CYP2D6 Poor Metabolizer: Medications to Avoid and Pharmacogenomics Results Explained
5–10% of people are CYP2D6 poor metabolizers. Codeine won’t work, antidepressants may reach toxic levels, and tamoxifen loses efficacy. Here’s the full list of affected drugs and what to do.
Category: Diagnostics · By GenomeNarrator Team · · 7 min read
Your liver breaks down most medications before they exit your body. One of the key enzymes responsible for this is CYP2D6, a protein encoded by the CYP2D6 gene that metabolizes roughly 25% of all prescribed drugs. Your genetic makeup determines how fast this enzyme works, and the clinical difference is significant enough that your prescriber should know about it.
Depending on which variants you carry, CYP2D6 places you into one of four categories.
Poor Metabolizers carry two non-functional gene copies. The enzyme is nearly absent, causing drugs to accumulate to potentially toxic levels. This affects approximately 5–10% of people with European ancestry.
Intermediate Metabolizers have reduced enzyme activity: one functional copy paired with one partially working or non-functional copy. They process drugs more slowly than average. This affects 10–15% of Europeans and is the most common abnormal phenotype.
Normal Metabolizers have two working copies and process drugs at the expected rate. Standard dosing guidelines were written for this group.
Ultrarapid Metabolizers carry extra copies of a functional gene (duplication). They clear drugs so rapidly that standard doses may never reach therapeutic levels.
Which drugs are affected, and why it matters
The clinical impact depends entirely on which drug is involved. There are two key mechanisms.
Prodrugs that require CYP2D6 activation: codeine and tramadol are the most critical examples. In poor metabolizers, these drugs are not converted to their active pain-relieving form. They simply don't work. In ultrarapid metabolizers, conversion is so rapid that toxic morphine levels can accumulate, a danger that led to codeine being banned in children under 12 in most countries.
Drugs that CYP2D6 inactivates: antidepressants (fluoxetine, paroxetine, many tricyclics), antipsychotics (haloperidol, risperidone), beta-blockers (metoprolol), and tamoxifen. In poor metabolizers, these accumulate, increasing both efficacy and side effects. In ultrarapid metabolizers, they may be cleared before reaching therapeutic levels.
Tamoxifen deserves special mention. It requires CYP2D6 to be converted to its active form, endoxifen. Poor metabolizers may have significantly reduced benefit from tamoxifen therapy, something that has direct implications for breast cancer treatment decisions.
Tell your prescriber. This is the most important step. Your doctor should know your CYP2D6 status before prescribing any affected medication. For most of the drugs involved, alternatives exist that don't rely on CYP2D6.
Carry documentation. A pharmacogenomics summary, like the medication report you can generate from your GenomeNarrator analysis, is useful to bring to any medical appointment or emergency room visit.
Don't adjust medications yourself. Being a poor metabolizer doesn't automatically mean you're on a wrong dose; it means closer monitoring may be warranted. Never change a prescribed dose without discussing it with your prescriber.
The CPIC evidence framework
The Clinical Pharmacogenomics Implementation Consortium (CPIC) publishes evidence-based dosing guidelines based on CYP2D6 phenotype. For most affected drugs, CPIC Level A evidence supports considering an alternative drug or dose adjustment in poor and ultrarapid metabolizers. GenomeNarrator cross-references CPIC guidelines when generating pharmacogenomics findings.
Consumer testing limitations
SNP arrays detect common CYP2D6 variants (including *2, *3, *4, *5, *6, *9, *10, *17, and *41) but cannot reliably detect copy number variants (gene duplications or deletions). This means ultrarapid metabolizer status from gene duplication, and some intermediate phenotypes, may be missed. Comprehensive CYP2D6 diplotyping requires clinical-grade testing. Any finding from consumer data should be confirmed before clinical decisions are made.
More from GenomeNarrator: all genomics articles, how the analysis works, or GenomeNarrator home.