Correction to: Drug Absorption Parameters Obtained Using the Isolated Perfused Rat Lung Model Are Predictive of Rat In Vivo Lung Absorption.

A Correction to sthis paper has been published: https://doi.org/10.1208/s12248-020-00545-x.

Pulmonary drug absorption and systemic exposure in human: Predictions using physiologically based biopharmaceutics modeling.

Systemic exposure of inhaled drugs is used to estimate the local lung exposure and assess systemic side effects for drugs with local pharmacological targets. Predicting systemic exposure is therefore central for successful development of drugs intended to be inhaled. Currently, these predictions are based mainly on data from in vitro experiments, but the accuracy of these predictions might be improved if they were based on data with higher physiological relevance. In this study, systemic exposure was simulated by applying biopharmaceutics input parameters from isolated perfused rat lung (IPL) data to a lung model developed in MoBi® as an extension to the full physiologically-based pharmacokinetic (PBPK) model in PK-Sim®. These simulations were performed for a set of APIs with a variety of physicochemical properties and formulation types. Simulations based on rat IPL data were also compared to simulations based on in vitro data. The predictive performances of the simulations were evaluated by comparing simulated plasma concentration-time profiles to clinical observations after pulmonary administration. Simulations using IPL-based input parameters predicted systemic exposure well, with predicted AUCs within two-fold of the observed value for nine out of ten drug compounds/formulations, and predicted Cmax values within two-fold for eight out of ten drug compounds/formulations. Simulations using input parameters based on IPL data performed generally better than simulations based on in vitro input parameters. These results suggest that the developed model in combination with IPL data can be used to predict human lung absorption for compounds with different physicochemical properties and types of inhalation formulations.

IMI - Oral biopharmaceutics tools project - Evaluation of bottom-up PBPK prediction success part 4: Prediction accuracy and software comparisons with improved data and modelling strategies.

Oral drug absorption is a complex process depending on many factors, including the physicochemical properties of the drug, formulation characteristics and their interplay with gastrointestinal physiology and biology. Physiological-based pharmacokinetic (PBPK) models integrate all available information on gastro-intestinal system with drug and formulation data to predict oral drug absorption. The latter together with in vitro-in vivo extrapolation and other preclinical data on drug disposition can be used to predict plasma concentration-time profiles in silico. Despite recent successes of PBPK in many areas of drug development, an improvement in their utility for evaluating oral absorption is much needed. Current status of predictive performance, within the confinement of commonly available in vitro data on drugs and formulations alongside systems information, were tested using 3 PBPK software packages (GI-Sim (ver.4.1), Simcyp® Simulator (ver.15.0.86.0), and GastroPlus™ (ver.9.0.00xx)). This was part of the Innovative Medicines Initiative (IMI) Oral Biopharmaceutics Tools (OrBiTo) project. Fifty eight active pharmaceutical ingredients (APIs) were qualified from the OrBiTo database to be part of the investigation based on a priori set criteria on availability of minimum necessary information to allow modelling exercise. The set entailed over 200 human clinical studies with over 700 study arms. These were simulated using input parameters which had been harmonised by a panel of experts across different software packages prior to conduct of any simulation. Overall prediction performance and software packages comparison were evaluated based on performance indicators (Fold error (FE), Average fold error (AFE) and absolute average fold error (AAFE)) of pharmacokinetic (PK) parameters. On average, PK parameters (Area Under the Concentration-time curve (AUC0-tlast), Maximal concentration (Cmax), half-life (t1/2)) were predicted with AFE values between 1.11 and 1.97. Variability in FEs of these PK parameters was relatively high with AAFE values ranging from 2.08 to 2.74. Around half of the simulations were within the 2-fold error for AUC0-tlast and around 90% of the simulations were within 10-fold error for AUC0-tlast. Oral bioavailability (Foral) predictions, which were limited to 19 APIs having intravenous (i.v.) human data, showed AFE and AAFE of values 1.37 and 1.75 respectively. Across different APIs, AFE of AUC0-tlast predictions were between 0.22 and 22.76 with 70% of the APIs showing an AFE > 1. When compared across different formulations and routes of administration, AUC0-tlast for oral controlled release and i.v. administration were better predicted than that for oral immediate release formulations. Average predictive performance did not clearly differ between software packages but some APIs showed a high level of variability in predictive performance across different software packages. This variability could be related to several factors such as compound specific properties, the quality and availability of information, and errors in scaling from in vitro and preclinical in vivo data to human in vivo behaviour which will be explored further. Results were compared with previous similar exercise when the input data selection was carried by the modeller rather than a panel of experts on each in vitro test. Overall, average predictive performance was increased as reflected in smaller AAFE value of 2.8 as compared to AAFE value of 3.8 in case of previous exercise.

Drug Absorption Parameters Obtained Using the Isolated Perfused Rat Lung Model Are Predictive of Rat In Vivo Lung Absorption.

The ex vivo isolated perfused rat lung (IPL) model has been demonstrated to be a useful tool during drug development for studying pulmonary drug absorption. This study aims to investigate the potential use of IPL data to predict rat in vivo lung absorption. Absorption parameters determined from IPL data (ex vivo input parameters) in combination with intravenously determined pharmacokinetic data were used in a biopharmaceutics model to predict experimental rat in vivo plasma concentration-time profiles and lung amount after inhalation of five different inhalation compounds. The performance of simulations using ex vivo input parameters was compared with simulations using in vitro input parameters, to determine whether and to what extent predictability could be improved by using input parameters determined from the more complex ex vivo model. Simulations using ex vivo input parameters were within twofold average difference (AAFE < 2) from experimental in vivo data for all compounds except one. Furthermore, simulations using ex vivo input parameters performed significantly better than simulations using in vitro input parameters in predicting in vivo lung absorption. It could therefore be advantageous to base predictions of drug performance on IPL data rather than on in vitro data during drug development to increase mechanistic understanding of pulmonary drug absorption and to better understand how different substance properties and formulations might affect in vivo behavior of inhalation compounds.

Pulmonary Dissolution of Poorly Soluble Compounds Studied in an ex Vivo Rat Lung Model.

Many inhaled drugs are poorly water soluble, and the dissolution rate is often the rate-limiting step in the overall absorption process. To improve understanding of pulmonary drug dissolution, four poorly soluble inhalation compounds (AZD5423 (a developmental nonsteroidal glucocorticoid), budesonide, fluticasone furoate (FF), and fluticasone propionate (FP)) were administered as suspensions or dry powders to the well-established isolated perfused rat lung (IPL) model. Two particle size distributions (d50 = 1.2 µm and d50 = 2.8 µm) were investigated for AZD5423. The pulmonary absorption rates of the drugs from the suspensions and dry powders were compared with historical absorption data for solutions to improve understanding of the effects of dissolution on the overall pulmonary absorption process for poorly soluble inhaled drugs. A physiologically based biopharmaceutical in silico model was used to analyze the experimental IPL data and to estimate a dissolution parameter ( kex vivo). A similar in silico approach was applied to in vitro dissolution data from the literature to obtain an in vitro dissolution parameter ( kin vitro). When FF, FP, and the larger particles of AZD5423 were administered as suspensions, drug dissolution was the rate-limiting step in the overall absorption process. However, this was not the case for budesonide, which has the highest aqueous solubility (61 µM), and the smaller particles of AZD5423, probably because of the increased surface area available for dissolution (d50 = 1.2 µm). The estimated dissolution parameters were ranked in accordance with the solubility of the drugs, and there was good agreement between kex vivo and kin vitro. The dry powders of all the compounds were absorbed more slowly than the suspensions, indicating that wetting is an important parameter for the dissolution of dry powders. A wetting factor was introduced to the in silico model to explain the difference in absorption profiles between the suspensions and dry powders where AZD5423 had the poorest wettability followed by FP and FF. The IPL model in combination with an in silico model is a useful tool for investigating pulmonary dissolution and improving understanding of dissolution-related parameters for poorly soluble inhaled compounds.

Ranking in Vitro Dissolution of Inhaled Micronized Drug Powders including a Candidate Drug with Two Different Particle Sizes.

Pulmonary dissolution of poorly soluble drug substances (DSs) may limit the drug absorption rate and consequently influence clinical performance. Dissolution rate is thus an important quality attribute, and its influence on in vivo drug release must be characterized, understood, and controlled early in the development process. The aim of this study is to establish an in vitro dissolution method with the capability to capture therapeutically relevant differences in the dissolution rate between drug batches and drug compounds. A method was developed by which a biorelevant aerosol fraction was captured on a filter using a sedimentation technique in a modified Andersen cascade impactor to avoid particle agglomeration. Subsequently, the filters were transferred to a commercial Transwell system where dissolution in 3 mL of phosphate buffer at pH 6.8 with 0.5% sodium dodecyl sulfate (SDS) occurred at sink conditions. Dissolved DS was quantified over time using UPLC-UV. Dissolution data was obtained on a series of micronized and aerosolized lipophilic DSs, budesonide, fluticasone furoate (FF), fluticasone propionate (FP), and AZD5423. The latter is a lipophilic AstraZeneca development compound available in two different mass median diameters (MMD), 1.3 (AZD54231.3) and 3.1 µm (AZD54233.1). Dissolution data were evaluated using a Weibull fit and expressed as t63, the time to dissolution of 63% of the initial dose. The following rank-order of t63 was obtained (mean t63 and MMD in brackets), budesonide (10 min, 2.1 µm) = AZD54231.3 (10 min, 1.3 µm) < AZD54233.1 (19 min, 3.1 µm) < FP (38 min, 2.4 µm) < FF (63 min, 2.5 µm). The method could differentiate between different drug compounds with different solubility but similar particle size distribution, as well as between the same drug compound with different particle size distributions. Furthermore, a relation between the in vitro dissolution rate ( t63) and mean pulmonary absorption time in man (literature data) was observed, indicating clinical relevance. It is thus concluded, that the method may be useful for the characterization and ranking of DSs and drug products in early development, as well as being a potential tool for the control of dissolution as a potential quality attribute.

Pulmonary absorption - estimation of effective pulmonary permeability and tissue retention of ten drugs using an ex vivo rat model and computational analysis.

Permeation of inhaled drugs across the pulmonary epithelium can regulate the rate and extent of local drug absorption and hence the pulmonary tissue concentration. Therefore, understanding pulmonary epithelial transport could be important for successful design of novel inhaled medicines. To enhance understanding of pulmonary epithelial transport, drug transport data were generated for a set of inhaled compounds (n = 10) in the single-pass, isolated perfused rat lung model. A compartmental in silico model was used to estimate pulmonary permeability and tissue retention. The theoretical model was also used to re-analyze previously obtained historical drug transport data from the isolated perfused lung (n = 10) with re-circulating buffer. This was performed to evaluate the re-circulating model for assessing tissue retention measurements and to increase the number of data points. The tissue retention was an important parameter to estimate to be able to describe the drug transport profiles accurately of most of the investigated compounds. A relationship between the pulmonary permeability and the intrinsic (carrier-mediated transport inhibited) permeability of Caco-2 cell monolayers (n = 1-6) was also established. This correlation (R2 = 0.76, p < .0001) suggests that intrinsic Caco-2 permeability measurements could offer early predictions of the passive transcellular permeability of lung epithelium to candidate drugs. Although, for some compounds a deviation from the correlation suggests that other transport mechanisms may coexist. The compartmental in silico model was successful in describing the pulmonary drug transport profiles of the investigated compounds and has potential for further development to investigate the effects of formulations with different features on the pulmonary overall absorption rate.

Reply to "Comment on 'In Silico Modeling of Gastrointestinal Drug Absorption: Predictive Performance of Three Physiologically Based Absorption Models'".

This is a reply to the comment on "In Silico Modeling of Gastrointestinal Drug Absorption: Predictive Performance of Three Physiologically Based Absorption Models" by Turner and other Simcyp associates. In the reply we address the major concerns raised by Turner et al. regarding the methodology to compare the predictive performance of the different absorption models and at the same time ensure that the systemic pharmacokinetic input was exactly the same for the different models; the selection of the human effective permeability value of fexofenadine; the adoption of model default values and settings; and how supersaturation/precipitation was handled. In addition, we also further discuss aspects related to differences in in silico models and the potential implications of such differences. Our original report should be viewed as the starting point in a thorough and transparent review of absorption prediction models with the overall aim of improving their application as validated tools for bridging studies of active pharmaceutical ingredients from various sources and origins in a regulatory context. With this reply we encourage other independent investigators to perform further model evaluations of commercial as well as other existing or recently implemented models. This will boost the overall progression of physiologically based biopharmaceutical models for predicting and simulating intestinal drug absorption both in research and development and in a regulatory context.

IMI - Oral biopharmaceutics tools project - Evaluation of bottom-up PBPK prediction success part 3: Identifying gaps in system parameters by analysing In Silico performance across different compound classes.

Three Physiologically Based Pharmacokinetic software packages (GI-Sim, Simcyp® Simulator, and GastroPlus™) were evaluated as part of the Innovative Medicine Initiative Oral Biopharmaceutics Tools project (OrBiTo) during a blinded "bottom-up" anticipation of human pharmacokinetics. After data analysis of the predicted vs. measured pharmacokinetics parameters, it was found that oral bioavailability (Foral) was underpredicted for compounds with low permeability, suggesting improper estimates of intestinal surface area, colonic absorption and/or lack of intestinal transporter information. Foral was also underpredicted for acidic compounds, suggesting overestimation of impact of ionisation on permeation, lack of information on intestinal transporters, or underestimation of solubilisation of weak acids due to less than optimal intestinal model pH settings or underestimation of bile micelle contribution. Foral was overpredicted for weak bases, suggesting inadequate models for precipitation or lack of in vitro precipitation information to build informed models. Relative bioavailability was underpredicted for both high logP compounds as well as poorly water-soluble compounds, suggesting inadequate models for solubility/dissolution, underperforming bile enhancement models and/or lack of biorelevant solubility measurements. These results indicate areas for improvement in model software, modelling approaches, and generation of applicable input data. However, caution is required when interpreting the impact of drug-specific properties in this exercise, as the availability of input parameters was heterogeneous and highly variable, and the modellers generally used the data "as is" in this blinded bottom-up prediction approach.

IMI - Oral biopharmaceutics tools project - Evaluation of bottom-up PBPK prediction success part 2: An introduction to the simulation exercise and overview of results.

Orally administered drugs are subject to a number of barriers impacting bioavailability (Foral), causing challenges during drug and formulation development. Physiologically-based pharmacokinetic (PBPK) modelling can help during drug and formulation development by providing quantitative predictions through a systems approach. The performance of three available PBPK software packages (GI-Sim, Simcyp®, and GastroPlus™) were evaluated by comparing simulated and observed pharmacokinetic (PK) parameters. Since the availability of input parameters was heterogeneous and highly variable, caution is required when interpreting the results of this exercise. Additionally, this prospective simulation exercise may not be representative of prospective modelling in industry, as API information was limited to sparse details. 43 active pharmaceutical ingredients (APIs) from the OrBiTo database were selected for the exercise. Over 4000 simulation output files were generated, representing over 2550 study arm-institution-software combinations and approximately 600 human clinical study arms simulated with overlap. 84% of the simulated study arms represented administration of immediate release formulations, 11% prolonged or delayed release, and 5% intravenous (i.v.). Higher percentages of i.v. predicted area under the curve (AUC) were within two-fold of observed (52.9%) compared to per oral (p.o.) (37.2%), however, Foral and relative AUC (Frel) between p.o. formulations and solutions were generally well predicted (64.7% and 75.0%). Predictive performance declined progressing from i.v. to solution and immediate release tablet, indicating the compounding error with each layer of complexity. Overall performance was comparable to previous large-scale evaluations. A general overprediction of AUC was observed with average fold error (AFE) of 1.56 over all simulations. AFE ranged from 0.0361 to 64.0 across the 43 APIs, with 25 showing overpredictions. Discrepancies between software packages were observed for a few APIs, the largest being 606, 171, and 81.7-fold differences in AFE between SimCYP and GI-Sim, however average performance was relatively consistent across the three software platforms.

In Silico Modeling of Gastrointestinal Drug Absorption: Predictive Performance of Three Physiologically Based Absorption Models.

Gastrointestinal (GI) drug absorption is a complex process determined by formulation, physicochemical and biopharmaceutical factors, and GI physiology. Physiologically based in silico absorption models have emerged as a widely used and promising supplement to traditional in vitro assays and preclinical in vivo studies. However, there remains a lack of comparative studies between different models. The aim of this study was to explore the strengths and limitations of the in silico absorption models Simcyp 13.1, GastroPlus 8.0, and GI-Sim 4.1, with respect to their performance in predicting human intestinal drug absorption. This was achieved by adopting an a priori modeling approach and using well-defined input data for 12 drugs associated with incomplete GI absorption and related challenges in predicting the extent of absorption. This approach better mimics the real situation during formulation development where predictive in silico models would be beneficial. Plasma concentration-time profiles for 44 oral drug administrations were calculated by convolution of model-predicted absorption-time profiles and reported pharmacokinetic parameters. Model performance was evaluated by comparing the predicted plasma concentration-time profiles, Cmax, tmax, and exposure (AUC) with observations from clinical studies. The overall prediction accuracies for AUC, given as the absolute average fold error (AAFE) values, were 2.2, 1.6, and 1.3 for Simcyp, GastroPlus, and GI-Sim, respectively. The corresponding AAFE values for Cmax were 2.2, 1.6, and 1.3, respectively, and those for tmax were 1.7, 1.5, and 1.4, respectively. Simcyp was associated with underprediction of AUC and Cmax; the accuracy decreased with decreasing predicted fabs. A tendency for underprediction was also observed for GastroPlus, but there was no correlation with predicted fabs. There were no obvious trends for over- or underprediction for GI-Sim. The models performed similarly in capturing dependencies on dose and particle size. In conclusion, it was shown that all three software packages are useful to guide formulation development. However, as a consequence of the high fraction of inaccurate predictions (prediction error >2-fold) and the clear trend toward decreased accuracy with decreased predicted fabs observed with Simcyp, the results indicate that GI-Sim and GastroPlus perform better than Simcyp in predicting the intestinal absorption of the incompletely absorbed drugs when a higher degree of accuracy is needed. In addition, this study suggests that modeling and simulation research groups should perform systematic model evaluations using their own input data to maximize confidence in model performance and output.

Binding processes determine the stereoselective intestinal and hepatic extraction of verapamil in vivo.

The aim of this study was to investigate the mechanisms that might explain the observed route-dependent stereoselective pharmacokinetics (PK) of R/S-verapamil (R/S-VER) following oral and intravenous (iv) administration, by using a novel pig-specific physiologically based pharmacokinetic (PBPK) model suitable for investigations of first-pass extraction in the gut (E(G)) and the liver (E(H)). The PBPK model consisted of eight tissue compartments and was designed to simultaneously model the plasma concentration-time (PCT) profiles from three sampling sites after intrajejunal (ij) or iv administration of VER. The PBPK model successfully described the observed PCT profiles and E(H) over time for R- and S-VER. Extensive tissue binding to gut mucosa, liver, and lungs was an important determinant of the observed PK data. The stereoselective PK of VER was explained by a combination of several processes, including enantioselective plasma protein binding, blood-to-plasma partition, and gut mucosa and liver tissue distribution. The absence of stereoselectivity after iv dosing indicates that the first-pass tissue binding effect is an important factor in determining the steroselective PK of R/S-VER after oral administration. Additionally a combination of extensive liver tissue binding and a metabolite inhibition mechanism explained the time-dependent E(H) for both R- and S-VER. An in vitro-in vivo correlation of absorption needs to consider these processes because tissue binding may confound analysis of a drug's biopharmaceutical properties when using classical deconvolution or convolution techniques. In conclusion, a combination of PK data from multiple plasma sampling sites and a PBPK modeling approach provided a mechanistic understanding of processes involved in the intestinal absorption and first-pass extraction of R- and S-VER.

Extensive intestinal glucuronidation of raloxifene in vivo in pigs and impact for oral drug delivery.

In this study an advanced multisampling site pig model, with simultaneous venous blood sampling pre- and post liver, was applied to quantify the role of the intestine in relation to the liver in first-pass glucuronidation of raloxifene in vivo. The pharmacokinetic of raloxifene (a BCS/BDDCS class II compound) in humans is characterized by extensive metabolism (>90%) and the major metabolite is the 4'-ß-glucuronide (R-4-G). Following intra-jejunal (i.j.) single dose administration in pigs raloxifene was metabolized in the gut (E(G)) during first-pass to more than 70% and a high concentration (AUC(0-6 h) ratio R-4-G/raloxifene >100) of R-4-G was reached in the portal vein. The hepatic extraction (E(H)) of raloxifene was ~50% and as in humans the bioavailability become low (~7%) in pigs. Interestingly the E(H) of raloxifene and R-4-G was time-dependent after i.j. administration. It is clear that the gut was the dominating organ in first-pass extraction of raloxifene in vivo in pigs. The quantification in this study support earlier human data and emphasize that intestinal glucuronidation should be considered early in the pharmaceutical development.

Drug metabolism of CYP3A4, CYP2C9 and CYP2D6 substrates in pigs and humans.

Pigs are becoming increasingly used as a test animal both in pharmacological and toxicological assessment of new drug compounds. For interspecies comparisons and predictions it is important to characterize the expression and function of membrane transport and enzymatic proteins in pigs, particularly at a mechanistic level which will make extrapolation of observation between pig and man to be made with more confidence. The major objective of this report was to increase the integrative knowledge of drug metabolism in pigs and to compare with corresponding data from human liver microsomes. This was done by using human substrates of CYP3A4 (verapamil and testosterone), CYP2C9 (diclofenac) and CYP2D6 (dextromethorphan). In addition, the mRNA expression of important drug metabolizing enzymes and carrier-mediated transporters were assessed in intestine and liver tissues from pigs. It was shown that CYP3A4 activity is quantitatively comparable between the two species but data suggest that qualitative differences may exist. Verapamil showed similar metabolism pattern as in humans and the CYP3A4 inhibitor ketoconazole was able to inhibit the depletion of both R- and S-verapamil. A correlation between individual pig CYP3A mRNA expression and in vivo hepatic extraction ratio was established which indicates that CYP3A is the major determinant factor in both pigs and humans. However, investigations of the metabolism of testosterone resulted in qualitative different metabolite pattern between pigs and humans. The metabolism of diclofenac was very low in pig liver microsomes and did not correlate to corresponding activity in human liver microsomes. In contrast dextromethorphan exhibited a very extensive and rapid metabolism in pig liver microsomes compared to human data. Together with previously determined gene expression data it confirms that CYP2D6 substrates will be very rapidly metabolized in pigs. The mRNA data increased the knowledge of the interindividual variability and the relative expression of different enzymes and transporters in pig intestine and liver. In conclusion, this study has increased the understanding of similarities and differences between pig and human biotransformation of drugs by providing new data for four different model compounds.

Different effects of ketoconazole on the stereoselective first-pass metabolism of R/S-verapamil in the intestine and the liver: important for the mechanistic understanding of first-pass drug-drug interactions.

In this acute study a pig jejunal intestinal perfusion model with multiple plasma sampling sites and three different administration routes was used to investigate the quantitative contribution of the intestine versus liver to the first-pass extraction of each enantiomer of verapamil (VER). A subclinical dose of ketoconazole (8 mg) was coadministered in the perfusion solution to selectively inhibit gut wall CYP3A. Both enantiomers of VER and its main metabolite norverapamil (NOR) as well as the inhibitor ketoconazole were quantified in all plasma compartments by liquid chromatography-tandem mass spectrometry. The overall first-pass metabolic extraction of VER and the metabolite NOR was shown to be stereoselective with the S-isomer being more extensively extracted. For VER the ratio of R- enantiomer to S-enantiomer was greater in the hepatic vein than in the portal vein (approximately 2.2 versus 1.4), indicating that the stereoselective metabolism of VER in pigs mainly occurs on the first pass through the liver and not in the intestine. Ketoconazole increased the area under the curve from time 0 to 6 h and C(max) of R- and S-VER at least 3-fold in the portal vein, most likely explained by inhibition of gut wall metabolism. Conversely, hepatic extraction was increased because the effect of gut wall metabolism was not observed at the peripheral sampling sites. In conclusion, this study provided novel and more direct information on the contribution of the intestine and the liver, respectively, to the overall first-pass extraction of racemic VER.