How to Read Your Pharmacogenomic Results

Your pharmacogenomic report arrived and it says something like: CYP2D6 *1/*4 — Intermediate Metabolizer — Activity Score 1.0.

Three pieces of information, none of which are self-explanatory. This article explains exactly how the lab got from your DNA to that label, what the label predicts, and — critically — the several well-documented reasons that label may not match what your body actually does.

Step 1: The star alleles

PGx reports do not list individual SNPs. They list star alleles — a naming convention like *1, *4, *17. Each star allele is a defined combination of variants that together produce a particular version of the enzyme.

You have two copies of each gene, so you get two star alleles. That pair is your diplotype — e.g. CYP2D6 *1/*4.

Each star allele is assigned a functional category:

CYP2C19*1 encodes a normal function enzyme, while CYP2C19*2 is the most common no-function allele followed by *3. The CYP2C19*17 allele is defined by a promoter-region variant resulting in enhanced gene transcription and therefore increased metabolic capacity.

Step 2: The activity score

For CYP2D6 specifically — the most complex and most clinically important of these genes — each allele is assigned a numeric activity value, and the two are added together.

The CYP2D6 activity score is the sum of the activity values assigned to each allele. Roughly: a normal-function allele scores 1, a decreased-function allele scores 0.25 or 0.5, and a no-function allele scores 0.

There is one extra wrinkle unique to CYP2D6: gene duplication. Some people carry more than two copies. If you have three functional copies, your score exceeds 2, and you metabolise these drugs unusually fast. CYP2D6 is highly polymorphic, and different alleles can lead to impacts ranging from null activity to dramatic increases through gene duplication.

Step 3: Genotype to phenotype

The activity score is then translated into the metabolizer phenotype — the label at the top of your report.

Within the CPIC guidelines, individuals with an activity score of 0 are poor metabolizers, those with 0.5 are intermediate metabolizers, those with 1.0, 1.5 and 2.0 are normal metabolizers, and those scoring above 2 are ultrarapid metabolizers.

CPIC — the Clinical Pharmacogenetics Implementation Consortium — is the body that publishes the evidence-based guidelines linking these phenotypes to prescribing recommendations. When a report cites CPIC, it is citing the field's standard.

What the phenotype actually predicts

Here is the part that trips people up, and it is not intuitive: whether "poor metabolizer" means too much drug or too little drug depends entirely on the drug.

Case A: The drug is active as given

Most drugs work in the form you swallow. The enzyme's job is to clear them.

• Poor metabolizer → drug clears slowly → accumulates → higher risk of side effects. May need a lower dose.
• Ultrarapid metabolizer → drug clears fast → may not reach effective levels. May need a higher dose or a different drug.

Case B: The drug is a prodrug

Some drugs arrive inactive and require the enzyme to convert them into the active form. Now everything inverts.

• Poor metabolizer → cannot activate the drug → the drug does not work.
• Ultrarapid metabolizer → converts too much, too fast → dangerous overdose from a standard dose.

The genes that matter most

Four reasons your phenotype label may be wrong

This is the section most PGx reports skip, and it is the most important thing on this page.

1. Phenoconversion — other drugs change your phenotype

Your genotype is fixed. Your actual enzyme activity is not, because other drugs you take can inhibit the same enzyme.

There are commonly used drugs that potently inhibit CYP2D6, causing "phenoconversion" — which has the potential to convert a genotypic normal metabolizer into a phenotypic poor metabolizer. CPIC guidelines recommend adjusting the genotype-based activity score based on the use of inhibitors, with the specific adjustment depending on whether the inhibitor is moderate or strong.

A concrete example: co-administering bupropion (a CYP2D6 inhibitor) with aripiprazole (metabolised by CYP2D6) can lead to excessively high aripiprazole concentrations and potential discontinuation due to adverse effects.

Your report says "normal metabolizer." The drug you started last month has made you a functional poor metabolizer. The report has no idea.

Phenoconversion to a lower metabolizer phenotype has been reported with concomitant CYP450-inhibiting drugs, increasing age, and cancer — so it is not only about drug interactions.

2. Labs disagree with each other — substantially

Read that again. Same DNA. Same genotype. Different phenotype label in nearly three cases out of ten — and sometimes a different drug recommendation as a result.

This is why standardisation efforts exist. CPIC and the Dutch Pharmacogenetics Working Group convened a panel of 37 international CYP2D6 experts using a modified-Delphi method specifically because translating CYP2D6 genotype to phenotype was not standardised across laboratories and guidelines, and the resulting discordance was causing inconsistent therapeutic recommendations and confusion among patients and clinicians.

The practical implication: ask which framework your report used. If it does not say CPIC, that is worth knowing.

3. Rare variants the test never looked for

If an individual carries a rare variant, the actual phenotype may differ from the predicted phenotype. A potential risk is the missed identification of rare or novel variants that are typically not interrogated on clinically used testing platforms.

A targeted PGx panel checks a defined list of known star alleles. If you carry something outside that list, the lab will typically default to calling it *1 — normal function — because absence of a known variant is treated as presence of the reference. That default can be wrong.

This is one genuine advantage of whole genome sequencing in this space: it reads the entire gene rather than a pre-selected set of positions, so novel and rare variants are at least present in the data, even if current tools cannot yet interpret them.

4. Everything else about you

An individual's metabolizer status also depends on factors including epigenetic variation, age, diet, comorbidities, smoking, pregnancy, and concomitant medications.

Genetics is one input. Liver function, kidney function, age, pregnancy and smoking all shift real-world drug metabolism, and none of them appear on your genotype report.

What PGx results genuinely change

Having spent that long on caveats, the technology is real and it matters. Legitimate applications:

• Avoiding a drug that will not work. Roughly 50% of people with major depressive disorder fail their initial SSRI therapy — PGx will not fix all of that, but it can remove some of the guesswork.
• Avoiding predictable toxicity. An estimated 25,000 patients per year in the United States seek emergency department treatment for adverse events associated with antidepressants.
• Dose adjustment. CYP2B6 poor metabolizers, for instance, have greatly reduced metabolism and elevated plasma concentrations, so a lower starting dose, slower titration, and a 25% reduction of standard maintenance doses should be considered.
• High-stakes drugs. Thiopurines, warfarin, clopidogrel, abacavir. Here PGx is not an optimisation — it is a safety check.

The bottom line

Your PGx phenotype is a prediction generated by a defined algorithm from a defined set of variants. It is useful. It is not a measurement of your body, it can be altered by the other drugs you take, and different labs will sometimes convert the same genotype into different labels.

Treat it as a genuinely valuable input to a conversation with your prescriber — not as an instruction, and never as a reason to change a medication on your own.