GENOME-WIDE ASSOCIATION STUDIES OF NON-RESPONSE TO ANTIDEPRESSANTS LEVERAGING HARMONIZED POPULATION-WIDE REAL-WORLD DATA FROM HEALTH REGISTRIES AND BIOBANKS

While recent genome-wide association studies (GWAS) have shown that antidepressant response has a genetic component, the robust identification of genetic associations in current studies is limited by insufficient sample sizes as well as variability in defining treatment-related phenotypes. Through the application of new standardized guidelines (Koch et al. 2025, Lancet Psychiatry), it is now possible to apply real-world biobank and registry data in pharmacogenomics studies. Leveraging the world’s largest population-based genotyped cohorts with longitudinal data (N=2.4 mill individuals) linked to drug prescription registry and other registry data relevant for precision psychiatry, this study represents the largest GWAS on antidepressant treatment outcomes and the first GWAS utilizing harmonized phenotypes across cohorts and countries. In population-based cohorts across nine countries, non-response to selective serotonin reuptake inhibitors (SSRIs) was defined as switching from an SSRI to another antidepressant between 6-14 weeks or augmentation or combination with another antidepressant or antipsychotic, while response to SSRIs was defined as treatment for ≥6 month with the same SSRI without switching or augmentation. Treatment-resistant depression was defined as ≥2 switches between antidepressants between 6-14 weeks, and the control group was defined by history of treatment with the same antidepressant for ≥6 month and < 2 switches. Using these phenotypes, we performed GWAS including more than 60,000 cases (non-response/resistance) and more than 200,000 controls (responders). The SNP-based heritability was significantly different from zero for both SSRI non-response (h2 = 0.0169, SE = 0.0051, p = 9e-4) and treatment-resistant depression (h2 = 0.0211, SE = 0.0047, p = 7e-6). We identified several novel genome-wide significant loci and genetic correlations with psychiatric traits. The proposed work is ongoing, and future analyses will include phenotype validation using data from self-reported medication response and the validation of genetic discoveries using comprehensive longitudinal clinical information from deep-phenotyped cohorts and large-scale clinical trials. Leveraging real-world data for well-defined and validated treatment outcome measures can increase sample sizes to improve the discovery of variants associated with non-response to antidepressants, with the potential to form the basis for precision psychiatry approaches.