GENETIC TESTING IN PSYCHIATRY
BY BENJAMIN MARGOLIS, M.D.
As a psychiatrist, if I had an available tool that would provide a tailored genetic map of an individual I'm working with that would allow me to prescribe exactly the right medication, I'd want it available for every patient I treat. Such is the promise of genetic testing to guide prescribing of psychotropic medications.
This technology is already being applied robustly in cancer treatment, where antibody-based therapies can be precisely targeted to the genetic makeup of a particular tumor, or we can look for genetic markers that directly correspond to response to certain standard chemotherapeutic agents. In oncology, the era of pharmacogenomics has arrived. In both neurology and psychiatry, testing for HLA genes can help predict risk for a potentially serious side effect from mood-stabilizing antiepileptic medications.
In psychiatry, the promise of pharmacogenetics is great. We can simply run a test where a cotton swab is gently swiped inside each cheek, sampling a few cells from the mouth, and then mail that swab to a lab. In a few days, a comprehensive report returns with a list of psychiatric medications and their potential "gene/drug" interactions. We should then be able to use this report to minimize the chance for side effects and maximize the possibility of therapeutic response. If only it were so simple!
When someone has depression or anxiety and requires treatment with a medication, in our field, we carefully consider the nature of the symptoms we're treating, the efficacy and tolerability of the medications we are suggesting, the other medications the person is taking, other medical problems they have, and integrate this into our decision-making tree before suggesting a medication. This is our standard, and considering gene/drug interactions is not yet part of the algorithm.
While genetic screening sounds like a wonderful tool to add to our armamentarium, in psychiatric clinical practice, the evidence for clinical benefit is still pending, and there are some potential pitfalls. To start with a brief description of the tests: in our bodies, genes are sequences of molecular code made of DNA that are translated by cellular machinery to create proteins. We now have sophisticated tools to be able to read and interpret variations in the DNA within particular genes. In the case of psychiatry, the genes we test for mainly code for a set of proteins within the liver called cytochrome P450 proteins, which are enzymes (molecular machines that speed along chemical reactions in the body) involved in processing and clearing medications. Those cytochromes have names like 2D6, 3A4, etc. When we take an oral medication, the medication is absorbed through the GI tract, and then can be either cleared and processed by the liver and excreted in stool, or by the kidneys and out through the bladder. Within the liver, those CYP450 enzymes, again with names like 2D6 and 3A4, take that drug and process it, creating metabolites. With each step, those metabolites themselves may have pharmacologic effects on the body and are themselves either processed by more CYP450s or excreted. The above process is a brief and hopefully not-too-baffling summary of the process of pharmacokinetics, which is what your body does to a medication.
For the majority of psychiatric medications, we have a detailed understanding of the entire pharmacokinetic journey from ingestion to metabolite to excretion. We know which particular cytochromes make which metabolites, what those metabolites do, and on average, what range of time they stick around in the body and where. So where does the genetic testing come into play?
Pharmacogenetic testing identifies variations in someone's genetic code that translate to different versions of those CYP450 enzymes. Some of them may work faster or slower to clear a medication or metabolite. We should then potentially be able to predict how quickly or slowly your body processes a medication. We may be able to predict how much or how quickly you may process metabolites associated with side effects, and then predict which medications might get along best with you. It sounds brilliant, and it is. The promise is undeniable. The problem is that pharmacokinetics is only part of the story when it comes to your body. In addition to pharmacokinetics, there are multiple other factors at play. There are interactions with the other medications an individual may be taking. There are considerations of age, other medical history, hydration status and weight. In addition to what your body does to clear a medication (pharmacokinetics), there are considerations of what the medication is doing to your body (called pharmacodynamics). The genetic screening companies are now offering tools for genetic profiles of some of the drug targets, profiling differences of receptors themselves, but these are not yet necessarily correlated to clinical response.
We can't yet say whether someone will respond to a medication based on these tests. In my clinics and in the care of individuals with I/DD, I have to a limited extent been using these tests for the past four years from the two largest commercial labs.
I don't yet offer these screens as standard of care before deciding In general practice, pharmacogenetic testing in psychiatry for treatment of anxiety or depression is not standard of care, and the APA does not yet support its widespread use. That leaves doctors working with the I/DD community still waiting for the tools to mature. to start a medication. If an individual or caregiver is interested in pursuing them, we talk about the process and the potentially limited information that these tests can provide. When we get test results back, there are color-coded results in green, yellow or red correlating to the extent of a particular predicted gene/drug interaction. However, that doesn't necessarily mean we would absolutely avoid a particular medication that may still have therapeutic value. That information may be helpful in predicting that someone may need a higher or lower dose of a medication, or that they may be more likely to have side effects at higher doses, but in terms of clinical practice, this does not change how we would go about prescribing a medication. In practice, regardless of genetic testing, we "start low and go slow," regardless of a genetic test result, monitoring for both clinical response and side effects carefully along the way. If someone wants to incorporate the results of their genetic testing as we decide what medication to utilize, I of course have no objection, and we will use that information together as part of a plan, but not necessarily as the main decision-making tool. There is the risk that a potentially helpful medication could be withheld and the benefit missed if too much weight is placed on these tests.
There is a financial consideration, as insurance does not necessarily cover these tests and they can run into hundreds of dollars. When it comes to limited resources available for some individuals with I/DD and their families, this can be a significant consideration. The practice arena I have found these screens to be most helpful is not in starting new medications, but in assessing which to taper or reduce. As a field, we are committed to minimizing polypharmacy, minimizing psychiatric medications as much as possible. Individuals with I/DD in psychiatric care have in the past been treated with two or sometimes even three antipsychotics at a time, a deep and complex discussion on its own, but here we will only discuss the universal effort to consolidate to antipsychotic monotherapy whenever possible. When I am assuming the care of an individual coming out of institutional care settings, or for whom limited history is available, and who may have been treated with two or three antipsychotics for 30 years, which medication do we choose to reduce first? Pharmacogenetic testing can potentially provide us with a suggestion on what we would target and can potentially guide our medication optimization.
Even in this circumstance, this comes with a potential financial cost and along with the caviat that there are multiple other factors at play in our medication optimization scenario above. Side effects, age and other medical considerations are often clear and help guide our reduction strategy.
In general practice, pharmacogenetic testing in psychiatry for treatment of anxiety or depression is not standard of care, and the American Psychiatric Association does not yet support its widespread use.1 The FDA has issued warnings against unproven claims from some manufacturers.2
So where does that leave us psychiatrists working with the I/DD community? It leaves us still waiting for the tools to mature. When prescribing medication, we make decisions in the context of the whole person, integrating all of the complex factors discussed earlier before proposing a plan and a trial of medication. Someday soon (but not quite yet), genetic testing will likely be an integral part of that process.•
ABOUT THE AUTHOR:
Benjamin Margolis, M.D. is board-certified in neurology and psychiatry, having completed the combined residency program at Brown University. He specializes in neuropsychiatric care of adults with I/DD in the New York Hudson Valley and is Senior Psychiatrist and Staff Neurologist at Access: Supports for Living. He attended medical school at the Albert Einstein College of Medicine after the postbaccalaureate premedical program at Columbia University, and holds a BFA from the School of Visual Arts in New York City. He is an active member of the American Academy of Developmental Medicine and Dentistry (AADMD) and is working to improve behavioral and mental health care of individuals with I/DD on local, state and national levels, along with serving on the Health Advisory Board for Special Olympics New York for the Strong Minds Program.
References
1. Zeier Z, Carpenter LL, Kalin NH, Rodriguez CI, McDonald WM, Widge AS, Nemeroff CB. Clinical Implementation of Pharmacogenetic Decision Support Tools for Antidepressant Drug Prescribing. Am J Psychiatry. 2018 Sep 1;175(9):873-886. doi: 10.1176/appi.ajp.2018.17111282. Epub 2018 Apr 25. PMID: 29690793; PMCID: PMC6774046.