QSP modeling to guide drug discovery and development
Quantitative systems pharmacology (QSP) is an approach that uses computational models to describe dynamic interactions between a drug and pathophysiology to understand the system at the cellular and biochemical network levels. QSP modeling aims to improve understanding of the biological system and disease, facilitate early and more thorough in silico testing of drug candidates, and support rational decision making to reduce development cost and time. QSP modeling can reduce late-stage attrition rates, one of the biggest current challenges in drug development.
The area of quantitative systems pharmacology has emerged at the crossroads between traditional PK/PD and mechanistic systems biology modeling. It combines a mechanistic model of disease pathophysiology, PK/PD of a therapeutic agent, and quantitative experimental data. The resulting model can be used to run simulations to understand how drugs modify cellular networks and how they are impacted by the (patho)physiology, the significant pathways, and the impact of drug parameters and biological variance on drug efficacy and safety.
Quantitative systems pharmacology is being increasingly used in drug discovery and development from exploratory research to late‐stage clinical trials to guide research and decision making on areas such as:
- Target feasibility and selection: Designing a therapeutic agent often starts with choosing from a list of potential candidates. Developing QSP models for each potential target leads to establishing affinity and dose requirements and predicting optimal drug parameters early on. This approach helps eliminate targets with low developability so you can pursue more promising candidates.
- Drug efficacy and safety: Most drugs fail in clinic because of low efficacy, and high efficacy levels in animal experiments may not translate to humans. QSP models have the potential to predict this behavior. In addition to predicting which drugs will be more efficacious, QSP can help identify which drugs might fail and for which reasons. Because QSP models can predict drug exposure at the organ level, they also provide insights into the mechanism of toxicity and potential side effects.
- Dose optimization: Complex diseases such as cancer, diseases of the central nervous system, and metabolic diseases typically involve combination therapy. Incorporating disease mechanisms via QSP models leads to important and often counterintuitive insights for deciding optimum dose levels and combination therapy approaches.
- Precision medicine: Many diseases exhibit heterogeneity where subpopulations are affected differently. Use of QSP models can incorporate the impact of biological variance on efficacy and safety and lead to rational decisions on which patient subpopulation to target during clinical trials.
The quantitative systems pharmacology workflow typically involves the following steps:
- Build models using a drag-and-drop block diagram or programmatic interface
- Import and visualize experimental time course data
- Fit data to estimate parameters using local and global optimization methods
- Simulate “what if” scenarios
- Identify key pathways and parameters by running sensitivity analysis
- Simulate virtual patients to explore variability and alternate hypothesis
- Perform large-scale computations, such as Monte Carlo simulations, in parallel using desktop, cluster, or cloud to improve performance
For more information on developing and using quantitative systems pharmacology models, see SimBiology®.
Examples and How To
See also: SimBiology, PK/PD model, computational biology, biotech and pharmaceutical, Statistics and Machine Learning Toolbox, Optimization Toolbox, Parallel Computing Toolbox, genetic algorithm