Institut de Neurosciences des Systèmes, Inserm UMR 1106, Aix Marseille Université, Marseille, France.
ExactCure, Nice, France.
J Psychopharmacol. 2024 Dec;38(12):1054-1062. doi: 10.1177/02698811241275630. Epub 2024 Sep 29.
Treatment optimization is mandatory in psychiatric diseases and the use of population pharmacokinetics (popPK) models through model informed precision dosing (MIPD) has the potential to improve patient medical care. In this perspective, meta-modelling methods could provide popPK models with improved predictive performances and most of covariates of interest. The aims of this study were to develop meta-models of clozapine and lithium, assess their predictability and propose optimized dosing regimens for both drugs.
Two popPK models for each drug were retained to develop the meta-models. For clozapine, the model with the best predictive performances and gender as a covariate and one with smoking status were retained. For lithium, the model with the best predictive performances and fat-free mass as covariate and one with glomerular filtration rate were retained.
Both meta-models showed improved predictability compared to the original models. Clozapine meta-model simulations allowed us to propose dosing regimen according to gender and smoking status. Steady-state doses ranged from 375 to 725 mg/day for clozapine once daily, and from 350 to 650 mg/day for clozapine twice daily. Lithium meta-model simulations allowed us to propose dosing regimen according to weight, body mass index, gender and GFR. Our steady-state dose propositions ranged from 625 to 1125 mg/day for males, and from 375 to 750 mg/day for females.
Both meta-models met the acceptability criteria for use in clinical practice on all subpopulations of interest. Those models could be used in the perspective of MIPD for clozapine and lithium.
在精神疾病的治疗中，必须进行治疗优化，而通过模型指导的精准给药（MIPD）使用群体药代动力学（popPK）模型有可能改善患者的医疗护理。在这种情况下，元建模方法可以为 popPK 模型提供改进的预测性能和大多数感兴趣的协变量。本研究的目的是开发氯氮平和锂的元模型，评估其预测能力，并为这两种药物提出优化的给药方案。
保留两种药物的每个药物的 popPK 模型来开发元模型。对于氯氮平，保留预测性能最佳且具有性别协变量的模型和具有吸烟状态的模型。对于锂，保留预测性能最佳且具有瘦体重作为协变量的模型和具有肾小球滤过率的模型。
与原始模型相比，两种元模型均显示出更好的预测能力。氯氮平元模型模拟允许我们根据性别和吸烟状态提出给药方案。氯氮平每日一次的稳态剂量范围为 375 至 725 毫克/天，每日两次的剂量范围为 350 至 650 毫克/天。锂元模型模拟允许我们根据体重、体重指数、性别和 GFR 提出给药方案。我们的稳态剂量建议范围为男性 625 至 1125 毫克/天，女性 375 至 750 毫克/天。
两种元模型均满足在所有感兴趣的亚人群中用于临床实践的可接受性标准。这些模型可用于氯氮平和锂的 MIPD 应用。
