Pharmacogenomics: Why the Same Drug Affects People Differently

Two people take 20 mg of citalopram for depression. For one, it's the right dose. For another, it's barely detectable. For a third, it's equivalent to 40 mg and causes side effects. The difference isn't placebo or psychology. It's written in their DNA.

Pharmacogenomics is the study of how genetic variation affects drug metabolism. Your genes determine which enzymes your liver uses, how efficiently those enzymes work, and therefore how quickly you break down drugs. This variation is enormous - between-person differences in drug metabolism can span 5-10 fold for the same drug. Standard dosing ignores this entirely.

The cytochrome P450 system: your drug-metabolising factory

Your liver metabolises nearly every drug through a family of enzymes called cytochrome P450 (abbreviated CYP). These enzymes are the workhorses. CYP2D6, CYP2C19, and CYP3A4 are the three most important for psychiatric and many other medications.

Genetic variation affects how many copies of these genes you have and how well they work. Most people are "normal metabolisers" - they have two functional copies and metabolise drugs at population average rates. But about 7-10% are "poor metabolisers" (slow or non-functional copies) and 5-10% are "rapid metabolisers" (extra copies or hyperactive variants).

Poor metabolisers accumulate drug in their system. Standard doses become excessive doses. They experience side effects at doses most people tolerate. Rapid metabolisers burn through drugs quickly. Standard doses have minimal effect - they need higher doses to get therapeutic levels.

CYP2D6: the most variable enzyme system

CYP2D6 metabolises about 25% of all medications, including many antidepressants, antipsychotics, and ADHD medications (atomoxetine, guanfacine). It's also the most genetically variable - people can have 0, 1, 2, 3, or even 4 functional copies.

Medications significantly affected by CYP2D6 metabolism include:

• Atomoxetine (Strattera) - metabolised almost exclusively by CYP2D6
• Most tricyclic antidepressants (amitriptyline, nortriptyline)
• Some SSRIs (fluoxetine, paroxetine, venlafaxine)
• Antipsychotics (aripiprazole, risperidone)
• Codeine - requires CYP2D6 to be activated, so poor metabolisers get no pain relief

If you're taking atomoxetine and getting no benefit at standard dose, poor CYP2D6 metabolism might be the answer - a higher dose could work. If you're getting severe side effects at standard dose, rapid metabolism is less likely, but poor metabolism of other substrates could amplify toxicity.

CYP2C19: the SSRI variable

CYP2C19 metabolises citalopram, escitalopram, sertraline, and some other SSRIs. About 2-5% of people are poor metabolisers (particularly common in East Asian populations, up to 10% prevalence). Poor metabolisers accumulate these drugs and experience excessive side effects at standard doses.

This matters clinically. If someone on citalopram 20 mg experiences significant sedation, nausea, or sexual dysfunction but hasn't been on it long, poor CYP2C19 metabolism is a likely explanation. Dose reduction to 10 mg might be all that's needed.

Rapid metabolisers conversely might need higher doses for therapeutic effect. The FDA labeling for citalopram recommends dose reduction in CYP2C19 poor metabolisers, but many prescribers don't know this or don't use testing to identify who they are.

CYP3A4: the giant and the wild card

CYP3A4 metabolises about 50% of all medications - it's the most important enzyme overall. But it's also less well-mapped genetically. Most variation in CYP3A4 activity comes from non-genetic factors (age, other medications, grapefruit juice) rather than genetic variation.

Genetic testing for CYP3A4 is less standardised and less useful than for CYP2D6 or CYP2C19. This is partly why pharmacogenomic testing typically focuses on the latter two.

DPYD: the cancer drug time bomb

One genetic test is genuinely critical: DPYD testing before starting certain cancer chemotherapy drugs (5-fluorouracil, capecitabine). Poor DPYD metabolisers can experience fatal toxicity from standard doses of these drugs. This is one of the few pharmacogenomic tests recommended universally before drug initiation.

For non-cancer drugs, DPYD testing is less critical, but poor metabolisers of certain medications might require dose adjustment.

Practical pharmacogenomic testing in the UK

Getting pharmacogenomic testing in the UK is frustratingly difficult. The NHS rarely orders it despite clear clinical evidence. Most hospital pharmacies don't offer it. Private tests exist but cost £200-400 and aren't well-integrated into NHS prescribing.

Some private psychiatrists and functional medicine practitioners order tests from companies like Genomind, Tempus, or Myriad. These give detailed metaboliser status and drug recommendations. The information is valuable, but access is limited to private practice.

If you're struggling with medication side effects or non-response, requesting pharmacogenomic testing from your prescriber is reasonable. If they refuse, you can order private testing and bring results to them - many will adjust dosing based on the data even if they didn't order the test initially.

How to use pharmacogenomic information

If you're a poor metaboliser of CYP2D6:

• Expect standard doses to cause more side effects
• Request dose reduction or slower titration
• Monitor carefully for accumulation over time
• Avoid drugs heavily metabolised by CYP2D6 unless necessary

If you're a rapid metaboliser of CYP2D6:

• Expect standard doses to have minimal effect
• Request higher doses or more frequent dosing
• Ensure you're at therapeutic level before concluding a drug doesn't work
• Monitor drug levels if possible

The principle applies to CYP2C19 with the same logic.

The future: personalised prescribing

Eventually, pharmacogenomic testing before starting psychiatric medications will be standard. You'll get testing, results will inform dosing, and "find the right dose" won't take six months of trial-and-error. This is the trajectory of medicine - from population-average dosing to individualised dosing based on biology.

For now, it's an informed conversation with your prescriber if you're struggling with medication response. The science is solid. The practical implementation lags behind.