous outcome definitions and a large patient population, making the results more generalizable to the broader population. Additionally, the follow-up duration in all the included studies exceeded 6 months, enhancing the credibility of our findings. We also conducted publication bias and sensitivity analyses to further emphasize the reliability of this meta-analysis.

There are several limitations to this study. All the studies included focused on Chinese populations, which may limit the generalizability of our findings to other regions of the world. But we believe that with the accumulation of further clinical evidence in China, the safety profile of indobufen as a potential alternative to aspirin will gain stronger support, paving the way for the development of larger multicenter studies with longer follow-up periods and more diverse populations across different regions—including global clinical trials. Additionally, because antiplatelet therapy is a long-term treatment requiring good patient adherence, the included trials were open-label, which might raise concerns about potential bias, although none of the studies carried a high risk of bias. Owing to the lack of available data, we were unable to perform subgroup analyses on the basis of individual patient characteristics, and the efficacy and safety of indobufen in specific populations require further investigation.

Our meta-analysis indicates that, compared with aspirin, indobufen offers a safer profile for antiplatelet therapy in patients with cardiovascular and cerebrovascular atherosclerosis, as it reduces the incidence of mucosal bleeding. In terms of efficacy, indobufen is associated with a lower incidence of adverse cardiovascular events and MI, while its impact on MACCEs and cardiovascular mortality is similar to that of aspirin. Indobufen demonstrated no overall inferiority in effectiveness, with superior results in some aspects (adverse cardiovascular events and MI). This provides additional robust evidence supporting the feasibility of using indobufen as an alternative antiplatelet therapy, particularly in patients who are intolerant to aspirin.