Precision Medicine: Matching Treatments to Individual Biology

The Day the Human Blueprint Was Read

When the Human Genome Project declared the first complete sequence of the human genome in April 2003, scientists celebrated the achievement with appropriate fanfare — and then immediately confronted a humbling reality. Knowing the three billion base pairs of the human genome was not the same as understanding it. But over the two decades that followed, the genomic revolution gradually transformed medicine, creating an approach now called precision medicine: the use of an individual's genetic, molecular, and environmental profile to tailor prevention, diagnosis, and treatment with far greater specificity than one-size-fits-all clinical practice allows.

From Population Medicine to Individual Biology

Traditional pharmacology assumed that if a drug worked on average in a trial population, it could be prescribed to all patients with the relevant condition. The variability hidden within those averages was accepted as noise. Precision medicine reframes that variability as signal — information about who will respond, who will experience toxicity, and who will derive no benefit at all from a given treatment.

This shift has practical consequences. Trastuzumab (Herceptin) is dramatically effective in women with HER2-positive breast cancer, which accounts for approximately 20% of breast cancers. In HER2-negative patients, it is essentially useless and carries cardiovascular risks. Prescribing it without knowing HER2 status would harm as many patients as it helps. The companion diagnostic — a test that identifies which patients have the target — is therefore as important as the drug itself.

Pharmacogenomics: Genes That Determine Drug Response

Pharmacogenomics examines how genetic variations — single nucleotide polymorphisms, copy number variations, structural variants — alter the absorption, metabolism, distribution, excretion, and target interactions of drugs. The CYP450 enzyme system is the most clinically validated domain of pharmacogenomics. Variants in CYP2D6, CYP2C9, CYP2C19, and other genes determine metabolizer phenotype and dramatically affect drug plasma concentrations and clinical outcomes.

Codeine provides a stark example. Codeine is a prodrug that requires CYP2D6-mediated conversion to morphine for analgesic activity. CYP2D6 ultra-rapid metabolizers — representing roughly 1–3% of Western populations and up to 29% of certain North African populations — convert codeine to morphine so rapidly that standard doses cause morphine toxicity, including potentially fatal respiratory depression. Several pediatric deaths attributable to codeine in ultra-rapid metabolizers prompted the FDA and WHO to restrict codeine use in children. The genetic mechanism was known; the clinical warnings came decades later. That gap cost lives.

Biomarker-Driven Oncology

Cancer genomics has been the most productive application of precision medicine to date. Next-generation sequencing of tumor tissue can identify the somatic mutations driving a specific patient's cancer, enabling selection of targeted therapies matched to those mutations rather than to the anatomical tissue of origin.

The FDA's 2017 approval of pembrolizumab for any MSI-H/dMMR solid tumor — the first tissue-agnostic oncology approval in history — marked a conceptual turning point: it is the molecular target, not the organ of origin, that determines treatment eligibility.

FDA Companion Diagnostic Requirements

The FDA requires co-development of companion diagnostics for drugs whose use is restricted to patients with specific biomarkers. A companion diagnostic is an in vitro diagnostic test — a tissue assay, blood test, or imaging technique — that provides information essential to the safe and effective use of the corresponding therapeutic. Trastuzumab cannot be prescribed without HER2 testing; crizotinib cannot be prescribed without ALK testing. The companion diagnostic and the drug are co-approved, and use of the drug without the test is considered off-label.

As of 2024, the FDA had approved more than 50 companion diagnostics, the vast majority in oncology. The regulatory co-development burden adds complexity to drug development but ensures that drugs reach the patients most likely to benefit.

Liquid Biopsy and Circulating Tumor DNA

Traditional tissue biopsy is invasive, painful, and provides a single snapshot of tumor biology — a challenge when tumors are heterogeneous and evolving. Liquid biopsy uses a simple blood draw to detect circulating tumor DNA (ctDNA): fragments of DNA shed by tumor cells into the bloodstream. ctDNA analysis can identify actionable mutations, detect minimal residual disease after treatment, and monitor for the emergence of resistance mutations — all without a surgical procedure.

Liquid biopsy is already FDA-approved for companion diagnostic use in certain non-small cell lung cancer settings (Guardant360 CDx, FoundationOne Liquid CDx). Its clinical role is expanding rapidly into screening applications and treatment monitoring across multiple tumor types.

Polygenic Risk Scores: Population to Individual

For complex diseases — cardiovascular disease, type 2 diabetes, breast cancer — genetic risk is not determined by a single gene but by thousands of common variants, each contributing a small probability increment. Polygenic risk scores (PRS) aggregate these variants into a single number that quantifies an individual's genetic predisposition relative to the population.

A high cardiovascular PRS — placing an individual in the top 5% of genetic risk — confers risk equivalent to a familial hypercholesterolemia mutation. Studies have shown that high-PRS individuals benefit more from statin therapy and lifestyle interventions. PRS-based risk stratification is beginning to be incorporated into clinical guidelines and preventive medicine frameworks.

The Equity Problem

The equity problem in precision medicine is serious. Most genome-wide association studies (GWAS) — the research generating the variant data that populates PRS algorithms — have been conducted predominantly in populations of European ancestry. The resulting PRS perform substantially less accurately in African, Asian, and Hispanic populations, both because of population-specific variant frequencies and because of less reference data from these groups. The NIH's All of Us Research Program, which aims to enroll at least one million US participants with emphasis on historically underrepresented groups, represents a direct attempt to address this gap.

Germline vs. Somatic Variants

A critical distinction in precision medicine is between germline variants (inherited, present in every cell of the body) and somatic variants (acquired mutations present only in the tumor or affected tissue). BRCA1 and BRCA2 mutations are germline — they are hereditary and confer lifetime cancer risk to the carrier and first-degree relatives. EGFR mutations in lung cancer are typically somatic — present in the tumor cells but not inherited. The clinical implications differ: germline findings trigger family testing and preventive interventions; somatic findings primarily guide treatment selection for the individual patient's current disease.

The FDA Table of Pharmacogenomic Biomarkers in Drug Labeling — updated continuously — now lists more than 300 drug-gene pairs for which genetic information is clinically actionable. The table spans every major drug class and represents the operational backbone of clinical pharmacogenomics practice.

This article is for educational purposes only and does not constitute medical advice. Genetic testing and precision medicine decisions should be made in consultation with qualified healthcare providers and, where appropriate, genetic counselors.