Reducing animal testing

Animal models are still indispensable in some research areas. They provide the only way of modelling the effects of a new medicine on an entire organism, with all its organs and the complex interactions that take place in the body. Long-term effects can often only be demonstrated in animal models, and that is also why authorities responsible for licensing drugs still insist on them as part of the approval process.

On the other hand, we know that animal models are not always transferable to humans. To reduce animal testing and improve translatability of results to humans, we are intensively researching various alternative methods.

One of these methods is to model human physiology outside the living organism, for example on a chip, in a lab, or to use computer models to test the behaviour of a medicine in the body. Using such methods, Roche has reduced the number of animals in experimental use by almost 40% over the past ten years.

Roche is dedicated to following the guiding principles of the 3Rs for animal research, and encourages the development of alternatives to the use of animals in research.

Roche is pioneering and promoting innovative approaches in preclinical research. These methods help us to identify promising drug candidates faster and with more precision and, at the same time, reduce the number of animals needed in research. They also lead to a better understanding of diseases and provide valuable insights for a more personalised approach to medicine.

Roche has developed three-dimensional cell systems to observe new molecules and their effects and side effects on human cells outside the body, at an early stage.

Multiple organs can be placed on one chip about the size of a USB stick and are interconnected to model the dynamics of a human organism.

This approach should allow researchers to replace a great deal of animal testing in the early preclinical development stage or future major advances.

Organ on a chip
The entire human organism on one single chip

Laboratory experimentation and animal testing give us information on the effect of a new medicine and its distribution in the body. Computer models enable us to transfer those results to humans using complex simulations, allowing us to predict dosages and effects before direct human testing begins. In the context of these models we also use real world data. It is difficult to predict effects in humans since every patient has different features of a disease. If we can build in specific characteristics of different patient populations into a test model, we will be better able to anticipate the effects of a medicine in the human body. Currently, animal data is still required as a reference for computer models. Once real world data sources are more established, we will be able to reduce animal experiments even further.

Real world data makes models smarter
Patient-derived organoids


No toxicity is observed in organoinds (above)


Organoid toxicity and cell death (above)

With the help of patient-derived organoids we are able to model important aspects of a human organ outside a living organism. In this case, we mimicked critical parts of the gut and predicted adverse events of potential medicines in the human body. This promising approach minimises the number of animal experiments and studies while making preclinical drug development programmes more precise and predictive for clinical trials. The pictures show an untreated organoid (top left) where no toxicity is observed, as evidenced by a lack of signal from a fluorescent probe for cell death (top right). In contrast, the two pictures at the bottom depict organoid toxicity and cell death (shown by the green signal on the right) based on drug intervention.

Organoids have the potential to revolutionise the way we discover and develop medicines. With the Roche Institute for Translational Bioengineering (ITB), Roche is breaking new ground and driving development in the field of organoids.

Stories of

Innovation

See all stories

Discover more