A Propensity-Matched Nested Case-Control Study of Acute Coronary Syndrome Patients Genotyped for CYP2C19

Introduction Ticagrelor is widely considered superior to clopidogrel however a pharmacogenetic substudy of PLATO indicated that the majority of this difference is due to genetic nonresponders to clopidogrel. We evaluated patient outcomes following genotyping for CYP2C19 in a propensity matched acute coronary syndrome cohort treated with either clopidogrel, ticagrelor or aspirin monotherapy. Methods ICD10 coding identified 6,985 acute coronary syndrome patients at Waitemata District Health Board over a five year period (2012-2016). Ticagrelor was subsidised by The Pharmaceutical Management Agency of New Zealand in July 2013. Patients were genotyped for CYP2C19 *2, *3 and *17 alleles using the Nanosphere Verigene analyser and treatment was tailored accordingly. Logistic regression and nearest neighbour propensity matching was employed in a 1:3 fashion with each treatment group to balance patient characteristics. Results A total of 146 patients were genotyped and compared with 438 matched patients taking either clopidogrel, ticagrelor or aspirin monotherapy. Post July 2013 clopidogrel was prescribed more often in responders than in those without genotype information (68 vs 39%, X2 9, 95% CI 4 to 34, p=0.003). Conversely, ticagrelor was used more frequently in clopidogrel nonresponders. Mortality with personalised treatment was equivalent to ticagrelor (HR 0.8, 95% CI 0.3 to 1.8) but higher in those treated with clopidogrel (HR 2.3, 95 % CI 1 to 5.3). Readmissions with ACS were higher in nonresponders treated with clopidogrel versus those treated with genotype appropriate dual antiplatelet therapy (HR 3.9, 95% CI 0.8 to 18, p =0.03). Conclusion Personalised antiplatelet management was equivalent to ticagrelor with respect to all-cause mortality and ACS readmissions. It also led to more appropriate use of both clopidogrel and ticagrelor and potential cost savings.