In the present meta-analysis, we found the following points: (1) Compared with S-DAPT, treatment with D-DAPT retains P2Y12 inhibitor was associated with lower risks of MACE and major bleeding complications; (2) Compared with S-DAPT, D-DAPT retains P2Y12 inhibitor was associated with a lower risk of major bleeding, but this was only observed among East Asians. Among non-East Asians, the rate of major bleeding was similar between two groups. (3) There were also no significant associations between D-DAPT retains P2Y12 inhibitor treatment with cardiovascular death, all-cause death and individual cardiovascular outcomes among East Asians and non-East Asians. Based on these results, compared with standard DAPT strategy, de-escalation of DAPT retains P2Y12 inhibitor treatment might be considered as an alternative DAPT strategy for East Asians.
It is well known that East Asians who received dual antiplatelet therapy after PCI have a higher risk of bleeding and a lower risk of ischemia[20]. This may be due to the lower body mass index (BMI) of the East Asians compared with non-East Asians. Previous studies have shown that obesity is associated with thrombosis[21], and it’s a prethrombotic state that could cause a series of changes in the body to promote the formation of thrombus[22]. This may be one of the reasons why the risk of ischemia in East Asians is lower than that in non-East Asians. Secondly, the genetic polymorphism between different races may also explain the profile of East Asians. Previous Multi-Ethnic Study of Atherosclerosis (MESA) studies [23]have shown that Blacks have highest levels of dysfunctional endothelial profile (such as factor VIII, D-Dimer, plasmin–antiplasmin, and von Willebrand factor), so they have the highest risk of thrombotic events, followed by Caucasians and Hispanics, and finally the Chinese participants. In addition, previous studies have shown that the level of inflammation can also affect thrombosis, and the level of inflammation in East Asians is lower than that in non-East Asians[24], which may also explain the low risk of ischemia in East Asians.
Previously, Bianco et al. (4 RCTS, 29089 patients) indicated that after short-term DAPT followed by P2Y12 inhibitor monotherapy was associated with a lower risk of clinically relevant bleeding [odds ratio (OR) = 0.70, 95%CI: 0.58–0.86] for patients undergoing PCI, as compared with 12 months DAPT, without an increasing risk of 1-year cardiovascular events (OR = 0.90, 95%CI: 0.79–1.03) [25]. Similarly results was observed in the analysis of Michelle et al. (5 trials, 32,145 patients) [26]. But none of them included the latest research. Moreover, the population of these studies is very wide, and they did not consider the differences of ethnicity. Our research confirmed that for East Asians, de-escalation strategy retains P2Y12 inhibitor was associated with a lower risk of major bleeding, but this was not observed among non-East Asians.
During one-to-three months after stenting, which belongs to the high incidence phase of ischemic events after PCI, the thrombotic risk outweighs the bleeding risk. While bleeding events generally occurred during a longer period after stent implantation, which belongs to the chronic phase. From the studies we included, it was noted that most of the de-escalation strategies start at one or three months after receiving dual antiplatelet therapy. As mentioned before, East Asians have a profile of high risk of bleeding. Therefore, in the chronic phase, downgrading of dual antiplatelet therapy, so as to reduce the degree of platelet inhibition, might reduce the risk of bleeding events among East Asians[7–10]. This is consistent with the results of our meta-analysis. Among non-East Asians, we found that the effects of de-escalation of DAPT and standard DAPT on major bleeding events were similar and no significant difference. Although the TWILIGHT study[12] which across both East-Asians and non-East Asians showed downgrading therapy could reduce bleeding events with BARC ≥ 2 in Asians and Caucasians, de-escalation of DAPT cannot reduce the risk of major bleeding (HR = 0.49, 95CI%: 0.33–0.74). What’s more, the study did not report the primary interest outcomes in East-Asians, the effect on major bleeding events between East-Asians and non-East Asians was not clear. And our sensitivity analysis showed that whether or not the study is removed, there is no significant effect on major bleeding outcomes of the non-East Asians subgroup. Therefore, the results of our meta-analysis can be considered reliable.
However, the results should be interpreted with caution. First, different definition might influence the incidence of outcomes, including MACE and major bleeding events. Second, various de-escalation strategies among different races included in this analysis might affect the pooled analysis results. Thirdly, the TWILIGHT trial included both East-Asians and non-East Asians. Due to lack of patient level data, we pooled all the data into non-East Asians group.
Limitations: This analysis has certain limitations. First, the rates of ischemic events were lower than anticipated in most of the included trials, resulting in limited statistical power for ischemia outcomes. Secondly, the included patients composed of both ACS and SCAD, and owing to lack of individual level data, we could not perform subgroup analysis for both groups. Thirdly, adherence of agents was not noted, which was a recognized and common factor associated with long-term outcomes. Fourth, the race was judged by the sites of the participating studies. Therefore, the possibility of the race mix-up was not excluded in those studies. And since patient category (East Asian versus other) is totally dependent upon study, it is possible that the racial differences are really just differences among the studies. Fifth, the current analysis just focused on major bleeding events, but minor bleeding was not reported, which more likely resulted in lower adherence of treatment in clinical practice. Finally, although we strictly performed the study searching and selection, some extent of potential publication or selection bias cannot be neglected.