P3841Real world outcomes for clopidogrel versus ticagrelor treated patients with acute coronary syndrome: a systematic literature review and meta-analysis

Introduction The ischemic benefit of ticagrelor over clopidogrel seen in randomized trials has not been fully reproduced in real world cohorts. The discrepancy between these two treatments may be influenced by differences in prescribing. Purpose To provide insights in patient care for patients with high levels of comorbidities. Methods A systematic literature review and meta-analysis were conducted to compare clinical outcomes for clopidogrel vs. ticagrelor in real-world populations with acute coronary syndrome. MEDLINE and Embase were searched from inception to January 2018. Pairwise meta-analyses were performed using unadjusted random-effects models for both outcomes and patient characteristics to test for differences. Heterogeneity was assessed using the I2 statistic. Results The search identified 17 studies for analysis from 1,771 publications (TOTAL N=66,198; CLO N=54,175; TIC N=12,023). Clopidogrel-treated patients were more likely to have diabetes (33% vs. 24%), hypertension (59% vs. 47%), hyperlipidaemia/dyslipidaemia (45% vs. 39%), and prior stroke/transient ischemic attack (TIA) (8% vs. 5%). A slightly higher risk of myocardial infarction (MI) was observed for clopidogrel (4.5% vs. 3.5%; OR 1.29 [1.15–1.46]) and included clopidogrel patients had significantly more comorbidities (Figure 1). No statistical difference between clopidogrel and ticagrelor was found for risk of MACE, defined as the composite of cardiovascular (CV) death, MI or stroke, (13.6% vs. 10.1%; OR 1.41 [95% confidence interval 0.93–2.14]), despite clopidogrel patients having significantly more diabetes and hypertension (p≤0.01). Similarly, the risks of CV death (2.3% vs. 2.1%; OR 1.11 [0.58–2.13]), stroke (1.2% vs. 1.1%; OR 1.11 [0.56–2.17]), and stent thrombosis (1.3% vs. 1.0%; OR 1.34 [0.88–2.06]) were not statistically differentiable, despite clopidogrel patients having significantly more hyperlipidaemia in studies analysed for CV death (p<0.05); significantly more diabetes, hypertension, peripheral arterial disease (PAD), hyperlipidaemia, and prior stroke/TIA in studies analysed for stroke (p<0.05); and significantly more hyperlipidaemia and prior stroke/TIA in studies analysed for stent thrombosis (p<0.01). There were no significant differences in all bleeding events (major or minor, as defined by study) (2.0% vs. 2.7%; OR 0.74 [0.54–1.00]) or major bleeding (2.9% vs. 2.7%; OR 1.07 [0.85–1.33]). Included clopidogrel patients had significantly more diabetes, hypertension, PAD, and prior stroke/TIA (p<0.05). Significant heterogeneity (I2 >60%) was observed for MACE, CV death, stroke and bleeding. Figure 1. Comorbidities in MI Conclusion This analysis suggests clopidogrel-treated patients had more comorbidities than ticagrelor-treated patients, which may introduce a bias with negative impact on their outcomes. Despite this circumstance, with the exception of MI, this analysis showed no statistical difference in efficacy and safety between clopidogrel and ticagrelor. Acknowledgement/Funding This analysis was funded by Sanofi.