Department of Chemical and Biological Engineering, Illinois Institute of Technology, Chicago, Illinois, USA.
Department of Biobehavioral Nursing Science, University of Illinois at Chicago, Chicago, Illinois, USA.
AAPS J. 2024 Aug 19;26(5):94. doi: 10.1208/s12248-024-00962-2.
Chronic kidney disease (CKD) is a complication of diabetes that affects circulating drug concentrations and elimination of drugs from the body. Multiple drugs may be prescribed for treatment of diabetes and co-morbidities, and CKD complicates the pharmacotherapy selection and dosing regimen. Characterizing variations in renal drug clearance using models requires large clinical datasets that are costly and time-consuming to collect. We propose a flexible approach to incorporate impaired renal clearance in pharmacokinetic (PK) models using descriptive statistics and secondary data with mechanistic models and PK first principles. Probability density functions were generated for various drug clearance mechanisms based on the degree of renal impairment and used to estimate the total clearance starting from glomerular filtration for metformin (MET) and dapagliflozin (DAPA). These estimates were integrated with PK models of MET and DAPA for simulations. MET renal clearance decreased proportionally with a reduction in estimated glomerular filtration rate (eGFR) and estimated net tubular transport rates. DAPA total clearance varied little with renal impairment and decreased proportionally to reported non-renal clearance rates. Net tubular transport rates were negative to partially account for low renal clearance compared with eGFR. The estimated clearance values and trends were consistent with MET and DAPA PK characteristics in the literature. Dose adjustment based on reduced clearance levels estimated correspondingly lower doses for MET and DAPA while maintaining desired dose exposure. Estimation of drug clearance rates using descriptive statistics and secondary data with mechanistic models and PK first principles improves modeling of CKD in diabetes and can guide treatment selection.
慢性肾脏病（CKD）是糖尿病的一种并发症，会影响循环药物浓度和药物从体内的消除。可能会为治疗糖尿病和合并症开多种药物，而 CKD 使药物治疗的选择和剂量方案复杂化。使用模型来描述肾脏药物清除率的变化需要使用大型临床数据集，这些数据集的收集既昂贵又耗时。我们提出了一种灵活的方法，使用描述性统计和二次数据与机制模型和 PK 第一原理相结合，在药代动力学（PK）模型中纳入受损的肾脏清除率。根据肾功能损害的程度，为各种药物清除机制生成了概率密度函数，并用于从肾小球滤过开始估算二甲双胍（MET）和达格列净（DAPA）的总清除率。这些估算值与 MET 和 DAPA 的 PK 模型集成进行模拟。MET 的肾清除率与估计肾小球滤过率（eGFR）的降低成正比，与估计的净肾小管转运率成比例降低。DAPA 的总清除率与肾功能损害变化不大，与报告的非肾清除率成比例降低。净肾小管转运率为负，部分解释了与 eGFR 相比，肾脏清除率较低的原因。估算的清除率值和趋势与文献中 MET 和 DAPA 的 PK 特征一致。根据估算的清除率降低水平进行剂量调整，相应降低了 MET 和 DAPA 的剂量，同时维持了所需的剂量暴露。使用描述性统计和二次数据与机制模型和 PK 第一原理估算药物清除率可改善糖尿病中 CKD 的建模，并可指导治疗选择。
