Pharmacogenetic Testing in Kids for Psychiatric Medications: Is it Good Medicine?
Let’s start this complicated and controversial conversation by reviewing some possible scenarios. Do any of these anecdotes feel familiar?
So here you are, a wonderful pediatric primary care provider, finally feeling somewhat comfortable prescribing psychiatric medications, and now a whole new minefield awaits you. You are not alone. Psychiatric providers are struggling to understand the emerging science of pharmacogenetics as well and its relations to psychotropic prescribing. Some psychiatrists may tell you to stay away as the testing is not yet ready for primetime, while others may tell you about the helpful information they have learned when using the test. What are you to do? While this newsletter is not the place for a full academic review of pharmacogenetics, we will go through the basics and tell you how to find more information.
First and foremost, it is important to know that the tests are not all they are purported to be. In fact, on November 1, 2018, the FDA released a consumer warning stating: “The FDA is alerting patients and health care providers that claims for many genetic tests to predict a patient’s response to specific medications have not been reviewed by the FDA and may not have the scientific or clinical evidence to support this use for most medications. Changing drug treatment based on the results from such a genetic test could lead to inappropriate treatment decisions and potentially serious health consequences for the patient.” And in April of this year, the FDA sent a warning letter to a specific laboratory for “illegally marketing certain genetic tests that have not been reviewed by the FDA for safety and effectiveness”
Yet, we cannot dismiss this field all together! On September 3, 2019, the FDA updated its “Table of Pharmacogenomic Biomarkers in Drug Labeling” and the updated labelling includes several psychiatric medications. Thus, some of the medications we prescribe may have specific written suggestions regarding prescribing based on pharmacogenetics in its drug labelling. As you probably know, it is now required to test for HLA-B1502 in Asian patients before starting carbamazepine as the allele in this population has been associated with a greater risk of Stevens Johnson Syndrome.
Of course, the research is complicated by the fact that many of the algorithms being used by the testing companies to determine their choice of medications for patients remains proprietary, limiting one’s ability to assess its validity. In addition, much of the published literature has been paid for by industry funding. Lastly, as usual, most of the information is from adult studies, not pediatric, and we know kids already have developmental differences in metabolism.
So, let’s go back to the beginning.
Pharmacogenetics, also known as pharmacogenomics, is the science of examining specific genes to help match those genes with specific types of medicines or dosages. The science is based on there being polymorphisms in DNA that effect how drugs pass through and interact in our body. Pharmacogenetic testing attempts to look at two aspects of DNA polymorphisms, pharmacodynamics and pharmacokinetics. Pharmacokinetics looks at how the medicines are absorbed, distributed, metabolized and excreted. Different pharmacokinetics can lead to different drug levels in someone’s body. Pharmacodynamics looks at how the drug interacts with the drug targets, so may influence how a medication works.
Let’s start with the pharmacokinetics, where there seems to be more data and evidence surrounding the psychiatric medications. Many psychiatric medications are metabolized in the liver, specifically in the cytochrome P450 system, and scientists have been able to identify P450 polymorphisms that affect psychiatric metabolism, such as differences in the 2C19 or 2D6 enzymes. Many of the pharmacogenetic tests on the market try to determine if someone is a rapid or slow metabolizer based on these polymorphisms and try to extrapolate one’s response to a medication based on that finding. For some medications and polymorphisms, the answer may be to use the drug but at a lower dose (perhaps a 50% reduction) or to use the drug at a higher dose if an ultra-rapid metabolizer. Sometimes, it may be suggested that a very slow metabolizer or a very fast metabolizer may do better if they stay away from a certain medication. Much of the hype around the testing implies that the testing will determine why your patients are having so many side effects, and there are claims that testing will help you find a medicine that will not cause side effects in your patients. Many of those assertions are overblown. While theoretically some side effects may be due to issues of differences in metabolism, many side effects have other explanations. Some adverse drug reactions are due to hypersensitivities and not absorption, while others are due to concomitant medications, smoking, medical issues, and psychological states. The CPIC (Clinical Pharmacogenetics implementation Consortium) guidelines, based on adult data, have been established to help with prescribing certain psychiatric medications based on several Cytochrome P450 polymorphisms, but it is essential to understand that the tests that are currently in use will not magically eliminate side effects for your patients. In addition, we do not have data on whether these modified dosages suggested by CPIC, if used in children, will lead to therapeutic benefit. The boxed warning remains as much of a warning, whether one has or has not had pharmacogenetic testing and whether the medication is in red or green!
Now let’s discuss pharmacodynamics, where the evidence is less clear. The thinking in this area of study is that genetic variations in the drug targets may affect drug outcome. For example, a person with a genetic polymorphism that results in decreased serotonin transporters may have a decreased response to an SSRI. Some studies of COMT (Catechol O-Methyltransferase gene) suggest that polymorphisms may cause a decreased response to methylphenidate or an increased response to amphetamines. However, much of this data is still theoretical and in its early stages. We cannot say that one phenotype tested can now predict drug response. As the FDA said in 2018, “The relationship between DNA variations and the effectiveness of antidepressant medications has never been established.” We do know that age, ethnicity, polypharmacy, and substance use all affect drug response, and we do not know how all the different phenotypes that affect a drug interact to make it more or less efficacious, as the research is still looking at one phenotype at a time.
Again, much of the outcome data is industry funded. In addition, much of it is done in adults. The limited data available looking at psychiatric treatment outcomes when using pharmacogenetic testing as a guide mostly examines mood disorders.
As described in the FDA warning and as seen in real life, patients and doctors may follow the red, yellow, and green suggestions of the reports without fully understanding how the company made those decisions. The clinical decision is then made to forego a first line medication in favor of a medicine suggested by the tests, a medicine that may not have as much evidence for efficacy, and in pediatrics, may not have even as much evidence for safety. In reality, the tests may be suggesting that dosage adjustments may be considered, rather than that the drug needs to be avoided at all costs. Therefore, as much as is possible, a careful review of the pharmacogenetic and pharmacokinetic data presented in the reports is more important than the visual aids. In addition, while some of the tests for the polymorphisms are well established (such as the 2D6 described above), others are in its infancy and may be pushed by the companies before they are ready. Again, this added information about polymorphisms that may or may not yet be accurate or helpful may then complicate the picture and steer one away from the right medication.
Fortunately, or unfortunately, the testing itself has become easier and more financially attainable as well. A phone call or email to one of the limited companies that perform this testing will result in a representative signing your practice up and sending you some starter kits. A simple cheek swab is all that is required of your patient along with their insurance information. While the costs used to be prohibitive, these tests now are often, but not always, either covered by insurance or guaranteed by the company itself not to exceed a certain out of pocket fee.
Patients and doctors need to understand what these tests can and cannot do. Nobody should be promising that a simple check swab will determine the medicine and dosage that’s right for the child, promising a medicine that will definitely work and a medicine that will definitely not have side effects. Unfortunately, we are not there yet in the world of psychiatry. Any discussion of testing needs to set realistic expectations.
It also seems a bit premature to incorporate these tests as a routine first step in primary care, especially in pediatric primary care. While at least one academic specialty mental health service has started using the tests for CYP2D6 and CYP2C19 routinely on their inpatient pediatric psychiatry population, we must remember the overall limitations and dangers. Given that the pharmacogenetic tests seem to be less helpful in ADHD treatment and are based mostly on adult data, and that the evidence and safety data in kids for psychiatric medications are already more limited, starting with a first line medication for children and adolescents for the diagnosed disorder seems to be the best starting point.
However, if a child has already had the testing, and you are presented with a report, it will be important to be able to understand the report and not just skip to the color-coded graphics. Pay attention to the pharmacokinetic results. See if they map onto any CPIC guidelines about medication avoidance or dosage adjustments. Read through any pharmacodynamic claims as well and assess whether they have clinical utility at this time. Explain to your patient that the report may help you avoid a medication that may cause problems, but it is less likely to help you find a medicine that will work.
Finally, if you have a patient for whom you have struggled to find the right medicine, a patient for whom medicines seem to never work or to cause way too many side effects, it may be appropriate to bring up the topic of pharmacogenetic testing. But as discussed above, the testing may not end up adding to your armamentarium. At this point, if you have not already done so, a call to Project TEACH is definitely in order!
