Cost-constrained optimal sampling for system identification in pharmacokinetics applications with population priors...

Area of Science:

• Pharmacokinetics
• Systems Biology
• Computational Pharmacology

Background:

• Pharmacokinetic (PK) models represent complex nonlinear systems requiring accurate parameter estimation for personalized medicine.
• Population PK data exists, but individual patient parameter determination is crucial for tailored therapy.
• Current methods often involve frequent blood sampling, which can be costly and burdensome.

Purpose of the Study:

• To develop an irregular sampling strategy for pharmacokinetic (PK) system identification.
• To minimize uncertainty in PK parameters for individual patients under cost constraints.
• To enable personalized drug therapy through optimized drug concentration monitoring.

Main Methods:

• Utilized Monte Carlo simulations to estimate the average Fisher's information matrix for PK models.
• Employed a minimax criterion to identify sampling points that minimize maximum parameter uncertainty.
• Applied a genetic algorithm for optimizing the irregular sampling schedule.

Main Results:

• Demonstrated the design of an adaptive sampling scheme tailored to individual patients.
• Showcased the strategy's ability to accommodate various dosing regimens.
• Validated the approach for flexible therapeutic strategy development.

Conclusions:

• The proposed irregular sampling strategy effectively reduces uncertainty in patient-specific PK parameters.
• This method supports personalized medicine by optimizing drug monitoring.
• The approach offers a flexible and cost-effective solution for therapeutic drug management.