A CYP2D6 and CYP2C19 psychiatric medication pharmacogenetic test examines inherited differences in two liver-enzyme genes that help process many antidepressants. The result may explain why a standard dose produces unusually high exposure, troublesome adverse effects, low exposure, or little benefit from certain medications. It can also help a prescriber choose a different drug, adjust a starting dose, slow titration, or monitor more closely. However, this is not a test for depression, anxiety, bipolar disorder, or another psychiatric diagnosis. It does not measure current drug levels, predict a person’s full response to treatment, or identify one antidepressant that is certain to work. The value of the result depends on the exact drug, the variants the laboratory tested, coexisting medications, age, liver and kidney function, prior treatment history, and clinical symptoms. Interpretation should therefore connect the laboratory phenotype—such as poor, intermediate, normal, rapid, or ultrarapid metabolizer—to a specific medication and an up-to-date prescribing guideline.
- CYP2D6 and CYP2C19 results mainly predict how quickly certain antidepressants are metabolized and how much drug exposure may occur.
- A “poor metabolizer” result does not mean poor treatment response; it usually means reduced enzyme activity and potentially higher exposure to an affected drug.
- A “rapid” or “ultrarapid” result can lower exposure to some medications, but the clinical effect varies by drug.
- Strong enzyme inhibitors can override the inherited result, creating a temporary functional phenotype called phenoconversion.
- Pharmacogenetic guidance is drug-specific: a result that matters for citalopram may have little or no prescribing relevance for another antidepressant.
Table of Contents
- What the Test Can and Cannot Tell You
- How CYP2D6 and CYP2C19 Results Are Generated
- Understanding Metabolizer Phenotypes
- Antidepressant-Specific Interpretation
- Drug Interactions and Phenoconversion
- When Testing May Be Most Useful
- Limitations and Test Quality
- Using Results With a Prescriber
What the Test Can and Cannot Tell You
CYP2D6 and CYP2C19 encode cytochrome P450 enzymes that participate in the metabolism of many medications. Genetic variation can reduce, eliminate, preserve, or increase enzyme activity. When an antidepressant depends substantially on one of these pathways, inherited enzyme activity can influence the concentration of the parent drug, an active metabolite, or both.
That information can be clinically useful because antidepressant exposure affects tolerability and, for some drugs, the probability of response. Excess exposure may contribute to nausea, sedation, agitation, sexual adverse effects, anticholinergic symptoms, blood-pressure changes, or cardiac risk, depending on the medication. Low exposure may contribute to inadequate response at an otherwise reasonable dose. The relationship is not universal, however. Some antidepressants use several metabolic pathways, some have active metabolites, and some have wide therapeutic ranges. For those drugs, a genotype-associated change in one enzyme may have limited clinical importance.
The test does not directly measure whether an antidepressant will relieve symptoms. Treatment response also depends on diagnosis, symptom pattern, illness severity, comorbid conditions, adherence, dose, duration, psychosocial stressors, psychotherapy, sleep, substance use, and other biological factors. A normal-metabolizer result does not guarantee efficacy or freedom from adverse effects. Likewise, an atypical phenotype does not prove that a past treatment failure or side effect was caused by genetics.
Pharmacogenetic testing also does not diagnose a psychiatric condition. It cannot distinguish unipolar depression from bipolar depression, identify suicidality, determine whether symptoms are medication-induced, or replace a clinical assessment. Screening for bipolar disorder before starting an antidepressant remains important because antidepressants can precipitate mania or hypomania in susceptible people regardless of CYP2D6 or CYP2C19 genotype.
Commercial reports sometimes place medications into colored categories such as “use as directed,” “use with caution,” or “consider an alternative.” These categories are summaries created by the testing company, not universal clinical standards. Two laboratories can test different variants, use different algorithms, or combine multiple genes differently and therefore produce different medication classifications. The most transferable parts of the report are the actual genotype or diplotype, the predicted phenotype, the tested allele list, and any copy-number findings.
How CYP2D6 and CYP2C19 Results Are Generated
Most laboratories report a pair of star alleles, called a diplotype. Examples include CYP2C19 1/2 or CYP2D6 1/4. Each star allele represents a defined combination of DNA variants associated with normal, decreased, absent, or increased function. The laboratory then translates the diplotype into a predicted metabolizer phenotype.
CYP2C19 testing is usually technically straightforward compared with CYP2D6, although the result still depends on which alleles were examined. Common no-function alleles include *2 and *3, while *17 is associated with increased expression. A person with two no-function alleles is generally classified as a poor metabolizer. A person with one no-function allele and one normal-function allele is generally an intermediate metabolizer. One or two increased-function alleles can produce rapid or ultrarapid metabolism, depending on the allele combination.
CYP2D6 is more complex. The gene can have single-nucleotide variants, small insertions or deletions, whole-gene deletions, duplications, multiplications, hybrid genes, and rearrangements involving the nearby CYP2D7 pseudogene. A result such as *1xN indicates multiple copies of an allele, but the clinical meaning depends on which allele was duplicated. Extra copies of a normal-function allele can increase activity; extra copies of a no-function allele do not restore function.
CYP2D6 laboratories commonly assign an activity value to each allele and add the values to calculate an activity score. The score is then translated into poor, intermediate, normal, or ultrarapid metabolizer status. Because professional groups periodically refine allele-function assignments and phenotype translation, an older report may not use the same terminology as a current guideline.
Test coverage matters. A limited assay may detect only a small set of common variants and may miss alleles that are more frequent in a particular ancestry group. It may also fail to resolve copy number or structural variation accurately. “No variant detected” on a limited panel does not prove that both genes are fully normal. The report should identify the alleles tested, the method used for CYP2D6 copy-number analysis, whether structural variants were assessed, and any limitations in ancestry representation.
Sample type usually has little effect on the inherited result. Blood, saliva, or a cheek swab can all provide DNA. The genotype generally remains stable throughout life, so a valid result can inform future prescribing. The clinical interpretation can change, however, as new evidence emerges, medication lists change, and guidelines are updated.
Understanding Metabolizer Phenotypes
A metabolizer phenotype describes predicted enzyme activity, not a person’s overall metabolism or health. It is also not a judgment about whether someone is a “good” or “poor” candidate for treatment.
Poor metabolizer
A poor metabolizer has little or no functional activity for the relevant enzyme. For a drug that is inactivated mainly by that enzyme, concentrations may rise at standard doses and adverse effects may become more likely. For a prodrug that requires enzyme activation, poor metabolism can instead reduce formation of the active compound. Most antidepressants discussed in CYP2D6 and CYP2C19 guidance are interpreted primarily through changes in parent-drug exposure, but active metabolites can complicate the pattern.
Intermediate metabolizer
An intermediate metabolizer has reduced activity. The effect is usually less pronounced than in a poor metabolizer and may be clinically important only for selected drugs or in combination with other factors. A strong enzyme inhibitor, older age, liver impairment, or another interacting medication can make reduced activity more consequential.
Normal metabolizer
A normal metabolizer has enzyme activity within the reference range used by the guideline. Standard starting recommendations usually apply when there are no other reasons to adjust treatment. This classification does not guarantee that a standard dose will be effective or tolerated.
Rapid and ultrarapid metabolizer
CYP2C19 distinguishes rapid from ultrarapid metabolizers. CYP2D6 commonly uses the ultrarapid category without a separate rapid category. Increased enzyme activity may reduce exposure to a medication that is cleared by that pathway, potentially lowering the chance of benefit. In other situations, active metabolites or multiple pathways make the consequence uncertain. Guideline recommendations may therefore range from choosing an alternative drug to using usual dosing with monitoring.
Phenotypes should always be interpreted for a named medication. “CYP2D6 poor metabolizer” is a lifelong genetic classification, but it does not mean every CYP2D6 substrate requires a dose reduction. The fraction of the drug cleared by CYP2D6, its therapeutic window, active metabolites, and available clinical evidence determine whether action is recommended.
Antidepressant-Specific Interpretation
Professional guidance links genotypes to individual antidepressants rather than treating all drugs in a class as equivalent. The following examples describe common patterns, not personalized dosing instructions.
| Medication or group | Primary result often considered | Typical interpretation issue |
|---|---|---|
| Citalopram and escitalopram | CYP2C19 | Rapid or ultrarapid metabolism may lower exposure; poor metabolism may raise exposure and adverse-effect risk. |
| Sertraline | CYP2C19, with CYP2B6 also relevant in current guidance | Reduced metabolism can increase exposure; recommendations depend on the combined evidence and phenotype. |
| Paroxetine | CYP2D6 | Poor metabolism can increase exposure, while ultrarapid metabolism can lower exposure; paroxetine also inhibits CYP2D6. |
| Fluvoxamine | CYP2D6 | Poor metabolism may increase exposure, but recommendations and evidence strength differ from those for paroxetine. |
| Venlafaxine | CYP2D6 | Poor metabolism changes the balance between venlafaxine and its active metabolite; an alternative may be considered in some guidelines. |
| Vortioxetine | CYP2D6 | Poor metabolism increases exposure and is associated with a lower recommended maximum dose in labeling and guidance. |
| Tricyclic antidepressants | CYP2D6 and CYP2C19 | Both formation and clearance of active compounds may be affected; therapeutic drug monitoring can be especially useful. |
For citalopram and escitalopram, CYP2C19 rapid or ultrarapid metabolism can produce lower concentrations and a greater chance of inadequate response at usual doses. Guidance often favors an alternative antidepressant that is not predominantly metabolized by CYP2C19 rather than automatically escalating the dose. CYP2C19 intermediate and poor metabolizers may have higher concentrations. Slower titration, a lower maintenance dose, or an alternative may be appropriate. For citalopram specifically, the FDA limits the maximum recommended dose to 20 mg daily in CYP2C19 poor metabolizers because higher exposure increases the risk of QT prolongation.
Sertraline is metabolized through several pathways. CYP2C19 remains clinically relevant, and current guidance also considers CYP2B6. A CYP2C19 poor-metabolizer result can support a lower starting dose, slower titration, a reduced maintenance dose, or selection of an alternative, especially when adverse effects have already occurred. A report limited to CYP2D6 and CYP2C19 may therefore provide only part of the pharmacogenetic picture for sertraline.
Paroxetine depends substantially on CYP2D6 and is also a strong CYP2D6 inhibitor. A poor metabolizer can have increased exposure, while an ultrarapid metabolizer may have lower exposure and reduced probability of benefit. The drug’s self-inhibition makes dose-concentration relationships nonlinear and reinforces the need for clinical monitoring rather than simple arithmetic based on genotype.
Venlafaxine is converted by CYP2D6 to O-desmethylvenlafaxine, an active metabolite. Poor metabolizers tend to have more parent drug and less metabolite, even though the combined active exposure may not change in a simple way. Some guidance suggests considering a different antidepressant for CYP2D6 poor metabolizers because tolerability or effectiveness may be altered. Desvenlafaxine, which is the active metabolite, is less dependent on CYP2D6 for its disposition.
Vortioxetine exposure is higher in CYP2D6 poor metabolizers. The FDA pharmacogenetic association table and prescribing information identify a lower maximum dose for this group. This is an example of a result with a clear drug-specific labeling consequence.
Fluoxetine is often included on psychiatric medication panels, but CYP2D6 genotype has a less straightforward clinical relationship because fluoxetine and norfluoxetine are both active and both inhibit CYP2D6. Current CPIC guidance does not provide a genotype-based dosing recommendation for fluoxetine because evidence has not established a reliable action that improves outcomes. A colored commercial report should not be treated as stronger evidence than the underlying guideline.
Tricyclic antidepressants such as amitriptyline, nortriptyline, imipramine, clomipramine, and doxepin can be influenced by both CYP2C19 and CYP2D6. One enzyme may contribute to conversion of a tertiary amine to an active secondary amine, while the other helps clear active drug. Interpretation may therefore require both genotypes, the particular tricyclic, clinical indication, electrocardiographic risk, and serum drug concentrations. Therapeutic drug monitoring can complement pharmacogenetics when a tricyclic is necessary.
Drug Interactions and Phenoconversion
Genotype predicts baseline enzyme capacity, but the medication list determines how much of that capacity is available today. Phenoconversion occurs when a drug or other factor changes actual enzyme activity enough that the person behaves like a different metabolizer phenotype.
CYP2D6 inhibition is especially important in psychiatry. Bupropion, fluoxetine, and paroxetine are strong inhibitors. A genetically normal CYP2D6 metabolizer taking a strong inhibitor may function clinically like a poor metabolizer for another CYP2D6 substrate. Moderate inhibitors can shift activity to an intermediate range. The effect usually resolves after the inhibitor is stopped and cleared, although the timing depends on the drug and its metabolites.
CYP2C19 inhibitors can similarly increase exposure to affected medications. Omeprazole and esomeprazole are common examples that may reduce CYP2C19 activity. Enzyme inducers can lower exposure by increasing metabolic capacity, though induction involves several pathways and is not captured by a static genetic report.
Drug interactions can also occur beyond CYP2D6 and CYP2C19. Antidepressants may interact through CYP3A4, CYP1A2, CYP2B6, transporters, protein binding, or pharmacodynamic effects. Combining serotonergic agents can increase serotonin-syndrome risk. Drugs that prolong the QT interval can add to citalopram-related cardiac risk. Anticoagulants and antiplatelet drugs can increase bleeding risk with serotonin reuptake inhibitors. None of these issues is ruled out by a favorable pharmacogenetic category.
Smoking, inflammation, pregnancy, liver disease, kidney disease, age, and nutritional status may also alter drug exposure. CYP1A2 induction from smoking is particularly relevant to some psychiatric medications, but it is outside a CYP2D6/CYP2C19-only result. A medication review should include prescription drugs, nonprescription medicines, supplements, cannabis products, nicotine use, and recent medication changes.
When Testing May Be Most Useful
Testing can be considered before treatment, but its practical value is often greatest when there is a specific prescribing question. Examples include repeated adverse effects at low or standard doses, unexpectedly poor response despite adequate adherence and duration, a history of several unsuccessful antidepressant trials, a need to use a medication with a well-established gene–drug guideline, or a complex regimen with significant interaction potential.
A prior result can also be useful when restarting a medication, changing clinicians, or considering a new drug years later. Because the genotype is stable, the original laboratory report should be retained. The interpretation may need updating if the test was performed years ago or if the report did not assess CYP2D6 copy number and structural variants.
Preemptive testing may reduce trial-and-error for selected medications, but evidence for broad commercial combinatorial panels is mixed. Clinical trials have sometimes found modest improvements in remission or response, particularly when prescribing avoids substantial gene–drug interactions. Other analyses have found small effects, methodological limitations, or uncertainty about which proprietary algorithm produced the benefit. Testing should be viewed as one decision-support tool rather than a replacement for careful follow-up.
Testing is not necessarily needed when a person is doing well on a stable regimen. A result that predicts atypical metabolism does not automatically justify changing an effective, well-tolerated medication. The observed clinical response already incorporates genotype, dose, interactions, adherence, and many other factors. Abruptly changing treatment solely because of a new report can cause relapse or discontinuation symptoms.
Children and adolescents require additional caution. Developmental changes in enzyme activity, age-specific evidence, pediatric dosing, and the higher importance of close monitoring for behavioral activation or suicidality can limit direct extrapolation from adult studies. The result may still be informative, but prescribing should follow pediatric expertise and drug-specific evidence.
Limitations and Test Quality
A high-quality report should state the genotype, phenotype, tested variants, analytical method, copy-number capability, interpretation standard, and date of interpretation. It should distinguish laboratory certainty from prescribing recommendations. An indeterminate or ambiguous CYP2D6 result may require a more comprehensive assay rather than forcing a phenotype assignment.
Ancestry affects allele frequencies but should not be used as a substitute for testing. Some uncommon or population-enriched alleles are absent from limited panels, increasing the chance of misclassification in underrepresented groups. Broad allele coverage and validated structural-variant analysis are more important than using race or ethnicity as a dosing proxy.
Laboratory phenotype calls may disagree because tests interrogate different alleles or because translation standards have changed. When results conflict, compare the raw diplotype, copy-number findings, and methodology. A clinical pharmacogenetics specialist, pharmacist, genetic counselor, or laboratory director may help resolve discrepancies.
Many psychiatric panels include genes such as SLC6A4 and HTR2A. Current CPIC evidence does not support using these two genes to guide serotonin reuptake inhibitor prescribing. Their presence on a commercial panel does not make their medication recommendations equivalent to well-established CYP2D6 or CYP2C19 guidance. Other genes may be relevant for selected psychiatric drugs—for example, HLA variants for severe cutaneous reactions with certain antiseizure medications used in psychiatry—but they answer different questions.
A report can also become outdated. Star-allele definitions, phenotype translation, and prescribing recommendations evolve. Electronic health records should preserve the genotype while allowing the clinical interpretation to be refreshed. Patients should avoid relying only on a screenshot of a color category without the underlying allele information.
Finally, pharmacogenetics cannot account for every source of variability. It does not measure present serum concentration, confirm adherence, predict placebo response, identify psychosocial contributors, or determine the best psychotherapy. Therapeutic drug monitoring, symptom scales, electrocardiography, laboratory tests, and close clinical review may provide information that genetics cannot.
Using Results With a Prescriber
Bring the complete laboratory report, not only the summary page, to the prescribing visit. The discussion should identify the current or proposed antidepressant, the relevant gene, the genotype and phenotype, the evidence level, interacting medications, prior responses, and the clinical goal.
Useful questions include whether the recommendation comes from CPIC, the Dutch Pharmacogenetics Working Group, an FDA label, or a proprietary laboratory algorithm; whether the test assessed CYP2D6 copy number and structural variants; whether any current medication is causing phenoconversion; and whether dose adjustment, slower titration, an alternative drug, or additional monitoring is most appropriate.
Medication changes should be individualized. A lower starting dose may reduce adverse effects but could delay benefit if titration is too cautious. Choosing an alternative avoids a gene–drug interaction but introduces a different adverse-effect profile and different interactions. Increasing a dose to compensate for rapid metabolism may be unsafe when evidence is limited. The best choice balances the pharmacogenetic result with the person’s treatment history, symptoms, preferences, comorbidities, pregnancy status, cardiac risk, and other medications.
Do not stop an antidepressant suddenly based on a test result. Abrupt discontinuation can cause dizziness, flu-like symptoms, insomnia, irritability, sensory disturbances, anxiety, and recurrence of the underlying condition. Tapering plans vary by medication, dose, treatment duration, and individual sensitivity.
Urgent clinical assessment is warranted for new suicidal thoughts, severe agitation, mania, confusion, high fever, muscle rigidity, fainting, a fast or irregular heartbeat, seizures, or symptoms suggesting serotonin syndrome. A pharmacogenetic result may help explain exposure, but it does not replace immediate evaluation of a serious reaction.
The most useful interpretation is concise and specific: this person has a defined CYP2D6 or CYP2C19 phenotype; this medication is substantially affected by that pathway; current drugs do or do not alter the functional phenotype; and an evidence-based guideline supports a particular action. When those links are weak, the result should be treated as supplementary rather than decisive.
References
- Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A Genotypes and Serotonin Reuptake Inhibitor Antidepressants — 2023 guideline.
- Dutch Pharmacogenetics Working Group Guideline for the Gene–Drug Interaction Between CYP2C19 and CYP2D6 and SSRIs — 2024 guideline.
- Dutch Pharmacogenetics Working Group Guideline for CYP2D6, CYP2C19, and Non-SSRI/Non-TCA Antidepressants — 2024 guideline.
- FDA Table of Pharmacogenetic Associations — 2024 regulatory resource.
- Celexa Prescribing Information — 2022 FDA label.
Disclaimer
This article is for general education and is not a diagnosis, prescription, or individualized medication recommendation. Pharmacogenetic results should be interpreted by a qualified prescriber or pharmacist together with the complete medication list, clinical history, and current treatment response. Do not start, stop, or change an antidepressant dose solely because of a genetic test result, and seek urgent care for suicidal thoughts or a severe medication reaction.
