Pharmacogenomics of antipsychotic-induced
weight gain & metabolic syndrome
Antipsychotic-induced weight gain (AIWG) is a common and severe side effect which can result in obesity and related medical conditions (Lett et al., Molecular Psychiatry 2012). The primary goal of our study is to develop a genetic risk model to predict AIWG, by combining hypothesis-free, genome-wide association study (GWAS), pathway analyses and Machine Learning in large, collaborative samples. Our secondary goal is to investigate the clinical relevance of AIWG risk models by examining weight gain, health care utilization, and treatment costs. Early identification of those at risk will facilitate the development of preventive interventions and guide clinical treatment algorithms to advance precision medicine.
Antipsychotic (AP) drugs (e.g., clozapine, olanzapine) are the mainstay pharmacological treatment for schizophrenia (SCZ) and related psychotic disorders. One serious and common side effect is AIWG which causes obesity, metabolic syndrome and premature death in many individuals in addition to be one of the leading causes of patient non-compliance. Family and twin studies strongly support the involvement of genetic factors at the onset of AIWG (Gebhardt et al., 2010). Prior efforts in our own group have identified several risk alleles associated with AIWG which were independently replicated, such as the melanocortin-4-receptor (MC4R) gene (Malhotra et al., 2012) and a variant upstream of the OGFRL1 gene (Brandl et al., 2016). Results to date are consistent with the notion that genetic risk plays a key role in weight gain caused by frequently used antipsychotics.
These previous studies have demonstrated the feasibility of identifying relevant, replicable gene variants associated with AIWG and extensively contributed to the existing understanding of AIWG. However, a more comprehensive approach that combines large well-phenotyped datasets, incorporates genome-wide data and leverages advanced statistical methods is warranted. In our current project, we aim to expand our investigations through the assembly and genotyping of various well-characterized samples with sufficient power to detect novel gene variants associated with AIWG.
Populations: Samples deriving from various projects will be combined to create a sample of well-characterized adults (n > 5,000) and children from the US, Canada and Europe, treated with antipsychotic medications. An important facet of our study is that we will include a large number of first episode patients who have not previously been exposed to antipsychotics.
Aim 1a: Conduct a GWAS analysis in a large combined sample of AIWG
To this end, we will run genome-wide association using linear mixed models corrected for: duration of prior treatment, drugs used and principal components from ancestry investigations. Additionally, we will conduct secondary exploratory analyses in ethnic subgroups (e.g., African-Americans) and pediatric populations, which are less studied in AIWG.
Aim 1b: Pathway analysis and variant prioritization
To better understand the biological mechanisms that underlie AIWG, we will perform pathway analysis, followed by prioritization of genes based on gene expression and reported GWAS hits to define genomic regions of interest for AIWG.
Aim 2: Develop a genetic risk model for AIWG prediction incorporating our identified susceptibility variants
For this step, an individual risk prediction for AIWG will be generated using identified risk genes from 1 and investigated in replication samples. By using Machine Learning algorithms, we will combine significant risk variants, gene-gene interactions, gene-environment interactions, clinical and demographic risk factors into proposed model. (See Figure 1 as exemplary and preliminary model).
Funding: Development, validation and benefits of a genetic risk model for antipsychotic- induced weight gain; Canadian Institutes of Health Research, Operating Funds Award (2015 - 2020)