부산대학교 도서관

There has been a lot of academic and regulatory interest regarding bridging clinical trials between different populations. The aims of this study were to:Develop a PBPK model for the US paediatric population (USPP) incorporating demographic and CYP3A5 phenotype frequency of different ethnic groups (White, Hispanic, Black and Asian).Apply the USPP to predict midazolam pharmacokinetics (PK) of a clinical study performed in the US.Demographic information, height for age and weight for height relationships, and CYP3A5 phenotype frequencies were established for each US ethnic group using NHANES and literature data. Four separate US paediatric PBPK populations were defined within the Simcyp Simulator (v21).Simulations of IV and oral midazolam PK were made in the USPP and a North European paediatric population (NECP) and compared with the clinical study. The reported trial design was matched as closely as possible and 400 subjects, 0.5 female, age 0.5 to 16y were run for each population.For a 0.25mg/kg oral dose, the predicted AUC0-inf was 143±109 and 225±136 ng/ml.h and Cmax was 57.4 and 79.4ng/ml for the USPP and NECP, respectively. The observed AUC0-inf and Cmax was 137±86 ng/ml.h and 55.6ng/ml, respectively. The predicted AUC was 195, 115, 150 and 135 ng/ml.h for the White, Black, Hispanic and Asian USPP and Cmax was 72, 48, 58 and 54ng/ml, respectively.Prediction of midazolam PK was improved by including the different ethnic groups for the USPP. However, significant differences can be observed between these groups for drugs where elimination changes due to phenotypic enzyme expression (e.g. CYP3A5) and it is important that clinical studies present this information.