Transport and binding of methotrexate in vivo

The Application of a Physiologically Based Toxicokinetic Model in Health Risk Assessment

Toxicokinetics plays a crucial role in the health risk assessments of xenobiotics. Classical compartmental models are limited in their ability to determine chemical concentrations in specific organs or tissues, particularly target organs or tissues, and their limited interspecific and exposure route extrapolation hinders satisfactory health risk assessment. In contrast, physiologically based toxicokinetic (PBTK) models quantitatively describe the absorption, distribution, metabolism, and excretion of chemicals across various exposure routes and doses in organisms, establishing correlations with toxic effects. Consequently, PBTK models serve as potent tools for extrapolation and provide a theoretical foundation for health risk assessment and management. This review outlines the construction and application of PBTK models in health risk assessment while analyzing their limitations and future perspectives.

A minimal physiologically based pharmacokinetic model for high-dose methotrexate

Purpose

High-dose methotrexate (HDMTX) is administered for the treatment of a variety of malignant tumors. Wide intra- and inter-individual variabilities characterize the pharmacokinetics of MTX, which is mostly excreted renally. HDMTX dosages are prescribed as a function of body surface area whereas dose adjustments depending on renal function are not well defined. We develop a population pharmacokinetic model with a physiological description of renal excretion as the basis for clinical tools able to suggest model-informed dosages and support therapeutic monitoring.

Methods

This article presents a minimal physiologically based pharmacokinetic (PBPK) model for HDMTX, which specifically accounts for individual characteristics such as body weight, height, gender, age, hematocrit, and serum creatinine to provide individualized predictions. The model supplies a detailed and mechanistic description of capillary and cellular exchanges between plasma, interstitial fluid, and intracellular fluid compartments, and focuses on an individualized description of renal excretion.

Results

The minimal PBPK model is identified and validated with a literature dataset based on Chinese patients suffering from primary central nervous system lymphoma. A comparison with a pharmacokinetic model from the literature suggests that the proposed model provides improved predictions. Remarkably, the model does not present any significant bias in a wide range of degrees of renal function.

Conclusion

Results show that model predictions can capture the wide intra- and inter-individual variability of HDMTX, and highlight the role played by the individual degree of renal function. The proposed model can be the basis for the development of clinical decision-support systems for individualized dosages and therapeutic monitoring.

Toxicology testing in drug discovery and development

The primary objective of toxicology studies in the drug discovery process is to evaluate the safety of potential drug candidates. This is accomplished using relevant animal models and validated procedures. The ultimate goal is to translate the animal responses into an understanding of the risk for human subjects. To this end the toxicologist must be aware of the international guidelines for safety evaluation as well as traditional and nontraditional toxicology models. As described in this unit, the typical toxicology profile consists of safety pharmacology, genetic toxicology, acute and subchronic toxicology, absorption, distribution, metabolism, and excretion (ADME) studies, reproductive and developmental toxicity, and an evaluation of carcinogenic potential.

Pharmacokinetics of methotrexate in solid tumors

The transport of methotrexate (MTX) into Walker 256 carcinoma (W256) and hepatoma 5123 (H5123) transplanted in rats was investigated after a pulse injection and continuous infusion of the drug. A mathematical model was developed which adequately described the distribution and transport of MTX in both solid tumors. In H5123 the uptake was limited by the amount of drug carried by plasma (flow-limited transport), but in W256 MTX uptake was limited by the rate at which the drug crossed the tissue barriers (tissue-limited transport). Relative uptake by the solid tumors was almost eightfold more efficient with low than with high doses. MTX concentration in tumor interstitial fluid equilibrated with that of plasma in about 50 hr using a micropore chamber with a diffusion coefficient of 0.5 microm/min as sampling device. MTX concentration was higher in resistant than in responsive tumors.

Pharmacokinetics of YM466, a new factor Xa inhibitor, in rats and dogs

The pharmacokinetics of YM466, a selective inhibitor for factor Xa, was investigated after single intravenous and oral dosing to rats and dogs. After i.v. dosing, plasma YM466 concentration declined in a bi-phasic manner with a terminal elimination half-life of 1.4 h in rats and 0.8 h in dogs. Total plasma clearance values were 884 and 1212 ml/h/kg in rats and dogs, respectively. After oral dosing, plasma YM466 concentrations reached maximum within 2 h and increased in a dose-proportional manner in rats while increase was nonlinear in dogs. The absolute bioavailability of YM466 was 2.7-4.5% in rats, almost constant regardless of the dose levels investigated, while it was 6.9-24.6% in dogs, indicating nonlinear pharmacokinetics. The plasma protein binding of YM466 was 54.7-56.5% in rats and 45.2-49.0% in dogs and almost constant regardless of the concentration. No metabolism of YM466 was observed in an in vitro liver microsome study. These findings suggest that the low bioavailability of YM466 is attributable to the poor absorption not to the extensive metabolism.

Uptake of rosuvastatin by isolated rat hepatocytes: comparison with pravastatin

1. The liver is the target organ for the lipid-regulating effect of rosuvastatin, a new 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor, and liver-selective uptake of this drug is therefore a desirable property. The uptake kinetics of rosuvastatin were investigated and compared with those of pravastatin using isolated rat hepatocytes. 2. Uptake for both drugs involved both active transport and passive diffusion processes. The Michaelis constant (K(m)) of uptake rate for rosuvastatin (9.17 micro M) was approximately half that for pravastatin (16.5 micro M). However, the maximum uptake rate (V(max)) and carrier-mediated uptake clearance (V(max)/K(m)) of rosuvastatin were significantly (p < 0.01) greater than those of pravastatin, and a larger contribution of carrier-mediated uptake clearance to total uptake clearance was shown for rosuvastatin (contribution ratio 0.903 versus pravastatin 0.654). 3. Sodium and chloride ions did not play a significant role in the uptake of rosuvastatin and pravastatin, but the uptake of both drugs was inhibited both by depletion of cellular ATP and by organic anions such as bromosulfophthalein. 4. Rosuvastatin competitively inhibited the uptake of pravastatin, with an inhibition constant (K(i)) (2.75 micro M) relatively similar to its K(m). 5. The results suggest that an organic anion transport protein is the main mediator of the hepatic uptake of rosuvastatin and pravastatin, which occurs in an ATP-dependent manner. Our results indicated that rosuvastatin was taken up by the hepatocytes via the same transport systems as pravastatin, but with a greater affinity and efficiency than pravastatin.

Carrier-mediated active transport of a novel thromboxane A(2) receptor antagonist [2-(4-chlorophenylsulfonylaminomethyl)indan-5-yl]acetate (Z-335) into rat liver

To elucidate the transport system by which [2-(4-chlorophenylsulfonylaminomethyl)indan-5-yl]acetate (Z-335) is taken up into the liver, we investigated the uptake characteristics of Z-335 in isolated rat hepatocytes. In addition, we estimated the hepatic uptake of Z-335 in intact rats under steady-state conditions and compared it with the in vitro uptake clearance. Uptake of Z-335 is highly concentrative (cell-to-medium concentration ratios were 21.2 at 0.5 min and 71.7 at 5 min), temperature-dependent, and sensitive to metabolic inhibitors, indicating that uptake is mediated by energy-dependent uphill transport. In the presence of metabolic inhibitors [carbonyl cyanide p-trifluoromethoxyphenylhydrazone and rotenone], uptake remained at 37 and 49% of the control value, respectively, suggesting that ATP-independent uptake contributes to the total uptake of Z-335. The concentration dependence of the initial uptake velocity indicated a two-component process, one saturable component, with a K(m) value of 45.6 microM and a V(max) value of 4.1 nmol/min/mg of protein, and a nonspecific diffusion clearance, with a P(dif) value of 8.3 microl/min/mg of protein. The contribution of the carrier-mediated uptake to the total uptake in a linear range was estimated as 91%. The in vivo hepatic intrinsic clearance (CL(int, app)) was comparable with that in vitro uptake clearance (PS(influx)) and indicated that the CL(int, app) of Z-335 at steady state is rate-limited by the uptake process. In conclusion, hepatic intrinsic clearance of Z-335 at steady state is rate-limited by the uptake process since Z-335 is efficiently taken up by an active transport mechanism, followed by metabolism or biliary excretion.

Application of the PKCYP-test to predict the amount of in vivo CYP2C11 using tolbutamide as a probe

Previous reports have shown that the determination of drug metabolism capacity can be made by the pharmacokinetic estimation of the quantity of cytochrome P450 (CYP) in vivo (PKCYP-test), in which an apparent liver-to-blood free concentration gradient in vivo (qg) is introduced, which is useful for evaluating fluctuations of CYPIA2 in rats. The aim of the present study was to examine the application of the PKCYP-test to evaluate the quantity of in vivo CYP2C11 by using tolbutamide as a probe, to confirm its validity using a physiologically-based pharmacokinetic rat model. Rats treated with carbon tetrachloride (CCl4-treated rats) were used as a model for low levels of CYP2C11 in the liver. In CCl4-treated rats, the total body clearance (CLtot) of tolbutamide and the amount of CYP2C11 fell to about a quarter and a third of that in control rats, respectively. The time-course of tolbutamide concentrations in serum in control rats could be simulated by a physiologically-based pharmacokinetic model. In CCl4-treated rats, take into consideration the qg value of control rats, the level of CYP2C11 was accurately predicted by the PKCYP-test, and the time-course of tolbutamide concentrations in serum could be predicted by the same physiologically-based pharmacokinetic model. In conclusion, we have shown that the PKCYP-test can be used to predict levels of CYP2C11. It was also demonstrated that the qg and amount of CYP are useful parameters in the PKCYP-test by constructing a physiologically-based pharmacokinetic model which was applied to the PKCYP-test.

Prediction of drug absorption using multivariate statistics

Literature data on compounds both well- and poorly-absorbed in humans were used to build a statistical pattern recognition model of passive intestinal absorption. Robust outlier detection was utilized to analyze the well-absorbed compounds, some of which were intermingled with the poorly-absorbed compounds in the model space. Outliers were identified as being actively transported. The descriptors chosen for inclusion in the model were PSA and AlogP98, based on consideration of the physical processes involved in membrane permeability and the interrelationships and redundancies between available descriptors. These descriptors are quite straightforward for a medicinal chemist to interpret, enhancing the utility of the model. Molecular weight, while often used in passive absorption models, was shown to be superfluous, as it is already a component of both PSA and AlogP98. Extensive validation of the model on hundreds of known orally delivered drugs, "drug-like" molecules, and Pharmacopeia, Inc. compounds, which had been assayed for Caco-2 cell permeability, demonstrated a good rate of successful predictions (74-92%, depending on the dataset and exact criterion used).

Effect of alpha(1)-acid glycoprotein on the pharmacokinetics of tamsulosin in rats treated with turpentine oil

The pharmacokinetics of tamsulosin (TAM) was investigated using male Sprague-Dawley rats in which plasma alpha(1)-acid glycoprotein (alpha(1)-AGP) levels were elevated by the subcutaneous injection of 0.2 mL/kg of turpentine oil. alpha(1)-AGP levels increased about eight times after turpentine oil treatment, causing a threefold decrease in plasma unbound fraction (f(u)) of TAM. When 0.3 mg/kg of TAM was dosed intravenously, total and nonrenal clearances (CL(tot) and CL(nr)) in turpentine-treated rats were 47% and 44% lower than those in nontreated controls, respectively. The area under the concentration-time curve of plasma unbound TAM (AUC(inf,u)) was lower than that in the control. When 1 mg/kg of TAM was dosed orally, oral clearance (CL(oral)) in alpha1-AGP-induced rats was 65% lower than in the control. The AUC(inf,u) and unbound oral clearance (CL(oral,u)) were nearly equal in both groups. Moreover, a positive correlation was observed between fu and CL(oral) of TAM (r(2) = 0.603, P < 0.01), whereas no correlation was observed between f(u) and CL(oral,u). The absolute bioavailability (BA) increased from 19.2% to 46.9% by induction of alpha(1)-AGP. These results suggest that decreased f(u) caused by the elevation of plasma alpha(1)-AGP level affects the pharmacokinetics of TAM, but does not affect the CL(oral,u,) which represents the hepatic metabolism of TAM.

Pharmacokinetic analysis of felodipine-grapefruit juice interaction based on an irreversible enzyme inhibition model

Aims Ingestion of grapefruit juice (GFJ) alters the pharmacokinetics of various orally administered drugs. Quantitative evaluation of this GFJ-drug interaction is required for the proper clinical management of patients. Methods Using felodipine as a model drug, we constructed a pharmacokinetic model based on irreversible inhibition of intestinal cytochrome P450 3A4 (CYP3A4) by GFJ. We fitted previously published data [5, 6] for felodipine ER (extended release formulation) to the ratio of CLGI,int before and after grapefruit juice ingestion by nonlinear least-squares regression analysis to estimate the reaction rate constant between GFJ and CYP3A4 (K) and the elimination rate constant of CYP3A4 (k ).

RESULTS

The model gave a turnover rate of CYP3A4 of 0.0849 h-1, corresponding to a half-life of 8.16 h, in agreement with reported values. The AUC-time profiles of felodipine ER in the case of different amounts and schedules of GFJ ingestion were simulated using the parameter values estimated from the model.

CONCLUSIONS

The modelling leads to the important conclusion that GFJ-felodipine interaction increases with increasing frequency and amount of GFJ ingestion, and that an interval of 2-3 days between GFJ intake and felodipine administration is necessary if GFJ-felodipine interaction is to be avoided.

Correlation of the intrinsic clearance of donepezil (Aricept) between in vivo and in vitro studies in rat, dog and human

1. Donepezil hydrochloride (Aricept) is used for the treatment of Alzheimer's disease. Here the correlation of the intrinsic clearance (Cl(int)) of donepezil between the in vivo and in vitro states was studied in rat, dog and human. 2. In an experiment with 14C-donepezil and human microsomes the routes of metabolism were identified as N-dealkylation and O-demethylation, and no unknown metabolites were detected. 3. The Cl(int) of donepezil in the male rat, female rat, dog and human liver microsomes were 33.7, 13.4, 37.0 and 6.35 microl/min/mg microsomal protein respectively, and sex difference in rat and interspecies difference in the estimated Cl(int) were found. 4. After a single intravenous administration to the male rat, female rat and dog, total plasma clearance (ClP(total)) was 78.6, 29.5 and 88.3 ml/min/kg respectively, and a sex difference was observed in rat. 5. After a single oral administration to the male rat, dog and healthy volunteer, ClP(total) was 140, 105 and 2.35 ml/min/kg respectively, and remarkable differences were observed between animals and man. 6. The contribution of renal clearance to blood clearance (Cl(r)) was low in all species. The predicted in vitro hepatic clearance (Cl(h-pre)) was in the rank order: male rat (15.91 ml/min/kg) > dog (7.96) > female rat (7.67) > human (1.04). Although Cl(h-pre) was underestimated, Cl(h-pre) was significantly correlated with that of ClB(total) in the different animal species and in man, indicating that the in vitro-in vivo ranking order was conserved.

Role of pharmacokinetics in the discovery and development of indinavir

The discovery of indinavir is a successful example in which pharmacokinetic and metabolic information were incorporated into drug design. The use of animal and in vitro human metabolic data in predicting the oral bioavailability and hepatic clearance in humans was critical in selecting indinavir as a drug candidate for development. In its development stage, pharmacokinetics continued to play an important role in identifying the key properties of indinavir in vivo, which allowed the characterization and prediction of the time course of drug action under physiological and pathological conditions. This review describes the role of pharmacokinetics and drug metabolism in the discovery and development of indinavir.

Differences in the hepatobiliary transport of two quinolone antibiotics, grepafloxacin and lomefloxacin, in the rat

The biliary excretion of grepafloxacin (GPFX) was compared with that of lomefloxacin (LFLX) in rats. The biliary clearances (Cl(plasma)(bile)) of GPFX was 2.9 times greater than LFLX based on the plasma concentration reached during constant intravenous (i.v.) infusion. The liver-plasma unbound concentration ratio, K(pu), of GPFX (1.7) was also higher than that of LFLX (0.7). The hepatic uptake clearance, assessed from an integration plot analysis, of GPFX was comparable with the hepatic blood flow rate, and 1.5 times that of LFLX, indicating that membrane transport in the uptake process is more efficient for GPFX. This was also supported by the difference between the uptake clearance of GPFX and LFLX in isolated rat hepatocytes. The bile-liver unbound concentration ratio of GPFX and LFLX was approximately 6 and 3, respectively, and the biliary clearance based on the unbound liver concentration of GPFX was 1.8 times that of LFLX. These results suggest that the concentrative transport of GPFX also across the canalicular membrane was more efficient than that of LFLX. Thus, the membrane transport activity via both sinusoidal and canalicular membranes determines the net excretion of each compound.

Determination of parameters responsible for pharmacokinetic behavior of TCDD in female Sprague-Dawley rats

2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is the most toxic member of a class of planar and halogenated chemicals. Improvements in exposure assessment of TCDD require scientific information on the distribution of TCDD in target tissues and cellular responses induced by TCDD. Since 1980, several physiologically based pharmacokinetic (PBPK) models for TCDD and related compounds have been reported. Some of these models incorporated the induction of a hepatic binding protein in response to interactions of TCDD, the Ah receptor, and DNA binding sites and described the TCDD disposition in a biological system for certain data sets. Due to the limitations of the available experimental data, different values for the same physical parameters of these models were obtained from the different studies. The inconsistencies of the parameter values limit the application of PBPK models to risk assessment. Therefore, further refinement of previous models is necessary. This paper develops an improved PBPK model to describe TCDD disposition in eight target tissues. The interaction of TCDD with the Ah receptor and with hepatic inducible CYP1A2 were also incorporated into the model. This model accurately described the time course distribution of TCDD following a single oral dose of 10 microg/kg, as well as the TCDD concentration on Day 3 after six different doses, 0.01, 0.1, 0.3, 1, 10, and 30 microg TCDD/kg, in target tissues. This study extends previous TCDD models by illustrating the validity and the limitation of the model and providing further confirmation of the potential PBPK model for us in optimal experimental design and extrapolation across doses and routes of exposure. In addition, this study demonstrated some critical issues in PBPK modeling.

Human immunodeficiency virus protease inhibitors. From drug design to clinical studies

The discovery of human immunodeficiency virus (HIV) protease inhibitors is an example in which pharmacokinetic evaluation was implemented early in the discovery phase to obtain optimal pharmacological and pharmacokinetic properties. Currently, three HIV protease inhibitors, saquinavir, indinavir and ritonavir are clinically available. As a family, these HIV protease inhibitors are characterized pharmacologically by their ability to inhibit the viral protease enzyme. Pharmacokinetically, they are quite different due to their dissimilarity in physicochemical properties. Bioavailability appears to be limited with saquinavir, but not with indinavir and ritonavir. Although all three drugs are metabolized extensively by cytochrome P-450, saquinavir and indinavir are high clearance drugs while ritonavir is a low clearance drug. Despite their significant differences in elimination clearance, all three HIV proteases are given at high oral doses (600-800 mg) either twice or three times daily. These HIV protease inhibitors show superior therapeutic activity and a more favorable safety profile than those reported for the established reverse transcriptase inhibitors. However, the potential for interactions with other drugs metabolized by the CYP 3A4 isoform appears to be considerable. In addition, repeated administration of enzyme inducers results in a substantial decrease of plasma concentrations of protease inhibitors. Therefore, co-administration of drugs, such as rifampicin and rifabutin, must be avoided. HIV protease inhibitors are promising in the treatment of AIDS. Although they are not a cure, they can significantly inhibit that viral replication and improve the quality of life for people who have HIV infection.

Pharmacokinetic studies of human urinary kininogenase in healthy volunteers and animals

Plasma concentrations of human urinary kininogenase (HUK) were determined in healthy volunteers during and after intravenous (iv) infusion by enzyme immunoassay (EIA). Plasma kinin concentrations were also determined by radioimmunoassay (RIA), and related to HUK concentrations. When HUK was infused [at 0.04, 0.075, 0.15, and 0.3 p-nitroaniline unit (PNAU)/body] over 30 min, plasma HUK concentration rapidly increased and reached a maximum at the end of dosing. Then, the concentration of HUK in plasma decreased biexponentially, and the elimination half-life of the terminal phase was found to be approximately 170 min. The area under the curve of concentration versus time from 0 to 180 min (AUC0-180min) and the maximum concentration (Cmax) increased in proportion to the dose, whereas the pharmacokinetic parameters [mean residense time (MRTinf) = 200-270 min, plasma clearance (CLp) = 2.5-3.3 mL/min/kg, volume of distribution at steady state (Vdss) = 470-730 mL/kg] did not very significantly within the dose range of the present study. On the other hand, when HUK was infused (at 0.15 PNAU/body), plasma kinin concentrations reached approximately 2 ng of bradykinin eq/mL 15 min after the onset of administration. This concentration was maintained during the dosing period, after which kinin was rapidly eliminated, and its concentration returned to baseline at 10 min after dose withdrawal. Plasma kinin concentrations at 15 to 30 min after the onset of dosing (at 0.075, 0.15, and 0.3 PNAU/body) increased in proportion to the dose. The pharmacokinetic parameters of HUK (MRTinf, CLp, Vdss) were compared with those of rats, rabbits, and dogs (log-log plots of body weight versus MRTinf, CLp, and Vdss). The Vdss value showed a good correlation (r = 0.996 for n = 4) with the body weight of respective animal species, the correlation with CLp was weak (r = 0.911), and MRTinf did not exhibit any correlation.

Inverse targeting of peritoneal tumors: selective alteration of the disposition of methotrexate through the use of anti-methotrexate antibodies and antibody fragments

We have hypothesized that antidrug antibodies (ADAb) may be employed to impart site-specific alterations in the disposition of drug molecules, potentially allowing for targeted drug therapy. We are specifically interested in minimizing systemic exposure to free drug and systemic toxicities resultant from regional chemotherapy through the intravenous administration of ADAb. In this report, we present the production and purification of anti-methotrexate Fab fragments, and we present investigations of the effects of anti-methotrexate Fab and anti-methotrexate immunoglobulin G on the disposition of methotrexate in the rat. Pharmacokinetic studies revealed that intravenous anti-methotrexate immunoglobulin G (anti-MTX IgG) and anti-methotrexate Fab (anti-MTX Fab) administration produced dramatic alterations in the plasma pharmacokinetics of methotrexate (MTX), following intraperitoneal MTX administration (area under the total MTX concentration vs time curve for anti-MTX IgG relative to control, 420 +/- 90 (p < 0.05); for anti-MTX Fab relative to control, 46 +/- 6.1 (p < 0.05); area under the free MTX concentration vs time curve for anti-MTX IgG relative to control, 0.64 +/- 0.16; for anti-MTX Fab relative to control, 0.45 +/- 0.20 (p < 0.05)). Additional studies conducted in anesthetized rats revealed no significant alterations in the area under the total peritoneal MTX concentration vs time curves, free MTX peritoneal concentration vs time curves, or peritoneal exit rate of MTX in anti-MTX Fab treated animals relative to controls. Therefore, our pharmacokinetic studies demonstrate that ADAb may produce site-specific alterations in drug pharmacokinetics, potentially enhancing the site specificity of drug distribution and drug action following regional chemotherapy.

Kinetic analysis of receptor-mediated endocytosis of G-CSF derivative, nartograstim, in rat bone marrow cells

To elucidate the mechanism of the receptor-mediated clearance of granulocyte colony-stimulating factor (G-CSF), we performed kinetic analyses of the receptor-mediated endocytosis (RME) processes using a human G-CSF derivative, nartograstim (NTG), and isolated rat bone marrow cells. The first-order rate constants involved in RME processes were obtained by computerized model fitting of the time courses of the ligand-receptor complex on both the cell surface and in the cell interior and the degradation products in the medium in the pulse-chase experiment. They were also calculated based on a kinetic model involving the ligand concentration dependence of the initial binding rate, the steady-state degradation rate, and the steady-state amounts of ligand on both the cell surface and in the interior. The rate constants for the RME processes after receptor binding determined in the different experiments were similar, that is, the half-times for the dissociation, internalization, and degradation of the ligand-receptor complex were 770, 10-30, and 20 min, respectively. However, the association constant obtained by measuring the initial binding was fivefold greater than that calculated under steady-state conditions. These kinetic analyses support the hypothesis that the internalization of the receptor may be accelerated by ligand binding, causing downregulation of the receptor on the cell surface. These overall kinetic analyses based on steady-state and non-steady-state data of the RME processes clarify the dynamics of the interaction between NTG and its receptor.

Superoxide dismutase derivative prevents oxidative damage in liver and kidney of rats induced by exhausting exercise

To prevent oxidative tissue damage induced by strenuous exercise in the liver and kidney superoxide dismutase derivative (SM-SOD), which circulated bound to albumin with a half-life of 6 h, was injected intraperitoneally into rats. Exhausting treadmill running caused a significant increase in the activities of xanthine oxidase (XO), and glutathione peroxidase (GPX) in addition to concentrations of thiobarbituric acid-reactive substances (TBARS) in hepatic tissue immediately after running. There was a definite increase in the immunoreactive content of mitochondrial superoxide dismutase (Mn-SOD) 1 day after the running. Meanwhile, the TBARS concentration in the kidney was markedly elevated 3 days after running. The activities of GPX, and catalase in the kidney increased significantly immediately and on days 1 and 3 following the test. The immunoreactive content of Mn-SOD also increased 1 day after running. The exercise induced no significant changes in immunoreactive Cu, Zn-SOD content in either tissue. The administration of SM-SOD provided effective protection against lipid peroxidation, and significantly attenuated the alterations in XO and all the anti-oxidant enzymes, measured. In summary, the present data would suggest that exhausting exercise may induce XO-derived oxidative damage in the liver, while the increase in lipid peroxidation in the kidney might be the result of washout-dependent accumulation of peroxidised metabolites. We found that the administration of SM-SOD provided excellent protection against exercise-induced oxidative stress in both liver and kidney.

Physiologically based pharmacokinetic models: mathematical fundamentals and simulation implementations

This review paper gives an overview of the building blocks of physiologically based pharmacokinetic (PBPK) models and their implementation using computer facilities. The approach focuses on the development of a PBPK model with the most important and appropriate limiting steps for the conditions and exposure scenarios under study. In this approach, the assumptions made in constructing the set of equations, as well as the fitting of variables to specific experimental results, need to be accounted for when making extrapolation to other conditions. A well-constructed PBPK model should account for all possible ranges of extrapolation from the development stages so that appropriate experimental studies and assumptions can be designed to handle the intended applications. Two common assumptions are revisited: the flow-limited assumption and the metabolic clearance using Michaelis-Menten kinetics assumption. Computer hardware and software requirements for implementing PBPK models are briefly reviewed.

A theoretical approach to the estimation of tissue flows using tritiated water as indicator

An alternative method for the determination of flow rates to individual organs or tissues that collectively comprises experimental isolated preparations is described. The isolated hindlimb constitutes a good example of such an experimental preparation. After injecting a bolus dose of tritiated water to the arterial side, the amount of marker (tritiated water) remaining in different tissues at various time is determined. From these experimental data, the regional flow rate to individual tissues can be estimated, applying either a physiological model approach or statistical moment theory.

Dose-dependent pharmacokinetics: experimental observations and theoretical considerations

Clinically, absorption and elimination of most drugs follow linear kinetics, and pharmacokinetic parameters describing absorption and elimination of a drug do not change over the therapeutic dose range. However, dose-dependent pharmacokinetics have been reported more frequently in preclinical studies, particularly in toxicity studies, where high doses are often employed. This review highlights the major types of dose-dependent pharmacokinetics with unique examples. Before setting out on a pivotal subchronic and chronic toxicity study of a new drug, a pilot study is often performed to establish a dose range in which a reasonable relationship between plasma AUC and dosage exists to ensure sufficient exposure of animals to the drug. Theoretical bases and possible causes of dose-AUC disproportionality are discussed. Factors affecting the distribution and elimination of drugs and causes of dose-dependent tissue distribution and elimination are also discussed. Often, the non-linear kinetics complicate the design of dosage regimens and prediction of efficacy and toxicity. Thus, an understanding of the influence of dose on the pharmacokinetics is important in the evaluation of the efficacy and toxicity of new drugs.

Distinction between the depletion of opsonins and the saturation of uptake in the dose-dependent hepatic uptake of liposomes

Opsonins play a role in the hepatic uptake of particles such as bacteria, lipid emulsion, and liposomes. The objective of this study was to distinguish between opsonin depletion and uptake saturation in the dose-dependent hepatic uptake of liposomes. The uptake of opsonized and unopsonized liposomes was determined in the isolated perfused liver. Serum (2.9 mL) was required to opsonize 1 mumol liposomes fully, indicating that a rat (250 g with 10 mL of serum) can opsonize 3.5 mumol liposomes. Next the dose effect on hepatic uptake of opsonized and unopsonized liposomes was examined. Saturation of uptake was found only for the opsonized liposomes. On the other hand, the hepatic uptake clearance decreased dose dependently from 4.31 to 0.79 (mL/min), with increasing doses from 0.075 to 17 mumol/250 g, respectively, after i.v. administration. Thus, the decrease in the hepatic uptake clearance at the medium dose was due to the saturation of uptake alone, and at the high dose it was due to opsonin depletion as well. These results show that the saturation of liposomal uptake in the liver and the depletion of opsonins occurred at different liposome dosage levels.

Decreased systemic clearance of diltiazem with increased hepatic metabolism in rats with uranyl nitrate-induced acute renal failure

The effect of uranyl nitrate (UN)-induced acute renal failure (ARF) on the pharmacokinetics of diltiazem (DTZ) was examined in rats through in vitro and in vivo studies. In vitro homogenate studies demonstrated that DTZ was metabolized to deacetyl diltiazem (DAD) predominantly in the liver. Metabolism in the small intestine, kidney, or blood pool was negligible compared with that in the liver. UN-induced ARF (UN-ARF) increased the in vitro hepatic clearance (CLvit) of DTZ 1.4-fold. In vivo pharmacokinetic studies following intravenous (iv) and portal venous (pv) administration revealed that UN-ARF increased the intrinsic clearance (CLi) of DTZ from 243.0 to 414.5 ml/min/kg but decreased its total plasma clearance (CLt) from 90.3 to 64.3 ml/min/kg. The increase in CLi was consistent with the increase in CLvit of the liver. The in vitro plasma free fraction of DTZ (fp) was decreased from 0.25 to 0.14 by UN-ARF, but the in vitro blood/plasma partition of DTZ (Rb) remained constant at unity. From the CLi and fp changes, the plasma intrinsic clearance for unbound DTZ (CLi') was calculated to be increased 2.7-fold, from 1104.5 to 2960.7 ml/min/kg, by UN-ARF. The fp decrease was also reflected in the steady-state distribution volume (Vdss) of DTZ, which was decreased significantly from 3595.5 to 2528.3 ml/kg. The absolute bioavailability of pv DTZ (Fpv) was decreased by UN-ARF from 37.5 to 15.5% but was still much higher than the reported oral bioavailability (6%), indicating poor absorption of DTZ from the GI tract. From the calculation based on a well-stirred pharmacokinetic model, DTZ was found to increase the hepatic blood flow (HBF) of the control rats more than twofold at doses of 3 mg/kg (iv) or 10 mg/kg (pv), possibly due to the vasodilating effect of DTZ. However, the effect of DTZ on HBF was not present in the UN-ARF rats. It is not clear at present whether this could be attributed to vasoconstricting effects of UN-ARF or blockade of the vasodilating effect of DTZ.

Prediction of the volume of distribution of 7-hydroxycoumarin in man from in vitro and ex vivo data obtained in rat

The essential parameter to estimate the first dose size of a drug in man is the volume of distribution. For a drug that has never been used in man before, estimates of the volume of distribution can only be obtained from animals and in vitro data. The purpose of this study was to compare various approaches presented in the literature for predicting the volume of distribution at steady state (VSS) and the terminal phase volume of distribution (Vd beta) in man. A lipophilic active metabolite of coumarin, 7-hydroxycoumarin (7OHC), was selected for this investigation. This compound is extensively metabolized in both the central and peripheral compartments. Of the six methods evaluated, only an empirical allometric approach yielded a reasonable estimate of VSS. All methods underestimated VSS and none of the applicable methods were able to predict Vd beta. The reason for this discrepancy may be due to the fact that the calculation of VSS in man was done assuming elimination from the central compartment.

Stereoselective disposition and tissue distribution of carvedilol enantiomers in rats

After intravenous bolus injection of rac-carvedilol at 2 mg/kg to the rat, the (+)-(R)- and (-)-(S)-enantiomer levels in the blood and tissues (liver, kidney, heart, muscle, spleen, and aorta) were measured by stereospecific HPLC assay. As compared with the (+)-(R), the (-)-(S) had a larger Vdss (3.32 vs. 2.21 liter/kg), MRT (33.4 vs. 25.6 min), and CLtot (96.1 vs. 83.8 ml/min/kg). AUC comparison after iv and po administration showed systemic bioavailability of the (-)-(S) to be about half that of its antipode, explained by the fact that the free fraction of the (-)-(S) in blood was 1.65-fold greater than that of the (+)-(R). Tissue-to-blood partition coefficient values for the (-)-(S) were 1.6- to 2.1-fold greater than those for the (+)-(R) in all tissues, showing that the (-)-(S) accumulates more extensively in the tissues. These results were consistent with the greater Vdss for the (-)-(S) estimated from systemic blood data. The stereoselective tissue distribution of carvedilol enantiomers results from an enantiomeric difference in plasma protein binding rather than in tissue binding.

Physiologically based models for bone-seeking elements. III. Human skeletal and bone growth

A model of skeletal and bone growth for the human from birth to maturity has been developed. Dry and hydrated bone, bone marrow, ash, and calcium are included in the model. Growth of the skeleton and its fractions is expressed as a set of allometric equations relating fraction volume or weight to body weight. Blood flow rates to mature bone and bone marrow are scaled from experimentally determined values in smaller animals, but bone and marrow volumes and growth patterns cannot be scaled directly from measurements or models in small animals. The growth model compares well with measured bone weights, ash weights, and bone and skeletal densities in humans. Its form is adaptable to physiologically based description of the kinetics of bone-seeking elements.

A biologically based toxicokinetic model for pyrene in rainbow trout

A biologically based toxicokinetic model was developed to stimulate the metabolic disposition of pyrene in trout with an average body weight of 450 g and dosed with a single bolus injection of the chemical (10 mg/kg). The model consists of a membrane-limited muscle compartment and six flow-limited compartments including the gills, liver, gut, kidney, carcass, and blood. The compartments are represented by mass balance equations including terms for the binding of pyrene to tissue and blood proteins, biotransformation, penetration rate into the muscle, blood flow rate, tissue mass, etc. The model also provides for nonsaturable and saturable clearances of pyrene by the liver and kidney. Michaelis-Menten constants for pyrene metabolism (Km, Vmax) were determined from in vitro experiments using isolated liver cells. Renal clearance of pyrene was very close to the glomerulus filtration rate of trout. Solution of the system of equations yielded the time courses of pyrene concentration in the tissues. Predicted concentrations of pyrene in the gills, liver, gut, kidney, muscle, and blood were consistent with experimental observations for at least 6 days. The model was validated by comparing the model predicted and experimental results of trout weighing 285 g and dosed with a single intraarterial dose (3 mg/kg) of pyrene. The predicted pyrene concentrations also were in adequate agreement with the empirical data.

Moment method for the estimation of mass transfer coefficients for physiological pharmacokinetic models

In vitro and in vivo techniques have been utilized to estimate mass transfer coefficients for physiological pharmacokinetic models. No single method has been adopted for estimating this parameter, in part, due to the different model structures with which this parameter may be associated. A specific method has been derived to calculate mass transfer coefficients for non-eliminating membrane-limited tissue compartments. The present method is based on observed concentration-time data, and requires the calculation of the areas under the zero and first moment curves for plasma, and the first moment curve for the tissue. A Monte Carlo simulation technique was used to determine the percentage biases of the method based on a published model for streptozoticin and adriamycin. For the latter model, the method was compared to a non-linear regression parameter estimation technique.

Cyclopentenyl cytosine: interspecies predictions based on rodent plasma and urine kinetics

A hybrid compartmental-physiological model for cyclopentenyl cytosine (CPE-C) is designed on the basis of early limited rodent pharmacokinetic data. Application of model independent pharmacokinetics and biochemical knowledge was first used to conceptualize such a model. The approach was to scale the physiological parameters of the model (compartmental clearances) and keep constant the anatomic parameters of the model (compartment volumes). Scaling of physiological mechanisms was based on body weight/surface area ratios. Using these principles, simulations with the model can reasonably anticipate the in vivo behavior of (CPE-C) in several species (mouse, rat, dog). The model is useful in understanding species differences in pharmacokinetic behavior of CPE-C.

A semiparametric approach to physiological flow models

By regarding sampled tissues in a physiological model as linear subsystems, the usual advantages of flow models are preserved while mitigating two of their disadvantages, (i) the need for assumptions regarding intratissue kinetics, and (ii) the need to simultaneously fit data from several tissues. To apply the linear systems approach, both arterial blood and (interesting) tissue drug concentrations must be measured. The body is modeled as having an arterial compartment (A) distributing drug to different linear subsystems (tissues), connected in a specific way by blood flow. The response (CA, with dimensions of concentration) of A is measured. Tissues receive input from A (and optionally from other tissues), and send output to the outside or to other parts of the body. The response (CT, total amount of drug in the tissue (T) divided by the volume of T) from the T-th one, for example, of such tissues is also observed. From linear systems theory, CT can be expressed as the convolution of CA with a disposition function, F(t) (with dimensions 1/time). The function F(t) depends on the (unknown) structure of T, but has certain other constant properties: The integral integral infinity0 F(t) dt is the steady state ratio of CT to CA, and the point F(0) is the clearance rate of drug from A to T divided by the volume of T. A formula for the clearance rate of drug from T to outside T can be derived. To estimate F(t) empirically, and thus mitigate disadvantage (i), we suggest that, first, a nonparametric (or parametric) function be fitted to CA data yielding predicted values, CA, and, second, the convolution integral of CA with F(t) be fitted to CT data using a deconvolution method. By so doing, each tissue's data are analyzed separately, thus mitigating disadvantage (ii). A method for system simulation is also proposed. The results of applying the approach to simulated data and to real thiopental data are reported.

Age-related change in tissue-to-plasma partition coefficient of cefazolin for noneliminating organs in the rat

Age-related changes in tissue distribution characteristics of cefazolin, a cephalosporin antibiotic, were examined for noneliminating organs of rats. The in vivo tissue-to-plasma partition coefficients (Kp,vivo) varied markedly among different ages and organs. In particular, muscle and skin acted as reservoirs for cefazolin distribution. There were also marked differences in interstitial fluid space (IS), determined using [14C]inulin, among different ages and organs. For muscle and bone, the magnitude of the age-related changes in Kp,vivo of cefazolin and IS was in the order of 1-week-old greater than 7-week-old = 100-week-old greater than 50-week-old rats. This is well correlated with the age-related changes in the volume of distribution at the steady state of cefazolin per body weight (Vdss/BW) and the extracellular fluid volume per body weight (Vecw/BW) determined previously using [14C]inulin. The predicted Kp value (Kp,pred) was estimated by incorporating the serum protein binding parameters of cefazolin, the IS values, and an interstitial-to-plasma albumin concentration ratio (AR) into equations derived from an extracellular fluid model. The Kp,pred values exhibited a fairly good correspondence with the Kp,vivo values determined for various organs, except gut, in rats of all four ages. These results suggest that the determinant of the age-related change in Vdss/BW is the difference in the IS value of muscle and bone, while the age-related change in serum protein binding plays only a modest role.

Kinetic analysis of the elimination process of human epidermal growth factor (hEGF) in rats

Pharmacokinetic study of human epidermal growth factor (hEGF) in rats was performed in vivo. The hepatic extraction ratio (EH) of [125I]hEGF, determined from the difference between the artery and the hepatic vein plasma concentrations at steady state, was 0.19. The hepatic clearance (CLH:7.56 ml/min/kg body wt), calculated by multiplying EH by the hepatic plasma flow rate (QP,H), was approximately 70% of the total body clearance (CLtot: 10.8 ml/min/kg body wt), which was determined from the steady-state arterial plasma concentration and the infusion rate. These results indicated that the liver is the main organ responsible for the removal of [125I]hEGF from the systemic circulation in rats. The renal extraction ratio (ER) of [125I]hEGF was half of that of [14C]inulin; this may have resulted from the plasma protein binding of [125I]hEGF, which was approximately 50% as determined by the charcoal adsorption method and the equilibrium gel-filtration method. The renal clearance (CLR:2.65 ml/min/kg body wt), calculated by multiplying ER by the renal plasma flow rate (QPR), was approximately 17% of the CLtot (15.6 ml/min/kg body wt), indicating a minor contribution of CLR to CLtot compared with that of CLH to CLtot. The CLR of [125I]hEGF calculated from the urinary excretion data was one-tenth of that calculated from the plasma concentration difference between the femoral artery and the renal vein at steady state. These results suggest that the bulk of [125I]hEGF cleared from the plasma by the kidney may have been metabolized further in the renal tubules before appearing in the urine.

Physiological pharmacokinetic models: some aspects of theory, practice and potential

Models are intellectual constructs that pattern selected relationships among the elements of one system to correspond in some way to elements of a second system. In pharmacokinetics, physiological models provide a clearly articulated, rational, explanatory basis for the integration of empirical data; they do this by partitioning the biological system into relevant components (tissues, organs, etc.) and linking them together through the circulatory system. Unlike conventional mammillary compartment models, there is a clear correspondence between model system elements and physiological entities. By virtue of their high degree of physical and biochemical relevance, these models can help provide deep insight into structure, function and mechanism. Pharmacokinetic (and potentially pharmacodynamic) response-time relationships can thus be understood in terms of interconnections and behavior of constituent subsystems. At their worst, these models provide stale or infertile views of reality and thus frustrate and alienate us with the triviality of their insights. At their best, they allow us to understand the accumulation of thought in pharmacokinetics and pharmacodynamics, and help with the integration of data and improvement of experimental design.

A simulation study on the effect of a uniform diffusional barrier across hepatocytes on drug metabolism by evenly or unevenly distributed uni-enzyme in the liver

The effect of a uniform diffusional barrier on hepatic extraction of the parent drug by evenly or unevenly distributed uni-enzyme was quantitatively determined by the present simulation study. Five models of enzymic distribution were defined with regard to the hepatic blood flow path, and the extraction ratios were calculated or simulated under the various conditions of average intrinsic clearances and diffusion clearances across hepatocytes. Differences in the extraction ratios among the five models were evaluated by the "relative extraction ratios," which are the extraction ratios in each model divided by that in the model where the enzymatic activity is evenly distributed. It was found that when a diffusion clearance was high compared to the intrinsic clearance, enzymic distribution was not an important determinant of the extent of hepatic extraction. By contrast, when a diffusional barrier across hepatocytes exists, i.e., the diffusion clearance is low or intermediate compared to the intrinsic clearance, extraction ratios differed widely among the models of enzymic distribution, especially at intermediate average intrinsic clearances. In the presence of a diffusional barrier, the more skewed the distribution of the enzymatic activity is, the lesser the amount of drug eliminated at steady state. The most efficient metabolism occurred when the enzymatic activity was evenly distributed.

Extravascular transport in normal and tumor tissues

The transport characteristics of the normal and tumor tissue extravascular space provide the basis for the determination of the optimal dosage and schedule regimes of various pharmacological agents in detection and treatment of cancer. In order for the drug to reach the cellular space where most therapeutic action takes place, several transport steps must first occur: (1) tissue perfusion; (2) permeation across the capillary wall; (3) transport through interstitial space; and (4) transport across the cell membrane. Any of these steps including intracellular events such as metabolism can be the rate-limiting step to uptake of the drug, and these rate-limiting steps may be different in normal and tumor tissues. This review examines these transport limitations, first from an experimental point of view and then from a modeling point of view. Various types of experimental tumor models which have been used in animals to represent human tumors are discussed. Then, mathematical models of extravascular transport are discussed from the prespective of two approaches: compartmental and distributed. Compartmental models lump one or more sections of a tissue or body into a "compartment" to describe the time course of disposition of a substance. These models contain "effective" parameters which represent the entire compartment. Distributed models consider the structural and morphological aspects of the tissue to determine the transport properties of that tissue. These distributed models describe both the temporal and spatial distribution of a substance in tissues. Each of these modeling techniques is described in detail with applications for cancer detection and treatment in mind.

Analysis of nonlinear tissue distribution of quinidine in rats by physiologically based pharmacokinetics

The nonlinear tissue distribution of quinidine in rats was investigated by a physiologically based pharmacokinetic model. Serum protein binding of quinidine showed a nonlinearity over the in vivo plasma concentration range. The blood-to-plasma concentration ratio (Cb/Cp) of quinidine also showed a concentration dependence. The steady-state volume of distribution (Vss) determined over the plasma concentration range from 0.5 to 10 micrograms/ml was 6.0 +/- 0.45 L/kg. The tissue-to-plasma partition coefficient (Kp) of muscle, skin, liver, lung, and gastrointestinal tract (GI) showed a nonlinearity over the in vivo plasma concentration range of quinidine, suggesting saturable tissue binding. The concentration of quinidine in several tissues and plasma was predicted by a physiologically based pharmacokinetic model using in vitro plasma protein binding and the Cb/Cp of quinidine. The tissue binding parameters were estimated from in vivo Kp values. The predicted concentration curves of quinidine in each tissue and in plasma showed good agreement with the observed values.

Physiological model for the pharmacokinetics of cis-dichlorodiammineplatinum (II) (DDP) in the tumored rat

A physiological model has been developed to describe the disposition of cis-dichlorodiammine-platinum(II) (DDP) following i.v. dosing in the female rat bearing the Walker 256 carcinoma. The model simulates concentrations of DDP and its mobile and fixed metabolites in plasma, liver, gut, skin, muscle, tumor, carcass, and kidney, and DDP and mobile metabolite excretion following a 4 mg/kg dose. In the kinetic model, DDP binds irreversibly to low MW nucleophiles and macromolecules (largely proteins) within the plasma and tissue compartments to form mobile and fixed metabolites, respectively. Reaction rates for the formation of each metabolite are tissue/organ specific. The rate constant for the biotransformation of DDP to fixed metabolite in plasma (k2P = 0.0082 min-1) was determined from in vitro incubation studies. This rate was used as the basis for estimating the biotransformation rate constants for DDP to fixed and mobile metabolites in other compartments. Both DDP and mobile metabolite are assumed to follow flow-limited transport, to freely traverse compartmental barriers, and to partition equally in all compartments. Both are excreted in the urine, the major route of Pt elimination. Urinary excretion is modeled as a linear process involving filtration only; an assumption based on a calculated renal clearance of 1.1 ml/min, a value very similar to the estimated GFR. Biliary excretion is a minor route of mobile metabolite elimination and is modeled as a linear process occurring in the liver. Four hours after dosing, approximately 60% of the administered Pt remains in the tissues and plasma. Of this, over 75% of the plasma Pt and 90% of the metal ion in every other compartment is fixed (protein bound). Fixed Pt can be eliminated from a compartment only after its biotransformation to mobile metabolite. In most compartments this rate of elimination corresponds closely to the average rate of protein turnover in that compartment.