Advocates of animal research say that animal testing is necessary to ensure that drugs are safe and effective for humans. Yet research has shown that less than 10-15% of human clinical trials are successful, even though these trials are based on seemingly promising preclinical animal data.
But a new study has called this data into question, noting concerns with the design and reporting of animal experiments.
The study, reported in PLOS earlier this month, focused on “investigator brochures,” documents that summarize preclinical data obtained for new drugs. The documents, which are produced to help Institutional Review Boards and regulatory agencies to decide whether moving forward with clinical trials is worth the risk, are often missing important data.
Researchers in this study examined 109 investigator brochures for clinical trials that had been approved from 2010 to 2016. The brochures included 708 efficacy studies conducted in animals.
Upon reviewing the efficacy studies performed in animals, the researchers found that:
- Fewer than 5% of the efficacy studies included in investigator brochures included important information about bias-reducing measures, including sample size calculations, blinded outcome assessment and randomization.
- Only 11% of preclinical efficacy studies cited in the investigator brochures were published in peer-reviewed reports, meaning that 89% of animal studies referenced in these brochures were not published at all. This means that individuals reviewing the brochures were unable to determine whether the preclinical data were critically and independently reviewed.
- 82% of investigator brochures included only preclinical studies with positive effects, which suggests a bias, as data from less flattering studies were likely omitted.
Although this study was conducted in Germany, the researchers feel that the results are applicable throughout Europe and the United States.
Studies like this call into question why the research community should have the luxury of continuing to work with animal models at all. In addition to issues with experimental design and transparency, there are more inherent problems, such as extrapolating data across species. If the scientific community really wants to improve human health and well-being through research, then efforts need to be refocused on human-relevant models.
Yasinski, E. “Clinical trials may be based on flimsy animal data.” Science. April 5, 2018