By Elisabeth Chai
“Medicine is a science of uncertainty and an art of probability.” The musings of Sir William Osler, the “Father of Modern Medicine,” have been well regarded by the medical community throughout the last century. Every day, clinicians are forced to reach a compromise between endless diagnostic possibilities and the need to make a decision. Psychiatry especially grapples with making judgments based on imperfect knowledge, and the complex interactions between genetics, culture, and the environment leave no room for universality in psychiatric illness or pharmacological efficacy. However, recent developments in molecular genetics and genomics may challenge this long-standing dichotomy.
Pharmacogenomics aims to systematically assess the simultaneous impact of gene variants related to drug metabolism. Methods such as genome-wide association studies (GWAS) are used to identify genetic loci associated with known outcomes, which are then compared to an individual’s genomic information to curate a personalized treatment regimen (1). When applied to psychiatric medicine, which observes wide variability in treatment response, there is significant potential in assisting the prescription and dosage of medications to avoid adverse side effects. For example, individuals taking paroxetine (Paxil), a commonly prescribed antidepressant, were found to be 3.26 times more likely to experience suicidal ideation (2). Furthermore, there appears to exist a correlation between the plasma concentrations of paroxetine and cytochrome P450 (CYP) 2D6 genotype (3). Delineating the relationship between paroxetine and CYP 2D6 polymorphisms could allow clinicians to predict which patients would benefit most from paroxetine and the dose at which it will be most effective.
Here’s how the testing would work: for $2,000 per test, a simple cheek swab, saliva, or blood sample is obtained and sent to a laboratory for genetic analysis, where results are returned within a few days. An individual’s phenotype is classified as a poor metabolizer, intermediate metabolizer, extensive metabolizer (normal), or ultrarapid metabolizer according to the speed at which they are able to metabolize a specific medication (3). Treatments are then tailored to phenotype: a poor metabolizer might be prescribed an alternative drug they are able to adequately metabolize without buildup, whereas an ultrarapid metabolizer might be given a higher-than-usual dose. In a study to determine the cost-effectiveness of pharmacogenetic-guided treatment in major depressive disorder, the pharmacogenetic-guided treatment group was found to produce more quality-adjusted life-years (2.07 over 3.00 years, versus 1.97), lower probability of death by suicide (0.328%, versus 0.351%), and $4,598 in savings compared to those given standard of care treatment (4). Results were further pronounced in those with severe or treatment-resistant depression.
Although pharmacogenomics is touted a promising tool for personalized pharmacological therapy, its clinical implementation remains extremely limited. Like the psychiatric disorders they are intended to treat, drug response is dynamic and affected by a number of genetic and environmental factors. Interactions between several genes and exogenous modifiers can create different phenotypes, thus, polymorphisms studied in isolation may not be effective predictors of clinical outcome. Genetic phenomena such as gene silencing, epistasis, genomic imprinting, and RNA interference may also influence gene expression in select environmental conditions (6). Most importantly, pharmacogenomic testing would only yield significant results if the patient’s particular gene variant had been previously studied — leaving those with rare mutations at risk for undesirable treatment results.
As we continue to observe advancements in pharmacogenomic technologies, biomarkers and novel targets for therapeutic interventions are being rapidly uncovered. While pharmacogenomics still faces many barriers to successful clinical uptake, evidence of its potential to bypass psychiatry’s traditional trial-and-error methodology is undeniable. Safe, sensitive treatment options are on the horizon — as well as the confidence to embrace the uncertainty and probabilities that made up the foundation of Osler’s philosophy.
References:
Srinivasan, B. S. et al. (2009). Methods for analysis in pharmacogenomics: lessons from the Pharmacogenetics Research Network Analysis Group. Pharmacogenomics, 10(2), 243–251. https://doi.org/10.2217/14622416.10.2.243
Huang, X. et al. (2022) Efficacy of psychotropic medications on suicide and self-injury: a meta-analysis of randomized controlled trials. Translational Psychiatry, 12, 400. https://doi.org/10.1038/s41398-022-02173-9
Charlier, C. et al. (2003). Polymorphisms in the CYP 2D6 gene: association with plasma concentrations of fluoxetine and paroxetine. Therapeutic Drug Monitoring, 25(6), 738–742. https://doi.org/10.1097/00007691-200312000-00014
McElroy, S. et al. (2000). CYP2D6 genotyping as an alternative to phenotyping for determination of metabolic status in a clinical trial setting. AAPS PharmSciTech, 2(4), E33. https://doi.org/10.1208/ps020433
Groessl, E. J. et al. (2018). Cost-Effectiveness of a Pharmacogenetic Test to Guide Treatment for Major Depressive Disorder. Journal of Managed Care & Specialty Pharmacy, 24(8), 726–734. https://doi.org/10.18553/jmcp.2018.24.8.726
Nebert, D. W. et al. (2003). Pharmacogenomics and "individualized drug therapy": high expectations and disappointing achievements. American Journal of Pharmacogenomics: Genomics-Related Research in Drug Development and Clinical Practice, 3(6), 361–370. https://doi.org/10.2165/00129785-200303060-00002
Comments