Population Pharmacokinetic Method to Predict Within-Subject Variability Using Single-Period Clinical Data.

Area of Science:

• Pharmacokinetics
• Clinical Trial Design
• Statistical Modeling

Background:

• Sample size determination in clinical trials relies on within-subject variability (WSV).
• Accurate WSV estimation is challenging without replicate study designs.
• Current methods often use total variability, potentially leading to suboptimal sample sizes.

Purpose of the Study:

• To develop and validate an efficient population-based method for predicting WSV from single-period clinical trial data.
• To assess the method's performance in accurately estimating WSV across various scenarios.
• To provide a reliable approach for sample size estimation in pharmacokinetic studies.

Main Methods:

• Simulated 1000 virtual pharmacokinetic trial datasets with varying sample sizes, WSV, and interindividual variabilities (IIVs).
• Employed population pharmacokinetic modeling to estimate residual variability (RV).
• Compared estimated RV with true WSV and evaluated bioequivalence using a real eperisone dataset.

Main Results:

• Residual variability (RV) approximated within-subject variability (WSV) well when WSV was 40% or less, for sample sizes >18 subjects.
• RV was underestimated at WSV of 50% or greater, even with low interindividual variability (IIV).
• Analysis of the eperisone dataset showed RV (44-48%) closely matched the true WSV (50%).

Conclusions:

• The developed population-based method accurately predicts WSV in single-period clinical trials.
• This method offers a validated approach for optimizing sample size estimation.
• Enhances the statistical rigor of pharmacokinetic study design.