|
|
|
The View From Here Disease Models The key-word is "predictive". We must increase the precision of the preclinical disease models to enable us to select compounds and targets in discovery that will be effective and safe in patients. We must have access to models that represent the clinical outcome that we are expecting to see in treated patients. No disease model will mimic the human pathology perfectly. What we need is a model that can predict how the compound will affect a certain parameter in humans. It may be a parameter related to treatment or prevention of a certain disease or it may be related to patient safety. Many people see disease models as animal models but in my mind it is much wider than this. Animal models are very useful because of their complexity, interactions between different cell types and organs, maintained organ three dimensional structure, and the wide range of phenotypic responses. There are, however, challenges with animal models that can be managed by using advanced cell culture systems (e.g. ethics, throughput, amount of compound needed, cost). Ideally, instead of using cells or animals, we would like to use computer models to predict efficacy and safety of compounds. The complexity of human biology and pathophysiology, however, limits the current usefulness of "in silico" predictions to certain areas. It is clear to me that we need to use all different types of models, in vivo, in vitro and in silico, in parallel. Only then can we enhance the models and be able to predict accurately which compounds will be effective and safe for patients. Enjoy the read! Jan Törnell Professor Jan Törnell (MD, PhD) is Vice President Translational Science DECS. In 1996 he was recruited to Astra to build and lead a core unit (Astra Transgenic Centre) to generate and analyse transgenic animals for the company worldwide. He was appointed VP Translational Science DECS in 2006 with global responsibility for technologies supporting the interphase between discovery and development. In 1992 he was appointed associate professor and received his PhD in 1990. In 1985 he completed his medical studies and earned his MD. |
In
this issue of Editor's Choice, I would like to draw your attention to
Disease Models. Why would I do that? - You may think that using disease
models for drug discovery is pre-historical and belongs in museums. In
my mind truly predictive disease models are absolutely essential
to increase productivity in drug discovery, reduce clinical attrition
rates and get more medicines to patients faster and at lower cost.