Prediction of Human Pharmacokinetics

Abstract

The prediction of human pharmacokinetics is an extremely difficult endeavor during the selection of drug candidates for further human clinical testing. Despite a variety of available in vitro and in vivo methodologies, successful predictions are still difficult when performing them prospectively. This chapter gives a general overview of in vitro and in vivo methodologies used to predict human pharmacokinetics.

Contents

7.1 List of Abbreviations 127

7.2 Basic Concepts 128

7.3 Prediction of Human Fraction Absorbed 129

7.4 Prediction of Human Clearance 130

7.5 Human Volume of Distribution Prediction 139

7.6 Physiologically Based Pharmacokinetic Models 141

7.7 Context of Confidence in Human

Pharmacokinetic Prediction 142

References 143

Additional Reading 144

7.1 LIST OF ABBREVIATIONS

ADME

Absorption, distribution, metabolism, and excretion

BrW

Brain weight

CLhepatic

Hepatic clearance

CLint

Intrinsic clearance

CLu

Unbound clearance

F

Bioavailability

Fa

Fraction absorbed from the intestine

S.C. Khojasteh et al., Drug Metabolism and Pharmacokinetics 127 Quick Guide, DOI 10.1007/978-1-4419-5629-3_7, © Springer Science+Business Media, LLC 2011

S.C. Khojasteh et al., Drug Metabolism and Pharmacokinetics 127 Quick Guide, DOI 10.1007/978-1-4419-5629-3_7, © Springer Science+Business Media, LLC 2011

Fg Fh fu fumic fut

HPGL IVIVE

MPPGL

P450

PBPK

1 microsome

Re/I

Vd,u

Vincubation Vmax vp Vr

Fraction that escapes intestinal metabolism Fraction that escapes hepatic metabolism Unbound fraction in blood/plasma Unbound fraction in microsomes Unbound fraction in tissues Hepatocytes per gram of liver In vitro-in vivo extrapolation

Michaelis-Menten constant (i.e., substrate concentration when v is V2 of Vmax) Maximum life span Microsomal protein per gram of liver Mean residence time Cytochrome P450

Physiologically based pharmacokinetic model Pharmacokinetic

Amount of microsomal protein in the incubation Hepatic blood flow

Ratio of binding proteins in extracellular fluid (except plasma) to binding proteins in plasma

In vitro half-life

Volume of distribution

Unbound volume of distribution

Extracellular fluid volume

Incubation volume

Maximum rate of the metabolic reaction Plasma volume

"Remainder" of the fluid volume

7.2 BASIC CONCEPTS

One of the key functions of DMPK scientists in drug discovery is prediction of human pharmacokinetics and the human dose. The general concept is outlined in Fig. 7.1. The three most relevant parameters are bioavailability (F), clearance (CL; clearance also contributes to F), and volume of distribution (Vd). As described in Chap. 3, F is determined by the fraction absorbed from the intestine (Fa), the fraction that escapes intestinal metabolism (Fg), and the fraction that escapes hepatic metabolism (Fh). Other pharmacokinetic (PK) parameters, such as half-life (i1/2), can be derived from these parameters. Numerous methods are available for prediction of these parameters, and they are listed in Table 7.1. The most commonly used methods are described in subsequent sections. More sophisticated predictions can be made with a physiologically based pharmacokinetic (PBPK) model using a range of preclinical in vitro and/or in vivo data.

Preclinical biomarker and/or efficacy data

PK/PD/efficacy modeling

Preclinical biomarker and/or efficacy data

PK/PD/efficacy modeling

PK in preclinical species

Allometry

Human PK prediction

In vitro ADME data

^ ■■■-^^'"iVIVE

Human dose prediction

Figure 7.1. Flowchart depicting methodology for human pharmacokinetic and dose prediction. ADME = absorption, distribution, metabolism, excretion; IVIVE = In vitro-in vivo extrapolation; PD = pharmacodynamic; PK = pharmacokinetic.

Table 7.1. Most commonly used methods to predict human pharmacokinetics

Pharmacokinetic parameter Methods

Fraction absorbed

Clearance

Volume of distribution

Solubility and dissolution data

In vitro permeability data from Caco-2 or MDCK cell or PAMPA studies Fa from preclinical PK studies In vitro-in vivo extrapolation using recombinant P450, microsome or hepatocyte data (with or without free fraction correction) Allometry (with or without the rule of exponents;

with or without free fraction correction) Single species (allometric) scaling (with or without free fraction correction) Single species liver blood flow method Tang and Mayersohn method

Allometry (with or without free fraction correction) Single species scaling (with or without free fraction correction) Oie-Tozer method

7.3 PREDICTION OF HUMAN FRACTION ABSORBED

The fraction absorbed is influenced by the intestinal solubility of the drug and the permeability across enterocytes. Solubility and dissolution rate studies can predict if absorption is limited by the solubility of the drug, and this is reflected by the maximum absorbable dose. Note that the magnitude of the anticipated human dose should be taken into consideration as well. In vitro permeability studies involving Caco-2 or MDCK cells or PAMPA provide a good idea about the intrinsic permeability of the drug. Details are provided in Chap. 4. Efflux by transporters can limit the fraction absorbed, but intestinal transporters can be saturated relatively easy. Finally, preclinical PK data can be used to predict Fa and the first-order absorption rate constant (ka). The fraction absorbed combined with knowledge of intestinal metabolism and systemic clearance can be used to predict F.

Preclinical Models for Human Drug Absorption

Although monkeys may be good models to predict Fh in humans, Fa x Fg is frequently substantially smaller in monkeys than in humans for drugs that undergo a significant degree of metabolism (Akabane et al. 2010). This probably reflects an increased capacity for intestinal metabolism in monkeys, because an earlier study showed that Fa in monkeys correlates well with Fa in humans (Chiou and Buehler 2002).

Although dogs are commonly used to study oral absorption, Fa in dogs is frequently larger than Fa in humans (Chiou et al. 2000). In addition, Tmax tends to be longer in humans than in dogs. The correlation between Fa in rats and humans is more robust (Chiou and Barve 1998).

7.4 PREDICTION OF HUMAN CLEARANCE

Clearance predictions can be based on either human in vitro ADME data or in vivo PK data from preclinical studies.

7.4.1 In Vitro-In Vivo Extrapolation

In vitro-in vivo extrapolation is the process by which organ clearance is scaled up using in vitro data. Since the liver is the main organ involved in metabolism of xenobiotics, this section focuses on the liver as the organ of interest. A flowchart of the process of in vitro-in vivo extrapolation is shown in Fig. 7.2.

Vitro Vivo Extrapolation Image
Figure 7.2. Flowchart depicting the process of in vitro-in vivo extrapolation. CLhepatic = hepatic clearance; CLint = intrinsic clearance; HPGL = hepato-cytes per gram of liver; MPPGL = microsomal protein per gram of liver.

As mentioned above, we will be covering only CLhepatic in this chapter. For drugs and compounds that are eliminated via organs other than the liver, estimated organ clearances can be summed together to get an estimate of total body clearance.

7.4.2 In Vitro Methods of Determining Intrinsic Clearance

Metabolic CLint is a measure of the ability of hepatocytes to eliminate a drug or compound irrespective of other external factors, such as protein binding and hepatic blood flow. Determination of metabolic CLint can be accomplished using traditional enzyme kinetic methodologies or substrate depletion methods.

7.4.2.1 Determination of Intrinsic Clearance Using

Michaelis-Menten Kinetic Parameters The relationship between the rate of a metabolic reaction and the substrate concentration is depicted in Fig. 7.3.

Figure 7.3. Plot of relationship between metabolic reaction rate (v) and substrate concentration (C). Km = Michaelis-Menten constant; Vmax = maximum rate of the metabolic reaction.

Rate

Figure 7.3. Plot of relationship between metabolic reaction rate (v) and substrate concentration (C). Km = Michaelis-Menten constant; Vmax = maximum rate of the metabolic reaction.

Rate

Concentration (C)

The Michaelis-Menten equation describes this relationship for many metabolic reactions and is presented below:

v — rate of the metabolic reaction.

Km = Michaelis-Menten constant (i.e., substrate concentration when v is V2 of Vmax).

C = concentration of the substrate (i.e., drug or compound of interest).

Michaelis-Menten kinetic parameters can be determined using in vitro data under conditions of linearity with respect to incubation time and either microsomal protein concentrations (for microsomal incubations) or number of cells (for hepatocyte incubations). Once Michaelis-Menten kinetic parameters are estimated, the metabolic CLint can be calculated as follows:

CLint = m* under linear conditions where C << Km Km

Based on the above equations, CLint is concentration dependent at high substrate concentrations approaching Km. For most drugs, the concentrations administered in vivo are under conditions of linearity with respect to CLint. CLint estimated from in vitro incubations are in units of volume/time/mg of microsomal protein for microsomes (e.g., mL/min/mg microsomal protein) and volume/ time/number of cells for hepatocytes (e.g., mL/min/106 cells).

Determination of CLint from Michaelis-Menten kinetic parameters can be labor intensive. As can be deduced from Fig. 7.3,

Vmax X C

Vmax = maximum rate of the metabolic reaction.

in vitro incubations must be performed at multiple substrate concentrations for a good estimate of Vmax and Km. In addition, the metabolite reaction that is being monitored must be the formation of a major metabolite for a primary metabolic pathway in order for a prediction to be accurate. Alternatively, for cases in which multiple major metabolic pathways are responsible for drug elimination, the formation rates of the primary metabolites from these multiple pathways must be monitored, and CLint of each pathway must be estimated and summed together to get the overall CLint. The substrate depletion rate may also be estimated at a range of substrate concentrations to estimate a hybrid overall apparent Vmax and Km. Regardless, every method utilizing Michaelis-Menten kinetic parameters requires that incubations be performed at a wide range of substrate concentrations. Due to the more resource intensive nature of this method of CLint estimation, substrate depletion estimation methods performed at one substrate concentration with C << Km have become common.

7.4.2.2 Determination of Intrinsic Clearance Using Substrate Depletion Method at a Single Substrate Concentration (In Vitro Half-Life Method)

Metabolic CLint can be estimated using the in vitro t1/2 of a microsome or hepatocyte incubation. Using this method, the substrate concentration must be less than Km. The following equation is used to estimate CLint from the in vitro t1/2:

0.693 iVincubation \

£1/2(invitro) \Pmicrosome/

ti/2(in vitro) - in vitro half-life Vincubation — volume of the incubation

Pmicrosome - amount of microsomal protein in the incubation

If tl/2(in vitro) is in minutes, Vincubation is in mL and Pmicrosome is in mg, then CLint is in units of mL/min/mg microsomal protein. It is important to keep track of units when performing in vitro-in vivo extrapolation.

The in vitro t1/2 method can also be applied to hepatocytes, in which case Pmicrosome would be replaced with "the number of hepatocytes in the incubation". Since the number of hepatocytes in an incubation is usually expressed as "X x 106 cells", the units for CLint in the preceding example would be mL/min/106 cells.

A common concentration used for estimation of CLint using the in vitro t1/2 method is 1 mM.

7.4.3 Scaling Factors for In Vivo-In Vitro Extrapolation of Intrinsic Clearance

In this section, we present scaling factors for the conversion of CLint estimated from in vitro incubations to CLint for the whole liver or body (Tables 7.2 and 7.3).

Table 7.2. Microsome and hepatocyte scaling factors for various species

Human

Rat

Dog

MPPGL (mg microsomal

32 (95% CI:

61 (95% CI:

55 (95% CI:

protein per gram liver)

29-34)

47-75)

48-62)

HPGL (x106 hepatocytes

99 (95% CI:

163 (95% CI:

169 (95% CI:

per gram of liver)

74-131)

127-199)

131-207)

CI = confidence interval

Adapted from Smith et al. (2008) and Barter et al. (2007)

CI = confidence interval

Adapted from Smith et al. (2008) and Barter et al. (2007)

A detailed analysis of MPPGL and HPGL is not available for monkeys and, therefore, the human values of these parameters should be used for monkeys (Table 7.2).

Table 7.3. Liver weight and liver scaling factor (g liver per kg body weight) for various species

Species

Mouse

Rat

Dog

Monkey

Human

(weight)

(0.02 kg)

(0.25 kg)

(10 kg)

(5 kg)

(70 kg)

Liver weight (g)

1.75

10.0

320

150

1,800

Liver scaling factor

87.5

40

32

30

25.7

(g liver per kg

body weight)

Adapted from Davies and Morris (1993)

Adapted from Davies and Morris (1993)

7.4.4 Liver Models of Hepatic Drug Clearance

Conversion of CLint for the whole liver or body to CLhepatic, which includes the impact of physiological factors such as blood flow and protein binding, involves the use of a liver model such as the well-stirred model, the parallel tube model, or the dispersion model.

None of the models has been shown to be superior to the others (Baranczewski et al. 2006); so, for simplicity, the equation for the well-stirred model is presented below:

CLhepatic — hepatic clearance fu — unbound fraction Q — hepatic blood flow

More details on the well-stirred model including values for hepatic blood flow are presented in Chap. 1.

In vitro-in vivo extrapolation is usually "validated" by determining if the in vitro ADME data correctly predict the clearance observed in preclinical studies. Although this is valuable, human clearance can involve pathways distinctly different from preclinical species, which can render this "validation" of limited value.

Microsomal Binding Considerations

Use of the well-stirred model for basic and neutral drugs or compounds often does not require the inclusion of fu in the calculation. For basic and neutral drugs or compounds, fu may cancel out with fumic (unbound fraction in microsomes) (Obach 1999). The well-stirred model equation including fumic is as follows:

f CLint fumT

fumic

In contrast to plasma protein binding, microsomal binding, fumic, is generally not species dependent, provided the micro-somal protein concentration is the same (Zhang et al. 2010). Equations have been developed to estimate fumic using drug lipophilicity (Hallifax and Houston 2006) f 1

/umic

C = microsomal protein concentration in mg/mL

log P/D = log P if molecule is a base (pKa >7.4) and log D74 if molecule is an acid or neutral compound (pKa<7.4).

Continued

This equation has also been extended to include calculation of the unbound fraction in hepatocytes (Kilford et al. 2008).

7.4.5 Allometry

Allometry was initially used to establish an empirical relationship between the body surface area of an animal and its body weight.

S = surface area a — coefficient of allometric equation W — body weight

Subsequently, allometry has been used to quantitatively relate a range of morphological parameters and biological functions to body weight, and the generic equation is:

Y — morphological parameter or biological function b — exponent of allometric equation

Currently, allometry is widely used to predict human clearance based on preclinical clearance values.

CL — a x Wb or log(CL) — log (a) + b x log(W) (7.8)

Body weight (kg)

Figure 7.4. An example of allometry to predict human clearance. CL = clearance.

Body weight (kg)

Figure 7.4. An example of allometry to predict human clearance. CL = clearance.

An example is presented in Fig. 7.4. Note that the units for CL in this equation are mL/min, which subsequently needs to be converted to mL/min/kg. It is worth emphasizing that allometry is of little value if different clearance mechanisms are operational across species. Several types of correction factors have been proposed to improve the quality of these predictions. These factors decrease the predicted human clearance values relative to standard allometry.

Maximum life span (MLP) correction:

Brain weight (BrW) correction:

To clarify when to use a specific correction factor, the "rule of exponents" has been proposed as follows:

b < 0.7 ! standard allometry 0.7 < b < 1.0 ! MLP correction b > 1.0 ! BrW correction

A comprehensive analysis suggested that "none of the correction factors resulted in substantially improved predictivity" (Nagilla and Ward 2004).

Two species allometry has been proposed as well, and Tang et al. (2007) indicated that these methods were as predictive as three species allometry with the "rule of exponents."

The unbound clearance (CLu) can also be assessed in the allo-metric equation:

fu — fraction unbound in plasma

Finally, Tang and Mayersohn (2005) proposed a method that enhances the predictivity of allometry, particularly in cases of "vertical allometry." Standard allometry is performed to determine the value of a (the coefficient of the allometric equation), and the human clearance is obtained using the following equation:

a — coefficient of standard allometric equation

The following aspects should be taken into consideration when applying allometry:

• Allometry is empirical.

• Allometry is of little value if different clearance pathways are operational.

• Allometry may be most useful if the clearance is predominantly renally mediated across species.

• The R2 value of the correlation should be considered when assigning a degree of confidence in the prediction.

• An exponent, significantly different from 0.7 (in particular <0.5 or > 1.0) is indicative of significant species differences and should reduce the confidence in the prediction (Hu and Hayton 2001).

• The preclinical species with the lowest body weight (usually mouse or rat) and the highest body weight (usually dog) have the greatest influence on the human prediction.

Although some level of success has been achieved with allome-try, some researchers have shown that the method is not more predictive than simpler approaches, such as single species scaling (Hosea et al. 2009) or the single species liver blood flow method (Nagilla and Ward 2004; Ward and Smith 2004).

7.4.6 Single Species Scaling

Single species scaling is based on direct extrapolation of the clearance in a particular preclinical species using a fixed exponent (usually 0.75), with or without a protein binding correction. In the former case, the equation is:

CLu,human CLu,animal x (Whuman/Wanimal) (7.15)

Opinions are divided about the most predictive preclinical animal model for humans. Hosea et al. (2009) observed good prediction of human pharmacokinetics using only rat data, which is also more practical than obtaining dog and/or monkey data. Other single species scaling methods were proposed by Tang et al. (2007) with the CL units being ml/min/kg.

CLh uman

7.4.7 Single Species Liver Blood Flow Method

Clearance, as percentage of liver blood flow in preclinical species can be a predictor of human clearance. Ward and Smith (2004) suggested that the monkey is the most predictive species.

CLhuman CLanimaI X (Qhuman/Qanim al) (7.19)

7.5 HUMAN VOLUME OF DISTRIBUTION PREDICTION

Vd is predominantly governed by physicochemical parameters, and, consequently, species differences are less pronounced for Vd than for metabolic clearance. Therefore, human Vd is easier to predict than CL, and the differences between various methods are generally limited unless the Vd is remarkably small or large.

7.5.1 Allometry

Allometry has been used with some degree of success to predict human Vd.

Vd = a x Wb or log(Vd) = log(a) + b x log(W) (7.20)

a — coefficient of allometric equation b — exponent of allometric equation

Figure 7.5. An example of allometry to predict human volume of distribution. Vd = volume of distribution.

Figure 7.5. An example of allometry to predict human volume of distribution. Vd = volume of distribution.

An example is presented in Fig. 7.5. Note that the units for Vd in this equation are mL or L, which subsequently needs to be converted to mL/kg or L/kg. Correction factors, such as MLS and BrW, are generally not used. Some researchers have proposed that the unbound volume of distribution (Vd,u) should be used in the allo-metric equation as follows:

Vd,u = Vd/fu fu = fraction unbound in plasma

7.5.2 Single Species Scaling

Single species scaling is based on the similarity of Vd across species. Some have advocated using an identical Vd in humans as in preclinical species, but a correction factor is commonly applied to account for differences in plasma protein binding.

Vd,u,human Vd,u,animal x (/u,human//u,animal) (7.22)

7.5.3 The Oie-Tozer Method

In the Oie-Tozer method, the unbound fraction in human plasma and the average unbound fraction in tissues from preclinical species (assumed to be equal to the unbound fraction in human tissues), combined with appropriate human values for plasma and fluid volumes, are used to predict the human Vd (Obach et al. 1997).

xVr fu,human)

fu,human

fut,species average

Vp — plasma volume

Ve — extracellular fluid volume

Vr — remainder of the fluid volume

Re/i — ratio of binding proteins in extracellular fluid (except plasma) to binding proteins in plasma fut,species average — average unbound fraction in tissues from preclin-ical species

7.6 PHYSIOLOGICALLY BASED PHARMACOKINETIC MODELS

PBPK models offer a sophisticated integrated framework to understand compound disposition in the body and predict human pharmacokinetic profiles. The concept is described in detail in Chap. 10.

Prediction of Human Pharmacokinetic Profiles Using Methods Involving the Normalization of Animal Profiles

Methods predicting human pharmacokinetic profiles involving normalization of animal plasma concentration-time profiles include the Wajima plot and the various Dedrick methods. These methods are based on the assumption that plasma concentration-time profiles from different species are superimposable upon appropriate normalization of concentration and time scales. For illustrative purposes, a representative concentration-time profile and a Wajima plot are shown below. Each line represents the concentration-time profile of one animal species.

Continued

Concentration-time plot Wajima plot

Concentration-time plot Wajima plot

Time Time/MRT

A Wajima plot involves normalization of the concentration scale by dividing it by the steady-state concentration (Css; calculated as Dose/Vss]. The time scale is normalized by dividing it by the mean residence time (MRT; calculated as Vss/CL). The composite profile of all preclinical species is used as the predicted human concentration-time profile. Ideally, normalized concentration-time profiles from various species superimpose upon each other, and predicted human profiles are obtained by back converting the normalized profile using predicted Css and MRT values for humans. The original set of compounds for which Wajima plots were applied were antibiotics with small Vd values that were mainly cleared renally in humans. A distinct advantage of this plot over the various Dedrick methods is that the CL and Vss values used to calculate predicted Css and MRT values for humans can be obtained from any method of prediction. In contrast to the Wajima plot, the various Dedrick methods utilize allometric principles to normalize concentration and time scales. Detailed information on the Wajima plot and Dedrick methods can be found in Wajima et al. (2004) and Mahmood (2005).

7.7 CONTEXT OF CONFIDENCE IN HUMAN PHARMACOKINETIC PREDICTION

Human PK predictions are performed extensively preclinically and can determine the fate of compounds under consideration for development. At an earlier stage, predictions are based on limited data sets (in vitro and rodent in vivo data), and the goal is usually to categorize compounds and identify those that are worth pursuing further in preclinical studies. Once a more complete data set is available, it is possible to make more refined human PK predictions using perhaps more than one method, but even at this stage the predictions are more a reflection of assumed risk. Prospective predication of human pharmacokinetics continues to be an extremely difficult endeavor. For example, the most successful clearance methods generally have success rates of 60-80% with success defined as being within twofold of the observed value in humans. Often, having more preclinical data does not improve the prediction success rate. Indeed, Beaumont and Smith (2009) commented that "Generation of further large amount of preclinical information on a compound with uncertain human pharmacokinetic prediction tends to add confusion rather than clarity." Finally, the predicted PK properties of each drug candidate must be carefully evaluated in the context of its other properties and liabilities (i.e., in vivo potency, toxicity, etc.).

References

Akabane T, Tabata K, Kadono K et al (2010) A comparison of pharmacokinetics between humans and monkeys. Drug Metab Dispos 38:308-316 Baranczewski P, Stanczak A, Sundberg K et al (2006) Introduction to in vitro estimation of metabolic stability and drug interactions of new chemical entities in drug discovery and development. Pharmacol Rep 58:453-472

Barter ZE, Bayliss MK, Beaune PH et al (2007) Scaling factors for the extrapolation of in vivo metabolic drug clearance from in vitro data: reaching a consensus on values of human microsomal protein and hepa-tocellularity per gram of liver. Curr Drug Metab 8:33-45 Beaumont K, Smith DA (2009) Does human pharmacokinetic prediction add significant value to compound selection in drug discovery research? Curr Opin Drug Disc Dev 12:61-71 Chiou WL, Barve A (1998) Linear correlation of the fraction of oral dose absorbed of 64 drugs between humans and rats. PharmRes 15:1792-1795 Chiou WL, Jeong HY, Chung SM et al (2000) Evaluation of using dog as an animal model to study the fraction of oral dose absorbed of 43 drugs in humans. Pharm Res 17:135-140 Chiou WL, Buehler PW (2002) Comparison of oral absorption and bioavail-

ability of drugs between monkey and human. Pharm Res 19:868-874 Davies B, Morris T (1993) Physiological parameters in laboratory animals and humans. Pharm Res 10:1093-1095 Hallifax D, Houston JB (2006) Binding of drugs to hepatic microsomes: comment and assessment of current prediction methodology with recommendation for improvement. Drug Metab Dispos. 34:724-726 Hu T-M, Hayton WL (2001) Allometric scaling of xenobiotic clearance:

uncertainty versus universality. AAPS PharmSci 3:1-14 Kilford PJ, Gertz M, Houston JB, Galetin A (2008) Hepatocellular binding of drugs: correction for unbound fraction in hepatocyte incubations using microsomal binding or drug lipophilicity data. Drug Metab Dispos. 36:1194-1197

Mahmood I (2005) Interspecies pharmacokinetic scaling: principles and application of allometric scaling. Pine House Publishers, Rockville, Maryland

Nagilla R, Ward KW (2004) A comprehensive analysis of the role of correction factors in the allometric predictivity of clearance from rat, dog, and monkey to humans. J Pharm Sci 93:2522-2534 Obach RS, Baxter JG, Liston TE et al (1997) The prediction of human pharmacokinetic parameters from preclinical and in vitro metabolism data. J Pharmacol Exp Ther 283:46-58 Obach RS (1999) Prediction of human clearance of twenty-nine drugs from hepatic microsomal intrinsic clearance data: an examination of in vitro half-life approach and nonspecific binding to microsomes. Drug Metab Dispos 27:1350-1359 Smith R, Jones RD, Ballard PG et al (2008) Determination of microsome and hepatocyte scaling factors for in vitro/in vivo extrapolation in the rat and dog. Xenobiotica 38:1386-1398 Tang H, Mayersohn M (2005) A novel method for prediction of human drug clearance by allometric scaling. Drug Metab Dispos 33:1297-1303 Tang H, Hussain A, Leal M et al (2007) Interspecies prediction of human drug clearance based on scaling data from one or two animal species. Drug Metab Dispos 35:1886-1893 Wajima T, Yano Y, Fukumura K et al (2004) Prediction of human pharmacokinetic profile in animal scale up based on normalizing time course profiles. J Pharm Sci 93:1890-1900 Ward KW, Smith BR (2004) A comprehensive quantitative and qualitative evaluation of extrapolation of intravenous pharmacokinetic parameters from rat, dog, and monkey to humans. I Clearance Drug Metab Dispos 32:603-611

Zhang Y, Yao L, Lin J et al (2010) Lack of appreciable species differences in nonspecific microsomal binding. J Pharm Sci 99:3620-3627

Additional Reading

De Buck SS, Mackie CE (2007) physiologically based approaches towards the prediction of pharmacokinetics: in vitro-in vivo extrapolation. Expert Opin Drug Metab Toxicol 3:865-878 Houston JB, Carlile DJ (1997) Prediction of hepatic clearance from microsomes, hepatocytes and liver slices. Drug Metab Rev 29:891-922 McGinnity DF, Collington J, Austin RP et al (2007) Evaluation of human pharmacokinetics, therapeutic dose and exposure predictions using marketed oral drugs. Curr Drug Metab 8:463-479 Obach RS (2001) The prediction of human clearance from hepatic microsomal metabolism data. Curr Opin Drug Discov Devel 4:36-44 Pelkonen O, Turpeinen M (2007) In vitro-in vivo extrapolation of hepatic clearance: biological tools, scaling factors, model assumptions and correct concentrations. Xenobiotica 37:1066-1089

0 -1

Post a comment