Within the context of model-based drug development, pharmacokinetic/pharmacodynamic (PK/PD) modeling plays a central role. However, despite the clear value of PK/PD modeling and its increased use in the pharmaceutical industry, the field continues to be challenged by the lack of flexible and intuitive software tools. Existing packages either require cumbersome programming or offer limited functionality that cannot be extended.

A case in point is the task of model building. Conventional tools typically include a library of PK/PD models, but creating a custom model or even tweaking an existing one usually requires coding up the equations—a laborious task, especially for a scientist without a programming background. Furthermore, the shift towards mechanistic and physiologically-based PK models has made building, maintaining, and communicating an equation-based representation of PK/PD models quite challenging.

click to enlarge SimBiology streamlines model-based drug development by providing tools for the complete PK workflow. (Image: The MathWorks)

But building a suitable PK/PD model is only half the story; the real value of a model is the predictive capability that it brings to quantitative drug development. PK/PD models characterize the temporal relationship between drug dose and concentration (PK), and the drug’s effects (PD). A well-characterized PK/PD model is a powerful tool in decision-making—it can help explain observations and guide design of future trials. Pharmacometricians can use models to ask “what if” questions via a simulation exercise.Integration of model-building, simulation, and analysis capabilities in a single environment provides researchers with the opportunity for a rapid, exploratory workflow that is not possible when using separate tools for modeling and analysis.

Technical software vendors are beginning to address these needs. For example, The MathWorks SimBiology implements the PK/PD workflow, from modeling to analysis, within a single graphical environment. Building a PK/PD model or carrying out analysis tasks such as parameter estimation or sensitivity analysis are point-and-click activities within SimBiology—this makes it a very accessible tool, even for researchers with no prior programming experience.

Because SimBiology is based on the MATLAB environment for technical computing, it has direct access to an industry-tested simulation solver suite, and is seamlessly integrated with other MATLAB-based functionality such as parallel computing, statistics, and optimization. This can help with both the pre- and post-processing parts of a PK/PD workflow. For example, following model selection and parameter estimation, researchers can simulate the model in SimBiology to test various dosing strategies in an in silico population, and make use of powerful MATLAB-based optimization algorithms to design an optimal strategy. The resulting time and cost savings from working in a single environment can be tremendous.

Bringing new drugs to market is very expensive and late-stage clinical failure due to safety and efficacy concerns is a key factor in escalating development costs. Industry experts and regulatory authorities believe that a systematic adoption of the model-based drug development paradigm would allow researchers to predict these failures earlier, and avoid the expense of failed late-stage trials.¹ The use of intuitive tools that integrate the entire workflow from modeling to analysis in a single environment can make a significant contribution to model-based drug development.

At A Glance
Company: The MathWorks
Product: SimBiology v3.0
Date Introduced: March 2009
Product Requirements: MATLAB v7.8 and Statistics Toolbox v7.1 (R2009a)
Recommended Products: Optimization Toolbox v4.2 (R2009a)
Supported Platforms: Windows, Windows x64, Mac OS X (Intel), Linux, Linux x86-64

About the Authors
Asawari Samant is a application engineer in the Computational Biology group at The MathWorks. Her background is in chemical engineering, with a focus on modelling of biological systems. Sam Roberts is a principal application engineer at The MathWorks, focusing on the Life Science sector. His background is in biological and pharmaceutical applications of statistics and machine learning.

References:
1. Department of Health and Human Services, Food and Drug Administration. Innovation or Stagnation? Challenge and Opportunity on the Critical Path to New Medical Products.(2004) http://www.fda.gov/downloads/ScienceResearch/SpecialTopics/CriticalPathInitiative/CriticalPathOpportunitiesReports/ucm113411.pdf Accessed July 2009.