Pharmacogenetic data helps inform individualized treatment decisions for bipolar depression
Session: Personalized Medicine for Bipolar Depression: Combining Evidence and Clinical Wisdom to Improve Clinical Outcomes
Presenter: Joseph Goldberg, MD, Clinical Professor of Psychiatry at the Icahn School of Medicine at Mount Sinai
In the world of cancer and infectious disease research, personalized medicine is all the rage, with progress zooming forward at terminal velocity. Today, psychiatry is gathering information in this area as well. While individualized mental health treatments are still hovering back on the horizon, there are “interesting studies” out there, explained Joseph Goldberg, MD, Clinical Professor of Psychiatry at the Icahn School of Medicine at Mount Sinai. He presented key focal points and a few of these studies during his Psych Congress 2021 session, Personalized Medicine for Bipolar Depression: Combining Evidence and Clinical Wisdom to Improve Clinical Outcomes.
To begin at the beginning, Dr. Goldberg offered up two crucial definitions.
- Precision-based medicine is patient-generated health data, meaning measurable phenomenology that can guide a patient’s trajectory in terms of treatment.
- Personalized medicine is a subset of precision medicine that primarily deals with genomics—those 20,000 or more genes in the human genome. “We’re looking at how a patient’s genes inform us about the chances of that patient having a better effect, a worse effect, or an adverse effect to a specific treatment,” Dr. Goldberg said.
Genomics ties into phenotypes—observable characteristics of an individual’s genetic constitution that directly results from interaction between a person’s genotype and their environment. “In our world, this could apply to suicide, psychosis, a nocturnal phenomenology, impulsive aggression, sensation seeking, cognitive deficits—so many issues can count as phenotypes with a genetic corollary,” Dr. Goldberg said. “What’s fascinating to me, as a researcher in bipolar disorder, is that this tells me that there are so many manifestations above and beyond mood.”
For example, the doctor explained, it may someday be possible to see beyond clinical disease features of a diagnosis and understand how an individual’s protein expressions inform treatment decisions and understand how phenomenology potentially influences the expression of genes. “If you come in contact with cigarette smoke, ionizing radiation, or other environment factors, these things could adversely affect gene function, expression, and structure,” he said.
All this patient data encompasses what Dr. Goldberg referred to as “omics.” These are various measures for the analysis of an individual’s genetic profile—measures that may also be able indicate if a gene has been modified by environmental factors.
- Exposome: the measure of all the exposures of an individual in a lifetime and how those exposures relate to health
- Epigenome: the complement of chemical compounds that modify the expression and function of the genome
- Microbiome: the microorganisms in a particular environment (including the body or a part of the body)
- Metabolome: the unique chemical fingerprints that specific cellular processes leave behind, the study of their small-molecule metabolite profiles
- Proteome: the entire complement of proteins that is or can be expressed by a cell, tissue, or organism
- Transcriptome: the sum total of all the messenger RNA molecules expressed from the genes of an organism
Personalized medicine in relation to psychiatry and psychopharmacology typically spotlights pharmacogenetics, which relates to the interaction between genetic predisposition and responses to therapeutic drugs, with the goal being to learn something about a particular outcome state.
There are two main concepts relevant to pharmacogenetics.
Pharmacokinetics, or the impact of the body on the drug. asks, “How does the patient metabolize a drug?” Pharmacogenetic testing looks at whether the patient has the genetic vulnerability to metabolize certain substrates rapidly or slowly, which introduces different—and sometimes undesirable—results, which would subsequently suggest the need for a different drug.
“An ultra-rapid metabolizer is going to burn through a drug so fast that they may not get an effect, so you have to either pick a different drug that goes through a different pathway, or you have to give a much higher dose to be able to achieve the desired effect,” Dr. Goldberg said.
A particular concern for slow metabolizers—those who cannot break down a drug—is a dangerous drug buildup, which can potentially lead to adverse effects as well as lack of efficacy.
While Dr. Goldberg does not recommend an automatic pharmacogenetic test for every patient, it’s a useful tool if desired results just aren’t happening. “If I’m giving a patient lumateperone or lamotrigine and not seeing a benefit. I could at least speculate, ‘Gee, I wonder if this is someone who’s a poor metabolizer of UGT1A4.’ So, I could do pharmacogenetic testing to confirm my suspicion, or I might just pick a different drug that doesn’t go through that pathway.”
In addition to pharmacokinetics, the second concept relevant to pharmacogenetics is pharmacodynamics, which is how the drug effects the body, and here, too, there are safety and efficacy implications.
“We expect the day will come when we can actually do pharmacogenetic testing and predict response,” Dr. Goldberg said. “As of now, we have limited data on the extent to which we can predict response based on pharmacodynamic pharmacogenetics. But we do have some important data about safety pharmacodynamics.”
One such safety insight involves carbamazepine, which is occasionally used to treat bipolar disorder when other medicines have not worked. “If someone is of Southeast Asian ancestry and is going to take carbamazepine, you have to measure the presence of a particular gene called HLA-B*1502 because that means there’s a higher risk for Stevens Johnson Syndrome.”
Another red flag might be a patient taking valbenazine, deutetrabenazine, or iloperidone, and they’re a poor metabolizer of 2D6. This could result in a drug build up that potentially leads to cardiac toxicity.
Again, this is not to say that every patient needs pharmacogenetic testing, but as Dr. Goldberg explained, what’s important is to know the metabolic pathway of drugs being prescribed. “So, if a patient comes to me and says, ‘Gee, Dr. Goldberg, I’ve taken venlafaxine and paroxetine and vortioxetine, and I haven’t gotten better,’ then pharmacogenetic testing would probably help affirm a hypothesis that this patient is a poor metabolizer of 2D6, and I should pick a different substrate. I should pick a drug that goes through a different class.”
Dr. Goldberg offered another example where pharmacogenetic testing might be helpful. “It turns out, the serotonin transporter gene has a short and a long variant, or allele. If you happen to get a copy of the short allele from both your mom and dad, that may have some genetic implications in terms of risk for depression and severity of depression.
“According to a meta-analysis,1 if you happen to have the short form of this gene, you’re about 1.35 times more likely to flip into a hypomania or mania if given an SSRI than someone with the long variant. Now again, let’s not rush out and start testing everybody’s serotonin transporter gene before we pick an SSRI. But there may be a signal here that speaks to a subtype of the illness. The more we can collect information like this, the more we can start to talk about applying pharmacogenetics to personalized medicine.”
But, Dr. Goldberg added, pharmacogenetics testing does have limitations. A drug response may connect to genetics, but it could also relate to the patient’s comorbidities, chronicity, psychosis, impulsivity, age at onset, etc.
With so many riddles still unsolved in the world of personalized medicine specific to bipolar depression, Dr. Goldberg emphasized the importance of what he calls deep phenotyping. “That’s describing everything you can think of that you feel might influence your decision about choosing a particular treatment for a given patient.”
That brings the discussion around to moderators and mediators.
Moderators are baseline characteristics that influence outcome.
- Age at onset
- Baseline severity
- Episode number
- Polarity proneness
- Rapid cycling
- History of childhood trauma
- Poor metabolizer genotypes
Mediators are things that can influence the outcome of treatment once it’s been started.
- Treatment adherence
- Drug–drug interactions
- Therapeutic alliance
- life events after treatment has begun
- Adverse drug effects or nocebo effects
In terms of deep phenotyping, consider a patient with a good response profile for a particular drug, but they’re adherence is poor. So, they’re not achieving the desired benefits. A personalized approach to this case would be to review the patient’s mediating and moderating factors to find a solution—which might be a long-acting injectable or a drug with a very long life. “That’s an example of personalizing medicine based on knowledge of the mediating effect of their habitual adherence.”
As a researcher in bipolar disorder, Dr. Goldberg looks forward to more—and repeatable—studies specific to personalized medicine in mental health. Until that day arrives, it’s time for clinicians to get up close and personal with the concept of tailoring individualized pharmacology to a patient’s clinical profile and presentation—rather than to a diagnosis.
1. Daray, Federico & Thommi, Sairah & Ghaemi, S. (2010). Daray FM, Thommi SB, Ghaemi SN. The pharmacogenetics of antidepressant-induced mania: a systematic review and meta-analysis. Bipolar Disord 12: 702-706. Bipolar disorders. 12. 702-6. 10.1111/j.1399-5618.2010.00864.x.