Pediatric physiologically based pharmacokinetic (PBPK) models facilitate the estimation of pharmacokinetic (PK) parameters in children under specific exposure conditions. In human health risk assessment, PBPK modeling has been used to determine a chemical-specific human kinetic adjustment factor (HKAF). Due to increased demands in regulatory assessment, model evaluation and qualification have gained growing attention. The aim of this study was to undertake model qualification of pediatric PBPK models for compounds that are primarily metabolized by cytochrome P450 (CYP) enzymes. The objectives were to determine the appropriateness of the virtual individual creating algorithm in PK-Sim® in predicting PK parameters and their variability in children and identify critical system-specific inputs. PBPK models in adults were constructed for several pharmaceuticals (grouped by major clearance process such as CYP3A4). Several age groups of virtual individuals were created to represent children in pediatric clinical studies. The mean and variance of clearance (CL) from virtual populations were compared to observed values. Sensitivity analysis on area under the curve (AUC) was performed. System-specific parameters of virtual children that contribute to inter-individual PK properties were assessed. Eighty-one percent of the comparisons between simulated and observed clearance values were within twofold error. The mean fold errors were 1.1, 1, 0.7 and 1.8 in adolescents, children, infants and neonates, respectively. CL variability was reasonably predicted for 70% of the comparisons with comparable coefficients of variation between observed and predicted. The sensitivity analysis revealed that fraction unbound in plasma, parameters related to CYP enzyme-mediated metabolism and liver volumewere most important in the estimation of pediatric exposure. A comparison of variabilities in weight, height and liver volume in virtual children showed reliable agreement with observed data. The presented results of predictive performance and properties of virtual populations provide confidence in the use of PK-Sim for pediatric PBPK modeling in toxicological applications including PBPK-based-HKAF derivation.

中文翻译：

---

用于CYP450代谢化合物的儿科暴露评估的PK-Sim®儿科模块的模型鉴定。

基于儿科生理学的药代动力学（PBPK）模型有助于在特定暴露条件下估计儿童的药代动力学（PK）参数。在人类健康风险评估中，PBPK模型已用于确定特定于化学物质的人类动力学调节因子（HKAF）。由于对法规评估的需求不断增加，模型评估和资格认证越来越受到关注。这项研究的目的是对主要由细胞色素P450（CYP）酶代谢的化合物进行儿童PBPK模型的模型鉴定。目的是确定PK-Sim®中虚拟个体创建算法在预测PK参数及其对儿童的变异性方面的适当性，并确定关键的系统特定输入。针对几种药物构建了成年人的PBPK模型（按主要清除过程分组，例如CYP3A4）。创建了几个年龄段的虚拟个体，以代表儿童进行儿科临床研究。将虚拟种群的清除率（CL）的平均值和方差与观察值进行比较。进行了曲线下面积（AUC）的敏感性分析。评估了有助于个体间PK特性的虚拟儿童的系统特定参数。模拟和观察到的间隙值之间的比较的81％在两倍误差内。青少年，儿童，婴儿和新生儿的平均折叠误差分别为1.1、1、0.7和1.8。合理地预测了70％的比较的CL变异性，观察和预测之间的变异系数相当。敏感性分析表明，血浆中未结合的部分，与CYP酶介导的代谢有关的参数和肝脏体积在估计儿科暴露中最为重要。对虚拟儿童体重，身高和肝体积的变异性进行比较后，发现与观察到的数据具有可靠的一致性。提出的虚拟人群预测性能和属性的结果提供了对在包括基于PBPK的HKAF在内的毒理学应用中将PK-Sim用于小儿PBPK建模的信心。与CYP酶介导的代谢和肝脏体积有关的参数在估算儿科暴露中最为重要。对虚拟儿童体重，身高和肝体积的变异性进行比较后，发现与观察到的数据具有可靠的一致性。提出的虚拟人群预测性能和属性的结果提供了在毒理学应用（包括基于PBPK的HKAF推导）中将PK-Sim用于小儿PBPK建模的信心。与CYP酶介导的代谢和肝脏体积相关的参数在估算儿科暴露中最为重要。对虚拟儿童体重，身高和肝体积的变异性进行比较后，发现与观察到的数据具有可靠的一致性。提出的虚拟人群预测性能和属性的结果提供了在毒理学应用（包括基于PBPK的HKAF推导）中将PK-Sim用于小儿PBPK建模的信心。