Pharmacogenomics

Pharmacogenomics is the study of how a person's genes affect their response to drugs.

8 frameworks, from 1957 to today · Pharmacology · Medicine & Health

still a working view ended · each leads to its entry below

Overview

Pharmacogenomics is the study of how a person's genes affect their response to drugs1. The field rests on a simple but consequential observation: the same dose of the same medicine can cure one patient, do nothing for another, and poison a third. Its practitioners ask which inherited differences explain that variation, how those differences work at the level of enzymes, transporters, and receptors, and how to turn the answer into a test that guides prescribing1. The stakes are practical—safer doses, fewer adverse reactions, and drugs matched to the biology of a tumor rather than to a diagnostic label. The history below follows the field from the first hints that heredity shapes drug response to the present era of panel testing and ancestry-aware guidelines.

Heredity and drug response before the gene

Long before anyone knew what a gene was, clinicians noticed that drug reactions ran in families. The modern field usually dates its founding to a short 1957 paper by the German pharmacologist Friedrich Vogel, who coined the term pharmacogenetics for the study of inherited differences in drug response. But the observations came earlier. In the 1930s, clinicians reported that some patients given the antimalarial primaquine developed severe hemolysis—destruction of red blood cells—while others did not, and that the sensitivity appeared familial. During the Second World War, military physicians noted that some soldiers given the antimalarial drug pamaquine suffered the same reaction. The explanation came in the 1950s, when researchers linked the hemolysis to a deficiency of the enzyme glucose-6-phosphate dehydrogenase (G6PD), a hereditary trait common in some populations. This was an early, concrete demonstration that a single inherited enzyme defect could turn a standard drug into a hazard1.

A second landmark came from anesthesia. In the 1950s, reports accumulated of patients who, after receiving the muscle relaxant succinylcholine during surgery, remained paralyzed for hours instead of minutes1. The cause was eventually traced to inherited variants of the enzyme butyrylcholinesterase, which normally breaks the drug down1. People with low-activity forms of the enzyme cleared the drug slowly and stayed paralyzed longer. These cases established the field's core logic: a drug's fate in the body depends on enzymes, and enzymes vary between people because of heredity.

The rise of pharmacogenetics as a science

Through the 1960s and 1970s, pharmacogenetics grew into a recognizable research program1. Investigators studied twins to separate genetic from environmental contributions to drug handling, and they mapped the inheritance of traits such as the acetylator status—the speed at which a person's liver acetylates certain drugs, governed by the enzyme N-acetyltransferase. The discovery in the 1970s that the anticonvulsant phenytoin and other drugs were metabolized by a family of liver enzymes, the cytochrome P450 system, gave the field its central cast of characters. Debrisoquine, an antihypertensive, became a classic probe: some people metabolized it rapidly and others poorly, and the poor metabolizers were later shown to carry variants in the gene CYP2D6. By the 1980s, the field could describe, for a growing list of drugs, which enzyme handled them and which inherited variants changed the handling.

The crucial conceptual step was the classification of genetic polymorphism—the existence of common variants in a population—into phenotypes. A person could be a poor, intermediate, extensive, or ultrarapid metabolizer of a given enzyme, and that status predicted drug levels. This phenotype language, still in use, made pharmacogenetics clinically legible: it translated a genotype into a prediction about how a patient would handle a drug. The field also learned that the same variant could matter differently depending on the drug and the enzyme's role—a poor metabolizer of a prodrug that must be activated might get no benefit, while a poor metabolizer of a drug that is inactivated might accumulate toxic levels.

From single genes to the genome

The term pharmacogenomics emerged in the 1990s, as the Human Genome Project and new sequencing technologies made it possible to study many genes at once rather than one enzyme at a time. Where pharmacogenetics had focused on single-gene traits, pharmacogenomics promised a genome-wide view: the variants that shape drug response across the whole of a person's DNA. The change was partly rhetorical and partly real. It brought the field into contact with genomics, bioinformatics, and the pharmaceutical industry, which saw in it a way to explain why drugs failed in trials and to rescue compounds that worked only in a subset of patients.

The 1990s and 2000s produced the discoveries that anchor clinical practice today. The enzyme thiopurine methyltransferase (TPMT) was shown to determine toxicity of thiopurine drugs used in leukemia and autoimmune disease; patients with low-activity TPMT accumulated toxic metabolites. Variants in CYP2C19 were linked to the activation of clopidogrel, an antiplatelet drug that is a prodrug, so that poor metabolizers got less protection against clots12. The HLA gene HLA-B\57:01 was found to predict a severe hypersensitivity reaction to the HIV drug abacavir, and HLA-B\15:02 to predict a dangerous skin reaction to the anticonvulsant carbamazepine in some populations. These were not merely associations; they were actionable predictions, and they set the stage for the biomarker era.

Biomarkers, companion diagnostics, and oncology

A pharmacogenomic biomarker is a measurable genetic feature that predicts drug response, and a companion diagnostic is a test marketed alongside a drug to identify the patients who should receive it3. The oncology field drove this model hardest. The breast cancer drug trastuzumab, approved in the late 1990s, was effective only in tumors that overexpressed the HER2 receptor, and testing for HER2 became standard. Later, targeted therapies were paired with tests for mutations in genes such as EGFR, BRAF, and KRAS, so that a lung or melanoma tumor's genotype, not just its tissue of origin, determined the drug. This was the "magic bullet" idea revived at the molecular level: a drug aimed at the specific abnormality driving a cancer.

The oncology model also exposed the field's central tension. Tumor genomics is somatic—mutations acquired by the cancer—while inherited pharmacogenomics concerns the patient's own germline. Both matter, and both require tests, but they answer different questions: whether a drug will hit its target, and whether the patient will handle the drug safely. The companion-diagnostic framework, built for the first question, had to be extended to the second.

Clinical implementation and guidelines

By the 2010s, the field's problem was no longer discovery but translation: how to get a genotype into a prescribing decision at the point of care. Professional consortia took up the task. The Clinical Pharmacogenetics Implementation Consortium (CPIC), formed in 2009, began publishing peer-reviewed guidelines that translate genotype into dosing recommendations for specific drug–gene pairs. The Dutch Pharmacogenetics Working Group produced similar guidance. These bodies did not decide whether to test; they assumed a genotype was available and told clinicians what to do with it1. That framing was deliberate, because the evidence for testing everyone was still thin, while the evidence for acting on a known genotype was stronger.

Implementation also required infrastructure: electronic health records that could store genotypes and fire alerts, laboratory standards for reporting, and reimbursement. Early institutional programs—at St. Jude Children's Research Hospital, which built pharmacogenomic testing into pediatric cancer care, and at several academic medical centers—demonstrated that the workflow could function. The field learned that the barrier was rarely the science; it was the plumbing.

Preemptive panel testing and population-aware practice

The most recent turn is preemptive panel testing: testing a patient for many pharmacogenes at once, before any specific drug is prescribed, so that the result is on file when a prescription is written. This inverts the older model of testing reactively, when a drug is already chosen. Panels covering CYP2D6, CYP2C19, CYP2C9, VKORC1, TPMT, and HLA genes are now offered by several health systems and commercial labs. The approach is attractive because a single test can inform many future prescriptions, but it raises questions about which variants to include, how to handle variants of uncertain significance, and how to keep results current as knowledge changes.

Running alongside this is a reckoning with ancestry. Many pharmacogene variants differ in frequency across populations—CYP2D6 ultrarapid metabolizers are more common in some African and Oceanian populations, HLA-B\15:02 is more frequent in parts of East and Southeast Asia, and CYP2C19* poor metabolizers are common in East Asia and Oceania. Guidelines increasingly warn against using self-identified race as a proxy for genotype, while acknowledging that ancestry-informed testing strategies can improve detection. The field is working toward panels and algorithms that account for population structure without reducing it to a crude category.

The present

Today pharmacogenomics is a mature translational science with a recognized set of drug–gene pairs, published guidelines, and growing clinical adoption. Its main approaches descend directly from the history above: phenotype classification from the early pharmacogenetic era, genome-wide discovery from the 1990s, biomarker-and-companion-diagnostic pairing from oncology, and guideline translation and preemptive testing from the implementation era. The open questions are practical and unresolved: how to generate the randomized evidence that convinces payers and skeptics, how to handle the many variants whose effects are unknown, how to make preemptive testing equitable across populations, and how to keep guidelines synchronized with a literature that grows faster than any clinician can read. The field's founding insight—that heredity shapes drug response—is no longer in doubt; what remains is the work of making that insight reliably useful at the bedside.

References

  1. Pharmacogenomics (introduction, Pharmacokinetics, Off-target sites, History, Applications, Clinical Pharmacogenetics Implementation Consortium (CPIC)), Wikipedia
  2. CYP2C19, Wikipedia
  3. Companion diagnostic, Wikipedia

The frameworks of Pharmacogenomics

Each framework is a way of working the field has used: when it held, and the question it answers.

  1. 1957-1995

    Classical Pharmacogenetics

    How do inherited differences shape drug response?

    Fly to its tower in Nousfarer

  2. 1970-Present

    Metabolizer Phenotype Classification

    How fast does a person metabolize a given drug?

    Fly to its tower in Nousfarer

  3. 1995-Present

    Pharmacogenomics and Genome-Wide Discovery

    Which genes across the genome affect drug response?

    Fly to its tower in Nousfarer

  4. 1998-Present

    Biomarker and Companion Diagnostic Model

    Which genetic test identifies patients for a drug?

    Fly to its tower in Nousfarer

  5. 2000-Present

    Pharmacogenomics of oncology and targeted therapy

    Which tumor mutations predict response to targeted drugs?

    Fly to its tower in Nousfarer

  6. 2009-Present
  7. 2010-Present

    Preemptive Panel Testing

    Should we test many pharmacogenes before prescribing?

    Fly to its tower in Nousfarer

  8. 2015-Present

    Ancestry-Aware Pharmacogenomics

    How do drug-response variants differ across ancestries?

    Fly to its tower in Nousfarer

How the ideas connect

Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.

Classical PharmacogeneticsMetabolizer Phenotype ClassificationPharmacogenomics and Genome-Wide Disc…Biomarker and Companion Diagnostic Mo…Pharmacogenomics of oncology and targ…Guideline Translation and Clinical Im…Preemptive Panel TestingAncestry-Aware Pharmacogenomics

Beyond the field

Where Pharmacogenomics's ideas came from in other fields, and what they led to there: the strongest 12 of 32 links.

A city in Nousfarer: a red-carpeted stair climbing terraces to a crowned tower on an acropolis, more towers round it, forest and snowy mountains beyond.
A city in Nousfarer, the game played in the browser.

Pharmacogenomics is a planet in Nousfarer

Its city has a tower for each of these 8 frameworks. Land, walk in, and climb one concept at a time.

Land on Pharmacogenomics

Related fields

Fields its ideas are linked to

More in Pharmacology

See where Pharmacogenomics sits on the Map of Human Knowledge