Genetic and epigenetic predictors of antidepressant response

Despite being the mainstay of treatment for major depressive disorder (MDD), antidepressants have highly variable efficacy, with fewer than a third of patients achieving remission after a first-line trial. This has fuelled a decades-long search for genetic and epigenetic biomarkers to guide personalised treatment. This narrative review critically synthesises the evidence, tracing the evolution from hypothesis-driven candidate-gene studies to large-scale, unbiased genomic and epigenomic approaches. We find that the initial promise of single candidate genes, such as SLC6A4, has largely dissolved under the scrutiny of larger, more rigorous studies, with most early findings failing to replicate. Consequently, attention has shifted to genome-wide association studies, which indicate that antidepressant response is a highly polygenic trait influenced by thousands of genetic variants, each contributing a very small effect. The primary clinical success has been in pharmacokinetics, where variants in CYP2D6 and CYP2C19, and more recently CYP2B6, are used to guide dosing and mitigate adverse effects, while evidence for predicting efficacy remains limited and appears to be medication-specific. Emerging epigenetic and transcriptomic markers offer dynamic insights, with DNA methylation and especially histone post-translational modifications helping to elucidate antidepressant mechanisms, while baseline inflammatory transcriptomic signatures are among the more reproducible correlates of non-response. However, their clinical translation is currently stalled by fundamental challenges, including tissue specificity and a lack of validated, replicable signatures. While panel-based pharmacogenetic testing provides modest clinical benefits, accurate prediction of therapeutic success remains an unmet goal. Future progress will require substantially larger, ancestrally diverse, medication-aware cohorts integrated with multi-omic analyses and anchored to disease-relevant tissue context through cell-type-resolved human brain studies, post-mortem validation, and patient-derived induced pluripotent stem cell models.