Efficient simulation of clinical target response surfaces.

Area of Science:

• Pharmacometrics
• Computational Biology
• Drug Development

Background:

• Simulating combination therapies presents computational challenges, often neglecting patient variability or limiting dose selection.
• Existing models struggle to bridge the gap between simulated concentrations and clinical doses, especially when accounting for interindividual variability (IIV) and parameter uncertainty.

Purpose of the Study:

• To develop efficient algorithms for identifying combination therapy doses that meet predefined efficacy targets.
• To incorporate interindividual variability (IIV) and parameter uncertainty into dose-finding simulations for combination therapies.

Main Methods:

• Devised novel algorithms for efficient simulation and dose searching in combination therapy.
• Integrated population pharmacokinetic/pharmacodynamic (PK/PD) modeling principles with uncertainty quantification.
• Developed a response surface of confidence levels for all dose combinations.

Main Results:

• The new algorithms efficiently identify optimal combination doses while accounting for IIV and parameter uncertainty.
• Generated a confidence level response surface, visualizing the probability of achieving efficacy targets across dose combinations.
• Demonstrated the importance of population-level simulation over individual-focused approaches.

Conclusions:

• Efficient simulation of combination therapies is achievable by accounting for patient variability and uncertainty.
• The developed method provides a valuable tool for optimizing experimental design and clinical trial strategies in drug development.
• Population-based simulation is essential for robustly predicting combination therapy outcomes.