Review Article Dose-Dependent Pharmacokinetics: Experimental Observations and Theoretical Considerations

Download as pdf or txt
Download as pdf or txt
You are on page 1of 31

BIOPHARMACEUTICS & DRUG DISPOSITION, VOL.

15, 1-31 (1994)

REVIEW ARTICLE

DOSE-DEPENDENT PHARMACOKINETICS:
EXPERIMENTAL OBSERVATIONS AND
THEORETICAL CONSIDERATIONS
JIUNN H. LIN

Drug Metabolism, Merck Research Laboratories, West Point, PA 19486, U.S.A.

ABSTRACT
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 pharmacokineticshave
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.

INTRODUCTION

The pharmacological response of a drug is generally believed to be directly


correlated to its concentration at the site of action, which is in turn dependent
on a number of processes, including absorption, distribution, biotransformation
and excretion. Some of these processes, such as drug metabolism, secretion and
protein binding, involve enzymes or proteins (carriers) with definite capacities
that are saturable.
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.
The theoretical bases and possible causes of dose-dependent pharmacokinetics
have been reviewed The interested reader is referred to these

0142-2782/94/01OOO1-3 1$20.50 Received 23 February 1993


0 1994 by John Wiley & Sons, Ltd. Accepted 16 July 1993
t

reviews for further details. This review highlights the major types of dose-
dependent pharmacokinetics with unique examples, some of which were
generated in our laboratory.

LINEAR AND NON-LINEAR SYSTEMS

Mathematically, the body can be considered to be an operator that transforms


the input (dose rate) into output (plasma concentration of drugs). A linear system
exists when input-output relationships are fully described by one or more linear
equations. The most important feature of a linear system is that it obeys the
law of superposition, which states that in a linear system the output produced
by several simultaneously applied inputs is equal to the sum of the outputs of
separately applied inputs.’ A system is said to be non-linear when it does not
conform.
Practically, dose-dependent pharmacokinetics are reflected most commonly
in a greater than or less than proportional increase in the area under the drug
concentration-time curve (AUC) with increase in dose. The proportionality
between the dose (input) and AUC (output) of a drug is an important criterion
used to determine whether the drug exhibits linear or non-linear kinetics in the
body. However, based only on the dose-AUC relationship, sometimes the non-
linearity of absorption and/or disposition processes of a drug may easily be
overlooked, depending on the complexity of these processes and the relative
contribution of saturable pathways to the overall processes. Thus, careful
analysis of the data is necessary to determine whether linearity or non-linearity
obtains.

DOSE-DEPENDENT PHARMACOKINETICS
Dose-A UC disproportionality after oral administration
The oral administration of a drug is by far the most preferable route for the
development of new drugs from virtually every viewpoint. In the toxicity studies,
whenever possible the test compound is administered orally in a solution or a
suspension to animals. Traditionally, it is assumed that the increasing dose levels
equate proportionally with a margin of safety. This assumption may be true
in some cases in which doses administered are relatively small, where linear
kinetics follow, but often this is an invalid assumption, because dosages used
in the toxicity studies are relatively high and can result in non-linearity of
absorption, distribution, metabolism, and excretion. Therefore, before setting
out on a pivotal subchronic and chronic toxicity study of a new drug, a pilot
study is needed to establish a dose range in which a reasonable relationship
between blood concentration (AUC) of the drug and dosage exists to ensure
sufficient systemic exposure of animals to the drug. From a toxicological point
DOSE-DEPENDENT PHARMACOKINETICS 3

of view, it is not meaningful to increase the dose, if a plateau in the AUC is


observed.
The AUC following oral administratior, can be described from the following
relationship:

AUC =Fx Dose -- F x Dose


CLtotal fuCLint

where F is the bioavailability, i.e. the fraction of the dose which reached the
general circulation after absorption; CLtotaland CLintare the total body and
intrinsic clearance, respectively; and f, is the fraction of unbound drug in
plasma. As indicated in equation (l), AUC after an oral dose is affected not
only by absorption but also by protein binding and elimination.

Less than proportional increases in AUC. Often high oral doses result in less
than proportional increases in AUC; the following cases are given as examples.
We first consider dose-dependent absorption. There are at least three processes
that can cause dose-dependent drug absorption-dissolution rate, transit time
of drugs remaining in the regions of the GI tract, and the ability or inability
of drugs to cross intestinal barriers. Dressman et al.697 have developed a
theoretical model to demonstrate that the dose-to-solubility ratio is an important
determinant in drug absorption. Both griseofulvin and digoxin have low aqueous
solubilities (1 5 pg ml- and 24 pg ml- l , respectively) but vastly different clinical
dose ranges (100-600mg and 0-15-1 mg, respectively). The absorption of
griseofulvin is predicted to exhibit pronounced dose-dependent absorption,
whereas absorption of digoxin is predicted to be proportional to dose due to
its low dose-solubility ratio. These predictions are in good agreement with
clinical observations.
Less than proportional increases in AUC with oral dose are commonly due
to limited solubility of the drug. As shown in Figure 1, the increase in AUC
of L-365,260, a potent CCKBreceptor antagonist, is less than proportional to

Dose, mgkg Dose, msnCs


Figure 1. Relationship between AUC or C , and dose of L-365,260 in dogs receiving a single
oral dose (0)or after the last dose of three months of daily oral administration ( 0 )(mean? SD).
(After Chen et uf.,*with permission)
4 J. H. LIN

the oral dose; a 50-fold increase in dose results in a 14-fold increase in AUC.*
Because L-365,260 has poor aqueous solubility ( - 2 pg ml- l), the dispropor-
tionality is likely to be attributable to this property.
It is well acknowledged that the stomach is not an important site of drug
absorption and that drugs are not absorbed significantly until they reach the
small inte~tine.~ Thus, the rate of gastric emptying markedly influences the
extent and rate at which drugs are absorbed irrespective of whether they are
acids, bases, or neutral compounds.
L-697,661, a potent and specific inhibitor of human immunodeficiency virus
(HIV) reverse transcriptase, was rapidly cleared in rats with a total clearance
of 65mlmin-l kg-I and a half life of 80min after i.v. administration
(2mg kg-l). Comparison of the concentration of L-697,661 in the systemic
circulation during portal and femoral vein infusion of the drug indicated that
L-697,661 is subject to extensive hepatic first-pass metabolism (- 70%). When
the drug was given orally as a suspension in 0.5% methylcellulose, absorption
was slow and incomplete. Bioavailability was less than 5%. Increasing the dose,
from 40mgkg-I p.0. to 160mgkg-1 P.o., yielded only a small increase in
C,, and AUC, from 96.5ngml-l and 17-4pgrninml-' to 180ngml-I and
35 - 6pg min ml- l , respectively. The mean T,, ranged from 173 min at
40 ng kg- p.0. to 260 min at 160 mg kg- p.0. The apparent terminal half life,
ranging from 200 min to 260 min, was about three times that after i.v. dosing,
suggesting that absorption continued after reaching the peak concentration (data
on file, Merck Research Laboratories).
In order to elucidate the underlying mechanism of the prolonged absorption
observed in rats, the kinetic behaviour of 14C-L-697,661 in the stomach was
studied. Following oral administration of radiolabeled drug (40 mg kg- l ),
approximately 57%, 6O%, 46% and 25% of the dose was found in the stomach
at 1 h, 4 h, 8 h, and 24 h after dosing. These results indicate that L-697,661
markedly decreased the rate of gastric emptying, leading to a prolongation of
its own absorption.
Furthermore studies with 14C-PEG4000, a poorly absorbable compound,
suggested that the influence of L-697,661 on gastric emptying time was dose
dependent; a higher dose of L-697,661 resulted in a longer gastric emptying time.
Thus, the less than proportionate increase in AUC of L-697,661 in rats could
be, at least partly, attributed to its dose-dependent effect on the gastric emptying
time.
Although most drugs are absorbed by passive diffusion, some are absorbed
from the small intestine by specific transport processes. When passive diffusion
is minor, a less than proportional increase in AUC may occur as the active
process is saturated. Examples of actively absorbed drugs are L-dopa,I2
methyldopa, l 3 cephalexin,I4 c y ~ l a c i l l i n , ~bestatin,
~ . ~ ~ l6 cephradine, l7 and
methotrexate.
In an in situ study, Tsuji et al. l9 have shown that absorption of cyclacillin
by rat intestine is concentration dependent. The absorption of cyclacillin can
DOSE-DEPENDENT PHARMACOKINETICS 5

be described by simultaneous Michaelis-Menten and first-order kinetics. The


apparent V,, and Km values for the active transport system are estimated to
be about 75 pg min- and 420 pg ml- respectively. Thus, a saturable
absorption is expected when cyclacillin concentration in the intestinal lumen
is greater than 1 mgml-I.
Similarly, absorption of methotrexate by the in situ perfused rat intestine has
been shown to be saturable and best described by Michaelis-Menten kinetics,I8
suggesting that the drug is absorbed by an active transport system. Consistent
with these results, the absorption of methotrexate has been reported to be dose
dependent in patientsZoand in animals.21
We next consider increased elimination clearance. Since the concentration
of drug in blood after oral dosing is a function of both absorption and
elimination, the AUC is a net result of both processes. Thus a less than
proportional increase in AUC may be due to an increase in elimination as well
as to a decrease in absorption.
For low-clearance drugs, binding is an important determinant in drug
elimination, and total body clearance is directly proportional to the unbound
~ ~ . ~ ~ of protein binding of a drug may result in
fraction in p l a ~ m a . Saturation
rapid drug elimination. Thus, a less than proportional increase in AUC with
dose may be observed after oral administration of high doses. Both blood AUC
and C,, for L-693,612, a potent carbonic anhydrase inhibitor, increased less
than proportionally to the dose after oral administration of 0-05-25 mg mg- '
in rats (Figure 2)." When the dose was increased 500-fold, AUC and C,,
increased only about 10- and 25-fold, respectively. The lack of oral dose
proportionality could not be attributed to non-linear absorption since a similar
pattern and degree of disproportionality was observed after i.v. administration
of the same range of doses (Figure 2). The total blood clearance of L-693,612
increased with increasing i.v. dose, due to saturation of carbonic anhydrase
binding sites in erythrocytes. The total blood clearance was 0.52ml h-I kg-
at 0*05mgkg-' and 28mlh-'kg-' at 25mgkg-I.
Similarly, a less than proportional increase in plasma AUC of cefixime, a
new oral cephalosporin, with dose was observed in dogs, when the drug was
given in a dose range of 6.25-200mg kg-I P . O . ~Within
~ the 6.25-50mg kg-I

Figure 2. Blood AUC of L-693,612 in rats after oral and intravenous administration of
0.05 mg kg-', 0.25 mg kg-', 5 mg kg-I, and 25 mg kg-I. (After Wong et u I . , ~ with
~ permission)
6 J. H. LIN

range, the bioavailabilty was about 55% and was not affected by the dose.
However, as the dose was increased from 50 mg kg-I to 200 mg kg-I, the AUC
increased less than proportionally to the dose and the bioavailability dropped
from 55% to 27%. Since the drug exhibited concentration-dependent protein
binding, with the unbound fraction ranging from 7% at 0.5 pgml-I to 45%
at 300pgml-', it is believed that the less than proportional increase in AUC
is due to saturation of protein binding sites in plasma, resulting in enhanced
clearance.
Runkel et af.26,27 reported that the AUCs of naproxen in plasma increased
linearly with dose for increments up to a regimen of 500mg twice daily, but
larger doses resulted in an apparent plateau of the AUC. The authors concluded
that a disproportionate increase in renal clearance caused by a disproportionate
increase in the fraction of unbound drug was the most likely explanation for
the observed plateau effect.
Autoinduction is a dose- and time-dependent phenomenon in which the
elimination clearance of a drug increases following multiple doses and the
increase in clearance is greater after a high dose than after a low d ~ s e . ~ * - ~ O
Bertilsson et af.29 have studied the time course of autoinduction of
carbamazepine in epileptic children. Drug clearances were measured during each
of dose administration on day 1, day 6, days 21-36, and days 146-162 using
stable isotope technique. Clearance started to increase on day 6, continued to
increase at days 21 -36, but showed no further increase thereafter. Recently,
studies with P-naphthoflavone, a widely used enzyme inducer, indicate that
autoinduction of the compound could occur as early as 1-2 h after a single i.v.
administration in the rat.31
L-699,392, a potent leukotriene biosynthesis inhibitor, is currently under
investigation for the potential utility in the treatment of asthma. In a five-week
toxicity study, Rhesus monkeys received daily oral doses of L-699,392 at
30mg kg-I, 60mg kg-I, and 120mg kg-I. On day 1, AUC values increased
with dose in an approximately proportionate manner: 115 pg h ml- I ,
301 pg h ml-l, and 443 pg h ml- respectively. After 22 d of drug treatment,
AUC values declined markedly to 30 pg h ml- l , 43 pg h ml- I , and 68 pg h ml- I,
respectively, suggesting that autoinduction of L-699,392 occurs in monkeys
(Tocco et af.,unpublished data). On day 22, a fourfold increase in dose resulted
in only a twofold increase in AUC. The ratio of the AUC on day 1 to that on
day 22 was about 3 . 8 for the lowest dose and 6.5 for the highest dose, indicating
that the autoinduction of L-699,392 is dose dependent. In an i.v. study designed
to show autoinduction, three monkeys received I4C-L-699,392(2 mg kg- i.v.)
on the day before (day 1) and the day after (day ll), nine oral doses of
60mg kg-' d-I. Plasma clearance was 2.5-fold greater at the end of the
treatment period. These results support the hypothesis of autoinduction.
Similar results have been reported by Batra and Y a ~ o b for i ~ ~monkeys
receiving compound D, 5 mg kg-I p.0. twice daily for 30 d. At this dose,
elimination of compound D was not induced, as indicated by comparison of
DOSE-DEPENDENT PHARMACOKINETICS 7

concentrations on days 1 and 30. When the dose was increased to 20 mg kg- I ,
the plasma concentrations on day 30 were significantly lower than those on day
1 . In addition, C,, and AUC, which increased in a dose-proportional manner
on day 1, increased in a much less than proportional manner on day 30.

Greater than proportional increase in AUC. An increase in AUC which is


greater than dose proportional is more likely to be attributable to a decrease
in elimination than to an increase in absorption. However, in some cases,
absorption can be increased with increasing dose due to some unique
mechanisms.
We first consider decreased elimination clearance. L-706,631, a potent
leukotriene D4 receptor antagonist is currently under investigation for potential
utility in a number of disease states, such as bronchial asthma, allergic rhinitis,
and allergic conjunctivitis. Following 5 mg kg-I and 25 mg kg-I P.o., AUC
values increased in a greater than proportionate manner. The AUC increased
from 38 pg min ml-I at 5 mg kg-l p.0. to 357 pg min ml-I at 25 mg kg-I in
rats, and from 716 pg min ml- to 5686 pg min ml- in Rhesus monkeys.33The
greater than proportional increase in AUC is likely to be due to non-linear
clearance. At 1 mg kg-I and 5 mg kg-I i.v., total body clearance was about
40mlmin-I kg-I and 14mlmin-'kg-' for the rat, and 5 mlmin-I kg-I and
2mlmin-I kg-I for the monkey. Insofar as a significant first-pass effect in
monkeys is not predicted by the clearance, the disproportionate increase in the
AUC is most likely to be attributable to decreased elimination clearance after
high doses. On the other hand, a significant first-pass effect in rats is expected.
Thus, the disproportionate increase in AUC in rats is due to both a decrease
in elimination and an increase in bioavailability; the latter is due to saturation
of hepatic metabolism during the first pass of the drug through the liver, which
effect we consider next.
Salicylamide is a classic example of a drug that saturates hepatic metabolism
during the first pass through the liver. The disproportionality in AUC with oral
doses is associated with an increase in bioavailability due to a decrease in first-
pass effect (Figure 3).34 The AUC-oral-dose relationship of propranolol is also
said to be non-linear for the same r e a s ~ n . ~ ~ . ~ ~
Finally, we consider other mechanisms. Alendronate, a potent antiosteolytic
bisphosphonate, is poorly absorbed in laboratory animals and humans (<2%).
The poor absorption of alendronate can be attributed to its very poor
lipophilicity (octanol/buffer partition coefficient = 0.0017), which prohibits
transcellular transport across the epithelial membranes, and to its negative charge
at the physiological intestinal pH (4-8), which hinders pericellular transport.
Over a wide dose range of 2-40mg kg-l, the extent of absorption in rats
increases with increasing dose, ranging from 0.6% to 3.5% of the dose.37
Since alendronate is a strong chelating agent, the increased absorption with dose
is believed to be attributable to the interaction between the drug and calcium
in the gastrointestinal fluid. At low doses, a major fraction of alendronate may
8 J. H.LIN

ci 25 r
s
1000

a 1
2.5
10

m
5
-
5# \
a I
, -
_
0 50 100 150 200 250 300
Time (min)

Figure 3. Plasma concentration and AUC increase disproportionately with the dose of oral
salicylamide: 0,0.3-1 S Og; 0 , 2 g. (After with permission)

be associated with calcium and the calcium complex cannot be absorbed, but
the fraction becomes less significant when the doses are increased. This
hypothesis is supported by the fact that co-administration of other chelating
agents, such as EDTA and citric acid, enhanced the absorption of alendronate.

Dose-dependent tissue distribution


It is generally believed that only the unbound drug can diffuse across
membranes that restrict distribution from the vascular compartment to the tissue
compartment and vice versa. Therefore, changes in drug-protein binding in
plasma and tissues can affect the distribution of drugs in the body. Kinetically,
drug distribution is often expressed in terms of the volume of distribution,
usually defined as the ratio of amount of drug in the body to the plasma
concentration.
The simplest quantitative expression relating the volume v d of distribution
to plasma and tissue binding is given by

where Vp is the plasma volume, Vt is the tissue volume, and fu andf, are the
fraction of unbound drug in the plasma and tissue, re~pectively.~**~~
Equation (2) assumes that binding of drugs to albumin takes place only in
the vascular compartment. Since approximately 59% of total body albumin is
located in the interstitial fluid of the extravascular space,4oOie and Tozer41
expanded equation (2) to the following:
DOSE-DEPENDENT PHARMACOKINETICS 9

where RE,, is the ratio of the amount of protein in the extracellular fluid
outside the plasma to that in the plasma and VE is the extracellular space minus
the plasma volume. It is evident that the volume of distribution is dependent
not only on plasma protein binding Cf,) but also on tissue binding Cft).

Saturation of protein binding in plasma. Assuming that a drug is bound to


n discrete sites on a protein molecule, that all these sites have the same affinity
for the drug, and that all sites are mutually independent, the fraction f, of
unbound drug can be expressed as

where Kd is the dissociation constant at equilibrium, Df and p t are


concentrations of unbound drug and total protein (bound and unbound),
respectively, and n is number of binding sites on the protein molecule.
As indicated in equation (4), the binding of a drug to protein is determined
by four factors -the drug concentration, protein concentration, number of
binding sites per molecule of protein, and the dissociation constant. Regardless
of the affinity, the maximum possible binding capacity is equal to the product
of the molar concentration of protein and the number of binding sites for the
drug per molecule. For example, the concentration of albumin in plasma in
healthy subjects and laboratory animals is about 6 5 0 ~ MTherefore,
.~ for a
drug with a molecular weight of about 300, a concentration of 200 pg ml- or
greater would be required to saturate a single binding site.
As indicated in equations (2) and (3), an elevated unbound fraction f, in
plasma will result in an increase in the volume of distribution. A significant
correlation between the unbound fraction and volume of distribution has been
reported for p r o p r a n ~ l oand have demonstrated
l ~ ~warfarin.43Sawada et al.44945
that the species differences in the distribution volume of drugs are mainly due
to species differences in plasma binding Cf,), and there is no interspecies
variation in the distribution volume of unbound drugs.
Diflunisal, a nonsteroidal anti-inflammatory drug, is extensively bound to
plasma protein (>99070)and in vitro binding studies revealed that albumin is
the major binding component for diflunisal in plasma.46 The binding of
diflunisal to plasma protein is concentration dependent; the unbound fraction
increased from about 0.25% to 4.2% when the drug concentration increased
from 5 pg ml- to 300pg ml- I . Since the binding of diflunisal is non-linear in
this concentration range, its volume of distribution is expected to be dose
dependent; the volume of distribution of diflunisal increased from 130 ml kg- '
at 3 mg kg-l i.v. dose to 260ml kg-I at 60mg kg-l i.v. dose.47
Children who ingested intoxicating doses of aspirin had a volume of
distribution of salicylate which was twofold greater than in patients receiving
10 J. H. LIN

normal low doses.48This dose-dependent increase in distribution is believed to


be due to an increase in the unbound fraction of plasma ~ a l i c y l a t e . ~ ~
The concentration of al-acid glycoprotein in plasma can vary considerably
in several physiological and pathological conditions. In healthy people, the
concentrations of al-acid glycoprotein range from 10 pM to 30 PM.~OIn some
diseases (e.g., nephritis), the levels can be as high as 1 3 0 ~ M .Although
~' it is
generally believed that basic drugs bind mainly to al-acid glycoprotein and acid
drugs bind to albumin, it has been shown that acidic drugs bind to al-acid
glycoprotein as ell.'^,^^
Theoretically, since the concentration of al-acid glycoprotein is much lower
than that of albumin in plasma, an increase in volume of distribution due to
saturation of binding sites is more likely to occur with basic drugs. Actually,
this is not the case because basic drugs may also bind to other plasma proteins
including albumin. The relative contribution of each plasma protein to the overall
binding may change as the drug concentration increases.s0To our knowledge,
there is no example of dose-dependent changes in volume of distribution due
to saturation of al-acid glycoprotein binding sites.
Glucocorticoids are transported in the circulation by transcortin, an a-
globulin.54This protein binds steroids non-covalently with a very high affinity
and a low capacity. The affinity constants of prednisolone for transcortin and
albumin in normal healthy subjects are approximately 3 x lo7M - and
2 x lo3M - l, respectively; and the respective binding capacities are about
6x 10-7M and 6x 10-4M.S5Binding of prednisolone to plasma protein is
concentration dependent; the bound fraction decreased from about 95% at
20 ng ml- I to 60% at lo00 ng ml- I . The steady-state volume of distribution in
normal subjects increased from 241/1-73 m2 to 52 1/1-73 m2 over the 5-40mg
i.v. dosage range.s5

Saturation of protein binding in blood cells. Drugs present in blood are


distributed not only in plasma but also to a varying extent in erythrocytes,
leucocytes, and platelet^.^^.^^ There are three major components in the
erythrocytes capable of binding drugs -hemoglobin, carbonic anhydrase, and
the cell membrane. For example, binding of phenothiazines, pentobarbital, and
diphenylhydantoin to hemoglobin has been d e m ~ n s t r a t e d .Binding
~ ~ ~ ~ ~of
MK-417, MK-927, and acetazolamide to carbonic anhydrase60-62and imipramine
and chlorpromazine to the membrane63.64have also been demonstrated.
Carbonic anhydrase, a zinc metalloenzyme, predominates in red blood cells
of all mammals, accounting for more than 90% of the enzyme in the body.65
The main function of the enzyme is to regulate C 0 2 and O2 exchange between
blood and both alveolar air sacs and peripheral tissues. The concentration of
carbonic anhydrase is about l00pM in rat erythrocytes and 150pM in human
erythrocytes.61.66 MK-417 and MK-927 are potent carbonic anhydrase
inhibitors currently under clinical investigation as topical ocular hypotensive
agents for the treatment of glaucoma.
Table 1. Pharmacokinetic parameters of MK-417 in blood of anemic, polycythemic, and normal rats after i.v. administration of various
doses of the drug (meankSD, n=4-7)
Normal rats Anemic rats Polycythemic rats

0.05 45.7k3.9 80.8k7.3 1.29k0.09 41.2k3.0 68.425.5 1.15+0.08 52.827.5 79*4&12.2 1*01+0*16
0.1 49.4k9.9 76-7514.2 1*02+0.19 35.352.3 57.5f4.8 1*08+0-12 54.359.0 77.456.0 0.98020.06
0.2 41'623.5 76.0k5.7 1.31+0.08 34'759.2 105.759.YSb 2.06k0.5Pb - - -
0.5 38.6 5.1
+ 113.0 l9.4'* 2.23 f0*28'-' 35.4 k 3.5 166.9 2 20a-' 3.63 0.33a-C39.3 k 6.7 73 * 0 f24-0
+ I -25 0.12
+
1 37 -2k 4.7 147-75 23 * 2"-' 2 a74 2 0.32'-' 32.8 6.3 328 * 42 46-6a-d 8 -02 2 1.4a-d 40.9 5 10.3 I 11.7 2 30. 78.b.d 1 .84 2 0.25a'b*d
2 38.3 5 3.2 408.0 5 98. P-' 9.07 1.94"-' 33.9 2 13.0 831.4 2 346"- 17* 7 2 2-2"-' 40.8 2 6.2 222.0 2 28 .2a.b.d.c4.37 - + 0.61a.b,d.e
Analysis of
variance NS p<O*OI p<o*o1 NS p<O-OI p<O*OI NS p<O*Ol p<O*OI
aSignificantly different from 0.05 mg kg-' dose at p<O.Ol.
bSignificantly different from 0.1 mg kg-' dose at p<O.O1.
'Significantly different from 0.2 mg kg-' dose at p<O*Ol.
dSignificantly different from 0.5 mg kg-' dose at p<O*Ol.
'Significantly different from 1 mg kg-l dose at p<O.O1.
12 J. H. LIN

Since the enzyme in the vascular space is limited, high doses of carbonic
anhydrase inhibitors can saturate the enzyme. In rats, the volume of
distribution measured in terms of whole blood concentration of MK-417
increases dramatically as the dose is increased, from 80 ml kg- at low dose
(0.05 mg kg- l) to 400ml kg- at high dose (2 mg kg-I) (Table@'.)1 Interestingly,
non-linear kinetics based on whole blood concentrations of MK-417 occurred
at different dose levels among anemic, polycythemic, and normal rats, when
the dose exceeded 0.2mgkg-l in anemic rats, and 0.5mgkg-I and
1 -0 mg kg- I in normal and polycythemic rats, respectively (Table 1).
Furthermore, dose-dependent changes in volume of distribution were greater
in anemic rats than normal and polycythemic rats. Each of these observations
can be explained by the strong and saturable binding of the drug to carbonic
anhydrase within erythrocytes. The volume of distribution (80 ml kg- I ) at low
dose, approximately equal to the blood volume of 70ml kg-', is consistent
with the fact that the enzyme is mainly located in red blood cells.
Kinetically, the blood cells form their own compartment in the circulating
blood. Since it is the blood flow that perfuses through the tissues and organs,
whole-blood clearance is generally a more appropriate measure of organ function
than plasma clearance. However, for analytical reasons, plasma concentration
of drugs rather than blood concentration is usually measured. Assuming rapid
equilibrium between blood cells and plasma, the total-blood clearance CLb can
be estimated from plasma clearance CLp as follows:23

Equation ( 5 ) is valid when the blood/plasma concentration ratio is constant


over the concentration range studied. However, the blood/plasma ratio may not
be a constant value, so that the extrapolation of total-blood clearance is invalid.
For example, the blood/plasma ratio was about 150 for MK-927 in rat blood61
and about 80 for MK-417 in rabbit when the drug concentration in
blood was less than 5 pg ml- l, and the ratios decreased dramatically to about
five as the drug concentration increased to 15 pg ml- l. Similarly, concentration-
dependent erythrocyte distribution was observed for cyclosporin in human
blood.68 The fraction of cyclosporin present in erythrocytes decreases from
about 70% at 1 pgml-' to about 15% at 10pgml-I. In these cases, a direct
determination of total-blood clearance based on blood concentrations is desirable.

Saturation of protein binding in tissues. In tissues, the components that are


able to bind the drugs include non-specific binding proteins, enzymes, and
receptors. Enzymes and receptors are probably the most important from a
pharmacological point of view since they support the therapeutic effect, but
it is very difficult to distinguish them from the non-specific binding sites, because
the binding capacities of enzymes or receptors are so low that they are masked
by the non-specific binding.
DOSE-DEPENDENT PHARMACOKINETICS 13

Specific binding was elegantly demonstrated by Tocco et al.69 in their


stereospecific binding studies with the P-adrenergic blocking agent timolol. The
1-isomer (timolol) was bound to stereospecific (20%) as well as non-specific
(80%) sites in the particular fraction (8500g) of the rat heart, while the d-isomer
of timolol was bound to non-specific sites only. Stereospecificity was not
observed in the cytoplasmic fraction where most of the drug was found.
A tracer radioligand technique using 3H-Ro 15-1788 was used to measure in
vivo benzodiazepine receptor occupancy in mice.70The non-specific binding of
the ligand to brain tissue was determined when the animals were treated with a
saturating dose of clonazepam ( 5 mg kg- i.v.). The difference in radioactivity
in brain in the absence and presence of a saturating dose of clonazepam was then
defined as the specific binding to receptors. This method was validated by com-
parison with in vitro and autoradiographic techniques. This approach uniquely
distinguishesthe specific and non-specific binding of benzodiazepinesto receptors.
Binding of drugs to tissue proteins, like plasma protein binding, can be
saturated with increasing dose (concentration), resulting in a change in the
volume of distribution. However, saturation of tissue binding and of plasma
protein binding can occur simultaneously. Since the volume of distribution is
directly related to plasma protein binding and inversely related to tissue binding
(equation (2)), the magnitude of one type of binding will offset the other,
complicated by the likelihood that not all tissues are saturated to the same extent.
Depending on the degrees of non-linear tissue binding, the volume of distribution
of a drug may not be sensitive to the changes in tissue binding. Therefore, it
is desirable to determine the non-linearity of tissue binding by examining the
ratio of the drug concentration in a given tissue to that in plasma. At equilibrium,
the tissue/plasma concentration ratio should be equivalent to the ratio of
unbound fraction in plasma to that in a non-eliminating tissue. The situation
in an eliminating tissue is more complex; the ratio may be affected by blood
flow to the tissue and the eliminating p r o c e ~ s e s . ~ ~
The liver plays a key role in drug metabolism and is often the site of
pharmacological action. Thus, the distribution and binding of drugs to liver
tissue can be pharmacokinetically and pharmacodynamically important.
L-154,819 and L-654,969 are active metabolites of the HMG CoA reductase
inhibitors, lovastatin and simvastatin, respectively. Tissue binding of L-154,819
and L-654,969 to liver was studied in rats using a single-pass isolated-liver
perfusion technique. When the ratio of steady-state liver concentration to output
concentration C,,,, was plotted as a function of C,,,, for L-654,969 a biphasic
curve was observed. Initially, the ratio decreased dramatically when C,,,,
increased, and then remained relatively constant at high C,,,, (Figure 4) (Xu
and Lin, unpublished data), indicating that liver-tissue binding of L-654,969
was concentration dependent and would be described by two binding classes:
one being of high affinity and low capacity, presumably the HMG CoA
reductase; the other being of low affinity and high capacity. Similar results were
observed for L-154,819 (Figure 4).
14 J . H.LIN
100
L-154,819

u?
& 40
>

20

'0 5 10 15 20 '0 5 10 15 20
C P S (rdml) C P S (rdml)
Figure 4. Liver/output perfusate concentration of L-654,969 and L-154,819 at steady state in single-
pass perfused-rat-liver studies

Studies in animals indicate that the concentration of warfarin is much higher


in the liver than in plasma at low concentration range. The liver/serum ratio
of S-warfarin was about 10 when concentrations were less than 0.2pg ml- in
serum, and about 0-25 when concentrations were greater than 2 pg ml- 1.72 The
non-linear distribution suggests that binding of warfarin to liver tissue involves
at least two classes of binding sites, one with high affinity and limited capacity
(specific binding) and the other with low affinity and relatively unlimited capacity
(non-specific binding). Although warfarin acts on the liver, the time course of
pharmacological effects correlates with drug in plasma rather than drug in
liver.73This interesting observation suggests that the protein binding sites of
warfarin in the liver are mostly non-specific.
Similarly, non-linear distribution of methotrexate was observed in the liver
of mice.74 At 0.01 pg ml- in plasma, the liver/plasma ratio was about 50,
decreasing to about 10 at 1 pg ml- l . The non-linear tissue binding is believed
to be associated with specific binding of methotrexate to dihydrofolate reductase.
Non-linear liver/plasma ratio was also observed with ouabain in rabbits .75
Skeletal muscle, the largest tissue of the body, accounts for approximately
40% of the body weight and is composed of numerous fibers of the polymerized
proteins, actin' and myosin. Normally, the binding of drugs to muscle tissue
is relatively low; however, the contribution of muscle-tissue binding is significant
because of its large mass.76
In an i.v. infusion study in rabbits, the muscle/plasma concentration ratio
of ouabain was found to be concentration dependent, decreasing from about
0-65 at a mean concentration of 0.3nM to 0.26 at a concentration of
50011M.'~
Similarly, concentration-dependent distribution of S-warfarin was observed
in rats. The muscle/serum concentration ratio was about 0 * 3 at concentrations
less than 0.2 pg ml- and about 0 - 1 at concentrations greater than 2 pg ml- 1.72
d-Tubocurarine, a classic neuromuscular blocking agent, is widely distributed
in all tissues, with the highest levels in muscle tissue. A high dose of the drug
given to dogs results in a higher fraction of the dose in the muscle than does
DOSE-DEPENDENT PHARMACOKINETICS 15

a low dose.77 Approximately 30% of the dose is presented in the muscle at


30min after an i.v. dose of 0-3mgkg-l, and about 60% after 1.0mgkg-l.
The reason for the large increase in distribution of d-tubocurarine in muscle
after high doses is not clear.
After administration of 3H-digoxin intravenously t o dogs, high
concentrations of the drug were present in heart tissue. The heart/serum
concentration ratio was about 50 at a serum concentration of 1 ng ml-’, while
the ratio was decreased to about 10 at serum concentration of 10ngml-1.78
Human heart tissue was analyzed for digoxin at post mortem at various times
after administration of a single dose of digoxin, and non-linear distribution of
digoxin between plasma and heart tissues was also observed.79
Non-linear distribution of ouabain was also observed in the heart tissue of
rabbits. The heart/plasma concentration ratio decreased from 15 at a plasma
concentration of about 0.5 nM to about two at a concentration of 500 nM.75
Both ouabain and digoxin have a large volume of distribution, approximately
five to 10 times that of body water (0.61 kg-1).80,81It is believed that the
specific binding of cardiac glycosides to Na+, K+-ATPase plays an important
role in the tissue distribution. The distribution volume of digoxin decreased from
about 61 kg-l in normal subjects to 41 kg-1 in patients with chronic renal
failure. Zannad et aLS2have shown that erythrocyte Na+, K+-ATPase activity
is diminished by renal failure, and they postulate that the reduced Na+,
K -ATPase activity may be responsible for decreased binding of digoxin and
+

the decreased volume of distribution.


Equilibrium of drug-protein binding is normally very rapid compared to rates
of distribution and elimination. For drugs with a binding dissociation constant
less than lo-’ M, the mean residence time on protein is less than a fraction of
a ~ e c o n dHowever,
.~ the binding of cardiac glycosides to Na+ , K -ATPase is
+

very slow .84985 The association and dissociation processes take place in times
of the order of hours depending on the tissues and animal species.86 By
incorporating a slow and saturable process in the ouabain binding to N a + ,
K -ATPase, Harashima et al. 87 have successfully developed a kinetic model
+

in predicting the time course of ouabain tissue distribution based on in vitro


binding data.
Unlike in other organs, the endothelial cells in brain capillaries are joined
by continuous tight intercellular junctions. This suggests that chemicals must
pass from blood to brain through cell membranes rather than through
pericellular spaces. The uptake of drugs by the brain is mainly by simple
diffusion and correlates well with lipophilicity.88
Most nutrients in the circulation are water-soluble compounds that would
not traverse the blood-brain barrier (BBB) in the absence of special carrier-
mediated transport systems. There are at least eight different transport systems
for nutrients and thyroid hormone on the BBB.89*90
Some drugs are reported to penetrate the brain by an active transport system.
For example, L-dopa is known to be transported through the BBB by a large
16 J. H. LIN

neutral amino -acid carrier system with K , and V,, of 0-43mM and
63 nmol min- g - l , r e s p e ~ t i v e l y . ~However,
~ since the therapeutic
concentration of L-dopa is relatively low (2-5 FM), the distribution of the drug
into the brain would not be expected to be dose dependent.
a-Methyldopa, a drug used widely in the treatment of hypertension, is also
believed to traverse the BBB by the large neutral amino-acid carrier system.92
The rise in levels of a-methyldopa in the brains of rats after an injection of
the drug was depressed when large neutral amino acids, but not acidic amino
acids, were co-administered with the drug.

Other mechanisms. Alendronate, a new bisphosphonate, is currently under


investigation as an antiosteolytic agent in the treatment of a broad range of
bone disorders. Following i.v. administration (1 mg kg-I), approximately
60-70'70 of the dose was taken up by the bone and 30-40% was excreted in
the urine in laboratory animals.93 Like other bisphosphonates, alendronate
binds to hydroxyapatite in the bone. The uptake by the bone is less than
proportional to the dose in rats receiving single i.v. doses greater than
lOmg kg-I i.v.94,95A 500-fold increase in dose (0-1-50mg kg-l) resulted in
a 300-fold increase in drug concentration in bone suggesting that accessible
hydroxyapatite in bone tissues is limited and saturabie. The uptake clearance
of alendronate by tibia in rats is 0.2mlmin-l/g tissue following 1 mg kg-l i.v.
When the dose is increased from 1 mg kg- to 20 mg kg- l, the uptake clearance
declines from 0.2 to 0*07mlmin-1g-1.
Although alendronate distributes mostly in calcified tissues, the drug appears
in all non-calcified tissues of rats as well. Following i.v. administration at low
dose (1 mg kg- l ) , about 63% was found in non-calcified tissues within 5 min,
declining to 5 % in 1 h. However, when the dose was raised to 30 mg kg-I i.v.,
alendronate accumulated in kidneys, liver, lung, and spleen. Approximately
20-30% of the dose remains in these non-calcified tissues 48 h after dosing.94
Accumulation was also reported by other laboratories when bisphosphonates
were given intravenously at high doses. Monkkonen et a1.96,97reported that
dichloromethylene bisphosphonate and 3-amino-1 -hydroxypropane-1 ,1-
diphosphonic acid accumulates in the liver, spleen, lung, and kidneys of mice
and rats. The dose-dependent phenomenon of accumulation was postulated to
be due to phagocytosis of large complexes of bisphosphonates with metal ions
at high concentrations after high d0ses.~8
Studies in the 1950s on the distribution of thiopental and polychlorinated
hydrocarbons (e.g. DDT) have led to the notion that drugs accumulate in adipose
tissue by partitioning more dependent on lipophilicity than on binding. Recent
studies by B i ~ k e have
l ~ ~ shown the absence of an expected correlation in
adipose uptake distribution and lipophilicity (octanol/water partition). Recently,
Cheung and Levy72reported that the fat/serum ratio of S-warfarin is highly
dependent on the unbound fraction of the drug, suggesting that only free drug
can partition into the adipose tissue. However, they also found that the
DOSE-DEPENDENT PHARMACOKINETICS 17

fatlserum concentration ratio is concentration dependent, suggesting that fat


tissue does contain binding site for warfarin.
Thus, the lipophilicity concept of adipose-tissue distribution should be re-
evaluated. If lipophilicityis a major determinant, uptake by adipose tissue should
be independent of dose, whereas if binding sites are the determining factor,
uptake should be dose dependent.

Dose-dependent elimination
Drugs are eliminated from the body by a number of processes, including
biotransformation and excretion. The efficiency with which the body eliminates
a drug is often expressed in terms of total clearance. The total clearance is equal
to the sum of individual and simultaneously occurring organ clearances. The
organ clearance CL can be expressed as

CL =
Wu CLint (for the well stirred model)
Wu CLint
CL = Q(1 - efu CLnt / Q ) (for the parallel-tube model) (7)

where Q is the blood flow rate to the eliminating organ, fu is the unbound
fraction in blood and CLint is the intrinsic clearance, a measure of the
intracellular removal of drug, described by the Michaelis-Menten equation

where V, is the capacity of the enzyme or carrier systems, K, is the Michaelis


constant and C,is the unbound drug concentration. It is evident that the CLint
is constant when the concentration is very small in respect to K, with a limited
value of V,,/K, . However, non-linearity occurs when the drug concentration
approaches the value of K,.
Since a drug is normally eliminated by a combination of a number of
elimination processes, the determination of linearity or non-linearity of
elimination becomes more complicated as the number of processes increases.
Depending on the relative contribution of saturable pathways to overall
elimination, non-linearity may easily be overlooked. The influence of non-linear
process on plasma profiles has been mathematically illustrated by van Rossum
et al.83 Assuming two competing pathways of elimination, only one of which
is saturable, the non-linearity is only apparent if the saturable pathway accounts
for more than 30% of the overall elimination (Figure 5). Thus, careful
examination of each individual elimination pathway is necessary in order to
determine whether linearity or non-linearity exists.
18 J. H. LIN
Plasma concentration(mgh; log scale) . Plasma Concentration (mgh: log scale) 1

(h) (h)
Figure 5 . Plasma concentration curves in the case of combined linear and non-linear kinetics of
elimination. (a) Dose variation for a theoretical drug that for 67% (f-0.67) is eliminated by a
potentially capacity-limited pathway. A straight part is observed at low but also at high plasma
concentrations. Non-linearity occurs at plasma concentrations around the K, value. (b) Variation
of the fraction of body clearance that represents elimination via a capacity-limited pathway. Non-
linearity is only apparent if the capacity-limited pathway accounts for more than 30% of clearance.
(Q, is the maximum metabolic conversion rate.) (After van Rossum er UI.,'~with permission)

Capacity-limited biotransformation. Ethanol oxidation by alcohol


dehydrogenase occurs primarily in the liver. The apparent K , value determined
in vitro is about 2.0mM for humans, monkeys, and rats.lOOl'O'This implies
that non-linear kinetics can occur when the concentration of ethanol in blood
approaches 0.1 mgml-l (2.0mM). In vivo studies have shown that non-
linearity of ethanol elimination occurs after administration of a 1 g kg-' dose
when the blood concentration is around the K , value.Io2
Ethoxybenzamide, an antipyretic agent, is eliminated by a single oxidative
pathway (deethylation) to form salicylamide. In vitro studies with rat liver
microsomes showed a K , value of 0.38mM (65pgml-').lo3 The plasma
concentrations of the drug were below 0.20 mM following 20 mg kg- i.v. and '
above 0.50mM up to 2 h after 100mgkg-' i.v. As expected, non-linear
kinetics was observed after the high dose, and total clearance declined from
7 - l m l m i n - ' k g - ' at 20mgkg-' to 3*8mlmin-'kg-' at 100mgkg-1.103
Phenytoin is also known to be eliminated primarily by a single oxidative
pathway to form the hydroxylated metabolite, 5-@-hydroxyphenyl)-5-
phenylhydantoin. K , values were estimated to be in the range of 1 pg ml- to
15 pg ml-l (4-60 pM).'" The consequences of capacity-limited metabolism are
most readily observed under steady-state conditions. Figure 6 shows the
relationship between the plasma phenytoin concentration at steady state and
the administered dose in epileptic patients. For each patient the steady-state
plasma concentration of phenytoin increases more than proportionally to the rate
DOSE-DEPENDENT PHARMACOKINETICS 19

Daily Phenytoin Dose (ma)


Figure 6. For each patient the plasma phenytoin concentration at steady state increases
disproportionately with an increase in the rate of administration. In all patients, the daily dose
required to achieve a steady-state concentrationof 20 mg I - ' is not much greater than that required
to achieve a value of 10 mg I-', within the therapeutic concentration range. These patients show
differences that are greater than those expected in random samples. Selection here was to show
the kinetic behavior in patients in whom several dosage adjustments were required. The lines are
computer fits of the data. (After Winter and Tozer,lMwith permission)

of administered dose. Also,there are considerable variations in the concentration-


dose relationship due to highly variable K , values between patients. lo5
Diflunisal is a fluorinated salicylate, 5-(2 ,4 -diflurophenyl) salicylic acid,
with anti-inflammatory properties, which is eliminated mainly by conjugation
as ether and ester glucuronides. The K, values obtained from in vitro studies
with rat hepatocytes are about 0-06mM (15 pg ml- l ) for both glucuronidations
while the V,, for ester glucuronidation is about three times that for ether
glucuronidation (1 -97 nmol min-I mg-1 microsomal protein versus
0.59 nmol min- mg- microsomal protein) (Lin and Lin, unpublished data).
A steady-state kinetic study in rats revealed that unbound intrinsic clearance of
diflunisal decreased from about 650 ml min- kg- at a steady-state concentration
of 5pgml-I to 40mlmin-I kg-I at a concentration of 300pgml-1.47
The elimination of salicylate can be described by two capacity-limited and
three linear processes. The formation of glycine and phenolic glucuronide
conjugates are saturable and follow Michaelis-Menten kinetics; the K , values
estimated from urinary data in man were 338 mg (expressed as amount of drug
in the body) for glycine conjugation and 629mg for the phenolic
glucuronidation.IM These values are equivalent to about 35 pg ml- and
65 fig ml- I , respectively, since the V, of salicylate is about 150 ml kg-1.22These
values are below the therapeutic concentration range (100-300 pg ml- I ) . The
non-linear kinetics of salicylurateand phenolic glucuronide formation may explain
why a small change in dosage often results in a greater than proportional increase in
plasma concentration of salicylate. In a clinical study,lo7a small increase in the
-
daily dose of aspirin from 1-0g to 1 2 g produced a pronounced therapeuticresponse
in patients with rheumatoid arthritis who had not responded to the lower dose.
Cilastatin, a potent inhibitor of renal dehydropeptidase I, was designed to
inhibit renal metabolism of the antibiotic imipenem in order to achieve
20 J. H. LIN

therapeutically relevant imipenem concentrations in the urinary tract. Cilastatin


is mainly eliminated both by N-acetylation and renal excretion. The elimination
clearance via the N-acetylation pathway was decreased from about 18 ml min-
kg-l at 5mgkg-l to 5mlmin-l kg-l at 200mgkg-l. The non-linear
formation of N-acetylation resulted in dose-dependent kinetics of cilastatin. lo8

Saturation of carrier-mediated transport. The renal excretion of drugs usually


involves three processes: glomerular filtration, renal tubular secretion, and
reabsorption from the renal tubular lumen. Glomerular filtration is a passive
process and is a function of unbound concentration of drug in plasma. The
relationship between renal clearance CLRand these processes can be expressed as
CLR =fuGFR + CL, - CLR, (9)
where GFR, CL, , and CLRaare glomerular filtration rate, secretion clearance,
and reabsorption clearance, respectively, and f , is the unbound fraction of
drug in plasma.
Rearrangement of equation (9) yields

When the CLRf,GFR of a drug is much greater than unity, this means that
renal tubular secretion of the drug occurs. Conversely, when the ratio is much
less than unity, reabsorption of the drug from the tubular lumen occurs. Renal
tubular secretion is a specialized process and saturable, whereas tubular
reabsorption can occur by passive diffusion or active transport. Active
reabsorption is also saturable.
Famotidine, a potent histamine H2-receptor antagonist, is mainly eliminated
renally; approximately 60-70% of the dose is excreted as unchanged drug in
the urine after i.v. dosing. Renal clearance of famotidine in normal human
subjects (4.44ml min-l kg-') is about three times the fuGFR, suggesting that
famotidine is secreted by an active transport system. lo9 In rats, the secretion
of famotidine was inhibited by quinine but not by probenecid, indicating that
farnotidine is secreted by an organic cation transport system. lo The renal
clearance of famotidine in rats was decreased from about 35 ml min-l kg-l at
a steady-state concentration of 0 - 2 p g m l - l to 8 mlmin-l kg-l at a
concentration of 75 pg ml-l.The secretory mechanism is saturable and exhibits
limitation of transport with an apparent T,,, (maximum transport) of
180 pg min- kg- 1.22 Other histamine H2-receptor antagonists, including
cimetidine, ranitidine, and mizatidine, are also actively secreted by the kidney,
and the renal excretion of these drugs is dose dependent. Itoh et al. l 2 have
studied the renal tubular transport of cimetidine in the isolated rat kidney by
means of the multiple-indicator dilution method. They concluded that the
transport on the luminal membrane has a much higher affinity for cimetidine
than that on the antiluminal membrane.
DOSE-DEPENDENT PHARMACOKINETICS 21

Both cephalexin and ampicillin, the widely used antibiotics, are known to
excrete by active renal ~ecretion."~ The V, values for the renal secretory
transport system are estimated to be about 0.24 pmol min-l kg-1 for
cephalexin and 0-04 pmol min - kg - for ampicillin, and the K , values are
about 17 pM and lOpM, re~pective1y.I~~ Since these K, values are of the order
of the therapeutical concentration of the antibiotics, dose-dependent renal
excretion of the antibiotics is expected clinically.
Enalaprilat, the active metabolite of enalapril, is a potent angiotensin-
converting enzyme inhibitor, and is almost exclusively excreted by the kidneys.
The ratio of renal clearance tof,GFR was about 2.7 after an i.v. dose in rats
(1 mg kg-I). Treatment with probenecid and p-aminohippuric acid caused a
profound decrease in the ratios to 1* 10 and 1-25, respectively, suggesting that
enalaprilat is secreted by the organic anion transport system. As expected, the
renal clearance of enalaprilat decreased with increasing dose.115
Renal excretion also plays a substantial role in the elimination of alendronate
in man and animals. Renal clearance is approximately 10-fold greater than
f,GFR, suggesting that alendronate is secreted by an active transport process.
The secretory mechanism in rats exhibits limitation of transport with an apparent
T, of about 25 pg min-I kg-I. The renal clearance of alendronate is dose
-
dependent, decreasing from 7 4 ml min - kg - at 1 mg kg - to
3-25ml min-I kg-1 at 20mg kg-1.116 Interestingly, the renal handling of
alendronate is not affected by treatment with high doses of cimetidine, quinine,
probenecid, or p-amino hippuric acid, suggesting that alendronate is not secreted
by conventional anionic or cationic transport systems. In contrast, etidronate,
a bisphosphonate that is actively secreted by the renal tubules,"' reduced the
renal clearance of alendronate in a dose-related manner without affecting GFR.
This interaction between alendronate and etidronate strongly suggests that they
compete for an uncharacterized renal transport system.
Carrier-mediated reabsorption from the lumen of renal tubules occurs for
many endogenous compounds that are the nutrients to the body, e.g. glucose
and amino acids. On the other hand, the reabsorption of drugs, if any, usually
occurs by simple diffusion. However, some exogenous organic anions, such as
rn-hydroxybenzoate, nicotinate, and pyrazinoate, have been reported to be
transported by tubular secretion as well as active reabsorption.118
Unlike secretory processes, saturation of the reabsorption process results in
an increase in renal clearance. The renal clearance of inorganic sulfate is 10-30%
of GFR under normal physiological conditions, suggesting reabsorption of this
endogenous substrate. The clearance approaches the value of GFR when serum
sulfate concentrations are increased substantially by administration of sodium
sulfate. 1 19
The clearance of drugs by the liver through biliary excretion is largely
determined by carrier-mediated transport at the sinusoidal and canalicular
membranes of hepatocytes. There are at least three classes of hepatobiliary
transport systems: for organic anions, cations, and uncharged drugs. 120 Both
22 J. H . LIN

the entry into the liver and biliary excretion can be saturated at high doses,
providing a clue to the involvement of carrier-mediated mechanisms in both
of these transport steps. However, examples of dose-dependent biliary excretion
of drugs are lacking.
Receptor-mediated endocytosis is a general mechanism in the uptake of
biologically important peptide hormones, and the liver plays an important role
in the disposition of a variety of polypeptides, including insulin, glucagon, and
epidermal growth factors (EGF).l2l Following i.v. administration, 1251-EGF
disappeared very rapidly from the plasma of rats in a monophasic manner. The
half life is about 20 s after a low dose (0.05 nmol kg-I) and about 100 s after
a high dose (20 nmol kg-1).122The dose-dependent kinetics of EGF in rats is
mainly attributed to the saturable receptor-mediated endocytosis in the liver.
Kinetic analyses of hepatic uptake data indicated that the V,, and K , values
of the transport system are about 27-40 pmol min- '/g tissue and 7-50 nM,
respectively.

Concentration-dependent diuresis. Theophylline, a potent bronchodlator, is


widely used in the treatment of acute pulmonary edema and asthma. The drug
is eliminated by hepatic metabolism and renal excretion and the overall clearance
is the sum of the renal and metabolic elimination processes. Although the overall
clearance of theophylline appears to be linear, each metabolic and renal pathway
is non-linear. This unique situation is a result of the reciprocal nature of the
non-linear processes. 123
Because theophylline induces diuresis, renal clearance of theophylline is
elevated at high plasma concentrations after a single dose due to its urine flow-
rate dependence and then decreases rapidly as urine flow rate decreases at low
concentrations. In contrast, the metabolic clearance of theophylline is relatively
slow at high concentrations and increases at lower concentrations of the drug.
Thus, the two processes tend to offset each other and the overall clearance
remains roughly constant.

Non-linear protein binding. As indicated in equation (4), binding of a drug,


and accordingly the unbound fraction, is dependent on the concentration of
protein, the number of binding sites, the affinity constant, and the drug
concentration. As the drug concentration increases, the binding sites on the
protein become increasingly occupied. Thus, the unbound drug and the unbound
fraction increase. A significant correlation between total-body clearance and
the unbound fraction in plasma has been reported for warfarin124 and
t01butamide.l~~Although tissue binding is a principal determinant of the
volume of distribution and half life of a drug, the elimination clearance of a
drug is independent of tissue binding.
Diflunisal is highly bound to rat plasma and the binding is concentration
dependent. A kinetic study was conducted to determine the relationships between
the steady-state concentration and protein binding, and total and unbound
DOSE-DEPENDENT PHARMACOKINETICS 23

1.8 -. .
.
&
1.6-8. ' 0

1.4-

1.2 -
. %
1.0-.

0.8 t 1 1 t
0 100 200 300
DiflunisalConcentation
(mW

3.0 5
0
0
0
2.0

.
0

08
1 -
200 300
Diflunisal Concentation
(mg/L)

Figure 7. Relationship between total-body clearance of diflunisal and plasma diflunisal concentrations
in rats (panel a). The intrinsic clearance of unbound diflunisal(0) and the corresponding unbound
fraction of the drug (0) are shown in relation to plasma concentrations of diflunisal (panel b).
Each point represents data from an individual rat. (After Lin et at. ,4' with permission)

clearance.47The total-body clearance decreases initially and then goes up as the


concentration of diflunisal increases, whereas the intrinsic clearance of unbound
drug decreases with increasing concentration (Figure 7). The former is a
consequence of the dual effects of saturable metabolism and saturable protein
binding; the latter is a reflection of saturable metabolism. The fraction of
unbound diflunisal is increased about 13-fold from 0.29% to 3.9% with a
steady-state concentration range of 5-300 pg ml- I.
Similarly, dual effects of saturable metabolism and plasma protein binding
were reported for salicylate. Furst et ai.49demonstrated that although the
intrinsic clearance of unbound salicylate decreased over the therapeutic
concentration range of 100-300 pg ml-I, total clearance was unchanged. The
consistency in total clearance is a result of the opposing effects of saturable
plasma protein binding and saturable metabolism.
Both MK-417 and MK-927 are carbonic anhydrase inhibitors. Elimination
As. ~indicated
kinetics of these inhibitors are dose d e ~ e n d e n t . ~ ~ ~ . ' ~ ~ in Table 1,
the total-blood clearance increased markedly when the dose of MK-417 was
24 J. H. LIN

raised, from 1 - 3 m l h - 1 k g - 1 at 0.05mgkg-l i.v. to 9 m l h - l k g - ' at


2 mg kg- I i.v. Not surprisingly, the dose required to saturate the binding is
different in anemic, polycythemic, and normal rats. Non-linear kinetics of
MK-417 were observed when the dose exceeded 0.2 mg kg-' in anemic rats,
and 0.5 mg kg- and 1-0mg kg- in normal and polycythemic rats. Furthermore,
dose-dependent changes in total blood clearance were greater in anemic rats
than normal and polycythemic rats (Table 1). However, the terminal half life
appears to be dose independent. The consistency in the half life of MK-417 is
the result of quantitatively similar changes in total clearance and apparent
volume and distribution affected by saturable binding of carbonic anhydrase.
The study of prednisolone in healthy male volunteers has shown that the
systemic plasma clearance of i.v. prednisolone is dose dependent and increases
from 11 1 ml min- '/1.73 m2 to 194 ml min- '/1-73 m2 over the 5-40 mg i.v.
dosage range.55The steady-state volume of distribution also increases, but little
change in mean residence time and half life is found. The concentration-dependent
binding of prednisoloneto plasma proteins, mainly transcortin, is largely responsible
for the dose-dependent kinetics of prednisolone. Similar changes in the binding
of hydrocortisone are thought to affect the clearance of hydroc~rtisone.~~'

Depletion of co-substrate. All the conjugations of drug are bireactant


reactions, which require substrate (drug) and co-substrate. The co-substrates,
such as UDPGA and PAPS, can be depleted when the demand is greater than
the supply. After i.v. injection of 15 mg kg-l, 30mg kg-I, l5Omg kg-I, and
300 mg kg- of acetaminophen to rats without concomitant administration of
inorganic sulfate or a sulfate donor, the metabolic formation clearances of
acetaminophen to sulfate conjugate are 39.9f 9.6 ml min- kg-I,
36.6 f9.3 ml min- kg - l , 6.0 f 0.5 ml min- kg- l , and 2 . 8 f0.4 ml min-
kg - l , respectively. 12* This dose-dependent behavior of acetaminophen is
thought to be partly due to the depletion of co-substrate PAPS (adenosine
3 ' -phosphate-5 ' -sulfatophosphate) caused by a depletion of inorganic sulfate.
An injection of 15 mg kg- of acetaminophen causes very little sulfate
depletion in rats whereas a 300 mg kg- dose causes endogenous inorganic
sulfate concentrations in serum to decrease from the normal level of about
0.9mM to about 0.1 mM for several hours.129To determine the effect of
sulfate depletion on the pharmacokinetics of acetaminophen, rats received an
i.v. injection and a continuous infusion of sodium sulfate as well as an i.v.
injection of acetaminophen, either 15 mg kg-I, 30mg kg-', 150 mg kg-I, or
300mg kg-l. The values of the metabolic formation clearance of
acetaminophen to sulfate conjugate in sulfate-repleted rats are 36 * 3 f
1-8mlmin-1kg-1, 29.8f6-lmlmin-1kg-1, 10.9k2-4mlmin-1kg-1, and
6 - 5 f 0 . 6 m l min-I kg-I, respectively.130 The metabolic fate of a 15 mg kg-'
dose of acetaminophen in sulfate-repleted rats is very similar to the value
reported for rats that had not received supplemental sodium sulfate. On the
other hand, the metabolic formation clearances of acetaminophen to sulfate
DOSE-DEPENDENT PHARMACOKINETICS 25

conjugate of 150mg kg-1 and 300 mg kg- dose are increased 1 -5- and
twofold, respectively, by concomitant administration of inorganic sulfate when
compared to rats that had not received supplemental sodium sulfate.128These
results indicate that the depletion of co-substrate has significant dose-dependent
effects on the pharmacokinetics of acetaminophen.

Product inhibition. Theoretical aspects of the effect of product inhibition


on elimination of drugs has been discussed by Ashley and Levy.I3l The
biotransformation of phenytoin was shown to be inhibited by its major
metabolite, 5-(p-hydroxyphenyl)-5-phenylhydantoinboth in V ~ V O Iand ~ ~ in
vitro.133Inhibition by the product is believed to be partly responsible for the
dose-dependent kinetics of phenytoin observed in rats132 and in Rhesus
m0nke~s.l~~
Similarly, inhibition of ethoxybenzamide demethylation by its metabolite,
salicylamide, has been demonstrated both in vivo and in ~ i t r 0 . Following
l~~ i.v.
administration of lOmg kg-l, 20mg kg-I, or 80mg kg-I, the elimination of
ethoxybenzamide in rabbits was dose dependent. The dose-dependent kinetics
of ethoxybenzamide could be partly attributed to product inhibition.

Drug-induced hepatotoxicity. Drugs that are metabolized in the liver may at


the same time destroy or damage metabolic enzymes, so that the elimination
of drugs may decrease in a dose-dependent manner. L-689,502 is a potent and
highly selective inhibitor of HIV protease. Following i.v. administration, total-
body clearance in rats decreased from about 180ml min-' kg-* at 1 mg kg-'
to 86mlmin-I kg-I at 20mg kg-I. Similar results were observed in dogs;
clearance fell from 29mlmin-l kg-l at 0.5mgkg-I to 17mlmin-l kg-1 at
10 mg kg- 1.136Bile flow in rats was retarded in a dose-dependent manner after
an i.v. injection of L-689,502. Consistent with the cholestatic effect, L-689,502
caused a substantial increase in serum levels of aminotransferase. Moreover,
pretreatment of rats with L-689,502 resulted in a significant decrease in the
elimination kinetics of antipyrine and diflunisal as well as L-689,502 itself. These
results suggest that the dose-dependent kinetics of L-689,502 in rats and dogs
are due to hepatotoxicity caused by the drug.

CONCLUSIONS

Although the pharmacokinetics of most drugs in their clinical dose range can
be adequately described by first-order or linear processes, there are a number
of drugs that exhibit non-linear kinetics, particularly at high doses. The non-
linearity sometimes complicates the design of dosage regimens and prediction
of efficacy and toxicity. For example, in cancer chemotherapy, the common
approach is to maximize the dose in order to enhance entry of drugs into resistant
cells.137As a consequence, non-linear kinetics may occur when metabolism
26 J. H . LIN

and/or excretion processes are saturated, and a small change in dose may result
in substantial changes in plasma levels, leading to no pharmacological effect
or toxic effect.
In preclinical studies, toxicologists also face a difficult dilemma when they
attempt to extrapolate observed high dose toxicity to the safety of low doses.
The underlying difficulty is that the kinetic behavior may be dose dependent,
resulting in a greater-than- or less-than-dose-proportional response in AUC with
unpredictable toxicologic consequences. Thus, an understanding of the influence
of dose on the pharmacokinetics is important in the evaluation of the efficacy
and toxicity of new drugs.

REFERENCES
1. G. Levy and M. Gibaldi, Pharmacokinetics. In Concepts in Biochemical Pharmacology:
Handbook for Experimental Pharmacology, Vol. 8. J . R. Gillette and J. R. Mitchell (Eds),
New York, Springer, 1975, Vol. 8, part 3, pp. 1-34.
2. M. Gibaldi and D. Perrier, Nonlinear pharmacokinetics. In Pharmacokinetics, Dekker, New
York, 1982, Ch. 7, pp. 271-318.
3. P. G. Dayton and J. E. Sanders, Dose-dependent pharmacokinetics: emphasis on phase I
metabolism. Drug Metab. Rev., 14, 347-405 (1983).
4. J. M. Van Rossum and G. Van Lingen. Dose-dependent pharmacokinetics. In Pharmacokinetics:
Theory and Methodology, M. Rowland and G. Tucker (Eds), Pergamon, New York, 1986,
Ch. 5, pp. 91-114.
5. D. S. Riggs, Linear vs. nonlinear operations. In Control Theory and Physiological Feedback
Mechanisms, Krieger, New York, 1976, Ch. 1, pp. 23-27.
6. J. B. Dressman, D. Fleisher and G. L. Amidon, Physicochemical model for dose-dependent
drug absorption. J. Pharm. Sci., 73. 1274-1279 (1984).
7. J. B. Dressman and D. Fleisher, Mixing-tank model for predicting dissolution rate control of
oral absorption. J. Pharm. Sci., 75, 109-116 (1986).
8. I.-W. Chen, J. M. Dorley, H. G. Ramjit, S. M. Pitzenberger and J. H. Lin, Physiological
disposition and metabolism of L-365.260, a potent antagonist of brain cholescystokininreceptor,
in laboratory animals. Drug Metab. Disposit., 20, 390-395 (1992).
9. W. G. Crouthamel, G. H. Tan, L. W. Dittert and J. T. Dolusio, Drug absorption i.v. Influence
of pH on absorption kinetics of weakly acidic drugs. J. Phurm. Sci., 60, 1160-1166 (1971).
10. R. R. Levine, Factors affecting gastrointestinal absorption of drugs. Digestive Diseases, 15,
171-188 (1970).
11. T. Z. Csaky, Intestinal permeation and permeability: an overview. In Pharmacology of Intestinal
Permeation, Vol. 1, T. Z. Csaky (Ed.), Springer, Berlin, 1984, pp. 51-59.
12. D. N. Wade, P . T. Mearrick and J. I. Morris, Active transport of L-dopa in the intestine.
Nature, 247, 463-465 (1972).
13. G. L. Amidon, A. E. Merfeld and J. B. Dressman, Concentration and pH dependency of a-
methyldopa absorption in rat intestine. J. Phurm. Pharmacol., 38, 363-368 (1985).
14. M. Hu, P. J . Sinko, A. L. J. DeMeere, D. A. Johnson and G. L. Amidon, Membrane
permeability parameters for some amino acids and @-lactam antibiotics: application of the
boundary larger approach. J. Theor. Biol., 131, 107-114 (1988).
15. A. Tsuji, E. Nakashima, A. Asano, R. Nakashima and T. Yamana, Kinetics of Michaelis-
Menten absorption of amino-penicillins in rats. J. Pharm. Pharmucol., 30, 508-509 (1978).
16. Y. Tomita, T. Katsura, T. Okano, K. Inui and R. Hari, Transport mechanisms of bestatin
in rabbit intestinal brush-border membranes: role of H+/dipeptide cotransport system. J.
Phurmacol. Exp. Ther., 252, 859-862 (1990).
17. K. Inui, T. Okano, H. Maegawa, M. Kato, M. Takano and R. Hori, Coupled transport of
p.0. cephalosporins via dipeptide carriers in rabbit intestinal brush-border membranes: difference
of transport characteristics between cefixime and cephradine. J. Phurmacol. Exp. Ther., 247,
235-241 (1988).
DOSE-DEPENDENT PHARMACOKINETICS 27

18. V. S. Chungi, D. W. A. Bourne and L. W. Dittert, Drug absorption VIII. Kinetics of GI


absorption of methotrexate. J. Pharm. Sci., 67, 560-561 (1978).
19. A. Tsuji, E. Nakashima, I. Kagami and T. Yamana, Intestinal absorption mechanism of
amphoteric @-lactamantibiotics 11: Michaelis-Menten kinetics of cyclacillin absorption and
its pharmacokinetic analysis in rats. J. Pharm. Sci., 70, 772-777 (1981).
20. W. E. Evans, W. R. Crom and J. C. Yalowich, Methotrexate. In AppliedPharmacokinetics-
Principles of Therapeutic Drug Monitoring, 2nd edition, W. E. Evans, J. J. Schentag and
W. J. Jusko (Eds), Applied Therapeutics Inc., Spokane, WA, 1986, Ch. 30, pp. 1009-1056.
21. K. B. Bischoff, R. L. Sedrick, D. S. Zaharko and J. A. Longstreth, Methotrexate
pharmacokinetics. J. Pharm. Sci., 60, 1128-1133 (1971).
22. J. H. Lin, D. M. Cocchetto and D. E. Duggan, Protein binding as a primary determinant of
the clinical pharmacokinetic properties of nonsteroidal anti-inflammatory drugs. Clin.
Pharmacokinet., 12, 402-432 (1987).
23. G. R. Wilkinson, Clearance approaches in pharmacology. Pharmacol. Rev., 39, 1-47 (1987).
24. B. K. Wong, J. H. Lin, P. J. Bruhin, Nonlinear pharmacokinetics for L-693,612, a carbonic
anhydrase inhibitor after oral administration in rats. Pharm. Res., 8 (Suppl.), S292 (1991).
25. M. Bialer, U. K. Batra, J. A. Morrison, B. M. Silber, Z. M. Look and A. Yocobi, Dose-
dependent pharmacokinetics of a new oral cephalosporin, cefixime, in the dog. Pharm. Res.,
4, 33-37 (1987).
26. R. Runkel, E. Forchielli, H. Sevalius, M. Chaplin and E. Segre, Nonlinear plasma level response
to high doses of naproxen. Clin. Pharmacol. Ther., 15, 261-266 (1974).
27. R. Runkel, M. Chaplin, H. Sevelius, E. Ortega and E. Segre, Pharmacokinetics of naproxen
overdoses. Clin. Pharmacol. Ther., 20, 269-277 (1976).
28. R. H. Levy, Timedependent pharmacokinetics. In Pharmacokinetics: Theory and Methodology,
M. Rowland and G. Tucker (Eds), Pergamon, New York, 1986, pp. 115-129.
29. L. Bertilsson, B. Hojer, G. Tybring, J. Osterich and A. Rane, Autoinduction of carbamazepine
metabolism in children examined by a stable isotope technique. Clin. Pharmacol. Ther., 27,
83-88 (1980).
30. P. A. Reilly, I. Tadnobu, D. Kadar and L. Endrenyl, Enzyme induction following a single
dose of amobarbital in dogs. J. Pharmacokinet. Biopharm., 6, 305-313 (1978).
31. A. Adedoyin, L. Aarons and J. B. Houston, Time-dependent disposition of 8-naphthoflavone.
Pharm. Res., 10, 35-43 (1993).
32. V. K. Batra and A. Yacobi, An overview of toxicokinetics. In Toxicokineticsand New Drug
Development, A. Yacobi, J. P. Skelly and V. K. Batra (Eds), Pergamon, New York, 1989,
pp. 1-20.
33. D. J. Tocco, B. K. Wong and J. H. Lin, Physiological disposition and metabolism of L-706,631
in rats and Rhesus monkeys. Data on file, Merck Research Laboratories.
34. W. H. Barr, Factors involved in the assessment of systemic or biologic availability of drug
products. Drug Information Bulletin, 3, 27-45 (1969).
35. D. G. Shand and R. E. Rango, The disposition of propranolol. I. Elimination during oral
absorption in man. Pharmacology, 7 , 159-168 (1972).
36. T. Walle, E. C. Convadi, K. Walle, T. C. Fagan and T. E. Gaffney, The predictable relationship
between plasma levels and dose during chronic propranolol therapy. Clin. Pharmacol. Ther.,
24, 668-677 (1978).
37. J. H. Lin, I.-W. Chen, F. A. de Luna, Factors affecting the absorption of alendronate in rats.
Drug Metab. Disposit. (1993a), in press.
38. J. R. Gillette: Factors affecting drug metabolism. Ann. New York Acad. Sci., 179,43-66 (1971).
39. G. R. Wilkinson and D. G. Shand, A physiological approach to hepatic clearance. Clin.
Pharmacol. Ther., 18, 377-390 (1975).
40. W. J. Jusko and M. Gretch, Plasma and tissue protein binding of drugs in pharmacokinetics.
Drug Metab. Rev., 5 , 43-140 (1976).
41. S. Oie and T. N. Tozer, Effects of altered plasma protein binding on apparent volume of
distribution. J. Pharm. Sci., 68, 1203-1205 (1979).
42. G. H. Evans, A. S. Nies and D. G. Shand, The disposition of propranolol. 111. Decreased
half-life and volume of distribution as a result of plasma binding in man, monkey, dog and
rat. J. Pharmacol. Exp. Ther., 186, 114-120 (1973).
43. G. Levy, C. M. Lai and A. Yacobi, Comparative pharmacokinetics of coumarin anticoagulants.
XXXII: Interindividual differences in binding of warfarin and dicumarol in rat lives and
implications for physiological pharmacokinetic modeling. J. Pharm. Sci., 67,229-23 1 (1978).
28 J. H. LIN

44. Y. Sawada, M. Hanano, Y. Sugiyama, H. Harashima and T. lga, Prediction of the volume
of distribution of basic drugs in humans based on data from animals, J. Pharmacokinet.
Biopharm., 12, 587-594 (1984).
45. Y. Sawada, M. Hanano, Y. Sugiyama and T. lga, Prediction of the disposition of nine weakly
acidic and six weakly basic drugs in humans from pharmacokinetic parameters in rats. J.
Pharmacokinet. Biopharm., 13, 477-492 (1985).
46. J. H. Lin, Species differences in protein binding of diflunisal. Drug Metab. Disposit., 17,221-223
(1989).
47. J. H. Lin, K. F. Hooke, K. C. Yeh and. D. E. Duggan, Dose-dependent pharmacokinetics
of diflunisal in rats: dual effects of protein binding and metabolism. J. Pharmacol. Exp. Ther.,
235, 402-406 (1985).
48. G . Levy and S. J. Yaffee, Relationship between dose and apparent volume of distribution of
salicylate in children. Pediatrics, 54, 713-717 (1974).
49. D. E. Furst, T. N. Tozer and K. L. Malman, Salicylate clearance, the resultant of protein binding
and metabolism. Clin. Pharmacol. Ther., 26, 380-389 (1979).
50. J . M. H. Kremer, J. Wilting and L. H. M. Janssen, Drug binding to human Alpha-1-acid
glycoprotein in health and disease. Pharmacol. Res., 40, 1-47 (1988).
51. M. Werner, Serum protein changes during the acute phase reaction. Clin. Chim. Acta, 25,
299-305 (1969).
52. S . Urien, E. Albengres, J. L. Pinquier and J. P. Tillement, Role of alpha-1-acid glycoprotein,
albumin and nonesterified fatty acids in serum binding of apazone and warfarin. Clin.
Pharmacol. Ther., 39, 683-689 (1986).
53. Z. H. Israili and H. El-Attar, Binding of some acidic drugs to alpha-I-acid glycoprotein. Clin.
Pharmacol. Ther., 33, 255 (1983).
54. A. A. Sandberg, H. Rosenthal, S. L. Schneider and W. R. J. Slaunwhite, Protein-steroid
interaction and their role in the transport and metabolism of steroids. In Steroid Dynamics,
G . Pincus, T. Nanao and J. R. Tait (Eds), Academic, New York, 1966.
5 5 . J. Q. Rose, A. M. Yurchak and W. J. Jusko, Dose-dependent pharmacokinetics of prednisone
and prednisolone in man. J. Pharmacokinet. Biopharm., 9 , 389-417 (1981).
56. M. Ehrnebo, Drug binding to erythrocytes. In Protein Binding and Drug Transport, J-P.
Tillement and E. Lindenlaub (Eds), Schattauer, Stuttgart, 1986, pp. 49-57.
57. M. Lemaire, Drug binding to leucocytes and platelets. In Protein Binding and Drug Transport,
J .-P. Tillement and E. Lindenlaub (Eds), Schattauer, Stuttgart, 1986, pp. 79-90.
58. M. Wind, A. Berliner and A. Stern, The binding of phenothiazines to oxyhemoglobin A and
S. Res. Commun. Chem. Pathol. Pharmacol., 5 , 759-766 (1973).
59. M. Ehrnebo, Influence of drug binding to blood cells on pharmacokinetics, Acta. Pharm. Suec.,
17, 81-82 (1980).
60. J . H. Lin, I.-W. Chen and F. A. deluna, Dose-dependent pharmacokinetics of MK-417, a
potent carbonic anhydrase inhibitor, in experimental polycythemic and anemic rats. Pharm.
Res., 8, 608-614 (1991a).
61. J. H. Lin, 1.-W. Chen and T. H. Lin, Species-dependent stereopharmacokinetics of MK-927,
a potent carbonic anhydrase inhibitor. Drug Metab. Disposit., 19, 816-822 (1991b).
62. S. M. Wallace and S. Riegelman, Uptake of acetazolamide by human erythrocytes. J. Pharm.
Sci., 66, 729-731 (1977).
63. W. 0. Kwant and P. Seeman, The membrane concentration of a local anesthetic chlorpromazine.
Biochim. Biophys. Acta., 183, 530-543 (1969).
64. M. H. Bickel, Binding of chlorpromazine and imipramine to red cells, albumin, lipoproteins
and other blood components. J. Pharm. Pharmacol., 27, 733-738 (1975).
65. T . H. Maren, Carbonic anhydrase: Chemistry, physiology and inhibition, Physiol. Rev., 47,
595-781 (1967).
66. J. H. Lin, T. H. Lin and H. Cheng, Uptake and stereoselective binding of the enantiomers
of MK-927, a potent carbonic anhydrase inhibitor, by human erythrocytes in vitro. Pharm.
Rex, 9, 335-340 (1992~).
67. J. H. Lin, I.-W. Chen, E. H. Ulm, J. R. Gehret and D. E. Duggan, Dose-dependent
pharmacokinetics of MK-417, a potent carbonic anhydrase inhibitor, in rabbits following single
and multiple doses. Drug Metab. Disposit., 18, 836-841 (1990).
68. A. J. Wood and M. Lemaire, Pharmacologic aspects of cyclosporin therapy: pharmacokinetics.
Transplant. Proc., 17, 27-35 (1985).
DOSE-DEPENDENT PHARMACOKINETICS 29

69. D. J. Tocco, K. F. Hooke, F. A. de Luna and A. E. W. Duncan, Stereospecific binding of


timolol, a beta-adrenergic blocking agent. Drug Metab. Disposit., 4, 323-329 (1976).
70. N. E. Goeders and M. J. Kuhar, Benzodiazepine receptor binding in vivo with ['HI
Ro-15-1788. Life Sci., 37, 345-355 (1985).
71. H A .G. Chen and J. F. Gross, Estimation of tissue-to-plasma partition coefficients used in
physiological pharmacokinetic models. J. Pharmucokinet. Biopharm., 7, 117-125 (1979).
72. W. K. Cheung and G. Levy, Comparative pharmacokinetics of coumarin anticoagulants XLIX:
Nonlinear tissue distribution of s-warfarin in rats. J. Pharm. Sci., 78, 541-546 (1989).
73. J. T. Slattery, A. Yacobi and G. Levy, Comparative pharmacokinetics of coumarin
anticoagulants XXV: warfarin-ibuprofen interaction in rats. J. Phurm. Sci., 66,943-947 (1977).
74. R. L. Dedrick, D. S. Zaharko and D. J. Lutz, Transport and binding of methotrexate in vivo.
J. Pharm. Sci., 62, 882-890 (1973).
75. H. Harashima, S. Sugiyama, T. Iga and M. Hanano, Nonlinear tissue distribution of ouabain
in rabbits. Drug Metab. Disposit., 16, 645-649 (1988a).
76. H. Kurz and B. Fichtl, Binding of drugs to tissues. Drug Metab. Rev., 14, 467-510 (1983).
77. E. N. Cohen, A. Corbascio and G. Fleischi, The distribution and fate of d-tubucorarine. J.
Pharmacol. Exp. Ther., 147, 120-129 (1965).
78. J. E. Doherty and W. H. Perkins, Tissue concentration and turnover of tritiated digoxin in
dogs. Am. J, Cardiol., 17, 47-52 (1966).
79. J. E. Doherty, W. H. Perkins and W. J. Flanigan, The distribution and concentration of tritiated
digoxin in human tissue. Ann. Intern. Med., 66, 116-124 (1967).
80. H. Harashima, M. Mamiya, M. Yamazaki, Y. Sugiyama, Y. Sawada, T. Iga and M. Hanano,
Significance of binding to Na, K-ATPase in the tissue distribution of ouabain in guinea pig.
Pharm. Res., 9, 474-479 (1992a).
81. R. H. Reuning and D. R. Geraets, Digoxin. In Applied Pharmacokinetics, W. E. Evans, J.
J. Schentag and W. J. Jusko (Eds), Applied Therapeutic Inc., Spokane, WA, 1986, Ch. 18,
~~

pp. 570-623.
82. F. Zannad. R. J. Rover and G. Issartel. Digoxin in the elderly and in renal failure. Contribution
of erythrocyte 86-ribidium uptake t e s t s . u r . J. Cardiol.; 12, 285-296 (1981).
83. J. M. Van Rossum, C. A. M. Van Ginneken, P. T. H. Henderson, H. C. J. Ketelaars and
T. B. Vree, Pharmacokinetics of biotransformation. In Kinetics of Drug Action, Handbook
of Experimental Pharmacology, Vol. 47, J. M. Van Rossum (Ed.), 1974, Springer, Berlin,
pp. 125-167.
84. 0. Hansen, Interaction of cardiac glycosides with ma+-K+) activated ATPase. A biochemical
link to digitalis induced inotropy. Pharmacol. Res., 36, 143-163 (1984).
85. H. Harashima, Y. Sugiyama, Y. Sawada, K. Shigenobu, Y. Kasuya, T. Iga and M. Hanano,
Kinetic analysis of the positive inotropic action (PIA) of ouabain in isolated perfused rabbit
heart. Slow onset of PIA and slow binding to Na+, K+-adenosine triphosphatase. J.
Pharmacobio-Dyn., 11, 533-540 (1988b).
86. L. Brown and E. Erdmann, Binding of digitalis derivations to beef, cat and human (Na+, K+)-
ATPase affinity and kinetic constants. Arch. Int. Phurmucodyn., 271, 229-240 (1984).
87. H. Harashima, M. Mamiya, M. Yamazaki, Y. Sawada, Kinetic modeling of ouabain tissue
distribution based on slow and saturable kinding to Na, K-ATPase. Pharm. Res., 9,1607-161 1
( 1992b).
88. W. M. Pardridge, Transport of protein-bound hormones into tissues in vivo. Endocrine Rev.,
2, 103-123 (1981).
89. W. M. Pardridge, Overview of blood-brain barrier transport biology and experimental
methodologies. In Peptide Drug Delivery to the Brain, Raven, New York, 1991, pp. 52-98.
90. T. Terasaki and. W. M. Pardridge, Stereospecificity of triiodothyronine transport into brain,
liver and salivary gland: role of carrier- and plasma protein-mediated transport. Endocrinology,
121, 1185-1191 (1987).
91. W. M. Pardridge and L. J. Mietus, Kinetics of neutral amino acid transport through the blood-
brain barrier of the newborn rabbit. J. Neurochem., 38, 955-962 (1982).
92. D. C. Markowitz and J. D. Fernstrom, Diet and uptake of aldomet by the brain: competition
with natural large neural amino acids. Science, 197, 1014-1015 (1977).
93. J. H. Lin, D. E. Duggan, I.-W. Chen and R. L. Ellsworth, Physiological disposition of
alendronate, a potent antiosteolytic bisphosphonate, in laboratory animals. Drug Metab.
Disposit., 19, 926-932 (1991~).
30 J . H. LIN

94. J . H. Lin, I.-W. Chen and D. E. Duggan, Effects of dose, sex and age on the disposition
of alendronate, a potent antiosteolytic bisphosphonate, in rats. Drug Metab. Disposit., 20,
473-478 (1992b).
95. J. H. Lin, I.-W. Chen and F. A. deLuna, Uptake of alendronate by bone tissue in hypocalcemic
and hypercalcemic rats. Drug Metab. Disposit., 21, 800-804 (1993b).
96. J . Monkkonen, P. Ylitalo, H. A. Elo and M. M. Alvaksinen, Distribution of 14C-clodronate
(dichloromethylenebisphosphonate) disodium in mice. Toxicol. Appl. Pharmacol., 89,287-292
(1987).
97. J. Monkkonen, H. M. Koponen and P. Ylitalo, Comparison of the distribution of three
bisphosphonates in mice. Pharmacol. Toxicol., 65, 294-298 (1989a).
98. J . Monkkonen, A. Urtti, P. Paronen, H. A. Elo and P. Ylitalo, The uptake of clodronate
(dichloromethylene bisphosphonate) by macrophages in vivo and in vitro. Drug Metab.
Disposit., 17, 690-693 (1989b).
99. M. H. Bickel, The role of adipose tissue in the distribution and storage of drugs. Prog. Drug.
Res., 28, 273-303 (1984).
100. F. Lundquist and H. Wolthers, The kinetics of alcohol elimination in man. Acta Pharmacol.
Toxicol., 14, 265-289 (1958).
101. A. B. Makar and G. J. Mannering, Kinetics of ethanol metabolism in the intact rat and monkey.
Biochem. Pharmacol., 19, 2017-2022 (1970).
102. R. L. Dedrick and D. D. Forrester, Blood flow limitations in interpreting Michaelis constants
for ethanol oxidation in vivo. Biochem. Pharmacol., 22, 1133-1 140 (1973).
103. J. H. Lin, M. Hayashi, S. Awazu and M. Hanano. Correlation between in vivo and in vitro
drug metabolism rate: oxidation of ethoxybenzamide in rat. J. Pharmacokinet. Biopharm.,
6 , 327-337 (1978).
104. M. E. Winter and T. N. Tozer, Phenytoin. In Applied Pharmacokinetics: Principles of
TherapeuticDrugMonitoring, W. E. Evans, J. J. Schentag and W. J . Jusko (Eds). Applied
Therapeutics Inc., Spokane, WA, 1986, Ch. 16, pp. 493-539.
105. A. Richens and A. Dunlop, Serum phenytoin levels in the management of epilepsy. Lancet,
2, 247-248 (1975).
106. G. Levy, T. Tsuchiya and L. P. Amsel, Limited capacity for salicyl phenolic glycyronide
formation and its effect on the kinetics of salicylate elimination in man. Clin. Pharmacol.
Ther., 13, 258-268 (1972).
107. J . J. Calabro and J. M. Marchesano, Fever associated with juvenile rheumatoid arthritis.
New Engl. J . Med., 276, 11-18 (1967).
108. J . H. Lin, I.-W. Chen and E. H. Ulm, Dose-dependent kinetics of cilastatin in laboratory
animals. Drug Metab. Disposit., 17, 426-432 (1989).
109. J . H. Lin, A. N. Chremos, K. C. Yeh, J. Antonello and G. A. Hessey, Effects of age and
chronic renal failure on the urinary excretion kinetics of famotidine in man. Eur. J. Clin.
Pharmacol., 34, 41-46 (1988b).
110. J. H. Lin, L. E. Los, E. D. Ulm and D. E. Duggan, Kinetic studies on the competition between
famotidine and cimetidine in rats: evidence of multiple renal secretory systems for organic
cations. Drug Metab. Disposit., 16, 52-56 (1988~).
111. J . H. Lin, Pharmacokinetic and pharmacodynamic properties of histamine H2-receptor
antagonists: relationship between intrinsic potency and effective plasma concentrations. Clin.
Pharmacokinet., 20, 217-236 (1991).
112. N. Itoh, Y. Sawada, Y. Sugiyama, T. Iga and M. Hanano, Kinetic analysis of rat renal tubular
transport based on multiple indicator dilution method. Am. J. Physiol., 251, F103-Fl14 (1986).
113. R. Hori, K. Okumura, A. Kamiya, H. Nihira and H. Nakano, Ampicillin and cephalexin
in renal insufficiency. Clin. Pharmacol. Ther., 34, 792-798 (1983).
114. A. Kamiya, K. Okumura and R. Hori, Quantitative investigation on renal handling of drugs
in rabbits, dogs and humans. J. Pharm. Sci., 72, 440-443 (1983).
115. J. H. Lin, I.-W. Chen, E. H. Ulm and D. E. Duggan, Differential renal handling of angiotensin-
converting enzyme inhibitors enalaprilat and lisinopril in rats. Drug Metab. Disposit., 16,
392-396 (1988a).
116. J . H. Lin, I.-W. Chen, F. A. deLuna and M. Hichens, Renal handling of alendronate in rats:
an uncharacterized renal transport system. Drug Metab. Disposit., 20, 608-613 (1992a).
117. U. Troehler, J. P. Bonjour and H. Fleish, Renal secretion of diphosphonates in rats. Kidney
Int., 8, 6-13 (1975).
DOSE-DEPENDENT PHARMACOKINETICS 31

118. I. M. Weiner, Organic acids and bases and uric acid. In The Kidney: Physiology and
Pathophysiology. D. W. Seldin and G. Giebisch (Eds), Raven, New York, 1985, pp. 1703-1724.
119. J. H. Lin and 0. Levy, Renal clearance of inorganic sulfate in rats: Effect of acetaminophen-
induced depletion of endogenous sulfate. J. Pharm. Sci., 72, 213-217 (1983).
120. D. K. F. Meijer, Current concepts on hepatic transport of drugs. J. Heputol., 4,259-268 (1987).
121. Y. Sugiyama and M. Hanano, Receptor-mediated transport of peptide hormones and its
importance in the overall hormone disposition in the body. Pharm. Res., 6, 192-202 (1989).
122. D. C. Kim, Y. Sugiyama, H. Satoh, T. Fuwa, T. Iga and M. Hanath, Kinetic analysis of
in vivo receptor-dependent binding of human epidermal growth factor by rat tissues. J. Pharm.
Sci., 77, 200-207 (1988).
123. D. D-S. Tan-Liu, R. L. Williams and S. Riegelman, Nonlinear theophylline elimination. Clin.
Pharmacol. Ther., 31, 358-369 (1982).
124. G. Levy and A. Yacobi, Effect of plasma protein binding on elimination of warfarin. J. Pharm.
Sci., 63, 805-806 (1974).
125. W. L. Schary and M. Rowland, Protein binding and hepatic clearance: studies with tolbutamide,
a drug of low intrinsic clearance, in the isolated perfused rat liver preparation. J.
Pharmacokinet. Biopharm., 11, 325-343 (1983).
126. J. H. Lin, E. H. Ulm and L. E. Los, Dose-dependent stereopharmacokinetics of
5,6-dihydro-4H-4(isobutylamino)thieno(2,3-b)thiopyran-2-sulfonamide-7,7-dio~de, a potent
carbonic anhydrase inhibitor, in rats. Drug Metab. Disposit., 19, 233-238 (1991d).
127. P. G. Zager. W. J. Burtis, J. A. Luetscher, A. J. Dowdy and S. Sood, Increased plasma
protein binding and lower metabolic clearance rate of aldosterone in plasma of low cortisol
concentration. J. Clin. Endocrinol. Metab., 42, 207-214 (1976).
128. R. E. Galinsky and G. Levy, Dose- and time-dependent elimination of acetaminophen in rats:
pharmacokinetic implications of co-substrate depletion. J. Pharmacol. Exp. Ther., 219, 14-20
(1981).
129. J. H. Lin and G. Levy, Sulfate depletion after acetaminophen administration and replenishment
by infusion of sodium sulfate or N-acetylcysteine in rats. Biochem. Pharmacol., 30,2723-2725
(1981).
130. J. H. Lin and G. Levy, Effect of prevention of inorganic sulfate depletion on the
pharmacokinetics of acetaminophen in rats. J. Pharmacol. Exp. Ther., 239,94-98 (1986).
131. J. J. Ashley and G. Levy, Kinetics of diphenylhydantoin elimination in rats. J. Pharmacokinet.
Biopharm., 1, 99-102 (1973).
132. J. J. Ashley and G. Levy, Inhibition of diphenylhydantoin elimination by its major metabolite.
Res. Commun. Chem. Pathol. Pharmacol., 4, 297-306 (1972).
133. P. Brondy, T. Chang and J. Glazko, Inhibition of diphenylhydantoin (DPH) hydroxylation
by 5-(p-hydroxyphenyl)-5-phenylhydantoin(p-HPPH). Fed. Proc., 31, 582 (abstr.) (1972).
134. A. J. Glazko, T. Chang, E. Maschewske, A. Hayes and W. A. Dill, Role of hydroxylated
metabolites of phenytoin in dose-dependency. In Microsomes and Drug Oxidation, V. Ullrich,
I. Roots, A. Hildebrant, R. W. Estabrook and A. H. Conney (Eds), Pergamon, New York,
1977, pp. 508-513.
135. J. H. Lin, Y. Sugiyama, M. Hanano and S. Awazu, Effect of product inhibition on elimination
kinetics of ethoxybenzamide in rabbits: Analysis by physiological pharmacokinetic model.
Drug Metab. Disposit., 12, 253-256 (1984).
136. J. H. Lin, I.-W. Chen and J. King, Dose-dependent toxicokinetics of L-689,502, a potent
HIV protease inhibitor, in rats and dogs. J. Pharmacol. Exp. Ther., 263, 105-111 (1992).
137. G. Powis, M. M. Ames and J. S. Kovach, Dose-dependent pharmacokinetics and cancer
chemotherapy. Cancer Chemother. Pharmacol., 6, 1-9 (1981).

You might also like