# Pharmacogenomics in Psychiatry

## Foundations of Pharmacogenomics

### Key Concepts

**Pharmacogenomics:** the study of how genetic variation influences drug response (efficacy and adverse effects) **Pharmacokinetic genes:** encode drug-metabolizing enzymes (CYP450 system) and drug transporters; determine drug levels. **Pharmacodynamic genes:** encode drug targets (receptors, transporters); influence drug efficacy at the target site. Genetic variation is typically characterized as single nucleotide polymorphisms (SNPs), gene deletions, or gene duplications. **Phenotype categories:** ultra-rapid metabolizer (UM), extensive/normal metabolizer (NM), intermediate metabolizer (IM), poor metabolizer (PM)

### CYP450 Enzyme System

The cytochrome P450 enzyme superfamily mediates Phase I oxidative metabolism of most psychotropics. **CYP2D6:** metabolizes ~25% of all drugs; highly polymorphic (>100 known alleles); substrates include most SSRIs, many antipsychotics, TCAs, atomoxetine. **CYP2C19:** metabolizes citalopram, escitalopram, sertraline, some benzodiazepines, voriconazole. **CYP3A4:** metabolizes quetiapine, aripiprazole, many benzodiazepines, carbamazepine; less polymorphic, more influenced by inducers/inhibitors. **CYP1A2:** metabolizes clozapine, olanzapine, fluvoxamine; induced by smoking (clinically significant) **CYP2B6:** metabolizes bupropion; less well-characterized clinically.

## CYP2D6 Polymorphisms

### Clinical Significance

The most clinically relevant pharmacogenomic gene in psychiatry. **Poor metabolizers (PM):** 5-10% of Caucasians; carry two non-functional alleles; dramatically elevated drug levels at standard doses; increased risk of side effects. **Ultra-rapid metabolizers (UM):** 1-10% depending on ethnicity (up to 29% in Ethiopian populations); gene duplication/multiplication; sub-therapeutic levels at standard doses; treatment non-response. **Intermediate metabolizers (IM):** one reduced-function allele; mildly elevated drug levels. **Normal metabolizers (NM):** two functional alleles; standard dosing expected.

### Drugs Most Affected by CYP2D6 Status

| Drug/Class | CYP2D6 PM Implication | CYP2D6 UM Implication | CPIC Guideline? |
|------------|----------------------|----------------------|-----------------|
| TCAs (nortriptyline, amitriptyline) | Markedly elevated levels; reduce dose 50% | Sub-therapeutic levels; increase dose or use alternative | Yes (strong) |
| Paroxetine | Elevated levels; increased side effects | Reduced efficacy | Yes |
| Fluvoxamine | Elevated levels | Reduced efficacy | Yes |
| Aripiprazole | Reduce dose by 50% | Consider dose increase | Yes |
| Risperidone | Elevated active moiety levels | Reduced levels | Yes |
| Haloperidol | Elevated levels; increased EPS | Reduced efficacy | Emerging |
| Atomoxetine | 5-10x higher plasma levels; start low | Sub-therapeutic; increase dose | Yes |
| Codeine | No conversion to morphine; ineffective | Rapid conversion; fatal overdose risk | Yes (strong) |

**TCAs (nortriptyline, amitriptyline, imipramine):** strongest evidence; CPIC guidelines recommend dose adjustments based on phenotype; TDM is complementary. **SSRIs:** paroxetine and fluvoxamine are CYP2D6 substrates; escitalopram more affected by CYP2C19. **Antipsychotics:** aripiprazole (reduce dose by 50% in CYP2D6 PM or with strong CYP2D6 inhibitor), risperidone, haloperidol. **Atomoxetine:** CYP2D6 PM have 5-10x higher plasma levels; start at lower dose. **Codeine:** prodrug requiring CYP2D6 activation to morphine; PM get no analgesic effect; UM at risk for fatal overdose (particularly relevant in postpartum)

## CYP2C19 Polymorphisms

### Clinical Significance

**PM:** 2-5% of Caucasians, 12-23% of Asian populations; higher levels of substrate drugs. **UM:** 20-30% of some European populations; lower levels, potential non-response. **Key substrates:** citalopram, escitalopram, sertraline (partial), clobazam, some PPIs. **CPIC guidelines:** recommend alternative drug or dose reduction for CYP2C19 PM taking escitalopram/citalopram; recommend dose increase or alternative for UM.

## Pharmacodynamic Genes

### Serotonin Transporter Gene (SLC6A4)

5-HTTLPR polymorphism (short/long alleles) historically linked to antidepressant response and depression vulnerability. Short allele associated with reduced SERT expression. Initially high-profile findings (Caspi et al. 2003: gene-environment interaction with stress) have not been reliably replicated. Current consensus: 5-HTTLPR testing has no clinical utility; should not guide prescribing.

### HLA-B*15:02 and HLA-A*31:01

**HLA-B*15:02:** strong association with carbamazepine-induced Stevens-Johnson syndrome/toxic epidermal necrolysis in individuals of Southeast Asian descent. FDA recommends testing prior to carbamazepine initiation in patients of Asian ancestry. **HLA-A*31:01:** associated with carbamazepine hypersensitivity across ethnicities; testing guidelines vary.

### Other Pharmacodynamic Targets

**COMT (catechol-O-methyltransferase):** Val158Met polymorphism affects dopamine metabolism; theoretical relevance to antipsychotic response but not clinically actionable. **HTR2A (serotonin 2A receptor):** associated with antidepressant response in some GWAS; not yet clinically useful. **MTHFR (methylenetetrahydrofolate reductase):** C677T variant affects folate metabolism; marketed on some commercial panels but clinical relevance to psychiatric treatment is unsupported.

## Commercial Pharmacogenomic Testing Panels

### Available Products

GeneSight, Genomind, Tempus, OneOme, and many others. Typically test 5-15 genes (pharmacokinetic and pharmacodynamic) Provide categorized drug recommendations (e.g., "use as directed," "use with caution," "use with increased caution")

### Evidence for Clinical Utility

**GUIDED trial (GeneSight):** largest RCT; pharmacogenomic-guided treatment did not significantly improve primary outcome (HAM-D response) vs. treatment as usual; secondary outcomes showed modest benefit. **Subsequent meta-analyses:** mixed results; some show small improvements in remission rates; methodological concerns limit conclusions. CPIC and DPWG guidelines provide evidence-based gene-drug pair recommendations, but most are limited to CYP2D6 and CYP2C19 for specific drugs. The pharmacodynamic gene components of commercial panels generally lack adequate clinical validation.

### Limitations

Panels test for known variants only; rare or novel variants may be missed. Gene-gene interactions and environmental factors (diet, smoking, drug interactions) are not captured. Results represent metabolizer phenotype under ideal conditions; actual phenoconversion can occur with concomitant CYP inhibitors/inducers. Overinterpretation risk: clinicians may avoid appropriate medications or make unnecessary switches based on panel results. Cost-effectiveness data are limited; insurance coverage is variable.

## Practical Application

### When Pharmacogenomic Testing is Most Useful

Patients with treatment-resistant depression who have failed multiple adequate trials. Patients experiencing unexpected side effects at standard doses. Prior to prescribing carbamazepine in patients of Asian ancestry (HLA-B*15:02) When prescribing TCAs (use alongside therapeutic drug monitoring) Patients on polypharmacy where drug interaction prediction is complex.

### When Testing is NOT Helpful

First-line treatment selection in treatment-naive patients (clinical factors remain more relevant) As a substitute for therapeutic drug monitoring when TDM is available. To predict efficacy of a specific medication (current evidence is insufficient for this)

<image>
A diagram of the CYP2D6 metabolizer phenotype spectrum. Show four patient silhouettes representing ultra-rapid, normal, intermediate, and poor metabolizers. For each phenotype, show the relative enzyme activity level (bar graph), the expected plasma drug concentration curve at standard dosing (pharmacokinetic curves overlaid), clinical implications (efficacy vs toxicity), and prevalence by ethnicity. Include example allele combinations (e.g., *1/*1 for NM, *4/*4 for PM, *1/*1xN for UM). Clean medical education style.
</image>

<image>
A flowchart for clinical decision-making with pharmacogenomic test results. Start with "Pharmacogenomic test result received." Branch by gene (CYP2D6, CYP2C19, HLA-B*15:02). For each gene, show the phenotype result leading to specific dose adjustment recommendations or drug avoidance. Include a decision box for "Is the drug being prescribed a substrate of the identified enzyme?" and a reminder to consider phenoconversion from drug interactions. Include CPIC guideline references. Color-coded by actionability level.
</image>

<image>
An infographic showing a sample commercial pharmacogenomic panel report. Display the gene list (CYP2D6, CYP2C19, CYP1A2, CYP3A4, SLC6A4, HTR2A, COMT, MTHFR) with the strength of evidence for clinical utility color-coded: green for strong evidence (CYP2D6 for TCAs), yellow for moderate evidence (CYP2C19 for SSRIs), red for limited/no evidence (SLC6A4, COMT, MTHFR). Include callout boxes explaining why certain genes lack clinical actionability. Educational reference format.
</image>

## Clinical Pearls

CYP2D6 and CYP2C19 genotyping has the strongest evidence base in psychiatry; most other pharmacogenomic markers on commercial panels lack clinical validation. Always check for CYP inhibitor/inducer drug interactions before attributing unexpected drug levels to genetics -- "phenoconversion" from drug interactions is far more common than genetic poor metabolism. HLA-B*15:02 testing before carbamazepine use is FDA-recommended and potentially life-saving in patients of Southeast Asian descent. Pharmacogenomic testing complements but does not replace therapeutic drug monitoring for drugs where TDM is available (lithium, TCAs, clozapine, valproate) The MTHFR gene is frequently included on commercial panels and marketed to patients, but current evidence does not support its clinical utility in guiding psychiatric treatment. CPIC guidelines (cpicpgx.org) are the gold standard, peer-reviewed resource for gene-drug pair recommendations -- use these rather than proprietary panel interpretations. Pharmacogenomic testing should never delay initiation of urgently needed psychiatric treatment.

## References

- Hicks JK, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants. *Clin Pharmacol Ther*. 2017;102(1):37-44.
- Greden JF, et al. Impact of pharmacogenomics on clinical outcomes in major depressive disorder in the GUIDED trial. *J Clin Psychiatry*. 2019;80(2):19m12723.
- Bousman CA, et al. Pharmacogenomic-guided treatment for depression: Clinical Pharmacogenomics Implementation Consortium guideline. *Clin Pharmacol Ther*. 2023;114(1):51-66.
- FDA Table of Pharmacogenomic Biomarkers in Drug Labeling. Updated periodically.
- Luzum JA, et al. The pharmacogenomics research network translational pharmacogenetics program. *Clin Pharmacol Ther*. 2021;109(6):1430-1440.
- Zeier Z, et al. Clinical implementation of pharmacogenetic testing in a health system: a practical guide. *Nat Rev Genet*. 2021;22(4):257-275.