Hartman Stan J F, Swaving Joost G E, van Beek Stijn W, van Groen Bianca D, de Hoop Marika, van der Zanden Tjitske M, Ter Heine Rob, de Wildt Saskia N
Department of Pharmacology and Toxicology, Radboud Institute of Health Sciences, Radboudumc, Nijmegen, Netherlands.
Department of Pharmacy, Radboud Institute of Health Sciences, Radboudumc, Nijmegen, Netherlands.
Front Pharmacol. 2020 Dec 16;11:592204. doi: 10.3389/fphar.2020.592204. eCollection 2020.
: Modeling and simulation is increasingly used to study pediatric pharmacokinetics, but clinical implementation of age-appropriate doses lags behind. Therefore, we aimed to develop model-informed doses using published pharmacokinetic data and a decision framework to adjust dosing guidelines based on these doses, using piperacillin and amikacin in critically ill children as proof of concept. : Piperacillin and amikacin pharmacokinetic models in critically ill children were extracted from literature. Concentration-time profiles were simulated for various dosing regimens for a virtual PICU patient dataset, including the current DPF dose and doses proposed in the studied publications. Probability of target attainment (PTA) was compared between the different dosing regimens. Next, updated dosing recommendations for the DPF were proposed, and evaluated using a new framework based on PK study quality and benefit-risk analysis of clinical implementation. : Three studies for piperacillin (critically ill children) and one for amikacin (critically ill pediatric burn patients) were included. Simulated concentration-time profiles were performed for a virtual dataset of 307 critically ill pediatric patients, age range 0.1-17.9 y. PTA for unbound piperacillin trough concentrations >16 mg/L was >90% only for continuous infusion regimens of 400 mg/kg/day vs. 9.7% for the current DPF dose (80 mg/kg/6 h, 30 min infusion). Amikacin PTA was >90% with 20 mg/kg/d, higher than the PTA of the DPF dose of 15 mg/kg/d (63.5%). Using our new decision framework, altered DPF doses were proposed for piperacillin (better PTA with loading dose plus continuous infusion), but not for amikacin (studied and target population were not comparable and risk for toxicity with higher dose). : We show the feasibility to develop model-informed dosing guidelines for clinical implementation using existing pharmacokinetic data. This approach could complement literature and consensus-based dosing guidelines for off-label drugs in the absence of stronger evidence to support pediatricians in daily practice.
建模与仿真越来越多地用于研究儿科药代动力学，但适合不同年龄段的剂量在临床中的应用却滞后了。因此，我们旨在利用已发表的药代动力学数据和一个决策框架来制定基于模型的剂量，并根据这些剂量调整给药指南，以危重症儿童使用哌拉西林和阿米卡星作为概念验证。
从文献中提取了危重症儿童的哌拉西林和阿米卡星药代动力学模型。针对一个虚拟的儿科重症监护病房（PICU）患者数据集的各种给药方案模拟了浓度-时间曲线，包括当前的每日推荐剂量（DPF）以及研究出版物中提出的剂量。比较了不同给药方案之间达到目标的概率（PTA）。接下来，提出了DPF的更新给药建议，并使用基于药代动力学研究质量和临床应用效益-风险分析的新框架进行评估。
纳入了三项关于哌拉西林（危重症儿童）的研究和一项关于阿米卡星（儿科重症烧伤患者）的研究。对307名年龄范围在0.1至17.9岁的危重症儿科患者的虚拟数据集进行了浓度-时间曲线模拟。仅对于400mg/kg/天的持续输注方案，未结合的哌拉西林谷浓度>16mg/L的PTA>90%，而当前DPF剂量（80mg/kg/6小时，30分钟输注）的PTA为9.7%。阿米卡星20mg/kg/d的PTA>90%，高于DPF剂量15mg/kg/d的PTA（为63.5%）。使用我们的新决策框架，针对哌拉西林提出了改变后的DPF剂量（负荷剂量加持续输注时PTA更佳），但对于阿米卡星未提出改变（研究对象和目标人群不可比，且更高剂量存在毒性风险）。
我们展示了利用现有药代动力学数据制定基于模型的给药指南以供临床应用的可行性。在缺乏更有力证据支持儿科医生日常实践的情况下，这种方法可以补充针对未标记药物的文献和基于共识的给药指南。
