Genomic Data as a Catalyst in Drug Development: Enhancing Clinical Trial Success Through Genetic Insights

For pharmaceutical companies, drug development has always been a contentious issue. If clinical trial data are satisfactory and the indications are clear, approval for market launch and substantial profits may be just one step away. However, if a single misstep occurs, the subsequent consequences are all too predictable. Therefore, drug development is a protracted, high-risk, and costly process.

Statistics show that more than half of the seemingly promising drug candidates in laboratories ultimately fail, with over a quarter proven ineffective in late-stage clinical trials. However, genomic information is becoming increasingly accessible through public databases—particularly specific genes identified as being associated with diseases. This existing knowledge is likely to significantly improve the success rate of drug development. According to the pharmaceutical industry and scientific teams, guiding experimental molecular target research with genomics will certainly enhance success rates.

We have noted several studies in this area, including June’s Genome Advance of the Month, which focuses on pharmacogenomics and leverages known gene associations to help select research targets, thereby improving the success rate of drug development. Also noteworthy is a recent study published in Nature Genetics by Dr. Matthew Nelson of GlaxoSmithKline at Research Triangle Park (RTP) and other scientists.

As is well known, in the United States, all new drugs must obtain FDA approval before being marketed, demonstrating their safety and efficacy. Clinical trials are designed to establish the foundational basis for such FDA approval. These trials are divided into three phases: Phase I, involving 20 volunteers to assess drug safety; Phase II, involving 100 or more participants to evaluate efficacy and side effects; and Phase III, typically involving thousands of patients to test efficacy and adverse reactions.

Among new drugs that pass Phase I trials, generally two-thirds proceed to Phase II trials, and one-third of those advance to Phase III. Ultimately, only one in ten new drugs receives FDA approval.

Given the aforementioned attrition rates, Dr. Nelson and his colleagues hypothesized that leveraging genomic information could help new drugs secure FDA approval. They sought to determine the importance of gene-disease associations in guiding the development of gene-based therapeutics. Their findings concluded that gene-disease correlations indeed steer experimental efforts in the right direction.

The research report estimates that “selecting targets supported by genetic evidence can double the success rate of clinical development, although success rates vary significantly across different diseases.”

Researchers utilized several databases to identify known gene–disease associations. These databases included the GWAS Catalog, which contains thousands of gene–disease associations identified through genome-wide association studies; the Database of Genotypes and Phenotypes (dbGaP); and Online Mendelian Inheritance in Man (OMIM), a compendium of human genes and genetic phenotypes. Subsequently, researchers manually matched disease characteristics with disease terms from the Medical Subject Headings (MeSH), an online medical vocabulary search tool, and collected information on pharmaceuticals at various stages of development.

The results indicate that genes associated with human traits offer more promising research targets for successful therapeutic drug development than genes unrelated to disease. Researchers hypothesize that mutations in genes influencing human physiological characteristics may heighten sensitivity to drug efficacy.

Dr. Nelson and his colleagues also pointed out that if these drugs could target genes known to be associated with diseases, particularly those related to the musculoskeletal system, metabolism, and blood disorders, they would perform better in the drug development pipeline. This finding underscores the necessity of a widely cited reference database capable of aggregating disease-associated genes using unbiased methods. Currently, such databases have received support from the National Institutes of Health (NIH).

An institute under the NIH—the National Human Genome Research Institute (NHGRI)—is currently undertaking several research projects aimed at incorporating data into a large public database to advance the application of genomic information in drug research. These initiatives include PhenX, which coordinates phenotype datasets and integrates them into a comprehensive large-scale database; the IGNITE Network, which provides clinical support through broader contextualization, proposes clinical interventions and recommendations, and identifies methods and outcomes for multiple pharmacogenomic actions; and ClinGen, a public database of genomic variants designed to improve human health and advance biomedical research.

Furthermore, among the pharmaceutical companies currently at the forefront of this field are Pfizer and Genentech, to our knowledge. They have chosen to partner with 23andMe, which shares genetic data from 650,000 customers and whole-genome sequences from 3,000 Parkinson’s disease patients for drug development.

In July this year, Calico, a subsidiary of Google, partnered with Ancestry, a DNA genealogy analysis company. Leveraging Ancestry’s extensive genealogical and genetic database, the collaboration employs analytical tools and algorithms to investigate the family histories of long-lived individuals and the life-related information encoded in their genes. The research findings are being modeled and commercialized to analyze and predict individual lifespan, as well as to develop drugs that delay or halt aging.

It is evident that as the associations between an increasing number of genes and diseases are elucidated, genomic information will play an increasingly significant role in drug development.

The researchers mentioned in the article include those from GlaxoSmithKline; the Department of Systems Biology and the Department of Biomedical Informatics at Columbia University in New York; and the State Key Laboratory of Brain and Cognitive Sciences, the Centre for Genomic Sciences, and the Department of Biochemistry at The University of Hong Kong.

Compiled by Zhou Yanxun
Editor: Mo Renying