The Enzymes That Control More Than You Think

The cytochrome P450 (CYP) enzyme family is responsible for metabolising approximately 70–80% of all clinically used drugs. Two members of this family — CYP2D6 and CYP2C19 — are among the most pharmacogenomically variable genes in the human genome. That variability has direct, clinically actionable consequences for how medications work in your body.

The problem: most patients have never been genotyped for these variants, and most prescribers don’t routinely order pharmacogenomic testing. According to a 2026 study published in NCBI/PMC analysing 15,000 patients, pharmacogenomic insights from CYP2D6/CYP2C19 profiling uncovered “novel actionability, allele-specific copy number variation, and phenoconversion” at meaningful frequencies — findings that changed prescribing decisions in a substantial portion of patients.

Let’s break down what this actually means.

CYP2D6: Metabolizer Categories and Why They Matter

CYP2D6 metabolises approximately 25% of commonly prescribed medications, including many antidepressants (SSRIs, SNRIs, tricyclics), antipsychotics, opioids, beta-blockers, and tamoxifen.

Based on which CYP2D6 alleles you carry, you’re classified as:

Poor Metabolizer (PM): ~5–10% of European populations; higher in certain ethnic groups. You have minimal CYP2D6 enzyme activity. Drugs that CYP2D6 metabolises accumulate to higher-than-expected plasma concentrations — increasing both efficacy and toxicity risk. Prodrugs that require CYP2D6 activation (like codeine → morphine, or tamoxifen → endoxifen) don’t work effectively.

Intermediate Metabolizer (IM): ~10–15% of the population. Reduced but not absent CYP2D6 activity. Sits between poor and normal metabolizers on the activity spectrum.

Normal Metabolizer (NM): The majority of the population. Standard dosing assumptions are calibrated to this group.

Ultrarapid Metabolizer (UM): ~1–5% (higher in North African and East African populations). Gene duplication of functional CYP2D6 alleles results in excess enzyme activity. These individuals metabolise drugs so quickly that standard doses are subtherapeutic. Notably, they can also convert prodrugs (like codeine) to their active form at dangerously rapid rates.

Clinical Examples

Tamoxifen and breast cancer: Tamoxifen, used in hormone-receptor-positive breast cancer treatment, is a prodrug converted to its active metabolite endoxifen by CYP2D6. CYP2D6 poor metabolizers have significantly lower endoxifen levels, which may compromise treatment efficacy. Current CPIC guidelines recommend considering alternative agents (such as aromatase inhibitors) for CYP2D6 poor metabolizers on tamoxifen.

Codeine toxicity in ultrarapid metabolizers: This is the example that made pharmacogenomics famous. CYP2D6 ultrarapid metabolizers who receive standard codeine doses can produce lethal morphine levels. This is documented in fatalities, including in children after tonsillectomy. The FDA has required black box warnings. Codeine is contraindicated in CYP2D6 UMs.

Antidepressants: Tricyclic antidepressants (TCAs) and several SSRIs (fluoxetine is itself a CYP2D6 inhibitor — the paradox of a drug that also inhibits its own metabolism) behave unpredictably in poor metabolizers. Higher plasma concentrations increase QTc-prolongation risk.

CYP2C19: The Proton Pump Inhibitor and Antidepressant Gene

CYP2C19 metabolises a partially overlapping but distinct set of drugs: proton pump inhibitors (PPIs like omeprazole, lansoprazole), clopidogrel (the antiplatelet drug), several benzodiazepines, and many antidepressants including SSRIs.

Population frequencies vary significantly by ethnicity:

  • Poor metabolizers (PM): 2–5% of European populations; 15–25% of Asian populations
  • Intermediate metabolizers (IM): ~25–30% across most populations
  • Rapid/Ultrarapid metabolizers (RM/UM): ~3–5%

Clinical Examples

Clopidogrel and heart attack risk: Clopidogrel is a prodrug that requires CYP2C19 activation to its active thiol metabolite. CYP2C19 poor and intermediate metabolizers have reduced platelet inhibition from clopidogrel — which means the drug meant to prevent stent thrombosis after cardiac stenting isn’t fully working. CPIC guidelines recommend alternative antiplatelets (ticagrelor, prasugrel) for CYP2C19 poor metabolizers in the acute coronary syndrome setting. This is one of the clearest pharmacogenomic actionability cases in cardiology.

Proton pump inhibitors: CYP2C19 poor metabolizers actually get more effective acid suppression from standard PPI doses — because the drug accumulates to higher levels. UMs may need higher doses for adequate acid suppression. This has implications for H. pylori eradication protocols as well as GERD management.

Escitalopram and citalopram: Both are CYP2C19 substrates. Poor metabolizers experience higher plasma levels, increasing QTc-prolongation risk. The FDA dose cap on citalopram (40mg/day, 20mg/day in the elderly) is partly informed by this pharmacogenomic risk.

Phenoconversion: The Complication Nobody Talks About

Here’s where it gets complex: your genotype and your phenotype can diverge — a phenomenon called phenoconversion.

Several drugs are potent CYP2D6 inhibitors: fluoxetine, paroxetine, bupropion, duloxetine. If you’re a CYP2D6 normal metabolizer but you’re taking paroxetine, you are functionally a poor metabolizer while on that drug. Your genotype says “normal” but your clinical behaviour mirrors “poor.”

This matters enormously for polypharmacy. A patient who is a CYP2D6 normal metabolizer genotypically, but is on paroxetine and tamoxifen simultaneously, may have inadequate endoxifen conversion — the same problem as a genetic poor metabolizer.

Drug-drug interactions at the CYP level are predictable from pharmacogenomic knowledge. This is one reason the 15,000-patient study found “novel actionability” that wasn’t visible from genotype alone — they were accounting for drug-gene and drug-drug-gene interactions.

What Testing Options Exist

CPIC-aligned pharmacogenomic panels from companies like GeneSight (Myriad), Genomind, and academic medical centre programs provide CYP2D6/CYP2C19 genotyping with phenotype translation and drug-specific recommendations. These are available with or without insurance coverage, and costs have dropped to $200–500 for comprehensive panels.

Consumer genomics (23andMe, AncestryDNA): These tests genotype some CYP variants but not comprehensively. CYP2D6 in particular has complex structural variation (gene duplications, deletions, hybrid alleles) that standard SNP arrays don’t fully capture. Do not rely on consumer genetics for CYP2D6 phenotype determination.

When is it worth testing? If you’re starting a new antidepressant, are on clopidogrel after cardiac stenting, are taking tamoxifen for breast cancer, have a history of adverse drug reactions or treatment failures at standard doses, or are managing a complex polypharmacy situation — CYP2D6/CYP2C19 testing adds genuine clinical value.

The Pharmacist’s Recommendation

If I had to pick one pharmacogenomic test that delivers the most actionable clinical value per dollar spent, a CYP2D6/CYP2C19 panel from a CPIC-aligned provider wins, particularly if you take any medications in the categories above.

The clinical utility of pharmacogenomic testing is no longer theoretical — it’s validated, guideline-supported (CPIC, DPWG), and increasingly available. The gap is implementation: most prescribers don’t order it, most patients don’t know to ask for it.

That’s a gap this site exists to close.

Sources: Pharmacogenomic insights in psychiatric care: CYP2D6/CYP2C19 in 15,000 patients. NCBI/PMC 2024; CYP2C19 pharmacogenetic variations and clinical implications. NCBI/PMC 2026; Meta-analysis of CYP2D6 and CYP2C19 worldwide variation. Translational Psychiatry; CPIC guidelines: cpicpgx.org; Dutch Pharmacogenetics Working Group (DPWG) dosing recommendations.