As is the case with many topics in genetics, I learn the most in my time away from the office, researching questions for friends and family. And the genetics questions du jour are almost all related to pharmacogenomics (PGx) testing. More specifically, PGx testing for psychiatric medications. Maybe you are getting these questions too? For me, they usually go something like this:
Are these genetic tests that promise to tell you what
antidepressant medication will work best for you a real thing?
And if they are, why isn’t everybody taking them?
I have witnessed friends and family members struggle with medication management for depression, anxiety and mood disorders. People in these situations are often desperate for help and quite vulnerable. I can fully appreciate the hope that a simple genetic test could provide the answers to ease the journey. But as we find with many things in genetics, the reality is much more complicated than the hopeful answer we wish we could give. Ultimately, I know my position on this will disappoint many who are looking for that silver bullet. While there are a handful of applications for pharmacogenomics in specific situations related to psychiatric medications that have evidence to support their use, there is little evidence that multigene panels in this area lead to better outcomes. There is concern that harm may come because of the use of this unproven information to guide important decisions with prescribing and dosing of psychiatric meds.
My conclusions regarding these tests are in part based on critiques from reputable sources on the current state of commercially available PGx psychiatric panels, including the American Psychiatry Association Workgroup for Novel Biomarkers and Treatments. This workgroup performed a detailed review of several commercially available tests and concluded that there is a lack of sufficient evidence to support the widespread clinical use of the proprietary combinatorial pharmacogenomic models used by these labs. There are many publications that highlight issues with existing studies about these tests including concerns related to conflict of interest and problematic study design. The financial sector is also following this topic closely given much has been invested based on the promise of these testing products being adopted broadly. Recognizing the strong commercial drivers at play here also causes me to view laboratory claims with a dose of skepticism.
But when it comes to the general public, I think the commercial push to see these tests more broadly adopted is drowning out the voices of the experts who are urging caution. It seems that since many in the field of psychiatry aren’t convinced that these tests are ready for prime time, the labs have decided to bypass the most relevant specialty, and go straight for patients and primary care providers. Additionally healthcare payers are banking on the promise that these PGx tests will more than pay for themselves by allowing for better precision in prescribing of expensive medications. Payer support is helping to move psychiatric PGx testing to the mainstream.
New pharmacy-laboratory partnerships are emerging to promote these tests. Last year, PGx lab Genomind® announced a partnership with Albertsons Sav-On, Jewel-Osco and Acme Sav-On pharmacies: pharmacists can discuss PGx testing with patients and, if the patient consents, the pharmacists will directly contact the prescribing provider to “suggest the Genecept® Assay. ” The sample can be collected right in the pharmacy. Last month, Myriad Genetics, Inc. announced a similar program with Kroger Prescription Plans to promote GeneSight® genetic tests in Kroger pharmacies. From the GeneSite® press release: “pharmacists at more than 2,300 Kroger stores will provide counselling about GeneSight® to eligible employer group members and facilitate testing with their prescribing healthcare professionals.”
Pharmacists are now direct marketing genetic testing to patients. And while members of the pathology and clinical laboratory space are taking some issue with this, there hasn’t been much public concern raised by the broader medical genetics community about this proposal to have pharmacists providing “genetic counseling” and facilitation of genetic testing.
It will be interesting to see how these programs evolve with greater attention from regulators. With several years of push and pull between labs marketing these tests and clinicians raising concerns about their clinical utility and safety, the FDA has recently started to flex their regulatory muscles in this space. In Oct 2018, the agency published a Safety Communication, warning patients not to change management based on PGx results without first discussing with their healthcare provider and to be aware that claims made by genetic testing laboratories about PGx tests are not supported by sufficient clinical evidence. The FDA cautions healthcare providers in the use of these tests and directs providers to FDA-approved drug and genetic test labels. Lastly, the communication advised test manufacturers not to include specific drug information that is inconsistent with FDA-approved drug labeling. In April, the FDA sent a letter to Inova Health System with concern that the clinical validity of their PGx tests had not been established for the reported intended uses. Shortly after this letter was issued, Inova elected to cease offering their MediMap® tests. The FDA has been in communication with several other laboratories and stated that “most firms addressed the FDA’s concerns by removing specific medication names from their labeling, including promotional material and patient test reports.”
Some of the critiques I have heard about the FDA’s engagement in regulating these tests is centered around whether or not the FDA should be the authority on the evidence required to support the relationship between certain variants and drug metabolism. A frequently referenced pain point is the difference between the PGx genes/variants that make the cut per FDA drug labeling and the evidence grade rating per the Clinical Pharmacogenetics Implementation Consortium of the Pharmacogenomics Research Network (CPIC). In looking at most of the commercially available PGx tests on the market today though, it is clear that many of the variant-drug connections included on lab reports are not consistent either with the FDA-approved list or with CPIC guidelines.
For example, the sample GeneSight® report available online as of the day of this posting, under the category of “Mood Stabilizers” shows three drugs in the red bucket, “Significant Gene-Drug Interaction” and no drugs in the green, “Use As Directed” bucket. Top of the red list is lamotrigine (Lamictal®), which has a footnote that reads “Use of this drug may increase the risk of side effects.” The justification given in the Gene-Drug Interactions table is a variant in the UGT1A4 gene. Search of the CPIC database gives the UGT1A4 – lamotrigine pair a “D” level rating. According to the website, the CPIC D Level is defined as follows: “There are few published studies, clinical actions are unclear, little mechanistic basis, mostly weak evidence, or substantial conflicting data. No prescribing actions are recommended.” How might this report affect a person with bipolar who is struggling to find the right medication? One can imagine that it may be difficult for both the patient and the prescribing provider to feel comfortable with a treatment plan when not supported by this genetic test report.
For the same GeneSite® sample report, under the “Antidepressants” heading, there are a total of 22 antidepressents for which analysis is available, with only three in the “Use as Directed” green bucket. Top of that long list of 22 drugs in the red, “Significant Gene-Drug Interaction” bucket is bupropion (Wellbutrin®). Bupropion is a medication commonly used to treat depression and has been approved by the FDA for use since 1985 with a generic version of this drug readily available. There is no data in the CPIC database to support the assertion made my GeneSight® of a “Significant Gene-Drug Interaction” with bupropion. And interestingly, the only three antidepressant medications that made it to the green “Use as Directed” category are expensive drugs for which no generic version is available: levomilnacipran (Fetzima®), desvenlafaxine (Pristiq®), and vilazodone (Viibryd®). If I received this report and didn’t know better, I might assume that these drugs would be worth the high price tag if they are genetically the most likely to treat depression without the potential for side effects. There is no gene-drug information in CPIC about any of these three preferred medications, and I didn’t have to look very far beyond the GeneSight® report to see the long list of side effects and contraindications associated with each of these medications. But imagine the difficulty a prescribing provider might have in convincing a patient to consider forgoing the expensive new drugs in the green bucket to consider a more affordable medication with a longer history of success in treatment from the red bucket. Lab reports are not often looked at as one piece of the puzzle, but rather as the *truth* by patients. And as I have previously written on a different topic, it is incredibly difficult to convince a patient that an expert assessment may be more trustworthy than what is printed on a test report.
Regulation of genetic testing is a big and thorny issue, and I don’t claim to have easy solutions for improving these challenges. But what I do hope to do is to begin a conversation with my fellow genetic counselors on what role we should have in the dissemination of information regarding PGx testing. I feel it is our professional obligation to understand, to the best of our abilities, the evidence or lack thereof when counseling our patients, consulting with other healthcare providers and discussing these tests with friends and family. When people first started asking me about these tests, my initial feeling was one of hope and optimism. Of course it would be wonderful if a simple genetic test could provide a clear path towards the best medication for those who are suffering. Now, after having spent hours down the rabbit hole to try to better understand the current state of this field, I remain hopeful that these tools may someday provide real benefit for the masses. Unfortunately, it seems to me that at the present time, this wild west of the competitive genetic testing marketplace has resulted in bigger but not necessarily better panels, including information that is often not evidenced-based. I worry that these reports could lead people down a wrong, and potentially dangerous path.
So for now when my loved ones ask me, “are these genetic tests that promise to tell you what antidepressant medication will work best for you a real thing?” I will give the more cautious and complicated answer. While this technology holds promise for the future, the evidence we have at this point does not support that these tests will help guide better care and lead to better outcomes for most people. And I will continue to do my best to support them on the journey forward, wherever the bumpy and winding road may lead.