Janthima Methaneethorn1,2, Manupat Lohitnavy3,4, Kamonwan Onlamai5, Nattawut Leelakanok6. 1. Pharmacokinetic Research Unit, Department of Pharmacy Practice, Faculty of Pharmaceutical Sciences, Naresuan University, Phitsanulok, 65000, Thailand. janthima.methaneethorn@gmail.com. 2. Center of Excellence for Environmental Health and Toxicology, Naresuan University, Phitsanulok, Thailand. janthima.methaneethorn@gmail.com. 3. Pharmacokinetic Research Unit, Department of Pharmacy Practice, Faculty of Pharmaceutical Sciences, Naresuan University, Phitsanulok, 65000, Thailand. 4. Center of Excellence for Environmental Health and Toxicology, Naresuan University, Phitsanulok, Thailand. 5. Department of Pharmacy, Bhumibol Adulyadej Hospital, Bangkok, Thailand. 6. Department of Clinical Pharmacy, Faculty of Pharmaceutical Sciences, Burapha University, Chonburi, Thailand.
Abstract
BACKGROUND AND OBJECTIVE: Tacrolimus is a narrow therapeutic index drug with high pharmacokinetic variability, and several tacrolimus population pharmacokinetic (PopPK) models were developed to guide individualized drug dosing. These models, however, may not perform well in other clinical settings. Therefore, we aimed to assess the predictive ability of published tacrolimus PopPK models using a dataset of Thai kidney transplant patients. METHODS: The external dataset was retrospectively collected from medical records of Bhumibol Adulyadej Hospital, Thailand. Published tacrolimus PopPK models were systematically searched from PubMed, Science Direct, CINAHL Complete, and Scopus databases. Models conducted using a nonlinear mixed-effects approach with covariate resemblance to our external dataset were selected. The external dataset consisted of Thai kidney transplant patients receiving oral immediate- or extended-release tacrolimus formulations twice or once daily, respectively. Accuracy and precision of predicted concentrations were evaluated using mean absolute prediction error (MAPE), root mean square error (RMSE), and goodness of fit plots. RESULTS: Only three models produced acceptable population predictions with the MAPE of < 50%. By using the Bayesian posthoc estimate of individual pharmacokinetic parameters, all models well performed with the MAPE and RMSE of < 30% and 40%, respectively, except two models; one could not successfully converge and the other substantially underpredicted tacrolimus concentrations. CONCLUSION: We evaluated ten tacrolimus PopPK models, and eight models resulted in satisfactorily individual predicted tacrolimus concentrations in Thai kidney transplant patients and may be used to aid tacrolimus dose adjustment along with a clinical judgment.
BACKGROUND AND OBJECTIVE: Tacrolimus is a narrow therapeutic index drug with high pharmacokinetic variability, and several tacrolimus population pharmacokinetic (PopPK) models were developed to guide individualized drug dosing. These models, however, may not perform well in other clinical settings. Therefore, we aimed to assess the predictive ability of published tacrolimus PopPK models using a dataset of Thai kidney transplant patients. METHODS: The external dataset was retrospectively collected from medical records of Bhumibol Adulyadej Hospital, Thailand. Published tacrolimus PopPK models were systematically searched from PubMed, Science Direct, CINAHL Complete, and Scopus databases. Models conducted using a nonlinear mixed-effects approach with covariate resemblance to our external dataset were selected. The external dataset consisted of Thai kidney transplant patients receiving oral immediate- or extended-release tacrolimus formulations twice or once daily, respectively. Accuracy and precision of predicted concentrations were evaluated using mean absolute prediction error (MAPE), root mean square error (RMSE), and goodness of fit plots. RESULTS: Only three models produced acceptable population predictions with the MAPE of < 50%. By using the Bayesian posthoc estimate of individual pharmacokinetic parameters, all models well performed with the MAPE and RMSE of < 30% and 40%, respectively, except two models; one could not successfully converge and the other substantially underpredicted tacrolimus concentrations. CONCLUSION: We evaluated ten tacrolimus PopPK models, and eight models resulted in satisfactorily individual predicted tacrolimus concentrations in Thai kidney transplant patients and may be used to aid tacrolimus dose adjustment along with a clinical judgment.
Authors: W Zhao; V Elie; G Roussey; K Brochard; P Niaudet; V Leroy; C Loirat; P Cochat; S Cloarec; J L André; F Garaix; A Bensman; M Fakhoury; E Jacqz-Aigrain Journal: Clin Pharmacol Ther Date: 2009-10-28 Impact factor: 6.875
Authors: R Venkataramanan; A Swaminathan; T Prasad; A Jain; S Zuckerman; V Warty; J McMichael; J Lever; G Burckart; T Starzl Journal: Clin Pharmacokinet Date: 1995-12 Impact factor: 6.447