Abstract
The disposition of seven marketed and two AstraZeneca acid (organic anion) compounds with a range of volume of distribution at steady state (Vss) and clearance have been profiled in rat and dog. Pharmacokinetic (PK) parameters along with liver and muscle tissue levels were collected, and their contributions to total Vss were calculated. The physiologically based prediction of Vss correlated (all predictions within 2-fold) with the Vss obtained from plasma PK analysis. The Vss of the acid drugs with atypically high values could be explained by significant sequestering of compound to the liver. A “media loss” in the in vitro hepatocyte assay that monitors loss of compound from the incubation media along with physiologically based PK (PBPK) modeling was assessed for its ability to accurately predict the impact of hepatic uptake on both clearance and Vss. This methodology significantly improved the prediction of metabolic in vivo clearance compared with standard hepatocyte scaling approaches that do not take into account hepatic uptake. Predictions of Vss from the media loss assay also correlate with the measured values from plasma PK analysis. However, hepatic uptake will have little overall impact on half-life, because of the concomitant impact on both Cl and Vss, as long as hepatic extraction is not high. The methodology described here is particularly useful when there is no allometric relationship between species as a result of interspecies differences in liver uptake. In this situation, the potential use of human hepatocytes combined with PBPK modeling avoids the question of which species pharmacokinetics is most predictive to humans.
Introduction
Within the drug discovery environment, the ability to accurately predict human in vivo pharmacokinetics leading to the successful progression of drug candidates is one of the most challenging aspects of drug metabolism and pharmacokinetics (PK). Of particular importance is the ability to be able to predict the clearance (CL) and volume of distribution (Vd) of a candidate drug, which when combined can be used to predict human half-life (t1/2). Therefore, inclusion of all factors that influence CL and Vd are of paramount importance in the accurate prediction of the human PK of a candidate drug.
For many drugs, the major route of clearance is via metabolism within the liver. Prediction of hepatic clearance from a variety of in vitro systems has formed a cornerstone of the design and selection of small molecule drug candidates within the pharmaceutical industry for more than a decade. For compounds whose clearance is predominantly via hepatic metabolism, the approaches taken have been relatively successful (Soars et al., 2002; Ito and Houston, 2004; Riley et al., 2005), although not without their failures (Soars et al., 2007). The realization that hepatic transporters may add additional layers of complexity has added to the challenge in terms of both the modeling required (Sirianni and Pang, 1997; Liu and Pang, 2005) and the in vitro assays used (Soars et al., 2007). This extra complexity has allowed the opportunity to reassess compounds for which metabolic clearance was poorly predicted via more conventional methods.
The PK parameter commonly used to characterize distribution of a drug is the volume of distribution at steady state (Vss). The Vss represents the ratio of the amount of drug in the body to the plasma concentration at steady state. The human Vss of a candidate drug is commonly predicted using an allometric scaling approach. This is generally regarded as a successful method in most cases, whether with use of one species, usually rat (Caldwell et al., 2004), or two or more species (Mahmood, 1999). Successful allometric scaling of Vss to humans assumes similar tissue binding between species. Therefore, any distribution differences between species due to mechanistic dissimilarity may lead to poor predictions to humans. Physiologically based PK (PBPK) models have been developed and used in the last 30 years for the prediction of Vss of drugs and chemicals. PBPK models strive to be mechanistic by mathematically transcribing anatomical, physiological, physical, and chemical descriptions of the phenomena involved in the complex absorption, distribution, metabolism, and excretion processes (Kawai et al., 1994; Charnick et al., 1995). A large degree of residual simplification and empiricism is still present in PKPB models, but they have an extended domain of applicability compared with that of empirical allometric methods.
There have been several publications highlighting the impact of hepatic uptake on the Vss of a drug (Paine et al., 2008; Grover and Benet, 2009) and the potential impact of drug-drug interactions. Acid drugs [which are typically characterized by low Vss (0.1–0.25 l/kg) because of their poor tissue distribution and high plasma protein binding] are particularly sensitive to transporter effects. Paine et al. (2008) have shown that the liver levels for the acid drug atorvastatin are many times higher than those of the blood levels at steady state in a rat bile duct-cannulated study. The high concentration of atorvastatin in an organ that has a significant contribution to body weight led to a Vss many times higher than that of a typical acid drug. This Vss value along with the liver levels could be quantitatively predicted by a PBPK model that uses in vitro hepatocyte uptake data in combination with various physiological parameters. Therefore, inhibition and induction of hepatic uptake transporters may lead to changes in Vss and species differences in the transporters may lead to poor predictions of PK in humans.
The aim of this work was to investigate seven marketed acid drugs and two AstraZeneca acid compounds and determine their PK in both rat and dog bile duct-cannulated studies. Bile duct-cannulated studies were used to assess the contribution of biliary and renal clearance to the overall clearance value. Tissue (blood, liver, and muscle) distribution for the compounds was investigated in both rat and dog, and a simple PBPK model was used to assess tissue contribution to Vss. The nine compounds were assessed in an in vitro hepatocyte assay for their cellular disposition, and data were fitted to a two-compartment model (Paine et al., 2008) to determine in vitro clearance values for active/passive transport and metabolic clearance in both rat and dog. These parameters were then used in conjunction with a seven-compartment PBPK model to predict the impact of hepatic uptake on PK parameters. Results were compared between the in vivo measurements and the predicted PK parameters.
Materials and Methods
Chemicals.
All chemicals and reagents used were of the highest available grade. Telmisartan was sourced from Sequoia Research Products Ltd. (Oxford, UK). Ramatroban and the AstraZeneca compounds were synthesized in house by AstraZeneca. S-Ibuprofen, indomethacin, losartan, tolmetin, and sulfisoxazole and all other chemicals were purchased from Sigma-Aldrich (Poole, Dorset, UK). Hepatocyte suspension buffer consisted of 2.34 g of Na-HEPES, 0.4 g of d-fructose, 2 g of bovine serum albumin, and 1 liter of powder equivalent of Dulbecco's modified Eagle's medium (Sigma-Aldrich, Gillingham, UK) diluted in 1 liter of water and adjusted to pH 7.4 with 1 M HCl.
Preparation of Hepatocytes.
Rat.
Rat hepatocytes were isolated from male Sprague-Dawley rats using the two-step in situ collagenase perfusion method of Seglen (1976) described in detail in Soars et al. (2007). Cells were resuspended in suspension buffer without bovine serum albumin, and an estimation of hepatocyte yield and viability was obtained using the trypan blue exclusion method. Only cells with a viability of >80% were used.
Dog.
Dog hepatocytes were isolated in house from male beagle dogs approximately 1 year old. The isolation procedure was based on the two-step in situ collagenase perfusion method described in more detail by McGinnity et al. (2004). Hepatocytes were resuspended for use in suspension buffer. Yield and viability were assessed using the trypan blue exclusion method, and only batches with viability >80% were used.
Determination of CLinc in Hepatocytes Using the Standard Incubation Method (“Cells and Media Assay”).
The rate of turnover in hepatocyte suspensions (CLinc) was estimated using a procedure that sampled the whole incubation. Ten-microliter aliquots of dimethyl sulfoxide (10 μl) stocks of ibuprofen indomethacin, losartan, ramatroban, sulfisoxazole, telmisartan, tolmetin, AZ1 and AZ2 (100 μM) were added to 490-μl aliquots of hepatocyte buffer (protein-free) and warmed to 37°C. Hepatocytes were diluted to 2 × 106 cells/ml in protein-free hepatocyte buffer and warmed to 37°C. At time 0, 0.5 ml of cells was added to the appropriate substrate solution, mixed, and placed in a shaking water bath (37°C and 80 oscillations/min). At 5, 15, 30, 45, 60, 75, and 90 min, 40-μl aliquots were removed and quenched with 80 μl of ice-cold methanol. Samples were mixed and then stored at −20°C to await analysis and to enhance protein precipitation. Before analysis, samples were spun at 2000g for 20 min at 4°C to pellet protein, and the supernatant was transferred to a 96-well Agilent plate for LC-MS/MS analysis. Parent peak area was plotted against time, and CLinc was estimated by multiplying the incubation volume by the elimination rate constant. Each compound was incubated on at least three separate occasions.
Determination of Loss from Media (CLmed) Using Hepatocytes (“Media Loss Assay”).
CLmed values were determined essentially as described above except that 2-ml incubations were prepared. Aliquots (80 μl) were removed at 0, 0.5, 1, 2, 4, 6 15, 30, 45, 60, 75, and 90 min and placed into centrifuge tubes. These aliquots were immediately centrifuged at 7000g for 30 s using a MSE MicroCentaur centrifuge (Thermo Fisher Scientific, Loughborough, UK), and 40 μl of the supernatant was pipetted into 80 μl of ice-cold methanol. Samples were then frozen for 1 h at −20°C, and centrifuged at 2000g for 20 min at 4°C. The supernatants were removed and analyzed as described below. Each compound was incubated on at least three separate occasions, and the CLmed was estimated as described under Data Analysis. Data were fitted to the two-compartment model shown in Fig. 1 and in vitro parameters were calculated using ModelMaker 3 (ModelKinetix, Oxford, UK).
Compartmental model describing hepatocyte incubation.
Determination of Plasma Protein Binding.
Plasma protein binding was measured using equilibrium dialysis at least twice for each compound. Ibuprofen, indomethacin, losartan, ramatroban, sulfisoxazole, telmisartan, tolmetin, AZ1, and AZ2 (final concentration 10 μM) were spiked into rat or dog plasma and placed on one side of a dialysis cell; the other side contained only buffer. The compounds were dialyzed through a 50-kDa membrane in a Dianorm rotating unit (Diachema, Langnau, Switzerland) for 18 h at 37°C. Aliquots from the buffer and dialysate side of the membrane were then quenched in methanol and analyzed via LC-MS/MS as described below. The extent of plasma protein binding was calculated by dividing the concentration of compound in the absence of plasma (buffer side) by the concentration of compound in the presence of plasma (dialysate side).
Determination of Fraction Unbound in the Incubation.
Unbound fraction in hepatocytes (1 × 106 cells/ml) was measured at least twice for ibuprofen, indomethacin, losartan, ramatroban, sulfisoxazole, telmisartan, tolmetin, AZ1, and AZ2 at 1 μM by equilibrium dialysis over a 3-h period at 37°C as has been described previously by Austin et al. (2005).
Determination of Fraction Unbound in Liver Tissue.
Unbound fraction in 25% liver homogenate was measured at least twice for ibuprofen, indomethacin, losartan, ramatroban, sulfisoxazole, telmisartan, tolmetin, AZ1, and AZ2 at 1 μM by equilibrium dialysis at 37°C for 18 h. Free fraction in whole liver was calculated by correcting for the homogenate dilution using the equation 1 outlined and validated by Austin et al. (2002):
where fu2 in this case is the free fraction in whole liver, fu1 is the measured free fraction in 25% homogenate, and C1 and C2 are the concentrations of the homogenate, i.e., 25 and 100%, respectively.
Determination of Blood/Plasma Ratio.
Ibuprofen, indomethacin, losartan, ramatroban, sulfisoxazole, telmisartan, tolmetin, AZ1, and AZ2 at 100× final concentration were spiked into aliquots of fresh rat or dog blood and plasma (0.5 ml) and incubated for 15 min at 37°C in a shaking water bath followed by centrifugation in a MSE MicroCentaur centrifuge at 7000g for 4 min. Aliquots of plasma from both the blood and plasma incubations were quenched with methanol and stored at −20°C for at least 1 h. Before analysis, samples were centrifuged at 2000g for 20 min, and the supernatant was transferred into HPLC vials for LC-MS/MS analysis. The blood/plasma ratio was calculated by dividing the peak area from directly spiked plasma by the peak area from plasma isolated from spiked blood.
Determination of kmem.
kmem for each compound was determined by measuring the instantaneous binding of compound to the cell by adding compound to freshly prepared hepatocytes (1 μM, 1 million cells/ml), immediately sampling followed by centrifugation (7000g), taking the supernatant, and comparing the compound levels with the total incubation concentration using LC-MS/MS analysis. This was done five times for each compound.
In Vivo Studies.
All in vivo work was subject to internal ethics review and conducted in accordance with Home Office requirements under the Animals Scientific Procedures Act of 1986. Healthy virus antibody-free male Sprague-Dawley rats were obtained from Charles River (Margate, UK). They were housed in a light-controlled room, kept at a temperature of 19°C ± 2°C and 55 ± 10% humidity. They received a Teklad 2021 diet (Harlan, Loughborough, UK) and had access to water ad libitum. Healthy male beagles (bred in-house) were housed in pairs in a controlled environment (temperature 18°C ± 2°C and humidity 55 ± 10%). They were fed with a SDS D3 (E) Dog Maintenance diet and had access to water ad libitum. They were fully vaccinated and wormed and had daily access to an external play area for exercise and socialization. After at least 4 weeks of acclimatization, a chronic bile duct cannulation was performed on two dogs following the technique described by Kissinger and Garver (1998). Dogs were allowed to recover from surgery for a minimum of 1 month.
Pharmacokinetics Studies in the Rat.
After at least 1 week of acclimatization, rats (250–350 g) were surgically prepared under isoflurane anesthesia. The bile duct was cannulated, and cannulae were also implanted into the jugular vein (dosing cannula) and carotid artery (blood sampling cannula). Dose solutions were administered via the intravenous (jugular vein) cannula. Doses administered were calculated by the weight difference of the dosing syringe before and after administration. Serial plasma samples (200–300 μl) were taken from the intra-arterial (carotid artery) cannula. Approximately 200 μl of blood was drawn through the cannula before a sampling aliquot was taken to ensure that circulating blood was sampled through the cannula. Serial blood samples were taken over a 7-h time course. Blood was centrifuged at 1110g for 10 min at 4°C. Plasma was transferred into plain polypropylene tubes, each prepared in advance to contain 1.5 μl of concentrated phosphoric acid to stabilize any potential acyl glucuronide metabolites. Samples were then immediately frozen upright on dry ice and stored at −20°C. Bile and urine samples were treated with phosphoric acid, frozen, and stored as described previously. Livers and sartorius muscle tissue were taken at termination for all compounds.
Pharmacokinetics Studies in the Dog.
Test compounds were dissolved in bicarbonate buffer (1 mg/ml) containing 10% ethanol to a final concentration of 1 mg/ml. The dose was administered to the dogs (n = 2) via a 30-min infusion in the cephalic vein. Approximately 2.5 ml of blood was collected on EDTA via the jugular vein 0, 15, 30, 60, 120, 180, 300, 420, 720, and 1440 min after the beginning of the infusion. Bile was collected over 24 h. Dogs were kept in metabolism cages over 24 h for urine collection on dry ice. Plasma, bile, and urine samples were prepared within 30 min of collection according to the following procedure. Blood was centrifuged at 1110g for 10 min at 4°C. Plasma was transferred into plain polypropylene tubes, each prepared in advance to contain a modicum of concentrated phosphoric acid to stabilize any potential acyl glucuronide metabolite. Samples were then immediately frozen upright on dry ice and stored at −20°C. Bile and urine samples were treated with phosphoric acid, frozen, and stored as described previously. When terminal studies in the dog were performed, they were carried out as above, but no samples were collected during the study except liver and sartorius muscle tissue, which were taken at termination. Because of a shortage of availability of terminal dogs, it was only possible to monitor tissue levels of three compounds: telmisartan, losartan and ibuprofen.
Tissue Measurements.
Tissues were homogenized, and all samples were analyzed as described below. Tissue concentrations were measured, and their contribution to total Vss using volume of distribution of tissue = Vt × Kp × (1 − E), where Vt is volume of tissue, Kp is plasma tissue partition coefficient, and E is the tissue extraction ratio. Other pharmacokinetic parameters were calculated using noncompartmental analysis in WinNonlin (version 3.2; Pharsight, Mountain View, CA). Metabolic clearance was calculated by subtracting any renal or biliary clearance from the total clearance.
In Vivo Sample Preparation.
Plasma was dispensed into 50-μl aliquots and 150 μl of methanol was added (containing internal standard) and mixed. Bile and urine were diluted in water before analysis.
Liver and muscle were weighed, diluted with water, and homogenized. Homogenate was dispensed into 50-μl aliquots in triplicate and processed as per plasma. Appropriate standard curves and quality control samples prepared in the equivalent blank tissue were used for each analysis.
Sample Analysis.
Samples were analyzed by LC-MS/MS with an HP1100 HPLC system (Hewlett Packard, Palo Alto, CA) linked to a Quattro Ultima mass spectrometer (Micromass; Waters, Milford, MA) in negative or positive electrospray ionization mode with data analysis on QuanLynx software (version. 4.0, Micromass; Waters). Cone voltage and collision energy were optimized for each compound. In these analyses, chromatographic separation was achieved using a Waters Symmetry C8 3.5-μm (2.1 × 30 mm) column using 10 μl of each sample. The mobile phase consisted of an aqueous phase of water with 0.1% (v/v) formic acid and an organic phase of methanol with 0.1% (v/v) formic acid. Samples were quantified using appropriate calibration curves and quality controls. When plasma, liver, and muscle levels were fitted to the seven-compartment model (Fig. 2), Berkeley Madonna software (version 8.3.18; University of California, Berkeley, CA) was used to estimate in vivo parameters.
Seven-compartment model describing sinusoidal bidirectional passive permeation and active uptake/efflux in the liver. I, interstitial fluid; LI, liver interstitial; T, tissue; B, blood; L, liver; LV, liver vascular; L, mem, liver cell membrane; L, cell, liver cellular.
Data Analysis.
In vitro.
A hepatocyte incubation model that was described previously by Paine et al. (2008) was used; this model includes medium, cellular, and cell membrane compartments (Fig. 1). After the addition of a test compound, CLinc is obtained from standard procedures by sampling from the hepatocyte-medium suspension. Clearance from the incubation can be estimated from the slope of the natural log (concentration in incubation) time plot. However, CLinc has limited applications when uptake occurs because the free concentration of drug is different between the medium and hepatocyte. A more appropriate in vitro clearance term is overall clearance from the medium (CLmed) and can be expressed as in eq. 1:
where the subscripts cell and med refer to the cell (excluding the membrane) and medium compartments, respectively. C is the total drug concentration, and fu is the free fraction of the corresponding compartment. CLmed is the clearance from the medium, and CLint, met is the unbound metabolic intrinsic clearance from the cell. Albumin was not present in the cell suspension, and therefore fumed was assumed to be equal to 1. Ψ represents the ratio of the free concentration inside the cell to the free concentration in the medium at steady state.
The amount of drug bound to the cell membrane (Xmem) is directly proportional to the concentration of drug in the medium as in eq. 2:
where kmem is the proportionality constant between the amount in membrane and free concentration in medium and has units of volume. In this model, the initial concentration in the medium is given by eq. 3:
where dose is the total amount of drug added to the suspension, Vmed is the volume of the medium, and Cmed(t=0) is the medium concentration at time 0. Hence, kmem can be estimated by sampling the medium as soon as possible after addition of test compound.
CLmed can also be expressed in terms of passive, uptake, efflux, and metabolic intrinsic clearances as in eq. 4:
Intrinsic clearances in eq. 4 were obtained by fitting the model in Fig. 1 to the concentration of drug in medium with time and the assumption that efflux was negligible. CLmed was then estimated for the test compounds by substituting the intrinsic terms into eq. 4.
Hence, from eqs. 1 and 4, the ratio of the free concentration inside the cell to the free concentration in the medium (Ψ) can be expressed as in eq. 5:
When Vmed approximates to the volume of the incubation (Vinc), CLmed also equals the incubational clearance (CLinc) divided by the fraction in the medium at steady state (fmed, ss) as in eq. 6:
When no uptake occurs, fmed, ss is equal to the fraction unbound in the incubation (fuinc).
In vivo.
A seven-compartment physiological model (Fig. 2) as described by Paine et al. (2008) illustrates sinusoidal bidirectional passive permeation and active uptake/efflux between the liver vascular and cellular compartments and is described by the in vivo intrinsic clearances: CLint, L, pass, CLint, L, uptake, and CLint, L, efflux. Drug is eliminated from the cellular compartment with in vivo intrinsic clearance (CLint, L). The model has been designed to be analogous to the in vitro model shown in Fig. 1.
Predicting in vivo from in vitro.
Equations for blood clearance and Vss for the above seven-compartment physiological model (Fig. 2) have been described by Paine et al. (2008). These equations were used to predict the clearance and Vss of the test compounds in rat and dog. Assumptions on physiological volumes, blood flows, and the fraction unbound in body tissues are also described in Paine et al. (2008).
The in vitro parameters CLint, pass, CLint, uptake, CLint, met, and kmem were scaled to their in vivo analogs CLint, L, pass, CLint, L, uptake, CLint, L, and kL, mem using standard rat and dog biological scaling factors (where kL, mem is the proportionality constant between free blood concentration and the amount of drug in the cell membranes of the liver). However, using standard biological scaling factors to scale in vitro CLint, met generated in isolated human hepatocytes suggests a systematic underprediction of in vivo CLint, L (Ito and Houston, 2005; Riley et al., 2005). A similar systematic bias of approximately 5-fold has been observed in the rat data in this laboratory (Grime and Riley, 2006) and can be found in the rat CLint, met data presented by some academic laboratories (Naritomi et al., 2001) but not others (Ito and Houston, 2004). Moreover, little is known when CLint, uptake in isolated hepatocytes is scaled to CLint, L, uptake in the intact liver, and this work allows the opportunity to investigate any discrepancies. Therefore, the plasma and liver levels of compounds that showed significant uptake in hepatocytes (fmed, ss < 0.5 × fuinc) were fitted to the seven-compartmental model as shown in Fig. 2, and values for CLint, L, pass, CLint, L, uptake, and CLint, L were determined. These values were compared with the scaled in vitro parameters using standard biological scaling factors, and correction factors were estimated on the basis of any discrepancies. These correction factors were then applied to all nine compounds, and predictions were made for both Vss and clearance.
Results
Physical Properties.
The physical properties of the nine compounds are shown in Table 1. Overall the range of physical properties for the compounds is typical of small molecule acid drugs. For example, the molecular weights range from 200 to 500, acid pKa values range from 3 to 5, and lipophilicity ranges from a logD7.4 of −1 to 2. AZ1 and AZ2 in most regards have properties that are very similar to those of the marketed drugs, albeit being the most acidic.
Physical properties
In Vivo.
The major PK parameters in rat and dog are shown in Table 2. In most cases metabolic clearance is the main clearance mechanism, with the exception of sulfisoxazole in the dog in which renal clearance is the major component. It was noticeable that the PK parameters changed very little between bile duct-cannulated and non-bile duct-cannulated animals (data not shown). In the rat, five of the compounds have Vss values outside the typical acid range (0.1–0.25 l/kg): losartan, telmisartan, AZ1, AZ2, and ramatroban. For the most part, the Vss values in the dog fall within the expected range for acids, with the exception of ramatroban (Vss of 2.2 l/kg) and to some extent telmisartan (Vss of 0.7 and 0.4 l/kg). The contribution of the different tissues to the total Vss is shown in Table 3 and Fig. 3. The Vss of the acid drugs outside the normal range in the rat are explained by the large contribution by the liver. Those compounds with higher Vss in the rat have high concentrations in the liver. The liver levels measured in the dog were low and were consistent with the Vss of these compounds. The physiologically based prediction of Vss correlates very well with the measured values (all predictions within 2-fold) (Fig. 4).
Observed in vivo parameters in bile duct-cannulated rats and dogs
All data represent individual animals or mean ± S.D. of a minimum of three animals.
Volume of distribution of tissues in rat and dog
Mean values of two animals (rat) and one animal (dog) from tissue homogenate prepared in triplicate.
Contribution from tissues to total Vss.
Correlation of a physiologically predicted Vss with Vss calculated from plasma (the solid line represents the line of unity, dotted lines signify 2-fold errors between the predicted and observed values, and statistics were calculated using log data). RMSE, root mean square error; AFE, average fold error.
In Vitro.
Hepatocyte assay media loss plots for each of the compounds in rat and dog are shown in Figs. 5 (rat) and 6 (dog). Indomethacin, losartan, ramatroban, AZ1, and AZ2 all show significant biphasic disappearance from the media in rat hepatocytes (Fig. 5). In dog, losartan, AZ1, and to a much lesser extent ramatroban and sulfisoxazole show biphasic disappearance from the media (Fig. 6). The data were fitted using the two-compartment model in Fig. 1 to obtain in vitro parameters. The measured in vitro parameters are shown in Table 4. These indicate that on the basis of fmed, ss and fuinc measurements the compounds that are most actively transported in hepatocytes are losartan, telmisartan, AZ1, AZ2, and ramatroban in the rat. Table 5 shows the in vitro and in vivo values of the individual intrinsic clearance parameters for these compounds in the rat. In addition, in Table 5, the parameter Ψ, which represents the ratio of the free concentration inside the hepatocytes relative to the concentration in the media, is >1 for all five compounds. In general, there was little active transport observed in dog hepatocytes and, therefore, dog data were not used for comparing intrinsic clearances between the in vitro and in vivo cases. With this limited data set, it appears that the passive correction factor from in vitro to in vivo is close to unity, suggesting that the value can be simply scaled on the basis of hepatocellularity. The correction factor for metabolic intrinsic clearance is 2-fold and is consistent with known literature differences between measured and scaled metabolic clearance from hepatocytes (see Data Analysis). However, the correction factor for hepatic uptake is more than 6 times that of the scaled value using hepatocellularity alone. These mean correction factors have been applied to the in vitro intrinsic clearances of all the compounds to predict the pharmacokinetics.
Typical media loss assay plots for each compound in rat and associated fits to a two-compartment model (Fig. 1).
Typical media loss assay plots for each compound in dog and associated fits to a two-compartment model (Fig. 1).
In vitro parameters
In vitro and in vivo intrinsic clearance parameters in rat
The correction factor between in vitro and in vivo values is defined as the factor over and above hepatocellularity (117 and 215 million cells/g liver for rat and dog, respectively).
Prediction of In Vivo Parameters.
In general, the media loss assay improves the prediction of clearance in the rat compared with the standard cells and media assay (average fold area 2.4 and 4.2, respectively) (Table 6; Figs. 7 and 8). Moreover, in contrast to the standard assay, the prediction of losartan, ramatroban, telmisartan, and AZ1 are predicted within 2-fold using the media loss assay. Inclusion of uptake tends to have less impact on the predictions of clearance in the dog, which are generally well predicted by both assays, with the exception of sulfisoxazole; this may be due to the difficulty of measuring the low turnover of the compound coupled with relatively high free levels in plasma. However, there is some evidence of a bias toward an overprediction in the correlation with the media loss assay (Fig. 8), especially with dog data. This may be a reflection of the correction factors applied being generated from rat data. Predictions of Vss from the media loss assay correlate well with the measured values (Fig. 9). Most predictions are within 2-fold with the exception of telmisartan in the rat and indomethacin in the dog, which are within 4-fold.
In vivo predictions in rat and dog
Correlation of predicted clearance using the cells + media assay and metabolic clearance calculated from plasma (the solid line represents the line of unity, dotted lines signify 2-fold errors between the predicted and observed values, statistics were calculated using log data). ▵, dog; ▴, rat. RMSE, root mean square error; AFE, average fold error.
Correlation of predicted clearance using the media loss assay and metabolic clearance calculated from plasma (the solid line represents the line of unity, dotted lines signify 2-fold errors between the predicted and observed values, statistics were calculated using log data). ▵, dog; ▴, rat. RMSE, root mean square error; AFE, average fold error.
Correlation of predicted Vss using the media loss assay and Vss calculated from plasma (the solid line represents the line of unity, dotted lines signify 2-fold errors between the predicted and observed values, statistics were calculated using log data). ▵, dog; ▴, rat. RMSE, root mean square error; AFE, average fold error.
Discussion
The results show that when an acid compound is actively sequestered in the liver, both clearance and Vss can be affected. Moreover, scaling methods such as allometry may fail because of species differences in the sequestering of drug within the liver and therefore predictions to human may be inaccurate. For instance, telmisartan and losartan have much larger Vss in rat compared with that in dog even though they have higher plasma protein binding in rat. This can be explained by the greater liver sequestration of drug in rat compared with that in dog. Therefore, an in vitro tool to predict the level of hepatic uptake in an individual species is required to understand discrepancies in species PK parameters and ultimately more accurately predict to man.
The in vivo metabolic clearance of compounds that undergo insignificant hepatic uptake are well predicted using the standard homogeneous sampling (cells + media) methodology (Ito and Houston, 2005; Riley et al., 2005). Whether a compound undergoes active uptake or not, this methodology should always be used for an initial understanding of hepatic metabolic turnover. If the in vivo metabolic clearance is well predicted using the standard cells + media assay, then this should be the method of choice because it has far fewer variables than the in vitro media loss assay coupled with PBPK modeling. However, when an underprediction of clearance occurs because of significant hepatic uptake, then the media loss assay coupled with PBPK modeling has been shown in this work to be successful for predicting the PK of a series of acid compounds with a range of active hepatic uptake in both rat and dog. The overall compound disposition is not related to the magnitude of active uptake alone but in its interplay with other factors such as passive transport and plasma binding. For example, Paine et al. (2008) showed that in the rat indomethacin has a large active uptake component but because of high passive hepatic transport and high plasma binding, the impact on distribution and clearance is low and it behaves like a typical acid compound (metabolic clearance 0.3 ml · min−1 · kg−1, Vss 0.1 l/kg). Furthermore, our data have shown that, in rat, ramatroban has a much lower passive component and lower plasma binding than indomethacin, resulting in a significant impact on clearance and Vss (metabolic clearance 14 ml · min−1 · kg−1, Vss 1.3 l/kg). The data also suggest that uptake will have little overall impact on half-life, because of the concomitant impact on both Cl and Vss, as long as hepatic extraction is not high. However, it does have the potential to raise free levels of substrate compounds in the liver, which may lead to adverse drug reactions.
Various laboratories have published data on the discrepancy between scaled hepatocyte data using hepatocellularity factors and measured metabolic clearance, and in most cases an underprediction of metabolic clearance is observed. We have observed a similar phenomenon for metabolic clearance but no discrepancy for the passive component. However, an in vitro/in vivo mean correction factor over hepatocellularity of approximately 6 is observed for the uptake component and suggests that there is a lower transporter capacity in the isolated hepatocyte relative to the in vivo case. Going forward, further refinement of the uptake correction factors would be desirable, and this should be possible as the number of data sets increase. Moreover, this would enable a more sophisticated statistical approach to model the discrepancy between isolated hepatocytes and the in vivo situation (e.g., regression analysis).
Of importance, the results presented herein suggest that this methodology could be applied to predicting the impact of hepatic uptake on human Vss as well as clearance using human hepatocytes. This is particularly useful when there is no allometric relationship between rat and dog as a result of interspecies differences in liver uptake. In this situation, the use of human hepatocytes combined with PBPK modeling avoids the question of which species PK is most predictive to man.
Authorship Contributions
Participated in research design: Paine and Gardiner.
Conducted experiments: Paine and Gardiner.
Performed data analysis: Paine and Gardiner.
Wrote or contributed to the writing of the manuscript: Paine and Gardiner.
Acknowledgments
We thank Roger E. Fessey for generating the plasma protein binding data.
Footnotes
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.111.039842.
-
ABBREVIATIONS:
- PK
- pharmacokinetic(s)
- CL
- clearance
- Vd
- volume of distribution
- Vss
- volume of distribution at steady state
- PBPK
- physiologically based pharmacokinetics
- CLinc
- clearance from the incubation
- LC-MS/MS
- liquid chromatography/tandem mass spectrometry
- HPLC
- high-performance liquid chromatography
- CLmed
- clearance from the medium
- kmem
- proportionality constant between amount in membrane and concentration in medium
- CLint, met
- unbound metabolic intrinsic clearance
- fmed, ss
- fraction in the medium at steady state
- fuinc
- fraction unbound in the incubation
- CLint, L, pass
- liver unbound passive diffusion intrinsic clearance
- CLint, L, uptake
- liver unbound sinusoidal uptake intrinsic clearance
- CLint, L, efflux
- liver unbound sinusoidal efflux intrinsic clearance
- CLint, L
- liver unbound intrinsic clearance
- CLint, pass
- unbound passive diffusion intrinsic clearance
- CLint, efflux
- unbound sinusoidal efflux intrinsic clearance
- CLint, uptake
- unbound sinusoidal uptake intrinsic clearance
- kL, mem
- proportionality constant between free blood concentration and the amount of drug in the cell membranes of the liver.
- Received March 31, 2011.
- Accepted July 21, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics
