Application of a physiologically based pharmacokinetic model for the prediction of mirabegron plasma concentrations in a population with severe renal impairment

Chronic kidney disease and physiologically based pharmacokinetic modeling: a critical review of existing models

INTRODUCTION

Physiologically based pharmacokinetic (PBPK) modeling is a paradigm shift in this era for determining the exposure of drugs in pediatrics, geriatrics, and patients with chronic diseases where clinical trials are difficult to conduct.

AREAS COVERED

This review has collated data regarding published PBPK models on chronic kidney disease (CKD), including the drug and system-specific input model parameters and model evaluation criteria. Four databases were used from 13th June 2023 to 10th July 2023 for identifying the relevant studies that met the inclusion/exclusion criteria. Alterations in plasma protein (albumin/alpha-1 acid glycoprotein), gastric emptying time, hematocrit, small intestinal transit time, the abundance of cytochrome (CYP) 450 enzymes, glomerular filtration rate, and physicochemical parameters for different drugs were explicitly elaborated from earlier reported studies. Moreover, model evaluation depicted that models in CKD for most of the included drugs were within the allowed two-fold error range.

EXPERT OPINION

This review will provide insights for researchers on applying PBPK models in managing patients with different levels of CKD to prevent undesirable side effects and increase the effectiveness of drug therapy.

Understanding Statin-Roxadustat Drug-Drug-Disease Interaction Using Physiologically-Based Pharmacokinetic Modeling

A different drug-drug interaction (DDI) scenario may exist in patients with chronic kidney disease (CKD) compared with healthy volunteers (HVs), depending on the interplay between drug-drug and disease (drug-drug-disease interaction (DDDI)). Physiologically-based pharmacokinetic (PBPK) modeling, in lieu of a clinical trial, is a promising tool for evaluating these complex DDDIs in patients. However, the prediction confidence of PBPK modeling in the severe CKD population is still low when nonrenal pathways are involved. More mechanistic virtual disease population and robust validation cases are needed. To this end, we aimed to: (i) understand the implications of severe CKD on statins (atorvastatin, simvastatin, and rosuvastatin) pharmacokinetics (PK) and DDI; and (ii) predict untested clinical scenarios of statin-roxadustat DDI risks in patients to guide suitable dose regimens. A novel virtual severe CKD population was developed incorporating the disease effect on both renal and nonrenal pathways. Drug and disease PBPK models underwent a four-way validation. The verified PBPK models successfully predicted the altered PKs in patients for substrates and inhibitors and recovered the observed statin-rifampicin DDIs in patients and the statin-roxadustat DDIs in HVs within 1.25- and 2-fold error. Further sensitivity analysis revealed that the severe CKD effect on statins PK is mainly mediated by hepatic BCRP for rosuvastatin and OATP1B1/3 for atorvastatin. The magnitude of statin-roxadustat DDI in patients with severe CKD was predicted to be similar to that in HVs. PBPK-guided suitable dose regimens were identified to minimize the risk of side effects or therapeutic failure of statins when co-administered with roxadustat.

Improved Pharmacokinetic Feasibilities of Mirabegron-1,2-Ethanedisulfonic Acid, Mirabegron-1,5-Naphthalenedisulfonic Acid, and Mirabegron-L-Pyroglutamic Acid as Co-Amorphous Dispersions in Rats and Mice

Mirabegron (MBR) is a β3-adrenoceptor agonist used for treating overactive bladder syndrome. Due to its poor solubility and low bioavailability (F), the development of novel MBR formulations has garnered increasing attention. Recently, co-amorphous dispersions of MBR, such as MBR-1,2-ethanedisulfonic acid (MBR-EFA), MBR-1,5-naphthalenedisulfonic acid (MBR-NDA), and MBR-L-pyroglutamic acid (MBR-PG), have been developed, showing improved solubility and thermodynamic stability. Nevertheless, the pharmacokinetic feasibility of these co-amorphous dispersions has not been evaluated. Therefore, this study aimed to characterize the pharmacokinetic profiles of MBR-EFA, MBR-NDA, and MBR-PG in rats and mice. Our results exhibited that relative F24h and AUC0-24h values of MBR in MBR-EFA, MBR-NDA, and MBR-PG rats were increased by 143-195% compared with the MBR rats. The absolute F24h, relative F24h, and AUC0-24h values of MBR in MBR-EFA and MBR-NDA mice were enhanced by 178-234% compared with the MBR mice. In tissue distribution, MBR was extensively distributed in the gastrointestinal tract, liver, kidneys, lung, and heart of mice. Notably, MBR distribution in the liver, kidneys, and lung was considerably high in MBR-EFA, MBR-NDA, or MBR-PG mice compared with MBR mice. These findings highlight the potential of these co-amorphous dispersions to enhance oral F of MBR.

Physiologically Based Pharmacokinetic Modelling of Glycopyrronium in Patients With Renal Impairment

Glycopyrronium bromide, a synthetic anticholinergic agent used to treat patients with chronic obstructive pulmonary disease (COPD), is eliminated from the body by renal excretion and therefore systemic exposure is expected to be increased in patients with decreasing renal function. Despite enrollment of patients with decreasing renal function to evaluate the impact of renal impairment on the pharmacokinetics of glycopyrronium in clinical studies, no patients with severe renal impairment were included. A physiologically based pharmacokinetic (PBPK) model was developed in patients with COPD with normal renal function and used to predict systemic exposure of glycopyrronium in patients with severe renal impairment. The model accurately predicted plasma concentration-time profiles in patients with normal renal function, and mild and moderate renal impairment; the predicted and observed AUC and C_max in these populations were similar. Compared to patients with normal renal function, a 1.20-, 1.45-, and 1.59-fold increase AUC was predicted in patients with mild, moderate, and severe renal impairment, respectively, suggesting dose adjustment is not necessary in patients with renal impairment. In conclusion, PBPK models, verified with clinical study data from patients with normal renal function, can potentially be used to predict the pharmacokinetics and recommended dose adjustment for patients with renal impairment.