High-dose statin therapy and the risk of haemorrhagic stroke in Asian patients with stable coronary artery disease: insights from the REAL-CAD study

Abstract

Abstract Background/Introduction The REAL-CAD study identified that aggressive lipid lowering with high-dose statin reduced cardiovascular events also in Japanese patients with coronary artery disease (CAD). However, data from the SPARCL trial found that the benefits of high-dose atorvastatin treatment were partially offset by an increase in haemorrhagic stroke (HS). Although meta-analysis showed statin does not increase HS in Western countries, the evidence about the relation between statin and HS in Asian countries is still conflicting. In addition, the CREDO-Kyoto score is one of the prediction scorings for bleeding after coronary revascularization and might be a useful tool for the prediction of HS in this cohort. Recognizing the risk of HS and predicting of HS in the Asian cohort is clinically important. Purpose This study examined the factors associated with HS using the REAL-CAD cohort. Furthermore, we evaluated the performance of the CREDO-Kyoto bleeding risk score to predict HS in this cohort. We also performed the corresponding analysis of ischaemic stroke for reference purposes. Methods We sub-analysed the REAL-CAD study, prospective, multicentre, randomized, open-label, blinded endpoint study, in which 13,054 Japanese patients with stable CAD were randomized to high-dose (4 mg/day) or low-dose (1 mg/day) pitavastatin. Associations for stroke were determined using competing risk models: the Fine and Gray subdistribution hazards model accounting for the competing risk of death in models of haemorrhagic and ischaemic stroke in REAL-CAD trial. Patients were categorized to low (score 0), moderate (score 1–2), and high (score>3) according to CREDO-Kyoto bleeding score for predicting of HS. Results The HS events in high-dose group tended to be higher than low-dose group (4mg vs. 1mg: 43 (0.7%) vs. 30 (0.5%)). The associated factors of HS on univariate analysis were non-prior myocardial (hazard ratio (HR): 0.62, 95% CI: 0.39–0.99) and non-prior cerebral (HR: 0.25, 95% CI: 0.09–0.70) infarction, atrial fibrillation (HR: 2.4, 95% CI: 1.2–4.7), prior HS (HR: 4.2, 95% CI: 1.5–11.8), anaemia (HR: 2.4, 95% CI: 1.4–4.1), and non-statins use before run-in period (HR: 0.52, 95% CI: 0.28–0.99). High-dose pitavastatin was not a correlate with HS. The multivariate analysis revealed anaemia might have a relation with HS (HR: 4.3, 95% CI: 0.90–20.6). The number of HS was the highest in the high CREDO-Kyoto bleeding score group (Figure 1, HR: 2.4, 95% CI: 1.3–4.6), whereas there was no significant difference in the number of HS between the moderate- and low-risk groups (HR: 1.4, 95% CI: 0.84–2.3). Conclusions High-dose pitavastatin was not associated with the incidence of HS in this large Japanese cohort with stable CAD. High CREDO-Kyoto bleeding score was associated with HS as compared with low or moderate scores, even each of the variables consisting of CREDO-Kyoto score was not associated with HS. Figure 1 Funding Acknowledgement Type of funding source: None