Characterising antidepressant usage based on electronic prescription records in the australian genetics of depression study

One in five people will be diagnosed with major depressive disorder (MDD) in their lifetime, and approximately one third of those are refractory to treatment. MDD is primarily managed with antidepressant medication. However, outcomes are variable and many individuals do not experience any remission of symptoms or cease treatment due to adverse side-effects. Several clinical, treatment and individual characteristics have been associated with poor depression treatment response or adverse side-effects but most variables still have a low sensitivity to distinguish between those who will and will not improve from certain types of treatment. Thus, the incorporation of genetic information alongside demographic and clinical predictors is a promising avenue to improve our understanding of antidepressant treatment response. In this study, we aim to characterise individuals with differential treatment response based on antidepressant prescription record data in the Australian Genetics of Depression Study (AGDS). To stratify individuals based on prescription patterns, we used Pharmaceutical Benefits Scheme (PBS) prescription data from ∼16,000 AGDS participants who consented to record linkage of 4.5-years (01/07/2013-31/12/2017) of their PBS data. We focused on selective serotonin reuptake inhibitors (SSRIs) or serotonin and norepinephrine reuptake inhibitors (SNRIs) and included information on self-report of electroconvulsive therapy (ECT) treatment (or recommendation for ECT treatment), self-report of bipolar disorder (BD) diagnosis and prescription of lithium. Participants with more than 20 (monthly) prescriptions of the same antidepressant were allocated to SSRI or SNRI groups and contrasted to a “treatment resistant” ECT and a BD group. Individuals in each group were characterised using demographic and self-reported course and treatment outcome data, other medication use, as well as genetically using polygenic scores (PGS) and CYP2C19 and CYP2D16 metaboliser profiles. Age of onset and self-reported response rates were very similar across all SSRI and SNRI groups. However, the median number of self-reported lifetime depressive episodes was higher in the SNRI (n=7) than SSRI response groups (n=5). In comparison, participants in the ECT and BD groups reported the highest median number of depressive episodes (n=13+) and were more likely to be male. This corresponded with the mean PGS for both depression and bipolar disorder being significantly higher in both these groups compared to the SSRI and SNRI groups. Further differences between the groups are being investigated using PGS for psychiatric disorders and related traits and CYP gene metaboliser profiles, and preliminary analyses show interesting differential patterns of association. Understanding the genetic influences on long-term pharmacological treatment usage is a promising field of research to reduce or mitigate the risk of poor treatment outcomes for individuals with depression. Characterising participants based on prescription data is a highly valuable avenue to better understand factors affecting response to treatment. Future work will examine differences between groups in other health outcomes such as length of treatment, discontinuation of medication, medication switching, and number of supervised mental health treatment plans.