A random sampling approach for robust estimation of tissue-to-plasma ratio from extremely sparse data

Area of Science:

• Pharmacokinetics and Pharmacometrics
• Biostatistics
• Computational Biology

Background:

• Accurate estimation of tissue-to-plasma ratio is crucial for drug development and understanding drug distribution.
• Traditional methods struggle with extremely sparsely sampled paired data, lacking robustness and uncertainty quantification.
• Existing approaches like independent time points and naive data averaging fail to provide reliable variability measures.

Purpose of the Study:

• To develop and validate a new nonparametric approach for robust estimation of tissue-to-plasma ratio.
• To address challenges posed by extremely sparsely sampled paired data in pharmacokinetic studies.
• To compare the performance of the proposed method against existing techniques, including robustness against outliers.

Main Methods:

• A novel 2-phase random sampling approach was developed for estimating tissue-to-plasma ratio.
• Comparison with independent time points approach, naive data averaging, and pseudoprofile-based bootstrap (PpbB).
• Robustness was evaluated by introducing outliers (10-40% inflation) into generated datasets.

Main Results:

• The proposed random sampling approach yielded robust tissue-to-plasma ratio estimates with uncertainty and variability.
• Independent time points approach showed wide ratio variations (0-50), while naive averaging lacked variability measures.
• The 2-phase random sampling method demonstrated superior robustness and required fewer replications compared to PpbB.

Conclusions:

• A 2-phase random sampling approach is proposed as a superior method for robust tissue-to-plasma ratio estimation.
• This method effectively handles extremely sparsely sampled paired data, providing reliable uncertainty quantification.
• The proposed approach enhances the accuracy and reliability of pharmacokinetic parameter estimation in drug development.