Application of Physiologically Based Pharmacokinetic Modeling in the Research of Anti-HIV Drugs.

Physiologically based pharmacokinetic (PBPK) models allow integrated representations of the time course of drug concentrations in the important organs and sites of drug action, toxicity, absorption, metabolism, and excretion. PBPK models have multiple compartments representing defined organs or physiological spaces, with parameters that are potentially directly measurable. They lie on a continuum of pharmacokinetic model types, which includes Compartmental (least complex), Semi-PBPK, PBPK, and Systems Biology (most complex) models. Compartmental models are typically parameterized using a “top-down” approach, where the model is fitted to a specific data set to estimate parameter values. PBPK models are typically parameterized using a “bottom-up” approach, where model parameters are derived from literature, in vitro data, or scaled data from another species. Although there is increasing availability and use of commercial software for PBPK modeling, it is feasible to construct bespoke PBPK for specific projects using general-purpose software platforms. The challenge here is often collecting, collating, and justifying the data used to parameterize the model. In this chapter, the fundamental equations for a simple PBPK models are presented with respect to key “submodels” of an example whole-body model. The current rise in the rate of publication and application of PBPK models is likely to be sustained in the foreseeable future. Contributing factors will likely be the wider use of commercial PBPK models, the advent of broader collective efforts to advance and coordinate PBPK and Systems Biology modeling, and the increasing ease with which bespoke PBPK models can be coded and shared.

Physiologically based pharmacokinetic (PBPK) modeling is a mechanistic dynamic modeling approach that can be used to predict or retrospectively describe changes in drug exposure due to drug-drug interactions (DDIs). With advancements in commercially available PBPK software, PBPK DDI modeling has become a mainstream approach from early drug discovery through to late-stage drug development and is often used to support regulatory packages for new drug applications. This Minireview will briefly describe the approaches to predicting DDI using PBPK and static modeling approaches, the basic model structures and features inherent to PBPK DDI models, and key examples where PBPK DDI models have been used to describe complex DDI mechanisms. Future directions aimed at using PBPK models to characterize transporter-mediated DDI, predict DDI in special populations, and assess the DDI potential of protein therapeutics will be discussed. A summary of the 209 PBPK DDI examples published to date in 2023 will be provided. Overall, current data and trends suggest a continued role for PBPK models in the characterization and prediction of DDI for therapeutic molecules. SIGNIFICANCE STATEMENT: Physiologically based pharmacokinetic (PBPK) models have been a key tool in the characterization of various pharmacokinetic phenomena, including drug-drug interactions. This Minireview will highlight recent advancements and publications around physiologically based pharmacokinetic drug-drug interaction modeling, an important area of drug discovery and development research in light of the increasing prevalence of polypharmacology in clinical settings.

Physiologically-based pharmacokinetic (PBPK) modeling has been widely used in human risk assessment and in early drug development to predict human PK from in vitro and/or in vivo animal data. Recently, the application of PBPK modeling has been extended to the evaluation of drug-drug interactions. For most xenobiotic agents, the PK event scale such as elimination is in hours or days. This is much longer than the transit time of the agent in the body, and a PBPK model can be significantly simplified through lumping based on the physiochemical properties, mass transfer, and biotransformation. However, for a xenobiotic agent with a short PK event scale, e.g. in minutes, such an approach is not applicable. In this manuscript, the authors used the observed PK data from an ultrasound contrast agent to illustrate the role of a short PK event scale in the development of a suitable PBPK model. The model development process showed that a PBPK model assuming uniform venous and arterial blood pools, with a static lung model including alveolar and tissue regions, was unable to adequately capture the characteristics of the PK of the agent. Detailed information describing the pulmonary and cardiovascular circulation, and a heterogeneous dynamic lung model became necessary for the model. This exercise once again demonstrates the importance of the principles and methodologies that have been established since the 1960s that need to be followed during PBPK model development.

Pharmacokinetic (PK) modeling is a useful method for investigating drug absorption, distribution, metabolism, and excretion. The most commonly used mathematical models in PK modeling are the compartment model and physiologically based pharmacokinetic (PBPK) model. Although the theoretical characteristics of each model are well known, there have been few comparative studies of the compatibility of the models. Therefore, we evaluated the compatibility of PBPK and compartment models using the lumping method with 20 model compounds. The PBPK model was theoretically reduced to the lumped model using the principle of grouping tissues and organs that show similar kinetic behaviors. The area under the concentration–time curve (AUC) based on the simulated concentration and PK parameters (drug clearance [CL], central volume of distribution [Vc], peripheral volume of distribution [Vp]) in each model were compared, assuming administration to humans. The AUC and PK parameters in the PBPK model were similar to those in the lumped model within the 2-fold range for 17 of 20 model compounds (85%). In addition, the relationship of the calculated Vd/fu (volume of distribution [Vd], drug-unbound fraction [fu]) and the accuracy of AUC between the lumped model and compartment model confirmed their compatibility. Accordingly, the compatibility between PBPK and compartment models was confirmed by the lumping method. This method can be applied depending on the requirement of compatibility between the two models.

Over the last decade, physiologically based pharmacokinetics (PBPK) application has been extended significantly not only to predicting preclinical/human PK but also to evaluating the drug-drug interaction (DDI) liability at the drug discovery or development stage. Herein, we describe a case study to illustrate the use of PBPK approach in predicting human PK as well as DDI using in silico, in vivo and in vitro derived parameters. This case was composed of five steps such as: simulation, verification, understanding of parameter sensitivity, optimization of the parameter and final evaluation. Caffeine and ciprofloxacin were used as tool compounds to demonstrate the “fit for purpose” application of PBPK modeling and simulation for this study. Compared to caffeine, the PBPK modeling for ciprofloxacin was challenging due to several factors including solubility, permeability, clearance and tissue distribution etc. Therefore, intensive parameter sensitivity analysis (PSA) was conducted to optimize the PBPK model for ciprofloxacin. Overall, the increase in Cmax of caffeine by ciprofloxacin was not significant. However, the increase in AUC was observed and was proportional to the administered dose of ciprofloxacin. The predicted DDI and PK results were comparable to observed clinical data published in the literatures. This approach would be helpful in identifying potential key factors that could lead to significant impact on PBPK modeling and simulation for challenging compounds.

Objective: Preclinical studies have demonstrated that CFAK-C4 has anti-tumor efficacy in a variety of malignancies. To maximize its efficacy, it is necessary to understand the pharmacokinetic (PK) properties of CFAK-C4 in the body. A PK study was conducted to characterize CFAK-C4 disposition in plasma and various tissues, including brain, heart, liver, lung, muscle, spleen, and sternum. Subsequently, a physiologically-based pharmacokinetic (PBPK) model was developed to simultaneously characterize and predict plasma and tissue CFAK-C4 concentrations and thus help guide future dosing strategies alone and in combination. Methods: Female CD-1 mice received a single IP injection of CFAK-C4 with a dose of 50mg/kg, and then plasma and tissue samples were collected at serial time points after injection. Three mice were sacrificed at each time point. CFAK-C4 concentrations were determined by a validated LC-MS/MS method. Noncompartmental PK analysis was performed using WinNonlin (Pharsight, Version 5.3) for PK parameters. CFAK-C4 concentration-time profiles were fitted with a PBPK model composed of compartments for plasma, all the measured tissues, peritoneum, and a remainder compartment which represented all other tissues where CFAK-C4 was not measured, using ADAPT5 (BMSR, USC). The PBPK model was assessed by goodness-of-fit plots together with agreement of estimated parameters with noncompartmental analysis. Results: CFAK-C4 concentrations followed a monoexponential decay in plasma, while there was a longer elimination phase observed in tissues. As a result, CFAK-C4 concentration-time profiles in plasma and tissues were simultaneously fitted into a plasma-flow-rate-limited PBPK model successfully. Partition coefficients (Kp), as a measure of the extent of tissue distribution, and plasma clearance (Cl) were estimated by the PBPK model, while the volumes of distribution and plasma flow rates of tissues were fixed to physiological values. The model predicts that CFAK-C4 can be well distributed to various tissues quickly, and Cmax is achieved within half an hour after IP injection. The estimated Cl, 0.111 (±7.4%) l/h, was similar to the value from non-compartmental analysis (NCA) (0.0925 L/h). The model also predicts CFAK-C4 has the highest penetration to lung, with a Kp of 28.1 (±15.2%), followed by brain, with a Kp of 16.6 (±11.6%), and it has the lowest penetration to muscle, with a Kp of 3.6 (±8.3%). Conclusions: The wide tissue distribution of CFAK-C4 provides a great advantage for maximizing its anti-tumor efficacy. The PBPK model predicts CFAK-C4 plasma and tissue concentrations reasonably well. This PBPK model will be used as a tool to build PK/PD models to characterize the temporal relationship between CFAK-C4 pharmacokinetics and its antitumor efficacy and thus help decide future dosing strategies for the treatment of various malignancies. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3786. doi:1538-7445.AM2012-3786

The application of PBPK models in estimating human brain tissue manganese concentrations

Intratumoral pharmacokinetic (PK) and pharmacodynamic (PD) heterogeneity contribute to variability in NB tumor response to chemotherapy and can be responsible for tumor relapse. Herein we propose to develop a whole body PBPK model with an individualized tumor compartment to derive individual tumor specific concentration-time profiles for the NB standard of care drug TPT. This model can then relate intratumoral heterogeneity in tumor blood flow to PD response and antitumor effects. PK studies of TPT (0.6, 1.25, 5, and 20 mg/kg, IV bolus) will be performed in CD1 nude mice (n = 3 mice/time point) bearing orthotopic NB (NB5) xenograft. Blood samples will be collected at predetermined time points using cardiac puncture, and plasma separated and stored until analysis. Animals will be perfused using saline solution to remove residual blood, and tissue samples including tumor, muscle, adipose, bone, liver, gallbladder, kidney, spleen, lungs, brain, heart, duodenum, and large intestine collected. TPT concentrations in plasma and tissue homogenate samples will be quantified using a validated HPLC fluorescence spectrophotometry method. Tumor samples will be divided into two sections each, one for TPT quantification and one for immunohistochemistry of PD markers for DNA damage (γ-H2AX) and apoptosis (CASP3). A cohort of mice will be used to quantify tumor blood flow using contrast-enhanced ultrasound (CEUS) using MicroMarker® microbubbles prior to dosing the mice for the PK study. TPT plasma and tissue concentration-time data will be used to develop the whole-body PBPK model with an individualized tumor compartment using NONMEM. Individual tumor perfusion data obtained using CEUS will be combined with the PBPK model to derive tumor specific concentration-time profiles. A preliminary study conducted in non-tumor bearing mice receiving TPT 5 mg/kg showed that TPT plasma and tissue concentration-time data were reasonably described by our PBPK model. As expected from our previous studies, the brain tissue was found to have the lowest exposure to TPT with a brain to plasma partition coefficient (Kp,brain ∼ 8%). We also observed high permeability of TPT (Kp > 1) into the gallbladder, duodenum, large intestine, spleen, liver and kidney. In future we will study the correlations between individual tumor concentrations based on our comprehensive PBPK model and γ-H2AX and CASP3 activity. Citation Format: Yogesh T. Patel, Megan O. Jacus, Abbas Shirinifard, Abigail D. Davis, Suresh Thiagarajan, Stacy L. Throm, Vinay M. Daryani, Andras Sablauer, Clinton F. Stewart. Development of a whole body physiologically-based pharmacokinetic (PBPK) model with individualized tumor compartment for topotecan (TPT) in mice bearing neuroblastoma (NB). [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4519. doi:10.1158/1538-7445.AM2015-4519

This was the first study to construct a physiologically-based pharmacokinetic (PBPK) model for mirabegron which incorporates the overall elimination pathways of metabolism by cytochrome P450 (CYP) 3A4, uridine 5'-diphosphate-glucuronosyltransferase (UGT) 2B7, and butyrylcholinesterase (BChE) and renal excretion. The objective was to assess the risk of drug-drug interactions (DDIs) by estimating the contribution of each elimination pathway and simulating the magnitude of the DDIs with UGT2B7 inhibitors.A PBPK model for mirabegron was constructed to reproduce the plasma concentration-time curves from a phase 1 study and the magnitude of the DDI with ketoconazole taking into account the overall elimination pathways. The PBPK model was subsequently verified using data from other DDI studies.The constructed PBPK model estimated the contribution for each elimination pathway: 44% and 29% for CYP3A4 and UGT2B7 in the liver, 1.6% for UGT2B7 in the kidney, 3.2% for BChE in plasma, and 22% for renal excretion.Co-administration of probenecid (an UGT2B7 inhibitor) or fluconazole (an UGT2B7 and CYP3A4 inhibitor) was predicted to increase area under the curve for mirabegron to 115% or 174%, respectively.In conclusion, PBPK modeling and simulation revealed a low DDI risk for mirabegron following co-administration with BChE or UGT2B7 inhibitors.

ObjectiveThe objective of this study was to compare the predictive performances of a glomerular filtration rate (GFR) model with a physiologically based pharmacokinetic (PBPK) model to predict total or renal clearance or area under the curve of renally excreted drugs in subjects with varying degrees of renal impairment.MethodsFrom the literature, 11 studies were randomly selected in which total or renal clearance or area under the curve of drugs in subjects with different degrees of renal impairment were predicted by PBPK models. In these published studies, drugs were given to subjects intravenously or orally. The PBPK model was generally a whole-body model whereas the GFR model was as follows: Predicted total clearance (CLT) = CLT in healthy subjects × (GFR in RI/GFR in H), Predicted AUC = AUC in healthy subjects × (GFR in H/GFR in RI), where H is the healthy subjects and RI is renal impairment. The predicted clearance or area under the curve values using PBPK and GFR models were compared with the observed (experimental pharmacokinetic) values. The acceptable prediction error was within the 0.5- to 2-fold or 0.5- to 1.5-fold prediction error.ResultsThere were 33 drugs with a total number of 101 observations (area under the curve, total and renal clearance in subjects with mild, moderate, and severe renal impairment). From PBPK and GFR models, out of 101 observations, 94 (93.1%) and 96 (95.0%) observations were within the 0.5- to 2-fold prediction error, respectively.ConclusionsThis study indicates that the predictive power of a simple GFR model is similar to a PBPK model for the prediction of clearance or area under the curve in subjects with renal impairment. The GFR method is simple, robust, and reliable and can replace complex empirical PBPK models.

The objective of this study was to compare the predictive performance of an allometric model with that of a physiologically based pharmacokinetic (PBPK) model to predict clearance or area under the concentration-time curve (AUC) of drugs in subjects from neonates to adolescents. From the literature, 10 studies were identified in which clearance or AUC of drugs from neonates to adolescents was predicted by PBPK models. In these published studies, drugs were given to children either by intravenous or oral route. The allometric model was an age-dependent exponent (ADE) model for the prediction of clearance across the age groups. The predicted clearance or AUC values from the PBPK and ADE models were compared with the experimental values. The acceptable prediction error was the percentage of subjects within an 0.5- to 2-fold or 0.5- to 1.5-fold prediction error. There were 73 drugs with a total of 372 observations. From PBPK and allometric models, 91.1% and 90.6% of observations were within 0.5- to 2-fold prediction error, respectively. For children ≤2 years old (n = 130), PBPK and allometric models had 89% and 87% of observations within the 0.5- to 2-fold prediction error, respectively. This study indicates that the predictive power of PBPK and allometric models was essentially similar for the prediction of clearance or AUC in pediatric subjects ranging from neonates to adolescents.

Applications of PBPK/PBBM modeling in generic product development: An industry perspective

Role of Transporters in Drug Development.

BackgroundTenofovir is a drug used in combination with other anti-HIV drugs to treat patients with HIV-1 infection. It is used during pregnancy to reduce the risk of HIV transmission to...

Buprenorphine plays a crucial role in the therapeutic management of pain in adults, adolescents and pediatric subpopulations. However, only few pharmacokinetic studies of buprenorphine in children, particularly neonates, are available as conducting clinical trials in this population is especially challenging. Physiologically-based pharmacokinetic (PBPK) modeling allows the prediction of drug exposure in pediatrics based on age-related physiological differences. The aim of this study was to predict the pharmacokinetics of buprenorphine in pediatrics with PBPK modeling. Moreover, the drug-drug interaction (DDI) potential of buprenorphine with CYP3A4 and P-glycoprotein perpetrator drugs should be elucidated. A PBPK model of buprenorphine and norbuprenorphine in adults has been developed and scaled to children and preterm neonates, accounting for age-related changes. One-hundred-percent of the predicted AUClast values in adults (geometric mean fold error (GMFE): 1.22), 90% of individual AUClast predictions in children (GMFE: 1.54) and 75% in preterm neonates (GMFE: 1.57) met the 2-fold acceptance criterion. Moreover, the adult model was used to simulate DDI scenarios with clarithromycin, itraconazole and rifampicin. We demonstrate the applicability of scaling adult PBPK models to pediatrics for the prediction of individual plasma profiles. The novel PBPK models could be helpful to further investigate buprenorphine pharmacokinetics in various populations, particularly pediatric subgroups.