Peng Yin1, Andrea L Jorgensen1, Andrew P Morris1, Richard Turner2, Richard Fitzgerald2, Rod Stables3, Anita Hanson2, Munir Pirmohamed2 1. Department of Biostatistics, University of Liverpool 2. Department of Molecular & Clinical Pharmacology, University of Liverpool 3. Liverpool Heart and Chest Hospital SNP-treatment interactions of cardiovascular medications and risk of acute coronary syndrome recurrence Introduction Results Table 1: Clinical risk factors for acute coronary syndromes (p < 0.05) To identify genetic loci associated with response to cardiovascular drugs, we have undertaken a genome-wide association analysis (GWAS) of 1470 patients recruited to a large, UK prospective pharmacogenetic study of acute coronary syndrome (PhACS), adjusting for patient demographics and clinical factors including drug type. Approximately 14% (201/1470) of the patients had a recurrent cardiovascular event during the follow up, including myocardial infarction (MI) or stroke, and in some cases death. Figure 3: Regional plot for EVC2 locus Hazard ratio (95% CI) P value Age 1.047 (1.032, 1.062) 2.1x10-10 BMI 1.022 (1.004, 1.040) 0.016 Prior MI 1.955 (1.444, 2.647) 1.43x10-5 ACEI pre-admission 1.571 (1.166, 2.117) 0.0030 Aldosterone 2.119 (1.300, 3.453) 0.0026 Non-ST elevation ACS admitted to hospital Patient Demographics MI/Stroke/CVA Death (201/1470) Drugs SNPs Baseline Hospital D/C Figure 4: Regional plot for CNR1 locus We defined the primary outcome as time to recurrence of any cardiovascular events (MI, stroke, death) during up to 5 years following hospital discharge. Demographics and clinical risk factors were first of all tested for association with this outcome in a Cox proportional hazard (PH) regression framework (Table 1). Patients were genotyped using the lllumina Omni Express array and the genotypes underwent quality control, using the following criteria: sample call rate > 95%; SNP call rate > 95%, minor allele frequency > 5%, Hardy-Weinberg test p value > 0.0001. We also checked gender discrepancies between clinical and genetic data, heterozygosity and relatedness between samples. After applying genotype quality control, the genotype scaffold was imputed up to the 1000 Genomes Phase I reference panel (all ancestries, March 2012 release). We tested for SNP-treatment interaction for each cardiovascular drug under an additive dosage model after adjusting for clinical risk factors, main effects of SNP and treatment, and principal components to account for population structure (Figure 1). Methods Figure 1: Manhattan plot of SNP-Clopidogrel interactions with time to acute coronary syndromes in 1357 individuals Implication We identified an intergenic variant (rs4936159) with genome-wide significant (p<5x10-8) evidence of interaction with clopidogrel: minor allele frequency: 0.05 p-value: 1.2 x10-8 Hazard ratio (95% CI): 19.3 (7.0-53.3) The variant maps near NCAM1, which is involved in left ventricular remodelling, and seems to be associated with ischemic damage, Calcium signalling in cardiomyocytes. Nominal evidence of clopidogrel interaction was also observed for variants mapping in/near: EVC2, CNR1, SV2B, LIG3/RFFL. Three of those candidate genes involved in G protein-coupled receptor (GPCR) signalling and may be worth of further investigation. Figure 2: Region plot for NCAM1 locus Our study highlights several biological candidate genes that are associated with response to cardiovascular drug Clopidogrel in ACS recurrence time: NCAM1, EVC2, CNR1, SV2B, LIG3/RFFL. Conclusions Acknowledgement We wish to acknowledge funding from Wellcome Trust (grant number WT098017); the UK Department of Health (NHS Chair of Pharmacogenetics). We also acknowledge all patients, clinicians, research nurses and researchers involved in the study. Correspondence about the subject matter of this poster is very welcome. Please e-mail Peng Yin peng.yin@liv.ac.uk Website: https://www.liverpool.ac.uk/translational-medicine/research/statistical-genetics/about/