ORIGINAL ARTICLE Genome-wide association study on antipsychotic-induced weight gain in the CATIE sample EJ Brandl 1,2,8 , AK Tiwari 1,8 , CC Zai 1,3 , EL Nurmi 4 , NI Chowdhury 1 , T Arenovich 1 , M Sanches 1 , VF Goncalves 1,3 , JJ Shen 5 , JA Lieberman 6 , HY Meltzer 7 , JL Kennedy 1,3 and DJ Müller 1,3 Antipsychotic-induced weight gain (AIWG) is a common side effect with a high genetic contribution. We reanalyzed genome-wide association study (GWAS) data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) selecting a refined subset of patients most suitable for AIWG studies. The final GWAS was conducted in N = 189 individuals. The top polymorphisms were analyzed in a second cohort of N = 86 patients. None of the single-nucleotide polymorphisms was significant at the genome-wide threshold of 5x10 - 8 . We observed interesting trends for rs9346455 (P = 6.49x10 - 6 ) upstream of OGFRL1, the intergenic variants rs7336345 (P = 1.31 × 10 - 5 ) and rs1012650 (P = 1.47 × 10 - 5 ), and rs1059778 (P = 1.49x10 - 5 ) in IBA57. In the second cohort, rs9346455 showed significant association with AIWG (P = 0.005). The combined meta-analysis P-value for rs9346455 was 1.09 × 10 - 7 . Our reanalysis of the CATIE GWAS data revealed interesting new variants associated with AIWG. As the functional relevance of these polymorphisms is yet to be determined, further studies are needed. The Pharmacogenomics Journal (2016) 16, 352–356; doi:10.1038/tpj.2015.59; published online 1 September 2015 INTRODUCTION Antipsychotic-induced weight gain (AIWG) remains one of the major challenges in treating patients with schizophrenia, schi- zoaffective disorder and other severe psychiatric disorders. Especially second-generation antipsychotics such as olanzapine or clozapine can cause severe weight gain leading to metabolic disorders and early cardiovascular death. 1 Despite the fact that some medications such as perphenazine or ziprasidone have a lower risk for AIWG and extensive efforts to develop treatment strategies such as finding new medications or using comedication with metformin and exercise programs, 2 there is currently no medication with high clinical efficacy without any risk to cause this serious side effect. Biological and clinical factors underlying AIWG are only partially understood. Marked interindividual variability in the amount of AIWG makes predictive tests highly desirable; however, to date, no test with sufficient predictive value for a meaningful risk estimate before treatment start could be established. Genetic factors are known to influence an individual’s risk for AIWG (for review, see Lett et al. 3 and Shams and Mueller 4 ). Although candidate gene studies have delivered some interesting findings such as the HTR2C gene, 5 other findings are yet to be replicated in indepen- dent samples limiting their translation into clinical care. A recent genome-wide association study (GWAS) in children and adolescents identified an important new risk variant near the melanocortin 4 receptor gene with replication in several inde- pendent samples. 6–8 Therefore, GWAS seem to be a promising approach to achieve a better understanding of genetics of AIWG. One of the largest clinical trials conducted for comparing the effectiveness of antipsychotic drugs is the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE). This study showed high rates of non-compliance, with 74% of patients discontinuing their medication within the 18 months study duration. Depending on the medication assigned in the trial, 10 (risperidone group)–18% (olanzapine group) of patients stopped taking the prescribed medication owing to side effects, and weight gain was among the leading causes of discontinuation, especially in patients taking olanzapine with discontinuation because of AIWG in 9%. 9 Several previous studies have investigated genetics of meta- bolic parameters in the CATIE sample. 10–12 However, the CATIE study was not primarily designed as a pharmacogenetic study to test metabolic parameters. Therefore, chronic patients of mixed ethnicities, previous medications with high risk for AIWG, patients with weight gain-inducing comedication and patients receiving medication with no or very low propensity for AIWG were included in the trial. All these clinical factors may lead to underestimation of AIWG and may limit the meaningfulness of genetic association findings in this sample. Need et al. 10 investigated association of 118 candidate genes with 21 phenotypes of antipsychotic response, including weight gain as a quantitative (maximum weight gain in % at any point in phase 1) or qualitative trait (weight gain 47% of baseline yes or no), in N = 756 individuals. Despite some suggestive findings in several genes, no variant showed significant association with AIWG after multiple test correction. Although the authors used EIGENSTRAT to control for population stratification and covaried for self-reported ethnicity, they included patients of mixed ethnicities in 1 Pharmacogenetics Research Clinic, Campbell Family Mental Health Research Institute, Centre for Addiction and Mental Health, Toronto, ON, Canada; 2 Department of Psychiatry and Psychotherapy, Charité Universitätsmedizin Berlin, Berlin, Germany; 3 Department of Psychiatry, University of Toronto, Toronto, ON, Canada; 4 Department of Psychiatry and Biobehavioral Sciences, University of California at Los Angeles, Semel Institute for Neuroscience, Los Angeles, CA, USA; 5 Department of Pediatrics and Adolescent Medicine, The University of Hong Kong, Hong Kong, China; 6 Department of Psychiatry, College of Physicians and Surgeons, Columbia University and the New York State Psychiatric Institute, New York City, NY, USA and 7 Department of Psychiatry and Behavioral Sciences, Northwestern Feinberg School of Medicine, Chicago, IL, USA. Correspondence: Dr DJ Müller, Department of Psychiatry, Centre for Addiction and Mental Health (CAMH), University of Toronto, 250 College Street R132, Toronto, ON M5T1R8, Canada. E-mail: daniel.mueller@camh.ca 8 These authors contributed equally to this work. Received 17 May 2015; accepted 1 July 2015; published online 1 September 2015 The Pharmacogenomics Journal (2016) 16, 352 – 356 © 2016 Macmillan Publishers Limited All rights reserved 1470-269X/16 www.nature.com/tpj