PI-021 - A RETROSPECTIVE GENOME-WIDE ASSOCIATION STUDY OF ANTIPSYCHOTIC DRUG RESPONSE IN PEDIATRIC PATIENTS WITH PSYCHIATRIC ILLNESS IN A RURAL HEALTHCARE SETTING.

J. Staples¹, S. Killam¹, K. Brown¹, E. Sather¹, R. Dalton¹, J. Jessop¹, J. Loveland², C. Schwanke², A. Elias², Q. Chen³, K. Bigos⁴, E. Woodahl¹; ¹University of Montana, Missoula, MT, USA, ²Shodair Children's Hospital, Helena, MT, USA, ³Lieber Institute for Brain Development, Baltimore, MD, USA, ⁴Johns Hopkins University, Baltimore, MD, USA.

Background: Risperidone is a commonly prescribed antipsychotic, yet discontinuation rates are high. To identify biomarkers of efficacy in pediatric patients, we conducted a GWAS with EHR phenotype data.

Methods: We obtained microarray and EHR data from 161 patients (128 male and 33 female; age < 18) admitted or initiated on risperidone at a pediatric psychiatric hospital serving rural and underserved patients. We tested 6.3M genotyped and imputed variants (MAF ≥ 5%) and non-genetic factors for association with study endpoints (days to risperidone discontinuation, days to readmittance, length of stay, and max risperidone dose).

Results: Overall, 41% of patients discontinued risperidone. We detected significant associations with the 4 endpoints (p < 3x10-07) in 8 independent loci: 1) days to risperidone discontinuation: 7.158034920.C>T in protein tyrosine phosphatase receptor N2 (PTPRN2); 2) days to readmittance: 1.74774706.AT>A in fucose 1 phosphate guanylyltransferase-troponin I3 interacting kinase (FPGT-TNNI3K) and 19.2459090.G>A in long intergenic noncoding RNA (LINC) 01775; 3) length of stay: 18.68676761.G>C in Gilles De La tourette syndrome chromosome region 1 (GTSCR1)-LINC01541; and 4) max risperidone dose: 2.170444877.C>A in peptidylprolyl isomerase G (PPIG), 15.54535310.G>A in unc 13 homolog C (UNC13C), 17.77582333.A>G in RNA binding fox 1 homolog 3-microRNA 4739 (RBFOX3-MIR4739), and 16.28731469.G>GA in eukaryotic translation initiation factor 3C-EIF3C-like (EIF3C-EIF3CL). In multivariate regression, we explained 25%, 53%, 35%, and 39% of variability in these endpoints, respectively.

Conclusion: These findings provide insight into the biological complexity of risperidone response and may be useful for improving mental health for underserved patients.