Stroke is a devastating clinical event associated with substantial mortality and morbidity [1]. Patient with coronary artery disease (CAD) always have a high prevalence of stroke due to concomitant atherosclerotic disease of the cerebral vascular system or cardiogenic embolism [2, 3]. Nevertheless, the pathophysiology and causes of stroke are more diverse than those in ischemic coronary syndromes. Either ischemic or hemorrhagic stroke can cause more deterioration in the quality of life compared with other ischemic or bleeding events, even if patients who survive in the acute period [4]. Any type of stroke is thought to result in a life-long reduction in utility and have a much greater impact on the quality of life, regardless of the severity of stroke [5‐9].

Antiplatelet and anticoagulant treatments play pivotal roles throughout the prevention of cardiovascular and cerebrovascular disease. Dual antiplatelet therapy (DAPT) has been recommended for patients with acute coronary syndromes (ACS) and those undergoing percutaneous coronary intervention given its benefit in the risk of stent-related and spontaneous recurrent ischemic events [10, 11]. However, considering the improved safety and efficacy of drug-eluting stents (DES) [12, 13] and advances in medical treatment [14‐16], the optimal duration of DAPT in patients with ACS remains controversial. While more powerful antithrombotic strategy might be beneficial to reduction of ischemic stroke (IS), it leads to a higher risk of hemorrhagic stroke (HS) or intracranial hemorrhage (ICH). Compared with spontaneous HS, oral anticoagulant (OAC) related intracerebral hemorrhage has a larger hematoma volume [17] and a worse prognosis [18‐20]. Therefore, it is essential to figure out the safety and efficacy of intensive antithrombotic therapy (ATT) for stroke of CAD populations. Additionally, clinical evidence supported antiplatelet for non-cardioembolic stroke prevention, while anticoagulant is more recommended for the prevention of most types of cardioembolic stroke [21]. However, whether antiplatelet therapy and OAC yield the same benefit in the risk of stroke among CAD population is not clear.

Therefore, we conducted a systematic review and meta-analysis to investigate whether escalation of ATT, including antiplatelet therapy and OAC, is beneficial in different types of the stroke among patients with CAD.

A total of 8259 articles were retrieved after duplication removal, of which 264 articles warranted full-text review for detail. We finally identified 42 studies (301,547 enrolled patients) that met the inclusion criteria and provided at least 1 endpoint of interest (Fig. 1). Among 42 RCT involved patients referred for CAD, 15 studies (153,856 enrolled patients) were randomized after diagnosis of ACS, which combined unstable angina, Non-ST-elevation myocardial infarction, and ST-elevation myocardial infarction; 2 studies (37,498 enrolled patients) were about stable CAD; 23 studies (89,569 enrolled patients) were patients with ACS or stable CAD undergoing percutaneous coronary intervention, including 18 with DES, 1 with bare-metal stent [25], 3 with DES or bare-metal stent [26‐28], and 1 was unreported [29]. The remaining 2 studies incorporated 1 specifically for CAD accompany heart failure (HF) [30], 1 for cardiovascular disease or multiple risk factors [31]. One study enrolled only veterans [32]. The quality assessment and characteristics of included studies are presented in Additional file 1: Tables S1 and S2.

Compared with conservative ATT, intensive ATT was associated with a significantly lower risk of all stroke (RR 0.86, 95% CI 0.80–0.94; P = 0.001) (Fig. 2). There was a high between-study heterogeneity within OAC group (I² = 59.1%, P = 0.023), but no heterogeneity within antiplatelet group (I² = 0.0%, P = 0.571). Intensive ATT also reduced the risk of IS (RR 0.80, 95% CI 0.71–0.91; P = 0.001) (Fig. 3), but increased the risk of HS (RR 1.36, 95% CI 1.00–1.86; P = 0.051) (Fig. 3) and ICH (RR 1.46, 95% CI 1.17–1.81, P = 0.001) (Additional file 1: Fig. S1). Subgroup analyses found that OAC yielded more benefit in all stroke than antiplatelet therapy (OAC: RR 0.73, 95% CI 0.58–0.92, P = 0.006; Antiplatelet therapy: RR 0.90, 95% CI 0.83–0.97, P = 0.004; P value for between-group heterogeneity = 0.030). There was no significant difference in the treatment effect of IS, HS, or ICH between antiplatelet and OAC subgroups (Table 1).

Subgroup analyses showed that the effects of antiplatelet therapy were mainly driven by long-term DAPT. Long-term DAPT was of obvious benefit on all stroke (RR 0.86, 95% CI 0.78–0.95; P = 0.002) and IS (RR 0.84, 95% CI 0.75–0.94; P = 0.002), but had no advantage over HS (RR 0.96, 95% CI 0.62–1.49; P = 0.858) and even increased the risk of ICH (RR 1.37, 95% CI 1.08–1.75; P = 0.011) compared with short-term DAPT (Table 1). Novel P₂Y₁₂ inhibitor did not show a significant benefit to any type of stroke. Pre-defined subgroup analysis based on the type of CAD showed that intensive ATT significantly reduced all stroke in both ACS (RR 0.89, 95% CI 0.79–0.99; P = 0.033) and non-ACS (RR 0.73, 95% CI 0.63–0.85; P < 0.001) populations, and there is no heterogeneity between two groups (P value for between-group heterogeneity = 0.066) (Table 2). Meta-regression found no study-level covariates which explained the variability of all stroke, IS or HS. No apparent systematic bias was found, and no individual study unduly influenced the effects estimates in the sensitivity analyses.

We presented a meta-analysis of all published RCT evaluating the stroke outcomes of intensive versus conservative ATT involving 301,547 CAD patients, with an average follow-up of 20.6 months. We found that among patients with CAD who have already received antiplatelet therapy, intensive ATT, either escalation of antiplatelet therapy or addition of OAC, significantly reduced the risk of all stroke and IS, but increased the risk of HS and ICH compared with conservative ATT. OAC was more effective than antiplatelet therapy in the prevention of all stroke.

Stroke has a profound impact on mortality and morbidity given its high risk of death and irreversible sequelae which explicitly decrease the quality of life. Furthermore, brain–heart interactions leading to post-stroke cardiac injury called “stroke-heart syndrome” (SHS) including acute MI, HF, AF, and sudden cardiac death [33]. Therefore, while it is a rare event, small absolute differences in stroke are clinically significant. In this analysis, we found that among CAD patients who have already received antiplatelet therapy, the escalation of either OAC or antiplatelet therapy significantly reduced the risk of IS, but was accompanied by an increase of HS and ICH. It is not surprising that more intensive ATT results in a lower risk of ischemia and a higher risk of bleeding.

Although it has been well established that OAC is more effective than antiplatelet therapy in the prevention of stroke among patients with AF [34, 35], the relative benefit of OAC versus antiplatelet in stroke among CAD patients is not yet clear. In a Cochrane review analysis, there was no difference between vitamin K antagonists versus antiplatelet therapy in the outcome of any recurrent stroke among patients with presumed arterial origin [36]. Similarly, in the two recent trials, neither rivaroxaban nor dabigatran was found to be superior to aspirin in preventing recurrent stroke after embolic stroke of undermined source [37, 38]. In our study, we found that among CAD patients who have already received antiplatelet therapy, although both OAC and enhanced antiplatelet therapy significantly reduced the risk of all stroke, OAC reduced an extra 17% occurrence of stroke events compared with antiplatelet therapy. The reason for this difference between OAC and antiplatelet therapy is not known. It has been well-established that OAC effectively prevents ‘red’ fibrin clots in areas of reduced or stagnant blood flow, such as the fibrillating left atrium, whereas antiplatelet drugs are effective in preventing ‘white’ platelet clots in areas of high shear stress, such as arterial atherosclerosis thromboembolism [39]. According to the constituent ratio of TOAST classification of stroke, cardioembolic stroke accounted for 21% which is more than large-artery atherosclerosis stroke accounted for 18% [40]. Although we excluded studies that exclusively enrolled patients with AF or other diseases who need to receive long-term OAC treatment, not all the studies completely excluded patients with AF. Golwala et al. [41] claimed in meta-analysis that dual (DATT) and triple antithrombotic therapy (TATT) are equivalent in preventing cardiovascular events with DATT being safer by approximately halving bleeding risk. While Gragnano et al. [42] did not concur with the conclusive statement since the heterogeneity between the duration of TATT in AF population. Therefore, whether the variety in the proportion of patients with AF as well as stroke sources can explain our finding warrants further study.

With the introduction and widespread adoption in the clinical practice of novel P₂Y₁₂ inhibitors, it has been speculated that clopidogrel may yield less additional inhibition of platelet aggregation and clinical benefit compared with more potent novel P₂Y₁₂ receptor inhibitors. In a post-hoc analysis of Assessment of Dual Antiplatelet Therapy with Drug-Eluting Stents (ADAPT-DES) trial, high platelet reactivity, indicated by clopidogrel responsiveness, was independently predicted increased risk for IS [43]. The magnitude of increase in the risk of IS was greater per lesser degrees of residual P₂Y₁₂ receptor inhibition, which implies that more potent inhibitors of platelet aggregation and activation would reduce the frequency of stroke. However, this assumption has not been supported by clinical evidence. In our subgroup analysis, neither prasugrel nor ticagrelor was found superior to clopidogrel for preventing any type of stroke. Therefore, further research is warranted to determine the optimal antiplatelet regimen for the prevention of stroke in patients with CAD. Additionally, subgroup analysis also demonstrated that the efficacy of antiplatelet therapy was mainly driven by long-term DAPT subgroup, which reduced all stroke by 14% and reduced IS by 16%. And we failed to find significant heterogeneity between ACS and non-ACS patients in any endpoint.

In conclusion, among CAD patients who have already received antiplatelet therapy, either enhanced antiplatelet or anticoagulant treatments significantly reduced all stroke. The therapeutic effect of OAC for all stroke was more obvious than antiplatelet. Whether this extra benefit of OAC versus antiplatelet therapy is consistent between patients with and without AF warrants further study.