Pharm Res DOI 10.1007/s11095-016-1982-5
RESEARCH PAPER
Characterization of Pharmacokinetics in the Göttingen Minipig with Reference Human Drugs: An In Vitro and In Vivo Approach Floriane Lignet 1 & Eva Sherbetijan 1 & Nicole Kratochwil 1 & Russell Jones 1 & Claudia Suenderhauf 2 & Michael B. Otteneder 1 & Thomas Singer 1 & Neil Parrott 1
Received: 4 March 2016 / Accepted: 21 June 2016 # Springer Science+Business Media New York 2016
ABSTRACT Purpose This study aims to expand our understanding of the mechanisms of drug absorption, distribution, metabolism and excretion in the Göttingen minipig to aid a knowledge-driven selection of the optimal species for preclinical pharmaceutical research. Methods The pharmacokinetics of seven reference compounds (antipyrine, atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline) was investigated after intravenous and oral dosing in minipigs. Supportive in vitro data were generated on hepatocellularity, metabolic clearance in hepatocytes, blood cell and plasma protein binding and metabolism routes. Results Systemic plasma clearance for the seven drugs ranged from low (1.1 ml/min/kg, theophylline) to close to liver blood flow (37.4 ml/min/kg, cimetidine). Volume of distribution in minipigs ranged from 0.7 L/kg for antipyrine to 3.2 L/kg for hydrochlorothiazide. A gender-related difference of in vivo metabolic clearance was observed for antipyrine. The hepatocellularity for minipig was determined as 124 Mcells/ g liver, similar to the values reported for human. Based on these data a preliminary in vitro to in vivo correlation (IVIVC) for metabolic clearance measured in hepatocytes was investigated. Metabolite profiles of diazepam and midazolam compared well between minipig and human. Electronic supplementary material The online version of this article (doi:10.1007/s11095-016-1982-5) contains supplementary material, which is available to authorized users. * Floriane Lignet
[email protected]
1
Pharmaceutical Sciences, Roche Pharmaceutical Research and Early Development, F. Hoffmann – La Roche, Ltd, Basel, Switzerland
2
Klinische Pharmakologie & Toxikologie, Universitätsspital Basel, Basel, Switzerland
Conclusions The results of the present study support the use of in vitro metabolism data for the evaluation of minipig in preclinical research and safety testing.
KEY WORDS in silico . metabolism . minipig . pharmacokinetics
ABBREVIATIONS ADME AUC CYP FCS GFR IVIVC NCA PBPK PK
Absorption, distribution, metabolism and excretion Area under the curve Cytochrome P450 Fetal calf serum Glomerular filtration rate In vitro-in vivo correlation Non-compartmental analysis Physiologically-based pharmacokinetic Pharmacokinetic
INTRODUCTION Evaluation of the safety of new pharmaceuticals relies on a combination of in vitro, in silico and in vivo animal data to provide an assessment of the risk to human health (1). Accordingly, drug regulatory authorities require the evaluation of new compounds in rodent and non-rodent species before the conduct of clinical trials in order to ensure safety in man (2). Dogs and non-human primates are the most commonly used non-rodent species due to a longer tradition and more extensive knowledge, but their use is restricted by certain physiological differences as well as societal constraints (3). In this context, interest in the minipig as an alternative preclinical species is increasing (4). In Europe, the most commonly used breed was developed in the 1960s by the Göttingen University, Germany, by crossbreeding Minnesota minipig, Vietnamese potbelly pig and German landrace pig.
Lignet et al.
Göttingen minipigs are now bred under controlled pathogenfree conditions, and particular care is taken to minimize inbreeding and genetic drift by management of the animal gene pool (5). The Göttingen minipig strain presents similarities to human in terms of anatomy, physiology and biochemistry and its small size facilitates handling, making it a practical animal for biomedical research (6). Relative to the body weight, the gastro-intestinal tract of minipig and human have comparable size (7). The pig stomach is mono-gastric as in humans but exhibits some differences, for example the diverticulum ventriculi, in which formulations given by gavage could be mis-dosed (8). Due to a gastric emptying time that is longer and more variable than in human, and a constant residue of food in the stomach making a fasted state difficult to attain, minipig seems not to be a good model for food effect studies (9,10). However, the species has proven useful in dermatology, toxicology and cardiovascular safety studies (11–16). In a recent review, Dalgaard collected data on drug metabolism and distribution in minipig (17), summarizing the similarities and discrepancies observed in metabolic enzyme and transporter activity compared to human. A major outcome of this study was that the minipig should be the preferred animal model when considering compounds metabolized by aldehyde oxidase, N-acetyl-transferases, or CYP2C9 enzymes. Studies that focused more precisely on cytochrome P450 mediated metabolism in minipig showed that CYP1A, CYP2A and CYP3A catalyzed reactions are very close to those in humans (18–20). The minipig has also been shown to be able to catalyze conjugation with glucuronic acid (21) and there is evidence that glucuronosyltransferase reactions in pigs and minipigs are elevated compared to human (22,23). In addition, minipig has been shown to have a high acetylating capacity (24,25) but a decreased level of sulphate conjugation compared to other species (26). Ontogeny of the protein expression and activity of CYP3A enzymes in minipig liver and intestine was found to be comparable to human, supporting the use of the minipig in juvenile studies (27,28). These results were confirmed at the transcriptome level, as RNA expression of metabolic enzymes in liver was comparable between minipig and human (29) with expression levels of most CYPs similar to human, while uridine 5′-diphospho-glucuronosyltransferases (UGT), Flavin containing monooxygenase 1 (FMO1) and aldoketoreductases (AKR/CR) levels were higher in minipig. Additionally, the advancement of transcriptional programs at equivalent age categories is analogous in human and minipig (29). Overall there are many good reasons to consider the minipig as a species for pre-clinical research. However knowledge gaps still exist and due to limited published examples on the use of minipig there can be a reluctance to deviate from the more traditional pre-clinical species. To fill these gaps and to further evaluate predictivity of minipig for human,
additional data on the absorption, distribution, metabolism and elimination (ADME) processes for well characterized drugs are needed. Such data will help to optimize the use of minipig in pre-clinical research and will also enable the development and validation of mathematical models (8,9). In the present study, we investigated similarities and differences in the ADME characteristics of reference drugs in minipig, human, rat and monkey and looked into the underlying processes in vivo and in vitro, to provide further understanding of minipig as a toxicological and pharmacokinetic model. We considered seven well studied small molecule drugs for which data on pharmacokinetics in other laboratory species and in human were available in the literature: namely antipyrine, atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline. These compounds were dosed intravenously and orally to male and female minipigs and their plasma pharmacokinetics (PK) were followed for at least 24 h. The resulting concentration-time profiles were analyzed using noncompartmental methods to derive PK parameters which could be compared to the literature-derived values in human and other laboratory species. Furthermore, in order to better understand the observed pharmacokinetics, blood-plasma partitioning, plasma protein binding and intrinsic clearances in minipig hepatocytes were measured in vitro. In this way in vitro hepatic intrinsic clearances and in vivo clearances could be compared and we established an in vitro—in vivo correlation (IVIVC). Additionally, to support the scaling of in vitro clearances in minipig, the liver hepatocellularity was measured. Finally, metabolite profiles for diazepam and midazolam were investigated in hepatocyte incubations and in minipig plasma to allow a qualitative comparison of the metabolism in human and minipig.
MATERIALS AND METHODS Selected Drugs Antipyrine, atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline were purchased from commercial suppliers. These drugs were selected based on their absorption, metabolism and elimination routes in humans. Thus, antipyrine is rapidly and completely absorbed through passive transcellular diffusion (30) and is a marker of general cytochrome activity (31). Around 50% of atenolol is absorbed by paracellular diffusion (32) and it is mainly eliminated by renal filtration, with 80 to 90% of the compound excreted unchanged, indicating a limited metabolism (33). Cimetidine permeates by paracellular diffusion, has a bioavailability of around 50% and is mainly cleared by the kidneys (32,34,35). Diazepam is very well absorbed and is mainly metabolized by
Characterization of the Göttingen Minipig Pharmacokinetics
CYP3A4 and CYP1A2 (36). Hydrochlorothiazide is more than 90% excreted unchanged in urine (34,37). It permeates by paracellular diffusion and its bioavailability after oral dosing is around 60% (30,37,38). Midazolam is metabolized by CYP3A (36) and has a bioavailability of 30 to 50% which is limited by first pass metabolism (39,40). Theophylline is completely absorbed and is a substrate of CYP1A2 (36,41). While these compounds were selected based on the reported behavior in human it is recognized that differences will exist in the minipig, for example other CYP isoforms may play a role in minipig metabolism. These drugs also cover a wide range of chemical properties. Table I presents an overview of drug properties based on human relevant data reported in the literature and from in house measurements.
In Vitro Studies Protein Binding Pooled and frozen plasma from minipig were obtained from commercial suppliers (BioreclamationIVT, USA, Product N u m b e r M I NI P I G P L E D T A 3 -G O T , l ot N u m b e r MPG13742). The Teflon equilibrium dialysis plate (96-well, 150microL, half-cell capacity) and cellulose membranes (12– 14 kDa molecular weight cutoff) were purchased from HTDialysis (Gales Ferry, Connecticut). Both biological matrix and phosphate buffer pH were adjusted to 7.4 on the day of the experiment. DMSO stock solutions for the compounds were prepared. The determination of unbound compound was performed using a 96-well format equilibrium dialysis device with a molecular weight cutoff membrane of 12 to 14 kDa. The dialysis membranes were conditioned as recommended by the supplier. Equal volumes of matrix samples containing substances and blank dialysis buffer (Soerensen buffer at pH 7.4) were loaded into the opposite compartments of each well. The dialysis block was sealed and kept for 5 hours at a temperature of 37°C and 5% CO2 environment in an
Table I
Intrinsic Clearance in Hepatocytes Cryopreserved minipig hepatocytes (BioreclamationIVT, USA, Product Number M00615, lot Number XNG) were used. For the suspension cultures, Nunc U96 PP-0.5 ml (Nunc Natural, 267245) plates were used, which were incubated in a Thermo Forma incubator from Fischer Scientific (Wohlen, Switzerland) equipped with shakers from Variomag® Teleshake shakers (Sterico, Wangen, Switzerland) for maintaining cell dispersion. The cell culture medium was William’s medium supplemented with 10% FCS. Incubations of test compound at 1 microM test concentration in suspension cultures of 1 Mcells/ml were performed in 96 well plates and shaken at 900 rpm for up to 2 hours in a 5% CO2 atmosphere and 37°C. After, 3, 6, 10, 20, 40, 60 and 120 minutes, 100 μL cell suspension in each well was quenched with 200 μL methanol containing an internal standard. Samples were then cooled and centrifuged before analysis by LC-MS/MS. Log peak area ratios (test compound peak area / internal standard peak area) or concentrations were plotted against incubation time and a linear fit made to the data with emphasis upon the initial rate of compound disappearance. The slope of the fit was then used to calculate the intrinsic clearance according to: Clint (μL/min/1x106 cells) = -slope (min-1) * 1000 / [1x106 cells].
In Vitro Properties of the Reference Compounds in Human
Measured (logP) or logD at pH 7.4 pKab Fraction unbound in plasma (%) Blood – plasma partitioning (Rb) BCS class Clearance route a
incubator. The seal was then removed and matrix and buffer from each dialysis was prepared for analysis by LC-MS/MS. All protein binding determinations were performed in triplicates. The percent unbound fraction (fu) was calculated by determining the compound concentrations in the buffer and matrix compartments after dialysis according to: %fu = 100 * buffer concentration after dialysis / matrix concentration after dialysis
Antipyrine
Atenolol
Cimetidine
Diazepam
Hydrochlorothiazide Midazolam
Theophylline
(0.38)a 1.5 (B) 93e 0.88 f Ic Several CYPs
(0.16)1 9.6 (B) 94e 1.07 g IIIc Renal elimination
0.232 7.11(B) 78e 0.97 h IIIc Renal elimination
2.712 3.6(B) 2,3e 0.58 h Id CYP3A4, CYP1A2
(–0.07)a 8.6(A), 9.9(A) 321 1.7 h; 2.7 i IIIc Renal elimination
(–0.02)a 8.7 (A) 61e 0.82 k Id CYP1A2
3.28b 5.8(B) 1,7e 0.55 j; 0.8 h Id CYP3As
value taken from DrugBank (www.drugbank.ca); b in house measurement; c Lennernas, 2007 (42); d Wu, 2005 (43); e Obach, 2008 (44); f estimated with SimulationPlus ADMET Predictor; g Taylor, 1981 (45); h Paixão, 2009 (46); i Akabane, 2010 (47); j Gertz, 2011 (48), k Mitenko, 1973 (49)
Lignet et al.
Bioanalytics of Intrinsic Clearance and Protein Binding Samples LC-MS/MS was used for the quantification of antipyrine, atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline. The HPLC system consisted of Shimadzu pumps. The analytical column was operated at 60°C. A 5500 Q TRAP AB Sciex mass spectrometer equipped with a TurboIonSpray source (IonSpray Voltage 5500 V in positive mode or -4500 V in negative mode for hydrochlorthiazide) and a HTS CTC PAL autosampler were used. 1 μL aliquots of the centrifuged sample solutions were injected and transferred at the analytical column at a flow rate of 600 μL/min. To elute the compounds, a high pressure linear gradient from 0 to 95% B in 40 seconds was applied. Total run time was 1.7 min. Data analysis was performed using peak area ratio between analyte and internal standard. Analyst 1.6.1 software was used for analytical data processing.
Hepatocellularity Commercially available, cryopreserved minipig hepatocytes (BioreclamationIVT, USA, Product Number M00615, lot Number XNG) were used. Hepatocyte suspension cultures were prepared by thawing and centrifugation (10 minutes at 100 g, Sorvall Legend X1R, ThermoScientific) in thawing media (Celsis IVT, Baltimore) The hepatocyte suspension cultures of either 2, 1 or 0.3 Mcells/mL were then washed twice with PBS. After centrifugation (5 min, at 100 g, Sorvall Legend X1R, ThermoScientific), the supernatant was discarded and the cell pellet kept on ice prior further analysis. Liver samples from six Göttingen minipigs (3 males, 3 females, 6 months old, purchased from Ellegaard A/S (DK)) were used. Three samples (50 mg) of each animal liver were put in 7 ml homogenization tubes (Precellys®24, Bertin Technologies) with 1 ml of AT extraction for DNA determination or 1 mL 1% Triton-X-100 (SURFACT-AMPS X-100) in water for protein content determination. Samples underwent three cycles of 20 seconds homogenization (Precellys®24-Dual, Bertin Technologies). Protein content of hepatocyte lysate and liver homogenate was determined with a BCA protein assay kit (Pierce Biotechnology, Rockford). To the hepatocyte cell pellets and liver homogenates, 1% Triton-X-100 (SURFACT-AMPS X-100) in water was added. After mixing, the cell lysates were kept for 1 h at 4°C and vortexed from time to time before they were stored overnight at -20°C. The samples were then thawed, centrifuged (2500 g, 30 min, 4°C) and the supernatant was used for the protein content determination. The protein content was measured by the Pierce BCA protein assay kit following the standard protocol. Absorbance was read at 562 nm (Multiskan GO, ThermoScientific).
The protocol for the DNA assay was adapted from Downs (50) and the Sigma fluorescence assay kit (DNA quantitation kit, Sigma, St-Louis) as follows. Liver homogenates were diluted 10 fold with the Sigma fluorescence assay buffer. An aliquot of 50 μl was then diluted in 450 μl fluorescent buffer. After centrifugation (2500 g, 30 min, 4°C), the supernatant was collected and kept on ice prior to analysis. For hepatocyte suspension cultures, 50 μl AT extraction buffer (1 M NH4OH, 0.2% Triton-X) was added to 2∙millions of cells. After 10 minutes incubation at 37°C, lysate was diluted in 1.95 ml of fluorescent assay buffer (DNA quantitation kit, Sigma, St-Louis), centrifuged (2500 g, 30 min, 4°C) and the supernatant was kept on ice prior to analysis. Standard curve was obtained from calf thymus DNA (250–10 μg/ml) diluted into the fluorescence assay buffer provided with the Sigma assay DNA kit. Hoescht dye solution of 2 μg/ml was used. Finally, 4 μl of either standard or sample were loaded in 200 μl of Hoechst dye solution on a 96-well OptiPlate (PerkinElmer). Fluorescence was then read at 355 nm excitation and 460 nm emission (FLUOstar Galaxy, BMG Labtech). Hepatocellularity or hepatocytes per gram liver (HPGL) was calculated from the protein or the DNA content in hepatocytes and liver tissue, based on liver-to-cell content ratio. The HPGL value derived from the protein content was corrected by a factor of 0.98, as hepatocytes contain 97% of the total protein in liver tissue and 99% of protein in an isolated cell suspension (51,52). The HPGL value derived from the DNA content was corrected for the percentage of binucleated hepatocytes, 7% for the minipig (51,53). Furthermore, a correction factor of 0.75 was applied taking into account the difference in DNA proportion coming from hepatocytes between the whole liver and the isolated hepatocytes suspension. Blood to Plasma Partitioning K2-EDTA blood was collected from a healthy adult minipig and pre-incubated for 30 min at 37°C in an incubator using an IKA Mixer. For each compound, a 0.5 mg/ml stock solution was prepared in DMSO. Each blood plasma partitioning study was conducted in triplicates. For each reference compound, 1 μl of stock solution was added to 1 ml pre-incubated blood that was transferred into a 2 ml Eppendorf tube (low protein binding tube), resulting in a final concentration of 500 ng/ml. Blood samples were mixed by gently turning the tubes five times, to avoid destruction of the erythrocytes, and incubated at 37°C in an incubator using an IKA mixer. After 30 min, 50 μl whole blood samples were drawn, 50 μl of water were added to lyse the erythrocytes and 50 μl of plasma were then added to have a matrix match for all samples. The remaining blood was further centrifuged (3000x g for 3 min) to collect 50 μl plasma samples. Fifty microliter of fresh blood and 50 μl of water were added to have matrix match for all
Characterization of the Göttingen Minipig Pharmacokinetics
samples. Drug concentrations in all samples were quantified by LC-MS. Standards and QC samples were measured together with unknown samples. For standards and QC, whole blood was spiked with water (1 part blood / 1part water) and this mixture was frozen on dry ice and thawed to room temperature 3 times. Final concentrations of the standards and QC samples were prepared in blood/water-mix from 0.125 ng/ml to 2500 ng/ml. For each compound, blood-plasma partitioning ratio was estimated comparing the mean concentration of drug in the blood samples to the mean concentration of drug in the plasma samples.
Metabolite Profiles Incubations with cryopreserved minipig and human hepatocytes were performed as described for intrinsic clearance determinations. Minipig plasma samples from 3 male animals were pooled across animals at selected time-points post dosing (1 mg/kg i.v. dose). Plasma time-point samples were precipitated by addition of an equal volume of cold acetonitrile (1:1, v/v) containing an internal standard, vortex mixed, cooled and centrifuged (13,000 x g for 20 min, 6°C). The resulting supernatant was diluted 1:1 (v/v) with water and analyzed by LC-MS. Metabolite formation was determined by LC-MS in the positive ion mode by using a Waters Synapt G2 HDMS QTof (Waters, Manchester, UK) coupled to a Waters Acquity UPLC system equipped with a CTC autosampler (Waters, Manchester, UK). Liquid chromatography separation of the metabolites was performed by using a linear gradient of acetonitrile optimized for each compound. A typical LC method is described as follows: Separation was achieved using a Waters acquity BEH C18 column (Waters, Manchester, UK) with the dimension 100 mm x 2.1 mm I.D., particle size 1.7 μm at a flow rate of 400 μl/min. The injection volume was set to 10 μl, and the column oven temperature was 30°C. The mobile phase consisted of water with 0.1% (v/v) formic acid (mobile phase A) and acetonitrile with 0.1% (v/v) formic acid (mobile phase B). The following gradient was applied: 0.5 to 10 min linear from 5 to 60% B, from 10 min to 10.5 min 60 to 95% B, from 10.5 to 11.5 min isocratic 95% B and 11.5 min to 15 min re-equilibration at 5% B. The SYNAPT G2 HDMS was operated at a resolution of approximately 18,000 full width half maximum at m/z 400 in the full scan MS mode. High-resolution MSE mass spectra were acquired in the range of m/z 100 to m/z 1200. Exact mass measurement was performed after external calibration by using a sodium formate solution according to the manufacturer’s methodology prior to LC/MS investigations. Data was acquired using a lock spray of leucine enkephalin (m/z 556.2771) infused at 10 μl/min.
Metabolites were identified by comparison with control samples. Metabolite identification was performed by comparison of product ion spectra of reference compounds with that of found metabolites and/or by interpretation of product ion spectra of found metabolites. The elemental composition of the product ions was matched with that of the proposed fragment structure. The semi-quantitative abundance of metabolites was assessed on the basis of peak areas of extracted ion chromatograms with a mass tolerance of 5 p.p.m.
In Vivo Studies In vivo studies were conducted in adult male and female Göttingen minipigs of at least 8 months old, purchased from Ellegaard A/S (DK). The animals were allowed to acclimatize for several weeks before any experiment. Under maintenance conditions, minipigs were fed twice daily with standard minipig pellet diet and given access to tap water ad libitum. Straw was given daily for bedding and environment enrichment. All animal experiments were conducted in accordance with the international ethical guidelines for the care and use of laboratory animals and were approved by the local veterinary authorities. Animals were weighed before each experiment and the dose adjusted accordingly. Weights ranged between 9.6 and 23.45 kg for the females and between 8.8 and 17.45 kg for the males. Intravenous dosing was performed by bolus injection in the ear vein and oral dosing by gavage. After i.v. dosing, 1 ml of venous blood was withdrawn at 0.08, 0.25, 0.5, 1, 2, 3 (or 4), 7 and 24 h. For the oral experiments, 1 ml of venous blood was sampled before gavage, and at 0.08, 0.25, 0.5, 1, 2, 3 (or 4), 7, 24, 32, and 48 h after dosing. Urine was collected during the first 24 h after intravenous dosing of all compounds but midazolam. At least one week wash out was allowed between experiments. All intravenous dosing were performed on 3 male and 3 female minipigs, except for hydrochlorothiazide where 4 animals of each gender were used and for antipyrine, where 16 profiles per gender were generated using 4 males and 6 females. All oral experiments were performed on 4 male and 4 female minipigs. The different numbers of animals used for different compounds was driven by practical considerations and availability of animals. For antipyrine a sex-related difference in PK was seen in the first studies and so this compound was administered again to confirm a statistically significant difference. The doses given were the following: antipyrine, 10 mg/kg i.v. and 20 mg/kg p.o.; atenolol 1 mk/kg and 3 mg/kg p.o.; cimetidine, 1 mg/kg i.v. and 20 mg/kg p.o.; diazepam, 1 mg/kg and 1 mg/kg p.o.; hydrochlorothiazide, 4 mg/kg i.v. and 4 mg/kg p.o.; midazolam, 1 mg/kg i.v. and 1 mg/kg p.o.; theophylline, 1 mg/kg i.v. and 3 mg/kg p.o.
Lignet et al.
Bio-Analytics The calibration standards were prepared in blank plasma. The concentration range was 0.1 to 2500 ng/ml. The QC samples were prepared in blank plasma at 1, 50, and 1000 ng/ml. Plasma samples (50 μL) were treated with 200 μL of Methanol containing internal standard. After centrifugation, 50 μL of the supernatant was diluted with 200 μL of water, an aliquot (1 μL) was injected into an HPLC system (Shimadzu, 2 pump LC30AD / autosampler SIL-20 AC + rack changer / Degasser DGU-20A5). The analyte and the internal standard were separated from matrix interferences using gradient elution from 10 mM ammonium acetate in water /acetic acid (1000:0.5, v/v) to methanol/acetic acid (1000:0.5, v/v). The analytical column was a Phenomenex Luna C18 column (50 x 1 mm, 5 μm) with column heating at nominal 60°C. Mass spectrometric detection was carried out on a Sciex API5500 mass spectrometer using selected reaction monitoring (SRM) in the positive ion mode.
the urine collected in the first 24 h after intravenous dosing. The renal clearance Clr was estimated as Clr = Clsys . Fe. Several values of the glomerular filtration rate (GFR) in minipig can be found in the literature: Suenderhauf et al. reported a GFR ranging between 1.7 and 2.5 ml.min−1.kg−1 (8), Cibulskyte et al. noted a GFR around 2.3 ml.min−1.kg−1 (48 ml.min−1 in animals weighing 20.2 kg on average) (54), while a value of 5.5 ml.min−1.kg−1 is reported in (55). We used a mean value of 3 ml. min−1.kg−1 for comparison to our measured renal clearances.
Comparison to Other Species To allow a comparison of the PK parameters in minipig to other animal models and human, PK characteristics of the reference compounds in rat, dog, monkey and human were extracted from the scientific literature. For each compound, an arithmetic mean of the reported values was calculated.
Data Analysis: non-Compartmental Analysis
Establishment of an IVIVC
Plasma concentration time profiles were analyzed by non-compartmental analysis (NCA) using the software Pheonix WinNonLin (version 6.4, Pharsight Corporation, Mountain View, CA, USA). The maximum observed plasma concentration (Cmax) and the time to reach it (Tmax) were obtained directly from the profiles, while terminal half-life (t1/2), systemic clearance (CLsys) and volume of distribution at steady state (Vss) were estimated by non-compartmental analysis with computation of the areas under the curve (AUC) by linear trapezoidal interpolation. Calculation details are described in the Phoenix WinNonlin 6.4 user’s manual. Gender-related pharmacokinetic differences were investigated by comparing male and female parameter values using a non-parametric Mann–Whitney test. For each compound, the bioavailability (F) was computed as
Intrinsic liver clearances (Clint,h,pred) were scaled from the measured hepatocyte intrinsic clearances (Clint) using the number of hepatocytes per gram of liver (HPGL), an average liver weight (LW) and an average body weight (BW) as follows : . Cl int;h;pred ¼ Cl int : HPGL :LW BW ð2Þ
F ¼
AU C po : D iv AU C iv : D po
ð1Þ
Where AUCiv and AUCpo are respectively the AUC for intravenous and oral dosing, and Div and Dpo are the intravenous and oral doses respectively. Cross-over data, allowing computation of individual bioavailability estimates, were only available for hydrochlorothiazide and antipyrine. For the other compounds, an average bioavailability was calculated using the mean of AUCiv and AUCpo across all animals. For all compounds but midazolam, the fraction of drug excreted by the kidneys, Fe, was estimated from the concentration of drug in
From in vivo data, total hepatic clearance Clh was computed for each compound assuming that the only two routes of elimination were hepatic metabolism and renal clearance. Therefore, Cl h ¼ Cl sys − Cl r Where Clsys is the systemic clearance estimated by NCA of the intravenous plasma PK data, and Clr is the renal clearance calculated from the fraction of the dose excreted in the urine. The intrinsic hepatic drug clearance (Clint,h,mes) was estimated considering the liver blood flow (Qh), based on the wellstirred model (56,57) as follows: Cl h : Q h Cl int;h;mes ¼ Q h −Cl h
ð3Þ
Correlation between intrinsic hepatic clearance estimated from in vitro data and intrinsic hepatic clearance calculated from in vivo measurements was then investigated by linear regression. Calculation of the IVIVC accounting for the measured fu and Rb was also considered and details are given in the supplementary material II.
Characterization of the Göttingen Minipig Pharmacokinetics
RESULTS In Vivo Pharmacokinetics Mean plasma concentration-time profiles of antipyrine, atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline after intravenous and oral dosing in male and female minipigs are presented in Figs. 1 and 2. For most of the compounds, the lower limit of quantification of the plasma concentration was reached before the end of the 48 h sampling period. Therefore, concentration-time profiles are shown only over the times for which concentrations were detectable. For each drug, pharmacokinetic parameters estimated by non-compartmental analysis are presented in Table II. Comparison of male and female parameter values, performed using a Mann–Whitney test, showed a significant difference in
Fig. 1 Mean ± standard deviation plasma concentrations of atenolol, cimetidine, diazepam, hydrochlorothiazide, midazolam and theophylline dosed intravenously (hollow circles) and orally (filled squares) in male (continuous lines) and female (dashed lines) Göttingen minipigs. 4 males and 4 females were used for all oral dosing and for hydrochlorothiazide intravenous dosing. The other intravenous studies were performed on 3 male and 3 female minipigs. The doses used were the following: atenolol 1 mg/kg and 3 mg/kg p.o.; cimetidine, 1 mg/kg i.v. and 20 mg/kg p.o.; diazepam, 1 mg/kg and 1 mg/kg p.o.; hydrochlorothiazide, 4 mg/kg i.v. and 4 mg/kg p.o.; midazolam, 1 mg/kg i.v. and 1 mg/kg p.o.; theophylline, 1 mg/kg i.v. and 3 mg/kg p.o.
clearance and terminal half-life only for antipyrine (p < 0.05). For all other compounds no gender-related differences could be demonstrated and so the mean was taken across both sexes (the mean estimates of PK parameters in males and females separately can be found in the supplementary material I). For all compounds, a large inter-individual variability was observed in the intravenous and oral plasma concentration versus time profiles. In particular, antipyrine showed a very large coefficient of variation (CV) for clearance, at 58% in males and 24% in females, while estimates of clearance of diazepam were much less variable with a CV at 11% (data in supplementary material I). Plasma systemic clearances ranged widely from 1.1 ml.min−1.kg−1 for theophylline to 37.4 ml.min−1.kg−1 for cimetidine while volume of distribution was more similar across drugs ranging from 0.7 to 1.5 L/kg for all drugs except for hydrochlorothiazide which showed an increased
Lignet et al. Fig. 2 Mean ± standard deviation of plasma concentrations of antipyrine after intravenous (left) and oral dosing (right), in male (hollow shapes and continuous lines) and female (filled shapes and dashed lines) minipigs. Intravenous dosing was performed on 16 animals of each gender and oral on 4 animals in each gender.
distribution into tissues with a Vss of 3.2 L/kg. Terminal halflives ranged from 0.8 h for cimetidine to 11.9 h for theophylline. Midazolam showed the lowest bioavailability of ~10%, while diazepam bioavailability was greater than 100% (this high value may be partly explained by the fact that the AUCinf for the oral profile was derived by extrapolation and ~25% of the final AUCinf was extrapolated). For comparison to the minipig pharmacokinetic parameter values, representative values for human, monkey, dog and rat were collected from the literature and are shown in Fig. 3 with full details provided in supplementary material I. For all compounds but hydrochlorothiazide, the values estimated from the intravenous dosing in minipig are within the range of values observed in other species. An inter-species comparison of hydrochlorothiazide could not be performed due to lack of available data in species other than the minipig. The collection of urine allowed evaluation of the fraction of dose excreted by the kidneys and thus the renal clearances could be estimated, as reported in Table III. Renal clearances ranged from 0.01 ml.min − 1 .kg − 1 for diazepam to 7.5 ml.min−1.kg−1 for hydrochlorothiazide. For the three compounds that are mainly (>80%) eliminated in the urine in human i.e. atenolol, cimetidine and hydrochlorothiazide, the fractions of the dose excreted in minipig urine were lower
Table II Pharmacokinetic Parameters in Minipig After Intravenous and Oral Dosing of Reference Compounds Estimated by Non-Compartmental Analysis of the Concentration vs Time Profiles, Presented as Mean ± Standard Deviation
at 37, 21 and 39% respectively. To allow assessment of the role of active secretion in renal elimination, Table III compares the renal clearances to estimated glomerular filtration of unbound drug. Thus atenolol appears to be cleared by renal filtration while both cimetidine and hydrochlorothiazide have renal clearances much greater than would be expected for filtration at the glomerulus and therefore seem to be actively secreted. Metabolic Profile of P450 Substrates Midazolam and Diazepam Comparison of the in Vitro and in Vivo Metabolite Profiles in Human and Minipig Midazolam and diazepam were selected for further investigation as they are classical examples of P450 and phase II metabolism which show high in vitro metabolic turnover in minipig. Midazolam. The in vitro metabolism of midazolam was investigated and compared between human and minipig hepatocytes. In human hepatocytes, 1-hydroxymidazolam was the predominant metabolite together with its corresponding
Intravenous dosing
Antipyrine in males Antipyrine in females Atenolol Cimetidine Diazepam Hydrochlorothiazide Midazolam Theophylline
Oral dosing
CL (ml/min/kg)
T1/2 (h)
Vss (L/kg)
Tmax (h)
Cmax/Dose (ng/ml)
F (%)
1.5 ± 0.9 4.8 ± 1.1 7.9 ± 2.9 37.4 ± 8.2 9.6 ± 1.1 11.8 ± 1.6 22.3 ± 8.6 1.1 ± 0.4
9.3 ± 5.8 1.7 ± 0.4 4.4 ± 1.5 0.8 ± 0.1 1.7 ± 0.5 5.2 ± 1.2 1.0 ± 0.4 11.9 ± 4.5
0.8 ± 0.3 0.7 ± 0.2 1.5 ± 0.1 1.5 ± 0.2 0.8 ± 0.3 3.2 ± 0.8 1.3 ± 0.5 0.9 ± 0.1
2.4 ± 1.3 0.7 ± 0.4 1.6 ± 0.5 1.7 ± 1 0.9 ± 0.5 2.3 ± 1.5 1.8 ± 1.2 2.6 ± 1.2
550.7 ± 137.1 371.5 ± 205 147.3 ± 77 32.2 ± 15.4 342 ± 176.1 87.6 ± 23.3 12.3 ± 9.0 1020.5 ± 222.4
36.3 ± 10 31.1 ± 5.3 34.0 32.6 144.6 62.2 12.0 108.2
Characterization of the Göttingen Minipig Pharmacokinetics
Fig. 3 Comparison of parameters values estimated by NCA analysis on the PK of antipyrine (ATP), atenolol (ATE), cimetidine (CIM), diazepam (DZP), hydrochlorothiazide (HCTZ), midazolam (MDZ) and theophylline (THP) in minipigs to values extracted from the literature for human, monkey, dog and rat. All data are available in Supplementary Material I.
glucuronide conjugate. In addition, a low level of 4hydroxymidazolam and an N-glucuronide of midazolam itself were also detected. In minipig hepatocytes, the most intense metabolites detected were consistent with the mono-oxidized metabolites, 1-hydroxymidazolam and 4-hydroxymidazolam, together with the formation of their corresponding glucuronide conjugates. In addition, a low relative level of an Nglucuronide conjugate of parent drug was detected, together with a third mono-oxidized metabolite, its corresponding mono-oxidized glucuronide conjugate. In general, minipig showed a slightly lower level of in vitro metabolism relative to human consistent with the observed intrinsic clearance determinations in human and minipig hepatocytes (data in supplementary material IV). However, all major human metabolites could be detected in vitro in minipig. Investigation of the major metabolites of midazolam detected in minipig plasma, identified the metabolites 4hydroxymidazolam and a 1,4-hydroxymidazolam, together with a 1-hydroxymidazolam glucuronide, a 4-
Table III Fraction of the Dose Excreted in the Urine in the First 24 h and Estimation of the Renal Clearance After Intravenous Dosing of Antipyrine, Atenolol, Cimetidine, Diazepam, Hydrochlorothiazide and Theophylline to Göttingen Minipig, Presented as Mean ± standard Deviation Fe in urine (%) Clr (ml/min/kg) fup (%) GFR * fup Antipyrine in males Antipyrine in females Atenolol Cimetidine Diazepam Hydrochlorothiazide Theophylline
2.7 ± 2.6 1.8 ± 1.5 37.1 ± 11.6 20.7 ± 8.2 0.1 ± 0.2 38.8 ± 11.8 13.6 ± 4.1
0.04 ± 0.07 0.09 ± 0.08 2.7 ± 0.5 7.5 ± 3.1 0.01 4.6 ± 1.4 0.15 ± 0.1
98 98 94 70.21 3.7 51.88 81.69
2.9 2.9 2.8 2.1 0.1 1.6 2.5
Comparison to GFR (glomerular filtration rate) and fraction unbound in plasma Bold characters indicate a renal clearance larger than the GFR*fup
hydroxymidazolam glucuronide and a 1,4-hydroxymidazolam glucuronide conjugate. The 4-hydroxymidazolam glucuronide was the most abundant metabolite detected in minipig plasma. In general, the minipig plasma metabolite profiles showed relatively good correlation with the profiles obtained from in vitro incubations in minipig hepatocytes.
Diazepam. The in vitro metabolism of diazepam was investigated and compared between human and minipig hepatocytes. In human hepatocytes, 3-hydroxy-diazepam and N-desmethyl diazepam were the predominant metabolites. The glucuronide conjugate of the 3-hydroxydiazepam could also be detected at a trace level. In minipig hepatocytes, the most intense metabolites detected were again 3-hydroxy-diazepam and Ndesmethyl diazepam, together with a glucuronide conjugate of the 3-hydroxy-diazepam. In addition, several metabolites of low relative abundance were also detected, including a second mono-oxidized metabolite and two further mono-oxidized glucuronide conjugates. In general, minipig showed a slightly higher level of in vitro metabolism relative to human, consistent with the intrinsic clearance determinations derived in hepatocytes (data in supplementary material IV). However, all of the major human in vitro metabolites were detected in minipig hepatocytes indicating good metabolite coverage. Investigation of the major metabolites of diazepam detected in minipig plasma identified the metabolites 3-hydroxydiazepam, N-desmethyl diazepam, and up to three monooxidized glucuronide conjugates. The resulting minipig plasma metabolite profiles showed good correlation with the metabolite profiles obtained from in vitro incubations in minipig hepatocytes, with all major minipig metabolites detected. It is interesting to note that for both compounds investigated, glucuronidation appeared to be a more pronounced metabolic pathway in vivo in minipig, with glucuronide conjugates
Lignet et al.
of the resulting phase I mono-oxidized species being the more predominant metabolites detected in minipig plasma. IVIVC Protein Binding, Blood to Plasma Partitioning, Hepatocellularity and Intrinsic Clearance Determinations To establish an in vitro – in vivo correlation (IVIVC) for the minipig, these supplementary in vitro data were generated. Protein Binding. Evaluation of the fraction of the reference compounds unbound in minipig plasma indicated significant binding for diazepam and midazolam, with values of 3.7, and 6.9% respectively. These may be compared with the values reported in the literature for human of 2.3 and 1.7% (44). The others compounds showed moderate to insignificant protein binding in minipig as in human (Tables I and V). Blood to Plasma Partitioning. The evaluation of the partitioning between blood and plasma for the reference compounds revealed values ranging from 0.61 to 1.1 for diazepam and atenolol respectively (Table IV). These values are very similar to those reported in the literature for human (see Table I for comparison), except for hydrochlorothiazide which showed a lower blood to plasma partitioning in minipig than in human (0.8 in minipig versus 1.7 or 2.7 in human depending on the source (46,47)). Hepatocellularity. The hepatocellularity of the Göttingen minipig was assessed based on the DNA and the protein quantification methods previously used for other species (53,58). For both methodologies, specific amounts of hepatocytes suspension cultures were compared to a specific amount of
Table IV Intrinsic Clearance Measured in in Minipig Hepatocytes (Clint), Presented as Mean ± Standard Deviation, Fraction of Unbound Drug Measured In Minipig Plasma (fu p ) and in the Culture Medium for Hepatocytes (fuinc), and Blood to Plasma Concentration Ratio (Rb) of the Reference Compounds in Minipigs Compound
Clint (μl/min/Mcells)
fup (%)
fuinc (%)
Rb
Antipyrine Atenolol Cimetidine Diazepam Hydrochlorothiazide Midazolam Theophylline
0.7 ± 2.4 1.6 ± 1.8 5.65 ± 0.05 4.05 ± 0.75 1.2 ± 1.7 10.8 ± 2.2 0.95 ± 0.95
98 94 70 3.7 52 6.9 82
79 100 86 47 79 16 72
0.79 1.1 0.93 0.60 0.80 0.61 0.83
The in vitro Clint values for antipyrine, atenolol, hydrochlorothiazide and theophylline could not be determined with confidence as they are below the assay sensitive range (Clint- values < 3 μl/min/Mcells)
homogenized liver tissue, assuming that the liver tissue is mainly constituted of hepatocytes (58,59). For reference, the DNA and protein contents of rat liver and rat hepatocytes were determined. DNA and protein contents in rat liver homogenate and hepatocytes were in agreement with the literature values and are given and compared to minipig in the supplementary material III. In the minipig, the DNA content was 2.9 ± 0.6 mg/g liver (mean ± standard deviation) and 17.1 ± 0.8 μg/106 hepatocytes. From these results, a HPGL of 119 ± 15 ∙106 cells/g liver for the minipig was calculated. The protein content of minipig liver homogenates and hepatocytes were 127.9 ± 13 mg/g liver and 0.97 ± 0.1 mg/106 cells, respectively. Thus, the HPGL calculated by the protein method was 129 ± 13 106 cells/g liver. The two HPGL values determined by either DNA content (119 ± 15 106 cells/g liver) or protein content (129 ± 13 106 cells/g liver) determinations are in good agreement with each other and are similar to the values reported for human and rat. The average of these two HPGL values for minipig is 124 ± 17 106 cells/g liver and this value was used for scaling of clearance. All results are summarized in Table V. Establishment of the Correlation Intrinsic hepatic clearances were estimated from the in vitro measurements using equation 2 and from the in vivo measurements using equation 3. The values of the parameters used are presented in Table VI, the resulting calculation in Table VII, and in vivo intrinsic hepatic versus intrinsic hepatic clearances in hepatocytes are represented in Fig. 4. For diazepam, midazolam and cimetidine, the in vitro clearance values could be determined with confidence. The other selected compounds showed very low turn-over with in vitro clearance values below the assay sensitive range of 3 μL/min/Mcells (hepatocyte suspension culture) and could therefore not be determined with confidence. Taking this limitation into account, an in vitro to in vivo correlation showed an under prediction for 6 of 7 compounds based on the in vitro measurements. Theophylline was the exception with a predicted value ~2-fold greater than the observed in vivo intrinsic clearance. For all compounds but hydrochlorothiazide (3.7 fold) and cimetidine (11 fold), the prediction was within three fold of the observed. Cimetidine showed an exceptionally large underprediction of the clearance (11 fold), even though the in vitro clearance could be determined with high confidence.
DISCUSSION Seven well studied marketed drugs were dosed intravenously and orally in male and female Göttingen minipig and noncompartmental analysis of concentration-time profiles was used to estimate species-specific pharmacokinetic parameters.
Characterization of the Göttingen Minipig Pharmacokinetics Table V Deviation
HPGL Derivation from DNA Quantification and Protein Quantification in Minipig Hepatocytes and Liver Tissue, Presented as Mean ± Standard DNA content of hepatocytes (μg/106 cells) 17.1 ± 0.8 Protein content of hepatocytes (mg/106 cells) 0.97 ± 0.08
DNA content of liver (mg/g liver) 2.9 ± 0.6 Protein content of liver (mg/g liver) 127.9 ± 13.44
These parameters were then compared to values extracted from the literature for human, monkey, dog and rat to highlight similarities and differences in pharmacokinetics between species (although a lack of reported data limited the comparison for hydrochlorothiazide). For systemic plasma clearance, values in minipig tended to be lower than in rat and dog and overall more comparable to clearance observed in monkey. An exception was cimetidine which showed a surprisingly high clearance in minipig. Cimetidine is mainly excreted in urine in human (32,34,35) and renal elimination was less important in the minipig (only ~20% of the dose excreted in urine). However, cimetidine clearance predicted from in vitro data was more than 11 times lower than the intrinsic clearance computed from in vivo data. In human, cimetidine is metabolized to N’-glucuronide, sulfoxide and hydroxymethyl metabolites (60,61) and the high clearance in minipig may be an indication of a high activity of these enzymes in this species. Indeed comparative studies indicated that minipig glucuronosyltransferases were much more efficient than in human, showing a higher activity (22) (this is further discussed below when we consider the qualitative metabolism of diazepam and midazolam). The renal clearance estimated in minipig is above the value expected due to glomerular filtration indicating active renal secretion. In human, cimetidine is secreted by P-glycoprotein (62) and as it has been shown that minipig and human present similar kidney transporter expression profiles (29) it is likely that a similar situation exists in minipig. A strong gender-related difference in pharmacokinetic was observed for antipyrine with female minipigs presenting a higher clearance and shorter terminal half-life. The renal clearance was very low and similar across genders, suggesting that this difference is due to increased metabolism in females. In human, antipyrine is metabolized by at least six hepatic cytochrome P450 enzymes (CYP1A2, CYP2B6, CYP2C8,
Table VI Parameter Values Used for the Computation of the Minipig Hepatic Clearance
HPGL (106 cells/g liver) 119 HPGL (106 cells/g liver) 129
CYP2C9, CYP2C18, and CYP3A4) (31) and a similar gender-related variability of antipyrine PK in human has been reported in the literature in two different studies with the halflife shorter in female subjects and metabolite production significantly different in males and females (63,64). For the other drugs known to be substrates of specific cytochromes in human such as diazepam (CYP3A, CYP2C19), midazolam (CYP3A4, CYP3A5) and paracetamol (CYP1A2, CYP3A4 and CYP2E1) no gender effects were seen in minipig pharmacokinetics although it has been reported by Skaanild et al. that CYP1A2, CYP2A6 and CYP2E1 show a higher metabolic activity in females compared to males (65). For atenolol the bioavailability in minipig was 34%, comparable to the value of ~50% reported in man (33). The ~40% of the dose retrieved unchanged in urine also matches closely the human situation (33) and as both the in vitro and in vivo metabolism of atenolol were low it also seems that as in man, the bioavailability in minipig is mainly is limited by incomplete absorption due to poor permeability. As in human, diazepam had high bioavailability in the minipig. The clearance of diazepam was lower in minipig than in the other pre-clinical species although still markedly higher than in human. The large interspecies differences in clearance are expected to be due to differences in the hepatic metabolism. Indeed intrinsic clearance is higher in rat and dog hepatocytes than in human hepatocytes (66), and our measurements showed a reasonable prediction of hepatic clearance from minipig hepatocytes. Variations in the metabolites produced are discussed in detail below. Unfortunately very few data were available to us from the literature on the PK of hydrochlorothiazide to allow a comparison of the minipig to other species although the bioavailability of ~60% is similar to the value reported for human (supplementary material I). However only ~40% of the dose was found in minipig urine while in human more than 90% is
Parameter
Meaning
Value
Unit
Source
HPGL BW LW Qh
Number of hepatocytes per gram of liver Body weight Liver weight Hepatic blood flow
120.106 14.2 237 38.9
Cells.g−1 kg g ml.min−1.kg−1
In house (8) (8) (8)
Lignet et al. Table VII Mean Intrinsic Hepatic Clearances and Standard Deviation Estimated from In Vitro Measurements (Cl int,h,est ) and From In Vivo Measurements (Cl int,h,mes ) for Seven Reference Compounds in the Göttingen Minipig Clint,h,est (ml.min−1.kg−1) Clint,h,mes(ml.min−1.kg−1) Compound
Mean
SD
Mean
Atenolol
1.4
4.8
3.5
SD 2.2
Antipyrine Cimetidine
3.2 11.3
3.6 0.1
6.1 128.1
3.0 10.4
Diazepam Midazolam
8.1 21.6
1.5 4.4
12.8 52.3
1.1 11.0
Theophylline
1.9
1.9
1.0
0.3
Hydrochlorothiazide 2.4
3.4
8.9
1.7
excreted in urine (37) indicating very limited metabolism. The in vitro metabolism of hydrochlorothiazide was also very low in minipig hepatocytes and tended to under predict the in vivo estimated hepatic clearance. The renal clearance was higher than expected due to glomerular filtration, which may be explained by the diuretic properties of the drug. It was noticed that the amount of urine collected after hydrochlorothiazide dosing was larger than after any other reference compound (data not shown). Midazolam bioavailability was rather low in minipig (5– 14%) and closer to the bioavailability in dog and rat than to the higher value of 34% observed in human. As systemic blood clearance was close to hepatic blood flow and extrahepatic clearance was assumed to be negligible, the low
Fig. 4 In vivo intrinsic hepatic clearances ± standard deviation versus intrinsic hepatic clearances in hepatocytes for 7 reference compounds in the Göttingen minipig. ATP: antipyrine, ATE: atenolol, CIM: cimetidine, DZP: diazepam, HCTZ: hydrochlorothiazide, MDZ: midazolam, THP: theophylline. The straight line is the line of unity while the dashed line and dotted line correspond to 2 and 3 fold differences respectively. For antipyrine, atenolol, theophylline and hydrochlorothiazide, measurements of intrinsic hepatic clearances in hepatocytes presented a standard deviation higher than the estimated mean. As a consequence, left error bars cannot be displayed on a log scale.
bioavailability may be largely attributed to high first pass effect in liver. The quantitative contribution of intestinal metabolism to the PK of midazolam in minipig remains unclear. Theophylline has a high bioavailability and low systemic clearance in minipig as in human and other pre-clinical species. In all species, the volume of distribution is inferior or equal to 1 and the bio-availability is complete. The fraction of dose excreted in the urine was only 14% which is similar to the human situation where theophylline is extensively metabolized and the renal clearance was less than filtration clearance indicating reabsorption of filtered theophylline. Theophylline was the only reference drug for which the in vitro hepatocyte clearance was over-predictive, however the large uncertainty in the in vitro measurement due to very low turnover does not allow any in vitro to in vivo extrapolation with confidence. For all drugs a large inter-individual variability was observed after both intravenous and oral dosing. In particular, times to maximal plasma drug concentrations after oral dosing were highly variable which could reflect the variable gastric emptying described in the minipig (9,10). High interindividual variability could also be linked to the gene expression variability in the species which was reported to be significantly higher than in human tissues by Heckel et al. (29) assuming that protein expression and consequent metabolic activity are correlated with RNA expression. The scaled intrinsic clearances measured in hepatocytes for these seven reference drugs tended to correlate roughly to the estimated in vivo hepatic intrinsic clearances. However it has to be acknowledged that the current study is not sufficient to establish a quantitative in vitro to in vivo correlation for hepatic clearance because the in vitro intrinsic clearance values for 4 of the tested compounds were too low to be determined with confidence in our assay. Therefore further work is needed here. However, it is noted that the tendency to under predict in vivo intrinsic clearances seen here for minipig has also been reported for human (67). The scaling of in vitro clearance reported here was made according to our usual in-house practice which assumes similar free fractions of drug in the in vitro incubations as in vivo. We acknowledge that a mechanistic approach should ideally consider the measured protein binding both in the in vitro incubation medium and in blood, and indeed such calculations were done (data in supplementary material II). A reason that these did not lead to a better correlation may arise from the fraction unbound in plasma and the fraction unbound in the hepatocyte medium containing FCS showed comparable values for this set of drugs. Only diazepam showed a significantly higher binding in the plasma than in the medium. A further investigation of the in vitro to in vivo correlation for more lipophilic and highly protein bound drugs is currently ongoing in our lab. The in vitro and in vivo metabolite profiles of midazolam and diazepam were also investigated here. For both examples the
Characterization of the Göttingen Minipig Pharmacokinetics
major human phase I metabolites were clearly detectable in minipig. Previous reports on cytochrome P450 mediated metabolism in minipig (18–20) have indicated that reactions catalyzed by the CYP1A, CYP2A and CYP3A enzyme families are very similar to those in human and the activity of CYP3A measured in vitro using midazolam as a marker were comparable in human and minipig (68). Our investigation of the in vivo metabolite profiles in minipig plasma indicated that glucuronidation is a major pathway. It is interesting to note that glucuronidation reactions both in vitro and in vivo in minipig were more pronounced than in equivalent human incubations. This observation is consistent with previous reports that glucuronosyltransferase reactions are catalyzed more efficiently in pigs and minipigs than in man (22,23). Overall our observations for these 2 drugs support that, from a metabolism perspective, minipig compares well with human both in vitro and in vivo. For the first time, the hepatocellularity was determined in the minipig and showed an average of 124 ± 17 106 cells/g liver, a value similar to the hepatocellularity reported in human and rat. The present dataset will be valuable for the building and validation of a physiologically-based pharmacokinetic (PBPK) model of the minipig. This type of models are based on a mechanistic understanding of drug ADME properties, and their use during drug research and development is now widespread. PBPK models in animals can help to design appropriate pharmacological and toxicological studies. Moreover, they have been shown to provide more accurate prediction of human PK than classical allometric scaling methods (69–71). PBPK models have been developed for common laboratory species such as rat, mouse, dog and monkey and have been implemented in software dedicated to PBPK modelling (72). More recently, the minipig is also being considered as a relevant animal model and a minipig PBPK model is available in PKSim (Bayer Technologies Services). Another PBPK model was implemented in GastroPlus (SimulationPlus Inc.) based on a published model of Suenderhauf et al. (8). However these preliminary models would profit from further refinement based on in vitro and in vivo data for well characterized drugs, such as our present database.
testing of new compounds, provided that the metabolic reactions are consistent with those in human. Further work is needed to confirm these findings with a wider range of compound types and to verify simulation of minipig pharmacokinetics using physiologically-based modelling. ACKNOWLEDGMENTS AND DISCLOSURES This work was funded by the Roche Post-Doc Fellowship (RPF) Program. We thank all our associates for their support, in particular Anthony Vandjour, Christelle Rapp and Claudia Senn for the in vivo measurements, Hamina Daff, Sandrine Simon, Isabelle Walter, Andreas Goetschi and Pierre Alexis Gonsard for their work on plasma protein binding and bioanalysis, Aynur Ekiciler for the hepatocellularity evaluation and in vitro intrinsic clearance determination, Peter Schrag for the blood to plasma portioning measurements, Martin Kapps for the LC/MS bioanalysis, and Michaela Marschmann for the investigation of the metabolite profiles.
REFERENCES 1.
2. 3.
4.
5.
6.
7.
CONCLUSIONS 8.
We have collected data on the ADME properties of the minipig for several reference drugs and have investigated the comparability to human. Our ability to predict in vivo behavior from in vitro measurements has also been explored and a satisfactory in vitro to in vivo correlation for hepatic metabolic clearance was established that should help with the species selection and study design. Overall, these results support the use of the minipig as a model for toxicological and safety
9.
10.
Chapman KL, Holzgrefe H, Black LE, Brown M, Chellman G, Copeman C, et al. Pharmaceutical toxicology: designing studies to reduce animal use, while maximizing human translation. Regulat ToxicoL Pharmacol : RTP. 2013;66(1):88–103. Zbinden G. The concept of multispecies testing in industrial toxicology. Regulat ToxicoL Pharmacol : RTP. 1993;17(1):85–94. Forster R, Bode G, Ellegaard L, van der Laan J-W. The RETHINK project on minipigs in the toxicity testing of new medicines and chemicals: conclusions and recommendations. J Pharmacol Toxicol Methods. 2010;62(3):236–42. Gutierrez K, Dicks N, Glanzner WG, Agellon LB, Bordignon V. Efficacy of the porcine species in biomedical research. Front Genet. 2015;6. Simianer H, Kohn F. Genetic management of the Gottingen Minipig population. J Pharmacol Toxicol Methods. 2010;62(3): 221–6. Swindle MM, Makin A, Herron AJ, Clubb Jr FJ, Frazier KS. Swine as models in biomedical research and toxicology testing. Vet Pathol. 2012;49(2):344–56. Sjogren E, Abrahamsson B, Augustijns P, Becker D, Bolger MB, Brewster M, et al. In vivo methods for drug absorption—comparative physiologies, model selection, correlations with in vitro methods (IVIVC), and applications for formulation/API/excipient characterization including food effects. Europ J Pharmaceut Sci : Off J Europ Fed Pharmaceut Sci. 2014;57:99–151. Suenderhauf C, Parrott N. A physiologically based pharmacokinetic model of the minipig: data compilation and model implementation. Pharm Res. 2013;30:1–15. Suenderhauf C, Tuffin G, Lorentsen H, Grimm HP, Flament C, Parrott N. Pharmacokinetics of paracetamol in Gottingen Minipigs: in vivo studies and modeling to elucidate physiological determinants of absorption. Pharm Res. 2014. Christiansen ML, Mullertz A, Garmer M, Kristensen J, Jacobsen J, Abrahamsson B, et al. Evaluation of the use of Gottingen minipigs to predict food effects on the oral absorption of drugs in humans. J Pharm Sci. 2015;104(1):135–43.
Lignet et al. 11.
12.
13.
14.
15.
16.
17. 18. 19.
20. 21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
Holzgrefe H, Ferber G, Champeroux P, Gill M, Honda M, Greiter-Wilke A, et al. Preclinical QT safety assessment: crossspecies comparisons and human translation from an industry consortium. J Pharmacol Toxicol Methods. 2014;69(1):61–101. Markert M, Stubhan M, Mayer K, Trautmann T, Klumpp A, Schuler-Metz A, et al. Validation of the normal, freely moving Gottingen minipig for pharmacological safety testing. J Pharmacol Toxicol Methods. 2009;60(1):79–87. Forster R, Bode G, Ellegaard L, van der Laan JW. The RETHINK project: minipigs as models for the toxicity testing of new medicines and chemicals: an impact assessment. J Pharmacol Toxicol Methods. 2010;62(3):158–9. Bode G, Clausing P, Gervais F, Loegsted J, Luft J, Nogues V, et al. The utility of the minipig as an animal model in regulatory toxicology. J Pharmacol Toxicol Methods. 2010;62(3):196–220. van der Laan JW, Brightwell J, McAnulty P, Ratky J, Stark C. Regulatory acceptability of the minipig in the development of pharmaceuticals, chemicals and other products. J Pharmacol Toxicol Methods. 2010;62(3):184–95. Mahl JA, Vogel BE, Court M, Kolopp M, Roman D, Nogués V. The minipig in dermatotoxicology: methods and challenges. Exp Toxicol Pathol. 2006;57(5–6):341–5. Dalgaard L. Comparison of minipig, dog, monkey and human drug metabolism and disposition. J Pharmacol Toxicol Methods. 2014. Skaanild MT. Porcine cytochrome P450 and metabolism. Curr Pharm Des. 2006;12(11):1421–7. Puccinelli E, Gervasi PG, Longo V. Xenobiotic metabolizing cytochrome P450 in pig, a promising animal model. Curr Drug Metab. 2011;12(6):507–25. Helke KL, Swindle MM. Animal models of toxicology testing: the role of pigs. Expert Opin Drug Metab Toxicol. 2013;9(2):127–39. Vree TB, Beneken Kolmer EW, Peeters A. Comparison of the metabolism of four sulphonamides between humans and pigs. Vet Quarter. 1991;13(4):236–40. Anzenbacherova E, Anzenbacher P, Svoboda Z, Ulrichova J, Kvetina J, Zoulova J, et al. Minipig as a model for drug metabolism in man: comparison of in vitro and in vivo metabolism of propafenone. Biomed Papers Med Faculty Univ Palacky, Olomouc, Czechoslovakia. 2003;147(2):155–9. Marini S, Longo V, Mazzaccaro A, Gervasi PG. Xenobioticmetabolizing enzymes in pig nasal and hepatic tissues. Xenobiotica. 1998;28(10):923–35. Baggot JD. Pharmacokinetic-pharmacodynamic relationship. Annales de recherches veterinaires. Ann Vet Res. 1990;21 Suppl 1:29s–40s. Martinez M, Amidon G, Clarke L, Jones WW, Mitra A, Riviere J. Applying the biopharmaceutics classification system to veterinary pharmaceutical products. part II. physiological considerations. Adv Drug Deliv Rev. 2002;54(6):825–50. Toutain PL, Ferran A, Bousquet-Melou A. Species differences in pharmacokinetics and pharmacodynamics. Handb Exp Pharmacol. 2010;199:19–48. Van Peer E, Verbueken E, Saad M, Casteleyn C, Van Ginneken C, Van Cruchten S. Ontogeny of CYP3A and P-Glycoprotein in the liver and the small intestine of the göttingen minipig: an immunohistochemical evaluation. Basic Clin Pharmacol Toxicol. 2014;114: 387–94. Van Peer E, De Bock L, Boussery K, Van Bocxlaer J, Casteleyn C, Van Ginneken C, et al. Age-related differences in CYP3A abundance and activity in the liver of the Gottingen minipig. Basic Clin Pharmacol Toxicol. 2015. Heckel T, Schmucki R, Berrera M, Ringshandl S, Badi L, Steiner G, et al. Functional analysis and transcriptional output of the Gottingen minipig genome. BMC Genomics. 2015;16(1):932. Koljonen M, Hakala KS, Ahtola-Satila T, Laitinen L, Kostiainen R, Kotiaho T, et al. Evaluation of cocktail approach to standardise
31.
32.
33. 34.
35. 36. 37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49. 50.
Caco-2 permeability experiments. Europ J Pharmaceut Biopharmaceut : Off J Arbeitsgemeinschaft fur Pharmazeutische Verfahrenstechnik eV. 2006;64(3):379–87. Engel G, Hofmann U, Heidemann H, Cosme J, Eichelbaum M. Antipyrine as a probe for human oxidative drug metabolism: identification of the cytochrome P450 enzymes catalyzing 4hydroxyantipyrine, 3-hydroxymethylantipyrine, and norantipyrine formation. Clin Pharmacol Ther. 1996;59(6):613–23. Collett A, Sims E, Walker D, He YL, Ayrton J, Rowland M, et al. Comparison of HT29-18-C1 and Caco-2 cell lines as models for studying intestinal paracellular drug absorption. Pharm Res. 1996;13(2):216–21. Reeves PR, McAinsh J, McIntosh DA, Winrow MJ. Metabolism of atenolol in man. Xenobiotica. 1978;8(5):313–20. Sitsen JMA, Maris FA, Timmer CJ. Concomitant use of mirtazapine and cimetidine: a drug–drug interaction study in healthy male subjects. E J Clin Pharmacol. 2000;56(5):389–94. Somogyi A, Gugler R. Clinical pharmacokinetics of cimetidine. Clin Pharmacokinet. 1983;8(6):463–95. Olkkola KT, Ahonen J. Midazolam and other benzodiazepines. Handb Exp Pharmacol. 2008;182:335–60. Beermann B, Groschinsky-Grind M, Rosen A. Absorption, metabolism, and excretion of hydrochlorothiazide. Clin Pharmacol Ther. 1976;19(5 Pt 1):531–7. Patel RB, Patel UR, Rogge MC, Shah VP, Prasad VK, Selen A, et al. Bioavailability of hydrochlorothiazide from tablets and suspensions. J Pharm Sci. 1984;73(3):359–61. Veber DF, Johnson SR, Cheng HY, Smith BR, Ward KW, Kopple KD. Molecular properties that influence the oral bioavailability of drug candidates. J Med Chem. 2002;45(12):2615–23. Thummel KE, O’Shea D, Paine MF, Shen DD, Kunze KL, Perkins JD, et al. Oral first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism. Clin Pharmacol Ther. 1996;59(5):491–502. Zhang Z-Y, Kaminsky LS. Characterization of human cytochromes P450 involved in theophylline 8-hydroxylation. Biochem Pharmacol. 1995;50(2):205–11. Lennernas H. Modeling gastrointestinal drug absorption requires more in vivo biopharmaceutical data: experience from in vivo dissolution and permeability studies in humans. Curr Drug Metab. 2007;8(7):645–57. Wu CY, Benet LZ. Predicting drug disposition via application of BCS: transport/absorption/ elimination interplay and development of a biopharmaceutics drug disposition classification system. Pharm Res. 2005;22(1):11–23. Obach RS, Lombardo F, Waters NJ. Trend analysis of a database of intravenous pharmacokinetic parameters in humans for 670 drug compounds. Drug Metab Disposition: Biol Fate Chem. 2008;36: 1385–405. Taylor EA, Turner P. The distribution of propranolol, pindolol and atenolol between human erythrocytes and plasma. Br J Clin Pharmacol. 1981;12(4):543–8. Paixao P, Gouveia LF, Morais JA. Prediction of drug distribution within blood. Europ J Pharmaceut Sci : Off J Europ Fed Pharmaceut Sci. 2009;36(4–5):544–54. Akabane T, Tabata K, Kadono K, Sakuda S, Terashita S, Teramura T. A comparison of pharmacokinetics between humans and monkeys. Drug Metab Dispos. 2010;38(2):308–16. Gertz M, Houston JB, Galetin A. Physiologically based pharmacokinetic modeling of intestinal first-pass metabolism of CYP3A substrates with high intestinal extraction. Drug Metab Dispos. 2011;39(9):1633–42. Mitenko PA, Ogilvie RI. Pharmacokinetics of intravenous theophylline. Clin Pharmacol Ther. 1973;14(4):509–13. Downs TR, Wilfinger WW. Fluorometric quantification of DNA in cells and tissue. Anal Biochem. 1983;131(2):538–47.
Characterization of the Göttingen Minipig Pharmacokinetics 51.
Seglen PO. Preparation of isolated rat liver cells. Methods Cell Biol. 1976;13:29–83. 52. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, et al. Measurement of protein using bicinchoninic acid. Anal Biochem. 1985;150(1):76–85. 53. Carlile DJ, Zomorodi K, Houston JB. Scaling factors to relate drug metabolic clearance in hepatic microsomes, isolated hepatocytes, and the intact liver: studies with induced livers involving diazepam. Drug Metab Dispos. 1997;25(8):903–11. 54. Cibulskyte D, Kaalund H, Pedersen M, Horlyck A, Marcussen N, Hansen HE, et al. Chronic cyclosporine nephrotoxicity: a pig model. Transplant Proc. 2005;37(8):3298–301. 55. Rowland M, Tozer TN. Clinical pharmacokinetics and pharmacodynamics : concepts and applications. 4th ed. Philadelphia: Wolters Kluwer Health/Lippincott William & Wilkins, c2011; 2005. 56. Rowland M, Benet LZ, Graham GG. Clearance concepts in pharmacokinetics. J Pharmacokinet Biopharm. 1973;1(2):123–36. 57. Yang J, Jamei M, Yeo KR, Rostami-Hodjegan A, Tucker GT. Misuse of the well-stirred model of hepatic drug clearance. Drug Metab Dispos. 2007;35(3):501–2. 58. Sohlenius-Sternbeck AK. Determination of the hepatocellularity number for human, dog, rabbit, rat and mouse livers from protein concentration measurements. Toxicol in vitro : Int J Publ Assoc BIBRA. 2006;20(8):1582–6. 59. Worboys PD, Bradbury A, Houston JB. Kinetics of drug metabolism in rat liver slices. II. comparison of clearance by liver slices and freshly isolated hepatocytes. Drug Metab Dispos. 1996;24(6):676– 81. 60. Larsson R, Erlanson P, Bodemar G, Walan A, Bertler A, Fransson L, et al. The pharmacokinetics of cimetidine and its sulphoxide metabolite in patients with normal and impaired renal function. Br J Clin Pharmacol. 1982;13(2):163–70. 61. Mitchell SC, Idle JR, Smith RL. The metabolism of [14C]cimetidine in man. Xenobiotica. 1982;12(5):283–92. 62. Dudley AJ, Brown CD. Mediation of cimetidine secretion by Pglycoprotein and a novel H(+)-coupled mechanism in cultured
renal epithelial monolayers of LLC-PK1 cells. Br J Pharmacol. 1996;117(6):1139–44. 63. O’Malley K, Crooks J, Duke E, Stevenson IH. Effect of age and sex on human drug metabolism. Br Med J. 1971;3(5775):607–9. 64. Teunissen MW, Srivastava AK, Breimer DD. Influence of sex and oral contraceptive steroids on antipyrine metabolite formation. Clin Pharmacol Ther. 1982;32(2):240–6. 65. Skaanild MT, Friis C. Cytochrome P450 sex differences in minipigs and conventional pigs. Pharmacol Toxicol. 1999;85:174–80. 66. Chenery R, Ayrton A, Oldham H, Standring P, Norman S, Seddon T, et al. Diazepam metabolism in cultured hepatocytes from rat, rabbit, dog, guinea pig, and man. Drug Metab Dispos. 1987;15(3):312–7. 67. Hallifax D, Houston JB. Evaluation of hepatic clearance prediction using in vitro data: emphasis on fraction unbound in plasma and drug ionisation using a database of 107 drugs. J Pharm Sci. 2012;101(8):2645–52. 68. Turpeinen M, Ghiciuc C, Opritoui M, Tursas L, Pelkonen O, Pasanen M. Predictive value of animal models for human cytochrome P450 (CYP)-mediated metabolism: a comparative study in vitro. Xenobiotica. 2007;37(12):1367–77. 69. Parrott N, Jones H, Paquereau N, Lave T. Application of full physiological models for pharmaceutical drug candidate selection and extrapolation of pharmacokinetics to man. Basic Clin Pharmacol Toxicol. 2005;96(3):193–9. 70. Jones HM, Parrott N, Jorga K, Lavé T. A novel strategy for physiologically based predictions of human pharmacokinetics. Clin Pharmacokinet. 2006;45:511–42. 71. Jones HM, Chen Y, Gibson C, Heimbach T, Parrott N, Peters SA, et al. Physiologically based pharmacokinetic modeling in drug discovery and development: a pharmaceutical industry perspective. Clin Pharmacol Ther. 2015;97(3):247–62. 72. Bouzom F, Ball K, Perdaems N, Walther B. Physiologically based pharmacokinetic (PBPK) modelling tools: how to fit with our needs? Biopharm Drug Dispos. 2012;33(2):55–71.