<img alt="" src="https://secure.lote1otto.com/219869.png" style="display:none;">
Skip to content

Asking the Right Questions of Your Data: Experiences in Model Informed Drug Development

Background image with system motherboard concept and question mark.jpeg

 At the Chief Medical Officer Summit earlier this year, Cytel's Director of Quantitative Pharmacology and Pharmacometrics Cecilia Fosser, and Senior Director, Business Development, Chuck Gelb presented on how model-informed drug development (MIDD) techniques can improve decision-making and Probability of Success ( PoS) of clinical trial programs.  In this blog, we share some highlights and the informative10-minute video replay of their talk which includes a modeling and simulation case study. 

At the outset of the 10 minute presentation, Gelb used the 'Rumsfeld Theorem' to characterize the issue of uncertainty in drug development. Rumsfeld said: 

 "There are known knowns. These are things we know that we know. There are known unknowns. That is to say, there are things that we know we don't know. But there are also unknown unknowns. There are things we don't know we don't know."

Use of Model-Informed Drug Development techniques can help  answer some of these "unknowns" by examining data from across clinical studies and allowing us to make more assumptions in a hypothesis generating fashion. By combining pharmacometrics with established statistical techniques,  data collected across clinical studies can be more effectively leveraged, providing quantitative guidance and helping to improve decision-making. Applying these approaches can help emerging drug developers gain confidence in the Probability of Success of the development plan.

As Fosser highlights, model-informed drug development may be helpful:

  • Whenever the team needs to make decisions based on a cumulative understanding of the  data, such as dose selections for drug development milestones or assessment of safety risks.
  • Whenever the team would like to extrapolate between doses that were observed into a dose that has not yet been tested (or a time that was not yet tested).
  • Whenever the team would like to change indications or study populations using healthy volunteer data as a link between the two.
  • Whenever it would be unsafe or unethical to put a proposed dose directly into humans. (for poor/extensive drug metabolizers)
  • For modeling a pediatric population.

Watch the short video to learn more:



Bonus! Click the button below to download the slides from the presentation.


 Further reading:

4 questions to explore in model-informed drug development

Case Study: Exposure response modeling in hematology indication

It's time to bridge the gap between pharmacometrics and biostatistics

contact iconSubscribe back to top