An Open-Source Framework for Virtual Bioequivalence Modeling and Clinical Trial Design.

Area of Science:

• Pharmacokinetics and Pharmacodynamics
• Computational Biology and Bioinformatics
• Drug Development and Regulatory Science

Background:

• Bioequivalence (BE) assessment traditionally requires extensive human testing.
• In silico virtual bioequivalence (VBE) offers a promising alternative to reduce human trials.
• Standardized computational workflows are needed for reliable VBE assessments.

Purpose of the Study:

• Introduce the VBEToolbox R package for standardized computational VBE workflows.
• Integrate in vitro and in vivo data for pharmacokinetic model training and VBE.
• Demonstrate the application of VBE for determining necessary study sizes and statistical power.

Main Methods:

• Developed the VBEToolbox R package within the Open Systems Pharmacology framework.
• Employed a nonparametric approach to handle parameter non-identifiability and uncertainty.
• Integrated in vitro data (skin permeation, dissolution) and in vivo data (absorption) into mechanistic and physiologically based pharmacokinetic models.
• Utilized clinical data for inferring inter-individual variability and establishing in vitro-to-in vivo extrapolations.

Main Results:

• Successfully demonstrated VBE for testosterone dermal formulations using an in vitro/in vivo dermal absorption model.
• Assessed VBE for oral bupropion formulations by integrating in vitro dissolution data into a physiologically based pharmacokinetic model.
• Illustrated the utility of VBEToolbox in streamlining VBE workflows and determining virtual study designs.

Conclusions:

• VBEToolbox provides a standardized framework for computational VBE assessments.
• In silico approaches, like those enabled by VBEToolbox, can significantly reduce the need for human bioequivalence studies.
• The presented case studies highlight key considerations for developing and applying VBE models in drug development.