In drug development, a familiar pattern often emerges. For example, a promising new oncology compound advances from preclinical studies to early-stage clinical trials, buoyed by years of research and encouraging laboratory data. Initial trial results may suggest high efficacy, sparking optimism among developers. But as larger datasets begin to surface, that promise fades, with some patients responding remarkably well, and others showing little to no benefit.

Increasingly, the underlying cause is pointing toward a deeper, biological explanation: human genetic variability. From the way a drug is metabolized to how it interacts with its molecular targets, inherited differences in our genomes are one of the most influential factors shaping therapeutic response (1).

As a result, precision medicine is reshaping how researchers develop and test new drugs and how clinicians deliver them (2). At its core is a growing understanding of the genome, and the realization that one-size-fits-all medicine may soon be a thing of the past. Scientists are now leveraging genomic insights to create more effective, targeted, and safer therapies, while grappling with the challenges of data integration, regulation, and real-world clinical application.

The genetic basis of drug response: why one size doesn’t fit all

People do not respond to medications in the same way, and this challenge complicates drug development and clinical care alike. Much of this variability is rooted in inherited genetic differences, which can affect how drugs are absorbed, metabolized, and eliminated from the body (pharmacokinetics), how they interact with their molecular targets (pharmacodynamics), and how likely they are to trigger adverse reactions.

At the pharmacokinetic level, enzymes responsible for metabolizing drugs vary between individuals due to genetic polymorphisms. These differences can lead to altered drug activation or clearance.

A textbook example is clopidogrel, an antiplatelet prodrug that must be converted into its active form by the cytochrome P450 2C19 (CYP2C19) enzyme in the body to exert therapeutic effects (3). “For a third or more of the population, this drug doesn’t work, “said Chad Bousman, a pharmacogeneticist at the University of Calgary. “It’s like a placebo.” These individuals carry certain variants of the CYP2C19 gene that impair their ability to metabolize clopidogrel, leaving the drug inactive and ineffective.

Understanding genetic variation in drug targets (the domain of pharmacodynamics) is equally critical. Some therapies only work when their targets are present in the right form or quantity. In breast cancer, for instance, only tumors that overexpress the human epidermal growth factor receptor 2 (HER2) respond to trastuzumab, a monoclonal antibody designed to bind to that receptor. Without HER2 amplification, the drug is largely ineffective (4).

Genetic variation can also heighten the risk of adverse drug reactions, sometimes with life-threatening consequences. For example, individuals who carry a specific genetic variant of the human leukocyte antigen B (HLA-B) gene, called HLA-B*57:01, are at high risk of a hypersensitivity reaction to abacavir, an antiretroviral drug used in HIV therapy (5). Hypersensitivity to abacavir is characterized by a wide range of symptoms including fever, rash, gastrointestinal symptoms, and malaise. After the association was discovered, genetic testing for HLA-B*57:01 became standard clinical practice. This intervention led to a considerable reduction in adverse events (5).

Using genome sequencing data to power precision medicine

None of these advancements would be possible without genome sequencing technologies, a set of technologies that determine the precise order of DNA bases in an individual's genome. The foundational effort of the Human Genome Project, which took more than a decade and nearly $3 billion to complete, produced the first comprehensive map of the human genome (6,7). This effort demonstrated both the power and limitations of early sequencing methods and sparked the development of high-throughput platforms collectively known as next generation sequencing (NGS).

Modern NGS can read millions of DNA fragments simultaneously, allowing researchers to decode an entire human genome in just a few days for under $1,000 (8). NGS works by fragmenting DNA, sequencing the fragments in parallel, and computationally assembling the data to identify genetic variants across the genome. These variants can influence disease risk, treatment response, and drug metabolism, providing the basis for more personalized approaches to medicine.

One major area of application is pharmacogenomics, the study of how an individual’s genetic makeup influences their response to drugs. By analyzing a patient’s genome, clinicians can predict whether a drug will be effective, whether it might cause harmful side effects, or whether an alternative treatment or dosage should be considered.

These technological breakthroughs have enabled critical discoveries in clinical oncology. One notable success came from identifying variants in dihydropyrimidine dehydrogenase (DPYD), an enzyme essential for metabolizing fluoropyrimidines, a class of chemotherapy drugs that includes 5-fluorouracil (5-FU) and capecitabine, both commonly used to treat solid tumors (9). Certain DPYD variants reduce the enzyme’s activity, leading to impaired drug clearance and severe toxicities such as neutropenia, mucositis, diarrhea, and cardiotoxicity (9). Understanding these genetic differences has been crucial in guiding personalized chemotherapy dosing.

The impact of genetic testing extends beyond clinical oncology. “We have a variety of therapeutic areas where there are data that would support using the genetic information in the clinical setting to decide whether to use that drug or whether to use a different dose of that drug,” said Julie Johnson, a pharmacologist at Ohio State University. “That ranges from cardiovascular disease, anti-platelet drugs, or anti-coagulant drugs, anti-depressants, pain medications, and proton pump inhibitors.”

These advances in genome sequencing have transformed the ability to tailor treatments based on an individual’s genetic makeup, moving precision medicine from concept to clinical reality. As sequencing technologies continue to evolve and become more accessible, their integration across diverse therapeutic areas promises to improve drug safety and efficacy.

Applying genetic insights to drug development

For drug developers, genetic information is increasingly shaping clinical trial design, informing regulatory decisions. Stratified medicine, which involves selecting clinical trial participants based on specific genetic markers, has drastically improved trial outcomes by focusing on patients most likely to benefit. For example, trials testing BRAF inhibitors like vemurafenib in melanoma only enroll patients with mutations in the BRAF gene (10).

Building on stratified trials, adaptive trial designs offer greater flexibility by using interim data — including genetic information — to modify trial parameters in real-time. This can include revising inclusion criteria or adjusting treatment arms based on emerging genetic insights (11). As Johnson noted, collecting genetic data during a trial enables researchers “to really arrive at an understanding of who in the population likely is going to benefit.”

Regulatory agencies are also evolving to keep pace with these innovations. The FDA now incorporates genetic information in dozens of drug labels and maintains an expanding list of pharmacogenomic biomarkers to guide clinical use (12). However, challenges persist, especially around harmonizing regulatory standards internationally (13). These inconsistencies can delay widespread adoption of pharmacogenomic testing and contribute to disparities in access to precision medicine worldwide.

The future of drug discovery is personal, but challenges remain

Designing drugs around a person’s genetic code is becoming a medical standard. Yet the road ahead has hurdles. One is integrating multiomics data. While genomics provides foundational insights, layers like transcriptomics, proteomics, and metabolomics offer additional context. Combining these data streams into cohesive models of disease and drug response is no small feat, and it demands advances in computational modeling and artificial intelligence (AI).

Still, some researchers remain cautious about AI’s current reliability. Julie Johnson noted that while there is optimism about AI’s potential to streamline multiomic data analysis, “we’re still early to answer the question about how reliable [AI is]” and that “we’re not quite there.” Bousman echoed this concern, highlighting a critical gap between model development and clinical usability: “If a clinician is going to use AI in practice, they should have some understanding of how it works,” he said, referring to the need for transparency in how AI programs generate specific answers.

In addition, implementation challenges persist. Many clinicians lack formal training in genomics, and access to pharmacogenomic tools is inconsistent. “Typically, the first major hurdle is ensuring that clinicians are aware that this opportunity exists,” Bousman said. Even when clinicians are aware, he said, they may dismiss its relevance due to a lack of guideline support, despite the fact that “clinical guidelines are at minimum a decade behind the evidence.”

Beyond technological and logistical barriers, the science itself remains a work in progress. While researchers have made significant advances in mapping genes involved in drug metabolism, the mechanisms driving efficacy and side effects — particularly those unrelated to drug levels — are less well understood. “A lot of times, we don’t fully understand all the genetic and non-genetic factors that contribute to how a person responds to a drug,” Johnson said. Unraveling these biological complexities is essential to realizing the full promise of pharmacogenomics.

References

1. Roden, D. M., Wilke, R. A., Kroemer, H. K. & Stein, C. M. Pharmacogenomics: The genetics of variable drug responses. Circulation 123, 1661–1670 (2011)..
2. Yamamoto, Y., Kanayama, N., Nakayama, Y. & Matsushima, N. Current Status, Issues and Future Prospects of Personalized Medicine for Each Disease. J. Pers. Med. 12, 444 (2022).
3. Clopidogrel (oral route). Mayo Clin.
4. Gajria, D. & Chandarlapaty, S. HER2-amplified breast cancer: mechanisms of trastuzumab resistance and novel targeted therapies. Expert Rev. Anticancer Ther. 11, 263–275 (2011).
5. Dean, L. in Med. Genet. Summ. (eds. Pratt, V. M. et al.) Abacavir therapy and HLA-B*57:01 genotype. National Center for Biotechnology Information (US), 2012.
6. NIH. The Cost of Sequencing a Human Genome.
7. NIH. Human Genome Project Fact Sheet.
8. Tafazoli, A., Guchelaar, H.-J., Miltyk, W., Kretowski, A. J. & Swen, J. J. Applying Next-Generation Sequencing Platforms for Pharmacogenomic Testing in Clinical Practice. Front. Pharmacol. 12, 693453 (2021).
9. Dean, L. & Kane, M. in Med. Genet. Summ. (eds. Pratt, V. M. et al.) Fluorouracil therapy and DPYD genotype. National Center for Biotechnology Information (US), 2012.
10. Castellani, G. et al. BRAF Mutations in Melanoma: Biological Aspects, Therapeutic Implications, and Circulating Biomarkers. Cancers 15, 4026 (2023).
11. Kaizer, A. M. et al. Recent innovations in adaptive trial designs: A review of design opportunities in translational research. J. Clin. Transl. Sci. 7, e125
12. Kim, J. A., Ceccarelli, R. & Lu, C. Y. Pharmacogenomic Biomarkers in US FDA-Approved Drug Labels (2000–2020). J. Pers. Med. 11, 179 (2021).
13. FDA. International Regulatory Harmonization.