Quantitative structure-pharmacokinetic relationships for the prediction of renal clearance in humans

Development of a predictive algorithm for the efficacy of half-life extension strategies

A challenge in development of peptide and protein therapeutics is rapid elimination from the body, necessitating frequent dosing that may lead to toxicities and/or poor patient compliance. To solve this issue, there has been great investment into half-life extension of rapidly eliminated drugs using approaches such as albumin binding, fusion to albumin or Fc, or conjugation to polyethylene glycol. Despite clinical successes of half-life extension products, no clear relationship has been drawn between properties of drugs and the pharmacokinetic parameters of their half-life extended analogues. In this study, non-compartmentally derived pharmacokinetic parameters (half-life, clearance, volume of distribution) were collected for 186 half-life extended drugs and their unmodified parent molecules. Statistical testing and regression analysis was performed to evaluate relationships between pharmacokinetic parameters and a matrix of variables. Multivariate linear regression models were developed for each of the three pharmacokinetic parameters and model predictions were in good agreement with observed data with r2 values for each parameter being: half-life: 0.879, clearance: 0.820, volume of distribution: 0.937. Significant predictors for each parameter included the corresponding pharmacokinetic parameter of the parent drug and descriptors related to molecular weight. This model represents a useful tool for prediction of the potential benefits of half-life extension.

The Trends and Future Prospective of In Silico Models from the Viewpoint of ADME Evaluation in Drug Discovery

Drug discovery and development are aimed at identifying new chemical molecular entities (NCEs) with desirable pharmacokinetic profiles for high therapeutic efficacy. The plasma concentrations of NCEs are a biomarker of their efficacy and are governed by pharmacokinetic processes such as absorption, distribution, metabolism, and excretion (ADME). Poor ADME properties of NCEs are a major cause of attrition in drug development. ADME screening is used to identify and optimize lead compounds in the drug discovery process. Computational models predicting ADME properties have been developed with evolving model-building technologies from a simplified relationship between ADME endpoints and physicochemical properties to machine learning, including support vector machines, random forests, and convolution neural networks. Recently, in the field of in silico ADME research, there has been a shift toward evaluating the in vivo parameters or plasma concentrations of NCEs instead of using predictive results to guide chemical structure design. Another research hotspot is the establishment of a computational prediction platform to strengthen academic drug discovery. Bioinformatics projects have produced a series of in silico ADME models using free software and open-access databases. In this review, we introduce prediction models for various ADME parameters and discuss the currently available academic drug discovery platforms.

How effective are ionization state-based QSPKR models at predicting pharmacokinetic parameters in humans?

Optimizing the pharmacokinetics (PK) of a drug candidate to support oral dosing is a key challenge in drug development. PK parameters are usually estimated from the concentration-time profile following intravenous administration; however, traditional methods are time-consuming and expensive. In recent years, quantitative structure-pharmacokinetic relationship (QSPKR), an in silico tool that aims to develop a mathematical relationship between the structure of a molecule and its PK properties, has emerged as a useful alternative to experimental testing. Due to the complex nature of the various processes involved in dictating the fate of a drug, the development of adequate QSPKR models that can be used in real-world pre-screening situations has proved challenging. Given the crucial role played by a molecule's ionization state in determining its PK properties, this work aims to build predictive QSPKR models for PK parameters in humans using an ionization state-based strategy. We divide a high-quality dataset into clusters based on ionization state at physiological pH and build global and ion subset-based 'local' models for three major PK parameters: plasma clearance (CL), steady-state volume of distribution (VDss), and half-life (t1/2). We use a robust methodology developed in our lab entitled 'EigenValue ANalySis' that accounts for the stereospecificity in drug disposition and use the support vector machine algorithm for model building. Our findings suggest that categorizing compounds in accordance with ionization state does not result in improved QSPKR models. The narrow ranges in the endpoints along with redundancies in the data adversely affect the ion subset-based QSPKR models. We suggest alternative approaches such as elimination route-based models that account for drug-transporter interactions for CL and chemotype-specific QSPKR for VDss.

In vitro dissolution considerations associated with nano drug delivery systems

Nano drug delivery systems (NDDS) offer promising solution for the translation of future nanomedicines. As bioavailability and therapeutic outcomes can be improved by altering the drug release from these NDDS, it becomes essential to thoroughly understand their drug release kinetics. Moreover, U.S. Food and Drug Administration requires critical evaluation of potential safety, efficacy, and public health impacts of nanomaterials. Spiraling up market share of NDDS has also stimulated the pharmaceutical industry to develop their cost-effective generic versions after the expiry of patent and associated exclusivity. However, unlike the conventional dosage forms, the in vivo disposition of NDDS is highly intricate and different from their in vitro behavior. Significant challenges exist in the establishment of in vitro-in vivo correlation (IVIVC) due to incomplete understanding of nanoparticles' in vivo biofate and its impact on in vitro experimental protocols. A rational design of dissolution may serve as quality and quantity control tool and help develop a meaningful IVIVC for favorable economic implications. Clinically relevant drug product specifications (critical quality attributes) can be identified by establishing a link between in vitro performance and in vivo exposure. In vitro dissolution may also play a pivotal role to understand the dissolution-mediated clearance and safety of NDDS. Prevalent in vitro dissolution methods for NDDS and their limitations are discussed in this review, among which USP 4 is gaining more interest recently. Researchers are working diligently to develop biorelevant in vitro release assays to ensure optimal therapeutic performance of generic versions of these NDDS. This article focuses on these studies and presents important considerations for the future development of clinically relevant in vitro release methods. This article is categorized under: Toxicology and Regulatory Issues in Nanomedicine > Regulatory and Policy Issues in Nanomedicine.

Predicting blood-to-plasma concentration ratios of drugs from chemical structures and volumes of distribution in humans

Abstract

Despite their importance in determining the dosing regimen of drugs in the clinic, only a few studies have investigated methods for predicting blood-to-plasma concentration ratios (Rb). This study established an Rb prediction model incorporating typical human pharmacokinetics (PK) parameters. Experimental Rb values were compiled for 289 compounds, offering reliable predictions by expanding the applicability domain. Notably, it is the largest list of Rb values reported so far. Subsequently, human PK parameters calculated from plasma drug concentrations, including the volume of distribution (Vd), clearance, mean residence time, and plasma protein binding rate, as well as 2702 kinds of molecular descriptors, were used to construct quantitative structure–PK relationship models for Rb. Among the evaluated PK parameters, logVd correlated best with Rb (correlation coefficient of 0.47). Thus, in addition to molecular descriptors selected by XGBoost, logVd was employed to construct the prediction models. Among the analyzed algorithms, artificial neural networks gave the best results. Following optimization using six molecular descriptors and logVd, the model exhibited a correlation coefficient of 0.64 and a root-mean-square error of 0.205, which were superior to those previously reported for other Rb prediction methods. Since Vd values and chemical structures are known for most medications, the Rb prediction model described herein is expected to be valuable in clinical settings.

Graphical abstract

Supplementary informations

The online version of this article (10.1007/s11030-021-10186-7) contains supplementary material, which is available to authorized users.

Antibacterial medicinal chemistry - what can we design for?

INTRODUCTION

The need for new antibacterial agents continues to grow, but success in development of antibiotics in recent years has been limited. To improve the chances that new compounds will progress into clinical trials and beyond, it is vital that we consider as early as possible in the process the various challenges that discoverers and developers of new antibiotics will face.

AREAS COVERED

The author looks at the factors that affect medicinal chemistry aimed at providing successful antibacterial agents. Target selection, target inhibition, accumulation in bacteria, and pharmacokinetics are all discussed, with a particular emphasis on how our current understanding should impact design and optimization strategies.

EXPERT OPINION

From the perspective of a medicinal chemist, the primary question when considering the various aspects of antibacterial drug discovery should be 'what can I design for?' It is important to be aware of the limitations of our understanding, and also the constraints and challenges that arise due to the diversity of the bacteria we try to address. Progress is needed to simplify approval pathways and to increase return on investment for the next generations of clinically useful agents to succeed.

Predicting tubular reabsorption with a human kidney proximal tubule tissue-on-a-chip and physiologically-based modeling

Kidney is a major route of xenobiotic excretion, but the accuracy of preclinical data for predicting in vivo clearance is limited by species differences and non-physiologic 2D culture conditions. Microphysiological systems can potentially increase predictive accuracy due to their more realistic 3D environment and incorporation of dynamic flow. We used a renal proximal tubule microphysiological device to predict renal reabsorption of five compounds: creatinine (negative control), perfluorooctanoic acid (positive control), cisplatin, gentamicin, and cadmium. We perfused compound-containing media to determine renal uptake/reabsorption, adjusted for non-specific binding. A physiologically-based parallel tube model was used to model reabsorption kinetics and make predictions of overall in vivo renal clearance. For all compounds tested, the kidney tubule chip combined with physiologically-based modeling reproduces qualitatively and quantitatively in vivo tubular reabsorption and clearance. However, because the in vitro device lacks filtration and tubular secretion components, additional information on protein binding and the importance of secretory transport is needed in order to make accurate predictions. These and other limitations, such as the presence of non-physiological compounds such as antibiotics and bovine serum albumin in media and the need to better characterize degree of expression of important transporters, highlight some of the challenges with using microphysiological devices to predict in vivo pharmacokinetics.

In Silico Prediction of Human Renal Clearance of Compounds Using Quantitative Structure-Pharmacokinetic Relationship Models

Renal clearance (CLr) plays an essential role in the elimination of drugs. In this study, 636 compounds were obtained from various sources to develop in silico models for the prediction of CLr. Stepwise multiple linear regression and random forest regression methods were employed to build global models and local models according to ionization state or net elimination pathways. The local models toward compounds undergoing different net elimination pathways showed good predictive power: the geometric mean fold error was close to 2, indicating the clearance of most compounds could be predicted within a 2-fold error range. Six classification methods were used to construct classification models. However, the performance of these classification models was less than satisfactory, and the mean accuracy of the top five models in test sets was 0.65. Moreover, qualitative analysis of physicochemical profiles between compounds undergoing different net elimination pathways revealed that compounds with higher lipophilicity tended to be reabsorbed more easily and showed lower CLr, while compounds with higher values of polar descriptors tended to secrete more easily and showed higher CLr.

Development of an in silico prediction system of human renal excretion and clearance from chemical structure information incorporating fraction unbound in plasma as a descriptor

Prediction of pharmacokinetic profiles of new chemical entities is essential in drug development to minimize the risks of potential withdrawals. The excretion of unchanged compounds by the kidney constitutes a major route in drug elimination and plays an important role in pharmacokinetics. Herein, we created in silico prediction models of the fraction of drug excreted unchanged in the urine (f_e) and renal clearance (CL_r), with datasets of 411 and 401 compounds using freely available software; notably, all models require chemical structure information alone. The binary classification model for f_e demonstrated a balanced accuracy of 0.74. The two-step prediction system for CL_r was generated using a combination of the classification model to predict excretion-type compounds and regression models to predict the CL_r value for each excretion type. The accuracies of the regression models increased upon adding a descriptor, which was the observed and predicted fraction unbound in plasma (f_u,p); 78.6% of the samples in the higher range of renal clearance fell within 2-fold error with predicted f_u,p value. Our prediction system for renal excretion is freely available to the public and can be used as a practical tool for prioritization and optimization of compound synthesis in the early stage of drug discovery.

Pharmacokinetic and Pharmacodynamic Properties of Drug Delivery Systems

The use of drug delivery systems (DDS) is an attractive approach to facilitate uptake of therapeutic agents at the desired site of action, particularly when free drug has poor pharmacokinetics/biodistribution (PK/BD) or significant off-site toxicities. Successful translation of DDS into the clinic is dependent on a thorough understanding of the in vivo behavior of the carrier, which has, for the most part, been an elusive goal. This is, at least in part, due to significant differences in the mechanisms controlling pharmacokinetics for classic drugs and DDSs. In this review, we summarize the key physiologic mechanisms controlling the in vivo behavior of DDS, compare and contrast this with classic drugs, and describe engineering strategies designed to improve DDS PK/BD. In addition, we describe quantitative approaches that could be useful for describing PK/BD of DDS, as well as critical steps between tissue uptake and pharmacologic effect.

Biotransformation and bioactivation reactions - 2018 literature highlights

In the past three decades, ADME sciences have become an integral component of the drug discovery and development process. At the same time, the field has continued to evolve, thus, requiring ADME scientists to be knowledgeable of and engage with diverse aspects of drug assessment: from pharmacology to toxicology, and from in silico modeling to in vitro models and finally in vivo models. Progress in this field requires deliberate exposure to different aspects of ADME; however, this task can seem daunting in the current age of mass information. We hope this review provides a focused and brief summary of a wide array of critical advances over the past year and explains the relevance of this research ( Table 1 ). We divided the articles into categories of (1) drug optimization, (2) metabolites and drug metabolizing enzymes, and (3) bioactivation. This annual review is the fourth of its kind (Baillie et al. 2016 ; Khojasteh et al. 2017 , 2018 ). We have followed the same format we used in previous years in terms of the selection of articles and the authoring of each section. This effort in itself also continues to evolve. I am pleased that Rietjens, Miller, and Mitra have again contributed to this annual review. We would like to welcome Namandjé N. Bumpus, James P. Driscoll, and Donglu Zhang as authors for this year's issue. We strive to maintain a balance of authors from academic and industry settings. We would be pleased to hear your opinions of our commentary, and we extend an invitation to anyone who would like to contribute to a future edition of this review. Cyrus Khojasteh, on behalf of the authors.

The Importance of Incorporating OCT2 Plasma Membrane Expression and Membrane Potential in IVIVE of Metformin Renal Secretory Clearance

Transporter expression, determined by quantitative proteomics, together with PBPK models is a promising approach for in vitro-to-in vivo extrapolation (IVIVE) of transporter-mediated drug clearance. OCT2-expressing HEK293 and MDCKII cells were used to predict in vivo renal secretory clearance (CL_r,sec) of metformin. [¹⁴C]-Metformin uptake clearance in OCT2-expressing cells was determined and scaled to in vivo CL_r,sec by using OCT2 expression in the cells versus the human kidney cortex. Through quantitative targeted proteomics, the total expression of OCT2 in HEK293, MDCKII cells, and human kidney cortex was 369.4 ± 26.8, 19 ± 1.1, and 7.6 ± 3.8 pmol/mg cellular protein, respectively. The expression of OCT2 in the plasma membrane of HEK293 and MDCKII cells, measured using an optimized biotinylation method followed by quantitative proteomics, was 30.2% and 51.6%, respectively. After correcting for percent of OCT2 expressed in the plasma membrane and the resting membrane potential (millivolts) difference between the OCT2-expressing cells and the renal epithelial cells, the predicted CL_r,sec of metformin was 250.7 ml/min, a value within the range of the observed CL_r,sec of metformin. These data demonstrate the promise of using quantitative proteomics for IVIVE of transporter-mediated drug clearance and highlight the importance of quantifying plasma membrane expression of transporters and utilizing cells that mimic the in vivo mechanism(s) of transport of drugs.

A quantitative threshold for high/low extent of urinary excretion of compounds in humans

In this study, a quantitative threshold was determined for the high/low extent of urinary excretion (UE) of compounds in humans, using a straightforward but robust statistical method known as receiver operating characteristic curve (ROC) analysis, and also 18 potential physicochemical determinants of UE were evaluated. Data on the percent of drug excreted unchanged into the urine, %Ae , were used to determine the threshold for high/low UE. Compounds can be divided into high/low UE groups using the threshold value of Ae  = 16.8%, namely those with Ae  > 16.8% are classified as high UE and those with Ae  ≤ 16.8% as low UE. The %Ae negatively correlated with cLogP (r = -0.56); however, cLogP could not quantitatively predict the value of %Ae (R(2) adj. = 0.32). Several determinants of the extent of UE, including cLogP, ACD labs cLogP and ACD labs cLogD(pH=7.4) , were successfully evaluated as a priori indicators of the extent of UE using two cut-off values for each parameter. Moreover, 87% of the 90 compounds in the external validation set were correctly classified using this approach. Analysis of the physicochemical spaces of compounds in these two groups showed significant overlap, which hinders the a priori determination of the extent of UE of compounds using a single threshold/cut-off value of simple physicochemical parameters. In conclusion, 16.8% is a quantitative threshold value to distinguish between high and low UE and new molecular entities with cLogP and ACD labs cLogP values of ≤0.7 and ≥1.0 and ACD labs cLogD(pH=7.4) values of ≤0.0 and ≥0.5 could be identified as exhibiting high and low UE, respectively. Copyright © 2016 John Wiley & Sons, Ltd.

Delineating the Role of Various Factors in Renal Disposition of Digoxin through Application of Physiologically Based Kidney Model to Renal Impairment Populations

Development of submodels of organs within physiologically-based pharmacokinetic (PBPK) principles and beyond simple perfusion limitations may be challenging because of underdeveloped in vitro-in vivo extrapolation approaches or lack of suitable clinical data for model refinement. However, advantage of such models in predicting clinical observations in divergent patient groups is now commonly acknowledged. Mechanistic understanding of altered renal secretion in renal impairment is one area that may benefit from such models, despite knowledge gaps in renal pathophysiology. In the current study, a PBPK kidney model was developed for digoxin, accounting for the roles of organic anion transporting peptide 4C1 (OATP4C1) and P-glycoprotein (P-gp) in its tubular secretion, with the aim to investigate the impact of age and renal impairment (moderate to severe) on renal drug disposition. Initial PBPK simulations based on changes in glomerular filtration rate (GFR) underestimated the observed reduction in digoxin renal excretion clearance (CL_R) in subjects with moderately impaired renal function relative to healthy. Reduction in either proximal tubule cell number or the OATP4C1 abundance in the mechanistic kidney model successfully predicted 59% decrease in digoxin CL_R, in particular when these changes were proportional to reduction in GFR. In contrast, predicted proximal tubule concentration of digoxin was only sensitive to changes in the transporter expression/ million proximal tubule cells. Based on the mechanistic modeling, reduced proximal tubule cellularity and OATP4C1 abundance, and inhibition of OATP4C1-mediated transport, are proposed as possible causes of reduced digoxin renal secretion in renally impaired patients.

Novel minimal physiologically-based model for the prediction of passive tubular reabsorption and renal excretion clearance

PURPOSE

Develop a minimal mechanistic model based on in vitro-in vivo extrapolation (IVIVE) principles to predict extent of passive tubular reabsorption. Assess the ability of the model developed to predict extent of passive tubular reabsorption (Freab) and renal excretion clearance (CLR) from in vitro permeability data and tubular physiological parameters.

METHODS

Model system parameters were informed by physiological data collated following extensive literature analysis. A database of clinical CLR was collated for 157 drugs. A subset of 45 drugs was selected for model validation; for those, Caco-2 permeability (Papp) data were measured under pH6.5-7.4 gradient conditions and used to predict Freab and subsequently CLR. An empirical calibration approach was proposed to account for the effect of inter-assay/laboratory variation in Papp on the IVIVE of Freab.

RESULTS

The 5-compartmental model accounted for regional differences in tubular surface area and flow rates and successfully predicted the extent of tubular reabsorption of 45 drugs for which filtration and reabsorption were contributing to renal excretion. Subsequently, predicted CLR was within 3-fold of the observed values for 87% of drugs in this dataset, with an overall gmfe of 1.96. Consideration of the empirical calibration method improved overall prediction of CLR (gmfe=1.73 for 34 drugs in the internal validation dataset), in particular for basic drugs and drugs with low extent of tubular reabsorption.

CONCLUSIONS

The novel 5-compartment model represents an important addition to the IVIVE toolbox for physiologically-based prediction of renal tubular reabsorption and CLR. Physiological basis of the model proposed allows its application in future mechanistic kidney models in preclinical species and human.

Key to Opening Kidney for In Vitro-In Vivo Extrapolation Entrance in Health and Disease: Part I: In Vitro Systems and Physiological Data

The programme for the 2015 AAPS Annual Meeting and Exhibition (Orlando, FL; 25-29 October 2015) included a sunrise session presenting an overview of the state-of-the-art tools for in vitro-in vivo extrapolation (IVIVE) and mechanistic prediction of renal drug disposition. These concepts are based on approaches developed for prediction of hepatic clearance, with consideration of scaling factors physiologically relevant to kidney and the unique and complex structural organisation of this organ. Physiologically relevant kidney models require a number of parameters for mechanistic description of processes, supported by quantitative information on renal physiology (system parameters) and in vitro/in silico drug-related data. This review expands upon the themes raised during the session and highlights the importance of high quality in vitro drug data generated in appropriate experimental setup and robust system-related information for successful IVIVE of renal drug disposition. The different in vitro systems available for studying renal drug metabolism and transport are summarised and recent developments involving state-of-the-art technologies highlighted. Current gaps and uncertainties associated with system parameters related to human kidney for the development of physiologically based pharmacokinetic (PBPK) model and quantitative prediction of renal drug disposition, excretion, and/or metabolism are identified.

Structure and function of multidrug and toxin extrusion proteins (MATEs) and their relevance to drug therapy and personalized medicine

Multidrug and toxin extrusion (MATE; SLC47A) proteins are membrane transporters mediating the excretion of organic cations and zwitterions into bile and urine and thereby contributing to the hepatic and renal elimination of many xenobiotics. Transported substrates include creatinine as endogenous substrate, the vitamin thiamine and a number of drug agents with in part chemically different structures such as the antidiabetic metformin, the antiviral agents acyclovir and ganciclovir as well as the antibiotics cephalexin and cephradine. This review summarizes current knowledge on the structural and molecular features of human MATE transporters including data on expression and localization in different tissues, important aspects on regulation and their functional role in drug transport. The role of genetic variation of MATE proteins for drug pharmacokinetics and drug response will be discussed with consequences for personalized medicine.

BDDCS Predictions, Self-Correcting Aspects of BDDCS Assignments, BDDCS Assignment Corrections, and Classification for more than 175 Additional Drugs

The biopharmaceutics drug disposition classification system was developed in 2005 by Wu and Benet as a tool to predict metabolizing enzyme and drug transporter effects on drug disposition. The system was modified from the biopharmaceutics classification system and classifies drugs according to their extent of metabolism and their water solubility. By 2010, Benet et al. had classified over 900 drugs. In this paper, we incorporate more than 175 drugs into the system and amend the classification of 13 drugs. We discuss current and additional applications of BDDCS, which include predicting drug-drug and endogenous substrate interactions, pharmacogenomic effects, food effects, elimination routes, central nervous system exposure, toxicity, and environmental impacts of drugs. When predictions and classes are not aligned, the system detects an error and is able to self-correct, generally indicating a problem with initial class assignment and/or measurements determining such assignments.