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. Finally, twenty-seven studies with 269,935 patients (87,988 and 181,947 in the ticagrelor and clopidogrel groups, respectively) were included for the meta-analysis. Nine studies were clinical trials [5, 7-9, 21-25], and 18 were observational studies [10-12, 26-40]. The patients were enrolled from 2003 to 2019, and the articles were published from 2007 to 2020. Among them, nineteen studies included ACS patients underwent PCI, and eight included those intended for PCI. 253,979 patients received PCI. The countries in which these studies were conducted were East Asian countries such as China, Korea, and Japan, as well as European and American countries such as the United States, Canada, Sweden, the Netherlands, and England. In addition, the ethnicities of the cohorts included both East Asians and Caucasians, and the duration of follow-up ranged from one month to 468 days. The main study and population characteristics are summarized in Table 1.
Efficacy endpoints
For the clinical trials, no significant difference was found in the primary efficacy endpoint (MACE) between ticagrelor and clopidogrel groups (OR 0.76, 95% CI 0.54-1.06, p = 0.11, I² = 66.74%; Fig. 2a). Similar results were shown in both the MA group (OR 0.97, 95% CI 0.82-1.15, p = 0.76, I² = 84.18%; Fig. 2b) and the PA group (OR 0.86, 95% CI 0.75-1.00, p = 0.05, I² = 72.32%; Fig. 2b) in observational studies.
The ticagrelor group demonstrated reduced secondary endpoints, compared to the clopidogrel group, regarding ST (OR 0.72, 95% CI 0.58-0.90, p = 0.00, I² = 0.00%; Table 2a) in clinical trials, all-cause death (OR 0.83, 95% CI 0.70-0.98 p = 0.03, I² = 69.89%) and CV death (OR 0.66, 95% CI 0.44-0.99, p = 0.04, I² = 70.59%) in the PA group, and CV death (OR 0.59, 95% CI 0.45-0.79, p < 0.001) in the MA group (Table 2b).
Safety endpoints
Ticagrelor led to significantly higher risks of bleeding (OR 1.49, 95% CI1.14-1.94, p = 0.00, I² = 63.97%) and minor bleeding (OR 1.57, 95% CI 1.08-2.30, p = 0.02, I² = 59.09%) over clopidogrel in clinical trials (Table 2a). The increased risks of bleeding (OR 1.39, 95% CI 1.06-1.83, p = 0.02, I² = 76.11%) and minor bleeding (OR 1.61, 95% CI 1.37-1.89, p = 0.00, I² = 0.00%) were also identified in the PA group of observational studies (Table 2b). However, only the minor bleeding risk (OR 1.21, 95% CI 1.14-1.72, p = 0.007; Table 2b) increased significantly in the MA group of observational studies.
Subgroup analysis
In the subgroup of PCI strategy, patients underwent PCI (OR 0.38, 95% CI 0.23-0.63, p = 0.00, I² = 0) benefited more from ticagrelor than those intended for PCI (OR 1.02, 95% CI 0.70-1.49, p = 0.93, I² = 68.99%) in MACE (Fig. 3a), but ticagrelor introduced a higher risk of bleeding in the patients either underwent PCI (OR 1.64, 95% CI 1.07-2.51, p = 0.02, I² = 0) or intended for PCI groups (OR 1.46 95% CI 1.05-2.02, p = 0.02, I² = 71.56%; Fig. 3a).
Subgroup analysis based on different ethnicity was performed to study the clinical outcomes of ticagrelor and clopidogrel between Caucasian and East Asian populations. Ticagrelor showed a superior MACE reducing effect over clopidogrel in Caucasian patients (OR 0.84, 95% CI 0.75-0.94, p = 0.00, I² = 0; Fig. 3b). Meanwhile, it was related to a lower risk of bleeding (OR 1.09, 95% CI 0.99-1.20, p = 0.07, I² = 0; Fig. 3b). However, the results were inconsistent in East Asian populations. Ticagrelor was comparable with clopidogrel regarding MACE (OR 0.67, 95% CI 0.36-1.25, p = 0.21, I² = 77.78%; Fig. 3b) and related to a higher bleeding risk (OR 1.81, 95% CI 1.43-2.29, p = 0.00, I² = 0; Fig. 3b).
Subgroup analysis based on different Asian countries showed that Chinese patients benefited more from ticagrelor treatment than those in Korean and Japanese, while the risk of bleeding significantly increased in all three Asian countries (Fig. 3c). Further subgroup analyses were conducted to analyze the safety and efficacy of ticagrelor in different follow-up duration classifications. It showed that bleeding risk and MACE was comparable between the two groups during the follow-up duration (Online Resource ESM_6).
Sensitivity analyses and publication bias
Sensitivity analyses were performed by including high quality RCTs. The results remained consistent, except for bleeding (OR 1.57, 95% CI 0.97-2.53, p = 0.06, I² = 84.86%; Online Resource ESM_8).
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-fill analysis showed that five studies were filled for MACE with the total results influenced, and two were filled for MI with the total results unaffected (Online Resource ESM_9, Table 4).