To the best of our knowledge, this is the first systematic review and meta-analysis to investigate the association between aspirin use and the prognosis of nonobstructive CAD. Based on the pooled data from thirteen studies with 34,463 participants, this study revealed that routine aspirin therapy did not decrease the risk of MACEs, cardiovascular death and MI, but reduced the risk of all-cause death.
The benefit of aspirin therapy is well established for secondary prevention in patients with atherosclerotic diseases, and a low-dose of aspirin is recommended by the guidelines of the European Society of Cardiology. Unfortunately, in contrast to previous randomized trials that majorly recruited patients with obstructive CAD, few randomized trials have probed the association between routine aspirin use and cardiovascular events in patients with nonobstructive CAD. Increasing evidence indicates that patients with nonobstructive CAD already develop slight atherosclerotic alterations with a higher risk of cardiovascular events[10, 11]. As a result, more aggressive management might be needed for these patients. At present, the recommendations of guidelines for aspirin therapy in patients with nonobstructive CAD are ambiguous. Indeed, the extreme variation in aspirin prescription rates (29.0–90.1%) in our study corroborated the dilemma of aspirin use in nonobstructive CAD. Therefore, evidence from our study that aspirin therapy failed to improve the prognosis of nonobstructive CAD might shed light on the management of this disease with aspirin in the future.
Several studies have assessed the impact of antiplatelet therapy on cardiovascular events in patients with nonobstructive CAD. In patients diagnosed with MINOCA, an observational study reported that aspirin use was not associated with a significantly reduced risk of the composite of adverse cardiovascular events (odds ratio 0.601; 95% CI 0.305–1.183) during 2-year follow-up. Similarly, another study with a 3-year follow-up showed that antiplatelet agents did not improve the prognosis of nonobstructive CAD (HR 1.089; 95% CI 0.642–1.847). Because of the different effect sizes and lack of detailed antiplatelet regimens, these two studies were excluded from our meta-analysis. Another two studies that enrolled participants without a medical history of MI also found that aspirin therapy had no beneficial effect[20, 28], which was consistent with our findings. However, these studies were insufficiently representative, as participants with MI were either included or excluded. Our combined data showed that aspirin therapy did not improve the prognosis of patients with nonobstructive CAD regardless of whether they suffered from baseline MI or not, which was first reported (Table 3).
Of note, according to current guidelines, once a diagnosis of MI is confirmed, aspirin as well as P2Y12 receptor antagonists should be routinely administered[36, 39]. However, there is a debate as to whether antiplatelet agents should be routinely prescribed to patients with MINOCA. Luis et al. reported that insufficient secondary prevention medications (e.g., aspirin) were responsible for the poor outcome among MINOCA patients with “normal” cardiac magnetic resonance images. On the contrary, a post-hoc analysis of the CURRENT-OASIS 7 trial showed that potent antiplatelet therapy might increase the risk of cardiovascular events. The pooled data from our study indicated that aspirin therapy was not associated with improved prognosis in patients with nonobstructive CAD. Consistently, the results of dual antiplatelet therapy data from four studies could further strengthen this conclusion.
Our study showed that aspirin use reduced the risk of all-cause death in patients with nonobstructive CAD, with no effects on cardiovascular death or MI. Post-hoc sensitivity analyses of all-cause death indicated a reversible result after excluding the study by Hwang et al (Supplementary Fig. S3). In their study, more than half of aspirin users did not receive statin therapy, which suggested that the beneficial effects of aspirin on all-cause death might be influenced by a lower prescription rate of statins (23.7%). Currently, statins are widely available at a low price, and the effect of statins is superior to that of aspirin in primary prevention. Multiple studies have shown that statins can perform well in the management of patients with nonobstructive CAD[20, 23, 28, 30, 31]. The efficacy of aspirin is likely to be attenuated or covered because of the gradually increasing prescription rate of statins. Hence, the association between routine aspirin therapy and all-cause death remains unclear in patients with nonobstructive CAD under the comprehensive management of risk factors, such as smoking, diabetes, dyslipidemia, and hypertension, which requires further investigation.
Nonobstructive CAD is a heterogeneous disease with multiple potential underlying conditions, including vulnerable plaque, CAS, microvascular dysfunction and myocarditis, which may explain why aspirin use was not associated with improved outcomes.
First, critical findings might be ignored when evaluating coronary stenosis with coronary angiography alone. Intravascular imaging showed that vulnerable plaques, such as plaque rupture and plaque erosion, were detected in a few patients with nonobstructive CAD, which could eventually develop into thrombi[42, 43]. It was reasonable to prescribe antiplatelet agents to patients with thrombi. Although antiplatelet therapy based on intravascular imaging should be the target, in actual clinical practice it is not readily available due to the cost–benefit ratio. Additionally, it was showed that plaque burden, rather than vulnerable plaque, was a stronger predictor for patients with nonobstructive CAD. The plaque burden was also associated with similar ischemic events, regardless of the presence or absence of obstructive CAD. Indeed, coronary stenosis is positively correlated with plaque burden. Overall, patients with nonobstructive CAD had a lower plaque burden than those with obstructive CAD. This might explain why aspirin therapy was not as effective as expected in patients with nonobstructive CAD. On the other hand, those with a higher plaque burden may benefit from antiplatelet therapy. For example, two studies showed that only patients with a coronary artery calcium score of ≥ 100 were likely to benefit from aspirin therapy in the primary prevention of atherosclerotic disease[46, 47]. Therefore, future research is needed to support our hypothesis that a personalized antiplatelet regimen based on plaque burden might contribute to reduction in ischemic cardiovascular events in patients with nonobstructive CAD.
Second, a previous study showed that CAS accounted for approximately 60% of patients with stable angina, indicating that it is a common condition in patients with nonobstructive CAD. Similarly, a systematic review revealed that approximately 30% of MINOCA might be induced by CAS. The presence of CAS was associated with poor outcomes in these patients[8, 49]. However, limited evidence is available regarding the effects of aspirin therapy on CAS. A high dose of aspirin (> 325mg daily) was considered as a potential inducer of coronary spasm, whereas a low dose of aspirin had the opposite effect. Conversely, another study found that aspirin therapy might lead to CAS even at a low dose. Hence, the impact of aspirin therapy on the prognosis of CAS was under ongoing debate. A recent meta-analysis showed that there was no significant association between aspirin use and clinical outcomes among patients with vasospastic angina without organic stenosis. In the present study, pooled data based on a subset of CAS also indicated that aspirin use was not associated with the risk of cardiovascular events (Table 3). Of note, this finding should be interpreted with caution because of the post-hoc nature of subgroup analysis.
Third, recent studies indicated that patients with nonobstructive CAD could present with angina-like chest pain, but there was no evidence of myocardial ischemia. Several non-ischemic diseases, including myocarditis, cardiomyopathy, and Takotsubo syndrome, can mimic MI in patients with MINOCA. Previous studies have shown that routine aspirin use does not reduce the risk of cardiovascular events in non-ischemia settings[55, 56]. In addition, a meta-analysis reported that aspirin therapy might worsen the long-term prognosis of patients with Takotsubo syndrome without any benefit.
There are several limitations of our study. First, the definitions of the primary endpoint differed slightly between the studies, which might be a potential source of heterogeneity in our results. Second, our results were extracted from observational cohort studies, most of which were retrospective. In addition, a small amount of the extracted data was unadjusted. Although subgroup analyses based on study design and adjustment revealed no interaction effects, our results might still be affected by potential confounding factors. Third, the lack of reported haemorrhagic events might affect the outcome in patients on aspirin therapy. Finally, only published studies were included in the meta-analysis. The results might be corrected by unpublished studies, despite no evidence of publication bias. Currently, a prospective, multicenter, randomized trial is in progress, in which 4422 women suspected of myocardial ischemia with nonobstructive CAD have been recruited. This trial may provide additional information about the efficacy of aspirin therapy (NCT03417388).