Interspecies Pharmacokinetic Scaling

One of the primary objectives of preclinical pharmacokinetics is to generate information describing the absorption, distribution, metabolism, and excretion (ADME¹) processes in animals that can then be used for the extrapolation to human ADME processes. The many intrinsic differences in the ADME processes between animals and humans make extrapolation of animal data very difficult. This section will address some aspects of these processes that can be used legitimately for the extrapolation of animal data as well as for situations in which the extrapolation cannot be justified.

Absorption.

Drug absorption is influenced by many physiological factors, but it depends also on the physicochemical characteristics of the drug itself. The physiological factors include gastric and intestinal transit time, blood flow rate, gastrointestinal pH, and first-pass metabolism, while the physicochemical factors are the drug’s intrinsic properties, such as (-logKa, or the ionization constant), molecular size, solubility, and lipophilicity. Whereas the physiological factors are subject to species variation, the physicochemical factors are species-independent.

The oral bioavailability of a drug is defined as the fraction of an oral dose that actually reaches the systemic circulation. Because the entire blood supply of the upper gastrointestinal tract passes through the gut wall and the liver before reaching the systemic circulation, the drug may be metabolized by the liver and gut wall during the first passage of absorption. Kinetically, the oral bioavailability (F) can be described as:F=fabs·(1−fg)·(1−fh) Equation 1where f_abs is the fraction of dose absorbed from the gastrointestinal lumen andf_g and f_h are the fractions of drug metabolized by the gut wall and liver, respectively, during the first passage of drug absorption (Lin and Lu, 1997). The f_abs of a drug is determined mainly by its permeability across the biomembrane, which is similar among species. Therefore, the f_abs of a drug is expected to be similar among species. However, thef_g and f_h of a drug may differ substantially from one species to another because of their intrinsic differences in metabolism, resulting in differences in bioavailability across species.

This was best exemplified by the observation of marked interspecies differences in the bioavailability of indinavir (MK-639) when the drug was given orally as a solution in 0.05 M citric acid. The bioavailability varied from 72% in dogs to 24% in rats and 19% in monkeys (Lin et al., 1996a). Ultimately, it was discovered that the low bioavailability observed in rats and monkeys was due to extensive hepatic first-pass metabolism. By comparing the drug concentration in the systemic circulation during portal or femoral vein infusion, hepatic first-pass extraction (f_h) was estimated to be 68% in rats. However, in situ studies with isolated intestinal loop preparations in anesthetized rats showed that intestinal first-pass metabolism (f_g) was minimal (<8%). Consistent with the in vivo and in situ studies,in vitro hepatic and intestinal first-pass extraction (f_h and f_g) for rats were estimated to be 55% and 5%, respectively, using the liver and intestinal microsomalV_max/K_M data. Although in vivo hepatic first-pass extraction was not determined for dogs and monkeys, the in vitro values were estimated to be 17% and 65%, respectively, using dog and monkey liver microsomes (Lin et al., 1996a). From eq. 1, and taking thef_g and f_h into account, the extent (f_abs) of indinavir absorbed from the gastrointestinal lumen was quite similar among species (55%–80%). Thus the observed species differences in the bioavailability of indinavir were due mainly to the differences in the magnitude of hepatic first-pass metabolism. When human intestinal and hepatic microsomes were used, the extent of intestinal and hepatic first-pass metabolism of indinavir in humans was estimated to be 5% and 26%, respectively (Chiba et al., 1997). With the extent of absorption (55%–80%) obtained from animal studies, we predicted that the bioavailability of indinavir in patients would be 40%–60%. As predicted, when clinical data became available, the bioavailability of indinavir was found to be approximately 60% (Yeh et al., 1998).

L-365,260, a potent CCK_B (cholecystokinin) receptor antagonist, is another example that shows species similarity in the fraction of drug absorbed (f_abs). The bioavailability of L-365,260 was 14% for rats and 9% for dogs when given orally as a suspension in 0.5% methylcellulose (Linet al., 1996b). Because the extent of hepatic first-pass metabolism was low and estimated to be 30% for rats and 14% for dogs (Lin et al., 1996b), the limited bioavailability was attributed mainly to poor absorption as a result of low aqueous solubility (<2 μg/ml). When L-365,260 was given as a solution in PEG 600, the bioavailability increased to 50% in rats and 70% in dogs. Taking hepatic first-pass metabolism into consideration, the extent (f_abs) of L-365,260 absorbed from the gastrointestinal lumen was similar between rats and dogs at approximately 80%. With this information at hand, L-365,260 was administered in capsules containing PEG 600 in the subsequent clinical studies. As expected, the formulation resulted in good absorption of L-365,260 in humans. The peak concentration (C_max) and AUC were, respectively, 2.3 μg/ml and 450 μg · min/ml for dogs and 0.5 μg/ml and 148 μg · min/ml for normal human subjects when the same dose (50 mg) of L-365,260 in PEG capsules was given orally to dogs (12 kg) and normal volunteers (70 kg).2 TheC_max and AUC values were comparable in dogs and humans when compared on a weight-normalized dose basis.

The examples of indinavir and L-365,260 suggest that drug absorption in humans can be extrapolated reasonably well from animal data when information on first-pass metabolism is also available. Indeed, Clark and Smith (1984) reported in a survey that the fraction of dose absorbed (f_abs) from the gastrointestinal lumen for a large variety of drugs is remarkably consistent between animal species and humans. The bioavailability, however, differs substantially among species, presumably as a result of species differences in the magnitude of first-pass metabolism.

Distribution.

The rate of distribution of a drug to the organs or tissues is determined by the blood flow perfusing the tissues and the ease with which the drug molecules cross the capillary wall and penetrate the cells. Like most physiological parameters, blood flow and circulation time can be extrapolated across species by use of an allometric equation (Edwards, 1975; Boxenbaum, 1980):y=awb Equation 2where y is blood flow or circulation time, wis body weight, a is the allometric coefficient, andb is the allometric exponent. The allometric relationship between blood circulation time (sec) and total body weight (kg) is 21w^0.21 (Stahl, 1967). The blood circulation time is 15 sec in a 250-g rat and 50 sec in a 70-kg human. This implies that a drug molecule circulates the body four times per min in the rat and only once per min in humans. Similarly, both hepatic and renal blood flow (expressed as ml/min/kg body weight) decrease as the animal size increases. The hepatic blood flow is approximately 70 ml/min/kg for the rat and 20 ml/min/kg for humans (Boxenbaum, 1980), and the renal blood flow is 55 ml/min/kg for the rat and 15 ml/min/kg for humans (Edwards, 1975). Clearly, the smaller animal species deliver drugs faster and more frequently to the organs of elimination: namely, the liver and kidneys. Thus it is expected that smaller animal species would eliminate drugs more rapidly than would humans, particularly for those drugs with high clearance, when compared on a weight-normalized basis.

It is generally believed that only the unbound drug can diffuse across membranes that restrict distribution of a drug from the vascular compartment into tissues and vice versa. Therefore, drug protein binding in plasma and tissues can affect the distribution of drugs in the body. It is well known that the extent of binding of drugs to plasma proteins differs considerably among species. The observed species differences in plasma protein binding may reflect differences in the affinity or the number of binding sites on the protein molecule. Albumin, the major drug binding protein in plasma, is composed of a single polypeptide chain of ∼590 amino acids. Although structural and functional homologies of albumin exist among species, there are small differences in the amino acid sequences between humans and animal species (Callan and Sunderman, 1973; Kragh-Hansen, 1981), leading to differences in the binding affinity and sites.

Diflunisal, a nonsteroidal anti-inflammatory drug, is bound extensively to plasma protein and eliminated mainly by conjugation as ester and ether glucuronides in humans and rats (Lin et al., 1985). Although diflunisal exhibits concentration-dependent pharmacokinetics in both rats and humans, the kinetics of the drug are different in these two species. The plasma clearance of diflunisal in rats remains relatively constant over the therapeutic concentration range of 50–150 μg/ml (150–450 μM) (Lin et al., 1985), while the clearance in humans decreases with concentration over this same range (Meffin et al., 1983). In contrast, in vitrobinding studies have demonstrated that diflunisal shows nonlinear plasma protein binding in rats over the therapeutic concentration range, but the unbound fraction of diflunisal in plasma remains unchanged in humans over this range (Lin, 1989). A detailed kinetic study has demonstrated that lack of changes in plasma clearance in rats is a consequence of the opposing effects of saturable metabolism and saturable plasma protein binding (Lin et al., 1985), whereas the nonlinear clearance in humans is attributed more likely to the saturable metabolism alone. Further studies with serum albumin revealed that the number of binding sites for diflunisal was different between human and rat albumin, whereas the affinity for albumin in both species was comparable. The number of binding sites, which was 3 for human serum albumin and 1 for rat serum albumin, respectively, varied, while the respective association constants were similar (approximately 4.5 × 10⁵ M⁻¹) (Lin, 1989). Consistent with these observations, displacement studies with binding markers specific to albumin (¹⁴C-diazepam,¹⁴C-warfarin, and³H-digitoxin) suggested that diflunisal bound to three discrete binding sites on human serum albumin but only one site on rat serum albumin (Lin, 1989). Although the albumin concentration in plasma (500–600 μM) is similar in humans and rats, these results strongly suggest that a much higher drug concentration is required to saturate plasma protein binding capacity of diflunisal in humans because of the larger number of binding sites.

The volume of distribution, a measure of the extent of drug distribution that is determined by the binding of drugs to tissue as well as plasma proteins, is an important determinant of half-life. It is, therefore, desirable if the volume of distribution in humans can be predicted from that in animals. Kinetically, the simplest quantitative expression relating the volume of distribution (V_d) to plasma and tissue binding (Lin, 1995) is given as:Vd=Vp+∑Vtfp/ft Equation 3where V_p is the plasma volume,V_t is the tissue volume, andf_p and f_t are the fractions of unbound drug in plasma and tissue, respectively. From this relationship, it is evident that theV_d increases whenf_p is increased and decreases whenf_t is increased.

Rearrangement of eq. 3 yields:Vf=Vd/fp=Vp/fp+∑Vt/ft Equation 4where V_f is defined as the volume of distribution of unbound drugs. From this equation, it is clear that a change in f_t has a greater impact thanf_p on V_fbecause ΣV_t is much greater thanV_p.

Although it is easy to determine the plasma protein binding of drugs, the study of tissue binding is hampered by methodological problems. The technical difficulties associated with the determination of drug binding to tissues are reflected by the very limited amount of published information on the subject. Fichtl et al. (1991)reported that there were striking species differences in plasma protein binding and the V_d of propranolol. The values for V_d varied by more than 20-fold, being lowest in monkeys and highest in rabbits. However, when theV_d was corrected forf_p, the volume of distribution of unbound propranolol, V_f, was virtually the same for all species. Consistent with this, Sawada et al. (1984a)reported that the V_f values of ten basic drugs were quite similar among species, including humans. Based on these results, Fichtl et al. (1991) proposed that theV_f of drugs should be similar in humans and other species. The authors suggested that with knowledge of theV_f from laboratory animals and off_p from human plasma protein binding determined in vitro, one can predict theV_d (V_f ×f_p) in humans before the initial clinical studies are initiated. Unfortunately, this approach is not valid for all drugs. Boxenbaum (1982) compared the pharmacokinetic parameters for 12 benzodiazepines in dogs and humans. Eight of the 12 benzodiazepines had quite different V_f values between the dogs and humans, the differences being as much as sevenfold for lorazepam. The large species differences in theV_f values were also reported for β-lactam antibiotics (Sawada et al., 1984b). Thus the species similarity in the V_f of propranolol observed by Fichtl et al. (1991) might simply be fortuitous. In conclusion, the results from the aforementioned examples suggest that the binding to plasma and tissue protein andV_d of drugs in humans cannot be readily extrapolated from animal data.

Metabolism.

From an evolutionary standpoint, all mammals are similar because they originate from a common ancestor, yet they have differentiated as a result of their dissimilar environmental adaptions. Biochemistry provides countless examples of similarities and differences between species, the most instructive of which is the structure of cytochrome P450s. Cytochrome P450s appear to have evolved from a single ancestral gene over a period of 1.36 billion years. To date, at least 14 P450 gene families have been identified in mammals (Nelson et al., 1996). Although all of the members of this superfamily possess highly conserved regions of amino acid sequences, there are considerable variations in the primary sequences across species. Profound differences in substrate specificity, however, can arise even with a small change in the amino acid sequences. As a result of the species differences in the amino acid sequences of the isoforms, both the rate of drug metabolism and the metabolite pattern may differ significantly among animal species.

In addition to the species differences in amino acid sequences and substrate specificity, the levels of P450 isoforms may also differ across species. For example, the hepatic enzyme levels of CYP1A, CYP2C, and CYP3A isoforms in rats are approximately 28, 638, and 165 pmol/mg microsomal protein, respectively (De Waziers et al., 1990), and the corresponding values for humans are 37, 55, and 87 pmol/mg microsomal protein (Guengerich, 1995). Antipyrine is the most widely studied probe used for assessing in vivo metabolic functions in animals and humans. The major metabolic pathways of antipyrine in humans are N-demethylation, 4-hydroxylation, and 3-methyl-hydroxylation (Eichelbaum et al., 1982). At least four human cytochrome P450 isoforms (CYP1A2, CYP2C9, CYP2C18, and CYP3A4) have been identified as involved in the metabolism of antipyrine (Engel et al., 1996). Knowing the complexity of antipyrine metabolism and the interspecies differences in the levels of P450 isoforms, it is not surprising that Boxenbaum (1980) failed to extrapolate the intrinsic clearance (V_max/K_M) of antipyrine from animal data to humans when using the allometric approach.

Stevens et al. (1993) compared Phase I and Phase II hepatic drug metabolism activities, using human and monkey liver microsomes. Of the eight P450-dependent activities measured, onlyN-nitrosodimethylamine N-demethylase activity was not significantly different between the two species. Coumarin 7-hydroxylase activity was higher in humans than in monkeys. In contrast, erythromycin N-demethylase, benzphetamineN-demethylase, pentoxyresorufin O-dealkylase, ethoxycoumarin O-deethylase, and ethoxyresorufinO-deethylase activities were significantly greater in monkey microsomes than those in human microsomes. Of the seven microsomal and cytosolic Phase II activities measured, only 17α-ethynyl estradiol glucuronidation was significantly higher in humans. These results are in contradiction with the popular belief that monkey metabolism is comparable to human metabolism.

Similar to cytochrome P450s, uridine diphosphate glycosyltransferases (UGTs) also show species differences. At least ten rat UGTs and eight human UGTs have been defined and characterized to date by cDNA cloning (Clarke and Burchell, 1994). Comparison of the amino acid sequences of all UGTs indicates that they share a common C-terminal domain, but that the N-terminal half of these isoforms is quite variable. Examination of each of the UGT isoforms has revealed that there are interspecies differences in UGT activities, both in quantitative and qualitative aspects.

Zidovudine (AZT), an HIV reverse transcriptase inhibitor, is extensively metabolized in humans but not in rats. Approximately 75% of an oral dose was recovered in human urine as the 5′-O-glucuronide, and 15% was recovered as unchanged drug (Blum et al., 1988). On the other hand, only 2% of an oral dose was recovered as AZT glucuronide in rat urine, whereas approximately 78% of the dose was excreted as unchanged drug (Goodet al., 1986). Consistent with the in vivo data,in vitro studies confirmed that human liver UGT catalyzed the glucuronidation of 0.1 mM AZT 10- to 25-fold faster than did rat liver UGT (Resetar and Spector, 1989).

In a recent study, hepatic and intestinal UGT activities in rats and rabbits were investigated by measuring the glucuronidation of 1-naphthol, 2-methylumbelliferone, 4-nitrophenol, 2-hydroxybiphenyl, and 4-hydroxybiphenyl (Vargas and Franklin, 1997). Generally, intestinal UGT activities were higher in rabbits when compared with those of rats, while hepatic activities were much higher in rats than in rabbits. In rats, the activities (nmol/min/mg microsomal protein) in the small intestinal mucosa were much lower than those in liver, with the activities in the intestine representing 5%–15% of hepatic levels. In contrast, the intestinal activities were comparable (70%–100%) to the hepatic activities for most aglycones in rabbits.

In addition to the species differences in catalytic activities of drug-metabolizing enzymes, interspecies differences also exist in enzyme inhibition and induction. Various mechanisms are known to underlie enzyme inhibition, including competition for the catalytic site of the enzyme, noncompetitive (allosteric) interaction with the enzyme, suicide destruction of the enzyme, and competition for cofactors. Among these mechanisms, competitive inhibition is probably the most common. If enzyme inhibition occurs by the interaction of two substrates competing for the same enzyme, the competitive nature of the inhibition will depend on the K_M value of the substrate and the inhibitory constant of an inhibitor (K_i) value of the inhibitor as well as their concentrations at the site of enzyme. BecauseK_M and K_ivalues can be different between species, it is expected that the degree of enzyme inhibition would be species-dependent.

Isoforms of the CYP2D subfamily have been isolated from rats and humans and have been shown to have similar substrate specificities. Debrisoquine 4-hydroxylation is specifically catalyzed by these isozymes. The inhibition kinetics of debrisoquine 4-hydroxylase activity by quinidine and one of its diastereoisomers, quinine, have been compared in human and rat liver microsomes (Kobayashi et al., 1989). Both quinidine and quinine are potent competitive inhibitors of debrisoquine 4-hydroxylation. However, quinidine is a more potent inhibitor of this activity in humans than in rats, whereas the reverse is true for quinine. The K_ivalues of quinidine for debrisoquine 4-hydroxylation in humans and rats were 0.6 and 50 μM, respectively, whereas with quinine, the values were 13 and 1.7 μM, respectively. Similarly, furafylline exhibits species-dependent inhibition of phenacetin O-deethylase activity of liver microsomes (Sesardic et al., 1990). Furafylline, a mechanism-based inhibitor of CYP1A2, is more potent in inhibiting phenacetin O-deethylation in humans than in rats, despite the fact that phenacetin O-deethylation is catalyzed exclusively by CYP1A2 in both species.

Although the fundamental mechanisms of CYP1A induction are qualitatively similar in different species, including mice, rats, rabbits, and humans (McDonnell et al., 1992), there are important quantitative differences in the effectiveness of inducer-receptor coupling. For example, the gastric acid-suppressing drug omeprazole is a CYP1A2 enzyme inducer in humans but has little inductive effect in mice or rabbits (McDonnell et al., 1992;Diaz et al., 1990). Important species differences also exist in the response of other inducible subfamilies of cytochrome P450s. Phenobarbital induces predominately the members of the CYP2B subfamily in rats, whereas in humans it appears that the major form induced belongs to the CYP3A subfamily (Rice et al., 1992). Furthermore, members of the CYP3A subfamily in rats are inducible by the steroidal agent pregnenolone-16α-carbonitrile but not by the antibiotic rifampin. The opposite is true in rabbits and humans (Strolin Benedetti and Dostert, 1994; Nebert and Gonzalez, 1990). Thus it should not be assumed that drugs that do not induce P450 enzymes in animals do not have enzyme-inducing capacity in humans, and vice versa. Despite well-known species differences in response to P450 inducers, mice and rats have been used routinely in most pharmaceutical companies to assess the risk of potential drug induction in humans. This type of risk assessment may be of little direct relevance for certain drugs.

In summary, these examples clearly demonstrate that extrapolation of drug metabolism from animals to humans often is fairly difficult, if not impossible, both in the qualitative and quantitative aspects. As will be discussed later, however, reliable extrapolation of drug metabolism can be made from in vitro experiments.

Excretion.

Drugs and their metabolites are usually eliminated from the bodyvia urine or bile or, sometimes, both. The relative contribution of biliary and urinary excretion to the overall elimination of drugs depends on the nature of the drugs and the animal species. Generally, biliary excretion predominates in drugs with relatively large molecular weights (>500). One striking feature of many drugs excreted in bile is that their structures are amphipathic in character (i.e. they contain both polar and nonpolar groups). Many lipophilic compounds are excreted into the bile at a higher rate after conjugation with glutathione or glucuronic acid, presumably because the reaction not only increases molecular weight but also adds a polar group.

The amount of an organic chemical that is excreted in bile varies widely among species. In general, mice, rats, and dogs are good biliary excreters, while rabbits, guinea pigs, monkeys, and humans are relatively poor biliary excreters. The species differences in biliary excretion become less marked when the molecular size of the drug being excreted exceeds 700 daltons. The underlying mechanism for the species differences is at present unclear. Species differences in hepatic blood flow and bile flow do not seem to correlate with the biliary excretion of compounds (Smith, 1971 and 1973). Thus it is difficult to predict the biliary excretion of drugs in humans from animal data.

Besides biliary excretion, many drugs are excreted mainly as unchanged drugs by the kidneys. The rate of renal excretion (renal clearance) is dependent on renal blood flow, glomerular filtration rate (GFR), and tubular secretion and reabsorption. The GFR values vary considerably among species because of species differences in the number of nephrons. The GFR values are 10, 8.7, 4.8, 4.0, and 1.8 ml/min/kg for mice, rats, rabbits, dogs, and humans, respectively, and corresponding values of nephron number (per kg body weight) are 5.0 × 10⁵, 2.9 × 10⁵, 1.6 × 10⁵, 0.9 × 10⁵, and 0.29 × 10⁵(Renkin and Gilmore, 1973). Both the GFR and number of nephrons show a good allometric relationship. Like the GFR, the renal excretion of drugs also shows a good allometric relationship across species. Thus the renal clearance of drugs in humans can be extrapolated from animal data by use of the allometric approach. Ceftizoxime and methotrexate are good examples. Both drugs are excreted mainly as unchanged drugs in the urine. The plasma clearance of ceftizoxime and methotrexate in humans has been extrapolated successfully from animal data (Chapell and Mordenti, 1991).

Although the renal clearance of a drug in humans can be predicted reasonably well by use of the allometric approach, this approach requires at least four or five animal species in order to obtain a proper allometric relationship, thus limiting its practical value in drug development. A more simplistic yet useful alternative to predict human renal clearance is to use the ratio of GFR between rats and humans. As shown in table1, the ratios of renal clearance of various drugs between rats and humans is roughly equal to the ratio of GFR between these two species (Lin, 1995). These results suggest that with knowledge of the GFR ratio and the renal clearance of a drug in rats, the renal clearance of the drug in humans can be estimated roughly.

In Vitro Extrapolation

In recent years, there has been a large expansion in both the range and use of in vitro systems to study drug metabolism. Because of the simplicity of in vitro systems, they are very useful in studying the factors that influence pharmacokinetics and drug metabolism. A trickier task is to use thesein vitro systems to predict in vivopharmacokinetics and drug metabolism quantitatively. The difficulty in extrapolating in vitro to in vivo data lies in the complexity of the interdependent biological processes and their dynamic nature. Therefore, it is important to carefully set up the in vitro experimental conditions that simulate thein vivo situations and to understand the interdependent factors between biological components that affect each. In addition, a good understanding of pharmacokinetic principles is necessary for the in vitro/in vivo extrapolation.

Metabolite Profiles.

In drug development, early information on human metabolism of a new drug is critical in predicting potential clinical drug-drug interactions and in selecting the appropriate animal species for the toxicity studies. For human risk assessment, regulatory agencies require that the systemic exposure of an unchanged drug and its major metabolites in the animal species used in the toxicity study exceeds that expected in humans to ensure a safety margin. It is important, therefore, to select animal species that have metabolite profiles similar to those of humans. However, the in vivo human drug metabolism study normally is not carried out until the later stages of drug development, which is often too late for animal selection. Fortunately, the increased availability of human tissues and advances in bioanalytical and biochemical technologies have provided opportunities for in vitro studies of human metabolism at the early stage of drug development before the toxicity studies are initiated (Wrighton et al., 1993).

In general, the metabolite profile of a drug obtained in vitro quite accurately reflects the in vivo metabolite pattern, although it is limited to qualitative aspects. From the physiological and biochemical points of view, precision-cut liver slices are especially useful when the complete in vitrometabolite profile of a drug is being obtained. This system retains the physiological conditions of enzymes and cofactors of both Phase I and Phase II reactions and, therefore, better simulates the in vivo situation (Dogterom, 1993). The metabolism of indinavir (MK-0639) illustrates this point. The major metabolic pathways of indinavir in humans have been identified as the following: (a) glucuronidation at the pyridine nitrogen to yield a quaternary ammonium conjugate, (b) pyridine N-oxidation, (c) para-hydroxylation of the phenylmethyl group, (d) 3′-hydroxylation of the indan, and (e)N-depyridomethylation. The metabolite profile of indinavir obtained from human liver slices accurately reflects the in vivo human metabolite pattern (Balani et al., 1995). Although all of the oxidative metabolites of indinavir were also found in human liver microsomes, the N-glucuronide was not detected when indinavir was incubated with native or Triton X-100–treated human liver microsomes in the presence of 10 mM uridinediphosphoglucuronic acid (Lin et al., 1996a). The reason for the inability of human liver microsomes to form theN-glucuronide is not clear. Nevertheless, these results suggest that the liver slice is a better in vitro model for the study of the metabolic pathways of drugs.

Although liver slices are valuable in identifying metabolic pathways, their use in obtaining kinetic parameters may be limited.Worboys et al. (1996) showed that the values ofCL_int(V_max/K_M) of a series of drugs in slices are consistently less than those in hepatocytes by a factor ranging from 2 to 20. These results strongly suggest that a distribution equilibrium is not achieved between all of the cells within the slice and the incubation medium because of the slice thickness (∼260 μm).

Isolated and cultured hepatocytes also are used often as in vitro models for identifying the metabolic pathway of drugs.In vitro metabolism of ketotifen, an antiasthmatic drug, by cultured rat, rabbit, and human hepatocytes was consistent with thein vivo metabolic pathways: namely, oxidation in rat hepatocytes; oxidation, glucuronidation and sulfation in rabbit hepatocytes; and reduction and glucuronidation in human hepatocytes (Le Bigot et al., 1987). However, the results obtained from hepatocytes should also be interpreted with caution when quantitative comparison is the purpose, since many enzyme activities decline spontaneously during hepatocyte isolation or culture.

One final consideration in metabolite profiling is the choice of drug concentrations for in vitro studies. The major metabolic pathway may be shifted, depending on the drug concentration used. The clinical studies indicated that N-demethylation is the major metabolic pathway of diazepam in humans. However, in vitrostudies in human liver microsomes showed that 3-hydroxylation was the major metabolic pathway of diazepam metabolism when a high (100 μM) drug concentration was incubated (Inaba et al., 1988). Thisin vitro and in vivo discrepancy is likely a result of differences in the substrate concentration used. Indeed, when an in vivo relevant substrate concentration (2–4 μM) is used (Yasumori et al., 1993), the major metabolic pathway of diazepam is N-demethylation in human liver microsomes.

It is clear that each in vitro system has its advantages and disadvantages. As long as their limitations are recognized and appropriate cautions and considerations are taken in the design of the studies, in vitro systems can aid in the selection of the animal species for toxicity studies as well as provide preliminary profiles of human metabolism.

Drug-Drug Interactions.

Whenever two or more drugs are administered over similar or overlapping time periods, the possibility for drug interactions exists. Because of the potential of adverse effects, drug interactions have always been an important aspect to consider during the development of new drugs. In the past, such drug-interaction studies were primarily conducted at a relatively late stage during Phase II and III clinical studies. With the availability of human tissues and recombinant human enzymes, in vitro systems have been used in recent years as screening tools to predict in vivo drug interactions at a much earlier stage before the drug is selected for the development. Because oxidative metabolism represents a major route of elimination for many drugs, inhibition of cytochrome P450s is one of the main reasons for drug interactions.

Drug metabolism is a complex process, very often involving several pathways and various enzyme systems. In some cases, all of the metabolic reactions of a drug are catalyzed by a single enzyme, while in other cases a single metabolic reaction may involve multiple isoforms or different enzyme systems. The metabolism of indinavir exemplifies the first scenario, in which a single isoform of P450, CYP3A4, catalyzes four oxidative metabolic reactions—N-oxidation, N-dealkylation, indan hydroxylation, and phenyl hydroxylation—to produce six metabolites in human liver microsomes (Chiba et al., 1996). On the other hand, the S-oxidation of 10-(N,N-dimethylaminoalkyl) phenothiazines in human liver microsomes is catalyzed by numerous P450 isoforms, including CYP2A6, CYP2C8, and CYP2D6 (Cashman et al., 1993). Similarly, amitriptyline is N-demethylated to nortriptyline in humans by numerous P450 isoforms, including CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP3A4 (Venkatakrishnan et al., 1998). Therefore, definitive identification of the P450 isoforms responsible for drug metabolism is essential in predicting potential for drug interactions.

Over the last 10 years, a great deal of information on human cytochrome P450s at the molecular level has become available. This information, along with available antibodies and chemical inhibitors, has made it possible to determine easily the P450 isoforms responsible for the metabolism of a drug. In addition to the identification of the P450 isoforms, it is also important to evaluate the relative contributions of the metabolic pathways being inhibited to the overall elimination of the drug. With the advent of commercial liquid chromatography/mass spectrometry instrumentation and the development of high-field nuclear magnetic resonance as well as liquid chromatography/nuclear magnetic resonance techniques, the relative contribution of the metabolic pathways can be readily obtained. It should be noted that a significant interaction occurs only when drugs compete for the same enzyme system and when the metabolic reaction is a major elimination pathway. Rowland and Matin (1973) developed a kinetic model to evaluate the relative contribution of the metabolic fraction on the degree of drug interaction. They concluded that a significant drug interaction occurs only when the metabolic fraction of a particular pathway being inhibited is greater than 50% of the total clearance.

Although the identification of P450 isoforms is relatively straightforward, the experimental design and interpretation of in vitro interaction studies can be complicated and tricky. One of the important criteria in in vitro drug interaction studies is the use of clinically relevant concentrations of inhibitor and substrate. The use of supratherapeutic drug concentrations may produce a drug interaction in vitro but not in vivo. In addition, the major metabolic pathway may be shifted, depending on the drug concentration used. As mentioned earlier, in vitrostudies in human liver microsomes showed that 3-hydroxylation was the major pathway of diazepam when a high drug concentration (100 μM) was utilized (Inaba et al., 1988), while the major metabolic pathway of diazepam was N-demethylation in human liver microsomes when a clinically relevant drug concentration (2 μM) was used (Yasumori et al., 1993). It should be noted that theN-demethylation of diazepam is mainly catalyzed by CYP2C19, and 3-hydroxylation is mediated by CYP3A4. This example illustrates the importance of the use of drug concentration in in vitro drug interaction studies in order to accurately define the involved P450 isoforms and predict the in vivo situation.

Another important criterion in in vitro drug interaction studies is the concentration of microsomal protein used. TheK_i values of an inhibitor may be overestimated in the presence of a high microsomal protein concentration because of the depletion of the inhibitor by nonspecific binding to the microsomal proteins and microsomal metabolism. The K_i values for ketoconazole in human liver microsomes were estimated to be approximately 8 μM when a high microsomal protein concentration (1.5 mg/ml) was used (Lampen et al., 1995), while the estimated K_ivalues were approximately 0.03 μM when a low microsomal protein concentration (0.25 mg/ml) was used (Bourrie et al., 1996;Von Moltke et al., 1996). A sixfold increase in the microsomal protein concentration resulted in a 270-fold increase in the estimated K_i values.

The choice of in vitro enzyme systems, such as liver microsomes, cDNA-based vector systems, and liver slices, is also an important factor in in vitro drug-interaction studies. For instance, the apparent K_M characterizing the hydroxylation of ritonavir, a potent HIV protease inhibitor, in β-lymphoblastoid–derived microsomes was similar to that obtained with human liver microsomes, whereas the apparentK_M characterizing theN-dealkylation and decarbamoylation of ritonavir was 30- to 300-fold lower in β-lymphoblastoid–derived microsomes than those obtained with human liver microsomes (Kumar et al., 1996). The reason for this discrepancy is unknown; however, it is clear that we should be cautious in interpreting the kinetic parameters obtained from different in vitro systems.

Furthermore, an understanding of the mechanism involved in enzyme inhibition is critical in providing a rational basis for designing experimental conditions and interpreting drug-interaction data. For example, a compound that irreversibly inactivates an enzyme will result in a decrease in the V_max but has no effect on the K_M. The pattern of the kinetic data is similar to that of a reversible noncompetitive inhibitor, which also causes a decrease in the V_max but not theK_M. Thus an irreversible inhibitor can be incorrectly referred to as a reversible noncompetitive inhibitor. The experimental results reported by Franklin (1977) are a good example. Depending on the experimental conditions, SKF-525A acts as a competitive inhibitor or metabolic intermediate (MI) complexation inducing agent. As shown in table2, SKF-525A increased theK_M values of substrates but had little effect on the V_max values when incubated with substrates without preincubation of the inhibitor. In contrast, SKF-525A decreased the V_max values of substrates and had little effect on the K_Mvalues when SKF-525A was preincubated prior to substrate addition. Thus preincubation of SKF-525A changed the kinetics of inhibition from the reversible competitive type to irreversible MI complexation.

Although it is relatively easy to assess in vitro drug interaction, it must be emphasized that the correct prediction and extrapolation of in vitro interaction data to the in vivo situation requires a good understanding of pharmacokinetic principles. The following discussion will provide a few of the basic tenets of enzyme inhibition on the pharmacokinetics. If a drug is mainly metabolized by the liver, the total clearance is approximately equal to the hepatic clearance (CL_H) that can be expressed as eq. 5 (Wilkinson, 1987):CLH=Qh·E=Qh·fb·CLintQh+fb·CLint Equation 5where Q_h is the hepatic blood flow,E is the hepatic extraction ratio,f_b is the unbound fraction of drug in blood, and CL_int, the intrinsic clearance, is a measure of the drug metabolizing activity (V_max/K_M) in the liver. Depending on the underlying mechanism of the inhibitor, theV_max value of a drug can be decreased or the K_M value can be increased. Regardless of the mechanism, enzyme inhibition always results in a decrease in the intrinsic clearance (V_max/K_M). Therefore, the concept of intrinsic clearance is the cornerstone for the extrapolation of in vitro data to the in vivosituation.

Kinetically, drugs can be classified by whether their hepatic clearance is “enzyme-limited” (low) or “flow-limited” (high). When the intrinsic clearance of a drug is very small relative to the hepatic blood flow (Q_h ≫f_b · CL_int), the hepatic clearance is low and the CL_H is directly related to f_b andCL_int, as shown in eq. 6:CLH=fb·CLint Equation 6Thus a decrease in the intrinsic clearance (CL_int) caused by inhibition will result in an almost proportional change in the clearance of “low clearance” drugs. On the other hand, if the intrinsic clearance is so high thatf_b · CL_int≫ Q_h, then the hepatic clearance is limited by the hepatic blood flow, as shown in eq. 7:CLH=Qh Equation 7Thus a decrease in the intrinsic clearance caused by inhibition has little effect on the hepatic clearance of “high clearance” drugs.

Because the hepatic first-pass effect reflects the hepaticCL_int, hepatic bioavailability (F) can be expressed as:F=1−E=QhQh+fb·CLint Equation 8and the area under the curve (AUC) after oral dosing can be described as:AUC=F·doseCLH=dosefb·CLint Equation 9As shown in eq. 9, a decrease in theCL_int caused by enzyme inhibition will yield an almost proportional increase in the AUC after oral dosing, regardless of whether it is a low- or high-clearance drug. In contrast, after iv administration, a significant decrease in theCL_int only affects the clearance andAUC of low-clearance drugs because theCL_H is independent of theCL_int for high-clearance drugs, as indicated in eq. 7.

The indinavir-ketoconazole interaction is a good example of an in vivo drug-drug interaction that is route- and drug-dependent (low- or high-clearance drug). Indinavir is a high-clearance drug that has a blood clearance of 80–90 ml/min/kg in rats and 15–17 ml/min/kg in AIDS patients (Lin, 1997). These values are greater than rat hepatic blood flow (60–70 ml/min/kg) or close to human hepatic blood flow (20 ml/min/kg). Indinavir is eliminated exclusively by CYP3A-mediated biotransformation in both rats and humans (Chiba et al., 1996; Lin et al., 1996a). In vitro studies with rat and human liver microsomes indicated that ketoconazole competitively inhibited the metabolism of indinavir, with aK_i value of approximately 0.25 μM for both rat and human liver microsomes (Lin, 1996). Coadministration of ketoconazole (25 mg/kg po) had little inhibitory effect on the clearance of indinavir and its AUC after the iv administration of indinavir (10 mg/kg iv) in rats. The clearance rate decreased from 87 ml/min/kg in control rats to 83 ml/min/kg in ketoconazole-coadministered rats. However, ketoconazole significantly increased the bioavailability of indinavir and its AUC after oral dosing. The bioavailability increased from approximately 20% in control rats to 89% in ketoconazole-coadministered rats (Lin, 1996). Ketoconazole, on the other hand, is a low-clearance drug with a clearance rate of 8–10 ml/min/kg in rats. In vitro studies with rat liver microsomes revealed that indinavir also competitively inhibited the metabolism of ketoconazole, with aK_i value of 4.5 μM. As expected, coadministration of indinavir (20 mg/kg po) in rats significantly increased the AUCs of ketoconazole twofold after both iv and oral administration of ketoconazole (Lin, 1996). The clearance of ketoconazole in rats decreased from 8.5 ml/min/kg when given alone to 4.5 ml/min/kg when coadministered with indinavir.

The ultimate goal of an in vitro drug interaction study is to predict the quantitative effect of drug inhibition in vivo. For competitive inhibition, the per cent of inhibition can be described as in eq. 10:Percentofinhibition=CLint,o−CLint,iCLint,o Equation 10=[I]/Ki1+([I]/ki)+([S]/KM) where CL_int,o andCL_int,i are the intrinsic clearances in the absence and presence of inhibitor, respectively.K_M is the Michaelis constant of the substrate, K_i is the inhibitory constant of the inhibitor, and [S] and [I] are the substrate and inhibitor concentrations, respectively.

As shown in eq. 10, the degree of inhibition depends not only on theK_M and K_ivalues of substrate and inhibitor but also on their concentrations ([S] and [I]). Both [S] and [I] continue to change as a function of time in vivo after drug administration unless they are maintained under steady-state conditions. Thus appropriate pharmacokinetic models are needed in order to obtain an accurate in vitro/in vivoextrapolation. Lin et al. (1984) successfully applied a physiologically based pharmacokinetic model to predict product inhibition. This model incorporated the K_M andK_i values together with the kinetic parameters of the plasma profiles of the parent drug and its metabolite to predict the quantitative effect of product inhibition of salicylamide on the elimination of ethoxybenzamide in rabbits after a single dose. Although the physiologically based pharmacokinetic approach can provide an accurate prediction of drug interaction, this approach is very costly and time-consuming when all of the parameters needed are being obtained. A closer examination of the literature reveals that in most cases, in vitro interaction studies are generally carried out to assess the potential of drug interaction, more or less in a qualitative sense, by comparing the relative affinities of the substrate (K_M) and inhibitor (K_i) with their concentration ranges in clinical studies. One of the most common approaches is the use ofin vitro K_i values together within vivo values of the peak plasma concentration of inhibitor to forecast the possibility of drug-drug interactions in vivo.

Even for the qualitative prediction, the in vitro/in vivoextrapolation of drug-drug interaction appears to be difficult and controversial. One of the controversies is whether the total (bound + unbound) or unbound plasma concentration of the inhibitor should be used to predict the in vivo drug interaction. A basic tenet of pharmacokinetics is that only unbound drug can diffuse across hepatocytes, and that unbound drug concentration in the blood is in equilibrium with that in the hepatocytes. Thus it is generally believed that only unbound inhibitor can compete with the substrate for the enzymes. However, there are reports that contradict this basic tenet. For example, instead of unbound inhibitor concentration, total plasma concentration of ketoconazole gave a good in vitro/in vivoextrapolation of the terfenadine-ketoconazole interaction (Von Moltke et al., 1994). Similarly, Tran et al. (1997) reported that the in vivo K_i values of stiripentol on the metabolism of carbamazepine were more consistent with the in vitro K_i values when total plasma concentrations of stiripentol, but not unbound concentrations, were used to estimate the in vivo K_i values. These authors speculated that the stiripentol concentration at the enzyme site was much higher than its unbound concentration in the blood because of a high liver/plasma partition. A similar observation has been reported for selective serotonin reuptake inhibitors (Von Moltke et al., 1998). A good in vitro/in vivo extrapolation of drug-drug interaction by selective serotonin inhibitors was obtained only when a liver/plasma partition ratio was taken into account. The issue of intrahepatic exposure of enzyme to inhibitor or substrate and its relationship with plasma concentration requires further investigation.

Sometimes, the failure of in vitro/in vivoextrapolation may originate from the nature and design of the in vitro experiments. Cimetidine, an H₂-receptor antagonist, has been well-documented to inhibit cytochrome P450-mediated drug metabolism in humans in vivo. However, the concentration of cimetidine required forin vitro inhibition of a cytochrome P450–mediated reaction is typically 100 to 1000 times greater than the plasma concentration of cimetidine associated with the inhibition of drug metabolism in patients (Knodell et al., 1991). Clearly, the in vitro data will falsely predict the potential in vivodrug interaction. Although the reason for the in vitro andin vivo discrepancy is not fully understood, studies byChang et al. (1991a, 1991b) have suggested that cimetidine may be a mechanism-based inhibitor. This may explain the in vitro/in vivo discrepancy. In vitro studies with rat liver microsomes revealed that cimetidine inhibited the activities of CYP2C11, CYP2B1/2, and CYP3A1/2, with 50% inhibitory concentration (IC₅₀) values in the range of 1.0 to 7.4 mM (Knodell et al., 1991). Preincubation of rat liver microsomes with a low concentration (0.05 mM) of cimetidine in the presence of NADPH resulted in a substantial decrease in the enzyme activities, suggesting that a mechanism-based inactivation is involved (Chang et al., 1991a). It is possible that cimetidine acts as an irreversible inhibitor in vivo but as a reversible inhibitor in vitro. Therefore, an understanding of the underlying mechanism involved in drug interaction is very important in order to provide a rational basis for the design of experimental conditions.

Prediction of In Vivo Metabolic Clearance.

One of the main objectives of in vitro metabolism studies is the quantitative prediction of in vivo metabolic clearance from the in vitro data. The prediction of metabolic clearance from in vitro systems, however, is difficult and highly controversial. Some scientists believe thatin vitro/in vivo extrapolation is possible, whereas others are less optimistic and believe that it is extremely difficult to predict in vivo clearance from in vitrometabolism data. Each group can cite examples from the literature to support its views (Sugiyama et al., 1989; Pang and Chiba, 1994; Houston, 1994; Gillette, 1984). Knowing that in vitroextrapolation is an approximation, we believe that quantitativein vitro metabolic data can be extrapolated reasonably well to the in vivo situation with a good understanding of the interdependent factors that are involved and the application of appropriate pharmacokinetic principles.

There are many examples of good quantitative in vitro andin vivo correlation. Ethoxybenzamide, an antipyretic agent, is exclusively metabolized to salicylamide by rat liver microsomes. Thein vitro V_max andK_M values (3.46 μmol/min/kg and 0.378 mM) obtained from rat liver microsomes are in good agreement with those obtained in vivo by application of a two-compartment model (3.77 μmol/min/kg and 0.192 mM) (Lin et al., 1978). Indinavir, a potent HIV protease inhibitor, exhibited marked species differences in hepatic clearance. This drug was metabolized mainly by isoforms of the CYP3A subfamily to form oxidative metabolites in all species examined (Lin et al., 1996a). The in vitro hepatic clearance obtained from incubations of rat, dog, and monkey liver microsomal preparations was in good agreement with the corresponding in vivo hepatic clearance of indinavir. Thein vitro hepatic clearances were 31, 25, and 7.8 ml/min/kg for rats, monkeys, and dogs, respectively, while the correspondingin vivo hepatic clearances were 43, 36, and 11 ml/min/kg (Lin et al., 1996a). Chiba et al. (1990)successfully predicted the steady-state concentration of imipramine and its active metabolite, desipramine, in rats by using theV_max and K_Mvalues obtained from in vitro microsomal studies. Felodipine, a calcium channel blocker, is primarily metabolized to its pyridine analog in rats, dogs, and humans. The hepatic clearances of this drug obtained from in vitro studies with hepatic microsomes were 16 liters/hr for rats, 39 liters/hr for dogs, and 259 liters/hr for humans and agreed reasonably well with those clearances observed in vivo (6.2 liters/hr, 88 liters/hr, and 321 liters/hr; Baarnhielm et al., 1986). Similarly, a goodin vitro and in vivo correlation of the clearance of cytarabine hydrochloride was reported by Dedrick et al.(1972). Furthermore, Iwatsubo et al. (1997)successfully predicted the in vivo clearance and bioavailability of YM796, a central nervous system drug for the treatment of Alzheimer’s disease, by using a recombinant system of human CYP3A4 together with knowledge of the content of this isoform in human liver microsomes. Recently, Houston and Carlile (1997) showed an excellent correlation between in vivo and hepatocyte intrinsic clearance for 21 drugs in rats. These examples clearly show that the in vivo metabolic clearance can be approximated from in vitro metabolic data if appropriate pharmacokinetic principles are utilized.

As mentioned earlier, although allometric scaling has successfully been used to predict the renal clearance of drugs in humans from animal data, the scaling usually failed to predict the metabolic clearance of drugs (Boxenbaum, 1980). To improve the prediction of metabolic clearance in humans, empirical correction factors, such as brain weight or maximum life span, have been proposed and used (Boxenbaum, 1984). Recently, Lave et al. (1997) proposed a new allometric approach that integrated in vitro metabolic data to improve the predictability of hepatic clearance of drugs in humans. For ten extensively metabolized compounds, correction of thein vivo clearance in the different animal species forin vitro metabolic clearance significantly improved the predictions in humans, compared with the conventional approaches in which clearance is extrapolated directly using body weight or correcting for brain weight.

A literature survey revealed that in some cases, in vitrometabolic data failed to predict in vivo clearance. Sources of inaccuracy in predicting the in vivo metabolic clearance may include the nature and design of in vitro experiments, presence of extrahepatic metabolism, and active transport in the liver. Unfortunately, the reason for the lack of in vitro/in vivocorrelation rarely has been understood or explained.

Recently, we carried out a study in our laboratory to examine the inductive effect of dexamethasone on the intestinal and hepatic first-pass metabolism of indinavir in rats. Pretreatment with dexamethasone (40 mg/kg/day po for 3 days) resulted in three- and tenfold increases in the V_max values in the intestinal and hepatic microsomes, while no effect was observed on theK_M values in either microsomal preparation. By using the intestinal and hepatic intrinsic clearance obtained fromin vitro V_max/K_M values and the mucosal and hepatic blood flow values, the in vitrointestinal and hepatic first-pass metabolism extraction ratios of indinavir were estimated to be 0.009 and 0.45 in control rats and 0.036 and 0.88 in dexamethasone-treated rats, respectively. For comparison, the in vivo intestinal and hepatic first-pass metabolism (extraction ratio) were measured in control and dexamethasone-treated rats. The intestinal first-pass metabolism was determined using the in situ intestinal loop technique, while the hepatic first-pass metabolism was estimated by comparing the indinavir concentrations in the systemic circulation during portal vein or femoral vein infusion of the drug. The detailed experimental procedures for the intestinal and hepatic first-pass metabolism studies have been described elsewhere (Lin et al., 1996a). As shown in table3, there is a reasonably good correlation between in vitro and in vivo hepatic first-pass metabolism extraction ratios both before and after dexamethasone induction, whereas a significant discrepancy between in vitro and in vivo intestinal first-pass metabolism extraction ratios was observed. The predicted in vitrointestinal first-pass metabolism is much lower than that determinedin vivo by approximately sixfold in control rats and tenfold in dexamethasone-treated rats (table 3). The reason for the observed discrepancy between in vitro and in vivointestinal first-pass metabolism is not clear at the present time.

One possible explanation for the discrepancy is the involvement of P-glycoprotein in the intestinal first-pass metabolism of indinavir. P-Glycoprotein, located in the apical brush border membrane of enterocytes of the small intestine, can act as an efflux transporter that extrudes a drug from inside the enterocytes into the intestinal lumen as the drug is being absorbed across the epithelial cells. A portion of the extruded drugs can be reabsorbed into the enterocytes. Consequently, P-glycoprotein increases the exposure of drugs to drug-metabolizing enzymes and hence enhances the intestinal metabolism of drugs by prolonging the intracellular residence time through the repetitive processes of extrusion and reabsorption. The effect of P-glycoprotein on intracellular residence time and intestinal metabolism in Caco-2 cells was reported by Gan et al.(1996), who used cyclosporin A as a model compound. Cyclosporin A was metabolized to a greater extent when the drug was dosed to the apical side than to the basolateral side. Recently, indinavir has been shown to be a substrate of P-glycoprotein (Kim et al., 1998). Thus it is possible that the effect of P-glycoprotein may be to increase the intestinal metabolism of indinavir by prolonging the intracellular residence time of the drug. If P-glycoprotein plays a role in intestinal metabolism, a more complex model must be developed in order to predict in vivo intestinal first-pass metabolism.

Conclusion

The use of animal models to predict pharmacological and toxic effects in humans has a long history, and it comes from the belief in the “unity life” of mammals. For years, scientists believed that the central physiological functions of circulation, respiration, and autonomic regulation were common in all mammals and that the major routes of endogenous metabolism, such as the citric acid cycle and oxidative phosphorylation, were similar in mammals. However, in the past 30 years, pharmacokineticists have failed to find an animal species in which the ADME processes of drugs are consistently the same as those in humans. In fact, it can be presumed that such an animal species will never be found.

With the recent breakthroughs in molecular biology, it is possible that in the very near future we can use transgenic (humanized) animals for studying human ADME. A number of genetically modified animals have been established as models for human genetic diseases, but the transgenic approach for evaluating the metabolism and transport of drugs has not been utilized yet to any large degree (Liggitt et al., 1992; Burki and Lederman, 1995). Using standard techniques, it may not be difficult to develop transgenic animals that express genes coding for both human CYP3A4 and P-glycoprotein in the liver and intestine. This type of transgenic animal would certainly provide valuable means for evaluating the intestinal and hepatic first-pass metabolism of drugs. Although it is ambitious, the dream of using transgenic animals in evaluating human ADME may become true within 10 years!

The concept that the properties of the whole are the sum of the properties of the parts has had a profound impact on the manner of conducting science. In their recent review article entitled “Complexity and Emergence in Drug Research,” Kier and Testa (1995)rightly pointed out that this concept has guided scientists in all different scientific fields to the belief that the route of understanding nature is through the dissection of a system into its parts, followed by the study of these parts. Once a system has been dissected into its parts, scientists attempt to reassemble the information about the parts to understand the interdependence of all controlling factors of the whole system. Drug metabolism scientists also apply this fashionable concept by using a broad spectrum of methods, from subcellular (microsomes) to cellular (hepatocytes) to organ levels (isolated perfused liver) to study drug metabolism. However, extrapolation of in vitro metabolic data to thein vivo situation is not always straightforward. As seen with the example of the intestinal first-pass metabolism of indinavir, the failure of in vitro extrapolation may be a result of the involvement of P-glycoprotein. Thus it is important to understand fully the interdependent factors that influence intestinal metabolism before a complex predictive model in which all controlling factors are incorporated can be developed.

Although there are many limitations in the applications of interspecies scaling and in vitro extrapolation of ADME data to humans, we believe, through our experience, that a good prediction of certain ADME processes in humans can be made when in vitro data are integrated with in vivo animal data. The experience with indinavir is a good example. Metabolism and pharmacokinetic studies with indinavir in animals, combined with data from human tissue preparations in vitro, allowed accurate predictions of the oral bioavailability in human subjects, while recognition of the CYP3A4-inhibitory properties of indinavir provided insight into the potential drug-drug interactions with indinavir and other drugs.