Pharmacokinetic Model Based on Stochastic Simulation and Estimation for Therapeutic Drug Monitoring of Tacrolimus in Korean Adult Transplant Recipients

• Published in: Therapeutic drug monitoring Vol. 44; no. 6; p. 729
• Main Authors: Choi, Suein, Hong, Yunjeong, Jung, Sook-Hyun, Kang, Gaeun, Ghim, Jong-Ryul, Han, Seunghoon
• Format: Journal Article
• Language: English
• Published: United States 01.12.2022
• Subjects: Adult; Cytochrome P-450 CYP3A; Drug Monitoring - methods; Humans; Immunosuppressive Agents - pharmacokinetics; Models, Biological; Republic of Korea; Tacrolimus - pharmacokinetics; Transplant Recipients; Republic of Korea
• ISSN: 1536-3694
• DOI: 10.1097/FTD.0000000000001006

Tacrolimus shows high variability in inter- and intraindividual pharmacokinetics (PK); therefore, it is important to develop an appropriate model for accurate therapeutic drug monitoring (TDM) procedures. This study aimed to develop a pharmacokinetic model for tacrolimus that can be used for TDM procedures in Korean adult transplant recipients by integrating published models with acquired real-world TDM data and evaluating clinically meaningful covariates. Clinical data of 1829 trough blood samples from 269 subjects were merged with simulated data sets from published models and analyzed using a nonlinear mixed-effect model. The stochastic simulation and estimation (SSE) method was used to obtain the final parameter estimates. The final estimated values for apparent clearance, the volume of distribution, and absorption rate were 21.2 L/h, 510 L, and 3.1/h, respectively. The number of postoperative days, age, body weight, and type of transplant organs were the major clinical factors affecting tacrolimus PK. A tacrolimus PK model that can incorporate published PK models and newly collected data from the Korean population was developed using the SSE method. Despite the limitations in model development owing to the nature of TDM data, the SSE method was useful in retrieving complete information from the TDM data by integrating published PK models while maintaining the variability of the model.