An “Omics” view of drug development

Russ B. Altman, Daniel L. Rubin, Teri E. KleinDrug 2004 · DOI

TL;DR

This paper argues that the pharmaceutical industry's adoption of pharmacogenomics, the study of how genes influence drug response, is crucial for improving drug development efficiency and patient outcomes. It highlights two major challenges: fully characterizing genes involved in drug metabolism (pharmacokinetics) and understanding how genetic variations affect drug targets like G-protein coupled receptors (pharmacodynamics).

What Problem Does This Paper Address?

The pharmaceutical industry faces a complex landscape where pharmacogenetics and pharmacogenomics offer immense potential to identify patient subgroups who will benefit most from specific drugs, thereby saving promising drug candidates and improving efficacy. However, this also introduces challenges: increased complexity in drug applications, market fragmentation, and the need for prescribers to master "personalized medicine." The core problem is the gap between the theoretical promise of genetics-based drug selection and the practical reality of having enough validated drugs and understanding to implement it effectively. The authors aim to outline the current state and future directions for integrating "omics" approaches into drug development.

Methods

The paper doesn't describe specific experimental methods or algorithms in detail. Instead, it focuses on a conceptual framework and a call to action. The authors advocate for a systems-level approach, akin to other "-omics" fields (genomics, proteomics, metabolomics), to comprehensively study gene-drug interactions. They emphasize the need for:

  • Comprehensive gene cataloging: Identifying all genes involved in drug metabolism (pharmacokinetics) and drug action (pharmacodynamics).
  • High-throughput experimental and computational assays: To gauge drug affinity and metabolism by allelic variants of enzymes and transporters.
  • Data integration and mining: Leveraging large datasets from initiatives like the HapMap project to identify common variations and their functional consequences.
  • Systematic study of gene families: Focusing on key drug targets like G-protein coupled receptors (GPCRs) and understanding how genetic polymorphisms affect their function and expression.

Datasets

The paper mentions the PharmGKB database as a curated list of published gene-drug relationships, which at the time of writing contained associations for 1,349 genes with approximately 300 drugs. It also refers to the PubMED database for literature scans and the HapMap project as an example of a large-scale effort to catalog genetic variation. No new datasets were introduced.

Key Insights and Ideas

  • Pharmacogenomics is not just about saving drugs; it's about efficiency and safety: By understanding individual responses, drug development can be streamlined, and patient safety improved.
  • The "systems biology" approach is essential: Studying gene-drug interactions in isolation is insufficient; a comprehensive, "omics"-level view is needed to capture complex relationships.
  • Two major frontiers exist:
    • Pharmacokinetics (PK): Fully characterizing all genes involved in drug metabolism and transport is a significant, yet achievable, challenge with the completion of the human genome.
  • Pharmacodynamics (PD): Understanding how genetic variations affect drug targets, particularly GPCRs, is more complex due to the diversity of targets and pathways.
  • Industry and academia have complementary roles: Industry focuses on high-impact drugs for common diseases, while academia can explore rarer variations and fundamental principles, often in less competitive areas.
  • Cost-effectiveness is key for clinical adoption: For pharmacogenomic interventions to be widely adopted, the cost of genetic testing must be justified by significant improvements in patient outcomes, especially in high-stakes scenarios.
  • A new research ethic is emerging: The success of other "-omics" fields points towards a future of inexpensive, high-throughput data collection, followed by database mining to generate testable hypotheses.

Key Results

The paper is primarily a perspective and a call for research priorities, rather than a report of specific experimental results. However, it highlights:

  • The existence of approximately 1,349 genes with documented gene-drug relationships in the PharmGKB database.
  • The dominance of cytochrome P450 genes (CYP2D6, CYP3A4, etc.) in current pharmacogenomic research, emphasizing their importance in drug metabolism.
  • The fact that G-protein coupled receptors (GPCRs) represent a significant class of drug targets, making them a critical area for pharmacodynamic research.

Limitations and Future Work

Limitations:

  • The authors acknowledge that the "logical case" for genetics-specific drug selection is more mature than the "practical list of drugs for which outcomes are demonstrably improved."
  • The complexity of drug action (pharmacodynamics) is harder to assess than drug metabolism (pharmacokinetics) due to the vast number of potential drug targets.
  • The current understanding of genetic variation's impact on drug response is still incomplete, especially for less common variations or complex gene interactions.

Future Work:

  • Complete characterization of all genes involved in drug metabolism (PK): This includes understanding their gene expression, protein translation, modifications, and localization.
  • Detailed study of GPCRs and their associated pathways: Characterizing the structure and function of GPCR genes and identifying important variations that affect drug response.
  • Systematic evaluation of gene homologs: Identifying and assessing the importance of gene products that are similar to known drug-metabolizing genes.
  • Development of high-throughput genomic and proteomic technologies: To efficiently assay gene expression, protein levels, and activities to understand functional consequences.
  • Prioritization of research efforts: Given the vast amount of work, identifying which targets and variations are most critical for drug development will be essential.
  • Establishing cost-effective strategies for genetic testing: To enable widespread clinical adoption of pharmacogenomic insights.

Summary updated

Abstract

Abstract The pharmaceutical industry cannot be blamed for having a love/hate relationship with the fields of pharmacogenetics and pharmacogenomics. At the same time that pharmacogenetics and pharmacogenomics promise to save pipeline drugs by identifying subsets of the population for which they work best, they also threaten to increase the complexity of new drug applications, fragment markets, and create uncertainty for prescribers who simply do not understand or have time to master “personalized medicine.” Most importantly, the logical case for genetics‐specific drug selection and dosing is much more mature than the practical list of drugs for which outcomes are demonstrably improved. Understandably, pharmaceutical developers and regulators have been careful in creating strategies for using genetics in drug development, and only recently has the FDA begun to establish preliminary rules for pharmacogenetic testing. A growing public academic effort in pharmacogenetics and pharmacogenomics is helping flesh out the basic science underpinnings of the field, and this should combine with extensive efforts of industry to create a solid foundation for future use of genetics in drug development. Two grand challenges to accelerate our capabilities include the characterization of all human genes involved in the basic pharmacokinetics of drugs, and the detailed study of the genes and pathways associated with G‐protein‐coupled receptors and how they are affected by genetic variation. Drug Dev. Res. 62:81–85, 2004. © 2004 Wiley‐Liss, Inc.
BibTeX
@inproceedings{altman2004omics,
  author = {Russ B. Altman and Daniel L. Rubin and Teri E. Klein},
  title = {{An “Omics” view of drug development}},
  booktitle = {Drug},
  year = {2004},
  doi = {10.1002/ddr.10370},
}
Figure