Woojin Jung’s research while affiliated with Chungnam National University and other places

Given the high prevalence of the population who have metabolic dysfunction‐associated steatohepatitis (MASH), interest is growing in MASH‐targeted treatments. However, currently, there has been only one regulatory approved drug for MASH (Rezdiffra). Pioglitazone, a commonly used type 2 diabetes mellitus drug, is currently used off‐label for the treatment of MASH. Our study aimed to perform a model‐based meta‐analysis to quantitatively examine the efficacy of pioglitazone in improving histological parameters and liver enzymes in patients with MASH. A comprehensive search was performed in Pubmed and clinicaltrials.gov. We collected histological outcomes (including steatosis, inflammation, ballooning, and fibrosis) and liver enzyme data. Due to sparse data, the gathered histological outcomes were used to generate virtual data. Next, model development for the virtual histological dataset was performed using a logistic model. In addition, Weibull and exponential models were tested to find the best fit for liver enzyme data. Model evaluations were carried out by visual predictive check, bootstrap method, and stacked bar plot. Eight studies with 540 patients were included. A logit model was used to analyze four outcomes. The results showed that using pioglitazone improved all four histological parameters. These effects are dose‐ and time‐dependent under the Emax‐time model for steatosis and ballooning, and under the linear relationship for inflammation and fibrosis. For liver enzymes, the Weibull model fitted well for both ALT and AST data. In conclusion, the developed models of pioglitazone may serve as a benchmark to assess the effectiveness of novel MASH‐targeted treatments.

Ritonavir (RTV) is a potent CYP3A inhibitor that is widely used as a pharmacokinetic (PK) enhancer to increase exposure to select protease inhibitors. However, as a strong and complex perpetrator of CYP3A interactions, RTV can also enhance the exposure of other co‐administered CYP3A substrates, potentially causing toxicity. Therefore, the prediction of drug–drug interactions (DDIs) and estimation of dosing requirements for concomitantly administered drugs is imperative. In this study, we aimed to develop a physiologically‐based PK (PBPK) model for RTV using the PK‐sim® software platform. A total of 13 clinical PK studies of RTV covering a wide dose range (100 to 600 mg including both single and multiple dosing), and eight clinical DDI studies with RTV on CYP3A and P‐gp substrates, including alprazolam, midazolam, rivaroxaban, clarithromycin, fluconazole, sildenafil, and digoxin were used for the model development and evaluation. Chronopharmacokinetic differences (between morning vs. evening doses) and limitations in parameter estimation for biochemical processes of RTV from in vitro studies were incorporated in the PBPK model. The final developed PBPK model predicted 100% of RTV AUClast and Cmax within a twofold dimension error. The geometric mean fold error (GMFE) from all PK datasets was 1.275 and 1.194, respectively. In addition, 97% of the DDI profiles were predicted with the DDI ratios within a twofold dimension error. The GMFE values from all DDI datasets were 1.297 and 1.212, respectively. Accordingly, this model could be applied to the prediction of DDI profiles of RTV and CYP3A substrates and used to estimate dosing requirements for concomitantly administered drugs.

Finerenone, a novel selective non-steroidal mineralocorticoid receptor antagonist, has been indicated in chronic kidney disease associated with type 2 diabetes mellitus. Considering the potential complications of diabetes, finerenone can be co-administered with various drugs, including fluconazole, diltiazem, and ritonavir. Given that finerenone is a substrate of cytochrome P450 (CYP) 3A4, the concurrent administration of finerenone with CYP3A4 inhibitors (diltiazem or fluconazole or ritonavir) could potentially lead to drug interactions, which may cause adverse events such as hyperkalemia. No studies have investigated interactions between finerenone and diltiazem or fluconazole or ritonavir. Therefore, this study aims to investigate the pharmacokinetic interaction of finerenone with diltiazem or fluconazole or ritonavir and to evaluate the impact of fluconazole on the pharmacodynamics of finerenone. The pharmacokinetic study included four rat groups (n = 8 rats/group), including a control group (finerenone alone) and test groups (finerenone pretreated with diltiazem or fluconazole or ritonavir) using both non-compartment analysis (NCA) and population pharmacokinetic (pop-PK) modeling. The pop-PK model was developed using non-linear mixed-effects modeling in NONMEM® (version 7.5.0). In the pharmacodynamic study, serum potassium (K+) levels were measured to assess the effects of fluconazole on finerenone-induced hyperkalemia. The NCA results indicated that the area under the plasma concentration-time curve (AUC) of finerenone increased by 1.86- and 1.95-fold when coadministered with fluconazole and ritonavir, respectively. In contrast, diltiazem did not affect the pharmacokinetics of finerenone. The pharmacokinetic profiles of finerenone were best described by a one-compartment disposition with first-order elimination and dual first-order absorption kinetics. The pop-PK modeling results demonstrated that the apparent clearance of finerenone decreased by 50.3% and 49.2% owing to the effects of fluconazole and ritonavir, respectively. Additionally, the slow absorption rate, which represents the absorption in the distal intestinal tract of finerenone, increased by 55.7% due to the effect of ritonavir. Simultaneously, a pharmacodynamic study revealed that finerenone in the presence of fluconazole caused a significant increase in K+ levels compared with finerenone alone. Coadministration of finerenone with fluconazole or ritonavir increased finerenone exposure in rats. Additionally, the administration of finerenone in the presence of fluconazole resulted in elevated K+ levels in rats. Further clinical studies are required to validate these findings.

Niclosamide, a potent anthelmintic agent, has emerged as a candidate against COVID‐19 in recent studies. Its formulation has been investigated extensively to address challenges related to systemic exposure. In this study, niclosamide was formulated as a long‐acting intramuscular injection to achieve systemic exposure in the lungs for combating the virus. To establish the dose–exposure relationship, a hamster model was selected, given its utility in previous COVID‐19 infection studies. Pharmacokinetic (PK) analysis was performed using NONMEM and PsN. Hamsters were administered doses of 55, 96, 128, and 240 mg/kg with each group comprising five animals. Two types of PK models were developed, linear models incorporating partition coefficients and power‐law distributed models, to characterize the relationship between drug concentrations in the plasma and lungs of the hamsters. Numerical and visual diagnostics, including basic goodness‐of‐fit and visual predictive checks, were employed to assess the models. The power‐law‐based PK model not only demonstrated superior numerical performance compared with the linear model but also exhibited better agreement in visual diagnostic evaluations. This phenomenon was attributed to the nonlinear relationship between drug concentrations in the plasma and lungs, reflecting kinetic heterogeneity. Dose optimization, based on predicting lung exposure, was conducted iteratively across different drug doses, with the minimum effective dose estimated to be ~1115 mg/kg. The development of a power‐law‐based PK model proved successful and effectively captured the nonlinearities observed in this study. This method is expected to be applicable for investigating the drug disposition of specific formulations in the lungs.

Machine learning techniques are extensively employed in drug discovery, with a significant focus on developing QSAR models that interpret the structural information of potential drugs. In this study, the pre-trained natural language processing (NLP) model, ChemBERTa, was utilized in the drug discovery process. We proposed and evaluated four core model architectures as follows: deep neural network (DNN), encoder, concatenation (concat), and pipe. The DNN model processes physicochemical properties as input, while the encoder model leverages the simplified molecular input line entry system (SMILES) along with NLP techniques. The latter two models, concat and pipe, incorporate both SMILES and physicochemical properties, operating in parallel and with sequential manners, respectively. We collected 5238 entries from DrugBank, including their physicochemical properties and absorption, distribution, metabolism, excretion, and toxicity (ADMET) features. The models’ performance was assessed by the area under the receiver operating characteristic curve (AUROC), with the DNN, encoder, concat, and pipe models achieved 62.4%, 76.0%, 74.9%, and 68.2%, respectively. In a separate test with 84 experimental microsomal stability datasets, the AUROC scores for external data were 78% for DNN, 44% for the encoder, and 50% for concat, indicating that the DNN model had superior predictive capabilities for new data. This suggests that models based on structural information may require further optimization or alternative tokenization strategies. The application of natural language processing techniques to pharmaceutical challenges has demonstrated promising results, highlighting the need for more extensive data to enhance model generalization.

Compartment modeling is a widely accepted technique in the field of pharmacokinetic analysis. However, conventional compartment modeling is performed under a homogeneity assumption that is not a naturally occurring condition. Since the assumption lacks physiological considerations, the respective modeling approach has been questioned, as novel drugs are increasingly characterized by physiological or physical features. Alternative approaches have focused on fractal kinetics, but evaluations of their application are lacking. Thus, in this study, a simulation was performed to identify desirable fractal-kinetics applications in conventional modeling. Visible changes in the profiles were then investigated. Five cases of finalized population models were collected for implementation. For model diagnosis, the objective function value (OFV), Akaike’s information criterion (AIC), and corrected Akaike’s information criterion (AICc) were used as performance metrics, and the goodness of fit (GOF), visual predictive check (VPC), and normalized prediction distribution error (NPDE) were used as visual diagnostics. In most cases, model performance was enhanced by the fractal rate, as shown in a simulation study. The necessary parameters of the fractal rate in the model varied and were successfully estimated between 0 and 1. GOF, VPC, and NPDE diagnostics show that models with the fractal rate described the data well and were robust. In the simulation study, the fractal absorption process was, therefore, chosen for testing. In the estimation study, the rate application yielded improved performance and good prediction–observation agreement in early sampling points, and did not cause a large shift in the original estimation results. Thus, the fractal rate yielded explainable parameters by setting only the heterogeneity exponent, which reflects true physiological behavior well. This approach can be expected to provide useful insights in pharmacological decision making.

Donepezil patch was developed to replace the original oral formulation. To accurately describe the pharmacokinetics of donepezil and investigate compatible doses between two formulations, a population pharmacokinetic model for oral and transdermal patches was built based on a clinical study. Plasma donepezil levels were analyzed via liquid chromatography/tandem mass spectrometry. Non-compartmental analyses were performed to derive the initial parameters for compartmental analyses. Compartmental analysis (CA) was performed with NLME software NONMEM assisted by Perl-speaks-NONMEM, and R. Model evaluation was proceeded via visual predictive checks (VPC), goodness-of-fit (GOF) plotting, and bootstrap method. The bioequivalence test was based on a 2 × 2 crossover design, and parameters of AUC and Cmax were considered. We found that a two-compartment model featuring two transit compartments accurately describes the pharmacokinetics of nine subjects administered in oral, as well as of the patch-dosed subjects. Through evaluation, the model was proven to be sufficiently accurate and suitable for further bioequivalence tests. Based on the bioequivalence test, 114 mg/101.3 cm²–146 mg/129.8 cm² of donepezil patch per week was equivalent to 10 mg PO donepezil per day. In conclusion, the pharmacokinetic model was successfully developed, and acceptable parameters were estimated. However, the size calculated by an equivalent dose of donepezil patch could be rather large. Further optimization in formulation needs to be performed to find appropriate usability in clinical situations.