Pharmacogenomics: How Your Genes Affect Drug Response

Why People Respond Differently to the Same Drug

When a physician prescribes a standard dose of a medication, they are relying on population-average data from clinical trials. But individual patients are not population averages. A drug that works well for 70% of patients may fail completely in 20% and cause serious adverse effects in 10%. Traditional medicine handles this through trial and error: start a drug, observe the response, adjust or switch if necessary. Pharmacogenomics aims to replace this trial-and-error approach with predictive testing.

Genetic variation affects drug response at multiple levels. Polymorphisms in genes encoding drug-metabolizing enzymes change how quickly or slowly a drug is activated, inactivated, or converted to toxic metabolites. Variants in genes encoding drug targets (receptors, enzymes, ion channels) change the sensitivity of the target to the drug. Variants in drug transporters alter the distribution of drugs into and out of cells and tissues. Variants in immune system genes (particularly HLA genes) determine susceptibility to immune-mediated drug hypersensitivity reactions.

The magnitude of these genetic effects can be enormous. A CYP2D6 ultrarapid metabolizer may have codeine-to-morphine conversion rates 50 to 100 times higher than a poor metabolizer, meaning a standard codeine dose that provides no pain relief in one patient causes respiratory arrest in another. A patient carrying the HLA-B*5701 allele has a nearly 50% chance of developing a severe, potentially fatal hypersensitivity reaction to abacavir, while a patient without this allele has virtually zero risk. These are not subtle differences that require statistical analysis to detect; they are clinically dramatic effects caused by identifiable genetic variants.

Drug-Metabolizing Enzyme Polymorphisms

Genetic variation in CYP enzymes is the most extensively studied area of pharmacogenomics. Individuals are classified into metabolizer phenotypes based on the predicted activity of their CYP enzyme variants:

Poor metabolizers (PMs) carry two non-functional alleles and have little or no enzyme activity. Drugs that are inactivated by the enzyme accumulate to higher levels, increasing the risk of dose-dependent toxicity. Drugs that are activated by the enzyme (prodrugs) have reduced or absent efficacy.

Intermediate metabolizers (IMs) carry one reduced-function or one non-functional allele paired with a functional allele. They have reduced but not absent enzyme activity and may need dose adjustments for sensitive drugs.

Normal (extensive) metabolizers (NMs/EMs) carry two functional alleles and metabolize drugs at the expected rate. Standard dosing recommendations are based on this phenotype.

Ultrarapid metabolizers (UMs) carry duplicated or amplified functional genes, producing enzyme activity well above normal. Drugs inactivated by the enzyme are cleared faster, potentially falling below therapeutic levels at standard doses. Prodrugs activated by the enzyme are converted to their active form more rapidly, increasing the risk of toxicity from the active metabolite.

CYP2D6: The Most Polymorphic Drug Enzyme

CYP2D6 exhibits more genetic variation than any other drug-metabolizing enzyme, with over 130 known allelic variants. Approximately 6 to 10% of Caucasians are poor metabolizers (carrying CYP2D6*4/*4 or similar null combinations), while 1 to 2% of Caucasians and up to 29% of certain Ethiopian and Saudi Arabian populations are ultrarapid metabolizers (carrying gene duplications, CYP2D6*1xN or *2xN).

The clinical impact is well documented for several drugs. Codeine is a prodrug converted to morphine by CYP2D6. Poor metabolizers get essentially no pain relief. Ultrarapid metabolizers produce morphine so rapidly that standard doses can cause respiratory depression, leading to FDA contraindications for codeine use in children under 12 and nursing mothers after reports of infant deaths. Tamoxifen, used for estrogen receptor-positive breast cancer, requires CYP2D6-mediated conversion to its active metabolite endoxifen. Poor metabolizers have significantly lower endoxifen levels and may have higher recurrence rates, though guidelines on prospective testing remain under debate. Atomoxetine (for ADHD) is primarily inactivated by CYP2D6; poor metabolizers have 10-fold higher plasma levels and need lower doses to avoid excessive side effects.

CYP2C19: Clopidogrel and Beyond

CYP2C19 polymorphisms have the most dramatic clinical impact on clopidogrel (Plavix), a widely prescribed antiplatelet drug. Clopidogrel is a prodrug requiring CYP2C19-mediated activation. The CYP2C19*2 and *3 loss-of-function alleles are common (approximately 25 to 35% of Caucasians and 55 to 70% of East Asians carry at least one). Poor metabolizers have significantly higher rates of major cardiovascular events (heart attack, stroke, stent thrombosis) after coronary stent placement. The FDA black box warning recommends considering CYP2C19 testing and using alternative antiplatelet agents (prasugrel, ticagrelor) in patients identified as poor metabolizers.

CYP2C19 also affects proton pump inhibitor (PPI) metabolism. Poor metabolizers have higher omeprazole levels and better acid suppression, while ultrarapid metabolizers may not achieve adequate acid control at standard doses. In H. pylori eradication therapy, CYP2C19 poor metabolizers have higher eradication rates because they maintain higher PPI levels throughout the treatment period.

CYP2C9 and VKORC1: Warfarin Dosing

Warfarin is metabolized primarily by CYP2C9 and acts by inhibiting vitamin K epoxide reductase (VKORC1). Genetic variants in both genes significantly affect the dose required to achieve a target INR. CYP2C9*2 and *3 variants reduce warfarin metabolism, requiring lower doses. A common VKORC1 promoter polymorphism (-1639G>A) reduces VKORC1 expression, making the target more sensitive to warfarin and again requiring lower doses. Together, CYP2C9 and VKORC1 genotypes, along with clinical factors like age and body size, account for approximately 40 to 50% of the variability in warfarin dose requirements.

The FDA-approved label for warfarin includes a dosing table based on CYP2C9 and VKORC1 genotypes. Clinical trials (COAG and EU-PACT) showed that genotype-guided warfarin dosing improved time in therapeutic range compared to standard clinical dosing in European populations, though results were less clear in more genetically diverse populations.

Drug Target Polymorphisms

Genetic variation in drug targets, not just metabolizing enzymes, also affects drug response. Beta-2 adrenergic receptor (ADRB2) polymorphisms influence the response to beta-agonist bronchodilators in asthma. The Arg16Gly polymorphism is associated with reduced bronchodilator response and increased risk of desensitization with regular short-acting beta-agonist use.

Serotonin transporter (SLC6A4) polymorphisms affect SSRI antidepressant response. The 5-HTTLPR long/long genotype has been associated with better SSRI response in some studies, though the effect sizes are modest and clinical implementation remains limited. OPRM1 polymorphisms (encoding the mu-opioid receptor) affect opioid analgesic requirements: the A118G variant has been associated with reduced morphine potency, requiring higher doses for adequate pain control in carriers.

HLA-Linked Drug Hypersensitivity

Some of the most serious adverse drug reactions, severe cutaneous reactions and drug-induced hypersensitivity, are linked to specific human leukocyte antigen (HLA) alleles. These reactions occur because the drug or its metabolite binds to a particular HLA protein on antigen-presenting cells, triggering a T-cell-mediated immune response.

HLA-B*5701 and abacavir: Abacavir, an HIV nucleoside reverse transcriptase inhibitor, causes a potentially fatal hypersensitivity syndrome in approximately 5 to 8% of Caucasian patients. The risk is almost entirely confined to carriers of HLA-B*5701. Prospective testing for HLA-B*5701 before prescribing abacavir has virtually eliminated this reaction and is now standard of care worldwide. This is one of the clearest success stories in pharmacogenomics.

HLA-B*1502 and carbamazepine: Carbamazepine, an anticonvulsant, causes Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) predominantly in carriers of HLA-B*1502, which is common in Southeast Asian populations (up to 15% prevalence) but rare in Caucasians. The FDA recommends testing for HLA-B*1502 in patients of Southeast Asian descent before starting carbamazepine.

HLA-B*5801 and allopurinol: Allopurinol, used for gout, causes severe cutaneous adverse reactions predominantly in HLA-B*5801 carriers. This allele is found in approximately 6 to 8% of African Americans and up to 8% of Han Chinese but only 1 to 2% of Caucasians. The American College of Rheumatology recommends testing prior to allopurinol initiation in high-risk populations.

Pharmacogenomic Testing in Clinical Practice

Pharmacogenomic testing is available through multiple commercial platforms and academic medical centers. Tests range from single-gene assays (testing for HLA-B*5701 before abacavir) to multi-gene panels that simultaneously genotype dozens of pharmacogenes. The Clinical Pharmacogenetics Implementation Consortium (CPIC) publishes evidence-based guidelines for 24 gene-drug pairs, providing specific dosing recommendations based on genotype.

The FDA has included pharmacogenomic information in the labeling of over 400 drugs. The level of recommendation varies: some labels require testing (abacavir/HLA-B*5701), some recommend testing (clopidogrel/CYP2C19), and some provide informational content about genetic effects without mandating testing. The number of drugs with pharmacogenomic labeling continues to grow as evidence accumulates.

Pre-emptive pharmacogenomic testing, genotyping a patient for multiple pharmacogenes before any drug is prescribed, is being implemented at several health systems. The idea is to have the genetic information available in the electronic health record so that when any relevant drug is prescribed in the future, the system can automatically alert the clinician to genotype-based dosing recommendations. Studies at institutions including St. Jude Children's Research Hospital and Vanderbilt University Medical Center have demonstrated the feasibility and clinical utility of this approach.

Limitations and Future Directions

Pharmacogenomics has clear limitations. Genetic variation explains only a portion of drug response variability; environmental factors (diet, smoking, other medications), disease state, organ function, adherence, and the microbiome all contribute. For many gene-drug pairs, the evidence is strong enough to inform clinical guidelines but not strong enough to justify universal pre-prescription testing. Cost-effectiveness data, particularly for multi-gene panels, is still being generated.

Ancestry-related disparities are a concern. Most pharmacogenomic studies have been conducted in populations of European descent, and the frequency and clinical significance of many pharmacogene variants differ across ethnic groups. Expanding research to include diverse populations is essential for ensuring that pharmacogenomics benefits all patients equitably.

Despite these limitations, the field is moving steadily toward broader clinical adoption. As whole-genome sequencing costs continue to fall (now under $200 per genome), the economic argument for pre-emptive pharmacogenomic testing becomes increasingly favorable. The integration of pharmacogenomic data into electronic health records with clinical decision support alerts is making it practical for busy clinicians to use genetic information at the point of prescribing. The ultimate goal, selecting the right drug at the right dose for each patient based on their individual genetic profile, is progressively becoming clinical reality.