Comparison of Clinical Outcomes Between Ticagrelor and Clopidogrel in Acute Coronary Syndrome: an Updated Meta-Analysis

Background: Ticagrelor is currently recommended for patients with acute coronary syndrome (ACS). However, recent studies have yielded controversial results. Objective: To compare the clinical outcomes of ticagrelor and clopidogrel in ACS patients. Methods: Three electronic databases were queried until April 1, 2021. Major adverse cardiovascular event (MACE) was the primary ecacy endpoint. The secondary ecacy endpoints included stroke, stent thrombosis (ST), cardiovascular (CV) death, all-cause death, and myocardial infarction (MI). The safety endpoints were (major and minor) bleeding. Odds ratios (ORs) and 95% condence intervals (CIs) and were calculated to represent the estimated effect sizes. Results: Nine clinical trials and 18 observational studies with 269,935 ACS patients were included. No signicant difference was detected in MACE (OR 0.76, 95% CI 0.54-1.06, p = 0.11, I² = 66.74%), but ticagrelor introduced a higher risk of bleeding (1.49, 1.14-1.94, 0.00, 63.97%) and minor bleeding (1.57, 1.08-2.30, 0.02, 59.09%) in clinical trials. The secondary ecacy endpoints differed between clinical trials and observational studies. Subgroup analysis demonstrated that ticagrelor showed better therapeutic effects in patients underwent PCI (0.38, 0.23-0.63, 0.00, 0) than those intended for PCI (1.02, 0.70-1.49, 0.93, 68.99%). Meanwhile ticagrelor showed different therapeutic effects on ACS patients of different ethnicities and from different countries. Conclusion: This meta-analysis demonstrated that ticagrelor is not superior to clopidogrel in MACE but is associated with a higher risk of bleeding in ACS patients. Different PCI strategies, ethnicities, and countries may be the factors that contribute to different therapeutic ecacy of ticagrelor.


Introduction
Currently, cardiovascular disease (CVD) is the largest contributor to the disease burden, accounting for approximately one-third of the global deaths [1]. Besides, acute coronary syndrome (ACS) as a common and serious CVD, its incidence increases dramatically with age, proposing a great challenge to public medical care. Dual antiplatelet therapy (DAPT) is the mainstay treatment strategy against ACS, with timely vascularization as needed [2,3]. As for the choice of antiplatelet agent, the 2016 American College of Cardiology (ACC) / American Heart Association (AHA) guidelines and the 2018 European Society of Cardiology (ESC) / European Association for Cardio-Thoracic Surgery (EACTS) guidelines recommend ticagrelor over clopidogrel for ACS treatment or the patients who have received percutaneous coronary intervention (PCI) [2,4].
As a novel generated adenosine diphosphate receptor antagonist, ticagrelor provides faster, more potent, and more stable platelet inhibitory effects than clopidogrel [5,6]. The large Platelet Inhibition and Patient Outcomes (PLATO) trial exhibited that compared with clopidogrel, ticagrelor reduced the incidence of stroke, myocardial infarction (MI), and cardiovascular (CV) death, without elevating the risk of major bleeding [7]. However, data from other large clinical trials [8,9] and observational studies [10][11][12] drew controversial conclusions. Meanwhile, meta-analyses published recently also reported inconsistent results [13][14][15][16]. Therefore, we conducted a meta-analysis to review previous relevant studies and compare the clinical bene ts of ticagrelor and clopidogrel treatments in the ACS population in the context of aspirin use to address these con icting conclusions.

Methods
We reported the current meta-analysis following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline [17] (Online Resource ESM_1), and have registered the study in an international prospective register of systematic reviews PROSPERO (ID: CRD42021251212).

Literature search
Three electronic databases, Cochrane, EMBASE, and PubMed library, were searched for eligible citations prior to April 1, 2021. The following relevant keywords were used: "ticagrelor", "clopidogrel", "myocardial ischemia", "ACS", "percutaneous coronary intervention", and "PCI". The detailed search strategy was summarized in Online Resource ESM_2. Additionally, reference lists in relevant meta-analyses were manually searched for potential eligibility.

Inclusion and exclusion criteria
We reviewed the full text of the potentially eligible literature to determine if they ful lled the inclusion criteria as follows: (1) Adult (≥18 years old) patients with ACS who underwent PCI (PCI strategy proportion equal to 100%) or intended for PCI (proportion less than 100%); (2) Clinical trials and observational studies that compared ticagrelor versus clopidogrel in the context of aspirin use; (3) One or more of the following outcomes reported during any follow-up period: MACE, all-cause death, CV death, MI, stroke, stent thrombosis (ST), bleeding, major or minor bleeding. Literature that met the following exclusion criteria would be excluded: (1) Studies with incomplete data, or with only reported unadjusted endpoints in observational studies; (2) Studies from the same sample source; (3) Studies not in English.

Study end points
The primary e cacy endpoints were trial-de ned primary MACEs or e cacy endpoints (described as death / CV death, MI, and / or stoke) (Online Resource ESM_3). The secondary endpoints included stroke, ST, MI, CV death, and all-cause death. The safety endpoints were trial-de ned bleeding (a composite of major or minor bleeding), major bleeding, and minor bleeding (Online Resource ESM_3). For the de nitions of safety outcomes, if not otherwise speci ed, we prioritized PLATO de nitions when available.

Data extraction
Data from included citations were independently populated with a standardized data extraction by two researchers (SMY and CWC), with any discrepancy resolved by a third researcher (LLP). The data systematically extracted in this meta-analysis included the authors' last names, study type, country (where the study was conducted), publication year and journal, disease subtype, sample size, age, the proportion of male and PCI strategy, dosing regimen, reported outcomes assessing e cacy and safety, and duration of follow-up.

Quality assessment
We assessed the risk of bias of randomized controlled trials (RCTs) using the Cochrane Risk Bias Assessment Tool, which covers six domains of bias, classi ed into three levels including "low bias", "unclear", and "high bias" [18]. Meanwhile, we assessed the quality of observational studies using the Newcastle-Ottawa Scale (NOS), which consists of eight items divided into three aspects including "selection", "comparability", and "outcome" assessment. The NOS adopts the semi-quantitative principle of star allocation to assess the literature quality, with a full score of nine stars [19].

Statistical analysis
We used the odds ratios (ORs) to represent the estimated effects, which along with the corresponding 95% con dence intervals (CIs) were obtained via the Stata 16.0 software (StataCorp, CollegeStation, TX, USA) [20]. Given the inclusion of heterogeneous populations, we chose the random-effects model to pool the effect sizes for this meta-analysis. Furthermore, we used the Higgins' I² statistics and Cochran's Q test to estimate heterogeneity across studies. A p value < 0.05 was considered as statistically signi cant.
All analyses were performed by analyzing data from clinical trials and observational studies separately to reduce the heterogeneity due to different study types. Meanwhile, subgroup analyses were performed to search for potential sources of heterogeneity. In brief, we performed two subgroup analyses, including propensity score-adjusted analyses (PA) group and multivariate-adjusted analyses (MA) group, in the observational studies according to different data types. Additionally, we conducted pre-speci ed subgroup analyses based on PCI strategy, ethnicity, country, and duration of follow-up.in the included RCTs.
The stability of our ndings was evaluated by sensitivity analysis, that is, calculating the effects by including high quality RCTs. When more than ten studies were included, the presence of publication bias was investigated by the Egger's test and displayed by visual estimation (symmetry) of contour-enhanced funnel plots.

Results
Eligible studies and patient characteristics The process of searching, retrieving, and screening in this meta-analysis is shown in Fig. 1. A total of 5,145 potentially relevant literatures were screened. Then 1,299 duplicates were excluded, and 3,767 were retrieved for title and abstract screening. Subsequently, the full texts of 79 articles were reviewed for eligibility.
Among them, forty-two articles did not meet the inclusion criteria, eight presented with incomplete data, and two employed same cohorts included in our study.
Subgroup analysis based on different Asian countries showed that Chinese patients bene ted more from ticagrelor treatment than those in Korean and Japanese, while the risk of bleeding signi cantly increased in all three Asian countries (Fig. 3c). Further subgroup analyses were conducted to analyze the safety and e cacy of ticagrelor in different follow-up duration classi cations. It showed that bleeding risk and MACE was comparable between the two groups during the follow-up duration (Online Resource ESM_6).
According to different data type, publication bias was investigated in the two groups: the composite of propensity score matched/adjusted studies and clinical trials, and the composite of multivariable adjusted studies and clinical trials. By contour-enhanced funnel plots (Fig. 4) and the results of the Egger's test (Table 3), we detected no publication bias except for MACE and MI in the composite of propensity score matched/adjusted studies and clinical trials. The results of nonparametric trim-and-ll analysis showed that ve studies were lled for MACE with the total results in uenced, and two were lled for MI with the total results unaffected (Online Resource ESM_9, Table 4).

Discussion
This meta-analysis, based on 27 studies, suggested that ticagrelor was not only inferior to clopidogrel in patients with ACS, but also related to an increased bleeding risk. Whereas, ticagrelor was more effective in the treatment of those who underwent PCI than clopidogrel, as it signi cantly reduced the incidence of MACE. Meanwhile the present meta-analysis revealed that Caucasians and East Asians had inconsistent safety and e cacy pro les under ticagrelor treatment. Among East Asian patients, Chinese bene ted more from ticagrelor compared to Korean and Japanese. Care should therefore be taken to screen the eligible population when applying ticagrelor, and the bleeding risk under ticagrelor treatment was of concern.
Recently, several studies have revisited the issues concerning clinical applications of ticagrelor and clopidogrel. Wu et al. [13] found that ticagrelor treatment was more bene cial than clopidogrel treatment in European, American and Asian populations, who had a higher risk of bleeding. However, this analysis mixed all studies together and ignored the heterogeneity between clinical trials and observational studies. Furthermore, Guan et al. [15] reported that the e cacy endpoint of ticagrelor was comparable to that of clopidogrel, but the safety endpoint of ticagrelor should be further studied. Similarly, they performed a mixed analysis while including studies involving patients with stable coronary atherosclerotic heart disease (CAD). Consistently, Fan et al. [16] reported that ticagrelor only exhibited a tendency to reduce MACE at the expense of bleeding. Although the current meta-analysis differentiated RCTs and observational studies, no further subgroup analysis based on other study (or population) characteristics was performed. In this meta-analysis, we performed pre-speci ed subgroup analyses for MACE and bleeding according to PCI strategy, ethnicity, country, and follow-up duration in RCTs. Additionally, we performed two subgroup analyses (the PA and MA groups) in observational studies according to different data types. The aforementioned meta-analyses on this issue reported different clinical e ciency of ticagrelor, but they consistently found increased bleeding risk. We conducted a comprehensive analysis from both clinical trial and real-world practice considerations to compare the clinical outcomes of ticagrelor and clopidogrel in different subgroups of patients. Therefore, some de nitive evidence can be provided for clinicians to choose from ticagrelor and clopidogrel.
We found similar primary e cacy results for ticagrelor treatment in both clinical trials and observational studies, though the MA group differed slightly. The secondary e cacy endpoints differed in clinical trials and observational studies, which can be attributed to inherent differences between study types. In brief, clinical trials (especially RCTs) match baseline characteristics well. And observational studies, although clinical factors associated with treatment selection can be matched by propensity score matched / adjusted analyses or multivariable adjusted analyses, there are still some unadjusted or incomplete adjustment variables that may affect the results. This may explain the difference in bleeding in MA group, as well as the difference in the secondary endpoints between clinical studies and observational studies.
In addition to pharmacological treatment with DAPT, the primary management of ACS patients involves an early invasive strategy consisting of coronary angiography, PCI, and even coronary artery bypass grafting (CABG) [2,4,41]. Considering that a high proportion of high-risk patients with ACS received PCI after coronary angiography, we included those who underwent or intended for PCI to better compare the clinical outcomes. Subgroup comparison in this study revealed that ticagrelor could signi cantly reduce the incidence of MACE in patients underwent PCI, but induced a higher bleeding risk than clopidogrel. Whereas, in patients intended for PCI, ticagrelor was not superior over clopidogrel regarding MACE and bleeding. In ACS patients intended for PCI, the proportion of PCI strategy ranged from 42% (Cannon 2007) to 84.8% (Goto 2015) in the ticagrelor group in clinical trials, and from 88.5% (Sahlén 2016) to 92.3% (Vercellino 2017) in observational studies. Ahn et al. reported no signi cant difference in the outcomes between PCI and CABG complete revascularization [42]. Thus, ACS patients who received PCI and CABG had different clinical characteristics from those receiving medical treatment alone. Ticagrelor could signi cantly reduce the risk of MACE in patients undergoing PCI, providing valuable suggestions for clinicians to choose from ticagrelor and clopidogrel according to the PCI strategy.
In terms of ethnicity, Caucasians made up 91% of the population included in the PLATO trial, whereas East Asians accounted for only 6%. Caucasians and East Asians differ substantially in phenotypes and genomics; Caucasian based recommendations do not necessarily apply to East Asians [13]. We conducted subgroup analyses between different ethnicities, and we found that ticagrelor in Caucasian patients statistically decreased the incidence of MACE, while not increasing the risk of bleeding. However, this was not the case for East Asian populations, where ticagrelor had a comparable effect to clopidogrel on MACE but was more likely to cause bleeding events. East Asians are thought to be more prone to bleeding events than Caucasians, but are relatively resistant to adverse ischemic outcomes after PCI (the so-called "East Asian paradox") [43][44][45]. In addition, cytochrome P450 2C19 loss-of-function alleles related to high platelet reactivity are more common in Asian populations [46]. In a nutshell, there are signi cant ethnic differences between Caucasian and East Asian patients in terms of thrombosis, platelet P2Y12 receptor inhibition, and predisposition to bleeding complications [43]. Those differences might partly contribute to this inter-population disparity. The number of RCTs included in this meta-analysis is relatively small, requiring more RCTs for further elaboration.
In East Asia, the MACE incidence was signi cantly reduced by ticagrelor in Chinese, but not in Korean and Japanese, which can, at least partially, be attributed to the different proportions of patients, underwent PCI. Speci cally, The Korean TICAKOREA trial was a multicenter randomized study, in which 81.5% of the ACS patients assigned to the ticagrelor group underwent PCI [8]. In another PHILO trial conducted mainly in Japan, 84.8% of patients treated with ticagrelor received PCI [9]. More importantly, data from four Chinese trials showed that up to 94% of patients underwent PCI. The proportion of patients underwent PCI might be partly contributed to the e cacy difference of the ticagrelor treatment among Asian countries. Additionally, baseline characteristic of included population, diagnostic criteria, and dosage regimen might be the reasons for the differences. The e cacy of ticagrelor among Asian populations needs to be validated by larger trials.
DAPT reduces the incidence of thrombotic events but exposes patients to an increased risk of bleeding. The optimal DAPT duration subsequent to a PCI remains unclear. The safety and e cacy of short-duration (≤ 3 months) DAPT in elderly patients were acceptable [47], and abbreviated-duration (≤ 6 months) DAPT in CAD population did not signi cantly increase the incidence of MACE, but dramatically reduced the risk of major bleeding [48]. In addition, Verdoia et al. reported that, compared to the standard one-year DAPT, a shorter-duration (3 or 6 months) could protect against major CV ischemic events in an improved manner [49]. In this meta-analysis, MACE and bleeding endpoints were comparable between 1-month and 6-month follow-up. The recommended DAPT duration has shifted from 12 months to a more exible approach on the basis of the individual's ischemia and bleeding risk. Thus, balancing the risk of ischemia and bleeding when using DAPT becomes a clinical challenge because patients with a low risk of ischemic events and a high risk of bleeding would bene t from shorter duration [50].
The current study has several limitations. First, the sample sizes of several RCTs and the number of RCTs are relatively small. Due to limited research data or controversial results, the results of bleeding are not robust in sensitivity analyses. Second, the follow-up duration subsequent to PCI varied from in-hospital to one year or longer. Although we performed subgroup analysis in terms of the duration, the total results might also have swayed to some extent. Third, we pooled trials with heterogeneous populations that varied in study design, disease subtype, treatment strategy, and endpoints de nitions. Fourth, possible adverse drug reactions, loss of tolerability, and discontinuation of treatment are not incorporated into the consideration.

Conclusion
We suggest that in ACS patients, ticagrelor is comparable in e cacy to clopidogrel, while it associates with a higher risk of bleeding. Clinicians should selectively adopt ticagrelor and clopidogrel according to different PCI strategy, ethnicities, and countries.

Declarations Acknowledgments
We thank TopEdit (www.topeditsci.com) for its linguistic assistance during the preparation of this manuscript.
Author contribution LLP contributed to the conception or design of the work. SMY, CWC and LLP contributed to the acquisition, analysis, or interpretation of data for the work.
SMY and CWC drafted the manuscript. LLP critically revised the manuscript. All authors gave nal approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Con ict of Interest
The authors declare no competing interests.

Ethical Approval
Not applicable.

Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.

Funding
No funding was received to assist with the preparation of this manuscript. Abbreviations: CV death, cardiovascular death; MI, myocardial infarction, CI, con dence intervals; OR, odds ratios.  Abbreviations: MACE, major adverse cardiac events; MI, myocardial infarction; CI, con dence intervals; OR, odds ratios. Figure 1 Flowchart diagram of searching and screening of studies.

Figure 2
Comparison the primary e cacy outcomes (MACE) between ticagrelor and clopidogrel treatment. Forest plots reporting outcomes in clinical trials (a) and observational studies (b). MACE, major adverse cardiac events.