Population Pharmacokinetics and Limited Sampling Strategy for Therapeutic Drug Monitoring of Polymyxin B in Chinese Patients With Multidrug-Resistant Gram-Negative Bacterial Infections

Polymyxin B is used as a last therapeutic option for the treatment of multidrug-resistant Gram-negative bacterial infections. This study aimed to develop a population pharmacokinetic model and limited sampling strategy, a method to estimate the area under the concentration curve (AUC) by using a limited number of samples, to assist therapeutic drug monitoring of polymyxin B in Chinese patients. Population pharmacokinetic analysis was performed using Phoenix(®) NLME with data obtained from 46 adult patients at steady state. Various demographic variables were investigated as potential covariates for population pharmacokinetic modeling. The limited sampling strategies based on the Bayesian approach and multiple linear regression were validated using the intraclass correlation coefficient and Bland-Altman analysis. As a result, the data was described by a two-compartment population pharmacokinetic model. Through the modeling, creatinine clearance was found to be a statistically significant covariate influencing polymyxin B clearance. The limited sampling strategies showed the two-point model (C(0h) and C(2h)) could predict polymyxin B exposure with good linear relativity (r(2) > 0.98), and the four-point model (C(1h), C1(.5h), C(4h), and C(8h)) performed best in predicting polymyxin B AUC (r(2) > 0.99). In conclusion, this study successfully developed a population pharmacokinetic model and limited sampling strategies that could be applied in clinical practice to assist in therapeutic drug monitoring of polymyxin B in Chinese patients.