SU102 - GENOME-WIDE ASSOCIATION OF TARDIVE DYSKINESIA

Abstract

Background Tardive Dyskinesia (TD) is a persistent and potentially irreversible condition associated with long term exposure to antipsychotics. Commonly observed in schizophrenia, TD is a movement disorder characterised by involuntary movements of orofacial muscles, trunk and limbs. TD appears to represent a genetically complex phenotype, with candidate gene and genome-wide studies identifying a collection of genes. The pathophysiology of TD is still unclear and the genetic factors predisposing TD across ancestries remain relatively unexplored. Here, we performed a GWAS of TD in a sample of East Asians, European and African-American ancestries. Methods Two available datasets were used: 780 schizophrenia patients of Chinese ethnicity from Singapore and 626 schizophrenia patients of European and African ancestry from the CATIE cohort. Dyskinesia was rated on the Abnormal Involuntary Movement Scale (AIMS) and TD case ascertainment followed the Schooler and Kane criteria. Standard GWAS QC procedures were carried out. Imputation was carried out on the Sanger imputation server; markers were phased via SHAPEIT and imputed via Minimac3 (MACH) to the 1000 Genomes Project Phase 3 reference panel (GRch37). To address complex ancestry effects, linear mixed model association was carried out via GEMMA. Fixed-effects meta-analysis was conducted via GWAMA. Results Meta-analysis revealed one locus on chr16 (rs11639774), which met the GWAS significant threshold (Z=5.551, p=3.09e-8). Three other loci of interests were at subthreshold significance; these were found on chromosomes 1 (rs499646, p=8.48e-8), 6 (rs6926250, p=2.59e-7) and 12 (rs4237808, p=1.1e-7) respectively. Further investigation revealed that these four loci harboured i) downstream gene variant (rs11639774) ii) intronic variants (rs499646, rs6926250) and iii) intergenic variant (rs4237808). Though MAGMA gene-set analysis did not reveal significant gene-based results, one of the top gene-sets appeared to implicate immunoglobulin production. Further gene prioritization of the top 200 associated genes revealed unique brain expression profiles. Notably, several pathways emerged that point to DNA damage and immune responses. Discussion Here we report the largest TD GWAS study to date and evidence for a significant GWAS locus for TD and three other candidate loci of interest. This suggests an inherent vulnerability to TD, which might be brought about by long term antipsychotic exposure. Of interest, the role of immune system in the aetiopathogenesis of TD warrants further research.