Media Release
I have co-authored a study that shows that up to nine in 10 potential medicines that progressed from preclinical testing into human trials failed during clinical development, despite earlier testing intended to assess their safety and effectiveness.
Before a new medicine is tested on people, it goes through laboratory and preclinical testing, which commonly includes testing in animals. Only drugs that perform well enough at this stage are allowed to progress to human clinical trials. However, a review has found that of the drugs that made it through this preclinical stage, up to nine out of 10 (or 91%), failed during testing in humans.
A study of 60 years of drug failures in human clinical trials worldwide has shown that 77% of failed drug candidates were discontinued for biological reasons, such as not working effectively in humans or having safety/adverse-effect problems.
Preclinical testing is intended to predict these sorts of outcomes.
Medicines intended for humans which go through animal testing often do not reliably predict how a drug will behave in humans, and this is often because of biological differences between species. Colleagues and I have long argued that there really needs to be greater use of human-relevant methods, such as approaches based on human cells, tissues and disease processes. These are available and would make results more reliable and reduce the unnecessary suffering of laboratory animals worldwide.
The United States is moving towards phasing out animal testing requirements for monoclonal antibodies (man-made proteins) and other drugs in an effort to better predict human responses to medicines. There are about 115 million animals used in lab experiments worldwide, which also includes testing for new household and personal products. Experiments often involved species such as mice, rats, dogs and monkeys.
The full study is available here.
Filipova D. Boomgaarden W. Fox L. Knight A. et al. (2026). Need for NAMs: A systematic evidence synthesis revealing over half a century of drug development failure. NAM Journal, 100119.